Digitized by the Internet Archive

in 2012 with funding from

University of Illinois Urbana-Champaign

http://archive.org/details/organicsemin1971752univ

/ a

ORGANIC SEMINAR ABSTRACTS

1971-1972

Semester II

School of Chemical Sciences

Department of Chemistry-

University of Illinois

Urbana, Illinois

3

SEMINAR TOPICS

II Semester 1971-72

Reactions of Alkyl Ethers Involving n- Complexes on a Reaction Pathway 125 Jerome T. Adams

New Syntheses of Helicenes 127 Alan Morrice

Recent Advances in Drug Detection and Analysis I36 Ronald J. Convers

The Structure of Carbonium Ions with Multiple Neighboring Groups 138 William J. Work

Recent Reactions of the DMSO-DCC Reagent ll+O James A. Franz

Nucleophilic Acylation 1^2 Janet Ollinger

The Chemistry of Camptothecin lU^ Dale Pigott

Stereoselective Syntheses and Absolute Configuration of Cecropia Juvenile Horomone 1U6 John C. Greenfield

Uleine and Related Aspidosperma Alkaloids 155 Glen Tolman

Strategies in Oligonucleotide Synthesis 162 Graham Walker

Stable Carbocations: C-13 Nuclear Magnetic Resonance Studies 16U Moses W. McMillan

Organic Chlorosulfinates 166 Steven W. Moje

Recent Advances in the Chemistry of Penicillins and Cephalosporins 168 Ronald Stroshane

Cerium (iv) Oxidations 175 William C. Christopfel

A New Total Synthesis of Vitamin D 18 William Graham

Ketone Transpositions 190 Ann L. Crumrine

- 125 -

REACTIONS OF ALKYL ETHERS INVOLVING n-COMPLEXES ON A REACTION PATHWAY

Reported by Jerome T. Adams February 2k 1972 The n-complex (l) has been described as an intermediate on the reaction pathway for electrophilic aromatic substitution and acid catalyzed rearrange- ment of alkyl aryl ethers along with sigma (2) and pi (3) type intermediates. 1 ' 2

+xR

Physical evidence for the existence of n-complexes of alkyl aryl ethers was found in the observation of methyl phenyl ononium ions by nmr 3 and ir observation of n-complexes of anisole with phenol. 4 In an effort to understand the rearrangement of alkyl aryl ethers, product studies have been used to investigate the mechanism. Two mechanistic pathways have been suggested, one involving intramolecular transfer of an alkyl group and the other intermolecular transfer. 1 ' 2 Evidence in support of the intramolecular transfer is (a.) the high ortho yield observed in the rearrangement of benzyl phenyl ether, 5 Gc-phenethyl phenyl ether, and sec-butyl phenyl ether,"7 which give respective yields of 100$,

7, and 92$ of the rearranged products, and (b) the partial retention of configuration observed in the rearrangement of optically active sec -butyl 2 ' 8 6 phenyl ether with and a-phenethyl tolyl ether with A1B ? . Evidence for intermolecular reaction is (a) the rearrangement of optically active sec -butyl phenyl ether with AlBr 3 to products with inversion of configuration8 and (b) the observation of alkyl side chain rearrangement in the products of the reaction of 2-, 3-, and 4-octyl phenyl ethers with 9 7 BF 3 . The reactions of 2-, j- , and k- octyl phenyl ethers result in product distributions of similar character with respect to both side-chain positional rearrangement and o:p ratios which are in the range of 2.2-3.0 with the ortho showing greater side-chain rearrangement. Alkylation of alkyl aryl ethers has been compared to the acid catalyzed rearrangement of alkyl aryl ethers and in the case of anisole, the n-complex has been used to explain the relatively high o:p ratio observed in the 10 alkylation of anisole with A1C1-,- methyl chloride in nitromethane. Evidence for the intermediacy of an n-complex in an electrophilic aromatic substitution is the isotope incorporation in the product and the starting material in the alkylation of anisole with methyl chloroformate-d ? and

3 -1 AgSbF6 in chlorobenzene. Scheme 1 is consistant with the formation of an n-complex (k) , and although an n-complex is involved in the mechanism a normal o:p ratio of 35:60 was observed. Recovered anisole contained 13$ anisole-0-methyl-d 3 from dealkylation of h, and methyl chloroformate-do recovered showed no incorporation of a protio-methyl. In further experiments with methyl chlor oformate-d.^, anisole, and AgSbF6 designed to investigate methyl-d 3 incorporation in the ortho and para products, the ratios of isotope incorporation for the ortho and para isomers were the same within the present limits of experimental error (Table l). 12 This suggests that the ortho and para products are derived from similar processes. If the para substituted 2 products come from intermolecular reaction as has been suggested, ' then it is likely in this case that the ortho products also come from intermolecular reaction. - 126 -

Table 1 Percent of Methyl Anisoles 6 7 8

ortho 52.7 33-0 9.7 5-2

para 4-y.b 3^.0;>4.u y.q- 7.0 This recent work is an example of an electrophilic aromatic substitution with n-complex formation that seems to involve neither high ortho yields nor intramolecular reaction as a major reaction pathway. In this light it seems as though the previous criterion for the formation of n-complexes on the reaction pathway of electrophilic aromatic substitutions is not sufficient,

-CD.

Scheme 1

CD. + CH 3 CD3OCOCI k, ^>

BIBLIOGRAPHY

1. H. J. Shine, "Aromatic Rearrangements", Elsevier Publishing Co., New York, N. Y., 1967* PP 82-89 and references cited therein. 2. a) M. J. S. Dewar, "Molecular Rearrangements", P. deMayo Ed., John Wiley and Sons, Inc., New York, N. Y., 1963* PP 313-318 and references cited therein; b) idem, p 31^- 3. W. P. Meyer, University of Illinois Seminar, p 262, Summer 1971 and references cited therein; C. M. Brouwer, E. L. Mackor, and C. Maclean, Rec. Trav. Chem. Pays-Bas, 85, 109, 11^ (1966). k. B. B. Wayland and R. S. Drago, J. Amer. Chem. Soc, Qk, 3817 (1962). 5. C. S. Tarbell and J. C. Petropoulos, ibid., jk, 2^"[l952). 6. H. Hart and R. J. Elia, 76, 3031 (195*0- 7. M. J. S. Dewar and N. A. Puttnam, J. Chem. Soc, ^090 (1959). 8. P. A. Spanninger and J. L. von Rosenberg, Chem. Commun., 795 (1970). 9. H. R. Sonawane, M. S. Wadia, and B. C. Subba Rao, Indian J. Chem., 6, 19^ (1968). 10. P. Kovacic and J. J. Hiller, Jr., J. Org. Chem., 30, 1581 (1965). 11. D. A. Simpson, Ph. D. Thesis, University of Illinois, Urbana, 111., 1970; D. A. Simpson, S. G. Smith, and P. Beak, J. Amer. Chem. Soc, 92, 1071 (1970). 12. P. Beak and J. T. Adams, unpublished data. - 127 - NEW SYNTHESES OF HELICENES

Reported by Alan Morrice February 28, 1972 INTRODUCTION Newman introduced the name "helicenes" to denote the series of ortho- condensed aromatic hydrocarbons where the benzene rings are angularly annelated in such a manner as to give a helically shaped molecule. The prefixes penta, r hexa, hepta are used to denote the ), 6, 7 ring compounds, respectively, and a shorthand way of representing this feature is, for example, [6]helicene for hexahelicene (l). Wynberg has introduced the term "heterohelicenes" in referring to those helicenes where one or more of the benzene rings have been replaced by heterocyclic rings. As examples of inherently dissymmetric OIQ chromophores these compounds are of great interest to chemists since their optical properties can O provide the basis for comparison between different theoretical treatments. We will consider primarily the syntheses that have been developed which utilize OIQ a photoinduced cyclodehydrogenation of an arylolefin to give the helicene framework; however, more Hexahelicene 1 classical methods of synthesis will also be considered. HELICENES

Although pentahelicene was synthesized as early as 1938? 1 stereochemical appreciation of the helicenes can be said to date from Newman's classical synthesis of hexahelicene in 1956. 2 The long sequences involved, however, did not lend themselves easily to the synthesis of the higher homologs. Wood and Mallory3 showed that the well known photocyclization of cis- stilbenes to dihydrophenanthrenes could be used to prepare the fully aromatic phenanthrene if carried out in the presence of an oxidant, e_. g. , iodine. R. H. Martin and his colleagues extended this procedure to the synthesis of helicenes by the cyclization of appropriate arylethylenes. Acylation of phenanthrene 4 with 2 under Friedel- Crafts conditions gave the ketone _3 and presumably the 2- isomer as well, 5 although no mention was made of the latter. Reduction

(LiAlH4 ) followed by dehydration (HCOgH) gave the diphenanthrylethylene k (Scheme l), without statement of the yields.

Scheme 1

CH 2CO CH=CH CH2C0C1

6 A much more general approach is shown in Scheme 2. The phosphonium salt 8 could be condensed with benzaldehyde, 2-naphthaldehyde, and 3- phenanthraldehyde to give, each in good yield, the arylalkenes 9, 10, and 11, respectively. Alternatively, advantage could be taken of some interesting work recently reported by Bestmann and co-workers. 7 Treatment of compound 8 with sodamide in liquid ammonia gave the corresponding alkylidenephosphorane - 128 -

Scheme 2

( + © CH2PPh 3Br

(a) R = CH3

(b) R = C0 2 CH3

H" H (a) 1. NBS Of lQ) (b) 1. LiAlH4 2. PPh3 2. PBr3 3- PPh3 ©0 ^ CH PPh Br 2 3

s

11 12 which, in the presence of molecular oxygen, underwent self condensation to the alkene 12. The formal mechanism is believed to involve initially oxidation of the ylid to the aldehyde which then condenses with another molecule of the ylid to give the alkene in a normal Wittig reaction. Diarylolefins could be synthesized either via a double Wittig reaction8 or by a Siegrist 9 ' 10 reaction (Scheme ~y). This latter reaction proceeded in good yield (89/ ) and could be extended to the synthesis of heterocyclic stilbenes. The cyclizations were carried out using the c is/trans mixtures obtained from the above reaction schemes in benzene solution containing a few crystals of iodine. A medium pressure mercury lamp was used with a Pyrex well. The - 129 -

Scheme 3

OHC

13

HoC Siegrist

g> 2PhCH=NPh

DMF, KOt-Bu

15

times of irradiation ranged from 2 to 8 hours. The cyclizations proceeded, presumably via the cis- isomers, to give the expected helicenes as products and are summarized in Table I. The synthesis of hexahelicene from lU via 15 is a good example of the efficiency of these syntheses, the product being obtained in 53% overall yield (yields given in Table I are only for final step),

The structures of the products were shown by a combination of nmr (both -""H 11 ' la ' and -^cO, uv, mass spectra, and deuterium labelling. 8 13 Also the high optical activities of those products that were resolved were an indication

Table I Photocyclizations

J.kene Helicene* 1o Yield

9 6 (1) 80

15 6 (1) 60

k 7 (16) 13

10 7 (16) 20

11 8 (IT) 62

12 9 (-) k8

13 13(18) 52

18 *See Note at end of discussion section - 130 -

of their structure. The same method has been used by Laarhoven 14 to prepare, inter alia, the interesting double helicene 19 in unspecified yield. This compound can exist in any of three forms of which the d_ (or l) and the meso are shown (Scheme h). In a detailed study of the products isolated ( from several of these reactions,15 ' ls Laarhoven found that photodehydrocyclization will occure only if, for the atoms involved in bond formation, the sum of the free valence numbers, calculated for the molecule in its first excited state " by the Hiickel method, 1 7 exceeds 1.0.

Scheme k

a (i) meso 19

An interesting extension of this work has been done by the Moradpour group who, by carrying out the irradiations in the 290-370 nm region with light circularly polarized at 313 nm> obtained products possessing a small though significant optical activity. 18 Similar results have been obtained independently by Calvin's group19 and represent an example of a true "asymmetric synthesis." HETEROHELICENES By a Wittig reaction of compound 8 with 2-formylpyridine and subsequent cyclization of the alkene in benzene solution, Martin and Deblecker synthesized 20 the heterohelicene 2_0 in l8yo yield. Wynberg and his group have synthesized a series of heterohelicenes containing thiophene rings, and a brief review of this work has recently been published. 21 Research in this area was stimulated by the discovery of Wynberg that dithienylethylenes will cyclize uniquely and in high yield to benzodithiophenes (Scheme 5)- The synthesis of the heterohelicenes 29 and 32 is illustrative of his approach (Scheme 6).

Naphtho [2,1- b] thiophene 2_2 could be lithiated and then formylated to give, as sole product, the aldehyde 23. The locus of thiophene metalation is preferentially the o-position, if this is free. Since no exception has yet been found, the reaction is a useful tool for unambiguous synthesis. Formation of the chloromethyl derivative 2h was

Scheme 5

/

S' CH=CH

21 131 -

Scheme 6

1. nBuLi

H C1 2. FhNMe CHO 2

I CHO 23 2k 1. LiAlH4 2. SOCl 2 21

POCl3 CHO Vilsmeier

I PhNMe © CHO e CH 2PPh 3 Cl

25

OHC CHO

26 +

25 + 26

27

OHC

1. nBuLi

2. PhNCHo

I CHO

29

hv

31 - 132 - achieved by initial reduction of 2_3 to the alcohol followed by hydroxyl substitution with thionyl chloride. This sequence was followed since direct chloromethylation of 22 gave a complex mixture of products. Formation of the phosphonium salt 25 was then effected with triphenylphosphine. Lithiation followed by formylation of 21 gave a mixture of the mono- and dialdehydes 26 and 27. A Vilsmeier reaction, on the other hand, gave 26 as sole product. Irradiation of the alkene 28 as a cis/trans mixture with a high pressure mercury lamp with a quartz filter gave the heptaheterohelicene 29 in h^o yield. Advantage could then be taken of the terminal thiophene ring to form the aldehyde JO which gave alkene 31, and cyclization as before gave the undecaheterohelicene _32. Note that the cyclizations are unambiguous; only one product is possible. A slightly different approach is exemplified by the synthesis of 3^ by a double photocyclization ol' the alkene 33 (Scheme 7). Further work has led to two more synthetic schemes which require a minumum number of steps and are adaptable to large scale production. 23

Scheme 7 s-~~\v

Wittig ^ © 27 CH 2PPh 3 Cl

5^

OTHER SYNTHESES

The most general of the non-photoinduced procedures is that of Bogaert- Verhoogen and Martin, -*PA who prepared a series of penta- and hexahelicene- carboxylic acids, e.g. hexahelicene-8-carboxylic acid 36, from the corresponding bromoarylacrylic acids, e>g. _35> "by a Hewett reaction (K fusion, 240-270°) as shown in Scheme 8.

Scheme 8

*

CO^H CO^ 35 5i

Tauber and Vogel 26 isolated, in unstated yield, the N- containing heterohelicenes 37 and 38 from the reaction mixture obtained upon treatment of 1, 4-cyclohexanedionebisphenylhydrazone with ethanolic sulfuric acid. Zander and Franke prepared, in 6$ yield, 5, 10- dihydrocarbasolo[3, 4-£]carbazole 39 by a double Bucherer reaction

27 ' on 2, 7-dihydroxynaphthalene. This compound was first reported in 2a 1927. The synthesis of the phenanthrofuran kO in 8 steps has been reported. 29 - 153 -

37

RESOLUTION AND RACEMIZATION Hexahelicene 1 was resolved by Newman using 2- (2,4, 5> 7-tetranitro-9-fluorenylideneaminob'xy)- propionic acid (TAPA) and was found to racemize partially upon melting. This same compound was also used recently to resolve pentahelicene, which was found to racemize in solution at around room temperature. 3 Martin6 managed to ko resolve heptahelicene by mechanical selection of individual crystals obtained by slow recrystallization from benzene and to confirm the separation by measurement of their optical activities. Wynberg and Groen 31 resolved several sulfur- containing heterohelicenes with the crystal picking techniques of Martin and also succeeded in resolving 29 by recrystallization from (- )-a-pinene. These heterohelicenes were found to be less optically stable than their benzo counterparts. Chromatography of 37 an<^ 3§_ on acetylcellulose, elution with benzene, gave fractions having low optical activity. 2S PHYSICAL PROPERTIES A large amount of work has been done on the emission and excitation 32-41 spectra of hexahelicene and, with the advent of convenient and rapid syntheses of the higher helicenes, hepta-, octa-, and nonahelicene have also been studied. 40 The mass spectrum of hexahelicene has been analyzed, 42 and polarography studies on the series hexahelicene through nonahelicene have been published. 43 STURCTURE AND ABSOLUTE CONFIGURATION The terms right-handed and left-handed are used to describe the sense of screw of a helix, and enantiomers with right- and left-handed helicity are labelled P and M respectively. 44 As example, the M-enantiomer is represented in structure 1. Fitts and Kirkwood, 45 using Kirkwood's polarizability theory of optical activity, showed that six benzene rings disposed in the form of a helix with right-handed chirality should have a positive rotation. Moscowitz 46 showed that the rotatory dispersion curve can be calculated from the absorption spectrum in an application of the Kronig- Kramers Theorem. His calculations, based on the 169 tt-tt* transitions expected from Hiickel theory, implied, however, that the P-isomer of hexahelicene should be levorotatory. Tinoco and Woody, using as a model a free electron constrained to move in a helix, found that a right-handed helix should exhibit a positive rotation. 47 Brewster, using a bond polarizability model, came to the same conclusion, 48 and furthermore Brown, Kemp, and Mason, by a method similar to that of - 13^ -

Moscowitz, concluded that the (+) -isomers of benzo [c Jphenanthrene ([^]helicene), 49 penta-, hexa-, and heptahelicene should have P-chirality. Although gross structural information on penta- and hexahelicene was available from 50 51 52-54 theoretical ' and X-ray data, it was not sufficiently accurate and ( the absolute configuration of hexahelicene was only confirmed recently 53 by Trueblood's group, who showed (- )-2-bromohexahelicene to have the M- configuration by X-ray crystallography. Since this compound could be debrominated to (- )- hexahelicene, this established that (- )- hexahelicene must have M-chirality. 31 Finally, Groen and Wynberg, using essentially the same method as Moscowitz, calculated that the (+)- isomers of the thiophene- containing heterohelicenes would have the P- con figuration, a prediction borne out by the X-ray structure of a crystal of j+1 that was composed of M-helices and exhibited a negative rotation. 21 NOTE

In structures 16, 17? 18, 19; 3i> and 32 the circles representing aromaticity have been omitted for easier visualization.

BIBLIOGRAPHY

1. H. A. Weidlich, Chem. Ber., 71, 1203 (1938). 2. (a) M. S. Newman and D. Lednicer, J. Amer. Chem. Soc, 78, Vf65 (1956)} (b) M. S. Newman, R. S. Darlack, and L. Tsai, ibid ., 89 6191 (1967). 3. C. S. Wood and F. B. Mallory, J. Org. Chem., 29, 3373^196^). k. M. Flammang-Barbieux, J. Nasielski, and R. H. Martin, Tetrahedron Lett., 7^+3 (1967). 5. E. Clar "Polycyclic Hydrocarbons, " Vol. 1, Academic Press, London, I96U. 6. R. H. Martin, M. Flammang-Barbieux, J. P. Cosyn, and M. Gelbcke, Tetrahedron Lett., 3507 (1968). 7. (a) H. J. Bestman, Angew. Chem., Int. Ed. Engl., k, 83O (1965); (b) H. J. Bestmann, R. Armsen, and H. Wagner, Chem. Ber., 102, 2259 (1969). 8. R. H. Martin, G. Morren, and J. J. Schurter, Tetrahedron Lett., 3683 (1969). 9. R. H. Martin, M. J. Marchant, and M. Baes, Helv. ChLm. Acta., 5^, 358 (l97l). 10. (a) A. E. Siegrist, ibid ., 50, 906 (1967); (b) A. E. Siegrist and H. R. Meyer, ibid ., 52, 1282 (1969)5 (c) A. E. Siegrist, P. Liechti, H. R. Meyer, and K. Weber, ibid ., 52, 2521 (1969). 11. R. H. Martin, N. Defay, H. P. Figeys, M. Flammang-Barbieux, J. P. Cosyn, M. Gelbcke, and J. J. Schurter, Tetrahedron, 25, ^985 (1969). 12. N. Defay, D. Zimmerman, and R. H. Martin, Tetrahedron Lett., I87I (l97l). 13. R. H. Martin and J. J. Schurter, ibid ., 3679 (1969). Ik. W. H. Laarhoven and Th. J. H. M. Cuppen, ibid ., 63 (l97l). 15. W. H. Laarhoven, Rec. Trav. Chim. Pays-Bas, 87, 687 (1968). 16. W. H. Laarhoven, Th. J. H. M. Cuppen, and R. J. F. Nivard, Tetrahedron, 26, U865 (1970). 17. C. A. Coulson, "Valence, " 2nd ed, Oxford University Press, London, 196l, p. 271. 18. (a) A. Moradpour, J. F. Nicoud, G. Balavoine, and H. Kagan, J. Amer. Chem. Soc, 92, 2353 (1971); (b) H. Kagan, A. Moradpour, J. F. Nicoud, G. Balavoine, R. H. Martin, and J. P. Cosyn, Tetrahedron Lett., 2^79 (l97l)j (c) G. Tsoucaris, G. Balavoine, A. Moradpour, J. F. Nicoud, and H. Kagan, Compt. Rendi Acad. Sci., Paris, Ser. B, 272 (22), 1271 (l97l). 19. W. J. Bernstein, M. Calvin, and 0. Buchardt, J. Amer. Chem. Soc, 9k, k9k (1972). We are grateful to Prof. Calvin for private transmittal of this information as well. - 135 -

20. R. H. Martin and M. Deblecker, Tetrahedron Lett., 3597 (1969). 21. H. Wynberg, Ace. Chem. Res., h, 65 (l9Tl) and references therein. 22. M. B. Groen, H. Schadenberg, and H. Wynberg, J. Org, Chem., 36, 2797 (l97l) and references therein. 23. P. G. Lehman and H. Wynberg, Rec. Trav. Chim. Pays-Bas, 90, 11 3 (l97l). 2k. D. Bogaert-Verhoogen and R. H. Martin, Tetrahedron Lett., 30I+5 (1967). 25. C. L. Hewett, J. Chem. Soc, 1286 0-938).

26. H. J. Teuber and L Vogel, Chem. Ber. , 103, 3319 (1970). 27. M. Zander and ¥. H. Franke, ibid ., 102, 2728 (1969). 28. W. Fuchs and J. Niszel, ibid ., 60, 209 (1927). 29. J. N. Chatterjea and K. D. Banerji, J. Indian Chem. Soc, 1+7, 576 (1970). 30. C. Goedicke and H. Stegemeyer, Tetrahedron Lett., 937 (1970). 31. H. Wynberg and M. B. Groen, J. Amer. Chem. Soc, 9^, 2968 (l97l). 32. C. C. Meridith and G. F. Wright, Can. J. Chem., 3H7 II77 (i960). 33. M. F. O'Dwyer, M. Ashraf El-Bayommi, and S. J. Strickler, J. Chem. Phys., 36 1395 (1962). 3I+. W. Rhodes and M. F. Amr El-Sayed, J. Mol. Spectrosc. , 9, 1+2 (1962). 35. N. C. Kneten, N. J. Krause, T. 0. Carmichael, and 0. E. Weigang Jr., J. Amer. Chem. Soc, 8k, 1738 (1962). 36. T. Azumi and S. P. McGlynn, J. Chem. Phys., 38, 2773 0-963 ). 37. 0. E. Weigang Jr., J. A. Turner, and P. A. Trouard, ibid ., 1+5, 1126 (1966). 38. R. Chang and S. I. Weissman, J. Amer. Chem. Soc, 89, 5968 "(1967).

39. R. D. Allendoerfer and R. Chang, J. Mag. Reson. , 5, 273 0-971). 1+0. E. Van der Donckt, J. Nasielski, J. R. Greenleaf, and J. B. Birks, Chem. Phys. Lett., 2, 1*09 0968).

1+1. 0. E. Weigang Jr. and P. A. Trouard Dodson, J. Chem. Phys., _j+9, 1+21+8 0-968). 1+2. R. C. Dougherty, J. Amer. Chem. Soc, 90, 5788 0-968). 1+3. W. H. Laarhoven and G. J. M. Brus, J. Chem. Soc B, 11+33 0-971).

1+1+. R. S. Cahn, C. K. Ingold and V. Prelog, Angew. Chem., Int. Ed. Engl., _5, 385 0-966). 1+5. D. D. Fitts and J. G. Kirkwood, J. Amer. Chem. Soc, 77, I+9I+0 (1955). 1+6. (a) A. Moscowitz, Tetrahedron, 13, 1+8 (l96l); (b) A. Moscowitz, Adv. Chem. Phys., k, 67 (1962). 1+7. I. Tinoco Jr. and R. W. Woody, J. Chem. Phys., 1+0, 160 (196I+). 1+8. J. H. Brewster, Topics in Stereochemistry, 2, T~(l967). 1+9- A. Brown, C. M. Kemp and S. F. Mason, J. Chem. Soc A. 751 0-971). 50. (a) F. Harnick, F. H. Herbstein and G. M. J. Schmidt, Nature, 168,158 0-951); (b) A. 0. Mcintosh, J. Monteath Robertson, and V. Vand, Nature, 169, 322 (1952); (c) A. 0. Mcintosh, J. Monteath Robertson, and V. Vand, J. Chem. Soc, 1661 (195^). 51. C. A. Coulson and S. Senet, j. Chem. Soc, 1819 (1955). 52. T. Hahn, Acta. Cryst., 11, 835 (1958); C. A., 5^, 15696d. 53« M. A. Herraez Zarza and F. Sanchez, An. Real Soc Espan. Fis. Quim. , Ser. B, 61, 953 (1965). 5I+. I. R. Mackay, J. Monteath Robertson, and J. G. Sime, Chem. Commun. , 11+70 (1969). 55. D. A. Lightner, D. T. Hefelfinger, G. W. Frank, T. W. Powers, and K. N. Trueblood, Nature Phys. Sci., 232, 12l+ 0-971). I

( ,

- 136 - RECENT ADVANCES IN DRUG DETECTION AND ANALYSIS

Reported by Ronald J. Convers March 2, 1972 The many needs for drug detection and analysis today include legal identification of suspected illegal drugs taken in police raids, monitoring of drug concentrations in body fluids of persons in therapeutic programs, identification of drugs taken in cases of overdose, and assay of "street drugs" for legal purposes and for prevention of overdoses in the persons who insist on using them. These categories have different time requirements. Ideally, drug identification and quantification should be as rapid as possible in cases involving comatose overdose victims, since positive treatments can vary with the causative drug. Detection and analysis of drugs in body fluids for a large- scale screening program by a government agency should also be rapid for the sake of efficiency. Detection and analysis of drugs in dosage form and of drugs in the body fluids and tissues of dead overdose victims need not be so rapid. Drug analysis in this last case has been reviewed. 1 4 Use of drug monitoring to obtain optimum therapeutic body concentration levels in long-term treatments has been reviewed by Vesell and Passananti 5 and further discussed by Gardner-Thorpe,

6 T et _al* ^ae PurPose f this seminar will be to outline some of the older techniques and then to examine some of the recent improvements made in drug identification and quantification, particularly in relation to "drugs of abuse," or "street drugs," in human body fluids. Application of thin layer (tlc) Y 12 and gas liquid (glc) 13 l9 chroma- tography, mass spectrometry, 20 22 and combined glc-mass spectrometry- computer

2 t systems to drug identification has led to increased speed and efficiency 25 2< recently. Recent progress in methods such as ultraviolet spectroscopy, paper chromatography, 28 and color tests 29 has been less pronounced. The need to monitor illicit drug use by patients in public drug addict rehabilitation programs has led to the publication of several analytical schemes involving 30 of body extracts. method of Dole _et al. 31 drugs are tic fluid In the , adsorbed from a 50 ml urine sample onto ion exchange paper, eluted at three different pH values, and thus separated into three classes for tic resolution and sprayed color reagent identification. The ion exchange adsorption separates drugs from pigments and other materials which might interfere with the tic identification, permits sample storage in minimal space, and permits cuts in cost of analysis when several samples from the same person might be tested at the same time. In 1970, Althaus, et al., 23 reported identification of the drug, Darvon, and its metabolites in the urine of a comatose overdose victim using a glc-mass spectrometer-computer system. Law, et al., 24 have discussed subsequently their own use of this analytical technique. Other recent developments have been the application of neutron activation analysis to drugs containing halogens, 39 the design of an automated system for analysis of barbiturate, morphine, or basic drug content in body fluids, 40 and the publication of results of "street drug" assays. J BIBLIOGRAPHY

E. G. C. Clark, "Isolation and Identification of Drugs," The Pharmaceutical Press, London, 1969- C. P. Stewart and A. Stolman. "Toxicology Mechanisms and Analytical Methods," Vol. I, Academic Press, New York, i960. C. P. Stewart and A. Stolman, "Toxicology Mechanisms and Analytical Methods," Vol. II, Academic Press, New York, 1961. I. Sunshine, "Handbook of Analytical Toxicology," The Chemical Rubber Company, Cleveland, Ohio, 1969.

E. S. Vesell and G. T. Passananti, Clin. Chem. , 17, 851 (1971). C. Gardner-Thorpe, M. J. Parsonage, P. F. Smethurst, and C. Toothill, Clin. Chim. Acta, 36, 223 (1972). - 137 -

Brown, L. Shapazian, and G. Griffin, , 7- J. K. D. J. Chromatogr . 6k, 129 (1972). 8. G. F. Phillips and J. Gardiner, J. Pharm. Pharmacol., 21, 793 (1969). 9. A. M. Heaton and A. G. Blumberg, J. Chromatogr., kl, 3&f (1969). 10. K. Genest and C. G. Farmilo, J. Pharm. Pharmacol., 16, 250 (196*4-). 11. L. A. Dal Cortivo, J. R. Broich, A. Dihrberg, and B. Newman, Anal. Chem., 3§, 1959 (1966). 12. M. C. Bastos, G. E. Kananen, R. M. Young, J. R. Monforte, and I. Sunshine, Clin. Chem., 16, 931 (1970) 13. J. T. Stewart, G. B. Duke, and J. E. Wilcox, Anal. Lett., 2, kkg (1969). Ik. P. Lebish, B. S. Finkle, and J. W. Brackett, Jr., Clin. Chem., 16, 195 (1970). 15. N. M. Ryabtseva, M. I. Kuleshova, B. A. Rudenko, and V. F. Kucherov, Izv.

Akad. Na.uk SSSR, Ser. Khim. , 2676 (1970). 16. J. MacGee, Clin. Chem., 17, 587 (1971). 17. E. P. van der Slooten, H. J. van der Helm, and P. J. Geerlings, J. Chromatogr., 60, 131 (1971). 18. B. S. Finkle, E. J. Cherry, and D. M. Taylor, J. Chromatogr. Sci., 9, 393 (1971). 19. R. Luders, P. O'Brien, and F. Tischler, J. Pharm. Sci., 60, 627 (1971). 20. A. H. Beckett, G. T. Tucker, and A. C. Moffat, J. Pharm. Pharmacol., 19, 273 (1967). 21. H. M. Fales, G. W. A. Milne, and T. Axenrod, Anal. Chem., k2, 1*4-32 (1970). 22. G. W. A. Milne, H. M. Fales, and T. Axenrod, ibid., k^, 1815 (1971). 23. J. R. Althaus, K. Biemann, J. Biller, P. F. Donaghue, D. A. Evans, H.-J. Forster, H. S. Hertz, C. E. Hignite, R. C. Murphy, G. Preti, and V. Rheinhold, Experientia, 26, 71^ (1970). 2k. N. C. Law, V. Aandahl, H. M. Fales, and G. W. A. Milne, Clin. Chim. Acta, 32, 221 (1971). 1*4-5 25. H. Kupferberg, A. Burkhalter, and E. L. Way, J. Pharm. Exp. Ther., , 2k7 (196*4-). 26. A. E. Takemori, Biochem. Pharmacol., 17, 1627 (1968). 27. E. Lang, Pharmazie, 26, 2*4-5 (1971).

28. S. N. Tewari, ibid . , 26, 163 (1971). 29. R. Mechoulam, Z. Ben-Zvi, and Y. Gaoni, Tetrahedron, 2*4-, 5615 (1968). 30. S.J. Mule, Anal. Chem., 36, 1907 (196*4-). 31. V. P. Dole, W. K. Kim, and I. Eglitis, J. Amer. Med. Ass., 198, 115 (1966). 32. S. J. Mule, J. Chromatogr., 39, 302 (1969).

33. J. G. Montalvo, Jr., E. Klein, D. Eyer, and B. Harper, ibid . , kf, 5^2 (1970). 3*4-. J. M." Fujimoto and R. I. H. Wang, Toxicol. Appl. Pharmacol., 1§7 186 (1970). 35. S. J. Mule, J. Chromatogr., _55, 255 (1971).

36. R. C. Baselt and L. J. Casarett, ibid . , 57, 139 (1971).

37. K. K. Kalstha and J. H. Jaffe, ibid. , 60, 83 (1971). 38. N. Weissman, M. L. Lowe, J. M. Beattie, and J. A. Demetriou, Clin. Chem., 17, 875 (1971). 39. M. Margosis, J. T. Tanner, and J. P. F. Lambert, J. Pharm. Sci., 60, 1550 (1971). kO. D. J. Blaokmore, A. S. Curry, T. S. Hayes, and E. R. Putter, Clin. Chem., 17, 896 (1971). *4-l.- Chem. Eng. News, 15 (1972). *4-2. F. E. Cheek, S. Newell, and M. Joffe, Science, 167, 1276 (1970). *43. "Drug Atlas," Midwest Research Institute, Kansas City, 1971. - 138 - THE STRUCTURE OF CARBONIUM IONS WITH MULTIPLE NEIGHBORING GROUPS

Reported by William J. Work March 6, 1972 While neighboring group participation in carbonium ion reactions has constituted one of the most widely studied phenomena in recent years, very little research has been done with carbonium ions which have more than one neighboring group. Carbonium ions with two neighboring groups should be of particular interest since it might be possible to observe them as an intermediate

N = some nucleophile

similar to 1, or, at least, they would serve as a good model for S 2 reactions. The bonding in 1, by analogy with inert gas halides, can be thought of as the overlap of the unoccupied p-orbital on carbon with the electron-rich orbitals of the nucleophiles. 1 Such bonding to carbon has been hypothesized for two different iron carbonyl carbides. 2 ' 3 Temperature dependent nmr has been used to determine the rate constants for rearrangement of carbonium ions with multiple neighboring groups. Nuclear 4 magnetic resonance studies of ortho- substituted trityl cations (2), acetoxonium 5 " 10 11_ and pivaloxonium ions (3), thioniaaceanthrene cations (4), and " " 7-substituted norbornadienyl cations, 12 1 7 have lead to the~~conclusion that they undergo degenerate rearrangements with varying degrees of facility. The rate of

SMe rearrangement was found to depend upon the relative stabilities of the positive charge on the carbonium carbon and on the neighboring groups. Cation stabilizing substituents on the carbon in question increased the rate of rearrangement, as did cation destabilizing substituents on the neighboring group. In the case of the 7-methoxynorbornadienyl cation, the stability of the C-7 positive charge was such that the energy barrier for "bridge- flipping" was was so low that the authors hypothesized that simultaneous coordination to both double bonds was being observed. 17 If the interpretation be accepted, then this is the first example of an organic molecule with a four- electron, three- center bond to carbon in it. This means that the pentacoordinate structure 1 can exist as an intermediate rather than the commonly accepted transition state. Molecular orbital calculations of the potential energy curve for the have shown the 7-norbornadienyl cation that most stable" conformation is one with C-7 tilted towards one of the double bonds. 18 20 Several studies have been made on the effect of a second neighboring group on the rate of solvolysis reactions. Comparison of the solvolysis rates for l-methoxypentane-3-brosylate and l,5-dimethoxypentane-3-brosylate in acetic - 139 - acid showed a rate retardation for the dimethoxy compound. 21 The authors explained this result as being due to a dipolar field effect of the second oxygen -which destabilized the positive charge developing at the methoxy group. However, the result could be equally well explained by steric hindrance to attack by the solvent molecule if the second methoxy group were coordinated to carbon undergoing attack. Similar results were obtained for the solvolysis of mono- and dimethoxybutane-4- tosylates. 22 For this case, it seems much more likely that the rate retardation observed for the dimethoxy compound is due to a dipolar field effect since the methoxy groups are in much closer proximity. Finally, solvolyses of derivatives of 7-hydroxynorbornadienes have been found to proceed about 10 3 faster than those of 7-bydroxynorbornenes. 23 This result was explained by the interaction of the second double bond with the developing positive charge at C-7 in either a symmetrical of unsymmetrical manner. The latter would seem to be the more likely explanation in view of the charge repulsion between the developing negative charge on the leaving group and the electrons in the second double bond.

BIBLIOGRAPHY

1. R. E. Rundle, J. Amer. Chem. Soc, 85, 112 (1963). 2. E. H. Braye, L. F. Dahl, W. Hubel, D. L. Wampler, ibid ., 84, 4633 (1962).

3. M. R. Churchill and J. Wormald, ibid . , 93, 3073 (1971). 4. R. Breslow, L. Kaplan, and D. LaFollette, ibid ., 90, 4056 (1968). 5. H. Paulsen and H. Behre, Angew. Chem. Int. Ed., 87~886 (1969) 6. H. Paulsen and H. Behre, ibid ., 8, 887 (1969). 7. H. Paulsen, H. Meyborg, and H. Behre, ibid., 8, 888 (1969). 8. H. Paulsen, Chimia, 24, 290 (1970). 9. H. Paulsen and H. Behre, Chem. Ber., 104, 128l (1971). 10. H. Paulsen and H. Behre, ibid., 104, 1299 (1971). 11. R. J. Basalay, Ph.D. Thesis, University of Illinois, 1971- 12. P. R. Story, L. C. Snyder, D. C. Douglass, E. W. Anderson, and R. L. Kornegay, J. Amer. Chem. Soc, 85, 363O (1963). 13. P. R. Story and M. Saunders, ibid ., 84, 4876 (1962). Ik. R. K. Lustgarten, M. Brookhart, and S. Winstein, ibid., 89, 635O (1967). 15. M. Brookhart, R. K. Lustgarten, and S. Winstein, ibid., "89, 6352 (1967). "89, 16. M. Brookhart, R. K. Lustgarten, and S. Winstein, ibid . , 6354 (1967).

17. R. K. Lustgarten, M. Brookhart, and S. Winstein, Tetrahedron Let., 1971 j 141 (1971). 18. R. Hoffmann, J. Amer. Chem. Soc, 86, 1259 (1964). 19. R. Hoffmann, J. Chem. Phys., 40, 2"£8o (1964). 20. H. Konishi, H. Kato, T. Yonezawa, Bull. Chem. Soc. Jap., 43, 1676 (1970). 21. D. S. Tarbell and J. R. Hazen, J. Amer. Chem. Soc, 91, 76*57 (1969)- 22. J. R. Hazen, J. Org. Chem., 35, 973 (1970). 23. S. Winstein and C. Ordronneau, J. Amer. Chem. Soc, 82, 2084 (i960). .

1 - 1^0 - RECENT REACTIONS OF THE DMSO-DCC REAGENT

Reported by James A. Franz March 9, 1972 This seminar -will be limited to a discussion of the reactions of sulfoxides and carbodiimides with phenols, oximes, hydroxylamines, carboxylic acids, hydroxamic acids, amides, sulfonamides and hydrazine derivatives, since the use of dimethyl sulfoxide (DMSO) and dicyclohexylcarbodiimide (DCC) or phosphorus pentoxide or acetic anhydride for the oxidation of alcohols, halides, tosylates, mercaptans, et _al. has been adequately reviewed. 1 The mechanisms invoked to explain the observed products are variants of the mechanism of the alcohol oxidation reaction, 2 which rests on firm ground and has withstood objections. 3 E Phenols react with the intermediate 1 to yield ortho-subs tituted 6 methylthiomethyl (-CH2SCH 3 ) phenols (3), recently observed benzoxathians 7 (k) , and methylthiomethyl ethers (_5). o rtho-Disubstituted or poly-substituted 8 phenols lead to dienones. The oxysulfonium intermediate 2 decomposes to products along paths not unlike the Pummerer, 9 Sommelet, 10 and Stevens 11 rearrangements

C H N:=C-NHC 6 1:L 6H1:L CgHnNHCNHCsHu CH3+ CH 3 (H+) H2P0 4 ^> > CH 3-S-CH 3 r^-.o-ioynoz

£ Nh OCH2SCH 3

Oximes of ketones react with the DMSO-DCC reagent to give N-methylthiomethyl nitrones (6) and methylthiomethyl ethers (7) (Table i). First- and second-order 12 Beckmann rearrangements (e.g., leading to "8 and 9) have also been observed. Carboxylic and hydroxamic acids yields methylthiomethyl esters (10) and N-acyl ureas (ll). In addition, hydroxamic acids yield products which are the result of Lossen-type rearrangements. Primary carboxamides react with DMSO-DCC 13 to form sulfilimines (12) and nitriles. Aryl and alkyl sulfonamides react to form sulfilimines in high yields. N-Alkyl sulfonamides form derivatives of DCC. N-Aryl sulfonamides are alkylated at the ortho positions of the aromatic ring or at nitrogen in a manner reminiscent of the reactions of DMSO-DCC with phenols. 14 Mildly basic amines give sulfilimines in good yield. Hydrazines lead to products expected from the formation of intermediate diazonium ions. 15 1 1 In some cases " these reactions with DMSO-DCC have been compared directly with reactions of the same compounds and DMSO activated by phosphorus pentoxide or acetic anhydride, in general with remarkably different results. These comparative reactions with phenols, 16 acids, 17 sulfonamides, amides, " carboxylic and amines 17 21 demonstrate the divergence of products obtained from phosphorus pentoxide- or acetic anhydride- activation of DMSO from those obtained from corresponding DMSO-DCC reactions. This is often due to a reaction of a substrate directly with DCC or directly with products of the decomposition of DMSO. - lUl - Table 1 Examples of Products of DMSO-DCC Reactions Reagent Products Reference cr

Ph C=NOH Ph2C=N-CH2SCH 3 Ph2C=N-OCH2SCH 12 2 + 3

6 (12-79$) 7 (2-hT?o)

>N f 12 1 8 (W) (hofft - II °2H CsNuNHC-N-CeHu C2N -( O 13 OCH2SCH3 o-c^y^X N02 10 ( hQffo) 11 (^)V7 PhCONH; PhC0NS(CH PhCN 3 ) 2 CeHnNCQNHCeHu PhC0NHC=NC6H i;L 13

12 ( kl%) ( 12$) CH2NHC0Ph fflC 6H11

BIBLIOGRAPHY

1. (a) J. G. Moffatt, "Techniques and Applications in Organic Synthesis: Oxidation/' Vol. 2, R. Augustine, Ed., Marcel Dekker, Inc., New York, 1969. (b) W. W. Epstein and F. W. Sweat, Chem. Rev., 67, 2^7 (1967). 2. A. H. Fernselau and J. G. Moffatt, J. Amer. Chem. Soc, 88, I762 (1966). 3- K. Torssell, Acta. Chem. Scand., 21, 1 (1967). k. K. Torssell, Tetrahedron Lett., khk5 (1966). 5. J. G. Moffatt, J. Org. Chem., 36, 1909 (1971). 6. K. Onodera, S. Hirano, N. Kashimura and T. Yajima, Tetrahedron Lett., 1^27 (1965). 7- (a) M. G. Burdon and J. G. Moffatt, J. Amer. Chem., Soc, 87, ^656 (1965). (b) K. E. Pfitzner, J. P. Marino and R. A. Olofson, ibid., 87, I+658 (1965). (c) M. G. Burdon and J. G. Moffatt, ibid ., 5855 (1966). (d) J. P. Marino, K. E. Pfitzner and R. A. Olofson, Tetrahedron, 27, 4l8l, ^195 (1971). 8. M. G. Burdon and J. G. Moffatt, J. Amer. Chem. Soc, 89, Vf25 (1967). 9. G. A. Russell and G. J. Mikol in "Mechanisms of Molecular Migrations," B. A. Thyagara.jan, Ed., Interscience Publishers, New York, N.Y., 1968, p. 57. 10. C. R. Hauser, S. W. Kantor and W. R. Brasen, J. Amer. Chem. Soc, ~~75, 2260 (1953). 11. H. E. Zimmermann in "Molecular Rearrangements," P. deMayo, Ed., Interscience Publishers, Inc., New York, N.Y., 1963, p. 378. 12. A. H. Fenselau, E. H. Hamamura and J. G. Moffatt, J. Org. Chem., ~~35, 35^6 (1970). 13. U. Lerch and J. G. Moffatt, ibid., 36, 3391 (1971). Ik. U. Lerch and J. G. Moffatt, ibid., 3E, 731^ (1971). 15. U. Lerch and J. G. Moffatt, ibid., 55, 3861 (1971). 16. (a) P. Claus and W. Vycudilik, Tetrahedron Lett., 3607 (1968). (b) K. Onodera,

S. Hirano and N. Kashimura, J. Amer. Chem. Soc, 87, 4651 (1965). ( 17. See footnotes 6 and 7 in reference lh. 18. D. S. Tarbell and C. Weaver, J. Amer. Chem. Soc, 63, 2939 (19^1). 19. P. Claus and W. Vycudilik, Monatsh. Chem., 101, 39c7 405 (1970). 20. H. Kise, G. Whitfield and D. Swern, Tetrahedron Lett., 176 (1971). 21. For a review of sulfilimines see F. Challenger, "Organic Sulfur Compounds," Ch. 29, N. Kharasch, Ed., Pergamon Press, New York, N.Y., 1961. s

- 1U2 - NUCLEOPHILIC ACYLATION

Reported by Janet Ollinger March 23, 1972 Processes which temporarily reverse the characteristic type of reactivity, e.g. nucleophilicity or electrophilicity, of an atom are synthetically useful as they broaden the utility of functional groups in synthesis. 1 The carbonyl carbon atom behaves as an electrophile in conventional acylating reactions, and a nucleophilic acylation involves reaction of a nucleophilic carbonyl carbon (R-C~=0) with an electrophile. Direct nucleophilic acylations have been observed with the reaction of the salt formed from alkyl lithiums and nickel carbonyl with a, ^-unsaturated 2 3 carbonyl compounds to give 1, *i-diones, ' and the reaction of beryllium 3 4a 3 4l carbonyls ' and lithium carbamide ' ° with ketones and aldehydes. These methods have been reviewed. 3 Indirect nucleophilic acylations involving carbanions of a "masked" carbonyl, 3 the products of which are converted to carbonyls upon hydrolysis, have been more useful. The most extensively studied of these are the 5 2-lithio-l, 3-dithianes and 2-lithio-l, 3, 5-trithianes which have been recently reviewed. 6 The steps of the nucleophilic acylation are shown below.

HS SH S S S y><^ BuLi electrophile j^Xf RCHO > R H RXLi BT R' 1 2 I

R^^^R' k hydrolysiSy 5

7" The dithiane jj>, which gains stability through d-orbital overlap, can also be prepared by the addition of alkyl lithium reagents to 2- methylene- 1, 3- dithiane6 ' 9 and to the 2-carbene. 9 It reacts with most of the electrophilic groups that normally combine with Grignard and organolithium reagents, such " 6 10 16 6 10 17~ 19 1:L 12 21 22 12 12 as halides, > carbonyls, > > epoxides, > > > imines, nitriles, s ls 6 6 19 15a~ c carboxylic acid derivatives, > oxetanes, carbon dioxide, ^ and also silyl, 15 1513 19 germanyl, ^ and stannyl halides. a, 3- Unsaturated carbonyls undergo 1, 2- addition, 24lD but nitrostyrene undergoes 1, ^-addition. Reaction of 3. in the presence of 24b 2, 4-dinitrobromobenzene causes oxidative coupling. Carbonyl compounds synthesized by this method include 1-deuterioaldehydes, 24

10a 17 ia- optically active aldehydes and ketones, ^° cyclic monoketones and diketones, > metacyclophanes, 14 acyclic monoterpenes, 13 sesquiterpenes, 20 21-ketosteroids, 22 and acyl-silanes 15 and germanes 15b 1:L ~ 5? 10^ " is 1T Reagents for hydrolysis of _3 "to h are mercuric chloride, mercuric 21 13 18 2S oxide-boron trifluoride etherate, silver nitrate, N-bromosuccinimide, ' 18 26 2S c N-chlorosuccinimide and a silver salt, > chloramine-T, "> oxidation with 1-chlorobenzotriazole followed by treatment with sodium hydroxide, 26d and bromine. ^^ 14 Conversion of h to the ketal and subsequent hydrolysis to jj has also been reported. Recently developed acyl carbanions include cyanohydrins of aldehydes protected as their ethers, which react with ketones and aldehydes; 2r nitronate anions, which give 1, 4-diones upon addition to enones and subsequent transformation of a nitro group to a carbonyl; 28 and methyl methylthiomethyl sulfoxide which produces aldehydes after alkylation and hydrolysis. 29 - 1U3 - BIBLIOGRAPHY

1. E. J. Corey, Pure and Appl. Chem. , lU, 19 (1967). 2. E. J. Corey and L. S. Hegedus, J. Amer. Chem. Soc, 91, k^ZJ (1969). I 3. D. Seebach, Angew. Chem. Int. Ed. Engl., 8, 639 (1969). k. (a) I. I. Lapkin, G. Y. Anvarova, and T. N. Povarnitsyna, J. Gen. Chem. U. F. (USSR), J56, 19^5 (1966); (b) Schollkopf and Gerhart, Angew. Chem. Int. Ed. Engl., 6, 805, 970 (1967).

5. P. Y. Johnson, Chem. Comm. , IO83 (l97l). 6. D. Seebach, Synthesis, 1, 17 (1969). 7. D. L. Coffen, T. E. McEntee, Jr. and D. R. Williams, Chem. Comm., 913 (1970). 8. R. M. Carlson and P. M. Helquist, J. Org. Chem., 33, 2596 (1968). , 9. R. M. Carlson and P. M. Helquist, Tetrahedron Lett., 173 0969 >. 10. (a) D. Seebach and D. Steinmliller, Angew. Chem. Int. Ed. Engl., 7, 619

(1968); (b) D. Seebach, D. Steinmliller, and F. Demuth, ibid., J_> 619 (1968). 11. D. Seebach, N. R. Jones, and E. J. Corey, J. Org. Chem., 33, 300 (1968). 12. E. J. Corey and D. Seebach, Angew. Chem. Int. Ed. Engl., IjJ H75 (1965). 13. C. A. Reece, R. A. Rodin, R. G. Brownlee, W. G. Duncan and R. M. Silverstein, Tetrahedron, 2k, 1+2^9 (1968). Ik. T. Hylton and V. Boekelheide, J. Amer. Chem. Soc, £0, 6887 (1968). 15. (a) E. J. Corey, D. Seebach, and R. Freedman, ibid ., 8£, kjk (1967); (b) A. G. Brook, J. M. Duff, P. R. Jones, and N. R. Davis, ibid ., 89,

kj>l (1967); (c) P. R. Jones and R. West, ibid . , 90, 6978 (1968). 16. E. J. Corey, N. H. Anderson, R. M. Carlson, J. Paust, E. Vedejs, I. Vlattas, and R. E. K. Winter, ibid ., 90, 32^5 0-968). 17. E. J. Corey and D. Crouse, J. Org. Chem., 33, 298 (1968). 18. E. J. Corey and B. W. Erickson, ibid., 36/ 3553 (l97l). 19. E. J. Corey and D. Seebach, Angew. Chem. Int. Ed. Engl., k, 1177 (1965). 20. J. A. Marshall and A. E. Greene, J. Org. Chem., 36, 2035 Tl97l). ( 21. E. J. Vedejs and P. L. Fuchs, ibid ., j>6, 366 (l97l). 22. (a) J. B. Jones and R. Grayshan, Chem. Comm., li-l (1970); (b) J. B. Jones and R. Grayshan, ibid., jkl (1970 ). 23. (a) D. Seebach and H. F. Leitz, Angew. Chem., 81, lOVf (1969); (b) W. H. Baarschers and T. L. Loh, Tetrahedron Lett., 3%Qj) (l97l). 2k. D. Seebach, B. W. Erickson, and G. Singh, J. Org. Chem., 31, 4303 (1966). 25. D. Seebach, Angew. Chem. Int. Ed. Engl., 6, kk2 (1967).

' 26. (a) R. B. Woodward, Harvey Lecture Ser. , 59, 31 (1965); (b) W. F. J. . . ;. :L Huurdeman, H. Wynberg, and D. W. Emerson, Tetrahedron Lett., 3^9 (l97l)j (c) D. W. Emerson and H. Wynberg, ibid ., 3kk5 (l97l); (d) P. R. Heaton,

, J. M. Midgley and W. B. Whalley, Chem. Comm. 75O (l97l); (e) W. H. Truce and F. E. Roberts, J. Org. Chem., 28, 96I (1963); (f) B. W. Erickson, Ph.D. Thesis, Harvard University, 1970. 27- G. Stork and L. Maldanado, J. Amer. Chem. Soc, 93, 5286 (l97l)- E. 28. J. McMurry and J. Melton, ibid . , 9^, 5309 (1971 )• 29. K. Ogura and G. Tsuchihashi, Tetrahedron Lett., 3I5I (1971 )• . - Ikk - THE CHEMISTRY OF CAMPTOTHECIN

Reported by Dale Pigott April 3, 1972

Camptothecin (l), a. novel pentacyclic alkaloid, was first isolated in 1 a tree native to mainland China, "by Wall et al. I965 from Camptotheca acuminata , "^ in a program seeking natural antitumor agents. Its structure was determined1 by X-ray crystallographic studies made on the iodoacetate derivative.

CH 3 2 ,

Because of the reported activity1 of camptothecin and its sodium salt against leukemia L1210 in mice and Walker 256 intramuscular tumor in rats, there has been intensive investigation of its antitumor properties in the past several years, 3 8 and the activity which was originally reported has been confirmed. 3 ' 4 Although no specific proposals for the mechanism of camptothecin action have yet appeared, several researchers 5 ,e have noted that camptothecin and its sodium salt inhibit

both RNA and DNA synthesis in vivo . Tests with analogues have thus far revealed no other compounds which possess this. activity. 6 BIOGENESIS

9 ' Possible biogenesis of this unique skeleton has received some attention. 10 It seems probable that camptothecin arises from an indole alkaloid precursor via a mechanism such as that shown in Scheme I. Geraniol has been shown to be a precursor for the non- tryptophan portion of the alkaloid ajmalicine (_2) which contains the same non-tryptophan backbone (dark lines) as camptothecin. Wenkert has suggested a conversion from an intermediate similar to ajmalicine which would 10 explain the origin of this terpenoid skeleton; an analogous conversion is known to occur between other alkaloids. 11

Scheme I *15S SYNTHESIS quickly appeared describing syntheses of analogues Numerous publications~ have of camptothecin10 > 12 18 and syntheses of intermediates in proposed total synthetic 19~29 pathways Recently two total syntheses of (+.) camptothecin have also appeared. 30 ' 31 The first of these is illustrated in Scheme II as reported by Stork and Schultz. 30 Danishefsky et al. have also published the total synthesis shown in Scheme III. 31

Scheme II

OAc

(±) 1 ->- -? > > > COoH — EtO COvMe EtO

! c EtO m, > I:CtcA^^

I COoMe MeO-.C

+ 1 -^ > ( ) COoEt

CO^Me BIBLIOGRAPHY

M. E. Wall, M. C. Wani, C. E. Cook, K. H. Palmer, A. T. McPhail, and G. A. Sim, J. Amer. Chem. Soc, 88, 3888 (1966). 2. A. T. McPhail and G. A. Sim, J. Chem. Soc. (B), 923 (1968). 3- J. M. Venditti and B. J. Abbott, Lloydia, 30, 332 (1967). ibid k. R. E. Perdue, M. E. Wall, J. L. Hartwell, and B. J. Abbott, ., 31, 229 (1968). Mol. Pharmacol., 632 ( 1971 J- 5- S. B. Horwitz, C. K. Chang, and A. P. Crohlman, 7, 6. B. Hacker, A. S. Kende, and T. C. Hall, Fed. Proc, 30, A627 (1971). 2h6, 7. D. Kessel, Biochim. Biophys . Acta, 225 (1971). io, and J. B. Block, 8. J. A. Gottlieb, A. M. Guarino, J. B. Call, V. T. Oliver Cancer Chemother. Rep., 5^ ]+6l (1970). (1968). 9- M. Shamma, Experientia, 2J£, 107 Chem. Soc, 10. E. Wenkert, K. G. Dave, R. G. Lewis, and P. W. Sprague, J. Amer. 89, 67^1 (1967). n n^^^.Qr.cj-? u t m^+ottti a, Wfll spv . pnd L. J. Durham, ibid., 00, 1792 / (1966). 12. Mc Shamma and L. Novak, Tetrahedron, 25, 2275 (1969). 13- J. A. Beisler, J. Med. Chem., Ik, lll6~~( 1971) • Uk E. Winterfeldt and H. Radunz, Chem. Commun., 37^+ (1971). National 15. R. H. Wood, P. G. Hoffman, and V. S. Waravdekar, Abstracts, l62nd Meeting of the American Chemical Society, Washington, D.C., Sept 1971* No. ORGN 088. 16. H. J. Teague, Diss. Abstr. Int., 31, l8to (1970). 17. C. V. Grudzinskas, ibid ., 32, 1U49B (1971). 18. M. E. Wall and M. C. Wani, Abstracts, 153rd National Meeting of the American Chemical Society, Miami Beach, Fla., April 1967; No. M^. Levine, 19. J. A. Kepler, M. C. Wani, J. N. McNaull, M. E. Wall, and S. G. J. Org. Chem., jk, 3853 (1969). 20. M. C. Wani, J. A. Kepler, J. B. Thompson, M. E. Wall, and S. G. Levine, Chem. Commun., kOk (1970). 21. M. Shamma and L. Novak, Coll, Czech. Chem. Commun., 35, 3280 (1970). 22. T. Kametani, H. Nemoto, H. Takeda, and S. Takano, Tetrahedron, 26, 5753 (1970). (1971). 23. T. K. Liao, W. H. Nyberg, and C. C. Cheng, J. Heterocycl. Chem., 8, 373 (C), 2k. L. H. Zalkow, J. B. Nabors, K. French, and S. C. Bisarya, J. Chem. Soc, 3551 (1971). l60th National 25. A. S Kende, R. W. Draper, I. Kubo, and M. Joyeux, Abstracts, Meeting of the American Chemical Society, Chicago, 111., Sept. 1970, No. URGN 10. 26. M. C. Wani, H. F. Campbell, G. A. Brine, J. A. Kepler, M. E. Wall, and Chemical S. G. Levine, Abstracts, lo2nd National Meeting of the American Society, Washington, D.C., Sept. 1971, No. MEDI 51. 27. J. B. Nabors, Diss. Abstr. Int., 51, 125B (1970). 28. T. A. Bryson, ibid ., 31, W7B (1971). 29. J. F. Eggler, ibid., £2, 38^81 (1972). hoik (1971). 30. G. Stork and A. G. Schultz, J. Amer. Chem. Soc, 93, ibid., fb p. vmkmann. S. Danishefsky, J. Eggler, and D. M. Solomon, 93, 55 U9f±;. - ll+6 -

STEREOSELECTIVE SYNTHESES AND ABSOLUTE CONFIGURATION OF CECROPIA JUVENILE HORMONE

Reported by John C. Greenfield April 13, 1972 INTRODUCTION AND SCOPE

The molting and metamorphosis of insects are controlled by the elaboration of specific hormones. One hormone, termed the juvenile hormone (JH), occurs in some insects and is responsible for the maintenance of juvenile characteristics. This hormone is produced during the larval stages of development, and the cessation of its production causes molting and adult development. The first JH to be isolated in pure form was obtained by Roller and Bjerke 1 in 1965 from the abdomina of the giant silk moth Hyalophora cecropia. Elucidation of its structure and its subsequent synthesis were 2 reported by Trost and coworkers who assigned the structure la,- although the absolute configuration of the cis epoxide was not determined"? Further work has determined that the material extracted from cecropia is a mixture of ca.

90 mole $ of methyl (E,E)-(lORJ,llS)-( + )-10,ll-epoxy-T-ethyl-3,ll-dimethyir2,6-

OCH3

la R = C 2H 5

lb R = -CH3

tridecadienoate la, amd 10$ of methyl (E,E)-(lOR,llS)-(+)-10,ll-epoxy-3,7,ll- trimethyl-2,6-tridecadienoate lb. Following the work of Trost, some 16 other syntheses of la and lb of varying stereoselectivity and utility have appeared in the literature. This seminar will present those synthetic routes, ©ften novel, to JH which are highly stereoselective and will indicate how the 10,11 oxirane absolute configuration has been determined. The early work regarding the isolation, structural determination, and synthesis of cecropia JH has been 3-5 extensively reviewed. Also, the synthesis and bioassay of structural analogs of insect juvenile hormones have recently been reviewed6 ' 7 and will not be presented. STEREOSELECTIVE SYNTHETIC APPROACHES The stereoselective assembly of the carbon skeleton of JH requires the formation of two trans trisubstituted unsaturation centers, 8 and the formation of the cis epoxide. The first stereoselective synthesis of la was by Corey, 9 who had shown that vinyl halides could be alkylated with lithium dimethyl copper. 10 Since vinyl halides could be prepared stereospecifically from acetylenes, a new stereospecific synthesis resulted. This alkylation appeared twice in the first Corey synthesis (Scheme I). The propargylic alcohol 2 was converted to the

vinyl iodide _3, which then was treated with lithium diethyl copper and ethyl

iodide to form k in ca . 50$ yield. Compound k, uncontaminated with any of the cis product, was extended by reaction with PBr 3 followed by 3-lithio-l-trimethyl-

silyl propyne and was then converted to the alcohol _5. Compound j? was converted

to the vinyl iodide and alkylated stereospecifically (97$ trans ) to form the allyl alcohol 6 in 50$ yield. Oxidation/esterification at C-l with Mn02 /kaCN in methanol was followed by selective cis epoxidation at C-10, 11 to yield d,l-la, which was identical with the natural material. A later modification11 of this procedure was published which simultaneously produced both of the trans trisubstituted olefinic centers of k from a diyne precursor. The use of diaLkyl 12 Cu-Li reagents was also used by Anderson and coworkers ' to prepare analogs of )

- 11*7 -

Scheme I Corey's First Synthesis

=C-CH2 OH >

3 X = I 5 X = -C 2PHn 5 4

v

-6 ^ ^A d, 1 - la

JH stereospecifically, and by Siddal, Biskup, and Fried11 to prepare the C-l8 JH from an acetylenic ester. 13 W.S. Johnson first approached the synthesis of JH with a. modification of the Julia rearrangement of cyclopropyl carbinols. 14 Johnson found that trans olefins could be prepared with 90-95$ stereoselectivity by rearrangement of cyclopropylalkyl bromides with a ZnBr2 catalyst. The ester 7 (Scheme II 0° when treated with ZnBr2 in Et2 at produced the homoallylic bromo trans ,

Scheme II Johnson Synthesis via Julia Rearrangement

OOMe » Br OOMe

V

COOMe OOMe <^

la OOMe -> &> 1 ~ - ll*8 -

trans ester 8 in 87$ yield, with 95$ of the product having the desired configuration. Compound 8 was converted to the iodide, and then treated the lithium salt of 3>5-heptadione to produce 9 in ^5$ yield. Conversion of 9 to the chloroketone 10 was effected with LiCl/CuCl2 followed by Ba(0H) 2 , which was then alkylated with excess methyl magnesium bromide to form the chlorohydrin 11 by the selective Cornforth15 method. The cis -d,l epoxide of _la was then formed in methanolic K2 C03 . The final product showed

contamination with 8$ of the trans , trans , trans isomer and with 0-5$^of

the trans, cis , cis isomer as determined by glpc. An identical synthetic sequence was later used to prepare the C-17 JH lb in corresponding yield. 16 A third approach to the synthesis was the stereospecific fragmentation of a cyclic precursor. 17 The bicyclic triol 12 (Scheme III) was selectively monotosylated to form 13, which when treated with NaH in THF at 20° was cleaved stereospecifically to the olefinic ketone lA in quantitative yield. The hydroxyketone Ik was similarly cleaved to the unsaturated ketone _15,

S cheme III Stereospecific Cleavage OH

R (*"

15 12 E=-H 13 R = -OTs

after first being alkylated with methyl lithium. Compound 15 . which contains the necessary JH skeleton, was an important intermediate in Trost's earlier unselective synthesis. 3 This method of preparing trisubstituted olefins stereospecifically is limited by the availability of the requisite cyclic precursor. 8 The Claisen rearrangement of allyl vinyl ethers can be used to produce predominately trans olefins. 18 Johnson and coworkers 19 used this method for a second stereoselective approach to JH. The hydroxy ester 16 (Scheme IV), when treated with the ketal 17 in toluene at 100° for 8 hr using 2,l*-dinitrophenol as a catalyst, produced the keto ester ]_9, presumably via the vinyl allyl ether intermediate 18. The keto ester 19 was obtained in 51$ yield and consisted 87$ of the trans isomer and 13$ of the cis . It was reduced with NaBH4 to the to the corresponding mixture of alcohols _20, which were separated by silica gel chromatography. Compound _20 was then elongated similar by treatment with J-7, and the ketone which was formed was immediately reduced to _21 in 70$ yield with no contamination of the trans , cis isomer. Compound _21 was then converted stereoselectively to _22 via an S™-i' reaction with S0C12 in hexane at 0° for 17 hr (85$ yield), but _22 contained 12$ of an impurity presumed to be the secondary C-10 chloride. The chloride mixture was treated with NaBH4 in DMS0 to displace the CI, leading to a trans , trans , cis ester which was converted to d, 1-la. Analysis of the final product by nmr and glpc showed no contaminants, and the yield of d, 1-la, from _22 was 35$. A modification of this procedure appeared later20 in which the produce in yield. chloroketal 23 was condensed with 20 (Scheme V) to 2J+ 7^$ Greater stereochemical control was attained, with 2h consisting of 98$ trans , trans isomer. However, reduction of 2h by NaBH4 to the chlorohydrin was unselective, and when treated with K2C03 /ke0H the epoxide was obtained in - 11*9 -

Scheme IV Johnson's Second Synthesis via. Claisen Rearrangement COOMe -COOMe w

16 IT

OOMe

21 22

d, 1 - la

a J:2 ratio of cis to trans isomers. This lack of selectivity decreased the overall yield of the desired cis product, although separation of the chlorohydrin diastereomers was possible by preparative tic. Scheme V

20 -> OOMe > —>a,i-ia Med OMe

23 2k 21 22 Corey's second synthesis utilized a modified selective Wittig reaction

to introduce a trisubstituted olefinic center stereospecifically. Compound 2_5 was converted to the ylid which was caused to react with the protected hydroxy aldehyde 26 (Scheme Vl) to form the normal Wittig betaine at -78°. Reaction - ISO - Scheme VI Second Corey Synthesis

CHoOTHP P(Ph) 3r+ OH 25 THP = tetrahydropyran-2-yl

CHpOTHP

d, 1-lb <- OOMe

of the betaine with 2 equivalents of sec-butyllithium in pentane at -25° followed by paraformaldehyde produced 27 in 50% yield after workup. Only the correct trans , trans, cis isomer was formed. Conversion of the 1° alcohol to a methyl by" a sequence involving LiAlH^., followed by removal of THP in 5 mM p-toluenesulfonic acid and esterification at C-l with methanolic Mn02 /faaCN, gave the triene ester 20 (6o^)« Selective terminal epoxidation gave d, 1-lb in 6Cffo yield. A Wittig reaction at the C-7 aldehyde of 27 with CH2*P"["Ph) 3 gave the vinyl derivative which could be reduced with diimide to ultimately form d, 1-la. Van Tamelen'^ 3 has recently published yet another stereoselective synthesis of C-18 JH. His strategy employed available terpenoids of the correct stereochemistry as starting materials. Geranyl benzyl ether produced the aldehyde 29 upon ozonolysis (Scheme VII), which was then treated with the Grignard reagent of 2-bromo-l-butene to form 30. Compound 30 was then treated with S0C12 in hexane at 0° for 2 hr to afford the rearranged chloride 31 of correct trans , trans geometry. VJhen 31 was treated with the anion of cis -2- methyl-2-ethylallyl phenyl, sulfide at -78° in THF for 1 hr, the trans, trans, cis

0CH2 Ph 0CH2 Ph Scheme VII Stereoselective van Tamelen Synthesis MgBr 0HC

29

Ph-S-

CH2 Ph CH2 Ph

d, 1-la 33 >

- 3.51 -

Okylation product 32 was isolated in T.> u Ld. Cleavage of the thioether md benzyl ether was L:i -78° effected with / b to form the trienol 33 md this was converted to the ester via Corey's method, followed by elective terminal epoxidation to form d, 1-la. Yields were reported as •?0?o for all steps except 31 •-> 32 (75^7, although the intermediates were iot purified or characterized. Finally, another stereoselective synthesis of racemic la has been 2 eveloped. This procedure utilizes s< of the above stereoselective eactions and is reported to b nenable to large scale preparations. Compound !0, prepared by a four-step route from t_] methyl geranate in high yield, ndergoes the Claisen rearrangement with trimethyl orthoacetate at 110° to 'orm 3h in 85$ yield (Scheme VIII ), 96°/o of the product being the desired all tbtis diester. Selective reduction of the unconjugated ester function was heme VIII COOMe

7 20 )K R = - COOMe ^5 R CH2 0H CHO XYA CH=C OMe

COOMe — OOMe

d, 1 la

ffected with LiAUi 4 in .Et 2 o/l'HF at -78°to the corresponding alcohol 35 lich in turn was oxidized with chromic acid-pyridine complex to xorrn the Ldehyde 36 (90$). Wittig reaction of 36 with the appropriate phosphonium Lide formed 37 in 6ofo yield, which was then chlorinated with N-chloro- iccinimide to form the chloroketone 38 in 6ofo yield. Compound 38 was treated Lth methyl magnesium bromide to give the chlorohydrin 39 (82$ threo; lg % erybhro ) . The previously reported high stereoselectivity (95/j"threo ) ruld not be attained. The diastereomeric chlorohydrin mixture was treated r Lth methanolic K2C0 3 to yield la (82$) plus the trans isomer (l8 ')„ sparation of these isomers via glpc was possible. An alternate preparation r 35 was also given, which led to the same isomeric product mixture. 2 A sequel " to the above procedure described the stereoselective synthesis

C-17 JH (lb) from trans , trans methyl farnesoate kO. This was

;heme IX

OOMe

OOMe

d, 1-lb - 152 -

transformed selectively into the terminal bromohydrin with NBS in aqueous THF and converted to the epoxide which was opened to the glycol kl. Cleavage with NaI04 gave the aldehyde k2. Compound k2 was converted by successive steps in analogy with Scheme VIII to yield racemic lb, again contaminated with ca. 1k% of the all trans isomer. Satisfactory analyses and spectra were reported for all intermediates in Schemes VIII and IX.

ASSIGNMENT OF ABSOLUTE CONFIGURATION OF THE C-10, 11 EPOXIDE 4 26 In 1970, Meyer and Hanzman reported that the natural cecropia JH

mixture exhibited a ( + ) optical rotation with [&]-q ~ + 7° ( +^° ) . An application 27 of Brewster's rules gave a tentative assignment of 10R, IIS for the cis epoxide, although this method of assignment is known to be unreliable for 28 29 strained rings. In a later paper, a sample of the natural C-l8 JH was found to be hydrated preferentially at the more highly substituted C-ll in

perchloric acid to form a vic-diol (5:2 thr eo : erythro ) with the configuration at C-10 retained. The diastereomeric glycols k^> were treated with + a-phenylbutyric anhydride and the esters were separated by tic on silica. The cv-phenylbutytric acid was recovered, and its ORD was measured. The recovered acid was dextrorotatory implying an R configuration at C-10 30 by Horeau's method. Since the oxirane was known to be cis , C-ll and C-12 must have the opposite chirality.and thus an S configuration at C-ll was assigned.

31 Nakanishi arrived at the same assignment in a different manner. Cis and trans epoxide isomers of synthetic JH were hydrolyzed to the glycolsTn aqueous H2S04 /THF. The cis epoxide formed the threo glycol, and the trans epoxide formed the erythro glycol, with the configuration at C-10 retained. This was established by position of 0-l8 incorporation from 18 32 H2 and from the nmr study of the [2 ]-acetonide H _ derivative kk of the glycol. A study of the ORD curves of the P?(dpm) 3 complexes of the glycols from natural JH, and comparison with model glycols of known configuration led to the assignment of 10R, IIS.

CD CD. ^*N) Me

kk

Finally, the absolute configuration of the cis epoxide has been established by synthesis 33 with asymmetric reagent?. Loew and Johnson started with an optically pure sample of chloride 2} (Scheme V) which led eventually to 2h with known configuration at C-ll. Reduction with NaBH4 gave corresponding diastereomeric chlorohydrins which were separated by preparative tic. Methanolic K2C03 produced an optically active JH. Optically purity was estimated at ca. 90^. A similar procedure by Faulkner and Peterson34 employed ,

- 153 -

OMe

hi

the condensation of optically pure 45 with 20, which led to diastereomeric glycols which were separated by tic on silica. The threo diols gave the cis epoxides, and the (+) cis epoxide of known 10R, IIS configuration was thus related to the natural isomer.

BIBLIOGRAPHY

1. H. Roller and J. S. Bjerke, Life Sci., 4, l6lj (1965). 2. K. H. Dahm, B. M. Trost, and H. Roller, J. Amer. Chem. Soc, 89, 5292 (1967). 3. B. M. Trost, Accounts Chem. Res., 3, 120 (1970). 4. C. E. Berkoff, Quart. Rev. (LondonJ, 23, 372 (1969). 5. Y. S. Tsizin and A. A. Drabkina, Usp. Khem. , 39, 1074 (1970): Russ. Chem. Rev., 6 i+98 (1970). 6. K. Slama, Annu. Rev. Biochem. , 40, 1079 (1971). 7. F. S. Doening, N. Punja, and C N. E. Ruscoe, Rep. Progr. Applied Chem., 55, 446 (1971). 8. For recent reviews of stereoselective olefin synthesis, see: (a) D. J. Faulkner, Synthesis, 175 (l97l). (b ) J. Reucroft and P. G. Hammes, Quart. Rev. (London), 25, 135 (1971). 9. E. J. Corey, J. A. Katzenellenbogen, N. W. Giljnan, S. A. Roman, and B. W. Erickson, J. Amer. Chem. Soc, 90, 5618 (1968). 10. a) E. J. Corey, J. A. Katzenellenbogen, and G. H. Posner, ibid ., 89, 4245 (1967). b) W. H. Robinson, Jr., University of Illinois Organic Seminars, November 13, I969. 11. E. J. Corey, J. A. Katzenellenbogen, S. A. Roman, and N. W. Gilman, Tetrahedron Lett., 1821 (l97l). 12. a) R. J. Anderson, C. A. Henrick, and J. B. Siddal, J. Amer. Chem. Soc, 92, 735 (1970). b) J. B. Siddal, M. Biskup, and J. Fried, ibid., ~91, I85U (1969). 13. W. S. Johnson, T. Li, D. J. Faulkner, and S. R. Campbell, ibid. , —90, 6225 (1968). 14. M. Julia, S. Julia, and S. Y. Tchen, Bull. Soc. Chim. France, 1849 (1969). 15. J. W. Cornforth, R. H. Cornforth, and K. K. Mathew, J. Chem. Soc, 112, 2539 (1969). 16. W. S. Johnson, S. F. Campbell, A. Krishnakumaran, and A. S. Meyer, Proc Nat. Acad. Sci., U. S. , 62, 1005 (1969). 17. R. Zurfltth, E. N. Wall, J. B. Siddall, and J. A. Edwards, J. Amer. Chem. Soc, 90, 6224 (1968). 18. S. K. Chung, University of Illinois Organic Seminars, April 27, 1970. 19. W. S. Johnson, T. J. Brocksom, P. Loew, D. H. Rich, L. Wertheman, R. A. Arnold, T. Li, and D. J. Faulkner, J. Amer. Chem. Soc, 92 4463 (1970). 20. P. Loew, J. B. Siddall, V. L. Spain, and L. Wertheman, Proc. Nat. Acad. Sci., U. S., 1462, 1824 (1970). 21. E. J. Corey and H. Yamamoto, J. Amer. Chem. Soc, 92, 6636 (1970). E. 22. J. Corey and H. Yamamoto, ibid ., 92, 226, 3523 "u970). 23. E. E. van Tamelen, P. McMurray, and U. Huber, Proc. Nat. Acad. Sci., U. S. 68, 1294 (1971). 24. C. A. Henrick, F. Schaub, and J. B. Siddall, J. Amer. Chem. Soc, in press. 25. R. J. Anderson, C. A. Henrick, J. B. Siddall, and R. Zurfltih, ibid ., in cress. . ,

- 15k -

26. A. S. Meyer and E. Hanzmann, Biochem. Biophys. Res. Commun. , kl, 891 (1970).

27. J. H. Brewster, J. Amer. Chem. Soc. , 8l, 5^75 0-959).

28. J. H. Brewster, Topics Stereochem. , 2, 1 (1967).

2'9. A. S. Meyer, E. Hanzmann and R. C. Murphy, Proc. Nat. Acad. Sci. , U. S. 68, 2312 (1971). 30. A. Horeau and H. B. Kagan, Tetrahedron, 20, 2U3I (196*0 31. K. Nakanishi, D. A. Schooley, M. Koreeda, and J. Dillon, Chem. Commun., 1235 (1971). 32. M. Koreeda, D. A. Schooley, K. Nakanishi, and H. Hagiwara, J. Amer. Chem. Soc, 93, 1+08*1 (1971). 33. P. Loew and W. S. Johnson, J. Amer. Chem. Soc, 93, 3765, U08U (l97l). 34. D. J. Faulkner and M. R. Peterson, ibid ., 93, 376£ 0-971).

- 155 -

ULEINE AND RELATED ASPIDOSPERMA ALKALOIDS

Reported by Glen Tolman April 17, 1972

A tryptamine residue (the thickened bonds in 3) appears almost invariably in combination with an ubiquitous monoterpene moiety in the Aspidosperma alkaloids as illustrated in the case of aspidospermine (3). Uleine (1) and apparicine (_2)> a3ialoids isolated from several specie's of

Aspidosperma , are part of the small, highly interesting group of Aspidosperma alkaloids that do not possess the tryptamine bridge so characteristic of most of the alkaloids in this family. Uleine(l) lacks completely the two-carbon tryptamine bridge between the B-position of the indole nucleus and the basic nitrogen, while apparicine ~{z) lacks only one carbon of the bridge. It is the purpose of this seminar to discuss the chemistry, the known aspects of the biosynthesis, and some of the synthetic pathways to this small group of alkaloids. CH 3

CH3O

The two-dimensional structure of uleine was published in 1959 by Buchi. 1 Uleine was assigned structure 1 mainly on the basis of its nmr spectrum and the results of Hoffmann degradations of uleine methiodide. 1 Buchi found that

under the conditions of Hoffmann degradations (15O ; potassium hydroxide in ethylene glycol) the 1,13 exocyclic double bond migrated into conjugation with the newly formed 3 A double bond to afford the substituted ca.rba.zole k. Results 4 reported by Djerassi showed that the CX^O-N-^ bond could be cleaved without isomerization of the exocyclic double bond. When uleine methiodide was treated with sodium methoxide in methanol at room temperature, the compound 5 "was produced, This type of nucleophilic displacement of the quaternized N-5 from benzylic C-^

N(CH 3^23 ) ^ (CH 33/2)

- /^n(ce 3 ) 3 P

is similar to those seen in the case of gramine methiodide (6). Presumably some of the facility of nucleophilic displacement is due to the strain in the D-ring, and experiments were carried out to see if the strain also facilitated nucleophilic displacement of Nb from C-1+ in the free base. When uleine was treated with sodium methoxide in refluxing methanol, no reaction occurred. 4 Interestingly, when uleine was treated with acetic acid-pyridine, the acetate salt of a quaternary pyridinium compound 7 was isolated. 4 Upon attempted recrystallization from methanol, the salt underwent attack by the solvent to - - 1 56

OCH.

N(Ac)CH : N(Ac)CH.

N(Ac)CH;

9 give 8. Treatment of 8 with dilute acid resulted in loss of methanol and migration of the exocyclic double bond to afford the carbazole 9« The mass spectrometry of uleine has been extensively studied by Djerassi. 4 The major peaks in the spectrum occur above m/e 167, which is presumed to be the carbazole ion. These major peaks in the fragmentation process can be rationalized on the basis of rearrangements following either the fission of the benzylic C(4)-Nb bond or fission of the allylically activated C(2)-C(l0) bond. The stereochemistry of the ethyl side chain at C-3 of the uleine remained in question until 1967' Because of the nature of the ring system, the hydrogens at C-2 and C-U of uleine must be cis to one another. The C-3 ethyl group can ( be situated either equatorially over the indole nucleus or axially toward the basic nitrogen. Since the rate of quaternization of alkaloids is dependent on the degree of steric hindrance about the basic nitrogen, Shamma and co-workers 5 examined the pseudo- first-order reaction rates of several alkaloids including uleine with methyl iodide in an effort to distinguish in which position the C-3 ethyl side chain resided. The rate of methiodide formation determined for uleine in acetonitrile at 25° was 6 X 10 /sec. 5 This rate is very fast compared with those measured in cases such as 10~ 4 aspidospermine (3) (1.7 x / sec )> which is known to possess some degree of steric hindrance about the basic nitrogen. It is also of the same order 2 of magnitude as that measured for apparicine (2) (7 X 10~ /sec), which is relatively free of steric hindrance about the basic nitrogen. From these data, it was assumed that the C (3)- ethyl side chain is situated equatorially over the indole nucleus and away from the basic nitrogen as shown in structure 10.

J$&

11 ( - 157 -

Also in that year, 5- epi- uleine (ll), the ethyl group of which is axially 6 situated, was isolated from Aspidosperma subincanum . The C-methyl triplet in the nmr spectrum of J-epi- uleine appeared at 6 l. 08 while it appeared at 6 6 0.88 in the spectrum of uleine. The upfield shift of the signal in uleine was attributed to the shielding effect of the indole nucleus. Also the rate of methiodide formation of 3- .£Pi- uleine was subsequently determined and found to be 20 times slower than that for uleine."7 The structure of apparicine was published by Djerassi and co-workers in e 9 19&7- ' The mass spectrum of apparicine showed a molecular ion two mass units lower than the molecular ion of uleine and a fragmentation pattern very similar to that of uleine. The similarity of uv spectra of apparicine and uleine suggested that they possessed the same chromophoric system. The N-methyl singlet observed in the nmr spectrum of uleine was absent in the spectrum of apparicine, which showed instead an AB quartet corresponding to the two protons at the benzylic position, C-6. The ethyl group evident in the nmr spectrum of uleine was replaced by the characteristic signals attributable to an ethylidene group in apparicine. These spectral data, in addition to the identification of pyridine- 3, ^--dicarboxylic acid from dehydrogenation and oxidation of apparicine, led to the assignment of structure 2 to apparicine. BIOSYNTHESIS The biosynthesis of Aspidosperma alkaloids has been very extensively 10 12 12 reviewed ' - and has been the subject of a previous seminar. Tracer studies 13 have shown that Aspidosperma alkaloids result from the combination of amino acid tryptophan (12) and a nine or ten carbon skeleton 15 which is biosynthesized from geraniol. This combination can be visualized in the 13 case of the Aspidosperma alkaloid, vindoline (l4). It is evident that

COpH

12 uleine and apparicine possess the ubiquitous C 10 monoterpene skeleton common to all Aspidosperma alkaloids but lack the two- carbon bridge biogenetically derived from tryptophan (or tryptamine). Wenkert proposed ten years ago 14 that uleine resulted from the condensation of a progenitor of tryptophan with the C 10 unit to give structure 15 which upon cyclization and elimination of the B-glycosyl group would give uleine (lO).

-> 10 <- > 2

H0 2C CH=0 HOoC CH=0 OH OH

R = CH-CH- CH 20P03H 15 16 - 158 - 9 Djerassi in 1965 extended this hypothesis to include apparicine (2) by- proposing that a prototropic rearrangement of 15 to 16 followed by cyclization and elimination of the _P-glycosyl unit would give apparicine (2). By this hypothesis, uleine and apparicine arise biogenetically from a common precursor and contain no carbon atoms from the side chain of tryptophan. Kutney and co-workers in I96915 tested this hypothesis set forth by

Wenkert and Djerassi. They fed Aspidosperma pyricollum , which contains both uleine and apparicine, tryptophan labelled with tritium in the indole nucleus. They fed tritiated tryptophan to the whole plant by the cotton wick method and also to just the roots by the hydroponic method and found that there was significant incorporation of tryptophan into apparicine but not into uleine. To investigate further the incorporation of tryptophan into apparicine, Kutney performed a double label experiment with tryptophan 14 tritiated in the indole nucleus and with a C label at either C-2 or C-3. The results showed clearly that the label at C-3 is incorporated into apparicine, presumably at C-6, and that over 97$ of the activity of radio 15 labelled C-2 is removed in the biosynthesis of apparicine. The results of Kutney and co-workers show that apparicine and uleine are not biosynthesized from a common precursor and that apparicine but not uleine arises from tryptophan. The negative results in the case of uleine, though by no means conclusive, seem to suggest that uleine is derived from a progenitor of tryptophan as proposed by Wenkert. 14 In an effort to elucidate further the biosynthetic pathway leading to 16 apparicine, Kutney and co-workers fed tritium- labelled stemmadenine (17). vallesamine (l8_), and 3_am i nomethylindole (19) to Aspidosperma pyricollum . Because of the structural similarities that these compounds share with apparicine, they are possible candidates for precursors. Also depending upon which, if any, of the three would be found to be incorporated, something could be said about whether the C-2 is lost early in biosynthesis or toward the end of biosynthesis. It was found that tritiated stemmadenine (17). (

n(ch3 )h

Me0 2C CHpOH Me0 2C CH 2OH IT 18 19 was incorporated into apparicine in astonishingly large amounts, while the other two alkaloids showed little incorporation. From these results, it can be concluded that stemmadenine (17) is of some importance in the biosynthesis of apparicine and that fragmentation of the tryptamine bridge with loss of C-2 occurs as one of the last steps in the biosynthesis of apparicine. Kim and Erickson17 recognizing the similarities in structure of condylocarpine (20) and uleine proposed that the two alkaloids may be biogenetically related. They postulated that oxidative cleavage of the tryptamine bridge from the basic nitrogen, followed by expulsion of the J-C2-unit, would interrelate these two groups of alkaloids. In an effort to show that the expulsion of a ^-C^-unit from an

o>- methylene- indoline to give an indole is chemically feasible, they synthesized several model compounds and tried the eliminations. They were not successful COpMe in cleaving the JJ-Cg-unit from an _o- methylene- indoline 20 system. s —

- 159- SYNTHESIS 7 18~ 25 The synthetic pathway to these alkaloid ' fall roughly into two groups: l) those whose key intermediate contains the suitably substituted piperidine attached through C-l to the opposition of indole set up for cyclization of the C-ring by formation of the bond between C-k and the J- position of indole; 2) those whose key intermediate contains the suitably substituted piperidine attached to the ^-position of indole set up for cyclization of the C-ring by formation of a bond between C-l and the o-position of indole. Because of the many similarities in the approaches reported, only two of the more elegant routes will be discussed in detail one from each group. The first group is best represented by the synthesis of uleine and 7 2^ PA 3-epi-uleine reported by Joule and co-workers. > The key intermediate in this pathway is 3- e"thyl-l, 2, 5, 6-tetrahydro-l-methyl-4-pyridyl-indol-2-yl ketone (25b), which was generated from methyl 3-ethylisonicotinate (21 ).

C02CH3

21

-CH<

jZ53P=CH DMSO Mg- amalgam CH2I 2

10 11 - 16 o - Claisen condensation of 21 with methyl propionate followed by acid hydrolysis and decarboxylation yielded 3-ethyl-U-propionylpyridine (22). The enamine of 2_2 was generated and treated with benzene diazonium chloride to give the phenylhydrazone 25. The Fischer indole synthesis of 23 afforded 2- (3- ethylisonicotinoyl)- indole (2j±). The methiodide of 2h was formed and subsequently reduced to 25a with sodium borohydride. The conversion of the allylic alcohol 25a to the conjugated ketone 25b was accomplished by oxidation with manganese dioxide. The conversion of 2 5b to 26 involves the kinetically controlled protonation of the enolate of 25b formed by treatment with sodium methylsulfonylmethanide in dimethylsulfoxide. Treatment of the enamine 26 with acid results in protonation at C-3 and electrophilic attack of the resulting iminium intermediate on the _P-position of indole to form a mixture of dasycarpidone (27) and 5~ epi- dasycarpidone (28). Dasycarpidone (27) was converted to uleine by treatment with methylene triphenylphosphorane. 5- Epi- dasycarpidone (28) was converted to 3- epi- uleine by reaction with magnesium amalgam and methylene iodide. The elegant stereospecific syntheses of uleine and 3-^pi_uleine reported by Buchi 18 illustrates the synthetic approach common to the second group. The key intermediates in the syntheses of uleine and 3-epi-uleine are the - cis , c is methyl ketone J50 and the trans , trans- methyl ketone yi, respectively. Both methyl ketones come from the same precursor, the acetoxy methyl ketone

32, which was synthesized starting from l-aminohexan-3-one prepared from l-chlorohexan-3-one via the phthalimide method. Mannich condensation of l-aminohexan-3-one with 3- formyl indole afforded the trans- piper idone 35 * The cyclization is postulated to proceed through the chair- like intermediate 3k with the large substituents on the two double bonds trans oriented so as to form a piperidone with two equatorial substiuents. Formylation of 33 with formic acetic anhydride was followed by ethynylation with postassium acetylide in tert- butyl alcohol- tetrahydrofuran. The nucleophilic acetylide should attack from the side opposite the indole ring to give the ethynylcarbinol <

35 » Mercuric acetate in acetic acid was used to convert J>5 to the acetoxy methyl ketone 32. Reductive elimination of the acetoxy group of 32 with

H H H jM R L h~ u* > 2° Y is H R = a $& lit

H

(

R

35 O - 161 -

lithium in liquid ammonia afforded the trans , trans- methyl ketone 31. Cyclization of the trans, trans- methyl ketone with boron trifluoride etherate in methylene chloride followed by reduction of the N-formyl group with lithium aluminum hydride afforded racemic 3- epi- uleine (ll). Pyrolysis of the acetoxy ketone 32 produced a mixture of products in which the a, _§- unsaturated ketone 36 was the only component of further use in the synthesis. Catalytic hydrogenation of 36 afforded the cis , cis- methyl ketone _30. Cyclization of the cis, cis- methyl ketone 30 with boron trifluoride etherate in methylene chloride followed by reduction of the N-formyl group with lithium aluminum hydride afforded racemic uleine (10).

BIBLIOGRAPHY

1. G. Buchi and E. W. Warnhoff, J. Amer. Chem. Soc, 8l, *4-*4-33 (1959). 2. J. Schmutz, F. Hunziker, and R. Kirt, Helv. Chim. Acta., ko, II89 (1957). 3. J. Schmutz and F. Hunziker, Helv. Chim. Acta., ^L, 288 (1958). h. J. Joule and C. Djerassi, J. Chem. Soc, 2777 (196*0. 5. M. Shamma, J. A. Weiss, and R. J. Shine, Tetrahedron Letters, 2*4-89 (1967). 6. A. J. Gaskell and J. A. Joule, Chem. Ind. (London), I089 (1967). 7. A. Jackson, N. D. V. Wilson, A. J. Gaskell, and J. A. Joule, J. Chem. Soc, c, 2738 (1969).

8. J. A. Joule, M. Ohashi, B. Gilbert, and C. Djerassi, Tetrahedron, 21 , 1717 (1965). 9- J. A. Joule, H. Monteiro, L. J. Durham, B. Gilbert, and C. Djerassi, J. Chem. Soc, *4-773 (1965). 10. A. I. Scott, Accounts Chem. Res., 3> 151 (1970), and references cited in this review. 11. I. D. Spenser, in "The Chemistry of Alkaloids", ed. S. W. Pelletier, Reinhold, New York, 1970, p. 706. 12. J. K. Kirkpatrick, Univ. of 111. Seminar Abstracts, October 2*4-, 1966, p. 6*4-, 13. A. R. Battersby, Pure Appl. Chem,, lk, 117 (1967). 1*4-. E. Wenkert, J. Amer. Chem. Soc, 8*4-, 98 (1962). 15. J. P. Kutney, V. R. Nelson, and D. C. Wigfield, J. Amer. Chem. Soc, 91, *+278 (1969). 16. J. P. Kutney, V. R. Nelson, and D. C. Wigfield, J. Amer. Chem. Soc, §1, 1+279 (1969). 17. I. K. Kim and K. L. Erickson, Tetrahedron, 27, 3979 (1971). 18. G. Buchi, S. J. Gould, and F. Naf, J. Amer. Chem. Soc, 93, 2*4-92 (1971).

19. T. Kametani and T. Suzuki, J. Org. Chem., 36, 1291 ( 1971)7 20. T. Kametani and T. Suzuki, Chem. Pharm. Bull., 19, 1*4-2*4- (1971). 21. T. Kametani and T. Suzuki, J. Chem. Soc, C, 1053 (1971). 22. A. Jackson, A. J. Gaskell, N. D. V. Wilson, and J. A. Joule, Chem. Commun., 36*4- (1968). 2*4-. N. D. V. Wilson, A. Jackson, A. J. Gaskell, and J. A. Joule, Chem. Commun., 58*4- (1968). 25. L. J. Dolby and H. Biere, J. Amer. Chem. Soc, £0, 2699 (1968).

STRATEGIES IN OLIGONUCLEOTIDE SYNTHESIS

Reported by Graham Walker April 20, 1972 The availability of oligonucleotides of defined length and sequence would make possible determination of the role of modifications in nucleic acids, clarification of the punctuation of the genetic code in protein syn- thesis, studies in nucleic acid-nucleic acid interactions, and would pro- vide an entry into many problems of current interest in biological systems. The problem of synthesizing oligonucleotides has attracted the attention of both organic chemists and biochemists with the result that several dif- ferent approaches are presently being explored. The object of this seminar is to evaluate critically the various strategies (chemical, enzymatic, and chemical- enzymatic) currently being used to surmount this difficult synthetic problem. CHEMICAL METHODS

Chemical methods which add a ribonucleotide to an oligoribonucleotide in a stepwise fashion can be divided into two classes: those which con- dense a phosphate on the monomer with a hydroxyl on the oligomer and those which condense a hydroxyl on the monomer with a phosphate on the oligomer. In the first class, techniques are available for extending the oligomer in 1 2 3 either the 3'">5' direction, ' or in the 5'~>3' direction. In the deoxy- ribonucleotide series, growth is usually in the 5'">3' direction with the non- 4-9 reacting end of the oligomer often terminating in a protected phosphate. ,-> f In the second class techniques have been developed for growth in the 5 3 10 11 direction. ' The weakest link in all chemical syntheses is the condensation step, since the commonly employed condensing agents, dicyclohexylcarbodiimide and arenesulfonyl chlorides, give such low yields as to render the synthesis of an extended oligomer extremely inefficient. Attempts to surmount the problem by forming the anion of the sugar hydroxyl12 or by condensing a phosphodiester 13 14 with the monomer instead of a phosphomonoester ' have met with limited success. The problem of separating the resulting oligonucleotides has led to the development of solid support systems where the oligomer is joined to the 15-18 polymeric support by either the 5' hydroxyl or by a 5' phosphate, 19 ' 20 as well as to techniques in which a "handle" is incorporated into the oligomer 6-8 to aid in its separation. ENZYMATIC METHODS

Polynucleotide phosphorylase is an enzyme which acts in vivo to break down ribonucleic acids to nucleoside diphosphates but which can also be used synthetically to add nucleotide units to oligoribonucleotides. If an excess of the nucleoside diphosphate is used polymers are formed, but by limiting the reaction time and the quantity of the diphosphate a distribution of products can be produced. 21 The addition of a specific ribonuclease to the polynucleotide phosphorylase reaction mixture allows the high yield addition of one monomer to an oligonucleotide but places severe restrictions on the composition of the oligomer. 21? 22 Ribonucleases have also been used synthetically for the addition 23-27 of the 5' hydroxyl of the monomer to a cyclic 2', 3 '-phosphate on the oligomer. The excessively high concentrations of monomer required lead to low yields of products in most cases. CHEMICAL- ENZYMATIC METHODS Recently a promising new high yield technique has been reported which promises to combine the ease and specificity of an enzyme reaction with the generality of a chemical method. Polynucleotide phosphorylase is used to catalyze the addition of a monomer to an oligomer but the nucleoside diphosphate - 163 -

is modified with a simple blocking group so that it will still add specifically to the oligomer but, once added, will prevent the addition of further units. Removal of the blocking group under mild conditions allows another monomer to be added, resulting in a stepwise growth of the oligo- ( nucleotide chain. The two blocking groups which have been successfully employed to date are the 2' (3' )-o-methoxyethyl group and the 2' (3' )-iso- 23 29 valeryl ester. > BIBLIOGRAPHY

1. R. Lohrmann, D. Soil, H. Hayatsu, E. Ohtsuka, and H. G. Khorana, J. Amer. Chem. Soc, 88, 819 (1966).

2. A. Holy, Collect. Czech. Chem. Commun., 3_5> 3^86 (1970). 3. B. E. Griffin and C. B. Reese, Tetrahedron, 2k, 2537 (1968). k. A. F. Cook, M. J. Holman, and A. L. Nussbaum, J. Amer. Chem. Soc, 91, 6U79 (1969). 5. E. Ohtsuka, M. Ubasawa, and M. Ikehara, ibid ., 92, 5507 (1970). 6. T. Hata, K. Tajima, and T. Mukaiyama, ibid ., 21; ^928 (1971 )•

7. K. L. Agarwal, A. Yamazaki, and H. G. Khorana, ibid . , 93, 275U (l97l). 8. S. Narang, 0. S. Bhanot, J. Goodchild, J. Michniewicz, R. H. Wightman, and S. K. Dheer, J. Chem. Soc. D, 516 (1970).

9. W. Freist and F. Cramer, Chem. Ber. , 103, 3122 (1970). 10. E. Ohtsuka, K. Murao, M. Ubasawa, and M. Ikehara, J. Amer. Chem Soc, 92, 31+1*1 (1970). 11. E. Ohtsuka, M. Ubasawa and M. Ikehara, ibid ., 93, 2296 (l97l). 12. R. G. von Tigerstrom and M. Smith, Science, l6j_, 1266 (1970). 13. J. H. van Boom, P. M. Burgess, G. R. Owens, C. B. Reese, and R. Saffhill, J. Chem. Soc D, 869 0-971 )• 11*. T. Neilson, ibid ., 1139 (1969). 15. K. F. Yip and K. C. Tsou, J. Amer. Chem. Soc, 93, 3272 (l97l). I 16. T. Kusama and H. Hayatsu, Chem. Pharm. Bull., lEf 319 (1970). 17. T. Shimidzu and R. L. Letsinger, Bull. Chem. Soc. Jap., kh, 1673 (l97l). 18. V. K. Potapov, 0. G. Chekhmakhcheva, Z. A. Shabarova, and M. A.

Prokof'ev, Proc Acad. Sci. USSR, Chem Sect., 196 , 60 (l97l). 19. G. M. Blackburn, M. J. Brown, and M. R. Harris, J. Chem. Soc. C, 2U38 (1967).

20. W. Freist and F. Cramer, Angew. Chem. Int. Ed. Eng. , _2, 368 (l970). 21. R. E. Thach, T. A. Sundarajan, K. F. Dewey, J. C. Brown, and P. Doty, Cold Spring Harbor Symp. Quant. Biol., 31, 85 (1966). 22. W. M. Stanley Jr., M. A. Smith, M. B. Hille, and J. A. Last, ibid ., 31, 99 (1966). 23. S. C. Mohr and R. E. Thach, J. Biol. Chem. jM, 6566 (1969). 2k. M. Saito, Y. Furuichi, K. Takeishi, M. Yoshida, K. Arima, H. Hayatsu,

and T. Ukita, Biochem. Biophys. Acta., 195 , 299 (1969).

25. M. J. Rowe and M. A. Smith, ibid ., 2^7 , I87 (l97l)- 26. A. P. Kavunendo, E. N. Moroyova, and N. S. Tikhomirova- Sidorova, J. Gen. Chem. USSR, kl, 220 (l97l). 27. T. Koike, T. Uchida, and F. Egami, J. Biochem. (Tokyo), JO, 55 (l97l). 28. J. K. Mackey and P. T. Gilham, Nature, 233, 551 (l97l). 29. G. Kaufmann, M. Fridkin, A. Zutra, and U. Z. Littauer, Eur. J. Biochem. 2k, k (1971). ,

- 161+ . STABLE CARBOCATIONS: C-13 NUCLEAR MAGNETIC RESONANCE STUDIES

Reported by Moses W. McMillan April 27, 1972 INTRODUCTION and SCOPE 13 The carbonium ion concept dates back to the early 1900' s; however, C 13 nmr investigations did not begin until 196k. The potential of C nmr as

1 5 a tool for structural elucidation was noted in I965, but the utility of carbon magnetic resonance is only now being widely appreciated. This seminar will deal with 13C nmr observations of stable carbonium ions (carbocations) in FS03H-SbF 5 (magic acid) solvent systems at low temperatures, conditions which stabilize the carbonium ion. Many of these observations were obtained by the use of internuclear double resonance (iNDOR). 4 To facilitate this presentation it will be necessary to subdivide it into the alkylhalide carbonium, three- and four-membered cycloalkylcarbonium, arylcarbonium, and the norbornylcarbonium ion systems. Although 13C nmr indicates that methylcarbonium ion is not formed in strong acid systems, 5 there is ample evidence of the existence of stable " alkylcarbonium ions 6 8 such as the propyl, butyl, and pentyl cation complexes in strong acid solvent systems. Observations in the alkylhalide carbonium ion systems have indicated the existence of dialkylhalonium ions 9 and 10 bridged halonium ions. The 13C magnetic resonance spectra of cyclopropylcar- binyl and cyclobutyl cations indicate an equilibrating set of C 3- and C - 2, p 4 cyclopropylcarbonium ions which involve a new tetra- and/or petacooridinated " bonding concept. 11 13 Olah has recently reported the detection of the tetramethyl-cyclobutenium dication, an aromatic 2-rr system. 14 Impressive arrays of o- and (3-phenylalkyl cations and a- phenyl- P-bromoalkyl cations have been investigated which distinguish between the various methods whereby these 15" ions are stabilized. 20 Carbon- 13 nmr studies of the norbornyl cation system in "magic acid" point to the existence of both trivalent ("classical")

carbonium ions and tetra- or pentacoordinated ("nonclassical" ) carbonium ions with other ions containing varying degrees of derealization between the two 13 21 22 extremes. > > CONCLUSION

Carbon- 13 nmr has been shown to be a powerful tool for structural studies in strong acid media. The correlation of observed chemical shift with electron density at the carbon atom has permitted investigators to make surprisingly accurate extrapolations to positively charged atoms in other cation systems.

BIBLIOGRAPHY

1. J. B. Stothers, Quart. Rev. (London), 19, ikk (1965). 2. G. A. Olah, A. M. White, and D. H. O'Brien, Chem. Rev., 70, 56I (1970). 3. P. S. Pregosin and W. E. Randall in "Determination of Organic Structures by Physical Methods," Vol. k, F. C. Nachod and J. J. Zuckerman, Eds., Academic Press, New York: London, 1971? Chap. 6.

k. E. B. Baker, J. Chem. Phys., _21, 911 ( 19^2). 5. G. A. Olah, J. R. DeMember, and R. H. Schlosberg, J. Amer. Chem. Soc. 91, 2112 (1969). 6. G. A. Olah, E. B. Baker, J. C. Evans, W. S. Tolgyesi, J. S. Mclntyre,

and I. J. Bastien, ibid . , 86, I36O (196U). 7. G. A. Olah and M. B. Comisarow, ibid ., 88, 361 (1966). 8. G. A. Olah and A. M. White, ibid ., 91, 58QI (1969). 9. G. A. Olah and J. R. DeMember, ibid . , 91, 2113 (1969); . 92, 718 (1970). 10. (a) G. A. Olah and J. M. Bollinger, ibid ., 8£> hjkk (1907); 90, 9*+7, 2587, 6988 (1968); (b) G. A. Olah and P. E. Peterson, ibid ., —90, 1*675 (1968). - 165 -

11. G. A. Olah, C. J. Jeull, D. P. Kelly, and R. D. Porter, ibid., 94, 146 (1972); 92, 2544 (1970). 12. M. Saunders and J. Rosenfeld, ibid ., 92, 2548 (1970). I * 13. G. A. Olah, ibid ., 94, 808 (1972). 14. G. A. Olah, J. M. Bollinger, and A. M. White, ibid ., 91, 3667 (1969). 15. G. A. Olah, R. D. Porter, and D. P. Kelly, ibid ., Q^T^U (1971).

16. G. A. Olah, C. J. Jeull, and A. M. White, ibid . , 91, 396I (1969). 17. G. M. Olah, R. H. Schlosberg, D. P. Kelly, and Gh. D. Mateescu, ibid ., 92, 25^6 (1970).

18. G. A. Olah and R. D. Porter, J. Amer. Chem. Soc. , 9^, 6877 (l97l). 19. G. A. Olah and R. D. Porter, ibid ., 92, 7627 (l9707T 20. V. Koptyug and A. Rezvukhin, Tetrahedron Lett., 4009 (1968). 21 G. A. Olah, J. R. DeMember, C. Y. Lui, and R. D. Porter, J. Amer. Chem. Soc, 93, 2hk2 (1971). 22. G. A. Olah, S. M. White, J. R. DeMember, A. Commeyras, and C. Y. Lui,

ibid ., 92, 4627 0-970 ).

<

i . "

- 166 -

ORGANIC CHLOROSULFINATES

Reported by Steven W. Mcge May 1, 1972 Chlorosulfinates (l) are most familiar to organic chemists as reaction intermediates in the conversion of alcohols to their corresponding chlorides with thionyl chloride, as depicted in eq. 1. It is the purpose of this seminar to discuss the synthesis and reactions of chlorosulfinates and their bromo and fluoro homologs *

-HC1 \R : -so2 ROH+ S0C1 ; > R0SC1- R 0SC1 RC1 (1)

Chlorosulfinates have been synthesized by the action of thionyl 1 ' 2 3 ' 4 chloride on alcohols and organic sulfites, alkoxy silanes on 5 ' 6 thionyl chloride, silicon tetrachloride on alkyl sulfites, 5 boron 7 trichloride on organic sulfites, and sulfuryl chloride on olefins. 8 Factors which increase chlorosulfinate stability include electronegative substituents, decreased chain branching, 1 and increased chain length. 1 Benzylic and allylic chlorosulfinates are less stable than alkyl chlorosulfinates, which are in turn less stable than aryl chlorosulfinates. 2 This order of stability is in accord with the ease of formation of the ion pair 2, which is thought to be initially formed when decomposition occurs. Chlorosulfinates are much more stable in non-polar than in polar solvents. 10 For R = norbornyl, the endo is more stable than the exo isomer. 11 Although the thermal- and base-induced decompositions of chlorosulfinates have received most attention in the literature, the exact nature of the intermediate or intermediates in the decomposition of 1 has still not been satisfactorily elucidated. The decomposition of secondary alkyl chlorosulfinates proceeds predominantly with inversion of configuration in pyridine or hydrocarbon solvents, and with retention in ether solvents. 12 ' 13 Allylic chlorosulfinates give rearranged chloride products when decomposed in ether solution. 14 A deuterium isotope effect is observed for secondary alkyl chloro- 15 l6 sulfinates. ' »-Trifiuorometbyl- substituted allyl chlorosulfinate gives 17 rearranged product, but the y isomer does not. The ion pair 2 is sufficiently long-lived in some cases to allow a 1,2 hydride shift. 18 Vicinal aryl assistance with rearrangement has been shown to occur in 3-phenethyl chlorosulfinate. 19 Recent studies with structural isomers of chlorosulfinates, each of which could potentially go through 2 as a common reaction intermediate, show that this does not occur. 20 ' 21 The dialkyl sulfinate-chlorosulfinate-thionyl chloride system is a complex one, and careful manipulation of the reaction conditions is necessary to ensure ' 4 ' 22 ' 23 formation and successful isolation of the desired product. 3 The dimethylformamide- thionyl chloride adduct, which has proven quite useful in chlorination and formylation, 24 was originally thought to exist as dimethylchloroformiminium chloride (3), but it has been more recently shown that this reagent can be isolated as a compound which has an elemental analysis close 2 to that required for the chlorosulfinate k. - "

,ci-| r II (CH 3 )2N=c( Cl" + ^sci (CH N=C 3 ) 2 s H

The reaction of k with aromatic hydrocarbons to yield sulfides is one of the 26 more recent synthetic reactions of this remarkable reagent. Nuclear magnetic resonance studies have shown that the a-methylene protons 28 of chloro- 27 and fluorosulfinates are magnetically nonequivalent. The signal doublet observed for the methylene protons in neopentyl chlorosulfinate remains a . .

- 167 -

over a 100 temperature range, but collapses -when trace amounts of pyridinium chloride are added. 27 Homologs of chlorosulfinates which have been recently synthesized and 29 studied include chlorothiosulfinates, bromosulfinates, 30 fluorosulfinates, 28>3l>32 and fluorothiosulfinates. 33 * BIBLIOGRAPHY

1. P. Carre', Compt. Rend., 194, 1835 (1932); Chem. Abstr., 26, 4301 3 (1932). 2. P. Carre', Bull. Soc. Chim. 3 , 53, 1075 (1933) J Chem. Abstr ~28, I658 (1934). 3- W. E. Bissinger and F. E. Kung, J. Amer. Chem. Soc, 69, 215"5~( I9I+7) 4. P. D. Bartlett and H. F. Herbrandson, ibid., 74, 5971T1952) 5. B. R. Currell, M. J. Frazer, W. Gerrard, E. Haines, and L. Leader, J. Inorg. Nucl. Chem., 12, 45 (1955). 6. B. R. Currell, M. J. Frazer, and W. Gerrard, J. Chem. Soc, 2776 (i960). 7. J. Charalambous , H. J. Davies, M. J. Frazer, and W. Gerrard, ibid., 1505 (1962). 8. A. Y. Yakubovitch and Y. M. Zinoviev, J. Gen. Chem. (USSR), 17, 2028 (1947); Chem. Abstr., 43, 1249a (1949). 9 9. P. Carre', Compt. Rend., 196, 1806 (1933); Chem. Abstr., 27, 4211 (1933). 10. S. H. Sharman, F. F. Caserio, N. F. Nystrom, J. C. Leak, and W. G. Young, J. Amer. Chem. Soc, 80, 5965 (1958). 11. J. K. Stille and F. M. Sonnenberg, ibid., 88, 4915 (1966). 12. C. E. Boozer and E. S. Lewis, ibid ., 75, 3IB2 (1953). 13. E. S. Lewis and C. E. Boozer, ibid ., 74, 308 (1952). 14. F. F. Caserio, G. E. Dennis, R. H. DeWolfe, and W. G . Young, ibid ., 77, 4182 (1955).

15. E. S. Lewis and C. E. Boozer, ibid ., 74, 6306 (1952). 16. C. E. Boozer and E. S. Lewis, ibid ., 76, 794 (1954). 17. J. A. Pegolotti and W. G. Young, ibid., 83, 3251 (1961).

18. C. C. Lee, J. W. Clayton, D. G. Lee, and A . J. Finlayson, Tetrahedron, 18, 1395 (1962). 19. C. C. Lee, D. J. Kroeger, and D. P. Thornhill, Can. J. Chem., —42, 1130 - (1964). 20. P. K. Freeman, F. A. Raymond, and J. N. Blazewich, J. Org. Chem., ~~34, 1175 (1969). 21. D. J. Cash and P. Wilder, Jr., Chem. Commun., 662 (1966). 22. W. Gerrard, G. Machell, and P. Tolcher, Research Correspondence, Suppl. to Research (London), 8, No. 2, S7 (1955). 23. W. Gerrard, J. A. Porter, and P. L. Wyvill, ibid., 8, No. 7,S35 (1955). 24. L. F. Fieser and M. Fieser, "Reagents for Organic Synthesis," John Wiley and Sons, Inc., New York, N.Y., 1967, PP« 286-289. 25. G. Ferre' and A. -L. Palomo, Tetrahedron Lett., 2l6l (1969); Chem. Abstr., 71, 70058q (1969). 26. G. Wolter and W. Kosler, Z. Chem., 10, 401 (1970); Chem. Abstr., _j4, 12752 n (1971). 27. H. R. Hudson, R. G. Rees, and G. R. DeSpinoza, Spectrochim. Acta, 2JA, 926 (1971). 28. F. Seel, J. Boudier, and W. Gombler, Chem. Ber., 102 (2), 443 (1969); Chem. Abstr., 70, 86946t (1969). 29. R. Steudel and G. Scheller, Z. Naturforsch. , B, 24(3), 351 (1969); Chem. Abstr., 71, 3172j (1969). 30. G. Bellucci, F. Marioni, and A. Marsili, Tetrahedron, _25, 4l67 (1969). 31. J. I. Darragh, A. M. Noble, D. W. A. Sharp, D. W. Walker, and J. M. Winfield, Inorg. Nucl. Chem. Lett., it, 517 (1968). 32. A. M. Noble and J. M. Winfield, J. Chem. Soc. A, 501 (1970). 33. F. Seel, R. Budenz, and W. Gombler, Z. Naturforsch., 25B, 885 (1970). - 168 -

RECENT ADVANCES IN THE CHEMISTRY OF PENICILLINS AND CEPHALOSPORINS Ma 8 Reported by Ronald Stroshane Y > 1972 The chemistry of penicillins (l) and cephalosporins (2) has been presented 1 2 3 4 in recent seminars ' and in detail in recent reviews. > This field, however, continues to grow rapidly, and it is the purpose of this seminar to update and expand these writings.

H H H H H H

R-N^ i I ^S S3 R-N

:h3

co^x -!ooh

" Studies on the mode of action of the penicillins and cephalosporins 5 10 have shown that opening of the labile B- lactam ring occurs during inhibition of bacterial cell wall synthesis. Since this action requires a close fit into an enzyme system, it is not suprising that alterations in the stereochemistry11-20 21 22 80 and changes in ring- system structure * > form analogs that are bacterially inactive. Changes from A3 to A 2 cephems also have a pronounced effect on structure p3 24 and hence on biological activity. ' In order to obtain active penicillin and cephalosporin analogs, then, it is possible to modify side chain substituents but necessary to leave the B- lactam- thiazolidine ring system intact. There are two routes to new analogs: (l) chemical synthesis from inactive compounds and (2) semi- synthetic means, employing the naturally occurring antibiotics as starting materials, followed by enzymatic or chemical alterations. " After many unsuccessful attempts by a number of researchers 25 32 the total 33 synthesis of penicillin was acheived by Sheehan in 1959 and that of cephalosporin C by Woodward in 1966. 34 Since then some modifications of these syntheses have 35 3S been suggested. > Although intellectually satisfying, the chemical syntheses prove to be much less financially rewarding than the semi- synthetic methods. Enzyme systems have been found that acylate 6-aminopenicillanic acid (6-APA) in the presence of 37 3S good acyl donors ' and a limited number of penicillin analogs can be prepared in this way. By far the most common method of obtaining new analogs is by chemical treatment of 6-APA or 7-aminocephaldsporanic acid (7-ACA). The route frequently used merely requires condensation of the acid chloride of the desired side chain with 6-APA or 7-ACA in solution (obtained from available penicillins or cephalosporins). This technique and analogous ones have been used to obtain many important antibiotics. Some of the more recent ones are presented in Table 1. Because a cephalosporin amidase has not been found which converts = = cephalosporin C (2, R a-aminoadicpic acid, X OCOCH 3 ) to 7-ACA, chemical hydrolyses are required and the yields of pure 7-ACA are often low. Therefore one of the more actively pursued methods of obtaining cephalosporin analogs has been through ring expansion of the readily available pencillin nucelus. This technique was reported by a Lilly group in 1969 5 ^ and has since received further applications. - 169 -

Table 1

Penicillin Analogs (l)

Name r y- Ref .

terpenyl 19 cmpds. Na 39

a<-azidobenzyl pen. Ph-CH-CO- jja l+O N3 ampicillin Ph-CH-CO- H pivampicillin Ph-CH-CO- -CH2OCC(CH3 ). NH P Ar .CO- k2

aminoalicyclic pens. (CHp) n <^C0- ^

carbenicillin Ph-CH-CO- H kk COOH e Y sydnonylmethyl pen. A (+) V CHCO- Na

Y=H, Br, Ph

" Ph-CH2 CO- H 90

<" <3^> ^>-CH CO- ^ ?- sulfocillin C Na h6,k7 S03Na

and others 81-85

Cephalosporin Analogs (2) Cephaloridine ± >CH^CO- V s <^=E> U8

Cephalothin n 1/ ^

1+8 \ s /^CH 2CO- -OCCH3 )

- 170 -

Table 1 (cont.

Name X Ref

sydnonylmethylcephalosporin -OCOCH3 h5

cephazolin C^s CH C0- 2 ^9-52 / N =N' cephalexin Ph-CH-CO- H 53 NH 2

cephaloglycin Fh-CH-CO- -OCOCH3 5^

NH 2

and others 86

It had been suggested that cephalosporin C is a metabolic transformation product of penicillin N (l, R = a-aminoadipic acid, X = H). However, there had been no indication that the thiazolidine ring could be opened and expanded without disrupting the labile ^- lactam bond. The approach taken by Morin et al. 55 was the activation of the sulfur atom by oxidation, a facile cleavage of the C 2- sulfur bond with the introduction of a double bond, and then reclosure of the ring in an alternate sense. Phenoxymethyl penicillin (pen V) sulfoxide methyl ester (3) was stable to acetic anhydride under mild conditions. Upon heating to reflux (Scheme l), the penicillin sulfoxide derivative was destroyed and new material was isolated in 60% yield, composed of a 2: 1 mixture of two components which could be separated chromatographically and base resulted in (6 J_). Treatment of J_ with mild 8, methyl 3- methyl-7- (2-phenoxyacetamido)-3-cephem-4-carboxylate (deacetoxy- cephalosporin V methyl ester). Later work by another Lilly group 56 demonstrated conversion of 8 into 2 cephalosporin V (l3b). Mild hydrolysis of 8 (Scheme 2) afforded A -deacetoxy- cephalosporin V (9a) which was transformed in good yield to its p-methoxybenzyl ester (9b). Bromination gave the allylic bromide 10 from which the Br was displaced to give 11, the compound was oxidized in turn to obtain the desired A3 cephem 12 exclusively, which was reduced (13a) and finally converted to the 53 desired cephalosporin V (13b) by ester cleavage. Chauvette et al. expanded this route further by synthesizing the orally active cephalexin [20 ) from penicillin V (Scheme 3). Starting with penicillin V or G (lUa, b) the trichloroethyl protecting 34 group used by Woodward was incorporated, forming 15a, b. Sulfoxidation led to l6a, b which upon heating in acetic anhydride formed the A2 cephem derivatives Scheme I

FhOCHoCOHN

PhOCHoCOHN-

COOCH

8

Scheme 2 -> 8 > -N CH. CH2Br :OOR 10 C0 CH C OCH -p 9a, R=H 2 2 6H4 3 b, R=CH C H OCH -p Si 2 6 4 3 g PhOCHoCOHN <^

N> CH2 OAc i CHpOAc CH2 OAc C02 CH2 C6H 4 OCH3 -p C02CH2 C6H4OCH3 -p

13a, R=CH2 C6H4OCH3 -p 12 11 i, R = H Scheme 3

H H RCH2 COHN RCHpCOffl^,

RCH COM ': ;^S^ 2 A "> CH3_ " / / --OR'

ljj-a, R=PhO l£a, R=FhO; R'=CH2CC13 l6a, b b, R=Rh b, R=Ph; R'=CH2 CC1 3 RCH2COHN PhCHCOHH ? «

NH ;

2 oul3

20 18 R = NH, 17a, R = PhOj R'=CH2CC13 = 19 R PhCHCOHN b, R = Rbj R'=CH2CC13 NHCOOCHpCCIPUW-LO , •

- 172 -

17a,b. Conversion of the acetamidodeacetoxycephalosporin (lj) trichloro- ethyl esters via their imido esters produced 7-aminodeacetoxycephalosporin (T-ADCA) trichlorethyl esters 18. Reacylation using a carbonyl derivative of D-a-phenylglycine gave the "doubly protected" cephalexin 19 which was converted to the desired cephalexin 20. A variety of other interesting reactions using this important 4 57 " penicillin-to- cephalosporin scheme have been reported 66 and much recent research has been done exploring the novel chemistry of these B-lactam 6 antibiotics In addition, the patent literature contains many syntheses of new drugs of this type.

BIBLIOGRAPHY

J. A. Secrist, University of Illinois Seminar, September 22, 1969. A. J. Playtis, University of Illinois Seminar, October 2, 1969. E. P. Abraham, Quart. Rev. (London), 21, 231 (1967).

E. P. Abraham, Pure Appl. Chem. , 28, 399 (1972). J. Ghuysen, J. L. Strominger, and D. J. Tipper, "Comprehensive Biochemistry", M. Florkin and E. H. Stotz, ed., Elsevier, New York, 1968, vol. 26 (A). D. J. Tipper, and J. L. Strominger, Proc. Nat. Acad. Sci. U. S., 5*., 1133 (1965).

K. Izaki, M. Matsuhashi, and J. L. Strominger, ibid . , 55, 656 (1966). J. L. Strominger, K. Izaki, M. Matsuhashi, and D. J. Tipper, Fed. Proc, 26, 9 (1967). E. H. Flynn, and C. W. Godzeski in "Biosynthesis of Antibiotics",

B. Gottlieb and P. D. Shaw, ed. , Springer-Verlag, Berlin, 1967> vol. 1, p.l. C. W. Godzeski, and E. H. Flynn in "Biosynthesis of Antibiotics", D. Gottlieb and P. D. Shaw, ed., Springer-Verlag, Berlin, 1967, vol. 1, p. 19. M. L. Sassiver, and R. G. Shepherd, Tetrahedron Lett., 3993 (1969)

S. Wolfe, and W. S. Lee, Chem. Commun. , 2^2 (1968). B. G. Ramsay, and R. J. Stoodley, ibid., ^50 (1971). J. R. Jackson, and R. J. Stoodley, ibid ., 6kf (1971)-

J. P. Clayton, J. H. C. Nayler, R. Southgate, and E. R. Stove, ibid . , 129 (1969). J. R. Jackson, and R. J. Stoodley, ibid ., Ik (1970).

A. K. Bose , C. S. Narayanan, and M. S. Manhas, ibid. , 975 (1970). S. Wolfe, W. S. Lee, and R. Misra, ibid ., 1067 (1970). G. V. Kaiser, C. W. Ashbrook, and J. E. Baldwin, J. Amer. Chem. Soc, 93, 23^2 (1971). A. Vlietinck, E. Roets, P. Claes, and H. Vanderhaeghe , Tetrahedron Lett., 285 (1972). D. Todd, R. J- Cornell, and R. T. Wester, J. Chem. Soc, C, 1969, ^08. E. G. Brain, A. J. Eglington, J. H. C. Nayler, M. J. Pearson, and R. Southgate, Chem. Commun., 229 (1972).

C. F. Murphy, and R. E. Koehler, J. Org. Chem., 35 > 2^29 (1970). G. V. Kaiser, R. B. G. Cooper, R. E. Koehler, C. F. Murphy, J. A. Webber, I. G. Wright, and E. M. Van Heyningen, ibid ., 35, 2^30 (1970) J. E. Dolfini, J. Schwartz, and F. Weisenborn, ibid ., 3^> I582 (1968). G. Stork, and H. T. Cheung, J. Amer. Chem. Soc, 87, 3785? (1965). B. M. Green, A. G. Long, P. J, May, and A. F. Turner, J. Chem. Soc, 766 (196U).

G. C. Barrett, S. H. Eggers, T. R. Emerson, and G. Lowe, ibid . 788 (196*0. E. Galantay, H. Engel, A. Szabo, and J. Fried, J. Org. Chem., ""29, 3560 (196*0. ,

- 173 -

30. J. C. Sheenan, and J. A. Schneider, ibid ., 31, 1635 (1966).

31. A. I. Meyers, and J. M. Greene, ibid . , 31, 556 (1966). 32. D. Todd, and S. Teich, J. Amer. Chem. Soc, 75, 1895 (1953). 33. J. C. Sheenan, and K. L. Henry-Logan, ibid . ,"^1, 3089 (1959). 3^. R. B. Woodward, K. Heusler, J. Gosteli, P. Naegeli, W. Oppolzer,

R. Ramage, S. Ranganathan, H. Vorbruggen, ibid . , 88 , 852 (1966). 35. N. H. Johansson, and B. Akermark, Tetrahedron LettT, ^785 (1971). 36. R. D. G. Cooper, and F. L. Jose, J. Amer. Chem. Soc, 9k, 1021 (1972). 37. M. Cole, Biochem. J., 115, 7U7 (1969). 38. T. Nara, R. Okachi, and M. Misawa, J. Antibiot. (Tokyo), Ser. A, 2k, 321 (1971). 39* G. Pala, S. Casadio, G. Coppi, E. Crescenzi, and A. Mantegani, Arzneim.-Forsch., 20, 62 (1970). kO. B. Sjoberg, B. Ekstrom, and U. Forsgren, Antimicrobial Agents and Chemotherapy— 1967, 560 (1968). ^1. W. vonDaehne, W. 0. Godtfredsen, K. Roholt, and L. Tybring, Antimicrobial Agents and Chemotherapy— 1970, ^31 (1971). k2. T. S. Naito, S. Nakagawa, K. Takahashi, K. Fujisawa, and H. Kawaguchi, Antimicrobial Agents and Chemotherapy— I968, ^k (1969). kj>. H. E. Alburn, D. E. Clark, H. Fletcher III, and N. H. Grant, Antimicrobial Agents and Chemotherapy— 1967> 568 (1968). kk. G. N. Rolinson, and R. Sutherland, Antimicrobial Agents and Chemotherapy— 1967, 609 (1968). k^. T. Naito, S. Nakagawa, K. Takahashi, K. Fujisawa, and H. Kawaguchi, J. Antibiotics (Tokyo), Ser. A. 21, 300 (1968). k6. K. Tsuchiya, T. Oishi, C. Iwagishi, and T. Iwahi, ibid . , 2k , 607 (1971) kj. T. Yamazaki, and K. Tsuchiya, ibid ., 2k , 620 (1971X kQ. J. L. Spencer, F. Y. Siu, B. G. Jackson, H. M. Higgins, and E. H. Flynn, J. Org. Chem, 32, 500 (1967). k9- K. Kariyone, H. Harada, M. Kurita, and T. Takano, J. Antibiotics, (Tokyo), Ser. A. 23, 131 (1970). 50. M. Nishida, T. Matsubara, T. Murakawa, Y. Mine, Y. Yokota, S. Goto, and S. Kuwahara, ibid ., 23, 137 (1970). 51. M. Nishida, T. Matsubara, T. Murakawa, Y. Mine, Y. Yokota, S. Goto, and S. Kuwahara, ibid ., 23, lQk (1970). 52. Y. Mine, M. Nishida, S. Goto, and S. Kuwahara, ibid ., 23, 195 (1970). 53* R. R. Chauvette, P. A. Pennington, C. W. Ryan, R. D. G. Cooper, F. L. Jose, I. G. Wright, E. M. Van Heyningen, and G. W. Huffman, J. Org. Chem., 36, 1259 (1971). 5^. E. Binderup, W. 0. Godtfredesn, and K. Roholt, J. Antibiotics, (Tokyo), Ser. A, 2h, 767 (1971). 55« R- B. Morin, B. G. Jackson, R. A. Mueller, E. R. Lavagnino, W. B. Scanlon, and S. L. Andrews, J. Amer. Chem. Soc, 91, 1^01 (1969).

56. J. A. Weber, E. M. Van Heyningen, and R. T. Vasileff, ibid . 91, 567^ (1969). 57. D. H. R. Barton, F. Comer, D. G. T. Greig, G. Lucente, P. G. Sammes, and W. G. E. Underwood, Chem. Commun., 1059 (1970). 58. D. H. R. Barton, D. G. T. Greig, G. Lucente, P. G. Sammes, M. V. Taylor, C. M. Cooper, G. Hewitt, and W. G. E. Underwood, ibid ., I683 (1970). 59. B. G. Ramsay, and R. J. Stoodley, J. Chem. Soc, C, 3859 (1971)- 60. B. G. Ramsay, and R. J. Stoodley, ibid ., 386^ (1971). 61. D. H. R. Barton, F. Comer, D. G. T. Greig, P. G. Sammes, S. M. Cooper, G. Hewitt, and W. G. E. Underwood, ibid ., 35^0 (1971). 62. D. 0. Spry, J. Amer. Chem. Soc, 92, 5006 (1970). 63. G. E. Gutowski, B. J. Foster, C. J. Daniels, L* D. Hatfield, and J. W. Fisher, Tetrahedron Lett., 3^33 (1971)- 6k. G. E. Gutowski, D. M. Daniels, R. D. G. Cooper, ibid., 3^29 (1971). ,,

- 17^ -

65. D. 0. Spry, J. Org. Chem. , 37, 793 (1972). 66- R. D. G. Cooper, J. Amer. Chem. Soc, 9k, 1018 (1972). 67. C. F. H. Green, J. E. Page, and S. E. Saniforth, J. Chem. Soc., 1595 (1965). 68. R. M. Sweet, and L. F. Dahl, J. Amer. Chem. Soc., 92, 5^89 (1970).

69. R. D. G. Cooper, P. V. DeMarco, J. C. Cheng, and N. D. Jones, ibid . 91, 1^08 (1969). 70. R. D. G. Cooper, P. V. DeMarco, 0. P. Murphy, and L. A. Spangle, J. Chem. Soc, C, jkO (1970). 71. I. McMillan, and R. J. Stoodley, Chem. Commun., 11 (1968). 72. R. A. Archer, R. D. G. Cooper, and P. V. DeMarco, ibid ., 1291 (1970). 73. R. Nagarajan, and D. 0. Spry, J. Amer. Chem. Soc, 93, 2310 (1971).

-Jk. N. C. Neuss, H. Nash, and P. A. Lemke, ibid. , 93, 2337 (1971). 75. R. Nagarajan, L. D. Boeck, M. Gorman, R. L. Hamill, C. E. Higgens, M. M. Hoehn, W. M. Stark, J. G. Whitney, ibid., 93, 2308 (1971). 76. J. M. T. Hamilton-Miller, G. G. F. Newton, and E. P. Abraham, Biochem, J., 116, 371 (1970).

77. H. Bundgaard ; Tetrahedron Lett., ^613 (1971). 78. H. Bundgaard, J. Pharm. Sci., 60, 1273 (1971). 79« K. H. Dudley, T. C. Butler, and D. Johnson, J. Pharmacol. Exp. Ther., 179, 505 (1971). 80. D. M. Brunwin, H. Lowe, and J. Parker, J. Chem. Soc, C, 3756 (1971). 81. R. G. Micetich, R. Raap, J. Howard, and I. Pushkas, J. Med. Chem., 15, 333 (1972). 82. R. A. Firestone, N. Scelechow, D. B. R. Johnston, and B. G. Christensen, Tetrahedron Lett., 375 (1972). 83. M. R. Bell, S. D. Clemans, and R. Oesterlin, J. Med. Chem., 13, 389 (1970). Qk. W. von Daehne, E. Frederiksen, E. Gundersen, F. Lund, P. Mj/rch,

H. J. Petersen, K. Roholt, L. Tybring, and W. 0. Godtfredsen, ibid . 13, 607 (1970). 85. R. C. Raap, G. Chin, and R. G. Micetich, J. Antibiotics (Tokyo), Ser. A, 2k, 626 (1971). 86. M. L. Sassiver, A. Lewis, and R. G. Shepherd, Antimicrobial Agents and Chemotherapy— 1968, 101 (1969).

87. E. S. Wagner, and M. Groman, J. Antibiotics (Tokyo), Ser. A, 2k , 6k7 (1971). 88. 0. K. Kovacs, B. Ekstrom, and B. Sjoberg, Tetrahedron Lett., 1863 (1969). 89. R. A. Archer, and B. S. Kitchell, J. Org. Chem., 31, 3^09 (1966). 90. J. P. Clayton, J. H. C. Nayler, R. Southgate, and P. Tolliday, Chem. Commun., 590 (1971). 91. R. B. Morin, B. G. Jackson, E. H. Flynn, R. W. Roeske, and S. L. Andrews, J. Amer. Chem. Soc, 91, 1396 (1969). 92. G. C. Barrett, V. V. Kane, and G. Lowe, J. Chem. Soc, 783 (1962). 93. N. J. Leonard, and R. Y. Ning, J. Org. Chem. 31, 3928 (1966). 9k. U. Golik, J. Heterocycl. Chem., 9, 21 (1972). 95- D. M. Brunwin, and G. Lowe, Chem. Commun. 192 (1972). 96. I. McMillan, and R. J. Stoodley, J. Chem. Soc, C, 2533 (1968). 97. R. J. Stoodley, ibid ., 2891 (1968). 98. B. G. Ramsay, and R. J. Stoodley, ibid., 1319 (1969). 99* J. P. Clayton, R. Southgate, B. G. Ramsay, and R. J. Stoodley, ibid ., 2089 (1970). 100. K. Butler, Kirk-Othmer Encyclopedia of Chemical Technology, Second Edition, Vol. Ik, p. 652 (1967). 101. L. D. Cama, W. J. Leanza, T. R. Beattie, and B. G. Christensen, J. Amer. Chem. Soc, ^k, ll+08 (1972). 102. S. Karady, S. H. Pines, L. M. Weinstock, F. E. Roberts, G. S. Brenner, A. M. Hoinowski, T. Y. Cheng, and M. Sletzinger, ibid., 9k, 1^10 (1972).

- 175 -

CERIUM (IV) OXIDATIONS

Reported by William C. Christopfel May 11, 1972 INTRODUCTION

Oxidations of organic compounds "with cerium (TV) salts were last reviewed in the mid 1960's. 1 ' 2 At that time the investigation of the oxidation mechanism was incomplete and the scope of useful cerium ClV) oxidations was very limited. Since then, however, considerable work has been done on these oxidations, making the topic worthy of the review in this seminar.

CERIUM (IV): A ONE ELECTRON OR TWO ELECTRON OXIDANT The common valences of the cerium ion are three and four. 3 The cerium (il) species has never been observed in solution. 4 This fact might lead to the immediate speculation that the cerium (IV) oxidation is of necessity a one-electron process. The possibility of a two-electron oxidation involving cerium (IV) has, however, been proposed in an early paper by Trahanovsky in order to explain the high yield of one product in certain cerium (IV) oxidations of toluenes. 4 Toluene and many substituted toluenes are oxidized to the corresponding benzaldehydes or benzyl acetates, depending upon the solvent system, with four equivalents of eerie ammonium nitrate (CAN). Although these products could be formed by a one or two

electron oxidative process, (Schemes I, II ) it has been suggested that the high yield of one product in certain cases favors a mechanism such as a two- electron- oxidation mechanism which does not involve radicals, because in the latter case free radicals would be expected to attack the products, especially toward the end of the reaction. 4

SCHEME I e ArCH 3 + Ce(lV) —> ArCH2 - + Ce(lll) + H ArCH2 - + Ce(lV)—^ArCH^© + Ce£lll) ArCHs© + H2 > ArCH2 0H + H&

GXC • • • •

SCHEME II

ArCH 3 + Ce (TV)->ArCH2® + Ce(ll) + H ArCH2© + H1 2P W —->7" ArCHpOHT\J. U11 2 V + H® Ce(lV) Ce (!!)-» 2 Ce (ill)

6 uC • • • •

Subsequent work on the CAN oxidation of 2-aryl-l-phenylethanols to benzaldehyde and products obtained from the cleaved benzyl moiety (benzyl alcohol and benzyl nitrate under the reaction conditions employed) suggested that the cleavage occurs by a one-electron oxidation (Scheme III). 5

Scheme III

CAN N ^> r \ CH=0 + -CH2 0/CH CN r^ H2 3 .

- 176 - When the oxidation is carried out in the presence of 10$ acrylamide, a known radical trap, the formation of benzaldehyde and a polymer is observed with no benzyl alcohol or benzyl nitrate formation. Therefore, the benzyl alcohol and benzyl nitrate were considered to arise from a cleaved benzyl radical. 5 In addition, when competition experiments are carried out by oxidizing a mixture of two different 2-aryl-l-phenylethanols with a limiting amount of CAN, an excellent Hammett relationship is found with a p = -2.0 5 when the log of the rates are plotted against a . Most radical reactions correlate better with o than with a, and give p's in the range of -O.69 6- to -1.46. The relatively large negative p of -2.0 for the cerium (iv) oxidative cleavage of 2-aryl-l-phenylethanols indicates that a fair amount of positive charge is developed on the p-carbon in the transition state. 5 None the less, a p value of -2.0 is closer to those reported for radical reactions than the -5.0 range reported for cationic processes. 9 Radical trapping experiments on a number of other systems have been performed with results generally similar to those obtained for the 2-aryl-l- phenylethanols. For example; the oxidation of bicyclohexyl-l,l'diol with two equivalents of CAN gives a 9^-$ yield of cyclohexanone in 75$ aqueous- acetic acid. The same reaction run in the presence of an added 10$ acrylamide gives a 48$ yield of cyclohexanone, thus supporting the one electron oxidative cleavage mechanism. 10 For the eerie sulfate oxidation of 1-methoxycyclopropanol, conclusive evidence for the radical has been obtained by esr experiments. 11 The oxidation of cycloheptatriene (CHT) with CAN, however, seems to be a two electron process. 1,3,5- Cycloheptatriene is oxidized with four equivalents of CAN to benzaldehyde, benzene and carbon monoxide in high yield. 12 Radical trapping experiments have no effect on the yield. Apparently no radical intermediate is involved and a two-electron oxidation is implicated. 12 On the basis of two experimental facts: l) the oxidation of CHT with CAN proceeds in good yield in anhydrous acetonitrile and the benzaldehyde-to-benzene ratio is similar to oxidations in aqueous solvents « and 2) the oxidation of CHT with eerie perchlorate in perchloric acid gives a low yield (3*2$) of benzaldehyde, a one-elelcton ligand involvement mechanism is proposed for this CAN oxidation. 12,ls It has been suggested12 that this type of ligand involvement may occur in other oxidations by CAN and may account for the fact that CAN sometimes behaves like a two-electron oxidant and brings about more successful reactions with organic compounds that any other cerium (IV) salt. Recently, Meyer and Rocek reported that cyclobutanol reacts with known one-electron oxidants to yield cleavage products, but with two-electron oxidants to yield the ketone. Therefore, they suggest the use of cyclobutanol as a probe for distinguishing between one-and two-electron oxidizing agents. The CAN oxidation of cyclobutanol yields cleavage products which can be accounted for by a one-electron oxidation (eq l). 14

y 0H + + + (l) -2^ 0CH(CH2 ) 6CH0 CH 3 (CH2 ) 2CH0 CH2 =CHCH2CH0 H0(CH2 ) 3CH0 H2 Cerium (IV) is therefore shown to be consistently a one- electron oxidant, certain salts of which can sometimes give two- electron- like oxidations due to ligand involvement.

CERIUM (IV) OXIDATIONS OF ALCOHOLS There are three known types of reactions of cerium (IV) with alcohols: l) the oxidation to carbonyl compounds; 2) oxidative cleavage and the formation of cleavage products; 3) cyclization to tetrahydrofurans Equation 2 represents an example of the oxidation of an alcohol to a carbonyl compound, and provides a convenient and efficient oxidative synthesis of cyclopropanecarboxaldehyde. 15 - 177 -

Oxidations of the general type ArCHOHR ->ArCOR, for example equation ~$, work well except when a very reactive ring system or an easily oxidized or acid sensitive functional group is present. In those cases where the aldehyde or ketone is obtained in high yield, the convenience of this oxidation seems to make it the method of choice. ls Oxidative cleavage reactions of alcohols (eq U-8) represent a considerable 5> 1T le portion of the cerium (iv) oxidations studies. > Relief of ring strain and radical stability are the factors determining the course of the oxidation reactions of alcohols. For the oxidations of cyclopropanols (eq k) and cyclobutanols (eq 5) relief of ring strain facilitates cleavage of the carbon- carbon bond, and in the oxidations of 1, 2-diarylethanols (eq 6), norbornanol (eq f), an^ bicyclo [2. 2. 2]octan-2-ol (eq 8) the cleavage yields a stable radical and cleavage products are observed. Inspection of the alcohol oxidations which lead to direct carbonyl formation (eq 2 and 3) reveals that cleavage would produce the unstable pyridyl and cyclopropyl radicals. Accordingly, direct carbonyl formation occurs in preference to cleavage. Trahanovsky has determined the stability of the radical needed for cleavage to occur. Cerium (iv) oxidations of a series of alkylphenylcarbinols have been studied, in which the alkyl groups were methyl, ethyl, isopropyl, and tert- butyl, with particular attention being paid to the relative rates of cleavage to 19. ketone formation. For methylphenylcarbinol, ketone formation is the main

CH 2OH + 2CAN -> CHO 67$ (2)

H 2OH HO CAN (3) S^J 6(

> R-C-(CH2 ) 4-C-R + CHsCH^-C- R 00

OH Ce(lV) Cr^CR^CK^C- CH3 (5)

OH Ce(lV) > + CH. ^H ' (6)

CHO CAN OH (7)

0N0 ; CHO CHO

CAN (8) OH 9 .

- 178 - pathway, while oxidative cleavage is the almost exclusive route for isopropyl and tert- hutlyphenylcarbinols. Both pathways are important for ethylphenyl- carbinol. This suggests that if a radical as stable is a secondary carbon radical can be formed, oxidative cleavage will be the main pathway for oxidations 19 of alcohols by cerium (iv). The CM oxidation of 1-pentanol gives 2-methyltetrahydrofuran as the 20 major product, and only traces of the corresponding aldehyde (eq 9).

CAN > J \ 20$ (9)

This reaction is mechanictically similar to the Hofmann-Loffler reaction and probably typical of primary aliphatic alcohol like compounds, although equation 3 is "the only such example reported. The generally accepted mechanism for the cerium (iv) oxidation of alcohols is through unimolecular disproportionation of an alcohol- cerium (iv) complex 1? 2 21"" 26 followed by a second one electron oxidation of the radical (Scheme IV ) >

SCHEME IV

K k , v ^ - , ^ i R-O-H + Ce(lV) ^ [R-O-Ce(iv)] > Products H

Considerable evidence in the form of kenetic and spectroscopic studies on the cerium (iv) oxidation of alcohols documents the formation of the cerium (iv)-alcohol complex. Formation constants have been measured for a number of alcohols, and changes in the formation constants with changes in alcohol structure, solvent, ionic strength etc. . . have been 1 reported. ' 2*21-24,26,27

CERIUM (IV J OXIDATION OF GLYCOLS

The cerium (iv) oxidation of glycols is similar to the oxidation of alcohols. The oxidation of 1, 2-glycols leads to cleavage with the formation of the 6 14 corresponding aldehydes in very high yield. , The oxidation of 1, 2- cyclohex- anediol gives adipaldehyde in >98$ yield. 21 Complex formation in the cerium (iv) oxidation of a 1,2- glycol may be either a chelate complex, or an acyclic complex in which only one hydroxyl is coordinated with the cerium (iv). From measurements of complex formation constants (k) for a series of alcohols, 1, 2-diols and 2-methoxyalcohols, it has been shown that an adjacent hydroxy group causes a substantial increase in the stability of the complex compared to compounds with one hydroxyl 21 (Table l).

TABLE I

Substrate Complex formation constant K(M )

cis- 1, 2-Cyclohexanediol 29.

trans- 1, 2-Cyclohexanediol 18. trans- 2-Methoxycyclohexanol 2.1

Cyclohexanol 2. - 179 - Hintz and Johnson 21 interpret the data as evidence for chelate formation with 1, 2-diols. The higher K's for cis and trans 1, 2-cyclohexanediols are accounted for by added stability due to chelate formation, and the similarity of the K's for cyclohexanol and trans- 2-methoxycyclohexanol suggest that the methoxy group is not involved in chelate formation. Also, if an acyclic complex is formed similar K's for the cis and trans cyclohexanediols would be expected. The higher K for the cis glycol is accounted for in terms of a more flexable cyclic complex. Trahanovsky, however, has pointed out that the change in K's in going from the alcohol to the 1, 2-diol amounts to less than one kcal difference, and might be accounted for by a monodentate complex with the second hydroxyl 22 hydrogen bonded to a ligand (i) instead of a true chelate structure (il).

0- / M\

(I) (II)

CERIUM (IV) OXIDATION OF HYDROCARBONS As already mentioned, toluene and substituted toulenes are oxidized by cerium (iv) to the corresponding aldehydes or acetates depending on the solvent. 4 The oxidation of o- xylene by CAN, however, produces 2-methylbenzyl nitrate as the major product. 28 Apparently the non- homogeneous aqueous CAN oxidations of toluenes in general proceed via two competing pathways, one giving the carbinol and the other the nitrate ester as the initial oxidation product. In most cases the nitrate ester is hydrolyzed and further oxidized to the aldehyde, but for o- xylene the nitrate ester is stable to hydrolysis and meterial deflected to this pathway is not available for further oxidation. 28 Certain cyclopropanes are readily susceptible to reaction with one- electron 29 oxidizing agents (eq 10 ). The reaction probably involves cleavage of the carbon- carbon bond to produce a benzyl radical which is further oxidized by ligand transfer.

CAN ^ CH3 CN 0N0 2 0N0. Ar (10) R CAN > HOAc

To arrive at a better understanding of cerium (iv) oxidations of hydrocarbons, the relative reactivity of cerium (iv) toward carbon- hydrogen bonds and non aromatic carbon- carbon double bonds has been determined. The CHT system would produce the cycloheptatrienyl radical from a one- electron removal from a methylene carbon- hydrogen bond. This is a very stable radical, thus the methylene carbon- hydrogen bond is considerably weakened in this system. - 180 -

The cerium (iv) oxidation of CHT-d 8 shows a complete absence of any kinetic isotope effect. Thus, CAN reacts with nonaromatic carbon- carbon double bonds in preference even to a very reactive carbon- hydrogen bond of a hydrocarbon. 13

MISCELLANEOUS CERIUM (iv) OXIDATION Oximes and semicarbazones are oxidized by CAN to the parent carbonyl compounds in good yield (eq ll). 3 This procedure for regeneration of aldehydes and ketones from their derivitives can be carried out on a preparative scale and may be preferable to other hydrolytic procedures in some cases. Phenyldiazomethane has been reported to decompose rapidly to cis and trans 31 stilbenes when treated with catalytic amounts of CAN (eq 12). The high yield of the cis isomer is noteworthy but unexplained. The formation of oazido-P- nitratoalkanes from olefins, sodium azide and CAN has recently been reported 32 (eq 13). Addition was found to be exclusively trans and regioselective, but the details of the mechanism are not clear.

CAN > (11)

(12) CAN ->

lh$ 13%

N3 0N0 2 R 2C=CR 2 + NaN3 + CAN- *RpC— CR p (13)

RECENT USES OF CERIUM (iv) IN ORGANIC SYNTHESIS o-Benzoquinone has been produced in high yield from the eerie sulfate oxidation of catechol (eq l4). 33 Previous oxidative conversions resulted in very poor yield. 34 Cerium (iv) oxidations analogous to those reported for toluenes have been used in the synthesis of steroids. A new route to 19-nor- steroids has been reported in which CAN was used in conjunction with tris (triphenylphosphine)rhodium to replace an aromatic methyl group at C-l with hydrogen. 35 In the absence of a 1-substituent, oxidation occurs in ring C. Thus, oestrone acetate (R=Ac, R'=0) is oxidized to give the corresponding 3S 9a, liP-diol 11-nitrate in 69$ yield (eq 15. )

<& \^ (lk) - 181 -

The cerium (iv) oxidation of certain ring A-aromatic steroids has been shown to provide a new and simple way of functionalizing C-6. 37 The CAN oxidation of ij— methyloestra-1, 3, 5(l0)-"trienes produces the 6-acetate as the major product (eq 16). Preliminary results with other ring A-aromatic steroids indicate that the reaction is a general one for the formation of a 6-acetate group. The development of a synthetic method based on the selective oxidation of 38 organic free radicals by cerium (iv) has been reported. Enolizable ketones react with olefins in the presence of cerium (iv) acetate leading to the formation of "Y-keto esters (eq 17 ). The Y-keto ester is the major product, but the relative yields of the three products shown can be changed by modifying reaction conditions.

CAN -> (15)

OCOCH3 OCOCH3

CAN -> HOAc (16)

R = OCOCH3 , H, CI

Ce(iv) CH3-C-CH3 + C 6 H13CH=CH. -> C H C-CH C H HoAc 9 19 3 + S 13CH=CHCH^C-CH3 +

C6 H13CHCH2CH^ C- CH3 (17) OAc

BIBLIOGRAPHY

1. W. H. Richardson in "Oxidation in Organic Chemistry," Part A, K. B. Wiberg,

Ed., Academic Press Inc., New York, N. Y. , I965, pp 2^3-276. 2. G. M. Milne, Jr., M. I. T. Seminar Abstracts, 179 (1966). 3. R. E. Connick, J. Chem. Soc. (Suppl. 2), S235 (19U9). k. W. S. Trahanovsky and L. B. Young, J. Org. Chem., 31, 2033 (1966). 5. P. M. Nave and W. S. Trahanovsky, J. Amer. Chem. Soc, 90, Vf55 (1968). 6. G. A. Russell and R. C. Williamson, Jr., J. Amer. Chem. Soc, 86, 2357 (196U). 7. P. D. Bartlett, J. Amer. Chem. Soc, 82, I756 (i960). 8. R. L. Huang and K. H. Lee, J. Chem. Soc, C, 935 (1966). 9. H. H. Jaffe' Chem. Rev., 53, 191 (1953 ). 10. W. S. Trahanovsky, L. H. M. H. Bierman, Chem. Young, and J. Org. , Jj4, 869 (1969). - 182 -

11. S. E. Schaafsma, H. Steinberg, and Th. J. De Boer, Rec. Trav. Chim. Pays-Bas, 8£, TO (1966).

12. W. S. Trahanovsky, L. B. Yound, and M. D. Robbins, J. Amer. Chem. Soc. , 91, JOQk (1969). 15. P. Muller, E. Katten, and J. Rocek, J. Amer. Chem Soc, 95, 711*4- (1971). Ik. K. Meyer and J. Rocek, J. Amer. Chem Soc, 9k, 1209 (1972). 15. L. B. Young and W. S. Trahanovsky, J. Org. Chem., _32, 23*4-9 (1967). 16. W. S. Trahanovsky, L. B. Young, and G. L. Brown, J. Org. Chem., 52, 3865 (1967). 17. S. E. Schaafsma, H. Steinberg, and Th. J. De Boer, Rec Trav. Chim. Pays-Bas, 85/ 73 (1966). 18. W. S. Schaafsma, H. Steinberg, and Th. J. De Boer, Rec. Trav. Chem. Pays-Bas, 5068 (1969). 19. W. S. Trahanovsky and J. Cramer, J. Org. Chem., 36, I89O (1971). 20. W. S. Trahanovsky, M. G. Young, and P. M. Nave, Tetrahedron Lett., 2501 (1969). 21. H. L. Hintz and D. C. Johnson, J. Org. Chem., 32, 556 (1967). 22. L. B. Young and W. S. Trahanovsky, J. Amer. Chem. Soc, 91, 506O (1969). 23. T. R. Bala sub ramanian and N. Venkatasubramanian, Indian J. Chem., £, 36 (l97l)« 2k. D. L. Mathur and G. V. Bakore, Bull. Chem. Soc Jap., kkj 2600 (l97l). 25. P. M. Nave and W. S. Trahanovsky, J. Amer. Chem. Soc, 9^, I+536 (l97l). 26. C. F. Wells and M. Husain, Trans. Faraday Soc, 6j_, 1086 (l97l). 27. D. L. Mathur and G. V. Bakore, J. Indian Chem. Soc, kQ, 7 (l97l). 28. L. A. Dust and E. W. Gill, J. Chem. Soc, C, I63O (1970). 29. L. B. Young, Tetrahedron Lett., 5IO5 (1968).

30. J. W. Bird and D. G. M. Diaper, Can J. Chem., *££, 1*4-5 (1969). 31. W. S. Trahanovsky, M. D. Robbins, and D. Smick, J. Amer. Chem. Soc, 93* 2086 (1971). 32. W. S. Trahanovsky and M. D. Robbins, J. Amer. Chem. Soc, 93, 5256 (l9Tl).

33. R. Brockhaus, Ann. Chem., 712 , 21*4- (1968). A. Chem. (190*4- ). 3k. R. Willstatter and Pfannenstiel, Ber. , 3J_, kTkk 35. S. B. Laing and P. J. Sykes, J. Chem. Soc, C, 2915 (1968). 36. P. J. Sykes and F. J. Rutherford, Tetrahedron Lett., 3393 (1979).

37. D. M. Piatak and L. S. Eichmeier, Chem. Commun. , 772 (l9Tl). 38. E. I. Heiba and R. M. Dessau, J. Amer. Chem. Soc, £3, 53*4- (l9Tl). . )

- 183 -

A NEW TOTAL SYNTHESIS OF VITAMIN D

Reported by William Graham May 15, 1972

INTRODUCTION

The chemistry of vitamin D and its isomers has received considerable attention during the past decades. The photochemical synthesis of vitamin D (2)* and the results_of elaborate 3 " investigations in this field 1 4 have been thoroughly reviewed. The purpose of this seminar is to concentrate on the recent total synthesis of a precalciferol, and also the first total synthesis of a vitamin D, by non-photochemical methods. 5 STRUCTURE

The geometric structure of vitamin D has 6 " been known since 1936, and X-ray analysis, degradation studies, 1 3 and nmr spectroscopy8 have defined its stereochemistry. 18

.7 28 16 27 Rp = 21- 26

H 52 24 27 21 R 3 : 20 23 .-'3 HO

R = R : (l) 2 ergocalciferol (vitamin D2 )

R = R : cholecalciferol 3 (2) (vitamin D 3 )

9 The precalciferols, 12 of which precalciferol (k) is representative, occupy a position of central importance among the 9>10-seco-steroids. They are the first isolable products of irradiation of the provitamins

(e_.g. ergosterol and 7- dehydrochole sterol) , they are converted by further irradiation into their 6, 7- trans isomers, the tachysterols, 13 ' 14 and they are 15 1 transformed reversible ' by mild heating in inert solvents into the corresponding vitamins D. Of the four planar conformations conceivable for a precalciferol, the C-7 to C-10 transoid, C-6 to C-9 cisoid form should be the least hindered one from geometric considerations. Recent application of the nuclear Overhauser effect 4 has shown this to be the case.

= R R2 : (3) preergocalciferol

( precalciferol = R R 3 : (k) precholecalciferol

( precalciferol3 )

* All structures represent absolute configurations, Racemates are denoted by the prefix (+) - 184 - SYNTHESIS

In previous work on the synthesis of model compounds related" to the tachysterols and precalciferols, acetylenic intermediates18 21 have frequently been used to provide the two bridging carbon atoms between 5 rings A and C. Therefore, Lythgoe and co-workers proposed that the simplest way of generating the central _cis-double bond in precalciferol (k) would be by semihydrogenation of a dienyne. To secure the proper location of the ring A double bond and the correct stereochemistry at C(3) in the dienyne. (-)-(s)-3-ethynyl-4-methylcyclohex-3-en-l-«l (lla) (Scheme i) was used as a starting material.

Scheme I

O CH3 COaH CH< C02H 8 \

lla,R=H b,R=Si(CH 3 ) :

9a R=C02CH3 b R=CH2 OH c R=CHO

Lithium-ammonia reduction of 5-methoxy-2-methylbenzoic acid yielded (5)" the (+)-dihydro-acid 6, and subsequent acidic hydrolysis produced (i)-keto-acid 7» The racemic mixture was then resolved with quinine into

tne (+)-a- and (-)-p-enantiomers of 7. Reduction (NaBH 4 ) of the (-)-enantiomer gave lactone 8, from which ester 9a "was then obtained by methanolic sodium methoxide treatment. Reduction (L1AIH4) of 9a produced (+)-(s)-3-hydroxymethyl-4-methylcyclohex-3-en-l-ol (9b) in l4$~yield from 22 ' 23 acid 5. The diol 9b was then converted to aldehyde 9c by treatment with Mn02 in acetone. Reaction of 9c with a large excess of chloromethylenetriphenylphosphorane then gave a mixture of trans -o>- chlorodiene 10 and its cis- isomer. Finally, dehydrochlorination of the optically active mixture with NaNH2 in ammonia gave lla in over 30$ yield from 9b. e4 The source of ring C and the attached methyl group was the (+)-l-benzoate (19)25 (scheme xi) f J4-methylcyclohex-3-ene-trans-l,2-diol.

Scheme II

..GOsH /^l.-COgH ->

.COsH ^5^^NC02H \A -COpCH^±± YOsH 2 3

12 13 14 15 - 185 -

•COsR;

Mes Mes *-C02R2

OR]. 16 17a,R 1 =R2 =H l8a,R=Ac b,R3_=Ac,R2 =H b,R=H c,R=CHOCH2 CH3

CH 3

Mes=OCC6H2 (CH 3 ) 3 8

The starting material, (±)-l-methylcyclohex-3-ene -cis -l,2- dicarboxylic acid (1*0 , was prepared by diene synthesis from methyl methacrylate (13) and trans -penta-2,^-dienoic acid (12) • In order to suppress the dimerization of 12 the reaction was conducted with a large excess of methyl methacrylate at 115°. The acidic products of the reaction were then hydrolized with alkali to give a 2:1 mixture of the (+)- cis -acid ]J+ and the (+) -trans -acid 15, respectively. Heating the mixture with acetic anhydride converted both acids into the cis- anhvdride, and hydrolysis them provided the (+)~ cis -acid Ik in 55-60^ yield from 12. The enantiomers were resolved with quinine, and the (-)-acid was established to have the absolute configuration 14. Epoxidation of (-)-lh with m-chloroperbenzoic acid took place stereoselectively cis to the C-methyl group. On treatment with acid the epoxide ring was opened, possibly by internal displacement by the tertiary carboxy- group, giving the hydroxy—lactonic acid 16. Mesitoylation, ' followed by mild treatment with alkali, gave the dibasic acid 17a . Acetylation of its dibenzyl ester and subsequent catalytic removal of the benzyl groups yielded the acid 17b . Oxidative bis decarboxylation with lead tetraacetate gave the 1-mesitoate 2-acetate l8a, and then hydrolysis in methanolic sodium methoxide produced the 1-mesitoate l8b. This was caused to react with ethyl vinyl ether in the presence of a trace of p- toluene sulfonic acid to give the acetal l8c. The mesitoate group was removed by treatment with LiA.TH 4 , and then replaced by a benzoyl group. Hydrolysis of the product with acetic acid in dioxane removed the acetal residue to give the crystalline 1-benzoate 19. The (+)-enantiomer of acid Ik was also converted into (+)-19 (Scheme III), The enantiomeric 1-mesitoate 2-acetate of l8a was obtained by methods similar to those described in Scheme II. Reduction (LiAlH4 ) gave diol 20, which was converted into a 3:1 mixture of the monobenzoates 21a and 21b~7 respectively. The mixture was separated and 2ia was converted into the the p-toluenesulfonate _22. When this was treated with disodium ethanediolate in ethylene glycol, an epoxide was formed, which was then opened by invertive displacement at the allylie center to give diol 23a, •

- 186 -

Scheme III S |/ ^S-^iC02H ->

HO" RnO (-0-1^ R2 20 21a , R2 =H,R 2^Bz b_, R1 =Bz,R 2 =H

^> 19 TsO AcO

OR OR

22 23a , R=CH2CH2OH 2k&, R=CH2CH2OTs b, R=CH2CH2 OTs b, R=CH2CH2 I c, R=H

The p-monotoluene sulfonate 23b was readily obtained by selective reaction with p-toluenesulfonyl chloride and was then converted successively by standard methods into the acetate 2^a and the iodide 2^£b, which was converted, as described for the corresponding 1-mesitoate l8b, into the (+)-l-benzoate 19. This reaction sequence was a valuable source of

19 , and raised its total over-all yield from (±)-acid Ik to over 12^. 27 Now that 19 had been synthesized, the formation of the D ring of I precaliciferol 3 (h) could be completed. The l6-ketone _29, which contains four asymmetric centers, was a key intermediate in the synthesis of 35 Its double bond provides for the future functional! zation of positions 8 and 9, and its keto-group, which should be formed by the cyclization of a derivative _28 of l-methylcyclohex-3-ene- trans -l,2-diacetic acid, would allow the configuration at position 17 to be adjusted. Furthermore, the available evidence 28 suggested that the required 17(3-configuration would be the more stable of those possible. The intended route to the diacetic acid 28,

(Scheme IV) was the result of a decision to have the important ( 20R"T- or 20p configuration pre-established in the starting materials, and to maintain it throughout the synthesis. The readily available (R)-dihydro-

citronellic acid ( shown as the orthoester 26) contains configurational and structural features corresponding to those of the

isooctyl side chain, together with C-17 and C-l6 of ketone 29 * 29 To prepare _29, citronellonitrile (25) was hydrogenated to give the dihydro derivative, which was converted by way of its imino-ether into the methyl orthoester 26. Reaction with 19 gave, after mild alkaline 30 hydrolysis, the hydroxy-ester 27. Treatment with 1-dimethylamino-l- methoxyethylene, 31 followed by vigorous alkaline hydrolysis, gave the dibasic acid 28. Cyclization of the dimethyl ester afforded a 6:1 mixture by glpc of two ketones, which were separated by chromatography. The major component _29 crystallized. By treatment of the minor component, the 17»-epimer, with p-toluenesulfonic acid in acetic acid, more of the 17P- epimer was obtained as the desired ketone 29. The ketone was then converted, by way of the crystalline 8a, 9<*-epoxide, into the 8B, 9^-diol 30. The corresponding diacetate reacted with ethanedithiol in the presence of boron trifluoride-ether complex to give the corresponding ethylene thioacetal. Desulfurization with Raney nickel, followed by deacetylation, gave the crystalline trans-diol 31 in 18$ over-all yield from 19. - 187 -

Scheme IV

-> ->

£>

H

^> \ C02 CH3

28

R.

\=(

I H OH 30

51

Since dehydration of 3^a would favor the unwanted 8(lU)-olefin by both hyperconjugative and conformational factors, the proper location of the ring C double bond in 37 "was introduced by the elimination of the elements of hypochlorous acid from the trans-diaxial chlorohydrin 36b • To prepare 32 36b 93-chloro-des-AB-cholestan-8-one ( 5«-chloro-l|3-[(lR)-l,5-dimethylhexyl]- 7a(3-methyl- trans -perhydroindan-4-one) (35) was needed as a starting material. (Scheme V). The corresponding 8(3,9a-diol 31 was caused to react with p- toluenesulfonyl chloride to give the crystalline 9a-tosylate 32, alkaline treatment of which yielded the epoxide 33* Its structure was apparent from

its reaction with LiAlH4 , which gave des-AB-cholestan-88-ol. Reaction of ,33 with hydrochloric acid provided the crystalline chlorohydrin ~$k, from which 35 "was obtained by oxidation with chromic oxide. The overall yield from 31 was 73$« The enyne 11a was now converted into the trimethylsilyl ether lib. - 188 -

and its lithium derivative was brought into reaction with 35 • Scheme V

> + lib - k

35

After removal of the ether group this gave crystalline chlorohydrin 36b. 5 33 ' 34 Treatment with bis-(ethylenediamine) chromium (il) gave dienyne 37 as an oil. In spite of adverse reports 21 ' 35 on the selectivity of the Lindlar catalyst36 38 when applied to the semihydrogenation of conjugated dienynes of the kind in question, semihydrogenation was accomplished by using about ten times the recommended quantity of Lindlar ' s catalyst and 3 a considerably lower concentration of substrate (5 x 10 mole %) than is normal. The uptake of 1.0 mol. of hydrogen was then complete in one hour. There was obtained after chromatography, in addition to some recovered dienyne and minor amounts of over-reduced material, a good yield of a colorless oil, with M , [^L, and T , characteristics closely similar to 16 rnpx those of (+)-precalciferol 3 (k) , The crystalline 3, 5-dinitrobenzoate, obtained in 21$ yield overall from _35_, had properties identical with those of precalciferol 3; 5-dinitrobenzoate obtained from vitamin D 3 (2) as described by Velluz. 16 Final confirmation of identity was provided by the thermal conversion of the synthetic material into the 3,S-dinitrobenzoate 3 40 of vitamin D 3 (2) by an antarafacial 1,7 hydride shift. Identification was also confirmed by direct comparison with authentic material. In this first total synthesis of (+)-precalciferol 3 (h) , the yield-limiting starting material was ( + )-l-methylcyclohex-3-ene- cis -l,2-dicarboxylic acid (l^-), and precalciferol^, 3 , 5-dinitrobenzoate was obtained from it in 0.3^$~overall yield. BIBLIOGRAPHY

1. H. H. Inhoffen and K. Irmscher, Fortschr. Chem. Org. Naturst., IT, TO (1959). 2; L. F. Fieser and M. Fieser, "Steroids," Ch. k, Reinhold Publ. Corp., New York, N.Y., 1959- 3. H. H. Inhoffen, Angew. Chem., T2, 8T5 (i960). ,

- 189 -

4. G. M. Sanders, J. Pot, and E. Havinga, Forts chr. Chem. Org. Naturst. (Eng.)i _27> 131 (19&9) and references therein. 5. T. M. Dawson, J. Dixon, P. S. Littlewood, B. Lythgoe, and A. K. Saksena, J. Chem. Soc. (c), 2960 (1971). 6. A. Windaus and W. Thiele, Ann., 521, 160 (1936). 7. (a) D. Crowfoot and J. D. Dunitz, Nature, 162, 608 (1948). (b) D. Crowfoot Hodgkin, M. S. Webster, and J. D. Dunitz, Chem. Ind. , 1149 (1957). 8. V. P. Delaroff, P. Rathle, and M. Legrand, Bull. Soc. chim. France, 1739 (1963). 9. L. Velluz, A. Petit, and G. Amiard, ibid ., 1115 (1948).

10. L. Velluz, A. Petit, G. Michel, and G. Rousseau, Compt. rend., 226 , 1287 (1948).

11. L. Velluz and G. Amiard, ibid . , 228, 692 (1949). 12. L. Velluz and G. Amiard, ibid., £gH, 853 (1949).

13. L. Velluz, G. Amiard, and B. Goffinet, ibid ., 240 , 21 56 (1955). 14. M. P. Rappoldt, P. Westerhof, K. H. Hanewald, and J. A. Keverling Buisman, Rec. trav. chim., 77, 24l (1958). 15. A. Verloop, A. L. Koevoet, and E. Havinga, ibid ., 76, 689 (1957)* 16. L. Velluz, G. Amiard, and A. Petit, Bull. Soc. chim. France, 501 (1949). 17. L. Velluz, G. Amiard, and B. Goffinet, ibid ., 882 (1957). 18. G. N. Burkhardt and N. C. Hindley, J. Chem. Soc, 987 (1938). 19. H. H. Inhoffen and G. Quinkert, Chem. Ber., 87, l4l8 (1954). 20. H. H. Inhoffen, K. Weissermel, G. Quinkert, and K. Irmscher, ibid ., 88, 1321 (1955). 21. J. L. M. A. Schlatmann and E. Havinga, Rec. trav. chim., ~~80, 1101 (1961). 22. P: R. Bruck, R. D. Clark, R. S. Davidson, W. H. H. Gunther, P. S. Littlewood, and B. Lythgoe, J. Chem. Soc. (c), 2529 (1967).

23. J. Dixon, B. Lythgoe, I. A. Siddiqui , and J. Tideswell, ibid . , 1301 (1971).

24. T. M. Dawson, J. Dixon, P. S. Littlewood, and B. Lythgoe, ibid . 2352 (1971). 25. I. J. Bolton, R. G. Harrison, B. Lythgoe, and R. S. Manwaring, ibid., 2944 (1971). 26. R. C. Parish and L. M. Stock, J. Org. Chem., 30, 927 (1965).

27. I. J. Bolton, R. G. Harrison, and B. Lythgoe, J. Chem. Soc. (c) f 2950 (1971). 28. E. g., A. Butenandt and G. Fleischer, Ber., 70, 96 (1937). 29. C. Herschmann, Helv. Chim. Acta., 32, 2537 (19^9). 30. W. S. Johnson, L. Werthemann, W. R. Bartlett, T. J. Brocksom, T. Li, D. J. Faulkner, and M. R. Petersen, J. Amer. Chem. Soc, 92, 741 (1970). 31. A. E. Wick, D. Felix, K. Steen, and A. Eschenmoser, Helv. Chim. Acta, 47, 2425 (1964). 32. P. S. Littlewood, B. Lythgoe, and Ac K. Saksena, J. Chem. Soc. (C), 2955 (1971). t , as 33. J. K. Kochi, D. Singleton, L. Andrews, Tetrahedron, 24, 3505 (1968). 34. J. K. Kochi and D. Singleton, J. Amer. Chem. Soc, 20, 1582 (1968). 35- E. N. Marvell and J. Tashiro, J. Org. Chem., 30, 3991 (1965). 36. H. Lindlar, Helv. Chim. Acta, 35, 446 (1952). 37- Org. Synth., 46, 89 (1966).

38. R. L. Augustine, ' Catalytic Hydrogenation, ' Edward Arnold, London, p. 71, 1965. n , r s 39. R. B. Woodward and R. Hoffmann, J. Amer. Chem. Soc, 8j, 2511 (1965). 40. M. Akhtar and C. J: Gibbons, J. Chem. Soc, 5964 (1965). I )

- 190 -

KETONE TRANSPOSITIONS

Reported by Ann L. Crumrine May 25, 1972 INTRODUCTION Transposition of a carbonyl with an adjacent methylene function was of initial interest in steroid chemistry. The methods developed have been applied to terpene systems but until recently were of little general synthetic use as they relied heavily on the stereochemical features peculiar to the steroid skeleton This seminar will cover both those routes applicable to specific transpositions and those of more general utility. SYMMETRICAL INTERMEDIATES Early workers used the stereoelectronic influence of the polycyclic steroid to induce asymmetrical reactions from "symmetrical" intermediates.

1 3 6 BothA^-olefins and 2,3-epoxides , produced from readily available 3-oxo-steroids, were converted via stereoselective sequences to the 2-oxo-steroids. However, the yields resulting from such routes tended to be low due to production of isomeric materials either prior to or after formation of the "symmetrical" intermediates. An unusual, short, but unfortunately still low-yield sequence depended upon the reaction of excess n-propyl mercaptan with 2c*-bromo-3-oxo-steroids to give a 2-oxo-steroid. This apparently proceeded through a "symmetrical" 2,3-dimercaptan which can

eliminate in a number of ways, causing the low yield of about 2Fffo for one step. 4 ' 7 Autoxidation to the diketone 8 and reduction to the readily enolizable a-hydroxyketo- steroid or formation of this hydroxyketone via the a-halide " ~ 9 ' 1

18 19 ' 20 The sequence of ©-nitration or nitrosation > carbonyl reduction, and conversion of the nitrogen functionality to a ketone has been of recent interest. Corey's route, shown below (Scheme i), has both general applicability and gives reasonable yields (70^ or better).

SCHEME I OAcNOAc 1) R0N0 Cr(ll) I II || ArCCH2CH 3 2) BH 4 ArC- C-CH 3 > ArCH 2CCH 3 ^ I 65° 3) Ac 2 H J)h hrs

Alternate, quicker methods of converting an oxime to a ketone have been developed 21,S2 but tend to require more vigorous conditions.'" 3 One of the earlier methods using condensation of a ketone with 24 25 26 benzaldehyde has been rejuvenated by Bridgeman et al. ' The crucial step of the sequence is the reduction of the &-oxo-benzilidene to the benzilidene compound. This can be done efficiently in the D ring of a steroid with mixed hydride reagent (LiAlH 4 "AlCl 3 [l:3 ]) , but not in the more flexible, cyclohexane-type, A ring. p6 ' 07 A somewhat longer but still reasonable path must be used. This route is illustrated here (Scheme II for the synthesis of 2gH-pinan-J+-one. 28 SCHEME II

Zn,HOAc i v

The use of dithianes in a non- oxidative method for transposition of ketones could prove to be very useful since the conditions are mild. 29 It has been used successfully on eremophilane sequiterpenes as shown (Scheme III)

SCHEME III 1) LAH l)HC02 Et Ac 2) 2 Ca ° 9> /NH 3 3) H2 0,HgCl2 > AcO

BIBLIOGRAPHY

1. G. H. Douglas, P. S. Ellington, G. D. Meakins, and R. Swindells, J. Chem. Soc, 1720 (1959). 2. H. L. Slates and N. Wendler, J. Amer. Chem. Soc, 78, 3749 (1956). 3- T. Nakano, M. Hasegawa, and C. Djerassi, Chem. Pharm. Chem. (Tokyo), 11, J+65 (1963). k. R. L. Clarke and S. J. Daum, J. Org. Chem., 30, 3768 (1965). 5- P. D. Klimstra, R. Zigman, and R. E. Counsell, J. Medicin. Chem., 9, 924 (1966). 6. J. E. Gurst and C. Drerassi, J. Org. Chem., 29_, 5542 (1964).

7- R. L. Clarke, J. Org. Chem., 28, 2626 ( 1963)7" 8. D. Lavi, E. Glotter, and Y. Shvo, Tetrahedron, 19, 1377 (1963). 9- A. Lablache-Combier, B. Lacoume, and J. Levisalles, Bull. Soc. chim. France, 897 (1966). 10. D. H. R. Barton, D. Giacopello, P. Manitto, and D. L. Struble, J. Chem. Soc. C, 1047 (1969). 11. E. B. Hershberg, H. L. Herzog, S. B. Coan, L. Weber, and M. Jevnik, J. Amer. Chem. Soc, 74, 2585 (1952). 12. J. Elks, G. H. Phillipps, T. Walker, and L. J. Wyman, J. Chem. Soc, 4330 (1956). 13. J. H. Chapman, J. Elks, G. H. Phillipps, and L. J. Wyman, ibid , 4344 (1956). Ik. M. N. Huffman, M. H. Lott, and A. Tillotson, J. Biol. Chem., 2l8, 565 (1956). 15. M. N. Huffman, M. H. Lott, and A. Tillotson, ibid . , 217_, 107 (1955). 16. D. Barech and J. Jacques, Bull. Soc. chim. France, 67, (1955). 17. F. H. Stodola, E. C. Kendall, and B. F. McKenzie, jTOrg. Chem., 6, 841 (19IH). 18. A. Hassner, J. M. Larkin, and J. E. Dowd, J. Org. Chem., 33, 1733 (1968) 19. G. Just and Y. C. Lin, Chem. Comm., I35O (1968). 20. E. J. Corey and J. Richman, J. Amer. Chem. Soc, 92, 5276 "1970). 21. G. H. Timms and E. Wildsmith, Tetrahedron Lett.,195 (1971). 22. A. McKillop, J. D. Hunt, R. D. Naylor, and E. C. Taylor, J. Amer. Chem. Soc, 93, 4919 (1971). 23- A. McKillop, J. D. Hunt, E. C. Taylor, and F. Kienzle, Tetrahedron Lett., 5275 (1970). 24. H. H. Zeiss and W. B. Martin Jr., J. Amer. Chem. Soc, 75, 5935 (1953). 25. J. E. Bridgeman, Sir E. R. H. Jones, G. D. Meakins, and J. Wicha, Chem. Comm., 898 (1967). 26. J. E. Bridgeman, C. E. Butchers, Sir E. R. H. Jones, A. Kasal, and P. D. Woodgate, J. Chem. Soc. C, 244 (1970). 27. M. Fetizon, J. -C. Gramain, and I. Hanna, Compt. rend., 265, 929 (1967). 28. R. A. Jones and T. C, Webb, J. Chem. Soc. C, 728 (1971).

29. J. A. Marshall and H. 'Rnpb'kp. J. Over. rh^m. . ^2i. Infta Ma£a^ ORGANIC SRMINAR ABSTRACTS

1972-1973

Semester II

School of Chemical Sciences

Department of Chemistry

University of Illinois

Urhana, Illinois

SEMINAR TOPICS

II Semester 1972-73

J The Limits of Bredt's Rule: A New Interpretation 10 I David Ley

. r. The Chemistry of Diaminomaleonitrile (DAMN) and Diiminosuccinonitrile (DISN)...112 Jerrold M. Liesch

Mass Spectrometry as a Structure Tool in Carbohydrate Chemistry llU Ronald M. Stroshane

Recent Studies of the Oxidation of Olefins by Chromyl Chloride 116 Terry Balthazor

Field Ionization/Eield Desorption Mass Spectrometry 12h William C. Christopfel

The Total Synthesis of DL-Tetrodotoxin 126 Philip L. Gravel

Pyramidal Inversion Barriers: Hybridization, Electronegativity, and Other Effects 128 Tyler Thompson f> Recent Uses of Polymeric Reagents 137 Marvin S. lloekstra

The Chemistry of 7-Norbornenyl Anions 139 Gary Buske

Synthetic Uses of Dihydro-l,3-0xazines 1^1 Ving Lee

Reaction Intermediates in Decarbonylation Reactions of Carbonyl Compounds by Transition Metal Complexes 1^3 Sherrill A. Puckett

Neighboring Group Participation in Peroxyester Decomposition: A Case of Simultaneous Neighboring Croup Participation 152 Michael M. Chau

Recent Studies of Hypervalent Organic Iodine Compounds 15^- Cary Astrologes

- 104 -

THE LIMITS OF BREDT's RULE: A NEW INTERPRETATION

Reported by David Ley March 1, 1973 The importance of steric strain as a structure limiting factor in bicyclic bridgehead alkenes was recognized by Bredt 1 who suggested that a double bond at the bridgehead atoms of camphenes, pinenes, and similary constituted bicyclic compounds would impose an unnatural spatial arrangement upon the atoms Involved and these compounds would therefore suffer prohibitive 2 ' 3 strain. In modern terms, the strain results from torsional forces on the double bond which tend to twist the p-orbitals of the doubly bonded atoms away from the parallel condition required for maximum overlap. There has been much interest in determining the minimum size of the rings which can accommodate the strain of a bridgehead double bond. 4 ' 5 Stated more fundamentally, there is interest in determining the consequences of minimal p-orbital overlap as is structurally required for the formation of a bridgehead double bond. Prelog and coworkers demonstrated that larger bicyclic ring systems can accommodate bridgehead double bonds. These ring systems can be considered to be outside the limits of the applicability of the rule. The first serious 33, attempt to define the limits of Bredt ' s rule was made by Fawcett who stated that the rule covers bicyclic(x.y. z)systems where x>y>z>0 and defined S as the sum of x, y , and z. On the basis of information available, mainly Prelog' s work on bicyclic(n.3 .l)systems, 4 Fawcett proposed that bridged bicyclic ring systems with S=9 are large enough to permit isolation of bridgehead alkenes while transient reaction intermediates with bridgehead double bonds should be possible for S=T« It must be emphasized however, that all of the compounds studied by Prelog were formed under equilibrating conditions and therefore demonstrates only that the bridgehead alkenes obtained were more stable than the nonbridgehead alkenes with which they are isomeric. Thus these reports do not constitute a rigor ous_ test of the possibility of actually preparing bridgehead alkenes in smaller systems. The stability of the bridgehead double bond is dependent upon which bridge of the system contains it. This may be seen in early studies of the decarboxylation of some B-keto acids thought to involve enol or enolate intermediates which would entail the formation of a bridgehead double bond. 6 For example, 1 resists decarboxylation when heated alone or with quinoline at 7 2^0°. In contrast, 2 loses four moles of carbon dioxide when heated with 8 water in an autoclave at 220° for one hour. Finally, 3 which is a homolog of 9 2, is reported to be resistant to decarboxylation up to temperatures of 3^6°. The difference in reactivity between 1 and 2 cannot be adequately explained by Fawcett 1 s proposal which states that the strain energy is characteristic of the total size of the bicylic system. C0 H C.°2H 2 Q K \ HOoC H0 C-( C °2H 2 ) Vo2H |]^

In an attempt to rationalize these and other inconsistencies, Wiseman10 postulated that the strain of bridgehead double bonds is closely related to the strain energy of the corresponding trans -cycloalkene . This appears to be a more fundamental criterion for the qualitative evaluation of the strain of bicyclic bridgehead alkenes than that of Fawcett. The criterion also allows one to predict which branches of a bicyclic ring system can contain a bridgehead double bond. Wiseman's approach may be applied to the generalized bicyclic system shown as k. The double bond is exocyclic to ring -

- 105 -

ab and endocyclic in rings ac and be. In this structure, the double bond is trans in ring ac and cis in be and therefore should have a strain energy related to that of the trans ac system. In general, a bridgehead double bond will be more stable when it is trans in the larger of the two rings in which it is endocyclic and its stability can be estimated from knowledge about the monocyclic system. Since trans - cyclooctene is a known compound, Wiseman hypothesized that a bicyclic bridgehead double bond should be isolablewhen the larger of the two rings containing the double bond is eight-member ed. He also suggested that certain bicyclic compounds may possibly be isolable with the double bond trans in a seven-membered ring. Decarboxylation studies of bicyclic (3-keto acids and studies on the exchange of the cv-hydrogen atoms of bicyclic ketones in systems in which a trans enol or enolate is presumed to be formed give qualitative support for this view. 11 15 Wiseman tested his hypothsis by attempting the synthesis of bicyclo(3«3»l) non-1-ene (j?)> the first bridgehead alkene in which both rings containing the double bond are smaller than nine atoms . The Hofmann elimination was chosen for the final reaction of the synthetic sequence to preclude electrophilic addition to the double bond or isomerization of the double bond to another location. The amine required for this reaction was synthesized from 1-ethoxy- carbonylbicyclo(3.3«l)nonan-9-one16 by standard synthetic reactions. Prolysis of the quaternary ammonium hydroxide 6 at 1^-0° produced bicyclo(3.3«l)non-l-ene

(5) in good yield {kQffo after glpc purification). The structure of _5 was determined by its spectra and its chemical conversion to bicyclo(3«3«l) nonan-l-ol by treatment with dilute perchloric acid in aqueous acetone.

_i>

"FTMe) I~ + 3 N (Me) 3 0H" 6 > -QMs

An independent synthesis of _5 was published simultaneously by Marshall and Faubl17 using a completely different synthetic approach. Starting with l-methoxycarbonylbicyclo(3.3'l)nonan-2-one, 1:L they generated the mesylate

acid 7. Treatment of 7 with sodium hydride in dimethyl sulfoxide gave _5 (15$) and the (3-lactone 8 (30$). The use of potassium t-but oxide in dimethyl sulfoxide as above increased the yield of 5 from 7 to 25$. 18 19 The chemistry of j> has been studied by both Marshall and Wiseman.

The chemical reactivity of _5 relative to normal olefins shows its highly

strained nature. Although _5 is stable to distillation without decomposition, it reacts readily with atmospheric oxygen which cleaves the double bond. Hydrogenation over tristriphenylphosphinerhodium chloride occurs in two hours . .

- 106 -

20 •whereas normal trisubstituted double bonds reduce slowly if at all.

Treatment of _5 "with glacial acetic acid quantitatively gave bicyclo(3*3»l)- nonan-1-acetate within 25 seconds of mixing. In contrast, 1-methylcyclohexene gave no detectable addition product after 12 days under the same conditions. A highly exothermic reaction occurs between 5-and m-chloroperbenzoic acid to " 21 give the epoxide 9« Hydroboration-oxidation of j> gave a mixture of three alcohols 10, 11, and 12 in yields of 30$, 6k%, and 6$ respectively. The low regiospecificity of the reaction in contrast to most olefins is suggested to

H0^ Ha- il 12

be due to the excessive strain in _5 which causes the transition state for hydroboration to become more reactantlike

Diels-Alder adducts can be formed between _5 and 1,3-diphenylisobenzofuran

or cyclohexadiene . Cycloaddition of diazomethane followed by photolysis gave

13 which can be viewed as a strained trans -bicyclo( 6 . 1 . ) nonane with an additional methylene bridge in the larger ring. As such, 13 (and 9) must retain a portion of the strain of the alkene._5. 22 Using the reaction of _5 with acetic acid, Lesko and Turner determined the strain energy of 5 (_ca. 12 kcal/mol) to be of the same order of magnitude as the strain energies of trans- cyclooctene (9*2 kcal/mol) and cis -di-j>-butylethy- lene (9. 3 kcal/mol). Since the synthesis of 5- several other strained bridgehead alkenes have been reported. 23 27 In addition, molecules with bridgehead double bonds have been postulated as intermediates in a variety of reactions involving relatively ' 2e 35 strained bicyclic sys terns. 5 The first reported compounds with bridgehead double bonds endocyclic in seven-membered rings, bicyclo(3-2.2)non-l-ene (lA) and bicyclo(3*2.2)non- 26 l(7)-ene (15) were made by Wiseman and Chong. The distillate from the pyrolysis of the quaternary ammonium hydroxide 16 was trapped at -70° and although a powerful odor similar to that of _5 was noted, lh and _15 could not be isolated by conventional work-up procedures. Analysis of the product obtained by mass spectroscopy, nmr, and ir gave data consistant with dimer formation. Addition of acetic acid to the distillate gave 3-^-$ of the 1-acetate 17 Pyrolysis carried out in the presence of 1,3-diphenylisobenzofuran gave l8$ of the Diels-Alder adducts of as yet undetermined stereochemistry. When an nmr spectrum of the distillate was run at -80 two resonances were found for

-?

N (Me) 30H" 16 15 Ik 17

the vinyl protons. The lower field resonance was a broad triplet centered at 65*77 and the higher field resonance was a quartet at §5. hk; the ratio of the resonances was 1.0:1.3. After the mixture was allowed to warm to 0° for kO minutes and recooled, the high- field resonance disappeared completely while the low-field resonance decreased about 30$. From this data the authors concluded that the less stable isomer is formed in greater amounts. . > .

- 107 -

The difference in chemical reactivity of lk and 1_5 with 5 substantiates the postulate that the strain energy of bridgehead olefins is largely determined by the size of the larger ring rather than the total size of the bicyclic system. Support for this hypothesis comes from the work of Buchanan and 36 Jamieson. Tfte isomeric a,8-enones l8 and 19 were prepared (Scheme i) in an attempt to compare the relative strain of the two bridgehead double bonds

Scheme I TsO. NaBH, p-TsCl -> —

18 CH3

C'(0CH3 ) 2 CH 3 v

MeO (MeO MeO p-TsOH ^>

A comparison of the ultraviolet spectra and chemical shift of the vinyl hydrogen atoms may be considered to show the effect of the strain on the system (Table l). According to Wiseman's hypothesis, l8 should have less strain than 19. The spectral and chemical data tend to substantiate this. It is known that rotation about a single bond in a conjugated system reduces e and for large angles cause a hypsochromic shift of X 38 Rotation about the double bond in conjugated systems intensifies e 'and causes a bathochromic displacement of X .38*39 ^q -ultraviolet spectrum of 19 shows

a large shift of X and a near normal e . This is taken to indicate severe warping of tne double bond. The low e and normal X of 18 indicates non-planarity of the en one system. In the latter case, this has been confirmed by X-ray crystallography of the p-chloroanilide derivative 21 which shows the carbonyl group is twisted more than 37° out of plane of the olefinic double bond in the solid state. 40 The upfield shift of the B-hydrogens in 18 and 20 is attributed to the non-planarity of the system which causes an increase in the electron density at the B-carbon for 18 and _20 as compared to the planar 2-methylcyclohex-2-enone. The downfield shift of hydrogen in 19 denotes a substantial polarization (increased contribution from the ~0-C=C-(£+ resonance form). This is consistant with the hypothesis that the double bond should be less strained at C^-C^o f than at C^-C^ EtoH X 255 'max 10000 'ppm 3.3^

C 6H4-p-Cl

Isomer 18 should be resistant to 1-k addition since it would involve

the formation of the strained C 1 -C11 double bond. Attempts to carry out a Michael addition of diethylmalonate to l8 were unsuccessful. Isomer 19 was found to be too unstable for chemical characterization. On the basis of the chemical and spectral data, the authors concluded that 19 was more strained than 18 as predicted by Wiseman's hypothesis. The synthesis of bicyclo(^.2.l)non-l(8)-ene (22) and bicyclo(^.2.l)- 25 non-1-ene (23) was reported by Wiseman and coworkers. The isomers were obtained in a 5:1 ratio by the pyrolysis of the quaternary ammonium hydroxide

2k . The major isomer was shown to be _22 by spectral analysis and chemical degradation. Models indicate that _22 is more strained than 23. The predominant production of the less stable isomer is the result of kinetic control during Hofmann elimination. 41 The alkenes are most likely formed by syn elimination

of the exo hydrogens . The exo hydrogen at C-8 is rigidly held in a syn coplanar orientation while the exo hydrogen at C-2 can be in various orientations due to the greater mobility of the four-membered bridge.

->

22 23

An interesting example of a bridgehead double bond trans in two eight-member ed rings was reported by Weinshenker and Greene^3 who synthesized 9,9'-dehydrodianthrance (25). In this compound the double bond carbons may lie outside of the plane of the four substituent carbons. .

- 109 -

The first reported example of a bridgehead double bond in the one carbon bridge of a bicyclic( h. 2.1) system was made by VJarner and coworkers. 34 They

found that the initial adduct 26 of dichlorocarbene and 3>6-dihydrobenz- ' cyclobutene is thermally labile when neat or dissolved in dipolar aprotic solvents, When the neat liquid was allowed to warm to room temperature, crystals were deposited in ^0-60^ yield. X-ray analysis showed the crystals to be 27.

>

26 28 29 27

The authors suggest the only rational pathway for the formation of 27 42 is the ring opening of 26 via 28 to the transient species 29 which has a bridgehead double bond in a seven-membered ring. Further evidence for the formation of 29 was provided in trapping experiments. When one equivalent of furan was added to neat 26 and the mixture warmed to room temperature, a mixture of 1:1 adducts was formed (shown by mass spectroscopy and elemental analysis ) The formation of the transient bridgehead olefins 30 and 31 was reported by Wiseman and Chong.5 Pyrolysis of the quaternary ammonium hydroxide 32 in the presence of 1,3-diphenylisobenzofuran produced a small amount (0.; of an oil which had the spectral characteristics expected for the Diels-Alder products 33 and 3^. An independent synthesis of these olefins was carried out using a modification of Corey's synthesis of olefins. 43 Reaction of the

th i onocarbonate 3_5 "with triethyl phosphite in the presence of 1,3-diphenyl- isobenzofuran gave a solid (6 2$) which had the same R value on thin layer chromatography as the adducts produced by the pyrolysis of 32. It was reported that although the adducts from the two routes have similar spectral properties, there were significant differences in the fine structure of the nmr and ir spectra. On this basis, the authors concluded that 31 is formed in good yield from 3,5 and that both 30 and 31 are probably formed in small amounts in the thermolysis of 32.

N (i:e) 0H" 3 ^™» mtc*m 2k 32 T

^•0

§ 21 - 110 -

31,44: Keese and Krebs have reported the generation of the highly strained 1-norbornene (36) and suggested that the intermediate possibly prefers a triplet ground state. Evidence for the existence of 36 comes from the fact that the reaction of 37 or 38 with butyllithium or sodium in furan gave a constant ratio of the two adducts 39 and kO. There is some question as to whether 36 would exist in a singlet or triplet ground state. Using extended Huckel calculations based on ethylene, Keese and Krebs determined that the energy difference between the highest occupied and lowest unoccupied molecular orbital is one-half for that of untwisted ethylene. This energy difference is on the order of magnitude assumed in the extended Huckel method to be an indication of a triplet ground state. 45 However, a previous attempt at generating and trapping triplet 1-norbornene was unsuccessful. 46

In conclusion, Wiseman's hypothesis that the strain energy of bridgehead double bonds is related to the strain of the corresponding trans -cycloalkene appears to be well supported by experimental evidence. The hypothesis provides a rational criterion for qualitatively predicting the stability of various bridgehead double bonds and the probability of the formation of bridgehead double bonds as transient reaction intermediates. It is of interest to determine whether 1-norbornene exists in a singlet or triplet ground state. Both Bredt and Wiseman have predicted that the torsional strain should be too great to allow sufficient p-orbital overlap for the formation of a bridgehead double bond in this system. The existence of 1-norbornene in a triplet ground state would provide an explanation of this apparent discrepancy between theory and experimental evidence.

BIBLIOGRAPHY

1. J. Bredt, J. Houben, and P. Levy, Ber., 35, 1286 (1902). 2. J. Bredt, H. Thouet, and J. Schmit, Ann., ^37, 1 (1924). 3« For review and discussions see: a) F. S. Fawcett, Chem. Rev., 47, 219 (1950); (b) V. Prelog, J. Chem. Soc, 420 (1950); (c) R. C. Fort, Jr. and P. von R. Scheyer, Advan. Alicyclic Chem., 1, 36*4-370 (1966); (d) H. H. Wasserman in "Steric Effects in Organic Chemistry," M. S. Newman, Ed.,

John Wiley and Sons, Inc., New York, N. Y. , 1956, pp. 351-354; (e) E. L. Eliel, "Stereochemistry of Carbon Compounds," McGraw-Hill Book Co. Inc.,

New York, N. Y. , I962, pp. 298-302. 4. V. Prelog, P. Barman, and M. Zimmerman, Helv. Chim. Acta., 33, 256 (1950) and references therein. 5. J. A. Chong and J. R. Wiseman, J. Amer. Chem. Soc, 94, 8627 (1972) and references therein. 6. F. H. Westheimer and W. A. Jones, J. Chem. Soc, 5764 (1965). 7. A. C. Cope and M. W. Synerholm, J. Amer. Chem. Soc, 72, 5228 (1950). 8. H. Meerwin, F. Kiel, and E. Schoch, J. Prakt. Chem., 104, 163, 166 (1922) 9. 0. Bottger, Ber., JOB, 3l4 (1937). .

- Ill -

(1967). 10. J. R. Wiseman, J. Amer. Chem. Soc. , 82, 5966 r " 11. J. P. Ferris and N. C. Miller, ibid . , 85, 1325 (WJ). (1966). 12. J. P. Ferris and N. C. Miller, ibid ., 88, 3522 4067 (i960). 13. C. D. Gutsche and T. D. Smith, ibid . , 8|, Chemical Applications," McGraw- Hill Book 14. K. Bieman, "Mass Spectrometry, Organic 246-247. Co. Inc., New York, N. Y. , 1962, pp. 1336 (1967). 15. J. P. Schaefer and J. C. Lark, J. Org. Chem., 30, (1965). 16. E. W. Colvin and W. Parker, J. Chem. Soc, 5764 Chem. Soc. 5965 (1967). 17. J. A. Marshall and H. Faubl, J. Amer. , 8g, and H. Faubl, ibid., 92, 948 18. J. A. Marshall (19J0). ibid . (1970). 19. J. R. Wiseman and W. A. Pletcher, , 92, 956 Wilkinson, J. Chem. Soc, A, 1574 20. F. H. Jardine, J. A. Osborn, and G.

22 (1963). '21. G. Zweifel and H. C. Brown, Org. Reactions, 13, Chem. Soc, 90, 6888 (1968). 22. R. M. Lesko and R. B. Turner, J. Amer. Greene, ibid. 505 (1968). 23. N. M. Weinshenker and F. D. , £0, Chim. Fr. 3321 (I90G). 24 J -C. Brail and M. Mousseron- Camet, Bull. Soc. , J. Amer. Chem. Soc, 91, 2812 25*. J. R. Wiseman, H. -F. Chan, and C. Ahola, * (1969). 1 c \ ibid . 26. J. R. Wiseman and J. A. Chong, , £1, 7775 W). and M. I. Qureshi, Chem. Commun. , 832 (1969). 27. W. Carruthers ,-,^ c x Tatlow, Tetrahedron, 21, 2997 (1965). 28. S. F. Campell, R. S. Stephens, and J. C.

Amer. Chem. Soc , 91, 2047 (1969). 29. P. G. Gassman and R. L. Cryberg, J. Lwowski, J. Org. Chem., _36, 2864 (1971). / 30. J. 0. Reed and W. lfV7, n Int. Ed. Engl., 10, 282 (1971). 31. R. Keese and E. -P. Krebs, Angew. Chem.

. 52 (1972). 32. G. Kobrich and M; Baumann, ibid , 11, Shudo, J. Amer. Chem. Soc, £4, 7600 \ 33. P. G. Gassman, R. L. Cryberg, and K. (1972). Clardy, ibid., 9^7607 (1972). 34. P. Warner, R. LaRose, C. Lee, and J. Hirata, Tetrahedron Lett., 335 (1972). 35. M. Toda, H. Niwa, K. Ienaga, and Y. Jamieson, Tetrahedron, 28, 1123, H2o 2 )- 36. G. L. Buchanan and G. (19J Org. Chem., 30, 3019 (1965). 37. J. A. Marshall and C. J. V. Scanio, J. Applications of UV Spectroscopy, 38. H. H. Jaffe and M. Orchin, "Theory and John Wiley and Son, Inc., London, 1962, Ch. 15- Dauben, G. H. Beasley, and D. A. Cox, 39. W. E. Thiessen, H. A. Levy, W. G. Chem. Soc, 93, 4312 (1971). . . J. Amer. t ++ R Wells, Tetrahedron Lett. , 2285 U970) 40. B. G. Cordiner, M. R.~Vegar, and R. J. Chem. Soc. ,90, 5556 (1980); 41. a) M. P. Cooke, Jr. and J. L. Coke, J. Amer. 556l (1968). (b) J. L. Coke and M. C. Mourning, ibid ., 90, Fr., (1969). 42. R. Barlet and Y. Vo-Quang, Bull. Soc Chim. 3729 Amer. Chem. Soc, 43. a) E. J. Corey and R. A. E. Winter, J. & *™^*3U A. E. Winter, (b) E. J. Corey, F. A. Carey, and R. i^ > Int. Ed. Engl., 11,^'^a

Reported by Jerrold M. Liesch March 8, 1973

Diaminomaleonitrile (DAMN, 1.) and diiminosuccinonitrile (DISN, 2) are small, highly reactive, polyfunctional compounds which can be readily prepared 1 ' 2 in high yield from hydrogen cyanide and cyanogen. They are important 2 2 3 1 ' 2 ' 4 for the synthesis of hydantoins, triazines, purines, and imidazoles and 5 have great potential for novel and efficient heterocyclic synthesis. ' 6 DAMN is synthesized in low yield (< 25$) by the polymerization of hydrogen 3t cyanide, " ' ' the decomposition of the sodium salt of 1-cyanoformamide-p-toluene- 9 sulfonylhydrazone, or the addition of sodium cyanide to aminomalononitrile-p- 10 toluene sulfonate. DISN can be prepared in high yield (>}6$) by the base 2 catalyzed addition of hydrogen cyanide to cyanogen, or by passing chlorine into 2 a toluene solution of hydrogen cyanide and trimethylamine. DISN is quantitatively converted to DAMN upon catalytic hydrogenation, and converted in high yield with most reducing reagents. 1 ' 2 The chemistry of DAMN (Figure l) is consistent with that expected for its 1 ,n" 14 functionality. ' DAMN reacts with electrophiles to form mono and di-N-acyl derivatives. If the electrophile is polyfunctional, cyclic derivatives can result. Aromatic or conjugated aldehydes react with DAMN to produce monoazomethine derivatives. Diazotization of DAMN gives h, 5-dicyano-l,2,3-triazole (3) as the sole product. DAMN is transformed by photochemically induced rearrangement into Vamino-5-cyanoimidazole (k) , which is an important precursor for purine synthesis. DAMN reacts with DISN to give heterocyclic products. The structures obtained can be controlled by varying the reaction conditions. If DAMN and DISN are com- bined in tetrahydrofuran with one equivalent of sulfuric acid, aminotricyano- 1 pyrazine (j5) is the quantitative product. If trifluoroacetic acid is the solvent, tetracyanopyrazine (6) is the sole isolable product.

Figure 1 The Chemistry of DAMN.

WC^W^ WC ^N^ CN

= S, Se, C=C

NC ^-Ny H

The chemistry of DISN (Figure 2) is more unique than that of DAMN. Acid hydrolysis yields oxalic acid; whereas, hydrolysis with p- toluene sulfonic acid monohydrate 1 ' 6 in benzene yields oxalyl cyanide. Acid chlorides and anhydrides form the dx-N-acyl 6 derivatives of DISN in yields less than 10$. DISN reacts 1 " 5 with phosgene to yield k, 5-dichloro-^, 5-dicyanoimidazolidone (T), which is — )

- 113 - unstable to water, and can best be isolated as its hydrogenated product, 2-oxo-^ S- imidazoline dicarbonitrile (8). Oxalyl chloride yields a similar series of compounds lf * Alcohols displace 6 cyanide from DISN to yield dialkyloxaldiimidates. Aromatic amines act similarly affording diaryloxamidines, e but aliphatic amines give only tars f- DISN reacts with dienophiles 1 to yield the expected Diels- Alder reaction product In the case of aromatic olefins, 5 aziridine formation also occurs. The products distribu- tion depends upon the electron- donating power of the aryl group, with electron-rich systems favoring the Diels- Alder product. 6 DISN and acetone afford a low yield of 2, 2- dimethyl- 4, 5-dicyanoisoimidazole (9). The yield can be significantly improved by using catalytic amounts of DAMN or by sub- stituting 2,2-dimethoxypropane 1 ' for acetone. However, DISN plus other ketones and ketals yields a varied 6 mixture of elimination and rearrangement products. For example DISN plus 2-pentanone gives primarily 2- methyl- 2- (n- propyl)- h, 5-dicyanoisoimidazole and (10 2-methyl-3-ethyl-5,6-dicyanopyrazine (ll). On the other hand, DISN reacts with 3- methyl- 2- but anone to form 2,3-dicyano- 5 A^trimethyl-l,6-dihydropyrazine (12) and 1- i sopropyl- 2- methyl- i+,5-dicyanoimidazole (13). DISN with isopropyl amine ancTa" ketone affords the 1:1:1 adduct 2, 2- di substituted- 5-oxo-lf- (N-j.-propylamino)-isoimidazoline. 1,s In the case of acetone, the 2, 2- dimethyl product (ik) results. Figure 2 The Chemistry of DISN. NC OT

H . \ / •N

< ArN

!H.2CH 3

BIBLIOGFAFHY 1. R. W. Begland, et al., J. Amer. Chem. Soc, 93, 4953 (1971). 2. 0. W. Webster, et al., J. Org. Chem., 37, 4133 (1972). 3- a) R. A. Sanchez, J. P. Ferris, and L. E. Orgel, Science, I53, 72 (1966); b) ibid., J. Amer. Chem. Soc, ~~88, 107^ (1966); c) ibid., jTmoI. Biol., 30, 223 (1967). h. D. W. Woodward, U. S. Patent 2,53^,331 (1950); Chem. Abstr., —1+5, 5191 (1951). 5. T. Fukunaga, J. Amer. Chem. Soc, cjk, 32^2 (1972). 6, R. W. Begland and D. R. Hartter, J~0rg. Chem., 37, I+136 (1972). 7. D. W. Woodward, U. S. Patent 2,^99,^1 (1950); Chem. Abst., kk, 58981 (1950). H. Bredereck, G. Schmotzer, and E. Oehler, Ann. 600, 8l (195F]". t 9. R. E. Mosner, et al., J. Amer. Chem. Soc, 89, 5^3 (1967). 10. J. P. Ferris and R. A. Sanchez, Org. Syn., TjB, 60 (1969). 11. P. S. Robertson and J. Vaughan, J. Amer. Chem. Soc. , 80, 2691 (1958) 12. D. Shew, Ph. Do Thesis (1959), Indiana University, Diss"! Abst., 20, 1593 (1959) 13- W. A. Sheppard and 0. W. Webster, 163rd National Meeting of the American Chemical Society, Boston, Mass., April 1972, Abstract 0RGN kj>. Ik. H. Bredereck and G. Schmotzer. Ann.. &C\0. aqfia^l - 114-

MASS SPECTROMETRY AS A STRUCTURE TOOL IK CARBOHYDRATE CHEMISTRY

Reported by Ronald M. Stroshane March 15, 1973 Carbohydrates are perhaps the most ubiquitous naturally occurring compounds. Their presence in a variety of forms as monosaccharides, polysaccharides, nucleosides, glycolipids and components of many antibiotics, as well as their functional importance in many biological systems necessitates the development of methods for detection and identification of sugars. Early studies of carbohydrates using mass spectrometry were not encouraging: molecular ions were of low intensity, if observed, and the fragment ions were prolific, suggesting complex breakdown pathways. 1;2 Due to the low volatility and thermal lability of carbohydrates, the work reported to date has been 1-3 concentrated primarily on the volatile derivatives and these recent results suggest mass spectrometry is one of the fastest and most accurate ways of obtaining detailed structure assignments of carbohydrates, replacing many classical chemical and spectral procedures. Much work has been reported on the methyl ethers of simple monosaccharides and their fragmentation pathways in electron- impact mass spectrometry have been elucidated and confirmed by deuterium labeling, metastable, and high resolution studies. 1 ' 4 Additional attention has been given to the more volatile acetate esters 5 and isopropylidene derivatives have been used to determine stereochemistry in a number of cases. By careful comparison to standard spectra, polyacetate spectra can be used to distinguish between aldose and ketose as well as between pyranoid and furanoid structure. 1 * 2 Indeed comparing relative peak intensities has been used to distinguish between

en and (3 anomers, but such assignments are not unambiguous and standards must be run under identical conditions in order to make reliable assignments. Gas chromatography.mass spectrometry (GO- MS) is particularly useful in studying volatile derivatives of carbohydrates when only a small amount of sample is available. 7>Q Although peracetates can be used, much recent work has employed trimethylsilyl (TMS) ethers, dimethylsilyl (DMS) ethers,

-1 3 and trifluoroacetates (TFA) due to their greater volatility. ' Unfortunately, each of the derivatives has its drawbacks. The methyl

glycosides and methyl ethers are amenable to GC but require relatively , long reaction times for preparation as well as acidic or basic reaction conditions which can cause hydrolysis of labile substituents. Acetates are more easily prepared but require high column temperatures for good separation and their spectra are more complex than methyl ethers. The volatile TMS and DMS derivatives can be rapidly prepared on a micro scale ; and are ideally suited for GC-MS, but their spectra give many rearrangement ' 5 ions. 1 TFA derivatives seem to hold much promise as they are quite volatile and recent results seem to indicate that their breakdown products are not complex. 1 Although isopropylidene and TMS derivatives generally show sizeable M-15 peaks, molecular weight determination is complicated by the scarcity of molecular ' 9 ions in any of the derivatives. Chemical ionization 1 and field ionization techniques show promise in this regard. Indeed, field desorption has been used to give a molecular ion of an underivatized, highly non- volatile aminoglycoside antibiotic. 10 These techniques should be useful supplements to electron- impact mass spectrometry of sugar containing natural products. ,

- 115 - BIBLIOGRAPHY

1. T. Radford and D. C. DeJongh in "Biochemical Applications of Mass

Spectrometry/' G. R. Waller, Ed. , John Wiley and Sons, New York, N. Y. , 1972, Chapter 12. £ 2. H. Budzikiewicz, C. Djerassi, and D. H. Williams, "Structure Elucidation of Natural Products by Mass Spectrometry," Volume II, Holden-Day Inc., San Francisco, 196^, Chapter 27- 3. K. L. Rinehart, Jr. and G. E. VanLear in "Biochemical Applications of Mass Spectrometry," G. R. Waller, Ed., John Wiley and Sons, New York, N. Y. 1972, Chapter if.

k. K. G. Das and B. Thayumanavan, Org. Mass Spectrom. , 6, 1063 (1972). 5. J. B. Westmore, D. C. K. Lin, K. K. Ogilvie, H. Wayborn and J. Berestiansky,

ibid . , 6, 12^3 (1972).

6. D. Horton, E. K. Just, and J. D. Wander, ibid . , 6, 1121 (1972). 7. R. D. Schmid, P. Varenne, and R. Paris, Tetrahedron, 28, 5037 (1972). 8. R. D. Schmid, ibid ., 28, 3259 (1972). 9. A. M. Hogg and T. L. Nagabhushan, Tetrahedron Lett., *027 (1972). 10. W. C. Christopfel, University of Illinois Seminar, March 29, 1973- 11. K. Axberg, H. Bjorndal, A. Pilotti and S. Svensson, Acta Chem. Scand. 26, 1319 (1972). 12. F. Kagan and M. F. Grostic, Org. Mass Spectrom., 6, 1217 (1972). 13. Y.-M. Choy, G. G. S. Button, K. B. Gibney, S. Kabir, and J. N. C. Whyte,

J. Chromatogr., 72, 13 (1972). .

«

r - 116 - RECENT STUDIES OF THE OXIDATION OF OLEFINS BY CHROMYL CHLORIDE Reported by Terry Balthazor March 22, 1973 : INTRODUCTION Early work concerning the action of chromyl chloride (Cr02 Cl2 ) on olefins TO , carried out by Etard. 1 Because of the variety of products obtained^ ilclnltZ e S ^lorohydrins, expoxides, and ^chlorinated carbonyl ^the oxidationnl/+ of7olefins; by chromyl founds chloride has been considered to be of synthetic value. Recently 3 S however, Freeman et al. and Sharpless et al 4 have shown this oxidant to be a useful reagent for^ne step, high yiefl fy^theslssyntnesis of aldehydes and ketones from olefins. Sufficient kinetic and product studies have been carried out to establish the reactions proceed by initial formation of a complex which subsequently yields oxidation products on decomposition with water. The exact structure of the complex is not known, but elemental analysis 5 shows it to consist of one equivalent each of olefin and oxidant. Cr0 Cl is a very strong 2 2 2 oxidizing agent and early work noted a high degree of polymerization accompanied its use. This synthetic problem can be minimized if the reaction is carried out 3 ' 4 at low temperatures and if the complex is decomposed immediately after formation. 3 STRUCTURE AND DECOMPOSITION OF THE COMPLEX A general equation describing the oxidations of olefins by Cr02 Cl2 is shown m Scheme I. Chromyl 2 chloride is highly reactive with protic solvents and, because of this, the reaction is normally carried out in inert chlorinated solvents, although acetone 4 has recently been used. In a typical experiment, when Scheme I H2 Cr0 Cl 2 2 + Olefin > Cr02 Cl2 - Olefin > Products + Cr(lV) 1 chromyl chloride is added to a solution of olefin maintained at 0° complex 1 is formed rapidly and ~ immediately precipitates from solution. A variety of products including epoxides, ketones, aldehydes, chlorohydrins and a- chlorinated carbonyl compounds are formed on decomposition of 1 with water. Sulima and Gragerov found that 18 hydrolysis with water labeled with gave products with no excess indicating 0, Cr02 Cl2 as the oxygen transfer reagent. Iodometric titrations indicate chromyl chloride is reduced to Cr(lV) species in these reactions. Cr(iv) compounds 2 are known to be strong oxidants and may cause further oxidation after the hydrolysis of 1. To prevent this over oxidation, zinc dust sulfur dioxide 4 , and sodium bisulfite are used to reduce Cr(lV) compounds to the less reactive Cr(lll) species. Although product formation can not be used to specify the structure of 1 structure is at 2 least consistent with product formation in these reaction!' although a structure such as 3 can not be ruled out by the available data 5 01 and E ar found the „ taction of monoalkyl olefins with two equivalents of Cr0r^ Cl^ gave ^ 2 2 chlorohydrins in isolated yields of 35-50$. The chlorohydrins corresponded to those with the chlorine at the more substituted carbon of the olefin. Oxidation of cyclohexene with one or two equivalents of Cr02 Cl2 , at m CCI4, gave - trans 2- chlorocyclohexanol as the only isolated product'in

R 3 ^ / 4

cr ci 3 - 117 - this case but a complete analysis of the reaction products was not undertaken Adduct 1 was isolated in this case and decomposed with cold aqueous sodium bisulfite. On the other hand, Stairs and co-workers 9 found reaction of f equivalent amounts of cyclohexene and Cr02 Cl2 produced both cis- and trans- chlorohydrin in yields of ihfo and 21$ respectively. In addition, cyclohexanone 2 (5.5$), A - cyclohexenone (17$), cyclohexene (21$) and a high boiling unidentified material (21.5$) were found. Adduct 1 was also decomposed in situ with aqueous sodium bisulfite. Under the same reaction conditions, cyclopentene gave both cis - and trans - 2- chlorocyclopentanol and a small amount of 2- chlorocyclopentanone. 10 More recently Sharpie ss and Teranishi have found that oxidation of -78° 1, 2- di substituted olefins with Cr02 Cl2 , at in CH2 C12 , yield chlorohydrins and epoxides as major products. Of special interest is the fact that the epoxides are formed stereospecifically and the chlorohydrins stereoselectively, the result of cis- addition. Scheme II illustrates how the products might arise through an intermediate such as 2. The stereospecific formation of the epoxides requires stereospecificity throughout the reaction. Selective formation of the cis - chlorohydrin indicates that transfer of a chloride ion in the open

Scheme II

R2 a R^ .R 4

OCrOCl s

OK R

R4 carbonium ion h is faster than rotation about the carbon- carbon bond at the low temperature employed by Sharpie ss and Teramishi. Epoxide _5 might easily be formed by removal of the CrOCl2 group from 2 on hydrolysis. 33'''5 ^ ,lx . Oxidation of tri- and tetra- substituted alkenes and aryl 3 11-13 olefins " with chromyl chloride give aldehydes and ketones as the major products. Nenitzescu and co-workers found oxidation of 1-methylcyclopentene, in CCI4 at , followed by stirring at room temperature gave 1- met hylcyclopent anon in an 85$ yield. Adduct 1 was isolated and decomposed with water. Under the same reaction conditions, phenylacetaldehyde was formed in a 76.5$ yield from styrene. 12 Scheme III shows how aldehydes and ketones might arise via intermediate 2. Formulation of h with migration yields the ketone or aldehyde. This scheme 14 parallels that of opening an epoxide with a Lewis acid. Scheme III

Ri R2 R -7-R4 2 O < -> -> ^ " R^_R2R3 R4 OCrOCl2 + Cr(lV) .

- 118 - 13 Nenitzescu et al. , have recently reported the oxidation of cis- and - trans st ilbene . Diphenylacetaldehyde, _6, and benzyl phenyl ketone _7 are the major products in these oxidations. The ratios of formation of _6 to 7 are significantly different for the oxidation of the cis- and trans- isomers, For cis- stilbene the ratio of 6 to 7 is 1.3:1 while for trans- st ilbene the 13 ratio is 5-2:1. It is suggested the difference in ratios may be explained by a steric influence of a large -OCrOCl2 group. If we consider ion 8_ arising from cis- stilbene and ion £ resulting from trans- st ilbene , these conformations would minimize interaction of the -OCrOCl2 group and the phenyl group on the adjacent carbon. Preferential formation of 7. from 8_ and

6 from 9. is consistant with the experimental results.

C6H5, CeH H x 5 " " ^* OCrOCl2 J> ^aOCrOCl2 H ^ C6 H5 8 £ 15 As an alternative to Scheme III, Freeman et al. suggest that a free epoxide such as _5 may be formed from intermediate _2 and might rearrange to aldehydes and ketones under the conditions of workup. Nenitzescu11 has shown this is not in fact the case. Trans- 1-phenylpropene and trans- it 2- epoxy- 1- phenylpropane were subjected to the same reaction conditions. Products from the two compounds were the same but the proportions of the various products were quite different. An alternative suggested by Stairs9 is that the aldehydes and ketones may be formed by elimination of hydrogen chloride from chlorohydrins. 4 9 Indeed some chlorohydrins are known ' to readily rearrange in this manner but this pathway can not readily explain the ratios of phenylacetaldehyde and benzyl phenyl ketone formed upon oxidation of cis- and trans- stilbene In addition to formation epoxides, ketones, aldehydes and chlorohydrins, products resulting from cleavage of the double bonds are also found. Decomposition of adduct 1 of trans- 1-phenylpropene with cold water produces 1"T benzaldehyde in a 29% yield. When 1 is decomposed with aqueous sulfur dioxide . the yield of benzaldehyde is reduced to 9°/o, indicating cleavage occurs after hydrolysis. The most promising synthetic use of this reagent is due to Sharpless and Teranishi. 4 Oxidation of olefins with one equivalent of chromyl chloride 8 produces a small amount of o-chloro carbonyl compounds. Nenitzescu, for an example, found a 4.2% yield of 1-chloro-l-methylcyclopentanone in the oxidation of 1-methylcyclopentene with one equivalent of Cr02 Cl2 . It is 2 8 9 known that alcohols are oxidized by Cr02 Cl2 and it is suggested ' that ex- chloro carbonyl compounds may be formed via a secondary oxidation of chlorohydrins during hydrolysis. Scheme IV illustrates this reaction and is the same as that 28, proposed for the oxidation of alcohols by Cr02 Cl2 . Scheme IV

CI -HC1 ^ + CrO(OH)Cl "POCrO(OH)Cl2 /^) H , I

- 119 -

Noting this advanced oxidation, Sharpless and Teranishi carried out the oxidation of 1,2-disubstituted alkenes with a two- fold excess of Cr02 Cl2 . Under these reaction conditions ochloro ketones are obtained in good yield. r Oxidation of trans - 5-deeene, for an example, gives 6-chlorodecan- 5-one in a 90% yield. The reaction is carried out in acetone, a solvent in which 1 is soluble, and aqueous sodium bisulfite is employed for workup. Addition of zinc dust results in reduction of the ochloro ketone to the corresponding ketone in high yield. FORMATION OF THE COMPLEX Kinetic studies of the formation of adduct 1 from styrenes and alkenes

have only recently been carried out and the formation of _1 is found to be 15-18 very rapid. Freeman et al. , found the second order rate constants _1 (10°, CCI4) for styrene and cyclohexene to be 26.9 and 1.22 M sec" 1 respectively by following the disappearance of Cr02 Cl2 at 4l5 and hko mu using stopped- flow techniques. The reactions were followed until adduct 1 began to precipitate, which was after at least two half- lives. Under pseudo- first- order conditions employing a greater than tenfold excess

of styrene, the pseudo- first- order rate constant, k 3 , does not change appreciably

over a sixfold range of concentrations of Cr02 Cl2 , indicating a first- order dependence on Cr02 Cl2 . Under the same pseudo- first- order conditions a plot of k]_ vs. a tenfold range of concentration of styrene gives a straight line that passes through the origin, indicating a first order dependence on styrene and an overall second- order reaction. Similar conclusions were drawn from kinetic studies of the oxidation of cyclohexene J and 2,^-,4-trimethyl-l-pentene, 18 consistent with a mechanism involving formation of the complex by direct reaction between Cr02 Cl2 and olefin in each case.

Second- order rate constants for the formation of 1, k2 , were determined by 16 f Freeman and Yamachika for seven ring- substituted styrenes. Values of k2 ranged^ 1 from I.69 to 150.6 M •'-sec (l^CCl^ with rate constants for para- nitro- and para- methylstyrene being the lower and upper limits respectively. Correlation + of the rate constants with a substituent constants gave a p of -1-99 with a correlation coefficient (r) of 0.99^1. Correlation with Hammett's a gave a p of -1.21 with r=0.9TT^-- This suggests a transition state which involves partial bonding of Cr02 Cl2 with one or both termini of the unsaturated system, + 19 with some development of positive charge. Values of p larger than -3 generally suggest a relative large degree of carbonium ion character. 1T Freeman _et al. suggest structures _10 and 11 as possible representations

of the transition state for the oxidation of styrene with Cr02 Cl2 . Structures 10 and 11 have a high degree of bonding to the P carbon with a small to negligible

CsHgJ c6 h5 _i£; ft C6H5 v -'

CrOCl; 6 CI" ^Cl CrOCl2

10 11 12

degree of bonding to the benzylic carbon. This is also in agreement with the inverse secondary kinetic isotope effects, shown in Table 1, for deuterium substituted styrenes. The different k^/kp values in Table 1 eliminate a < symmetrical structure such as JL2, as a symmetrical structure would be expected to yield values of kVk^ nearly equal regardless of where the deuterium was located - 120 -

" Relative rates and activation parameters have been determined 1 7 for the formation of 1 for a series of substituted styrenes. Some of these are shown in Table 2. The large negative entropies of activation are consistent with a highly ordered transition state suggested by 10 and 11 .

Table 1 Secondary Deuterium Kinetic Isotope Effects

1 _1 Substrate k2 , M~ sec k k f/ r)

Styrene 2^.01 oDeutero styrene 2^.^-3 O.98

P, P-Dideutero styrene 27-18 0.88 a CCl 4 , T=0.0- .02 C

Activation parameters have been determined for twenty- two 15 18 cyclic and cyclic alkenes. > Entropies of activation ranged from ca. 20 to ^0 eu, indicating a high degree of order in the transition state. 18 For twelve acyclic alkenes, Freeman etal. found a linear correlation of the log of the relative rates with Taft's^ 1 inductive substituent constants, a . A p of -2.63 was noted with a correlation coefficient (r) of 0.991- Similarly, a p of -1.88 was obtained from a plot of o vs. log k2 for

1,2-dimethylcyclopentene, 13, 1- methylcyclopentene, Ik, and cyclopentene, 15 *

Table 2 Relative Rates and Activation Parameters

a Styrene Rel Rate AJ-f^ kcal/mol - AS^, eu

Styrene 1.0 8.1 23.8 4-Methylstyrene ^.1 k.G 32.6 o- Met hylstyrene 7 • 3 5.8 29-9 cis - P-Methylstyrene 1.9 8.0 20.1 trans - (3-Methylstyrene 8.3 5.8 29.3 cis-Stilbene 0.29 8.96 37-9 trans - Stilbene 0.l6 8.17 26.5 1,1-Diphenylethylene 12.8 3.3 35-3 Triphenylethylene 0.^-0 8.6 23.3 Tetraphenylethylene 0.01 a CCl 4 , 10°C To determine the symmetry or asymmetry of the transition state for aliphatic olefins, Freeman examined the relative rates of 13 , 1^ , and 15. In a transition state in which there is equal bonding and equal development of charge on the carbon atoms of an olefin, there will be a correlation of the log of the relative rates with the number of methyl groups substituted on the olefin. 22 The relative rates rates for 13, ik, and 1_5 are 66, 11 and 1 respectively. For a linear correlation, a value of 121 would be expected for 13. If the transition state is un symmetrical, a value of 22 would be expected for 13 from statistical considera- tions. The value of 66 does not correlate with either extreme, suggesting some asymmetry in charge distribution in the transition state for the oxidations of unsymmetrical alkenes. Freeman and Arledge 15 suggest structure l6 as a representation of the transition state for the oxidation of alkenes and cycloalkenes based on their - 121 - interpretation of the rate data for the relative reactivities of cyclohexene, "bicyclo [2.2.1] hept-2-ene and cyclopentene summarized in Table 3- Cyclohexene (IT), bicyclo [2.2.1] hept-2-ene (l8) and cyclopentene (15) were selected for /- study because of the expected influence of the ring size23 of the transition state on the relative rates of reaction. Table 3 shows an increased rate of reaction for _15 and l8 for reactions involving five-membered cyclic transitions states. This increased rate is thought to occur due to release of strain in 24 2E these reactions. > The value of k.l for the relative rate of 15 to IT is

1 ,+ i 6

i ,- b o

7 x ci ci

suggested to be in accord with the three- membered cyclic activated complexes that have values of 1.3, 1-5, and 1.25- Freeman suggests that l8 fits better with the reactions involving five-membered cyclic transition states. If this reasoning is correct, it follows that JL5 and YJ_ involve three- membered transition states, while with 1_8 there is a change of mechanism to a five-membered ring transition state. If this is correct, it should be noted that comparing the value of 511, for the ratio of l8 to IT, with other reactions involving a five-membered transition state is invalid. The change in mechanism

Table 3 Relative Reactivities of Cyclopentene, Cyclohexene and Bicyclo [2.2.1] hept-2-ene in Reactions Involving Three-

and Five-Membered Cyclic Transition States '

Reaction Size of Relative Rates Transition 15/17 18/17 State

Chromyl chloride15 k.l 511

S5> Chromic acid oxidation 3 1 -\ 5-5 b C Epoxidation 3 l-5 1.2 Dibromocarbene Addition28 ^ 3 1.25 29 e Picryl azide cycloaddition > 5 k2 8000 23 Benzonitrile oxide cycloaddition ^ 5 19 1800 30 Phenyl azide cycloaddition ^ 5 56 6500

0.002 M sulfuric acid in 96$ w/w acetic acid. Peracetic acid in acetic acid. c d Perlauric acid in chloroform. Potassium tert- butoxide and bromoform. e f Chloroform solvent. Ether solvent.

is suggested by Freeman, 15 also, to be indicated by the fact that a plot of log

(k2 , alkene/k2 , 1-hexene) vs. the ionization potentials for nine cyclic and acyclic alkenes is found to be nearly linear while compound l8 does not fit this correlation but rather its relative rate is high by a factor of _ca. 100.

The interpretation of the data in Table 3 and the implication of a change . in mechanism for l8 can be questioned. The value of k.l for the relative rate v of Cr02 Cl2 oxidation of 15 to 17 differs by factors of _ca. 10, 5, and 12 from the values for the cycloaddition reactions. Factors of ca. three separate the - 122 -

value of 4.1 from those involving three- membered transition states. It is at least questionable whether such small factors can eliminate a five- membered transition state such at 19, especially in view of the great spread in rates of the reactions compared in Table 3- The rate of reaction of Cr02 Cl2 oxidation is a factor of 10 5-107 faster than for phenyl azide and picryl 15 29 so azide cycloadditions. > > Possibly Cr02Cl2 oxidations are less selective than the cycloadditions and therefore show a small dependence of rate on the strain of the alkene. The smaller values of relative rates for Cr02 Cl2 oxidations then for cycloadditions shown in Table 3 are consistent with this postulate. 31 Erickson and Clark have recently shown that analysis of the size of a transition state by comparison of the relative rate of _15, IT and l8 must be done with caution. The relative rates, with espect to IT, for five- membered

osmium tetroxide addition to 1_5 and l8 are 21.9 and 72.3 respectively. Similarly, the relative rates for ozone addition are 3»9 and k. 3; indicating a greatly reduced selectivity for these rapid reactions. Considering the gross differences in relative reactivities found by Erickson and Clark, 31 conclusions drawn by Freeman 15 on the size of the transition state for the formation of

1. is at least questionable. CONCLUSION 15" 18 The kinetic data of Freeman, et al., at this time can not distinguish a three- or five-membered transition state leading to 1. The data does indicate a rapid, unsymmetrical electrophilic attack by chromyl chloride on the olefin. Although the structure of 1 is not known, stereochemistry of product formation indicate 2 as a reasonable working model. Future studies should

be centered on the analysis of the transition state and of the structure of _1 Recent synthetic studies by Freeman, et al., 3 and Sharpless, et al., 4 have cleared the path for use of chromyl chloride as a reagent for a one step, high yield synthesis of aldehydes and ketones from olefins.

BIBLIOGRAPHY

1. A. Etard and M. Moissan, Compt. rend., 116, kjh (1893). 2. For reports on chromyl chloride chemistry see: (a) W. H. Hartford and M. Darrin, Chem. Revs., 58, 1 (1958): (b) C. D. Nenitzescu, Bull. Soc. Chim. France, I3U9 (1968). 3. (a) F. Freeman, P. J. Cameron, and R. H. DuBois, J. Org. Chem., 33,

39TO (1968) j (b) F. Freeman, R. H. DuBois and H. J. Yamachika, Tetrahedron, 25, 3^1 (1969). k. K. B. Sharpless and A. T. Teranishi, J. Org. Chem., 38, 185 (1973)- 5. S. J. Cristol and K. R. Eilar, J. Amer. Chem. Soc, 72, ^353 (1950 ). 6. L. V. Sulima and I. P. Gragerov, J. Gen. Chem. U. S.S.R. (Eng. Trans.), 3T8T (1959). 7. R. A. Stairs, D. G. M. Diaper and A. L. Gatzke, Can. J. Chem., 46, 3695 (1968). 8. V. Psemetchi, I. Necsoic, M. Rentea and C. N. Nenitzescu, Rev. Roum. Chim., 14, 1567 (1969). 9. R. A. Stairs, D. G. M. Diaper and A. L. Gatzke, Can. J. Chem., Ul, 1059 (1963). 10. K. B. Sharpless and A. T. Teranishi, private communication. „T Nenitzescu,„ ., 11. C N. Rentea, I. Necsoiu, M. Rentea, A. Ghenciulescu and C N. Tetrahedron, 22, 3501 (1966). Roum. Chim., 12. C. N. Rentea, M. Rentea, I. Necsoiu and C. D. Nenitzescu, Rev. 12, 1^95 (1967). , ibid . 14, 13. A. Ghenciulescu, I. Necsoiu, M. Rentea and C. N. Nenitzescu, , 15^3 (1969). n , Ik. H. 0. House, J. Amer. Chem. Soc, 77, 5083 (1955). - 123 -

15. F. Freeman and K. W. Arledge, J. Org. Chem. , 37, 2625 (1972). l6. F. Freeman and N. J. Yamachika, ibid . , 92, 3730 (1970). F. and N. J. IT- Freeman Yamachika, ibid ., 9J+, 12l4 (1972). 18. F. Freeman, P. D. McCant, and N. J. Yamachika, ibid . , 92, 4621 (1970). Fo: 19- For an example see: H. C. Brown and Y. Okamoto, J. Amer. Chem. Soc. , 79, 19131Q' (1957). 20. w. F. Bayne and E. I. Snyder, Tetrahedron Lett., 2263 (1970). 21. R. W. Taft, Jr. in "Steric Effects in Organic Chemistry," M. S. Newman,

Ed , J. Wiley and Sons, Inc., New York, N. Y. , 19 56. 22. P. D. Bartlett and G. D. Sargent, J. Amer. Chem. Soc, 87, 1297 (1965). 23. R. Huisgen, R. Grashey and J. Sauer in "The Chemistry of Alkenes," S. Patai,

Ed , Interscience Publishers, London, 195^> Chapter 11. 2k. E. W. Garbisch, Jr., S. M. Schildcrout, D. B. Patterson, and C. M. Spreecher, J. Amer. Chem. Soc, 87, 2932 (1965).

25- A. K. Awasthy and J. Rocek, ibid . , £1, 991 (1969). 26. J. Boeseken and J. Stuurman, Reel. Trav. Chim. Pays-Bas, _56, 103^- (1937).

27- K. D. Bingham, G. D. Meakins, and G. H. Whitham, Chem. Commun. , kk^ (1966). 28. P. S. Skell and A. Y. Garner, J. Amer. Chem. Soc, 78. 5 J+30 (1956). 29. A. S. Bailey and J. E. White, J. Chem. Soc. B, 819 ~fl966). 30. R. Huisgen, G. Szeimies and L. Mobius, Chem. Ber., 100, 2^9^ (1967). 31- R. E. Erickson and R. L. Clark, Tetrahedron Lett., 5997 (1969). - 124 - FIELD IONIZATION/FIELD DESORPTION MASS SPECTROMETRY

Reported by William C. Christopfel March 29, 1973 Many organic compounds do not give satisfactory mass spectra by conventional electron impact (El) mass spectrometry due to thermal instability and/or the absence of observable high molecular weight ions. Although such compounds can sometimes be converted to more volatile derivatives, and mass spectra obtained at low eV, mass spectroscopists have sought more general solutions to volatility and ionization difficulties. Field ionization (Fl) and field desorption (FD) mass spectrometry offer considerable promise in this regard, and in many cases these techniques provide good mass spectra where EI fails. FI differs from EI in that the ions are produced by high electrostatic fields and not by electron bombardment. The ionization of molecules by a high electric field proceeds via electron tunneling from a ground state electronic energy level. During this ionization process the molecules do not acquire 1>2> electronic excitation energy which could lead to fregmentation. 3> 4 > 5 As a result, in many cases these relatively low energy ions undergo little or no fregmentation, frequently giving the molecular ion or M+l as the only major peak. In many cases a molecular ion is observed by FI where the EI mass spectra gave none. 4 '' >7>7a,a,9, io, n For example ^ in the EI mass spe ctra of D-glucose the molecular ion intensify is negligible while in the FI mass spectra the quasi molecular ion (M+l) is the base peak. 7 ' 73, FD mass spectra are obtained by the ion emitter dipping technique recently 12 introduced by Beckey. The emitter is dipped into a solution of the compound on which a mass spectrum is to be run, and after the solvent evaporation the solid compound residual on the emitter is the source of ions which are produced by a high electrostatic field. The FD technique eliminates the need to introduce the sample into the ion source in a gaseous state and thus thermal decomposition of organic materials prior to ionization is greatly reduced. Fragment ion peaks in the FD spectra are much smaller than in the FI spectra because of less thermal excitation. The main feature of the FD spectra is the very large relative molecular ion intensity, which is even larger than that observed in the FI spectra. For some thermally unstable compounds, molecular ions are obtained by FD but not by T ' 7a,8, FI mass spectrometry. io> 13,14 The FD masg spectra of phe-Asp-Ala-Ser-Val gives a very intense peak at (M+l) while the FI mass spectra contains no molecular ion. 10 FI and /or FD do not replace conventional electron impact mass spectrometry, but are complementary. Comparisions of FI, FD, and EI mass spectra of : nucleic acids, sugars, hydrocarbons, antibiotics, etc. . , have been reported and . point out the complementary nature of these different ionizing 6 ' 7 > 7a a,io,n, i3> techniques. 14,17 For e^^ie, the FD mass spectra of the permethylated tripeptide Ac-MeVal-MeGly-MeLeu-OMe shows only the M« peak which provides information about the molecular weight. On the other hand, the FI mass spectra of this tripeptide gives only a very small molecular ion peak, but there are characteristic fragment ion peaks which may be used for a sequence analysis. 10 Difficulties in the use of Fl/FD mass spectrometry include problems in: activation of the ionizing filament or blade, a short lifetime of the ion - producting filament or blade, difficulties focusing the ions and calibration problems for high resolution spectra. With the recent introduction of high temperature activated emitters some of these problems seem to be SOlved. 5 > 6 > 7 >7a; 14; 15>16> IT - 125 -

1. H. M. Fales "Mass Sepctrometry: Techniques and Applications/' G. W. A. Milne, Ed., Wiley- Inter science, New York, 1971, 795-8l6. 2. M. G. Inghram and R. Gomer, J. Chem. Phys., 22, 1279 (195*0- 3- R. Gomer and M. G- Inghram, J. Amer. Chem. Soc, 77/ 500 (1955). k. H. D. Beckey in "Advances in Mass Spectrometry," Vol. 2, R. M. Elliott,

Ed. , The MacMillan Company, New York, 1963, pp 1-2^. 5. J. Block in "Advances in Mass Sepctrometry," Vol. h, E. Kendrick, Ed., The Institute of Petroleum, London, 1968, pp 791-815. 6. H. D. Beckey, H. Knoppel, G. Metzinger and P. Schulze in "Advances in Mass Spectrometry," Vol. 3, W. L. Mead, Ed., The Institute of Petroleum, London, 1966, pp 35-67- 7- H. D. Beckey, Angew. Chem. (int. Ed.), 8, 623 (1969). 7a. H. D. Beckey in "Biochemical Applications of Mass Sepctrometry," George Waller, Ed., Wiley- Inter science, New York, 1971, pp 795-8l6.

8. H. U. Winkler and H. D. Beckey, Org. Mass Spectrom. , 6, 655 (1972). 9- W. K. Rohwedder, Lipids, 6, 906 (1971).

10. H. U. Winkler and H. D. Beckey, Biochem. Biophys. Res. Commun. , 46, 391 (1972). 11. J. C. Tou, Org. Mass Spectrom., 6, 833 (1972). 12. H. D. Beckey, Int. J. Mass Spectrom. and Ion Phys., 2500 (1969). 13. H. D. Beckey, G. Hoffmamm, K. H. Maurer and H. U. Winkler in "Advances in Mass Spectrometry," Vol. 5, A. Quayle, Ed., The Institute of Petroleum, London, 1971, PP 626- 631. Ik. H. R. Schulten and H. D. Beckey, Org. Mass Spectrom., 6., 885 (1972). 15- H. D. Beckey, M. D. Migahed, and F. W. Rollgem in "Advances in Mass Spectrometry," Vol. 5, A. Quayle, Ed., The Institute of Petroleum, London, 1971, pp 622-625. l6. F. W. Kollgem and H. D. Beckey, Ber. Bunsenges. Phys. Chem. J_6, 661 (1972). 17- K. L. Rinehart, Jr., The University of Illinois, personal communication, 1973-

' o

- 126 -

TIIE TOTAL SYNTHESIS OF DL- TETRODOTOXIN

Reported by Philip L. travel April 9, 1973 Since antiquity, various parts of the puffer fish, a Japanese delicacy, were known to be extremely toxic. Although attempts have been made since 1 the last quarter of the 19th century to identify the toxic component, it wasn't until 1950 that the toxin, which had come to be known as tetrodotoxin, was isolated. 2 In 196^, three groups, working independently, elucidated the structure of tetrodotoxin as octahydro- 12- (hydroxymethyl)-2-imino- 5,9:7, 10a- 3 dimethano-lOaH- [l,3]dioxocino[6, 5-d] pyrimidine- ^,7,10,11, 12- pentol (lj. Tarichatoxin, the toxin found in the skin of the California salamander, was 6 shown to be tetrodotoxin in the same year. In 1970, the chemically elucidated structure of tetrodotoxin was unequivocally confirmed by an X-ray crystal structure. The total synthesis of dl- tetrodotoxin has been accomplished by Kishi and coworkers. The strategy of their synthesis involves the initial development of a Diels- Alder reaction which produces an adduct that has the same stereochemical relationship at the ring juncture as C-^a and C-8a in _1, and that provides a functionality at C-8a which can later be converted into a 8 guanidino group. In the next stage of the synthetic plan, a model system pathway is developed, in which six asymmetric centers are stereo selectively formed in a cyclohexane ring, such that they will have the same configurations 9 as C-ha., -5, -6, -7, -8, and -8a in I. During the third stage, all asymmetric centers, except for the one corresponding to C-K, are introduced and a lactone 10 ring and a dihydrofuran moiety are formed. The lactone incorporates three- quarters of the hemilactal function and the dihydrofuran contains part of the Gguanidino ring and the other quarter of the hemilactal. The fourth and final stage completes the synthesis of tetrodotoxin by forming the cyclic guanidine and the hemilactal. 11

The first key intermediate is 2, a Diels- Alder adduct of 5- ( oximinoethyl)toluquinone and butadiene, which has a substituent that is readily convertable into a protected amine. Beckmann rearrangement of adduct 2 affords the desired amine as amide _5. A seven step reaction sequence was in amide is transformed into ether k, a model compound developed which J> which has the same asymmetric centers as the cyclohexane ring of 1. Paramount in this sequence is the early formation of the ether linkage, which causes the molecule to assume a cage- like conformation, thereby controlling the stereochemical course of the subsequent reactions. Key to the third stage

is the synthesis of epoxylactone _5 from amide 3. This is accomplished by using a series of reactions analogous to those used to synthesize ether k. Solvolysis

of _5 causes an intramolecular opening of the epoxide, forming lactone 6_. This new lactone ring bridges C-7 and C-8a, thus embodying three-quarters of the hemilactal group in 1_. Acetylation and pyrolysis of _5 gives dihydrofuran J. With key intermediate J at hand, the stage is set for the final elaboration into tetrodotoxin. In the first pathway developed, the carbon- carbon double bond is transformed into a protected diol, after which the amide is converted into a protected guanidine. Removal of the protecting groups and oxidative cleavage of the diol resulted in the formation of dl- tetrodotoxin. A second route, which gives slightly higher yields and more reproducible results, began by forming the protected guanidine first. Oxidative cleavage of the double bond and removal of the protecting groups also resulted in the formation of dl- tetrodotoxin. Keana and coworkers are also working on a total synthesis of tetrodotoxin, 12 but which, in contrast to the work of the Kishi group, begins with the formation of the guanidino group. ,

- 127 -

11 CflpOH ii,>n

."•H HoN 12 .ClIoOH

CHpOAc

OAc

2. AcO 7—0. H /;

BIBLIOGRAPHY

1. "Pharmacology and Toxicology of Naturally Oc curing Toxins." Vol. II.

H. Raskova, ed. , Pergamon Press, New York, 1971? p. 65. 2. A. Yakoo, J. Chem. Soc. Japan, 71, 590 (1950). 3. R. B. Woodward, Pure" Appl. Chem., % k<9 (196k). k. K. Tsuda, S. Ikuma, M. Kawamura, R. Tachikawa, K. Sakai, C. Tamura, and 0. Amakasu, Chem. Pharm. Bull., 12, 1357 (1964). 5- a. T. Goto, Y. Kishi, Takahashi, and Y. Hirata, Tetrahedron Lett., 779 (1964). b. T. Goto, Y. Kishi, S. Takahashi, and Y. Hirata, Tetrahedron, 21, 2059 (1965).

6. Mosher, F. A. Fuhrman, H. D. Buchwald, and H. G. Fischer, Science, l44 , 1100 (196^). 7- A. Furusaki, Y. Tomiie, and I. Nitta, Bull. Chem. Soc, Japan, 43, 3332 (1970). 8. Y. Kishi, F. Nakatsubo, M. Aratani, T. Goto, S. Inoue, H. Kakoi and S. Sugiura, Tetrahedron Lett., 5127 (1970).

9- Y. Kishi, F. Nakatsubo, M. Aratani, T. Goto, S. Inoue, and H. Kakoi, ibid . 5129 (1970). 10. Y. Kishi, M. Aratani, T. Fukuyama, F. Nakatsubo, T. Goto, S. Inoue, H. Tanino, S. Sugiura, and H. Kakoi, J. Amer. Chem. Soc, 9b, 9217 (1972). 11. Y. Kishi, T. Fukuyama, M. Aratani, F. Nakatsubo, T. Goto, S. Inoue,

H. Tanino, S. Sugiura, and II. Kakoi, ibid . , Sb, 2919 (1972). 12. J. F. W. Keana and C. U. Kim, J. Org. Chem., 36, 118 (1971). -

- 128 - PYRAMIDAL INVERSION BARRIERS: HYBRIDIZATION, ELECTRONEGATIVITY, AND OTHER EFFECTS Reported by Tyler Thompson April 12, 1973 Much of the recent interest in the process of pyramidal inversion in amines, phosphines, and ar sines has been stimulated by the possibility of introducing reasonable order into the ever increasing number of reported inversion barriers. Qualitative substituent effects have been recognized for some time, but only recently has quantitative predictive power been gained. In general, it has been observed that (l) electronegative substituents decrease bond angles and increase the s- character in the lone pair hydrid orbital, thus increasing the energetic difference between the pyramidal ground state and the planar transition state, in which the lone pair occupies a pure p orbital; 1 (2) angle strain, as in the three membered ring of aziridine, greatly increases the inversion barrier; (3) substituents that can conjugate with the lone pair lower the barrier, both by making the ground state more nearly planar, and by stabilizing the transition state; (h) bulky substituents decrease the inversion barrier by destabilizing the ground state; (5) substituents like nitrogen, oxygen, and halogens may raise the barrier due to lone pair repulsions that destabilize the transition state. 2 ' 3 The literature through 1970 has been reviewed critically, and a comprehensive tabulation of the data is available. 4 This review will cover the literature since 1970, with special focus on areas in which substantial advances have been made: (l) the effect of electronegative substituents on inversion barriers; (2) the possibility of (p— d) tt conjugation as a major contributing factor to low inversion barriers in pyramidal centers bound to a second or third row substituent; (3) the problem of distinguishing pyramidal inversion from hindered rotation around C —N bonds in acyclic amines; and (h) some moderately successful empirical correlations of inversion barriers with molecular structure. Data from the current literature, arranged according to structure, are presented in the Table. The techniques of determining inversion barriers, of which dynamic 5 nuclear magnetic resonance is by far the most important, have been reviewed sufficiently. 2 ' 6 For simple spin systems, calculation of inversion rates at the coalescence temperature T by the approximate formula k = tt hVp^/Jz has been shown7 to give values of the barrier AG* which agree with, values determined by complete line shape analysis within +0.1 kcal/mol In many cases. Observation8 of the inversion barrier of aziridine, 1, in the gas phase, 9 and comparison of the barriers of some N-alkyl aziridines in the gas phase and in several different solvents has shown that barriers determined in solvents of low polarity, like cyclohexane and toluene, are close to those determined in however, the gas phase (see 3, k, and _5, Table). The observed barrier increases, as the polarity of the solvent increases, and is typically 2— h kcal/mol higher 10 in protic solvents. The lack of significant solvent effect on the barrier of chloramines kO, kl, and k2, Cl-N(CH2R)o, correlates with their low basicity, indicating more s-character in the lone pair orbital and less ability to hydrogen bond to protic solvents. 1- Hybridization and Electronegativity 1 In a far-reaching review article in I96I, H. A. Bent developed the thesis that orbital hybridization provides a unifying principle underlying a broad range of correlations observed between chemical phenomena and such physical features as bond angles and lengths, electronegativity, and nmr chemical shifts and coupling constants. It is within this framework that correlations between structure and pyramidal inversion barriers can be best understood. In the model inversion process, a pyramidal molecule R 3M:, MM M

- 129 - with R —M—R bond angles close to 109° and the lone pair in an sp3 hybrid orbital, becomes planar in the transition state, with the R — 2 bond orbitals becoming sp and the lone pair occupying a pure p orbital. The inversion barrier can be thought of as arising from the energy necessary to rehybridize the lone pair from sp3 to p hybridization. Constricting the bond angles to less than 109°, as in three and four membered rings, effectively increases the amount of p-character in the R —M bonds, and increases the 11 s-character of the lone pair orbital. The inversion barrier is thus raised, because more energy is required to rehybridize the lone pair to a pure p orbital. Similarly, as the electronegativity of R increases, the R— bond gains p-character, and more s-character is channeled to the lone pair orbital. This increases the extension of the bonding electrons in the R— bond, putting greater electron density closer to the electronegative R; at the same time it puts the s-character in the lone pair orbital where it is more effectively used in stabilizing the molecule. Although this model is generally successful in rationalizing the trends observed in pyramidal inversion barriers, quantitative application of the principle is complicated by the difficulty of separating out other effects, such as steric interactions or various kinds of conjugation with the lone pair. To get quantitatively useful correlations one must usually compare strictly analogous compounds, in which only the property of interest is varied. With this point in mind, the effects related to hybridization can be illustrated by data from the Table. The effect of angle strain on the inversion barrier is readily seen by comparing 3? 27, 20, 22, and 2k. The order listed is the order of decreasing inversion barriers, and parallels

N-CH2 Ph

N-R -CIL W-R N-R

22, CH, 2k, CH3 3 a, Cl 23, CI 25, CI the order of increasing R--N--R bond angles. The electronegativity effect is seen in the N-chloro analogs of these compounds, 21, 23, and 25, and in the acyclic chloramines %, 37, and 40--_^2, in which the inversion barrier is raised 2-- 3 kcal/mol over the barrier for the N- methyl compounds. The high barrier in benzyl- 1- += butylfluoramine k3 ( AG 15 kcal/mol) compared with benzyl- t-butylmethylamine 31 ( &T"= 6.2 kcal/mol) gives a dramatic demonstration of the effect of a strongly electronegative element on the barrier. The anomalously low inversion barrier for the methoxyamine 39 (AG* = 9.9) compared with the hydroxylamine 38 (AG* - 12.1+) can be ascribed to the increased steric bulk of the 0-methyl group. 12 However, the barrier is still about 3 kcal/mol higher than that of dibenzylmethylamine 29. In many cases, e.g., l-tosyloxy-2,2-bis(trifluoromethyl)aziridine, the electronegativity of substituents sufficiently raises the inversion barrier to prevent observation " ' of inversion on the nmr time scale. 13 1 7 As we have seen, it is easy to find examples that give a qualitative demonstration of the electronegativity effect. But successful quantitative correlations are much more rare. The problem is not just the difficulty of distinguishing electronegativity effects from a number of other effects, but the concept of electronegativity is itself somewhat qualitative, and there are no commonly accepted electronegativity values which appear to apply to a wide variety of effects or observations. The thermochemical method of determining electronegativity of elements, originated by Pauling, has been reviewed recently, 18 and the values revised in light of the best thermochemical data available. 19 A series of approximate group electronegativities has been proposed, 20 but —

- 130 - correlations between the suggested values and some of the phenomena which may he related to electronegativity, such as nmr chemical shifts and coupling constants or Hammett suhstituent constants, are extremely limited. 2 Mislow has found a good correlation between the electronegativity of the group IVA elements as reported by Allred and Rochow21 and the 22 23 inversion barrier in phosphines kh — Vf, and arsines 60— 6H. However, these values for the electronegativity, derived from bond dissociation energies and nuclear quadrupole resonance of MC1 4 , and from nmr chemical shifts of M(CH 3 ) 4 , are unusual and seem to run counter to chemical intuition. The order of electronegativity was reported to be C>Pb>Ge>Sn>Si. Drago objected to this new scale of electronegativity, 24 attributing the observations to the increasing bulk of M and poorer bond overlap with CI and C, and 13 presented C —H coupling constants for M(CH 3 ) 4 which are inconsistent with 25 26 Allred' s proposed electronegativity scale. Allred and Rochow replied, vigorously defending their position, partly on the basis that bond overlap is a reflection of electronegativity. The debate illustrates the difficulties attending the definition of electronegativity. On balance, the Allred Rochow electronegativity values successfully correlate a large number of observations, so it appears that whatever combination of electronic properties the Allred—Rochow scale represents, these properties influence pyramidal inversion barriers, and a number of other phenomena, in a way that we may portray as the "electronegativity effect". The inversion barrier in hydrazines represent a serious exception to the general correlation of barrier height with electronegativity. Although we might expect a barrier in the range 10— 11 kcal/mol, in between that for chloramines and hydroxylamines, the observed barrier in N,N-dibenzylhydrazine and tetrabenzylhydrazine, jk and 35, is only 8 kcal/mol. This is only 1 to 1.5 kcal/mol higher than dibenzylmethylamine. Dewar and Jennings 27 attribute this small increase to lone pair repulsion, and explain the very low barrier in

76, for which no splitting of the methylene singlet was observed down to -130 ,

-° PhCH2 /CH2Ph PhCH2\ N-NH >/fl -N(CH2Ph) 2 W™* PhC /" \ H CH2ph it 35 8 76 by the absence of this lone pair repulsion. It seems likely, though, that the lack of splitting in 76- may be the result of accidental coincidence of the chemical shifts of the diastereotopic protons."" No explanation has been proposed to account for the apparent unimportance of the electronegativity of the nitrogen substituent. The Unimportance of (p— d)n Conjugation There are many examples of compounds which have "anomalously" low inversion barriers because of a second or third row substituent attached directly to the inverting center. The explanation has usually invoked greater transition state conjugation of the lone pair on the inverting atom with low- lying empty d orbitals on the substituent, typically Si, Ge, P, As, or S. 31 In most of these cases, it now appears to be unnecessary to invoke such conjugation, since the barriers can be correlated with electronegativity alone. 22 ' Besides the successful electronegativity correlations reported by Mislow, 23 there is other evidence that (p— d)rr conjugation is not important in pyramidal inversion. In the series of compounds PhEH2 , PhEH-M(CH 3 ) 3 , and ( (CH 3 ) 3Si) 2EH, 15 31 where E is N or P, and M is C, Si, Ge, or Sn, the N-H and P-H coupling constants have been found to reflect the electronegativity of M, but to exhibit a trend opposite to what would be expected if (p— d)n interaction were important. 32 A Hammett plot of the barriers of sulfenylaziridines , '

- 131 -

][, n CH3

12-18

12— 18 gives a p value close to zero, rather than positive as would be expected for a conjugative interaction with sulfur. 33 The inversion barrier of N-tosylazetidine _19 is considerably lower than that of N-methylazetidine _20, but this can be partially accounted for by the steric bulk of the tosyl group. Theoretical studies give further evidence of the lack of importance d)rr conjugation. Ab initio calculations 34 ,35 on of (p— silylamine, NH2S1H3 , 36 which has been found to be nonplanar, show that involvement of the d orbitals on Si does not lower the calculated inversion barrier significantly, and is not responsible for flattening the pyramidal angle. The reliability of these calculations is indicated by the very large basis sets used, the agreement of calculated optimum geometries with experimental geometries, and the agreement of calculated inversion barriers with observed barriers, when known. Trisilylphosphine and trigermylphosphine are both pyramidal 3 ' 3 which would be rather surprising if (p— d)TT interactions were significant, especially in trisilylphosphine, since phosphorus and silicon are both second row elements. One case in which (p— d)rr conjugation does seem reasonable, though, to is found to is when the PF2 group is attached nitrogen. (CH 3 ) 2N-PF ;? be 33 planar about nitrogen, and PF2 (CH3 )N—N(CT£3 )PF2 is either planar or at least has an immeasurably low inversion barrier at nitrogen. 40 In certain phosphine systems (2p-3p) TT conjugation of the phosphorus lone pair can lower the inversion barrier. That conjugation is significant in acylphosphines is demonstrated by _50 and 51., Ph( i-Pr)P-C(o)R for which ( the inversion barrier is about 15 kcal/nol lower than that for typical alkyl phosphines. 41 Phosphines have been found to be much less sensitive to conjugation with benzene rings, and in fact, photoelectron spectroscopy has indicated that the P lone pair is practically uncoupled to the aromatic rr-system 42 in phenyl dime thylphosphine. Conjugation appears to be especially important, however, in the phospholes 52— 57> in which the phosphorus lone pair completes

Ph H 3C

P CH(CH 3 ) 2

~H3 53 59 43 a six-electron heteroaromatic system. The inversion barrier in 53 is some 20 kcal/mol lower than that of the unconjugated analog _59. The same effect 44 is seen when arsenic is the inverting center, in arsindole 66 . Rotation vs. Inversion The barrier to Inversion of acyclic amines is considerably lower than that of cyclic amines, in which angle strain makes a substantial contribution to the barrier. In the case of an acyclic amine which contains a fairly bulky group and no electronegative substituents, the inversion barrier may be about the same as the N— C bond rotation barrier, so there is a definite 45 ambiguity in the nature of the process observed by nmr. In only one case has both inversion and rotation been observed by nmr in a single molecule. N-Benzyl-t-butylchloramine, 36? has an inversion barrier AG* = 9«0 kcal/mol and a t-butyl rotation barrier AG* = 8.3 kcal/mol. In the related chloramines ( k0 , kl j and k2, C1M(CH R) , inversion is indicated to be the observed process 2 2 10 by the fact that the barrier decreases with increasing substituent bulk. - 132 -

The ambiguity is illustrated by the case of N,N-dibenzyljnethylamine, 29. The appearance of the diastereotopic benzyl methylene protons as an AB quartet could be caused either by slowed inversion of nitrogen or by hindered 46 rotation around the N— CH2 bond. This possibility led Ingold to ascribe the process in _29 to hindered rotation, on the basis of similar barriers found in N,N-dibenzyl-t-butylamine, and especially in 68, for -which two separate barriers were observed. AG = 7^8 kcal/nol for coalescence of the benzyl

.CHaPh

( CH 3) 3C- CH2W,

methylene protons, but AG = 8.9 kcal/mol for the neopentyl group. Furthermore, complete line shape analysis revealed that, although AS* for the neopentyl group was near zero, for the benzyl group AS+ = -36+13 eu, giving an enthalpy of activation APT =3.1 kcal/mol. This large negative AS* for the benzyl group could be rationalized by restriction of the phenyl ring during rotation, to minimize interaction with the neopentyl group. It should be recognized, however, that these examples may not be germane to the problem of 29, since all of Ingold' s examples bear bulkier substituents which should lower~the inversion barrier and raise the rotation barrier. In addition, 68 contains a phenyl group attached directly to nitrogen, lowering 47 the inversion barrier drastically through lone pair conjugation. I5 fact, 18 complete line shape analysis" of _29 reveals an activation entropy AS' near zero {k + 3 eu) as expected for inversion. The first order inversion rate of 49 ' 5 29 derived from second order protpnation studies in water ' also gives a "similar barrier for inversion ( AG^ =6— 7 kcal/mol), and is not complicated by the rotation problem. The difficulty in applying Ingold' s results to the problem of the barrier in 29 illustrates the pitfalls involved in trying to separate the contributions of several different, opposing effects on inversion barriers. A common inversion—rotation transition state has been proposed for the t-butylamines 30--33; In which AG for coalescence of the methylene protons

= CH , CH0CD3, CD(CD )o, or CH Ph (CH 3 ) 3C-N Ri,R2 3 3 2 ^ Rs 30-33 is identical to that for coalescence of the three overlapping t-butyl methyl singlets. These results, however, are completely consistent with rapid inversion but slow rotation about the N— CH2 bond, the latter in concert with rotation about the N C(CH bond. — 3 ) 3 53 It should be noted in this context that the recently reported barrier 4 for the phosphorylated aziridine 9 is some 8 kcal/mol higher than the value

I t>-PPhj 69

for the analog 69. The possibility that the observed process in 9 is actually hindered rotation about the N— P bond should be considered. Quantitative Correlations 54 It has been suggested that a correlation may exist between the inversion barrier of amines and the N— CO rotation barrier in the corresponding amides, but the example of the high inversion barriers in the bicyclic amines 26, 27, and 28, together with normal rotation barriers of 17 to 18 kcal/mol in /

their N-acyl derivatives, demonstrates the shortcomings of such a correlation. 55 56 J Kessler " has reported a useful empirical correlation for predicting the inversion barrier in amines. He recognized that the ratio between the . free energies of activation for any two substituents is generally constant \ from system to system. Thus the ratio AG^(N-Cl)/AG^(N-CH3) is close to 1.28 for a wide variety of nitrogen systems. Using N-CH 3 as the standard, the relationship takes the form ^*(N-X)/A(r (N-CH 3 ) = Z . Mislow pointed out that, after some algebra, Kessler' s relationship proves to be an approximate expression of the standard Hammett free energy relationship, in which Z is 59 just the ratio of the appropriate p values. Mislow also recognized that inversion barriers could be reasonably well correlated with the bond angles of pyramidal amines. 60 Although molecular orbital calculations are in principle capable of giving reliable estimates of inversion barriers, 61 in practice the successes have been few. Ab initio calculations on small molecules are capable of giving excellent results. The recently calculated barrier62 of 6.9 kcal/mol for ammonia compares favorably with the reported value of 5.8 kcal/mol. For molecules of practical interest, though, semiempirical schemes must be used. Despite the usual claims of each author that his calculations produce numbers which are close enough to reality to be "useful", 47 ' 63 ' 64 only two studies have given really practical results. Dewar, 65 using the MLNDO/l scheme to calculate the inversion barriers of a number of amines, found generally the correct relative order of barrier heights, with the absolute magnitude of the barrier usually correct within 3 kcal/mol. The most successful calculations 6S however have been done by Mislow, et al . using the Pople CNDO/2 method especially parametrized for inversion barriers. For many molecules the calculated and observed barriers agree within 0.5 kcal/mol, and for most molecules the agreement is within 2 kcal/mol. Large substituent effects, such as angle strain and electronegativity, are reasonably well reproduced. Relying heavily upon these calculated values, but also drawing data from the literature where available, Mislow has discovered several linear free energy relationships that correlate the inversion barriers of one series of compounds, e.g.phosphines, with the barriers for another series, e.g. amines, arsines, 67 or sulfoxides. This work appears to hold the greatest promise for putting the prediction of pyramidal inversion barriers on a solid quantitative basis.

TABLE. RECENTLY DETERMINED PYRAMIDAL INVERSION BARRIERS.

Substituent _ _ , _ o~a ._ .% . _ _ No. / x Solvent T, C Kcal/mol Ref. R2 • \ % > } • •

1 H, H gas 68 17.3IT. 3 8 R2 2 D, H gas 79 17.917-9 8 v-Ar2

CH , H gas 3 3 117 19.0 9 1 R 1-5, 10-18 toluene 112 19.1 9 x Me BSi4 4 112 19.2 9 DMSO-d 125 20.1 6 9 R2 Ri DoO >150 -53 9 ' H^ C Me gas 53 17.1.7.1 9 2 toluene 60^ 17.5 6 b 9 R3 >7,8 Me 8Si 4 4 58 17.6 9 acetone 72 18.0 9 f Ph H CDCI3 76 18.7 9 methanol 97 19.5 9 H-V^Me i-Pr, CH 3 gas kj>hj> 16.U16.1+ 9 Ph2PO toluene 38 16.6 9 a) In most cases, the coalescence temperature. For CLS or CK method, T is the temperature at which the rate was determined, b) Complete line shape analysis. C 3 d

- 134 -

Me aSi 4 4 41 16.8 9 CDC13 60 17.7 9 methanol 82 18.6 9 <^-so2 -/oVh 106' D2 20.1 9 19 6 H, Me, Me C c Dg 4o 16.0 68 7 II, Me, CH2Ph CH2C12 25 15.0 68 -R 8 Me, H, Ph CSo -58 10.7 68 9 15.6 53 20, 21

10 SCC1 3 , CH 3 -86 9.2 69

11 SCF3 , CH 3 -61 10.4 69

12 SPh-4-0CH 3 , CH 3 CHoC lp -20 12.6 33

13 SPh-4-CII 3 , CH 3 CHoCIjd -21 12.5 33 ii+ SPh, CH 3 CH C 1 -23 12.4 33 2 2 R 15 SPh-4-Cl, CH 3 CH2C12 -21 12.5 33 0~ 16 SPh- 4- Br, CH3 CH2C12 -24 12.4 33 IT SPh-l+-N02 , CH 3 CH2C12 -17 12.8 33 22, 23 24, 25 -10 18 SPh-2,4-(N02 ) 2 , CH3 CH2C12 13.8 33 19 CHC1F 6.2 2 70 Ph -69* N- CH2 20 CH 3 CHC1F2 10.0 71 21 CI CHC1F2 -20 13.4 71 22 CH &-d. CDClFo -98* 8.3 71- 3 , " 23 CI CFClo -68* 10.3 71 26 N- CH Ph 24 CH 3 CHC1F2 -125 6.8 71 2 25 CI CHC12 F -87b 9-0 71 b CH2 C1 2 -86 9.2 71 C02H CD 30D -91 8.9 71 26 CDC1 26 14.8 C0 2H 55 N- CH2Ph 27 CDC1 3 -18 12.6 55 28 C5D5N 17 14.3 55 b 29 (PhCHa ) 2NCII 3 H2C=CHC1 -I35 6.7 72 27 -l4ib 6.6 48 d HoO 25 ~6-7 49 ,Ri CD CH CD 30 3 , 2 3 -130 7.2 51 t-bu-N 30-33 31 CH 3 , CII2Ph -138 6.2 52 32 both ai2 CD 3 -160 5-8 51 ~ 33 CH2 CD 3 , CD(CD 3 ) 2 l8 <4.4 51 34 (PhCH \13 2 ) 2NNH2 CC12 F2 -106 "~8.o 27 Ph,

35 [(PhCH2 ) 2N] 2 CHC1 F -105 8.2 27 2 P-M(CH3 ) 3 44- 47 b / 36 ClN(CH2Ph)(t-bu) -84 9.0 45 R b 37 ClN(CH 3 )CH2Ph acetone- 6 -66 10.3 12 b 38 II0N(CH 3 )CH2Ph acetone- d6 -35 12.4 12 Ph \ 39 CH 0N(CH )CH Ph acetone- -83b 3 3 2 6 9.9 12 k8 31 ^ 011 ^ 3 4o i--?^- C1N(CH2 CH 3 ) 2 CS2 -71 10.2 10 SI(0CH 4i ClN(CH2Ph) 2 CHFC12 -71.5 9-9 10 / 3 ) 3 t-buSi(CH 42 3 ) 2-Px ClN(i-PrCH2 ) 2 CHFC1 2 -81 9-5 10 \ Si(CH 3 ) 3 43 FN(t-bu)CH2Ph acetone- d.6 34 15.1 54 49 Ph cir C 44 3 , c Decalin 130 32.7 22 45 i-Pr, Ge P- CCH< 50 CSD 6 109 21.4 22 i Pr/ 46 i-Pr, Sn C6D6 72 19.3 22 47 i-Pr, Si Dioxane 62 18.9 22 b 48 57 17.1 73 49 <-8o < 10.4 74 /P-C-A-Ph 51 50 l-C 10 H7Br 110 19.4 4l H 51 1- 10 H7Br 100 19.3 4i OCHq c) Classical kinetics (CK). d) Derived from second order protonation rate C 38

135 -

52 i-Pr, Ph, Ph CFC1 3 1.8 >15 kl b 53 CH 3 , i-Pr, Ph CFCI3 42. 16.1 43 5'+ CII3, i-Pr, CH2 CH2Ph CFCI3 11* 16.1 43 55 Cn;., CD 2 CH(0Me), Ph CFCI3 1-3 15.3 43 Ph

56 P, CDpCH(0Me)Ph C6H 5 CN 153-8 23.3 43 C 57 CSH6 59- 26.3 k3 C 58 P CqDq 130 35-3 k3 c 59 CbH6 170 36.5 43 c CD2 CH(OMe)Ph 6o CH3As(Ph)CH2 CH3 decalin 2l8 43.1 75 6l Ph(i-Pr)As-SiMe neat i8ib 25.1 23 3 . ^-CH , -^CHs b ; 62 Ph(i-Pr)As-CeMe 3 neat >200 >27 23 19lb V / \ 59 63 Ph(i-Pr)As-SnMe 3 neat 25.9 23 -Ph

64 PhAs(SiIMe 2 ) 2 neat 17.7 23 58,65 \ e Ph 65 As k6-kQ 27 lkyl C 66 As, CD2 CH(0Me)Ph toluene- da 151 35.2 27 b 67 As, SiHMe 1- 10 n7Br l6l 2k. 3 27 2 Ph ^ 56,66,67 BIBLIOGRAPHY

1. H. A. Bent, Chem. Rev., 6l, 275 (l96l).

2. J. M. Lehn, Fortschr. Chem. Forsch. , 15, 311 (1970). 3. A. Rauk, L. C. Allen, and K. Mislow, Angew. Chem. Intl. Ed. Engl., £, ^00 (1970). Ik J. B. Lambert, Topics in Stereochemistry, vol. 6, N. L. Allinger and E. L.

Eliel, ed. , Wiley- Interscience, New York, N. Y. , 1971, P- 19-

5- G. Binsch, ibid . , vol. 3, 1968, p. 97- 6. C. T. Mabuni, Univ. of Illinois Organic Seminar Abstracts, spring 1968, p. 32.

7. D. Kost, E. H. Carlson, and M. Raban, Chem. Commun. , 656 (1971). 8. R. E. Carter and T. Drakenberg, ibid ., 582 (1972). 9- T. Drakenberg and J. M. Lehn, J. Chem. Soc. Perkin II, 532 (1972). 10. W. B. Jennings and R. Spratt, Chem. Commun., 54 (1971)- 11. M. Randic and Z. B. Maksic, Chem. Rev., 72, 43 (1972). 12. D. L. Griffith, B. L. Olson, and J. D. Roberts, J. Amer. Chem. Soc, 93, 1648 (1971). 13. R. G. Kostyanovskii, G. K. Kadorkina, and A. A. Fomichev, Izv. Akad.

Nauk SSSR, Ser. Khim. , 1672 (1972). Ik. R. G. Kostyanovskii and G. K. Kadorkina, ibid., 1676 (1972). 15- R. G. Kostyanovskii, Yu. I. El'natanov, L. M. Zagurskaya, and K. S. Zakharov,

ibid . , 1893 (1972). 16. J. Cantacuzene and J. Leroy, J. Amer. Chem. Soc, _93, 5263 (1971).

17. D. K. Kang, K. C. Servis, and A. B. Burg, Org. Mag. Resonance, 3., 101 (1971). 18. V. Gutmann and U. Mayer, Struct. Bonding, 10, 127 (1972). 19- A. L. Allred, J. Inorg. Nucl. Chem., 17, 215 (1961). 20. P. R. Wells, in Prog. Phys. Org. Chem., A. Streitwieser and R. W. Taft, ed., 6, 111_(1968).

21. A. L. Allred and E. G. Rochow, J. Inorg. Nucl. Chem., j?, 269 (1958). 22. R. D. Baechler and K. Mislow, J. Amer. Chem. Soc, 93, 773 (1971). 23- R. D. Baechler, .J. P. Casey, R. J. Cook, G. H. Senkler, Jr., and K. Mislow,

ibid . , 9k, 2859 (1972). 2k. R. S. Drago, J. Inorg. Nucl. Chem., 15, 237 (i960).

25- R. S. Drago, Rec Chem. Progr., 26, 157 (19^5) • 26. A. L. Allred and E. G. Rochow, J. Inorg. Nucl. Chem., 20, 167 (l96l). 27- M. J. S. Dewar and W. B. Jennings, J. Amer. Chem. Soc, £5, 1562 (1973). 28. S. Wolfe, A. Rauk, L. M. Tel, and I. G. Csizmadia, J. Chem. Soc. B, 136 (1971). 29- S. Wolfe, Accts. Chem. Res., 5, 102 (1972). . ,

- 136 -

30. Ref. k, p. 35-36. 31. Ref. k, p. J+3 32. P. G. Harrison, S. E. Ulrica, and J. J. Zuckerman, J. Amer. Chem. Soc. £3, 2307 (1971). 9k, 33. D. Kost, W. A. Stacer, and M. Raban, ibid . , 3233 (1972). 3k. J. M. Lehn and B. Munsch, Chem. Commun., 99^"Tl970 )

35. G. Szeimies, Tetrahedron Lett . , 5073 (1970). 36. C. Glidewell, D. W. H. Rakin, A. G. Robiette, and G. M. Sheldrick, J. Mol. Structure, 6, 231 (1970 ). 37. R. Beagley, A. G. Robiette, and G. M. Sheldrick, J. Chem. Soc. A, 3002 (1968), 38. H. Siebert and J. Eints, J. Mol. Structure, k, 23 (1969). 39. P. Forti, D. Damiani, and P. G. Favero, J. Amer. Chem. Soc. 95, 756 (1973). kO. T. T. Bopp, M. D. Havlicek, and J. W. Gilje, ibid . , 93, 305l"Tl97l). kl. E. Egan and K. Mislow, ibid . , 93, 1805 (1971). k2. W. Schafer and A. Schweig, Angew. Chem. Intl. Ed. Engl., 11, 836 (1972). k-3. W. Egan, R. Tang. G. Zon, and K. Mislow, J. Amer. Chem. Soc, 93, 6205 (1971) kk. R. II. Bowman and K. Mislow, ibid ., 9^, 286l (1972). l<-5. C. H. Bushweller and J. W. 0'Neil, Tetrahedron Lett., 3^71 (1971). k6. D. Brownstein, E. C Hor swill, K. U. Ingold, J. Amer. Chem. Soc, 92, 7217 (1970). k"7. W. J. Hehre, L. Radom, and J. A. Pople, Chem. Commun. , 669 (1972). 48. C. H. Bushweller, J. W. O'Neil, and H. S. Bilofsky, Tetrahedron, 28, 2694 (1972). k-9. T. P. Pitner and R. B. Martin, J. Amer. Chem. Soc, 93, 4400 (1971). 50. D. E. Leyden and W. R. Morgan, J. Phys. Chem., 75, 3190 (1971). 51. C. H. Bushweller and W. G. Anderson, Tetrahedron Lett., 129 (1972). 52. C. H. Bushweller, J. W. O'Neil, and H. S. Bilofsky, J. Amer. Chem. Soc, 93, 5^2 (1971). 53. K. D. Berlin, S. Rengaraj'u, and P. E. Clark, J. Heterocycl. Chem., 7, 1095 (1970). 54. J. Cantacuzene, J. Leroy, R. Jantzen, and F. Dudragne, J. Amer. Chem. Soc, 94, 7924 (1972). 55. W. J. Deloughry and I. 0. Sutherland, Chem. Commun., 1104 (1971).

56. II. Kessler and D. Leibfritz, Tetrahedron Lett., 4289 (1970). 57. H. Kessler and D. Leibfritz, ibid ., 4293 (1970). 58. H. Kessler and D. Leibfritz, ibid ., 4297 (1970). 59. J. Stackhouse, R. D. Baechler, and K. Mislow, Tetrahedron Lett., 344l (1971). 60. J. Stackhouse, R. D. Baechler, and K. Mislow, ibid., 3^37 (1971). 61. L. C. Allen and J. Arents, J. Chem. Phys., 57-, 1818 (1972). 62. R. E. Kari and I. G. Csizmadia, ibid ., _5§, "4337 (1972). 63. M. Shanshal, Theor. Chim. Acta, 21, l49 (1971).

64. II. Petersen, Jr., and R. L. Brisotti, Jr., J. Amer. Chem. Soc, 93, 346 (1971). 65. M. J. S. Dewar and M. Shanshall, J. Chem. Soc, A, 25 (1971). 66. A. Rauk, J. D. Andose, W. G. Frick, R. Tang, and K. Mislow, J. Amer. Chem. Soc, 93, 6507 (1971).

67. R. D. Baechler, J. D. Andose, J. Stackhouse, and K. Mislow, ibid . , 9k , 8060 (1972). 68. J. L. Pierre, P. Baret, and P. Arnaud, Bull. Soc. Chem. Fr., 3619 (1971). 69. M. Raban and D. Kost, J. Amer. Chem. Soc, 9k, 3234 (1972). 70. J. B. Lambert, B. S. Packard, and W. L. Oliver, Jr., J. Org. Chem., 36, 1309 (1971). 71. J. B. Lambert, W. L. Oliver, Jr., and B. S. Packard, J. Amer. Chem. Soc, 93, 933 (1971).

72. M. J. S. Dewar and W. B. Jennings, ibid . , 93, 401 (1971). 73. R. D. Baechler and K. Mislow, Chem. Commun., 185 (1972). 74. 0. J. Scherer and R. Mergner, J. Organometal. Chem., 40, C64 (1972). 75. G. H. Senkler, Jr., and K. Mislow, J. Amer. Chem. Soc, 9k, 291 (1972). ( .

- 137 -

RECENT USES OF POLYMERIC REAGENTS

Reported by Marvin S. Hoekstra April 23, 1973 The use of polymeric supports in chain-building reactions was established by Merrifield's introduction of solid-phase polypeptide synthesis in I960. This approach has been used since then by a number of groups for oligonucleotide1 and oligosaccharide 2 syntheses. However, solid-phase methods had found little application in other areas of synthesis until recently. Polymer supports offer solutions to a number of problems encountered in the workup of synthetic reactions. Thus a reaction may be designed in such a way that a product, undesirable by-product, or excess of reagent may be selectively removed via its linkage to an insoluble polymer. In reactions such as cyclizations, in which intermolecular interactions are undesirable, the substrate can be attached to the polymer in such a way that infinite dilution is approximated, 3 allowing only the intramolecular cyclization to proceed. In many cases polymeric reagents can be regenerated and reused repeatedly. With few exceptions the polymers on which the reacting species has been mounted are functionalized polystyrenes, cross-linked with divinylbenzene to provide the requisite bulk mechanical properties and insolubility. The amount of cross-linking has been shown to influence the size of molecules which 4 can be admitted into the thus formed "solvent channels". Functional groups used for attachment reside on the phenyl ring and typically allow for linkage via nucleophilic displacement. Weinshenker and Shen have reported the use of a polymeric carbodiimide as the dehydrating agent in the Moffat oxidations of primary and secondary alcohols. 5 The urea by-product, which is generally difficult to separate from the product in the monomeric reaction, is retained on the polymer. Water- soluble carbodiimides had previously been used to solve this workup problem, but the polymeric carbodiimide is superior in that it may be used with water- sensitive compounds Some less advantageous reactions include oxidations of secondary alcohols to ketones using N-chloropolyamides, giving yields comparable to those previously 6 reported for the analogous N-bromoacetamide and N-bromohydantoin oxidations, and epoxidations of olefins with a percarboxylic acid polymer which can be regenerated and reused, but gives yields generally inferior to those available using m-chloroperoxybenzoic acid. A~~number of reductions have been performed employing catalysts such as compound 1 and its analogs. Such catalysts, patterned after the homogeneous

catalyst tr is ( triphenylphosphine ) chlor orhodium but incorporating the advantage 4 , a , 9 of recoverability by filtration, have been used in the hydrogenation of olefins. The rate of reduction decreases with increasing olefin molecular weight, which limits the usefulness of this method in the case of such olefins as those in the steroid series, an area in which tris( triphenylphosphine) chlor orhodium was most useful. A polymer-mounted titanocene di chloride has been found to be an effective hydrogenation catalyst, whereas the monomeric compound has little 10 activity due to rapid aggregation in solution. Other reductions using 1 and its analogs include hydros ilylation and hydroformylation of olefins in varying yields. In Wittig olefinations, p-(diphenylphosphino) -polystyrene has been used in place of triphenylphosphine. 1T_13 Yields are comparable to those obtained in the conventional procedure. In another methylene-transfer reaction a polymer-attached dimethyl sulfonium methylide was used to convert benzaldehyde 14 to styrene oxide in 65% yield. A 75?o yield has been reported in the corres- ponding monomeric reaction. 15 Bromination of cumene with cross-linked N-bromopolymaleimide in refluxing carbon tetrachloride was shown by Katchalski and coworkers to give 16 a ,B,B'-tribromocumene (85$). With NBS under the same conditions, only a-bromocumene is produced. However, in polar acetonitrile NBS gives a,3,B'-tribromocumene also. The authors attribute the behaviour of the . .

succinimide residues. polymer to the polar environment provided by the polymer esters of C-labelled Crowley and Rapoport have reported that Dieckman cyclization 1-triethylcarbinylpimelates (2) undergo unidirectional 2-carboalkoxycyclohexanones (3). 1 In contrast to Give the corresponding Other workers have performed, monomeric pimelates close in both directions. polymer esters of acids containing more in pood yields, monoalkylations of 18 ' l9 dialkylation and ester condensation than one ^-hvdropen. The usual ^ from intermodular reactions by-products were not produced, since these result Similarly acylations of polymer which are minimized on the polymeric support. followed by cleavage, esters of acids containing more than one crhydrogen, products.-' give exclusively the ketone monoacylation Harrison and Harrison, by Polymer support of a cyclic ketone enabled 1,10-decanediol, to produce a repeated treatment with trityl chloride and structure. - stable macrocycle and threaded chain polymer- supported asymmetric sugar ^phenylglyoxalic When esterified to a , iodide to atrolactic acid which is acid can be converted with methylmagnesium the analogous monomeric 6<# optically pure, compared to 53^ optical purity in reaction. 21 OCE J + Et 3C" K £ | > .0 OCEt 3 toluene

= (f) polystyrene

Ph2P RhCl(PPh 33/2)

BIBLIOGRAPHY ' (1966); Amer. Chem. Soc. , 88, 5319 R L Letsinger and V. Mahadevan, J. M. Rubinstein and h! Ko*ster, Tetrahedron Lett., 1535 (1972)> Tetrahedron Lett., 288l (1972). A. Patchornik, Lett., P. Utille, and M. Vignon, Tetrahedron 2. G. Excoffier, D. Gagnaire, J. 5065 (1972), and references therein. 1970 j Kraus, J. Amer. Chem. Soc. > 75»7 3. A. Patchornik and M. A. J§, Chem. Soc, 3062 (1971 ) H. Grubbs and L. C. Kroll, J. Amer. 93, k. R. 32oT C. M. Shen, Tetrahedron Lett., U9 f ^ i 5. N. M. Weinshenker and Tetrahedron Lett., 3285 (1972). Angew. Chem. Internat. Ed. Engl. 6. H. Schuttenberg and R. C. Schulz, 10, 856 (1971). ^ V B, 1031 (1967). 7. t7 Tagaki, J. Polym. Sci.^ Part 5, and J. Hetflejs, Tetrahedron Lett., 8. M. Capka, P. Svoboda, M. Cerny, 8 ( Amer. Oil Chem.Soc, kg, 533 (1972). Q. H! S. BrIner and J. C. Bailar, J. D. Bonds, and C H. Brubaker, 10. R. H. Grubbs, C. Gibbons, L. C Kroll, W.

J. Amer. Chem. Soc, 95, 2373 (1973)- \ Lett. 1715 U9fi;.hQ71 F. Vela, Tetrahedron , 11. F. Camps, J. Castells, J. Font, and Commuc, 13^ (1972). 12. S. V. McKinley and J. W. Rakshys, Chem. Chem. Internat. Ed. "Engl. 11, 29» ^972). 13. W. Heitz and R. Michels, Angew. Oda, Kogyo Kagaku Zasshi, 70, 1269 (1967). Ik. S. Tanimoto, J. Horikawa, and R. Amer. Chem. Soc, 84, 37»- ^9b2j. 15. E. J. Corey and M. Chaykovsky, J. E. Katchalski, Tetrahedron Lett., 3629 16. C. Yaroslavsky, A. Patchornik, and

Rapoport, J. Amer. Chem. Soc, 92, «3 U970). IT. J^ Crowley and H. J. Chem., 269, U9fl>. 18. M. A. Kraus and A. Patchornik, Israel 9, Ferrando, and J. Font, Tetrahedron Lett, 19. F. Camps, J. Castells, M. J. Amer. Chem. Soc, 89, 5723 (1967). 20. I. T. Harrison and S. Harrison, J. Lett., 4855 (1972). 21. M. Kawana and S. Emoto, Tetrahedron - 139 - THE CHEMISTRY OF 7-NORBORNENYL ANIONS

Reported By Gary Buske April 26, 1973 Investigations of the interaction between carbon - carbon double bonds and non-conjugated radical or ionic centers has been a matter of considerable interest. An orbital at the 7 position of norbornene is geometrically well suited for such interactions. Indeed, the cation has been shown to bond with the double bond to form a nonclassical carbonium ion 1 (l), and such an interaction has also been proposed for the 7-norbornenyl radical as well.^ Similar interaction in the 7-norbornenyl anion is expected to exert a destabilizing effect due to the Hu.ckel antiaromaticity of the bishomocyclopropenyl anion structure (l) that would be formed. 3 Theoretical calculations have been 4 carried out on these systems. To determine if in fact such a destabilizing interaction exists, proton exchange rates in anti-7-cyano (3) and anti - (h) and syn-7-carbomethoxynor- 5 ' 6 bornenes (5) were studied. A slower exchange rate for 3 relative to 7-cyanonorbornane (2) would be predicted by antiaromatic destabilization of 3> but the inductive effect of the double bond is expected to oppose this effect and tend to increase the exchange rate of 3* Although 3 I s found to exchange lok times faster than 2, U-cyanocyclopentene exchanges ten times faster than cyanocyclopentane. 5 If the latter two compounds are used as a model for inductive effects, a small conjugative destabilization can be deduced.

2 X CN

J5 X = CN Y = H = Jt X C02 CH 3 Y - H X = H Y - CO CH i P 3 9 X = H Y = CI = = 6. X H Y NHNH2 10 X = CI Y = H XX = NHNHP Y = H 8 X = H Y = Br

Alternatively, bond angle strain at the 7 position may also be invoked to explain the data. Since the initial anion formed by proton abstraction from k is in a structurally favorable position to interact with the double bond

while the anion from _5 is not, antiaromatic destabilization of k predicts 4 to exchange slower than Opposite to this expected order, k actually exchanges J».6 1.2^ times faster than _5. It is interesting to note that this is the order expected by considering steric hinderance to the approach of the base. .

- Ik) -

7 Stille and Sannes have investigated the 7-norbornenyl anion more directly by basic oxidative cleavage of syn- (6) and anti- 7-norbornenylhydrazine (7), p— a reaction proposed to proceed through a carbanionic intermediate. 8 BottTsyn- and anti - compounds give approximately 9k to 6 anti to syn deuterium incorporation. These results were explained by the formation of rapidly equilibrating syn- and anti - pyramidal anions followed by deuterium capture. If this is the~case, the product distribution reflects only the relative quench transition state energies and not the anion distribution unless their rate constants of capture are fortuitously equal or much faster than their rates of equilibration. If the quench rates -are equal, then it can be implied that the syn- carbanion ."" is approximately 1.7 Kcal/mol less stable than its anti-epimer Treatment of syn-7-bromonorbornene (8) with magnesium followed by C02 quench forms a 2 to 1 mixture of anti - to syn-7-carboxynorbornene. 9 Rapid epimerization of the presumed alkylmagnesium halide intermediate is questionable in light of recent pmr studies/ and the possibility that product stereochemistry is determined by a radical intermediate during the 11 formation of the organometallic can not be ruled out. Thus interpretation of this data is difficult. Studies with .syn- (9) and anti -7-chlorobenzo- norbornadiene ( 10 ) suggest further that the stereochemistry is determined .prior to or during carbon - magnesium bond formation. 12 Treatment of 10 with magnesium followed by D 2 gives deuterium incorporation only in the anti- position while treatment with sodium naphthalene radical anion In the presence of magnesium chloride, a reaction also postulated to form an 13 alkylmagnesium halide species, yields approximately a 3 to 2 ratio of 1P anti- to syn - deuterium incorporation upon treatment with D 2 0. If the alkyl magnesium halide intermediate undergoes rapid inversion, both methods of preparation should have given identical results. Thus the stereochemistry of 7-benzonorbornadienylmagnesium halide is determined during its formation f and therefore it is not informative about equilibrated anion studies.

BIBLIOGRAPHY 1. (a) P. D. Bartlett, "Nonclassical Ions," W. A. Benjamin, Inc., New York, N. Y., 1965; O) H. Tanida, Accts. Chem. Res., 1, 239 (1968);

( c) S. Winstein, Quart. Rev. (London), 23, lUl 7JL969) 2. (a) J. Warkentin and E. Sanford, J. Amer. Chem. Soc, 90, 1667 (1968); (b) G. A. Russell and G. W. Holland, ibid ., 91, 3968 (T969); (c) S. J. Cristol and A. L. Noreen, ibid ., 91, 3969 (I969)j (d) J. K. Kochi, P. Bakuzis, and P. J. Krusic, ibid ., 95, 15l5 (1973). 3. For a review on cyclopropenyl anions, see: R. Breslow, Angew. Chem. Int'l. Ed. (Engl.). 7, 565 (1968). k. (a) H. 0. Ohorodynk and D. P. Santry, J. Amer. Chem. Soc, 9~L, 1*711 (1969); (b) M. J. S. Dewer and W. W. Schoeller, Tetrahedron, 27, 1*1+01 (1971). 5. R. Breslow, R. Pagni, and W. N. Washburn, Tetrahedron Lett., 5I+7 (1970). 6. D. D. Davis and W. B. Bigelow, ibid ., ll*9 (1973). 7. J. K. Stille and K. N. Sannes, J. Amer. Chem. Soc, 9k, 8U89 (1972). 80 D. J. Cram, "Fundamentals of Carbanion Chemistry," Academic Press, New York, N. Y., 1965 and references cited therein. 9« E. R. Sauers and R. M. Hawthorne, Jr., J. Org. Chem., 29, 1685 (196*0. 10. A. Maercker and R. Geuss, Angew. Chem. Int'l. Ed. (Engl.), 10, 270 (1971) and references cited therein. 11. C. Ruchardt and H. Trautwein, Chem. Ber., 95, 1197 (1962) and references cited therein. . 12. W. T. Ford and G. R. Buske, unpublished results. 13. S. Bank and J. F. Bank, Tetrahedron Lett., 1*533 (1969). lJ+1

SYNTHETIC USES OF DIHYDRO-l,3-OXAZIN

Reported by Ving Lee May 3> 1973 1 The chemistry of dihydro-l^-oxazines has been of' interest in the past two decades, but only recently their unusual synthetic utility as carbonyl precusors has been realized and utilized. The usefulness of the procedure may be illustrated by the two carbon extension of electrophiles 2} 3? 4? 5j6 by the carbanion la to give aldehydes or acids, a scheme which, in effect, provides the aldehyde equivalent of the malonic ester synthesis. The intermediate dihydro-l,3-oxazine _2 may be reduced with sodium borodeuteride to give the tetrahydro-l,3-oxazine _3 which on hydrolysis yields the aldehyde- 1-d. Likewise 2 may be hydrolyzed to give the corresponding acid.

R CHCOOH .,)

l) NaBH 4 or NaBD 4

H 30© R> ^0 ^CH-C

7 The imine bond in 1 ± s inert to additions by organometallics, but conversion 8 ' to the N-methylquaternary salt h makes nucleophilic addition a facile process and thus a ketone synthesis arises.

R' R'MrX H3O © CII-,1 > > T^N R' Vc)=0 N CHpR HpR RCH2 /

CH 3 ] GIL.

Ketones with a- quaternary carbons are synthesized by the stepwise addition of organometallics and alkylating agents to 2,4, 4,6-tetramethyl- 5,6-dihydro-l,3- oxazine ~_1. This scheme, in contrast to above, does not require a quaternarized nitrogen but involves a ketenimine 11 intermediate. The scope of this approach is shown by a complementary aldehyde 12 ' 13 and ketone synthesis which provides for the three carbon extension of nucleophiles. For example, 2- vinyl ic- 4, 4,6-trimethyl- 5,6-dihydro-l,3-oxazines 5? or 6 17 undergo facile 1, 4- additions by organometallics to give aldehydes. Addition of a second equivalent of organometallic results, on hydrolysis, in ketone 8 formation. ,

1) R'MgX. H.oO -> -> TV 2) R"Li -Tv

2 R=CfI; - - 6 R=C6H5

1) NaBH 4 or NaBD 4 -> © 2) 1I 3 T

The utility of the dihydro-oxazJ tie ketone synthesis is well examplified in the synthesis of the natural product, 1- Acetoxy-10-propyltrideca-trans-

x '' 1 5,9-d-iene, propylure, from ^-heptanone in 35 /?' yields. This is in contrast to a 10$ yield by other routes. 16 l) 2-cyclohexenone © Nail -> -tei CIU *N 0CH-

I H3 -0 CH3 RCHp( k

An extension of dihydro-oxazine chemistry17 provides for the synthesis of aldehydes and kel.o- acetic acid esters with the aid of the highly nucleophilic

ketene N,0-acetal 9., obtained from k (R=H) on treatment with sodium hydride.

BIBLIOGRAPHY

1. Z. Eckstein and T. Urban ski, Advances Heterocyclic Chem., 2_, 311 (19^3) • 2. A. I. Meyers, A. Nabeya, H. W. Adickes, and I. R. Politzer, J. Amer. Chem. Soc, 91, 763 (1969). 3- A. I. Meyers, A. Nabeya, 11. W. Adickes, J. M. Fitzpatrick, G. R. Malone,

and I. R. Politzer, ibid . %h , 91, (1969).

k. A. I. Meyers, H. W. Adickes, I. R. Politzer, and W. N. Beverung, ibid . 91, 766 (1969).

. 5- A. I. Meyers, II. W. Adickes, and I. R. Politzer, ibid , 91, 2155 (1969). 6. A. I. Meyers, A. Nabeya, II. W. Adickes, I. R. Politzer, G. R. Malone, A. C. Kovelesky, R. L. Nolen, and R. C. Portnoy, J. Org. Chem., 38, 36 (1973). 7- A. I. Meyers, I. R. Politzer, B. K. Bandlish, and G. R. Malone, J. Amer. Chem. Soc, 91, 5886 (1969). 8. A. I. Meyers and E. M. Smith, J. Amer. Chem. Soc, 92, 108*J- (1970). 9- A. I. Meyers and E. M. Smith, J. Org. Chem., 37, k-2B§ (1972).

10. A. I. Meyers, E. M. Smith, and A. F. Jurjevich, ibid . , 93, 231^ (.1971). 11. A. I. Meyers, and E. M. Smith, Tetrahedron Lett., ^355-3^ (1970). 12. A. I. Meyers and A. C. Kovelesky, Tetrahedron Lett., 1783 (1969). <

13. A. I. Meyers and A. C. Kovelesky, ibid . , '1-809 (1969). 1^. A. I. Meyers and A. C. Kovelesky, J. Amer. Chem. Soc, 91, 5887 (1969). 15. A. I. Meyers and E. W. Collington, Tetrahedron, 27, 5979 (1971). 16. G. Pattenden, J. Chem. Soc. (c), 2385 (1968). 17. A. I. Meyers and N. Nazarenko, J. Org. Chem., 38, 175 (1973)- .

- iU3 -

REACTION INTERMEDIATES IN DECARBONYIATION REACTIONS OE CARBONYL COMPOUNDS BY TRANSITION METAL COMPLEXES

Reported by Sherrill A. Puckett May 7, 1973 INTRODUCTION Conversions of aldehydes to alkanes and alkenes by decarbonylations " caused by group VIII transition metals have been known for seventy years. 1 3 However, only with the recent work of Tsuji and Ohno4 have these reactions become a synthetically useful means of eliminating carbonyl groups as carbon monoxide from various organic molecules. Complexes of rhodium, ruthenium, and, most recently, iridium of the general form MXL3 or MXL2 CO with L=PPhc,, = = AsPh 3 , PEtPho; X CI; M Rh, Ru, Ir, are active decarbonylation reagents. These complexes are low valent, reduced, and coordinatively unsaturated and thus can undergo oxidative addition 5 with aldehydes or acyl halides and can coordinate carbon monoxide strongly to form a new carbonyl complex. Although transition metal complexes have been most useful for decarbonylations of aldehydes and decarbonylation-dehydrohalogenations of acyl halides under mild conditions in homogeneous systems, the general reaction has a broader scope. Complexes of this sort have been used to extract carbon monoxide from acetylenic 6 7 8 Cj ketones , ketenes , acetylated organometallics , carboxylic acids , aroyl 10 11 cyanides , and anhydrides . Alcohols have also been decarbonylated by 12 transition metal complexes by probable first being oxidized to aldehydes . Yields of these reactions vary and the products are mixtures so that they appear at present, to be of little synthetic value in their application. STRUCTURE AND OXIDATIVE ADDITION OF TRANSITION METAL COMPLEXES

Oxidative addition involves conversion of complexes of metals with d8 electronic configuration to dG octahedral complexes by the addition of a covalent molecule. In these reactions the metal complex behaves simultaneously as a Lewis acid and Lewis base and the reaction is generally written as

' 13 LyM + XY -* LyM (x)(y). Criteria for oxidative addition reactions are (l) non-bonding electron density on the metal atom M, (2) two vacant coordination sites on the complex LyM to allow formation of two new bonds to X and Y, and (3) a metal M with its oxidation states which differ by at least two units The reactivity of d° complexes in oxidative addition depends markedly on the nature of the central metal atom and its associated ligands. The tendency of group VIII metals to form oxidized adducts of d6 configuration increases on 14 descending a triad or passing from left to right within group VIII . For, example, d8 square planar iridium complexes are oxidized to d6 octahedral complexes more easily than the corresponding da rhodium complexes are converted to d6 complexes probably because the filled f«orbitals of the iridium complexes do not effectively shield the increased nuclear charge. Ligands which increase electron density at the central metal atom enhance the tendency of the metal to undergo oxidative addition because they make the d8 complexes better Lewis bases. 15

8 6 Conversion of square planar d complexes to octahedral d complexes by oxidative addition is a useful generalized reaction for consideration of the structures of complexes which are active decarbonylation agents. For example, 6, T 8 chlorotris(triphenylphosphine) rhodium (i)i i 'i (1) exists as a crystalline, coordinatively unsaturated, planar, d8 complex. In solutions in organic solvents, 16 however, 1. exists as the formally three coordinate solvated complex 2..

The weakly bound or vacant site of 2_ makes it an exceedingly effective homogeneous decarbonylation reagent because of its ability to coordinate carbon monoxide strongly. Wilkinson and coworkers have isolated the dimeric form 3_ from solutions of 1_ and found that 2_ reacts irreversibly at low temperatures - .1 ih - Xid t0 aff° rd ( ChGme I} 0ther £*? f ^ 5 * compounds analogous T«S n° 6 bSen PrePared,^5, 26 ? - but 1 and 1+ have proven to b2 the most™i'wxdelyfa usedfreagents for stoichiometric and catalytic decarbonylationl Scheme I

Ph 3 P PPh. Ph.,P \ (S) Pb.,P CL PPh, Rh \ solvent 'Rh CI / PPh. CI PPh. y Ph^P CL PPh.,

3

CO

\k Ph., no \y Rh \ /CL PPh.,

k

Under mild homogeneous conditions, 1 • reacts with acvl erou-os ,+m^^ + -.-. to form, at least Initially, acyl complies of type 5a *ich rearrange to give'the^ hexa coordinate complex 5b An isolable complex wUaT-the structure 4 is 4ichloro-(j-phenylpropanoyl)-bis-(triphenylphosphine) rhodium (l.e.~R = CH ra r n Y = CI). Complex 5b undergoes facile molecular rearrangements to give ktrt*^' organxc products (Scheme II). Production of k during the stoichiometric" reaction Scheme II

Ph P PPh. Ph ^P 3 (S) Ph.P Y Ph.,P Y CO \ / \ / pc-y \| Rh Rh \ > Rh C_p -> ^Rh CI PPh. CI PPh. /\VT CI31 PPh.I Cl R PPh.

1 5a 5b A

v Ph 3 P CO

i 200° - CO Rh / \ + RY Cl PPh.,

h and under mild conditions is deactivating. Only at higher temperatures (c.a. 200 does k function ) as a source of a reactive decarbonylation catalyst by loss of carbon monoxide, regenerating the central reactive species 2.*° DECARBOXYLATION OF ACID HALIDES ^Acid halides add oxidatively and stochiometrically under mild homogeneous conditions to transition metal complexes, such as 1, to produce olefins from a process of decarbonylation followed by denydrohalogenat ion. If, for structural reasons, olefins cannot be formed, the products are the corresponding halides. For example, aroyl halides give the corresponding aryl halides in excellent yield. iy,2U,r - 1 )

- 145 -

Transition metal complexes JL_, 2_, U f and other analogous metal complexes are reactive decarbonylation catalysts at elevated temperatures. Early attempts by Tsuji and Ohno^ to substantiate complex h_ as the catalyst has since come under attack from Blum and coworkers. 10 Blum demonstrated in a rather simple way that the carbonyl complex 4 is not always an essential

intermediate in a catalytic decarbonylation with JL_ studying the catalytic

reaction of <*-naphthoyl chloride with JL and with k. If h_ is the essential

reactive species, the reaction should proceed faster with h_ than with 1_. Blum found, however, that Q--naphthoyl chloride is converted into a-chloro- napthalene by k_ in thirty minutes, and by JL in seven minutes. Moreover, during the reaction of k_ with aroyl chlorides, Blum was unable to detect even transient carbonyl bands in the infrared spectrum which would be 2 characteristic of a complex with carbonyl groups (i.e. Cl2 (PPh 3 ) 2Rh(CO)(COAr) which would arise from oxidative addition of an aroyl chloride to 4. This argues strongly for an alternate scheme by which the catalyst, whatever it is, can be regenerated. Blum proposes a pathway (Scheme III) which accounts for his results, and at the same time leaves open the possibility that h_

could be the essential intermediate in other cases. By heating JL with aroyl Scheme III

- + CI Rh(PPh 3 ) 3 Cl Rh(PPh 3 )o PPh. 1 2

PPh ? J>Ph M / II Ar C-Rh-Cl Ph.,P CO Ar —Rh—CI pph ci - Rh Cl PPh 3 Cl TPh. 8 6

V CO Ar PPh.

Rh Cl ci

PPh.

7

and 8_ which were characterized chlorides Blum was able isolate intermediates 6_, J_, by analyses and infrared spectra. Compounds of structure 6_ have only aroyl r metal carbonyl absorptions in the range I67O-I715 cm ; compounds J_ show only 1 peaks in the range 2055-2080 cm" ; compounds 8_ did not show any carbonyl absorption in the infrared. In one case Blum followed the transition from 6 -• 7 (e.g. Ar = B-naphthyl) by observing the characteristic carbonyl absorptions ~* in the infrared spectrum. Likewise, the transition X it was observed for 19 several compounds. About this same time Tsuji and Ohno isolated several = = Cl) ocyl complexes of type 5_ (e.g. R CH 3 (CH2 ) 4 , CH 3 (CH2 ) 5 , CH 3 (CH2 ) 14 j y and showed that these complexes were, in fact, intermediates by heating the - 1U6 - solid complexes of type 5, distilling the olefin formed from the reaction. The scope of the catalytic decarbonylation using catalysts JL and k_ is shown in Table 1. Simple aroyl chlorides, bromides, or iodides give the corresponding halides and aroyl cyanides are decarbonylated to aryl cyanides in good yields. The effect of aromatic substitution has only been qualitatively studied, demonstrating that electron- donating groups (i.e. ^-methoxy, ^-methyl) tend to decrease the rate of decarbonylation and electron-withdrawing groups (i.e. 3,4-dichloro, 3-chlorocarbonyl) enhance the rates and yields. ~ ( Aliphatic acid chlorides are decarbonylated and dehydrohalogenated by 1 and h_ to give mainly mixtures of olefins. Distillation of the olefins as 00 soon as they are formed lowers the yield of the non-terminal olefins. ' 22

Certain additives, such as triphenylphosphine and iodine , also inhibit double bond migrations, however, these tend to be specific for certain complexes. The role of these additives is not clearly understood, but it could be that the additives compete with the metal hydride complex formed by the reaction of HC1 with h_ for the olefin produced by decarbonylation-dehydrohalogenation.

Table 1 Scope of Decarbonylation Reactions of Acid Halides

" Acid Halide Complex Temp. , , Product composition) hnr min {% ^ Benzoyliodide 1 175-250 ^ ' Iodo benzene 62 Benzoylchloride k 200 k Chlorobenzene 85 20 Benzoyl bromide k 220 1.5 Bromobenzene 87 20 ^-Chlorobenzoyl I - 3-5 ^-Cyanochlorobenzene 95 10 cyanide

h-Methoxybenzoyl it 2^0-250 12 4-Chloroanisole k6 20 chloride U-Methylbenzoyl k 250 11 U-Bromotoluene k~$ 20 bromide

3 ,^4-Dichlorobenzoyl It 250 1.5 1>3>^* Trichlorobenzene 96 20 chloride Isophthaloyl chloride h 2^0-250 k 1,3 Dichlorobenzene (75) 92 20 3-Chlorobenzoyl Chloride (25) Octanoyl chloride h 200 1 1-Heptene (7l)? trans-2- heptene(2^) cis -2- Heptene (5) 90 20

Unstable n-olefin complexes (e.g. (RCH=CH2 )lrHClo(C0)(PPho) ) have been suggested to be involved in analogous double bond migrations (Scheme IV).

Scheme TV PPh.

PPh 3 CI CI CI. ,C1

^r PhsP >H mj ^\ Jj H C 2 CH< R s

- lk-7

Attempts to isolate intermediates, and so describe the catalyst, have recently been aided by the study of ligands other than triphenylphosphine, 23 ' 30 30 as -well as other metal complexes. Kubota has found that benzoyl chloride, which does not form a stable complex with h, adds oxidatively to

the iridium complexes 9. a-c to give the stable hexacoordinate complexes 10 a-c. About this same time Blum reported the observation of lOd but he was L

L CO CI.. ,co Ir + PhCOCl > lv rS \J>ci^'CPh L

9 10

a, L=AsPh 3 ; b, L=P(p_-tolyl) 3 ; c, L=PMePh2 and d, L=PPh 3

unable to isolate the complex. 23 A reexamination of the rhodium catalyst h with more basic ligands would clarify the role of ligand basicity in the stability of decarbonylation intermediates by perhaps making it possible to isolate rhodium intermediates. Care must be taken in interpreting Kubota' data, for two variables, metal and ligand, have been changed. Previously h 2C was thought to lose carbon monoxide to produce the reactive species 2_.

Kubota' s data seems to indicate that the rhodium catalyst j+ may not necessarily lose carbon monoxide first, but it may form transient intermediates similar to 10 which have gone undetected. Another interesting complex studied by Kubota is the nitrogen complex 3° 11 . Transient acylated intermediates of type 12 have been observed for

R-C2H5-, CH 3 (CH2 ) 10- » cyclopropyl, cyclobutyl, vinyl, and benzyl. The phenylacetyl compound 12 (i.e. R=benzyl) has been isolated and characterized and all of the complexes 13 studied.

PPh 3 ]Dph3 PhsP jj 0] CO ,1-R s / N2 . I y + RCC1 -^> > Ir. CI- Ir' -rls \PPh- cr , Vi y ^R PPh 3 PPh 3 11 12 13

Blum in his latest study, has been able to better describe the mechanism of decarbonylation-dehydrohalogenation using chorocarbonylbis(triphenylphosphine) 23 iridium, 9d. It had been suggested for some time that the P hydrogen atoms were transfered to the metal to produce olefins from acid halides (Scheme V). Hydride complexes like JL5 have been known for some time from the reaction of 1S Qd with HC1. Blum and coworkers have provided convincing evidence of f3 ? Scheme V HC-. CH 2 H Ph CO PhsP. CO -> .Jtt + Olefin N PPh. cry 'PPh. Cl Cl Ik 15 - 148 - hydrogen involvment in the olefin forming step of catalytic decarbonylation of acid halides by an isotope labelling experiment. Distillation of deuterated and non- deuterated 3-phenylpropanoyl chlorides (i.e. l6 a-d) over 9d which under these conditions does not cause significant hydrogen- deuterium exchange, showed that l6a and l6b evolved deuterium- free HC1 while l6c and l6d gave only DC1. Olefins obtained from l6 a-c were isotopically pure, but olefin 17d had partial terminal Hi exchange with benzylic deuterium atoms. Obviously this benzylic exchange could not be detected with 17 a-c and may, in fact, have been occur ing in 17 a-c. Because alkyl halides are not dehydrohalogenated

C6 H 5 CH2 CH2 C0C1 -> CSH5 CH=CH2 + HC1 l6a 17a

C6D 5 CH2 CH2 C0C1 -=> C6D 5 CH=CH2 + HC1 l6b 17b

C6H 5 CD2 CD2 C0C1 -> CGH 5 CD-CD2 + DC1 l6c 17c

- C6H 5 CD2 CH2 C0C1 ^ C6 HrjCD=CH£ DC1 l6d 17d

when refluxed several hours with 9.d, Blum concluded that hydrogen halide is extracted while the organic moiety is still bound to the iridium as a ligand, and proposed two related mechanisms to account for the mild catalytic decarbonylation of acid halides with the iridium complex 9d (Scheme VI). The Scheme VI

Cl(C0)lr(PPh 3 ) 2 9d -HX Irch2 ch2 co X CI X Ph oP PPh- PhaP CI PhaP^ CI jf ; CO olefin Ir -> .Ir > OC^I >PPh. 0C 1 \H OC ^CHpCHpR CIIJ-CH-R H 18 Ik 15

PPh.

CO CO x CO ^ PPh. ^Ph; H PPh- '/ RCH CfI v -_C_0 I^H -olefin 2 2 Ir S Ir' Ir 1 \ X CI X CI

2£ 20 21

-HX

RCH2 CH2 COX - lk-9 - pathway l8 — lk -*TL5 is proposed to be the major reaction pathway because Blum was able to isolate olefins and complex 15, in a number of cases (e.g. 9d R=CH 3 (CH2 ) 9- ; X=C1). Heating 15 causes elimination of HC1 to give which can react with another molecule of the acyl chloride. The pathway 19 ~* 20 — 21 has been proposed for the reaction of bulky acyl bromides (e.g. R=CH 3 (Ch17J"9- ; X=Br). Blum suggests steric crowding to be the major factor in determining

""20 -* - -»15 ^- s i. whether pathway 12 21 or l8 1J+ use( Aroyl chlorides and bromides react with 9d to give hexa and pent a coordinate aroyl complexes which rearrange with loss of carbon monoxide to the corresponding hexa and penta coordinate aryl- iridium complexes. These aryl- iridium complexes are extremely stable and do not eliminate aryl chloride even when heated above their melting points. 23 This feature of 9d has added some measure of selectivity to the decarbonylation catalyst 9d , making it possible to selectively eliminate one carbonyl function from an acyl- aroyl dichloride. An example is the dichloride of k- (p_-carboxyphenyl) butyric acid which gives exactly 0.5 moles of each HC1 and carbon monoxide. The unstable p_ allylbenzoyl chloride was isolated as the methyl ester (Scheme VII).

Scheme VII CH2 CH2CH2 C0C1 21 ci lOl -> ci + £ CO + -c £ HC1

CH 3

DECARBONYLATION OF ALDEHYDES Whereas many reaction intermediates from acid halides and transition metal complexes have been isolated, the same is not true for the reactions of aldehydes. As a consequence, many of the mechanisms proposed for aldehyde decarbonylations draw on analogy of the acid halides. Alphalic and aromatic aldehydes react with

complexes like _1 stoichiometrically or with complexes like 1 or k catalytically to produce alkanes 20> 31 Primary aldehydes react smoothly in benzene or dichoromethane while more sterically hindered aldehydes react with 1 at higher temperatures in solvents such as toluene, xylene, or benzonitrile. Recently Walborsky and Allen 32 demonstrated the intramolecularity of the decarbonylation reaction by treating 1- methyl- 2,2- diphenylcyclopropane- carboxaldehyde-_d, which contained 96 + 3% deuterium at the aldehyde carbon, with 1. The product hydrocarbon, obtained in 70% yield, contained 93 +_ yjo deuterium in the 1-position, demonstrating that this reaction may be synthetically valuable for the production of deuterated hydrocarbons.

C-D > ;

- 150 - The stereochemistry for stoichiometric, as well as catalytic, decarbonylation of aldehydes with transition metal catalysts is retention (Table 2). The mechanism is presumed to be similar to that of acid halides (Scheme II) and to involve intermediates similar to j3a and j?b which disproportionate into either the carbanions or the caged radical pairs both which have shown a high degree of retention of configuration in solvents of low dielectric constant such as those used in decarbonylation reactions. 33 ' 34> 3S Aldehydes (E)-2_5, (-)-(R)-2_3 and 2J3a-d have demonstrated retention of configuration but Walborsky was unable to distinguish if carbanions or radicals were responsible for the retention.

CH- Ph ph CH- x ,H2 CH 3 .CH2 CII 3 x ,// >/\„/H CHsCHg^C -^ C ^ i'[ vT CH3CH2 ^ Ph Ph

(E)-21 (Z)-22 C-)-(R)-23 (+)-(S)-24

Ph ji

-> Ph v X

25a, X-CH3, (+)-(R) 26a, X=CH 3 , (+)-(S)

b, X=d, ( + )-(s) b, X=C1, ( + )-(s) c, X=F, (-)-(S) c, X=F, (-)-(s) + d, X=0CH 3 , (-)-(s) d, X=0CH 3 , ( )-(S)

Table 2

St ereospecific Decarbonylation of Aldehydes with 1

'

Optical Yie Aldehyde Solvent Temp. Time Product Purity i 1o (E)-21 benzene reflux 1.5 hr (Z)-22 100 —

(-)-(R)-23 benzonitrile 160 1.5 hr ( + )-(S)-2i£ 81 51

(+)-(R)-25a xylene reflux l6 hr ( + )-(s)-26a 9k TO + — ( )-(s)-25b benzonitrile 160 1.5 hr (+)-(S)-26b 29

(-)-(s)-25c benzonitrile 160 1.5 hr (+)-(s)-26c 73 62

(_).( S )-25d benzonitrile 160 1.5 hr ( + )-(S)-26d 6 - 151 - BIBLIOGRAPHY

1. P. Sabatier and J. B. Senderens, C. R. Acad. Sci. Paris, 136 , 738, 921 (1903).

2. Y. Suen and S. Fan, J. Amer. Chem. Soc. , 65, 12^3 (19^3). 3. H. W. Eschinazl and H. Pines, J. Org. Chem., 2b, 1369 (1959). b. J. Tsuji and K. Ohno, Tetrahedron Lett., 3962T1965). 5. J. P. Collman and W. R. Roper, Advan. Organometal. Chem., _7, 53 (1968). 6. E. Muller, A. Segnitz, and E. Langer, Tetrahedron Lett., 1129 (1969). 7. P. Hong, K. Sonogashira, and N. Hagihara, J. Chem. Soc, Japan, 89, 7^- (1968). 8. J. J. Alexander and A. Wojcicki, Inorg. Chem., 12, 7^ (1973).

9. 0. Prince and K. Raspin, Chem. Commun. , 156 (19^). 10. J. Blum, E. Openheimer, and E. Bergman, J. Amer. Chem. Soc, ~"89, 2338 (1967). 11. J. Blum and Z. Lipshes, Progress in Coordination Chemistry, Elsevier Publishing Co., Amsterdam, 1968, p. bQ. 12. A. Emery, A. Oehlschlager, and A. Unrau, Tetrahedron Lett., M+01 (1970). 13. F. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry" Interscience

Publishers, New York, N. Y. , 1973, p. 772. lb. R. Nyholm and K. Vrieze, J. Chem. Soc, 5337 (1965). 15. J. Chatt and S. Butter, Chem. Commun., 501 (1967). 16. J. Osborn, F. Jardine, and G. Wilkinson, J. Chem. Soc, [A], 1711 (1966) 17- M. Bennett and P. Longstaff, Chem. Ind. (London), 81+6 (1965). 18. M. Baird, J. Mague, J. OSborn, and G. Wilkinson, J. Chem. Soc. [A], 13 Vf (1967). 19. J. Tsuji and K. Ohno, J. Amer. Chem Soc, 88, 3I+52 (1966).

20. K. Ohno and J. Tsuji, ibid . , 90, 99, (196877 21. J. Blum, Tetrahedron Lett., 1605 (1966).

22. J. Tsuji and K. Ohno, ibid . , Vfl3 (1966). 23. J. Blum and S. Kraus, J. Organometal. Chem., 33, 227 (1971). 2b. J. Mague and G. Wilkinson, J. Chem. Soc. [A], 1736 (1966). 25. J. Blum, H. Roseman, and E. Bergman, Tetrahedron Lett., 3665 (1967). 26. D. Clement and J. Nixon, J. Chem. Soc. [Dalton], 195 (1973). 27- J. Blum and Y. Pickholtz, Isr. J. Chem., 7, 723 (1969). 28. R. Coffey, Tetrahedron Lett., 3809 (1965). 29. R. Augustine and J. VanPeppen, Chem. Commun., 571 (1970). 30. M. Kubota and D. Blake, J. Amer. Chem. Soc, 93, 1368 (1971). 31. J. Tsuji and K. Ohno, Synthesis, 157 (1969). 32. H. Walborsky and L. Allen, J. Amer. Chem. Soc, 93, 5^66 (1971)- 33- D. J. Cram, "Fundamentals of Carbanion Chemistry," Academic Press,

New York, N. Y. , 1965, p. 153. 3b. P. Landsburg, et. al., J. Amer. Chem. Soc, 88, 78 (1966). 35- U. Schollkopf, Angew. Chem., Int. Ed. Engl., 9, 763 (1970).

152 -

NEIGHBORING GROUP PARTICIPATION IN PEROXYESTER DECOMPOSITION : A CASE OF SIMULTANEOUS NEIGHBORING GROUP PARTICIPATION

Reported by Michael M. Chan May IT, 1973 Neighboring group participation at incipient carbonium ions has been known for some time 1 and analogous stabilization of free radical centers has been 2 widely sought. The first examples of anchimeric acceleration in radical reactions, reported by Bentrude and Martin, 3 involved the participation of neighboring substituents (Entries 1, 2, and 6, Table I), in the large rate enhancements observed in the decomposition of ortho substituted _1 relative to the unsubstituted peroxybenzoate 1 (X~H). These results together with *

•Products

L 2a OY

1 Y=OC(CH 3 ) 3 , 2 CAr

X-SAr, SCH 3 , I, CH-CAr2 , C-CAr

Table I. DECOMPOSITION OF o-X- SUBSTITUTED t- BUTYL PEROXYBENZOATE IN CHLOROBENZENE

: Entry X ReL rate at 6o° AfT^-kcal/mole AS*' e.u. Ref . 4 1 -SCaH5 2. +5 x 10 23.0 + 0.03 -3.4 + 0.2 3b 4 2 -SCH 3 l.+l x 10 22.6 + 0.2 -5-5 + 1-3 3b 4 3 -2,6-(SCbH5 ) 2 k.3 x 10 ^ 7 k -SOCeHrs 72.7 29.^3 + 0.14 k. 5+0.38 6a + 5 -CH-C(C6 H 5 ) 2 67.0 26.3 + 0.7 -5.0 1.6 6b 6 -1 9 28.0 + 0.3 -0.8 + 1.7 3b - CH SCoH> 7.2 7 2 j ^ 32.3 7 8 -c(cn 3 ) 3 3.8 3^.2 + 0.5 12.5 + 3-5 3b 9 -C^CC6H5 1.8 37.37 + 1.80 1^.99 + 1.10 5b 10 -H 1.0 3^.10 10.0 ka 11 - S02 CH 3 38.0 19.5 ka

a. Calculated from the absolute rate equation of Eyring using the least square

method, b at k-0°. jc at 120°.

those from subsequent studies are summarized in Table 1. Studies of solvent polarity, substituents, and salt effects favor a dipolar transition state 2b, however, product and scavenging studies indicate that the 4 decompositions produce free radicals. The effects of leaving groups, Y, on the rate of decomposition correlate semi quantitatively with the K^ value of the

4a ' 5 60 corresponding acid HY. Labelling experiments show that JL (X=CH:CPh2 , 18 Y-0C(CH 3 ) 3 ) with labeled at the carbonyl position decomposes to give 3-benzhydrylphthalide with 88% of the label retained in the carbonyl position suggesting that the neighboring group participates at the peroxy oxygen. The fact that 3a shows little acceleration (Table I, entry 7) as compared to ^j-a 3 (K^a/K^ -T.02 x 10 at ^0° in chlorobenzene ) implies that the number of rotational degrees of freedom lost in the transition state has a large effect 8 on the rate of decomposition 7 as was predicted by Bartlett and Hiatt. If the neighboring sulfur of 4a participates at the 1>butoxy oxygen, a seven membered ring transition state with four bonds frozen would be formed which is not as likely as participation at the carboxy peroxy oxygen in a six membered ring transition state with only three bonds frozen, although further work is needed 4a to establish this firmly. Steric studies with 3b and d which show no - 153

X=CH2SC6H5 , Y=H

b X=S02CH 3 , Y=H a X=SC6 Hn " OQ X, Y=SC6H4 b X=H X, Y=SC6 H 4C=0 X '0-0 X

anchimeric acceleration suggest that the neighboring sulfur participates only if the groups attached to sulfur can attain a pyramidal configuration. Evidence for simultaneous neighboring group participation at a radical 7 center was first sought in the decomposition of 5. If both sulfurs were to

participate simultaneously, the transition state _6 would be expected.

Experimentally it was shown that _5 decomposed only 0.66 times as fast as 1 (X-SPh, Y~0C(CH3)3) giving no evidence for simultaneous neighboring group 9 participation. Recently we have shown that diperoxyester J_ undergoes thermolysis to give sufurane 8., t- butyl alcohol, and acetone as the major 4 x products with first order kinetics (k=1.57 x 10 sec ) a rate enhancement 2 factor of ca 10 relative to 1 (X=SPh, Y=0C(CH 3 ) 3 ) at 9 in chlorobenzene. In addition, the decoloration of radical scavenger galvinoxyl by a process which is zero order in galvinoxyl and the observation of a CIDNP emission Ph

2 PhS SPh

signal 67.8 during the thermolysis of give at ppm J_ further evidence for the radical nature of the reaction. This evidence is consistent with simultaneous neighboring group participation at both peroxyester centers in the decomposition

of J_. Before this can be accepted, however, an alternative explanation that the steric nonbonded interaction between the second peroxyester group with the phenylthio group hinders the free rotation of the latter, thus lowering the entropy of the starting diperoxyester and increasing the rate of decomposition, has to be evaluated by a model study.

BIBLIOGRAPHY

1. a) G. Baddely and G. M. Bennett, J. Chem. Soc, 26l (1933) ; b) H. H. Jaffe, Chem. Rev., _53, 191 (1953). 2. a) N. Lowry, M. I. T. Organic Seminar Abstract, p. 210, 1962 and reference therein; b) J. W. Wilt in "Free Radicals" Vol. I, J. K. Kochi, Ed., John

Wiley and Sons, Inc., 1973 > p-333'

3- a) J. C. Martin and W. G. Bentrude, Chem. and Ind. , 192 (1959); b) J- C. Martin 'and W. G. Bentrude, J. Amer. Chem. Soc, 8k, 1561 (1962).

k. a) D. L. Tuleen, W. G. Bentrude, and J. C. Martin, ibid . , 85, 1938 (1965); b) J. C. Martin, D. L. Tuleen, and W. G. Bentrude, Tetrahedron Lett., 1963, 229. 5. a) W. Honsberg, J. E. Leffler, J. Org. Chem., 26, 733 (I96l); b) Der- shing Huan£

Ph.D. Thesis, Univ. of Illinois, 1973- I 6. a) R. J. Arhart, Ph.D. Thesis, Univ. of Illinois, 1971; b) T. W. Koening and J. C. Martin, J. Org. Chem., 29, 1520 (1964); c) J. C. Martin and T. W. Koenig, J. Amer. Chem. Soc, ^ 1771 (1964). 7. T. H. Fisher and J. C. Martin, ibid . , 88, 3382 (1966).

8. P. D. Bartlett and R. R. Hiatt, ibid . /Ho, 1398 (1966). 9. J. C. Martin and M. M. Chau, unpublished results. . .

- l^k - RECENT STUDIES OF HYPERVALENT ORGANIC IODINE COMPOUNDS

Reported by Gary Astrologes May 21, 1973 INTRODUCTION trivalent organoiodine compounds (l) have been known as Tricoordinate, T isolable species for sometime. Phenyliodine diacetate 2 and o-Iodosobenzoic acid2 3 were first reported over 80 years ago although the structure of 3 3 has only recently been confirmed by x-ray crystal studies. This structure for 3 had been suggested initially by early workers because the compound differed from normal iodoso compounds in its smell land relative stability 5 2 ' 4 to oxidation. A similar inorganic compound IC1 3 was first made in l8lU. 61 7 The recent development of hypervalent bonding theory and its successful application to rare gas halides^has led to a renewed interest in this class of compounds

R o-ccn

R Ph-

B ? 0-CCH 3

BONDING THEORY

Early attempts to understand the bonding in 1 used sp'^d hybridization of the iodine atom to explain the ability of iodine to form stable compounds having three bonds and two electron pairs. According to the theory equatorial bonds use sp hybridized iodine orbitals while axial bonds use p d 2 orbitals Pimental6 proposed another bonding scheme, which has been called hypervalent bonding theory more recently by Musher. 7 In this theory a three-center four-electron bond is formed using the p orbitals of the iodine atom and two of its ligands while the third ligand forms a normal covalent bond to another iodine p orbital (Figure i). The hypervalent bonding theory predicts that the 10 bond should be linear, and that the two ligands should have appreciable electron

Figure I

Molecular Orbital Diagram of Hypervalent Bonding electron densities Antibonding Orbital OO Ci s Nonbonding Orbital

Bonding Orbital densities on them due to the electrons in the hybrid occupied nonbonding orbital. Therefore, this type of bonding would be stabilized by two electronegative ligands on the iodine in agreement with the existence of 211 as a more thermally stable compound than triphenyliodine. 12 However, thermal stability is difficult to treat quantitatively since differences in decomposition pathways or in crystal lattice energies can have large effects on decomposition temperatures. - 155 - The theories using d orbitals or hypervalent bonding both predict a trigonal bipyramidal structure of the three bonds and two electron pairs around the iodine atom. X-ray crystal studies of 13 14 15 PhICl2 , Fh2 ICl, p-ClCel^ICl^,, 3 and 3 agree closely with this expectation with the more electronegative atoms 16 in the apical positions. McCu Hough noted that the I-Cl bond distances in FhICl2 (h) are 0»13A greater than would be predicted from single bond covalent radii. This is consistent with the hypervalent theory which predicts a bond order of 0.5 for these bonds. Attempts have been made to explain this data 3 with sp d theory. The compact, electronegative s orbital of iodine used in its equatorial bond would be expected to form shorter bonds and prefer more 2 17~ lD electropositive ligands than the p, d hybridized apical bonds. However, 3b r a Cl nuclear quadrupole resonance study of the IC1 ~ ion is not compatible 3 2 with sp d bonding. Cornwall and Yamasaki20 concluded from their data that d orbitals were not very important in the bonding and that most of the negative charge was on the chlorine atoms also as hypervalent bonding theory would predict. A low carbonyl frequency of 1660 cm 1 observed by Bell and Morgan21 in the infrared spectrum of 2 is also consistent with the presence of more negative charge on the oxygen atoms than for most esters. 22 This theory led 7 Musher to suggest the synthesis of two new hypervalent iodine compounds which were later made by Agosta. 23 TRIARYLIODINES On the basis of the hypervalent bonding theory triaryliodines would be expected to be less stable than their acetate analogs due to the lower electronegativity of the apical phenyl groups. Triphenyliodine was first 12 prepared by Wittig from the reaction of PhICl2 or Ph2 I I~ with phenyl lithium

at -80° . This compound is unstable, decomposing sometimes explosively to biphenyl and iodobenzene at -10°. Using a similar procedure Clauss 24 made

a cyclic analog, 5-phenyl-5H-dibenziodole _5 and found it to be much more thermally stable than triphenyliodine decomposing only slowly at room

-^ CI ( 188P ;'o * ®

86c ci«

temperature but explosively at 105°. This enhanced stability of cyclic hypervalent compounds has been noted by other workers using sulfur25 or 26 phosphorous as the central atom. The thermal decomposition of _5 has been studied by Beringer to give the products shown in Table 1 in the yields indicated. 27 Table 1. Products from the Thermal Decomposition of 5 o Products I I

% yield at room 10-15 10-15 10-15 10-15 10-15 10 temperature °/o yield in 2,2 3.9 5-7 TT "Bo" refluxing hexane - 156 -

The products suggest a radical mechanism of decomposition as shown in Scheme I. At high temperatures the reaction statistically should go 2/5 through Scheme I

7

path b and l/$ through path a. From the instability of acyclic triaryliodines, path a might be predicted to be lower in activation energy, thus being favored at low temperatures. Beringer suggested that a third pathway, direct migration of the phenyl group, might be necessary to explain the high yields of 28 2-iodo-o-terphenyl 6. Recently Trost has suggested a similar fragmentation of 8, the hypervalent sulfur analog of j>. Scheme II shows a possible concerted 26 * 29 mechanism consistent with orbital symmetry arguments , which predict that Scheme II

->

8 6

bipyramidal structure a concerted axial-equatorial fragmentation from a trigonal is unlikely due to is forbidden. Since concerted axial-axial fragmentation fragmentations are the large distance between ligands, only diequatorial possible effects of likely to occur by a symmetry-allowed process. Although yet, this theory can still it orbital participation have not been analyzed because has to pass explain differences between the reactions of _5 and 8 5 geometry for through high energy intermediate 9 to achieve the diequatorial already coupling thereby lowering its reactivity. In contrast, sulfurane 8 to give has an equatorial phenyl group and decomposes below room temperature 28 mechanism the analogous terphenyl as the only product. Instead of a - 157 - involving direct migration of the phenyl group, the higher than statistical yield of 6 could be the result of radical 7 attacking the benzene formed. 27 to give a near Acid~catalyzed cleavage of _5 was found by Beringer statistical reaction (only 1/6 cyclic iodonium ion 10 ) with strong acids like (Scheme III). e HN03 in THF or A1C13 in benzene

Scheme III

10 HA HA ^>

Bulky or weak acids like triphenyl boron or acetic acid gave much higher yields phenyl ring of also of 10. Para or meta substitution of chlorine on the _5 increased the yield of 10 suggesting partial negative charge on the benzene ring in the transition state. Triaryliodines have also been proposed as intermediates in several reactions. Beringer M° suggested the presence of 11 in the electr ore duct ion of dibenziodolium I

11 G

ions to explain two of the products. He also postulated that compound 12 was formed in the reaction of trans-chlorovinyliodoso dichloride with dilithium compound and that it subsequently lost acetylene to give the observed 1-13 product/' Recently32 triaryliodines have been reported to react with aryl lithium

compounds to give iodonium ions, presumably via tetraryliodate ions (IAr 4 ) ~ analogous to the well-known IC14 ion. In one case (Scheme IV ) large concentrations of ion 1^- were shown to be present. This conclusion was reached since reaction with carbon dioxide to give a 2% yield of benzoic acid showed that little phenyl salts lithium remained. On the other hand a good yield ( Jk%>) of dibenziodolium 15 could be obtained upon acidication. The formation of this anion would require

thBu IClc I Ph 2n-BuLi CI' PhC^CPh -> 78° *Ol,LLi J

13 .

- Scheme IV - 158 - Li

_i> v-

+ PhLi

NH4CI C0 ;

H2 II V PhCOOH 2$

2 PhH +

Ihfo 15

either participation of unfilled d orbitals of iodine or the formation of a second three-center four-electron bond. The latter is consistent with nuclear » 20 quadrupolfi resonance data for the IC14 ion by Cornwall and Yamasaki who concluded that iodine 5d orbitals could not be contributing much to the bonding in that ion. ARYL IODINE DIESTERS The more stable hypervalent iodine compounds have halogen or oxygen atoms as two groups on iodine, and compounds of the form ArIX2 (X = CI, OBz, OAc) have been studied for many years. The usual methods of synthesis and the reactions of these compounds are summarized in Scheme V. 33 37 Phenyl iodine diesters can be prepared by oxidation of iodobenzene with peracids, 37 esterification 34 of iodosobenzene, 35>36 or by reaction of an acid with another diester.

Scheme V

-> Clc NaOH PhC02H <- Fhl(0 CFh) Phr" Phi -^> PhICl 5 PhIO > 2 2

OC(CH3 )3 (CH3 ) 3C0C1 A; A NaOHi CH 3 C02H 16 CH3CO3H •V PhC02H

> Phl(02CCH 3 ) 2

They are easily hydrolyzed by aqueous base to iodosobenzene. 32 Iodobenzene can be oxidized by chlorine or t -butyl hypochlorite to give phenyliodine 33 dichloride or t-butoxy chloride 16. The structure of 16 was assigned on the basis of its nmr, elemental analysis, and its quantitative reaction with iodide ion to give iodobenzene, t-butanol, and iodine which can be titrated with thiosulfate to give a calculated molecular weight within 2% of the theoretical value Cyclic analogs of 2 can be made by similar methods sucn as acetyiation of o-iodosobenzoic acid to give VJ4 or peracetic acid oxidation of 18 to give 1923 (Scheme Vl). The presence of an AB 2 septet in the nmr of 19 and a single carbonyl band at 1686 cm" 1 in the infrared spectrum was used to show that 19 is a symmetrical compound. Cyclic ester 21 was formed by the thermal )

- 159 -

q Scheme VI „ COpH \V X^ (CH CO) / 3 2 CH3CO3H -? >(oV t :o Nr=o r

17 18 19

21

38 decomposition of diaroyl peroxide _20. The structure of 21 was confirmed by- its synthesis from o-iodobenzoic acid dichloride and silver~p-nitrobenzoate. Both methods gave a product with an identical infrared spectrum similar to that of 17. Another analog of 2, polystyrene iodine diacetate, was made by Yamada and Okawara to prepare other polymers. 3 " (See Scheme VII

Scheme VII -(CH2CH)— l2 5 /l2 -(CH2 Cil) - Ac^O/koOo — (CHoCH) n > ' n I H S0 6 2 4 D\

:(oac) ;

Considerable work has been done on the use of Phl(OAc) 2 2 as an oxidizing agent. It has been shown to cleave mandelic acid40 or 1,2 glycols in a manner " similar to lead tetra-acetate, but 200 to ^-0,000 times slower. 41 43 Anilines 44,4!5 are usually oxidized to azo compounds by Phl(0Ac) 2 , acetanilides give 46 complex mixtures of products , and phenol gives benzoquinone or ^-substituted ' phenols depending on conditions.'1 7 Recently Balaban found that substituted pyrocatechols upon treatment with 2 give o-quinones in good yields (ca. 90%) in what should be a synthetically useful reaction. 11 The possible presence of radical intermediates in these oxidations of phenols is suggested by the work of Kresta and Livingston. 48 They found that the oxidation of 2,4,6-tri-tert-

butylphenol gave a stable free radical J22, detectable by esr while 2,6 dimethylphenolate anion gave a dimer 23 and a polymeric radical 2k presumably from an intermediate similar to 22. On the other hand cyclic

22 0— -f#-°' J 2k

intermediates have been suggested in the oxidation of 1,2 glycols 4" 1 or pyrocatechols. 11 Phenyl iodine dibenzoate, which might be expected to react similarly to the diacetate has not been studied as extensively although it does cleave benzpinacol. 35 - i6o -

The reaction of iodosobenzene with perbenzoic acid produces a very reactive 49 oxidizing agent, (dibenzoyldioxyiodo) benzene 25.

S Thy* ^ + + + ArlO Ar'COoH Arl(OOCAr')p _ > ArI02 Ar'C02H (Ar'C02 ) 2 28L 25

pyridine This compound is capable of oxidizing sulfoxides or sulfides to sulfones, its to pyridine N-oxide, iodobenzene to iodoxybenzene, triphenyl phosphine to of these oxide, and styrene to an epoxide. Evidence concerning the mechanisms oxidations was not presented, but the thermal decomposition of _25 gives products consistent with a homolysis of the oxygen-oxygen bonds. Iodoxybenzene, carbon dioxide or a carboxylic acid, and an aroyl peroxide are found but not may be biphenyl. A homolytic mechanism is not necessarily ruled out since _25 49 has been a good trapping agent for aryloxy radicals. A similar compound 26 suggested as an intermediate in the reaction of 2 with t-butyl hydroperoxide at -80°. 50 Phi (00 t-Bu) 2 Phi Phl(0Ac) + 2 t-BuOOH - 26 2 hrs + ; > ^o 2 + 2 HQA.C -80° 2 -00 t-Bu" :t-Bu0000 t-Bu I* decomposition products

The decomposition of phenyl iodine diacetate 2 in chlorobenzene has been 51 ' 52 3 shown extensively studied by Leffler and Merkushev.^ Compound 2 has been and to be covalent in acetic acid solution by both freezing point depression 54 22 these conductivity studies, but more recent conductivity work suggests that 5 warmed slightly above solutions contain very small (1G~ ) amounts of ions when good room temperature. The decomposition of 2 in chlorobenzene at 126.8 shows and first order kinetics giving products consistent with competing homolytic ionic dissociations leading to two pathways with the major products, methyl from iodide, phenyl acetate and carbon dioxide (Table 2, first column) coming

Table 2. Products of the Decompos it ion of 2. at 126.8° in chlorobenzene Product % yield % when \S) added CH3I ToJT" 1 Iodobenzene 26.0 81.2 Phenyl acetate 7^.6 20 Acetic acid 22.7 121.4 Chlorotoluene 10.8 2 co 72.5 13.4 2 2? 7.7 2-Iodocyclohexyl acetate I >

the ionic path (Scheme VIII, R = CH 3 ). Added cyclohexene, as shown in the second column in Table 2, intercepts the postulated iodine acetate intermediate giving 2-iodocyclohexyl acetate 27 and other products at the expense of the methyl iodide and carbon dioxide. The weakness of this mechanism lies in the absence of a reasonable process for formation of the iodine acetate. Of the three possibilities, nucleophilic attack on the phenyl ring, formation of phenyl cations, and benzyne formation, Leffler 51 favors the first. The fThViPT" TYiTirh"] <">+. 3 . inrin'hpri7Prip . nr>p>+.nr> nr>nrl. mnrt phlnTnf.nlnpnp ny^ pYnlm'nprl .

- 161 - by a radical mechanism (Ccheme IX, R = CH3 ). Following homolytic dissociation, fragmentation to methyl radicals occurs, but no cage recombination product, methyl acetate, is seen. Evidence for the presence of methyl radicals e Scheme VIII Ionic Dissociation

Phi 27

RI + C02 Scheme IX Homolytic Dissociation

I /OCR RC H C1 + Phi Phl^ > Phi 6 4 "OCR EC 6H5C1- mS^ + RCOOH

II 2 C 6H 5C1 + R* + COo includes a chlorotoluene isomer ratio similar to that given by methyl radicals generated in other ways The introduction of phenyl substituents into the acetate group of 2 greatly increases the radical reaction at the expense of the ionic pathway 52 and also causes radical cage recombination products R02CR to be formed. The results summarized in Table 3 show the yield of phenyl ester, presumably from the ionic mechanism, decreases with increasing phenyl substitution and reaches zero when R = triphenylacetate . In this decomposition triphenylmethyl radical was detected by its visible and esr spectra. The products in this 4i extreme .case were entirely those expected from radical cage recombination, iodobenzene in 99^ yield, COo in 98^ yield, and trityl triphenylacetate in

Sh% yield. The rate acceleration by phenyl substitution, shown in Table 3 > is consistent with the first step in the radical mechanism (Scheme IX ) where the carbonyl - R bond is weakened in the transition state.

-Die direct or inenyi auosc l "cut; ion on Keiacive Kaces 3. % Yield of Number of phenyl substituents Relative decomposition rate phenyl ester 1.0 75 1 l.k 11+ 2 26 5 3 1000 The decomposition of phenyl iodine dibenzoate in chlorobenzene has been 55 ' 56 34 35 studied by Leffler , Pausacker, , and Hey and shown to give iodobenzene, benzoic acid, chlorobiphenyl, phenyl benzoate, and carbon dioxide as products. Leffler 56 believes that initially the reaction involves an ionic dissociation and an internal nucleophilic displacement similar to Scheme VIII forming iodine benzoate and phenyl benzoate. Some of the iodine benzoate decomposes to iodobenzene and carbon dioxide, but the rest is thought to react with benzoic anhydride, added to remove traces of water, to give a radical chain carrier (maybe benzoyl radicals) that after an induction period accelerates the decomposition of the dibenzoate via a radical chain mechanism. Due to the complexity of the reactions involved, many details of this mechanism are highly speculative. ,

- 162 - BIBLIOGRAPHY

1. C. Willgerodt, J. Prakt. Chem., 55, I5U (l886). 2. V. Meyer and W. Wachter, Chem. Ber. , 25, 2652 (1892). (196U). 5. E. Shefter and W. Wolf, Nature, 205 , 512 k. P. Askenasy and V. Meyer, Chem. Ber., 26, 155^ (l893)- (181U). 5. J. L. Gay-Lussac, Ann. Chim. Physique (17, 91, 5 6. G. C. Pimental, J. Chem. Phys. , 19, M*6 (1951 )• 7. J. I. Musher, Angew. Chem., Int. Ed. Eng. , 8, 5^ (1969). 8. R. E. Rundle, Survey Progr. Chem., 1? 81 (1963). Pauling, Nature of the Chemical 9. G. Kimball, J. Chem. Phys., 8, 188 (19^0); L. Bond, Cornell University Press, Ithaca, 19^0, p. 92-95, 109-111. 10. E. E. Havinga and E. H. Wiebenga, Reel. Trav. Chim. Pays-Bas, 78, 72^ (1959).

11. A. T. Balabam, Revue Roumaine de Chimie, 1J+, 1281 (1969). and M. Rieber, 12. G. Wittig and K. Clauss, Ann., 578 , I56 (1952); G. Wittig Ann., 562, 187 (19^9). 15. E. M. Archer and T. G. van Schalkwyk, Acta. Cryst., 6, 88 (1953)-

Ik. T. L. Khotsyanova, Dokl. Akad. Nauk. , 110 , 71 (1956). 15. D. A. Bekoe and R. Hulme, Nature, 177, 1230 (1956). 16. J. McCullough, Acta. Cryst., 6, 7^6^1953). 17. F. H. Westheimer, Accounts Chem. Res., 1, 70 (1968). 18. D. P. Craig, R. S. Nyholm, A. Maccoll, L. E. Orgel, and L. E. Sutton, J. Chem. Soc, (195*0. 332 no . ... 19. P. C. Van Der Voorn and R. S. Drago, J. Amer. Chem. Soc, 88, 3255 (1966). 20. C. D. Cornwall and R. S. Yamasaki, J. Chem. Phys., 27, 1060 (l95T). 21. R. Bell and K. J. Morgan, J. Chem. Soc, 1209 (1960X 22. W. D. Johnson and J. E. Sherwood, Aust. J. Chem., 2^, 228l (l97l)- 23. W. C. Agosta, Tetrahedron Lett., Jl, 2681 (1965). 2k. K. Clauss, Chem. Ber., 88, 268 (1955). 25. E. F. Perozzi and J. C. Martin, J. Amer. Chem. Soc, 9k, 5519 (1972). 26. C D. Hall, J. D. Bramblett, and F. F. S. Lin, ibid ., 9k, 9261+ (1972). 27. F. M. Beringer and L. L. Chang, J. Org. Chem., 56, 1+055 (l97l). 28. R. W. LaRochelle and B. M. Trost, J. Amer. Chem. Soc, 95, 6077 (l97l). 29. R. Hoffman, J. M. Howell, and E. L. Muetterties, ibid ., 9k, 501+7 (1972). 50. F. M. Reringer and S. Messing, J. Org. Chem., 57, 21+81+ (1972). 51. F. M. Wringer, P. Ganis, G. Avitabile, and H. Jaffe, ibid ., 57, 879 (1972). 52. F. M. Beringer, and L. L. Chang, ibid ., y[, 1516 (197217" 55. D. D. Tanner and G. C. Gidley, ibid ., ^3, 38 (1968). 3I+. B. M. Lynch and K. H. Pausacker, Aust. J. Chem., 10, 329 (1957). 55. D. H. Hey, C. J. M. Stirling, and G. H. Williams, J. Chem. Soc, lU75 (1956). 36. H. J. Lucas, E. R. Kennedy, Org. Syn. 22, 69 (19^2). 57. J. G. Sharefkin and H. Saltzman, ibid ., 1+3, 62 (1965). 58. W. Honsberg and J. E. Leffler, J. Org. Chem., 26, 753 (l96l). 59. Y. Yamada and M. Okawara, Makromol. Chem., 152, 155 (1972); ibid ., 152, 165 (1972).

kO. K. Vaidyanathan and N. Venkatasubramanian, Curr. Sci. , 39, 255 (1970).

1+1. R. Criegee and H. Beucker, Ann., 5I+I , 218 (1959). k2. L. K. Dyall and K. H. Pausacker, J. Chem. Soc, 5950 (1958). 1+5. K. H. Pausacker, ibid ., 107 (1953)- kk. K. H. Pausacker, ibid ., I989 (1953). I4-5. R. Neu, Chem. Ber., J_2B, 1505 (1939). 1+6. W. D. Johnson and J. E. Sherwood, Aust. J. Chem., 25, 1215 (1972). 1+7. A. R. Fox and K. H. Pausacker, J. Chem. Soc, 295 7l957)- 1+8. J. Kresta and H. K. Livingston, J. Macromol. Sci. Chem., Al+(8 ), 1719 (1970). 1+9. B. Plesnicar and G. A. Russell, Angew. Chem., Int. Ed. Eng., 9, 797 (1970). 50. N. A. Milas and B. Plesnicar, J. Amer. Chem. Soc, 90, 1+1+50 (1968). 51. J. E. Leffler and L. J. Story, ibid ., 89, 2553 (1967T. 52. J. E. Leffler, D. C. Ward, and A. Burduroglu, ibid ., 9k, 5359 (1972). 55. E. B. Merkushev, A. N. Novikov, and L. F. Kharitonova, Soviet J. Org. Chem. 7, 526 (1971). 5U. W. D. Johnson and N. V. Riggs, Aust. J. Chem., 8, I+57 (1955). 55- T. T. Wang and J. E. Leffler, J. Org. Chem., 56, I55I (l97l). 56. J. E. Leffler, W. J. M. Mitchell, and B. C. Menon, ibid., 51, II55 (1966). e ORGANIC SEMINAR ABSTRACTS

1973-74

Semester II

School of Chemical Sciences

Department of Chemistry 3^ University of Illinois -v Urbana, Illinois ^v

ns r,

© St 7 <- h -

SEMINAR TOPICS

II Semester 1973-74

Recent Advances in Multiple (p-p)-rr Bonding to Silicon 115

Leonard J. Adz i ma

Reduction-Elimination of 3-Dicarbonyl Enolates by Lithium Aluminum

Hydri de 1 24 Thanin Utawanit

An Investigation of the Nature of Nitrenium Ions 126 Gaylen R. Brubaker

Steric and Electronic Interactions of Small Ring Cyclophanes 128 E. Robert Fretz

Recent Studies on the Biosynthesis of Vitamin B, 130 ? Patricia L. Cavender

Pseudorotation in Sulfur Compounds 132 Gary W. Astrologes

Synthetic Utility of Mercuration-Demercuration Reactions 134 David Sikkenga

Some New Carbocyclic Ring Expansion Reactions 143 Ikram Said

Organoselenium Chemistry 152 May D. Lu

Progress in Synthesis of Prostanoids 154 Alex M. Nadzan

Effects of Molecular Complexing on Chemical Reactions 156 Gautam Desiraju

Copper (I)-Catalyzed Photocycloaddition Reactions 158 Larry D. Martin

Solvolysis Mechanisms of Vinyl Triflates 167 Leonard J. Adzima r,

© - 115 -

RECENT ADVANCES IN MULTIPLE (p-p)TT BONDING TO SILICON Reported by Leonard J. Adzima January 21, 197^

Introduction One of the most profound differences between the bonding properties of carbon and silicon is the extreme reluctance of silicon to form multiple bonds having pT-pTT overlap. Over the years, a number of chemists have tried 1 2 3 to explain this seemingly unusual behavior. > > Although there have been many attempts to isolate a compound containing a double bond to silicon, no examples exist which are stable under ordinary conditions of temperature and pressure. However, recent work indicates that species containing silicon- carbon double bonds may be formed as intermediates in the pyrolysis of certain cyclic organosilicon compounds. In general, it is found that the reluctance of multiple bond formation is not limited to silicon. Other second -ow elements such as phosphorus and sulfur also show this type of behavior. However, this seminar will be limited to a discussion of (p-p)rr bonding of silicon. Theoretical Considerations The carbon-carbon double bond in ethylene consists of a o bond formed 2 from two sp hybridized orbitals and a tt bond formed from the overlap of two p orbitals. The formation of a silicon-silicon double-bond can be constructed

Figure 1 in a similar manner. However, no clear cut examples have been shown where (p-p)TT bonding to silicon occurs. It should be made clear, however, that multiple bonding in which silicon uses its vacant 3d orbitals for (d-p)ir A,s interactions is well known. For example, a review of bond length data shows that all Si-0 bond lengths are within 0.03A of I.63A. Since the sum of their covalent radii is 1.91A, it is generally agreed that partial multiple bonding is responsible for the short Si-0 bond length. Partial multiple bond character results from the "back coordination" of the lone pair pTr electrons of the oxygen atom into the vacant 3d orbitals of silicon. 6

Si — 6^ (d-p)TT

Pitzer1 provides one explanation for the absence of (p-p)tt bonding in second and subsequent rows of elements by postulating that an important repulsive interaction can come about between a p-orbital of one atom and an inner shell of another atom bonded to it. He calls interactions of this type "inner shell repulsions". By comparing nitrogen with phosphorus, it is seen - lib -

that the inner shell of nitrogen contains only two electrons, whereas in phosphorus there are ten electrons. Therefore, it is reasonable to suppose the inner shell repulsive effects in are N2 smaller than in P2 . As seen pictorally in Figure 2, the large repulsive effect in P2 does not allow a significant decrease in the P-P bond distance for substantial overlap of p-orbitals. Hence without good p-orbital overlap a stable tt bond cannot form.

(a) ~(b) Figure 2 (p-p)tt overlap in (a) N2 and (b) P2 7 Extended Huckel (EHMO) calculations have been done on H2CSiH2 (silaethylene) and H2SiSiH2 (disilaethylene) to try to ascertain the reasons for the low stability of these molecules. Overlap integrals, although not an infallible measure of bond strength, are consistant with the low stability of multiple bonds to silicon as shown in Table 1. Table 1 Overlap Integrals

S S 2pTT-2pTr 0.270 JP^-JP^ 0.228 2pTT-3prr 0.182 3pTr-3drr O.kkQ fl 2prr-3dTT O.362 3drr-3dn O.303 7 Calculations also suggest that the C-Si r\ bond is exceedingly polar and reactive and can be thought of as a carbanion-siliconium ion combination. To clearly define why silicon shows a reluctance to undergo multiple bonding is difficult. The lack of good (p-p)tt overlap is probably important but as to whether this is due to inner shell repulsions or other reasons is presently unknown. Experimental Evidence Over the years, a number of workers have tried to prepare compounds containing multiple bonds to silicon. One example where (p-p)tt bonding is likely is the silicon monoxide (Si-0) molecule which is formed in the gas phase from silicon dioxide and elemental silicon at high temperatures. 8 9 There is spectroscopic and mass spectrometric evidence in support of the structure. However, upon cooling, the vapor polymerizes with probable destruction of the multiple bonds. In I96I Fritz and Grobe10 announced the isolation of the compound

Si2C6H16 (structural formula: Me2Si=CH-SiMe 3 ) from the products of pyrolysis of tetramethylsilane. However, two years later they reported that this compound had identical physical and spectralspe properties with 1,1,3*3- tetramethyl-l,3-disilacyclobutane (l).1). Table 2 shows other examples which /CH 2 X Me2Si SiMe 2 x CHp/ - 117 - were reported to contain (p-p)tr "bonds involving silicon but later found to be incorrect.

Table 2 Proposed (p-p)tt Bonded Organosilanes12 Proposed Structure Established Structure

Ph2Si=CH2 (MePh2Si) 2 Fh .Ph x Si=Si^ (PhSi-Et)n EtX XEt I

(Ph 3 Ge) 2Si=Si(GePh3 ) 2 (Ph3GeSiGePh3 )n H Si=0 f— Si-0 -)n H / H

13 In 1966, Gusel'nikov and Flowers observed that 1,1-disubstituted- 1-silacyclobutanes undergo pyrolysis in the range ^00-^-60° C to give the corresponding substituted 1,3-disilacyclobutanes and ethylene. They suggest a mechanism that involves the initial formation of an unstable intermediate

Rx CH2 R ,CH, R 2 sr \CH2 y 2CH2 =CH2 + \si \Si/ R'" ^CHp/ R f/ \C!Hp/ ^R 1

containing a silicon- carbon double bond. This intermediate then undergoes rapid dimerization to form 1,3-disilacyclobutanes. They also proposed the formation of a bimolecular complex based on the polarity of the silicon- carbon bond, but considered it as less likely. In order to decide between the mechanisms they undertook a kinetic study of the pyrolysis of 1,1-dimethyl- 1-silacyclobutane. The only products over the temperature studied are ethylene and 1,1,3 >3-tetramethyl-l,3-disilacyclobutane. Up to ^Qffo reaction, the kinetic data show that the reaction is first order. The manometric first-order rate plot is curved at high percentage conversion. This is believed to be caused by inhibition by ethylene. Since the reaction is first order, only unimolecular mechanisms were considered. Rate constants and activation parameters for this decomposition are very similar to those for the decomposition of a number of cyclobutanes. Because the silicon- carbon double bond is unstable, rapid dimerization occurs in this case, though not in the carbon analog.

Me 400-J180 > MeCH=CH2 + C 2H 4 Ea=6l,200 cal/mole

Me2Si. ^00-14-60 *Me 2Si=CH2 + C2H4 Ea=63,800 cal/mole

Me2 Si' 2Me2Si=CH2 SiMe, - 118 -

Kinetic schemes 1 and 2 are most realistic.

* A > + A ^ B + C B C k:

k3 2B » D

k3 k4 2B * D B + C > X —*• nonvolatile compounds

Scheme 1 Scheme 2 Scheme 1 considers a reversible unimolecular decomposition of mono- silacyclobutane (A) to form a silicon- carbon double bonded intermediate (B) and ethylene (c) followed by cyclodimerization of (b) to give disilacyclo- butane (d). Scheme 2 involves the reaction of (b) with (c) giving nonvolatile side products as well as cyclodimerization of (b). Upon kinetic analysis of these two schemes, it was found that Scheme 2 predicts a time dependence of the apparent first order rate constant. Experimentally, this is not found. Therefore, Scheme 1 best describes the reaction. Presumably, ethylene inhibition is caused by the reversibility of the first step. They tested this hypothesis in the following manner. They reasoned that if the inhibition is due to the reversibility of the primary reaction it should be easy to prove using other olefins as inhibitors. Experiments were done using propene as inhibitor. Product analysis by gas chromatography shows an additional peak which could be l,l,3(or 2)-trimethyl- 1-silacyclobutane. Also, the mass spectrum of the reaction mixture after concentrating this compound gives a peak with m/e llU which corresponds to

1,1, j5( or 2)-trimethyl-l-silacyclobutane. This intermediate is also capable of undergoing a number of chemical reactions which lend support to a double bond structure. The last step in the thermal decomposition, cyclodemerization to form disilacyclobutanes, is an example of a 1,2-cycloaddition reaction of unsaturated compounds. This type of reaction becomes more favored when there are electron-withdrawing substituents attached to the double bond. Because of the longer C-Si bond, the amount of (p-p)fr overlap may be quite small which would decrease the stability of the double bonded intermediate and increase its tendency to undergo rapid cyclodimerization. Further work by Gusel'nikov and Flowers 14 provides more concrete support for the intermediate formation of Me 2Si=CH2 . Addition of water to the reaction mixture inhibits the formation of 1,1,3 ,3-tetramethy1-1,3- disilacyclobutane and the only additional products formed are trimethyl- silanol and hexamethyldisiloxane. It should be noted that under the same conditions, water and disilacyclobutane do not react. These results seem to indicate that the dimerization of Me 2Si=CH2 and its reaction with water are much faster than reaction with ethylene. It was also found that ammonia gives products characteristic of addition to a double bond. 15 Sommer and Roark have found that intermediates of the form R2Si=CH2

1 react with aldehydes and ketones. The intermediate Me 2Si=CH2 , generated by the pyrolysis of 1,1-dimethyl-l-silacyclobutane at 600° in a stream of (Me SiO) in yields N2 , reacts with excess heptanal to give 1-octene and 2 3 of 35-^0 percent. Additionally, Ph2Si=CH2 (generated by pyrolysis of 1,1-diphenyl-l-silacyclobutane) also reacts with an excess of heptanal to H

- 119 -

Me 2 Si" Me2Si-

SiMe,

/Me 2PSi=CH

[Me2Si=CH2 ] + CH2 =CH2 HoO ML

MeoSiQH Me 3SiNH2

Me SiOH | Me SiNH ; Scheme 3 3 3

Me 3SiOSiMe 3 + H2 Me 3SiNHSiMe 3 + NH 3 give 1-octene (35$) « The substitution of a phenyl for a methyl group on silicon causes no significant effects. They postulate that the reaction proceeds by the following mechanism.

Ri H2 Ri 1 HoC-Si'/ H C=Si / R 2 ff 2 Ri \ 3 R2 R; + R -C-0 ^ . c + \Si/ (RjR^iO). X V R4 » NC=0 R'

Scheme JT~ However, the reaction of 1,1-dimethyl-l-silacyclobutane with 2-pentanone does not give the products expected from Scheme h. Instead, the two major products are 2 and 3

6oo° ,0SiMe~ 0SiMe 3 H f 7 Me Si 1 + MeCPr * H C=C + \c = C + H C=CH 2 No flow 2 2 2 Me/ xEt

Also, reaction with cyclohexanone produces the analogous l-(trimethylsiloxy)- cyclohexene (>35$)« These two reactions indicate that another pathway (Scheme 5) is available for reaction with enolizable ketones. Evidence for this was provided by the reaction of benzophenone, a non-enolizable ketone Me HO H R2SiO H II R2 Si=CH2 C \c=c/ > R3 CH2R 4 Rs^ X R4 R3/ \R,

Me R Si R 2 3 ^° R3 X H SiR A sir! \C^ + 2 I II R4HC— CH CH„ R X 2 R4HC^ 3 ^ R4 \H'

Scheme 5 H

- 120 - ketone with 1,1-dimethyl-l-silacyclobutane. As expected, the products obtained were 1,1-diphenylethylene (>70$) and (Me2SiO) 3 . Therefore, we see that the double bonded intermediate reacts by two different pathways to give either a Wittig-type product or a silylenol ether. Sommer, et. al. 16 also found that silicon- carbon multiple bonded (p-p)tt intermediates react with acetonitrile and alkyl and aryl- substituted acetonitriles to form, in most cases, the corresponding a-trimethylsilyl- substituted acetonitriles. They view the general reaction as proceeding via a four centered transition state. They note that the reaction may not be

Me Sir^^CH Rl 6+ 6- 2 2 I 1 2 Me2Si=CH2 + R R -CH-CN ->Me 3SiC-CN 2 R X R C-- —

I CN completely concerted in going through a mechanism where the silicon atom becomes attached to the a-carbon of the nitrile and the carbon atom of the si lie on- carbon double bond becomes attached to the a-hydrogen of the nitrile. Reaction with acetonitrile-d3 with 1,1- dimethy1-1-silacyclobutane at 610° C yields deuteriotrimethylsilyldeuterioacetonitiles N=C-CD2-S i( CH2D )Me 2 . Spectroscopic evidence for a species containing a silicon carbon double bond has been provided by Barton and Mcintosh17 in low temperature trapping experiments of the gas phase pyrolysis products from 1, 1- dimethy1-1-silacyclo- butane . Depositing the pyrolysis products on a sodium chloride plate at -196° allowed the infrared observation of an unstable product which they suggest is Me2Si=CH2 . The ir of the products at -I96 is essentially that of C

Me2Si 600 Mi Me 2Si- it ^Si=CH ; 7 C2ri4 Me SiMe. the substrate, but, in addition they note a new sharp band at l^OTcm."1 In order to observe the behavior of this new peak, they recorded spectra at 10° intervals. They found that at about -120° the band at l^OJcm" 1 disappeared. When the cell was recooled this band did not reappear and the spectrum was that of 1,1-dimethyl-l-silacyclobutane. These observations are consistant with the reaction of Me2Si=CH2 with CH2 =CH2 to reform 1,1-dimethyl-l-silacyclo- butane . The photolysis of 1,1-diphenylsilacyclobutane at 2537A and 55° forms the intermediate Ph2Si=CH2 which can be trapped with MeOD in 95$ yield. The reaction is 80$ complete after 2k hours with the intermediate Ph2Si=CH2 trapped as Ph2 (CH2D)SiOMe when the photolysis is carried out in cyclohexane-methanol-di solution with high specificity of isotopic label. Proof that the reaction is

Ph2Si hv:55u MeOD C2H4+[Ph Si=CH Ph Si-CH X 2 2 ] > 2 2 2k hr. OMe

a) x=D (ca. 90$) b) x=H (ca. 10$) photoinduced is provided by the "dark" reaction of 1,1-diphenylsilacyclobutane and MeOD in refluxing cyclohexane. The only product is Ph2Si(OMe)CH2CH2CH2X (X=H,D), which is the expected cleavage reaction for an angle- strained silicon atom in a four-membered ring. 19 ^^

- 121 -

The photolysis of 1,1-diphenylsilacyclobutane represents a convenient low temperature method of generating these silaalkenes in high yield with few side products since ethylene is evolved as a gas. In some of these studies various silanones are involved as possible intermediates. A more thorough study has been done by Davidson and 20 Thompson. They postulate dimethyls ilanone, Me2Si=0, as an intermediate in the gas-phase thermal decomposition of octamethylcyclotetroxane. They suggest the following mechanism for the overall decomposition of (Me2SiO) 4 (Scheme 6).

(Me2SiO) 4 ^—^ Me 2Si=0 + (Me2SiO) 3 (l) (-1)

Me2Si=0 + (Me2SiO) 4 ^— (Me2SiO) 5 (2) (-2)

Me 2Si=0 + (Me2SiO) 5 ^— (Me2SiO) 6 (3) (-3)

(Me2SiO) 6 > 2(Me 2SiO) 3 (k) Scheme 6

The final product is mainly (Me2SiO) 3 . When the pyrolysis of (Me2SiO) 4 goes to 25$ completion, the only products are (Me2SiO) 3 and (Me2SiO) 5 . Up to k% decomposition, the reaction is kinetically simple with the decomposition of (Me 2SiO) 4 accurately following a first order rate law. The first k% of the decomposition can be described quite well by reactions (l), (2) and (-2). The other reactions cause the fall-off that is observed thereafter and explain the end product, (Me2SiO) 3 . Good evidence that reactions (l), (2), and (-2) are important in the early stages was obtained by trapping experiments. Upon the addition of ethylene the formation of (Me 2SiO) 5 is found to be almost completely suppressed whereas (Me 2SiO) 3 is not affected. Also, from their kinetic data, they calculated a lower limit for the -1 silicon- oxygen tt bond strength in dimethyls i Ian one to be > 37^ 8 Kcal mole. This value seemed to be reasonable because by taking the dissociation energy 1 of silicon monoxide (187 Kcal mole" ) in consideration with the silicon- oxygen 1 1 tt c bond energy (106 Kcal mole" ) a value of kO.k Kcal mole" for each bond is obtained. Hence, from this data, it seems reasonable that Me2Si=0 is important in theses reactions and could be formed as a reaction intermediate. Ando and co-workers21 have invoked the formation of an unstable inter- mediate with a silicon carbon double bond to explain certain products from the photolysis of ethyltrimethylsilyldiazoacetate. When the photolysis was carried out in alcohols, four products are obtained.

SiMe3 H H

N2C + t-BuOH -^»Me 3Si-C-C02Et + Me 3Si-C-C02t-Bu +

I I I C02Et Ot-Bu Ot-Bu k 17% H Me Me 3Si-C-C02t-Bu + Me2Si C-C02Et

AeJEt Ot-Bu H

Scheme 7 5 8$ 6 35$

The formation of 6 involving a methyl migration is rationalized by the formation of a silicon-carbon double bonded intermediate which may be resonance stabilized as shown in Scheme 8. —

- 122 -

SiMe t-Bu © H < 3 and ^ X MesSi^ C02Et

e -i Me C02Et Me Q C02Et Me © ° ^Si=CT < > ^Si-C^ <- ^)Si-C=C-OEt Me' NMe Me / XMe Me Mei

I ,, _. _ _ . t-BuOH Me 3Si-C=C=0 > * t-BuOH OEt Scheme 8 Roark and Peddle23 have investigated the reactions of 7,8-disilabicyclo- [2.2.2]octa-2,5-dienes in order to obtain evidence for an intermediate containing a silicon-silicon double bond, a disilene. Their experiments centered around the pyrolysis of Z> & a^d %.

Me^i -SiMe2 Me2Si- SiMe 2 Me2Si- SiMe2

8

All three of these compounds (% } 8 and 2,) were found to thermally decompose at 500°, 36O and 260° respectively when heated in sealed tubes. A quantitative amount of the expected aromatic hydrocarbon was obtained along with a trace of trimethylsilane and a nonvolatile viscous white oil. The mechanism of decomposition of these compounds is pictured to be a retro-Diels-Alder reaction. The reactivity J < 8 < is consistant with ( 9) such a reaction, though not uniquely so. Low pressure pyrolyses of 8 and 9 gave the same volatile products. Table k Volatile Pyrolysis Products of 8 Name % yield

Trimethyls ilane 3 3,3-Ditnethyl-l,3-disilabutane 0.5 2-Methyl- 2,k- dis ilapentane 2

1 , 1-D imethyl- 1,3- dis ilacyclobutane 30 l,3-Dimethyl-l,3-disilacyclobutane 10

It was found that exchange of the tetramethyldisilene bridge occurs to give a statistical mixture of monodeuterated and undeuterated J when 7 is pyrolyzed at 500° for 3 hours with 7 equiv of 9-deuterioanthracene. It is . —

- 123 -

MeoSi -SiMe. 500° + anthracene 3 hrs

interesting to note that at their respective cl ^composition temperatures, both

& and 2, react with anthracene to give almost quantitative yields of £. The rate constant for the reaction was shown to he independent of anthracene concentration, in accord with a mechanism involving the trapping of a disilalkene intermediate. Further evidence for the formation of a disilene comes from a mass spectral study of £, 8 and %• The mass spectra of 7> 8 and 9 all gave significant

1" + metastable peaks for the (parent)" > ll6 ~transition (Me2Si=SiMe2 j MW=ll6). The mass spectra were all quite simple with a major fragment occurring at m/e 115. The evidence presented for the formation of tetramethyldisilane is quite strong indeed. It is seen that the disilene is quite unstable but that it could be formed in situ to prepare compounds that would otherwise be difficult to prepare Conclusion

It is seen that progress is being made to understand more about (p-p)tt bonding to silicon. Although no stable compounds exists at standard tempera- tures and pressures, new methods have been developed for generating double- bonded silicon species. These methods may help to find other synthetic pro- cedures for the isolation of a species containing a (p-p)tt bond to silicon which may have considerable stability. Bibliography

1. K. S. Pitzer, J. Amer. Chem. Soc , 70, 21^0 (19^8).

2. R. S. Mulliken, J. Amer. Chem. Soc , 72, kk9& (1950.

3. I. R. Beattie and T. Gilson, Nature , 193, 10M (1962). k. F. G. A. Stone and D. Seyferth, J. Inorg. and Nuc Chem ., 1, 112 (1955).

5. B. J. Aylett, Adv. Inorg. Chem. & Radiochem . , 11, 249 (I96B). 6. W. E. Dasent, "Nonexistent Compounds," Marcel Dekker, New York, N.Y., 1965* 7. M. D. Curtis, J. Organometal^ Chem ., 60, 63 (1973). 8. H. C. Rowlinson and R. F. Barrow, J. Chem. Phys ., 21, 378 (1953). 9. R. F. Porter, W. A. Chupka and M. G. Inghram, J. Chem. Phys ., 23, 216 (1955). 10. G. Fritz and J. Grobe, Z. Anorg. Allgem. Chem ., 311, 325 (I961TT 11. G. Fritz, W. Kemmerling, G. Sonntag, H. J. Becher and E. A. V. Ebsworth, Z. Anorg. Allgem. Chem ., 321 10 (1963).

12. C. J. Attridge, Organometal. Chem. Rev. A , j>, 323 (1970).

13. L. E. Gusel'nikoy and M. C. Flowers, J. Chem. Soc , B 419 (1968). Ik. L. E. Gusel'nikov and M. C. Flowers, Chem. Commun ., QGk (1967). 15. L. H. Sommer and D. N. Roark, J. C. S. Chem. Commun ., 167 (1973).

16. L. H. Sommer, R. D. Bush, C. M. Golino and D. N. Roark, J. Organometal .

Chem . , 59, C17 (1973). 17. T. J. Barton and C. L. Mcintosh, Chem. Commun ., 86l (1972).

18. P. Boudjouk and L. H. Sommer, J. C. S. Chem. Comm . , 5^ (1973).

19. R. Damrauer, Organometallic Chem. Rev . (A), 1972, 8, 67. 20. I. M. T. Davidson and J. F. Thompson, Chem. Commun ., 251 (1971).

21. W. Ando, T. Hagiwara and T. Migita, J. Amer. Chem. Soc . , 95, 7518 (1973). 22. D. N. Roark and G. J. D. Peddle, J. Amer. Chem. Soc, 9k, 5837 (1972).

- 12^ -

REDUCTION- ELIMINATION OF p-DICARBONYL ENOLATES BY LITHIUM ALUMINUM HYDRIDE

Reported by Thanin Utawanit February 21, 197^ The lithium aluminum hydride (LAH) reduction of enolizable 1,3-dicarbonyl compounds is accompanied by an elimination reaction of an intermediate (presumably

a p- alanoxyenolate ) to produce allylic alcohols rather than 1,3-diols as the major reaction products. 1 The basic sequence of reaction is reduction- elimination- reduction. " This reduction- elimination has been observed with p-ketoesters, 2 5 2 6 7-11 12-14 p-ketoaldehydes, ' p-diketones, and malonic esters, and its synthetic utility is exemplified by applications in the syntheses of onoceran-8,8'-diol, 4 cedrol, 8 alantolactone, 15 costol and costic acid. 17" 18 Dreiding and Hartman2 found that both 2- carbethoxycyclohexanone (l) and 2-hydroxymethylenecyclohexanone (2) were reduced by LAH to give mainly the

unsaturated alcohols 2- methylene cyclohexanol (3) and 1- cyclohexenemethanol (J+), and a small amount of 2- hydroxymethylcyclohexanol (5,). It was postulated that (Scheme I) the enolic form of 1 and 2 reacts with LAH to form the enolate salts la and 2a, the latter could suffer a hydride attack to A or B. These are resistant to further reduction, but might be expected to undergo elimination in analogy to the base- catalyzed dehydration of aldols. The nonenolic form is

reduced in the normal fashion to j>. It was subsequently shown that the enolate, prepared by the treatment of a B-ketoester with NaH, underwent reduction nearly exclusively by the elimination pathway. 4 The stable keto form and enol form of 1,1-dibenzoylethane were also demonstrated to afford the diol and the unsaturated alcohol, respectively. 7 Mono substituted malonic esters are less subject to elimination than other 8-dicarbonyl compounds when the free esters are reduced; the preformed enolate s, 12 14 however, are effectively transformed to substututed allylic alcohols. ' a-Carboalkoxy- Y-butyrolactone enolates were found to be useful precusors of 15 16 18 a- methylene- y- butyrolactones by this method. ' ' 14 Marshall and co-workers found that LAH reduction of diethyl cyclohexylmalonate afforded 95$ yield of 1,3- diol, whereas its sodio derivative gave a mixture of

2-cyclohexyl-2-propen-l-ol (62-86$), 2- cyclohexylpropanal , and 2- cyclohexylpropan- l-ol. A more extended mechanistic pathway for the formation of all three products was provided. A particularly favorable case for elimination of arylaminomethylene- malonate esters was reported earlier. 13 One interesting feature of this reduction is its stereospecificity. Frankenfeld 11 and Tyler showed that LAH reduction of acetylacetone produces trans- 3- pent en- 2- ol in 82$ yield. The unsymmetrical 2,^4-hexadione gave two unsaturated alcohols trans- 3-hexen-2-ol (4l$) and trans- 2- hexen- k- ol (17$). No cis isomers were produced in either case. The high degree of stereospecificity of the reaction derives from the elimination step of an aluminum alkoxide in p-alanoxy enolate intermediate, postulated to occur preferentially via the 6-membered cyclic transition state. p-Ketoester enolates were also found to give the stereospecific reduction 3 5 products. ' For example, a- substituted acetoacetate was reduced to the trans- 2- substituted- 2- buten-1-ol and 2- substituted- 1- but en- 3- ol, in a ratio of 5 lT^T The lack of regio- selectivity resulted from preferential hydride attack at the ester carbonyl in the enolate as had been previously observed in the case of cyclic ketoester. 2 This lack of positional selectivity can be overcome by elaboration of the crucial p- alanoxyenolate intermediate by indirect means. We have found that treatment of pr hydroxy esters with aluminum alkoxides, followed by the addition of a strong base, causes enolization of the p-alanoxy esters, which is followed by an analogous elimination. As such a reaction is conducted in a non-reducing medium, the final products obtained are a, p- unsaturated esters. • •

125

aCH2OH

5 (10^)

CH2- 0A1H3 SO

la 2a

I V CH20H CHO CH2

^^0A1H 3 Ctr-OC

it (20$) B I (50$)

BIBLIOGRAPHY

1. H. 0. House, "Modern Synthetic Reactions," Benjamin, 1972, pp. 82-85. 2. A. S. Dreiding and J. A. Hartman, J. Amer. Chem. Soc, 75, 939 (1953). 3- V. M. Mico'vic' and M. Lj. Mihailovic', Bull. Soc. Chim. , Belgrade, 1£, 329 (195*0 1*. E. Romann, A. J. Frey, P. A. Stadler, and A. Eschenmoser, Helv. Chim. Acta.,

40 , 1900 (1957). 5. J. A. Marshall and S. B. Litsas, J. Org. Chem., 37, 1840 (1972). 6. J. C. Richer and R. Clarke, Tetrahedron Lett., 935 (1964). 7. A. S. Dreiding and J. A. Hartmen, J. Amer. Chem. Soc, 75, 3723 (1953).

8. G. Stock and F. H. Clarke, Jr., ibid . , 83, 311*+ (I96l). 9- L. A. Pohoryles, S. Sarel, and R. Ben-Shoshan, J. Org. Chem., 2j+, 1878 (1959). 10. A. Luttringhaus and N. Engelhard, Angew. Chem., 73, 2l8 (1961). 11. J. W. Frankenfeld and W. E. Tyler, III, J. Org. Chem., 36, 2110 (1971). 12. W. J. Bailey, M. E. Hemes, and W. A. Klein, ibid., 28, 1724 (1963)- 13. w. F. Gannon and E. R. Steck, ibid., 27, 4-137 (1962"). 14. J. A. Marshall, N. H. Andersen, and A. R. Hochstetler, ibid., j>2, 113 (1967) 15. J. A. Marshall and N. Cohen, J. Amer. Chem. Soc, 87, 2773 (1965)- 16. J. A. Marshall and N. Cohen, J. org. Chem., 30, 3475 (1965) IT. J. A. Marshall and R. D. Carroll, Tetrahedron Lett., 4223 (1965). 18. J. A. Marshall, M. Pike, and R. D. Carroll, J. Org. Chem., ^1, 2933 (1966). - 126 -

AN INVESTIGATION OF THE NATURE OK NITREN ION£

Reported by Gaylen R. Brubaker March 11, 197^

Although nitrenium ions have been postulated in mechanistic pathways for almost a decade, some question remains as to their existence as discrete intermediates. Inspite of this question, the nitrenium ion model lias provided a basis for the extension of this reaction type into a variety of leaving groups, x ' 2 and has been used to explain a number of ring closures,'^ 4 condensation reactions, '' and the formation of ortlio alkyl aromatic amines. A general review of N-chloro and N-bromoamine reactions, 7 as well as an extensive discussion of the evidence which supi^orts the existence of nitrenium s ions has been available since 1970. Since that time, studies of related systems have raised serious questions about Passman's conclusions and may allow reinterpretation of previous results. This seminar will concentrate on work reported since 1970, and examine its relationships to the nitrenium ion proposal. The mechanistic controversy can be examined by considering the following possibilities. Path a asserts that nitrenium ion 3 is a reaction intermediate

a X - 0CH 3

b X : CI

which can either rearrange to form carbonium ion h, or be converted to the triplet state 6, abstract two hydrogen atoms from solvent, and form the parent amine 1 by the loss of a proton. An alternative pathway results if rearrange- ment occurs with the concerted loss of chlorine. This route would require that the formation of the amine 1 occurs via homolytic cleavage, followed by hydrogen abstraction from solvent. Work by Edwards and coworkers lends support for this type of homolytic cleavage in a study o£ medium ring chloroamines. This study shows that these amines are inert to Ag , but not to Ag°, in the absence of + unchloronated amine, and that Ag is reduced to Ag° in the presence of base. This along with other evidence for a radical mechanism supports a radical

chain reaction in which the parent amine 8_ can be formed by reaction of radicals 5 9 or 10 with solvent (see Scheme 2). Although Edwards has provided support for homolytic cleavage as an alternative to the proposed triplet nitrenium ion for this system, it is difficult to extend this mechanism to a bicyclic system. - 127 Scheme 2 A& -> N

I CI AgCl 8

The existence of the nitrenium ion 3 is further supported by the report of 8 a 59$ yield of parent amine 1, in the carbontetrachloride/methanol, a system in which chain propogation seems highly unlikely. A number of studies have been carried out which help to elucidate the nature of carbon migration. Kovacic has observed the migration of several nonstrained systems including the reaction of l-N,N-dichloro and 1-N-chloro-N-ethylaminoadaman- 9 ' 10 ' 11 tane with A1C1 3 . He observes that under the conditions used, the substitu- tion of Ag for AICI3 severely retards the rate of reaction. In an effort to further understand the migration tendencies, N,N-dichlorocarbinamines were treated with AICI3 and hydrolyzed to the corresponding amines and ketones. The authors suggest that since the observed migration patterns are consistent with those reported for the Schmidt and Baeyer- Villager reactions, it is reasonable to expect that migration is concerted with chloride loss in this system. -1 Several attempts to detect the nitrenium ion by uncatalyzed solvolysis of

chloroamines have also been reported. > Kinetic data from the hydrolysis of 14 methyl substituted N- chloroaziridines, and a Hammett study of the reaction of N- chloro-N-t-butylanilines, are consistent with a heterolytic cleavage with considerable electron deficiency at nitrogen. 15 results cannot, however, be used to distinquish a partial position charge in the transition state, from a positively charge intermediate. It seems that neither proposed mechanism fully explains all of the available data. Perhaps future work will permit an estimation of the effects of different solvent and reactant systems and explain the differences that have been observed. The fact remains, however, that the nitrenium ion has provided a valuable model for advancing the synthetic utility of chloroamines.

BIBLIOGRAPHY

1. P. G. Gassman, K. Shudo, R. L. Cryberg and A. Battisti, Tetrahedron Lett., 875 (1972). 2. P. G. Gassman and G. D. Hart man, J. Amer. Chem. Soc, 95, ^9 (1973). 3. P. G. Gassman and A. Carrasquillo, Tetrahedron Lett., 10$ (1971). k. H. H. Wasserman, H. W. Adickes and 0. E. deOchoa, J. Amer. Chem. Soc, 93, 5586 (1971). 5. 0. E. Edwards, D. Vocelle and J. W. ApSimon, Can. I. Chem., 50, II67 (1972) 6. P. G. Gassman and T. J. van Berger, J. Amer. Chem. Soc, 95? 590 (1973)* 7. P. Kovacic, M. K. Lowery and K. W. Field, Chem. Rev., 70,"T39 (1970). 8. P. G. Gassman, Accounts Chem. Res., 3? 28 (1970). 9- P. Kovacic, J.-H. Liu, E. M. Levi and P. D. Roskos, J. Amer. Chem. Soc, 93, 5801 (1971). 10. P. Kovacic and S. J. Padegimas, J. Org. Chem., 37, 2672 (1972). 11. P. Kovacic, R. D. Fisher and T. D. Bogard, J. Amer. Chem. Soc, 95, 361*6 (1973).

12. P. Kovacic, T. A. Kling and R. E. White, ibid . , 9k, 7^l6 (1972) 13- P. Kovacic, R. E. White, M. B. Wazereno and M. R. Gleissner, J. Org. Chem. 38, 3902 (1973). ilk P. G. Gassman, D. K. Dygos and J. E. Trent, J. Amer. Chem. Soc, 92, 208i* (1970).

15. P. G. Gassman, G. A. Campbell and R. G. Frederick, ibid . , $k, 388^ (1972).

16. P. G. Gassman and G. A. Campbell, ibid . , %h, 3891 (1972). - 128 - STERIC AND ELECTRONIC INTERACTIONS OF SMALL RING CYCLOPRANES

Reported by E. Robert Fretz March 25, 197^ The unusual geometries of meta and para bridged cyclophanes are responsible for the steric and electronic interactions found in these compounds. Since studies of these bridged aromatic compounds have been known for quite some time, 1 this seminar will concentrate on recent developments in this field. SYNTHESIS

Of particular use in the synthesis of cyclophanes is Boekelheide' s recent procedure for transforming a sulfide linkage into a carbon- carbon double bond. This transformation utilizes both a Steven's rearrangement and a Hoffmann 2 elimination. By this method [2.2.2] (1,3,5) cyclophane -1, 9- diene 1 was prepared as shown in Scheme I. Sulfur was also extruded photochemically forming carbon- carbon

Scheme I

st> 1. (MeO) 2 CHBF 4 ^

2. KOt-Bu

Figure 1

3 single bonds in the bridges of metacyclophanes. Recently [8] paracyclophane has been prepared by cycloaddition of dispiro [2.2.2.2] deca-4,9-diene with 1,3- butadiene. CYCLOPHANES CONTAINING ONE AROMATIC RING

The rigidity of the methylene bridge in [7] metacyclophane has been studied and an energy barrier of 11. 5 kcal/mole for inversion of the bridge has been 5 > G>7 8 9 ' l0 found. Recently, [6] metacyclophane and [7] paracyclophane have also been prepared. These contain the shortest bridges and the most steric constraints of this class of cyclophanes. CYCLOPHANES CONTAINING TWO OR MORE AROMATIC RINGS The X-ray structure of a variety of cyclophanes has been determined showing 11>12 in detail the deformations of the aromatic rings. Benzene rings in [2.2] paracyclophane are bent by an angle of 12°, and the benzene rings are held 2.78 % apart, a distance which is closer than the normal intermolecular distance between stacked aromatic rings in their crystalline state. The structure of 13 a quadruple layered cyclophane is shown in Figure l. Of particular interest are the electronic and steric interactions which arise when two aromatic rings are held in close proximity. The energy barrier to the conformational flipping of the meta bridged ring in [2.2] metaparacyclophane has been found to be decreased in both [2.2] metaparacyclophane (1,9)- diene and 2,6-pyridino [2.2] paracyclophane. 15 The structural details of this latter compound have also recently been described where the two aromatic rings were found to be perpendicular to one another. 16 Transannular electronic interactions -12 9-

have produced shifts to higher wave lengths in the uv absorptions of certain cyclophanes, as well as increases in their TT-basicities. llD ' 1 ^' 17 Transannular interactions have been studied in multi- layered cyclophanes where it was found that the transannular tt-tt interactions between aromatic rings held between two other aromatic rings are increased over the interactions found in [2.2] € cyclophanes. l8_23 Recently transannular effects have been found to be important in determining the course of reactions involving cyclophanes. Cram has demonstrated that transannular stabilization of carbonium ions formed in the solvolysis of optically active 1-tosyloxy- [2,2] paracyclophane could be occurring. 24 Also actual bond formation between the benzene rings of both [2.2.2] (1,3,5) 2a 25 cyclophane and [2.2] metaparacyclophane have indicated transannular stabili- zation of positive charges. BIBLIOGRAPHY

1. For reviews see: (a) B. H. Smith, "Bridged Aromatic Compounds," Academic Press, New York, 1964; (b) D. J. Cram, Rec Chem. Progr., 20, 71, (1959); (c) R. Griffin, Chem Rev., 63, ^5 (1963); (d) F. Vogtle and P. Neumann, Chimia, 26, 64 (1972); (e) F. Vogtle and P. Neumann, Angew. Chem. Int. Ed., 11, 73 (1972); (f) D. J. Cram and J. M. Cram, Ace. Chem. Res., k, 204 (1971). 2. Ta") v. Boekelheide and R. Hollins, J. Amer. Chem. Soc, 95,3201 (1973); (b) V. Boekelheide and P. Anderson, J. Org. Chem., 38, 3928 (1973); (c) V. Boekelheide and R. Hollis, J. Amer. Chem. Soc, 92, 3512 (1970); (d) R. Mitchell and V. Boekelheide, ibid ., £2, 3510 (1970); (e) R. Mitchell and V. Boekelheide, Tet.

Lett., 1197 (1970); (f) V. Boekelheide and J. Mondt, ibid . , 1203 (1970);

(g) V. Boekelheide, P. Anderson, ibid . , 1207 (1970); (h) V. Boekelheide and

J. Lawson, Chem. Comm. , 1558 (1970).

3. V. Boekelheide, I. Reingold and M. Tuttle, ibid . , 4o6 (1973). k. T. Tsuji and S. Nishida, J. Amer. Chem. Soc, 95, 7519 (1973). 5- S. Fujita, K. Imaniura, and H. Nozaki, Bull. Chem. Soc. J., l88l (1972). g" 6. S. Fujita, S. Hirano, and H. Nozaki, Tet. Lett., 403.'(1972).

7- S. Fujita, Y. Hayashi, and H. Nozaki, ibid . , 16^5 (1972). 8. S. Hirano, T. Biyama, S. Fujita, and H. Nozaki, Chem. Lett., 707 (1972). 9- N. Allinger and T. Walter, J. Amer. Chem. Soc, 9k, 9267 (1972).

10. A. Wolf, V. Kane, R. Levin, and M. Jones, ibid . , 95, 1680 (1973). 11. K. Trueblood, J. Bernstein, and H. Hope, Acta. Cryst., B28, 1733 (1972) 12. C. Coulter and K. Trueblood, Acta. Cryst., l6, 667 (196577 13. H. Mizumo, K. Nishiguchi, T. Otsubo, S. Misumi, and N. Morimoto, Tet. Lett., 4981 (1972). lk. D. T. Hefelfinger and D. J. Cram., J. Amer. Chem. Soc, 92, 107^ (1970). 15. (a) V. Boekelheide, P. Anderson and T. Hylton, ibid .,

(b) S. Sherrod, R. Da Costa, R. Barener and V. Boekelheide, ibid . , 96, 1565

(1974); (c) V. Boekelheide, K. Galuszko and K. Szeto, ibid . , 96, 15W (1974); (d) S. Sherrod and V. Boekelheide, J. Amer. Chem. Soc, £4, 5513 (1972).

16. L. Weaver and B. Mathews, ibid , 96, 1581 (197*0. 17. T. Otsubo, S. Mizogami, Y. Sakata and S. Misumi, Chem. Comm., 678 (1971). 18. T. Otsubo, S. Mizogami, Y. Sakata and S. Misumi, Tet. Lett., 2^57 (1973)-

19. T. Umemoto, T. Otsubo, Y. Sakata and S. Misumi, ibid . , 593, (1973).

20. N. Kannen, T. Umemoto, T. Otsubo and S. Misumi, ibid . , *+53T (1973).

21. T. Otsubo, Z. Tozaka, S. Mizogami, Y. Sakata and S. Misumi, ibid . , 2927 (1972). 22. T. Otsubo, S. Mizogami, Y. Sakata and S. Misumi, Chem. Comm., 678 (1971). 23. T. Otsubo, S. Mizogami, Y. Sakata and S. Misumi, Tet. Lett., 4803 (1971). 2k. R. Singler and D. Cram, J. Amer. Chem. Soc, 93, 44*+3 (1971). ^ 25. D. Hefelfinger and D. Cram, ibid . 475*+ t , 93, (1971). 130

RECENT STUDIES ON THE BIOSYNTHESIS OF VITAMIN B 12 Reported by Patricia L. Cavender March 28, 197^

First isolated in 19^8, vitamin B 12 was synthesized in 1972 by R. B. 1 2 3 Woodward. The biosynthetic pathway, ' especially the condensation of four porphobilogen 2(PBG) to form uroporphyrinogen (uro'gen) III J5 and the subsequent ring contraction and reduction to the corrin 6, has not yet been completely elucidated.

The first step of the biosynthesis of vitamin B 12 , the condensation of glycine and succinic acid to form 6- aminolevulinic acid 1 (ALA), is 2> 3) 4 catalysed by ALA synthetase. Two molecules of ALA are combined to form J PBG 2 by ALA dehydrase. Many mechanisms2>6-H have been proposed to explain

the polymerization of four PBG to uro'gen III _5 by the enzymes uro'gen I " synthetase and uro'gen III cosynthetase. Recent progress in the synthesis 12 14 of dipyrrylmethanes 3 (DPM) has made these specifically labelled compounds available for incorporation studies. Bf9f 15-2C There has been some question as to the role of uro'gen III as an intermediate 21 in the biosynthesis. " Several groups have shown that seven methyl groups, z< 2T including the Ci methyl are derived from methionine. It has been suggested by Dolphin that the cobalt atom would facilitate methylation and may be inserted at this point rather than after corrin formation. 2

X)0H + H00C- s H00( NHr cooh Nllo' ^M NH2 ^N^ H DPM PBG 2 3 CH, CH, CH, H,NOC-CH2 A=CH2 C00H V-ch 2 ch 2 £nh, P=CH2 CH2C00H

H,N0CHp i h

.7 CM CH CH CONH >'1^1m : O'C-CHaCHj^CH CH, 2 2 3 i * NH IPA ' -COp CH, DMBI H,C-CH CH3 a P- /

Corrin 6 k uro' gen I R 2=P Vitamin B ia 7 R2=A 5 uro'gen III R X=A HO-CHj R2-P

In the conversion of cobyrinic acid to vitamin B l2 7, there are 190 possible intermediates. It has been found that amide formation occurs concurrently with the addition of the isopropanolamine (IPA) and phos- phate- ribose-dimethylbenzimidazole (DMBl) moieties. The top ligand, 5'- 2 3 deoxyadenosyl in the coenzyme, is added last. ' L- threonine is the 29 ' 30 precursor of IPA Labelling studies point to a common biosynthetic pathway for dimethylbenzimiiazole and riboflavin, both deriving from 31" 33 6,7- dimethyl-- 8- ribityl- lumazine (DMRL ) . The recent report of a stable cell- free cobyrinic acid producing preparation from P. Sherman ii promises rapid progress in further elucidation 34 of the biosynthesis of vitamin B 12 . - 131

BIBLIOGRAPHY

1. R. B. Woodward, Pure and Applied Chem., 33> 1*1-5 (1973). 2. B. F. Burnham, "Metabolic Pathways," ed. I). M. Greeriberg, Academic f Press, New York, 1969, Vol. Ill, 3rd edn. p. 403. 3- H. C. Friedman, Ann. Rev. Micro., 24, 159 (1970). k. P. L. Scholnick, L. E. Hammaker, and II. S. Marver, J. Biol. Chem. ?h'( 4l32 (1972). I. 5. A. Scott, C. A. Townsend, K. Okada, and M. Kajiwara, 'Trans. N. Y. Acad. Sci., 35. 72 (1973). 6. J. Dalton and R. C Dougherty, Nature, 223 , 1151 (1969). 7- A. R. Battersby, 23rd International Congress of Pure and Applied Chemistry, Special Lectures, _5, 1 (1971). 8. E. B. C. Llambias and A. M. C. Ratlee, .J., Biochem. 121 , 327 (P/fl). 9. R. Radmer and L. Bogorad, Biochem., 11, 904 (1972). 10. C. S. Russell, J. Theor. Bio., 3_2' 2T( (1972). 11. A. R. Battersby, E. Hunt, and E. McDonald, Chem. Commun., 442 (1973). 12. A. R. Battersby, J. F. Beck and E. McDonald, J. Chem. Soc, Perkin Trans. I, 160 (1974). 13- B. Prydman, G. Buldain, and-). C. Repetto, J. Org. Chem., 38, 1824 (1973). 14. A. M. A. R. Gonsalves, G. W. Kerner and K. M. Smith, Tetrahedron Lett., 2203 (1972). 15- A. M. Stella, V. E. Parera, E. B. C. Llambias and A. M. C. Batlle, Biochim.

Biophys. Acta, 252 , kdl (1971). 16. R. B. prydman, A. Valasinas and B. Prydman, U-S Z. Physiol. Chem., 554, " 839 (1973). 17. R. B. Frydman, A. Valasinas and B. Frydman, Biochem., 12, 80 (1973). 18. R. C. Davies and A. Neuberger, Biochem. J., 133, 4fl (1973). 19. A. R. Battersby, II. K. Gibson, L. N. Mender, ,1. Moron and L. N. Nixon, g Chem. Commun., 768 (1973). V 20. J. Plusec and L. Bogorad, Biochem., 9, V736 (1970). 21. A. I. Scott, C. A. Townsend, K. Okada, M. Kajiwara, and R. J. Cushley, J. Amer. Chem. Soc, 9k, 8269 (1972).

22. B. Franck, D. 1 '. Cantz, F. P. Montforts and J Schmidt chen, Aug. Chem., Int. Ed., 11, 421 (1972). 23. G. Muller and W. Dieterle, H-S Z. Physiol. Chem., 352 , 143 (.1971). 2k. A. R. Battersby, M. Ihara, E. McDonald, ,J . R. Stephenson and B. T. Golding, Chem. Commun., kok (1973). 25. C E. Brown, J. J. Katz and D. Shemin, J. Biol. Chem., 248, 8015 (1973). 26. A. I. Scott, C. A. Townsend, K. Okada, M. Kajiwara, P. J. Whitman and R. J. Cushley, J. Amer. Chem. Soc, £4, 8267 (1972).

27. A. I. . Scott, C A. Townsend, and R. J. Cushley, ibid , 95, 5759 (1973)- 28. D. Dolphin, Bioorg. Chem., 2, 155 (1973;. 29. D. A. Lowe and J. M. Turner, J. Gen. Micro., 64, 119 (1970). 30. G. Muller, R. Gross, and D. Siebke, H-S Z. Physiol. Chem., 352, 1720 (1971). 31. S. H. Lu and W. L. Alworth, Biochem., 11, 608 (1972). 32. P. Renz and A. J. Bauer-David, Z. Naturforsch, 27B, 539 (1972). 33- P- Renz and R. Wehhenmeyer, FEBS Letters, 22, 124 (1972). 34. A. I. Scott, B. Yagen, E. Lee, J. Amer. Chem. Soc, 22> 5'f6l (1973). ' -132- PSEUDOROTATION IN SULFUR COMPOUNDS Reported by Gary W. Astrologes April 8, 19T^ 1 19 Gutowsky and Hoffman showed that the F nmr of PF 5 consists of only a doublet. To explain this, Berry2 suggested an intramolecular mechanism called pseudorotation (see Figure l) in which small movements interconvert two equatorial and two axial substituents of a trigonal bipyramid. Whites ides and coworkers 3 showed that the Berry mechanism or some permutationally equivalent mechanism4 ' 5 was responsible for the ligand exchange in phosphoranes. 5 ' 6

u*

1 .J Figure 1 c C- _) 5*

In his papers 2 Berry suggested that his pseudorotation mechanism could apply to SF4 , which has a trigonal bipyramid structure with an equatorial electron pair. 7 ' 8 Early workers 9 ' 10 detected fluorine exchange in SF by low temperature nmr studies but favored mechanisms 4 ' involving dimeric intermediates 10 13 or catalysis by HF impurities. 11 To get around impurity and association problems Levin and Harris studied the far- infrared spectrum of gaseous SF 4 and found evidence for a double minimum potential function for intramolecular fluorine exchange with a barrier of 10.0 + 0.5 kcal/mole. 14 Nmr studies of purified SF 4 shewed that it is really an A2B2 system with the additional splittings observable at -101°. 15 Using this observation, Klemperer16 has recently analyzed variable temperature nmr spectra of SF4 and compared the results to that expected from the hO)hQ\ possible bimolecular and intramolecular fluorine exchange processes, Computer simulations showed that only a two pair exchange process 4 ' 5 such as the Berry mechanism2 (Figure l) could fit his spectra. 17 Many other substituted analogs of SF4 (sulfuranes) prepared or postulated as intermediates 18 have shown no evidence for ligand exchange.

C CF3 D B D CF • CF3 A 3 CF 3 CF 3

x0 'CFo B B CF3 I *« S mil O- x CF r o D 3 © CF CF CF 3 3 ^"»CF A 3 3 C A C s CF 3 B D CF 3 CF,

l9 (CH 3 )2NSF 3 has a temperature dependent F spectrum but whether this is of 19 intermolecular or intramolecular origin is unknown. Fast Pseudorotation is consistent with the nmr spectra of several isolated compounds . ^ 21 19 22 One of these l20 has been studied further by variable temperature F nmr. It's nmr at 25° shows two multiplets, one of which broadens below -^0° and splits into two peaks. AG41 for this process is 8.5+ 1 kcal/mole at -80° .

The other multiplet only broadens down to -150° . Pseudorotation of 1 would -133-

interchange endo groups A and C and exo groups B and D leading to the observed spectra if the endo groups have a larger difference in chemical shifts than the exo groups. Mechanisms involving sulfonium ions can be eliminated since all groups become equivalent after several ion- izations as shown by the plane of symmetry in 2. Neither ionization nor inversion of 1 can be detected by nmr at 200° indicating that Ag^1 for these processes must exceed 23 kcal/mole. This can be compared to inversion barriers of 25-29 kcal/nole observed in sulfonium ions by 23 others *

References ,

1. H. S. Gutowsky and C. J. Hoffman, Phy. Rev. 80, 110 (1950)j H. S. Gutowsky and C. J. Hoffman, J. Chem. Phys.> 19 1 1259 (1951)* 2, R. S. Berry, ibid ., 32, 933 (I960)* R. S. Berry, Rev. Mod. Phys., 32, 447 (1960*7^ 3« G. M. Whitesides and W. M. Bunting, J. Amer. Chem. Soc, 89, 6801 (I967)j G. M. Whitesides and H. G. Mitchell, ibid ., 91, 53B4 (I969). 4. J. I. Musher, ibid ., $k, 5662 (1972). 5» I. Ugi, D. Marquarding, H. Klusacek, P. Gillespie, and F. Ramirez, Accounts Chem. Res., 4, 288 (1971). 6. F. H. Westheimer, ibid ., 1, 70 (1968); K. Mislow, ibid., 3, 321 (1970)1 I* Ugi, D. Marquarding, H. Klusacek, G. Gokel, and P. Gillespie, Angew. Chem., Int. Ed. Engl., 9, 703 (1970). 7. R. E. Dodd, L. A. Woodward, and H. L. Roberts, Trans. Faraday Soc, 52, 1052 (1956).

8. W. M. Tolles and W. D. Gwinn, J. Chem. Phys . , 36, 1119 (1962).

9. F. A. Cotton, J. W. George, J. S. Waugh, ibid . , 28, 994 (1958). 10. E. Li Muetterties and W. D. Phillips, J. Amer. Chem. Soc, 8l, 1084 (1959). 11. E. L. Muetterties and W. D. Phillips, J. Chem. Phys, 46, 2861 (I967). 12. R. L. Redington and C. V. Berney, ibid., 43, 2020 (1955).

13. R. L. Redington and C. V. Berney, ibid . , t&, 2862 (1967)1 14. I. W. Levin and W. C. Harris, ibid ., _55, 3048 (1971). 15. J. Bacon, R. J. Gillespie, and J. W. Quail, Can. J. Chem., 41, 1016 (I963). 16. W. G. Klemperer, reported at the 166th American Chemical Society Annual Meeting, Chicago, Illinois, August 1973* 17. W. A. Sheppard and D. W. Ovenall, Org. Magn. Resonance, 4, 695 (1972)) D. B. Denney, D« Z. Denney, and Y. F. Hsu, J. Amer. Chem. Soc, 95, 4064 (1975)* J. C. Martin and R. J. Arhart, ibid., 93* 2339 (197l7j

R. J. Arhart and J. C. Martin, ibid . , 94, 4997 (197?)) R. M. Rosenburg and E. L. Muetterties, Inorg. Chem., 1, 756 (1962)* Jj_ I. Barragh and D. W. A. Sharp, Angew. Chem., Int. Ed. Engl., 9, 73 (I970). 18. R. Tang and K. Mislow, J. Amer. Chem. Soc, 91, 56"44 (1969); H. K. Kwart and H. Omura, J. Amer. Chem. Soc, 93 7250 (1971). 19« G. C. Demitras and A. G. MacDiarmid, Inorg. Chem., 6, 1903 (1967). 20. M. Allen, A. F. Janzen, and C. J. Willis, Chem. Commun., 55 (1968); M. Allen, A. F. Janzen, and C. J. Willis, Can. J. Chem., 46, 3671 (I968). 21. W. A* Sheppard, J. Amer. Chem. Soc, 93, 5597 (1971). 22. J. C. Martin and G. W. Astrologes, unpublished results. 23. K. K. Anderson, M. Cinquini, and N. E. Papanikolaou, J. Org. Chem., 35> 706 (1970); R. Scartazzini and K. Mislow, Tetrahedron Lett., 2719 U967); D. Darwish and R. L. Tomilson, J. Amer. Chem. Soc, 90, 2~ t 5938 (1968). - 13^ - SYNTHETIC UTILITY OF MERCURATION- DEMERCURAT ION REACTIONS

Reported by David Sikkenga April 11, 197^ INTRODUCTION Mercuration of carbon- earbon double bonds was first used in 1900 to synthesize a number of organomereurials. For the next fifty years, the interest in this field was centered around the structure of the organo- mercurial and the mechanism of its formation. Chatt in his 1951 review of the addition of mercuric salts to olefins only mentions briefly the possibility of replacing the mercury of the organomercurial with hydrogen to give the organic compound. 3a The stereochemistry of the addition reaction was also a topic of debate and much work as illustrated by Zefirov's 39 review in 1965 dedicated entirely to this topic. Even by 1968, little work had been done in utilizing the organomercurial as the intermediate of additional reactions by following its preparation with its reduction. Kitching's 6 review of oxymetalations (1968) only touches on this possibility, However, in the last five years, many papers have appeared illustrating various organic compounds prepared by the mercuration- demercurat ion sequence. The synthetic utility of this method is the focus of this seminar. The topic will be restricted in that it will deal only with additions to carbon- carbon multiple bonds, Additions to mono unsaturated alkenes is reviewed first. However, several papers have appeared on reactions of other multiple carbon- carbon unsaturated systems and will also be discussed. REACTIONS OF ALKENES Oxymercuration (Synthesis of Alcohols, Ethers and Acetate) Most olefins readily react with mercuric salts in the presence of water, alcohols or acetic acid to give the corresponding oxymercuration product which can then be reduced with alkaline solutions of NaBH4 to give the alcohol, ether or acetate (Scheme I) by Markovnikov addition, A variety of mercuric salts

Scheme I HgX OC ^ -C-i- —21 ROH L I NaBH4

R = H, alkyl or acetyl have been employed including the acetate, trifluoroacetate, nitrate, and halide mercuric salts, Brown and coworkers have developed a general synthetic method " 4 for hydration of olefins to give high yields of a variety of alcohols 1 without rearrangement, Using mercuric acetate in a 50% THF/H2O solvent, reaction of many olefins occured at room temperature in a matter of minutes. Reaction yields and rates were found to depend on the degree of olefin substitution. Reactivity trends are illustrated in Scheme II. For any degree of olefin

Scheme II R R R R v ^C=CH2 > R-CH=CH2 > R-CH=CH-R > R-C=CH-R > C=C R ( cis > trans ) R K substitution, increased branching of the alkyl group decreased yields and rates. Alkenes with a high degree of substitution gave side reactions which could be minimized by running the reaction at lower temperatures. Strained double bonds such as in norbornene react more readily than unstrained olefins. This may also be the reason for cis olefins reacting faster than trans olefins. - 135 - Ethers from a number of ole fin/alcohol combinations were also reported by 5 20 Brown. ' Methanol, ethanol, and isopropyl alcohol reacted smoothly to give Markovnikov addition products, Again, the degree of olefin substitution affected reaction rates and yields. Brown also reported that an increase in branching of the alcohol also reduced rates and yields. Using t-Butyl alcohol, yields were low partially due to competition of the acetate anion of the mercuric acetate with the alcohol to give the acetate. Using the less nucleophilic trifluoroacetate salt prevented this competition. The mechanism of the oxymercuration reaction involves electrophilic attack of the mercury on the least substituted carbon to give a cat ionic intermediate which is then attacked by the nucleophile (scheme III), The OAc Scheme III <5 Hg-OAc R / H HgOAc V/ R HgOAc S\ > R-&-C-Hu

6 7 structure of the cat ionic intermediate has been a subject of much debate. ' Some consider it a bridged mercuriura ion with the positive charge on the mercury. Brown 3 sees no conclusive evidence for the mercurium ion and thus postulates a "carbonium ion with a large fraction of the charge remaining on the mercury moiety." Pasto 7 finds only trans diaxial addition to substituted cyclohexenes much as are found for bromination where a bridged brominium ion exists and cites this as evidence for the mercurium ion. Yet Brown notes that norbornene oxymercuration gives ci s- exo products and concludes that a cyclic intermediate is not involved. 8 The mechanism of the NaBH4 reduction is also a matter of uncertainty,41 Whitesides has given good evidence for a free radical reaction9 involving breaking of the carbon- mercury bond to form the radical. However, what this radical abstracts hydrogen from is uncertain. An intermediate alkylmercuric hydride or mercurous hydride generated in situ have been suggested. Acetic acid as the solvent of a oxymercuration- demer cur at ion reaction will yield the acetate. The acetate has also been noted as a byproduct by workers using other solvents. A modified demercuration method has led to the 10 11 formation of allyl acetates in 85-100$ yields. ' Oxymercuration- demercuration has also been utilized to induce optical activity. Using optically active carboxylic acid-mercuric salts in place of mercuric acetate, 1-decene was hydrated to 2-decanol in 78$ yield and with 17.^0 enantiomeric excess. Assymetric hydrations of other olefins are also 12 13 reported, > Synthesis of Alkyl- and Hydro- Peroxides Bloodworth and coworkers have synthesized secondary t-Butyl peroxides 14 15 16 using oxymer curation- demer curation ' ' as illustrated in Scheme IV. Using a one fold exce,ss of tert - Butyl peroxide in methylene chloride under nitrogen the peroxymercurial was obtained in yields of around 50$ after purification using column chromatography. Also found was 15-20$ of the acetoxy mercurial formed by competition of acetate with the peroxide for the cationic intermediate. Results are given for several olefins containing either C2 , C 4 , C6 and Cs alkyl groups. The demercuration was accomplished by halogenation to yield the p haloperoxide or by the usual NaBH 4 reduction. The latter yielded a mixture of products as Scheme IV illustrates. Both the expected peroxide and an epoxide were obtained. The relative amount of the - 136 - Scheme IV

OOt-Bu

R- CH- CH3 +

OOtBu Hg(OAc) 2 R-CH=CH ->RCH-CH 2 t-BuOOH 2 I CH2 C12 HgOAc

OOt-Bu

I R- CH- CH2

X = Br, I two was dependent upon the temperature of reduction and the amount of alkyl substitution on the carbon containing the peroxide. Higher temperature and higher substitution favored formation of epoxide up to as much as 87$ of the product from a- methyl styrene. Olefins containing an aryl and an alkyl group on the a- carbon yielded specially large amounts of epoxide. Bloodworth postulates homolytic cleavage of the mercury- carbon bond to give an alkyl radical which can then either react with a hydrogen atom to give the peroxide or react with the oxygen of the peroxide to form the epoxide with expulsion of a tert- Butyl alkoxide radical (t-BuO)- Typical yields of alkyl peroxides from mono substituted olefins after reduction with NaBH4 are around hO°Jo. Although several cases are known of hydroperoxide- mer curat ion to form 1V 13 19 R2C(OOH)CH2HgOAc, > > there have been no reports of demer curat ions with NaBH 4 to produce the alkyl hydroperoxide. Synthesis of Amines, Amides, and Alkyl Azides Nucleophilic nitrogen containing compounds can also be used to attack the cationic intermediate created by mercuration of an olefin. For example, Brown and coworkers20 utilized acetonitrile as the nucleophile in the mercuration- demer cur at ion reaction (Scheme v). Brown found that mercuric nitrate Scheme V

Hg(N0 NaBH 3 ) ; 4 RCH^CH2 -> R-CH-CHp-HgNOa > R-CH-CH3 CH 3C=N OH 1 hr H-N-CCH3 Room temp CH^

0N0; was the reagent of choice since the nitrate ion did not form the nucleophilic addition product as both acetate and trifluoroacetate did. A number of amides were synthesized in this manner (Table I). As with the oxymercurations, the more highly substituted olefins were less reactive to amination. Sokolov and coworkers have used other nitriles to aminate olefins using mercuration techniques. 21 However, they did not demercurate to obtain the amides. Hydrolysis of the amides would afford the amines providing a method for Markovnikov amination of olefins. Certain amines have been reported to add to double bonds via aminomercuration methods. 2 Generally, much longer reaction times are needed and mercuric halides are used. Pyrrolidine reacts with propene in THF using .

- 137 -

Table I

OLEFIN ACETAMIDE YIELD (%) 1- Hexene N- 2- Hexyl 92 1-Decene N-2-Decyl 86 3,3-Dimethyl-l-butene N-Pinacolyl 90 Styrene N-a-Phenethyl 50 Cyclopentene N- Cyclopentyl 70 Cyclohexene N-Cyclohexyl 95 mercuric chloride to yield 75% of isopropylpyrrolidine after hydrogenolysis Acetoxymercuric compounds formed in an oxymercuration reaction were found to react with piper idine to yield the piperidino- mercuric compound by displacement of the acetate by piperidine in 65% yield. Alkyl azides have recently been synthesized by the mercuration-demercura- 29 tion method. Heathcock reports moderate yields (50-70%) of the alkyl azides derived from addition of azide ion to the mercurated cation followed by sodium borohydride reduction. Only terminal or strained olefins reacted. The general proceedure was very similar to Brown's proceedure for oxymercuration in that a 50% THF/H2 solvent and mercuric acetate were used. A three molar excess of sodium azide was employed. REACTIONS OF DIENES Oxymer curat ions Brown's hydration method has also been applied to dienes. Using one mole of mercuric salt to one mole of diene, a variety of enols were obtained. 4 Reactions were run at 0°C in a 80% THF/h2 solvent. Unconjugated symmetrical dienes gave the Markovnikov enol in relatively low yields (40-50%) after several hours of reaction at 0°C using mercuric trifluoroacetate. The enol was accompanied by the diol and unreacted diene. Brown justifies his low yields by noting that since one mole of mercuric salt is used, the statistical yields are 25% diol, 50% enol, and 25% diene. For unsymmetric dienes the possibility exists of greater reactivity of one double bond than the other. For the dienes used, reactivities agreed with those cited for hydration of olefins (Scheme II). Thus limonene was converted to its enol by hydration of the more reactive terminal double bond in 70% yield (Scheme Vl) in 30 minutes. Scheme VI

1) Hg(0Ac) 2 THF/H2

2) NaBhVOH'

In cases of dienes containing double bonds of similar reactivity (i.e. ^-vinyl- 1-cyclohexene) mercuric trifluoroacetate was found to be more selective than mercuric acetate. The presence of silver acetate in the mono hydration of

^4-- vinyl- 1- cyclohexene using mercuric acetate was reported to give low yields of the ketone that would be formed by oxidation of the alcohol formed in the normal mercuration-demercuration sequence. ?p The use of two moles or more of the mercuric salt extends the use of dienes to the formation of diols. 1,5 and 1,6 dienes react to give tetra- hydrofurans and tetrahydropyrans respectively. 2 ' 21 Unsaturated alcohols also give the diols or the ethers depending on the relative positions of - 138 - e£f double bond and hydroxy group. The ethers are formed by intramolecular hydration: The hydroxy group adding to the first double bond serves as the nucleophile for attack on the cation formed by mercuration of the second double bond (Scheme VII ). Brown reported again that only the Markovnikov Scheme VII OH

Hg( °Ac) 2 THF/ H2° CH =CH-(CH ), -CH=CH2 > CH2- CH- (CH2 ) - CH=CHS 2 2 NT M HgOAc

N=3 Hg(OAc) 2 AcOHgH2C CH2HgOAc V

OH CH CH-(CH -9H-CH ^\ 2 2 ) N 2 NaBH 4 I I V L 1 HgOAC OH H Hg AcOHgCH ^ 2 ^H2HgOAc OAc

OH OH NaBH 4 NaBH 4

cis and trans CH3-gH-(CH -nH-CH 2 ) N 3 H OH

i cis and trans products are formed. The reactions were reported to give high yields with short reaction times. l,lj~pentadiene give 83% yield of 2,k pentanediol in lj-5 minutes, and the reactions in Scheme VII took place in 30 minutes. Cyclic dienes react much as the acyclic dienes giving the ether products when a five or six membered ring is possible. Mixtures of the possible ether products are usually obtained as is the case for the cyclic diols. 1,*+- cyclohexadiene gives both cis and trans isomers of both 1,3- and 1,^- cyclohexanediol. 1, 5-cyclooctadiene gives a 77% yield of a mixture of the 13 two ethers (Scheme VIII ). Finally, Grundon used optically active mercuric Scheme VIII

1) Hg(0Ac) 2 H20/THF -> 2 ) NaBHVoir

19$ carboxylic acid salts to induce optical activity in the dihydrobenzofuran found from cyclization of 0-allylphenol. The amount of assymetric induction was small (5% enantiomeric excess) but reproducible. }

- 159 -

Aminomercuration of Dienes away from the double Olefins containing an amino group three carbons § much as unsaturated alcohols form bond are reported to form cyclic amines 24 25 been extended to the reaction of dienes cyclic ethers. > This has 26 27 1,5- ' thus, aniline reacts with cis with p- substituted anilines cyclooctadiene as shown in Scheme IX. Scheme IX HgBr

Hg(OAc) 2 THF 0-NH 2 25°C 1 hr KBr

b yield

with aniline in a similar Non.cyclic 1,5- dienes (eg. 1, 5-hexadiene) react compounds. manner to form the N- phenyl pyrolidine AND ESTERS REACTIONS OF a, 6- UNSATURATED KETONES rise to a new class of Work by Bloodworth and coworkers has given ^ unsaturated ketones or esters. oxymercuration reactions - those involving Markovnikov addition pattern of simple These reactions do not follow the found to add to the alpha olefins and dienes, The oxy- nucleophile is only in th alpha Y position for compounds with substituents e ££?? it adds to the beta.fposition.<°^fThis and for all other substitution patterns, either methanol or ter£-butyl pattern holds for both ketones and esters using of the oxymercurials with hydroperoxide as the nucleophiles. Reduction or beta methoxy and tert-butyl sodium borohydrode gives the expercted alpha oxymercurials yield large amounts peroxide ketones and esters although some ° f eP other 2!alkoxy-2-acyiaminopropionic acids, difficult to synthesize by of oxymercuration- demercuration of methods, have been formed as th| product 1 in the presence of alcohols.

H Q

CH2=C-N-C-R 1 =CH(CH CH COOH CH2 2 ) 5 3 2 1 MULTIPLE BOND SYSTEMS MERCURATION REACTIONS OF OTHER CARBON- CARBON Alkynes react ions recently reported a number of synthet ically £f£X Hudrlik has 1-ene 33 Using 1-octyne he has made three 2- substituted- of terminal alkynes. -as ether 3 and enamine k The enol -etf^ °compouX enofacetate 2, enol "* prepared using a catalytic amount of mercuric represent^^V^ff^asethe two stage boron trifluoride etherate and thus does not the enol ether and mercuration-demercuration synthetic procedure. However,

OR

CH3-C-(CH2 ) 5-CH3 CH2=CH(CH2 )sCH5^n 3 CH2=CH(CH2 ) 5 CH 3

R=CH 3 , C.H 5

h . s

- iko - enamine were synthesized using one molar equivalent of mercuric acetate follow by demercuration by sodium borohydride. The enol ether was prepared using methanol {6% yield) and ethanol (36/0 yield). Tert -butyl alcohol did not give the desired product. Addition follows Markovnikov ' rule and there was no rearrangement to form internal double bond isomers. The reaction of 1-octyne with aziridine produced the expected addition product h in only 17$ yield. When the same reaction was run in pyrrolidine, the enamine was not obtained but rather the saturated N-2-octylpyrrolidine

was obtained. The iminium salt _5 was suggested as the intermediate under- going reduction based on other known reductions of iminium salts by sodium borohydride, 34 The smaller three membered ring aziridine iminium salt would not be as stable and would probably not be formed. Allenes A number of allenes have been alkoxymercurated-demercurated to form the 3- alkoxy- 1- ene systems. Although this reaction proceeds in moderate yield, the synthetic utility is questionable due to the unavailibility of the starting allenes, The main interest in recent years has been the stereospecificity of the reaction. 35; 36; 37 Recent work has shown that stereospecificity depends on the structure of the allene and the mercuric salt used.

Cyclopropane s

Although saturated, cyclopropane s react much as olefins to give oxymercurated products. Sokolov reports the hydroperoxy-mer curat ion of 19 phenyl cyclopropane in 56% yield to form the hydroperoxy-mer cur ial 6 The relative reactivities of one cyclopropyl system and a olefin system can be compared by the C6H5- CH- CH2- CH2 product distribution of hydroxymer curat ion- demer cur a- 40 tion ° f carene ;(Scheme X.) The reaction produced (!)0H HsOAc > in low yield a mixture of 7, 8_, and £. Although the 6 double bond is tri substituted and of low reactivity by Brown's standards (Scheme II), it appears to be more reactive in that mole ratios of _T :8_ were k.k:l with only a trace of £ produced.

Scheme X

OH

1) Hg(0Ac) 2 THF/h2

2) NS1BH4/OH"

SUMMARY The mercuration- demercuration sequence has a wide variety of synthetic uses. Figure I illustrates some of the known possibilities. Other nucleophilic additions may be adde'd to the list as more work is reported. These reactions proceed in most cases with little if any of the rearrangements characteristic of other addition reactions involving cationic intermediates. Mild conditions are another feature of this sequence. Room temperature is sufficient for most reactions and lower temperatures have been used without causing excessively long reaction times. High yields are obtained with mono and disubstituted olefins although yield decreases as degree of substitution increases. Figure I Addition Possibilities Using Mercuration-Demercuration

Ethers Alcohols Acetates ROH

ROOH \ / RChN Alkyl peroxides Amides

R2NH

HOOH N< Amines Hydro peroxides Alkyl Azides

BIBLIOGRAPHY

1. H, C. Brown and P. J. Geoghegan, Jr., J. Org. Chem., £5, l8Wi- (1970). 2. H, C, Brown, P. J. Geoghegan, Jr., J. T. Kurek and G. J. Lynch, Organometal Chem, Syn,, 1, 7 (1970/1971). 3. H, G, Brown and P. J. Geoghegan, Jr., J, Org. Chem., ,37, 1937 (1972). k, H, C, Brown, P. J. Geoghegan, Jr., G. T. Lynch and J. T. Kurek, J. Org. Chem,, 37, 19^1 (1972). 5. H. C, Brown and M. H. Rei, J. Amer. Chem. Soc, 8£, 1^22 (I967).

6. W, Kitching, Organometal Chem. Rev., 3,, 6l (1968)7 7. D, J. Pasto and J. A. Gontarz, J. Amer. Chem. Soc, £3, 6902 (1971)- 8. H. C, Brown and K. T. Liu, J. Amer. Chem. Soc, £2, 3502 (1970). 9. G, M. Whitesides and J. San Filippo, J. Amer. Chem. Soc, £2, 66ll (1970). 10. Z. Rappoport, L. K. Dyall, S. Winstein and W. G. Young, Tetrahedron Lett., 3^83 (1970). 11. Z. Rappoport, P. D. Sleezer, S. Winstein and W. G. Young, Tetrahedron Lett., 3719 (1965). 12. R. M, Carlson and A. K. Funk, Tetrahedron Lett., 3661 (1971).

13. M. F. Grundon, D. Stewart, and W. E. Watts, Chem. Commun. , 573 (1973). lh, D. H. Ballard, A. J. Bloodworth, and R. J. Bunce, Chem, Commun., 815 (I969). 15, D. H. Ballard and A. J. Bloodworth, J. Chem. Soc, C, 9^5 (1971), 16, A. J. Bloodworth and G. S. Sylina, J. Chem Soc, Perkin Trans, 1 2^-33 (1972).

17, V. E. Schmitz, A. Rieche, and 0. Brede, J. Prakt. Cyem. , 312, 30 (1970).

18, V. I. Sokolov and 0. A. Reutov, Zh. Org. Khim. , 5_ Ilk (I9W), 19, V, I. Sokolov, Izv. Akad. Nauk SSSR, Ser. Khim., IO89 (1972),

20, H, C, Brown and J. , T. Kurek, J. Amer. Chem. Soc, $1, 56^7 (19^9 ). 21, V, I. Sokolov and A. Reutov, Izv. Akad. Nauk SSSR, Ser Khim,, 1, 225 (1968), 22, S, Moon and B. H, Waxman, J. Org. Chem., 3k, 1157 (1969). 23, S, Moon, B. H. Waxman, and Takakes, J. Org. Chem., 3^, 2951 (19^9). 24, J, J. Perie, J, P, Laval, J. Roussel and A. Lattes, Tetrahedron, 675 (1972). I 25, A. Dobrev, J. J. Perie, and A Lattes, Tetrahedron Lett., ^013 (1972). - 142 -

26. V. G. Aranda, J. B. Mur, G. Asensio, and M. Yus, Tetrahedron Lett., 3621 (1972). 27. V. G. Aranda, J, Barluenga, and M. Y. Astiz, Anales De Quimica, 68, 221 (1972). 28. J. J. Perie and A, Lattes, Tetrahedron Lett., 2289 (1969). 29. C. H, Heathcock, Angew. Chem. Int. Ed., 8, lj>k (1969). 30. A. J, Bloodworth and R. J. Bunce, J. Chem. Soc, C, 1^+53 (1971). 31. A. J, Bloodworth and R. J. Bunce, J. Chem. Soc, Perkin Trans. 1, 2787 (1972). 32. C. Gallina. M. Maneschi, and A. Romeo, J. Chem. Soc, Perkin Trans. 1, 113^ (1973). 33. P. F, Hudrlik and A. M. Hudrlik, J. Org. Chem., 38, k25k (1973). 3^, R. D. Back and D. K. Mitra, Chem Commun., 1^33 (1971). 35- R. D. Back, J. Amer. Chem. Soc, 91, 1771 (1969). 36, W, S. Linn, W. L. Waters, and M. C. Caserio, J. Amer. Chem. Soc, 92, i+018 (1970), 37i R. Vaidyanthaswamy and D. Devaprabhakara, Tetrahedron Lett., 915 (1971). 38. J. Chatt, Chem. Review, kS, 7 (1951). 39. N, S. Zefirov, Russ. Chem. Rev., 3k, 527 (1965). hO. B. A. Arbuzov, V. V. Ratner, Z. G. Isavea and E. Kazakova, Izv. Akad.

Nauk. SSSR, Ser. Khim. , 385 (1972).

Ul f D. J, Pasto and J, A. Gontarz, J. Amer. Chem. Soc, 91 » 719 (1969). 1 .

- 143 -

SOME NEW CARBOCYCLIC RING EXPANSION REACTIONS Reported by Ikram Said April 25, 197^ The use of carbocyclic ring expansion reactions in organic synthesis is well known and various ring expansion reactions have been developed. 1 Some of the classical methods applied for ring homologation by one carbon atom are the Demjanov rearrangement, 2 the Tiffeneau-Demjanov rearrangement2 and the pinacol rearrangement. 3 Well known ring homologation methods which incorporate a heteroatom into the ring are the Bayer Villiger 4 reaction (oxygen) and the Beckmann 5 rearrangement. Some examples of ring homologation 7 by two or more carbon atoms are also known. This seminar will review some ring expansion reactions that have been developed recently, and where possible, comparisons with some of the more established ring expansion reactions shall be made.

THE DIAZOMETHANE REACTION The most widely used methods of ring expansion are the reaction of cyclic ketones with diazoalkanes 6 and the Tiffeneau-Demjanov rearrangement of amino alcohols derived from the cyclic ketones. The diazoalkane ring expansion of cycloalkanones have been extensively studied and its great utility in organic synthesis have been shown. Some disadvantages of these reactions are: formation of epoxides (3) as a side reaction, formation of isomers (1 and 2), over homologation, low overall yields and reduced reactivity with large ring ketones. These difficulties have been overcome

r~\ RCHH. S .0© <\=0 c.v, (C (CH2 ) -, lN " n-3 ¥W/ / CHN 2

(CH 2 )

10 11 by the use of catalysts such as alcohols (methanol or ethanol ) or Lewis 12 13 + 14 acids such as A1C13 and BF 3 , and Et 3 BF4 , In most cases yields are usually high and formation of epoxides is greatly reduced or completely eliminated. The catalysts are believed to coordinate strongly with the carbonyl oxygen, increasing the activity of the carbonyl carbon to nucleophilic attack and preventing the formation of epoxides. As a result weakly nucleo- philic diazoalkanes like ethyldiazoacetate can be used15 and ring expansion 16 17 of 01, (3-unsaturated ketones, which normally give pyrazolines in the absence of catalysts', can be carried out. The effect of the catalysts in the diazoalkane ring expansion reaction can be compared to the Tiffeneau- Demjanov rearrangement of amino alcohols which does not form any epoxides. The deamination reaction have been resorted to in cases where ring expansion by the diazoalkane reaction failed. 18 However it has the disadvantage of having low overall yields - ikh -

THE CYANOGEN AZIDE REACTION 19 In 196^4- Hermes and Marsh reported, the synthesis of Cyanogen azide (N3CN) and showed that cyclopentanone can he obtained from the reaction of cyclopentene and the azide. It is believed that the olefin and cyanogen

/Vncn % CNN. ^ h^

azide first forms a cyanotriazoline (h) by 1,3 dipolar addition which then decomposes into products with evolution of nitrogen. Hermes and Marsh have further shown the reaction of cyanogen azide with numerous olefins 20 to form alkylidenecyanamides (_5) and N-cyanoaziridines (6). In all cases addition is such that the nitrogen bearing the cyano group becomes attached to the more substituted olefinic carbon and that 5 arises by rearrangement from the carbon bearing the -NCN group. McMurry2 ! has found that when the

7N CN •>s S NC-N CNN 3 „ V% } (- + y x ±-k h cynogen azide is allowed to react with an exocyclic olefin an alkyl migration occurs to give the ring expanded ketone. The ring expanded ketone was obtained in moderately high yields and the reaction worked well with large ring ketones too (Table i).

TABLE I Ring Expansion of Methylenecycloalkanes with Cyanogen Azide. 21

CH2

II c

(CH ^> 2* n n 3 k 5 6 7 11

The ring expansion reaction was also carried out with alkylidenecyclohexanes and the corresponding 2-substituted homologous ketones were obtained in good yields (Table II). The reaction of cyanogen azide with some unsymmetrically substituted methylenecycloalkanes is shown in Table III. The results in Tables II and III show the formation of mixtures which indicate that there is no migrational selectivity i.e. primary, secondary and tertiary ring bonds all seem to migrate with approximately equal ease. - i4 5 ~

TABLE II Ring Expansion of Alkylidenecyclohexanes with Cyanogen Azide. 21 '3

Yield %

8o

+ 85

+ 85

44 56

+ 76

The reaction was also extended to some exocyclic dienes to test its usefulness since ring expansion of the corresponding a, P-unsaturated ketones with diazo- methane have not been too successful. Low yields and mixtures were obtained (Table IV). House and coworkers 13 have shown that in the reaction of acyclic TABLE III Comparison of Ring Expansion of Unsymmetrically Substituted Cycloalkanones with Diazomethane and Cyanogen Azide Yield % +

55 CH2N 2 31:69 OWa. 52:48 8o X=CH2

+

CHoN1M 27:73 x=o 2 2 5 C-Ma. 41; 59 59 X=CH2 +

CH2N2 o

CM. .. 80 X=CH : 55:^5 ketones with diazomethane the order of migration is C 6H 5 ~(CHr,) 2 =CH->CH3~ CH 3CH2CH2>(CH 3 ) 2 CH'vinyl group to migrate better than the alkyl group. The opposite seems to occur in the case of 11 , again indicating little migratory selectivity. Better yields were obtained when the reaction was carried out in the presence of Ag as catalyst.

TABLE IV Ring Expansion of Dienes with Cyanogen Azide 21

Yield % + CD 10 CM. 70:30 50 CM. 72:28 8k

AgBF4

jCD 4- 11 cm: 13:87 60 CM. AgBF, 17:83 95 The cyanogen azide reaction probably involves a diazonium zwitterion intermediate, as indicated by the similar product distribution of the reaction of unsymmetrical eye loalkan ones (Table III) with CM 3 and diazomethane. However such an intermediate is uncertain since in most carbonium ion rearrangement there is some migratory aptitude which is not observed here. If a carbonium intermediate were involved one would expect much aziridine (6) to be formed but none was found here. For a concerted mechanism resulting in the loss of nitrogen little aziridine should be formed because of the difficulty of back side displacement. This assumes that rotation about the central C-C bond is slow since dipolar attraction between the two charge-bearing groups would tend to hold the system rigid and prevent rotation. The ring expansion of methylenecycloalkanes have also been carried out with arenesulfonyl azides. 22 Quantitative yields of the imide were

NCN R

R?r R R

CN Ry^ - ik-j - obtained and greater than $0% of the ketone after hydrolysis. Arenesulfonyl azide has the advantage of being less hazardous to handle and reactions can be carried out at higher temperatures. TABLE V

Arenesulfonimides from methylenecycloalkanes22a

ArS0 2N 3 C=CH2 > C=IW~S0 2Ar (CH (CH 2 ) n+1

Yield %

3 99 k 99 5 98 6 100

DECOMPOSITION OF MAGNESIUM SALTS OF VARIOUS 1- (l-BROMOALKYL)-l- CYCLOALKANOLS

The rearrangement of magnesium salts of halohydrins to ketones have been known for some time. Studies have shown23 that the halo and hydroxyl groups must be cis to effect rearrangement, the trans isomer gives polymeric products. Primary halides do not rearrange unless a good migrating group is involved (eg phenyl), while secondary and tertiary halides rearrange to the 24 corresponding ketones. Sisti, applied this reaction as a method for ring enlargement on a series of compounds depicted in 12 to form the correspond- ing ring expanded ketone. All the ketones obtained were at least > 90% pure (by vpc), the main contaminant being the isomeric ketone lM_ based on the same retention time and nmr with authentic samples. The isomeric ketones could have arisen from the isomeric halohydrins lj? whose magnesium salts would yield lh upon decomposition via a hydride or methyl shift. Or it could have been formed by the breakdown of the expoxides 16

TABLE VI Decomposition of the Magnesium Salts of Some Bromohydrins

1) MgBr/Et 2 (CH (CH b 2 ) -> 2 ) n \Rp 2) C 6H6 , A Ri

12 13

n=k n=5 n=6 Ref.

: "25 1 R2=CsH5 72% = r R2 Cri3 60% 25 =R2 : =CH 3 *M 57% 26 Br

!— OR; (CH2 )

Ro R]_

Ik 15 16 .

- 148 - formed from 12. The decomposition of the magnesium salts of a series norbornane halohydrins IT produced 18 by migration of the C2-C3 bond, none of the isomeric ketone 1^ , via migration of the more substituted C;l-C2 bond, was obtained. There is torsional strain in IT due to the eclipsed non-bonded interactions between the substituents on C-2 and the protons in C-3> non- bonded interactions between substituents on C-2 and the bridgehead proton are much less since the dihedral angles are kk° (H-OH) and T9° (H-CR^Br). As a result there is a preference for C2-C3 migration for relief of the eclipsing strain. The reaction with halohydrins having a primary halide was shown by the decomposition of the magnesium salt of 1-bromomethyl-l cyclohexanol 12 (n=5, Ri=R2-H) which produced cyclohexanecarbonaldehyde and no ring expanded product. This confirmed earlier work23 which showed that primary halides do not rearrange unless a good migrating group is present.

\ \-MgBr/Et x 2 > 2) C6H6 , A

Ri=H, R2-C6H5 Rj_ = H, R2= CH3 Rl =R2~CH3

Practical wise, the ring expansion reaction can be carried out conveniently under anhydrous conditions in refluxing benzene, so far only isopropylmagnesium I bromide have been used in all cases to obtain the magnesium salt. Yields are reasonably high and the reaction would be useful for obtaining 2- substituted cycloalkanones. However, polymeric materials are sometimes obtained due to unreacted halohydrin 12_, some of which possess extremely labile bromine. The halohydrins were usually obtained from the corresponding olefins by reaction with aqueous N-bromosuccinimide (NBS). RING OPENING OF BIS (ALKYLTHIO) CYCLOPROPANES

Bis (alkylthio) cyclopropanes 20 can be readily obtained from a, P 28 2 31 unsaturated aldehydes, epoxides, "^~"or from the dibromocarbeneadduct. The hydrolysis of 20 would produce the ring enlarged alkoxy ketone 21. When this was tested with compounds 22 and 23 under the usual hydrolysis conditions, ring opening did not occur except when 23 was heated m formic acid. A milder and more effective hydrolysis have since been developed, 31 the results are shown in Table VIII, the thioacetal was prepared mainly from the dibromoadduct

s _- .CHO 20 21 14-9 -

C02Me \_/^oh HgCl2/HgO,CH3OH 2 days, 65° 56$

22 = R-R=(CH2 ) 3 SR i3 (r CH 3 )

HgCls/HgO HgCls/HgO CH 3OH, k days CH3OH, 3.5 days 65° 65° COpMe OCH3 HgCl SCH3

dk-i

TABLE VIII Conversion of Some Bis (Methylthio) Cyclopropanes to Ketones. 31

H CH y (CH SCH3 2 ) * '*'. r>< CH SCH3

20 TFA-H2 (1:1) room temp. 5-10 hr. CH;

(cafe) n CHfe'

n Yield

3 55 if 90 6 50 10 TO - 150 - CONCLUSION The carbocyclic ring expansion reactions that have been discussed here are only some of the many ways that one can expand a cyclic compound. The cyanogen azide. reaction has great potential since it is applicable to a variety of cycloalkanones, both substituted and un symmetrical, and also to exocyclic dienes. Yields are high although mixtures are obtained in some cases. The main disadvantage of using cyanogen azide is that it is sensitive to thermal and mechanical shock and is potentially explosive. Using arenesulfonyl azide would reduce such problems, yields in this case are usually > 80% from the corresponding methylene cycloalkanes. The decomposition of magnesium salts of halohydrins is limited to those with secondary and tertiary halides, furthermore these halohydrins are quite labile and polymeric products are obtained. Ring opening of bis (alkylthio) cyclopropanes would be an attractive procedure, however one has to go through a long reaction sequence before the bis (thioalkyl) cyclopropane can be obtained. BIBLIOGRAPHY

1. C. D. Gutsche and D. Redmore, Carbocyclic Ring Expansion Reactions,

Academic Press, New York, N.Y. , (1968). 2. P. A. Smith and D. R. Baer, Organic Reactions, vol. 11, Wiley, New York, N.Y., p 157, (i960). 3- Y. Pocker in Molecular Rearrangements, Part 1, P. de Mayo, Ed., Wiley, New York, N.Y., p 1, (196*0. 4. C. H. Hassal, Organic Reactions, vol. 9, Wiley, New York, N.Y., p 73, (1957). 5. L. A. Donaruma and W. Z. Heldt, Organic Reactions, vol. 11, Wiley, New

York, N.Y. , p 1, (i960).

6. C. D. Gutsche, Organic Reactions, vol, 8, Wiely, New York, N.Y. , p 364 (195*0.

7- R. G. Carlson and E. L. Biersmith, Chem. Comm. , 18, 1049 (1969). 8. W. F. Berkowitz and A. A. Ozorio, J. Org. Chem., 36, 3787 (1971). 9. M. Higo, T. Sakashita, M. Toyoda and T. Mukaiyama, Bull. Chem. Soc, Japan, 45, 250 (1972). 10. A. Fachinetti, F. Pietra, and A. Marili, Tetrahedron Lett., 393 (1971). 11. a) P. K. Kadoba, Syn., 2, 71 (1973); b) J. A. Marshall and J. J. Partridge, J. Org. Chem., 33, 4090 (1968). 12. E. Muller, H. Kessler and B. Zeeh, Forschr. Chem. Forsch., 7, 128 (1966).

13. H. 0. House, E. J. Grubbs, and W. F. Gannon, J. Amer. Chem. Soc, 82 , 4099 (i960). 14. W. L. Mock and M. E. Hartman, J. Amer. Chem. Soc, £2, 5767 (1970 ). 15. W. T. Tai and E. W. Warnhoff, Can. J. Chem., 42, 1333 (196*0. 16. a) E. Muller and R. Heischkeil, Tetrahedron Lett., 1023 (1962); b) W. S. Johnson, M. Newman, S. P. Birkeland and N. A. Fedruk, J. Amer. Chem. Soc, 84, 989 (i960); c) E. Enzell, Tetrahedron Lett., 185 (1962). 17. G. W. Cowell and A. Ledwith, Quart. Rev., 2*+, 119 (1970). 18. a) J. T. Lumb and G. H. Whitham, Tetrahedron, 21, 499 (1965); b) H. Velgova and V. Cerny, Coll. Czech., 35, 24o8 (1970). 19. F. D. Marsh and M. E. Hermes, J. Amer. Chem. Soc, 86, 4506 (196*4-). 20. F. D. Marsh and 'M. E. Hermes, J. Org. Chem., 37, 29^9 (1972). 21. a) J. E. McMurry, J. Amer. Chem. Soc, 9±, 3670" (1969); b) J- E. McMurry and A. P. Coppolino, J. Org. Chem., 38, 2821 (1973). 22. a) R. A. Wohl, J. Org. Chem., 38, 38^2 (1973); b) R. A. Wohl, Helv. Chem.

Acta., 56, 1826 (1973) • 23. J- A. Geismann and R. T. Akawie, J. Amer. Chem. Soc, 73, 1993 (1951). 24. A. J. Sisti, J. Org. Chem., 33, 453 (1968). 25. A. J. Sisti, J. Org. Chem., 35, 2670 (1970). 26. A. J. Sisti and M. Meyers, J. Org. Chem., 38, 4*4-31 (1973). •

- 151 -

27- A. J. Sisti, Tetrahedron Lett., 5327 (196?)

28. T. P. O'Brien, A. T. Rachlin and S. Teitel, J. Med. Chem. , 12, 1112 (1969). 29. D. Seebach, Syn., 1, 17 (1969). 30. D. Seebach and M. Braun, Angew Chem., Qk, 60 (1972). 31. D. Seebach, M. Braun and N. Du Preez, Tetrahedron Lett., 36, 3509 (1973). < . -152- ORGANOSELENIUM CHEMISTRY

Reported by May D. Lu April 22, I97U 1>p In spite of the considerable attention - it has received, the mechanism of allylic oxidation of olefins by selenium dioxide remains controversial. Recently, Sharpless and coworkers 3 ' 4 suggested a general mechanism for the oxidation process as shown in Scheme I. In

order to support the presence of a [2,3 1 sigmatropic shift, they

Scheme I

Hv OH Se ') rJ -OH T^JT* l X' X Y X X

3 studied the oxidation (by H ? 2 ) of allyl phenyl selenides ^ and g, and found high yields of the rearranged alcohols. They also reported that

with appropriate substrates , the inter- mediate formed from the initial ene addition (step a) can be trapped as 4 Sc}sePh seleninolactones PhSe The selenoxide analog of the sulfoxide syn- elimination reaction was 2_ 2 ~ ~ discovered 5 three years ago. Since then it has attracted considerable atten- tion3,6 ' 7,a on account of its reaction condition being milder and the reaction proceeds faster than the sulfoxide analogs. Sharpless and Lauer8 reported that alkyl phenyl selenoxides bearing a P-hydrogen undergo stereospecific syn- elimination to form olefins at room temperature. Recently a number of reports on the synthetic applications 8 ' 9 ' x x ,13, i 4 of this facile syn- elimination process have appeared (Scheme II).

Scheme II HO excess 10 hr EtQH H20p 5 HOSePh 2 hr 0-2^ r.t. SePh SePh r.t.

NaI0 (1) LiN-i-Pr , THF, -?8° 4 f^V^^ 2 [j -78p CH 0H-H e^O Kj? (2) PhSeBr, , 3 2 / NaHC05 Ph if - 25° PhSeOH 1 hr, 89$

SePh STePh . ,0 VL&

R X X

>0 o HI SePh ( n W , -153-

In the conversion of epoxides to allylic alcohols, 9 generally, the elimination of the |3-hydroxy selenides occurs away from the hydroxyl group. The a-phenylselenocarbonyl compounds used in the synthesis of a^P-unsaturated carbonyls are readily formed from the nucleophilic

selenium reagent PhSe Na or from the electrophilic selenium reagents ( PhSeCl and PhSeBr. The yields of the above synthesis were reported to 10 ,l3 be better than procedures previously available j the reaction conditions were milder , permitting the presence of other labile functionalities. The synthetic utility of this gentle olefin forming process is limited only by the methods 1 5,16,1Y available for introducing selenium, as for example, PhSe into substrate molecules. As can be seen, the use of organoseleniums in synthesis has great potentials. However, a word of caution , selenium containing compounds are very toxic.

REFERENCES

1* E. N. Trachtenberg in "Oxidation," Vol. 1, R. L* Augustine, Ed., Marcel Dekker, New York, N. Y., 1969> Chapter 3, p. 119. 2* R. A. Jerussi in "Selective Organic Transformations," Vol* 1, B. S* Thyagarajan, Ed., Wiley- Interscience, New York, N. Y.» 1970, Chapter 6, p. 301. 3* K. B. Sharpless and R. F. Lauer, J. Amer. Chem. Soc, 94, 715^ (1972).

hi D. Arigoni, A. Vasella, K. B. Sharpless, and H. P. Jensen, ibid . 95, 7917 (1973). 5« D. N» Jones, D. Mundy, and R. D. Whitehouse, Chem. Commun., 86 (1970). 6. R. Walter and J. Roy, J. Org. Chem., 36, 2561 (1971). 7. K. B. Sharpless and R. F. Lauer, ibid ., 37, 3973 (1972). 8. K. B. Sharpless, M. W. Young, and R. F. Lauer, Tetrahedron Lett., 1979 (1973). 9. Ki B. Sharpless and R. F. Lauer, J. Amer. Chem. Soc, 95, 2697 (1973). 10* (a) A. C. Cope, H. H. Lee, and H. E. Petree, ibid., 80, 28U9 (1958). (b) A. C. Cope and J. K. Heeren, ibid., 87, 3125 (19o3). 11* H* J. Reich, I. L. Reich, and J. M. Renga, ibid., 95, 5813 (1973). 12. K. B. Sharpless, R. F. Lauer, and A. Y. Teranishi, J. Amer. Chem. Soc, 95, 6137 (1973). 13* (a) T. A. Spencer, R. A. J. Smith, D. L. Storm, and R. M. Villarica,

ibid . « 93, ^856 (I97l)> (b) W. G. Dauben, M. Lorber, and D. S. Fullerton,

J. Org. Chem., 3h, 3587 (1969) J (c) R. J. Teisen, ibid . . 36, 752 (1971). lUi P* A. Grieco and M. Miyashita, ibid ., 39, 120 (197^7^ 15. H* J. Reich, ibid .* 39, ^28 (19WJT 16* D. J. Clive* Chem. Commun., 695 (1973).

17. K» B* Sharpless and R. F* Lauer, J. Org. Chem., 39, ^29 ( 197*0 • . -154- PR OGRESS IN SYNTHESIS OF PROSTANOIDS

Reported by Alex M, Nadzan April 29, 197^ Prostaglandins (PG's) are a family of naturally occuring carboxylic acids Which exhibit a broad range of biological activity. 1 They occur in five series, designated by the letters E, F, A, B and C, depending on the sub- stitution within the prostanoid ring. Due to their limited availability from natural sources, considerable effort has been directed toward the synthesis of prostanoids over the past decade. 2 Synthesis of prostaglandin analogs has also been actively pursued in hope of separating the diverse biological properties associated with each of these fascinating natural products. This seminar will focus on significant advances in prostanoid synthesis during the last few years One of the most elegant, versatile and potentially practical syntheses developed has been that of E. J. Corey and coworkers. 3 All of the primary prostaglandins (PGE and PGF series) were synthesized in a highly stereo- controlled manner from a single key intermediate (i). Resolution at an early stage of synthesis afforded optically active I which permitted generation of prostanoids in naturally occuring form. Current refinements in Corey's synthesis have been directed at large scale production. 4 Considerable study has been concerned with the stereoselective conversion of intermediate II to III. 5 Employment of suitably introduced exogenous directing groups enabled reduction to proceed in almost quantitative yield and with a high degree of stereoselectivity; the desired 15S epimer (ill) was obtained in up to 92$ 6 yield upon treatment of II with bulky borohydrides at -I3O . OH I, X- 0} R, = CHOj R2 = Ac t \v»^ CH2CH2 0H II, X = 0} R = CH^CH-C-nCsHni R = Ac ^ x 2

'0 x Ill, X == 0; Rx = CH^CH-C^-nCsHni R 2 =

0R2 H 'OH - IV, X H,0CH 3 } R x = CHO; R 2 = H

Corey has recently modified his general route to permit direct stereo- controlled synthesis of PGA's 7 and PGC's.8 This pathway has become quite valuable since PGA's may now be transformed into PGE's (and thus PGF's) through stereoselective epoxidation to the 10, 11 a-oxide with gk% stereo- 9 selectivity by use of a remotely placed exogenous directing group. Numerous researchers have attempted to establish alternate routes to Corey intermediates, however, in most instances reactions are not as selective and/or yields are not as favorable. 10 One notable exception is the exceedingly novel route devised by Woodward, et. al. 11 Hydroxyaldehyde IV was constructed stereoselectively from cis-cyclohexane-l,3,5-triol in either racemic or optically active form and was converted to PGF2a in a comparable number of steps and yield to that of Corey's method. A completely different route to prostanoids has been developed by J. Fried and coworkers. 12 A novel stereospecific epoxide ring opening of

intermediate V with ; an optically active alane afforded diastereomeric products which were separated chromatographically and converted to optically active prostaglandins. Employment of this method of resolution has been 13 used in other recent prostanoid syntheses. Recently, Fried' s intermediate has been made available from cyclopentadiene in optically active form without resorting to chemical resolution, 14 thus, this route may become commercially feasable. , "155-

Other novel syntheses have also "been devised, however, most are less general and do not appear to be as efficient as the above routes. 13 * 15 The multitude of synthetic paths available to prostanoids and the advancements in study of physiological activity and metabolic fate of prostaglandins has led to generation of a variety of synthetic analogs, 16 A some of which are quite potent and demonstrate a higher degree of specificity in their physiological effects than the natural compounds. 17 Continuing efforts in synthesis of such compounds may result in synthetic prostanoids for the treatment of hypertension, gastric disorders, and infertility as well as agents for labor induction, therapeutic abortions, and fertility control. BIBLIOGRAPHY

1. For a recent review see: J. S. Bindra and R. Bindra, Progress Drug Res., IT, 410 (1975). 2. For reviews see: (a) P. H. Bentley, Chem. Soc. Rev. (London), 2, 29 (1973); (b) N. M. Weinshenker and N. H. Andersen in "The Prostaglandins," Vol. 1, P. W. Ramwell Ed., Plenium Press, New York, N. Y., 1973, Ch. 2j (c) C. R. Frihart, University of Illinois, Organic Seminar, Part 1, 1970-71, p. 12. 3. E. J. Corey, H. Shirahama, H. Yamamoto, S. Terashima, A. Venkateswarlu and T. K. Schaaf, J. Amer. Chem. Soc, 93, 1^90 (1971), and references therein. k. (a) E. J. Corey, Koelliker, V. and J. Neuffer, ibid . , 93, 1489 (1971); (b) E. J. Corey and C. U. Kim, J. Org. Chem., 3B~Tl233^1973 ) i (c) W. M. Weinshenker and R. Stephenson, ibid . , 37, 37^1 (1972). 5. E. J. Corey, S. M. Albonico, U. Koelliker, T~K. Schaaf, and R. K. Varma, J. Amer. Chem. Soc, 93, 1^91 (1971). 6. E. J. Corey, K. B. Becker and R. K. Varma, ibid., 9k, 86l6 (1972). 7. (a) E. J. Corey and G. Moinet, ibid., 95, 6F5I (1973); (b) E. J. Corey and J. Mann, ibid ., 95, 6832 (19737. 8. (a) E. J. Corey and G. Moinet, ibid ., 95, 7185 (1973); (b) E. J. Corey g" and P. A. Grieco, Tetrahedron Lett., 107 (1972). 9. E. J. Corey and H. E. Ensley, J. Org. Chem., 38, 3187 (1973).

10. (a) J. S. Bindra, A. Grodski, T. K. Schaaf and E. J. Corey, ibid , 95, 7522 (1973); (b) R. C. Kelly, U. VanRheenen, I. Schletter and M. D. Pillai, ibid., _95, 27^6 (1973); R. Peel and J. K. Sutherland, J. Chem. Soc, Chem. Commun., 151 (197*0. 11. R. B. Woodward, J. Gosteli, I. Ernest, R. J. Friary, G. Nestler, H. Raman, R. Sitrin, Ch. Suter, and J. K. Whitesell, J, Amer. Chem. Soc, 95, 685U (1973).

12. (a) J. Fried, C. H. Lin, J. C. Sih, P. Dalven, and G. F. Cooper, ibid . 9k, 43U2 (1972); (b) J. Fried, J. C. Sih, C. H. Lin, and P. Dalven, ibid ., 95, 43^3 (1972); (c) J. Fried and J. C. Sih, Tetrahedron Lett., 3899T1973). 13. C. J. Sih, P. Price, R. Sood, R. G. Salomon, G. Peruzzotti, and M. Casey, J. Amer. Chem. Soc, 9k, 36^3 (1972).

Ik. J. J. Partridge, N. K. Chadka and M. R. Uskokovic, ibid. , 95, 7171 (1973). 15. (a) N. Finch, L. Vecchia, J. J. Fitt, R. Stephani and I. viattas, J. Org. Chem., 38, 4412 (l973)j M E « J* Corey and B. B. Snider, Tetrahedron Lett., 3091 (1973); (c) E. J. Corey, G. W. J. Fleet and M. Kato, ibid., 3963 (1973). 16. (a) M. P. L. Cahton, E. C. J. Coffee and G- L. Watkins, ibid., 5°5, 773

Chem., \ (1972)} (b) M. Hayashi and T. Tanouchi, J. Org. \ (1973); (c) P. Rosen, S. Kwoh, J. L. Jernow, G. W. Holland, and F. Kienzel, ibid ., 38, 3^-0 (1973). 17. Ta) E. W. Yankee and G. L. Bundy, J. Amer. Chem. Soc, 9k, 3&51 (1972); (b) E. J. Corey and H. S. Sachdev, ibid., 95, 8483 (19737; (c) A. P. Labhsetwar, Nature, 238, 400 (1972)>"Td) J. F. Bagli, T. Bogri and % S. N. Sehgal, Tetrahedron Lett., 3329 (1973); (e) J. Fried and C. H. Lin, J. Med. Chem., 16, 429 (1973)- - 1 56 - EFFECTS OF MOLECULAR COMPLEXING ON CHEMICAL REACTIONS

Reported by Gautam Desiraju May 2, 197^ Although charge transfer and other, molecular complexes are very well 17 known and the phenomenon has been extensively reviewed, the effects of such complexing on the rates of reactions in solution have been studied in detail only recently. An early indication of these effects was the anomalous- 1 ''2 ly low acidity of picric acid in benzene, measured by Br^nsted and Bell. Yet these workers suggested other explanations for these effects. In general, if a reactant has donor (or acceptor) properties, addition of a non- reacting acceptor (or donor) will increase the 'reaction rate if the donor (acceptor) properties of the reactant are increased in the transition state. Conversely, a decrease in the reaction rate may be observed. Reactions, where the transition states are appreciably more electron deficient or electron rich than the reactant s, should, in theory be highly sensitive to complexation effects, provided that the other factors governing complex formation are favorable (e.g., steric factors). Here it is attempted to discuss only some of the types of reactions which are affected by the addition of non- reacting donors and acceptors* ALIPHATIC NUCLEOPKILIC SUBSTITUTION 3 Following a preliminary study by Smith and Leffler, Colter and co-workers carried out the first systematic investigation of these reactions in their work 4 e>>7> xo ' 12 on the relative rates of acetolysis of substituted $- fluorenyl tosylates 1. '

la R 1=R 3=N02 ; R 2=H lb R 1 =N0 2 ; R2=R 3=H Ts = -S02 -/0)-CH lc R2=N02 ; R 1 -R 3=H Id R 3=N02 ; R 1 =R2=H

The rate of acetolysis of l_a is increased by the addition of aromatic donors (hexamethylbenzene, phenanthrene, anthracene). Typically, first-order rate constants are five to ten times as great as those for the uncomplexed substrate. The following four- way reaction Scheme was postulated:

j^ ROTs d: ROTs.D -> 9- fluorenyl aromatics

K D = donor AcOH AcOH u V A + ROAc D ^T" ROAc.D +TsOH TsOH

The values ^of Km determined spectroscopically and kinetically are in agreement to within 10$. The possibility of formation of q- complexes in this reaction cannot be excluded since 9- fluorenyl aromatics are formed as by-products 5 with ^ the better donors (e.g., phenanthrene). The kinetic data can be analyzed to give the activation parameters for the overall reaction. The apparent entropy and enthalpy of activation are found to decrease with increasing donor concentrations. The entropy of activation for the acetolysis of the 1:1 complex (AS =-^ e.u./mole) is greater than that for the uncomplexed substrate (AS U=-12 e.u./mole); the rate enhancement is thus seen as an entropy effect. Acetolyses of the dinitro fluorenyl tosylates, lb, lc and Id give similar 8 kinetic data. The rate enhancements (measured as k /k~~T ave less (k /k =11 1 \ C* U C 11 to 14); than the similar enhancements observed for the acetolysis of la - - 157 -

(k c/ku=21 to 27). This would seem to indicate that the trinitro. transition state forms a stabler complex than the dinitro transition states. A semi- quantitative correlation has been made between the catalytic effective- ness of a donor and the energy of its highest occupied molecular orbital. :j It was found that increased reaction rates are paralleled by better donor properties (as measured by the energy of the highest occupied MO). The stereochemical aspects of these reactions have also been studied. 5 Acetolysis of the optically active 2,4,7-trinitro-9-fluorenyl(+)camphorsulphonate in the presence of donors led to a decreased optical purity in the corresponding 9- fluorenyl acetate, as compared with the optical purity of acetate formed in the absence of donor. The effect of pressure on the 'acetolysis of la in the presence and absence of donors has been studied briefly. -11 Rates of reaction at 1000 atm. are about four times greater than the corresponding rates at 1 atm. Volumes of activation have been estimated for the acetolyses of complexed and uncomplexed tosylates. RACEMIZATION OF SUBSTITUTED BIPHENYLC

Racemization of an unsymmetrically substituted biphenyl is effected by rotation about the central C- C bond. The barrier to rotation is largely steric in nature and the rate of racemization depends mainly on the potential energy barrier between the enantiomeric forms. Graybill and Leffler 13 observed that the racemization of 2,2'-dimethoxy- 6,6'-dicarbomethoxybiphenyl, 2, was accelerated by the addition of trinitro- toluene. Colter and Clemens found that the rate of racemization of (+)- 9,10-dihydro-3,^; 5,6-dibenzophenanthrene, 3, and of (+)-l,l ? binaphthyl, k, are increased by the addition of acceptors such as trinitrobenzene, picryl chloride or 2, 4,7-trinitrofluorenone, the largest effect being a twofold 14 increase observed with the last acceptor. Recently, it was found that activated charcoal increases the rate of racemization of k in acetone, by a factor of approximately 100. 1 ~ INTERCONVERSION OF GEOMETRICAL ISOMERS % 16 Von Gustorf and Leitich, found that the two isomeric allotfcimenes, 5a and 5b are interconverted in the presence of maleic anhydride at 160 . It was postulated that the interconversion occurred through cr complexes of maleic anhydride and the ofe fin, this complex being generated from an initially formed tr- complex. BIBLIOGRAPHY

v 1. R. P. Bell, Proc Roy. Soc, (London), Alk ) , 377 (193*0- 2. J. N. Br/Snsted and R. P. Bell, J. Amer. Chem. Soc, _53, 2*ff8 (1931). 3. B. B. Smith and J. E. Leffler, ibid ., 77, 2509 (1955T- k. A. K. Colter and S. S. Wang, ibid ., 85, 115 (1962). 4l, 5. K. Okamoto, I. Nitta and H. Shingu, Bull. Chem. Soc, Japan, 1*03 (1968). 6. A. K. Colter and S. S. Wang, J. Org. Chem., 27, 1517 (1962). 7. F. F. Guzik and A. K. Colter, Can. J. Chem., 51, l*+*hL (1965)- 8. A. K. Colter, F. F. Guzik and S. H. Hui, J. Amer. Chem. Soc, 88, 575 ^ (1966). 9. A. K. Colter and S. H. Hui, J. Org. Chem., 33, 1935 (1968). 10. A. K. Colter, S. S. Wang, G. H. Megerle and P. S. Ossip, J. Amer. Chem. Soc, 8?, 3106, (196IO. 11. R. K. Williams, J. J. Loveday and A. K. Colter, Can. J. Chem., 50, 1303 (1972). 12. A. K. Colter and M. R. J. Dack, "Molecular complexes," ed. R. Foster (1973). 13. B. M. Graybill and J. E. Leffler, J. Phys. Chem., 63, l*+6l (1959)- Ik. A. K. Colter and L. M. Clemens, J. Amer. Chem. Soc, 87, 8*4-7 (1965)-

, Chem. Soc, 6477 71973)- 15. R. E. Pincock, et. al. J. Amer. £5, f 16. K. E. von Gustorf and J. Leitich, Tetrahedron Letters, 4689, (I968). 1 17. G. Briegleb, "Elektronen Donator- Acceptor Komplexe," Springer, Berlin (1961). -158-

COPPER (i)-CATALYZED PHOTOCYCLOADDITION REACTIONS

Reported by Larry D, Martin May 6, 197^ INTRODUCTION The first photoisomerization of a Cu(l)-olefin complex, first character- 1 2 3 ized in 1963, ? was reported in I963. The Cu( I ) -catalyzed photocyclo- additions are of interest since in many cases they yield different stereo- chemistry than the triplet sensitized photocycloadditions. 4 Although the 5 6 7 exact role of the Cu(l) as the catalyst is not known, it is known ? that the more strained the olefin, the more strongly it complexes with metals. 8 9 Kochi ? has found that changing the anion associated with the Cu(l) can have large effects upon product yields. Photocycloaddition of cis,_cis-l,5-cyclooctadiene(l) Haight, et. al.i found that the octadiehe" 1 yielded a 1:1 complex with 2 CuCl which upon X-ray analysis was found to have structure 2. Complexes 10 with a 2:1 ratio of diene 1 to Cu(l), ^, have been obtained using Cu(l)BF4

CI /,. JSL >C1 ^:

^ e e X® BF 4 , OTf a

and Cu(l)trifluoromethanesulfonate (triflate, OTf). 11 The 2:1 complex can form with BF 4 and "OTf as the anions because they are poor ligands, but CI® coordinates strongly with Cu(l). 12 The triplet sensitized (benzene as sensitizer) photolysis of diene l3 yields only a polymer. However, the photolysis of complexes 2 or Jb yield h as the major product with the yield 13 14 13 ? from 2_, being solvent dependent. In ether a 2$% yield of tricyclic 14 compound ^ is obtained and in pentane (Equation l) a ^0% yield of ^ from the complex 2 is obtained. Scheme I shows the pathways considered by Srinivasan for the photolysis of diene 1 in ether saturated with CuCl.

SCHEME I

+ CuCl complex (2?)

complex + hv + CuCl

/\ X~\ + hv

CuCl CuCl -> h + CuCl -159-

For path A, the Cu(l)-diene complex absorbs the irradiation and then forms

the product, ji,. In path B the diene is activated by the light and then reacts with the Cu(l) to yield the product, ^. From the rate of product formation, the intensity of the irradiation and absorbance of complex 2 and diene 1, Srinivasan calculated quantum yields for paths A and B. For path A the calculated quantum yield is 10 and for path B it is 0.10. Therefore, Srinivasan reasons that path A is not the pathway of the reaction. Equation (l) shows the results that Whitesides obtained when he photo-

lyzed complex g in pentane. Since the trans , trans - diene 6 (The conformation

hv + pentane O 6* o or i 6b it 6a 62$ I9f

of is not known; however, on the basis its photolysis £, of product, J+,, it has been assumed to be 6b.) yields the tricyclic product h^ upon photolysis in the absence of CuCl but not upon thermolysis, and diene 2 neither the nor £, yield j^ i*1 either photolysis in the absence of CuCl or thermolysis, Whitesides proposes that the diene X, isomerizes to the trans , trans -diene 6 by photolysis in the presence of Cu(l), and then is isomerized to by a 2 + 2 photo- £, ^ [TT S TT S ] cycloaddition reaction as outlined in Scheme II. The Cu(l) catalyzed photo- SCHEME II

hv pentane v_y i

.^ + CuCl CuCl T hv

% ' CuCl hv A hv 5 • CuCl v \ + CuCl

& • CuCl

Jn» hv CuCl ~ 6 + CuCl 15 " isomerization of cis -olefins to trans -olefins is known. 17 Baldwin and Greeley18 propose that the formation of 4 from 2 can come from the radical 7 as shown in Scheme III. Baldwin and Greeley irradiated / \ hv

the mono-deuterated compound ga in ether and found that both the product U and the recovered diene ga had lost deuterium. However, Haller and Srinivasan* 9 .

-i6o-

repeated Baldwin and Greeley's work with the per-deuterated compound 9b and found no loss of deuterium. However, since the product U is only 28%of the converted diene and no label distribution study was made on the ramaining 72$ of the products no proof or disproof of the radical pathway is conclusively made

D D DD

9a

although Srinivasan's results do make the radical pathway appear less likely. Therefore, the mechanism for the isomerization, whether radical or concerted, is still uncertain.

Dimerization of Norbornene (10)

) The Cu( I -catalyzed photodimerization of norbornene ( 10 ) gives different stereochemical results from the triplet sensitized photodimerization as shown in Equations (2) and (3). In the photosensitized reaction a 27>% yield of dimers

hv (2) 0t0CH 3 or 10 12:88

hv (5) 11 + 12 CuX 10 >9T:3

X=cr%e BrD aOTf

11 and 12 is obtained with acetophenone and no change is noted in the ratio of the dimers 11 and 12 when several different sensitizers are used. The yields of % from the Cu(I) -catalyzed reactions are dependant upon the counter- ^° 5 8 ion employed. The yields of h when Cl" , Br and ^OTf are the counter- 5 ions are Ik. 5%, ^Q.k% and 88%respectively. Trecker concluded that the inverse of the quantum yield, $, has a linear relationship to the inverse 8 of the concentration of norbornene squared j however, when Kochi used a concentration range approximately 10-fold larger he discovered that there is 11 a linear relationship between l/i and l/[norbornene ] (See Figure l). Kochi discovered that although only a 1:1 complex of norbornene to Cu(l)0Tf can be isolated there is' more than one complex in solution according to the nmr. If only one complex with rapidly exchanging ligands is in solution then in excess norbornene the difference between the 6 value for the free and complexed norbornene should follow the linear relationship shown in Equation (k) A6 = +6 - 11 X c (& f C ), X c [Cu]/[norbornene]total (1+)

6-p = 6 of free norbornene 5 C = 6 of complexed norbornene -161-

Copper (It TnfUi* - ci*fm

8 Figure l Figure 211

As shown in Figure (2) a linear plot is not the result j therefore, more than one complex exists in solution. The reaction sequence shown in Scheme TV would have a quantum yield as defined in Equation (5). If k1 «k3 then l/$ has a linear dependance upon l/fnorbornene ] as shown in Equation (6). SCHEME IV8

NB = dimer =

« K Cu-NB + KB ^ Cu(NB)

CuNB + hv A * CuNB*

Cu(NB) 2 +hv &> Cu(NB), CuNB* + NB _^L-> dimer + Cu

Cu(NB) 2* i-^dimer + Cu CuNB* -i-> CuNB

Cu(NB) 2* -^->Cu(NB) 2

kx TNB ] \ -JA {4k [NB ] (5)' k [NB ]+k. x k2+k 4 ) K[NB>1

k2+k 4 1 l/$ (6)' k2

The quantum yield data indicates that both of the reacting norbornenes must be bonded to the Cu(l) in order to achieve the reaction. This mechanism can also explain the stereochemistry of the reaction which yields almost exclusively the exo- anti - exo dimer 11$ whereas the triplet sensitized reaction yields predominantly the endo- anti - exo dimer 12 as shown in Scheme V. In norbornene the Cu(l) complexes from the exo side; therefore, in the activated 2:1 complex the two norbornenes have their exo sides toward each other and this arrangement would yield the exo-anti-exo dimer 11. -162- SCHEME V

+#:och.

Dimerization of dicyclopentadienes

Exo-dlcyclopentadiene , J^, reacts similarly to norbornene in both its Cu( I ) -catalyzed and sensitized photo-reactions. Equations (7) and (8) illustrate the reactions observed. Endo-dicyclopentadiene, 14, is not so

hv (7)' CuBr, Et2

(8)

h* (9) CuOTf V I) Wo 15

h v 15 sens (10) j

-163-

well behaved. When CuBr is used as the catalyst no reaction occurs 5 however •when CuOTf is used the reaction goes as shown in Equation (9). The photo- sensitized reaction of endo-dicyclopentadiene, Ik, is shown in Equation (lO), The difference between the products of the Cu(l7^catalyzed and sensitized photolyses of ^ can be rationalized by analogy to the norbornene reaction, in which both olefins had to be bonded to the Cu(l) in order to have a reaction. If an analogous reaction is occuring in this case, then the formation of 1£ could indicate a small portion of the chelate compound l£ in the solution.

Cu OTf

16

Dimerization of cyclopentene

The Cu( I ) -catalyzed photodimerization of cyclopentene is a very clean reaction in comparison to the photosensitized reaction. The Cu( I ) -catalyzed - - - reaction yields the cis-anti cis , 1^, and cis syn cis , 18, dimers in a 10:1 19 ratio as shown in Equation (ll). Dimer 1£ is the only compound from the

photosensitized reaction which may not come from a radical intermediate, although a radical pathway for its formation can be formulated also. Dimerization of cyclohexene and cycloheptene Both cyclohexene and cycloheptene give good yields of dimeric products when photolyzed in the presence of CuOTf as opposed to the triplet sensitized photodimerization of cyclohexene which gives very low yields of dimer as 9 shown in Equations (13)-(15), The major difference between dimers 1%, ££,, and 2^ from the other cyclobutane products observed earlier is that the cyclobutane ring junction is no longer cis , which is the thermodynamically more stable form. 22 Although photolysis in the presence of ketones 23 24 sensitizers or HgBr2 is known to epimerize tertiary carbons, the photo- 9 lysis of a solution of dimer lg, and CuOTf does not result in epimerization. Therefore, the trans -fused rings do not come from the epimerization of cis - fused rings. The formation of 1% could be obtained by a [rr2 s+n2a ] cycloaddition % involving trans -cyclohexene and cis -cyclohexene as dhe antarafacial reactant. Such a reaction could result in two isomers, the d,l dimer 1% and the me so -16U-

hv + (13) CuOTf \

2k% hv >s-s-00 oo oo (1*0

CH3 0CCH2(5CH3 1# 1.5# 0.6$ 0.9% 0.9%

hv (15) CuOTf

form 2£. However, steric hindrance could be responsible for the lack of formation of 2g,. The same mechanism involving cycloheptene can explain the formation of 2^. There is evidence for the generation of trans -cyclohexene as a reaction intermediate, 16,25,2S and trans -cycloheptene has been trapped by a Diels-Alder reaction. 2 '' The isomerization of cis-cyclooctene to trans - cyclooctene by photolysis with Cu(l) catalyst also lends support to this mechanism. 17 Cycloaddition of norbornadiene

Norbornadiene-CuCl has been shown by X-ray analysis 28 to be a mono- dentate tetramer, 2^, in its crystalline form. The photolysis of 2^, in ether 5 yields 2j+ as the only volatile component. The same product is obtained from the photolysis of norboradiene in ether. 29

CI - Cu hv | (16) / ^Cu —„. CI ^ C- EtoO C!L Cu - Cu 2k /

23 -165-

Reactions Involving Cyclooctene While cyclooctene does not dimerize when photolyzed in the presence of 8 Cu(l), Kochi and Salomon have found that when norbornene is present in the same amount as Cu(l) a crossed- eye loadduct is formed according to Equation (rffr

hv CuOTf + (IT)

ko% However, since the analogous reaction between cyclooctene and 1-hexene does not proceed to form a dimer but yields only starting material and some trans - 9 cyclooctene, it appears that at least one member of a reacting pair in a Cu( I) -catalyzed cycloaddition reaction must be more strained than an acyclic olefin. Conclusion The use of Cu(l) as a catalyst in photocycloaddition reactions may be of some synthetic utility since it can afford stereochemical dimeric products that are not achievable in good yields by photosensitization. The use of CuOTf is preferred over the use of Cu(l) halides since it gives better yields and easier reactions. Although the exact role of the Cu(l) in the reaction is not known, there is evidence in the dimerization of norbornene and dicyclopentadiene that both of the reacting olefin units must be bonded to the metal before they will react. References

1. H. L. Haight, J. R. Doyle, N. C. Baenziger and G. F. Richards, Inorg.

Chem . , 2, 1301 (1965).

2. J. H. van den Hende and W. C. Baird, Jr., J. Amer. Chem. Soc , 85, 1009 (1963).

3. R. Srinivasan, ibid . , 85, 30U8 (1963).

k. D. J. Trecker, Org. Photochem . , 2, 63 (1969).

5. D. J. Trecker, R. S. Foote, J. P. Henry and J. E. McKeon, J. Amer. Chem . Soc, 88, 3021 (1966).

6. J. G. Traynham and J. R. Olechowski, ibid . , 8l, 5 71 (1959).

7. M. A. Muhs and F. T. Weiss, ibid ., 847TE97 IJ.962) .

8. R. G. Salomon and J. K. Kochi, ibid ., 96, 1137 ( 197*0 •

9. R. G. Salomon, K. Folting, VI. E. Streib and J. K. Kochi, ibid ., 96 , 11^5 (197*0. 10. S. E. Manahan, Inorg. Nucl. Chem. Lett ., 3, 383 (1967).

11. R. G. Salomon and J. K. Kochi, J. Amer. Chem. Soc , 95, I889 (1973). 12. S. E. Manhan and R. T. Iwamoto, Inorg. Chem ., ^,"1^09 (1965).

13. R. Srinivasan, J. Amer. Chem. Soc , 86 , 3318 (196U). Ik. G. M. Whitesides, G. L. Goe, and A. C. Cope, ibid ., 91, 2608 (I969). 15. H. Nozaki, Y. Nisikawa, Y. Kamatani, and R. Noyari, Tetrahedron Lett ., 2161 (1965).

16. R. S. H. Liu, J. Amer. Chem. Soc , 89, 112 (1967). 17. J. A. Deyrup and M. Bethowski, J. Org. Chem ., 37, 356l (1972).

18. J. E. Baldwin and R. H. Greeley, J. Amer. Chem. Soc , 87, kjlk (1965). 19. I. Haller and R. Srinivasan, ibid., 8b, 508^ (1966). -166-

20. D. R. Arnold, D. J. Trecker, and E. B. Whipple, ibid ., 87, 2596 (1965). 21. H. D. Scharf and F. Korte, Chem. Ber ., 97 2^25 (1964).

22. E. J. Corey, R. B. Mitra and H. Uda, J. Amer. Shem. Soc , 86 , 485 (1964). 25. N. C. Yang and D-D. H. Yong, ibid ., 80, 2913 (1958). 24. M. Gorodetsky and Y. Mazur, ibid ., 90, 6540 (1968). 25. J. A. Marshall, Accounts Chem. Res ., 2, 33 (1969).

26. P. J. Kropp, J. Amer. Chem. Soc , 88 "4091 (1966). 27. E. J. Corey, F. A. Carey and R. A. E. Winter, ibid ., 87, 934 (1965). 28. N. C. Baenziger, H. L. Haight and J. R. Doyle, Inorg. Chem ., 3, 1535 (1964) 29. W. G. Daubea and R. L. Cargill, Tetrahedron, 15,' 197 (1961).

3 -167- SOLVOLYSIS MECHANISMS OF VINYL TRIFLATES

Reported by Leonard J* Adzima May 9, 197^ Solvolysis reactions involving vinyl cation intermediates have been known for a number of years. 1 5 But most of these studies deal with generating 6 " 9 10 11 vinyl cations -which are stabilized by aromatic or cyclopropyl * rings or by neighboring double bonds. 12 The recent preparation of alkyl vinyl triflates 13 have led to significant progress in elucidating solvolysis mechanisms which may involve the formation of simple alkyl vinyl cations. 13 In 1969> Stang and Summerville found that the solvolysis of vinyl triflate 1 in 80$ aqueous ethanol at 76 gave 98+3$' dimethylacetylene . The cis isomer 2 at 100° gave 58$ dimethylacetylene, y$% 2-butanone and 9$ methylallene . The formation of 2-butanone from 2 indicates that solvent CH OTf IL ^OTf H^ OTf 3 ^ \„ S \ + c=c c=c C=C-CH H CH3 CH3 OH CH 3

capture of an intermediate vinyl cation ^ may be occurring. This finding coupled with evidence based on relative rates and deuterium isotope effects suggests the first example of the formation of a simple alkyl vinyl cation. Molecular orbital calculations indicate that vinyl cations prefer a linear structure and can be considered to have an sp-hybridized carbon and a vacant 14 15 p orbital. ' Evidence for a linear structure is provided by comparing the relative solvolysis rates of cyclic and acyclic vinyl triflates. 16 Since cyclic vinyl cations are rigidly held in less favored bent geometries, cyclic vinyl triflates should solvolize slower than similar acyclic systems. Experi- mentally it is found that 2 solvolizes about 3300 times faster than cyclic triflate J£, as would be expected for a process leading to a bent vinyl cation. A more complete study of the effects of alkyl substituents, solvents and added nucleophiles on the solvolysis of vinyl triflates has been made by Schleyer and co-workers. 17 They found that the effect of changes in solvent nucleophilicity at constant ionizing power is significant especially with substrates with trans fJ-hydrogen as opposed to those with trans P- alkyl groups. However, in contrast p* is large (-^.8) for alkyl substitution indicating high carbonium ion character. There is also no significant effect on rates or products upon adding strong nucleophiles such as Br. These somewhat con- tradictory results lead the authors to believe that these solvolysis reactions are proceeding through ion-pair pathways with no direct evidence for Sn2 or E2 mechanisms. Ion pair mechanisms have been supported by recent work involving the stereochemistry of various alkyl substituted vinyl triflates. 18 " 21 Bergman and co-workers18 have demonstrated that reactions of simple alkyl substituted vinyl substrates proceed with a significant amount of inversion of configura- tion at the vinyl center. The solvolysis of 3-methyl-2-heptenyl triflates (z)-

£, and (e)-5 in dry trifluoroethanol gave vinyl trifluoroethyl ethers (z)-6, and (E)-6, and 1-methyl-l-n-butylallene 7. The product data in Table 1 show that (z)-£ and (E)-£ give rise to different ratios of products. Therefore the mechanism of this reaction cannot be completely a free dissociation to a carbonium ion nor can it be completely a direct Sn2 displacement. The pro- posed mechanism involves the intervention of ion pairs (Scheme l) . The triflate leaving group in the ion pair shields the molecule somewhat from attack at that side, thus giving rise to the observed different ratios of products. 2 -16b- TABLE 1 Substrate r Products, % Ratio, (z)-6 (e)-6 7 (E)-6/(z)-6 (z)-5. 15.2 70.6 14. k.6 <">-£ 23.9 58.6 17.5 2.k t SCHEME I

R OTf R /CH 3 'C=C' C=C-CH 3 C=C CH. *CH. CH. CH3 OCH2CF3 (z)-5. (E)-6 i K + C=C-CH 3 CH.

R K R OCHpCFo c=c C=C-CH< ^c=c ^ ^ CH3 ^OTf CH3 + CH3 CF3 "OTf (E)-5 = (z)-6 R CH2CIi2CH2CH 3 References

1. P. J. Stang, Progr. Fhys. Org. Chem ., 10, 205 (1972).

2. G. Modena and U. Tonellato, Advan. Fhys. Org. Chem ., £_, 185 (1971). Z. Rappoport, T. 3. Bassler and M. Hanack, J. Amer. Chem. Soc , 2£» ^5 (1970). !«.. M. Hanack, Accounts Chem. Res ., 3, 209 (1972). 5. H. G. Richey and J. M. Richey in "Carbonium Ions," Vol. II, G. A. Olah and P. V. R. Schleyer, Ed., Wiley, New York, N. Y. 1970, p. 899.

6. C. A. Grob and G, Cseh, Helv. Chim. Acta. , kjj I9U (I96U).

7. G. Cappozzi, G. Melloni, G. Modena and M. Piscitelli, Tetrahedron Letters , 4039 (1968).

8. L. L. Miller and D. A. Kaufman, J. Amer. Chem. Soc , £0, 7282 (1968). 9. Z. Rappoport and J. Kaspi, ibid .*7~96, 586 (1974). 10. S. A. Sherrod and R. G. Bergman, ibid ., <£L, 2115 (1969).

11. M. Hanack and T. Bassler, ibid . , 91, 2117 (1969).

12. C. A. Grob and R. Spaar, Tetrahedron Letters , 1439 (1969). 13. P. Stang J. and R. Summerville, J. Amer. Chem. Soc , £L> ^600 (1969). Ik. R. Sustmann, J. E. Williams, M. J. S. Dewar, L. C. Allen and P. V. R.

Schleyer, ibid . , 91; 5350 (1969)? and references cited therein. 15. W. A. Lathan, W. J. Hebre and J. A. Pople, ibid., 2£, 808 (1971). 16. W. D. Pfeifer, C. A. Bahn, P. V. R. Schleyer, J. Bocher, C. E. Harding, K. Hummel, M. Hanack and P. J. Stang., ibid ., %L> 1516 (1971). 17. R. Summerville, C. A. Senkler, P. V. R. Schleyer, T. E. Dueber and

P. J. Stang, ibid . , 36, 1100 (1974). 18. T. C. Clark, D. R. Kelsey and R. Bergman, ibid ., 9jb 3^26 (1972). 19. T. C. Clark, and R. Bergman, ibid ., £4, 361FTT1972T 20. R. Summerville and P. V. R. Schleyer, ibid ., 94, 3529 (1972). 21. R. Summerville, P. V. R. Schleyer, ibid ., 96, 1110 (197*0.

« ORGANIC SEMINAR ABSTRACTS

1974 - 75

SEMESTER II

THE LIBRARY OF THE

JUL 1 6 1975

AT URBANA-O' ' 1PAFGN

School of Chemical Sciences Department of Chemistry University of Illinois Urbana, Illinois

SEMINAR TOPICS

II Semester 1974 - 75 ) Singlet Oxygen Reactions with Sulfur and Nitrogen Compounds 73 Larry D. Martin

Direct Amination on Nitrogen 75 Jane Berlin

New Masked Acyl Carbanions In Synthesis 84 Brian Holmes

Transition Metal Homogeneous Catalysis: Recent Applications in Fiber, Plastic and Other Industries 94 Robert W. Mason

Reaction of 1-Azirines 103 David B. Reitz

Synthetic Enzymes: Application of Solid Phase Peptide Synthesis to the Preparation of Selectively Blocked Presursors 105 H. Mitchell Rubenstein

Aromaticity and the Bridged Annulenes 112 Lance A. Christell

Synthesis and Reactions of Aromatic K-Region Epoxides 120 Richard Anos

Photoaff inity Labeling with Carbonyl Compounds 122 Richard L. Neeley

Intramolecular Friedel-Crafts Reactions of Olefins 124 Karl E. Weigers

Cathodic Electro-Organic Reactions : Principles and Applications 126 Gary Nickel

Heterocycles in Synthesis - Annelation Reactions 135 Marvin Reich

Some Recent Developments in Cyclopropanone Chemistry 139 Kenneth Berger

The Synthesis and Chemistry of Thioketones 148 Robert J. McGorrin

- T) - SINGLET OXYGEN REACTIONS WITH SULFUR AND NITROGEN COMPOUNDS

Reported by Larry D. Martin January 20, 1975 INTRODUCTION

The possible role of singlet oxygen('02 ) in the photodynamic effect, the dye sensitized photolytic damage of living tissues, has initiated interest in its reaction. The generation of ' 2 and its reactions with 2 olefins has been reviewed by Foote 1 and Kearns. There are only two chemically important excited singlet states of oxygen, and of these the first excited state is the one generally involved in '02 reactions. The electronic configurations and energies above the ground state are shown in Figure 1. REACTIONS WITH SULFUR COMPOUNDS Singlet oxygen reacts with sulfoxides, 3 sulfides, 4 b disulfides, 7 1C> and thiophenes 11 ' 12 to yield sulfones, sulfoxides, thiolsulfinates and sulfines, respectively. Foote 4 has demonstrated that the oxidation of diethylsulfide goes through the zwitterionic intermediate J.. Intermediate X subsequently reacts with a mole of diethylsulfide to yield two moles of 6 diethylsulfoxide; the 2:l(sulfide: '0 2 ) ratio has been demonstrated. Zwitterionic intermediates similar to 1 have been proposed for the reaction 7 12 4 of '02 with the sulfur compounds mentioned above. Foote found that a

CH3CH2 0-0

22 Kcal S w CH3CH2 /(3 1-4-

Figure 1 mixture of diethyl and diphenylsulfides yielded a larger diphenylsulfoxide: diethylsulfoxide ratio from the reaction with '0 2 than would be expected from '0 the ratio of their individual reaction rates with 2 . This result indicates '0 that X i s less selective toward diphenyl sulfide than is 2 . A similar result was obtained for mixtures of diet hyldi sulfide and diphenyldi sulfide. l0 5 Diethylsulfide has been shown to be an effecient '0 2 quencher, also. In benzene the sulfide quenches '0 2 twenty times faster than it reacts; however, in methanol very little quenching takes place. Foote 5 has shown that X is not responsible for the quenching. Diethylsulfide could quench ' 2 by a 13 charge- transfer mechanism as is proposed for the quenching of '0 2 by amines. REACTIONS WITH AMINES The photosensitized reaction of oxygen with amines has been shown to give aminehydroperoxides. 6 ' 14 ' 15 in Other reactions observed the" photosensitized oxidation of amines are oxidation of the alkyl side chain 16 1 and dealkyla- tion. 20 22 Davidson 17 ' 22 has shown that many of the reactions may not be the '0 result of 2 , but of reactions between the amines and the sensitizer. In 23 only one case, that of triethylamine, has it been proven that '0 2 is the 13 ' 26 reactive intermediate. Amines, 24 especially 1, ^4-diazabicyclo- [2.2.21- 23 ' 24 octane (DABCO), have been found to be efficient '02 quenchers. )

- Jk - 13 As mentioned earlier, the quenching is thought to go through a charge- 13 transfer complex between amine and. ' 2 . The linear correlation of quenching rates to ionization potential supports this theory. CONCLUSION

The reaction of '02 with sulfur compounds has been shown to go through zwitterionic intermediates similar to 1. Sulfides also act as ' 2 quenchers in a manner that may be similar to that of amines. Amines undergo a variety of reactions with ' 2 or with sensitizer followed by reaction with triplet" oxygen. Many of these reactions are quite useful synthetically. Further studies are needed to determine the reactive species in these reactions. However, these investigations have given some insight into possible reactions that are involved in the photodynamic effect. BIBLIOGRAPHY

1. C. S. Foote, Accts. Chem. Res., 1, 10'+ (1968). 2. D. R. Kearns, Chem. Rev., 71, 395 0-97.1). 3. G. 0. Schenck, Chem. Ber., 96, 517 (1963). h. G. S. Foote and J. W. Peters, J. Amer. Chem. Soc, £3, 3795 (1971). 5. C. S. Foote, R. W. Denny, L. Weaver, Y. Chang, and J. Peters, Ann. N. Y. Acad. Sci., 171, 139 (1970 ).

6. K. Gollink, Advan. Phot o chem. , 6, 1 (1968).

7- J. A. Barltrop, P. M. Hayes and M. Calvin, J. Aner. Chem. Soc, 76 ,

h3k8 (195*0 •

8. R. W. Murray, R. D. Smetana, and E. Block, Tetrahedron Lett., 1971 , 299 (1971).

9- R. W. Murray and S. L. Jindal, Photochem. Photobiol. , l6, ikj (1972). 10. R. W. Murray and S. L. Jindal, J. Org. Chem., 37, 35l6~7l972).

11. C. N. Skold and R. H. Schlessinger, Tetrahedron Lett., 1970 , 791 (1970

12. H. H. Wasserman and W. Strehlow, ibid . , 795 (1970 ). 13. E. A. Ogryzlo and C. W. Tang, J. Amer. Chem. Soc, %2_, 503^ 0-970 ). Ik-. H. Gaffron, Chem. Ber., 60, 2229 (192Y). 15. G. 0. Schenck, Angew. Chem., 69, 579 (1957).

16. a) M. H. Fisch, J. C. Gramain and J. A. Oleson, Chem. Commun. , 1970 , 13

(1970). b) ibid . , 1971 , 663 (197H).

17. R- F. Bartholomew and R. S. Davidson, ibid . , 1970 , 117^ 0-970). 18. R. D. Stipanovic, C. R. Howell and A. A. Bell, J. Heterocyc Chem., £, 1^53 (1972). 1 19. N. Kulevsky, C-H Niu, and V. I. Stenberg, J. Org. Chem., 38, 115 )- (1973).

20. J. H. E. Lindner, H. J. Kuhn, and K. Gollnick, Tetrahedron Lett., 1972 , 1705 (1972).

21. K. Gollnick and J. H. E. Lindner, ibid . , 1973 , 1903 (1973). 23*1-2 22. a) R. F. Bartholomew and R. S. Davidson, J. Chem. Soc. (c), 1971 ,

(1971). b) ibid . , 23^7 0-971). Chem. 186 23. W. F. Smith, Jr., J. Amer. Soc, gjfc, (1972). 24. C. Ouannes and T. Wilson, J. Amer. Chem. Soc, 9£, 6527 (1968).

25. D. Bellus and H. Lind, J. Chem. Soc Chem. Commun., 1972 , 1199 (1972). 26. R. S. Atkinson, D. R. G. Brimage and R. S. Davidson, J. Chem. Soc.

Perkin I, 1973 , 9o0 (1973). - 75 - DIRECT AMINATION ON NITROGEN

Reported by Jane Berlin January 27, 1975 The first example of direct amination on nitrogen appeared in 1907 in Raschig's synthesis 1 of hydrazine from ammonia and chloramine, NH2CI. Since that time N- amination has been applied to a wide variety of compounds, and several additional] aminating reagents of the general formula NH2X have been developed, where X represents an electronegative atom or electron-withdraw- ing group. Leaving group, X, affects the stability of the reagents, their solubility properties, and most important their synthetic utility. It is the purpose of this seminar to compare the most important direct aminating reagents with respect to these properties, to briefly consider their applications, and to illustrate the utility of selected synthetic intermediates which may be produced.

N- AMINATION REAGENTS

There are two principle methods for formation of chloramine, NH2 C1; neither is satisfactory. Raschig 1 orginally prepared the reagent in aqueous solution by reaction of ammonia with sodium hypochlorite. However, even in

NH3 + NaOCl > NH2C1 + NaOH dilute solution the reagent slowly decomposes to nitrogen, ammonia, and hydro- chloric acid; in concentrated solution decomposition is accelerated.

3 NH2C1 >N2 + NH3 + 3 HC1 The alternative gas phase preparation, by reaction of chlorine and ammonia, results in a gaseous mixture of chloramine and ammonia which is bubbled through a solution of the substrate. 2 Reasonable yields of reagent are only obtained with a large excess of ammonia, which also increases the rate of undesirable, 3 N2 forming side reactions. 8 NH + Cl N + 6 NH4CI 3 ) 3 2 > 2( ) ( g ( g ) g ( s )

There are numerous examples of direct amination on nitrogen by chloramine. 4 Schiff bases may be aminated in ether to give diaziridines, which are cleaved by dilute aqueous acid to an aldehyde and alkylhydrazine.

H N

/ \ TT + RCHO RCH-NC4H9 + NH2 C1 > R- CH- NC 4H9 H2 C4H9-NHNH2

Triazanium salts are generated from a gaseous mixture of ammonia and chloramine with either 2-dialkylamino-l,3,2-dioxaphospholane or 1,1-dialkyl- hydrazine.s The proton magnetic resonance spectra of the triazanium chlorides support the symmetrical structure shown.

CH2-0^ + I PNR2 NH2 C1 CHo-0 ^ [H2NNR2NH2 ] COM

R2NNH2 + NH2 C1

The action of chloramine on ooximinoketones with the formation of odiazoketones, the Forster reaction, has been applied to the oximes of acetophenone, benzophenone, benzaldehyde, and formaldehyde. 6 , - 76 - o-Pyridone undergoes N-amination in the presence of aqueous NaOH in 28$ yield. 7

^ H I NH2

Dialkylamine azides are generated by treatment of N-chloramines with metal azide in methylene chloride at room temperature. 8

Me 2NCl + NaN 3 > Me 2N-N 3 + NaCl The reaction of chloramine with ammonia and amines to form hydrazine has received greatest attention. 4 A recent review by Kovacic, et.al . summarizes the reactions in Table I.

Table I 4

Reactions of Chloramine with Amines

NH2C1 + 2NII 3 >NH2NH2 + NH 4 C1

NH2 C1 + 2RM2 > RNHNHg + RNIIoCl

NH2 C1 + 2R2NH > R2]MFfe + R2NH2 C1 + NH2CI R3N > R r lW!I2 cP

NH2 C1 + 2NH 3 (aq) >N2H4 + NH 4C1

NH2 C1 + NH 3 (aq) + OiT > N2H4 + CI + H2

NIIoCl + RNH2 (aq) + OlP > RNIiNllo + Cl° + H2

NH2 C1 + R2ffll(aq) + OR > R2MII2 + CI + H2

7 An S-nr2 mechanism has been proposed for direct amination." Analogy may be drawn to methylation in which nucleophilic attack by the substrate on the reagent causes simultaneous loss of an electron- withdrawing group. In aqueous solutions the formation of hydrazine has been found to be first order in both chloramine and ammonia and independent of hydroxide ion concentra- tion from pH 10 to Ik. Substitution of alkyl groups for hydrogen in ammonia increases the rate constant, as would be expected for an Sj^2 reaction. Reactions of tertiary amines and chloramine also follow second- order kinetics in organic solvents. 10 There are two important limitations to these reactions. First, hydrazine may further react with chloramine to form N2 and ammonium chloride. The decomposition may proceed as fast or faster than hydrazine production. There- fore, a large excess of ammonia or amine must be used to obtain reasonable yields. 11 Addition of gelatin to the reaction mixture inhibits the decom- position. 12 Second, although chloramine reacts with tertiary amines to form

N2 H 4 + 2 NK2 C1 > N2 + 2 NH 4C1 the 1,1,1-tri substituted hydrazinium chlorides, it has not been possible to obtain N- amino compounds when the tertiary nitrogen is part of an aromatic ring, as in pyridine. 13 - 77 - 14 In 1919> Sommer, Schulz, and Nassau synthesized hydroxylamine- 0- sulfonic acid (HSA, l) from chloro sulfonic acid and hydroxylamine sulfate. They discovered HSA, H2N0S020H, reacts similarly to chloramine with ammonia and amines. Reactions are generally carried out by mixing a methanol, ethanol, or water solution of HSA with a similar solution of the amine. 15 In most cases products crystallize out of the reaction mixture on standing at room temperature. Although aqueous solutions of HSA are not extremely stable, its II HO- S- ONH2 hydrolysis to hydroxylamine is much slower than the 11 II decomposition of chloramine. Of greatest interest was the finding that heterocyclic aromatic amines can be aminated with HSA. Though reaction 1 will occur at room temperature if a mixture is allowed to stand several days, reaction mixtures are usually heated. 15 For example, 72% yield was obtained by reaction of a 3- fold excess of pyridine with HSA at 90° for 20 minutes. 16 Although Sommer and co-workers 14 stated that one could work with equivalent quantities of amine and HSA, the yields obtained under such conditions were non- 15 reproducible. Studies by Sisler, et. al ., showed that many tertiary amines reacted vigorously with solid HSA or its solution in MeOH, but in all cases, excess amine was required for reasonable yield. Similarly, Gtisl and Meuwsen11 were able to significantly increase the yield of hydrazine only by increasing the ratio of amine to HSA; changes in temperature, reaction time and base concentration were ineffective. (Figure i) The amination of uracil with HSA by Klotzer 17 ' 18 prompted Broom and 19 Robins to try to aminate purines with HSA. (Figure II) In the presence of base it should be possible to remove the proton adjacent to a keto group in the purine ring and introduce an N- amino group; they were able to directly aminate inosine, guanosine, and 2'-deoxyguanosine at position 1. The only

11 Figure I Figure II

1 1 1

i !

0) I d •H 60 ^*r<~\ N : u IW 1 ! i P T* >> ml\ I : i ! Si ! 1 H 40 —J /

•H >5 ; 1

1 1 Purine V/ 1 l c 1 20 1

Mole ratio n- butylamine : HSA

I 70°, mole ratio amine: HSA:K0H - X:l:2 II 0°, the same mole ratio III room temperature without K0H previously reported N- amino derivatives of purines, where the amino group was attached directly to a ring nitrogen, were 9- N- amino purines. In all cases the amino group had been introduced into the purine by means of ring closure of preformed pyrimidine intermediates or by ring closure of a 1- amino deriva- tive of imidazole. On the basis of its increased stability, ability to aminate heteroaromatic amines, and commercial availablity, hydroxylamine- 0- sulfonic acid is superior to s , 78 chloramine as an aminating agent. However, its insolubility in organic solvents such as methylene chloride, benzene, and acetone, severely limits 20 ' 21 its utility. Carpino and co- worker directed their attention toward the synthesis of other O-acyl and O-sulfonyl derivatives of hydroxylamine which might be soluble in organic solvents. Their synthetic method was designed to take advantage of the ease of cleavage of the carbo-t-butoxy group. Although successful, the method is impractical; the starting material, t^ butyl azido formate (2), though commercially available is rather costly. (Scheme i)

Scheme I

II 11 I' NH2° H RCC1 k r ^ N 3C- OC (Me ) 3 > HON- C- OC (Me ) 3 RCONH-C-OC(Me) 3 H (RS02C1 t

HC10.

II H2 II RC-ONH2 [RC-ONH2 . HCIO4] Zinner22 described a more economical preparation of O-aroyl and 0- carbalkoxyhydroxylamines from the readily accessible ethyl acetohydroxamate 23 (3). By modification of this method Tamura, et al . were able to prepare the desired 0-arylsulfonyl and O-nitrophenylhydroxylamines (jjr£) quite easily. (Scheme II) Scheme II

Me Me HC1° 4 Hg°> H0N=C -£^> R0N=C > [R0NH2«HC10 4 ] R0NH2 X0Et OEt 3

•Me N02

s0 -- SO; R = Me >O- O- Me

5.

.N02

2N

1 8 £

The reactivity of these 0- substituted hydroxylamine s was compared by their ability to aminate on nitrogen in various substrates such as tri-n- butylamine and pyridine, as well as on sulfur and phosphorus in diphenyl sulfide, diphenyl sulfoxide, and triphenylphosphine. The reactions were generally carried out in methylene chloride solution at room temperature. (Table II). On the basis of the amination results, together with relative - 79 -

2 ' Table II. Comparison of (

Starting material it 5 1 8 2 (n-Bu)3N 87 72 85 15

C5H5N 80 68 55 80 trace

Ph2S 90 79 61 87

Ph2 SO 65 ko <20 <30 PhaP 86 76 92 91

stability and solubility properties, O-mesitylenesulfonylhydroxylamine,

MSH (h) , is the most effective and generally applicable of these reagents. O-Mesitoylhydroxylamine (£), first synthesized and studied by Carpino, 21 is useful in the amination of compounds, such as sulfonamides, which can be easily converted to the corresponding anions by treatment with alkali. However, the extremely low yields resulting from reaction with simple amines prohibit its use in many cases. 0- (2,^Dinitrophenyl)-hydroxylamine, DNPA CD, first synthesized by Sheradsky24 by condensing t- butyl N-hydroxycarbamate with 2,^-dinitrochloro- benzene and treating with trifluoroacetic acid, offers no apparent synthetic advantage. Although more stable than some 0-sulfonyl derivatives and thus, somewhat easier to handle, DNPA reacts easily only with strong nucleophiles 25 such as anionic nitrogens. Although Tamura reports storing MSH for several weeks at 0°, preparations 26 27 by his exact procedure have been known to explode. In 197^-, Johnson and co-workers published a modified procedure which results in small MSH crystals. Caution is yet advised. The advantages of MSH are quite clear when comparison is made to reactions using chloramine or HSA. For example, HSA and chloramine are known to react with tertiary amines to give the corresponding hydrazinium salts, but in both cases, a large excess of the amine must be employed. In contrast, in a typical amination, a solution of MSH in methylene chloride is added to an equimolar, stirred, cooled solution of the amine in methylene chloride. 28 The mixture is allowed to stand at room temperature for 5 minutes, and the product is filtered off as a white precipitate. The technique is equally applicable to aliphatic and heteroaromatic amines. In addition, N- amino derivatives of cyano- , nitro- , carboxy-, and ethoxycarbonylpyridines, which cannot be synthesized using HSA, may be obtained in high yields with MSH. 2S Similarly, adenine, which fails to aminate with HSA, yields 65^ 1- aminoadeninium me sitylene sulfonate when treated with MSH in methanol. 30 The ability of MSH 23 ' 27 31 to effectively aminate both sulfur and phosphorus is also noteworthy. ' INTERACTIONS OF N-AMINATED NUCLEIC ACIDS

The ability to aminate directly a nucleic acid base suggests an interest- ing application of these reagents. The introduction of methyl groups into RNA produces profound alterations in nucleic acid structure. 32 The bulky, hydrophobic methyl group may give rise to structural changes due to steric or electronic factors or a combination of both. Introduction of an amine group at the same site would provide a substituent of comparable size. 19 The amine group is hydrophilic, however, and may act as a hydrogen bond donor or acceptor; its placement at selected sites could be important in studies of tertiary structure and hydrogen bonding interactions of nucleic acids. - 80 -

THE N- AMINO GROUP AS A SELECTIVE BLOCKING AGENT Recent research shows a tremendous increase in the quantity of methylated nucleotides in cancerous tissue tRNA when compared to normal tissue tRNA. 33 Specifically, 7- methyl- 8- oxoguanine CLQ), which may arise from the oxidation of 7-methylguanine, has been found in substantially elevated levels in the urine of leukemia patients. 34 In an attempt to synthesize these natural methylated purine nucleosides, 35 Rizkalla, Robins, and Broom found the N- amino function useful as a blocking agent for selective methylation. Methylation of guanosine under nonaqueous

basic conditions (anhydrous K2 C0 3 , DMSO or DMF) leads to 1- methylguano sine, whereas methylation under essentially neutral conditions (DMF, DMSO, or DMA) leads only to T-methylguanosine. These are general procedures in which the most basic site is alkylated; that is, if a proton is removed by base from a ring nitrogen in nonaqueous media, alkylation occurs at that site. In neutral media alkylation occurs at the most nucleophilic nonprotonated site. By analogy with guanosine it was believed that direct methylation of 8- benzyloxyguanosine in neutral medium would yield the 7- methyl derivative which could readily be debenzylated to the desired compound. The reaction failed, however, and gave only 8-oxoguanosine. Alternatively, direct amination of 8-benzyloxyguanosine with HSA in aqueous NaOH followed by debenzylation afforded 1- amino- 8-oxoguanosine. Treatment of this compound with methyl iodide in the usual basic, nonaqueous conditions gave 1- amino- 7- methyl- 8-oxoguanosine. The N- amino group, having acted as a blocking agent at position 1, could then be selectively removed with HN02 to provide the desired compound (lQ).

CH- H2N ^L

H2N^^N C6H5

ribose ribose ribose

I

ribose

AMINE- N-IMINES

36 ' 3T Amine- N-imines (li) (also called N-ylides ) are the most synthetically important intermediates available from the products of direct amination on 38 nitrogen. These compounds, which have been recently reviewed, are analogous to the N- oxides (12) and the ammonium- ylides (l^). © © © © © • — N-Ol ^N-N-R ^N-Ctl

12 11 U ,

- 81 -

Both aliphatic and heteroaromatic amine- N- imine s are commonly prepared by deprotonation of the corresponding quaternary ammonium salts. If R is an acyl, sulfonyl, cyano, or nitro group, deprotonation results in formation of 39 an aminimide. This class of compounds, recently reviewed "by McKillip, et al . has been found useful in adhesives, detergents, photographic materials, water repellents, antistatic agents, and pharmaceuticals. The heteroaromatic- N-imines, in which charge may be delocalized throughout the ring system, are more stable than the unsubstituted aliphatic N-imines, in which charge is localized. 38 The following diagrams illustrate the nucleophilic and electrophilic reactions characteristic of the dipolar amine- N-imines. b C-H :B

c C-N-^N-R H ^CHa B: B: B J-

The nucleophilic imino group may be protonated, alkylated, or acylated (paths a and i); protonation reforms the quarternary ammonium salt. Electrophilic reactions are numerous. The acidic proton alpha to the quaternary ammonium group will exchange quite easily (paths b and f); some heteroaromatic N-imines will undergo ring opening on treatment with aqueous base (path g); and an aliphatic N- imine with a side chain of two or more carbons may eliminate an olefin leaving a substituted hydrazide (path c). In addition, the N-N-bond is thermally unstable (paths e and j); the N- acylimines give the corresponding tertiary amine and an isocyanate on heating. 1,3- Dipolar addition (path h) is the most synthetically useful of the 40 amine- li- imine reactions. Huisgen, Grashey, and Krischke first studied 1,3 additions as a route to interesting heterocycles using pyridine- N- imine,

quinoline- N- imine , isoquinoline-N- imine, and phenanthridine-N- imine. These reactions may be visualized as addition to the azomethine- imine form in which the C-N double bond is part of an aromatic ring. 41 The driving force of the addition is quite strong since the aromatic resonance must be sacrificed.

OH RC3C-C02 CH3 -> C02 CH3 DMF, 20°

3,^Dihydro-N-iminoisoquinolines, which undergo reaction with a wide variety N-iminoquinolines of dipolarphile s , give higher yields than the corresponding and isoquinolines presumably because destruction of an aromatic system is unnecessary. 42 Alkenes react to give pyrazolidines, while reactions with alkynes give 3-pyrazolines. Reactions with the esters of fumaric and maleic acid and with cis- and trans- stilbene are stereospecific, giving the cis and trans configurations of the resultant pyrazolidines. In solution, the N-iminoquinolines and isoquinolines exist in equilibrium with their dimers, the hexahydrotetrazines. 41 The existence of a reversible thermochromism indicates a mobile equilibrium. The hexahydrotetrazine is stable and offers a convenient source for azomethine imine s. , , • - 02

t*

1,3- Dipolar additions are also effective in building sulfur- containing heterocycles. The addition of the dipolarophile , diphenyl thiocarbonate, to isoquinoline-N-imine is accompanied by elimination of phenol and gives 8l$ 2-phenoxy-9H- [l,3,4]-thiadiazolo- [2,3-a]-isoquinoline.

S ^ / C-0C6H5 CO H5 C 6

*0C6H5 The photochemistry of the amine- N-imines has recently received much attention Balasubramanian, et. al . , have reported the formation of 1,2- diazepine in good yields by irradiation of aromatic amine-N-imines. (path k) In most cases small amounts of the parent amines were also isolated. This work is complementary to earlier studies by Streith43 and Sasaki. 44 Tamura, 45 et. al . have indicated, however, that quinoline-N-benzoylimine gives 2- benzamidoquinoline. (path l) Comparable results have been obtained with the isoquinolines. 37

hv hv

This interesting result may be compared to the photochemistry of the heteroaromatic amine N-oxides. Irradiation of adenine- 1- oxide yields two major products; the rearrangement product, isoguanine, and adenine. Brown, 47 et . al . have suggested that transfer of the oxygen to the adjacent carbon may proceed through a three- membered oxaziridine intermediate; an analogous diaziridine intermediate may be suggested for the N-imines. Work is in progress to elucidate the reaction mechanism and those factors determining product distribution. SUMMARY Development of reagents suitable for direct amination on nitrogen has provided new approaches to useful intermediates; interesting heterocycles such as pyrazolidines and pyrazolines may be reached by 1,3 dipolar addition to the amine-N-imines, methylation may be directed by an N- amino blocking group, and tertiary structure and hydrogen bonding interactions of nucleic acids maybe explored. The synthetic utility of direct amination reagents undoubtedly wi±l be exploited in future research. BIBLIOGRAPHY

1. F. Raschig, Chem. Ber., kO, ^580 (1907). 2. R. Mattair and H. H. Sisler, J. Amer. Chem. Soc, 73, lol9 (1951). J. Amer. Chem. Soc, 3- H. H. Sisler, F. T. Neth, R. S. Drago, and D. Yaney,

76 3906 (195^)' V , , , k. P. Kovacic, M. K. Lowery, and K. W. Field, Chem. Rev., 70, 639 (1970 J. 5. K. Utavy and H. H. Sisler, Inorg. Chem., _5, 1835 ( 196677 6. w. Rundel, Angew. Chem., Int. Ed. Engl., 1, hOJ> (1962). (195?) 7. K. Hoegerle and H. Erlenmeyer, Helv. Chim. Acta, 32,, 1203 (196*0. 8. H. Bock and K. L. Kompa, Z. anorg. allg. Chem., j£2, 238 (1962). 9. G. Yagil and M. Anbar, J. Amer. Chem. Soc, 8^, 1797 _ — - 83 -

10. G. L. Braude and J. A. Cogliano, J. Chem. Soc. , 4172 (1961). 11. R. Gtfsl and A. Meuwsen, Chem. Ber., 9£, 2521 (1959)-

12. E. Colton and M. M. Jones, J. Chem. Educ. , 32, 485 (1955). 13. G. M. Omietanski and H. H. Sisler, J. Amer. Chem. Soc, 78, 1211 (1956). 14. F. Sommer, 0. F. Schulz, and M. Nassau, Z. anorg. allg. Chem., l47, i42 (1925). 15. H. H. Sisler, R. H. Bafford, G. M. Omietanski, B. Rudner, and R. J. Drago, J. ..Org. Chem., 24, 859 (1959). 16. R. Gosl and A. Meuwsen, Org. Syn., 43, 1 (1963). IT- W. Klotzer and M. Herberz, Monatsh. Chem., %6, 1731 (1965).

18. W. Klotzer, ibid . 11J (1966). , 2L -T 19. A. D. Broom and R. K. Robins, J. Org. Chem ., 34, 1025 (1969). 20. L. A. Carpino, C. A. Giza, and B. A. Carpino, J. Amer. Chem. Soc, 8l, 955 (1959).

21. L. A. Carpino, ibid . , 82, 3133 (i960).

22. G. Zinner, Arch. Pharm. (Weinheim), _293, 657 (i960); 303 , 317 (1970). 23- Y. Tamura, J. Minamikawa, K. Sumoto, S. Fujii, and M. Ikeda, J. Org. Chem., 38, 1239 (1973).

2k. T. Sheradsky, J. Heterocycl. Chem. , k, 4l3 (1967). 25. T. Sheradsky, Tetrahedron Lett., 1909 (1968). 26. R. Y. Ning, Chem. Eng. News, 36 (Dec 17, 1973)- 27. C R. Johnson, R. A. Kirchhoff, and H. G. Corkins, J. Org. Chem., 3>2, 2^58 (191k). 28. Y. Tamura, J. Minamikawa, Y. Kit a, J.H. Kim, and M. Ikeda, Tetrahedron, 22, 1063 (1973). 29. Y. Tamura, J. Minamikawa, Y. Miki, S. Matsugashita, and M. Ikeda, Tetrahedron Lett., 4133 (1972). 30. D. F. Wiemer and N. J. Leonard, J. Org. Chem., 32, 3^8 (197*0. 31- (a) Y. Tamura, K. Sumoto, J. Minamikawa, and M. Ikeda, Tetrahedron Lett., 4l37 (1972); (b) Y. Tamura, H. Matsushima, M. Ikeda, and K. Sumoto, Synthesis, 277 (197*0. 32. P. R. Srinivasan and E. Borek, Science, l45, 5*1-8 (1964). 33« E. Tsutsui, P. R. Srinivasan, and E. Borek, Proc Natl. Acad. Sci. U.S., 56, 1003 (1966).

3k. 7a) K. Fink and W. S. Adams, Arch. Biochem. Biophys., 126 , 27 (1968); (b) R. W. Park, J. F. Holland, and A. Jenkins, Cancer Res., 22, 469 (1962). 35« B. H. Rizkalla, R. K. Robins, and A. D. Broom, Biochim. Biophys. Acta,

195 , 285 (1969).

36. A. Balasubramanian, J. M. Mcintosh, and V. Snieckus, J. Org. Chem., 35., 433 (1970). 37- J. Becher and C. Lohse, Acta Chem. Scand., 26, 4o4l (1972). 38. H. J. Timpe, Z. Chem., 12, 250 (1972). 39« W. J. McKillip, E. A. Sedor, B. M. Culbertson, and S. Wawzonek, Chem. Rev., 73, 255 (1973). kO. R. Huisgen, R. Grashey, and R. Krischke, Tetrahedron Lett., 387 (1962). kl. R. Huisgen, Proc. Chem. Soc, 357 (1961). k2. C. G. Stuckwisch, Synthesis, k69 (1973). k3. (a) J. Streith and J. M. Cassal, Angew. Chem. Int. Ed. Eng., 7, 129 (1968); (b) J. Streith and J. M. Cassal, Tetrahedron Lett., k^kl (1968); (c) J. Streith and J. M. Cassal, Bull. Soc. Chim. Fr., 2175 (1969). kk. S. Sasaki, K. Kanematsu, and A. Kaheki, Chem. Commun. , k32 (1969). *+5- Y. Tamura, H. Ishibashi, N. Tsujimoto, and M. Ikeda, Chem. Pharm. Bull. (Tokyo), 12, 1285 (1971). k6. A. R. Katritzky and J. M. Lagowski, "Chemistry of the Heterocyclic N- Oxides," Academic Press, New York, 1971, p. 320. kl. G. B. Brown, G. Levin, and S. Murphy, Biochemistry, 3, 880 (1964).

. -6V NEW MASKED ACYL CAEBANIONS IN SYNTHESIS

Reported by Brian Holmes February 3> 1975 During the last decade considerable effort has been directed toward the development of synthetic transformations that invert the inherent polarity of functional groups. Such transformations are referred to as symmetrization, 1 dipole inversion, charge affinity inversion, and umpolung. 2 Transformations that invert the polarity of the carbonyl function from its usual role as an electrophile in a polar condensation reaction to a nucleophile are referred to as nucleophilic acylation methods and the nucleophiles are referred to as masked acyl carbanions, masked carbonyl anions, and acyl carbanion equivalents. Nucleophilic acylation methods were thoroughly reviewed by Seebach in 1969« 3 The lithio dithiane reagents are perhaps the best known masked acyl carbanions and have proven to be extremely valuable synthetic tools. The 4 5 chemistry of these reagents has also been reviewed. > Although much of the recent work in this area has been devoted to the improvement of hydrolytic methods, 6 the lack of simple, mild, and inexpensive hydrolytic conditions remains the principal limitation of this method. A great deal of imagination has recently been devoted to the design of new masked acyl carbanions which are both more economical and more easily hydrolyzed to their corresponding carbonyl compounds. In addition, many of these new reagents carry out new, synthetically important transformations. This seminar will discuss new masked acyl carbanions in terms of their use, synthetic applications and limitations. The material is organized according to reagent in the following categories: l) metallo aldimines, 2) protected

cyan ohydr ins , 3) thioacetal monoxides, k) vinyl lithiums, and 5) miscellaneous reagents Metallo aldimines

Metallo aldimines (3) are most easily prepared by the addition of an 7 organometallic reagent (2) to an isocyanide (l) (Scheme i)." The isocyanide in this process must meet two requirements. First, it must have a trisubstituted

Scheme I

X-|_N=C + RM v ^ ^4- N=C^ ^M 12 3 a-carbon since the hydrogens a to isocyanides are acidic. Secondly, the alkyl substituent must not have more than one aromatic group since this leads to an isocyanide-exchange process (Scheme II). 8 The isocyanide most commonly used is 1,1,3,3-tetramethylbutyl isocyanide (TMBl)(l), because it meets the above requirements and is not particularly malodorous.

Scheme II _, /R Ph 3 C-N=C > Ph 3 CM + 1FC-R

The organometallic is generally an organolithium. Primary, secondary and tertiary alkyllithiums react well with TMBI. Organolithiums whose conjugate acids have a pKa less than 37 on the MSAD scale do not react appreciably due to the equilibrium nature of the addition reaction. 9 Phenyl lithium (conjugate acid, pKa:r37) reacts to about ^Qffo completion when added to an equimolar quantity of TMBI, while sodium diethyl malonate and lithium phenylacetylide do not react at all. Although the use of Grignard reagents extends the scope of this reaction by making available a greater variety of organometallics, yields are generally lower presumably due to unfavorable equilibria. ) -8 5 - Metallo aldimines react with a wide variety of electrophiles (Scheme 9 10 III). ' Scheme III r Li MgBr i D 2 -» R-C-D 50-95 48-67

*t- N=C CO, -> R-C-COH 52-80 26-48 M < R'X (R'=Me,Et) > R-C-R' 90 -

R=n-Bu, Et, CICOEt * R-C-COEtu 6k -

sec-Bu, or Hi i? (CH3 ) 3SiCl > R-CSi(CH3 ) 3 ko OH 1 I -> R-C-CH2 -C 90 —, CH<

Although a good method for the preparation of aldehydes and a-keto acids and esters, the greatest synthetic utility of this method lies in the preparation of C-l deuterated aldehydes. Deuteration of lithium aldimine derivatives proceeds in excellent yield with deuteration occuring exclusively at C-l (97-98% purity by nmr analysis). Deuteration of Grignard reagents is sluggish and some scrambling of the label occurs between C-l and C-2. Although lithium aldimine derivatives may be alkylated with primary halides, reactions with secondary halides give elimination. More hindered ketones can be prepared by the use of chlorodialkylboranes (Scheme IV), 11 which require a two- fold molar excess of the olefin. Scheme IV cl + BH2 C1- AO?

HS-CH2 C02H

HOoC-CHp-S (90% overall based on the metallo aldimine The Protected Cyanohydrin Method The protected cyanohydrin method of generating masked acyl carbanions utilizes the acidifying effect of the nitrile and the reversibility of cyanohydrin formation. 12 The protected cyanohydrin is generated by conversion of an aldehyde to a cyanohydrin followed by treatment with a base stable protecting group such as ethyl vinyl ether. The carbanion h, generated by the action of lithium diisopropylamide (IDA) in THF, reacts readily with primary or secondary halides to give ketones after acidic hydrolysis. The entire process is outlined in Scheme V with some representative yields given in Table I. $ -86-

Scheme V OEt EtO—U

1) KCN LDA RX | HoO II 4 CH3CH > CHoCH — > CH3 C-R -> CH 3 CR 2)=\ I CN OEt

Table I A protected aminonitrile has been used for the synthesis of nicotine (8) and myomine (9) RX Yield 13 (Scheme Vl). The derivative 6 was prepared in

n-C 4H9Br 80-85$ 92$ yield from pyridine-3- aldehyde (_5) by treatment (CH3 ) 2CHBr 80$ with morpholine perchlorate followed by potassium cis-Ch%CB>CHCHP CHpI 6l$ cyanide. The key step was the 1, ^-addition of the CH2 =CHCH2Br 76$ carbanion of 6 to acrylonitrile to give the adduct 7. The synthesis was completed by acidic hydrolysis of the cyanohydrin followed by reductive cyclization to give a mixture of 8 and 9 in a combined yield of 67$ from pyridine-3-aldehyde. Scheme VI 0-

HC104 9 ? CHO H |^S— CCN 1) KOt-Bu CCH2 CH2 CN L H 2) KCN I 2) CH 2=CHCN OrN' + 3) H 5 6 7 (90$)

Raney Nickel/k2

A protected cyanohydrin can be generated in one step by treatment of an aromatic or heterocyclic aldehyde with trimethylsilylcyanide (10) to give the 14 adduct 11 (Scheme VII ). Scheme VII

0Si(CH33/3) (CH3 ) 3SiCN 1) LDA ArCHO -> ArCH-CN ' ~* ArCR 2) RX 0" 10 11 3) H3

The carbanion of 11 , generated by treatment with LDA in THF at -78°, reacts with primary or secondary halides to give ketones in high yield after hydrolysis. By this procedure benzaldehyde and furfural can be converted to isopropyl phenyl ketone and isopropyl 2- fury1 ketone by reaction with isopropyl iodide in yields of 95$ and 80$ respectively. Protected cyanohyarins may also undergo conjugate addition. Increasing the steric bulk of the cyanohydrin or hindering the carbonyl has a 1,4- directing effect. 15 A general method achieving conjugate addition uses cyanohydrins derived

from or, P-unsaturated aldehydes which make the 1,2-addition process reversible. -Of- , , The cyanohydrin of crotonaldehyde adds 1,4 to cyclopentenone, cyclohexenone 3-methylcyclohenenone, and acrylonitrile in yields of 75-85$ (Scheme VIII ). Scheme VIII

OEt -< EtO

1) LDA H 3 CH-,CH=CH

CN 2)

Another example of conjugate addition with a protected cyanohydrin is found in the synthesis of the tetracycline analog 13 16 Treatment of the protected cyanohydrin of 12 with an etheral solution of potassium t-butoxide gave upon hydrolysis and chromatography the tetracycline 13. The overall yield from 12 was 35$ (Scheme IX ). Scheme IX

1) KCN 2) O 3) KOt-Bu + k) H3

The principle advantage of the protected cyanohydrin method over other acyl carbanion equivalents is the ease with which protected cyanohydrins may be generated in complex molecules. Moreover, secondary halides give mainly alkylat ion rather than elimination. Thioacetal Monoxide

Although similar in reactivity to the dithianes, the thioacetal R-S-CH2 -S-R monoxides (Jh) offer the advantage of easy hydrolysis without the necessity of mercury salts. 17 The only side product of the 14a R=Me ; hydrolysis is dimethyl disulfide. Methyl methylthiomethylsulfoxide Hb$ R=Et (l^a) is commercially available while other reagents may be prepared by the method of Scheme X. 18 Scheme X

Hg °2 RS-CH2 -SR > R-§-CH2SR ( 76-96$) AcOE 17 The carbanions 15 can be generated with sodium hydride, n-butyl lithium, or IDA, 19 usually in THF at 0°. These reagents can then be used for the synthesis of aldehydes, 17 symmetrical20 and unsymmetrical ketones, 19 1,2-diketones, 21 o>hydroxyketones, 21 arylacetic acids, 22 and a-dialkylaminoaldehydes 23 (Scheme XI). -88-

Scheme XI

N^ CHCHO (75- (550)

II HCR' R'X RSCH-SR CH^CH P )pCOp,E t

CH 3 ( CH2 J 2 C-CH (80-90^) (75-* 15a; R=Me Triton 15b; R=Et PhCOCl ROH <- Ph-C-CH H* < C02R (TO- 80$) CHO (T¥)

The reaction of nitriles with the carbanion 15a to produce enaminosulfoxides provides a convenient route to amino acids (Scheme XII). 24 The rearrangement of Scheme XII

H ? AcN AcNH P 1)/-CNv. H2N SMe MeS ACoO 1) MeOH 15a "* / QMe 2) HoO SMe SMe 2) Raney Ni \ 16 (720) 17 (820 from 16)

16 to 17 is rationalized on the basis of a Pummerer rearrangement with concomitant migration of the methylthio group. The generality of the method is demonstrated in the synthesis of 5-hydroxytryptophan (Scheme XIII ). Scheme XIII NHAc H2N SMe 15a HSS (MeO) 2CHCH2 CH2 CN > >=< ^U ^/-{ (MeO) 2CHCH2 CH2 SOMe (MeO) 2CHCH2CH2 SMe (770) (730)

:l MeOH 1)2 2) Raney NHNHo-HCl Ni (Jhfo) -oy-

The diethyl derivative 15b is the only carbonyl anion equivalent that both alkylates and undergoes conjugate addition in a synthetically useful manner. This "limch-pin" method of 1,4-diketone synthesis is demonstrated in the synthesis 19 of dihydrojasmone (l8) (Scheme XIV"). Monoalkylated derivatives of the type 19 Scheme XIV

15b I TsO-n-C6H13 EtS

HO HgClg/^O*

18 (76$ overall) are valuable for the preparation of Y-keto esters (Scheme XV). 19 Scheme XV

Et 2 OEt 0^/°\_-OEt EtS Et \ e xy + ( SOEt EtS EtS SEt 19 (82$

Although the initially formed dithioketal monoxides can be hydrolyzed a to their corresponding carbonyl compounds by treatment with dilute sulfuric acid,™ distilled aldehyde products are often contaminated with dimethyl disulfide. 19 This difficulty can be avoided by carrying out the hydrolyses with mercuric chloride and a catalytic amount of 'JQffo perchloric acid. Vinyllithium Reagents Vinyllithium reagents that are used as masked acyl /XR carbanions have the general formula 20. Although CH2 =C reagents of this type have been postulated for some time, 4 generation of this type of anion in a synthetically 20> X=0, S, or Si. useful manner has only recently been accomplished. Three new procedures hold synthetic promise. Methoxyvinyllithium (MVL) (21) can be generated from methyl vinyl ether 25 26 by the action of n-butyllithium or IDA in THF at -78°. > The reaction products of MVL with electrophiles contain the vinyl ether function which may be further elaborated or, more usually, converted to their corresponding carbonyl compounds by treatment with aqueous methanolic 0.02 N HC1 at 25°. Aldehydes may serve as electrophiles for MVL with the ultimate products being exclusively cehydroxy- ketones. The reactions of MVL are summarized in Scheme XVI. A similar technique employs alkylthiovinyllithium reagents which are 2y obtained by treatment of 22 with sec-butyllithium in THF-HMPA at -78°. Alkyl halides, aldehydes and epoxides can serve as electrophiles (Scheme XVII ). Hydrolysis of the resultant vinyl sulfides can be accomplished by treatment with mercuric chloride in refluxing acetonitrile-water, or by 28 treatment with titanium tetrachloride in water at 20°. A -90- Scheme XVI

1/2 RC02Et n-CsHv/I ' «* £ v ^ n-C8HiyCCH 3 (80$) R OH / OMe CH P =C (75-90^) •Li Vn MeO 21 l) Simmons- -vh• /=v Smith

(*o$) c H

Scheme XVI'E / SEt CH2 =C \ H '^2 sec-BuLi

SEt ,1) RX 1) RCHO CH3CR CH2 =C^ R-CH-C-CH,n

(65-90^) 2) HgCl, Li 2) HgCl£ (51-6^)

Epoxysilanes are efficient precursors of carbonyl compounds under acidic 29 30 conditions. * Trimethylsilylvinyllithium 23 has been used for the preparation 31 of the hydrindanone _25. Conversion of intermediate 2k to 2_5 requires ketaliza- tion epoxidation of the vinyl silane and acidic hydrolysis (Scheme XVIIl). Scheme XVIII

1) H0CH2 CH2 0H l) sec-BuLi 2) m-CiPhC03H^ » Cul A ) A 3) H2S0 4 Br SiMe. Li SiMe ^V 23 SiMe, (56$ 2I1

Miscellaneous Reagents The Pummerer rearrangement converts sulfoxides to sulfides with simultaneous oxidation of the a-carbon. This reaction has been used for the synthesis of a-hydroxyaldehydes, as shown in the synthesis of DL-threo-O-methylphenylserine 32 (26) (Scheme XIX). Scheme XIX

OH l)Mel/ OMe OAc 1) NaCN NH S

PhSCTL AgO \ I 2) m*C1 PhCHO- ±k> PhCHCHpSPh —> PhCH-CH > Ph-CH-C-C02H 5) *** Me0 NaOAc L A ( 100^) 26 . -91-

Treatment of 3,3-diethoxy-l-methylthiopropyne (2j) with n-butyllithium in hexane at -78° gives an acyl acetate carbanion equivalent. 33 Alkyl halides or carbonyl compounds can serve as electrophiles to give acetoacetic esters (28) or methylthiobutenolides (29) respectively (Scheme XX) Scheme XX

° CH3S-C=C-CH(OEt) 2 - > CH 3S-C=C=C(OEt) 2 27 HgClg R-C-CH2C02Et < 28 (80-90^) 3) p-TsQH

29 (6>90% Summary

Although the masked acyl carbanions discussed here undergo a variety of different reactions, many comparisons are possible. Table II is a summary of the general methods with comparisons to the lithio dithiane reagents. The entries were selected on the basis of the most similar examples available in each of the main categories. Many reagents not included in Table II are useful for specific purposes. The vinyl silane reagent _23 may provide a synthesis of hindered 1,^-diketones if hydrolysis can be simplified and yields improved. The aryl sulfoxide method of a-hydroxyketone synthesis is comparable to those in Table II. The acyl acetate reagent is comparable in both simplicity and yields with other reagents which introduce this functionality. o V S to O K O w Sd W w td P c+ O ^ CD o=o 1 i 1 i o a c+ P 3 P. 3 1 V>=o o^=o o=o o=o o=o O P CD CD 1 1 4 P. 4 o 1 1 3 4 < 4 4 d- H- O 4 o > w O— o^=o td W W o H- H- c+ i-> P ro o=< = i w 1 o o P 3 td w 3 3 W bdrAJ — — \->

3 ti P IV) |m|3 td M |ml3 H 4 CD 1 = d- CD 1 d- to g <*t~ 1 111 (t IB H- (J CD bo S O bd II lo a K, 4> (D O 4> d H 4 i 3 I 3 > S It It HcKD P ~ CD O bd K C d- H CD 4 4 D. O 4 4 3 4> X 3 3 v; 13 Pi d- cd cd CD CD 3 hd d- 1 td W H P « W J H 3 3 O H- a Pi 03 H M 3 =<{ bd Hj 6 g'H O O o o rf" X H« H 3 H . h ID ID CD O 3 O c+ •r) VO VO CD -^ M|V>J tf dl-p- CD H*P ro OS |vn|-p-fl> 4 O O ^ K' V Kj Ul H . p H- VO VO H« H p. c+ 3 & vn oo 00 00 -q O0 CD H«« O 4) WW ID xS >I3 ^ d- 3* 03 CD P> ID H- c^-^ H 1 pi ID a id d- 3 bd 4 P. 1 H w d- m c+ 3 3 9)9 d- << — H- H 3" 3 CD ID CD CD H ID H 3 ro tn •- ^ ^— 4 CD 4 i 3^d: s, 3 4 «, X H« Pi CD CD \ P CD CD - H H £ Hl3 t; o >.. d- o o 8* ^ ft IV) 4 CD K H hd W hd td CD fD 3' W ^ ^ P -£©-= O H CD 7^ d ^ P ^ Kj H C7\ -q H H o \ H- -si ~\- O I 1 H- O MD V-M Oo—] 0A 00 CO CD 1 4 &^& ON VO ro vn vn H ro d- i Pi CD Q.

Mi CD 3 o w S 13 2 S H S • d- CD = CD I CD CD td H' CD ^ o II hd d- d- 3 1 CD 3^ ro CD H. 3 3* - i 4 ^ 1 3 3 oi|3 H« O H ID d- O 1 3 X CD w •d o o H td fi H £ § 9- p. 3 13 d-«

•d 3 i-3 ro o M W td hd rol3 4 3 3' vn 4> ro crq = .--^ 1 CD td CD 03 CD o 3* o o II 13 O, W d- 03 -3 "=« 03 CO CAVO H ro vn O Pi OD

-ee-H Id- g s s w S S S Ih- hd ro S td t4 J J * CD i ro 1 ."-o o n oq CD ID CD 3 ID CD CD I 3 CD td H> H« • v» CD a 00 O O to II 13 d- d" J vn 1 3 u 1 3^ 3 cr\ Pi td ^ p ro 13 hd 01 H- H- \^ H« CD O 4 c+\ 1 £ CD CD P o Hj d- • hd S hd 2 3 O 3 P • 3"a 3' 4 W CD CD IZf ID d- I td CD ^# P H CD CD (K) 3 VO CO ^ bd 3 P Ion 4 i-b O d- vn ro H 3 3 H\ =<; CD 3 3 ^ -q-q H VO VO <«• o o CD 00 K| s.

Eeferences

1. E. J. Corey, Pure Appl. Chem., 14, 19, (1967). 2. D. Seebach, and M. Kalb, Chem. Ind., (London), 687 (1974). 3. D. Seebach, Angew. Chem., Int. Ed. Engl., 8, 639 (1969). "A 4. D. Seebach, Synthesis, 17 (1969). 5. J. Ollinger, University of Illinois Seminar, May 2 (1972). 6. E. J. Corey, and B. W. Erickson, J. Org. Chem., 36, 3553 (1971)} T. L. Ho, h. C.

Ho, and C. M Wong, Chem. Commun., 791 (1972; M. Fetiz-cn^, and M. Jurion, 1 Chem. Commun., 382 (1972); T. Oishi, K. Kamemoto, and Y. Ban, Tetrahedron Lett., IO85 (1972); T. L. Ho, and C. M. Wong, Synth. Commun., 561 (1972); Y. Tamura, K. Sumoto, S. Fujii, H. Satoh, and M. Ikeda, Synthesis, 312 (1973). 7- H. M. Walborsky, and G. E. Niznik, J. Amer. Chem. Soc, 91, 7778 (1969). 8. H. M. Walborsky, G. E. Niznik, and M. P. Periasamy, Tetrahedron Lett., 4965 (1971)} M. P. Periasamy and H. M. Walborsky, J. Org. Chem., 39, 6ll (1974). 9- G.E. Niznik, W.H. Morrison, III, and H.M. Walborsky, J. Org. Chem., 39, 600 (1974 10. H. M. Walborsky, W. H. Morrison, III, and G. E. Niznik, J. Amer. ChemTSoc, 92, 6657 (1970). 11. Y. Yamamoto, K. Kondo, and I. Moritani, Tetrahedron Lett., 793 (1974). 12. G. Stork, and L. Malanado, J. Amer. Chem. Soc, 93, 5286 (1971). 13. E. Leete, M. R. Chedekel, and G. B. Boden, J. Org. Chem., 37, 4465 (1972). 14. K. Deuchert, U. Hertenstein, and S. Hunig, Synthesis, 777T1973). 15. G. Stork, and L. Malanado, J. Amer. Chem. Soc, 96, 5272 (1974). 16. E. Aufderhaar, J. E. Baldwin, D. H. R. Barton, D. J. Faulkner, and M. Slayton, J. Chem. Soc (c), 2175 (1971). 17. K. Ogura, and G. Tsuchihashi, Tetrahedron Lett., 3151 (1971). 18. K. Ogura, and G. Tsuchihashi, Bull. Chem. Soc. (japan), 45, 2203 (1972). 19. J. E. Richmann, J. L. Herrman, and R. H. Schlessinger, Tetrahedron Lett., 3267, 3271, 3275 (1973). 20. G. S chill, and P. R. Jones, Synthesis, 117 (1974). ^ 21. J. L. Herrman, J. E. Richmann, P. J. Wepplo, and R. H. Schlessinger, V Tetrahedron Lett., 4707 (1973). 22. K. Ogura, and G. Tsuchihashi, Tetrahedron Lett., 1383 (1972). 23. L. Duhamel, P. Duhamel, and N. Mancelle, Bull. Chem. Soc. Fr., 2, 331 (1974). 24. K. Ogura, and G. Tsuchihashi, J. Amer. Chem. Soc, 96, i960 (1972).

25. J. E. Baldwin, G. A. Hofle, and 0. W. Lever, Jr., J. Amer. Chem. Soc, 96 , 7125 (1974). 26. U. Schollkopf, and P. Hanssle, Justus Liebigs Ann. Chem., 763, 208 (1972). 27. K. Oshima, K. Shioji, H. Takahashi, H. Yamamoto, and H. Nozaki, J. Amer. Chem. Soc, 95, 2694 (1973). 28. T. Mukayama, K. Kamio, S. Kobayashi, and H. Takei, Chem. Lett., 3723 (1972). 29. J. J. Eisch, and J. T. Trainor, J. Org. Chem., 28, 487, 2870 (1963). 30. G. Stork, and E. Colvin, J. Amer. Chem. Soc, 93, 2080 (1971). 31. R. K. Boeckman, Jr., and K. J. Bruza, Tetrahedron Lett., 3365 (1974). 32. S. Iriuchijama, K. Maniwa, and G. Tsuchihashi, J. Amer. Chem. Soc, 96, 4280 (1974). 33. R. M. Carlson, and J. L. Isidor, Tetrahedron Lett., 4819 (1973). 34. T. Mukaijama, K. Navasaka, and M. Furusato, J. Amer. Chem. Soc, 94, 864l (1972). 35. D. J. Bennett, C. W. Kirby, V. A. Moss, Chem. Commun., 218 (1967). - 9 k- TRANSITION METAL HOMOGENEOUS CATALYSIS: RECENT APPLICATIONS IN FIBER, PLASTIC AND OTHER INDUSTRIES

Reported by Robert W. Mason February 10, 1975 INTRODUCTION An extensive fiber and plastics industry has developed over the past thirty years based on homo- and copolymers containing such monomeric units as vinyl acetates and chlorides, ethylene and propylene glycols, as well as esters and amides. The tonnage of such polymers produced by U. S. industry far exceeds that of all other synthetic organic chemicals combined, indicative of their impact in the American economy. The precursors to the monomers have historically been produced by heterogeneous, fixed bed catalyzed olefin oxidations and additions. Only in the past ten to fifteen years has homogeneous catalysis been applied to industrial organic synthesis, primarily involving transition metal complexes dissolved in aqueous or acetic acid solutions. This seminar will discuss some of the newer homogeneously catalyzed syntheses and the commercial applications of the products. 1 ' 2

HETEROGENEOUS VS HOMOGENEOUS CATALYSIS When considering the economics of a heterogeneous or homogeneous catalytic system, one must examine efficiency, reproducibility, specificity, and separa- tion and recovery of the catalyst. 2a The first three points favor a homogeneous system, as one is dealing with a molecular species as the catalyst, rather than a surface. One can change ligands, metal oxidation states, or even the metal itself in order to maximize the yield of the desired product. However, the fourth point, catalyst recovery, favors a heterogeneous system, where simple coarse filtration is sufficient to separate the catalyst and the reaction solution. Homogeneous catalysts require an expensive distillation, and any loss of catalyst through inefficiency will add to the overall cost of the product. A third type of catalyst system, consisting of polymer supported transition metal complexes, 3 combines the advantages of heterogeneous and homogeneous catalysts, and will undoubtedly find increased industrial use. HISTORICAL

The first homogeneously catalyzed process to be commercialized on a large scale was Wacker Chemie's two- stage oxidation of ethylene to acetaldehyde, developed by Smidt 4 and coworkers:

a) PdClo + CH2 CH2 + H2 —> Pd + CH 3CH0 + 2HC1 (l)

Pd + 2CuCl2 —>PdCl2 + 2CuCl (2)

b) 2CuCl + 2HC1 + f02 —> 2CuCl2 + H2 (3)

CH2CH2 + ^02 —> CH3CHO (4) In the first stage, palladium chloride reacts with ethylene, producing the acetaldehyde. The insoluble palladium metal is oxidized back to palladium (il) 100- 110° with copper (il) chloride, the co- catalyst in the process. At an£ ] q_ j_± atm, 99$ of the ethylene reacts in a single pass through the reactor. After pressure release, the crude acetaldehyde is distilled. The aqueous reactor solution is recirculated and the copper (i) chloride oxidized back with air or oxygen to copper (il) chloride, the second stage of the process. A fresh batch of ethylene can then be fed in, and the process continued. One must be careful in applying the term "catalyst" in such homogeneous reactions, for the active catalyst is not a metal surface, but rather a series of related metal complexes. One must think in terms of catalytic intermediates, 5 as demonstrated by the current hypothesis, based on extensive evidence collected by several workers, for the mechanism of the Wacker process, as reviewed by Jira and Freisleben: b 95 - a. Olefin absorption:

CH ]~ 2 _^> 2 f CH2CH2 + [PdCl 4 *c- -> Pd- CI + CI (5) CH2 CI J b. Hydrolysis:

1 n - CH2 f CH2 ^s> f- + H >Pd-OH + CI || —>Pd-Cl 2 II— 2 (6) CH= CH l CI 2 x

CI "I CH: 2 |

|| >Pd-OH + HpO —>Pd-OH + KrO 2 •^ || (T) CH CH 2 ^ 2 CI

c. Trans- cis isomer izat ion (reaction 8 same as 6 and T)

CI 1 CH CH 2 | 2 | + -i* + || >Pd-OH 2 HpO «r —>Pd-OH H^O +C1 (8) CH CH 2 CI 2 OH

1 CI1 "* CH2 f + ch2 y + || —> Pd-OH + H^0 +C1~ II—>Pd-ci 2HpO (9) CH I CH 2 OH 2 OH d. Olefin insertion

C1 "• CH2 _^ ]" ~$ [HOCH CH -PdCl II 2 2 2 (10) CH: I 12 OH J

e. Hydride transfer via hydrido olefin complex: OH

H OH I \/ CI I ji. CH C-PdCl C I _ 3 2 (11) _i». ^~ ]' || —>Pd-Cl! [HOCH2CHp-PdCl2 CHo | H H

f. Rate determining cleavage of a- complex;

OH OH

]" CH 3C-PdClo| + H2 -> 'CHa-O-OHpj + [PdCl2 (12)

i L H i J

y

CH.3CHO + H 3 Pd + 2C1 )) .

- 96 - With the success of the acet aldehyde synthesis, work turned to develop- ing homogeneously catalyzed processes for the production of vinyl acetate and chloride monomers (VA and VCM). Polyvinyl acetate finds application in paint emulsions, adhesives, and acrylic fibers (consuming over 50% of the VA) Copolymers of VA and ethylene are used in milk carton coatings and hot-melt adhesives. 7 Polyvinyl alcohol, from polyvinyl acetate, is used in the pro- duction of Vinylon, a synthetic fiber. 8 Vinyl chloride, previously made by the hydrochlorination of acetylene, is an extremely important compound in the plastics industry. The brittle homopolymer is used for rigid plastic applica- tions, such as molded parts, panels, and pipes, or can be combined with plasticizers for soft plastic uses, such as flexible and low temperature plastics, electrical wire coatings, and transparent films. Traditionally, the route to VA was based on acetylene:

Hg0 ^°2 HCCH HCCH > CH3CH0 > CH3COOH > CH3C00CH=CH2 (13 Hg Mn Salt Salt va As ethylene became cheaper than acetylene, the switch to a Wacker-type process for the production of VA began in the early 1960's, with a $30/ton 9 savings in product cost. The reactions, based on a Russian patent, are analogous to the acetaldehyde process, 2 but use an acetic acid, rather than aqueous, solvent:

a) CH2CH2 + Pd(00CCH3 ) 2 > CH 3 C00CH=CH2 + CH3C00H + Pd (lh)

Pd + 2Cu(00CCH3 ) 2 > Pd(00CCH 3 ) 2 + 2Cu(00CCH3 ) (15) b) 2Cu(00CCH3 ) + |02 + 2CH3C00H > 2Cu(00CCH3 ) 2 + H2 (l6)

|02 + CH2CH2 + CH3C00H > CH3C00CH=CH2 + H^O (17)

Despite severe corrosion problems, two processes have developed, one vapor phase, which yields VA in 95%> and the other liquid phase, which yields an acetaldehyde/VA mixture. The key to the two processes is the acetic acid. By adjusting the acetaldehyde/VA ratio to 1.1^, the liquid phase process becomes self-sufficient in acetic acid. The VA and acetaldehyde are separated by distil- lation, the acetaldehyde is oxidized with oxygen to acetic acid, and the acetic acid is recycled to the VA reactor. The vapor phase process requires continuous addition of fresh acetic acid. The Japanese have minimized this expense by hydrolyzing polyvinyl acetate to polyvinyl alcohol and acetic acid. The poly- alcohol is used to manufacture Vinylon fiber, and the acetic acid is recycled to the VA plant. A newly developed vinyl chloride monomer synthesis converts ethylene to 1,2-dichloroethane (DOE), and then thermally eliminates PIC1 from the DCE to yield VCM: 2d

a) CH2CIi2 + 2CuCl2 > C1CH2CH2C1 + 2CuCl (18)

2CuCl + 2HC1 + -§0 2 > 2CuCl2 + H2 (19

CH2CH2 + 2HC1 + |02 > C1CH2 CH2C1 + H2 (20 A b) C1CH2CH2 C1 > CH2-CHC1 + HC1 (21

The DCE is prepared in 98.7?^ yield, employing a CuCl/CuCl2 aqueous catalyst at 150° and 250-275 psig, in a steady state reactor in which DCE formation and catalyst regeneration occur simultaneously. The HC1 formed in the thermal - 97 - cracking of the DCE is recycled to the DCE reactor, reducing the amount of fresh HC1 needed in the process. CARBONYLATION10 Faced with an increasing demand for polyvinyl alcohol and acetate in the late 1960' s, the Monsanto Company decided to developed its own acetic acid plant. With ethylene costs increasing, little chance was seen for improving the acet aldehyde process. In examining alternative syntheses, Monsanto was first attracted to Badische Anilin- und Soda-Fabrik' s (BASF) methanol carbonyla- tion procedure, 11 employing a cobalt octacarbonyl- iodide promoter catalytic system:

CH3OH + CO > CH3COOH (90/0) (22) The reaction conditions however are harsh (210°, 7500 psi) and side reactions due to hydrogenation yield a complex product mixture. As Wilkin son- type rhodium complexes had been shown to be effective carbonyla- tion catalysts, 12 Roth and coworkers at Monsanto pursued the application of such complexes to the carbonylation of methanol. 11 ' 13 The first plant went into pro- 14 duction in Nov., 1970. Four days after start-up, acetic acid was being pro- duced in 99% yield with a purity exceeding Analytical Reagent specifications, and an overall savings of 15% as compared to the acetaldehyde process. Unlike the BASF process, Monsanto' s reaction conditions are very mild (l75°> 1 atm), eliminating both hydrogenation and corrosion problems. The catalyst employed maybe RhCl 3 -3H20, Rh2 3 - 5H20, RhCl(CO) (PPh 3 ) 2 , or [Rh(C0) 2 Cl] 2 , with an iodide promotor. Although Monsanto has not said which complex is actually used in the process, an RhCl(C0)(PPh 3 ) 2-HI catalyst is mentioned most often in the literature. After extensive kinetic work, Roth proposed the following reaction mechanism:

CH30H + HI ^=^ CH3I + H2 (23) + CH3I [RhL4 ]^==^ [CH3-RhIL4 ] (2k) { N [CH3-RhIL4 ] + CO ^ [CH3-Rhl(C0)L 3 ] + L (25)

CO N [CH3-RhIL 3 ] + CO ^ [CH 3C-Rhl(C0)L3 ] (26)

+ [CH 3C-Rhl(C0)L 3 ] v [CH3CI] [RhL4 ] (27)

[CH3CI] + H2 > HI + CH3COOH (28) 8 In the second step, oxidative addition of methyl iodide to the d - square 6 planar catalyst, yielding a d - octahedral complex, is rate limiting. The carbonyl insertion reaction (26) is believed to proceed by a mechanism similar 16 to that for the Co2 (C0) 8 catalyst, which is well documented:

CO

RCH CH ^Co(C0) + CO ^=^ RCH CH C-Co(C0) (n=2 or 3) (29) 2 2 n 2 2 n+1 Although reactions 27 and 28 may be combined, Roth prefers the stepwise reduc- tive elimination- hydrolysis sequence. In late 1973? Monsanto licensed its low pressure acetic acid process to the 17 USSR, for construction of a 150,000 ton/year plant in Ukranian Republic 4 ) - 98 - GLYCOLS Another ethylene based product of major commercial importance is ethylene 18 glycol. It finds application as a paint solvent in the automobile industry, antifreeze, in some explosives, in unsaturated polyester resins, and in poly- esters such as Dacron fibers and Mylar sheets, a polymer of ethylene glycol and diethyl terephthalate. As ethylene glycol is toxic, it cannot be used in foodstuffs. Propylene glycol however is not toxic, and finds application as a solvent in cosmetics, pharmaceutical preparations, soft drink syrups, as a tobacco humectant, not to mention a major use in unsaturated polyesters (as 19 reinforced glass fibers and plastics and a safety glass surface coating). Conventionally, glycols have been manufactured by oxidation of the corresponding olefin to the oxide, followed by hydrolysis of the oxide to the glycol:

RCH=CH2 > RCH-CH2 > RCH(OR")CH2OH (30)

The current disadvantages to this route are threefold. First, the olefins are increasing in cost, due to the increased prices for crude petroleum and natural gas. Second, the oxide synthesis produces considerable amounts of waste material or by-products. The chlorohydrin route yields large quanti- ties of calcium salts, creating a water pollution problem, while the oxirane and peracetic acid routes produce tert- butyl alcohol and acetic acid, respec- tively, in large amounts. Markets must be found for any by-products in order to make such processes economical. Finally, the oxide hydrolysis yields a

complex mixture of mono-, di- , and triglycols, which must be separated by a costly fractional distillation. 18 > 19 The Kuraray Company of Japan, in an effort to avoid some of these problems, has developed a new glycol synthesis based on a two- stage, palladium- catalyzed, homogeneous olefin oxidation. 20 The olefin in first oxidized to the glycol 21 monoacetate in the presence of a PdCl2/LiCl/LiN0 3 catalyst in acetic acid at 60°:

RCH=CH2 + is0 2 + HOAc > RCIi(OH)CH2OAc (31 In the absence of oxygen, the glycol monoacetate and nitric oxide are produced. In the presence of oxygen, no nitric oxide is observed, suggesting that the nitrate ion is the actual oxidixing agent in the reaction, with oxygen reoxidiz- ing the nitric oxide. After distillation of the monoacetate in 95* 8% yield and recycling of the catalyst solution, the monoacetate is hydrolyzed with 22 steam and an acid catalyst, yielding 95% of the glycol. The resulting aqueous acetic acid is recycled to the monoacetate reactor for further use:

RCH(OH)CH2OAc + H2 > HOAc + RCH(OH)CH2OH (32) Similar processed developed by Du Pont 23 and Shell rely on a PdCl^Li Salt/ Cu Salt catalyst in acetic acid, with oxygen as the oxidizing agent. Considerable amounts of by-products result with this catalytic system, requiring distilla- tion of the complex mixture. The Kuraray process avoids this problem, and produces only the glycol in 90% overall yield. PPG has licensed the process 24 from Kuraray and may use it in its olefin plant complex in Puerto Rico. HYDR0F0RMYLATI0N

The hydroformylation reaction, commercially known as the Oxo-process, is formally the addition of H and CHO, as H2 and CO, across a double bond. With a cobalt octacarbonyl catalyst, and a 1:1 H,p/C0 gas mixture, 1-alkenes are hydroformylated and then reduced, yielding straight chain alcohols:

RCH=CH2 > RCH2CH2CH0 > RCH2CH2CH2OH (33) At 3000 psi and 110-150°, a 3:1 mixture of straight chain and branched 25 aldehydes is produced. With a 2:1 Hs/cO gas mixture, and a temperature of 150-180 , alcohols can be produced directly, along with some alkane by- product. The alcohols however are more difficult to distill than the 26 aldehydes. Additional problems occur with double bond migration. For a example, the hydroformylation of methyl oleate yields a complex mixture of aldehydes, linear alcohol and branched alcohols with the formyl group attached at positions 6-13 along the carbon chain. 26a With the discovery that HRh(CO) (pph 3 ) 2 catalyzes hydroformylation at 25° and 1 atm, yielding only aldehyde product, 12 interest switched to develop- ing such complexes as industrial, homogeneous, hydroformylation catalysts. Frankel and coworkers have developed such catalysts, and initially investigated the hydroformylation of methyl oleate to a 50:50 mixture of methyl 9-formyl- stearate and methyl 10-formylstearate: 27

CH3 (CH2 ) 7CH=CH(CH2 ) 7 C00CH3 > CH3 (CH2 ) 7 CHoCH(CHo) T 000CH3 + (3*0 "I CH0

CH3 ( CHo ) 7CIIC1I2 ( CII2 ) 7C00CH3 CH0

The reaction is performed with a 1:1 Ho/C0 gas mixture, at 120° and 85O-9OO psi for 3-5 hr, using 500 g of methyl oleate, 2.5 g of 5$ Rh on A12 3 or CaC0 3 and 2.2 g of triphenyl phosphine, with or without 300 ml of a solvent, such a toluene. Atomic absorption and infrared spectroscopy indicated the presence of a soluble Rh-PPh 3 complex in the resulting product mixture. This led Frankel to suggest that the active catalyst is probably the soluble Rh-PPh 3 complex, possibly HRh(cO) (PPh 3 ) 2 , and not the supported rhodium. The reaction mechanism is probably similar to that for the Co2 (C0)a catalyst, involving 16 HCo(C0) 3 as the active catalyst:

Co2 (C0) 8 + H2 ^=^ 2HCo(C0) 4 (35)

IICo(C0) 4 ^=^ HCo(C0) 3 + 00 (36)

CH2

HCo(C0) 3 + RCH-CH2 ^=^ || —>CoH(C0) 3 (3?) RCH

CH2 f

1 —> Co (CO) ^== RCH CH -Co(C0) (38) 1 3 2 2 3 RCH

CO

RCH2CH2-Co(C0) 2 + CO ^=^ RCH2CH2C-Co(C0) 3 (39)

RCH2CH2 C-Co(C0) 3 + Ho > RCH2CH2CH0 + HCo(C0) 3 (ho) The active catalyst in the rhodium system is unknown at present-. Although little use has been found for the formyl derivatives of the fatty esters, several derivatives of the formyl group have been prepared and the products examined for commercial application. Although expansion of the laboratory scale hydroformylation procedure to a full scale plant opera- tion has yet to be undertaken, the costs of a methyl formylstearate plant have I been projected, 28 such a plant to be built as an adjunct to an existing vegetable - 100 - oil processing facility. The only major change in the synthetic scheme would "be the conversion from triphenylphosphine to triphenylphosphite, $L0/lb as compared to 35^/lb. Soybean, safflower and linseed oils, containing complex mixtures of palmitate, stearate, oleate, linoleate and linolenate methyl esters, have been hydroformylated and distilled to a mixture of mono-, di- , and triformyl 2S esters. Reaction of the mono-, di- , and triformyl esters with HC1 and

CH(0CH3 ) 3 in CH3OH or KHS0 4 and ethylene glycol in benzene yielded the di- methyl or ethylene acetal derivatives: 30

0— CP -CH0 > -CH(0CH or -CH 3 ) 2 \ (41) CH;

These acetals show good characteristics as polyvinyl chloride plasticizers. Previously used plasticizers, such as di(2-ethyloctyl) phthalate, can be leached from the plastic and present an environmental problem, as they accumulate in vital animal tissue. The triacetal esters, when used at a 30$ plasticizer concentration level, show less tendency to leach from the plastic than thephthalates, but exhibit poor heat stability and low tempera- ture flexibility characteristics. The diacetals have poorer leaching properties, but exhibit better strength and low temperature characteristics. Obviously, more work remains to be done in this area before the phthalates can be economically replaced.

Oxidation of the mono- , di- , and triformyl esters with O^/WkiO^ in cold 27 acetone or 02/5% Ca suspended in naphthenate, a mixture of alicyclic hydrocarbons, 31 yield a mixture of mono-, di-, and tricarboxylic acid deriva- tives. The acids can be used as plasticizers, 31 or in polyester and polyamide applications. Esterification of the acid mixtures also yield useful deriva- tives. Butyl esters of oleate, linoleate, and linolenate oil mixtures exhibit superior wear, low temperature viscosity and oxidation resistance characteristics as lubrication additives. 32 The additives can meet the rigorous new military specifications (Mil- L- 23699 ) for aircraft turbine engines. Plasticizer applications of the esters have also been proposed. 31 Hydroformylated castor (a mixture of ricinoleic, oleic, linoleic, stearic, and dihydroxystearic acids), safflower, linseed and oleic safflower methyl ester oils have been hydrogenated to a mixture of poly- alcohols with Ha/Raney Nickel?3 Reaction of the polyols with polymethylene polyphenyl isocyanate and a triisopropanol amine catalyst yield polyurethane foams which exhibit excellent humidity shrinkage resistance and compressibility 33 ' 34 strengths. The castor oil foam has demonstrated flame resistance. 35 263 36 Additional polyol applications related to lubrication - and plasticizers have also been proposed. CONCLUSION

Transition metal complexes have become very important as homogeneous catalysts, as they catalyze reactions in an efficient, reproducible, and specific manner. As convertional heterogeneously catalyzed processes have become more expensive, due to increasing costs of starting material, homo- geneous processes, with their inherent distillation costs, have become economically more competitive. One can expect homogeneous catalysts, and polymer supported, transition metal complexes, to find increased application in industrial organic synthesis and catalysis. , - 101 - BIBLIOGRAPHY

1. For a review of homogeneous catalysis see a) J. Halpern, Chem. Eng. News, Oct. 31, 1966, p 68; b) C. A. Tolman and J. P. Jesson, Science, l8l, 501 (1973); c) E. W. Stern, Catal. Rev., 1, 73 (19&7); d) G. N. Shrauzer, "Transition Metals in Homogeneous Catalysis", Marcel Dekker, Inc., New

York, N. Y. , 1971. 2. For a review of industrial homogeneous catalysis see a) F. R. Hartley,

J. Chem. Educ, _5_0, 263 (1973); E. W. Stern in "Origin and Refining of Petroleum", Advances in Chemistry Series, No. 103, American Chemical

Society, Washington, D. C. , 1971, pp 150-152, 198-207; c) G. Szonyi in "Homogeneous Catalysis", Advances in Chemistry Series, No. 70, American Chemical Society, Washington, D. C, 1968, Ch. k; d) L. Friend, L. Wender

and J. C. Yarze, ibid . , Ch. 8. 3- a) J. C Bailar, Jr., Catal. Rev., 10, 17 (197*0; b) G. J. K. Acres, A.

J. Davidson, Chem. Eng. (London), 283 , 1^5 (197*0- 4. J. Smidt, Chem. Ind. (London), $k (1962). 5. F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry", 3rd ed.

Wiley Interscience, New York, N. Y. , 1972, p 785. 6. R. Jira and W. Freisleben in "Organometallic Reactions", Vol. 3? Wiley Interscience, New York, N. Y. 1972, pp 10-27-

7- a) T. Reis, Hydrocarbon Proc, 45_(l0), 171 (1966); a) Chem. Eng. (N. Y. ), Sept. 12, 1966, pp 101-2. 8. P. Remirez, Chem. Eng. (N. Y. ), Aug. 12, I968, pp 9^6. 9- I. I- Moiseev and M. N. Vargaftik, USSR Patent 137,511; Chem. Abstr., 56, 136% (1962). 10. R. A. Gardiner, U. of I. Organic Seminars, I965-I966, Part I, p 9k. 11. J. F. Roth, J. A. Craddock, A. Hershman and F. E. Paulik, Chem. Tech., 600 (1971). 12. J. A. Osborn, F. H. Jardine and G. Wilkinson, J. Chem. Soc, A, l8ll (1966).

13. F. E. Paulik and J. F. Roth, Chem. Commun. , 1578 (1968). Ik. Chem. Eng. (N. Y. ), Nov. 29, 1971, p 64. 15- Chem Eng. News, Aug. 30, 1971, p 19- 16. a) C. W. Bird, Chem. Rev., 62, 283 (1962); b) Ref. Id., pp 152-3- 17- a) Chem Eng. (N. Y. ), Nov. 2^, 1973, P 18; b) Chem. Eng. News, Dec. 3, 1973, P 6. 18. A. J. Gait, "Heavy Organic Chemicals", Pergamon Press, New York, N. Y. 1967, pp 107-110. 19- A. J. Gait in "Propylene and Its Industrial Derivatives", John Wiley

and Sons, New York, N. Y. , 1973, PP 273-290. 20. A. Mitsutani, Chem. Econ. Eng. Rev., _5(3), 32 (1973)- 21. M. Tamura and T. Yasui, Chem. Commun., 1209 (1968). 22. Japan. Patent 19928/1971. 23. E. I. Du Pont de Nemours and Co., Brit. Patent 1,058,995; Chem. Abstr., 66, 9^702t (1967). 2k. Chem. Eng. News, Oct. 25, 1971, p 13- 25- A. J. Chalk and J. F. Harrod, Advan. Organometal. Chem., 6, 117 (1968). 26. a) E. N. Frankel, S. Metlin, W. K. Rohwedder, and I. Wender, J. Amer. Oil Chem. Soc, k6, 133 (1969); b) F. Piacenti, S. Pucci, Mi Bianchi, and P. Oino, J. Amer. Chem. Soc, 90, 6847 (1968); c) M. Bianchi, P. Fredian and F. Piacenti, Chim. Ind. (Milan), _55, 798 (1973)- 27. E. N. Frankel, J. Amer. Oil Chem. Soc, 48, 248 (1971).

. 28. J. P. Friedrich, G. R. List and V. E. Sohns, ibid , 50, 455 (1973)-

. 29- a) E. N. Frankel and F. L. Thomas, ibid , _*4£, 10 (1972); b) E. N. Frankel, F. L. Thomas and W. K. Rohwedder, Ind. Eng. Chem., Prod. Res. Develop., 12, 47 (1973)- - 102 -

30. R. A. Awl, E. N. Frankel, E. H. Pryde and G. R. Riser, J. Amer. Oil

Chem. Soc, 51, 22^ (197*0 • 31. A. W. Schwab, E. N. Frankel, E. J. Dufek and J. C. Cowan, ibid., k%., 75 (1972).

• 32. E. J. Dufek, W. E. Parker and R. E. Koos, ibid . , 51, 351 (197*0

33. C. K. Lyon, V. H. Garrett and E. N. Frankel, ibid . , 51, 331 (197*0-

3*+. a) T. H. Khoe, F. H. Otey and E. N. Frankel, ibid . , k$ , 6l5 (1972);

. b) T. H. Khoe, F. H. Otey, E. N. Frankel and J. C. Cowan, ibid , j>0, 331 (1973). 35- C. K. Lyon and T. H. Applewhite, J. Cell. Plast., 3, 91 (1967). 36. G. V. Ferguson, Chem. Ind. (London), k-51 (1965).

- 103 - REACTIONS OF 1-AZIRINES

Reported by David B. Reitz February 2k, 1975 " 1-Azirines (2H-azirines) have been known since 19 32 1 and have been reviewed. 2 6 2-Azirines have been postulated as reaction intermediates, 7,a however, none have been isolated. 1-Azirines have been prepared by vapor phase pyrolysis of vinyl azides, 2 ' 9 ' 10 oxidation of N- aminophthalimide in the presence of alkynes, 7 pyrolysis of 8 l-phthalimido-l,2,3-triazoles, photolysis of vinyl azides, 11 ' 12 photolysis of 13 14 isoxazoles, thermolysis of isoxazoles, and most conveniently by the thermolysis of vinyl azides employing a tertiary amine catalyst in toluene at reflux. 15 1-Azirines react intra- and intermolecularly with dipolarophiles, 1,3-dipoles, dienes, and electrophiles (Table i). They are valuable intermediates for the preparation of heterocyclic compounds.

A or hv sensitized or direct (via triplet direct (via singlet) N + - [PhC=CHR] [PhC=N-CHR] 'C=X vinyl nitrene nitrile ylide

Table I. Reactions of 1-Azirines

(eactants Conditions Product Yields References

1, R=CH0 hv, G<313 nm) oxazole >85$ 16-19 1, R=CH0 hv, (33^ nm) isoxazole >95$ 16-19 1, R=CH0 200°, benzene isoxazole 77-85$ 18,19 1, R=CHNPh hv, (Vycor) imidazole "high" 19

1, R=CHNPh I38 , xylene pyrazole "high" 19 1, R-CHCHR-l hv, ( Vycor) pyrrole 55-98$ 19,20, (te) 1, R-CHCHPh hv, ( Cor ex) benzazepine 80-85$ 21 1, R1CCR2 hv, (Pyrex) pyrrole 40$ 22 1, R^CO hv, (Pyrex) oxazoline 35-95$ 2>28 2 3 1, R^RgCS hv, (Pyrex) thiazoline 25-30$ A =26, A 27 1, co2 hv, (Pyrex) oxazoline- 5-•one 4o$ 28 1, CHRiCHRs hv, (Pyr_e_x) pyrroline 80-85$ 26,29- 31 hv. (Ni -Co dimer 20- 32-33 h i ) k% dimer, NaOCH3 65 , methanol pyrimidine 80$ 33 1, PhCSNCO 25°, xylene pyrimidone 3^ 6 5H JL, diphenylisobenzofuran 111 , toluene tetracyclic adduct 65-90$ 35 1, cyclopentadienones 111°, (-CO) 3H- azepine 48-87$ 36 1, diphenylcyclopropenone 111°, toulene 4-pyridone 27- 50$ 37 1, acetone HCIO4 oxazoline 39$ 38 1, PhCOCl 5°, acetone oxazole 75$ 39

1, aziridines 80 , benzene diazabicyclohexane 73-81$ ko 1, PhCClNCH2Ar 0°, PhH,NEt 3 diazabicyclohexene 28$ hi J[, IN H2S0 4 75$ dioxane pyrazine 35$ 12 1- k- 1, benzoquinone s hv, (Pyrex) isoindole-dione 33-^3$ k-2 ,

- 104 - BIBLIOGRAPHY

1. P. W. Neber and A. Burgard, Ann., kgl, 28l (1932). 2. G. L'abbe, Chem. Rev., 6£, 361 (1969). 3- F. W. Fowler, Advan. Heterocycl. Chem., 1£, 4 5 (1971). h. N. R. Bertoniere and G. W. Griffin, Org. Phot o chem. , 3, 169 (1973). 5- A. Padwa, M. Dharan, J. Smolanoff, and S. I. Wetmore, Jr., Pure Appl. Chem., £5, 269 (1973). 6. P. Claus, et al. , ibid . , 33, 3V7 (1973). 7- D. J. Anderson, T. L. Gilchrist, G. E. Gymer, and C. W. Rees, J. Chem. Soc, Perkin I, 550 (1973). 8. T. L. Gilchrist, G. E. Gymer, and C W. Rees, ibid . , 555 (1973). 9. G. Smolinsky, J. Amer. Chem. Soc, 83, 4483 (1961). 10. G. Smolinsky, J. Org. Chem., 27, 3557 (1962). 11. J. Ciabattoni and M. Cabell, Jr., J. Amer. Chem. Soc, 22* 1**82 (1971). 12. T. A. Foglia, P. A. Barr, and G. Maerker, J. Amer. Oil Chem. Soc, ig., 4l4 (1972). 13- T. Nishiwaki and F. Fujiyama, J. Chem. Soc, Perkin I, 1456 (1972). 14. T. Nishiwaki, T. Kitamura, and A. Nakano, Tetrahedron, 26, 453 (1970 )• 15. M. Komatsu, S. Ichijima, Y. Ohshiro, and T. Agawa, J. Org. Chem., ^8, k3kl (1973). 16. B. Singh and E. F. Ullman, J. Amer. Chem. Soc, 8£, 69II (1967).

17. B. Singh, A. Zweig, and J. B. Gullivan, ibid . , ^C H99 (1972).

18. M. Maeda and M. Kojima, Chem. Commun. , 539 (19737- 19. A. Padwa, J. Smolanoff, and A. Tremper, Tetrahedron Lett., 29 (1974). 20. K. Isomura, M. Okada, and H. Taniguchi, Chem. Lett., 629 (1972). 21. A. Padwa and J. Smolanoff, Tetrahedron Lett., 33 (1974). 22. H. Giezendanner, M. Marky, B. Jackson, H.-J. Hansen, and H. Schmid, Helv. Chim. Acta, 5£, 745 (1972). 23. H. Giezendanner, H. Heimgartner, B. Jackson, T. Winkler, H.-J. Hansen, and H. Schmid, ibid., 56, 26ll (1973).

24. B. Jackson, M. Marky, H.-J. Hansen, and H. Schmid, ibid . , ££, 919 (1972).

25. W. Sieber, P. Gilgen, S. Chaloupka, H.-J. Hansen, and H. Schmid, ibid . 56, 1679 (1973). 26. A. Padwa, D. Dean, and J. Smolanoff, Tetrahedron Lett., koQf (1972). 27. B. Jackson, N. Gakis, M. Marky, H.-J. Hansen, W. von Philipsborn, and

H. Schmid, Helv. Chim. Acta, _5£, 916 (1972 ). 28. A. Orahovats, B. Jackson, H. Heimgartner, and H. Schmid, ibid . , j>6, 2007 (1973). 29. A. Padwa, M. Dharan, J. Smolanoff, and S. I. Wetmore, Jr., J. Amer. Chem. Soc, ££, 1945 (1973).

1

30. A. Padwa and J. Smolanoff, ibid . , 22; 546 (1971). 31. A. Padwa and S. I. Wetmore, Jr., J. Org. Chem., 32, 1396 (197*0-

32. A. Padwa, J. Smolanoff, and S. I Wetmore, Jr., ibid . , 38, 1333 (1973). 33* A. Padwa, M. Dharan, J. Smolanoff, and S. I. Wetmore, Jr., J. Amer. Chem. Soc, £5, 195^ (1973). 34. V. Nair and K. H. Kim, J. Org. Chem., 3£, 3763 (197*0.

. 35- A Hassner and D. J. Anderson, ibid , 39, 2031 (197*0. 36. D. J. Anderson and A. Hassner, J. Amer. Chem. Soc, 93, 4339 0-971)- 37- A. Hassner and A. Kascheres, J. Org. Chem., J57, 2328~Tl972). 38. N. J. Leonard and B. Zwanenburg, J. Amer. Chem. Soc, 8£, 4456 (1967).

39. F. W. Fowler and A. Hassner, ibid . , £0, 2875 (1968).

40. K. Matsumoto and K. Maruyama, Chem. Lett., 759 0-973) • 41. N. S. Narasimhan, H. Heimgartner, H.-J. Hansen, and H. Schmid, Helv. Chem. Acta, 56, 1351 (1973). 1+2. P. Gilgen, B. Jackson, H.-J. Hansen, H. Heimgartner, and H. Schmid,

ibid . 263*)- , 5L (197*0. *+3- R. Huisgen, H. Gotthardt, and H. 0. Bayer, Angew. Chem. Int. Ed. Engl., I, 135 (1964). -105-

SYNTHETIC ENZYMES: APPLICATION OF SOLID PHASE PEPTIDE SYNTHESIS TO THE PREPARATION OF SELECTIVELY BLOCKED PRECURSORS

Reported by: H. Mitchell Rubens tein March 5; 1975 INTRODUCTION Preparation of enzymes with known primary sequences is one of the most challenging problems in synthetic chemistry. If the synthetic enzyme obtained has specific activity comparable to that of the native enzyme, then chemical methods can be employed to synthesize analogues which enable determination of the effects of substitution or deletion of amino acids on catalytic activity. The formation of peptide linkages between two amino acids requires protected amino acids to prevent multiple additions of amino acid to a. poly- peptide when only one new peptide linkage is desired. Side change functional groups (residues) must also be protected from unwanted side reactions. Many procedures have been developed, but since i960, Merrifield's method of solid " phase peptide synthesis (SPPS) 1 3 has been most successful. 'This general 4 method is outlined in Scheme 1. SPPS has been automated to allow faster polypeptide synthesis and relieve experimenters from thousands of manipulations.

II R

I i BN-C-COOH + Cl-CH—resin H J STEP 1. INITIATION l BN-C-C-O-CHo-resmML .

I II STEP 2. DEPR0TECTI0N OF n'^-terminal U * II. .N-C-C-O-CHo-resm '"

I H STEP 5. ELONGATION H R'O H R III .BN~C-C-N-C-C-0-CH.,-resin

H STEP h. TERMINATION R'O H R

II ! II H2PN-C-C-N-C-C00H + X-CHo-resin

I I H H

Scheme 1. SPPS. B = PROTECTING GROUP, R = RESIDUE SPPS can be applied to the synthesis of enzymes in at least two different ways. The procedure shown in Scheme 1 is known as the stepwise addition 5 G 7 " 10 method. > A newer method, convergent fragment condensation, can use SPPS to form fragments which could then be coupled to form the enzyme. This review will discuss the progress on the fragment condensation method. STEPWISE ADDITION VS. FRAGMENT CONDENSATION As of this writing, two enzymes have been synthesized by the stepwise 5 11 addition of amino acids (AA), ribonuclease A (124 AA) * and lysozyme (129 PA). 6 This method requires manual esterification of the first amino acid (C-terminus) to the resin and addition of the remaining amino acids, -106-

one at a time, as directed by an automated program, until the molecule is complete. Its main advantage is simplicity of operation, but it possesses two significant disadvantages. 1) In the syntheses of ribonuclease A and lysozyme, the chain was allowed to grow to a length of over 100 AA prior to any purification other £ than simple washing. Thus high molecular weight impurities (incomplete chains) were present after cleavage from the resin. For smaller poly- peptides (10-20 residues) countercurrent distribution is an effective means of purification, but it has not been successful in separating large molecules. Merrifield was able to purify ribonuclease A by the action of trypsin, an enzyme which hydrolyzes most peptide bonds but does not hydrolyze ribonuclease A. A substantial amount of impurities was removed by the action of trypsin. Ammonium sulfate precipitation of impurities and column chromatography were also utilized to purify the synthetic enzyme to a specific activity i&fo that of native ribonuclease A. If chromatography had been the sole means of purification, the specific activity would have been only $fo that of the native enzyme. In the synthesis of enzyme analogues, purification techniques such as tryptic digestion and ammonium sulfate precipitation would not be applicable. The synthesis of lysozyme was also plagued with purification problems. In addition to the high molecular weight impurities, due to incomplete chain formation, other high molecular weight by-products were formed by degradation of certain residues during deprotection. The synthetic material was purified to a specific activity 2-~5% that of the native material. 2) In the synthesis of an analogue or analogues, the ability to substitute amino acids in any location in the molecule would be desirable. Substitution close to the C-terminus by the stepwise addition method would require a separate synthesis for each analogue, making the study of a series of substitutions exceedingly laborious. Because of these problems with the stepwise addition method, convergent fragment condensation, has been suggested as an alternative. In this procedure smaller, more easily purified fragments would be prepared and then # condensed to build the protein. Substitution of amino acids would be required only in small, easily prepared fragments. Much current research is directed toward methods for efficient synthesis of these fragments. SYNTHESIS OF SELECTIVELY BLOCKED FRAGMENTS The synthesis of fragments by SPPS requires their removal from the support without loss of protecting groups, since the loss of residue pro- tecting groups promotes side reactions during coupling. Previously, fragments were prepared in solution by classical methods. SPPS has not been used for preparation of fragments since conditions required for cleavage from the resin (hydrogen bromide in trifluoroacetic acid (TFA), or liquid hydrogen fluoride) would also remove blocking groups. It is not practical to utilize blocking groups which are more resistant to these conditions. Instead, new methods are being developed for removal of the blocked fragment, and for different linkages to the resin. The linkage of the fragment to the resin must be weaker than that to the residue protecting a group, yet stronger than that to the N - protecting group. NEW RESINS The first method introduced for the synthesis of selectively blocked fragments is based on the development of new resins 12 prepared from the chloromethylated Merrifield resin (Scheme 2). The p-alkoxybenzyl resin (l) has two advantages over the conventional Merrifield resin. -107-

1) The first amino acid is attached to the polymer backbone by the action of dicyclohexylcarbodiimide (DCC). Previously, the first amino acid was attached to the Merrifield resin by refluxing the dicyclohexylamine salt in ethanol or ethyl acetate for 2k to 65 hours. The current method requires only 150 minutes at 25°. 2) The peptide chain can be removed by the treatment of the resin with 50$ TFA for 30 minutes. The carbobenzoxy protecting group (z) is stable under such conditions (but is removed by hydrogen bromide in acetic acid or hydrogenation) and less than 0.1$ racemization was found to occur. The resultant fragment has a free C-terminus. With the p-alkoxybenzyloxy- carbonylhydrazide resin (2) an amino acid can be added in the presence of DCC and the peptide can be removed by 50$ TFA. The resultant fragment has a C-terminal hydrazide.

methyl-U-hydroxybenzoate

/-^ Na0CH3 y—v __ ®-Q)-CII2 Cl + ®-

h-hydroxybenzyl alcohol LiAUi 4 Na0CH 3

®-<^>-CHoO -^)- CH2 0H 1

(o)-oocci

®-©-CH2 0-@-CH ,,00C0^)

H2NWH 2

®-u)/- a{2 o-/oVcH 2 oocN(H)NH;

Scheme 2.

However, these resins have a serious flaw: The N-terminal end of the fragment requires the use of the carbobenzoxy protecting group since other protecting groups are labile to the cleavage conditions. Since this group is not easily removed, further addition at the N-terminus is not possible. Therefore, the fragments produced in this fashion are only suitable as precursors to the N-terminal end of a polypeptide. TFANSESTERIFICATION Transesterification with dimethylaminoethanol at room temperature 13 provides a mild way to cleave the ester bond which holds the fragment to the resin. The dimethyaminoethyl ester can be hydrolyzed with aqueous dimethyl- formamide (DMF), allowing the use of easily removable N^-protecting groups (e.g. 2-(4-biphenylyl) isopropyloxycarbonyl (Bpoc), removed by 1$ TFA in . . -108- methylene chloride, and t-butyloxycarbonyl (Boc), removed by anhydrous HCl) Although transesterification is a significant step towards the synthesis of selectively blocked fragments, it has drawbacks. Even peptides with unhindered C-terminal amino acids such as alanine are cleaved from the resin in low yield (65$), and hindered amino acids (e.g. proline, valine) undergo little or no esterification after four days. Racemization of C-terminal phenylalanine during transesterification is also a problem. Racemization is at acceptably low levels for the other amino acids (Table l) . Since simple alpha-hydrogen removal does not explain the different levels of racemization, a second mechanism has been advanced 14 15 to explain the lability of phenylalanine. ' This mechanism, which has been proven in other systems, 15 involves the formation of an oxazolone derivative under basic conditions (Scheme 3)«

H racemization

I + C-R QHRO -HX.-HX^ ? C-R -H N N ^ T~ -C-N-C-C-X C C=0 +H ~™C. C-0 +HX — ^0 ^0^ H

Scheme 3- X = leaving group

Phenylalanine, cysteine, tyrosine, serine, threonine, and aspartic acid are, thought to be labile to racemization via this mechanism. Recently, the problem of slow transesterification with dimethyl- aminoethanol has been resolved by the addition of a catalytic amount of 3 T thallous ethoxide16 (Sodium eth oxide does not work as well). Various model peptides were employed m the tests and the results can be seen in Table 1.

TABLE 1. Transesterification with 2-Dimethylaminoethanol

Thallous Protected peptide-resin alkoxide Time % Yield Racemization (equiv) (hr) (0

Boc-L-Leu-L-Ala-O-resin None 72 65+5 0.5+0.1- 0.1 5 .00 0.6

Boc-L-Ala-L-Phe-O-resin None h't> 37 2.8 0.1 5 80 k.-5

Boc-L-Leu-D-Val-0-resin 0.15 2.5 70 0.6 0.1 16 66 0.8 0.25 16 85 0.8 Boc-L-Pro-0-resin None 96 - 0.1 21 96

The data indicate a greatly improved reaction rate with thallous ethoxide, and steric hindrance no longer appears to be a problem as evidenced by valine and proline. Racemization remains a problem for phenylalanine, and presumably precludes the use of this method for phenylalanine, cysteine, tyrosine, serine, threonine, and aspartic acid as C-terminal residues. However, these amino acids are not racemized in other positions in the fragment -109-

Transesterification cleaves peptides from the resin in high yields

(Table l) , and racemization can be controlled to acceptable levels with the correct choice of a C-terminal amino acid. Moreover, more easily removed N^-protecting groups than carbobenzoxy may be used, which allows a synthesis of fragments deprotected at the N position. APPLICATION These new methods for synthesis and release of selectively blocked fragments have had limited applications. The new resins have been used to produce selectively blocked nonapeptides and to form a tetrapeptide, which was condensed with a dipeptide to form the hexapeptide. Table 1 contains all of the selectively blocked fragments that have been produced by the transesterification method. More applications have not yet been reported because the original report appeared in August, 197*+ • The possible applications are many. For example, the transesterification and p-alkoxybenzyl resin methods might be used for a docosapeptide corresponding to positions 18 through 39 of human ACTH, adrenocorticotropin hormone, (human type corticotropin-like intermediate lobe peptide, CLIP) by condensation of selectively blocked fragments. An entirely solution phase synthesis of this docosapeptide was reported recently. 9 It has the sequence: Arg-Pro-Val-Lys-Val-Tyr-Pro- Asn-Gly-Ala-Glu-Ser-Ala-Glu-Ala-Phe-Pro-Leu-Glu-Phe-OH. Kawatani and 9 co-workers dissected this polypeptide into four subunits (Scheme h) .

Z-Arg(B)-Pro-0H \ B-Val-Lys(z)-Val-Tyr-Pro-Asn-Gly-OH k B-Ala-Glu(B)-Asp(B)-Glu(B)-Ser-Ala-Glu(B)-Ala-Phe-Pro-OH £ H-Leu-Glu(B)-Phe-OBenzyl 6

Scheme k.

Fragments k and ^ were synthesized by the laborious condensation of smaller subunits (Schemes 5 and 6).

B-Asn-OH + H-Gly-OBenzyl

1 H-Asn-Gly-OBenzyl

Z- Pro- OH—

H-Pro-Asn-Gly-OH

B-Val-Tyr-NHNH2

H-Val-Tyr-Pro-Asn-Gly-OH

B-Val-Lys ( Z ) -NHNILj 4

S cheme 5 • -110-

Z-Ala-Fhe-NHNH2 \- H-Pro-OH

H-Ala-Phe-Pro-OH

B-Glu(B)-0H

H-Glu(B)-Ala-Phe-Rro-OH

B-Ser-Ala-NHNH2

I-Ser-Ala-Glu(B)-Ala-Phe-Pro-OH

B-G1u(b)-0H

H-Glu(B)-Ser-Ala-Glu(B)-Ala-Phe-Pro-GH

B-Asp(B)-0H

H-Asp(B)-Glu(B)-Ser-Ala-Glu(B)-Ala-Phe-Pro-OH

B-Glu(B)-0H

H-Glu(B)-Asp(B)-Glu(B)-Ser-Ala-Glu(B)-Ala-Phe-Pro-OH

B-Ala-OH

1

Scheme 6.

However, the methods discussed earlier in this review might simplify the synthesis of these fragments. Fragments 2. and J+ might be synthesized as one unit. The p-alkoxybenzyl resin could be used since this nonapeptide would be an N-terminal fragment ; this would eliminate the complicated

synthesis employed by Kawatani and co-workers. Since S_ contains proline as the C-terminal AA, this fragment could be formed on the Merrifield resin with subsequent removal by transesterification. The synthesis of Jo could be accomplished by solution phase techniques, and the fragments could then be combined as previously reported. 9 OTHER CONSIDERATIONS

Although it is not the purpose of this review to discuss in detail methods for coupling amino acids and peptides, or procedures for deprotection, some mention should be made of the difficulties in these steps of peptide synthesis. They are well illustrated in the solution synthesis of CLIP.

Coupling of J| to Jo produced a tridecapeptide, which was then coupled to h to produce an eicosapeptide. Although the products were formed in good yield with respect to the larger fragment (80^ and lh%) , large excesses of 6 (~ 10X) and h_ (~ hx) were required. The requirement of a large excess of one of the two peptides which are to be coupled is a major disadvantage to the selectively blocked fragment method. Further work is needed to overcome this difficulty. Although deprotection is not limited to the fragment synthesis method, it is a low yield step which severely reduces the final yield of a protein. In the synthesis of the docosapeptide, deprotection produced the desired peptide (CLIP) in ~$6% yield from the protected material, the worst . step in the entire synthesis. More easily removable residue protecting groups are needed. -111-

CONCLUSION The stepwise addition method fails in the synthesis of large poly- peptides (e.g. enzymes) due to serious purification difficulties. Emphasis has shifted to the synthesis of selectively blocked fragments, which can then be condensed to form the complete molecules. Fragment condensation should be the favored method in the synthesis of enzyme analogues. To date, these fragments have been produced by very long, costly, and complicated solution phase techniques. This review has explored the possibility of the synthesis of selectively blocked fragments by the solid phase technique. The introduction of new resins and a new method for removal of the peptide from the resin should facilitate the synthesis of these fragments. Although an alternative route to the synthesis of CLIP has been suggested in this paper, the viability of these techniques still must be determined. However, the results to date have been encouraging, and further development of these methods should make easier the chemical synthesis of enzymes and their analogues.

BIBLIOGRAPHY

1. R. B. Merrifield, Advanc. Enzym. , 32, 221 (1969). 2. J. M. Stewart and J. D. Young, Solid Phase Peptide Synthesis, W. H. Freeman and Co., San Francisco, Calif. (1969).

) 3. M. Fridkin and A. Patchornik, Ann. Rev. Biochem. , _j+3, 419 (197 +). h. R. B. Merrifield, J. M. Stewart, and N. Jernberg, Anal. Chem. , ~~38 , 1905 (1966).

5. R. B. Merrifield, Intra-Science Chemistry Reports , _5, 183 (1971). 6. J. J. Sharp, A. B. Robinson, and M. D. Kamen, J. Amer. Chem. Soc, 95, 6097 (1973). 7. R. Hirschmann, Intra-Science Chemistry Reports, _5, 203 (l97l)»

8. H. Watanabe, et al . , Chem. Pharm. Bull., 22, 1889 (197U). 9. H. Kawatani, F. Tamura, and H. Yaj'ima, ibid., _22, 1879 (I97U). 10. R. G. Denkewalter, D. F. Veber, F. W. Holly, and R. Hirschmann,

J. Amer. Chem. Soc, 91, 502 (1969); R. G. Strachan, et al ., ibid . , 91,

503 (1969); S. R. Jenkins et al . , ibid., 91, 505 (1969); D. F. Veber, et al ., ibid., 91, 506 (1969); R- Hirschmann, et al ., ibid., 91, 507 (1969). 11. B. Gutte and R. B. Merrifield, J. Amer. Chem. Soc, 91, 501 (1969). 12. S. S. Wang, J. Amer. Chem. Soc, 95, 1328 (1973).

13. M. A. Barton, R. U. Lemieux, and J. Y. Savoie, J. Amer. Chem. Soc, 95 , 4501 (1973). 1^. N. F. Albertson, Organic Reactions, 12, 168, John Wiley and Sons, Inc., New York, 1962. 15. M. Goodman and C. Glaser, Peptides: Chemistry and Biochemistry, Marcel Dekker, Inc., New York, 1970, p 267.

16. J. Y. Savoie and M. A. Barton, Can. J. Chem., _5_2, 2832 (197*0. 17. A. Loffet, Experentia, 23, ^06 (1967).

- 112 - AROMATICITY AND THE BRIDGED ANNULENES

Reported by Lance A. Christell March 13, 1975

1 ' 2 3 The present emphasis on the synthetic and theoretical aspects of nonbenzenoid aromatic chemistry has involved work with the annulenes. Bridged annulenes are polycyclic compounds in which the largest ring (or the largest

tt- cycle) contains ten or more tt electrons in a fully unsaturated periphery. These compounds have been synthesized to test the concepts of aromaticity. The basis of the work is simple Hiickel MO theory and the 4n+2 rule. It

states that fully unsaturated, monocyclic hydrocarbons with 4n+2 tt electrons will be more stable than the analogous acyclic polyenes. The corollary kn rule predicts these cyclic systems will be less stable than the acylic analogues due to a triplet ground state. (More refined MO theories predict a Jahn- Teller distortion will split the nonbonding orbitals and provide the molecule a singlet ground state and an increase in stability). 4 Many methods have been used to test the predictions of these rules with varying degrees of success. Comparing the reactivity of cyclic polyenes with the reactivity of benzene has proven unsatisfactory due to benzene's unique stability. The relative heats of hydrogenation of cyclic polyenes are excellent measures of their stabilities, but the method is laborious and has been used sparingly. 5 The electronic properties of the [4n+2] and [^njannulenes are predicted to be very different. The derealization of electrons in the [kn+2] annulenes gives rise to a diamagnetic ring current, which introduces a magnetic anisotropy to the molecuJes. The anisotropy can be measured by the diamagnetic suscepti- 7 bility exaltation, 6 the molecular Zeeman effect in microwave spectroscopy, or pmr chemical shifts. 8 Conversely the [^n]annulenes have a paramagnetic contribution to their electronic fields. This affects the chemical shift in a fashion opposite to that of the [ 4n+2] annulene' s diamagnetic current. The pmr chemical shifts are more sensitive to remote electronic effects than cmr chemical shifts. The cmr peak positions are dominated by local electronic effects rather than through- space interactions. 9 Originally the bridged annulene structure was designed to remove the severe steric interactions due to the interior hydrogens present in the higher annulenes. Pi- trans- [10]annulene, (l), possesses this interaction and is as yet unknown. Substitution of a methylene bridge across the 1,6- positions removes this interaction and the resulting compound, 1, 6- methano [10] annulene io (2)"' is well documented.

H H

[10]ANMJLENES The majority of work involving bridged [10]annulenes has been done by Vogel and co-workers. In the pmr spectra of 3- substituted 1,6- methano [10]-

annulenes, where the R substituent was varied from CH20H to C00CH3 and NH2 , (Scheme l) an exact parallel can be traced in the pmr chemical shifts of the corresponding protons of the analogous mono substituted benzenes and the proton of the [10]annulene. 1:L Scheme 1

:CC1; 1) Br2 m 2) KOH

(h3c) 3cok

Replacement of the methylene bridge protons with methyl and cyano groups has a great effect on the molecular tt framework. With variable temperature cmr Vogel has observed the tautomerization depicted below:

R R R R 1 C-1,6 0-11 T°C H-2 H-3 H H 11^.6 34.8 35 7.27 6.9 H H 113-7 3^.2 -110 CH 3 CH3 79-8 16.2 107 6.11 6.2; CH3 CH3 89.3 20.0 -110 CH 3 CN 67.I 13.6 35 CH 3 CN 62.6 11-3 -80

18.1 21.6 //

Cyanide is considered a good stabilizer of three- membered rings, and the cmr spectra indicate that it causes the electronic environment of carbons 1 and 6 to become more like the cyclopropyl tautomer shown. [12]AMfULEM;S

The existence of definite bond alternation or a shift in the pmr spectrum to higher field due to a paramagnetic ring current is considered another confir- mation of the theory of aromaticity. Vogel and co-workers have synthesized two 13 14 compounds, l,6-methano[12]annulene (3), and l,7-methano[12]annulene (j+), which illustrate these different effects. The first of these compounds exhibits definite bond alternation and the predominate molecular conformation is the syn-boat structure £&. This was determined from the pmr spectrum in which proton 'a 1 appears at 6 7-00, and proton 'b' at 6 2.29-

b a - 111+ _

The synthesis of ^ is shown in Scheme 2. Scheme 2

NpCHCOpHs NaOH y Cu H3COH COOH

100-120°

In the 1,7-annulene k, the ring is slightly more rigid and the pmr spectrum exhibits the effect of a paramagnetic ring current. The ring protons appear at higher field than the bridge protons (6 5*1- 5* 8 compared to 6 6.1). 15 Trost and co-workers have prepared the more rigid pyracylene (j?). Comparison of its pmr spectrum with that of the dihydro analogue 6 shows an

average upfield shift of approximately 1 ppm for the hn rr system.

6 6.01 6 3.^9

6 6.52

6 6.01 6 7-0^

An analogous compound with a dinitrogen central bridge has been. prepared by 16 Paudler and co-workers as depicted in Scheme 3« Compound 8 is again a 12 n system with the outer carbon periphery planar to within 0.002 A.

Scheme 3

HpfflH; ^

Pd/C;PhN0p ^ or # DDQ - 115 - The insertion of nitrogen bridges in two tricyclic systems has been accomplished 17 in the compounds 9 and 10, synthesized by Paudler and co-workers, and by Flitsch and MuterT8 respectively Both £ and 10 exhibit pmr spectra which imply a paramagnetic ring current. 6 4.77

6 3-86

6- 5-22-6.52 3-25

Quite a few doubly-bridged [l^]annulenes have been synthesized and have 1 been proven to be aromatic. The dioxa-bridged (ll) and various joined 20 ' 21 dimethano- bridged annulenes (12, l^) have been shown to remain relatively planar.

11

The crystal structure of a carbethoxy derivative of Ik has been solved, 22 and it proves that the nonaromatic character of lk is not caused by its nonplanarity, but by the increased angles between tt bond axes, a,, is the standard deviation 2 of the perimeter bond lengths. Dmax is the largest deviation of any sp carbon atom from the plane through C-2, C-5> C-9, C-12, lax is "the largest misalignment angle between axes of adjacent p-orbitals.

J D (A) $ (deg) Cmpd s max max 11 .006 0A9 2k 12 .013 0.57 28 13 .025 0.75 35 Ik .062 O.69 75

The fully unsaturated compound 15 was obtained from the corresponding ethylideno- bridged compound23 (see Scheme k). Scheme k H

H ' - lib -

The pmr spectrum of 15 exhibits the expected AA'BB 1 pattern at 6 8.66 and 6 8.08, and a singlet at 6 8.06 indicative of aromatic character. Boekelheide and co-workers have developed the general method illustrated in Scheme 5 to synthesize various tetracyclic bridged annulenes of the

general structure l6 in which R can be H, CH3 , C2H5 , or 11-C3H7. The dihydro compound l624 is easily oxidized to pyrene in the air or upon further ^^

Me-S + (MeO)gCHBF, > + cis 1 quant ^/ r^N ^N

+ (MeO) Me S- 2 CHBF4 ^ 2 hv CsHi2 H2 CCl2j,-30° THF, sealed 35$ tube

E

0, -> =^Sz3 <=&=> or ^2> hv R 16

irradiation. Its pmr spectrum was obtained by computer subtraction of the pyrene spectrum from the pmr of the irradiated solution. The two inner diamagnetic protons exhibit a singlet at 5 -5.^9, verifying the presence of a ring current. 25 Jutz and Schweiger have reported an improved synthesis of 17, originally 26 reported by Anderson, MacDonald, and Montana. They react b, 5-trimethylene- azulene with the per chlorate salt formed from 1- dimethylaminopropene to give IT in 95$ yield. The characteristic eight line pattern for an AB2 system was revealed with a 50- cycle sweep width pmr spectrum.

6 9.08 or 6 9.66

6 9.16

6 9.50 6 8.62

6 9. *K) 9-00 17 6 " - ±±l - Two monocation homologues of pyrene have been recently reported by Murata and co-workers. They found 18 to be a significantly delocalized 27 system in agreement with MO calculations. However, lg, does not behave as a delocalized system, but as a phenanthrene unit coupled with an allylic fragment. 2€ This was deduced from the large range of chemical shifts of 12 compared to 18, especially the downfield shift of the 'allylic' protons. Also the spectrum of Ig. uv was quite unlike that of 18 , but agreed well with the calculated spectrum. Paudler and co-workers have synthesized a [l4]annulene with two nitro- gen atoms in the tt periphery (7, Scheme 3)« The l4rr system fits well in this series, exhibiting a pmr spectrum with peaks at 6 6.82 (H-7,8), 6 6.32 (H-1,2,5,6), and 6 3*28 (E-3,k). Analogously, pyrrole shows pmr absorptions 29 at 6 6.62 (H-2,3) and 6 6.05 (H-l,^-). [l6]ANMJLEWE 30 Ogawa and co-workers have prepared a rigid [l6]annulene with two oxygen bridges of furan rings to force planar ity as outlined in Scheme 6. Scheme 6

LiOMe -> DMF, 70° 25

1) NBS,hv 2) (h3c) 3cok 2 min. 80° (H3C) 3C0H:THF 1:1

The pmr spectrum, which demonstrates a paramagnetic ring current, consists of a peak at 6 11.7 (lH) and a series of absorptions at 6 3«^^-8 (llH). The spectra are identical at k0° and at -78°, indicating that the trans structure does not undergo conformational change on the pmr time scale in that temperature range. [l8]ANmJLEMES

Boekelheide and co-workers, again using the. general method used to 27 synthesize l6, have accomplished the synthesis of 2JL as illustrated in Scheme 7 Scheme 7 1) n-BuLi 20a 2) DMF 20b 68 1o

NaBH 20b - * > 20c " quant .

KBr 20c -> 20d HOAc 93f - 118

Scheme 7 (continued)

Base 20d + HS- HpC H2- SH EfcOH W^W

" // \V// w 1) (MeO) 2 CHBF, " ^ 1) (MeO) 2 CHBF4 2) (H3C) 3C0K 2) (H3C) 3COK 22$ 19fo MeS 3Me \\ //

/-

H II hv / H H // VJ y \ V-

^d6 i // ^ // \ // \ // 21

The pmr spectrum of the inseparable stereoisomers of 20 appears as a series of peaks at 6 8.^+6 to 8 8.8k- for the outer protons and at 6 -5.0 to 6 -5.69 for the four inner protons. The compound is unstable in air, oxidizing to thiacoronene, The results reviewed here for the [12] and [l6]annulenes agree well with the predicted properties. The pmr results indicate that the compounds are either antiaromatic or nonaromatic depending on the particular effect of the bridging atom or atoms. This parallels the results seen in the monocyclic

[tai]annulene series. Hetero atoms inserted in both bridge and tt- cycle positions produce predictable perturbations in the molecules. The [4n+2] series including [10], [lk], aXi ^- [l8]annulenes are aromatic by the pmr chemical shift criterion. The use of various types of bridging groups and the inclusion of hetero atoms in bridge or peripheral positions again produce bridged annulenes which obey the kn+2 rule. The pmr spectra of the bridged annulenes resemble closely those of the corresponding monocyclic annulenes. - 119 - Further work using bridged systems can follow many directions. Synthesis of larger rings will enable tests of whether aromaticity is limited by ring size and if so what occurs at this limit. Variations in the planar ity of the molecules and/or the overlap angles between the rr orbitals in the rings of new bridged annulenes will provide more information about aromaticity. BIBLIOGRAPHY

1. F. Sondheimer, Chimia, 28, 163 (197*0. 2. W. D. Ollis, Chem. Soc. Spec. Publ., 21 (1967); W. 0. Milligan and Robert A. Welch Found., Conf. Chem. Res., XII (1969). 3- A. J. Jones, Rev. Pure appl. Chem., 18, 253 (1968). k. A. Streitweiser, "Molecular Orbital Theory for Organic Chemists," J. Wiley, New York, N. Y., 1961, p 53- 5- R. B. Turner, W. S. Lindsay, and V. Boekelheide, Tetrahedron, 27, 33*kL (1971); C G. Frye and A. W. Weitkamp, J. Chem. Eng. Data, lk, 372 (1969).

6. H. J. Dauben, J. D. Wilson, and J. L. Laity, J. Amer. Chem. Soc, ^0_,

811 (1968), and ibid . , 91, 1991 (1969). 7. W. H. Flygare, Chem. Rev., Jk, 653 (197*0. 8. R. C. Haddon, V. R. Haddon, and L. M. Jackman, Fortsch. Chem. Forsch., 16, 113 (1971). 9. H. Gunther, H. Schmickler, H. Konigshofen, K. Recker, and E. Vogel, Angew. Chem. Int. Ed., 12, 2*0 (1973).

10. E. Vogel and W. A. Boll, Angew. Chem., 76, 78*+ ( 196*0-

11. E. Vogel and J. Sombroek, Tetrahedron Lett., 197*+ , 1627. 12. E. Vogel, H. Gunther, H. Schmickler, and J. F. M. 0th, Angew. Chem. Int. Ed., 12, 570 (1973). 13- E. Vogel, M. Mann, Y. Sakata, and J. F. M. 0th, Angew. Chem. Int. Ed., 13, 283 (197*0- lk. E. Vogel, H. Konigshofen, K. Mullen, and J. F. M. 0th, Angew. Chem. Int. Ed., 13, 281 (197*0. 15. B. M. Trost, G. M. Bright, C Frihart, and D. Britteli, J. Amer. Chem. Soc, 93, 737 (1971). 16. W. W. Paudler, J. L. Atwood, D. C. Harneir, and C. Wong, J. Amer. Chem. Soc, 26, 6132 (197*0. 17- W. W. Paudler and E. A. Stephan, J. Amer. Chem. Soc, %2, kh6Q (1970). 18. W. Flitsch and B. Muter, Angew. Chem. Int. Ed., 12, 501 (1973). 19- D. Ganis and J. D. Dunitz, Helv. Chim. Acta, 50, 2369 (1967). 20. A. Gavezzotti, A. Mugnoli, M. Raimondi, and M. Simonetta, J. Chem. Soc, Perkin Trans. 2, k25 (1972). 21. CM. Gramaccioli, A. Mugnoli, T. Pilati, M. Raimondi, and M. Simonetta, Acta Crystallogr. Sect. B, 28, 2365 (1972). 22. C M. Gramaccioli, A. S. Mimum, A. Mugnoli, and M. Simonetta, J. Amer. Chem. Soc, 95, 31*^9 (1973). 23. E. Vogel and H. Reel, Angew. Chem. Int. Ed., 11, 1013 (1972). 2k. R. H. Mitchell and V. Boekelheide, J. Amer. Chem. Soc, %2, 3510 (1970). 25. C. Jutz and E. Schweiger, Angew. Chem. Int. Ed., 10, 808 (1971). 26. A. G. Anderson, A. A. MacDonald, and A. F. Montana, J. Amer. Chem. Soc, 22, 2993 (1968).

27. I. Murata, K. Yamamoto, and Y. Kayane, Angew. Chem. Int. Ed., 13 , 807 (197*+). 28. I. Murata, K. Yamamoto, Y. Kayne, and H. Ori, Tetrahedron Lett., 135 (l975)« 29. A. Gordon and R. Ford, "The Chemist's Companion," J. Wiley, New York, N. Y., 1972, P. 265. 30. H. Ogawa, M. Kubo, and I. Tabushi, Tetrahedron Lett., 36l (1973).

31. V. Boekelheide, J. Lawson, and R. DuVernt, J. Amer. Chem. Soc, 95 ? 957 (1973). 120 -

SYNTHESIS AND REACTIONS OF AROMATIC K- REGION EPOXIDES

Reported by Richard Amos March IT, 1975 K- region epoxides are compounds possessing an oxirane ring at the 9,10 bond of a phenanthrene type structure (l). The carcinogenicity of polycyclic aromatic hydrocarbons has been recognized for many decades, but only recently have such arene oxides been implicated as the activated metabolites responsible for carcinogenesis. Early workers recognized the importance of the electron rich K-region in determining reactivity and began to use, with reasonable success, MO and valence bond approaches to predict the relative carcinogenic 4 activity of aromatic hydrocarbons. 1 Boyland, in 1950, was the first to propose the arene oxide as a metabolite, 5 and since that time a number of epoxides have been isolated from in vitro metabolism of the parent hydrocar- 6~9 bons. The biological aspects of the problem have been thoroughly re- viewed1 " 3 and will be discussed only briefly here. Chemical models of the enzyme systems have also been discussed. 14

-> i

00CCH3 -C(CH3 )0CH3 7 The synthesis of arene oxides divides conveniently into two areas: K-region 15 and non- K-region. The preparation of the latter has been reviewed. 11} The first 16 reported synthesis of a K-region epoxide was by Hewett in 19^+0, treating 5,6-dihydroxy- 5,6-dimethylchrysene with HC1 gas to give the corresponding epoxide. No further use of this method has been reported. Perbenzoic acid oxidation of aromatic hydrocarbons has been reported to give very low yields of K-region epoxides. 17 ' 18 Newman and Blum reported the first synthetically useful route 19 in 1964. Treatment of o_,o'-biphenyldialdehydes (2) with tris- (dimethylamino)- phosphine gave generally high yields of the K-region epoxides. The method has 2° been extensively applied, 11 most recently in the synthesis of pyrene-^-, 5-oxide. The dialdehyde is accessible by ozonolysis of the parent hydrocarbon or by NaI0 4 oxidation of the cis diol (3) obtained by 0s0 4 oxidation of the hydrocarbon. Harvey has reported21 a complimentary route from the cis diol. Oxidation to the o-quinone (k-) with SO3- pyridine in DMSO, followed by LiALH4 reduction gave the trans diol Tj5)» Treatment with the dimethyl acetal of dimethylformamide afforded the epoxide. The initial 0s0 4 oxidation step can be avoided by a selective reduction of the K-region, followed by Na2Cr2 7 oxidation of the - 121 - dihydroaromatic compound to the o-quinone. 22 A final method consists of the formation of a 2-alkoxy-l,3-dioxolane (6) from a K- region cis diol with

trimethyl orthoacetate, conversion to the halohydrin ester (T) with (CH3 ) 3SiCl, and cyclization to the arene oxide with base. 23 The two principle reactions of arene oxides are isomerization to phenols 1Cr 12 and ring opening by nucleophiles. K- region epoxides generally are much more non-K- region stable than and show slower rates of isomerization ~to the phenol, particularly at pH>7- Acid will rapidly isomerize both. 11,24 25 However, both classes of compounds do show the NIH shift, i.e. migration and retention of ring substituents during isomerization to the phenol. 10 Reactions with nucleophiles closely resemble the nucleophilic opening observed in aliphatic epoxides. 11,s6 in biological systems, arene oxides can undergo isomerization to the phenol, enzymatic hydration to the dihydrodiol, conjuga- tion with glutathione, or covalent attachment to proteins or nucleic acids. 12 The greater the biological lifetime of the arene oxide, the more likely is the last reaction. Since K- region epoxides are more stable, their role in carcinogenesis is more strongly implicated, particularly in light of the recent demonstration of their binding to DNA and RNA. 27 BIBLIOGRAPHY

1. C. A. Coulson, Adv. Cancer Res., 1, 1 (1953). 2. A. Pullman and B. Pullman, Adv. Cancer Res., 3, 117 (1955). 3- L. B. Kier, "Molecular Orbital Theory in Drug Research," Academic Press,

New York (1971 ), pp. 129-136, lkh-lk-7. 4. W. C Herndon, Trans. N. Y. Acad. Sci., 36, 200 (1974). ). 5. E. Boyland, Biochem. Soc. Symp. , j?, 40 (1950 6. D. M. Jerina, J. W. Daly, B. Witkop, P. Zaltzman-Nirenberg, and S. Udenfriend, Biochemistry, £, l47 (1970 ). 7« J- K. Selkirk, E. Huberman, and C. Heidelberger, Biochem. Biophys. Res. Comm., 43, 1010 (1971 )• 8. P. L. Grover, A. Hewer, and P. Sims, FEBS Letters, 18, 76 (l97l). 9- P. L. Grover, A. Hewer, and P. Sims, Biochem. Pharmacol., 21, 2713 (1972). 10. J. W. Daly, D. M. Jerina, and B. Witkop, Experientia, 28, 1129 (1972). 11. D. M. Jerina, H. Yagi, and J. W. Daly, Heterocycles, 1, 267 (1973)- 12. D. M. Jerina and J. W. Daly, Science, 1§5, 573 (1974). 13. F. Oesch, Xenobiotica, 3, 305 (1973). Ik. D. M. Jerina, Chem. Technology, 4, 120 (1973). 15. E. Vogel and H. Giinther, Angew. Chem. Int. Ed., 6, 385 (1967). 16. C L. Hewett, J. Chem. Soc, 293 (l94o). 17. B. L. Van Duuren, I. Bekersky, and M. Lefar, J. Org. Chem., 2£, 686 (1964). 18. E. Boyland and P. Sims, Biochem. J., 90, 391 (1964). 19. M. S. Newman and S. Blum, J. Amer. Chem. Soc, 86, 5598 (1964). 20. B. L. Van Duuren, G. Witz, and S. Agarwal, J. Org. Chem., 39, 1032 (1974). 21. S. H. Goh and R. G. Harvey, J. Amer. Chem. Soc, 95, 242 (1973)- 22. H. Cho and R. G. Harvey, Tetrahedron Lett., 1491 (1974). 23. P. Dansette and D. M. Jerina, J. Amer. Chem. Soc, %6, 1224 (1974). 24. P. Y. Bruice, T. C. Bruice, H. G. Selander, H. Yagi, and D. M. Jerina,

ibid . , 2§, 68l4 (1974). 25. A. J. Swaisland, P. L. Grover, and P. Sims, Biochem. Pharmacol., 22, 1547 (1973). 26. A. M. Jeffery, H. J. C. Yeh, D. M. Jerina, R. M. DeMarinis, C H. Foster, D. E. Piccolo, and G. A. Berchtold, J. Amer. Chem. Soc, 96, 6929 (1974). 27. P- L. Grover and P. Sims, Biochem. Pharmacol., 22, 66l (1973)- - 122 - PHOTOAFFINITY LABELING WITH CARBONYL COMPOUNDS

Reported by Richard L. Neeley April 7, 1975 Many important biological processes involve the interaction of a small organic molecule with a macromolecule. Photoaffinity labeling is the process of covalently attaching a modified ligand or a sufficiently reactive natural 3 ligand to the macromolecule by a photochemical reaction. 1 (Scheme I) Early workers used photogenerated carbenes or nitrenes, generally obtained from odiazoketones or aryl azides, to form the covalent bond. 4 While these methods are still of value, more recent methods have been introduced which give fewer rearrangement products than odiazoketones and less selective insertion reactions than the nitrenes derived from aryl azides. 5 Among these new methods is the use of carbonyl- containing ligands both as covalent attachment agents 7 11 and as sensitizers for photo- oxidative modification of the macromolecule. 10 15

Scheme I

hr^ ,-n\ L-x, L-x /7 7r w /T rr /7T rr * W W0 ^ & reactive receptor ligand

Several factors in the design of active- site- directed reagents favor carbonyl- containing compounds as photoaffinity labeling reagents. 16 Many natural compounds can be studied without modification. This is advantageous since high binding affinity of the ligand for the macromolecule is essential and even small changes in the structure of the ligand can result in greatly lowered affinity. 17 High non- selective reactivity of the photoactivated ligand ~ 1,e ' is essential. 18 2° Triplet state carbonyls can insert into C-H bonds of amino acids and thus do not require the presence of reactive regions in the a,P1 25 macromolecule itself. (Scheme II ) Photolytic destruction of proteins can be prevented by using carbonyl compounds which absorb at sufficiently long wavelengths. 7 9,11> 25 Carbonyl compounds are stable in their ground 2 states and do not react with water in their excited states. ~ In cases where covalent bonds are not formed, photosensitized oxidation of nearby amino acid residues often occurs. Identification of the specific residues " modified allows determination of the amino acids present at the active site. 13 15

Scheme II

Cil A»-Y 3 **» ,

- 123 - BIBLIOGRAPHY

1. J. R. Knowles, Accounts Chem. Res., 5, 155 (1972). 2. D. Creed, Photochem. Photobiol., 1£, 459 (197k). 3. J. A. Katzenellenbogen, Ann. Rep. Med. Chem., 9, 222 (1974). 4^ 4. G. W. J. Fleet, J. R. Knowles, and R. R. Porter, Nature, 224, 511 (1969);

T. C. French, E. G. David, and J. M. Buchanan, J. Biol. Chem., 238 , 2171, 217S 2186 (1963); T. L. Gilchrist, Chem. Ind. (London), 88l (1973); W. G. Han stein and Y. Hatefi, J. Biol. Chem., 249, 1356 (197*0; H. J. Kiefer, J. Lindstrom, E. S. Lennox, and S. J. Singer, Proc. Nat. Acad. Sci. U. S., 67, 1688' (1970); B. Levenberg, I. Melnick, and J. M. Buchanan, J. Biol. Chem., 255, 163 (1957); R. W. Rosenstein and F. F. Richards, J. Immunol., 108, lk6j (1972); A. Singh,

E. R. Thornton, and F. H. Westheimer, J. Biol. Chem., 237 , PC 3006 (1962); P. Talalay and A. M. Benson, "Enzymes," Vol. 6, 3rd edition, Academic Press,

Inc., New York, N. Y. , 1972, p. 591. 5- R. Breslow, A. Feiring, and F. Herman, J. Amer. Chem. Soc, 96, 5937 (197*0; N. Burridge and J. E. Churchich, J. Biol. Chem., 2_4^>, 6258 U970); A. M. Frishchauf and K. H. Scheit, Biochem. Biophys. Res. Commun., 5£, 1227 (1973); S. S. Hixson and S. H. Hixson, J. Org. Chem., 37, 1279 (1972); F. Sawada and N. Kanbayashi, J. Biochem., jh, 459 (1973); R. A. G. Smith and J. R. Knowles, J. Amer. Chem. Soc, 21) 5072 (1973); E. Scoffone, G. Galiazzo, and G. Jori, Biochem. Biophys. Res. Commun., 38, 16 (1970). 6. A. E. Ruoho, H. Kiefer, P. E. Roeder, and S. J. Singer, Proc. Nat. Acad. Sci. U. S., 70, 2567 (1973). 7. R. S. Antonof and J. J. Ferguson, Jr., J. Biol. Chem., 249, 3319 (1974). 8. R. E. Galardy, L. C. Craig, J. D. Jamison, and M. P. Printz, J. Biol. Chem.,

249 , 3510 (1974). 9. R. J. Martyr and W. F. Benisek, Biochemistry, 12, 2172 (1973). 10. N. Sonenberg. A. Zamir, and M. Wilchek, Biochem. Biophys. Res. Commun., 52, 693 (1974). 11. G. I. Glover, P. S. Mariano, T. J. Wilkinson, R. A. Hildreth, and T. W. Lowe, Arch. Biochem. Biophys., 1§2, 73 (1974). 12. L. C. Davis, G. Gibereau-Gayon, and B. L. Horecker, Proc. Nat. Acad. Sci. U. S., 68, 4l6 (1971). 13. P. Greenwell, S. L. Jewett, and G. R. Stark, J. Biol. Chem., 248, 5994 (1973). 14. F. Hucho, U. Markau, and H. Sund, Eur. J. Bioch., 32, 69 (1973)-

15. M. Rippa and S. Pontremoli, Arch. Bioch. Biophys., 133 , 112 (1969). 16. B. R. Baker, "Design of Active- Site- Directed Irreversible Enzyme Inhibitors,"

John Wiley and Sons, New York, N. Y. , 1967; B. R. Baker, Cancer Chemotheropy

Reports, 4, 1 (1959); B. R. Baker, J. Pharm. , 53, 347 (1964); E. Shaw, "Enzymes," Vol. 1, 3rd edition, Academic Press, Inc., New York, N. Y. 1970, p. 91- 17- R- 0. Robin, Jr., Chem. Rev., 38, 255 0-946); P. Ottolenghi and 0. F. Denstedt, Can. J. Biochem. Physiol., 36, 1075 (1957); J- L. Webb, "Enzyme and Metabolic Inhibitors," Vol. 1, Academic Press, Inc., New

York, N. Y. , 1963. 18. C. Walling and M. J. Gibian, J. Amer. Chem. Soc, 87, 336l (1965). 19. N. C. Yan and D.-D. H. Yan, J. Amer. Chem. Soc, 80, 2913 (1958). 20. N. J. Turro, J. C. Dalton, K. Davies, G. Farrington, R. Hautala, D. Morton, M. Niemczjk, and N. Schore, Accounts Chem. Res., 5, 92 (1972).

21. R. 0. Kan, "Organic Photochemistry," McGraw-Hill, New York, N. Y. , 1966. 22. R. Breslow, Chem. Soc. Rev., 1, 553 (1972). 23. J. Sperling and D. Elad, J. Amer. Chem. Soc, 93, 967 (1971 )• 24. D. Elad and J. Sperling, J. Chem. Soc, C, 1579 (1969)- g, 25- R- E. Galardy, L. C Craig, and M. P. Printz, Nature, New Biology, 242, 127 (1973). s _ - 124 - INTRAMOLECULAR FRIEDEL- CRAFTS REACTIONS OF OLEFINS

Reported by Karl E. Wiegers April Ik, 1975 The Friedel- Crafts reaction is normally thought of as involving electrophilic attack by an alkylating or acylating agent upon the pi electrons of an aromatic 1 2 "V ring. However, analogous reactions may also take place with olefinic substrates. This seminar will deal with Friedel- Crafts- type cyclizations of olefins. ACYLATIONS USING LEWIS ACID CATALYSTS The intramolecular acylation of cyclic enoyl chlorides leads to bicyclic ketones possessing differentiated functionality in two bridges. This route has 3 e been used in syntheses of the bicyclo[4.2.2]decan-8-one ; bicyclo[4A.0]decan-2-onef" 9" lx bicyclo[2.2.2]octan-2-one, T bicyclo[3.2.1]octan-8-one, 8 and bicyclo[3-3.1]nonanone systems. Both chloroketones and unsaturated ketones, in total yields of 50-75$, are found as products of the cyclizations using SnCl 4 or A1C1 3 as catalysts (Scheme 12 ' 13 I). The reaction has also been used in the synthesis of hydrophenanthrene and steroid14 ring skeletons. Scheme I ^ Marshall and co-workers 15 QOCl at \ ,9 have cyclized unsaturated aldehydes related to 1 in benzene in the

1— + of stannic chloride as a quqi^ ^ \ / / \ 7 / presence route to hydroazulenes. Good Yjoj yields of the cycloheptanols are obtained. The related bicyclo- 16 [5- 4.0]undecanone system was also investigated. CHO CYCLIZATION OF OLEFINIC ACIDS 17

a, B- Unsaturated ketones have been obtained by unsaturated carboxylic acids in excess poly- heating " phosphoric acid (PPA). 18 20 n-Alkenoic acids yield mixtures of cyclohexenones,and cyclopentenones, regard- 18 less of the initial position of the double bond, " as PPA catalyzes the rearrangement of the double bond to an internal position suitable for facile ring closure. y- and 6- lactones also lead to good yields of unsaturated

21 -' 22 cyclic ketones upon treatment with hot PPA. This result is thought to be a consequence of the known lacto-enoic acid tautomerism. 23 The cyclodehydration of cyclic lactones has been used by Dev in the synthesis of guaiazulene and 24 25 27 26 28 Se-guaiazulene. Both olefinic acids and Y- lactone > can be cyclized to unsaturated ketones by treatment with zinc chloride and glacial acetic acid in refluxing acetic anhydride. POLYOLEFIN CYCLIZATIONS Johnson and co-workers have extensively investigated the Lewis acid- catalyzed cyclization of polyolefinic acetals29 ' 3° and alcohols 31 ' 32 as synthetic pathways to steroids. It is found that these cyclizations lead stereo specifically to

polycyclic products of "natural" ( trans , anti , trans ) configuration at the ring 30 junctions (Scheme II). These reactions have most frequently been conducted using excess stannic chloride, although protic acids (HCOOH, CF3COOH.) also " have been utilized. 32 34 Investigations of the mechanism of the cyclization 31 34 suggest ' that the ring closures take place in a concerted fashion, without the intervention of incompletely cyclized intermediates.

Scheme II

SnCl 4 £benzene^

\_V HOCH2 CH2 - 125 - BIBLIOGRAPHY

1. For a review, see 'G. A. Olah, ed., "Friedel- Crafts and Related Reactions," Wiley, New York, 1963.

Ibid . 2. (a) . Vol. 3, pp. 1033-1152; (b) J. K. Groves, Chem. Soc. Rev., 1, 73 M (1972). f 3- S. Moon and T. F. Kolesar, J. Org. Chem., 39, 995 (197*0. 4. J. W. Cook and C. A. Lawrence, J. Chem. Soc, 1637 (1935). 5- C.-K. Chuang, Y.-L. Tien, and C.-M. Ma, Ber., 6£, 1*4-94 (1936). 6. (a) C. D. Nenitzescu, E. Cioranescu, and V. Przemetzky, Ber., 73, 313 (19**0)« (b) C. D. Nenitzescu and V. Przemetzky, Ber., Jk, 676 (l9*+l). 7. S. A. Monti and G. L. White, J. Org. Chem., 40, 215 (1975). 8. G. N. Fickes and K. C. Kemp, Chem. Commun. , cW (1973). 9- W. F. Erman and H. C. Kretschmar, J. Org. Chem., 33, 1545 (1968). 10. E. N. Marvell, G. J. Gleicher, D. Sturmer, and K. Salisbury, J. Org. Chem., 33, 3393 (1968). 11. E. N. Marvell, R. S. Knutson, T. McEwen, D. Sturmer, W. Federici, and K. Salisbury, J. Org. Chem., 35, 391 (1970 ). 12. R. Robinson and J. Walker, J. Chem. Soc, 60 (1937). 13. W. E. Bachmann and N. L. Wendler, J. Amer. Chem. Soc, 68, 258O (1946). 14. C. D. Nenitzescu and E. Cioranescu, Ber., 75, 1765 0-9427. 15. J- A. Marshall, N. H. Anderson, and P. C. Johnson, J. Org. Chem., 35, 186 (1970). 16. J. A. Marshall, N. H. Anderson, and J. W. Schlicher, J. Org. Chem., 35, 858 (1970). 17. For a review, see M. F. Ansell and M. H. Palmer, Quart. Rev., 18, 211 (1964). 18. (a) M. F. Ansell and S. S. Brown, J. Chem. Soc, 3956 (1958); 7b) M. F. Ansell, J. C. Emmett, and R. V. Coombs, J. Chem. Soc, C, 217 (1968); (c) M. F. Ansell and T. M. Kafka, Tetrahedron, 25, 6205 (1969); (d) M. F. Ansell and S. S. Brown, J. Chem. Soc, 2955 (1958). 19. (a) R. L. Cargill and A. M. Foster, J. Org. Chem., 35, 1971 (1970); / (b) R. L. Cargill, A. M. Foster, J. J. Good, and F. K. Davis, J. Org. Chem., 38, 3829 (1973). 20. S. Dev, J. Indian Chem. Soc, 32, 255 (1955). 21. (a) C Rai and S. Dev, J. Indian Chem. Soc, 34, 178 (1957); (b) T. M. Jocob and S. Dev, J. Indian Chem. Soc, 36, 429 (1959). 22. J. Ficini and A. Maujean, Bull. Soc Chim. Fr., 4395 (1972). 23. W. S. Johnson and R. H. Hunt, J. Amer. Chem. Soc, 72, 935 (1950). 24. T. M. Jacob, P. A. Vatakencherry, and S. Dev, Tetrahedron, ""20, 2815, 2821 (1964). 25. P. A. Plattner and G. Buchi, Helv. Chim. Acta, 2£, 1608 (1946). 26. W. S. Johnson and J. W. Peterson, J. Amer. Chem. Soc, 67, 1366 (1945). 27. B. Riegel, S. Siegel, and D. Kritchevsky, J. Amer. Chem. Soc, 70, 2950 (1948)- 28. G. Blichi, W. D. McLeod, Jr., and J. Padilla 0., J. Amer. Chem. Soc, 86, 4438 (1964). 29. W. S. Johnson, A. Van der Gen, and J. J. Swoboda, J. Amer. Chem. Soc, 89, 170 (1967). J. Amer. Chem. 30. W. S. Johnson, K. Wiedhaup, S. F. Brady, and G. L. Olson, therein. Soc, 2§, 3979 (1974), and references . 31. (a) P. A. Bartlett and W. S. Johnson, J. Amer. Chem. Soc, 95, 7501 (1973); (b) P. A. Bartlett, J. I. Brauman, W. S. Johnson, and R. A. Volkmann, J. Amer. Chem. Soc, 91, 7502 (1973). 32. R. L. Carney and W. S. Johnson, J. Amer. Chem. Soc, 96, 2549 (1974). 33. K. H. Harding, E. J. Leopold, A. M. Hudrlik, and W. S. Johnson, J. Amer. Chem. Soc, £6, 2540 (1974). if 34. W. S. Johnson and T. K. Schaaf, Chem. Commun., 6ll (1969)- • - 126 - CATHODIC ELECTRO- ORGANIC REACTIONS: PRINCIPLES AND APPLICATION: Reported by Gary Nickel April 21, 1975 Within the last fifteen years there has been an explosion of activity in the area of electro- organic reactions as evidenced by the number of recent reviews 1 and monographs2 on the subject. An electro- organic reaction may be defined as an organic reaction carried out by passage of faradaic current through a solution. In this seminar, we will examine on synthetic and mechanistic bases some practical electro- organic reductions including l) electro- hydrodimerization of activated olefins, 2) "mixed coupling" of carbonyl compounds with olefins or alkyl halides, 3) electrogenerated reagents from "small stable molecules", and k) intramolecular cyclization of dihalides. EXPERIMENTAL PRINCIPLES 3 The equipment and procedure for an electrochemical experiment need not be any more complex than that for a catalytic hydrogenation. Simple electrolysis cells designed for batch-type or continuous operation may consist of a modified flask containing a working cathode (commonly of Hg or Pb), and anode (Ft, Pb, or carbon), and a reference electrode (e.g., saturated calomel electrode, SCE) relative to which the cathode potential is measured. A porous diaphragm (ceramic, glass frit, gel) may be required to separate the products of the cathode reaction from the anode, where oxidation of the electrolyte or even solvent may be in progress. The substrate and supporting electrolyte are dissolved in a protic or dipolar aprotic solvent that is stable in the desired potential range. Among the factors that may be varied in electrochemical reactions are temperature, pH, concentration, solvent, electrolyte structure, and electrode material. A source of direct current is required, and meters to measure potential and current are desirable. For many preparative scale reactions, when selectivity is not necessary, the current density may be set at a convenient level and the desired number of faradays passed (constant current electrolysis). A more controlled technique is constant potential electrolysis (cpe), in which a predetermined electrode potential is maintained until the current flow drops to a low level. The use of cpe allows the selective reduction of one of two functionalities 4 or compounds whose reduction potentials differ by at least 0.15 to 0.20 V. For example, alkyl iodides may be reduced in the presence of alkyl bromides, and alkyl bromides may be reduced in the presence of alkyl chlorides. Reduction potentials may vary with a number of factors including the medium, the electrode material, and substituent effects. 5 Cpe provides different and frequently better selectivity in reduction of functional groups than chemical reducing agents. As another example of selectivity, Lingane, Swain and Fields 6 reduced 9- (o-iodophenyl)acridine (l) by cpe at -1.32 V to give a 90% isolated yield of the desired acridane (2j without loss of iodine. All chemical methods gave only compound (3), which could also be obtained by cpe at -1.67 V (95%)

H -1.67V > cpe - 127 - Polarography (current intensity- potential curves), cyclic voltammetry (reversible potential sweeps), and coulometry (measure of faradays used) are 3 three simple microscale technoques (10 M substrate) useful for a qualitative investigation of an electrochemical reaction. 7 Polarography indicates the number and height of reduction waves, the half-wave potentials (Ei), and the limiting current levels. Cyclic voltammetry allows the observation of the reversible oxidation or reduction of intermediates and products. These techniques yield valuable information, yet the preparative scale identifica- tion of products is very important for understanding electro- organic reduc- tions because differences in the duration of electrolysis or local variations in the concentration, pH, electrode surface or potential distribution may yield results unexpected from the microscale tests. The preparative scale experiment employs concentrated solutions of the substrate, which favor bimolecular reactions, and higher current levels, which may result in the generation of heat. At the electrode, adsorption phenomena and fouling of the metal are much more important. The depletion of protons at the cathode, and the discharge of the electrolyte anion at the anode may necessitate periodic addition of more acid or electrolyte. The rate of electroreduction may depend on the electrode potential, and on the rates of charge transfer (surface effects) and mass transfer (of reactants to the electrode). A practice in electrochemical literature that is frustrating for the organic chemist is that current efficiency (ce) is often reported in place of chemical yield. The current efficiency may be lower than the the chemical yield of current is wasted

. . /,, , -, , in the production heat (because of i (theoretical current ..-. . . n ,. \ . ,. ' , . , , . \ cell resistance) or in discharging the for desired reaction) , . .. , . . ' medium, or the ce may be much greater ce = . /, , , . \ than the yield if the reaction is

1 (total current , , . -. . . ) electrocatalytic. EFFECTS OF THE DOUBLE LAYER The electrical double layer, a specific ordering of ions and molecules at the electrode surface acting as a capacitor, was proposed to explain the small current flow observed when no electroactive species was present. The accepted model8 pictures a compact layer of adsorbed, unsolvated ions at the electrode surface (inner Helmholtz Plane, IHP), covered by a layer of polarized solvent and solvated, nonadsorbed ions (Outer Helmholtz Plane, OHP). Beyond this is a "diffuse layer of ions and polar molecules extending to the 4 7 8 bulk solution. The total layer is estimated to be 10 to 10 cm thick. The specific structure and composition of this layer, which may create conditions at the electrode surface vastly different from the bulk solution, has been shown to affect reactions profoundly. Breslow and Drury9 attempted to generate the anti- aromatic cyclohepta- trienyl and cyclopropenyl anions by electrochemical reduction of the corres- ponding cations. Polarography showed two le waves at mercury for (k) at -0.3 V and -1.5 V (SCE) corresponding to reduction to the the anion. Preparative cpe at either wave, even with 1:1 v CH3CN/CH3COOH as solvent, gave only reduced dimer and no evidence for the anion. Dimerization of two radicals, or at more negative potential by rapid transfer of an electron from the anion to a neighboring cation followed by coupling, presumably occurred at the IHP of the double layer where solvent was not present to protonate the anion. The use of guanidinium per chlorate as electrolyte resulted in partial protonation of the anion. Specific adsorption of the guanidinium ion at the IHP, where it acted as a proton source was proposed as an explanation. , ^ 1

- 128 -

y © 0CH3CN or l/l CH3CN, CHaCOOH^ Hg, O.H-M (BUJ4NCIO4' / tl BFP -l.U or - 1.7V

(it) 98-99$ isolated

CH.gCN, Hg, -1.6y (it) Ot "1=5— O.^M 'NH2 UC10 4 29/o If H2N-C-NH2

1Q In I967 Grimshaw et al . reported the reduction of coumarins to optically active dihydro coumarins in 17$ optical yield with chiral alkaloid electrolytes. Horner and coworkers 11 then studied the asymmetric reduction of ketones and Schiff bases to alcohols and amines with various conditions and alkaloids, obtaining optical yields up to 15$. More recently Kariv, Terni, 12 and Gileadi studied the reduction of acetophenone with quinidine sulfate to give (s)-l-phenylethanol preferentially (21$ optical yield). The asymmetric

pH k-B H CH3 CH3 quinidine aq. acetate buffer C6H5CCH3 _ N sulfate "^ CH3- C - C6H5 + C6H5- C6H5 CH3OH, Hg, E == -1.9V' +7C 2.3 x 10" 2M OH OH OH

1 2

-9A6 inactive °b

reduction could involve a layer of adsorbed alkaloid at the electrode or some preferential complexing of the alkaloid with the ketone. The asymmetric step could be either transfer of a proton to an anion or transfer of a hydrogen atom to a radical derived from the ketone. The best optical yield of 21$ was obtained by causing stronger adsorption of the alkaloid or complex, as influenced by more negative potential down to -1.8 V and more alkaloid up to 10" 2M. Better optical yields were also obtained at lower temperature and when the nitrogen of the alkaloid was free rather than quaternized. The sign of rotation of the resulting alcohol could be switched by use of alkaloids of different absolute configuration. ELECTROHYDRODIMERIZATION OF ACTIVATED OLEFINS

Electrohydrodimerization (EHD) 13 of activated olefins and electrohydro- 14 cyclization (EHC) of bis- activated olefins to give 33 coupled products can proceed in high yields (and high ce) in preference to simple hydrogenation or polymerization, as summarized on the following page. Various metal and ,

- 129 -

.CH- CHC02Et CH= CI^COaEt t 1 2e~ Mil 2e~ 2 -=^ x-c-c-c-c-x Y C=C »y 1 v 2He° 1 1 2H2 1 111 CH= CHC02Et ^— CH-CH2C02Et X H H

Current efficiency Current Olefin of dimer Ref diolefin Y efficiency Ref CH2=CHCN m 15 I. C(Et)2 98$ 19 CH2=C(CH3 )CN i% 15 II. (CH2 ) 2 4l$ 19

CH2=CHCOOEt 7*4-87$ 15 III. (CH2 ) 3 100$ 19

(CH3 ) 2C=CHCOOEt 66$ 16 IV. (CH2 ) 4 90$ 19 CH2=CHC0CH3 1+0-70$ IT V. 0CH2CH2 89$ 19

CH3CH=CHCONEt 2 6l$ 15 VI. (CH2 ) 5 10$ 19

C 6H5 CH=CHCOOEt 52$ 18 VII. (CH2 ) 6 0$ 19 amalgam reducing agents give dimer yields comparable to the EHD process 20 The EHC processing is an excellent route to three to six-membered rings with func- tional groups on the side chains. Baizer and coworkers at Monsanto have studied olefin coupling and developed a commercial process for coupling of acrylonitrile (AN) to give adiponitrile (ADN), an intermediate of nylon 6,6. Adiponitrile and polyacrylonitrile (anionic polymerization) 21 were major products in aprotic solvents, but in water with alkali metal salts as supporting electrolytes the initial radical anion was protonated, followed by further reduction and protonation to give propionitrile. The use of tetraethylammonium p_- toluene- sulfonate as electrolyte in water improved the solubility of AN and gave high yields of hydrodimer. Gillet22 and

Et^OTs (56$ aq.. soln) CH^CHCN -> NCCH2 CH2CH2CH2CN CH3CH2CN 25°, pH 7-9 Pb 98-99$ 30 A/dm2 E = -10V Feoktistov23 independently proposed that because the large organic cation was poorly solvated by water in comparison to alkali ions, its adsorption at the IHP caused a water poor zone at the cathode surface. This thin zone inhibited protonation of intermediates, and allowed reaction between two olefins while preventing polymerization. Feoktistov et_al. 23 demonstrated that the water- Et^OTs system provided the same proton availability at the cathode as DMF, 5$ H20. The same solubility or double layer effects might explain the improved yields of hydrodimer izat ion (70-90$) by sodium amalgams (no electrolysis) noted upon addition of quaternary ammonium salts. 2 Despite extensive study, no one mechanism for EHD has been established, and the mechanism may even vary with the experimental conditions. Lamy, Nadjo, and Saveant24 have shown at least for systems of low acidity and water content (DMF, Et^OTs), and millimolar concentration ranges, that electrolysis of activated olefins at the first of two waves (when resolved) resulted in rate- determining radical coupling of two radical anions to give a dimer dianion. Under most conditions mono- activated olefins have shown only one irreversible 25 2e wave by cyclic voltammetry and no esr signal, so that *t he initial radical anion must be reduced further or react rapidly. An anionic Michael- type coupling seemed possible under Baizer' s conditions because coupling improved for better Michael acceptors and some cross coupling was observed between two 26 different olefins at a potential where only one was reduced. Among the f —

- 130 - possible mechanisms are the coupling of two radicals or radical anions (adsorbed at the 3 position), addition of the radical anion or its protonated form to a neutral olefin molecule, or attack of an anion on a neutral olefin. The radical anions of diactivated olefins are more stabilized, and may represent the coupling ^) species in that case.

CH2=CHCN- > • CH2CHCN H • CH2 c"HCN- > • CH2 CH2 CN

• CH2 CH2CN- -> -CH2 CH2 CN 2H 2 •CH2 CHCW- ->NCCHCH2 CH2CHCN- ->ADN + CfkCHCN + CH2=CHCN- > NCCHCH2CIfeCHCN ? > ADN H CH2 CHCN + -CH2 CH2CN- > NCCHCH2CH2CH2CN' ->ADN

2 -CH2 CH2CN ->ADN

CH2 CH2 CN + CH2=CHCN- -> NCCH2CH2CH2 CHCN £ ADN H CH2CH2 CN + CH2=CHCN- ->NCCH2 CH2CH2CHCN > ADN

Diolefins such as 2,6-octadiene-l,8-dioate(ll) that undergo electrohydro- cyclization are reduced in two le waves at a less negative potential (easier) than diolefins such as 2,10-dodecadiene-l,12-dioate(VTl) that do not cyclize. 13 The fractional height and the positive shift in Ei of the first wave correlate to the yield of cyclic product (best yield by cpe 2at the first wave). The yield improves if the double bonds can be parallel and close. Cyclization is also observed at a potential where only one of the bis- olefins was reducible (because of different substitution), and coupling of two radical anions is impossible. This has been explained by a concerted le reduction- ring closure to a cyclic radical anion delocalized over both activating groups, that is rapidly reduced to the dianion. 19 e CH=CHX CH-CH-X CH-CHX 2H2^ (CH (CH ) (CH product 2 ) n 2 ^ 2 ) n \ CH=CHY CH-CH-Y 'CH-CHY

MIXED COUPLING The reductive coupling of carbonyl compounds to give gemdiols (pinacols) has been studied for many years. 27 Recently, Sugino and Nonaka28 have coupled ketones with acrylonitrile, maleic acid, and pyridine in 20$ sulfuric acid (see below). The dimers of AN with acetone, methyl ethyl ketone, and diethyl ketone 29 were produced at Hg with TO, 60, and 30% current efficiencies (ce). Frohling improved the technique and coupled a series of methyl ketones with ethyl acrylate and AN to give lactones (by in situ hydrolysis of nitrile and lactonization) in 30 high yields (70-90$). Nonaka and Odo coupled acetaldehyde and propibnaldeyde with AN to give the v-substituted y'butyrolactones with 18 and 15$ ce. This method offers an effective route to lactones from simple, mono- functional precursors. v - 131 -

3 % .20% H2S0 4 , CH3OH f* CH =CHCN > HO-C-CH CH CN 2 Hg, -1.26V SCE 2 2 s CH3 R 2.5 mol 0.1 mol R - CH3-C6H13

The Ei of -1.3 V SCE in the ketone mixed couplings indicated a le Traduc- tion of the protonated ketone to give an adsorbed radical, consistent with the reduction products of acetone in 0. 5 M H^SC^ (2-propanol, diisopropyl mercury, pinacol). 31 Fast sweep (lOOOV/sec) cyclic voltammetry detected the reversible reoxidation of the radical, and the reoxidation peak decreased after addition of ethyl acrylate. The coupling was thought to involve addition of this radical to the 3 position of the olefin, followed by rapid reduction (which should occur at -O.lOv for (CH C0HCH2CHCN) and protonation. 31 3 ) 2 H R5 X

OH C u* g°feso 4 1 j- , /\^ ^c_> /\ ^^i_yX R2' R 3 »i Ra Ri R2 ////// /

Choice of the correct electrolysis conditions has allowed the preparation of tertiary alcohols from ketones and weakly polarized olefins. Reduction of 32 acetone and allyl alcohol at Hg activated graphite gave 95-98% yield (70% ce) of 2-methyl-2,5-pentanediol (Grignard yield 30-40%). Reduction of acetone also proceeded in aq. DMF, TsOH, Et^NOTs to give good yields of coupled + 2H ? (CH3 ) 2C=0 CH2=CHCH20H > (CH3 ) 2C(0H)CH2CH2CH20H (95-98%)

33 alcohols with styrene (42%), butadiene (30%), and trans- stilbene (45%). The reduction- protonation- reduction of ketones to their anions in the presence of alkyl halides gave low yields of alcohols. 34 Electrolysis at the Ei of the alkyl halide in the presence of a more difficultly reduced Schiff 35 base resulted in coupling to give, in one case, amino acid derivatives (40-86%),

CH3 . CH3 DMF, Et^flBr c_ C6H5 CH2N=C-C00R + R'X > CsH5CH2ra_ C00R tig, -1.; to -1.9 I R'

These syntheses are simple and should be considered as complementary to the Grignard reaction. ELECTROGENERATED REAGENTS Oxygen was reduced at the dropping mercury electrode (DME) in aprotic solvents in two le" steps at -0.73 V to the superoxide radical anion, and at -2.40 V to the dianion. 36 The reduction to the radical anion was quasi-

reversible, and the esr spectrum could be recorded at -I96 . Solutions of the radical anion (0.1 M) were prepared by cpe at Hg (-1.0 V, 0.1 M BU4NCIO4) of an aerated solution of DMF (5% 2 ). Superoxide had a half life of 30 min 36 at 25°, and was probably present as the tetraalkylammonium salt. The addition of an alkyl halide increased the cathodic current and destroyed the 4 37 reversibility. Peroxides were formed in good yield (80%) with aliphatic or — ~ - 132 - benzylic halides and some esters and anhydrides by a nucleophilic mechanism:

* 2 RX ^R0> R02 " X-

'

R02 + e (fromi eleeelectrode or 2 )c- R02 ROo" RX- ->ROOR X

Diacyl peroxides were further reduced at the cathode potential. Relative reactivities (n-BuBr > _s-BuBr - i_-BuBr » t-BuBr, and n-BuBr > n-BuOTs > n-BuCl) indicated S-J2 reactions of a powerful nucleophile. By a similar scheme S02 was reduced (CH 3CN, 0.2M Et4NBr, Hg) to the radical 38 anion, which reacted with alkyl halides to give sulfones in high yield (90^>^ Some success was also obtained in the production of unsymmetric sulfones. The reduction of C02 at -2.2 V S'CE in the presence of alkyl halides (RX) gave very low yields of esters (RCOOR). 40 REDUCTION OF DIHALIDES

The electrochemical reduction of dihalides is easily as efficient as Zn or amalgam reductions, without the need for large amounts of expensive and dangerous metals. The electrolysis of 1,2-dihalides to give elimination was recently employed by Rieke and Hudnall 41 to generate benzocyclobutadiene (90$ yield of Diels- Alder dimer) from 1,2-dibromobenzocyclobutene at a 2e 42 wave at -1.3 V (Hg, CH2CN, Et4NC10 4 ). Rifi demonstrated that electrolysis of 1,3 and 1,4- dihalides gave cyclic products including cyclopropane s, cyclo- butanes, spiropentane, bicyclobutanes, and bicyclopentanes. Wiberg and 43 Epling have improved the electrochemical synthesis of cyclopropane (91$) and cyclobutane (90$). Phenylcyclopropane (70$) was prepared from 1,3- dibromo-1-phenylpropane, and cyclopropanol (60$) from l,3-dibromo-2- 44 D propanol. Fry and Scoggins*,45 trapped tetramethylcyclopropanone as its hemiketal (100$) in methanol at 0° by electrolysis of 2, ^-dibromo- 2, ^dimethyl-

3-pentanone. Wiberg et al . , claimed to trap [2.2.2]propellane ) at -15 by electrolysis of l,*<-dibromo-bicyclo[2.2.2]octane, followed by addition

of Cl2 . CH3

Br Br 2e ^H CH3OH

Although questions still exist, a good body of evidence has accumulated in favor of a stepwise intramolecular anionic displacement in most cases.

Q ' Br- CH2CH2 CH2- Br —^-> • CH CH CH Br + Br® CH CH CH Br — tj> Br 2 2 2 T& j£> 2 2 2 A

The cpe of 1-chloro- 3- bromo cyclobutane at a potential only sufficient to cleave the C-Br bond consumed 2e~ to give the anion that cyclized to bi cyclobutane. 42

2e CI-/ \-Br + Br + CI •2.0V - 133 -

Reduction of l-bromo-3- (bromomethyl ) cyclobutane in DMF resulted in protonation of the monoanion (by DMF or traces of water) and reduction of the second C-Br bond to give methylcyclobutane, but in dried HMPA gave bicyclopentane. 42

DMF Br-<^N/\-CH2Br- 1 -CT% ,-p^^© > ^ , ^/^

47 Fry and Britton have shovm that me so- and dl-2, ^-dibromopentane both gave mixtures of cis- . and trans- 1,2- dimethyl cyclopropane at -2.2 V in nearly the

same ratios {kl : ^5 and hk : 39*5), so that the diastereomeric carbanions must equilibrate before cyclization. A concerted cyclization by a 2e displacement of 2 Br and radical coupling has been proposed in some cases, but cyclic voltammetry (0.3 V/sec) in DMF at a Hg- coated Pt cathode showed for 1,3-dibromopropane two irreversible le" 43 waves at -1-75 V and -2.0 V^Hg). Faster sweeps merged the waves. Irreversibility (50 v/sec) and the absence of propyl mercury compounds indicated that the adsorbed bromopropyl radical must be reduced to the anion, desorbed, and cyclized rapidly. The adsorption of alkyl halides (at both ends for dihalides) and alkyl radicals at a Hg cathode has been credited with facilitating cyclization. In contrast, at Al cathode the intermediate is not held at the surface, and radical dimerization gave higher bromides. 48 The increased yield of cyclopropane at more negative potentials (lh% at -1.^-5 V and 9l/o at -2.65 V") might be explained by stronger adsorption of the ends to give geometries favoring cyclization, and by direct 2e transfer to give the anion and cyclization. CONCLUSION Electrochemistry offers interesting opportunities for the study of reaction mechanism and the development of either simple or selective syntheses, some of which may be industrially important. Although it is

a technique not commonly considered for use by organic chemists, some of . it advantages include mild conditions, low operation costs after the initial investment, experimental control of many variables, and some novel results. Only a limited area has been considered in this seminar, yet the whole field is vastly undeveloped. BIBLIOGRAPHY

1. (a) L. Eberson and H. Schafer, Fortschr. Chem. Forsch. , 21, 113 (l97l); (b) A. J. Fry, ibid., %k, 1 (1972); (c) F. Beck, Angew. Chem. Int. Ed., 11, 760 (1972). 2. 7a") "Organic Electrochemistry: An Introduction and Guide," M. M. Baizer

ed., Marcel Dekker, Inc., New York, N. Y. , 1973; (b) "Synthetic Organic

Electrochemistry," A. J. Fry, Haper and Row, New York, N. Y. , 1972; (c) "Introduction to Organic Electrochemistry," M. Rifi and F. Covitz,

Marcel Dekker, New York, N. Y. , 197^. 3« H. Lund and P. Iversen, ref. 2a, pp 166-236. 4. A. J. Fry, M. Mitnick, and R. G. Reed, J. Org. Chem., 35, 1232 (1970 ). 5. P. Zuman, "Substituent Effects in Organic Polarography, " Plenum Press,

New York, N. Y. , 1967. 6. J. J. Lingane, C. G. Swain, and M. Fields, J. Amer. Chem. Soc, 65, 13^ (19^3). 7- G. Cauquis and V. D. Parker, ref. 2a pp 93-1^9- • ) •

. «- i 8. D. M. Mohilner in "Electroanalytical Chemistry," (A. J. Bard, ed.

Vol. 1, Marcel Deleter, Inc., New York, N. Y. , 1966, p 24l. 9. R. Breslow and R. F. Drury, J. Amer. Chem. Soc, %6, 4702 (1974). 10. R. N. Gourley, J. Grimshaw, and P. G. Millar, Chem. Commun. , 1278, 1967; 2317, 1970.- 11. (a) L. Horner and D. Skaletz, Tetrahedron Lett., 3679 (1970); (b) L.

Horner and D. Degner, ibid . , 1245 (l97l); (c) L. Horner and R. Schneider,

ibid . , 3133 (1973); (d) L. Horner, D. Degner, and D. Skaletz, Chem. Ing. Tech., 44, 209 (1972). 12. E. Kariv, H. A. Terni, and E. Gileadi, Electrochim. Acta, 18, 433 (1973)- 13. M. M. Baizer and J. P. Petrovich in "Progress in Physical Organic Chemistry," Vol. 1, 1970, pp 189-227. 14. J. P. Petrovich, M. M. Baizer, and M. R. Ort, J. Electrochem. Soc, 116, 749 (1969). 15. M. M. Baizer, J. Electrochem. Soc, 111, 215, 223 (1964). 16. J. Wiemann and M. L. Bouguerra, Compt. Rend., C265, 751 (19^7) 17. M. M. Baizer, J. D. Anderson, J. Org. Chem., j50, 3138 (1965). 18. L. H. KLemm and D. R. Olson, J. Org. Chem., 38, 3390 (1973). 19. J- D. Anderson, M. M. Baizer, and J. P. Petrovich, J. Org. Chem., 31, 389O, 3897 (1966). 20. F. Matsuda, Tetrahedron Lett., 6193, (1966). 21. M. M. Baizer and J. D. Anderson, J. Org. Chem., 30, 1351 (1965)- 22. I. E. Gillet, Bull. Soc. Chim. Fr., 2919 (1968). 23- L. G. Feoktistov, A. P. Tomilov, and I. G. Seva st yanova, Soviet Electrochem., ' 1, 1165, 1300 (1965). 24. E. Lamy, L. Nadjo, and J. M. Seveant, J. Electroanal. Chem., 42, 189 (1973). 25. F. Beck, Chem- Ing. Tech., 37, 607 (1965). 26. M. M. Baizer and J. L. Chruma, J. Electrochem. Soc, 118, 450, (1971 ) 27- S. Swann, Jr., "Electrolytic Reactions" in "Technique of Organic Chemistry" (A. Weissberger, ed. ), Vol. 2, Wiley- Inter science, New York, N. Y., 1956. 28. K. Sugino and T. Nonaka, Electrochim. Acta, 13, 613 (1968). 29. A. Froling, Rec trav. chim. Pays-Bas, £3, ^7 (1974). 30. T. Nonaka and K. Odo, Denki Kagaku, 4l, 662 (1973)- 31. 0. R. Brown and K. Lister, Disc. Faraday Soc, 45, 106 (1968). 32. A. P. Tomilov and B. L. Klyuev, J. Gen. Chem. USSR, 39, ^+6 (1969). 33- M. Nicholas and R. Pallaud, Compt. Rend., 267C, l834"Tl968). 34. H. Lund and J. Simonet, Bull. Soc Chim. Fr., 1843 (1973). 35- T. Iwasaki and K. Harada, Chem. Commun., 338 (1974). 36. D. L. Maricle and W. G. Hodgson, Anal. Chem., 37, 1562 (1965). 37* R- Dietz, A. E. Forno, B. E. Larcombe, and M. E. Peover, J. Chem. Soc, B, 8l6 (1970). 38. D. Knittel and B. Hastening, J. Appl. Electrochem., 3, 291 (1973)- 39- D. Knittel and B. Kastening, Ber. Bunsenges. Phys. Chem., 77, 833 (1973 )• 40. J. H. Wagenknecht, J. Electroanal. Chem., j?2, 489 (1974). 41. R. D. Rieke and P. M. Hudnall, J. Amer. Chem. Soc, 95,2646 (1973)- 42. (a) M. Rifi, Tetrahedron Lett., 1043 (1969; (b) M. Rifi, J. Amer. Chem. Soc, 8£, 4442 (1967). 43. K. B. Wiberg and G. A. Epling, Tetrahedron Lett., 1119 (1974). 44. R. Gerdil, Helv. Chim. Acta, 53, 2100 (1970 ).

45. A. J. Fry and R. , Scoggins, Tetrahedron Lett . 4079 (1972). 46. K. B. Wiberg, G. A. Epling, and M. Jason, J. Amer. Chem. Soc, 96, 912 (1974). 47. 0. R. Brown and E. R. Gonzalez, J. Electroanal. Chem., 43, 215 (1973 )• 48. A. J. Fry and W. E. Britton, J. Org. Chem., 38, 40l6 (1973)-

. -135- HETEROCYCTES IN SYNTHESIS - ANNELATION REACTIONS

Reported by Marvin Reich April 28, 1975 Heterocyclic compounds have been used as reactants, intermediates, and reagents in organic synthesis. 1 Annelation reactions lead to a variety of important products and this area has been chosen to illustrate the usefulness of heterocycles as synthons. The diversity of transformations which these substances can undergo will be demonstrated by several reaction types: (l) ring opening and reclosure, 2) Diels-Alder reactions and subsequent modi- fications, 3) extrusion reactions, h) reactions of heterocycles as masked carbonyl compounds 2,ij-,6,-Trisubstituted pyrylium salts react with a variety of nucleophiles to give ring-opened products which subsequently close to new heterocycles or substituted benzene rings 2 which are difficult to make by other routes.

CH 3 Et20, MeOH,or Ref. 3 -> H2 Rp,C H 3C A

Other nucleophiles which can be used for similar conversions include azo- methines, ylides, carbanions of 3-dicarbonyl compounds, and or ganomet allies. a- and y-py^ones, which are resonance hybrids between neutral and charge- separated forms, give analogous reactions. 4 Some heterocyclic compounds have the ability to function as either the diene 5 or dienophile 6 in the Diels-Alder reaction with acyclic compounds. The resulting adduct is then further modified to the desired product. Some intermediates can be subjected to a retro Diels-Alder reaction yielding a new ring system. The latter may provide aromatic compounds possessing unusual I) substitution patterns. 7 The following sequence provides a stereospecific route to substituted hydrindanones, unavailable via homologs of trans- 2- met hyl- 2-butenoic esters.

C02CH3 - . H3C02C -jfRaNi CH3OH, A J r 2 - 0^^ Claisen ^ [ Ref. 6 C2H5 3- hydrolyze, H C02CH 3 decarboxylate Loss of heteroatoms from cyclic systems may lead to intermediates which can undergo further reaction to yield new alicyclic systems. Pyrolysis 8 of pyrazolines has long been a source of cyclopropanes. Both 1- and 2- pyrazolines undergo the reaction, the latter requiring a basic catalyst to promote tautomerization to the 1- isomer.

KOH Ref. 9 2to-2^0 >

k-k-io isolated yield The dot-.:": >!. mechanism of this reaction is unknown, and product stereochemistry is -:< difficult to predict ^:. 'a fused ring systems. Vicinal diaryloxiranes undergo photo fragment at ion in solution to give arylcarbenes and carbonyl com- 10 pounds. In the presence of olefins, a useful synthesis of cyclopropanes is achieved. Addition to olefins is stereospecific but with noncentrosymmetric substrates, epimeric mixtures are produced, similar to carbenes generated from other sources. The de compos it ion is apparently an a- elimination of a carbonyl fragment from the more stable of two zwitterionic intermediates. hv 0 0/\h 25^ nm -f ®f <-> H p CH3 CH3 CH3 CH3

Isoxazoles and pyridines have been used as masked carbonyl synthons which may be activated by hydrogenation or dissolving metal reduction respectively. Stevens has applied isoxazoles to the construction of the ring-bridging vinyl- ogous amidine system of semi- corrins 11 and will extend these results to the synthesis of corrins. Stork has employed isoxazoles as methyl vinyl ketone equivalents in the synthesis of fused polycyclic systems 12 including steroids.

H ,Pd 2 -> 10$ KOIL

These heterocycles have been used in a similar manner to construct pyridines 13a and substituted phenols. Bis annelation onto cyclic ketones may be conducted via 6-methyl-2-vinylpyridine. l4

("diethylene glycol l.< diethyl ether •0"| 1. Na/NH3 0-1 2. NaOH I A w 2. HO OH, H 3. H 3 !+. TsOH

BIBLIOGRAPHY

I. A. I. Meyers, "Heterocycles in Organic Synthesis", Wiley- Interscience, New York, (197*0. 2. K. Dimroth and K. H. Wolf, Newer Methods of Preparative Organic Chemistry, 3, 357 (196*+).

3. C. Toma and A. T. Balaban, Tetrahedron Suppl. , 7, 9 (1966). k. See for example, S. J. Norton and E. Sanders, J. Med. Chem. , 10, 961 (1967). 5. G. Markl and R. Fuchs, Tetrahedron Lett., 4695 (1972); E. E. Harris, R. A. Firestone, K. Pfister, R. R. Boettcher, F. J. Cross, R. B. Currie, M. Monaco, E. R. Peterson, and W. Reuter, J. Org. Chem., 27, 2705 (1962). 6. G. Stork and P. L. Stotter, J. Amer. Chem. Soc, %1, 7780 (1969). 7. See for example, D. L. Fields, J. Org. Chem., 36, 3002 (1971). 8. R. Fusco in "The Chemistry of Heterocyclic Compounds", Vol. 22, A. Weissburger (ed), Wiley- Interscience, New York, (1967), p 209- 9- R. M. Coates and J. E. Shaw, J. Amer. Chem. Soc, %2, 5657 0-970). 10. N. R. Bertoniere and G. W. Griffin, Org. Photochem. , £, 115 (1973). II. R. V. Stevens, C G. Christensen, W. L. Edmonson, M. Kaplan, E. B. Reid, and M. P. Wentland, J. Amer. Chem. Soc, 92, 6629 (l97l); R. V. Stevens, L. E. DuPree, W. L. Edmonson, L. L. Magid, and M. P. Wentland, J. Amer. Chem. Soc, 93, 6637 (1971 )• 12. G. Stork, S. Danishefsky, and M. Ohashi, J. Amer. Chem. Soc, 82, 5*+59 (1967), and subsequent papers. 2.3- a) G. Stork, M. Ohashi, H. Kamachi, and H. Kakisawa, J. Org. Chem., 36, 2784 (1971); bj M. Ohashi, T. Maruishi, and H. Kakisawa, Tetrahedron Lett., 719 (1968). 14. S. Danishefsky, P. Cain, and A. Nagel, J. Amer. Chem. Soc, SJ_, 380 (1975). -

SOME RECENT DEVELOPEMENTS IN CYCLOPROPANONE CHEMISTRY

Reported by: Kenneth Berger May 1, 1975

1. Introduction Cyclopropanones were first synthesized in 1965 and 1966 with the prepara- tion of tetramethylcyclopropanone and 2,2-dimethylcyclopropanone. 1 Prior to 1965 cyclopropanones were suggested as intermediates in Favorskii rearrangements. High reactivity due to strain and possible tautomeric structures has stimulated substantial chemical and theoretical interest in cyclopropanones. 1 * 5 1 2 For background information refer to the reviews by Turro * and by Subra- hmanyam. 3 The third review unfortunately has several typographical errors and missing functional groups in the diagrams. This paper covers developments in cyclopropanone chemistry since Turro's 1969 review. 1-

II. Tautomerization and Spectral Properties That cyclopropanones exist as the classical closed ring form is not readily apparent. It is speculated that a labile equilibrium exists between tautomers 1, 1 " 3 2, and 3* Tautomer h can also be envisioned. However, little evidence exists 1 2 for a major contribution from k. >

s 9" A- A \ 1 2 3 i classical oxyallyl allene enol form species oxide form

Early molecular orbital calculations predicted 2 to be more stable than 6 7 8 9 1. > However, microwave studies have shown 1 to be the predominate species. > After the microwave results were known, MINDO/2 calculations "predicted" 1 to be the most stable form. 10 Using INDO and SCF calculations Liberies11 predicts singlet 2 to be 83 Kcal/nol less stable than 1 and 3 to be 21 Kcal/mol less stable than 1. Other conclusions he draws are that the 03^03 angle in 2 is 105°, and the dis rotatory ring opening of 1 to 2 is lower energy than the conrotatory 12 with 2 near or at the energy maximum in the disrotatory opening. Another group looking into isomerization between 1 and 3 suggests that all previous calculations give too much weight to derealization stabilization. They12 suggest a distorted non-planar transition state may be much lower energy than the generally postulated planar oxyallyl intermediate or transition state. Bond additivity schemes are also used to calculate relative energies. 13 A sample calculation is shown below.

AH (l) = AH + AH (MeC0Me) - (MeCHMe) f f(A) f H f -3 Kcal/mol = (239) + (-51.8) - (190)

AH (l) = AH ( methylene cyclopropane) + AH (MeC0Me) - AH (CH C(Me) f f f f 2 2 )

Kcal/mol = (kQ) -f- (-51.8) — (-k.Ok)

Heats of formation calculated in this manner indicate that: 2 is 5k Kcal/mol higher in energy than 1 and 3 is 23 Kcal/nol higher in energy than 1. III. Chemical Properties Synthesis of Cyclopropanone Cyclopropanones tend to be unstable at room temperature and polymerize when concentrated. 1 They are also sensitive to acids and bases. Consequently, they are generally prepared under neutral conditions as dilute solutions at low temperatures. The following methods have been used in cyclopropanone syntheses: (l) reaction of diazoalkanes with ketenes, (2) photolysis of 1,3-cyclobutadiones, 1 3 (3) isomerization of allene oxides, (k) dehydrohalogenation of a-haloketones. > -140-

Masked cyclopropanones are useful synthetic intermediates for cyclo- propanone generation. The example below illustrates the trapping of free tetramethylcyclopropanpne (jj) from an equilibrium mixture with its hemiketal 6,

HO OMe furan

Unlike free cyclopropanones, many of the masked derivatives can be purified readily. Among the masked cyclopropanones that have been prepared are cyano- hydrins, 14 cyclopropanone tosylhydrazide adducts, 15 a-bromocyclopropyl tri- 16 1 "7 18 13 fluoroacetates, hemiketals, ' and ketals. Most of them have been pre- pared directly from the free cyclopropanones. The reaction of a diazoalkane with a ketene is the most general synthesis of free cyclopropanones, with typical yields of 70-90%. 1

H + Rn 1, 90%, R = R = H CHoCl. x 2 .c-c=o 2- 9 8, 80%, R = Me, R = H Ri 78° 2 2 Y-r 9, 90%, R = H, R = Me x 2 R2 Ri

Turro and co-workers"' have discussed two mechanisms for reaction of diazomethane and ketene. O C=0

.. ). •N^N (b)

They prefer route (a) because of lesser steric hindrance and the favorable arrange- ment for rapid internal nucleophilic displacement of nitrogen. Route (b) has resonance stabilization to lower the relative energy of the intermediate, but in order to displace the nitrogen it would be expected that the sp2 carbon would have to rotate to a conformation resembling the intermediate in route (a). If loss of nitrogen is the rate determining step, the allyl anionic intermediate would resist or hinder formation of cyclopropanone. Dimethyl ketene reacts 2 rapidly at -78° indicating that route (a) would be most likely.

Photolysis of 10 gives rise to _5, but generalization of this path to other 1 2 2 cyclopropanones has not been promising. > > ° With a trapping agent such as methanol yields as high as 90% of the hemiketal 6 can be obtained. However, most cases yield mixtures, or when trapping agents are left out, decarbonylate a second time to the corresponding olefin. 20

^OOnrn hv * MeoC=CMeo + CO Ph-H 10 25° Me OH 1 Peracid treatment of allenes is a possible method of producing allene oxides. 21 "23 Allene oxides with bulky substituents can be isolated and 23 isomerized to the corresponding cyclopropanone. It is believed that the t-butyl groups prevent diepoxidation and protect the carbonyl from attack. When allenes with less bulky substituents are used, as in the cases of 1,1-dimethyl- allene and 1,2-cyclononadiene, no allene oxide or cyclopropanone is isolated. 24

R .t-Bu hexane P, XC=C=C^ / C—C' k H R' p-ClPhC03H t- H R' H R«

= = t-Bu R = H, R' = t-Bu 11 R = t-Bu, R' = H 12 R H, R' 15 15 R = t-Bu, R' = H 16 R = t-Bu, R 1 = H lit R - H, R' = t-Bu

Compound 16 has also been generated under Favorskii rearrangement conditions, by action of potassium t-butoxide on a-bromodineopentyl ketone, 25 Excess t-butoxide converts 16 to ester l8. 25 As in the allene oxide case above, cyclo- propanones can be isolated only when bulky substituents are present. For this reason the synthesis of 16 is a special case. Most Favorskii reactions lead 1 3 to the Favorskii products. >

II t-Bu % t-Bu -Bu t-BuOK excess COOt-Bu t-BuOH t-BuOK t-Bu-CHCH2 -t-Bu Br H t-Bu

IT 16 75$ 18 single product

Of the methods available the diazoalkane route appears to be the most general method of generating cyclopropanones in high yields . Masked cyclo-

} ropanones have the advantage of being isolable and generally easier to handle 'han cyclopropanones, and these cyclopropanone precursors can do many cyclo-

1 ropanone reactions.

R eactions of Cyclopropanones

1. Addition of HX to the CO Bond Several reactions of cyclopropanones involve the conversion of the sp2 3 carbonyl carbon to sp . Some of these are shown below. HX may be HOH (30$), HCN, HC1 (65$), HQAc (100$), HOMe (100$), PhNHMe (100$), or PhNH2 ('

HX Y*

PhNH^ I ^^MepM 1

FhHN .NHPh x Me. .N NM6 M 2. Ring Expansion Addition of diazoalkanes to cyclopropanones yields cyclobutanones. The mechanism of this addition was studied by Turro and Gagosian. 26

R CH2NH< CH2NH2 \ + c=c=o -• K-Hf R—J3 1 1 R' -78 -78 12 R' " a) R=R'=Me 20 21

b) R=t-Bu, R'=H - product distributions

c) R=cyclopropyl, R'=H a = 65$ 35$

b = -JQffo 3<$ c = 65$ 35$

Their work demonstrates that cyclopropanones are intermediates in the ring expansion of ketenes by diazoalkanes. The diazoalkane used influences which cyclopropane bond migrates to form cyclobutanone, and the product ratios can be rationalized using steric arguments. In the reaction of 22 with diazomethane, the most substituted carbon preferentially migrates. # CH NH2/CH2C1 • -A 2 2 V 22 23 2k 37$

In the reaction of 22 with diazoethane the least substituted carbon pre- ferentially migrates, a result which can be rationalized by the following scheme,

Bottom Attack (Favored) major

(A) v^K-^ H K %, C=N2

\ minor

(B) + H H + ''^C=Np H i l

26 \~Y 30$ A-\ 28 , -2*5-

Top Attack (Unfavored) + £n2 (C)

H H ^-° H C-°

I W 25

*^C=N2 (D) N. H

I i 28 tf

Since the rate determining step in the addition of diazoalkanes to ketones 27 is the addition of the diazoalkane, steric interactions as the diazoalkane approaches and steric interactions in the zwitterionic intermediate control the product ratio. An alternate mechanism using a carbonium ion intermediate,

29 , predicts migration of the most substituted group to be prefered, which contradicts the experimental results. 26

Below are a few examples of ring expansions using masked eye lopropan ones demonstrating the usefulness of such masked reagents in synthesis. r _ .. s

o SO^r HO HHNHSO^r v m TsNHNH, MnO, 16 . X t-Bu t-Bu CHC1. t-Bu t-Bu 25'o

(ref. 16) OCOCF3 OH

Ir IAH .•< ^r AgOCOCF 3 — 'O mild n /) & base Br

I NaBH 4 H (T (ref. IT)

CI + I HONHR R-N. m R-N OH RNHj x Ag Y Me~COCl -d CH3CN 38-65$

1 3 3. Cycloadditions > As of now the parent cyclopropanone has not been shown to undergo cyclo- additions. However, substituted cyclopropanones undergo a variety of these additions. The diagram below illustrates some of the reactions possible.

CC1< X' A 0+00+ 00 — o # 1 €&&

X - CH2 , 35$ X = CH2CH2

A/^<& X = nch , 50$ 3

X = 0, 100$

X = C=CMe2 , 72$

These reactions can be classified by the Huisgen method or by the Woodward- Hoffmann method. It appears that cyclopropanones can generally undergo cyclizations by the mechanisms below, two of which involve the oxyallyl species,

oIj-TT

-* + M Y Xj

+ Me P CC=0 (Probably a polar -A mechanism) -145-

k. Resolution and Racemization of trans -Di-t-butylcyclopropanone, Intermediacy 25 28 29 of Oxyallyl > > As discussed earlier 16 can be formed by action of potassium t-butoxide on o-bromodineopentyl ketone in t-butyl alcohol, Favorskii reaction conditions. Resolution of 16 was accomplished by asymmetric destruction with d-amphetamine 7 to 9 1% optical purity. Optical purity of 90-100$ was obtained by partial reduction with (+)-diisopinocampheylborane. but the product was much harder to purify. The (+)-l6 can be racemized at 80°. At 150° or uv irradiation at 25°, 16 decarbonylatesTo trans -di-t-butyl ethylene. The allene oxide 15 isomer izes

to racemic 16 upon heating to 100° . The decarbonylation proceeds thermally and photochemically, while racemization proceeds only thermally. 29

t-Bi 80° 150c (+) - 16 dl-16 + CO

C«H6n6 or hv 25° t-Bu 100c

It has been proposed that oxyallyl is intermediate in the racemization. The rate of racemization is first order in 16, shows a small increase with increased solvent polarity, and is weakly accelerated by benzoic acid. Racemiza- tion in t-butyl alcohol-0-d does not lead to deuterium incorporation in recovered 25 16, thus eliminating an enolization mechanism. In an attempt to demonstrate the oxyallyl proposal, 16 was heated to 125° in the presence of furan, a selective trapping agent for oxyallyl. 28 Either 30 or 51 but no 32 was formed in the 28 reaction, and independently synthesized 32 was stable to the reaction conditions. o 16 125°

Therefore, the exo , exo oxyallyl intermediate must be the species which undergoes cycloaddition. pH 3 6- -, 80° 125c (-0-16 •6¥?

i 5 o° J exo-endo exo-exo oxyallyl oxyallyl O.0 + dl-16 CO EZO cis -J.4-0-

Woodward-Hoffmann rules predict a disrotatory opening to the exo-endo oxyallyl intermediate. A conrotatory opening would produce the exo-exo oxyallyl inter- mediate in one step. Various attempts to trap the "exo-endo" species have failed. 28

5. Cyclopropanones and Favorskii Rearrangement Conditions Cyclopropanones in closed or oxyallyl form have long been implicated as intermediates, in the Favorskii reaction, conversion of a-haloketones to rearranged carboxylic acids or esters with base in protic solvents. Alternative mechanisms such as the one below have also been suggested.

P. - "0 OR R0 jj H ^U ^X > s v ^J^s^X * R0CCH2CH 3 a' a y^fi *-*

This is known as the semibenzylic mechanism, and is favored if ot 1 -hydrogen is hindered, not acidic, 30 or in some manner inaccessible. 31 The importance of <*' -hydrogen accessibility is seen in the path using cyclopropanone as an inter- mediate. The base must abstract that hydrogen to form the cyclopropanone intermediate. 1 «xA^1

H(H i CH3CH2CO2R

Greene and co-workers 25 recently subjected 16 to Favorskii conditions using t-butoxide and t- butyl alcohol-0-d. The products were do, d x , d2 , and d3 compounds, indicating some base-catalyzed exchange of the a-hydrogens prior to ring opening. The product esters were stable to reaction conditions and did not deuterate under these conditions. In MeONa^feOD and sodiumethylene- glycoxide/dideuteroethyleneglycol only monodeuterated products are found.

C00t-Bu COOtBu COOt-Bu t-BuO" R=t-Bu 16 > R-C-CHDR + R-Q-CjB. + RCDCD2R t-BuOD 2D 98$ 50$

Rappe, Turro and co-workers 32 convincingly pinned cyclopropanone, or one of its tautomers, as intermediates in the following case. Product ratios were essentially the same, depending on the base used. The haloketones gave yields of 25-65$ depending on the base used, while the cyclopropanone and the hemi- ketal gave yields in the 90-100$ range.

IV. Conclusion Cyclopropanones are generally believed to be in a labile equilibrium with % their other tautomers. The carbonyl functionality, along with the strained ring, . . -147-

^C— .^S^->^„. \ka i Br ma j or minor 33 2h ~80$ ~20$

cause these compounds to be highly reactive. This reactivity creates problems in synthesizing and isolating cyclopropanones, which can be overcome by synthesis of masked cyclopropanones. These precursors can be purified and stored, and, by careful selection of the precursor and conditions, reactions similar to those with free cyclopropanones can be carried out. The most common and useful of their reactions are cycloadditions and ring expansions to cyclobutanones and P- lactams Bibliography

1. N. J. Turro, Accounts Chem. Res., 2, 25 (1969). 2. N. J. Turro, R. B. Gagosian, S. E. Edelson, T. R. Darling, J. R. Williams, and W. B. Hammond, Trans. N. Y. Acad. Sci., 396 (1971). 3. G. Subrahmanyam, J. Sci. Ind. Res., 32, 34<9 (1973). 4. N. J. Turro and W. B. Hammond, Tetrahedron, 24, 6017 (1968). 5. N. J. Turro and W. B. Hammond, ibid., 24, 6029 (1968). 6. J. G. Burr, Jr. and M. J. S. Dewar, J. Chem. Soc, 1201 (195*0. 7. R. Hoffmann, J. Am. Chem. Soc, 90, 1475 (1968). 8. J. M. Pochan, J. E. Baldwin, and W. H. Flygare, ibid ., 90, 1072 (1968). 9. J. M. Pochan, J. E. Baldwin, and W. H. Flygare, ibid ., 91, 1896 (1969). 10. N. Bodor, M. J. S. Dewar, A. Yarget, and E. Haselbach, ibid ., 92, 3854 (1970). 11. A. Liberies, A. Greenberg, and A. Desk, ibid ., 94, 8685 (1972). 12. M. E. Zandler, C. E. Choc, and C. K. Johnson, ibid ., 96, 3317 (197*0- 13. J. F. Liebman and A. Greenberg, J. Org. Chem., 39, 123 (197*0. 1*+. G. Stork, J. C. Depezay, and J. d'Angelo, Tetrahedron Lett., 6, 389 (1975).

15. F. D. Greene, R. L. Camp, V. P. Abegg, and G. 0. Pierson, ibid ., 42 , 4901 (1973)

16. J. T. Groves and K. W. Ma, Tetrahedron Lett., 11, 909 ( 1974 ) 17. H. H. Wasserman, H. W. Adickes, and 0. E. de Ochoa, J. Am. Chem. Soc, 93, 5586 (1971). 18. B. H. Bakker, G. J. A. Schilder, T. R. Bok, H. Steinberg, and T. J. deBoer, Tetrahedron, 29, 93 (1973). 19. G. Giusti, Bull. Soc. Chim, Fr., 11, 4335 (1972). 20. N. J. Turro, P. A. Leemakers, H. R. Wilson, D. C. Neckers, G. W. Byers, and G. F. Vesley, J. Am. Chem. Soc, 87, 2613 (1965). 21. J. K. Crandall and W. H. Machleder, J. Am. Chem. Soc, 90, 7347 (1968). 22. J. K. Crandall and W. H. Machleder, ibid ., 90, 7292 (1°£B). 23. R. L. Camp, F. D. Greene, ibid., 90, 7349 (1968). 24. J. K. Crandall, W. H. Machleder, and S. Ac Sojka, J. Org. Chem., 38, 114-9 (1973) 25. J. F. Pazos, J. G. Pacifici, G. 0. Pierson, D. B. Sclove, and F. D. Greene, ibid ., 29, 1990 (1974). 26. N. J. Turro and R. B. Gagosian, J. Am. Chem. Soc, 92, 2036 (1970). 27. J. N. Bradley, C. W. Cowell, and A. Ledwith, J. Chem. Soc, 4334 (1964). 28. F. D. Greene, personal communication. 29. D. B. Sclove, J. F. Pazos, R. L. Camp, and F. D. Greene, J. Am. Chem. Soc, 92, 7488 (1970).

30. A. C. Cope, M. E. Synerholra, and E. S. Graham, ibid ., 73, 4702 (1951). 31. E. W. Warnhoff, C. M. Wong, and W. T. Tai, ibid ., 90, 514 (1968). 32. C. Rappe, L. Knutsson, N. J. Turro, and R. B. Gagosian, ibid., 92, 2032 (1970). » -J.40-

THE SYNTHESIS AND CHEMISTRY OF THIOKETONES .

Reported by Robert J. McGorrin May 12, 1975 INTRODUCTION Thioketones, 1,z also referred to as thiones, are a class of bivalent sulfur impounds possessing a carbon- sulfur double bond. Despite a formal similarity, thioketones have different reactivities from their oxygen counterparts, and are com- parably rare in nature because of their inherent instability. The necessary overlap of the carbon 2p-orbitals and sulfur 3p-orbitals is less efficient than the 2p-2p over- lap of the C=0 bond, owing to the larger size of the sulfur 3p-orbitals. Carbon- of the thio- sulfur double bond energies, however, indicate that the resonance energy carbonyl group is large enough to compensate the unfavorable energetics of 2p-3p tt 3 overlapping. . The differences between a carbonyl and a thiocarbonyl group i can be explained on the basis of the disparity in size, polarizability and electronegativity of the hetero- 4 atoms. The TT-bond energy is smaller for the thiocarbonyl group than for the carbonyl group, and hence canonical forms 2 and 2 are important contributors to the bond re- presentation. By a comparison of the dipole moments of carbonyl and thiocarbonyl >=s >-s" >-s' I £ 2

= 5 co- compounds, such as acetone (n. = 2.8o D) and thioacetone (n 2Al D), Julg and workers 6 concluded that the carbon- sulfur double bond moment of 2.0 D is due to the sum of a C-S a- bond moment of 2.8 D directed toward sulfur and a C-S tt- bond moment of 0.8 D directed toward carbon. Aryl thiones with electron withdrawing substituents have smaller dipole moments than corresponding ketones, while conversely those with electron donating groups exhibit larger dipole s.

PREPARATIVE METHODS 4\ A general method for the synthesis of aromatic thiones involves treatment of the gaseous hydrogen sulfide and hydrogen chloride or hydrogen corresponding ketone with q fluoride in an nnert atmosphere. Best results have been obtained with polar solvents at temperatures below 0°. Benzophenone can be converted to thiobenzophenone by treatment with hydrogen sulfide in alcoholic hydrogen chloride, but will not react if hydrogen fluoride is used. ^ It is difficult, however, to obtain aryl-alkyl thioketones such as thioacetophenone using acid catalysts since both the intermediate hydroxythiol polymerization^ Under h and the thione_ under go_ rapid trimerization, dimerization or Ar S Ar OH H2 _RRCILY" \.Ar Y^HRR' , Ar^ ? + ^CHRR' + aA:HRR''^= S^S RRCH>\S RRCH SH HS(-C-S^-H ~ n k ' I isolated and then carefully controlled conditions, the pure trithiane can be readily and tri- pyrolyzed to produce the monomer. Pyrolyses of aryl sulfides', mercaptoles 113 10 Phosphorous penta- . thiolanes have also been observed ' to produce thioketones. heteroaromatic thione §ulfide 1(i > " has been used to prepare aryl thiones, the 5, from the corresponding carbonyl 0- amino thione 6, and aminovinyl thioketone -JJ

C—CH=CH—rNMe2 u- II S

£: R = H 6 1 8: R = CH 3 derivatives in toluene, xylene, or pyridine at reflux temperatures. Attempting to 12 prepare thioketones from sensitive pyrones, Dean and co-workers found that silicon ketone s_ into disulfide and the more active boron sulfide convert non-enolizable in 8ofo yield with silicon thiones in refluxing chloroform, and thus obtained 8 -149- di sulfide. The reaction of cyclic diaryldichloromethanes with substituted thiobenzophenones is a convenient 13 ' 14 method developed by Schonberg for preparing thiones in average 90% yields as shown in Scheme I. Diaryldichloromethanes can also be directly converted SCHEME I » Ar^ • CI Ar ci CsHs :S + Ar' CI Ar^ci R' = Xanthyl to the respective thiones upon treatment with sodium sulfide, sodium hydro sulfide 15 " potassium _ethyl xanthate or thioacetic acid in an ineri_atmosphere. Aromatic thiones were prepared with thiophosgene using the Friedel- Crafts reac- 16 17 tion initially by Gatterman in 1895. Mayer and co-workers obtained five thio- ketones with aryl thiophosgenes in reasonable (40-60%) yields by this method. Thio- phosgene has also been used to prepare aromatic derivatives, but this method gives products of questionable purity and is not generally applicable. 213 In situations where the aromatic ring is activated by electron donors, a methylene 213 or amino- 18 methylene group will react with elemental sulfur to form the thione directly, as exemplified by the conversion of £ and leukauramine 10 into Michler's 19 thioketone 11. Tarbell used the imine of Michler's ketone to obtain a 6cf yield of 11 by treatment

(2-Me NC H CHR 1 2 s 4 ) 2 (£-Me2NC6H 4 ) 2 C= S HpR 190 2: R=H, R'=S 11 10: R-NHp, R'=NH R A new general synthesis of thiobenzophenone derivatives developed R \ N Li by Ahmed and Lwowski20 involves the reaction of ketimine anions with dimethylthioformamide or carbon disulfide, and proceeds through -NR2" a postulated four- membered cyclic intermediate 12. A similar cyclic intermediate 14 has been postulated for production of dimethoxybenzo- 12 phenone from the corresponding diarylket imine 1^ and carbon disulfide. 21

Ar. CS; S—C=S HNCS CHBr 3 Ar' C—NH Ara 2 Ar : £-CH 30C 6 H 4 Ik 22 While hydrogen sulfide was proclaimed by Sen as a general reagent for the conver- sion of ketones into thiones, it is usually unsuitable for the preparation of mono- meric thioketones, since trimers or gem- dithiols are formed predominantly. la Hydrogen sulfide in the presence of amines and polar solvents also strongly promotes the formation of gem- dithiols from ketones. Under acid catalysis, acetone, methyl ethyl ketone and other low molecular weight aliphatic ketones yield trimeric thioketones, while cyclic and higher aliphatic ketones preferentially form gem- dithiols. Previously 22 reported methods using H2 S and MgBrSH for preparation of aliphatic thiones have la since been shown to be incorrect. ' 23 ' 24

Whereas trimeric thioacetone is stable at 200 , other aliphatic and alicyclic gem- dithiols can be converted into pure monomeric thioketones by elimination of hydrogen sulfide under reduced pressure and condensation of the distillate at -4o°C. (Scheme II) SCHEME II Rv^ /SH 200°/l° torr. X=S H2 S R^^SH cat. R %

Sand, active carbon and firebrick have been used as catalysts. Thiocyclopentanone 2 was prepared in 86% yield by this route. " In most instances, low molecular weight monomeric thiones with a—substituents such as 3-methylpentane-2-thione and 2,3- dimethyl-2-cyclopentenethione can be synthesized with H2 S at temperatures ranging -150- 26 thiofenchone 1- phenyl- 2- indane- from -80° to -too. In addition, thiocamphor, &$)> 4 hydroxy- A*-androstene-3-thione U&) thione (l6), A -cholestene-3-thione (lj), 170- from the and other hindered or non-enolizable thiones were easily prepared^at^ corresponding ketones with hydrogen sulfide and acidic catalysts. CN > , —c=c— CSNH2 OC2H5 12 i5

to form Hydrogen sulfide reacts with enol ethers or ketals in acidic media obtained in monomeric thioketones (Scheme IIl)?8,29 Dx-n- propyl thioketone was % a catalytic amount yield from k- ethoxy- 3- heptene in glacial acetic acid containing similar conversion of an c^cyano- j3-ethoxythiocarboxamide 1£ of sulfuric acid. A 3 proceeds using alkaline conditions with sodium hydrosulfide. The cleavage SCHEME III

1 H S,AcOH R_ R-CH2N /OR H2 S,AcOH 2 >0R -R'OH -2R'0H, R/

of diethyl ketals with H2 S enables preparation of thiones in average yields of 50$ under mild conditions in the presence of either ZnCl2 or an anhydrous organic acid. 29 Cyclic ketals derived from glycols are less efficient precursors owing to the difficulty of continuously removing the glycol distillate. Cyclic pr-oxocarb- oxylic esters are especially easy to convert into the corresponding thioesters ia 31 with acidic hydrogen sulfide. Another route to steroidal thiones with quater- nary a carbons is dissolving metal reduction of a dibenzylmercaptal. 32 As with aromatic thioketones, phosphorous pentasulfide is a convenient i 4Veaeent for the preparation of non-enolizable aliphatic thiones, but yields are considerably poorer. The scope of the reaction has been studied by Scheeren, and was found to be favored by moderate temperatures (30-50°), addition of sodium carbonate or bicarbonate, and the use of more polar solvents. In general, rates are higher in acetonitrile, but some thioacetamide by-product is obtained, as acetonitrile is not completely inert to reaction conditions. In addition to gem- thiols, the pyrolysis of a variety of sulfur derivatives may lead to the formation of thiones. Bornyl thione (28$) and cyclohexane thione (31$) were prepared by flash thermolysis at 900 K of the corresponding 35 allyl sulfides. 34 ' Recently Fraser prepared norbornane thione by pyrolysis at reduced pressure of the corresponding trithiane in 76$ yield, but had minimal success in preparing the relatively unstable cyclopentane (2$) and cycloheptane (15$) thiones.

PHYSICAL PROPERTIES Thioketones that are not stabilized by electronegative substituents or aryl groups tend to form equilibrium mixtures by enethiolization (Scheme IV). The ratio

SCHEME IV H S SH

I I iC— c—II of thione to enethiol depends on the temperature and solvent, as higher temperatures 36 and more polar solvents favor the thione tautomer. The characteristic intense color of thiones (blue, red, violet) has resulted in ultraviolet absorption. |) number of spectroscopic studies relating tautomerism to Colorless compounds, or those which exhibit weak chromophores indicate that the major species present has a OS single bond rather than a double bond, as a result of extensive enolization or trimerization. The extent of enolization can be determined 37 ' 38 by ir, nmr and polarographic methods. -151- exhibit between 200 and 600 nm, and Tne ultraviolet spectra of thioketones \aBiX thiones possess at least three absorption bands, one in the visible (n- n* tran- aryl 3 sition) and two in the uv region of the spectrum (tt- tt* transitions). Aliphatic thiones have a characteristic visible absorption near 500 nm, while the enethiol form can be recognized by its intense absorption at 220 nm. Infrared spectroscopy is of less value in the detection and measurement of thioketones than is electronic spectroscopy. There has been considerable difficulty in determining the thione absorption frequency because of its low intensity and its in a very crowded part of the spectrum. In the aliphatic series, ^the^position position 1 of the C= S stretching vibration has been located between 1000 and 1300 cnf . 1 Various thiobenzophenone derivatives have a strong band at 1207 to 122^ cm" that is not present in comparable carbonyls. Analysis of thione- enethiol mixtures obtained from sulfuration of ketones is easily performed by nmr because the chemical shift of the thiol proton is concen- tration dependent. Conversion of a carbonyl group to a thiocarbonyl group results bonded in downfield shifts of 0.5 ppm for the signals of methyl or methylene protons thione to the thione carbon, corresponding to a greater deshielding effect for the 26 ' 43 as compared to the ketone. A few studies of aliphatic thioketones have been conducted using mass spectro- 44 molecular metry. Thiocamphor and thiofenchone give rise to a greater fraction of ions than the corresponding ketones. Aromatic thiones can fragment to produce 45 46 fluorenyl cations by splitting off an SH group. Heller observed the radical the anion of thiobenzophenone in basic ethanol by light excitation (577 nm) within n-rr*band using esr. ADDITION REACTIONS makes it The strongly polar character of the carbon- sulfur bond in thioketones thiocarbonyl quite susceptible to both nucleophilic and electrophilic addition. The with amines to group is easily hydrolyzed in acidic or basic solution. Thiones react and semi- form Schiff bases, and with hydroxylamine and semicarbazides to form oximes and the carbazones. The reactions take place with both the thiocarbonyl compounds 24 isomeric enethiols, and in contrast to ketones, often proceed readily without

catalysts. . Aliphatic thiones react with nucleophilic compounds by addition- elimination oxygen exchange reactions, resulting in the loss of hydrogen sulfide, analogous to the condensation reactions of ketones. Treatment with simple alcohols affords ketals, and SCHEME IV R^JDR" R"0H H2 C(C0 2R") 2 -(CN) 2 C (C0 2R") 2 ~(CN) 2 R' OR" H2 S 47 4* 5 with active methylene compounds yields olefinic products. Yoshida has recently reported the formation of phosphoric ester derivatives by reaction of trialkyl phos- phites with cyclic thiones and gem- thiols (Scheme V). The thiol- and thiolether- intermediate, type products obtained are compatible with the formation of a betaine "" SCHEME V HS. .SH P(0R); P(0R);

followed either by proton transfer or alkyl rearrangement. The related reaction of l,3-dithiacyclohexane-2-thione 20 with trimethyl phosphite affords the phosphorane 21 which is converted to the phosphonate derivative 22 by thermal rearrangement of the methyl group. 49

'

P(0Me) s 3 Me P(0Me) 2 A 2 P(0Me) 3 A 70-80 X S S S S

hr. 2k hr. 22 2Q u 55 , 3 21 R N -152- Diaryl thiones are rapidly converted to gem- di chlorides by benzyl chloride, and to ethylmercapto diarylacetonitriles by ethyl iodide in the presence of cyanide 2a ion. Under basic conditions, p_-nitrobenzyl chloride will condense with the thione to give an unsymmetrical stilbene derivative 2^ either through a postulated thiirane H kr2 C = S + p_-N02C 6H4CH2 Cl CsH^-NOg | Ar2^/H Ar2 C— CH-CsH^-NO^ > Ar2 C=CHC6H4p_- N02 S or \ Q1 22 2h 25 50 51 2J> or chloroethyl thiol 2J+ intermediate. Linn and Ciganek demonstrated the facile nucleophilic ring- opening of tetracyanoethylene oxide with thiobenzophenone to afford tie trithiolane 26 and the dicyanothylene 2J in acetonitrile at 25°- They suggest that 0. (CN) 2 ^(CN); 2 Arp C= S Ar CN Ar2 C = S Ar- "7(CN)j \ CH 3C=N, 25^ iriS Ar CIT + co(cn) 2 26 £1 the reaction proceeds through a dicyano episulfide. Fluorinated analogs of aliphatic thioketones are highly susceptible to attack by nucleophilic reagents because of the strongly electron- withdrawing nature of fluorine. In general, fluorinated thiones are considerably more stable than their hydrogen counterparts because of the strengthened C- S bonding and their inability to undergo 52 enolization. Campbell 53 has observed that Chloramine T 28 reacts with aliphatic thioketones (Scheme VI ), acting both as nucleophile and oxidizing agent to produce SCHEME VI Ar CI RCHo' C=S Ar s— S Ar p-CH 3C sH4SC 2— X Ar' Na H V R' R ! 28 Ar Ar unsaturated disulfide dimers, which can undergo stereospecific Cope-type rearrange- ments to yield 1,^-dithiones. Aryl thioketones such as thiobenzophenone were found to yield trithiolane products only. The reactivity of the carbon- sulfur bond is further exemplified by the reaction of thioketones with electrophilic reagents. The base- catalyzed reaction of aliphatic thiones with anhydrides and alkyl and acyl halides leads to the S-acyl enethiols. Sulfur- alkylated enethiols can be prepared with excess methyl iodide in good yields, 1£ 54 and also from unsaturated methylene compounds to give Michael-type adducts. '/- In basic media carbon disulfide and carbon oxysulfide add to the thiocarbonyl in a reversed manner, and afford trithiocarbonates. lc Middleton and Sharkey55 report SCHEME VII S(0)

C— S— C— SCH 3 *OW/>,

C' CN, (C00R) CH N2 2 or <^ fr- S / .H c-s-(ch cn,(coor) (thione) } 2 ) 2 >/:-; \j&* ^^f * P-S-CH3 — S-C— 1) » the formation of thiiranes by the reaction of fluorinated thiones with diazoalkanes via a thiadiazoline intermediate. The mechanism appears to be a 1,3- dipolar cyclo- addition to the thiocarbonyl group, rather than a carbene insertion as previously — s ' suggested. Diazomethane reacts with aliphatic thiones to produce thiiranes 56 via thiadiazoline intermediates, while in contrast, it reacts with the enethiol form to yield S- methylated derivative 57a (Scheme VII ). Aromatic thioketones dimerize to form 1,3-dithiolanes 2^ by condensation with diazomethane. With disubstituted diazomethane s, both the thiirane and the dimeric 13 57b 58 olefin product can be obtained. > Corey developed an efficient route to the

^ £** PhV7

22 30 21-32

thiirane ^0 which employs the reaction of dimethyloxosulfonium methylide with thio- benzoph|none in 70$ yield. The carbene mechanism has been invoked by Seyferth and SJ Tronich to explain the formation of a dichlorodiphenyl thiirane from the reaction of carbene precursors with thiobenzophenone. OXIDATION AND REDUCTION

Passage of oxygen through a benzene solution of a diaryl thioketone generally produces the corresponding ketone along with sulfur, sulfur dioxide and trithiolane a 6 by-products. > ° The intermediate sulfines formed can be conveniently synthesized 61 62 63 by controlled oxidation with peracids. Latif and Schonberg produced spiro- dioxolanes ^1 in high yield by refluxing aryl thiones with tetrachloro-o-benzoquinone. Sodium borohydride and lithium aluminum hydride are convenient reagents for reduction of both aliphatic and aromatic thiocarbonyl groups to thiols. -In contrast, hydrogenation over transition metal catalysts gives considerable dimeric and olefinic by-products in addition to the mercaptan. Thiobenzophenone s have been completely desulfurized with Raney nickel and by Clemmensen reduction. 6 4

ORGANOMETALLIC REACTIONS The reactions of organolithium and -magnesium compounds with thioketones afford different products that with ketones. 65 Beak and Worley66 noted that the thiophilic addition of phenyllithium and phenylmagnesium bromide to aryl thiones proceeds in accord with a caged radical mechanism to give S- alkylated products. The extent of thiophilic 67 addition has also been investigated .by Paquer and • Vialle. With acyclic

CsHs^ A^C6H5Li(to/ ) C 6H5 R l)R"MgBr R SR" R SH R /C=" S ~ - ' CE S C S >SH y ^ > ,THF » + r tt 7~A X Y CeH 5 B,C 6H 5MgBr(lO/o)C6H5 R' 2)^ R^ H R' H R

R = t-Bu R" - Et Q% 13% 2$ R' = Me R" = i-Pr R" = t-Bu 20/o 1% % aliphatic thiones mixtures of S- addition, reduction, and enolization products result from Grignard reagents. Product ratios depend on the solvent, the type of Grignard reagent used, and the thioketone. Allylic Grignard reagents undergo C- addition with thiobenzophenone, thiocamphor, thiofenchone, and other a- disubstituted ketones giving a 0, y unsaturated thiols which can thermally cyclize to thiolanes £2, while ' vinylic Grignards only react with aryl' thiones to yield vinylic sulfides after rearrangement. 6813 65a Stable alkyldiphenylmethyl radical intermediates have been detected by esr. > 68b COUPLING REACTIONS he di e £ic products formed upon treatment of thiocarbonyl compounds with ^, ™ ? ? metals nave been studied particularly by Campaigne and co-workers. Olefin formation is ob- served by reaction of aromatic thioketones with Raney nickel or copper, which promote rapid de sulfur ization and coupling. 113 Heterocyclic thiones undergo similar coupling lb reactions upon heating, irradiation with uv light, or treatment with triethyl phos- -154- 69 phite. It is possible to generate dimer, trimer, thiirane and olefin derivatives of with or oxide. 52 thiones by treatment sodium acetylides with tetracyanoethylene % DIELS- ALDER REACTIONS Thiocarbonyl compounds are more reactive Diels-Alder dienophiles than the oxygen nalogs. Thiobenzophenone 70 and thiofluorenone 71 produce thermally the expected di- nydrothiopyran adducts with conjugated dienes. Alternatively, 1,2 adducts are formed with dienes when irradiated with uv light, in agreement with Woodward- Hoffmann con- siderations. Aliphatic thiones such as hexafluorothioacetone 55 and thiophosgene also have been observed to act as dienophiles in Diels-Alder reactions with cyclic dienes. PHOTOCHEMISTRY Despite the high reactivity of thioketones and their relatively intense ultra- violet absorptions, relatively few photochemical reactions have appeared in the litera- ture. Ohno72 has extensively reviewed photocycloaddition reactions of thiobenzophenone with olefins and acetylenes. Olefins substituted by electron releasing groups react in the visible region (589 nm) and result in the formation of lj^dithiane derivatives J£i by addition (n -* tt*) of two molecules of thiobenzophenone, whereas olefins with electron withdrawing groups are irradiated at 366 nm (tt- n )to yield thietane adducts J$k. Some olefins stabilized by aryl electron releasing groups may also form thietane 73 derivatives exclusively by phetocyclization with thiobenzophenone s at 589 nm. The

C6H5.

^ I ^C=N—C— CH "Ns^te 3

22 2k 35 2z reactions are highly stereoselective and proceed through triplet state diradical intermediates. Thiobenzophenone undergoes a 1,4 cycloaddition at 340 nm with cyclo- >ctatetraene to give a thiopyran, but with 6,6-diphenylfulvene and acenaphthylene 74 4 it yields 1:1 thietane adducts. Acetylenes condense with an n- TT* triplet state 72 thione to form an isothiochromene ^5, unlike the corresponding ketone. Thio- benzophenone is known to be photoreduced by alcohol or hydrocarbon donors to produce thiols and mixtures of disulfides via the n- tt* triplet state diphenyl sulfhydryl- methyl radical. 75 Photoaddition of thiobenzophenone to allenes has also been observed. Among the few aliphatic' photochemical reactions reported is the recent photo- 78 cycloaddition of adamantanethione 77 or thiophosgene to various olefins to give thietanes. Irradiation of arylalkylthiones in acetonitrile at 470 nm affords N- 79 thioacetyl ketimines $6 in 64-88% yields. Aralkylthiones with y- hydrogens under- go intramolecular Norrish type II photocyclizations in benzene solution to produce 80 81 cyclopropane # arylcyclopentane thiols. Blackwell and de Mayo have obtained a thiol upon 254 nm irradiation of thiofenchone, presumably via intramolecular B- hydrogen abstraction. Analogous reactions are rare in carbonyl photochemistry.

BIBLIOGRAPHY

1. Reviews: a) R. Mayer, J. Morgenstern, and J. Fabian, Angew. Chem. Int. Ed., 2, 377 (1964); b) E. Campaigne in "The Chemistry of the Carbonyl Group," S. Patai, Ed., Interscience, New York, 1966, pp. 917-959; c) D. Paquer, Int.

J. Sulfur Chem, 87, 269 (1972); d) D. Paquer, ibid . , 8, 173 (1973). 2. For earlier reviews: a) E. Campaigne, Chem. Rev., 39, 1 (1946); b) A. Schonberg in Houben-Weyl, "Methoden der Organischen Chemie," Vol. 9, Georg Thieme Verlag, Stuttgart, 1955, P- 704; c) E. E. Reid, "Organic Chemistry of Bivalent Sulfur," Vol. 3, Chemical Publishing Co., New York, i960, pp. 148-180. 3- T. L. Cottrel, "The Strength of Chemical Bonds," 2nd. Ed., Butterworths Pub., London, 1959, PP- 275-276. 4. R. Mayer, S. Scheithauer, and D. Kunz, Chem. Ber., 99, 1393 (1966). 5. M. Demuynck and J. Vialle, Bull. Soc. Chim. Fr., 27^ (1967). 6. A. Julg, M. Bonnet, and Y. Ozias, Theor. Chim. Acta., 17, 49 (1970). 3o"5 7_-__.A 1_Luttringhaus and J. Grohman, Z. Naturforsch. , 106 , (1955)- . • . • ,

-155- • - Sikstrom, J. Org. Chem. 8. R. M. Elofson, L. A. Baker, F. P. Gadallah, and R. A. 29 1355 (1964). • (1962). 9. D. J. Pasto and M. P. Serve, J. Org. Chem., 27, 4665 Commun., 1063 (1971;. 10. N. Sugiyama, M. Yoshida, H. Aoyama, and T. Wishio, Chem. 11. P. M. Weintraub, Int. J. Sufur Chem., 8, 321 (1973)- C, 12. F. M. Dean, J. Goodchild, and A. W. Hill, J. Chem. Soc, 2192 (1969). 7l6 (1968) 13. A. SchonbergandE. Frese, Chem. Ber., 101, 694, 701, 14. A. SchonbergandE. Frese, Tetrahedron Lett., 697 (i960). 15. k'."scBoriberg and E. Frese, Angew. Chem., 76, 98 (1964). 16. L. Gattermann, Chem. Ber., 28, 2869 (l895)7

. 101, (i960). 17. H. Viola, S. Scheithauer, .and R. Mayer, ibid , 3517 (1902). 18. R. Mohlau, M. Heinze, and R. Zimmermann, ibid . , 35, 375 2110 (1946). 19. D. S. Tarbell and V. P. Wystrock, J. Am. Chem. Soc, 68, 20. R. Ahmed and W. Lwowski, Tetrahedron Lett., 36ll (1969). V. Wells, Tetrahedron, 21. H. B. Williams, K. N. Yarborough, K. L. Crochet, and D. 1 , 268 l4, 2l4 (1937). 22. D7 C. Senfj? Indian Chem. Soc', 12, 647 (1935); 13, (1956); Angew. Chem. Int. Ed., 2, 23- R- Mayer, G. Hiller, M. Nitzschke, and J. Hentzsch, 370 (1963). • Ed., Interscience,, New 24. a) R. Mayer in "Organosulfur Chemistry," M. J. Janssen, Livingstone, Aust. York, 1967, pp. 219-240; b) M. Leban, J. Fresco, and S. E. J. Chem., 27, 2353 (197*0 Mayer, Chem. Ber., 25. S. Bleisch and R. 99, 1771 (1966). , , 1fV.M Ser. C, 1046 (1972 26. a) D. Paquer and J. Vialle, C. R. Acad. Sci., Paris, 2J4, j; b) P. Metzner, Bull. Soc. Chim. Fr., 3138 (1972). (1962). 27. E. Campaigne and B. E. Edwards, J. Org. Chem., 27, 3760 b) J. Goerdeler and W. 28. a) R. Mayer and P. Barthel, Chem. Ber., 93^ 428 .(i960); Soc, 4664 (1962). Mittler, ibid., 26, 944 (1963); c) J. F. Cavalla, J. Chem. 29. R. Mayer and H. Berthold, Chem. Ber., 96, 3096 (1963). 30. M. A. McCall, J. Org. Chem., 27, 2433 71962). 31. C. Djerassi and D. Herbst, ibid., 26, 4675 (1961). 32. J. W. Greidanus, Can. J. Chem., 48, 3530 (1970 ). Synthesis, 149 (19 D)- 33. J. W. Scheeren, P. H. Ooms, and R. J. Nivard, Commun., 409 (1974). 34. H. G. Giles, R. A. Marty, and P. de Mayo, Chem. Org. Chem., 2509 (1974). 35. P. S. Fraser. L. V. Robbins, and W. S. Chilton, J. 39, Chemistry," Vol. A. Senning, Ed., 36. R. Mayer in "Sulfur in Organic and Inorganic 3, therein. Marcel Dekker, New York, 1972, pp. 325-337, and references Soc, 80, 6367, 6373 (1950 J; 37. a) Z. Reyes and R. M. Silverstein, J. Am. Chem. b) S. Bleisch and R. Mayer, Chem. Ber., 100, 93 (1967). b). D. Paquer and J. 38. a) F. Duus and S. 0. Lawesson, Ark. Kern., 2£, 127 (1968); )• Vialle, Bull. Soc. Chim. Fr., 4407 (1971 . . o, Boer, Rec Trav. Chim. Pays-Bas., 04,L 39. a) 0. Korver, J.- V. Veenland, and T. J. de Tetrahedron, 10, 110 (I960 J. 289 (1965); b) P. Brocklehurst and A. Buraway, (1964). 40. J. Fabian and R. Mayer, Spectrochim. Acta., 20, 299 Acad. Sci. Paris, Ser. B, \l. a) C Andrieu, M. Demuynch, and Y. Mollier, C R. Tetrahedron Lett., 15(3 U9fi;. 265 127 (1967); b) C Andrieu and Y. Mollier, Trav. Chim. Pays-Bas., 84, 310 42. 0. korver, J. V. Veenland, T. J. de Boer, Rec

Acad. Sci. Paris, Ser. C, 2j4, 77, 43. M/Dagonneau. D. Paquer, and J. Vialle, C R. (1972 ) Brooks, Org. Mass Spec , 5, 44. a) M. M. Campbell, G. M. Anthony, and C J. W. Chem. B, 243 (1973) 297 (1971); b) J. W. Greidanus, Int. J. Sulfur 0, Chem. Ber. 102, 3192 (1969)- 45. 0. Schumann, E. Frese, and A. Schonberg, , 46. J. Heller, J. Am. Chem. Soc, 89, 4288 (1967). C ( Tn£\\ Chem. 76, 157 (19o^- 47. R. Mayer, J. Morgenstern, and J. Fabian, Angew. Lett., 235 (1975)- 48. Z. Yoshida, T. Kawase, and S. Yoneda. Tetrahedron (1967). 49. E. J. Corey and G. Markl, ibid . , 3201 B, 8 (1929); b) G. Hahn, ibid., 50. a) E. Bergmann and J. Hervy7~Chem. Ber., 62 9 3

62, 2485 (1929). , , , r , 2146 51. W. J. Linn and E. Ciganek, J. Org. Chem., 3^, (1969;; Bivalent Sulfur, E. E. Reid, m. 52. W. J. Middleton in "Organic Chemistry of Chemical Publishing Co., New York, 1965, PP- 359, 368. ^ ...... 5

-156- 53- M. M. Campbell and D. M. Evgenios, J. Chem Soc, Perkin Trans. I, 2866 (1973)-

5^. J. Morgenstern and R. Mayer, J. Prakt. Chem. , 3*_4, H^ (1^66).

55- W. J. Middleton and W. H.' Sharkey, J. Org. Chem.., 30, 138*4, 1390 (1965)-

56. A. P. Krapcho, M. P. Silvon, I. Goldberg, and E. G. Jahngen, ibid . , 32, 860 (197*0. a) D. Paquer and J. Vialle, Bull. Soc. Chim.Fr., 3328 (1969); b) D. H. R.

Barton, F. S. Guziec, and I. Shahak, J. Chem. Soc, Perkin Trans. I, 179*4 0-97*0 • 58. E. J. Corey and M. Chaykowsky, J. Am. Chem. Soc, 87, 1353 (19^5)-

59. a) D. Seyferth and W. Tronich, ibid . ,.91, 2138 (190*9); b) D. Seyferth, W.

Tronich, R. S. Marmor, and W. E. Smith, J. Org. Chem. , 37, 1537 (1972). ). 60. J. W. Lown and J. W. Maloney, ibid , 35, 1716 (1970 6l a) B. Zwanenberg, L. Thijs, and J. Strating, Tetrahedron Lett., 3*4-53 (1967); b) Zwanenberg, L. Thijs, and J. Strating, B. Quart. Rept. Sulfur .Chem. , £, 79 (1970). 62. N. Latif and I. Fathy, Can. J. Chem., 37, 863 (1959). 63. A. Schonberg and E. Singer, Chem. Ber., 96, 1256 (1963).

64. R. Mayer, S. Scheithauer, and D. Kunz, ibid . , 99, 1393 (1966). 65. a) M. Dagonneau and J. Vialle, Tetrahedron Lett., 3017 (1973); c) M. Dagonneau^

P. Metzner, and J. Vialle, ibid . , 367*4 (197*4-); C) M. Dagonneau and J. Vialle, Tetrahedron, 30, 3119 (197*4-7^""" 66. a) P. Beak and J. W. Worley, J. Am. Chem. Soc, 2*b 597 (1972); b) P. Beak ibid *4l*42 ). and J. W. Worley, ., 92, (1970 , 67. D. Paquer and J. Vialle, C R. Acad. Sci. Paris, Ser. C, 275, 589 (1972). 68. a) M. Dagonneau and J. Vialle, Tetrahedron, ^0, *4l5 (±91%J7 b) M. Dagonneau, J. Organomet. Chem., 80, 1 (197*4).

69. G. Scherowsky and J. Weiland, Chem. Ber., 107 , 3156 (197*4-). 70. a) A. Ohno, Y. Ohnishi, and G. Tsuchihashi, Tetrahedron, 25, 871 (1969); b) G. Scherowsky and J. Weiland, Chem. Ber., 107, 3156 (197*4-).

71. J. Martens and K. Praefcke, ibid . , 107 , 2319 (197*0- 72. A. Ohno, Int. J. Sulfur Chem., Part B, 6, 183 (l97l).

73- H. Gotthardt, Chem. Ber., 107 , 1856 (197*4-). 7k. T. S. Cantrell, J. Org. Chem., 39, 853 (197*4-). (l9?l); b) ibid. I . a) A. Ohno and N. Kito, Int. J. Sulfur Chem., Part A, 1, 26 , 8, 427 (1973). 76. H. J. Bos, H. Schinkel, T. C. Wijsman, Tetrahedron Lett., 3905 (1971 )•

77. C C Liao and P. de Mayo, Chem. Commun. , 1525 (1971 )•

78. a) H. Gotthardt, Chem. Ber., 10 , 2007 (1972); b) ibid . , Tetrahedron Lett., ~ - -1221, (1973). '

79- D. S. L. Blackwell, P. de Mayo, and R. Suau, ibid . , 91 (197*4-). 80. P. de Mayo and R. Suau, J. Am. Chem. Soc, 96, 6807 (197*0- 81. D. S. Blackwell and P. de Mayo, Chern. Commun., 130 (1973 )•