Presented to the Faculty of the Graduate Division by Robert John Rosscup in Partial Fulfillment of the Requirements for the Degr

REACTIVITY AM) REACTION MECHANISM S

OF mummy CHIAORIDES

A 'ES IS

Presented to the Faculty of the Graduate Division by

Robert John Rosscup

In Partial Fulfillment of the Requirements for the Degree

Doctor of Philosophy in the School of Chemistry

Georgia Institute of Technology

March, 1960 g 4

AND REACTION ItECHAEISMS

OF METHOXY CHLORIDES

Approved:

• - 1-1[ - 117 {] Er ii

ACKNOWLEDGEMENTS

I would like to extend my sincere appreciation to Dr. Jack Hine for his suggestion of this-problem and for his continued interest in its imPlementation. In addition a debt of gratitude is due him for the benefits I have derived from his willingness to share his knowledge and understanding of chemistry. I would also like to express my thanks to

Dr. Erling Grovenstein, Jr. and Dr. Henry Neumann for having, served on the reading committee. To the American Viscose Corporation, the Rayonier Corporation and the Alfred P. Sloan FOundation I am very

grateful for financial aid that supported my graduate study. Above all I gratefully acknowledge the encouragement and assistance from my wife, without which this undertaking could not have been completed. iii

TABLE OF PATENTS

Page ACENOMEDGEMENTS LIST OF TABLES

LIST OF ILLUSTRATIOES 0 0 0 0 0 0 . 0 0 0 0 0 0 0 OOOOO • 0 vii

SUMMARY 0 OOOOOO 0 0 Ik 0 0 • 0 viii Chapter

I. IDIRODUCTION 1

II. PROCEDURE O OOOO 0 O 0 OOO 0 12 Attempted Preparation of Trichloromethylamine 12 Preparation of Methyl Dichloromethyl Ether...... 15 Preparation of Bis-Nethoxytblocarbonyl)Disulfide . . • • 15 Preparation of Methyl Trichloromethyl Ether 16 Preparation of 21.-Methoxybenzal Chloride • • . . OO . OO 16 Preparation of 2-Metboxybenzyl Chloride 17

Preparation of Formic-d-Acid-d. . . • . • 0 0 OO 18 Preparation of Methyl Formate-d 19 Preparation of Methyl Dichloromethyl-d Ether 19 Reactivity of Methyl Monochloromethyl Ether 21 Reactivity of Methyl Trichloromethyl Ether 22 Reactivity of Methyl Dichloromathyl Ether ...... 23 Comparative Reactivity of Methyl Dichloromethyl Ether and Methyl Dichloromethyl-d Ether 27 Reactivity of 2-Methoxybenzyl Chloride 28 Reactivity of 21.-Methczybenzal Chloride 30 Reagents. 31 Infra -Red Measurements. . . OO . • . OOOOOO 0 33

III. RESULTS .... . 0 0 9 OOOOOO a 0 0 0 0 0 0 • • 35 Solvolysis of Chioromethyl Ethers 35 Reaction Rates of Methyl Dichloromethyl Ether in Isopropanol . 37 Comparative Reactivity of Methyl Dichloromethyl Ether and Methyl Dichloromethyl-d Ether 41 Reaction Rates of 24Metboxybenzyl and 2.-Methoxybenzal Chloride 42 Identification of Products in the Attempted Preparation of Trichloromethyl Arnim. . 0 . . • • • . . . 44 Page

IV. DISCUSSION AND CONOLUSIONS 47

Solvolysis of Methcocy Chlorides 47 The 0(-Elimination Reaction of Methyl Dichloromethyl Ether. • ...... • ...... 62

V. RECOMMENDATIONS 69

APPENDIXA.. . OOOOOOOOO • • 73 APPENDIX B • 67

APPENDIX C. C . 110000000400601,000 00 O0 95

BIBLIOGRAPHY. ••••••••••••• 101;

VITA• • • • • ...... • • ..... • 109

LIST OF TABLES

Table Page

1. Rates of Hydrolysis at 30 0 2

2. Rate Constants for Solvalysis of Chloromethyl Ethers in Ethanol-Ether at 00 ..... . . . . . . , 36

Comparison of the Quantities of Acid and Chloride Ion Produced in the Solvolysis of Methyl Dichloromethyl Ether . 37

4. Comparison of the Quantities of Acid and Chloride Ion Produced in the Solvolysis of Methyl Dichloromethyl Ethpr . 38

5. Reaction of CH3OCHC12 with Potassium Ipopropoxide 40 6. Comparative Reactivity ofCH3CCDC12 and oycga2 with Potassium Isopropoxide 42

7, Solvolysis of rCH30C6H4CHC12 in 5:1 Acetone Water at 30° 4 • 44

8. Infra Red Absorption Maxima in Microns. . .. . . 45 9. Solvolysis of CH30CH2C1 in Ethanol-Diethyl Ether at .., • • 74

10. Solvolysis of pyp2c1 in EthanolDiethyl Ether at 6° . • . . 74

II. Solvolysis of CH OCC1 3 ini Eh tanol-Eh ter at 00 3 75 12. Solvolysis of CH3OCHC12 in Ethancil-Ether at 0 ° 76

13. Solvolysis of CH OCHC1 it Ethanol-Diethyl Ether at 0° . . 76 3 2 14. Solvolysis of cycnci2 in Isopropanol at -14 f 0.5° 77

15. Solvolysis of CH3OCHC12 in Isopropanol at -12 ± 0.50 77 16. Solvolysis of CyCHC12 in t-Amyl. Alcohol at 0° 78 17. Reaction of CRq0CHC12 with Potaesiumt-Amyloxide in t-Amyl. Alcohol at 0° ' 79 18. Comparative Reactivity of CH OCDC1 and OCHC12 with Potassium Iscpropoxide 3 80

19. Solvolysis of 2.--CH3OCACH2C1 in 5:1 Acetone Water at 30° . 81

Table Page

20. Solvolysis of 21.-CH30C6E4CHC12 in 5:1 Acetone Water at 300 . . 82

21. 30C6H4CHC12 Solvolysis of 2-CH in 5:1 Acetone Water t 30o . 83

300 22. 30C6H4CHC12Solvolysis of 2701 in 5:1 Acetone Water at .

23. 30C6H4C4C12 Solvolysis of 2-CH in 5:1 Acetone Water at 300 . 85 24. Solvolysis of g-01 10C6Hh HC12 in 5:1 Acetone Water in the Presence of Added Chloride 86 25. First. Order Rate Constants at Zer Ionic Strength and Mass -law Constants

26. First Order Rate Constants at Zero Ionic Strength and Mass-law Constants 98 vii

LIST OF ILLUSTRATIONS

Figure Page 1. _Potential'Energy Diagram. 48 2. Infra-red Spectrum of Hofmann Reaction Product...... 88 3. Infra-red Spectrum of Hofmann Fraction 2 ...... 88 4. Infra-red Spectrum of Methyl Trichloromathyl Ether 89 5. Infra-red Spectrum of 21:-Methoxyben;a1 Chloride 89 6. Infra-red Spectrum of 11.-Methozybenzaldebyde,. 90 7. Infra-red Spectrum of pi:MethacYbenzyl Chloride 90 8. Infra-red Spectrum of 2.-Mathoxybenzyl Alcohol 91 9. Infra-red Spectrum of Partially Deuterated Methyl Formate 91 10. Infra-red Spectrum of Methyl Formate 92 11. Infra7red Spectrum of Partially Deuterated. Methyl Dichloromethyl Ether, ... o . ...... • • • • 92 12. Infra-red Spectrum of Partially Deuterated Methyl

DichloromethyI Ether. , . , . • , 0 0 0 - 0 .. 0 ... . 93 13. Infra-red Spectrum of Methyl Dichloromethyl Ether 93 14. Infra-red Spectrum of Methyl Dichioromethyl Ether 94 15. Infra-red Spectrum of Trichlorobromomethape . 0 . . • . • • . 94 16. Graphical Determination ofce and lc! • . • . . . . . , • • . 102

17. Graphical Determination ofce and . . .. * .. . 103 viii

SWIM'

The effect of 0(-halogen on the reactivity of other halogen atoms attached to the same carbon atom has been elucidated by previous workers for some phenylhalomethanes. It was found that added chlorine or bromine increased the Sigl reactivity. This enhanced reactivity was attributed to a resonance interaction between the O(-halogen and the electron deficient carbon atom of the incipient carbonium ion that tended to lower the energy of activation. Thua a structure with a positive charge on halogen contributes to the total structure of the carbonium ion.

/X -C-H - C - H

It was established that chlorine was better than bromine at enhancing the Snl reactivity in spite of the fact that chlorine is more electro- negative. Therefore, it appeared that the predominate electrical effect of the added halogen was a resonance interaction of electron donation rather than an inductive withdrawal of elections. Since chlorine has orbitals which are more similar energetically and comparable in size than bromine to the unoccupied orbital of the positively charged carbon atom of the carbonium ion, C-chlorine would be expected to exhibit the larger resonance interaction.

In the present investigation, however, it has been found that the

SNI reactivity, in la diethyl ether-ethanol, of Methyl chlorOmethyl ether and methyl dichloromethyl ether is reduced by thOubstitution of ix

0(-chlorine for O(-hydrogen. For example, methyl dichlotomethyl ether was found to be 80 times less reactive, per chlorite, than was methyl chioromethyl ether. The methazy group, which is common to both compounds, can affect the rate of ionization several ways, by an electron withdrawing effect which will tend to decrease reactivity, by resonance stdbilization of the carbonium ion which will tend to increase the reactivity and by resonance stabilization of the reactant which will tend to decrease the reactivity. The reactivity of all the chloromethyl ethers shows that the inductive effect of the methoxyl group is not predominate and that the resonance stabilization of the carbonium ion is important. But these two effects should be the same for both the ethers. Therefore, resonance stabilization of the reactant may explain the observed decrease effected by O(-chlorine. Thus contributions of structures such as

,H CH3 0= C\ Cl - Cl 2 equivalent forms to the total structure of the reactant are of importance in determining changes in reactivity. The first order solvolyses of 11.-metboxybensyl chloride and 2-methoxybenzal chloride have been investigated in a solvent composed of acetone and water present in a volume ratio of 5:1. In this case, OEchlorine did not effect a decrease in reactivity, but the increase was somewhat less than the increase observed in going from benzal chloride to benzotrichloride. By analogy with the chloromethyl ethers, it is x

suggested that contributions of ionic structures to the total structure of the reactant is responsible for this smaller increase in Sill reactivity.

The base catalyzed solvolysis of methyl dichioromethyl ether has also been investigated. This compound is of interest because it is a possible intermediate in the methoxide catalyzed methanolysis of chloroform. If the manner by which methyl dichioromethyl ether undergoes base catalyzed solvolysis can be ascertained, it could lend support to the suggestion that dihaloethers or alkoxyhalomethylenes are intermediates in the base catalyzed solvolysis of haloforms. The two possible reaction paths are (1) nucleophilic attack by base on carbon with the concurrent loss of chloride ion, commonly referred to as Se, or (2) nucleophilic attack by base on hydrogen with the concurrent, or subse- quent, loss of chloride ion to form methoxychloromethylene which then reacts rapidly with the solvent. The latter path has been designated as

0k-elimination. Previous workers have shown that the substitution of

0(-chlorine for (X-hydrogen results in a decrease in second order reactivity in those cases where the mechanism is known to be S n2. The observed decreases have varied from 30 to 1,000 fold. However, when the introduction of (A-chlorine results in a compound that can react by the alpha- elimination mechanism, it is accompanied by a considerable increase in second order reactivity. In the present investigation it has been estimated that potassium isopropcccide reacts with methyl dichloromethyl ether with a second order rate coefficient not more than 4 times smaller than in its reaction with methyl chioromethyl ether. While this decrease is considerably less than has been observed before, this is not sufficient evidence to prove that the dichloroether is reacting by an xi

alpha-elimination mechanism and the monochioroether is reacting by an se path. Therefore, partially deuterated methyl dichloromethyl ether was prepared and its reactivity toward potassium isoProPoxide was investigated. The second order rate constant for the reaction of the partially deuterated methyl dichloromethyl ether with potassium isopropoxide in isopropyl alcohol at -U° was found to be 0.371. 0.01 1. mole -1min. while-1 for the nondeuterated ether the second order rate constant was 0.88 * 0.034 1. mole -1 min. -1 . Allowing for the hydrogen content on the C4-carbon atom (0.16 to 0.33 atom fraction), the kinetic isotope effect is at least 3.2 and not greater than 7.6. An effect of such magnitude for an Se reaction, the only plausible alternative to an(X7eUmination mechanism) is quite unlikely in view of reports by other workers which indicate that only a slight isotope effect, varying from none to 1.04, it observed in Se reactions. From the .isotope effect observed in the present investigation it folloVs that a proton transfer is occurring in the rate determining step. It 3s also saki ent that the reaction does not involve the rapid reversible formation of an intermediate carbanion. If such were'the case, the methczydichlorocarbanion formed could abstract a proton from the solvent and revert to reactant and no isotope effect would have been observed. While the present results show that the base catalyzed isopropantlysis of methyl dichloromethyl ether is, initiated by the elimination of hydrogen chloride, they do not preclude the possibility that a carbanion is being formed as an intermediate that reverts to reactant at a rate comparable to which it loses a chloride ion. Therefore, it is rec ed that methyl dichloromethyl ether be prepared with an q-deuterium content of at least 0.98 atoms per molecule and its base catalyzed solvolysis studied. CHAPTER I

INTRODUCTION

The effect of a-halogen atoms on the reactivity of other halogen atoms attached to the same carbon atom has been extensively investigated. Studies have been carried out to elucidate these effects on V reactivity, on se reactivity and on the reactivity of compounds that undergo nuclepphilic substitution on carbon by an alpha-elimination mechanism (1). The effect of a-halogen on Sill reactivity has received considerable attention. Since this reaction proceeds by the rate determining ionization of a carbon halogen bond, Hughes (2) has suggested that an electromeric release of electrons, by a second halogen, toward the reaction center should accelerate this ionization. Stabilization of the resultant carbonium ion by a resonance interaction between the filled orbitals on the remaining halogen and the orbitals on the electron deficient carbon atom is a consequence of this electrameric release. In order to lend to the understanding of these concepts, Hine and Lee (3) studied the rates of hydrolysis, in 50% aqueous acetone, of some of the a-chloro and 00)ramo derivatives of toluene. Table 1 contains a list of the rate constants obtained for the various compounds.

(1) For significance of the terms SN1„ SN2 and alpha-elimination, see J. Hine, Ph sical Organic Chemistry, McGraw-Hill Book Co., Inc., New York, N. Y., 19 Chap. 5. (2) E. D. Hughes, Transactions of the Eamatla Society, 37, 603 (1941). (3) J. Hine and D. E. Lee, Journal of the American Chemical Society, 21 22 (1951).

Table 1. Bates of Hydrolysis at 306

Compound k x 104 min.' k x 104 min.-1 per per Br Cl

C6H5CH2C1 0.2231 ± 0.0041 0.2231 C6H5CH2Br 5.684 t 00085 5.684 2.214 t 0.039 c6115cm12 1.107 C6H5CHC1Br 31008 t 0.19 31008 C6H5CHDr2 6.847 ± 0.047 3.423 ceve13 110.5 ± 0.9 36.8 C6H5CC12Br 2122 ± 41 2122

C6H5CC1Br2 1803 ± 59 901 C6H5CBr3 1131 ± 21 377 a ce5CF201 0.0419 0.0419

aDatum on this compound is from (6).

It is apparent from these data that the influence of a-chlorine has been one of acceleration on both the C-Cl linkage and the C-Br linkage. In those cases where two halogens are already present, it i8 also evident that the replacement of a-hydrogen by bromine causes increased reactivity of the C-Br linkage. These effects were provision- ally attributed to carbonium ion stabilization.

X X+ /1 C m H C m H 3

Since chlorine has orbitals which are more similar energetically and more comparable in size than bromine to the unoccupied orbital on the positively charged carbon atom . of the carbonium ion, ofr.chlorine , wouldle expected to exhibit the larger resonance interaction. This explains the fact that the rate is increased more when a-chlorine is introduced into benzal bromide than when a-bromine is the added halogen. In the former case the. rate is increased 901/3.423 fold per bromine, while the latter change only causes a 377/3.423 fold increase, per bromine, in rate. Bensley and Kohnstam (4) have studied the effect of a-chlorine on the ionization of the C-Cl linkages in benzal chloride and in chlorodi- phepylmethane. Their results on the hydrolysis of benzal chloride and benzotrichloride in 50% aqueous acetone are in good agreement with those reported by Hine and Lee (3). They also found that the increase in rate effected by the added a-chlorine in benzal chloride was due to a decrease in the energy of activation rather than to changes in the entropy of activation and concluded that this is in agreement with the concept that, a-chlorine acts as an electron donor in the transition state. The o native reactivity of chlorodiphenylmethane and dichlorodiphenylmethane was determined in absolute ethanol at 0° and 19.97°. The introduction of a-chlorine was again found to result in an increase in the rate of heterolysis of the C-Cl linkage already present. Due to the uncertainty involved as to the intimate mechanism by which benzyl halides react with a nucleophilic reagent in the solvent systems employed, it has been impossible to ascertain quantitatively the

(4) B. Bensley and G. Kohnstem„ Journal of The Chemical Society, 3408 (1955) exact nature of the acceleration in heterolysis of the carbon halogen bond by the substitution of a-halogen for hydrogen, at least for benzyl chloride (5). For example, benzal chloride has been found to be five times as reactive per chlorine as benzyl chloride while benzotrichloride is 36.8 times as reactive as benzal chloride. In addition benzal bromide is less reactive, per bromine, than is benzyl bromide. If the benzyl halides, but not the benzal halides or benzotribalides„ react with the nucleophile bimolecularly„ then the only conclusion concerning 5141 reactivity that can be drawn from available data is that when two ca-halogens are already present the introduction of a third usually resultS in an

increase in the rate of ionization of a carbon halogen linkage . Hine and Lee (6) later reported a seemingly anomalous result. Benzodifluorochloride was found to be considerably less reactive than benzal chloride or even benzyl chloride. It was suggested that the decreased reactivity could.. be attributed to the'bigh electronegativity of fluorine as compared to chlorine (4.0 compared to 3.0) which appears to be more important, at least in this case, than the ability of the halogen atom to share electrons with the carbonium ion by a resonance interaction. However, Hine and Ehrenson (7) recently ttributed the law

(5) For discussions of the complication involved in the case of benzyl chloride, see E. Tommila, E. Paakala 0 U. K. Virtanen, A. Erva and S. Varila, Annoles Academie Scientiarum Fennicae„ Series A II, 1959 0 a. Kohnstam, Journal of fte , Chemical Societ No. 91; also B. Bensley and 4747 (1952); and P. ivort and P. J. C. Fierens, Bulletin des soci t s chimique8 Beiges, a, 975 (1956). (6) J. Hine and D. E. Lee, Journal of the American Chemical Society, 3182 (1952). (7) J. Hine and S. J. Ehrenson, ibid., fs, 824 (1958).

L 5

Sill reactivity of benzodifluorochloride to contributions of structures such as

2 equivalent forms • 4 equivalent forms

2 equivalent. forms 2 equivalent forms to the normal state of the Molecule. There is other evidence from which it might have been deduced that the strength of carbon halogen bonds could be increased by the interaction of substituent groups. Brockway .(8) has suggested that the observed shortening of the carbon fluorine bond in Polyfluorohalotethanes is due to contribution of structures such as F E C Cl F¢ to the normal state of the molecule. Since stabilization such as this is not operative in a carbonium ion, it could be a factor in reducing the ease of heterolytis of a carbon halogen bond. The reported synthesis of trichlortmethyl amine by Ascher (9), using a variation of the Hofmann (10) reaction, also suggested the

(8) Le O. Brockway, Journal of Muist21. Chemistry fl L42, 187 (1937). (9) K. R. S. Ascher, Journal of The Chemical Society, 2209 (1951). (10) For information concerning this type of reaction see Organic Reactions, _ed. in chief R. Adam: John Wiley and Sons, Inc., New York, X. Y. 0 1946, Volume III, page 261. 6

possibility that an interaction between nitrogen and chlorine 'might decrease the reactivity of the carbon chlorine bonds enough to make isola- tiOn of the amine possible. There do not appear to be any other reports in the literature concerning the isolation of compounds of the type H2ECX3 . Bbwever Henderson and Nhcbeth (11) concluded that compounds such as H2NCH2Br, H2EC 2 and. N2NCC13 were formed by the action of titanous chloride on the corresponding nitromethanes. No attempt was made to isolate any of these compounds and so the evidende is of an indirect nature. The reaction of oc--haloacetamides with hypohalites has been studied recently. It was found by Stevens and coworkers (12) that fiLen.., dihalides were formed in the Hofmann degradation of mono-ahaloacetamide. For example, la-chloroisobutyramide on treatment with sodium hypobromite gave 2,chloro2-braMopropane.. IdOe recently, Mank and Stevens (13), presented evidence that reactions of this type were optically stereo- specific. Optically active oichlorohydrocitnamamide was allowed to react with sodium hypobrcmite and the resultant tellrdihalide 1-bramo-2-pheny17 ethyl chloride, was shown to possest optical activity. This would support a mechanism in which the alkyl group is migrating to bromine without the intermediate formation of a carbanion. Furthermore Rusted and Kohlhase (14) have found that perfluoroaIky1acetamides react with sodium hypohalite

(11) T. Henderson and A. K. Macbeth, Journal of The Chemical Society, Eli 892 -0922). (12) Co L. Stevens, P. H. Mukherjee and V. J. Traynells Journal of the American Chemical Society, 7§„ 2264 (1956). (13) M. E Nhnk and C. L. Stevens Abstracts of Papers Presented at American Chemical SocietyMeetin& Atlantic City, New:Jersey, September 13-18 1959,. Page 53P. (14) D. R., Hosted and W. L. Kbblhase, Journal of the American Chemical'Societyp 761 5141 (1954). 7

to form a perfluoroalkyl halide. The first' member of the series, tri- fluoreacetamide„ did not react with hypdhalite but NA0promotrifluoroace- tamide when treated with base gave bromotrifluoromethane as the reaction product. In view of the preceding reports, a reinvestigation of the reac7 tion of trichloroacetamide under Hofmann-like conditions seemed warranted. Because of the close proximity of the toiling point reported for trichloromethyl amine, 109-110%, and that of bromotrichloromethane„ 104 ° , it seemed possible that the latter was the actual product isolated. There are, however, known compounds containing a strong resonance electron donor and a halogen attached to the sme.caibon atom. Of consid- erdble interest is the series methyl chloromethyl ether (15), methyl dichloromethyl ether (16) and methyl trichloromethyl ether (17). The monochloro compound (15) is known to be very reactive toward 3N - solvolyeis so it was thought to be of interest to deterMine the effect of the substitution of a-chlorine for a...hydrogen on the reactivity of the carbon chlorine linkage. If. resonance interactions between oxygen and chlorine, at least in the'dichIpro and trichioro ptherS,, are significant then the introduction of a-chlorine could result in a smaller increase in reactivity than was observed on going from benzal chloride to benzotrichloride. A further study on the effect of a-chlorine on the Sil reactivity of activated carbon chlorine linkage was also undertaken. TWO compounds,

(15) P. Ballinger, P. B. D. de la Mare, G. Kohnstem, and B. Prestt, Journal of The Chemical Society, 3641 (1955). (16) L. R. Evans and R. A. Gray, Journal, of 01_2E___lic Chemistry, 45 (1958). ?2, 7 (17) 1. B. Douglass and O. H. Warner, Journal of the American Chemical Society, 21 6070 (1956). 8

k-methoxybenzyl chloride (18) and k-methoxybenzal chloride (19), were

prepared in order to measure their rates of hydrolysis in a solvent

composed. of acetone and water in a ratio of 5 to 1. The benzyl chloride

had previously been investigated by Simonetta and Favini (20). No refer-

ence concerning the rate of reaction Of the benzal chloride could be found

in the literature.

The effect of a-halogen on the reactivity of compounds undergoing

_ nucleophilic substitution by an alpha-elimination mechanism has been

investigated by Rine and coworkers (21). This reaction involves the

initial removal of hydrogen and halogen from the same atoms and has been

established for the base catalyzed reaction of haloforms. Hine (22)

proposed that the base catalyzed solvolysis of haloforms proceeds by the

following mechanism where k2 is rate controlling.

k1 C'EX + Cit 3 + ROE 3 k -1

k2 Cx2 X-

fast CX2 Products OR or ROE

(18) J. lee, A. Ziering, L. Burger and S. D. Heineman, Jubilee Vol. Emil Barell, 264 (1946); Chemical Abstracts, Lil l, 6252 b (1947).

(19) H. Schmidt:, Berlohte der deutschen chemischen Gesellschaft, 121., 2331 (1906).

(21 K. Simonetta and G. Favini„ Journal of The Chemical Society :, 1840(1954..

(21) For the most recent paper, see 3. Eine and D. C. Duffey, Journal of the American Chemical Society, 81, 1131 (1959).

(22) J. Hine ibid., Ep 2438 ( 1950). Hine and Ehrenson (23) have correlated the relativezeactivities of ten haloforms, none of which contained more than one fluorine, toward hydroly, sis in aqueous solution in terms of an equation based on the above mechanism. From this correlation they concluded that fiuOrine appears to faCilitate dihalomethylene formation better than any other halogen. Previous work (24) 0 deuterium exchange, had indicated the existence of the initial equilibrium for haloforms containing fewer than two fltorine atoms but that the presende of even one fluorine atom was sufficient to dedrease the rate of exchange even though the observed rate of hydrolytis was increased. Hine and Langford (25) have shown that when two fluorines are present, at least for bromodifluoromethane, the loss of hydrogen and bromine is a Concerted process. That is, the reaction proceeds directly from reactants to the difluoramethylene, the carbanion never having any existence, and subsequently to products. They found that deuterobraMo- difltoromethane did not exchange hydrogen with the solvent, water, and that its rate of hydrolysis was approximately one-half of the rate of the non-deuterated haloform. The fact that this difluorohaloform undergoes hydrolysis by a concerted mechanism was attributed to the destabilizing influence fluorine, relative to the other halogens, would have on the carbanion and the stabilizing influence it has on the divalent carbon intermediate.

J. Hine and S. 3. Ehrenson, Journal of the American Chedical Society, §2, 824 (1958). (24) For leading references, see (25). (25) 3. Hine and P. Langford, Journal of American Chemical 5497 (1957). 10

Methyl dichlorcmethyl ether, a possible intermediate in the meth-

oxide catalyzed methanolysis of chloroform, might also be capable of

undergoing nucleophilic substitution by a base catalyzed alpha-elimination mechanism. There are several ways to determine whether methyl dichloro-

methyl ether undergoes nucleophilic substitution by an Se mechanism or by an alpha-elimination mechanism. The first of these is by a comparison

of the rates of reaction of methyl dichloramethyl ether and methyl chloro-

methyl ether with nucleophilic reagents. Thomas (26) has already evalu-

ated the effect of a-halogen on the rate of displacement of halogen in the methyl halides and has found them to decrease the rate quite markedly. His work could then serve as a guide for the present study. Another method is to determine the effect of added base on the reaction of thio-

phenoxide with methyl dichloramethyl ether. For example, thiophenoxide reacts at a negligible rate with chloroform (22) but in the presence of a base strong enough to effect alpha-dehydrohalogenation, the disappearance of thiophenoxide (triphenylthioorthoformate is the resultant product) is increased tremendously. However, Duffey (27) has reported that potassium isopropoxide does not appreciably accelerate the reaction of thiophenoxide with methyl dichloramethyl ether, at least in isopropyl alcohol-benzene

solution. He also observed that the second order rate constant for the

reaction of thiophenoxide with methyl dichloramethyl ether was approximately

(26) C. H. Thomas, The Effect of Halogen Atoms upon. the . Sn2 Reactivity of Other Halogen Atoms Attached to the Same Carbon Atm, Part I, Ph. D. Thesis, Georgia Institute of Technology, Atlanta, Georgia, 1953. (27) D. C. Duffey, Some Reactions of Methyl Dichloramethyl Ether, Part Four, Ph. D. Thesis, Georgia Institute of Technology, Atlanta, Georgia, 1959. 11

equal to the second order rate constant for the reaction of isopropoxide with methyl dichloramethyl ether.

If methyl dichloramethyl ether does react via an alpha-eliMination„

it might be expected that the influence of the oxygen would-be such that the reaction would proceed with the initial loss of hydrogen and chloride

ions by a concerted process. In order to ascertain the mechanism by which

this ether reacts with nucleophilic reagents, it was decided to prepare methyl dichlorcanethyl-d ether and determine its rate of reaction with potassium ieopropoxide under conditions such that a comparison could be made between it and the nondeuterated ether. Since the tondeuterated ether can be prepared in good yields by the action of phosphorous penta-

chlbride on methyl formate (27), it seemed reasonable to assume that the

deuterated ether could be prepared in this same way using methyl formate-d.

The ester can be obtained. by the esterification of formic-d-acid-d prepared from the decomposition of oxalic acid in the presence of

deuterium oxide. CHUM II

PROD

Attempted Preparation of Trichlotamethylamine.--To 16.5 g. (0.1015 mole) of trichloroacetamide in a 1 liter three-neck round. bottom flask l fitted with a stirrer, reflux condenser and a dropping funnel, was added with stirring and cooling 16 g. (5.2 ml. - 0.1 mole) of bromine. When cold aqueous potassium hydroxide (40 g. in 280 ml.) was introduced into the reaction vessel until the solution turned from red-brown to yellow. (This required 140m1. of the potassium hydroxide solution.) At this time the cooling bath was removed and a heating mantle attached to the reaction. vessel. The reaction mixture was heated to approximately 75 ° and additional potassium hydroxide (]6 g. in 30 ml.) introduced into the reaction vessel. The solution became cloudy after being heated at 75° for 10 minutes; however, it still had a slight yellowish tinge. The heating mantle was removed. and the reaction vessel was immersed in an ice-water bath. After the contents of the flask. had cooled, the stirring was stopped and the reaction vessel was removed from the cooling bath. At this time two phases were discernible in the reaction vessel. The

bottom phase, thought to be the desired product, had a volume , estimated between 2 and 3m1, The major portion of the upper phase was decanted and the remainder was extracted with ether and placed. over sodium sulfate to dry. The decanted portion vas extracted several times with ether and these extracts were placed over sodium sulfate. After sitting overnight the etheral solutions were filtered and the ether was removed by 13

distillation. A small portion of the residue was taken up in a one-fourth ml. syringe and injected into dry carbon disulfide. The infra-red spectrum of this material was then recorded using carbon disulfide in the reference cell, Fig. 2. The remainder of the residue was dissolved in methylene chloride, placed in a 50m1. flask and attached to a fractiona- tion column for distillation. After the head temperature reached 60 ° the residue was cooled to roam temPerature and decalin was added as a chaser. No material could be isolated in the reported boiling range of the desired product. To 33 g. (0.203 mole) of trichloroacetamide was added 32 g (10.4 ml. -'0.2 mole) of bromine with stirring and cooling. Following this, cold aqueous KOH (80 g. in 560 ml,) was added to the reaction vessel. During this addition of base, the color of the solution Changed from red- brown to yellow. This color change occurred after 290 nal. of the base had been added to the reaction vessel. The remainder of the base solution, 270 ml.„ was added with no apparent change in the solution color. The reaction mixture was then heated to 75° and an additional quantity of aqueous potassium hydroxide (32 g. in 60 ml.) that had been warmed to 75° was introduced rapidly into the reaction mixture. The resultant solution became cloudy after heating for 10 minutes at 75e but was ,still pale yellow in color. An adapter was placed in one neck of the reaction vessel replacing the condenser and a 500 ml. round bottam flask was

attached to the adapter. A small amount of, material was , collected by distillation under reduced pressure (water-aspirator) in a flask that was cooled in an ice-water bath during the distillation. The distilla-

tion was stopped when , the , reaction mixture became clear. The distillate consisted of two liquid phases which were separated by use of a separatory funnel. The bottam phase (2 ml.) was placed over silica gel to remove any water present. The previously removed condenser was then reinserted in the reaction vessel and the temperature of the mixture was maintained at 75° for one-half hour. The solution became more and more cloudy as the heating progressed and at the time the heat was removed the mixture was intensely clouded and had only a faint yellow cast. The mixture was cooled in an ice-water bath until its temperature reached 5 ° . Two phases were then discernible in the reaction vessel* Nbst of the upper phase was decanted and the remainder of the mixture was then poured into a small separatory funnel and the bottom. phase (2.5 ml.) was drawn off and placed over silica gel to remove any water present. This material had an odor similar to that of carbon tetrachloride. The dried liquid was com- bined with the liquid dried previously and distilled from a 10 ml. Claisen

flask into a test tube, cooled in an ice-water bath , and a liquid having b.p. of 108° was isolated. There was a strong odor of phosgene associated with this sample. A few drops of material having a b.p. of approximately 84° was also isolated. This latter material had an odor similar to chloroform. The infra-red spectrum of the material having a b.p. of —108* was recorded using carbon disulfide as the solvent and reference, Fig. 3. A sodium fusion (28) was run on this material and the following results obtained: nitrogen - negative, chlorine - positive and bromine - positive. Another experiment was carried out using 16.5 g. of trichloroacetamide. This experiment differed from the preceding in that the distillate

(28) For the procedures used in these determinations see option C under nitrogen and option E under halogens in R. i,. Shriner and R. C. Fuson, The Systematic Identification of Organic Compounds, 2nd ed., John Wiley and Sons, Inc., New York, N. Y., 1940, pp. 112 ff. was collected at 100° under atmospheric preseure. The refractive index of this organic material was recorded, 45 1.5033. Preparation of Methyl Didhloromethyl Ether (29).--To 113.5 g. (0.546 mole) of phosphorous pentachloride in a 250 ml. round bottom flask was added dropwise 32 g. (0.534 mole) of methyl formate (Eastman, White Label). During the addition of the ester the reaction flask was cooled with an ice-water bath. After the addition of the ester was complete, the ice bath was removed and the reaction mixture at overnight (8 hours) at roam temperature. Then a heating mantle was attached. to the flask and the mixture was heated for 5 hours. A few drops of material were collected as the heat temperature rose to 84° Instillation was continued until the bead temperature reached 95°. This crude distillate was then rectified through a.200 mm. vacuum jacketed column packed with tantalum beligrids.

The portion boiling at 82-83° was collected, 50.8 g. (80 yield)/ and found to have n24ip 1.4274. This agrees fairly well with that reported by Duffey (29). 45 1.4264. Preparation of Bis-(MethoxYthioCal'honY1)Disulfide (17).--Over a one hour period 120 ml. of carbon disul4da'llaa added to a stirred solution of 80 g. of sodium hydroxide, 300 ml. of water and Iwo ml. of methanoll cooled by an ice-water bath. The mixture was then stirredfor 30 minutes at which time it was bemogeneous. Then 0.1 g. of Potassium iodide was added and chlorine gas was bubbled through the solution until an iodine color appeared. This color was discharged with a 10 per cent sodium

(29) D. C. Duffey, Same Reactions of Methyl Dichloramethyl Ether, Part Ftour, Ph. D. Thesis, Georgia Institute of Technology, 1959; of H. Fischer and G. Weaker, Roppe-Seyler's Zeitschrift fur siologische cbemie, 2 1 (1941); Chemical Abstracts, 3102 mv3 16

bisulfite solution and the xanthate disulfide was separated, washed repeatedly with water and placed over calcium chloride to dry. The weight of the crude disulfide was 110.5 g. (45% yield). Preparation of Methyl Trichloramethyl Ether (17).--Binety-seven and four tenths g. of crude bis-(methoithiocarbonyl) disulfide was placed in a 200 ml. flask and immersed in an ice-salt water bath while chlorine was passed into the material at such a rate that the temperature did not rise above 10° . Chlorine was passed into the disulfide for 35 hours and 10 minutes during which time the weight of material in the reaction vessel increased by 1745 g. (total theoretical increase was 224 g.). The reaction vessel was then removed from the ice bath and allowed to stand at roam temperature for 22 hours. During this period a condenser with a calcium chloride drying tube attached was kept in the mouth of the reaction vessel to exclude moisture. The reaction miwture was then distilled through an 18 inch vigreaux column. Fraction 1: b.p. - 55-60°9 volume - 50 ml. of SC12 (red liquid). Fraction 2: b.p. - 105-110°, volume - 15 ml. The second fraction was slightly yellow in color. To remove this coloration : the material was cooled in an ice- water bath and cyclohexene was added slowly. The resultant solution was then attached to an 18 inch column packed with glass helices and distilled. Approximately 8 mi. of material were collected having b.p. 108-109° , 25 1.4500. The reported refractive index of methyl trichloramethyl ether is nr 1.4520. The infra-red spectrum was recorded neat: Fig. 4. PreparationgP-Metboxybenzal Chloride (19).--In a 100 ml. three-neck round. bottom flask fitted with a ing funnel, and a condenSer was placed 42 g. (0.202 mole) of phosphorous pentachloride and 20 ml. of 17

absolute diethyl ether. Then 23.8 g. (0.175 mole) of anisaldehyde in 20 ml. of diethyl ether was added to the flask. The mixture was refluxed for 14 hours in a 70° bath. The ether was removed at atmospheric pressure and the phosphorouis oxychloride formed was removed under reduced pressure. A heating mantle was attached to the flask and the remainder of material distilled, yielding 26 mlo of p ■-methoxybenzal Chloride, b.p. 120-121V 8 mm. Reported boiling point was 130V14 mm. The infra-red spectra of the product, Fig. 5, and also of the starting aldehyde, Fig. 6, were recorded using carbon disulfide as the solvent and the reference. Pi'e ration cUIo 1orideChl (18) .- In a 300 ml. three-neck round. bottom flask fitted with a stirrer, a gas inlet tube and a gas exit tube was placed 200 ma- of anhydrous benzene and 20 g. (0.157 mole) of anisyl alcohol. The flask was then lowered into an ice-brine bath at -10° and stirred vigorously while anhydrous hydrogen chloride was passed through the solution. After one hour the hydrogen Chloride gas was no

longer being absorbed and was , flowing freely out of the exit 'ham. The passage of hydrogen chloride into the reaction mixture was discontinued and the contents of the flask were transferred to a 500 ml. separatory funnel, the water separated and the organic layer placed in a 500 ml. Erlenmeyer flask over calcium chloride to remove any other moisture. After drying for one hour the benzene solution was separated from the drying agent by filtration, the benzene removed under reduced pressure and the residue distilled through an 18 inch vigreaux column . After 5 ml. of material had distilledl b.p. 110° 4/10 mm., the residue in the pot solidified. There appeared to be about 15 ml. of fluid present before 18

24 solidification, np 1.5466 Reported (30): nm20, 1.5482, b.p. 110° ' 10 mm. The infra-red spectra of the product Fig. 7, and of the alcohol, Fig. 80 were recorded neat. Preparation of Formic-d-Acid-d.--In a 250 mi. flask, which was attached to a 6 inch vigreaux column, was placed,125 ml. of diethylene glycol diethyl ether and 90 g. (1 mole) of anhydrous oxalic acid. TO this was added 27 ml. of 78 per cent deuterium oxide and 9 m1. of 96 per cent deuterium oxide. A heating mantle was attached to the reaction flask and the temperature of the flask was raised to 125° . The temperature of the reaction mixture was slowly increased over a 2 hour interval to 130 ° while

33 ml. of distillate, b.p. 97-101° 0 was removed. The reaction flask was then cooled to 85° and 36 ml. of 96 per cent deuterium oxide was introduced into the flask. The temperature was again increased to 130 ® and

35 ml. of material, b.p. 97-101° 0 was removed. The reaction mixture was again cooled and 31 ml. of 99.5 per cent deuterium oxide was introduced into the flask. The reaction mixture was then allowed to cool to roam temperature overnight. When heating was resumed, the temperature was increased to 130P and 32 naa of distillate, b.p. 96-101°, was collected over a two and one-half hour interval. The tetperature was then increased over a period of 3 hours to 178P. During this period 16 ml. of distillate b.p. 101-102° , thought to contain formic-d-acid-d was collected.

It was , found that 0.0479 g. of this material dissolved in 25 ma.. of water required 0.683 mmole. of sodium hydroxide when titrated to the phenolphthalein endpoint.

(3C) Mme. 14. A. Briers, Pb Bivort and P. J. C. Fierens Bulletin des socibtes chimiques Beiges, 501 (1956) . L9

Formic acid was also prepared in this manner fram oxalic acid dihydrate. In this experiment 27 ml. of distillate was collected and found to contain 65 per cent formic acid, by weight, by titration of a weighed sample with standard sodium hydroxide. Preparation of Methyl Formate-d.--To a 50 ml. flask containing 16 ml. of 68 per cent formic-d-acid-d, by weight, was added 20m1. of methanol and the flask was attached to a 6 inch vigreaux column fitted with a distillation head. A heating mantle was them attached to the reaction flask and the flask was warmed using a variac setting of 15. The material having b.p• 30-33 ° was collected (15.6 g., 97% yield) as methyl formate-d. The infra-red spectrum of this material was recorded : Fig. 9, using a 10 am. gas cell. A sample of methyl formate was prepared using the 27 ml. of 65 per cent formic acid and the same procedure as for methyl formate-d and its infra-red spectrum recorded, Fig.q 10. This spectrum was found to be identical to that of methyl formate purchased from Eastman (White Label). Preparation of Methyl Dichloromethyl-d Ether.--To 60 g. (0.288 mole) of phosphorous pentachloride in a 125 ml. flask which was attached to a 6 inch vigreaux column was added dropwise 15.9 g. (approximately 0.26 mole) of methyl formate-d. After the addition of the ester was complete a heating mantle was attached to the reaction flask and the reaction mixture was refluxed for 6 hours. A few drops of material were collected as the head temperature rose to 82°. The distillation was continued and an unweighed fraction, b.p. 82-96° , was collected. This fraction on recti- fication through a 10 inch column packed with glass helices yielded 19.3 g. (66% yield), bop. 82-83'. The infra-red spectrum of this 20

rectified material was recorded neat, Fig. 11, and also in carbon disulfide solution, with the solvent as referenee„ Fig. 12. This same method was also used to prepare methyl dichloromethyl ether from methyl formate that had been synthesized in this laboratory. The material boiling in the range 82.5-83e (68.5% yield) was collected to be used in comparative kinetic runs With methyl dichloromethyl-d ether. The infra-red spectrum of this rectified material was recorded neat, Fig. 13, and also in carbon disulfide solution, Fig. 14. The isotopic purity of the methyl dichloromethyl-d ether can be

estimated , in two ways. From the deuterium content of the heavy water samples used in the decomposition of oxalic acid, it can be shown that the dich1oroether should contain 0.84 deuterium.atoms per mole. This value will tend to be too high, however, if any oxalic acid decomposed to formic acid and was not removed:, or if any of the previously added water was not removed by distillation prior to the final treatment with heavy water, and too low if formic acid produced earlier is removed before the last heavy water addition. Since the first two sources of error are at least as probable as the third and since they will have a greater effect on the isotopic content, the value 0.84 is regarded as a maximum. The other method consists in measuring the infra-red spectra of the nondeuterated ether and the partially deuterated ether and assigning the absorption at a given wave length for both to an interaction due solely to a-hydrogen. The spectra were measured in "iso-octane" and the wave length chosen was 13.6 microns. The optical density of the nondeuterated ether (0.123 NO was 1.21 while the optical density of the contaminated deuterated ether (0.122 NO was 0.4. Thus the methyl dichloromethyl-d 21

ether contains at least 0.67 deuterium atoms per mole and if the pure deuterium compound absorbs at all at 136 microns, the deuterium content of this sample must be greater than 0.67 atoms per moles Reactivity of Methyl Monochloromethyl Ether.--Two experiments were carried out at 0° . In the first run the procedure was as follows. Into each of four 100 ml..flasks was pipetted„ at room temperature, 25 na— of anhydrous diethyl ether and 25 ml. of absolute ethanol. (The same pipette was used for both liquids.) These flasks were then placed in an ice-water bath at 0° and allowed to cool for one hour. Then 0.1 ml. of methyl monochloromethyl ether was introduced into one of the flasks using a one-fOurthml. hypodermic syringe. A stop watch was started at the same time the reactant was injected into the flask. When the desired reaction time was approached, the flask was withdrawn from the ice-water bath and swirled in a dry ice- acetone bath at -75°. The watch was stopped as the reaction flask was introduced into the quenching bath (it was found that the temperature of the reaction solvent was less than -200 after 15 seconds and still dropping rapidly) and the elapsed time interval was recorded. After a ehort interval the flask was removed from the dry ice-acetone bath and titrated to the methyl red endpoint with ethanolic sodium ethoxide. The reaction mixture was then allowed to came to roam temperature : to permit the remaining methyl chloramethyl ether to react, and was again titrated to the methyl red endpoint with sodium ethoxide. Each of the four samples was treated in this manner. A fifth sample flask was immersed in the dry ice-acetone bath for several minutes and then 0.1 ml. of methyl monochloromethyl ether was introduced and the flask swirledl removed from the bath and titrated with sodium ethoxide. The initial concentration of methyl monochloramethyl ether was taken to be equivalent to the total amount of base added during the two titrations of sample 4 less the amount required to neutralize the fifth sample. Rate constants calculated from these data are listed in Table 9. In the second run at OP the procedure differed from the first run in that the ground glass etoPpers of the reaction fleas were coated with silicone grease, the reaction flasks were kept in the dry ice-acetone bath during the titration, bra cresol purple was used as the indicator instead of methyl red and each sample was titrated after warming to room temperature. In an independent observation it was found that the base titre of a sample left in the dry ice-acetone hath l with the flask unstoppered, for a period.0 15 minutes increased by approximately 0.02 ml. Presumably this was daze to carbon dioxide absorbed by the reaction mixture. The rate constants calculated from this run are listed in Table 10. Reactivity of Methyl Trichloramethyl Ether.--When 0.1018 g of material thought to be the trichloroether was dissolved in 100 ml. of water and titrated with aqueous sodium hydroxide, it was found that 1.984 mmole.

were necessary to reach the phenolphthalein endpoint. . The sample was then titrated with silver nitrate, according to the method of Mohr, and

1.95 mole. were required. In another experiment . 0.0951 g. of material required 1.873 mmole of base and 1.80 mole of silver nitrate. The equiValent weights of the reactant obtained from the acidimetric measurements were 51.3 and 50.8 and from the halide titrations02.2 and 52.8. The theoretical equivalent weight of methyl trichloramethyl ether is 49.8 (ignoring the contributions of any acid but EC1 in the acidimetric determination 23

One kinetic experiment was carried out at 0° in a solvent composed of 50 per cent ethanol and 50 per cent, diethyl ether by volume as measured at roam temperature. In this experiment the solvent was placed in four 100 ml. flasks and the flasks were than put in an ice-water bath for 30 minutes. Then 0.05 ml. of methyl trichloramethyl ether was injected into each flask and the flasks were swirled in the bath. At various time intervals the flasks were withdrawn l the reaction quenched using a dry ice-acetone bath and the developed acidity titrated to the bramcresol PurPle endpoint with sodium ethoxide. The reaction mixture was then removed from the dry ice-acetone bath, allowed to warm to room temperature and titrated to the endpoint with sodium othoxide. - Each of the four samples was treated in this manner. No of the samples were also titrated for chloride using Mohrts method. In, calculating the rate constants (Table 11), sample one was used as zero reaction. Reactivity of Methyl Dichioromethyl Ether. -Expo. were carried out in a variety of solvents at various texTeratures. o exPeriments were carried out at 0° using the same solvent system (ethanol-ether) as vas used for :methyl monochloromethyl ether. In these experiments approximately 0.05 ml. of methyl dichaoromethyl ether was injected into each of 5 flasks containing 50 ml. of absolute ethanol and 50 ml. of anhydrous diethyl ether that had been cooled in a 0° ice-water bath. The rate of appearance of acid was followed by withdrawing the reaction flasks fraa the ice-water bath, quenching the reaction in a dry ice-acetone bath and titrating to the bromoresol purple endpoint with sodium ethoxide. The flasks were kept in the dry ice-acetone bath during the titration. After this titration was complete, the flasks were allowed to warm to room temperature and titrated again to the bromcresol purple endpoint. A

separate flask containing the solvent was cooled in the dry ice-acetone

bath and 0.05 ml. of the dichloro ether introduced. This sample required .

approximately 0.05 ml. of sodium ethozide for neutralization. In order

to establish the stoichiometryof the reaction of the dichloro ether with

ethanol the samples were diluted with water and titrated with standard

silver nitrate using the method of Mohr. A sample that required 1.22 mole. of base to neutralize the acidity developed by approximately 0.6 moles of ether required 1.24 mmole, of silver nitrate to titrate the

chloride ion liberated. First order constants calculated from these data are listed in Tables 12 and 13.

An experiment was carried out in 50 ml. of isopropyl alcohol at approximately -14 ± 0.5 ° in an ice-salt water bath. The procedure employed was the same as that used for the ethanol-ether run at 0 ° . The

rate constants calculated from these data are listed in Staple 14. Another experiment also was carried out in 25 ml. of isopropyl alcohol atapproxi- mate -12 t 0.5° . The cooling bath used. consisted of diethylene glycol- water that was kept at approximately the desired temperature by use of a

refrigeration unit. In this experiment 25 ml. of uS isopropyl alcohol were added to each of four 100 ml„ flasks and the flasks were

placed in the cold bath for 4 hours. Then 0.25 ml. of a benzene solution .

(8 ml. of methyl dichloromethyl ether and 35 ml. of benzene) of the dichloro ether was introduced into each of the flasks a using a 0.25 ml, hypodermic syringe, and the time of addition noted. The first flask, vas withdrawn from the bath after 2 minutes, swirled in a dry ice-acetone bath `and. titrated to the bromcresol purple endpoint with potassium isopropoxide.

The other three flasks were withdrawn at noted time intervals and treated in the same manner as the first flask. The flasks were then allowed to warm to room temperature and were again titrated to the bromcresol purple endpoint. Rate constants,. Thble 15 0 were then calculated fram these data using sample one as zero reaction. The solvolysis of methyl dichloramettwl ether was also studied in t-amyl, alcohol in an ice-water bath at 0 ® . To each of four 100 ml. flasks was added50 ml. of anhydrous t-amyl alcohol and the flasks were then placed in a 0° bath for one hour. Then 0.05 ml of methyl dichlorameth,y1 ether was introduced into :each flask and the time of introduction was recorded. As the desired reaction time approached, the flasks were individually withdrawn fram the 06 bath, swirled in a dry ice-acetone bath for 30 seconds and titrated with potassium isopropoxide. Because of the relatively high melting point of the t-amyl alcohol, -le, it was necessary to remove the flasks from the dry ice-acetone bath during the titration. However, the flasks were returned periodically during the titration to the bath in order to keep them as cool as possible. The endpoints were found to be stable for only 3 minutes when the titrations were completed. A fifth sample was run by cooling 50 ml. of t-amyl alcohol in an ice-salt water bath at -10e, injecting 0.05 ml. of the dichloro ether into the solution and titrating it immediately with potassium isopropoxide. The initial concentration of the reactant was determined by

allowing all of the flasks to , come to room tegperature and titrating the developed acidity. Rate constants, Table 160 were calculated from these data in the same manner as theywerh for the runs in isopropanol. The second order reactivity of methyl didhloramethyl ether with potassium t-amyloxide at 0 was also investi !tr.: ted. In these exPeriments 26

60 mi. of 0.0314 14 potassium t-amyloxide was placed in a 100m1. flask and the flask was pladed in an ice-waterbath to cool. The 0.1 ml. of methyl dichloramethyl ether was injected into the flask with a 0.25 mle hypodermic syringe and a stop watch started at the ttne of injection. At noted time intervals 5 ml. samples were withdrawn with a pipette and transferred into flasks containing 25m1. of anhydraus acetone that had been cooled in a dry ice-acetone bath. The resulting solutions were then titrated with 0.234 M 11.-toluenesulfOnic acid. From the data obtained, Table 17, it appears that the reaction is too rapid to measure conveniently under these conditions. An experiment was carried out in isopropyl alcohol that ;ermitted the calculation of second order constants for the reaction of methyl dichloramethyl ether with potassium isopropoxide at -12 1 0.5° . In this experiment 100 ml. flasks which contained 25 ml. of 0.0303M Potassium isopropoxide were immersed in the dietbylete glycol cooling bath for 4 hours. Then. 0.25

nl , of a benzene solution (40 ml. of benzene and Blal. of methyl , dichloro-

methyl ether) of the dichloro c. rind was injected into each , reaction flask. At the end of a predetermined reaction time, a flask:vas withdrawn, swirled in a dry ice-acetone bath, titrated with an excess of 0.022 M 2-toluenesulfonic acid and back titrated to the bramcresol purple endpoint with potasSium isopropoxide. After this titration each flasX was removed from the quenching bath, allowed to came'to roam temperature and titrated

once again to the endpoint. A sample was , also roan in which the_flaak containing the 25 ml. of potassium isopropoxide was placed diredtly in the

dry ice-acetone , bath: cooled for 5 minutes and 0.25 :ml, of the benzene solution introduced into it by means of the syringe. Following this the 27

sample was immediately titrated with an excess of acid and back titrated to the brow purple endpoint with potassium iscPrcPcxide. From these data, Table 5, Chap. III, it was possible to calculate approximate second order rate constants. The values obtained in this way were then corrected for the incursion of first order solvolysis. Comparative Reactivity of Methyl Dichloromethyl Ether and Methyl DichlOro- methyl-d Ether.--Duplicate runs were made using the methyl dichloromethyl ether prepared from synthetic methyl formate and methyl dichloromethyl-d ether as prepared in this laboratory. In these experiments an attempt was made to run the two compounds under identical conditions. Since the cooling bath used diethylene glycol, would only hold k samples conveniently, the reactions were carried out using 2 samples of each compound. In the first run the procedure was as follows. To each of four 100 ml. flasks was added 25 ml. of a 0.02845 VI potassium isopropoxide solution. These flasks were then placed in the cooling bath for 6 hours. Then 0.25 ml, of a solution consisting of 5 ml. of benzene and.1 ml. of methyl dichloromethyl ether was injected into the first flask, a stop watch started and the flask swirled in the bath. As the desired reaction time approached, the reaction flask was removed fram the bath, swirled vigorously in a .dry ice-acetone bath, titrated • with an excess of 0.0363 Itztoluenesulfonic acid and back titrated with potassium isopropoxide to the bromcresol purple endpoint. The flask was then removed from the dry ice-acetone bath, allowed to sit at room temperature and later again titrated to the endpoint with potassium isopropoxide. To the nett flask was then added 0.25 ml. of a solution consisting of 5 ml. of benzene and 1 ml. of methyl dichloramethyl-d ether, a stop watch started and the flask swirled in 28

the bath. As the reaction time used in the preceding sample approachedl this sample was removed from the bath and treated in the same manner as was dapple one This procedure was then repeated for the other two flasks' in the bath and later for four other samples. Each set of 2 samPlas was allowed to react during different time intervals. In order to calculate the initial concentration of base and of the dichioro ethers, two fladks containing 25 of the potassium isopropoxide solution were cooled in a dry ice-acetone bath and 0.25 ml. of the benzene solutions of the reactant ethers was introduced into each flask. Each sample was, then titrated with an excess of.11.-toluenesulfonie acid and back titrated with standard potassium isopropoxide. From this data, and the infinity titres, the initial concentrations can be calculated. In the second run the procedure was like - that- followed in the first run. However, the concentrations of the reacting materials were different. In this run the concentration of the potassium isopropoxide used was 0.0129 R[ and the solution injected into the reaction flasks was' made up from 101ml. of benzene and 1 ml. of the appropriate dichloro compound. From the collected data it was possible to calculate approximate second order rate constants. These values,. uncorrected for the incursion of first order solvolysis: are listed for run 1 in Table 6 0.

Chap. III and for run . 2 in Table 18, Appendix A. Reactivity' of p-Methoybendyl Chloride.--First order rate constants for the reaction of rmethoxybenzyl chloride were determined in a solvent composed of 83 1/3 per cent acetone and 16 2/3 per cent water, by volume: . as measured at room temperature. In' the first experiment 25 ml..of acetone and 6 ml. of water were placed in each of four 100m1. flasks 29

and the flasks were stoppered and placed in a water bath for 30 minutes. Then to each of the flasks was. added 5 ml. of an acetone solution containing the benzyl chloride (approximately 0.7 g. in 40 ml. of acetone) which also had been in the water bath for 30 minutes. As the reactant was introduced into the first flask, a stop watch was started and as the reactant was put in other flasks: the time was noted. Each flask was swirled for a few seconds after the reactant was admitted. At various time intervals flasks were withdrawn, swirled in a dry ice-acetone bath and titrated with ethanolic sodium ethoxide to the bromoresol purple endpoint. The flask was then withdrawn from the quenching bath and allowed to warm to roam temperature. Periodically the developed acidity was titrated until the reaction had apparently gone to completion. The reaction mixture was then left in the flask for at least eight hours to be sure reaction was complete. The reactivity of this compound was quite low and the first two samples differed so little in base titre that sample 2 was not used to calculate a rate constant and sample 1 was considered as zero reaction in calculating rate constants from samples 4 and 5. The same procedure was used for another experiment under the same conditions. The reactant solution contained 0.52 g. of 2-methoxybenzyl chloride dissolved in 50 na, of acetone., In addition one sample was run by cool.. ing the reaction flask in the dry ice-acetone bath and introducing the reactant. It was found that one drop of acid was required to change the indicator, bromcresol purple, from blue to yellow. This was the same amount that was needed when no reactant was present. Apparent first order rate constants calculated from these data are listed in Table 19. Reactivity of p-Nethoxybenzal Chloride -The solvolysis of this compound

was studied in the same solvent mixture and under the same conditions as was 2-methoxybenzyl chloride. In the first run s howevers aqueous sodium hydroxide was used to titrate the reaction mixture but this method 'was abandoned due to the high reactivity of the benzal chloride. The results

of this experiment are listed in Table 20. In the second experiment the procedure was as follows.. In each of four 100 ml. flasks was placed 25 ml. of acetone and 6 ml. of water and the flasks were placed in a 30° water bath for 30 minutes. Then to each of the flasks was added 5 ml. of an acetone solution (0.9 g. of 20.7.methoxybenzal chloride in 50 ml.. of acetone) of the diohloro compound and each flask was swirled after introduction of the reactant. A stop watch was started when the reactant

solution was put in the first flask and, the time of addition to the other flasks was recorded. The flaaks were then individually withdrawn from

the water baths swirled in a dry ice-acetone bath ana titrated with sodium ethoxide to the bromcresol purple endpoint. Following this the flasks were allowed to warm to room temperature and 'the developed acidity titrated. The zero titre was obtained. by immersing the reaction flask

in the dry ice-acetone bath s introducing the reactant solution and titrating the sample with sodium ethoxide. Apparent first order rate constants

were calculated from these data and are listed in Table 21. . Unfortunately the endpoint of sample 1 was exceeded and no rate constant could be calculated. In the third experiment the procedure was identical to that of the second experiment. The first sample, however ? was withdrawn after 6 minutes, as opposed to 12 minutes in run 2 and the zero titre was 31

assumed to be the same as in run 2 The rate Constants calculated fram these data are listed in Table 7. In the fourth experiment the procedure and initial concentration of reactant were essentially the same as in runs 2 and 3; however, one change in procedure was made. The first three samples in this experiment were started and, stopped individually rather than being all started at essentially the same time. This method permitted the taking of points in as short a time as 30 seconds and 2 minutes and 30 seconds. The 30 second sample was taken as zero reaction for the purpose of calculating rate constants (Table 22). Another experiment was carried out at 30° using the same soltrent system and procedure as in the precedingexperiment; however, the amount ofkmethoxybenzal chloride used in preparing the reactent4cetone solution was 0.414 g. as opposed to 0.9 g. The supple with a reaction time of two minutes was taken as zero reaction. The apparent first order rate constants calcUlated from these data are listed in Table 23. The base used in titrating these samples was potassium isopropoxide. An additional experimentwas carried out using the same procedure and initial concentration of reactant as in the preceding experiment but in the presence of 0.0272 M sodium chloride. Apparent first order rate constants calculated from these data are listed in Table 24. The base used in titrating these samples after the reaction period was potassium isopropoxide; however, aqueous sodium hydroxide that was 1.056 times as concentrated as the potassium isopropoxide was used as the titrant after the solution had warmed to roam temperature.

Reagents.--Trichloroacetamide, m ap. 140* , prepared by treating tht'acid, obtained from Merck and Co., with phosphorous pentachloride and then aqueous ammonium hydroxide was used in the attempts to prepare trichloromethylamine. Eastman White Label methyl formate, and phosphorous pentachloride from Merck and Co. and the Baker Chemical Co. were used in the preparation of methyl dichloromethyl ether. Commercial absolute ethyl alcohol from the U. S. Industrial Chemical Co. was dried using Smith's (31) method as modified by Manske (32). Diethyl ether was dried with sodium metal. Isopropyl alcohol,. 99%, from the Will Corp. was allowed to react with metallic sodium. Isopropyl benzoate vas then introduced into the reaction flask and the alcohol was collected by•distilIation. However in later experiments metallic potassium was used instead of sodium as the apparently low solubility of sodium benzoate caused gelation of the alcoholic media. t-Amyl alcohol was distilled from potassium t-amyloxide. Acetone was treated according to the method of Conant and Kirner (33). C. P. sodium chloride from the Baker Chemical Co. was used in the acetone-water solution. Eastman White Label anisyl alcohol and anisaldehyde were used to prepare 21 ,methoxybenzyl and

-metboxybenzal chloride without further purification. Diethylene glycol diethyl ether and %so-octane" were both distilled. from metallic sodium. Carbon disulfide, "Baker's AnalYzed," that was dried over silica

(32.) Et, L. Smith, Journal of The Chemical Society, 1288 (1927).

p2 ) R. R. Manske Journal of the. American Chemical Society, 23, 11 , (1931). (33) J. B. Conant and W. R. Eimer, ibid., L4b1, 245 (1924). 33

gel was used as solvent and reference in some of the infra-red measure-

ments. Anhydrous. hydrogen chloride obtained from, the Matheson Co. was used in the preparation of k-sethoxybenzyl chloride. Methyl chloromethyl

ether from the Matheson Co. was used without further purification. The potassium alkoxide solutions were prepared from the dried alcohols and

purified metallic potassium obtained from the Baker Chemical Co.,

Eastman White Label cyclohexene and "thiophene-free" benzene, Matheson Co., were dried with metallic sodium. 2 .-Toluenesulfonic acid, m.p.

102-104' 0 was obtained from the monohydrate, Eastman White Label by heating it under reduced pressure. Chlorine gas was obtained from Tesco Chemicals Inc. Oxalic acid was obtained. from the dihydrate, Merck

"Reagent" grade, by removing the water as an azeotrope with carbon tetrachloride. Two of the deuterium oxide samples used in the preparation of formic-d-acid-d were obtained from solutions that had been used for other purposes by distillation from basic potassium permanganate and were analyzed using density. measurements. The third sample, > 99.5 per cent, was used as obtained from the Stuart Oxygen Co.

Infra-Red Measurements.--All of the spectra were recorded using a Perkin-Elmer Recording Infra-Red Spectrophotameter, Model 21. In those cases where the spectra were measured in solution; a cell containing just the solvent was placed in the reference beam. When the spectrum of a compound was recorded "neat," an empty cell was placed in the reference beam. The spectra of gaseous materials were recorded using a 10 am. gas cell and no dell was placed in the reference beam. Sodium chloride optics were used for the measurement of all the spectra. 34

In all cases the following machine settings were used; resolution 927, response 2, gain 5, suppreision 5 scale 2, inches per micron and for the products of the reaction of acetamide scanning speeds of 2 and 3 were used, with speeds of 4 and 5 being used in all the other cases. 35

CHAPTER III

RESULTS

Solvolysis of Chlorometbyl Ethers.--The initial concentration of methyl chloromethyl ether was taken to be equivalent to the sUm of the base - required to neutralize the acidity developed during the reaction time and that required after the reaction was allowed to proceed to completion. In the reactions of methyl chloromethyl ether and methyl dichloro-

Methyl ether, the initial concentration was corrected by running a blank sample in which the reactant was introduced into solvent Cooled' to dry ice-acetone temperatures and titrating the acidity.vith sodium et4Oxide. In the case of methyl chloromethyl ether it was found that

0045 ml. of sodium ethoxide was needed to neutralize this aciditY•in a sample whose infinity titre was 40095 ml. With methyl dichloromethyl' ether it was found that 0.05±0.01 ml. was required in a sample -with an infinity titre of 43 ml. of sodium ethoxide. In the reaction of methyl triehloromethyl ether the first sample titrated was taken as zero re = action and the initial concentration of the other, samples were corrected Proportionally according to their infinity titres.

First order rate constants calculated from equation (1) are listed in Table 2

2.303 i;,,„ k t- * A CL) where A 00 is the corrected infinity titre and A is the difference between 36

A ae) and the corrected titre of the sample after the noted time interval, t seconds).

Tol.e R. Rate Constants for Solvolysis of Chloromethyl Ethers in EthanalrEther at 00

Reactant ki.(sec:)

=rope]. 12.1 ± 0.48 x 10-4 ma3-0-ac12 3.0 ± 0.06,x 10 -5

CH:3-0-CC13 0.461 ± 0.021 x 10-6

The first order rate constants for methyl chloromethyl ether were cam= dated from the data in Tables 9 and 10 by assuming that the amount of acid generated during the reaction time was equivalent to the amount of ether that had reacted. In the experiments with methyl dichloromethyl ether, the quantity of ether reacting during any time interval was equivalent to one-half the quantity of the base added to neutralize the developed acidity. This was ascertained by comparing the total moles. of base required to neutralize approximately 0.6 mmole. of ether to the total mole. of silver nitrate required after all the methyl dichloromethyl ether had reacted. The results of such titration on the 'samples listed in Table 13 are shown in Table 3. For methyl trichloromethyl ether, the amount of acid generated during the reaction interval was equivalent to about three times the amount of ether that reacted during this time. This was established by titrations which showed that for samples requiring 1.62 and 3..52 mole. of base, 37

1.6 and 1.54 mmole. of silver nitrate were required to titrate the

chloride formed and that 1.52 mole. of base was required to neutralize

the acidity developed from approximately 0,49 mole. of ether (34).

Table 3. Comparison of the Quantities of Acid and Chloride Ion Produced in the Solvolyvis of MethYlDiObloromethyl Ether

Sample Nhole. of RtONa Mnole. of AgRO3

1 1.22 1.24

2 1.25 1.29 3 1.29 1.31 4 1.13 1.23

Reaction Rates of Methyl Dichloromethyl Ether in Isopropanol.--First

order rate constants for the reaction of methyl dichloromethyl ether in isopropanol were calculated from experiments carried out at several

temPeratures. In one experiment the reaction was carried out at -14 ± 0.5° using a brine-ice bath to attain the desired temperature. In this experiment the reactant was injected into the reaction vessels with a hypodermic syringe and at the desired time intervals, flasks were withdrawn and titrated with sodium ethoxide. The average first order rate constant was calculated from the data in Table 14. The ratio of acid to

(30 However, X. 3, Douglass and, 0. H. Warner, Journal of American Chemical Hociety, 21 6070 (1 allowed methyl triehlorOMethyl Wii7e7.7115:g.) to react with ethanol 9 9g.) under reflux and obtained a 28.25 per cent yield of diethylcarbonate and a small quantity of methyl chloride. 38

chloride formed was ascertained by titrating the completely reacted

samples with sodium ethoxide and with silver nitrate. The results are

shown in Table 4.

Table 4. Comparigon of the Quantities of Acid and ghloride

Ion . Produced in the Oolvolysis of Methyl Dichloromethyl Ether

_Sample Nhae. of Et0Na Mmole, of AgNO3

2 1.34 1.29 3 1.35 1.29 4 1.36 1.31

5 1.17 1.13

An experiment was also carried out at -12_t 0.5° that differed

from the run at -14 ± 0.5° in that , a benzems solution of methyl dichloro-

methyl ether was injected into the cooled solvent rather than injecting

the pure ether. The rate constants were calculated as in the experiment at -14 ± 0.5°. The values obtained for ki(sec:1 ), using equation (1), at -14 ± 0.5° and at -12 t 0.5° were 1.27 x 10 -5 and 1.69 x 10-5, respectively. The rate of reaction of methyl dichloromethyl ether with potassium isopropoxide was studied in an experiment at -12 ± 0.5° in isopropyl

alcohol. The method of introducing the reactant ether was the same as for the solvolysis of the ether, that is, a benzene solution of the ether was injected into the isopropoxide solution cooled in the glycol-water bath. At recorded time intervals a flask was withdrawn, the reaction quenched by swirling it in a dry ice-acetone bath, an excess of 39

p:toluenesulfonic acid in isopropyl alcohol added and then the sample was back titrated to the bromcresol purple endpoint with standard potassium isopropoxide. The fraction of reactive impurities in the methyl dichloromethyl ether was established by cooling a reaction flask containing 25 ml. of 0.0383 M potassium isopropoxide (1 mmole.) in the dry ice-acetone bath, injecting the reactant solution into the flask, adding an excess of 0.02202 Mk-toluemegulfonic acid (S mmole.) and back titrating with potassium isopropoxide (13 mmole.). The mixture was then allowed to sit at room temperature to permit the reactant ether to react completely and titrated again with potassium isopropoxide The fraction of impurities present is then

A - + C A - S+C+D

The concentration of reactive imsurities in every other sample was obtained by multiplying the above fraction times the initial methyl dichloromethyl ether concentration and it was assumed that these reactive impurities reacted instantaneously at the start of the run. The second order rate constants were calculated using equation (2),

2.303 log 11=Ei1 r k2 t(rE-13) E vo-rx where 3 is the initial concentration of methyl dichloromethyl ether in moles/liter, B is the initial concentration of base in moles/liter, x is the concentration of orthoformate produced by time, t (in minutes), and r is the number of moles of potassium isopropoxide consumed per mole of ether. The ratio, r, was assumed to be two in these calculations and 40

was checked by comparing the number of mole. of base and silver nitrate required at infinite reaction. The average value of k2 obtained at -12 0.5° from the values listed in Table 5 was 0.792 1. mole -1 min-1 .

Table 5. Reaction of CH3 OCBC12 with Potassium Isopropoxide

Temperature -12 .17 0.5° Initial i-PrOK 0.0383 M Tos0H 0.02202 M Volume 25 ml.

Reaction Volume of Volume of Final Titre 101 k2 (1. Time (mam.) TosOH (a.) i-PrOK Ga.) of Ether 6110 of i-PrOf) mole-1 min-1 )

44.7 0.8 30.9 a 3 41.0 0.8 29.16 5.6b 6 35.7 0.37 31.63 8.3 9 35.0 0.86 29.80 7.44 12 33.0 1.87 30.89 7.96 25.17 25.0 2.1 32.00 7.83 36.83 18.6 0.8 30.56 8.08 51.75 16.0 1.9 32.28 7091.3

aConsidered to be zero reaction.

Nbt used in computing average or in computing corrected I.

In order to correct for the concommitant first order s4volyeip during the experitents in potassiuM isoProPexide at -12 005 °, an equation (3), which has previously been derived from the rate expression 41

for concurrent first and second order reactions, was used (35).

2.30 1 5 4- (ri/k2)] k2t log (3 ) rE - B - (ki/k2) . E. Csi jk2 rx)

In this equation the kl and kg are the first and second order rate constants, respectively, and all other quantities are the same as in equation (2). Values for k2 were then estimated from the data it Table 5, assuming that kl under these conditions was 1.014.x 10-3 min 1, by the method of successive approximations. The average value obtained for

k2 was 0.755 1• mole -1 min;1 . Comparative Re .tivit off Methyl Dichloromethy1 Ether and Methyl Dichloro- methyl-d Ether.--In these experiments, carried out at ,,I1 05=4: 0.50, the

initial concentration of ether present was calculated in the same manner as it was in the reaction of methyl dichloromethyl ether with isopropoxide at ± 0.50 . Second order rate constants were calculated from

the data obtained using equation (2) and are listed in Table 6. Titration of the reaction samples with silver nitrate again indicated that the ratio

of chloride formed to base consumed was 1:1; therefore, a value of 2 was assigned to r in these calculations. Values of kg, corrected for first Order solvolysis, were then calculated from the data in Tables 6 and 18 using equation (3) and setting ki equal to 1.02 x 10-3 min-; for both ethers. The average values of k2 obtained in this way for methyl dichloro- methYl ether and the partially deuterated methyl dichloromethyl ether were 0.88 and 0.37 1. mole -1 min71 , respectively. The rate constants for Methyl dichloromethyl-d ether after correction for partial deutera- tion of the methyl dichloromethYl-d ether were 0.116 1. mole-1 min:1

(35) D. C. Duffey, 22. cit., page 101. assuming the minimum fraction (0.67) of mono-deutero ether present and 0.273 1. mole-1 min :1 assuming the Maximum fraction (0,84) of mono- deutero ether was present.

Table 64. Comparative Reactivity of CH3OCDC12 and CH3OCHC12 with Isopropoxide

Temperature -11.51 , 0.5° Initial 1 -PrOK 0.02845 M Tos0H 0.0363 14 Volume 25 ml.

Reaction Volume of Volume of Final Titre Final Titre 1& kg 1 Time Tos0H i-PrOK CH3CCH012 CH30CDC12 (L. mole- rmin.) (ml.) (m1.) min-1 )

0 21.00 3.50. 3704a im me •••■•

0 20.00 2.08 36.92b ••11 6 16.50 1.30 29.82 9.29 6 18.13 1.70 34.03 4.28

12 15.83 3.80 34 .12 8.59

12 16.50 1.40 AMIN. 35.09 4.01 24 11.30 2.93 35.46 9.09 24 15.10 2.77 33.86 4.26

50 7.8o 3.52 35.89 1=, 9.10 50 11.28 2.08 35.43 3.86

a''bUsed in calculating fraction of impurities.

Reaction Rates of 141ethoocybenzy1 and kNethoxybenzal Chloride. !The reactivity of 2-methoxybenzyl chloride, in a solvent composed of acetone and water present in a ratio of 5:1, was investigated at 300 . This was done by adding an acetone solution of the reactant to a flask containing 43

the solvent mixture, allowing the reaction to proceed for a measured time interval, quenching the reaction in a dry ice-acetone bath and titrating the developed acidity with sodium ethoxide to the brompresol purple endpoint. First order rate constants were calculated from the data listed in Table 19 by the use of equation (1). The average ki was found to be

4.96 x 10-5 sec:'1 . Experiments were also carried out under the same conditions on p7methoxYbenzal chloride. In the first of these experiments sodium hydroxide was used as the titrant; however, difficulties were encountered using this material as the reaction mixture would freeze during the titration making the observation of endpoints unreliable. In latter runs an alcoholic base was used to titrate the developed acidity. In studying the reactivity of rmethoxybenzal chloride, a steady decrease of the rate constants as calculated from equation .(1) was observed. Bate constants were calculated from the data collected in one experiment (Table 7) in two ways, by using as zero reaction the titre of a sample in which the reactant was introduced into the solvent cooled in a dry ice-acetone bath, and by using as zero reaction the titre of the first point taken. In another experiment, Table 22, in which the initial concentra-

tion of p7methoxybenzal chloride was 0.013 . M and the point taken after 30 seconds was considered to be zero reaction, first order rate constants varying from 1.39 x 10 -3 to 0.905 x 10-3 secrl were obtained using equation CO. When the initial concentration' of the reactant was lowered from 0.013 M to 0.006 P4 it was found that the integrated rate constants decreased less rapidly (ar, Table 23, Appendix A) than they did at the 44

higher concentration. First order rate constants at zero ionic strength were calculated (see Appendix C) from the data in Tdbles 22 and 23. The values Obtained were 2.9 x 10 -3 and 1.9 x 1D-3 sec:11 respectively.

Table 7, Solvolysis of 2.-CH3OC6H4CHC12 in 5:1 Acetone Water at 300

Initial Concentration 0.013 M

Time Volume of Total Volume 103111 103k1 (sec.) Et0E4 EtONa 61.) (m1.) (sec. -1 ) .(sec. -1 )

0 0.80 24.20 a __b 360 10.93 24.58 1.54 822.5 15,7 24.80 1.18 0.896 1,165 17.6 24.85 1.03 0.800 1,492.5 19.4 24.92 0.988' D.812 1,823 20.6 24.85 0.951 .0.805 2,105 21.55 25.00 0.870 0.797

a,blised as zero reaction in calculating the following rate constants.

When the reaction was followed using an initial concentration of 0.006 . M in the presence of added sodium chloride, 0.0272 m6 the integrated first order rate constants remained fairly constant after ca. 22 per cent reaction. These values are listed in Table 24 of Appendix A.

IdentificationofProthicsinh t .eatiaronofTrichloro- methyl Amine.--During the experiments carried out in an attempt to prepare trichioromethyl amine, it was possible to isolate organic material that had a boiling point corresponding to the reported value for the 45

amine. When the crude product, separated from the aqueous phase by means of a separatory funnel, was distilled, two fractions were obtained;

Fraction 1, with an approximate b.p. of 80 0, and Fraction 2, with a b.p.

Table 8. Infra-Red Absorption Maxima in Microns

Crude Product Fraction 2d C13CBre el3 CH Cl2C0

3.4 ma 5.6 wc =m 3.5m 5.6 vs 8.0 sb 7.65 m 8.0 7.58 8.4 m 8.3 m 8.4 8.3 s -_

8.8 m-s 8.4 s Mae M. 111. 9.3 s 8.8 s 8.8 s 9.9 s 9,9 9.9 vr -_ 10.4 10.4 s 10.4 s 10.8 m 10 vs s 11 s 11 d 12 vs

13-14 s 13-14 s 13-14 13-14 s ADM

aMedium absorption. dSee Figure 3.

bStrong absorption. eSee Figure 15. cWeak absorption. of 108° . Infra-red spectra were recorded for both the crude product and Fraction 2. The maxima from these spectra are listed in Table 8. Also listed in this table are the maxima from the spectra of chloroform and trichlorobromomethane and phosgene (36). By comparison of the absorption

(36) The maxima for phosgene obtained from, C. R. Bailey and J. B. Hale, Philosophical Magazine and Journal of Science, 25, 98 (L938). 46

maxima listed aboie, it appears that the material isolated is actually a mixture of chloroform and trichlorobromomethane. While the spectrum of Fraction 2 was not identical to that of trichiorobromomethane, it is entirely possible that the discrepancies at the shorter wave lengths are due to phosgene formed in the synthetic material during distillation. To further elucidate the composition of the reaction product, a sodium fusion was carried out and positive tests were obtained for only bromine and chlorine. The refractive index of a known sample of trichlorobromomethane was measured under the same conditions as that of the reaction product with the following results: Reaction product 11 21it 1.5033., C13nr ni 1.5032. 47

DISCUSSION AND CONCLUSIONS

Solvolysis of Nethory Chlorides.--Previous studies have shown that the

substitution of a:hydrogen increases the SIT1 reactivity of the carbon chlorine linkage in the cases which have been studied, all of which were

pthenylhalamethanes. In this investigation, however, it has been found

that this same substitution in chloromethyl ethers causes a decrease in SIT1 reactivity. The first order rate constant for the reaction of methyl chloromethyl ether in 1:1 ethanol-diethyl ether was found to be 12 x

l0 sec. -1 which is in agreement with the value reported by Ballinger, et al. (37), 1203 x 10 -4 sec. -1$ in the same solvent. Two other compounds, methyl dichloromethyl ether and methyl trichloramethyl ether,

investigated under the same conditions as the monochloro ccaupound were found to have first-order rate coefficients of 3 x 10-5sec. -1 and 0.461 -6 -1 0.021 x 10 sec. $ respectively. In this series of compounds the

introduction of a:chlorine has not increased the rate of reaction but bas in fact decreased it quite markedly. This might have been the expected result on going from chlorodimethyl ither to methyl dichloromethyl ether if the former reacted with the solvent bimoleaularly, but it would certainly be unexpected that the introduction of another a-chlorine would further decrease the S ill reactivity. However, it has been demonstrated that the solvolysis of chlorodimethyl ether proceeds (37) Ballinger, et al. loc. cit. by an essentially unimolecular path (37). The app rent decrease in reactivity caused by the introduction of a-chlorine must, therefore, be the result of an electronic effect which is manifested to a greater extent in the reactant than it is in the carbonium ion intermediate or the transition state that precedes this intermediate. Before discussing this point. it might be beneficial to describe qualitatively some contem- porary views of transition state theory (38). In order to present these conceptions, use is often made of a plot with potential energy as the Ordinate and the reaction coordinate as the abscissa,

Energy

A Reaction Coordinate----, Fig. 1

This graph is peculiar to thOse reactions in which a meta-stable intermediate is formed dewing the course of the reaction. The small valley in the curve is a consequence of stabilization ,of this intermediate by electroniceffects„ 1, e., resonaneeinteractions. The peak preceding this valley is thought to be representative of the potential energy of the transition state and the intersection of the curve with the ordinate is then the potential energy of the reactant. The difference between

(38) For a more comprehensive discussion the reader is referred to A, A. Frost and. R. G. Pearson, Kinetics and. Mechanism, John Wiley and Sons, Inc., New York: N. Y. 1956p 2nd ed., Chap. 5. 49

these two points is said to be the energy of activation. This change in energy affects the free energy of activation which is directly related to the rate coefficient (39),

(umi = ve, RI

Now if the rate coefficients for the reaction of several compounds that differ only slightly in structure are determined under a specific set of conditions and are found to be different, it can be concluded that this is due to a change in the free energy of activation. That is, if a modification in structure is made such that the interval A-B in the above figure is decreased in magnitude, the effect will be to increase the rate coefficient Whereas a modification that increases the interval A-B would result in a decreased rate coefficient provided of course, that the difference in reactivity of the two compounds is primarily due to a change in the energy of activation and not the entropy of activation. While it is obvious that any structural modification would necessarily change the fXee energy content of a molecule, the important factor is not this change but rather how the interval A-B is affected.

In the case of the chlorodimethyl ethers, it has been sboWn that the first member of this series, monochlorodimethyl ether; react&via a unimolecular path (37).

slow CH H- -0-CC1 ail f Cl- 3 e 3

(39) B. Glaastone, K. 47% Laidler„ and H. Eyring, The Theory of Rate Processes McGrax-Hill Book Co., Inc., New York, IL Y. 0 1941,'p. 14. fast , CH3-0H--C e R 3 0-CH--0-R-

It might be expected that the dichloro and trichloro ethers would also react unimolecularly. However- since a:chlorine has been found to reduce S32 reactivity in those cases studied, the polychloro ethers should be even less Se reactive than the monochloroether. The possibility also exists that methyl dichloramethyl ether can react by an 'a-elimination mechanism with the solvent abstracting a proton from the ether in the rate determining step. That such is the case seems improbable in view of the fact that haloforms s which are known to undergo solvolysis rapidly in basic media via an a-elimination„ appear to be unreactive in neutral and acidic media where the only base present is the solvent. In any case, if the polychloro ethers react to same extent by-mechanisms other than SN18 it must be because th.e y reactivity has decreased even more than would be indicated simply by the change in the solvolysis rate constants. For methyl dichloromethyl ether the Sul mechanism would be represented by the following

/R CH -0-CHC12 ...411:V4 CH + Cl- 3 3 N Cl

cg -0-c+ 2 R-OR ---->fast CH -0- CH(OR)- 2 H+ Cl 3 \ el with the last stage actually being a series of rapid steps. The difference in reactivity of the monodhloro and the didhloro ether, kel/kC12

80, per chlorine, must be explicable on the ease , of ionization of the falter as compared to the latter. These compounds both have the methoxy group in common which can affect the rate of ionization several ways' (l) an electron withdrawing inductive effect which will tend to decrease reactivity, (2) resonance stabilization of the carbonium ion which will tend to increase the reactivity, (3) resonance stabilization of the reactant which will tend to decrease the rate, The reactivity of all of the chloramethyl ethers shows that the inductive effect is not predominate and that resonance stabilization of the .carbonium ion is important. But these two effects should be the same for all the ethers. Therefore!, resonance stabilization of the reactant may explain the observed decrease effected by a-,ehlorine. The ability of oxygen to increase its covalency suggests that the following resonance interaction would contribute (4o) to the stability of monochlorodimethyl ether.

$3-U-CHC1 H CH---CECf

In addition resonance interactions would account for stabilization of the transition state. This is probably best described in terms of the intermediate carbonium ion which is stabilized by the interaction of contribtting structures.

4-4 CH -0=CH 3 2

The oxonium (right hand) Structure would be expected, because of lack of charge separation and more favorable geometry, to contribute more to the total structure of the carbonium ion than to that of.the reactant. - The resultant greater stabilization of the carbonium ion explains the

(40) For a discussion of the relative importance of contributing structures, see 3. Hine, Mr_iEal Organic Chemistry, Chap. 1.

%1 activating influence of the a-methoxy substituent. When these same effects are operative in dichloramethyl methyl ether, the following structures would contribute to the total structure of the reactant. a of of / + CH -75-C-H ‹.-- CH -0=C-H < CH -15-c-n 3 Cl el a* 2. equivalent 2 equivalent forms forms

The carbonitmi ion would be stabilized by structures such as

H H H CH -ZI-C + <74 CH -00e_ 4-4 CH,Z-c 3 a — \\ Cl Cl C14-

The Sill-activating influence of a-chlorine observed in the benzyl halide series shows that the added contributions of the chloronium ion type (right hand) structure to the total structure of the transition state leading to the carbonium ion more than offsets the added stabilization of the reactant by chloronium ion structures and the additional structures of the type

+ CX

made possible by the presence of the added a-chloro SUbStituent. Since the methoxy group is a much more powerful resonance eiedtron donor than is phenyl, the analogous, but greater, stabilization of polychloromethyl ethers due to added oxonium structures (shown in the center: above) would

cause added a-chloro substituents to stabilize the carbonitm ion relative

to the reactant less in the case of the chloroethers than in the case

of the benzyl halides. While it thus would have been predicted that achlorine would increase Sill reactivity less in the case of chloramethyl methyl ether than in the case of benzyl halides, it perhaps could not have predicted that this effect would, be so large as to produce an actual decrease in reactivity, as observed. The suggestion that resonance interactions such as those described for dichloromethyl methyl ether are operative is not in itself unique. Brockway (41) has explained the observed shortening of carbon halogen . bond lengths, as compared to those expected-from covalent bond radii,

in some polyhalomethanes by of a similar argument. For example, the

shortening of the C-F bond in compounds containing two or more fluorine atoms was attributed to contribution of the following resonance struc tures to the normal state of the molecule. F- fl H -C=F+ (---) H-C F _ H

Same shortening of the C-Cl bond was observed in CHF2C1 and CHFC12 but only in the former was there any shortening of the C-F bond. While the C-Cl shortening in these compounds is very slight, possibly within the experimental error, no decrease is reported at all for chloroform. The absence of any change in the C-F bond length in the monofluorochloro- methanes, and presumably the lack of change in the C-Cl distance on going from methyl chloride to chloroform, was attributed to the fact

(41) L. 0. Brockway, Journal of Physical Chemistry, Il e 187 (1937) 1

that "chlorine does not have the great electronegativity which would allow the assumption of an extreme ionic form in these compounds" (41). The extension of Brockway's explanation for the shortened bond lengths and decreased reactivity in the halomethanes to account for the 'decreased reactivity of dichloramethyl methyl ether can be rationalized in the following way. Forlthe interaction of two or more groups, or atoms„ attached to the same atom to result in the contribution of ionic structure to the normal state of a molecule, it is necessary that at least one of the participants has a minimal capacity to donate electrons

by a mesomeric effect (42) and at least one a capacity to accept electrons by virtue of its electronegativity. Fluorine, which can donate electrons

by a mesomeric effect ( 13) and is said to have the greatest electronegativity (14)9 should be singularly adept in these interactions while chlorine in which both effects are diminished would interact to a lesser

extent. . However, various combinations of the halogens or halogens and other groups or atoms which can exhibit, these same electrical effects may produce interactions to change the reactivity Of a molecule from that "normally" expected. Therefore, it might be expected that when a methoxyl group, which is known to possess a strong mesomeric effect, and two chlorine atoms are attached to the same carbon tato% as in CH 3OCHC12„ ionic structures would contribute to the normal state of the molecule.

(42) For significance of the term "mesameric effect," see C. K. Ingold, Structure and Mselyasis in Organic Chemistry, Cornell University Press, Ithaca, N. Y., 1953, p. 63. (43) R. H. Jaff4„ Chemical Reviews, a 191 (1953). (44) L. Mauling, The Nature of the Chemical 1,24 Cornell Univer- sity Press, Ithaca, N. Y., 19 5s ed., p17.77:75. 55

In addition, Hine and coworkers have found difluoromethyl methyl ether (45) and difluoromethyl isopropyl ether (46) to be particularly stable to solvolysis in methanol and isopropyl alcohol, respectively, while

Thomas 07) has found evidence that methyl fluoromethyl ether solvolyzes at a noticeable rate in aqueous methandl at room temperature. In view of the marked decrease in the rate of ionization brought about by the substitution ofq-chlorine for Cr-hydrogen in the methyl chloromethyl ethers, which was attributed to an oxygen-chlorine resonance interaction, it was thought that the contributions of structures such as .

Cl - 91 CH C - CH -. 0=- -H 3 •61 3 Cl - might analogously influence the reactivity of 2.-methoxybenzal chloride relative to that of 2,-methOxybenzyl chloride. However, because of the greater separatidnOf'unlike charges (48),'It would be expected that — their contributions to the normal state of 2,-methoxybenzal chloride would be somewhat less than the contribution of similar structures to the normal 'State of methyl dichloroMethyl ether. In addition, the effec7 tiveness of the methoxy group in the benzene derivatives could also be diminished by contributions from the following structures.

(45) J. Hine and J. J. Porter, Journal of the American Chemical Society, 79, 5493 (1957). (46) J. Hine and K. Tanabe, Journal of the American Chemical Society, 79, 2654 (1957) (47) C. H. Thomas, loc. cit. (48) J. Hine, Rhysica1, 01/Eas Chemistry, p. 9. 56

CE3 - cc 12H (7--> CH3 0 CC1 2

In order to determine what effect, if any, the para methoxy group does exert on the rate of hydrolysis of p-methoxybenzal chloride, relative to p-methoxybenzyl chloride, these compounds were investigated under the same Conditions.

The first order hy&rolysis of p-methoxybenzyl chloride in 5:1 acetone-water at 30° has been found to have a rate constant of 4.96 x -1 10-5 sec. . This is in fairly good agreement with the value reported by Simonetta and Favini (49), 4.33 x 10 -5 sec. -I . The difference between these two values is probably due to the fact that the initial concentration of reactant was 0.1 M in the solutions investigated by these workers while in the present investigation the initial concentration was approximately 0.0093 M. While first order rate constants are independent of

initial reactant concentrations, they are susceptible to diminution by a• 1 sass-law" effect (50). In fact, Kohnstam, Queen ard Shillaker (49) in a preliminary report stated that 2 .-sethoxybenzyl chloride was analogous to diphenylmethyl chloride in its behavior in the presence of "common-ion" salts. The rate constants obtained in the present investigation did not progressively decrease with time. The chloride ion concentration, however, was never larger than 0.0046 M and since the 'ass-law" constant

is presumably 11.5 (51), not much variation would be expected. Further- more, it is possible that an ionic strength effect, which would facilitate

(49) M. Simonetta and G. Favini , Journal of The Chemical Society, 1840 (1954). (50) For the significance of this term, see Ingold, op. cit., p. 360. (51) G. Xbhnstam A. Queen and B. Shillaker, Proceedingsof The Chemical Society, 157 (1959). 57

the rate determining ionization , was compensating for diminution due to a small t mass-law* effect. No experimental data other than average rate constants were reported in the paper by Simonetta and Pavini (49) so the

initial chloride ion concentration due to hydrolysis of reactant prior

to the 'zero-time' measurement ie unknown. However, from data given for the hydrolysis of o-methoxybenzyl chloride, it can be estimated that the initial chloride ion concentration was at least 0.003 M in their study of

p-methoxybenzyl chloride. If the next to the last point in Table 19 is taken as *zero ° reaction, the chloride ion concentration is approximately

0.0037 NI; the rate constant calculated from the last point is 4.39 x 10 -5 -1 . This is in good agreement with the value reported by Simonetta sec.

and Favini(49).

In the solvent, 5:1 acetone-water, at 30° the rate constant for the hydrolysis of k-methoxybenzal chlOride has been estimated as 1.9 x 10 -3

-l . This shows that the introduction of O(-chlorine into EmethOxy- sec. benzyl.chloride brought about an approximately CO fold increase, per chlorine, in reactivity. The corresponding change in benzal chloride, give benzotrichloride, has been found to result in a 34 fold increase, per chlorine, in reactivity. Thus it appears that as expected the substitution of an O(-chlorine into a 2 .-kethoxybenzyl chloride increases the reactivity less than when the substitution is made in the un-methoxylated

compound. It is possible, though, that this is a solvent effect since

the methoxy compounds were studied in 83 1/3 per cent acetone and the

uhsubstituted compounds in 50 per cent acetone. Since the ratio of the

reactivity of benzOtrichloride to benzal chloride appears to change with the nature of the solvent in a somewhat unpredictable 58

manners; we cannot be sure what it would be in 83 1/3 per cent acetone.

It is believed that the substitution ofOechlorine for in p-methoxybenzyl chloride has brought about a smaller increase in the rate of heterolysis of a carbon chlorine bond than normally anticipated, by virtue of contribution of structures such as the following to the

2 equivalent fOrms normal state of 2.-methoxybenzal chloride. As was the case with the chloromethyl ethers, structures involving a resonance interaction between oxygen and chlorine have no counterpart in the resultant carbonium ion. The results obtained on this investigation illustrate the compli- cations inherent in assigning increases in 8N1 reactivity, due to the substitution of 0(-chlorine for O(-hydrogen, to enhanced carbonium ion stabilization. For in addition to this effect the added chlorine may, by resonance interaction, also increase the stability of the reactant.

The net result would then be a smaller, increase in reactivity than would : be predicted from existing data, or an actual reversal of reactivity, as was observed for the chloromethyl ethers.

1FOr example, it was found (52) that, the ratio of the first order rate constants decreased by 1.3 fold on going from 50-per cent acetone , to the better ionizing medium, 50 per cent ethanol. In contrast to this they also found that the ratio increased by 1.3 fold when the solvent was changed from 80 per cent ethanol to the better ionizing medium, 50 per cent ethanol. It seems likely, however, that benzotrichloride would be more reactive than benzal chloride in any solvent in which the 3 N1 mechanism was operative, (52) IL Beasley and G. Kohnstam, Journal of The Chemical Society, 287 (1956). 59

Further., the effect of the substitution of chlorine for hydrogen on a carbon atom other than the one involved in the heterolysis may also result in larger changes in reactivity than would be expected from

inductive effects if resonance interactions can contribute more to the

stabilization of the reactant than to that of the resultant carbonium

ion. For example, Brown, Kharasch and Chao (53) have found t-butyl chloride to be four thousand times more reactive than isobutylene chloride in 4:1 ethanol-water, whereas Saloroa (54) has found methyl chloromethyl ether to be twenty thousand times more reactive, per chlorine, than bis-

(chloromethyi) ether in ethanol. If an inductive effect alone were responsible for the decreased reactivity in bis-(chloromethyl) ether, as

it probably is for the decreased reactivity of isobutylene chloride, 1 it

would have been expected that methyl chloro-methyl ether would only have been 20 times as reactive as the dichloroether. This estimate was made in the following way. Branch and Calvin (56) have suggested that the inductive effect of a substituent group on the as, of an alkane carboxylic

acid is diminished by a factOr of 112.8 from its value when it is on the a-carbon for every carbon it is farther removed. Since the rate constant

..•■•10. 1 It should be noted that the intervention of -neighboring group participation could- in part account for this decrease being only 4,000 fold. However, it has been suggested (55) that chlorine is not an effective participant. (53)H. C. Brown, M. 8. Iharasch and T. H. Chao, Journal of the American Chemical Society, 62, 3435 (1940). (54)P. Salonoa„ Annales Universitatis Turkuensis A XIV, 1953. Data obtained from P. Ballinger, et, al., op. cit., p. 3645.

(55)S. Winstein and E. Grunwald, Journal of the American Chemical Society 70, 828 (1948).

(56)G. K. K. Branch and K. Calvin, The Theory of Organic Chemistry, Prentice-Hall, Inc., New York, N. Y., 1941, Chap. VI. for the first order solvolysis of t-butyl chloride was decreased 4,000

fold, then the effect of chlorine'could be expressed by the foll6wing,

16g kla = log (2.8)n log Woo,

where n is the number of additional atoms 1the chlorine has been removed from the 0(-carbon. The predicted ratio of the rate constants for methyl expression by setting n equal to one and solving for Icji46 1 . This result is then the change in rate expected per chlorine.

The faCt that a considerably greater decrease in reactivity was observed, 20,000 fold compared to 20 fold, can presumably be attributed

to resonance interactions between oxygen and, chlorine resulting in contribution of structures such as

eB Cl - - o =-C,s Ci- s 2 equivalent forms

to the normal state of the dichldroether.

It might be suggested that since the positive charge in the transition state of the solvolysis of the chloromethyl ethers is largely on oxygen and thus the charge on carbon is less than in the transition

states for the benzyl halides, that the 0(-chlorine atoms should supply

Due to the distribution, of the positive charge between oxygen and carbon in ario(-methoxy carbonium ion the substituent chlorine in the metho±Y cOtpoUnd may not be a full atom more distant from the center of positive charge than the one in isobutylene chloride but it should be somewhat more distant. 61

electrons by a resonance effect to a lesser_extent in the case of the chloronethyl ethers. Thus 0(-chlOrine should have less tendency to stabilize the carbonium ion-like transition state by resonance. Of course, in such a case the change in the character of the positive charge should also result in less destabilization of the transition state by the inductive effect of the chlorine. Nevertheless if the decrease in

inductive effect, C(-chlorinecould have a net result of deactivation. The importance of resonance of the type suggested here for the nethoxy chlorides seems to be strongly demanded in certain cases by several types of evidence; for example, the shortened bonds in certain fluoromethanes and the fact that for the trifluoromethyl group o is greater than(Ti(57). In addition, such Stabilization appears to be the most plausible explanation for the greater nucleophilicity of the fluoride ion than of chloride ion toward difluoromethylene and of fluOride ion than of bromide ion (58) toward chiorodif luoroacetate anion and of the fact that the reactivity ratio kbE2yBrikdB4sr2 is so much larger for methoxide ion that for thiophenolate (59). For this reason alone we would prefer to attribute the present observation to this kind of resonance interaction, which appears to be particularly important with such compounds asq-haloethers. In addition, however, there appear to be several specific arguments that can be advanced against the alternative . explanation described. First, it does not seem to explain the fact that the

(57) J. D. Roberts, R. L. Webb and E. A. McElhill, Journal of the American Chemical Society, 72, 408 (1950). (58) J. Bine and D. C. Duffey, ibid., 81, 1131 (1959). (591 J. Hine, S. J. Ehrenson and W. H. Brader, Jr., ibid., 78, 2282 (1956) . 62

more distant chlorine in bis-(chloromothyl) ether causes more SN1 deactivation than the one in isobutylene chloride. Second, it does not explain the well known striking inertness of organic polyfluorides.1

The data on the rates of basic hydrolysis of haloforms shows that when the electron demand is large enough the resonance effect of fluorine can outweigh its inductive effect. And yet, unlike the case of the methoxy chlorides, with the simple payfluorides there is noC(-substituent to take the positive charge largely onto itself.

The C(-Elimination Reaction of Methyl Dichloromethyl Ether. --One argument that the second order reaction of methyl dichloromethyl ether proceeds via an 0(-elimination rather than an SR2 mechanism can be based on the magnitudes of the rate constants found. The second order reaction of methyl dichloromethyl ether with potassium isopropoxide in isopropyl alcohol at -11+ 0.5° has been found to have a rate constant of 0.88+ 0.034 1. mole-1 min. -1 . The corresponding value for the reaction of methyl chloromethyl ether with sodium ethoxide at 0 ° in 1:1 diethyl ether-ethanol was found by Ballinger (60), et. al., to be 6.6 1. mole-1 -1 . min. Before comparing the second order reactivity of methyl dichloromethyl ether and methyl chloromethyl ether, several comments are necessary because of the difference in the conditions under which these 0 observations were made. Bimolecular reactions between a charged species

1While there is not a great deal of quantitative data on this point, among the many qtalitative examples is the fact that benzotri- fluoride is a quite stable compound, while benzyl fluoride is rather unstable, tending to undergo what appears to be a Friedel-Crafts polymerization on standing. (60) P. Ballinger, et. al., loc. cit. 63 and a neutral species are susceptible to changes in the polarity of the

solvent. Therefore , the change from isopropyl alcohol to 1:1 diethyl ether-ethanol is of importance. The effect of this change can be rationalized from available experimental data. For example, in this investigation it has been found that the first order rate constant for the solvolysis of methyl dichloronethyl ether in 1:1 diethyl ether-ethanol at-0° was 3 x 1075 sec. -1 . The rate constant in isopropyl alcohol at -11 ° -5 sec. -1 from which a value of from 3 to 5 x 10 75 at 0° was 1.69 x 10 would be estimated (61). Since this first order solvolysis proceeds with a rate determining ionization, it should exhibit large variations in rate with small changes in solvent polarity. Thus the relatively small change found in the 3N1 reactivity of methyl dichloronethyl ether would indicate that isopropyl alcohol is little, if any better a solvating medium than is 1:1 diethyl ether-ethanol. Bimolecular reactions are usually consier- ably less susceptible to change in solvent polarity than are unimolecular reactions. For example, the second order rate constant for the reaction of isopropyl bromide with hydroxide is. changed from 4.9 x 1075 1. mole -1

-1 to 3.0 x 10-5 1. mole -1 sec. -1 while the first order rate constant min for the solvolysis of ,t-butyl chloride is changed from 9.14 x 10 -6 sec. -1 when the solvent was changed from 80 per cent aqueous ethanol to 60 per cent aqueous ethanol (62). It therefore appears that the difference in plarity between 1:1 diethyl ether-ethanol and isopropyl alcohol would have only a slight effect on the bimolebular rate.

(6a) D. C. Duffey, • ---cit. °9, 0 100,100 has, reported a value of 5. 4 x 10- sec. -1 in isopropyl alcohol at 0d71- . (62) Data cited in Ingold, op. cit., p. 349. In addition to the solvent change, the change in nucleophilic

reagent also requires comment. This is necessary since the reactivity of isopropoxide toward methyl chloronethyl ether is not known. In com-

paring the relative nucleophilicity to two reagents whose nucleophilic atom is the same, the most important factors to consider seem to be basicity and steric hindrance. The greater basicity of isopropoxide (63) should tend to make it more nucleophilic than ethoxide but its greater size

should tend to make it less nucleophilic. There does not appear to be any data in the literature that show which of these two opposed effects is the larger. However, data does exist for some alkoxides from which an estimate can be made. For example, Shiner (64) found, that the second order reaction of ethoxide with isopropyl brordde, in ethanol at 25° , proceeded with a rate coefficient of 2.82 x 10 -6 1. mole -1 sec. -1 . This rate coefficient refers to both the elimination reaction, to give propylene, and the substitution reaction, resulting in ethyl isopropyl ether; but since the olefin fraction was found to be 0.625 the rate coefficient for the displacement reaction

is 1.06 x 10 -6 1. mole -1 sec. -1 . With the more eterically hindered base, t-butoxide Brown, Floritani and Okamoto (65) found that its second order reaction with isopropyl bromide, in t-butyl alcohol at 25° , proceeded with a total rate coefficient of 2,35 x 10 -6 1. mole -lsec. -1 . Since the olefin yield was at least 90 per cent, the rate coefficient for the displacement reaction is less than 0.235 x 10 -6 1. mole - 1 sec. -1 . The )4.5 fold decrease in the nucleophilicity of t-butoxide, compared to ethoxide, toward isopropyl bromide must be due to steric hindrance since t-butoxide is much

(63) J. Hine and M. Hine, Journal of the American Chemical Society, 74, 5266 (1952).

(64) V. J. Shiner, Jr., ibid., 74, 5285 (1952).. (65) H. C. Brown, I. Moritani and Y. Okamoto, ibid., 78, 2193 (1956). 65

more basic than is ethoxide. In view of this result it seems likely that

isopropoxide would be less nucleophilic than etho2ide l being structurally

intermediate between ethoxide and t-butoxide, at least toward, a halide with steric requirement as great as those of isopropyl bromide. Since methyl dichloromethyl ether would be expected to be in this category, it might be expected to react with sodium ethoxide in 50 per cent ethanol- ether with a rate constant larger than 0.88 1. mole -1 min. -I at -11° and probably larger than 2 1. mole -1 at 0° . It therefore appears that the substitution of0(-chlorine for

a-hydrogen resulted in a decrease of less than 4 fold in second order reactivity. This decrease is considerably less than the decrease observed by other workers for a similar substitution in halides believed to react by the SN2 mechanism. Thus Fells and Nbelwyn-Hughes (66) found that methyl chloride is 150 times as reactive toward hydroxide ion as methylene chloride. The deactivating influence of 0(-chlorine (compared to O(-hydrogen) has been found to range from 40 to 3,000 in other cases (67).

The most reasonable explanation for the expected Sgt reactivity is that the compound is reacting by the a-elimination nechanism (68). Airther evidence comes from data on the reactivity of the dichloromethyl ether toward potassium t-amyloxide in t-amyloxide. This reaction was found to have a rate constant of around 40 1. mole -1 min. -I. In view of the low nucleophilicity of the t-butoxide ion already described, this reaction rate

(61 I. Fells and E. A. Moelwyn-Hughes, Journal of The Chemical Society, 3 8 (1959).

(67) J. Eine, C. H. Thomas and S. J. Ehrenson, Journal of the American Chemical Society, 77, 3886 (1955); J. Hine, S. J. Ehrenson and W. H. Erader, Jr., -ibid., 7±3, 2282 (1956) and references cited therein.

(68) cf. J. Hine, ibid., 72, 2438 (1950). 66

seems to be much too fast for an Se reaction. In the present investigation more convincing evidence. concerning the mechanism of the reaction of methyl dichloramethyi ether with base was obtained. The second order rate:constant for the reaction of partially detterated methyl dichloromethyl ether with potassium isopropoxide in isOpropyl alcohol at was found to be 0.37 ± 0.01 14 mole -1 min. -1 . Allowing for the hydrogen content of the a-carbon atom (0.16 (fl per mole

4(0.33) the value for the second order rate constant of the deuterium Compound is 0.194 ± 0.078 1. mOle°1min. °1 . Thus there is a distinct isotope effect (69) 0 ykni, 5.42 ± 2:2. An effect of such magnitude tor an Se reaction, the only plausible alternative to the taelimination mechanism. is` quite unlikely. 1 From this isotope effect it folloVs that a proton transfer is occurring in, the rate determining step.

i-PrO° Cl2CHOCH i-PrOH Cl° CH3OCC1

The reaction does not involve the rapid reversible formation of an intermediate carbanion since the methoxydichlorptarbanion formed could abstract a proton from the solvent and revert to reactant and no isotope effect would have been observed.

IOnly small effects have been observed in S112 reactions. A value of 0.96 was observed in the reaction of methyl 2-bromobenzensulfonate with methoxide (70), while a value of 1.0 was observed in the reaction of isopropyl bromide with ethoxide (71). (69) K. B. Wiberg, Chemical Reviews, 25, 713 (1955). (70) R. R. Johnson and E. S. Lewis, Proceedingt of The Chemical Society, 52 (1958). (71) V. J. Shriner, Jr. loc. cit. 67

A concerted a-dehydrohalogenation might have been anticipated from the correlation equation of Hine and Ehreneon (72). This equation correlated the effect of a-halogen on haloform hydrolysis with respect to the halogen's ability to stabilize the trihalocarbanion and the dihalathethylene. The correlation parameters obtained from ten haloforms r none of which contained more than one fluorine, indicated that fluorine destabilized the carbanion, with respect to the other halogens, but that it was best at stabilizing the dihalamethylene. This latter stabilization was attributed to the greater ability Of fluorine to supply electrons by a resonance effect to the electron-deficient divalent carbon atoms. In fact when two fluoritet are present in the haloform, the dihalaaethylene stabilization relative to the stabilization of the trihalocarbanion is so great that the latter has no real existence, but decomposes to difluoramethylent and halide ion as it 18 formed. Thus Hine and Iangford (73) have obserVed - that the hydroxide catalyzed hydrol- ysi8 of deuterobromodifluoromethane has a rate constant approximately one-half as large as the nondeuterated haloform. In addition, they found that the deuterium content of the haloform did not decrease during the course of the reaction, as would be empepted if the trihalocarbanion were formed reversibly but that it actually increased. This result is compatible with a rate controlling concerted dehydrohalogenation.

(72)•J. Hine and S. J. Ehrenson, Journal of the American Chemical Society: 80 824 (1958),

(73) J. Hine and P. B. Iangfordr Journal of the American Chemical Society, 221, 5497 (1957). 68

In the case of methyl dichloramethyl ether then the presence of oxygen which can donate electrons by a resonance effect considerably better than fluorine should result in enhanced methylene stability. Thus as the proton is being removed by base : the concurrent loss, of halide ion is inevitable. It is of interest to note with regard to the result of this investigation that Hine and Tanabe (74) have proposed that the attack of isopropoXide on difluoromethylene would proceed with the expulsion of fluoride ion resulting in isopropoxyfluoromethylene to explain the formation of triisopropylorthoformate despite the observed unreactivity of difluoromethyl isopropyl ether : which was also isolated under their reaction conditions. In conclusion it appears that in the base catalyzed solvolysis of haloforms both alkyl dihalomethyl ethers and alkoxyhalamethylenes can' be intermediates. The ethers maybe formed by the attack of solvent on the dihalamethylene with concurrent proton donation by solvent. x 6r+ 1 cr_ 2 ROH + CX2 R - 0---C----H -0-R --R-O-C-H+ OR I I I t H X __ H X X + I R 0 C - H +- OR ROCX0 + ROE I - H X

The partial positive charge developing on the oxygen atom in the transition state should decrease its ability to donate electrons by a resonance effect thereby decreasing the possibility of halide ion expulsion. The alkoxyhalomethylene could then result either from the attack of base on the dichloromethylene or by the concerted dehydrohalogenation of the dihelo ether : where possible.

(74) J. Hine and K. Tanabe, ibidop 80, 3002 (1958). 69

CHAPTER 7

ElpommEnpAT100

There is another series of compounds that are of interest with regard to the influence of(0(-chlorine on reactivity. Truce, Birum and McBee (75) have prepared, and characterized, methyl chloromethyl fide, methyl dichloromethyl sulfide and methyl trichioromethyl sulfide by the action of thionyl chloride or sulfuryl chloride on dimethyl sulfide. The rates of solvolysis of these compounds should be measured, in a suitable solvent, to determine whether or not*.chlorine has a deactivating influence on the heterolysis of the carbon chlorine linkage.

However, there is good reason to believe that sulfur is not as strong a resonance donor of electrons as is oxygen, mrs its from Hammett (76) studies indicated CE3S = -0.01 and CH34 = -0.27, and consequently, *chlorine may not be as deactivating as it was in the oxygen ethers. Nevertheless, it would be expected the effect of 0(-chlorine would be slightly deactivating or at least not as activating as is in phenylhalomethanes. In addition to the chloromethyl thioethers, Truce, Birum and McBee (75) also prepared methyl chlorodifluoromethyl ether and methyl trifluoromethyl thioether. The rates of solvolysis of these compounds

(75) W. E. Truce, G. H. Birum and E. T. McBee, ibid., 711., 3594 (1952).

(76) L. P. Hammett, Physical Organic Chemistrz, McGraw-Hill Book Co., Inc., New York, N. Y., 1940, p. 184. (77) F. G. Bordwell and P. J. Boutan, Journal of the American Chemical Society, 71, 854 (1956). 70

should be measured since this would provide information concerning the effect ofC4-fluorine on the ease of heterolysis of the carbon chlorine linkage and the carbon fluorine linkage. It is further recommended that the solvolysis of methyl dichioromethyl thioether in basic media be studied to determine whether the reaction proceeds by an Se path or by an O(-elimination mechanism. Should the reaction proceed by an O(-elimination mechanism, it is probable that the reaction would proceed with the intermediate formation of a carbanion whose rate of reversion to reactant by proton abstraction from the solvent could be considerably faster than the rate at which it loses chloride ion. This would be expected by a comparison with methyl dichloromethyl ether since sulfur shoul&be better at stabilizing the carbanion than is oxygen, since sulfur pen increase its covalency by use of d-orbitals, and poorer at stabilizing the thiomethozychloro- methylene due to its reduced ability to donate electrons, compared to oxygen, by a mesomeric interaction. With regard to the present investigation, it is recommended that an attempt be made to prepare methyl dichloromethyl ether that has a minimum deuterium content of 0.98 atoms per molecule, and its reactivity toward isopropoxide in isopropyl alcohol be examined. These results would establish the intimate details of the mechanism with respect to the intermediate formation of a dichloromethoxycarbanion. It is also thought that the solvolysis of methyl chlorodifluoromethyl ether would be worthy of study to determine the effect of 0(-fluorine on the heterolysis of the carbon chlorine linkage. A possible method of synthesis of this compound would be to allow the trichioroether to react with antimony 71

trifluoride under conditions similar to those used by Truce, Birum and

McBee (75) for the thioether.

There also exists another method of determining the importance of ionic contributions to the ground state of methoxy containing alkyl chlorides. Previous studies have indicated that pi.-methoryphanyl phenyl- chloromethane deviates considerably from the best straight line for the solvolysis of benzhydryl chlorides (78), when the sigma value used for the pera-methary group is the one found from substituted benzoic acids. (Ws would be expected, however, since the electron withdrawing ability of a carbonium ion should be considerably greater than that of the carbonyl function of benzoic acid.) If interactions between oxygen and chlorine such as have been proposed in-this work are operative, then in a Hammett study of substituted diphenyldichloromethames the Para-methoxy compound should show a less positive deviation from the best straight line than was found in the benzhydryl series. Another approach to the establishment of the oxygen chlorine interaction would be to carry out a Hammett study on substituted benzal chlorides. It is also proposed that the reaction of methyl dichloromethyL- sulfone with base be investigated to determine if this compound undergoes solvolysis by an O(-elimination mechanism. Preliminary results from this laboratory (79) indicate that phenyl difluoromethylsulfone reacts with base by this mechanism with the initial loss of a proton and benzenesul- finate ion. Since chloride is a better leaving group than fluoride, it is possible that methanesulfonylchloromethylene would be an intermediate if the solvolysis of the dichlorosulfone proceeds via this reaction path.

(78) J. Hine, Physical Organic Chemistry, P. 144.

(79) J. J. Porter, private communication. 72

Dr. Grovenstein of the reading committee has suggested that it would be of interest to determine the apparent first order coefficient

for the solvolysis of methyl dichloromethyl-d ether. Be proposed that it is possible that this ether undergoes solvolysis in neutral media by

an *elimination mechAnism and consequently a primary isotope effect would be expected.

Finally, it is recommended that the products from the reaction of

other alkoxy dichloromethyl ethers with alkoxides be investigated to determine if the same product distribution is observed as in the alkoxide catalyzed alcoholysis of chloroform. For example should ethyl, isopropyl and sec-butyl dichloromethyl ether, which have all been prepared by Laato (80), react with the corresponding alkoxides to give product distribu- tions identical, or nearly so, to those observed from chloroform, this would be of aid in elucidating the intimate details of product formation. Additional information can also be obtained by determining product distribution from the reaction of such alkyl dichloromethyl ethers with

alkoxides such as t-amyl and t-butYL.

(80) H. Laato, Suomen Kemistilehti, B 3.2., 67 (1959). APPEMIX A

Table 9. Solvolysis of CH36CH 2C1 in Ethanol Diethyl Ether at 00

Time Volume of Total Volume 104 ki (Sec .) EtONa (ml.) EtOlid (m1.) 0.0279 X 0.0279 Ni (Sec. -1)

0.45 40,95a 122 5.60 40.95a 11.22 303 12,50 40.95a 11.65 .480 17.26 40.95a 11.14

30.60 40.95 14.15 Ave. 12.04

°Assumed to be the same as sample 5.

Table 10. Solvolysis of CH 3OCH2C1 in . Ethanol-Diethyl Ether at 00

Time Volume of Total Volume (Sec.) EtONa (ml.) EtONa 611.1

0.0279 . 1 0.0279 M

0.51 46.60

122. 7.24 46.64 12.83

303 14.85 47.15 12.13

1195 20450 46.15 11.63 960 31.30 46.10 11.62 Ave. 12.10 4- 0.48

Table 11. Solvolysis of CH OCC1 in 3 3 Ethanol-Ether at 0 0

Tits Volume of Total Volume 107 ki (Sec.) Et011ala1.) EtONa 610 0.02052 M 0.02052 M -1 (Sec.)

a 5,820 1.3o 78.90

9,28o 1.42 74.15 7.8oc

50,400 3.10 85.70 4.38

139,540 6.05 79.275+ 3.975b 4.68 + 0.26 Ave. 4.61 + 0.21

aThis was used as zero reaction in calculating the rate constants. bTbe endpoint was exceeded in titrating this sample and the average of the titres of the first three samples was used to calculate the rate constant.

°This was not included in computing the average. 76

Table 12. Solvolysis of CH3OCHC12 in

Ethanol-Ether at 0 0

Time Volume of Total Volume 105 ki (Sec.) Et0Na (ml.) EtONa (ml.) 0.0279 M 0.0279 M (Sec. -1 )

0 .05 40.0o -- 1,900 2.50 43.32 3.13

5,945 6.60 41.40 2.92 10,900 11.33 42.00 2.89 16,490 16.72 43.05 2:.96 24,590 23.72 45.60 2.99 Ave, 2.98+0.06

Table 13. Solvolysis of eg OCHC1 in 3 2 Ethanol Diethyl Ether at 0 °

Time Volume of Total Volume 105 ki (Sec.) Et0Na Et0N4 (m1.) 0.0279 M 0.0279 M -1 (Sec .)

5,069 6.21 43.64 3.03 7,664 9.07 44.82 2.95 18,059 19.70 46.27 3.0.7

24,214 14.04a 27.04a 3.015 Ave, 3.01.61E 0.036

aThe base used in these titration VAS 0,0418 M.

Table 14. Solvolysis of p OCHC1 in 2 Isopropanol at -14-i 0.5 0

Time Volume of Total Volume 105 ki (Sec.) i-PrOK (ml.) i-PrOK (m1.) 0.0381 M 0.0381 X (Sec.-1 )

0 0.65 41.65 -- 2,040 1.30 35.00 1.07 9,105' 4.45 35.15 1.30 17,255 7.80 35.40 1.36

21,142 8.00 30.50 1.35 Ave. 1.27 -1- 0.1

Table 15. Solvolysis of CyCHC1 2 in Isopropanol at -12 ± 0.5 0

Time Vblume of Total Volume

(Sec 9 ) 4m10 i-PrOK (ml.) 0.0381 M 0.0381 M

120 0.85 31.45 a 1,800 1.64 30.10 1.72 6,120 3.60 28.90 1.75

18,825 7.75 28.00 1.59 Ave. 1.69± 0.06

.Considered to be zero reaction. 78

Table 16. Solvolysis of CH OCHC1 in 3 2 t-Amyl Alcohol at 0°

Time "Volume of Total Volume (Sec.) i4rOK (m1.) i-PrOK (m1.) 0.0381M 0.0381 M

0 0.30 44.00 10,320 3.35 44.15 7.00 13,860 4.50 45.15 7.10 17,160 4.60 37.52 7.22 21,060 5.67 40.00 6 Ave. 7.07 79

Table 17. Reaction cif cH30pc12 with Potassium

t7Am7loxide in t7AMyl Alcohol at 00

Initial. Ether 0.0142 M Initial t-AmylOK 0.0314 M

Tos0 0.02204 M Initial Volume 60 m1.

KLapsed Volume . Time (Mini TosaH (a.) — (1. mole-liMin s 1)

1.58 2.20 57.5 6.75 1.33 31.1 13.05 0.87 19.08 0.85 25.20 0.76 31.58 0.76 43.55 0.60 - _ 0.70 1440.00 - - Ave . 44.3 ± 13.2

These volumes were required to neutralize 5 mi. of the reaction mixture, 80

Table 18. Comparative Reactivity of CH OCDC1 and CH OCHC1 with ISopropoxide 3 2 3 2 Temperature -11.5+0.50 Initial i-PrOK 0.0129 1.4 Tos0H 0.0363 M Volume 25 ml.

Reaction Volume of Volume of Final Titre Final Titre 101 k2 Time TosCH i-PrOK CH3OCHC12 CH3OCDC12 tl. mole4 Gmin.) (m1.) (m1.) min.)

o 8.50 2.95 41.98 a o 8.50 2.95 -- 41.98 b 6 10.95 11.95 37.97 9.24 6 10.25 8.78 -- 36.18 4.79 12 8.00 5.80 37.98 -- 10.30 12 9.30 7.30 38.26 4.50 24 6.25 4.32 38.26 -- 10.80 24 7.80 5.20 -- 39.90 4.6o 50 4.00 3.10 40.70 ..... 10.90 50 5.90 3.75 -- 42.50 4.72

a bUsed in calculating fraction of impurities. 81

Table 19. Solvolyais of 2-030C6H4CE2C1 in 5:1 Acetone Water at 30°

Time Volume of Total volume 105 Ic1 (Sec.) EtONa (m1.) EtONa (0..) 0 . 0384 M 0.0384 M (Sec. -1 )

600 0.70 13.45 a

4 ,590 '3.03 13.56 5.01 9,480 5.00 13.55 4.59b 3,210 1.15 7.85 4.94

62 480 2.20 8.17 4,84 9,930 3.25 8.13 5.14 13,890 4.00 8.07 4.93. Ave. 4.96± 0.09

aThis sample was used as zero reaction in calculating rate constants for samples 2 and 3.

bealcUlated as noted in footnote above; not included in calculating the average rate constant. Table 20. Solvolysis of 2.-CH30C6H4CHC12 in

0 5:1 Acetone-Water at 30 0

Time Volume of Total Voltms 103 ki (Sec.) NaOH 61.) NaOH (m1.) 0.03M 0.03 M (Sec . -1)

0a 1.0 32.4 540 19.8 32.6 1.67 865 24.6 32.8 1.57 1,255 32.2 36.0 1 .77 1,645 29.5 32.8 1.38

aConsidered to be zero reaction. 83

Table 21. Solvolysis of 11.-CH30C6H4CRC12 in

5:1 Acetone-Water at 30° Initial Reactant Concentration 0.013 M

Time Volume of Total Volume 103 Ili (Sec.) EtONa .(M1.) EtONa (a.) 0.0384 M 0.0384 M (Sec.. 7-)

0.80 24.2 a b 485 00, 01,

735 14.00 24.2 1.13 1,530 18.90 24.4 0.952 2,025 20.55 24.4 0.896

aConsidered to be zero "reaction. bUceeded the endpoint. 84

Table 22. Solvolysis of 2:-.CH30C6R4CHC1 in

5:1 Acetone Water at 30°

Initial Reactant Concentration 0.013 M

Time Volume of Total Volume 103 kl (aec.) Et0Na (01.) Et0R4 (m.) 0.0384 M 0.0384 14 (Sec. -1)

3 0 3.95 24.45

150 7.05 24.30 1.39 360 10.80 24.45 1.23 68o 14.00 24.47 1.04

1,095 17.10 24.40 0.968

1,635 19.65 24.45 0.905

aConsidered to be zero reaction. 85

Table 23. Solvolysis of 20:-CH30C6114CHC12 In

5:1 Acetone Water at 300

Initial Reactant Concentration 0.006 M

Time lirolnme of Total Volume 103 kJ. (Sec'.) 1.*'1.01C (m1.) i-PrOK (ml.) 0.0324 m 0.0324M (Sec . '1)

a 120 4.10 13.37

240 5.45 13.25 1.37

240 5.50 13.25 1.42 420 7.10 13 .4 0 1.30

417 7.05 13.35 1.29

633 8.60 13.45 1.28

865 9.55 13.58 1.14

1; 084 10.75 13.35 1.32

1 ; 419 11.35 13.32 1.19

aConsidered to be zero reaction. 86

Table 24. Solvolysis of 2-CH30C6H4CHC12 in 5:1

Acetone-Water in the Presence of Added Chloride Temperature 300 Initial Sodium Chloride 0.0272 M

Initial Reactant Concentration 0.006 M

Time Volume of Total Volume 104 kl (Sec.) i-PrOK (m1.) i-PrOK (m1.) 0.0324 M 0.0324 M (pec. -1)

60 2.20 13.43 --a 60 2.20 13.41 -- 120 2.67 13.45 7.07 180 3.00 13.39 6.26 240 3.30 13.52 5.61

539 5.00 13.49 5.93 685 5.60 13.53 5.69 821 6.25 13.55 5.85 999 7.06 13.55 5.94 1,245 7.93 13.47 5.99 1,465 8.40 13.55 5.61

'Considered to be zero reaction. APPEIW]X B

I I • . M m I . lila. Mid "Mai 111111111M1111111111111111111 NNE ITV!! II t MIN kill1111, ,1111111111111111116111111111111111111111111111111111111111111111111111111ME111111111 MIN III I 111111111111119 11■■■■ 11111111111■■1111111■■■■■ 1111=11111111111 inieMENIMEN ii CAME IRE UM MININERMI A11NII 111111111111111111ENIMMEINEINE 111111111111111 ■1111=1111 11111 IFIIIIIMEIMEN111111111111111 111111111111111111111111111111111111111 REMENEEMEMENEIMMEMENOMME Milli UNIIIMEIMMIIMINEMIIIIMMEM11111111111111111111•111111111111111111111111111111111111111111•11MMIMEMIIIIIIMEIMMIN 111111111111011H i RllKOf011INI01•1■■101•IMINIIIMMI10IIIIIMINIMIIIIIIIIIIIIIMEM1111111111111•111IMUMINIIIIMINEA1EMEll MEM ■■■ ■ 111111111111111111111111111111111111111111 ■ ■ ■ ■ EMMENm MENNMEEMMEMr M■■ ■IEM•■■■OMME■ME OHME WM=M E 11111111111111110111111111#111.111111111111MI111EMIIIIMINEMM111111•11111MMUMMI11NIIIII111MINIIIMIIIIIIMINIMMIRINI NIP immanammartimmumammummemmommiwinummumlimermmummommot MMI. ikiiininnmuninsimmisimemommiumNiiiiin".1111111111111111111111111EIMENEEMITIE RnmEmmirgum munnemmurmmneuenomminivismmomminnwmusnommaraIr'1111M11111111111111613111111111 1

Figure 2. Infra-red Spectrum of Hofmann Reaction Product. (carbon disulfide solution, 3 mm. cell thickness)

'1101UMMEM=ME■M•111711111 ■11MO

' ' EN Moinni INT P HN MINIEIROM MINEMINIUMMINNUMEMENNIMI 111,111111111111111111111MOINOMMONONVEM MINN NEMININIMMI 1111111111111

Figure 3. Infra-red Spectrum of Hofmann Fraction 2. (carbon disulfide solution, 3 mm. cell thickness)

uwu n1 nnm iiiiiiiiiiiAiiiiiliiiiiiiiiinniin 1ii1n1iiilis ^'NN 1iNN"1°"'liii"i "i1 i t '' n' inliiiiA' ' 1 Nii'i ' sm' 1s1^ ^^ ' 1 n T i i 1M'° ii1 ^ ^jl ^ umn11' nmm uu ^ÎII^^ •.il = 1^l " IIIAIAIIillIIIIIIIIIIAllllliiRilliIIIIIûN11111`lil^'1iNINW{Il ` ÎNNiiI ^ î'il i911i ^ 111111 lilt ilil IÎWMUI^ itINNIII I®INi m{ Nlim ii11NIN ^msnÎ liiifIN1i' A"^1NI1111; 1 Illiill inm nnm i u i i i i i nm imi m "n u iu n m i i nm"nm ilin.nn"miii i clAlli: : i ,nI iniii 1 I1 N 1An "A 16 11 m 1Ain A n i II i n u ul an ::: JHw^ o^Pe^ 1ncA w^ iin ^ ' uH!Osgli l "'!io l " 16i II „ nnali i " iin , .... P iln..^®r „ 3 ÊLE^^^ ÊÊP^ 1'EEÎEEI ^ !a E,l^ E®E9' ' E.E' Ev'E°1hi 11'1 ^1 q1 i1 11î E 1.Y 'i11.1 S 1, ^Gtl 1•, •1^.1,ulîliEEEfEi lai 'dI: IC .f l'^1si::.....1. 1 C1 111L: S®IÔiI^LPJi 07 111 .9111111 wPiC L'1:L.1NLI I:S 171Y+ .1i7@ilP.°Rl W I:i°"" 1•• «er i N 9 î!nili I::::: : :::I . ^^: 1 :::I € » 'I^ fir:' s wlsg ::::sSs..s ..r.:. i = I :::ww'î I ss s :Ta ^g c : I°1 al : : : s : °;• ii,...r.:,., _.::^'r'Eli^^;y i7 au^—..tx:••?3? s 11E::"^sî?sei1^Ê^g.i:°^l^F.:e•ys^^^li? ^ll ' ^ sideEI^c^^^_.°iGiE:^Q1lny^s I" ,..;.,..._.. - II':I:;EIEEEii IiI ÊE3 E`•.I^g ^€EEE EEgI^ 1liEs°ÊÊ S î', 1 î 1 1r' .'E `r1EEg -s €'=.

S LAIN"> "r F!t îll^ 3lblHlll^¢11^ Y Ind: rii9^9i+ÎiiIPtY: î 16i ' 11i"I€i!il^l °"'=° Tic"^ F.iii ^„ : - s sr asap pit :per p mss:s••^ ^• ^,^• _ îil9eîs•Era E21,"rsgeEF !.J_ a §ne.sr ls6c r— î!= 5 7s3ir + Êr.ÊijLe_ .: ^ ^g p^r, 'e^ :•aa •: 13?i??,z_'aE?e ?E ?E? ^ ! ^ ^-1:y t : -s,'9F' S*ayugN!--u a NEE='.°°•.4.i • i••• Q^ ^ EEC ^ i_ ^^ cl•-- ^ra _ ;4^ t î!^zyf:.w t^^ "+, ?'.^ 7tc êe• ^J ^^^- ^ ; ^ ^ '- t^ ....'- _i1.' !I 'FNig(.; s1 ir . ::; i- ir?4 C1RN+'fu:i yr4:: ^ e+. 1:Ltd:.. :',N a..-t ^ +y .R]1 .1';tq :a:'¶Ee,ipjsT7 ^^ II&i.Y )IY*: ...^ 9T ^P4:ntf^: :ST - `n- ®'OC i^.'lp,iiiotl I Ill III11 ,n1i. 1. ,. inl11^,1, I m1 ^C^ I ^ 1 +IrY - a N wn 1 1,1,1 1 •,__....g^ml Srr.:rÊEE^ _,..:".^, ,;.i3q ;ti ^?^m^^t x±Î^^r^^^:n_as.` - g t'=tt nniio,ifieinii3iililiiWiliN?',ii ^__ .-..,1n.1 „r

Figure 4. Infra-red Spectrum of Methyl Trichloromethyl Ether. (pure liquid, 0.1 mm. cell thickness)

Figure 5. Infra-red Spectrum of p-Methoxybenzal Chloride. (carbon disulfide solution, 0.025 mm. cell thickness) 90

■ 140 MO .00 11••••••Emmilimmi...... i.....E.E.E.B.•••0•••••••...... •••••••••••10... .E. mEspziour...... 9....r....frimommoirmem•iimmurEENLISLIMIIIIIIIIIIIIIIIIEIIBIIIILEELII iminummummumma NI.. 'NEM ■•11•11111111111•1111=11111MIE IIIINV 11111111EIMMINNEEMINIMIIIMEN111111111/M11111• 11•11111111111111111111111111111 P 111111-110•1111111A1111•111111•111m1111M11111111m1111111111 11111111111111111r s iremonirimmouralsomme I riIII II 1.111111111111111111 11111111111111111111 ME II 1111111110111.11111111 M11111111111•1111/11111111111111111111111111 q: . rprif, 2--,,,w.-- mmotimismniummmounummosmonimus smoimmosommom ■umenrill EMPENIPOrd IIIEIIIIIIIIIIIIMIIIIIIIIIIIIIIIIIIENIVAIIIOIM EPERIPIMPIVI E =rill 11 ENE 1111111111111•111111111111111IN NMI= EN MIK MP 'A HIBPIE11111111111 EN 111111111.1111111111111011 MUM MEP .7111NPANIESIMMIBIBLICIPPA' 1 EMI IR ERINIP ■ .111121 := mida k.IiitIAligglianligN[1-1P-AliEson.9 mon EimmEEN 1111 Milif? 1111: as p PREI ! 1 NI HER OR IQ , . 'W.gniiitailifilI MiLlimmimilliMPLL.11111N . Ilibionfflumhilh in igiblIP aii MN ii m 3 4

Figure 6. Infra-red Spectrum of p-Methoxybenzaldehyde. (carbon disulfide solution, 0.025 mm. cell thickness)

4., , 1 jaimillii [ 4 l'1 Ir1 li' Li1 1 71l' , , , 1 , -I-' -I [i. 1., 1 1 , ' .. " _ 1 ;[LI ^= ^

r - 9 , . [ =^

11 111 1 It '4 I+111 H- '

111111111 1.111.1111111111IN 1,11 .-WHIMINIMIENNIEMINNEMIMINNIIIMIIIIIIIIP=111111=11=1•1111•1111MINNIMEINE9111111MIERMINWAMILIMILjILi iLMIIM111111111 111111111111 11111111111■11111M11111111111111=11111 ■ IN11111111NEENII IIIIIIIMM11111111•1111111111•11111111111111 OMN■= MEI 111

Figure 7. Infra-red Spectrum of p-Methoxybenzyl Chloride. (pure liquid, 0.1 mm. cell thickness)

MANNIBIENNIONIMMIIIIIMIUMMINIMENEMEMEMIBI11111111.1118ER ■017NOMMENNIIMOMMINEEMMEIMMEI monuninunnremollEMORMIIIMMIIIMMIONITSMIIRSI•RIMUMMUMEIMWINIED•UffInIRRIBUOIRIM EINERSINE -9llimithalais onnonEsammosiellie warnmilimionasiatrainamosP rt

Figure 8. Infra-red Spectrum of p-Methoxybenzyl Alcohol. (pure liquid, 0.1 mm. cell thickness)

I 11 111 ilk AI fi kirf 1111 1111 111111111111111111111111.11111M1 11111111111YZEll

Figure 9. Infra-red Spectrum of Partially Deuterated Methyl Formate. (gas at 745 mm., 10 cm. cell thickness) 92

1.1111111101111111111111111111111111111111111111161111111111 IIIIIMMIIIIIIIIMII1MM111111111. 1■1 11■■1■1 1111■■■ 21113MINEINIMEIMMEMIEEMININIMENERIMMINI► ,141 AMIE EMEMONNIMOREMNIIMOMOMMOMENONIMO MO ■■■■■■ Huo ' EZE MMUMMILMOMMEN BilliMENVEMENOMMEMEMMIN it = AMSEEMEMMOMMEMEMEME IMMIUMENNMEMEMMEMEMEMMEMEMEN 14114 ,44: ENEfflifilifiliMMEMEN IMMINEUREINEAMMINEWEINIMMENMIliiiii --EiV5=1.111911111111111111111111111111111=1"911111111111151 NEAMMAMMEMEMOOMEHMMOOMMMOMMMOOMOMMMENMEM .14QuIMMEMEMEMMEMMIVAMMEMMEMMMUMMEMMEMEMMEMEMMEMEMEMEN - 1111•WENEMBEREMIEBEINIF IMILWEIMENIENINENEENIEVIMMEIMMENIMMEMINIEMENIENI mcmmegmmommmmmummummismormilmmmmommmommummarnmmmmismwsomm reitlE rilibmmimummilapNimmimmmffindirnmummmimmwmovmmmmurmmilmmmmwmmmummilmammmulamiliggeRneilla MENNEWNWEIrlardalliardil

Figure 10. Infra-red Spectrum of Methyl Formate. (gas at 71.5 mm., 10 cm. cell thickness)

MINEINIE zun. phunFir m. l llllllllllllllllllllllll llllll 96omplaamaimmommommisimaimmumommintoioAmllllllllllllllllllllllll llllllllllllllll lllllllllllllll llllllll williMINMENIEMBEIMMEMMEENEEMBEIMIEREMENIMIrifil MMRINEMMOMMEMN MOO MMEMNOMEMEMEMEMEMMENNOMMO INERMINingeinellr iNEMERMINNEMMEMEMINNEMIMMINIEEMEEMEMEIT, JIUMMUM EM MINER: MUMMMMMMMMMMMMMMMMMMMENMENVOIMMEMMERE MMOMMEMMEMMEMEMEMM M MMMMMMMMMMM ■■■■■ MME jgRIMMOMMWROMMEMMEIWAMMMEIMMWMIEMMMMMENUMEMEIMMWMMEIMMMINTgLiuMMENIMME tIL4WEI CE VIUMOINM OMMOVI N■■■■OM■■ OMMOOMMEMOMOMINIMMANN ligEMMENAMMEMEMEMEMM MMOMMEMMEMEMEMMUMMEMMEMEMMEMOINCNN fflIMMMEMMEMMIMUMMIMMEAMMEMMEMMOMMUMMEMMMEMMEMEMWMUMMEMEM '91/111thimullmmmilmm ilAMMEMZOMMIWZMIMEMEMErnarl m"mMennEIMMETIWOM NimmmammmaimmimmaimirimmimmmsummimmommmwqmmilimmtrAmmumilmin5Winairiii6011110111011110111111.1111111NIMIMMENIONNONNE

Figure 11. Infra-red Spectrum of Partially Deuterated Methyl Dichloromethyl Ether. (pure liquid, 0.025 mm. cell thickness)

Figure 12. Infra-red Spectrum of Partially Deuterated Methyl Dichloromethyl Ether. (carbon disulfide solution, 0.025 mm. cell thickness)

II; . lllll lllll ifflogoil MI " M 111111111 um, 1 mmummon. l lamp 11111 hillgilliN1111111111 inGillinfill1111111111Plim ,...111111111 iii111111111111111111111111 IMMIA111111111111111 Ail r . I I : j MIIIMMIE111111N 011111111111" 111 MK 1 „ lll . . . . , 1 04114mMOINNIBMIIIIIIMINININIIIMININ121111110" VIO NY NEWAMOMOMMIONOMONE OMMO 3 MOMMEMM IL 4d16445gOINMEWNEMONFI EMMEN&. rOMEN=E NNEMONEm 1.1111111MENE 1 %M EM BM MEM dinffilliMMEIESEIRIENIMEMEMEMEEM 5"1! Ilionlit'ffliliOtadlilillhailiMINEIMEREN ISEPDERNIENNEMIERE4MENTIMI . HiiiiiiiffieSilltiNNOMMEIHRIMID INELINIENEMENEMIIIIMENSIMMIEMEN imiiiiirovioliii1411111117111111101111NO IMERTAISIEMIUMBIlieMENEIBILMIFEMMT91111EprE l „. l ' VJLI.111 1111651111MINEERM IMINtikEZIEMEMIREMELV l ,,illiMEMEMNIVIN MENWNEEMENEWOMEMEM2M MEMMt2EL_ NM "iiiixar§WIKTAWAINICUPPINBWViWEINIWAIRMITTAMMIESILWVANAMIVE NU jilt ■ L RV : 74. :rn 5.r...- . vamp rogamaraormemitavertm . .- .71W/10M 7q 01011/1,11PL.::, .., 0 T" T ' IV.' i e77-,...,- d pr: .1,11r:r.,72,rw.61;pfirrl fV,111:,c1IIIRIME.,rlf rssg Aairwmirimmoret, :---Ivay.ury If :.reemicmr 4torer.:1•1X4F :Ierrno !!,r, • ,-,.•!- • 114,77 mmoLgi... canon •, ak I IR M MEMPVIIMMEVERIBBISIMPIN -, *ENEM r A llllllllllllll ll ii lll iiiiiiiiiiiNiiiiiiiiiiiiiiii- pkimiraiii owitnananiminallamvioutimur4amiiimmmiammaiAzwiamitmw._ iii.,., - llll - .ii

Figure 13. Infra-red Spectrum of Methyl Dichloromethyl Ether. (pure liquid, 0.025 mm. cell thickness)

M llll MIMM llllll MMMW PM II s oT INE pllllll mmumlimymn I 11 111 1 II I% 11 i11111N11 NINI on e 01 1 lll ;"I 171EASIM111111111111111 01111111111311111111 1111 11111111111111. CROMIONIMINEMOI MVO Immommoisvi1111111111MAN 111 IMMOOMOOMOOMOO j samommoskommd H. NEENCEIMENEMINNIMMINI ■■m■I III mgmosommsmol MiliONMISIMEIMMENIMEMEN EMEM I INEMililffiNtilaEiL laiiiMEINIMINEIBMIII I HI EMI= 1 WI■■R■IE■■■■■Et OPPEDEMNIMMEMEIMILMINEMIMMINIE11911 1111111111117213111MBIMMREN REOMOMMEMOO OMNI= IIIMMEEMENNEEN: 'MEE INIEN1■■■■■111■111■■■■ll INN ■I ■■MEI IM■■■■■E■E■■■IE ■I ■BEEIN ffiNI■RI ■■MEI■ Ej 0111191P'11111111/1111111MMINIIIMIEIBVIIIMINIIIIMIENNIE11 IIIMMEMININEWEIMUMAIMER mmmissummammmuumgwommu, Nib NWOOMOVINGERS

Figure 14. Infra-red Spectrum ofMethyl Dichloromethyl Ether. (carbon disulfide solution, 0.025 mm. cell thickness)

A A millesimensonmothI 1111111k. immumms off _,AIREEKOMME OM MOMMOONNMEMOOOMMEMMOO Off lllll TMNEMEMBEEMEMAINEEMMUNOWNIMMI MMUMMME M E iiiiMEINERNIEMIENUMMEINIMUNIMMINIMPEW EliEWEINMENEBEIEMEMBIERVERIMMIIIIIMIRNMENINIMINIMMININIMMMINEMFM Off!. EIEDIECFNMEIMIERMINPRIMOINIEMNIMEITEMMEMEMIPEIVIIMINMMIUMMINUMMIWINEIR MIPMI mgaimmemm BIM IMINMMIIIMMOMMIIIIMMIMMIMMI .4 MAM■ MMEMEREMUMNPREM nil OMMIMMEMMERM . ,i5■0YiWIEIEMNINNINNERIEARVOCIIMINIMMIIIW: " ..-,kWv!ri?VnllalltreAMIPIAMIMEMMAMAMENWATMEMEMMEIMMNMWMAMUW" 74EMENOMMEMEMENENEVE MEN MOMENOMO HiNNIMMEHEMMINOMMillEIROMMIIII ■IMPIEffli iiiiiMil

Figure 15. Infra-red Spectrum of Trichlorobromomethane. (carbon disulfide solution, 3 mm. cell thickness) APPEMM C APPENDIX 'C

Evaluation of the First Order Rate Constants for the Hydrolris of NI-Methotzbenzal Chloride at Zero Ionic Strength.-For the hydrolysis of rmetboxybenza1 chloride the equation derived by Bateman, Church, Hughes, Ingeld and Taher (81) for the effect of ionic strength and a common-ion on the rate of appearance of product can be written as follows.

dx— Ir2s 611" dt t,40 (c 4 2x) antilogio(ia 2) antilogio (Bo-,4A,)

In this expression, kf is the rate constant at zero ionic strength for the ionization of kmethoxybenzal chloride, Ce (the rate constant for the reaction of the carbonium ion with chloride ion divided by its rate

constant for reaction with water) is A measure of the susceptibility of the reaction to a "mass-law" effect, the initial concentration of reactant is denoted by 2, the concentration of the product at time, t, is represented by x, the initial chloride ion concentration by c and the Chloride ion concentration at time t is represented by c +2x. The ionic Otrengthlp,, is then equal to c 12x. The constants A and B are functions of the dielectric constant and the absolute temperature and are equal to -1.815 x 106 (010 -312 and -0.912 x 1016 OW -2, respectively. The parameter 0— is a measure of the charge separation in the transition state; thus it is a measure of the effect of ionic strength on the rate determining ionization.

(81) L. C. Bateman, N. GO Church, E. D. Hughes, C. g. Ingold and No A. Taber, Journal of The Chemical Society, 979 (1940). Bensley and Kohnstmn (82) have written the integrated form of equation (4) as k 0e) (J/t) (5) where

iic I =. antiloft ez z . cc + 2x )dx = (jilt - 4 0606Bax )antilogl. 0B Cr (e ÷ 2a) (6)

?c and J 11172dx (7) a - /0 The right hand side of equation (6) repreSentsthe first two terms of a series expansion for the preceding integral. The integral I is evaluated by substitution of the appropriate qtantities into this relationship. The values of the integrated first order rate constants, were calculated from equation (.). The integral J is evaluated by graphical

integration. The quantities Ce and k ) were obtained by the method of least squares. In the present investigation the procedure of Beasley and. Eamstsm was utilized. The dielectric constant of 5;1 acetone-water at 303° K is 28.74 (83), therefore A= -2.2337 and II= -1.2026 x 10 8 . BO attempt was made to evaluate the parameter Cr by measurement using inert salts but wide variations in its value have only alight effects on 4. From the data in Table 23 the first order rate constant and the mass-law constant were evaluated ror several values aSsigned The values of 4/t and

(182) Bensley and G. Eanstam, ibid., 3408 (-955). (83) Estimate from data listed by G. Akerlof, journal of the American Chemical Society, 5.1, 4125 (032). TA from which the values Ce= 130 and 43. = 1,90 x 10-3 sec. -1 were

obtained are shown in Fig. 16.

Table 25. First Order Rate Constants at Zero

Ionic Strength and Mass -law Constantsa

x 4 x 103 sec. 1

0 120 1.88

1.1 13 0 1.90 2.0 129 1.88

tae values of Vt and Iit evaluated from the integrated rate constants 1.32 and 1.14 x 10 -3 , denoted by,,a and b respectively in Fig. 16, were not used in palculating these values'as their inclusion resulted in the improbable values of 75 for CX° and 1.64 x 10-i for q.

It can be seen from the values listed in Table 25 that the rate constant is relatively insensitive to large changeS inq50 -!

The data in Table 22 was also used to cognate values of O( ° and

k.1°1* These values are listed in Table 26.

Table 26. First Order Rate Constants at Zero Ionic Strength and Mass-law Constants

0e0 x 103 sec . 1

0 341 3.23

1.1 295 2.90

2.0 353 3.24 99

The value that has been assigned to le.f from the data obtained in this investigation is 1.90 x 10-3 sec. -1 and the value assigned toCk ° is 130. There are several reasons for choosing these values rather than those listed in Table 26. In the derivation of the rate expression Bateman, et. al., assumed that the ionic activity coefficients obey the limiting Debye-Huckel relationship (84). Thus in an experiment where the chloride ion concentration exceeds ca. 0.01 M early in the reaction, and ions other than those produced in the reaction are absent, the integrated rate constants will fall more rapidly than would be predicted by the limiting law. Consequently the calculated values of O( ° and isID. 25 were calculated from experi- will be too large. The values in Table mental data where the initial and final chloride ion concentration was 0.00369 M and 0.0102 II, while for the values in Table 26 they were 0.0042 M and 0.021 M. In addition, Cp1 -:]=0.0115 Mwhen the second point was taken in this last experiment. The values assigned toe and correspond to cr.: 1.1 x 10-8 . This value of cr., which is identical to its value for diphenyldichloro- methane (82), was chosen because Kohnstam, Queen and Shillaker (85) have indicated that benzhydryl chloride and 2-methoxytenzyl chloride have identical Cr.'s. In addition the value assigned to(r in the present study, 130, is more similar to the corresponding value for dichlorodi- Phenylmethane in 85 per cent acetone at 24.76°, 142.5, than is the value 295.

(84) F. Daniels and. R. A. Alberty, Physical Chemistry, John Wiley and Sons, Inc., New York, N. Y., 1955, p. 484. (85) G. Kohnstam, A Queen and B. Shillaker, Proceedings of The Chemical Society, 157 (1959). 100

For experiments at higher ionic strengths, the values assigned to

0 and k? would predict integrated rate constants by the relationship,

1.90 x 10-3 i0 (BTA) 130 antilogi 0 ;(2:14 ) antilog (8) while the relationship using the higher values is as follows.

2.90 x 10 -3 (9) euitilogi0 (.13(9..)f 1J 295 antilogi 0 (A AV2 )

At an ionic strength of 0.03 equation (8) gives an integrated first order coefficient of 7.6 x 10-3 sec. -1, and from equation (9) the value is 6.46 x 10-3 sec. -1 . From the values listed in Table 24, it can be seen that the rate constant predicted by equation (8) is too high. This is expected for the limiting law should underestimate ionic activity coefficients at high ionic strength. Thus the value antilog10 A/ 2 in equation (8) will be too small and the calculated rate coefficient will be too large.

Furthermore, if any impurities which are more reactive than

2-methazytenzal chloride, for example 11.-methoxybenzo3l chloride, are present, the integrated rate constants will fall for this reason as well as because of the "mass-law" effect. Thus at the initial reactant concentration of 0.013 11, the integrated rate constants would be affected more by a reactive impurity than they would when the initial concentration was 0.006 M since they would continue to react to a higher ionic strength in the former case. From the data listed in Table 22 an integrated first order rate constant of 2.57 x 10-3 sec. -1 can be calculated for the point considered to be zero reaction, whereas the corresponding 1CIL

value calculated from the data in Table 23 is only 1.32 x 10 -3 sec. -1 .

In Fig. 17 is a plot of Jit vs. IA calculated from the data at 0.013 144 The solid line is the best straight line calculated from the corrected values and corresponds toce:= 295, 14!== 2.90 x 10 3 sec. -1 and C7-=1.1 x 108 . The dotted line is the best straight line from Fig. 16 with C)(° = 3 sec. -1 and cr.= 1.1 x 108 . While the solid line 130, k = 1.90 x 10

fits the data accurately, the deviations of the plotted points from the dotted line are such thatit also appears to be a fair repreOentation

of the data.

Finally, the value chosen for 142;1.90 x 10-3 sec. -1 , closely approximates the value for the first order rate constant, 14.6'2c 10 -3 sec. -1 1 obtained by plotting the .integrated rate constants against the per cent reaction and extrapolating to zero reaction. It Should be pointed out, er, that the value assigned to kf could easily be in error by 10 per cent. However, the original Objective of this study was to ascertain 'whether or not the stbstitution of 0(- chlorine forC-hydrogen in 2-methoxyben*yl chloride brought about a decrease in SE1reactivity, as was observed in going from plycyl to cg3=42„81.nce kcitkI was found to be at least twenty, and it is unlikely that the value assigned to kci is too small by more than twenty fold, a study to determine more accurately the first order rate constant, ka, for the hydrolysis of.2-methoxYbenzal chloride was not undertaken. 3.02

2.3

2.2 -

2.1 -

2.0 -

0 1 2 3 it 5 x 106

Figure 16. Graphical Determinition af0e and Icf

103

3.5

3.2

2.9

2.6

2.3

x103 2.0

1.7

1.4

1 .1

0.8 . 0 1 2 3 4 5 7 8.

x 106

Figure 17. Graphical Determination ofJX° and el BIMIOGRAPHI

1. Hine, J., Physical Organic Chemistry, McGraw-Hill Book Co. Inc., New York, N. Y., 1956, Chap. 5. 2. Hughes, E. D., Transactions of the Faraday Society, 32, 603 (1941). 3. Hine, and D. E. Lee, Journal of the American Chemical Society, 73, 22 (1951). Betel:ego' B., and G. Kohnstam, Journal of The Chemical Society, 3408 (1955). 5. Itimmila, E., E. Paakala, 11,K. Virtanen, A. Erva, and S.; Varna,

Annales Academie Scientiarm6Fennicae, Series . A II, 1959, No. 91; Bensley, 13., and G. Kohnstam, Journal Of The Chemical Societ 4747 (1952); Bivort, P., and P. J. C. Fierens,ITIlletin, des Semi tio Ohtiiques, Beiges, .61, 975 (1956). Hite, J.., and D. E. Lee, ,Journal of the American Chemical Society, 71, 3182 C1952).

7. Hine, J. , and S. J. Ebrensan, ibid., 80 824 (1958). 8. Brockway, L. 0., Journal of Dysical -Chemistry, 114,, 187 (1937). 9. Ascher, K. R. S., Journal of The Chemical Society, 2209 (1951). 10. Organic Reactions, ed. in. chief R. Adams, John Wiley and Sons, In., New York, N. Y., 1946, p. 267. 11. Handerson, T., and A. K. Macbeth, Journal of The Chemical Society, 121, 892 (1922). 12. Stevens, C. L. T.. K. Mnkherjee and V. Tranelis, Journal of the American Chemical Society, 2264 (1956). 13. Mank, M. E., and C. L. Stevens, Abstracts of Duets Presented at American Chemical Society Meeting{, Atlantic City, Nev Jersey, Septether1.3-18„ 1959, p. 53P. 14. Rusted, D. R., and W.I.. Kbhlhase, Journal of the American Chemical Society, 765 5141 (L954). 15. Ballinger, P. P. B. D. de la Mare, G. Kohnstam, and B. M. Prestt, Journal of The Chemical Society, 3641 (1955). 105

16. Evans, L. R., and R. A. Gray, Journal of erratic Chemistry, gi 745 (1958). 17. Douglass, I. B., mill. H. Warner, Journal of the American Chemical Society, 38, 6070 (1956). 18. Lee, J., A Ziering, L. Burger, and S.D. Heineman, Jubilee Vol Emil Barell, 264 C1946); Chemical Abstr ctip 41, 6252b (194777 19. Schmidt, H. Berichte der deutschen chemischen Gesellschaft, 2331 (L908): 20. Simonetta, M., and G. Favini, Journal of The Chemical Society, 1840 (1954). 21. Hine, J., =ID. C. Duffey, Journal of the American Chemical Society, 82, 1131 (1959). 22. Hine, J., ibid., 72, 2438 (7950). 23. Hine, J., and S. J. EhremSon, ibid., 801., 824 (1958). 24. Hine, J., and P. B. Langford, ibid., 71 5497 (1.957). 25. Ibid.

26, Thomas, C, H., The Effect of Halogen Atoms upon the Eze Reactivity of Other Halogen Atoms Attached to the Same Carbon Atom, Part I, Thesis, Georgia Institute of Technology, Atlanta, Georgia, 1953. 27. Duffey( D. C., Some Reactions of Methyl, Dichloromethyl Ether, Part Four, Ph, D. Thesis, Georgia Institute of Techncilogy, Atlanta Georgia, 1959. 28. Shriner, R.L., and R. C. Fuson The Systematic Identification of Organic Co o s 2nd ed., John Wiley and Sons, Inc., New York, 1 0, pp. 112 ff. 29. Duffey, D. C., loc. cit.; cf,Fiacher, H., and G. Weeks:Dr:0 gopPe- sodloili,ZeitaChrift fur 1.610 ache Chemie, gg!„ 1 (4.9)+1); Chemical Abstracts, 31, nce 3 . 30. Briers, Ifte.14.;4., P. Bivort, and P. J. C. Fierens, Bulletin des socit6s chimiques Beiges, 25 501 (1956). 31. Smith, E. L., Journal of The Chemical Society, 1288 (1927). 32. Manske, R. H. Journal of the American Chemical Society, 0, 1106 (1931). 33. Conant, J. B., and W. R. Kilmer, ibid., 46, 245 (1924). 106

34. Douglass , I. B., and G. H. Warner, ibid., 78 6070 (1956).

35. Duffey, 22. cit., p. 101. 36. Bailey, C. R., and J. Hale, Philosophical Magazine and Journal of Science, ga, 98 (1938). 37. Ballinger, et. al., loc. cit.

38. Frost, A. A., and R. G. Pearson, Kinetics and Mechanism, John Wiley and Sons, Inc., New York, N. Y., 2nd ed., Chap. 5.

39. Glasstone, S., K. J. Laidler, and H. Eyring, The Theory of Rate Processes, McGraw-Hill Book Co., Inc., New York, N. Y., 1941, p. 14. 40. Hine, Physical Organic Chemistry, Chap. 1.

41. Brockway, loc. cit.

42. Ingold, C. K., Structures and Mechanism in Organic Chemistry, Cornell. University Presi, Ithaca,'11. Y., 1953; p. 63.

43. Jef4, H. H., Chemical Reviews, 23, 191 (1953). 44. Pauling, L., The Nature of the Chemical Bond, 2nd ed., Cornell

University Press, Ithaca, N. Y., 1945, pp. 58 - 75. 45. Hine, J., and J. J. Porter, Journal of the American Chemical Society, 72, 5493 (1957). 46. Hine, J., and K. Tanabe, ibid., 71 2654 (1957). 47. Thomas, loc. cit. 46. Hine, Physical Organic Chemistry, p. 9. 49. Simonetta and Favini, loc. cit. 50. Ingold, cit. cit., p. 360.

51. Kohnstam, G., A. Queen, and B. Shillaker, Proceedings of The Chemical Society, 157 (1959). 52. Bensley, B., and G. Kohnstam, Journal of The Chemical Society, 287 (1956). 53. Brown, H. C. 11. S. Iharasch„ and T. H. Chao Journal of the American Chemical Society, 62, 3435 (1940). 54. Saloroa, -P., Annales Universitatis TUrkuensis A XIV, 1953; data in Ballinger, et. al., 22. cit., p. 3645. 107

55. Winstein, S., and E. Grunwald, Journal of the American Chemical Society, Eli 828 (1948).

56. Brandh, G. E. K., and M. Calvin, The Theory of Organic Chemistry, Prentice-Hall, Inc., New York, N. Y., 1941, Chap. VI.

57. Roberts, J. D. 1 . R. L. Webb, and E. A. McElhill , Journal of the American Chemical Society, E, 408 (1950). 58. Hite, J., and D. C. Duffey, ibid., gl, 1131 (1959). 59. Hite, J., S. J. Ehrenson, and W. H. Brader, Jr., ibid., 78, 228 (1956). 60. Ballinger, et. al., loc. cit. 61. Duffey, a. cit., p. 100. 62. Ingold, 922. cit., p. 349. 63. Hine, J., and M. Hine, Journal of the American Chemical:SocietY, 5266 U3521• 64. Shiner, V. J., Jr., ibid., 4t, 5285 (1952).

65. Brown, H. Q., I. Meritati, and Y. Okamoto, ibid., 71 2193 (1956). 66. Fells, I. and E. A. Moe1wyn4Hughes, Journal of The Chemical Society, 398 (1959). 67. Hite, J., C. H. Thomas, and S.; J. Ebrenson, Journal of the American Chemical Society, 73, 3886 (1955); Hine i Ehrenson, and W. H. Brader, Jr., ibid., 71, 2282 (1956). 68.. Hine, J1, ibid., la; 2438 0.9501!

69. Wiberg , K. B., Chemical Reviews, M, 713 (1955)• 70. Johnson, R. 114'amd E. S. Lewis, Proceedings of The Chemical Societx, 52 (1958). 71. Shiner, loc. cit. 72. Hine, J., and S. J. Ehrenson, Journal of the American Chemical Society, 024 (1958). 73. Hine, J., and P. B,JJangforel, ibid., 72, .1497 (1957). 74. Hine J., and K. Tanabe, ibid., 82; 3002 (1958). 75. Truce, W. E., a. H. Birum, and E. T. McBee, ibid., /lb 3594 (1952). 108

76. Hammett, L. P. Physical Organic Chemistr , McGraw-Hill Book Co., Inc., New York, N. Y., 1940, p. 1 . 77. Bordvell, F. G., and P, J. Boutan, Journal of the American Chemical Society, 11, 854 (1956). 78. Hine, J., Physical Organic Chemistry, p. 144. 79. Porter, J, J., private ponummicatiqu. 80. Laato, H. .Suomen Kemistilehti, B 31, 67 4959). 81. Bateman, L. C., M. G. Church, B. D. Hughes, C. g. Ingold„ and N. A. Taher, Journal of The Chemical §991117.0 979 (1940).

82. Beasley, B., and G. Ebhmstam, ibid., 3408 (1955).

83. Akerlof, G., Journal of the American Chemical Society, 11.0 4125 (1930.

84. Daniels, F., and B. A. Alberty, ical Chemistry, John Wiley and . Oons, Inc., New York, N. T.! , 1955, P. 84.

85. rohnstam, G., A. Queen, and B. Shillaker, Proceedings of The Chemical Society, 157 (1959). 109

VITA

Robert John Rosscup was born August 22, 1929, in Rolla, North

Dakota, the son of Susan (ale Copeland) and Frederick W. Rosscup. Be received his primary and secondary education in the Rolla public schools and upon graduation in 1947, enlisted in the Army. After receiving his discharge from the Army, he enrolled at the University of North Dakota in Septetber, 1949. He received a degree of Bachelor of Science (with honor) in Chemistry in June, 1953 and a Master of Science degree in August, 1954. Following graduation, he accepted a position with the Standard Oil Company at their Research Laboratories in Whiting, Indiana and worked there from September, 1954 to September, 1956. At that time he took a leave of absence in order to enter the Graduate Division of the Georgia Institute of Technology. While at the Georgia Institute of Technology, he received grants to support his research from the

American Viscose Corporation, the Rayonier Corporation and the Alfred P. Sloan Foundation.

He married the former Phyllis Ann Lahren in June, 1950. This union has been blessed with three children, Shawn Elise, Nancy Ann and Jean Lahren.