Reactions of Organometallic Compounds with Alkyl Halides. I. The

Reactions of Organometallic Compounds with Alkyl Halides. I. The Action of. Sodium Ethyl on (-)2-Bromooctane. BY NORMAN G. BRINK,^ JOHN F. LANE AND ...
0 downloads 0 Views 776KB Size
ETHYLON ( - - ) ~ - B R O M O ~ C T A N E THEACTIONOF SODIUM

May, 1943

943

Anal. Calcd. for C4hH~03: C, 80.76; Ii, 12.06. C, 80.81; H , 11.91.

(2 cc.), a t room temperature. After standing a t room temperature for twenty hours, the reaction mixture was poured into ether and the ether extract was washed successively with dilute hydrochloric acid, 0.5 N aqueous potassium hydroxide and water. The ether was removed by distillation leaving a white solid containing free palmitic acid. Palr:.itic acid was removed by adsorbing the crude ester frotn petroleum ether (15 cc.) on a column of Doucils (9'' X 12 n.m.). 'The column was washed with petroleum ether (7B cc.) and filtrate and washings were evaporated leaving a residue (0.302 g . ) which was crystallized frorn ethyl alcohol (6 cc.) a t 5". The antioxidant palmitate formed lath-like, white crystals (0.18 g.), m. p. 40.5-41.B0, which were identified as a-tocopherol palmitate. -

A natural antioxidant has been isolated from Mangona shark liver oil and identified as natural a-tocopherol. Evidence was obtained indicating that a-tocopherol is the major antioxidant present in this fish liver oil. This finding is of interest since i t indicates that the tocopherols may act as natural antioxidants in fish liver oils as well as in vegetable oils.

(9) Doucil (American Doucil Company, 121 South 3d Street. Philadelphia, Pa.) is a sodium aluminum silicate.

ROCHESTER, NEW YORK

[CONTRIBUTION FROM

THE

Found:

summary

RECEIVED FEBRUARY 8, 1943

FRICK CHEMICAL LABORATORY, PRINCBTQN UNIVERSITY]

Reactions of Organometallic Compounds with Alkyl Halides. I. The Action of Sodium Ethyl on (-)2-Bromooctane BY NORMAN G. BRINK,^

JOHN

F. LANEAND EVERETT S. WALLIS

Previous studies2 in this Laboratory have shown that the Wurtz reaction of an optically active alkyl bromide with sodium leads to a completely optically inactive product. Thus, when (+)2-bromobutane is treated with sodium a t room temperature, the resulting 3,4-dimethylhexane shows no detectable optical activity. This result would be consistent with a formulation3 of the Wurtz synthesis involving initial formation of free radicals Na

+ R X +R' + NaX

(1)

followed by their combination t o give the expected coupling product

+R-R

(2) or by their disproportionation to alkane and al2R'

kene by-products 2R'

----f

RH

+ olefin.

+

(Na+)R,-

(1) Sayre Fellow in Applied Chemistry, 1942-1943. (2) Wallis and Adams, Tms JOUKNAL, 66, 38.98 (1933). (3) Cf. Hbckel, Kraemer and Thiele, J . prakl. Chcm., 14% 207 (1935); Bachmann and Clark, TRIBJOURNAL,49, 2089 (1927); Richards, Trans. Faradcy Soc., 86,956 (1940). (4) Baughn. Evan. and Polanpi, Trans. Faraday Soc.. 81, 877

+

-

+ R-CHZ-CH~X RIH + R-CH=CHZ + (Na+)X-

(5)

In addition, when X is bromine or iodine, metalhalogen i n t e r ~ h a n g e ~ ~ NaRl

+ R z X --+

NaRz

+ RIX

(6)

can occur. It has been further suggestedsb that the initial stage of the Wurtz synthesis may involve direct formation of a sodium alkyl

(3)

Secondary alkyl radicals would be expected to show extreme optical instability, since they possess numerous planar resonance states comparable in energy to the normal state. In this they resemble the corresponding carbonium ions more closely than the corresponding carbanions, the planar resonance states of which are considerably higher in energy than the normal state.4

(1941).

More recently,6sodium alkyls have been shown to act on alkyl halides in a manner typical of the salts of very weak acids, either to effect substitution of the carbanion for halogen (Na+)RIR2X +(Na+)XR1- R1 (4) or by their strongly basic action to remove the elements of hydrogen halide with the formation of an olefin

2Na

+ R X -+

NaR

+ NaX

(7)

and that subsequent stages of the reaction involve the action of the sodium alkyl on additional alkyl halide according to equations (4) and (5). If this be granted, the observed products of the Wurtz synthesis may be accounted for without recourse to the concept of free radicals as critical reaction intermediates. In order to account for the complete optical inactivity observed in the formation of 3,4-dirnethylhexane from (+)%bromobutane and sodium, however, such an interpretation must be ampli(5) (a) Whitmore and Zook, Tmn JOW~NAL, 64, 1788 (1942): Morton, Davidion and Haksn, ibid., 64, 25142 (1942).

(b)

944

NORMAN G. BRINK,JOHN F. LANEAND EVERETT S. WALLIS

Vol. 65

stereochemical nature might be expected, were a previously prepared sodium alkyl allowed to act on an optically active alkyl halide. For this reason, and also in the hope of obtaining data which might serve to reduce the number of possible explanations of the detailed mechanism of the action of sodium on (+)2-bromobutanej we have investigated the action of sodium ethyl on a comparable halide, (-)2-bromooctane. Since the results of this study provide a basis for the immediate exclusion of the first two hypotheses enumerated in the preceding paragraph, and since they are of importance in estimating the suitaX-C~HB4- HC(CH:)(CzHs)(s-c4Ho) bility of the last three, we present them in detail a t C4Ko [C(CHd(CzHd(s-CJh) I- ( 8 ) (c) The sodium s-butyl initially formed is racemic, this point before proceeding with the discussion. Sodium ethyl was prepared from mercury diand the subsequent substitution (4) proceeds by ethyl and sodium according to Whitmore and way of a preliminary ionization (entailing raceZ00k.~" (-)2-Bromooctane, -30.7', was mization) of the optically active halide; i. e., the then added to the metal alkyl under pentane a t substitution is of the extreme SNltype6 -10'. After the initial reaction had subsided, RX +R + + X- (slow) (9) the mixture was stirred for five hours, during ( X - ) + H + + R-(Na+l+ R-R 4- (Na+ X-)(fast) which time i t was allowed to reaeh the tetnpera (10) (d) The sodiuiii s-butyl preserves the initial (d) ture of the room. Stirring was continued for configuration of the halide and reaction (4)pro- twelve hours, a t the end of which time 88..5% of ceeds by the alternative substitution inechanisni the alkyl bromide had reacted. The unreacted 2-bromooctane still present caused immediate 9N26 precipitation of silver bromide when test samples I I K - + C-X--tR & - ... ...&-X--+R-C+X(11) of pentane or ether solutions of the products were A A A treated with alcoholic silver nitrate. That i t was which involves complete Wdden inversion on the still optically active was evidenced by the high carbon atom a t which the substitution takes place negative rotations of solutions of the product a t and hence the exclusive formation of meso-3,4-cli- this stage of the operations. The mixture, havmethylhexane. (e) The substitution (4) pro- ing been freed from mercury diethyl by the action ceeds by a process of the type (11), involving of gaseous hydrogen chloride a t 0') was fractionWalden inversion, but the s-butyl carbanions ated through a seven-plate column until products arising from sodium s-butyl are very easily race- boiling below 150' had been removed. These mized. Further, i t is presumed that the d-form consisted mainly of octane and octylene in apof the carbanion reacts with d(+)-s-butyl bromide proximately equal amounts. The residue was much more rapidly than does the I form. Again digested with alcoholic silver nitrate, taken up in meso-3,4-dimethylhexane will be the product pentane and washed many times with cold concenformed trated sulfuric acid. Fractionation then gave [a]*O5g93 (-)3-methylnonane, [ Q ] ~ w -0.20'; d-S-CaHBI-s-CdHp(12) -0.25'; [ a ] z o-0.32'; ~ ~ [ 0 1 ] ? ~ -0.23'; ~ [iM]26D (Inversion) d-s-CJ3p- + d-s-CdHpBr meso-(s-C,H& -0.34' (homogeneous, 2-dm. tube); b. p. 166.8'(13) 167.1' (769 mm.),? and a small amount of optically (Inversion) inactive 7,8-dimethyltetradecane. The yield of l-s-CJ30- f d-s-C4HgRr ~ - ( I - C , H ~ )(11) ~ k l s > > kid pure (-)3-methylnonane from the reaction of There remains, of course, the possibility that in (-)2-bromooctane with sodium ethyl was 25% this instance, a t least, the Wurtz synthesis pro- (based on reacted alkyl bromide). The approxiceeds primarily by way of free radicals (equations mate molar ratio of the products octylene, octane, (7) Calhgaert and H. Sorow, THISJOURNAL, 68, 635 (1938). 1, 2, 3). In this event results of quite different

fied by one (or more) of the following hypotheses: (a) the sodium s-butyl initially formed is incapable of optical activity and the metal-halogen exchange reaction is reversible and sufficiently rapid to effect complete racemization of the halide before substitution by (4)has proceeded to an appreciable extent. (b) The hydrocarbon initially formed is optically active but is racemized by the metalating action of the sodium alkyl present, which attacks the tertiary hydrogens of the optically active centers to give an optically unstable carbanion

+

CI

____f

( 6 ) Hughes, Trans. Furaday SOL, 87, 603 (1941).

report 167.8' (760 mm.); Levene and Taylor, J. R i d . Chcm.. 64, 851 (19221, report 165-166.5° (751 mrn.).

May, 1943

THEACTIONOF SODIUM ETHYLON ( -)2-BKOMO&ThNE

3-methylnonane1and 7,8-dimethyltetradecane was 9 :12 :16 :1, respectively. I n order to establish whether a sodium alkyl could attack the tertiary hydrogen of optically active hydrocarbons such as (-)3methylnonane, a 10% solution of this hydrocarbon in pentane was stirred for eighteen hours a t room temperature with an excess of sodium ethyl. The hydrocarbon was recovered from this treatment with no appreciable loss of optical activity. Since the presence of the sodium amdgam (about 10% sodium by weight) arising from the preparation of sodium ethyl from mercury diethyl and sodium might conceivably have had some effect on the alkyl halide, we investigated the action of such an amalgam on (+)2-bromoOctanel [a]*'D +32.0', in pentane solution for eighteen hours a t the temperature of the room. The alkyl bromide was quantitatively recovered from this treatment with unchanged optical rotation. An analysis of the results of this study leads to the following conclusions. First, from the value *12.5'* for the maximal rotation of 3methylnonane, and the value [ a ] 2 0 D *33.809 for the maximal rotation of 2-bromooctane, i t follows that if no racemization had occurred during the reaction, the product isolated by us should have Evihad a molecular rotation [ikf-jZ5D -11.3'. dently racemization to the extent of 97% has occurred in the formation of 3-methylnonane. Moreover, i t is complete in the formation of 7,8dimethyl tetradecane. The presence of the latter hydrocarbon, as well as that of octane, among the reaction products is apparently to be connected with the presence of sodium s-octyl arising from a metal-halogen exchange reaction of type (6). Its complete optical inactivity strongly suggests that there exists a real parallel between the method of its formation here and the formation of the corresponding hydrocarbon, 3,bdimethylhexane from (+)2-bromobutane and sodium.2 The cause of the raceniization in both cases may very well be the same. Second, the presence of unreacted (-)2-bromooctane among the products obtained in the present investigation precludes the possibility that a halide of this type is rapidly racemized by an exchange reaction of type (6). Furthermore, we have shown that a sodium alkyl does not attack the tertiary hydrogen of an optically active hydro(8) Levene and Rothen, 1.Or.z. Chrm., 1,85 (1936). ( 0 ) Hughes, Ingold and Nnstumnn,

J . Chem. SOC..1106 (lOa7h

94.5

carbon of the type RlRSRJCH, where R1,Rz and Rs are alkyl groups. Therefore an interpretation of the reaction of sodium with (+)2-bromobutane, which presumes reaction (7) as the initial stage, followed by reactions of types (4),( 5 ) and (6) may not be supplemented by the first or second of the hypotheses enumerated above to explain the optical inactivity of the product. Now of the three remaining hypotheses, which are concerned with the detailed mechanism of the substitution process (4),we can reject the first. This requires that (4)proceed by the mechanism S N l , involving preliminary ionization of the C-Hal bond with a sufficient time lag between the ionization and the subsequent combination of the resulting carbonium ion, and the entering (racemic) carbanion to permit complete racemization of the former. Secondary paraffin halides, however, undergo such preliminary ionization only in media possessing a relatively high dielectric constant as well as a high capacity for solvating ions. Moreover, even under conditions most favorable to such a process, racemization occurs to an extent of no more than 70%.819~10In the present investigation, complete optical inactivity resulted only in the reaction which involved the replacement of bromine by s-octyl. The replacement of bromine by ethyl in the same system led to an optically active product. An interpretation of these results which required that they both proceed through an initial stage involving the Same carbonium ion would, however, also require that the stereochemical results of the two reactions be the same, i. e., both products should have been optically active or both inactive. Since this was not the case, such a n interpretation must be discarded. It is much more likely that these replacements will occur by a mechanism of the type sN26 (equation 11). Such a mechanism is favored by the low dielectric constant and solvating capacity of the medium, as well as by the high basicity of the entering anions. On this basis the important substitution reactions for the system sodium ethyl, (-)2-bromooctane may be written

+ s-CdhBr +CHs(C*Hs)CHC&ls + Br(15; CHa(GHs)CHCeHis + Brs-GHn- + CIHSBr (16) s-Cd&- + s-CdIaBr -+ CI&U + Br(17) C,Hs-

together with the corresponding elimination reactions (10) Huphea, TTEUS. Pnreder Sor.. I& 202 (1910).

940

NORMAX C,. BRISK,JOHX F. TANK

+

+

C?H:,--3- s - C S H I--f ~ B ~C2H6 CaHla Br-- 118) s-CsH17- C?HSl3r+CRHla -I- C2Hr Rr(19) ~-CRHI;- .x-CRHl,Rr --+- CRHls 4-CsHla Br- (20)

+ +

+

S.WALLIS

.?d EVERETT ~

1701.

65

on the basis of this fomiulation that (16) proceeds

about thirty-two times as fast as (13). Actually with ethyl bromide the specific rates of substitution of other anions for bromine by the mechWhether thew reactions occur iii the solution or a t misni SX2 are found to be niuch larger than those the juncture of the soluticm with the ionic crystrl lattice of the sodium alkyl cannot yet be decided, of corresponding substitutions involving secondsiiice without 3 knowledge of the energies, or a r y alkyl bromides. l 2 Moreover, previous iniiiore properly the free energies of activation of vestigatinns6-’2on the relative amounts of eliminathese reactions, no lower limit can be set for the tion and substitution in the reactions of basic anconcentration of carbanioil necessary to effect iotis with ethyl and s-alkyl halides indicate that rapid substitution and elirnination in solution. the reaction nf CzHs- with s-octyl bromide should If the additional assumption is made (hypothe- favor elimination over substitution by a factor sis (d) above) that the sodiuAi s-alkyl formed of about three or four, while the reaction of seither by process (I) or by process (6) retains the CaHI7-with ethyl bromide will greatly favor subinitial configuration of the halide from which it is stitution m-er elimination. The total production produced, then a replacement reaction such as of p s e s would then be small, since the contribu(1‘7) which involves JValden inversion will produce tinns nf (1 q) a n d (19) mould he siiiall fractions of a meso hydrocarbon (equation 11 above). Since, the main re:ictinn (16). In the reaction of sounder this assumption, (15) and (16) will now dium ethyl v i t h n-hexyl chloride studied by lead to enantiomorphic modifications of 3-methyl- Whitmnre and Zonk,s on the other hand, a coniionane, the extensive racemization of this hydro- sidera tion nf the previous i n v e s t i g a t i o n ~ on ~ s ~the ~ carbon is to be interpreted as arising from nearly elimination and substitution reactions of other equal contributions of the two processes. We basic anions on higher n-alkyl halides would are inclined, however, to reject this interpreta- lead to the prediction that about equal amounts tion, since it is improbable that a sodium alkyl of the gas ethane (elimination) and of n-octane formed from an optically active alkyl halide by (substitutioti) wnuld be formed. This was in(6) ur (7) will be optically active or that it could deed the case (here the metal-halogen exchange furnish optically stable carbanions for a replace- did not play a significant part, so that the prinment reactioii of the type (17). S o attempt to cipal reaction was that of C?H6- with n-hexyl prepare a11 optically active Crignctrd reagent by chloride). the actioii of magiiesium on CLII optically active ’The assuni j,tion upon which this interpretation alkyl halide has ever succeeded”; neither is i t rests, mmely, that reaction (22) is a tnuch more possible to obtain optically active amines of the rapid process than (23) remaills unproved, of type R1R2NH which are iso-electronic with the course, until the diastereomeric 7,8-dimethyltetracarbanioils RlRzCH - involved here. decanes c ~ i be i synthesized in the pure state and This difficulty is satisfactorily avoided, how- their propert;es compared with those of the prodever, when the hypothesis (e) above is adopted uct obtained bv this method. I t nlight be exrather than (d). .%gain the substitution reaction pected, hnwever. that the energy of the transition (17) leads to meso-’i&dirnethyltetradecane, since state (and hence the energy of activation) of the the conditions prncess (22) would be considerably lower than that for (!23), since the distances between the d-S-CgH1.lL-s-C&17- (rapid) (21) larger alkyl ,groups, which tend to repel one an1-s-CRHl7- + 1( - )-s-C8Hl,Br meso-(s-CsHlr-)2 other, will be larger in the former than in the (22) d-s-CBH17- + I ( -)-~-CgH17Br-+- d - ( ~ - C g H l l - ) 2 (23) latter transition state.I4 Similar considerations kz2 > > k u would apply to the processes (13) and (14). The are presumed t o obtain. Now process (15) leads R,-k-dirnethylhexane obtained by Wallis and (12) Hu,hes. Ingold a n d their co-workers find for ethyl bromide to (-)3-methylnonane while process (16) leads with sodium hydroxide (alcohol solation) 105 ~ S N Z 171; for ito r-3-methylnonane. From the extent of race- propyl bromide under iden:ical conditions 101 ksai 4.75 (ref. 6, mization of this hydrocarbon, it is to be concluded p. 612).

+

--

(11) Pickard and Kenyon, 1.Chcm. Soc., 99, 65 (1911); Schaarz and Johnson, THIN8 JOURNAL, 61, 1063 (1831): Porter, ibid., M,1436 (19351,

(13) Hughes and Ingold. Trans. Faraday Soc., 84, 657 (1941); see espeeinlly pp. 677480. (14) This waa pointrd @ut t o LIB by Dr. R . B. Pnrcll. of thin

taboratom.

May, 1943

THEACTIONOF SODIUM ETHYLON (-)2-BROM06CTANE

947

Adams,' therefore, may well have been the meso (+)2-bromobutane becomes fonn. l 6 2Na s-C4HgBr +s-CdHgNa NaBr I t is to be noted that if the foregoing formulas-C4HpNa Na s-CIH~. tion is correct, the formation of (-)3-methyl2-s-C4Ho*+ (~-CdHg)t rionane from (-)2-bromooctane by reaction (15) \iC4HS C ~ H W must have involved the Walden inversion. APparently inversion of configuration has occurred. Since the COI~C'CTI tration of ethyl radicals will since (-)2-bromooctane is configurationally iden- be very small, antl since the reaction of anions tical with (-)octanol-2, which in turn has been with ethyl halicles is a rapid process, equations shown by Levene8 very probably to possess the (15), (lG), (18) and (19) above require no modifisame configuration as (+)3-methylnonane, cation. The reaction scheme for the present inFinally, it is of interest to consider an alterna- vestigation then becomes tive interpretation, which is consistent 7 ( - ) C ~ H ~ C H ( C H ~ ) ( C ~-I-HBrI ~ ) (15) (-)s-CsH17Br with known facts. The calculations of C2H5C2H6 CSHM Br(18) Baughan, Evans and Polanyi4 have shown 7 ~ - C Z H L C H ( C H ~ ( C I HBrI ~ (16) that-for sodium methyl there exists a bar(r)s-CaHlr- + C J ~ & rier of about 25 kcal. between the normal C8HlS 4- C2Ha Br(19) ionic state, Na+CH3-, and the state of hornoJ , CllH.34 polar dissociation into sodium atoms and 2s-CsH17Na (2Na 2S-C8Hl7.) (24) C&s C&e methyl radicals, Na' , , .CHa'. For any other sodium alkyl this barrier will be lowered by Again the formation of 3-methylnonane by (16) an amount, R - R-, where R is the resonance is much more rapid than by (15),and (15) leads to energy of the alkyl radical and R- that of the a Walden inversion. The 7,s-ditnethyltetradeccorresponding carbanion. The latter will be ane produced by (24) would be optically inactive quite small for carbanions such as ethyl, i-propyl, for reasons cited a t the beginning of this paper, als-butyl, etc., since the energies of the resonaiice though now it might be supposed to consist of states lie considerably above those of the normal comparable amounts of meso and dl forms. state. The resonance energy R of an ethyl radiAn interesting conclusion to be drawn from this cal is about 7 kcal.; however, that of a secondary analysis is that a sodium s-alkyl should show a alkyl radical is about 12 kcal. It would thus considerable tendency to decompose into sodium, appear that while for a primary sodium alkyl a di-s-alkyl, alkane and alkene, even at compararather high barrier exists for dissociation into tively low temperatures. We hope to conduct a sodium atoms and free radicals, for a secondary study of the properties of such metal alkyls in the sodium alkyl it will be rather low (-13 kcal.). near future. We also contemplate a study of the : I secondary sodium alkyl may well be in equilib- action of a sodium alkyl on ail optically active srium with a free radical concentration sufficient alkyl chloride. This should reduce the metalto effect appreciable formation of coupling antl halogen exchange to a minimum, so that the prindisproportionation products (equations 2 and 3 cipal reactions would be the substitution and above). If this is more rapid than the alternative elixniiiatioii reactions corresponding to (12) and substitution and elitnitiation reactions of the (IS). While considerable olefin formation would corresponding carbanion with the alkyl halide be anticipated, the substitution product should the reaction scheme for the action of sodium on show little if any racemization.

+

+

+

+

+

' '

+

(15) We wish to emphasize at this point that the argument here presented applies only to the action of sodium s-alkyls on s-alkyl halides and not, for example, to the action of sodium (or of the corre. sponding sodium alkaryl) on optically active alkaryl halides. Indeed, in reactions of the latter type the mutual attraction of the aryl groups would be experted considerably lo lbwer the activation energy for the formation of the optically active hydrocnrbon. (Alk) (Ar)CHCH(Alk)(Ar). In agreement with this interpretation, Ott (Bcr., 61, 2124 (1928))has observed the formation of ( --)~.3-diphenyIbutanr Irom the uction of sodium ou (+)u-phenethyl chloride. Wallis nnd A d a m (ref. 2), on the uther hand. obLaiued i i u uplieally inasti\= product from the action of shiiuru on an optically nctive alkaryl bromide. The difiertnor in these results 15 due in rll probability LO the extreme susceptibility of the bromide to tacemization by a halupen-metal exchange reaction (equation 8 ) .

+

+

+

'

+

+

Experimental Part Materials.-( +)2-Bromoijctane,

[n]"D

f32.4' (homo-

gcricous), b. p . 81-82" (19 inm.), was prcpared in go($; yield accordiug to thc iiiethod of Shriner and Youiigle from (--)0~ta1101-2,[n]% -8.95' (honlogeileous), b. p. 85" (21) i i i n i . ) , obtaiiied froiri 111c resolution of dl-ovtaiiol-2 (East-

Kotl;ik Cu., Whit? I.nbti) at:c.ortliiig 10 K.c~iyoii.~'III likt: iIimiIicr ( -...)L'- Lruiric,,jctnii~, ( aI1('u -;