~TILLIAM A BONNER
1382
proximately 5 g. of calcium carbonate was added t o t h e silver solution before the reaction was initiated. T h e addition of potassium carbonate after termination of the reaction mas omitted. Control Experiments.--.?, number of control esperiments ( t e n ) were performctl in order t o determine the efficiency of the techniques einpl(iyed ancl the extent of solvolysis not attributable t o reaction with silver ion. Since the esperiments were relatively straightfurivard, detailed descriptions o f all of them will not be presented. One example will he given. In order t o determine whether any appreciable amount of solvolysis of the butenyl nitrates occurred during the workup procediire, and, furthermore, to eitimate the efficiency of t h e ether cxtraction technique employed, a mixture of the following composition (gii-en in weight per cent.) was prepared: tram-crotyl alcohol, 27.84% CY19,295; n-decane, 31 methallyl alcohol, 12. crotyl nitrate, 9.4656 a-meth:rliyl nitrate. T h e mixture was added to the heterogeneous system normally employed, ccntaiiiing sodium chloride, silver chloride and potassium carbonate. T h e inisture was shaken aiit! subseqceutly iT-orkeci up and an:ilyzed by the standard technique. T h e rcsults of thc analysis were: 20.59?o 12decane, 31.OOCyC t;.ans-crotyl alcohol, 2i.56cyG a-methallyl alcohol, 11.93:-i crotyl nitrate, 8.91% a-methallyl nitrate. These data indicate t h a t no medsurabie amount of solvolysis occurred during the workup procedure and t h a t t h e ether extraction was essentially 100tt eiiicient. Vapor Phase Chromatography.-The instrument employed in this work x i s a Perkin-Elmer model l54B vapor fractometer. The two columns used were 2 mcters in length and were filled with Perkin-Elmer packing ‘‘-1’’ and “ B , ” respectively. Helium \vas employed as the carrier gas. Two separate analyses were performed on each of the ether solutions resulting from the partial solvolysis experiments. Analysis of a Xkmicroliter sample a t 65’ (gas pressure, 12 p.s.i.; flow mcter re:i.tling 4.55) using column A provided data on the percentage composition of the mixture with respect to a-methallyl alcohol, c~-methylallylchloride and crotyl chloride. Under these conditions, cis- and transcrotyl alcohols could not be resolved satisfactorily, nor could t h e area of the peak produced by crotyl nitrate be measured quantitatively. Accurate measurement of t h e relatii-e percentages of the p , % m r y and secondary nitrate esters, cisand trans-crotyl alcohols, and a-niethallyi alcohol was possible when columns d and B were cniployed in series a t 1203
[COSTRIEL-TIOSFROM
TIiE
1701. 82
(gas pressure, 30 p.s.i.; flow meter reading, 4.72). A l tthis temperature, however, a considerable amount of rearrangement of the butenyl chlorides was observed. Consequently, a combination of the data obtained from analyses a t both temperatures was necessary for t h e calculation of the relative percentages of all compounds present. Under no conditions could cis- and trans-crotyl chlorides be resolved. No evidence of a separation of t h e tv-o crotyl nitrates was observed. Peak arcas were calculated by evaluation of the product of t h e carefully measured values of the peak height and t h e width of each peak at half the altitude. Control experiments indicated t h a t this method is very accurate. Since recorder response, and consequentl1- peak area, cannot be correlnted directly either with weight or mole per cent., a correction factor, R,was required. This factor was determined empirically b y injecting samples of known composition into the instrument, measuring t h e peak areas, and calculating the factor by which t h e measured area of a peak need he multiplied in order t o give a number representative of the weight per ceiit. of the currespoilding compound present in the mixture. T h e following tahlc gives the factor R for cach of the eight c,mipounds studieJ. Cumpuund
R
a - ~ I e t l i a l l yalcohui l Crotyl alcohol (cis- and t i t r i z s - ) cu-Rletliallyl chloride Crotyl cliloride a-,l.letha! 1yl nit rat e CrotyI nitrate u-Decane
1 000 1.009 1 122 1.045 1 242 1.239 0 961
Ciiritrol experiments performed with mixtures of known composition and measurements of t h e reproducibility of analyses have shown t h a t , unless otherwise stated, t h e percentage composition figures are accurate t o + l.6yc (absolute). In some cases, as noted in tables and the text, t h e accuracy is considerably greater. Hvdrolvsis with Dilute Aqueous Sodium Hvdroxide.The-methbd uf Young and h ~ d r e w s 3was employed for reactions mitli 0.8 aqueous sodium hydroxide. Analysis was b>-v.p.c. uqing cblumns .1itnd R a s described above.
Los AUGEIGS 21,CALIF
DEPARTYEST OF CI-IEXISTRT A S D CHEMICAL ESGISEERISG, S T A S F O R D UNIVERSITY]
Deuterium Isotope Effects During the Raney Nickel Catalyzed Cl-C2 Cleavage of 2-Phenylethanoll BY L
1 , 7 ~ A~. B ~ o ~ s~x m ~ h ~
RECEIVED J U L Y 27, 1959
In order to gain information as to the rate determining step in the Raney nickel catalyzed C1-C:! cleavage of P-phenylethanol into toluene, this reaction has been studied from the i-iewpoint of its deuterium isotope effects, analyzing the mole ratios of cleavage product (toluene) t o dehydroxylation product (ethylbenzene) by vapor liquid partition chromatography. SVhen l,l-dideuterio-2-phen~-lethanolreacted with ordinary Ranel- nickel in refluxing ethanol the toluene/ethylbenzene ratio was identical with t h a t of a control using stock 2-phen~.lethanol. However when 2-pheriylethanol reacted with deuterated Raney nickel in 0-deuterioethanol solvent the tolueiie/cthylbenze~~e ratio decreased markedly, indicating a normal deuterium isotope effect of about 2.1 for the cleavage reaction. Mechanistically these observations are most redsonably interpreted as indicating that the attack on the su!xtrate molecule by hydrogen adsorbed on the catalyst surface is the rate determining step in such heterogeneous C1-C2 hydrogenolyses.
In 1957 we reported the observation that the action of excess Raney nickel in refluxing ethanol on a variety of 2-substituted 2-aryl-ethanols resulted both in simple C 1 dehydroxylation producing alkylaromatic hydrocarbons as well as Cl-C2 carbon bond cleavage yielding alkylaromatics one homolog lower.2 Since that time we have studied (1) This constitutes communicatiun X I V in t h e series “ T h e Stereochemistry of Raney Nickel Action”; ior XI11 see ref. 6. (2) J. A. Zderic, W. A. Bonner and T. W. Greenlee, THISJ O I X N A L ,
7 9 , 1696 (1957).
this hydrogenolytic cleavage reaction extensively with the aid of stereochemical as well as deuterium and radioactive carbon tracer techniques, with the additional findings (a) that the single carbon fragment produced in such C1-C2 cleavages consists of carbon monoxide strongly adsorbed on the nickel catalyst surface, (b) that in neither dehydroxylation nor C1-C2 fission are intermediates produced which in their catalytic environment show an 13) W. A. Ronner and T. 1%’. Greenlee, i b i d . , 81, 2122 (1959).
March 20, 1960
C1-c.2 CLEAVAGE
appreciable tendency to undergo molecular rearrangernentl3s4(c) that extensive hydrogen exchange occurs between the catalyst and substrate molecule undergoing cleavage, such exchange being most rapid a t C2 and occurring there with predominant retention of configuration,6 (d) that the cleavage,' dehydroxylation ratio is increased by the presence of alkyl or aryl substituents a t C2, but is not markedly affected by ortho, @ra or meta-directing groups in the para-position of the nucleus of the 2-arylethanol undergoing cleavage6 and (e) that the C1C2 cleavage process occurs with substantial retention of optical activity and predominant reientim of configuration a t C2.6 (In our earlier publication5 we erroneously concluded from our optical and chemical data that the cleavage of ( - 1 -2-methyl-2-phenyl-1-butanol to (+) -2-phenylbutane proceeded with inversion rather than retention of configuration. LTe wish herewith to correct this misinterpretation and to express our gratitude to Prof. D. J. Cram for calling this oversight to our attention.) These observations now may be tentatively r a t i o n a l i ~ e d ~by, ~a, ~mechanism such as (l), which is in qualitative accord with both the chemical
I
Ar-C-H ____f
iC=O
I
I
R2
I
I
(Si) I11
and stereochemical aspects of such cleavage reactions. Examination of equation 1 discloses that several bond-breaking processes are involved in the over-all mechanism, namely, those of C-H a t C1 and 0-H in I, those of C-H a t C1 and C-C a t Cl-C2 in I1 and lastly that of Ni-H as adsorbed hydrogen leaves the catalyst surface during its attack upon 11. In order to find out which of the bond-breaking processes in (1) might be involved in a rate-determining step, we have now undertaken to examine the Raney nickel catalyzed cleavage of 2-phenylethanol for possible hydrogen-deuterium isotope effects under varying sets of conditions. When Raney nickel in refluxing ethanol acts upon 2-phenylethanol two products result,2 ethylbenzene from simple dehydroxylation a t C l and toluene from cleavage between C1 and C2. U'e have examined this system for deuterium isotope effects by determining the cleavage/dehydroxylation ( i e . , toluene,'ethylbenzene) ratio by means of quantitative vapor liquid partition chromatographic (v.1.p.c.) analysis in cleavage experiments where both the Raney nickel catalyst surface was deute(4) W, A. Bonner, THIS J O U R N A L , 81, 1181 (1959). ( 5 ) W. A. Bonner and T.W. Greenlee, ibid., 81, 3336 (1959). (6) T.W. Greenlee and W. A. Bonner, i b i d . 81, 4303 (1959).
OF
2-PHENYLETHANOL
1383
rated and the two hydrogens a t C l were replaced by deuterium ( i e . , with ll1-dideuterio-2-pheny1ethanol). The results of these experiments are summarized in Table I. TABLE I THE CLEAVAGE
2-PHENYLETHAPjOL I N DEUTERIUM LABELED SYSTEMS
OF
Toluene, mole
No. Catalyst
1 2 3 4 5
Substrate
N ( H ) Ph-CHz-CH2-OH S i ( D ) Ph-CH2-CHZ-OD X ( H ) Ph-CH2-CHa-OH h-l(H) Ph-CHy-CDZ-OH S i ( H ) Ph-CHZ-CH?-OH
7"
52 5 36 9 54 9 47.3 46 0
Ethylbenzene, mole 70
47 63 45 52 53
5 1 1
7 4
Mole 5% toluene/ Mole % ethylbenzene
1 105 0 585 1 217
0 898 0 872
KO.1 in Table I shows the toluene 'ethylbenzene ratio for the cleavage of stock 2-phenylethanol with fresh Raney nickel catalyst. No. 2 shows this ratio for the cleavage of stock 2-phenylethanol using deuterated Raney nickel catalyst in refluxing 0-deuterioethanol, while No. 3 represents the results of identically conducted control experiments run simultaneously. No. 4 summarizes the cleavage results using ordinary Raney nickel in ethanol with l,l-dideuteri0-2-phenylethanol, while No. 5 gives the data for control experiments conducted concurrently with the experiments in No. 4. Ideally, the toluene 'ethylbenzene ratios in Nos. 1, 3 and 5 would be constant. The fact that they are not emphasizes (cf. Experimental) the importance on the cleavage-dehydroxylation ratio of such unknown factors affecting the catalyst surface as the technique of catalyst preparation and the age of catalyst. Thus the data in No. 2 should be compared only with those in the control No. 3, and those in No. 4 only with its control in No. 5 . Comparison of Nos. 4 and 5 , where the analytical data are identical within experimental error, indicates that no deuterium isotope effect results in the cleavage of 2-phenylethanol under these conditions when the two hydrogens a t C1 are replaced by deuterium. Comparison of Kos. 2 and 3, however, shows that a substantial isotope effect favoring dehydroxylation over cleavage occurs when either (a) the adsorbed hydrogen on the catalyst surface is replaced by deuterium and/or (b) the hydroxylic hydrogen of 2-phenylethanol is replaced by deuterium (assuming rapid and complete OH + OD exchange under the reaction conditions employing 0-deuterioethanol). The absence of a deuterium isotope effect in the cleavage reaction of 1,l-dideuterioethanol argues either that C-H bond rupture a t C1 in 2-phenylethanol is not rate determining, or that the rate of H-D exchange between C1 and the catalyst surface is fast compared to the rate of Cl-C2 cleavage, yielding, for practical purposes, nondeuterated substrate. We have a t present no basis for a choice between these alternatives. The occurrence of an isotope effect decreasing the cleavage/dehydroxylation ratio when the catalyst surface is deuterated and the 2-phenylethanol substrate is 0-deuterated suggests, on the other hand, three possible rate-determining alternatives : (a) 0-H bond breaking in I of equation I , (b) rapid pre-
-
1384
WILLIAM A. BOXNER
liminary H D exchange a t C1 in I of equation ( l ) , followed by rate-determining C--D fission a t C1 or (c) Ni-H rupture in I1 of equation 1. The first alternative appears improbable because of the known lability of the 0-E1 bond in a hydroxyl environment, while the second appears soincwhat unlikely because of the known ease with which alcohols are dehydrogenated in the presence I I Iianey nickel.' The third alternative, that Hdesorption from the catalyst suriace is rate determining in the above Cl-C2 cleavage, finds some support in our previously observed difficulty in desorbing hydrogen from a Raney nickel surface, a process which requires baking the dried catalyst in This tentative conclusion that the rate determining step in the cleavage reaction is tlic desorption of hydrogen froni the catalyst surface during attack on the aldehyde intermediate (11 in equation 1) is in accord with other observations regarding isotope effects during heterogeneous catalytic processes. Thus the hydrogenation of ethylene in the presence of various metal catalysts including nickel has been found to be generally slower with deuterium than with hydrogen.3 Similarly, since normal isotope effects have been observed in the adsorption of hydrogen on metal surfaces, it is reasonable to assume that such isotope effects should also prevail in desorption processes. Any preliminary H-l3 exchange reactions a t C211 in KO. 2, Table I, would yield a deuterated substrate whose Cl-C2 cleavage woultl presumably be subject to only a smaller second order isotope effect. From the data in Table I i t is possible to get some idea of the magnitude of the present deuterium isotope effect. The reaction under consideration consists of two competing processes, a and b in equation 2 . If one assumes that both the cleavage (a) and dehydroxylation (b) 9r10
" / Ph-CHt-j-CH?-i-OH
i
ka
,--+
Ph-CHa
(2)
kb
\-
Ph-CH2-CIT3
paths are kinetically of the same order (presumably pseudo first order in 2-phenyIethanol), then it follows t h a t the observed toluene 'ethylbenzene ratios in Table I are also the ratios of the corresponding rate constants, i.e.
kH = 1,217 and kazD = 0,SXj kb-n
kb-D
Furthermore if the dehydroxylation path (b) iri equation 2 , being mechanistically different".!,' from cleavage, is nof subject to a normal deuterium isotope effect, i.e., if kb-13 = k b - ~ then , the isotope JOURNAL,74, 1033 (1052); a '. Reeves arid (7) W. A. Bonner, THIS H. Adkins, ibid., 6'2, 2874 (1940); M. L. Wolfram and J. V. Karabinos, i b i d . , 66, 909 (1944); E. C. Kleiderer and Z . C. Kornfeld, J . Or&.Che!?!., 13. 455 (1948); R. Paul, Bull. SOL. chim., 8, 507 (1941); Combt. r e n d . , 208, 1319 (1939); L. Palfray and S. Sabetay, ibid., 208, 109 (1939). ( 8 ) K. Wiberg, Chem. Revs., 65, 713 (19531. (9) R. M. Barrer, T Y Q M X Faraday . Soc., 32, 481 (193G). (10) E. B. Xlaxted and C. H. hIoon, J . Chcm. S a c , , 1542 (1036). 76, , 6330 (l9,34); \v. A . Bonnrr (11) W. A. Bonner, THISJ O U R N A L and J. A. Zderic, ibid., 78, 4360 ( 1 9 5 G ) . (12) W. A Bonner, J. A. Zderic a n d G A . Casaletto, i b d . , 74, ;IOSlG ( 1%2). ( 1 3 ) U'. A . Llmner and J. A . Zderic, ib7,/., 78, 3218 ( I % X ) .
~
Vol. 52
effect in the cleavage process is given by k a - H k , - ~= 1.2117'0.5X5 = 2.08. If, 011 the other hand, pat11 L in ( 2 ) is subject to a nornial isotope effect of tiiiknowii niagtiitutle, such that ki) H = &-D , then it follows that the isotope effect in the cleawtge path a is expressible as k , - ~ ' k , - ~ = n X 2.0s. 'Thus while the value 2.nS for the isotope effect in the cleavage rcaction is a reasonable one in the light oi other tleuterium isotope effect studies,7 it must be looked upon as a minimzrm value, since we currently haye no experimental data on the presence or absence of isotope effects in Raney nickel catalyzed dehydroxylation reactions. This latter question is currently under investigation.
Experimental Raney Nickel .-This was prepared in the usual fashion ,I4 then stored uiitler water until used. Immediately prior to use the catalyst slurry was rapidly filtered damp-dry on a iiiitered glass funnel and quickly transferred t o a tared xvatch glass. The desired weiglit of damp catalyst was iinrncdiatelv treated with the requisite quantity of solvent thanol iir 0-deuterioethanol) for each cleavage experiment. lie same catalyst batch was used lvitliiii a one-month period for all of the experiments described below. Deuterated Raney Nickel.-The deuterated catalyst waq prepared by equilibrating the requisite weight of tlie above damp-dry catalyst with %ITcdeuterium oxide as previriudy tlescr.ibed.'j Four 21-lioiir equilibrations with fresh dcuterium oxide w x e employed, filtering the catalyst damp-dry betweell each rinse. Deuterated catalyst samples were prerp'irecl i t i duplicate, along witli duplicate samples of catal1,st "equilibr~tecl" h y an identical procedure with ordinary wxter :is cotitrol. For ziiiiuli;itie~iusiibe in one of the cleavage eslwrirrient.; descrihcd l x l o w . 0-Deuterioethanol was prepared by the action of 99' i, deuterium oxide o i i a suspenqion of sodium ethylate in ailIiJ-drous ether a i tlewrihed earlier.15 Exarni twice-distilled product, h.p. 77.5-78.8", by niques indicated it t o bc better than 96:'c cliei geiieous. I t \vas accinrtliiigly used as solvent in one of tlic elc,rvage expcrirneiits hclow with no additional purificatii J I I . l,l-Dideuterio-2-phenylethanolwis prepared hb- tliereduction of phenylacetic acid with lithiurn nluniinurn deuteride ill ether s(ilution. The crude product, wliich had been freed o f tinreacted acid by nasliing with sodium hydroxide solution, was p,urified 1))- tlistillotion, h . p . 123-12io(5il 1nm.i; v.1.p.c. analysis of tlic distillate showed the presence of only a single ciimponent. ?\:ass spectrometric analJ-sis (kindly performed by tlie California Research Corporation) of the distillate ino f the desired product. dicated it tri contain ahrive Vapor Liquid Partition Chromatographic Analyses .-Tlie percentages of tolucne, ethylbetizetie and residual siilvetit i:i tlie crude ciearage reaction products tlescribeti below ivert' determilied by v.l.p.c. atialyii.;. A 1.2 X 120 em. v.1.p.c rrilunin packed wtth a 2 : 1 Celite-siliccitie packing was employed. Operating conditiiins enti-ded a column ternpv'ature of l t 3 0 and a Iwliuni f l i i w rate of 70 inl./niin. The o u t put signal from t b e platinum thermal conductivity deteet(ir i n tlie ciilumn awmibly w a s recorded on a X'ariaii G-10 rcciirder, atiri the bridge circuit a n d iample size injected i t i t o the criiurn;i were selected to give nearly full wale recortlir deflection as compouents lcft the criiuiiiii. I'titler these C ~ I I I tlitioiis solvent, toluene atid et!i>-lbeuzene were cciinplctv1~r?so!ved, tlie latter two components enicrging from tlic column about two minutes apxrt . Annlytical data were oht3incd f r m i tlie v.1 .p.c. traces b y measuring the nw:i Linder the curvc for each component using the product of peak hcigiit times half-peak width. The former parameter was measured with dividers and the latter with the aid of a traveling microscope. .\ series of known mixtures r i f tniuene anti ethy1l)enzene was prepared anti analyzed as above t(J eitabli41 the precision of tlie inetliiitl and t o provide a seric.; of standards which could later be analyzed alongside product mixtures of unknown crmcentrat i r n for control purpoie.;. T:ible 11 intlicates tlie results of t i i t w pre1iniin:iry t1uplic:ite analysts (except for s a l ~ i ~ i 11, le -~
: I I > I I o ~ i n g < >O, r ? . S > n i i : r c i , s , 31, 1: ( l ! l M l ) . (I:!, \V,A . 13
