[Reprinted from the .Journalof Organic Chemistry'.{0. i]r77 ( 1975).l Copvright 19?5by the American Chemical Societvand reprinted bv permissionof the copyright owner.

The Oxidation of Terminal Olefins to Methyl Ketones by Jones Reagent Is Catalyzed,by Mercury(Il)t

Harold R. Rogers, Joseph X. McDermott,2 and George M. Whitesides*

Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

ReceiuedJune 11. 1975

The oxidation of terminal olefins by Jones reagent in the presence of a catalytic quantity of mercury(II) affords good yields (>lO"t") of the corresponding methyl ketones. Similar oxidations of 1,2-disubstituted olefins gives fair (20-70%) yields; in the caseof unsymmetrically substituted olefins, mixtures of ketones are produced.

The Wacker process for oxidation of olefins to ketones products) different from those of the Wacker oxidation. has three mechanistically distinct parts:3 first, activation of Mercury(Il) is an obvious candidate for the catalyst for the olefinic double bond toward nucleophilic attack by new oxidation reactions: it resembles palladium(Il) in its coordination with Pd(II) and addition of a hydroxide moi- ability to activate olefins for nucleophilic attack,a but dif- ety to this electrophilic double bond; second, conversion of fers in that decomposition of the oxymercuration products the resulting 2-hydroxyethylpalladium(Il) compound to normally generates cations by loss of mercury(0) rather ketone and a (formally) Pd(O) atom by a series of palladi- than olefins by loss of mercury hydride.s Unfortunately, um(II) hydride addition-eliminations involving vinylic al- neither we nor others6 have been able to discover a satisfac- cohol intermediates; third, reoxidation of the palladium(0) tory solution to the principal problem in developing a mer- to palladium(Il) by copper(Il). Wacker oxidation is an ex- cury(II)-catalyzed analog of the Wacker oxidation: viz., an tremely useful and general reaction. It is, nonetheless, efficient regeneration of mercury(Il) from mercury(0). In worthwhile to try to develop procedures for oxidizing ole- the absence of a solution to this problem, there are, how- fins that use as catalysts metals less expensive than palla- ever, ways of involving mercury(II) in catalytic oxidation of dium, and which involve reactions (and possibly generate olefins other than in a direct analog of a Wacker oxidation. 3578 J. Org.Chem., Vol.40, No.24, 1975 Rogers,McDermott, and Whitesides

Table I Oxidation of Terminal Olefins by Jones Reagent Catalvzed bv Mercury(II)

Isolated

Registry no. Reqistry no. yield, oi

1-Octene 111-66-0 2-Octanone 111-13 - ? 82 Undecylenic acid 112-38-9 10-Oxoundecanoicacid 676-00-6 83 3, 3-Dimethyl- I -butene 558-3 ?- 2 3, 3-Dimethyl- 2-butanone 75-9?-B B6 o 2 - Allylcyc Iododec anone 32539-89- 2 t:-Oxo- 2-pr opylcyclododec anone 56666- 10- 5 70 Styrene 100-42-5 Acetophenoneb 98-86-2 26 1, 3-Hexadiene 592-48-3 Polymer d In addition. a ?7oyield of 2 -al lyl- 2 -hydroxycyclododecanone was obtained- b Benzoic acid (16%) was also isolated, together vtith pol,\/mer.

One, explored in this paper, utilizes mercury(II) in oxymer- too curation of an olefin, oxidizes the hydroxyl moiety of the resulting 2-hydroxyalkylmercury(II) compound to an acid- labile 2-ketoalkylmercury(Il) derivative, and regenerates 50 mercury(II) by proteolysis of the carbon-mercury bond of this substance (eq 1). Thus, the mercury(Il) performs the OH o RCH:cH, * fHgoacl**9 nJucu,H*oa. too

I Itot (i) t-- V I 950 oo o llH*ll ; RCCH, + [HgOAc]* ." RCCHTHgOAc essential function of olefin activation, but is regenerated o without leaving the mercury(II) oxidation level. This cycle roo is, in a sense,one in which mercury(Il) catalyzesthe hydra- tion of the double bond, and in which the reaction is driven in the direction of the thermodynamically less stable hy- 50 drated form by trapping this form by oxidation to ketone. Results and Discussion Jones reagent (CrO:rHzSO+-HzO) oxidizes alcohols to o ketones efficiently, and is relatively unreactive toward ole- o.o o.5 l.o fins.? When Jones reagent is added to an acetone solution Hq(tt/olefin. of an olefin at 20", a slow, nonselective oxidation takes place. Addition of mercuric acetate or mercuric propionate Figure l. The vield of'ketonedepends on the ratio of equiva- lentsof mercury(II)to olefinpresent at the startof the reaction:A, o/obased on olefin) to the solution results in a rapid ?ZOmol (EICO2)2Hg-catalyzedoxidation of 1-octeneto 2-octanone(25o, Terminal olefins are converted consumption of the oxidant. acetone,Jones reagent, t8 hr); B, (EICO2)2Hg-catalyzedoxidation (Table to methyl ketones in yields of 80-90o/o I); 1,2-disub- of 1-octene to 2-octanone (25', dioxane, NazCr:Oz'2HzO- stituted olefins react readily, but give low yields of ketones CF:rCOuH, 18hr); C, (FltCO2)2Hg-catalyzedoxidation of a mixture (25o, under these conditions. The yield of methyl ketones result- Jones oxidation of terminal olefins dioxane,NazCr.:O;.2H2O-CF3COzH, mum yield shownin A doesnot correspondto an optimizedyield is relatively insensitive to the amount of mercuric salt (seethe text for a discussionof this point). added (Figure 1). The catalyzed oxidation of terminal ole- fins by sodium dichromate-trifluoroacetic acid solution (Table II); showed similar insensitivity to the amount of mercuric salt be imprcatalysis by mercury(Il) and to provide a qualitative es- tosylate were used ions other than mercury(Il) were explored timation of the absolute activity of mercury(Il) as a cata- Several metal be unsatisfactory as catalysts. No re- lyst in reactions based on Jones reagent and dichromate briefly, and found to place treating 2-octene with sodium dichro- ion as oxidants. action took on acid solution in the presence of thalli- The vield of ketones from 1,2-disubstituted olefins can mate-trifluoroacetic Oxidation of Terminal Olefins to Methvl Ketones J. Org.Chem., Vol.40, No.24, 1975 3579

Table II Oxidation of 1,2-Disubstituted Olefins bv Sodium Dichromate-Trifluoroacetic Acid Solution Catalyzed by lllercury( II )

Isolated

Olefin Registry no, Product Registry no, yield, %

cis-2-Octene 7642-04-8 2-Octanone (64Vd+ 560 3-octanone (9670) 106- 68- 3 trans-2-Octene 13389-42-I 2-Octanone (Og7o)+ 540 3-octanone (37Va) Cyclohexene 110-83-8 Cyclohexanone 108-94- 1 4lb Cyclododecenec 1501 -82-2 Cyclododecanone 830-13-? 36D Norbornene 498-66-8 Norcamphord 497-38- 1 20" a2'3-Cholestene 15910-23- 3 No reaction d Oxidation was car ed out in the presenceof0.2 equiv of mercuricpropionate. b Oxidation was canied out in the presenceof 0.5 equiv of mercuricpropionate. ' A mixture ofcis and trans isomers.d The product was isolatedas the 2,4-dinitrophenylhydrazone. um(I), suggesting that olefin activation was slow. Similar cipitates, but no metallic mercury is formed and no organic treatment of 2-octene in the presenceof gold(III), palladi- solvolysis products can be detected. Thus, it appears that um(II), and rhodium(Ill) afforded mixtures of 2- and 3- the rapid disappearance of 1 when treated with Jones re- octanone in yields of 30, 10, and 296,respectively. Gold(0) agent may be an oxidative reaction (eq 2). Cyclohexanol and palladium(0) deposited on the walls of the reaction can be detected in ca. 5oloyield after 5 min reaction time. vessel in substantial amounts during the oxidation. Evidence for the mechanism of the mercury(Il)-cata- ( J ones ( FHgoAc- ;:-rhagenr FHgocro,H lyzed oxidations (eq 1) is inferential. Treatment of cyclo- \_J \_J hexene with aqueous mercury(Il) acetate gives trans-2- 3J hydroxycyclohexylmercuric acetatee (1), which is oxidized to cyclohexanone in yields very similar to those obtained [,2-f I from cyclohexene under the same reaction conditions. Sim- *l * Hgrrrr* crrrVr(2) ilarly, 2-(chloromercuri)cyclohexanone (2) is converted L\_/ rapidly and in high yield to cyclohexanone by aqueous sul- lir o furic or trifluoroacetic acid. Compound I is itself relatively +-

..lones (3%, ( Fu<- ( FOH Jones reagent' acebone) reagent -/-,-,//O \--l \--i (-'{'n + | T G)8/,, NarCrpr, aqumus V"n*oo. \-/ d.ioxane.H+) Conclusions 1 Jones reagent or trifluoroacetic acid-sodium dichromate fln solution oxidizes olefins to ketones in the presence of cata- aY" :tt1H,S(), or | | | | (100/,) lytic quantities of mercury(Il); of the various metals tried r'q (-f 'Co-H \--\, -HeCl \-/ as catalysts for the oxidation-thallium(I), gold(III), palla- dium(II), rhodium(III), and mercury(II)-mercury(II) 2 gives the best yields. Qualitative evidence described above suggests that these transformations occur by the sequence stable toward the acid conditions encountered in the oxida- of reactions outlined in eq 1. This oxidation provides a use- products tion. Oxidation arising directly from the ti-hy- ful alternative to several of the procedures presently used droxyalkylmercury(II) cation therefore seem unlikely. Mer- to convert olefins to ketones. It is less complex than Wack- cury(0) has been observed to form in small amounts only at er oxidation: the problems that arise in applying Wacker elevated temperatures; no Hg(0) was observed under the oxidation to high molecular weight, water-insoluble sub- conditions described. The B-hydroxymercury(Il) cations stances do not seem to be important, and it is unnecessary previously have been shown to be stable in acid solution.l0 to have present the large excess of copper salts normally yields The difference in the of ketones from 1- and 1,2-di- used to make the Wacker oxidation catalytic. It can be ap- (Jones substituted olefins under strongly acidic reagent) or plied to unprotected olefinic carboxylic acids, where dibo- (NazCrzO?-CF3CO2H) weakly acidic conditions is reason- rane-chromic acid results in destruction of the carboxylic ably attributed to differences in the oxidative and/or solvo- acid moiety. It is more direct and more economical than the primary lyticll stabilities of and secondary carbon-mercu- several procedures (oxymercuration-reduction, epoxida- ry bonds. We have tested these stabilities under the condi- tion-reduction) that generate an alcohol preliminary to a tions of these reactions by examining the behavior of cyclo- Jones oxidation. hexylmercuric acetate (3) and n-hexylmercuric acetate (4) toward Jones reagent in acetone. Compound 3 reacts rapid- Experimental Section ly; oxidation is complete in 30 min at 20o, yielding cyclo- General. Melting points,determined on a Thomas-Hoovercap- hexanone in 65% yield. Under the same conditions, the car- illary meltingpoint apparatus,are not corrected.GLC analysiswas bon-mercury bond of 4 does not react appreciably: n-hex- performedusing a 10 ft X 0.125in. 15%Carbowax 20M columnon ylmercuric sulfate can be recovered in good yield. Com- a F & M Model 810gas chromatograph equipped with a hydrogen pound I does not solvolyze appreciably when treated with flame detectorand a Hewlett-PackardModel 33738electronic in- tegrator.All solventswere reagentgrade and were used without the components of the Jones reagent without the chromi- furtherpurification. 1-Octene, 2-octene (a mixtureof cisand trans um trioxide (that is, acetone, water, and sulfuric acid): a isomers),cis-2-octene, and f rcns-2-

decene(a mixture of'cisand trans isomers).3,ll-dimethyl-1-butene, Ice was added as necessary to maintain the temperature at 25 + and undecylenicacid (Aldrich) were used as supplied. Mercuric ac- 5o. The dark greenish-brown solution was stirred for an additional etate and sodium dichromate dihydrate (Mallinckrodt), trifluo- 4 hr and then poured into water (200 ml) and extracted with di- roacetic acid (Matheson Coleman and llell), gold(III) chloride, pal- ethyl ether (il x 75 mll. The combined extracts were washed with ladium(II) chloride, and thallium(I) acetate (Fisher Scientific). water (3 X 50 ml). saturated sodium chloride solution (1 X 50 ml). and rhodium(III) chloride (Alfa) were used without further purifi- and water (l X 50 ml) and dried (MeSOq). cation. Elemental analyseswere performed by Robertson l,abora- References and Notes tory, Florham Park, N..J. Mercuric Propionate. Red mercuric oxide (108 g) was added in (1) Supported by the National Science Foundation, Grant MPS74-20946 l0-g portions to 100 ml of hot propionic acid. The oxide dissolved. (2) John M. Lyons Fellow, 1972-'1974. giving a slightly yellowish solution which was filtered and allowed (3) Studies of the mechanism and synthetic applications of the Wacker oxr- dation have been reviewed: E. W. Stern in "Transition Metals in Homo- to cool toxygen nucleophrle General Procedure for the Mercury(Il'l-Catalyzed Oxida- add to the olefin cis or trans to the palladium center? What are the de- tails of the sequence of palladium hydride elimination-addition reactions tion of Olefins. Method A. To a 500-ml erlenmeyer flask was that generates product? ls Pd(0) ever present in the catalytic Wacker (74 dihydrate, 50 ml of added 22.0 g mmol) of sodium dichromate oxidation, or is it palladium(ll)hydridethat is oxidized by Cu(ll)? water, and 300 ml of dioxane. With stirring, 6.8 g (20 mmol) of (4) Oxymercuration reviews: W. Kitching, Organomet. Chem. Fev., 3, 61 mercuric propionate and 35 ml of trifluoroacetic acid were added. (1968); W Kitching in "Organometallic Reactions", Vol. lll, E. l. Becker The dark orange-red solution was stirred until the salts had dis- and M. Tsutsui,Ed., Wiley-lnterscience,New York, N.Y., 1972, p 3'19 ff. L G. Makarova, ibid.,Vol. l, 1970, p 1.|9 ff; Vol. ll, 1971, p 335 fl solved (ca l0 min), and the flask was placed in a water bath. With (5) Stoichiometric oxidation by Hg(ll): H. Arzoumanian and J. Metzger, Syn- continued stirring, 100 mmol of olefin was added. The solution be- thesis,527 (1971). came dark and warm; ice was added as necessaryto maintain the (6) H. Arai, K Uehara. and T. Kunugi,J. Catal.,33,448 (197a); H. Arai, K temperature at25 * 5o. The solution was stirred for l8 hr, poured Uehara, S. l. Kinoshita, and T. Kunugi, lnd. Eng. Chem., Prod. Res. Dev., into water (300 ml), and extracted with hexane (3 X 75 ml). The 11, 308 (1972). (7) K. Bowden, l. M. Heilbron, E. B. H. Jones, and B. C. l. Weedon, J. Chem. combined extracts were washed with water (3 x 50 ml), saturated Soc.,39 (1946): A Bowers, T. G. Halsall,E. R. H. Jones, and A. J. (1 (1 X ml) sodium chloride solution X 50 ml), and water 50 and Lemin. ibid.. 2555 ( 1953) dried (MgSOa). (8) H. C. Brown and P. Geoghegan, Jr., J. Am. Chem. Soc.. 89. 1522 General Procedure for the Mercury(Ill-Catalyzed Oxida- ( 1967) tion of Olefins. Method B. To a 500-ml erlenmeyer flask was (9) A. N. Nesmeyanov and R. K. Freidlina,Zh Obshch. Khim., 7.43 ( 1937). (10) P. Abley, J. E. Byrd, and J. Halpern, J. Am. Chem. Soc., 95, 2591 200 ml of acetone,5 ml of water, and 6.8 g (20 mmtil) of added ( 1973). placed in water bath and, mercuric propionate. The flask was a (11) F. R. Jensen and R. J. Ouellette,J. Am. Chem. Soc., 85, 363 (1963); F. with stirring, 100 mmol of olefin was added to the bright vellow so- R. Jensen and B. Rickborn, "Electrophilic Substitution of Organomercu- .1968, lution. Jones reagentT(2M, i5 ml) was added dropwise during '1hr. rials", McGraw-Hill, New York, N.Y., Chapter 3.