Electrophilic Substitution at Saturated Carbon. XIII. Solvent Control of

Donald J. Cram, Bruce Rickborn, Charles A. Kingsbury, and Paul Haberfield. J. Am. Chem. Soc. , 1961, 83 (17), pp 3678–3687. DOI: 10.1021/ja01478a029...
0 downloads 0 Views 1MB Size
Download PDF
3678

D. J. CRAM,BRUCERICKBORS, C. A. KINGSBURY AND PAI'LHABERFIELU

cyclopropylmethylcarbinol, 2- and 3-methylcyclobutanol, crotylcarbinol and allylmethylcarbinol. With 45% fluoroboric acid a t room temperature, both cyclopropylcarbinols rapidly rearranged to allylmethylcarbinol and variable amounts of unidentified material. The other alcohols were stable. With 22% fluoroboric acid, heated under reflux for 30 min., 3-methylcyclobutanol showed approximately 80% starting material and 20y0 allylmethylcarbinol. Under the same

Vol. 5 3

conditions, 2-methylcyclobutanol was totally rearranged to allylmethylcarbinol and the olefinic alcohols were unaffected. Allylmethylcarbinol gave only polymer and unrearranged alcohol when heated under reflux for 1 hr. with 25% SUIfuric acid. Under the same conditions it is reportedla that allylmethylethylcarbinol is totally rearranged to 4methyl-3-hexen-1-01, This corresponds to rearrangement of allylmethylcarbinol to crotylcarbinol.

[CONTRIBUTION FROM THE DEPaRTMENT O F CHEMISTRY OF THE UNIVERSITY OF CALIFoRNIA AT L O S 24, CALIF.]

ASGELES,L O S A ~ N G E L E ~

Electrophilic Substitution at Saturated Carbon. XIII. Solvent Control of Rate of Acid-Base Reactions that Involve the Carbon-Hydrogen Bond's2 BY DOKALD J. CRAM,BRUCERICKBORN, CHARLES A. KLYGSBURY AND PAULHABERFIELD RECEIVED FEBRUARY 27, 1961 The base-catalyzed hydrogen-deuterium exchange reaction a t carbon alpha t o the nitrile, amide and ester groups has been found to occur with complete racemization in a number of systems. The kinetics of racemization of these systems has been examined with solvent composition as the main variable, and with methanol, methanol-dimethyl sulfoxide mixtures, ethylene glycol, t-butyl alcohol and n-butyl alcohol as solvents. Rate constants, kinetic order and thermodynamic activation parameters were evaluated. With potassium methoxide as base and (+)-2-methyl-3-phenylpropionitrile as substrate, the rate constant for racemization in dimethyl sulfoxide possessed a value -109 times that observed in methanol. The reactivity-depressing effect of hydrogen bonds on alkoxides is emphasized.

In our previous studies of electrophilic substitution a t saturated carbon,3 carbon4 and oxygen6were employed as leaving groups, and proton or deuteron donors were used as electrophiles. This and the two succeeding papers report the results of investigations of the stereochemistry and kinetics of basecatalyzed hydrogen-deuterium exchange reactions a t saturated carbon. In a formal sense, these acidbase reactions are electrophilic substitutions with hydrogen or deuterium as leaving groups, and deuteron or proton donors as electrophiles. Two previous studies of the stereochemistry of base-catalyzed hydrogen-deuterium exchange have appeared in the literature. Wilson et al. ,6 observed that the rates of hydrogen-deuterium exchange and of racemization of optically active l-phenyl-2methyl-1-butanone in a basic solution of dioxanedeuterium oxide were equal. A similar identity of rates was observed when optically active phenyl-ptolyldeuterioacetic acid was heated in aqueous base.? These results were interpreted as involving proton abstraction by base in the rate-controlling step, followed by deuteration6 (protonation7) of a symmetrical arnbident anion on either oxygen6 or carbon.' Of the many investigations of the kinetics of proton abstraction from carbon, only leading refer(1) Preliminary communications of some of the results of this paper have appeared previously: (a) D. J. Cram, B. Rickborn and G. R. Knox, J. Am. Chcm. Sac., 82, 6412 (1960): (h) D. J. Cram, W. D. Nielsen and B.Rickborn, ibid., 82, 6415 (1960). (2) This work was supported by a grant from the National Science Foundation. (3) D. J. Cram, J. L. Mateos, F. Hauck. A. Langemann, R. R . Kopecky, W. D. Nielsen and J. Allinger, J . Am. Chem. Sac., 81, 5774 (1959), is a summarizing paper. (4) Previous paper D. J. Cram and P. Haherfield, ibid., 88, 2363 (1961). ( 5 ) D. J. Cram, C. A. Kingsbury and A. Langemann, i b i d . , 81, 5785 (1959). ( 6 ) (a) S. K. Hsu, C. K. Ingold and C. L. Wilson. J . Chcm. Soc., 78 (1938); (b) C. L.Wilson, ibid., 1550 (1936). (7) D. J. G. Ives and G. C. Wilks, i b i d . , 1455 (1938); see also D. J. G . Ives, i b i d . , 81 (1938).

ences are mentioned here. 9mong compounds studied are nitroparaffins,8 tris-p-nitrophenylmethane,g acetylene,1° haloforms,Il and a number of ketones, nitriles, carboxylic acids, sulfones and amides.12 I n many of these studies, the rates of halogenation were measured, and the rates of proton abstraction and halogenation were presumed to be identical, since in most base-catalyzed halogenations of carbon acids anion formation is rate controlling. l 3 The present study involved use of cyano, amido and ester groups to acidify hydrogen attached to asymmetric carbon, The racemization and exchange reactions of optically active compounds I-V were examined with the possibility in mind that the stereochemistry and mechanism of these electrophilic substitution processes might be solvent dependent as were those that involved carbon leaving group^.^,^ The main objective was to obtain quantitative data concerning the dependence of the catalytic activity of bases on the character of the solvent.

Results The preparations of optically active nitriles 1'" and 114have been described previously, as have the optically active acids16,4from which amides 111 and (8) 0. Reitz, 2. pkysik. Chem., 8176, 363 (1936); (b) R. P. Bell "The Proton in Chemistry," Cornell University Press, Ithaca, N. Y., 1959, p. 109. (9) E. F. Caldin and G . Long, Proc. Roy. Soc. (London), M a s , 263 (1955). (10) L. H. Reyerson, J . A m . Chew. Sac., 87, 779 (1935). (11) J. Hine, N. W.Burske, M. Hine and P. B. Langford, i b i d . , 79, 1409 (1957). (12) (a) R . G. Pearson and R. L. Dillon, i b i d . , 75, 2441 (1953); (b) K. F. Bonhoeffer, K. H. Geib and 0. Reitz, J . Chem. Phys., I , 661 (1939). (13) For instance, Hsu and Wilson, J . Chem. Soc., 623 (1936). demonstrated t h a t optically active I-phenyl-2-methyl-1-butanone was racemized and brominated a t identical rates, hoth processes being catalyzed by acetate ion. (14) D. J. Cram and P. Haberfield, J. Am. C h e m Sac., 83, 2531 (1961). (15) C L. Arcus and J. Kenyon, J , Chem. Soc., 016 (1039).

Sept. 5 , 1961

REACTIONS AT THE CARBON-HYDROGEN BOND

A4CID-BASE

3679

TABLE I STEREOCHEMISTRY OF BASE-CATALYZED HYDROGEN-DEUTERIUM EXCHANGE REACTIONS O F NITRILE, AMIDEAND ESTER CHa CHI

*I

*I I

C O H ~ C H ~ C= - R

--

Run5

C6HsC-

I H

SubstratNature

Concn., N

= R’

H

Solvent

-Base

Concn., N

Nature

T,

OC.

Tm., hr.

%

%

rac.

exch.b

RCNC 0.140 (CH3)&ODbsd (CHshCOK 0.0130 25 1.3 20 19 RCN’ .140 DOCHZCHZOD*,~ COCH2CHz0Na .140 84 19 52 52 R’COP\T(C?H5)2’ .120 (CH8)3CODbmd (CHa W O K ,170 70 1.5 54 51 4 R’CO2C(CHa),‘ .0051 25 0.25 25 23 .150 (CH8)3CODb*d (CHs)aCOK 5 R C O N ( C ~ H ~ ) ~ ~ .140 (CH3)3CODbsd (CHshCOK ,450 i~ 31 49 47 .140 DOCH~CHIOD~*‘ DOCHzCHIONa ,450 155 25 72 71 6 RCON(CtH5)2* Runs are numbered consecutively from table to table. Analyzed by combustion and falling drop method by J. Nemeth, Urbana, Ill. Values have been corrected for incompletely deuterated solvent and for solvent protonation by substrate as the reaction proceeds. Almost optically pure, a Z 64-37.4’ ~ ( I 1 dm., neat). d 97% 0-D. e 99% (OD). Essentially optically pure, a Z 5 D +118.1” ( I 1 dm., neat). 0 Essentially optically pure, $ 5 ~ +-35.56’ ( I 1 dm., neat). h Essentially $75.04’ (1 1 dm., neat). optically pure, CPD

1

2 3

--

IV, and ester V were prepared (see Experimental). Compounds I-V were all obtained in either an optically pure or in a state close to optical purity. Preparations of t-butyl alcohol (99% 0-D) and deuterated ethylene glycol (99% (OD)*) have been described previously. CHJ

*I

CsHjCH2CCN I

I H

CzHs

*I

C6H,CCN

I

CH3

*I

C6HsCCON( CZHS)?

I

I1 H CHJ

*I

111

H CHI

*I

C6HjCHzCCON(CzHs)z

c 6 H , c c o , c ( CH3)s

H IV

V

I

I H

Table I records the results of experiments in which optically active substrates were submitted to the action of alkoxide bases in deuterated solvents. After racemization was 20-50% completed, the reaction was interrupted, the substrate isolated, and its optical activity and deuterium content determined. Nitrile I was run in deuterated t-butyl alcohol (a typical “retention solvent” when carbon is leaving group) and in deuterated ethylene glycol (a typical “inversion solvent” when carbon is leaving group). Amide I V and ester V were racemized only in deuterated t-butyl alcohol. Within experimental error, the percentages racemized and exchanged were identical. Thus the hydrogendeuterium exchange reaction occurred with racemization for the three substrates and the two solvent-base systems studied (runs 1-6). Table I1 records the results of a kinetic study of the racemization of (+)-2-methyl-3-phenylpropionitrile ((+)-I) in methanol, dimethyl sulfoxidemethanol, ethylene glycol and t-butyl alcohol, with the corresponding potassium, sodium or lithium alkoxides as catalysts. From the observed pseudofirst-order polarimetric rates were calculated the second-order rate constants, k2, which are listed. In Table I11 are listed the second-order rate constants, kz, for the racemization kinetics of (-)-2phenylbutyronitrile (( +)-11) a t 25’ in methanol, ethylene glycol and 1-butanol. I n each solvent, potassium or lithium alkoxide or tetramethylammonium hydroxide was used as base. (IS) D. J. Cram and B. Rickborn, J . Am. Chcm. (1961).

Sac., 88, 2178

The second-order rate constants, k2, for the racemization kinetics of (+) -N,N-diethyl-2-methyl-3phenylpropionamide (( +)-IV) and (+)-N,N-diethyl-2-phenylpropionamide (( +)-111) were determined in dimethyl sulfoxide with potassium and sodium t-butoxides as catalysts, respectively. Similarly, the rate constants for the latter amide ((+)111) and for (+)+butyl 2-phenylpropionate ((+)V) were determined with t-butyl alcohol and potassium t-butoxide as base, The results are found in Table IV. Discussion Stereochemistry of Hydrogen-Deuterium Exchange Reaction.-The nitrile, amides and ester used here to study the steric course of the hydrogen-deuterium exchange reaction are all weak acids, capable of producing ambident anions under the influence of base. These anions can in principle protonate either on oxygen or nitrogen to give the thermodynamically unstable tautomers, or on carbon to produce the stable tautomeric forms directly. In terms of the kinetic scheme

\

C=C-

/- I

: 1;

I --CCEN--+ HI k- a stable potentially asymmetric tautomers

C-C=Nt--,

symmetrical ambident anions

‘

C=C=N-

/

;I

OH

--+-

\

,-=”

unstable symmetric tautomers

formulated, if k-b/k-, > 1, then the unstable tautomer predominates in the initial product. However, Since kb/ka >> 1, then K = kak-b/kbk-a < 1, and the stable tautomer would ultimately dominate in the equilibrium mixture. Alternatively, if k-b/ k-a < 1, then the stable and potentially asymmetric tautomer would be produced without having to pass through the symmetrical unstable tautomer. Stereospecificity in the hydrogen-deuterium exchange reactions in these systems would be ob-

3650

TI. J. CRAM,BRUCERICKBORN, C. -4. KINGSBURY . ~ N DPAUL HABERFIELD

\’ol, 83

TABLE I1 RATESOF BASE-CATALYZED RACEMIZATION OF (+)-2-METHYL-3-PHENYLPROPIONITRILE(+)-I Run

7

8 9 10 11 12 13 14 15 16‘ 17b

18 1gn 20 21 22 23 24 25 26 27 28 29‘ 30 31 32 33 34 35 36 37

38 39 40 41’ 42 43 44 45 46‘ 47

48 49 50 51 52 52 54 65 56 57 58 59 60 61’ 62 63 64 65 66 67 68

Solvent

Nitrile concn., N

CHaOH 0.135 CHIOH .153 .162 CHJOH CHaOH .135 ,157 CHsOH CH:OH .162 ,128 CHiOH .144 CHIOH .155 CHIOH .122 CHsOH .135 CHIOH .157 CHIOH CHaOH .157 .180 25Yod (CH1)&0-75%’ CHsOH ,180 25%’ (CH:)2SO-75Yod CHIOH ,180 50%’ (CHi)rSO-50%’ CHaOH ,180 50%’ (CHt)zSO-50~’ CHIOH .180 76. 5%’ (CHt),SO-23.5%’ CHsOH .180 76.5%’ (CH:)pSO-23. 5yod CHIOH .159 90%’ (CHi)pSO-lO%’ CHIOH .159 (CHt)sSO-lO%d CHIOH .182 90%’ (CHa)L40-10%’ CHiOH .201 90%’ (CHp)zSO-lO%’ CHaOH .168 90%’ (CHi)kD-lO%’ CHpOH ,204 90%’ (CH‘)2SO-lOO/od CHzOH .I54 90%’ (CHa)zSO-lO%’ CH3OH ,154 90%’ (CHa)zSO-lO%’ CHsOH .168 90%‘ (CHa)zSO-lO%’ CHsOH .168 90%’ (CHi)2SO-lOO/od CHxOH .180 97%’ (CH:)zSO-3%’ CHIOH .I80 97%‘ (CH:)zSO-3%’ CHIOH .180 CHaOH ~ 97%’ ( C H S ) Z S O - ~ % .180 97%’ ( CHt)zSO-3%‘ CHIOH .180 97%’ (CHa)&O-3%’ CHIOH .180 CHsOH 97%’ ( C H I ) Z S O - ~ % ~ 97%’ (CH,)FSO-~%’ CHaOH .I80 ,180 97%’ (CH:)rSO-30/od CHaOH .180 97%‘ (CH:)zSO-3’%’ CHsOH .180 97%’ (CHajESO-3%’ CHaOH .180 97%’ (CHo)zSO-3%’ CHaOH .180 97%‘ (CHs)zSO-3%’ CH3OH .180 CHaOH 98.5%‘ (C&)zSO-1.5%’ .180 98%’ (CHz)S&-2%‘ CHtOH HOCHzCH20H .170 HOCHzCH20H .158 HOCHzCHzOH .170 HOCHZCHIOH .170 HOCHzCHiOH .153 HOCHzCHiOH ,170 HOCHzCHzOH .170 HOCHzCHiOH .170 HOCHzCHpOH .162 HOCHzCHzOH .150 HOCHzCHzOH .187 HOCHL!H*OH .170 HOCHzCHiOH .153 .162 HOCHzCHiOH (CH3)gCOH’ .180 (CH3)rCOH’ .180 (CHt)aCOH* .180 (CH8)ICOH’ .180 .180 (CHs)jCOII‘

Base

I

Nature

CH30K CH30xa CHiOLl CHaOK CH30Nd CH30Li CHiOK CHaOK CHIOK CHsOK CH OK CHTOLi CHsOLi CHrOXa CHaONa CHjOSd CHaONd CHBOLi CH30Li CHIOK CHIOK CHIOK CHrOK CHsOiTa CHlONa CHiONa CHlOh’a CHS0Li CH30Li CHsOK CHsOK CHsOK CHsOK CHBOK CHsOK CHaONa CHIONa CH:ONa CHaONa

CHI ON^ CHaOLi CIlaOK CHsOK HOCH&!HoOK HOCHzCH20K HOCH~CHZOK HOCHZCHzONa HOCH&HnOiTa HOCH2CHiONa HOCHzCH20Li HOCHtCHZOLi HOCHzCHZOLi HOCHZCHzOLi HOCHtCHlOLi HOCHnCHSOLi HOCH,CH20Na HOCHpCHtOLi (CHahCOK (CHI):COR (CHI )aCOK (CHi)rCOK (CHI):COK

-

T,

Concn., A

oc.

h,I . m.--L set.-'

0.2997 .3189 . 1804 ,2997 .3189 ,1804 .0095 ,0735 .5909 .2609 .2602 ,6067 .6067 .306 ,302 ,117 ,117 ,0876 ,2430 ,0049 ,0049 .0251 .0252 .00360 ,0250 ,1207 .On36 ,0112 ,0112

80.23 g o , 20 80.20 54.04 54. ti0 54.00 85.30 85.30 85.30 8.5.30 85.30 83.88 83.88 X I , 00 50.00 70.00 50.09 25.00 25.00 50.41 31.61 34.62 34. G2 31.61 31.61 31.61 54.35 54.35 31.61 25.00 25.00 25.00 25.00 25.00 25.00 2 5 . c0 25.00 25.00 25.00 26.00 25.00 26.00 25.00 83.85 83.82 83.85 83.85 83.84 83.85 83.85 83.85 83.82 83.88 8.7.88 83.85 60.28 60.28 25.00 25.00 25.00 25.00

0.02 x 10-4 1.2.5 f .01x 1 0 - 4 0 887 f ti.025 x 1 0 - 4 4 . 2 7 f 0.(13 x 1 0 - 4 3 . 4 %i. .10 x 10-6 3.00 f .06 X 10” I 84 rt .05 x 10-4 2.20 c .o; x 1 0 - 4 2.66 f .os x 10-4 2.43 f .06 X 2.41 i. .03 x 10-4 1.32 f . O I x 10-4 1 . 3 2 i .02 x lo.-‘ 9.37 i .47 X 6 . 1 3 rf .31 X IO-” 3.62. i .18 x 10-4 2.67 C .13 X 1.84 rt .08 X 1.94 f .G8 X 4 . 5 3 f .08 x 10-2 1.30 i .02 X lo-’ 8 . 2 9 f .08 X lo-‘ 7.02 f .20 X lo-‘ 1.16 f . 0 4 X 1 . 1 0 2z . 03 x 10-8 1 . 6 3 zk .04 X 2.16 f . 0 2 x lo-* 2.22 f .u4 x 10-2 1.29 f .03 X 1.60 f .04 X 1.59 f .03 X 1.95 f .04 X 2.47 zk .05 X 10-z 3 . 9 1 f .08 X 2 . 2 5 f .04 X lo-’ 1.67 f .03 X 1.98 i .04 x 10-2 2.89 f .06 X 4.74 f .20 x 10-2 1.31 f .03 x 1.62 f .04 X 1.92 .20 x 10-1 2 . 9 6 f .09 X 6.02i . 6 i x 10-5 5.70 i .23 X l G - 6 6.02 f .09 X 6.45 f .27 X 5 . 6 8 i .07 X 5.55 f: .08 x 1 0 - 6 6.08 i .34 x 10-5 5.99 f .17 X 4.58 .09 x 10-5 4.07 f .07 X 3.46 i .12 X 4.70 f .17 X IO-‘ 2.28 f ,05 X IOd 2 . 4 8 f .12 X 101.57 f .03 X IO--’ 7.S3 f . I 6 X lo-* 5.20 rt .07 x 10-3 5.06 f .10 x 10-8

,0021; ,0094 ,0201 .0469 .0943 ,0463 .0075 ,0210 ,0722 ,145 0 2 10 .00”h

,0498 ,0720 ,00583 ,0742 .231 .0103 .lo37 .256 .00954 .0751 .1481 .305 ,6642 .0751 .lo37 ,1481 ,0016 .I3161 .0099 .0207

1.31

+

ACID-BASEREACTIONS AT THE CARBON-HYDROGEN BOND

Sept. 5, 1931

3681

( CH3)aCOH' .I80 (CIIa)sCOI( ,229 25.00 1 . 0 4 f .02 X (CH&COH~ .180 (CH8)aCOK .439 25.00 1.35 i .07 X 71 ( CHa)3COHt .180 (CHs)sCOK .OZ~ 25.00 1 . 2 2 . I x 10-4 Solution was 0.180 M in Solution was 0.353 M in potassium iodide. 0 Solution was 0.170 M in potassium iodide, Solution was 0.2 M in HzO. lithium bromide. d Percentage by weight. e Solution was 0.229 M in potassium iodide. i Solution 0.192 M in lithium h Upper limit of solubility of lithium methoxide. 0 Solution was 0.15 M in sodium iodide. perchlorate. i Less than 0.003 M in water. U'ater 0.01 M. Water 0.22 M . 69

70

*

TABLE 111 RATES O F BASE-CATALYZED RACEMIZATION OF 0.100

M

CONCENTRATIONS OF (

r-

Run5

Solvent

Nature

Base-

- )-2-PHENYLBUTYRONITRILE AT 25.00'

Concn., N

kt, 1. m.-1 sec.-I

CHiOH CHsOK 0.00548 10.6 f 0.3 X lo-' 11.0 . 2 x 10-2 .0376 CHIOH CHsOK 11.1 f .2 x 10-2 CHaOH .00895 N( CH&OH& 9.75 f .22 x 10-2 .00824 CHBOH CH:OLi CHaOH CHpOLi .olio 9.70 f .16 X lo-* CHIOH" CHaOLi .0202 10.5 , I . 2 x 10-2 .OOi15 8.57 f .08 X HOCHzCHzOH HOCHzCHL?K .0448 3.67 j= .05 X HOCHzCHzOH HOCH~CHZOK 3.72 i .04 X loea .0267 N( C I - I ~ ~ O H ~ 80 HOCHzCHzOH ,00442 3.33 =t .10 x 1081 HOCHzCHzOH HOCHzCHzOLi 3 . 2 9 f .04 X 10rZ HOCHEHZOLi .03io 82 HOCHZCH~OH HOCHzCHzOH HOCHzCHzOLi ,0340 2 . 7 1 f .04 X lo-* 84e CHa(CHz)10H KOH .00123 3.75 i. .08 .5.5E CH3(CH2)3OH KOH ,00348 3.95 f .20 .000706 4.55 f .63 86" CH3(CHzhOH N( CH3),0Hb CH3(CHi)aOH N( CH&OH~ .00165 7.13 f .62 88. CHs(CHz)sOH LiOH .000706 1.85 .26 89" CHa(CH2)zOH LiOH .00794 1.55 f .05 a Runs are numbered consecutively from table to table. Actually N(CH3)40H.2H20.0.27N(CH&. Solution was 0.100 M in water. Solution was 0.200 M in lithium perchlorate. * Solution was 0.100 M in water. 72 73 74 75 76 77 78 79

+

+

TABLE IV RATESOF BASE-CATALYZED RACEMIZATION OF AMIDESAND ESTER AT 25.00' CHa CHI

*I

*I

C ~ H ~ C H ~ C=- R

C~HKC- = R'

I

I

H

H Run"

90 91 92 93 94

d u b s t r a t Nature Concn., N

-Base------

Solvent

RCON(C2Hs)t 0.150 (CH:)zSO, 0.004 M (CH3)sCOH R'CON(CzH5)z .150 (CHr)zSO, 0.0003 M (CHi)aCOH R'CON( CtH5)t .150 (CHs)sCOH R'CO&( CHa)r .150 (CH3)aCOH R'COzC( CH8)s .150 (CH8)sCOH Runs are numbered consecutively from table to table.

served only if two conditions were fulfilled. First, k-b/k-, would have to be less than unity. Second, the symnietrical anionic intermediate would have to be asymmetrically solvated and consumed before it passed into a symmetrical solvent shell. Some hope that these conditions would be a t least partially fulfilled was found in the fact that optically pure 2cyano-2-phenylbutanoic acid was found in a previous investigation to undergo base-catalyzed decarboxylations to give 2-phenylbutyronitrile with from 16% net retention to 11% net inversion of configuration, depending on the s01vent.'~ The 2pheriylbutyronitrile anion produced during this decarboxylation was even more stable than the 2methyl-3-phenylpropionitrile anion which was produced as intermediate in runs 1 and 2 of Table I. The results indicate that within experimental error, the exchange and racemization rates for the 6 runs of Table I are equal to one another, and that the exchange reaction is non-stereospecific.lI The

Nature

Concn., N

(CH3)3COK (CH3)zCOKa (CH3)rCOK (CHi)aCOK (CHs)aCOK

0.0870

.0103 .124 .0099 .127

kr, 1. m

-1

sec. -1

3.96 f 0.19 X lo-* 1.04 =k .03 X lo-' 4.59 f .15 X 10 3.05 i .09 : :10-9 r3.82 f .23 X lo-'

preferred interpretation of these results resembles that used by Ingold, et to explain the equality of rates of racemization and exchange of optically active l-phenyl-2-methyl-l-butanone, namely, that protonation or deuteration occurs more rapidly on the more electronegative center of the ambident anion ( k - b l k - a > 1). Thus the present results do not test the possibility that symmetrical carbanions might provide optically active products by virtue of asymmetric solvation. The lack of stereospecificity in the racemization reactions of systems I, 111,I V and V in two solvents as different as ethylene glycol and t-butyl alcohol strongly suggests that the reaction would be nonstereospecific in any solvent, and that racemization and exchange rates can be presumed almost identical. This assumption greatly simplifies interpretation of the kinetic data in the next section. (17) Stereospecificity of less than 15% net inversion and less t h a n 7 % net retention could not have been detected by t h e methods used.

D. J. CRAM,BRUCERICKBORN, C. X.KIXGSBLRY AND PAULHABERFIELD

-6 c

-20

-10

0

VOl. s:3

I i

Loq Molality o t M O W 3 .

Fig. 1.-Order plot of logarithm of base concentration against logarithm kobs for racemizatioii uf (+)-2-methyl3-phenylpropionitrile in methanol a t 85.3": 0 , KOCHa as base; 0, XaOCHs as base; 0,LiOCH3 as base.

Kinetic Order in Base for Racemization Processes.-A large number of kinetic runs described in Tables I1 and I11 were made to establish the kinetic order in base for the racemization process, and to identify as far as possible the structure of the kinetically active catalytic species. At low base concentrations (0.00548 to 0.0376 M) and a t 25', the rate of racemization of (- )-2-phenylbutyronitrile in methanol was first order in base, as shown by the similarity in values of the over-all secondorder rate constants for runs 72-76 of Table 111. Two of these runs involved potassium methoxide, two lithium methoxide, and one tetramethylammonium hydroxide as base. In run 77, the methanol solution was made 0.10 .dlin water with only a slight increase in the second-order rate constant. The reaction order coupled with the lack of dependence of rate on the nature of the anion of the base suggest that with the potassium and quaternary ammonium bases, the base is completely dissociated a t these concentrations, and methoxide anion is the active catalytic species. With lithium methoxide, the slightly lower values for kz are attributed to iricomplete dissociation of the base.ls Figure 1 is an order plot of logarithm of the base concentration against the logarithm of the pseudofirst-order rate constant for the rate of racemization of (+)-2-methyl-3-phenylpropionitrilein methanol a t 85.3'. With base concentrations ranging from about 0.01 to 0.60 M , 5 points that involve potassium methoxide and one sodium methoxide fall on a straight line with a slope of about 1.09. The single point for lithium methoxide falls somewhat below the line. Thus an order somewhat above unity is observed for the base in the racemization process a t (IS) J. D. Reinheimer. W. F. Kieffer, S. W. Frey, J. C. Cochran and E. W. Barr [ J . A m . Chcm. Soc., 80, 184 (1958)lstudied the Kinetics of the reaction of lithium, sodium and potassium methoxides with 2,4dinitrochlorobenzene in methanol at 25'. These authors from a study of the effect of added common ion and non-common ion salts preferred an explanation of their results that methoxide ion was the active nucleophile, and that ion-pairing became important particularly with lithium methoxide.

Fig. 2.-Order plot of logarithm of base concentration against logarithm of k o b s for racemization of (+)-%methyl~o (by 3-phenylpropionitrile in 97% ( C H ~ ) ~ S O - 3CHIOH weight) a t 25.00': 0 , KOCHa as base; 0, NaOCH3 as base; 0 , LiOCH3 as base.

this temperature and for these base concentrations. The lack of a common ion effect or even a marked salt effect is demonstrated by the following comparisons. A value of 2.38 X 1. m.-I set.-' can be calculated for kp (from Fig. 1) for racemization in methanol a t 85.3' with potassium methoxide a t 0.260 -11 concentration. Runs 16 and 17 were conducted a t this base concentration and temperature, but enough potassium iodide was added to each reaction medium to give concentrations of 0.170 and 0.353 JI of the salt, respectively. The rate constants, k l , were not altered by the presence of the salt, and were found to be 2.43 x 1 0 F and 1. set.-', respectively. Less 2.41 X striking is the fact that addition of lithium bromide to give a concentration of 0.1S0 121 of the salt in run 19 gave the same k2 as that of run 18 conducted with the same concentration of lithium methoxide (0.6067 M), but in the absence of added salt. In runs 2 0 4 8 of Table I1 with (+)-2-methy1-3phenylpropionitrile as substrate, the solvent composition was varied between 25% and 98.5% (by weight) of dimethyl sulfoxide with methanol as the second component. 9t 76.5% dimethyl sulfoxide, the value of the second-order rate constant did not change when lithium methoxide concentration was changed from 0.087G to 0.243 1l (runs 24 25 conducted a t 25'). I n 90% dimethyl sulfoxide a t 31.61', a t concentrations ranging from 0.0049 to 0.0250 M, lithium, sodium and potassium inethoxides exhibited values for k3 which are essentially constant (runs 27, 30, 31 and 35). I n runs 28 and 29 conducted in the same solvent mixture a t 34.62' with 0.0251 M potassium methoxide as catalyst, only a 16y0reduction in rate was observed when the reaction mixture was made 0.229 M in potassium iodide. Figure 2 is an order plot for base in 97% dimethyl sulfoxide a t 25'. At concentrations below 0.01 M (runs 36, 37, 42 and 47), lithium, sodium

+

A A ~ ~ ~ REACTIONS - B ~ k ~ AT ~

Sept. 5 , 1931

THE

CARBON-HYDROGEN BOSD

3683

and potassium methoxides gave the same values for k2. Above this con~entrationl~values increase until above 0.0G M concentration of base the reaction becomes about second order in base. In runs -5. 43 and 46, with 0.0210 M sodium methoxide as base, the effect of added sodium iodide on the rate was evaluated. A rate depression of about 3570 kobs. was observed. Comparison of the results of runs 39 and 41 indicates that the presence of 0.02 M water had only a small effect on the rate. - 6. With respect to the effect of base concentration and character on the order in base for the racemization reaction, methanol and dimethyl sulfoxide exhibit surprisingly similar behavior. At base concentrations below 0.01-0.02 AI, the reactions I -20 -10 0 are first order in base and almost independent of the nature of the cation. At higher concentrations, Lo9 Molality of MOCH,CH,OH. the order in base for potassium and sodium methFig. 3.-Order plot of logarithm of base concentration oxides increases, and decreases for lithium meth- against logarithm of hob. for racemization of (+)-2-methyloxide. Addition of common ion salts or small 3-phenylpropionitrile in ethylene glycol a t 83.8' : 0, KOCH*amounts of water to the media have if anything CHzOH as base; 0, NaOCH2CH20Has base; E,LiOCHZonly a small depressing effect on k2. These gen- CHzOH as base. eralizations support the explanation that a t low concentrations, the bases are fully dissociated, and associated and dissociated anions are probably methoxide anion is the catalytic species. At higher catalytically active, the data strongly suggest that concentrations, ion pairs are present and are also free glycoxide ions are the catalytic species, and catalysts, with lithium methoxide less and potas- that essentially all the base is dissociated. sium methoxide more active than free methoxide As might be expected in n- and t-butyl alcohols ion. Superimposed on this general picture is a as solvents, the dependence of kz on base concentrasmall salt effect, negative for free methoxide anion tion and character is more marked. I n the raceand negligible for associated base.20 mization a t 25.0' of (-)-2-phenylbutyronitrile in I n ethylene glycol a t 25.0°, the racemization of n-butyl alcohol which was 0.100 M in water (base ( - ) -2-phenylbutyronitrile follows good over-all concentrations less than 0.008 the values of kz second-order kinetics for base concentrations rang- for the bases decreased in the order (CH3)hNOH > ing from about 0.007 to 0.045 M,with either potas- KOH > LiOH (runs 84-89). Except for the quasium ethylene glycoxide or tetramethylammonium ternary ammonium base, only small changes in the hydroxide as base (runs 78-80). As in methanol, value for k2 are visible with changes in base conthe lithium alkoxide over the same concentration centration. With both variables operating, the range gave rate constants about 10% lower (see maximum spread in values for kz is less than a factor runs 81 and 82 of Table 111). Run 83, conducted in of 5. the presence of 0.200 M lithium perchlorate (soluRuns 64-71 were conducted a t 25.0' with (+)-2tion was 0.034 M in lithium ethylene glycoxide) methyl-3-phenylpropionitrile as substrate, t-butyl gave about a 2070 reduction in second-order rate alcohol as solvent and potassium t-butoxide as constant as compared to the run 82 made without base. Over a range of base concentrations from added common ion salt. 0.0016 to 0.439 M and water concentrations from I n ethylene glycol a t 83.5', the racemization of less than 0.003 to 0.01 M , k2 varied by only a factor (+) -2-methyl-3-phenylpropionitrile exhibits good of about 3. I n the two runs in which water was over-all second-order kinetics with Concentrations excluded as much as possible (runs 64 and 65), a of sodium or potassium ethylene glycoxide that tenfold increase in base concentration decreased the range from 0.00583 to 0.256 M , and with lithium value of kz by only a factor of 2. At about 0.01 M ethylene glycoxide up to 0.0751 M (runs 50 through base concentration, an increase in water concentra57, Table 11). Higher values of this base gave lower tion from less than 0.003 M to 0.01 Air decreased the values for k2. Figure 3 is an order plot of the data, value of kz by only about 3570 (run 65 slightly adand all but the points for lithium alkoxide a t higher justed and run 66). This small decrease points to concentration fall on a line with a slope of unity. potassium alkoxide as being the active species in These data indicate that ethylene glycol as a solvent both runs, rather than potassium alkoxide in one provides a simpler kinetic picture than methanol, and potassium hydroxide in the other. However, and this fact is attributed to the greater dissociating when water concentration was increased to 0.22 M power of the glycol. Except for those runs made a t (base concentration was 0.033 M ) , the rate dropped higher lithium alkoxide concentrations where both by a factor of almost 50 (run 67 slightly adjusted and run 71). I n this case it seems probable that (19) The solubility of lithium methoxide in 97% dimethyl sulfoxide most of the base is in the form of potassium hyis only 0.0027 m. 1. (see Experimental). (20) A. Brandstrom [Arbiu. K e m i . 13, 51 (1959), and earlier articles] droxide, whose catalytic activity is much less than following the early work of S. F. Acree [c.g., J. Am. Cksm. SOC.,87, that of the alkoxide. I n the presence of 0.01 M 1909 (1915)1 devised a method for measuring the relative importance water, the value of kz increased by a factor of about of associated and dissociated salts (bases) in a given reaction through 2.5 as base Concentration was increased from about use of dissociation constants obtained from cunductivity data. I

1

I

I

iw

-1

3684

D. J. CRAM,BRUCERICKBORN, C. A. KINGSBURY AND PAUL HAHEKPIELD

I

I

YO].

83

1

/

The data of Table VI offer striking evidence for the importance of medium in the activity of methoxide anion as a base. In passing from methanol to 98.5Oj, dimethyl sulfoxide, the rate of racemization by methoxide anion increased by a factor of 5 X 10'. Extrapolation of the curve of Fig. 4 to 1 0 0 ~ dimethyl o sulfoxide gives a rate increase over that in methanol of an estimated nine powers of ten. (21) Although kinetic runs were conducted in &butyl alcohol as SOL vent with lithium I-butoxide as base, the solutions were always cloudy due to the extreme insolubility of traces of lithium carbonate that alweyi formed. However, the rate data obtained left llttle doubt that lithium ;-butoxide was less active than potassium I-butoxide by about a power of ten.

These two solvents have dielectric constants that do not differ widely from one another (34 for methanol and 49 for dimethyl sulfoxidez2),and the

ably derives from so1;ation of the metal catfon,22 which leaves the methoxide anion relatively poorly solvated and highly reactive. The shape of the curve of Fig. 4 is informative. The first 10 mole 70of dimethyl sulfoxide increased the rate by a factor of 18, whereas the second 10 mole % increment only by an additional factor of about 2.5, after which the additional factor increased steadily with each additional 10 mole ?' & of dimethyl sulfoxide. Finally in passing from 93 to 9f3.5YGmole yG dimethyl sulfoxide, the rate increased by a factor of about 35. Thus two effects (22) H. L. Schlafer and W. Schaffernicht, Angrus. Chem., 78, 618

(1960).

ACID-BASE REACTIONS AT

Sept. 5, 1961

THE

CARBON-HYDROGEN BOND

3685

ing the activity of methoxide ion. In run 49 of Table I1 is reported k2 for racemization of ( + ) - 2 methyl-3-phenylpropionitrile in 93 mole % sulfolane, 7 mole % ' methanol a t 25.0'. Although a rate enhancement of 7.8 X lo4 (methanol = 1) is observed for the solvent mixture, the rate factor is over two powers of ten short of the enhancement found for dimethyl sulfoxide a t the same mole % concentration. This difference in activating ability CH$. . . ,HOCK* (CH:),SO between the two solvents is attributed to the widely different abilities of the two solvents to form hydroCH& * &(CH& 4-CHIOH gen bonds with methanol. Sulfones are inferior in I this respect.26 A B 0A number of other interesting rate relationships R-H are found in Table VI. Although k2 for potassium t-butoxide in t-butyl alcohol is 4 X lo6 as great as CH3 CHI that of potassium methoxide in methanol, k2 for I I 0-11 O---S-C€I, potassium methoxide in pure dimethyl sulfoxide is :+ several powers of ten greater than that for potasI