Effect of solvent on .beta.-arylalkyl solvolysis - American Chemical

228

Journal of the American Chemical Society

1 1OO:l /

January 4, 1978

Effect of Solvent on P-Arylalkyl Solvolysis Frank L. Schadt III,2a C. J. Lancelot,2band Paul v. R. Schleyer*2c Department of Chemistry, Princeton University, Princeton, New Jersey 08540. Received January 26, 1976

Abstract: Solvolysis rate constants, k,, for a series of (3-arylethyl and 1-aryl-2-propyl tosylates in solvents ranging from EtOH to CF3C02H were dissected into their aryl assisted ( F k a ) and aryl unassisted ( k , ) components in order to determine the effect of solvent on each reaction pathway. Correlations of k , and k a for PhCH2CHzOTs and PhCH2CH(OTs)CH3, and of k t for appropriate model substrates, EtOTs and 2-PrOTs, by the full Winstein-Grunwald equation, log ( k / k o ) = mY + IN, demonstrate quantitatively the different sensitivities evidenced by the k , and k a mechanisms to solvent ionizing power ( Y )and solvent nucleophilicity ( N ) :P ~ C H ~ C H ~ Ok T, (Sm, = 0.33, / = 0.78), ka ( m = 0.67, / = o), and EtOTs, k t ( m = 0.36, / = 0.82); PhCH2CH(OTs)CH3, k , ( m = 0.50,l = 0.46), k a ( m = 0.82, / = 0), and 2-PrOTs, k t ( m = 0.60, / = 0.47). Combinations of these constants allow the calculation of the kt(PhCH2CH20Ts)/kt(EtOTs) ratios, which vary 7500-fold in going from CH3CH2OH to C F ~ C O Z Hto, within a factor of 2 of the experimental values and the kt(PhCH2CH(OTs)CH3)/kt(2-PrOTs) ratios to within factors of 2-4. Increased solvent ionizing power does not appear to change the kinetically significant degree of aryl participation in transition states leading to the phenonium intermediate appreciably, as is shown by the slight solvent variation of p ( k a ) and by linear maY plots. All available evidence supports the interpretation that primary and secondary parylalkyl substrates solvolyze by competition between discrete aryl assisted and aryl unassisted pathways, each leading to distinct sets of products. Nucleophilic solvent assistance plays a dominant role in the k , pathway.

Solvolyses of all primary and most secondary P-arylalkyl systems proceed through discrete aryl assisted ( k J and/or aryl unassisted ( k , ) pathways (Scheme I).3 Although C r a m adopted this formulation when phenonium ions were first proposed as solvolysis intermediate^,^,^ two decades of intense investigation were required before the implications were quantitatively understood and all challenges could be answered satisfactorily. These studies also helped lead to the realization that simple6a secondary solvolysis, termed "borderline" in the S N I - S Nspectrum ~ of Hughes and Ingold, occurs with substantial nucleophilic solvent assistance.6b Symmetrical phenonium ions were postulated to account for dominant retention of product stereochemistry in solvolyses of the 3-aryl-2-butyl system (I).4a However, the observed rate constants ( k , ) of I were lower, in some cases, than those of a nonparticipating model, 2-butyl ( k t ( t h r e o - I - O T s ) / k t ( 2 butyl-OTs) = 0.6, acetic acid).7 When corrected by even the most liberal factors for decelerative phenyl inductive effect@ and internal return from the phenonium ions,9 only moderate relative rate ratios were obtained (24,5a435c). If the model system, 2-butyl, solvolyzed by simple ionization ( k , ) , i.e. involved the formation of a nonnucleophilically solvated "open" ion ( k , E k c ) ,the corrected rate enhancements would indicate too small a driving f 0 r c e ~ 5 'for ~ participation in ionization (1.5-2.2 kcal/mol) to be consistent with the formation of a a-bonded phenonium intermediate. As a n "escape from the dilemma",' I b Brown proposed weak x participation during ionization leading to a pair of rapidly equilibrating ions.' appeared Beginning in 1967, a series of which employed substrates containing deactivated phenyl groups as better models for the anchimerically unassisted route ( k , ) . R a t e enhancements observed with activated phenyl groups would result from the incursion of the ka process. This provided, for the first time, methods for the accurate dissection of the total rate constant ( k , ) into its k , and Fk*l6 components, and allowed a clear test of mechanism. Solvent-vs.-aryl competition would predict quantitative agreement between values of Fk4/kt determined independently from rate data and from product stereochemistry, while the interpretation involving equilibrating ions would not predict such a correlation. Since remarkably exact rate-product correlations were f o ~ n d , ~ , ' ~ the k , and k A pathways must be discrete, implying strong assistance by solvent or neighboring group compared to a limiting ionization process ( k , ) . Subsequent studies using the hindered 2-adamantyl substrate as a model for limiting (k,) secondary 0002-7863/78/1500-0228$01.00/0

Scheme I

solvolyses quantitatively confirmed that simple secondary substrates a r e substantially assisted by nucleophilic solvent participation in all but the most ionizing, least nucleophilic media. 7-20 To understand better the nature of k,-ka competition, we analyzed the effect of solvent on P-arylalkyl solvolysis. Changing the solvent from ethanol to trifluoroacetic acid causes a dramatic variation in the relative rate ratio (eq 1: the solvolysis rate, k t , of a substrate capable of neighboring-group assistance relative to that of a nonparticipating model) for 0-phenethyl tosylate (7500-fold, R = H ) and for l-phenyl2-propyl tosylate (100-fold, R = CH3). ~kt(PhCH2CHROTs) .'~ kt(CH3CHROTs) -- F k a ( P h C H 2 C H R O T s ) kS(PhCH2CHROTs) (1) kS(CH3CHROTs)

+

0 1978 American Chemical Society

Schadt. Lancelot, Schleyer -5.5

/

229

Effect of Solvent on /3-Arylalkyl Soluolysis

X

-i.8

II2C

-5.1

-CH2

I

OTs -j.L

-j.7 l o & kt

-6.0

/

-6.3

-6.6

-6.9

-,

v

r .C

I .eo

I

I

.TO

.60

I -50

I

I

.:c

.LO

I .PO

I

I

I

.13

0.3

-.LO

I -.2O

I -.10

0

Figure 1. Plot of log k,(d-arylethyl tosylates), acetic acid, vs.

0,75 O C .

Scheme I1 These large changes, indicating a shift in mechanism from aryl unassisted (k,) to aryl assisted ( k J , could be caused by either of two effects or their combination:21 (1) by a decrease in ks(PhCH2CHROTs) and kS(CH3CHROTs) relative to FkA as solvent nucleophilicity is decreased and/or (2) by a n enhancement of the relative magnitude of FkA as solvent ionizing bridged % power is increased, caused by (a) different responses of the two ion pair dissimilar pathways to solvent ionizing power ( M A> mJ21 or (b) a n increase of aryl bridging in transition states (TSJ Primary 8-Arylethyl Tosylates. The total rate constants, k t leading to the phenonium i r ~ t e r m e d i a t e .Dissected ~ ~ , ~ ~ values (Table I), in all solvents except trifluoroacetic acid were dissected into their k s and FkA components by the Hammett of k, and kA for P-phenethyl tosylate (11) and l-phenyl-2propyl tosylate (111) were correlated with measures of solvent treatment3*l3-l5(Figure 1). Log k, values for deactivated 0aryl substrates were plotted us. u constants to define a k, line. nucleophilicity ( N ) and solvent ionizing power (Y), and other parameters to determine which factors dominate. Differences between log k, for activated compounds and the With regard to the more intimate questions of mechanistic extrapolated k, line yield values of FkA. When several comdetail, it has been suggested that tight ion pairs are the species pounds defined the k, line, a n iterative procedure13ewas emfor which neighboring group and solvent compete in the solployed to calculate and eliminate small percentages of FkA volyses of secondary systems14b,15,23-26 or substrates generfrom the rates of these deactivated substrates. Initially, this a11y.26-28This m e ~ h a n i ~ m ~(Scheme ~ ~ , ~11)~is,kinet~ ~ ~ ,was ~ accomplished ~ , ~ ~ by constructing a plot of FkA for activated ically consistent with the original one involving attack on compounds vs. u+ and extending it through the u+ constants neutral substrate if the rate of internal return of the intimate of the k, substrate^.^^^,^^ Extrapolated FkA rate constants for these deactivated substrates were then deducted from their k t ion pair (k-1) is fast compared to k2 or k, (k-1 2 20 k2 and values to give refined k, points from which a new k, line was k,14b). The wide range of solvent data presented here will be analyzed in light of these two mechanistic alternatives. constructed. Since slight curvatures have been noted in Fka(P-aryl)-u+ plots which include both activating and Results deactivating groups,1b,5c,29,32-34 a scale of substituent constants Kinetics. A series of substituted P-arylalkyl tosylates were more appropriate for @-aryl participation was devised with solvolyzed in a range of solvents possessing greatly different neophyl as ~ t a n d a r d . ' ~Using . ~ ~ the , ~ ~Yukawa-Tsuno correnucleophilicities and ionizing powers. Ethanolyses, acetolyses, l a t i ~ of n ~F ~k a (= k,) for substituted neophyl brosylates in and formolyses were studied by standard techniques; trifluoacetic acid (75 0C),33,34,36 eq 2 (correln coeff 0.997), values roacetolyses were determined by a special procedure, involving of u(neophy1) were calculated from experimental data (log complete evaporation of the acidic solvent, developed here and (kX/kH)/-3.96) or from known values of u and u+ (u described in the Experimental Section. 0.43(u+ - u)).

+

Journal of the American Chemical Society

230

1 1OO:l 1 January 4, 1978

Table I. Summarv of Kinetic Data for B-Arvlethvl Tosvlates Solvent Ethanol

Substituent P-H p-CH3 p-CH3O

50% (v/v) aq ethanol

m,m'-(CF3)z

m-CF3 m-C1 m-F p a P-H

m-CH3

p-CH3O

Acetic acid

2-Fluorenyl

Formic acid p-c1

TemD. " c

75.04 75.01" 75.01 75.01 75.01" 75.03 100.6 75.0b 75.05 100.6 75.0b 75.08 100.6 75.0b 100.6 75.05 100.6 75.0b 74.99 100.6 75.0b 75.02 100.6 75.0b 75.04 100.6 75.0b 90.19 103.37 75.0b 75.01 100.6 75.0b 75.05 100.7 75.0b 75.03 90.14 75.0b 75.05 84.43c 100.6 75.0b 75.0 b.d 75.0bsd 75.0 b.d 75.0bsd 75.0bsd 75.0bsd 75.0b.d 75.0a.b.d.e 103.60 123.63 75.0b 75.0bte 103.56 123.70 75.0b 103.85c 124.09c 75.0b 75.0bed 75.24c 84.43' 75.0b 60.82 75.55, 75.0b 75.M 60.82 75.54 75.0b

k,. S-' (7.52 f 0.09) X (7.08 f 0.06) x (8.10 f 0.10) X (1.38 f 0.02) x 10-5 (1.35 f 0.03) X (3.47 i 0.10) x 10-5 (2.64 f 0.07) X 3.46 x 10-5 (4.09 f 0.06) X (3.17 f 0.08) X 4.07 x 10-5 (4.16 0.02) x 10-5 (3.42 f 0.28) X 4.13 x 10-5 (3.17 f 0.05) X lod4 (4.26 f 0.09) X (3.16 f 0.05) X 4.24 x 10-5 (4.48 f 0.03) X lo-' (3.24 f 0.04) X 4.48 x 10-5 (4.40 f 0.02) x 10-5 (3.36 f 0.01) x 10-4 4.39 x 10-5 (5.39 f 0.06) x 10-5 (4.20 f 0.14) X 5.37 x 10-5 (1.87 f 0.01) x 10-4 (5.04 f 0.04) X 5.44 x 10-5 (5.48 f 0.04) X (4.45 f 0.05) x 10-4 5.48 x 10-5 (7.39 i 0.17) x 10-5 (6.61 f 0.04) X 7.36 x 10-5 (8.88 f 0.10) X (3.53 f 0.07) x 10-4 8.86 X (2.87 f 0.1 1) x 10-4 6.40 x 10-4 (2.56 f 0.07) X 2.81 x 10-4 1.52 x 10-7 1.66 x 10-7 1.62 x 10-7 1.90 x 10-7 1.86 x 10-7 1.87 x 10-7 1.94 x 10-7 2.85 x 10-7 (5.75 f 0.02) x 10-6 (3.42 f 0.02) x 10-5 3.18 x 10-7 8.46 x 10-7 (2.12 f 0.02) x 10-5 (1.31 f 0.02) x 10-4 1.12 x 10-6 2.53 x 10-5 1.54 x 10-4 1.35 X 8.53 X lov6 2.24 x 5.22 X 2.19 X (1.91 f 0.01) X lov6 (9.41 f 0.32) X 8.89 X 4.29 x 10-5 (1.84 f 0.03) X (8.75 f 0.19) X 8.28 x 10-5

AH*.kcal/mol

hs*. eu

19.8

-22.4

20.0

-21.5

20.6

-19.6

19.6

-22.7

19.3

-23.4

19.8

-21.8

20.1

-20.8

19.7

-21.7

20.4

-19.6

21.4

-16.4

22.3

-13.5

21.5

-13.4

23.8 23.7 24.6 23.4 23.9 24.2 24.6 24.8 25.7

-21.7 -21.7 -19.4 -22.5 -21 .o -20.2 -18.9 -17.7 -14.8

25.6 26.1

-13.0 -11.1

25.8

-11.6

25.1 22.1

-10.0 -21.3

24.4

- 12.0

24.2 23.8

-9.4 -9.1

Schadt, Lancelot. Schleyer

/ Effect of Solvent on &Arylalkyl Solvolysis

23 1

Tsble I (Continued)

Solvent

Trifluoroethanol

Trifluoroacetic acid

Substituent

TemD. " c

m-CH3 P-CH~ m,p-(CH&

75.w 75.w 60.78 75.56 75.0b 75.0" 75.Ob.g 75.09 75.0g 75.0g 49.66 75.16 75.0b 49.75 75.1 1 75.0b 75.0' 75.0b.J 35.07 49.76 75.0b 49.79c 59.85c 75.0b 36.09 49.81 75.0b 36.03 49.76 75.0b 36.08 49.85 75.0b 35.99c 49.85c 75.0b 35.09 49.77 75.0b 30.08 49.76 75.0b

p-CH3O P-NO2 P-H P-CH~ p-CH3O p-CI P-H

P-CsHs 2-Naphthyl m-CH3 P-CH~ mp-(CH3)2 2-Fluorenyl p-C6H5O p-CH3O

a

38.

kt,S-'

M ,kcal/mol

7.93 x 10-5 2.94 x 10-4 (1.30 0.02) x 10-4 (5.75 f 0.05) x 10-4 5.45 x 10-4 1.77 x 10-3 1.68 X 4.82 X 5.31 x 10-5 3.39 x 10-4 (9.59 f 0.69) X (6.85 f 0.29) X 6.78 x 10-5 (7.09 f 0.20) x 10-5 (4.49 0.38) x 10-4 4.46 x 10-4 (3.95 0.11) x 10-4 3.22 x 10-4 (4.31 f 0.20) x 10-5 (1.67 f 0.26) X 1.32 x 10-3 1.75 x 10-4 3.90 x 10-4 1.20 x 10-3 (4.15 f 0.37) X (1.48 f 0.07) X 1.18 x 10-3 (1.61 f 0.07) X (5.57 f 0.31) x 10-4 4.24 x 10-3 (3.46 f 0.15) X (1.23 f 0.02) x 10-3 9.67 x 10-3 2.13 x 10-4 6.23 x 10-4 3.53 x 10-3 (1.65 f 0.07) X (5.54 f 0.45) x 10-4 3.51 x 10-3 (9.70 f 0.20) x 10-5 (4.87 f 0.12) x 10-4 2.97 x 10-3

24.1 23.0 22.6

-8.4 -8.8 -8.9

21.7 19.3h 20.5

-9.3 -28.2' -24.4

16.6

-30.3

15.6

-29.4

19.7 20.2 17.6

-17.8 -16.8 -21.5

16.4

-25.2

17.8

-21.2

17.3

-20.0

17.7

-17.4

14.7

-27.7

15.7

-25.0

15.3

-26.4

LU*, eu

Reference 10. Calculated from data at other temperatures. One run. Reference 13e. e Reference 32. f Reference 22. g Reference For ethyl nosylate; see ref 38c. Reference 42. j Reference 72,

log ( k X / k H = ) -3.96[a

+ 0.43(a+ - a)]

(2)37

Table 11. Values of Substitutent Constants

Substituent Ua U+" u(neophy1) These are listed in Table 11. The P-arylethyl data for 50% aqueous ethanol and acetic acid (Figure 1) were analyzed by 0.84 1.04 (0.93)c m,n1'-(CF3)2~ the iterative dissection method, using the deactivated substrates 0.78 0.7801 0.79 p-NO2 (0.57)c m.m'-(CF& to m - F to define the k , line. T h e k , lines for 0.54 0.61 P-CF3 0.52 (0.46)c m-CF3 0.42 formic acid and trifluoroethanol were constructed using the 0.39 0.41 (0.40)c m-Br kt(= k,) value for the p-NO2 compound and label scrambling 0.37 0.40 (0.38)c m-C1 data for the parent substrate ( k , = 0.1kt).38,39 A similar two (0.34)c 0.34 0.35 m-F point k, line was obtained for ethanol using label scrambling 0.23 0.1 1 0.18d p-CI data for the p - H ( k , = 0.994kt)39andp-CH3O ( k , = 0 . 5 6 l ~ ~ ) ~ 0.0 0.0 0.0 H compounds. T h e k , pathway makes a negligible contribution -0.07 -0.07 -0.07 m-CH3 in trifluoroacetic a ~ i d . Resultant ~ ~ - ~ ~ FkA values for each P-CH3 -0.17 -0.3 1 -0.22 solvent are given in Table 111. -0.38 -0.29 m.~-(CH3)2~ -0.24 A preliminary observation from the data in Table I11 is the -0.27 -0.78 -0.50 p-CH3O reduced influence in trifluoroacetic acid of the highly activating a L. M. Stock and H. C. Brown, Adu. Phys. Org. Chem., 1, 35 ether substituents. Both the p-CH3O and less basic p - C 6 H 5 0 Deter(1963). * Additivity of substituent effects is compounds are slower than p-CH3. Ample evidence exists for mined from eq 2 and values of u and u+. Determined from tosylate this type of deactivation through hydrogen bonding of oxygen data.32 and nitrogen-containing substituents in trifluoroacetic a ~ i d $ ~and , ~in~ other acidic solvents (i.e., H C 0 2 H 4 5 and tions involving the effect of solvent on the anchimeric pathway CF+2HzOH45-47). Therefore, the most activated, most k A (vide infra). Brown, at one time, interpreted the slight downprone aryl groups are unfortunately eliminated from correlaward curvature of a kt(= F ~ A ) - uplot + for activated j3-aryl-

232

J o u r n a l of the American Chemical Society

/

1OO:l

/

J a n u a r y 4, 1978

Table 111. F k a Values for P-Arylethyl Tosylates in Various Solvents, 75 "C, s-'

Solvent Substituent P-NO2 m-Br p-CI P-H m-CH3 P-CH3 m9p-(CH3)z p-CH30 P-C6H50 2-Fluorenyl F (P-W

EtOH

50% aq EtOH

4.53 x 10-9 4.52 X 6.7 x 10-7 6.35 x -1

d

4.24 x 4.13 x 2.12 x 3.49 x 2.27 x

a

10W 10-5 10-5

10-4

FSld

AcOH 3.53 x 10-11 7.49 x 1.03 x 6.23 x 8.91 x 8.30 x

10-8

10-7 10-7 10-7

1.14 X 0.32e

HCOzH

CF3CH20H

5.34 x 3.86 x 7.47 x 2.89 X 5.40 x 1.77 x

10-6

4.34 x 10-6

10-5

5.21

X 10-5

10-4

10-3

0.91f

CF3C02H 2.14 X b,c 4.80 X 10-6 b , c 6.78 x 10-5 4.46 x 10-4 1.18 x 10-3 4.24 x 10-3 9.67 x 10-3 2.97 x 10-3 3.51 x 10-3 3.53 x 10-3 0.42,h 0.22,' 0.281

0

3.38 x 10-4 0.94g

Calculated by the iterative method. Reference 43. Corrected for leaving group (ONs/OTs = 6.0) and contribution from References 40 and 41. e Reference 32,90 OC. f J. W. Clayton and C. C. Lee, Can.J . Chem., 39, 1510 (1961), 74 "C. g Reference 38,75 OC. * Reference 42, 75 "C. ' Calculated for m-Br-I1 nosylate, 70 OC.b1 Calculated for p-N02-I1 nosylate, 110 OC.6 ks.43372

ethyl tosylates in formic acid as evidence for considerably netic methods, 39 f I O ( A c O H ) and 80 f 10 ( H C 0 2 H ) a r e weaker aryl charge delocalization in Fka than in aromatic in close agreement, considering the different assumptions substitution reactions.22 Since it is now agreed that k a tranmade, and correspond reasonably well with results obtained sition state is strongly bound,3 the observed nonlinearity from stereochemical studies, 25 ( A c O H ) and 68 ( H C 0 2 H ) . probably results from deactivation of the important p-CH30 Acetolyses of threo-3-aryl-2-butyl brosylates for which there compound by hydrogen bonding in formic acid.45 a r e more kinetic and product data, when analyzed by the itAn initial attempt to extend the range of correlation by erative Hammett method, agree very well with the results combining data for meta- and para-substituted compounds determined from product stereochemistry. T h e success of these with those for polynuclear aromatic substrates48 was abantreatments provides strong e ~ i d e n c e ' ~that ~ - ~the 1-aryl-2doned for several reasons. While uo constants are available for propyl system solvolyzes by discrete aryl assisted and aryl the latter substrate class,49 u+ values depend on the method unassisted processes. of d e t e r m i n a t i ~ n . ~In~ addition, ~ . ~ ~ * ~the ~ most activated site Influence of Solvent Nucleophilicity ( N ) and Solvent Ionizing for attachment to the polynuclear aryl is the cy position, inPower (Y). Winstein emphasized the discrete nature of the aryl troducing peri or ortho steric interaction^.^^ There does exist unassisted ( k , ) and aryl assisted ( k a ) pathways by plotting a tert-cumyl chloride derived o+ constant for 2 - f l ~ o r e n y 1 , ~ ~k , ( P h C H z C H R O T s ) vs. kt(CH&HROTs) and and a good fit is obtained to the p+ line defined by the log FkA k a ( P h C H 2 C H R O T s ) vs. kl(neophyl-OTs) in a variety of values for other activated substituents in acetic acid.13e solvents for primary ( R = H)41and secondary (R = CH3)35 However, F k A for P-(2-fluorenyl)ethyl tosylate, approximately P-aryl substrates. These correlations can be obtained directly twice that of t h e p - C H 3 substrates in acetic acid, is lower than against measures of solvent nucleophilicity ( N ) and solvent the value for t h e p - C H 3 compound in trifluoroacetic acid. The ionizing power (Y) in the general Winstein-Grunwald equafluorenyl derivative is the only member of the series to form tion for solvolytic processes (eq 3):1'~919,573~8 a colored solution (b1ue-gree1-1~~) initially in trifluoroacetic log (klko) = m Y IN acid, perhaps indicating radical cation formation in this solvent. (3) The correlational range was extended somewhat with the Values of N have been determined19 for a wide range of solm p d i m e t h y l substituted compound. The B constants for such vents by evaluating eq 3 for methyl tosylate (eq 4). T h e acelimited alkyl substitution have been shown to be additive.55 tic-formic ~ A value F (0.3) was used as a measure of the subSecondary 1-Aryl-2-propyl Tosylates. Table IV presents strate's sensitivity to ionizing power, and l = 1.O in view of the total rate constants, kl, for a series of 1-aryl-2-propyl tosylates extremely high sensitivity of methyl derivatives to solvent which were dissected into k , and F k A components by H a m nucleophilicity (eq 4). mett3,13aand Taft treatment^.^,'^^ Table V lists kt values for 1,3-diaryl-2-propyl tosylates which were analyzed using a N = log ( ~ P o ) c H ~ o-TO.3Y ~ (4) multiple substitution t e c h r ~ i q u e ~to x lestimate ~~ the magnitude of aryl assisted and aryl unassisted pathways for each l-arylChoice of a proper solvent ionizing power scale for use in eq 3 to correlate the tosylate rate d a t a of the present study and 2-propyl unit. Values of F k a for each 1-aryl-2-propyl tosylate calculated by the three treatments (Table VI) show good in eq 4 posed difficulties. T h e standard measure of Y, based agreement. on the solvolysis of tert-butyl c h l ~ r i d e ,generally ~ ~ . ~ ~ produces Table VI1 compares the percent of aryl assisted reaction dispersion when employed to correlate limiting ( k , ) substrates (1 00 F k J k l ) calculated by the three treatments with experihaving other leaving group^.'^,^^ This standard has proved mentally observed percent yield of ester with retained condeficient in certain fluorinated ~ o l v e n t (the s ~ ~Y~value ~ ~ of 1.84 figuration. T h e values for 1-phenyl-2-propyl tosylate increase measured for trifluoroacetic acid on the tert-butyl chloride 3-1 1% on going from one to 25 iterations in the H a m m e t t scale64 is below that of formic acid, 2.05), because of ion pairing effects.62a W e have chosen 2-adamantyl tosylate as the treatment. Although this reduces somewhat agreement with the observed percentage of F k i as determined from product reference c ~ m p o u n d . ~ ~Although ~ ~ * * ' a~ secondary substrate, it ionizes through a limiting k, mechanism.17 R a t e constants data,56 we believe the iterative procedure ideally to be the more for 2-adamantyl tosylate plot linearly (correln coeff 0.999) with appropriate method of analysis. Small inaccuracies or lack of kinetic data to sufficiently define the k s line can cause signifthose for p-methoxyneophyl tosylate, a kA substrate suggested by Winstein as a standard to measure ionizing power uncomicant variations in the calculated value of F k a , especially for compounds where this contribution is small. T h e percentages plicated by internal return,20 for compounds containing sulfonate leaving groups.6' While literature data for the latter of F k i ( 1-phenyl-2-propyl tosylate) obtained by the three ki-

+

Schadt, Lancelot, Schleyer

/

233

Effect of Solvent on P-Arylalkyl Solvolysis

Table IV. Solvolysis Rate Constants of 1-Aryl-2-propyl Tosylate, X C ~ H ~ C H ~ C H ( O T S ) C H ~

Solvent

X

Ethanol

P-H p-CH30

80% v/v aq ethanol

p-NO2 P-CF3 m-CF3 m-CI p-CI P-H p-CH3 p-CH3O p-NO2

Acetic acid

m-CF3 m-CI

Temp, "C 50.0" 49.60a 50.0b 50.0'~~ 50.0emd 50.0C.d 50.0C,d 50.0',d 50.0C,d 50.0C,d 50.0',d 50.OC.' 1OO.OC~' 100.4' 50.0f lOO.0.f 75.12 100.30' 50.0' 100.0' 75.13 100.0e 5O.Oc 50.0C,e 100.0' 75.1 1 99.92 5O.Oc

p-CH30

100.0' 101.1' 101.1 50.0",' 1OO.O".'

Formic acidg

p-NO2 p-CF3 m-CF3 m-CI p-CI P-H p-CH3 p-CH3O

Trifluoroacetic acid

P-H

50.0e 75.0' 75.04' 50.0 50.0C,' 75.0C,e 50.0',' 75.0C,e 50.0',' 75.0Cfe 50.0e 75.Oc,' 50.0' 75.0',' 50.0C,' 75.O 50.0' 50.0c,J,k

k,, s W 1

(1.41 f 0.03) X (8.28 f 0.13) x 8.65 x 4.39 x 10-6 5.25 X 6.11 X 5.24 X 6.20 x 9.42 X 1.85 x 10-5 9.55 x 10-5 1.21 x 10-7 3.22 x 10-5 (5.68 f 0.06) X 1.92 X lo-? 5.46 x 10-5 (4.77 f 0.07) x (6.86 f 0.14) X 2.22 x 10-7 6.66 X (5.72 f 0.08) X (7.82 f 0.07) X 2.72 x 10-7 6.36 x 10-7 (1.83 f 0.02) x 10-4 (3.27 f 0.06) X (4.09 f 0.06) X 1.72 X 4.12 x 10-4 4.57 x 10-4 (4.72 f 0.03) X lov4 1.24 x 10-5 2.21 x 10-3 (1.32 f 0.03) X (2.22 f 0.10) x 10-4 (2.55 f 0.20) x 10-4 1.67 x 10-5 2.33 x 10-5 3.23 x 10-4 3.22 x 10-5 4.86 x 10-4 6.74 x 10-5 9.48 x 10-4 (3.04 f 0.01) x 10-4 3.99 x 10-3 (1.66 f 0.05) X 1.97 x 9.06 x 10-3 1.24 x lo-' (3.47 f 0.02) x 10-3 5.11 x 10-3

AH*, kcal/mol

AS*,eu

21.4 21.6 20.1 21.9 22.1 22.3 22.1 22.6 26.1

-17.1 -15.9 -20.4 -15.0 -14.2 - 12.8 -11.8 -7.2 -9.7

26.7

-6.7

26.4

-6.9

26.5

-5.0

25.6

-5.9

24.1

-6.4

24.8

-4.4

22.9

-9.1

23.6

-6.2

23.0

-6.7

22.4

-5.6

21.7

-4.4

22.7 19.2 19.5

2.3 -10.5 -8.9

a S. Winstein, M. Brown, K. C. Schreiber, and A. H. Schlesinger, J. Am. Chem. Soc., 74,1140 (1952). b Calculated using the divisor 1.05 decrease in rate for the p-CH3O substrate (80% v/v ethanold) in going from 50 to 49.60 "C. Calculated from data at other temperatures. References 13f. e Reference 13a. f Calculated using divisors of 1.04 and 296 f 18 decreases in rate in going from 100.4 to 100.0 and 50.0 "C, respectively, for other k , substrates (AcOH) p-NO2, m-CI, and p-CI. g Determined by a conductometric method, as were all formolyses. Calculated using the divisor of 15.3 f 1.4 decrease in rate in going from 75 to 50 "C for other k , substrates (HC02H),p-NOz, m-CF3, and m-CI. Reference 35. J Reference 74. Buffered with sodium trifluoroacetate.

compound are limited to the ethanol-formic acid range, rate data for 2-adamantyl tosylate are also linear (correln coeff 0.994) with those for neophyl tosylate in the complete ethanol-trifluoroacetic acid s p e c t r ~ m . ~Fortunately, ~ ? ~ ~ ~ , ~the ~ difference in Y value between acetic and formic acid in the tert-butyl chloride and 2-adamantyl tosylate scales is virtually the same (3.69 vs. 3.65 Y units), thereby allowing calculation of N values for both Y scales using the same m A F value, 0.3, in eq 4. Primary 8-Phenethyl Tosylate. Values of k, (75 "C) for P-phenethyl tosylate for the solvents in which k , was observable

(ethanol, 50% ethanol, acetic acid, and formic acid) were correlated by eq 5 (correln coeff 0.998). The value of k, in C F 3 C H 2 0 H was not included because of the difficulty in obtaining accurate rate data, as well as N and Y values in this anhydrous m e d i ~ m ~ ~(see a ,Table ~ ~ ~VIII, , ~ footnoten. ~ The k a rate c o n s t a d 8 (75 "C) in these same four solvents and trifluoroacetic acid were correlated by eq 6 (correln coeff 0.997). Comparison of eq 5 and 6 clearly shows the very different sensitivities to solvent nucleophilicity and ionizing power of the k , and k A pathway, as Winstein's results anticipatedS4'

234

Journal of the American Chemical Society

AcOH

75.0"

P-H

100.0",b

125.0" p-CH3O

100.0",b 100.0b

P-NO2

125.0 75.13

P-H

100.0b 100.0b~C

P-H

125.20 75.13 100.40b.C 100.0d 50.21 75.0h 24.12

p-CH30 HC02H"

p-H p-CH30

50.10

75.0b,d 75.0h 84.6 24.12 50.20 75.0b,d 50.12 75.10b 75.0d 24.12

P-NO2

P-H P-H p-CH30

50. I O

75.0h.d

(5.68 f 0.02) x (8.98 i 0.05) x (9.65 f 0.1 1) x (1.68 f 0 . 1 3 ) X (3.71 f 0.06) x (3.90 f 0.05) X (6.58 f 0.16) X (8.09 f 0.49) X (2.10 f 0.02) x (1.82 f 0.03) X (1.33 f 0.03) X (1.90 0.02) x 1.83 x (1.71 f 0.05) X (2.56 f 0.08) X (1.70 f 0.05) x (3.96 f 0.10) x 5.24 x (1.14 f 0.04) X (3.15 f 0.10) x (8.21 f 0.24) X (1.88 f 0.06) X 2.40 x (8.48 f 0.40) x (1.77 f 0.02) x 1.75 x (1.29 f 0.04) X (3.63 f 0.14) X 5.60

/

1OO:l

/

January 4,1978

27.5

-3.7

27.0

-11.5

25.3

-5.2

24.5

- 14.6

26.5

-5.1

23.7

-2.5

10-5

10-4

10-5

10-4 10-4

22.5

10-4

0.0

10-3 26.9

-4.3

22.3

-2.1

10-5

I 0-3

26.5

0.2

10-4

10-4 23.9

-0.5

x 10-3

a J. J. Harper, Ph.D. Thesis, Princeton University, 1968. Reference 13b. Determined by an automatically recording conductometric method, as were all formolyses. Calculated from data at other temperatures.

Table VI. Values of F k A of I-Aryl-2-propyl Tosylates Determined by Various Methods, 50 O c a , s-l

Solvent Substituent

Method

p-CI

Hammetth

P-H

Hammetth

p-CH3

Taft Mult subst Hammett

p-CH30

Taft Hammettb

EtOH

1.0 X IO-'

2.56 X

1.10

7.04 X 10-6

x 10-5

0.62 x 8.75 X 10-5 8.32 x 10-5

Taft Mult subst F values

80% EtOH

=I

%ld

AcOH (temp, "C)

3.24 x 10-7 8.56 x (100) 7.6 x 10-5 (loo) 5.02 X (100) 1.34 x 2.87 x 10-4 (100) 2.83 x 10-4 (100) 1.20 x 10-5 2.07 x 10-3 (loo) 2.08 x 10-3 (loo) 2.02 x 10-3 (loo) 0.19e

HCOzH (temp, "C) 4.70 x 6.52 x 2.79 x 3.64 x 3.12 x 2.68 x 1.63 x 1.93 X 1.86 X 9.03 x 1.24 X 1.23 X 1.19 X

CF3C02H

10-5 10-4 10-4

(75)

3.46 x 10-3

(

10-3 (75) 10-3 (75) 10-3 (75) 10-3 (75) (75) 10-3 10-I (75) 10-I (75) IO-' (75) 0.09f

1

(D-H) m,),causing a preferential increase in the numerator of eq 9. Values of k, for P-phenethyl tosylate, in which the k a pathway is available, generally increase with increased ionizing power (Table VIII, columns 3 and 7 ) . Also, for solvent pairs of similar nucleophilicity but different ionizing power (Table VIII, columns 6 and 7 ) , the 1 0 0 k d ( k a t k,) ratio is greater in the solvent of higher Y; 6-PhEtOTs (75 "C): acetic acid (53) and formic acid (91); ethanol (0.6) and 50% aqueous ethanol (8). Recently, we employed a treatment which quantitatively reproduced the observed 8 - P h E t O T s l E t O T s rate ratios by assuming only these effects were operative.Ib Incorporation of eq 5 and 6 for P-phenethyl tosylate and eq 10 for k , of ethyl tosylate into eq 9 yields eq 11 for calculation of the relative rate ratio in a variety of solvents.

236

Journal of the American Chemical Society

/ 1OO:l / January 4, 1978

Table IX. Plots of Log ( k a ) @-ArylalkylTosylates at 75 “C vs. Y o

Substrateb

Solvents 1,‘ EtOH

P-H

m-CH3 P-CH3 m,p-(CH3)2 Neophyl-OTs

I,‘ EtOH, CF3COOH I‘ I,‘ CF3COOH I,? EtOH I,‘ EtOH, CF3COOH I‘ I,‘ CF3COOH I, 80% aq EtOH, EtOH I, 80% aq EtOH, EtOH, CF3COOH

md

Correln coeff

0.63 f 0.01 0.67 f 0.03 0.66 f 0.05 0.75 f 0.06 0.56 f 0.04 0.65 f 0.06 0.64 f 0.04 0.75 f 0.07 0.59 f 0.02 0.67 f 0.04

0.999 0.997 0.998 0.993 0.996 0.989 0.998 0.991 0.999 0.993

Based on 2-adamantyl tosylate ~ c a l e . l ~ See ~ ~ Table * , I ~ VIII. Designation refers to 6-arylethyl tosylates except as noted. Group I solvents refer to 50% aqueous ethanol, acetic acid, and formic acid. Error limits indicate standard deviation of slope.

+ 0.82N

(corrected for internal return from the phenonium ion), respectively, for the tosylates. T o the extent that increased ionk,(P-PhEtOTs) - F ~ A O (ktO( 100.36Y+0.82N) izing power causes a specific enhancement of leaving group k,( EtOTs) activity which increases the kinetically significant magnitude of aryl bridging in the transition state leading to the phenonium intermediate, the effect proposed by Cramsc and Brown22 Agreement with experimental values (Table VIII, columns 4 could be observed. and 5 ) is remarkably good. W e find that this is of only minor importance. T h e /3-PhThis approach can be applied generally to neighboring group EtOTs/EtOTs rate ratio can be calculated accurately (eq 11; participation, although few mechanistic studies possess the Table VIII, columns 4 and 5 ) without assuming this effect. In wealth of solvent data amassed for @-arylsubstrates. Peterson addition, plots of log k A vs. Y over a wide solvent range a r e and Kamat’s data76 for the 6-octyn-2-yl tosylate (IV)/2-pentyl fairly linear for each activated (an,, hyl I0) 0-arylethyl subtosylate (V) system in acetic, formic, and trifluoroacetic acids strate in Table Ill except p - C H 3 0 R P ( e q6 for p - H , Table IX, when correlated by eq 12 and Figure 2). The value of m does increase (-14%) when data ~,0(100.14A’+0.61Y ) t kA0100,79y for trifluoroacetic acid are included, but the degree of corre-k,lV (12)77 lation decreases only slightly and the new value of m is genktO( 1 oO.I6N+0,70Y ktV ) erally within the combined error limits of m calculated with and without the trifluoroacetic acid point. Nevertheless, some gave the following agreement for observed (calculated) values curvature can be seen in Figure 2 plots which suggests that the of the relative ratio k,(IV)/k,(V): acetic 2.0 (2.2), formic 1.8 k A pathway might be responding modestly to changes in sol(2.8), and trifluoroacetic 16 (12). Raber et al. have successfully vent ionizing power, as proposed by CramSCand Brown.22 calculated percentages of cyclized product from 5-hexen- 1-yl p+ treatments provide a measure of transition-state positive p-nitrobenzenesulfonate in various solvents by this treatment .63c charge development in reactions affecting the ?r electrons of The relative importance of changes in solvent ionizing power the aromatic ring. Values of pneophyl (FkA,P-arylethyl tosylates) relative to the value for neophyl brosylates in acetic acid ( Y ) and nucleophilicity ( N ) on the value of k,(P-PhEtOTs)/ (Table X ) are remarkably constant (1.02 f 0.12, omitting the k,(EtOTs) can be determined from m and I coefficients (eq 1 I ) , the range of Y and N values being comparably large. T h e C F 3 C H 2 0 H result) with variation in solvent. The slope for anhydrous trifluoroethanol, as the Hammett p for deactivated relative rate ratio does increase with increased solvent ionizing k, substrates, seems high and might result from the lack of data power (ma 2m,); however, it is more sensitive to solvent (only H-, and p-CH3 a r e available),82 or experimental difnucleophilicity since the difference between I , (0.78,0.82) and ficulties in determining rate^.^^^,^^^,^^ Positive charge devel1~ (-0) coefficients is even greater. This result emphasizes the opment in the aryl ring is thus indicated to be quite similar for fact that neighboring aryl group and solvent are in competition FkA in solvents of greatly differing ionizing power. It is difficult to displace the leaving group. By decreasing the nucleophilicity to make detailed comments concerning the geometry of the k A of the latter (and the denominator of eq 1 l ) , the kinetic intransition-state complex from this pneophy] treatment. Medium fluence of the phenyl group becomes more noticeable, thereby effect^,^^,*^ such as hydrogen bonding to the aryl ringgs and increasing the magnitude of the relative ratio.21 solvation of positive charge,g6must be considered, and might Cramsc and Brown22have suggested that such relative ratios explain the increasing variation of Pneophyl (Table X ) in the might also be enhanced by “increases in the electrophilic more acidic, less nucleophilic solvents. In addition, indisparticipation of solvent which, in turn, results in increases in criminate use of such linear free-energy relationships to dethe participation of aryl in the transition states leading to the termine the “structure of the transition state” has legitimately bridged ion”.sc The fact that more reactive leaving groups tend to promote the intramolecular cyclization pathway favors this been criticized recently,g7 especially in conjunction with the interpretation. In acetic acid, slightly more rearrangement is Hammond postulate88 which has itself been questioned a t obtained for P-phenethyl brosylate than for the tosylate (2various t i m e s . 5 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 3%),799g0 and a slightly greater 0 - P h E t X I E t X rate ratio is Isotope effects, on the other hand, should provide a more obtained with nosylates than tosylates (buffered media: 0.49 detailed measure of the structural features of the transition vs. 0.46).71Bergman et al. observe more cyclized product from state92-94than d o Pneophy] (or p + ) values. The data in Table X I silver catalyzed acetolysis of 3,4-pentadien- 1-yl iodide than indicate a transition state for the k A pathway of primary &aryl for buffered acetolysis of the tosylate.81 “Super” leaving groups substrates in which the aryl ring has substantial electronic involvement with the reaction center but has migrated only dramatically increase this trend. 0-Phenethylmercury perchlorate solvolyzes 8.3 times faster than ethyl mercury perslightly toward it.95 For k A of 0-arylethyl substrates, a-Dz chlorate in acetic acid and 30 times faster in formic a ~ i d . ~ ~ , values ’ ~ (in trifluoroethanol and formic and trifluoroacetic acids) This is compared with the reduced ratios of 0.58 and 2.5 a r e intermediate in magnitudes (1.17-1.27) relative to the log ( k , / k t o ) = 0.36Y

ikSo(10°.33Y+0.78N)

-

Schadt, Lancelot, Schleyer

/

Effect of Solvent on @-ArylalkylSolvolysis

Table X. Plots of Log ( F ~ & ) s o 8-Arylethyl H Tosylates at 75 OC vs. u(neophy1) Solvent Re1 p a Correln coeff

EtOH 50% aq EtOH AcOH HC02HC CF3C02H' CF3CH2OHf

0.89 0.93 1.06 1.06 1.14 1.26

(1 .OOO) 0.990 0.995 0.999 0.999 (1.000)

237

No. of points

Points includedb

2 5 5 5 5 2

p-CH3 and p-CH3O All except p-CI All except p-CI All except p-CH30d All except p-CH30d H- andp-CH3d

a Value of p is relative to that of substituted neophyl brosylates in acetic acid at 75 OC, -3.96. b Substituents includep-Cl, H, m-CH3, p-CH3, m,p-(CH3)2, and p-CH3O. c Includes data from ref 22. d Deactivated by hydrogen bonding. See Kinetics section. e The slope changes somewhat to 1.19 (correln coeff 0.999) when data (corrected for leaving group) for a deactivated substrate such as m-Br are included (available data for p-NO2 were not used d).43f Reference 38.

-0.8

-1.6

-2.4

-3.2

-5.6

-6.L

-7.2

-8.0

Figure 2. Plots of log k~(P-arylethy1tosylates) vs. Y o T ~75 .

OC.

projected limiting value 1.49-1.56 (1.22-1.25 per D)96397and suggest a n activated complex in which the a carbon is somewhat less congested than is the solvent assisted Bonding to the primary a carbon by aryl is significant as evidenced by l-aryl-14C effects for @-p-anisylethylin all solvents, and for @-phenethylin all solvents except acetic acid ( F k d k t = 0.262,75 0C);9g small @-D2effects (0.97-1.00) indicate only minor transition state changes a t the @ carbon. For neophyl substrates, long consideied to possess a strong driving force for p a r t i ~ i p a t i o n , ~intermediate ~ ' ~ ~ ~ ~ ~ ~ ~a-D2 (1.21-1.25) and low P-l4C (1.014) isotope effects indicate a n early, unsymmetrical transition complex, while l-aryl-14C (1.023-1.035), a-D2, and a-14C effect^^^^^^ demonstrate strong perturbation a t both the origin and terminus of k a bonding. When the solvent is changed to t h e more ionizing, less nucleophilic trifluoroacetic acid, a-D2 isotope effects for @phenethyl substrates increase from 1.17 (formic acid, Fka sz 0.9kt) and 1.21 (trifluoroethanol, Fka FZ 0.9kt) to 1.27 (trifluoroacetic acid, Fka = kt), and for neophyl brosylate ( k a

k , ) from 1.21 (acetic acid) to 1.25 (trifluoroacetic acid) indicating less congestion a t the a-carbon. However, electronic involvement increases somewhat92 as shown by the increase in 1-aryl-l4C effects for /3-phenethyl and neophyl, and a-I4C effects for neophyl. This coupled with the increases observed in m and Pneophy] in going to the highly ionizing trifluoroacetic acid also indicates a small increase of aryl stabilization in t h e k a pathway with increasing the solvent ionizing power, as CramSCand Brown22have proposed. Role of Ion Pairs. All of the mechanistic criteria for primary &aryl systems indicate that solvent and neighboring aryl compete in rate determining displacement of the leaving group.3 W h a t is the nature of the species being attacked? Modern solvolysis theory indicates that ion pairs, but not dissociated carbenium ions, a r e intermediates in some of the solvolyses of simple secondary substrates. Whether or not these ion pairs or the transition states leading to them a r e nucleophilically solvated and the relative magnitude of internal return to neutral substrate remain as topics of controversy.20,26Sneen

238

Journal of the American Chemical Society

/

1OO:I

/

January 4, I978

Table XI. Summary of H/D and I2C/l4C Isotope Effects for Primary Alkyl Substrates

h / k D (temp, "c)

Substrate

Solvent

j3-@-Nitrophenyl)ethyIb.c

CH3COOH CF3COOH CF3COOH (buffered)d j3-(m-Bromophenyl)ethy16,c CH3COOH CF3COOH CF3COOH (buffered)d j3-Phenethyl CH3COOH HC02H CF3CH20H CF3COOH CF3COOH (buffered)d j3-@-Methoxyphenyl)ethyl CH3COOH HC02H CF3COOH Neophyl CH3COOH CF3COOH Ethyl H2 0 CH3OH CH3COOH CF3CH2OH CF3COOH 96% H2S04 HS03F

B

a

1.044 (100) 1.14 (110) 1.11 (110) 1.041 (100) 1.19 (70) 1.24 (70) 1.03 (93.9)ef 1.17 (75.3)ef 1.21 (75)e.g 1.27 (75.0)e,h

1.061 (100) 1.08 (110) 1.09 (110) 1.052 (100) 1.01 (70) 0.99 (70) 1.04 (93.9)ef 1.00 (75.3)e-f

1.18 (75)e,i 1.20 (50)e.j

1.00 (75)C.i 0.97 (50)e.J

1.21 (75)O 1.25.10)O 1.04 (54.3)e,' 1.04 (56.2)k,m 1.09 (100) k . m 1.13 (35)" 1.09 (1 25y.O 1.18 (30)e,p 1.30 (0)e,q

(temp, "C)

k12~/k14~

1-Arvl-Ca

a"

B"

1.093 (75) 1.141 (0)

1.014 (75) 1.014 (0)

1.002 (100)b 1.023 (60) 1.029 (45) 1.036 (30)b 1.028 (60)b 1.022 (30) 1.039 ( 0 ) b 1.023 (75) 1.035 (0)

1.01 (116.8)ksm 1.09 (35)" 1.16 (125)e*0 1.20 (30)e,p 1.58 ( 0 ) e * q

" References 98 and 99. b p-Nitrobenzenesulfonate leaving group. c Reference 43. d Buffered with CF3C02Na. e Tosylate leaving group. Reference 95a buffered with appropriate RC02Na. g Reference 38b,c. /I Reference 42. Actual values of 1.08 (0.4 M), 1.17 (0.02 M), 1.23 (0.008 M), and 1.21 (0.003 M) were extrapolated to "0" ROTSconcentration to obtain the tabulated value. Reference 95b, 0.06 M LiC104. Reference 95b, 0.055 M HC02Na. p-Bromobenzenesulfonate leaving group. K. T. Leffek, J. A.,Llewellyn, and R. E. Robertson, Can. J . Chem., 38, 1505 (1960). E. S. Lewis, J. C. Brown, and W . C. Herndon, ibid., 39,954 (1961). Triflate leaving group, CF3SO3. G. A. Dafforn and A. Streitwieser, Jr., Tetrahedron Lett., 3159 (1970). 0 I. Lazdins Reich, A. Diaz, and S. Winstein, J . Am. Chem. Soc., 91,5635 (1969). P P. C. Myhre and K. S. Brown, ibid., 91, 5639 (1969). 4 P. C. Myhre and E. Evans, ibid., 91,5641 (1969).

f J

has suggested "all reactions of nucleophilic substitution a t a saturated carbon atom proceed via an ion pair even methy1.27a,c,100 Scheme 11, an abbreviated version of the full ion-pair scheme,l0' is appropriate for simple primary systems where free ions and solvent separated ion pairs are not likely owing to their instability. Equation 13

provides the relevant kinetics expressions for Scheme 11. In order to explain the fact that simple primary substrates (eq 14, ~ behavior, no neighboring group) formally observe S Nkinetic Sneen postulated that k - l l k z m,27c in which case eq 14 simplifies to eq 15.

-

We disagree with this i n t e r p r e t a t i ~ n . ' Several ~ , ~ ~ lines of reasoning indicate that the species which is partitioned between the solvent and aryl assisted pathways does not possess a significantly different amount of positive charge than that on neutral substrate.26cThe very low p's (4to -0.34) observed for the k , route show this directly. Classical primary alkyl cations are highly energeticIo2 (-30 kcal less stable than typical tertiary ionsIo3), and there is considerable doubt if these species can exist at a11.Iw The ethyl cation is not observable directly in super acid the label scrambling found is probably due to intervention of intimate ion-pair intermediates.Io5Such species, or even nucleophilically solvated ion pairs, are very unlikely energetically in the less ionizing, more nucleophilic solvolysis solvents. Sneen27 proposes that primary ion-pair formation must

occur many times before capture by solvent, k - I >> k2. T h e presence of a neighboring group allows the determination of a lower limit for this inequality. According to the ion-pair interpretation, Scheme 11, any rate enhancement ( k l / k , ) produced by introduction of a neighboring group generally derives not from enhancement of k 1 but from a decrease in internal The maximum observable rate ( k J , return (k-1).23,26c,27,28 when all ion pairs are captured by neighboring-group participation, is equal to kl (derived from eq 13 when Fk, >> k-l >> k2). Under these circumstances, the maximum rate enhancement is given by eq 16.

A lower limit for ( k t / k s ) m ain x the /3-arylethyl system is provided by the =IO6 difference in ethanolysis rates (25 "C) of the conjugate base of P-(p-bydroxypheny1)ethyl bromide (VI), ka, and fl-(p-methoxyphenyl)ethyl bromide (VII), ks.106a Correcting by a factor of -10 for more favorable ion-pair formation in VI than VI1 ( ~ ~ - 0 U,,-CH~O = -0.25; p(primary ion pair) = -4, p(solvo1ysis) = -0.93 in 80% aqueous E t O H for tertiary 1-aryl-2-methyl-2-propyl chlorides3), ( k t / k s ) m a x = k-l/k2 0 lo5 is obtained. Charged aryl groups should not be excluded from Scheme 11 in view of the ubiquity of ion-pair intermediates proposed by advocates of this m e c h a n i ~ m ,extending ~ ~ ~ ~ ~even ~ ~ to~ the ~ displacement of methyl derivatives by charged nucleophiles in aqueous media." Substitution of the ratio ( k - l / k 2 = lo5)into eq 15 for the aryl unassisted pathway of fl-(p-anisyl)ethyl bromide yields kl 105kt.Since kt = 2 X kl is-2 X (ethanol, 25 "C),45 times greater than the rate of ionization of tert-butyl bromide (4.40 X 10-6)106b under the same conditions. Formation of a primary alkyl carbenium ion (kl') should be 2.5 X slower than for a tertiary ion assuming that the former

-

Schadt, Lancelot, Schleyer

/

239

Effect of Solvent o n @-ArylalkylSolvolysis

Table XII. Solvent Effects on Arvl Assisted and Solvent Assisted Pathwavs k,( 1-Ph-2-Prop)/kt(2-Prop)

kt X lo6, S-’ (50 “C)

Solvent EtOH 80% aq EtOH AcOH HCO2H CF3C02H

CH~CH(OTS)CH~

PhCH*CH(OTs)CH3

8.16b 54.4c 2.12d 405 e 303f

1.41 9.42 0.636 304 511Of.g

Exptl

Calcd (eq 8)

0.17 0.17 0.30 0.75 16.9

0.21 0.16 0.24 2.97 17.6

N O

Y O

0.00 0.00 -2.35 -2.35 -5.56

-1.75 0.00

-0.61 3.04 4.51

a Y and N are measures of solvent ionizing power and nucleophilicity,respectively, in the general Winstein-Grunwald e q ~ a t i o n . ~They ~.~* were determined19using 2-adamantyl tosylate, a limiting ~ubstrate,~’ as a standard to measure solvent ionizing p ~ w e r . ~ ~ R. . l E. * Robertson, ~~~ Can. J . Chem., 31, 589 (1953). Reference 20. Reference 13c. e Calculated using the extrapolated rate constant of isopropyl brosylate at 50 OC2’and the (OB~/OTS)~.,,,,~~” rate ratio of 2.57 in formic acid (ref 21 and 139). f Reference 74. g The unbuffered results of W i n ~ t e i n ~ ~ were used in eq 8 .

species, which is some 30 kcal/mol less stable than a tertiary cation in the gas phase,lo3 is a t least 20 kcal/mol less stable in solution (AG = -(1.99) (298) (2.303) log kl’/ktert = 2 X lo4). The >10l6 increase of the “observed” kl’/ktert ratio compared with the calculated value strongly implies that primary “ionpair formation” involves very little, if any, positive charge. There is no compelling evidence for the existence of ion-pair intermediates in the solvolyses of primary alkyl substrate^.'^,^^ Consideration of such stretched-bond for primary systems only introduces needless mechanistic complications. lo7 W e conclude that P-arylethyl solvolysis is best interpreted in terms of concerted inter- and intramolecular displacement processes.

Secondary 1-Phenyl-2-propyl Tosylate. Effect of Solvent. The secondary 1-phenyl-2-propyl/2-propyltosylate rate ratio (eq 1, R = CH3), as the primary system (eq 1, R = H), varies substantially on going from ethanol to trifluoroacetic acid (100-fold). T h e relative rate ratio can be calculated from measures of solvent ionizing power (Y) and solvent nucleophilicity ( N ) with eq 18, obtained by inserting eq 7 and 8, and eq 17 for 2-propyl tosylate into eq 1 (R = CH3). log (k/kto) = 0.60Y

+ 0.47N

(1 ?)lop

k d 1- P h - 2 - P r o ~ - O T s ) k t (2-Prop-OTs) FkAO(100.82Y) + ksO( 100.50Y+0.46N) (18) ktO( 100.60Y+0.47N

1

Agreement of calculated and observed values (Table X I I , columns 4 and 5) is quite good-within a factor of 2 except for formic acid which is within a factor of 4. For this secondary system, the sensitivities of the k, pathway to solvent ionizing power and nucleophilicity a r e nearly balanced (2-Prop-OTs, m, = 0.60, ls = 0.47; l-Ph-2-Prop-OTs, m, = 0.50, I , = 0.46). Changes in solvent ionizing power (Y) favor intramolecular displacement (kA) since mA 1.3- 1.5m,; however, solvent nucleophilicity is slightly more important in determining the overall relative rate ratio (eq 18) since the difference between I , and I A (0.46,0.47) is greater than that between m a and m, (0.32, 0.22). W e found the effect suggested by Cram5c and Brown22that increased ionizing power increases the degree of aryl participation in transition states (TSA) leading to the bridged ion-to be of only minor importance in this secondary system, as was found for the primary P-arylethyl system. Firstly, quite good agreement of calculated and observed relative ratios is obtained without this assumption (Table X I I ) . In addition, plots of log kA vs. Y a r e linear from ethanolic solvents to formic acid for the p - H , p-CH3, and p-CH3O substituted l-aryl-2propyl substrates (Table XIII). Addition of trifluoroacetic acid d a t a for the p-H compound does increase the m value -19%, similar to the average 14% increase observed for the P-arylethyl substrates, as predicted by the proposed effect. However, as

-

Table XIII. Plots of Log ( k A ) 1-Aryl-2-propylTosylates at 50 OCVS. Y O

Substituent P-H

Solvents

mb

Correln coeff

I‘ I,‘ CF3C02H“ IC

0.69 f 0.06 0.82 f 0.07 0.67 f 0.05 0.63 f 0.04

0.99 1 0.988 0.997 0.996

p-CH3 D - C H ~ O IC

Based on the 2-adamantyl tosylate ~ c a l e . l ~ ~See ~ *Table . ’ ~ XI1 and Results. Error limits indicate standard deviation of slope. Group I solvents refer to EtOH, 80% aqueous EtOH, AcOH, and HC02H. Unbuffered data from ref 35. (I

Table XIV. Plots of Log ( F k A ) s 1-Aryl-2-propyl ~~ Tosylates at 50 “ C vs. u(neophy1)

Solvent EtOH 80% aq EtOH AcOH HC02H

Re1 p a Correln coeff 0.93 0.78 0.80 0.85

(1.000)

0.97’

0.998

0.999 0.999 0.995

Points included p-H, p-CH30 p-H, p-CH3, p-CH30 p-H, p-CH3, p-CH30 p-CI, p-H, p-CH3, p CH30 p-CI, p-H, p-CH3

a Value of p is relative to that of substituted neophyl brosylates in acetic acid at 75 OC, -3.96.1b336 Substituents such as C H 3 0 might be deactivated by hydrogen bonding in certain acidic solvents. See Kinetics.

with the primary system, the degree of correlation is not decreased significantly and the m values calculated with and without the trifluoroacetic acid point are within the combined error limits. Finally Pneophyl values (Table XIV), measuring positive charge delocalization to the aryl ring, do not increase substantially (0.87 f 0.10) in going from ethanol to formic acid. These combined results indicate that the kinetically significant magnitude of aryl stabilization in the k A transition state (TSJ is relatively constant with variation in solvent. Although the wealth of isotope effect data for P-arylethyl derivatives does not exist for the 1-phenyl-2-propyl results for the related threo-3-phenyl-2-butyl system are in a range of solvents (Table XV). In formic and trifluoroacetic acids, where the reaction proceeds predominantly by the k~ process, intermediate a - D (1.142, 1.133), low P-PhCH(D) (1.040, 1.009), and higher @-methyl(1.05 I , 1.054) than y-methyl (1.005, 1.003) isotope effects provide evidence for a n unsymmetrically bridged transition state, similar to the situation for primary P-arylethyl substrate^.^^-^^ In acetic acid and 50% aqueous ethanol, the aryl unassisted pathway is a significant component of the total rate constant: 32% ( A c O H , iterative Hammett treatment, Table VII) and 22% (50% aqueous ethanol, determined from products30), and observed isotope effects should be a composite. a - D effects

240

J o u r n a l of the American Chemical Society

/

1OO:l

/

J a n u a r y 4, 1978

Table XV. Summary of H/D Isotope Effects for Secondary Alkyl Substrates Substrate

Solvent (temp, "C)

2-Propyl brosylate" (25 "C)

50% aq EtOHb CH3C02H CF3C02H 98% aq CF3C02Hd 50% aq EtOH (25)e CH3C02H (75.1)f HCOzH (25.02)f CF3C02H (-7.9)f

1-Phenyl-2-propyltosylate (25 "C) threo-3-Phenyl-2-butyl aryl sulfonates

CY

0-Internal /3- Terminal (PerD) (Per D)

1.114

yperD

k,.o/k,.H

1.059

1.14c 1.22

1.13 1.135 1.104 1.142 1.133

0.986

1.144

1.094 1.040 1.009

1.024 1.051 1.054

1.021

1.005 1.003

Tabulated in ref 23a. Volume percent. Extrapolated from value at 70 "C (1.12, K. Mislow, S. BorEic, and V. Prelog, Helu. Chim. Acta, 40, 2477 (1957)) using a 0.01 increase in magnitude for every 20 "C decrease in temperat~re."~ Reference 91. e Reference 30, brosylate leaving group. f References 109, 110, polarimetric rate constants, tosylate leaving group. (1.104and 1.135) areintheexpectedrangeforeither k,or kA mechanisms in these solvent^,^^^,^^ while the aryl p - D effect (50% aqueous ethanol) evidences the major anchimeric pathway; enhanced P-PhCH(D)15 (1.094, 1.144), and similar &methyl (AcOH, 1.024) and y-methyl (AcOH, 1.02 1) isotope effects indicate the presence of elimination and hydride shift which are included in the aryl unassisted p a t h ~ a y . ~ ~The l ~ small ~ , * ~change in a - D in going from formic to trifluoroacetic acid does not indicate a large change in the degree of participation by the phenyl group in transition states leading to the bridged phenonium ion. Role of Ion Pairs. The observation of excellent rate-product correlations for acyclic secondary P-aryl system' 3d,14 confirmed that the aryl unassisted ( k , ) and aryl assisted ( k A ) mechanisms are discrete processes which do not "crossover". Attack by solvent or neighboring group on neutral substrate was considered to be rate l i m i t i ~ ~ g ,although ~ ? ' ~ ~ it was acknowledged that formation of intimate ion pairs which were partitioned between k , and FkA would be kinetically consistent with the rate-product results.3 Several versions of the ion-pair scheme have been applied to secondary P-arylalkyl solvolysis, and these are evaluated below. Ramsey and Das29have correlated rate data for primary and secondary P-arylalkyl substrates with ionization potentials of substituted benzenes, values which are related to the chargetransfer transitions in the ultraviolet spectra of aryl-substituted tribenzylboranes, VIII. The fact that better linear plots were

Scheme 111 Products

Phenonium

PrOdUclS

Products

R H C C "

Y

x

X

bb--B.--R

hr

\R

\

,

\ '.. -B L

~

*

R

.HC

\R

VI11

obtained for Fka than k t , of 0-arylethyl tosylates in acetic acid (correln coeff 0.999, 0.932, respectively), was suggested to provide additional confirmation that these primary substrates solvolyze by two totally independent pathways, k , and kA. However, very good linear correlations were obtained for the secondary systems 1-aryl-2-propyl and threo-3-aryl-2-butyl using total rate constants, k t (correln coeff 0.99), even though aryl-deactivated derivatives yield almost no products from the aryl assisted pathway. It was interpreted that secondary P-aryl substrates undergo rate-limiting ionization ( k l ) to an unsymmetrically a-bridged intimate ion pair, which is captured by solvent (k2) or neighboring group ( k p ) much faster than return to substrate ( k 2 and k , > k-I), Scheme IIIA. W e disagree with the latter conclusion. Firstly, although the spectral method provides an interesting treatment of P-aryl solvolysis, the ultraviolet and ionization potential data are not as precise or readily available as the preferred HammettI3-I5 and neo-

phy11b,32,35 constants for correlations of k , and FkA, respectively. Furthermore, since k , and FkA processes have the same rate-determining step in this scheme, a satisfactory explanation for the rate-product stereochemistry correlations determined for threo-3-aryl-2-butyl brosylates (AcOH) cannot be provided. Two other ion-pair schemes, illustrated by the 3-aryl-2-butyl system, are consistent with the implications of rate-product correlations for secondary P-aryl solvolyses. Brown and employing the basic formulation of Shiner23 and sneer^,^^,^^ propose a steady-state scheme involving formation of an intimate ion pair from neutral substrate, followed by rate-limiting attack by aryl or solvent ( k - l / k z (or k p ) 2 20, Scheme IIIB). Cramer and Jewett30 interpret their isotope effect data (Table XV) as indicating two distinct ion-pair processes (Scheme 1IIC)-rate-determining formation of a

Schadt, Lancelot, Schleyer

/ Effect of Solvent on P-Arylalkyl Solvolysis

24 1

Table XVI. Physical Constants for @-ArylethylAlcohols”

Substituent Bp, “C (mm) Obsd Lit.

m,m’-(CF3)2

m-C1

m-F

p-c1

m-CH3

p-CH3

54-56‘

79 (0.3)

90.3 (2.6)

135 (13) 110 (0.5)?

70 (0.5) 87-88 (2.5)d

82-82.5 (0.6) 90-94 ( 3 ) d

Substituent Bp, “C (mm) Obsd Lit.

m,p-(CH3)2

p-CH3O

P-CsHs

142.5-145 (15) 109-110 (3)g

127-128 (2.8) 139-141 (14)h

96-97.5 94.5-95.56,‘

’

2-Naphthyle

2-Fluorenylf

68.8-70.2’ 66-67

131.1-133.1’ 136- 137b J

Thep-CF3, m-CF3, andp-CsHSO alcohols were used to prepare the tosylates (Table XVII) directly. @-Phenethylalcohol was purchased from Aldrich. Melting point, “C. G. Baddeley and G. M. Bennett, J . Chem. Soc., 1819 (1935). d Reference 22. e 0-(2-Naphthyl)ethyl alcohol. fP-(2-Fluorenyl)ethyI alcohol. g G. W. Pope and M. T. Bogert, J . Org. Chem., 2, 276 (1937). Reference 10. ’ Reference 48. JE. Profft and K . Steinhaus, J . Prakt. Chem., 22,47 (1963). phenyl bridged species ( kA) and rate-determining elimination stereochemistry, but that solvolysis occurs more slowly than the “model”, 2-butyl, posed the “dilemma”11bwhich required from a n open pair (k2).lI1 It is difficult to experimentally distinguish these alternatives almost 20 years of intense effort to solve. T h e resultant description of simple secondary solvolysis as proceeding by from the original Cram-Winstein description of direct aryl and substantial nucleophilic solvent participation provides a sigsolvent attack on neutral substrate. Such tight ion pairs having nificant advance in solvolysis theory, and recent years have seen a slightly stretched or weakened bond27ato the leaving group a r e not very different conceptually from covalent starting extensive efforts to refine our understanding of the implications,3,14,23,26-28 material. In either formulation, concerted displacement or ion pair, substantial bonding of solvent or neighboring aryl must Experimental Section occur in the rate-determining step to prevent crossover of the General. All boiling points are uncorrected. Melting points, which k , or k a pathways. are uncorrected, were determined using a Mettler FPI apparatus. We agree that ion pairs can be formed in simple secondary NMR spectra were taken on Varian A60-A spectrometers using tetsolvolysiszo but suggest that the usual formulations a r e defiramethylsilane as an internal standard. Infrared spectra were detercient in specifically excluding nucleophilic solvation.19~20~26c~93 mined using a Perkin-Elmer 237B grating spectrometer. Ultraviolet S u c h a “quantized” approach soon becomes kinetically comspectra were recorded on a Cary 14 spectrometer. Microanalyses and plex. In describing 2-propyl solvolysis, Shiner comments “one Karl Fischer determinations were performed by Hoffmann-La Roche, of the classical examples of borderline solvolyses seems to be Inc., Nutley, N.J. borderline in a bewildering number of ways! As many as four Materials. fl-Arylethyl Derivatives. Usually the corresponding ardifferent steps, . . . , can be made the dominant rate controlling ylacetic acids, purchased from Pierce @-CF3, m-CF3, m-F) or Aldrich influence depending on the choice of Analysis of (m-CI, p-CI, m-CH3, p-CH3,p-CH30,p-C& (as the nitrile), and the vast amount of d a t a a t hand seems to suggest a gradual @-naphthyl),were reduced to the alcohols by lithium aluminum hyspectrum of nucleophilic a t t a ~ h m e n t , ~ - l ~l 2-ranging ~ ~ , ~ ~from ~ * * dride. The following procedure for @-(m-chloropheny1)ethanol illustrates the method. A solution of (10 g, 0.059 mol) of @-(m-chlolimiting secondary (2-adamantyl17) to strongly solvent assisted ropheny1)acetic acid in ether was added slowly to a suspension of primary (methyl) s o l v o l y ~ i s . In ~ ~such - ~ ~a n~ interpretation, ~~ lithium aluminum hydride (4.85 g, 0.128 mol) in ether and refluxed unhindered secondary ion pairs, if formed, are stabilized from overnight. Excess lithium aluminum hydride was decomposed by the the rear by neighboring group or solvent. normal basic procedure.llsaThe ether solution was dried and evapok , and k~ Pathways. Perspectives and Conclusions. O u r rated to yield 7.30 g (79%) of the alcohol. Distillation gave the pure studies of the @-arylethyl’,13eand l - a r y l - 2 - p r 0 p y l ~ systems ~ ~ - ~ ~ ~ product. confirm by all criteria that the solvolyses of primary and secThe m,m’-di-CF, and p-C6H50 alcohols were produced by reaction ondary @-aryl substrates a r e grossly similar, proceeding by of the Grignard reagent of the appropriate bromobenzene (Pierce and Aldrich, respectively) with ethylene oxide. The m,p-di-CHj derivative discrete k , and k A pathways3 0-Arylethyl tosylates react via was prepared from ethylene oxide and the Grignard prepared from competitive displacement by nucleophiles, solvent or aryl. m,p-dimethylbromobenzene (bp 104-105 “C (16 mm), lit.Il6 bp Secondary substrates, such as l - a r y l - 2 - p r 0 p y l , ~ ~ threo~-~,~ 206-208 “C), obtained by diazotization of m,p-dimethylaniline 3 - p h e n y l - 2 - b ~ t y 1 , ’and ~ trans- and cis-P-arylcyclopenty1l5 (Aldrich, melting point from hexane 50.5-51.5 “C, lit.’I7 mp give, in some cases, significant amounts of rearranged tertiary 47.3-49.2 ”). P-(2-Fluorenyl)ethyl alcohol was produced by lithium derivatives and e l i m i n a t i ~ n . ’ ~ ~These . ’ ~ , ~products ~ can also aluminum hydride reduction of the arylacetic acid, obtained by a arise from strongly bound pathways, hydride shift, or elimiWillgerodt sequence from 2-acetylfluorene’ (Aldrich). nation from a nucleophilically solvated ion pair.20 Since these The corresponding tosylates were prepared by the normal pyridine processes d o not appear to introduce kinetic complexities (Le., method.IlsbFor some derivatives it was necessary to add chloroform to the recrystallization solvent in order to effect solution at room in the k , line of H a m m e t t plots), they a r e operationally intemperature. cluded in the k , term.3,13d%14 This rate constant, therefore, Physical constants and analytical data are reported in Tables XVI should be designated more appropriately “aryl unassisted” for and XVII. IR and NMR spectra for all compounds were consistent secondary systems. l I 3 , l 14 with the structures proposed. T h e prolonged uncertainty about the mechanism of 0-ar1-Aryl-2-propano1s.l l9Commercial 1-phenylacetone (Aldrich) was ylalkyl solvolysis arose because of a lack of definition conreduced with lithium aluminum hydride to the alcohol, bp 96-97 “C cerning the “borderline” position of the solvolysis of secondary (10 mm), lit.120bp 95 “ C (10.5 mm). The l-(p-CH3-, p-CI-, m-CI-, systems in the SNl-SN2 spectrum of Hughes and Ingold. and m-CF3-) phenyl-2-propanols were prepared’ l 9 by the base-catCram’s observations that the neighboring phenyl group in the alyzed condensation of the corresponding benzyl cyanidel2l with ethyl 3-phenyl-2-butyl system predominantly controls product acetate, followed by hydrolysis, decarboxylation,I2I and reduction

242

Journal of the American Chemical Society

/ 1OO:l / January 4, 1978

Table XVII. Physical Constants and Analytical Data for @-ArylethylTosylates

Anal. Mp, "Ca Substituent m,m'-(CFh p-NO2 p-CF3 m-CF3 m-CI m-F p-CI H

Obsd 72.1-73.1 131.8-133.1 87.5-88.7

Calcd, % Lit.

Found, %

C

H

C

H

49.52

3.42

49.64

3.27

55.81

4.39

56.12

4.40

131.5-1 32.5'

C

46.5-47.6 37-38.2 78.5-79.7 38.8-40.3

m-CH3 P-CH3 m,p-(CHdz p-CH30 P-C6H5 P-CsHsO 2-Naphthylg

C

2-Fluorenyll

103.4-104.6

68.4-69.3 40.6-42.0 57.8-58.8 94.1-95.3 59.7-61.2 63.5-65

35.5-36.6d 38.8-39.8e 39-40f 69-70f

C

C

57.97 61.21 57.97

4.86 5.14 4.86

58.21 61.37 57.99

5.16 5.20 4.89

66.18

6.25

66.32

6.38

67.08

6.62

67.10

6.52

68.46

5.47

68.34

5.43

72.50

5.53

72.34

5.49

57-58d 94.5-95.5g.h 63-64h 75-76'

Uncorrected. Reference 5c. The tosylate was a solid which melted below room temperature, even after four recrystallizations at -78 "C. Acetolysis infinity titers were within 4% of the theoretical value. Reference 10. e Reference 72. f Reference 22. g @-(2-Naphthyl)ethyl tosylate. Reference 48. I C. C. Lee and A. G . Forman, Can.J . Chem., 44, 841 (1966). 1 @-(2-Fluorenyl)ethyltosylate. with sodium borohydride:'22p-CH3, bp 99-100 "C (3 mm), lit. bp 66-67 0C123(0.4 mm), 97 0C124(2 mm);p-CI, bp 94-95 "C (1.5 mm), lit.123bp 94-95 "C (0.1 mm); m-CI, bp 109-1 10 "C ( 5 mm); m-CF3, bp 66-67 "C (1.5 mm). l-(p-Nitr~phenyl)acetone~~~~~~~ was reduced with sodium borohydride122to the alcohol, mp 70-70.6 "C, lit.122mp 68-69 "C. I-(p-Ani~yl)acetone~~~ was prepared by the same procedurel2*,1*6 and reduced with sodium borohydride122to the alcohol: bp 114-1 15 "C (4 mm), lit. bp 158-161 0C123(15 mm), 121 0C128(3 mm). l-(p-Trifluoromethylphenyl)-2-propanolwas prepared from the corresponding arylacetic acid by methylation with methyllithium followed by reduction, bp 73 "C (3 mm). 1,3-Di-p-nitr0phenyl-2-propanol.'~~ 1,3-Diphenyl-2-propyI acetate was nitrated;'29 hydrolysis gave the dinitro alcohol, mp 148.4-149.8 "C. Anal. Calcd for C15H1405N2:C, 59.60; H, 4.64; N, 9.27. Found: C, 59.71; H, 4.56; N, 9.21. I-p-Anisyl-3-p-nitr0phenyl-2-propanol.~~~ I-p-Anisyl-3-p-nitrophenyl-2-propanone was prepared according to House and Berkowitzi30and reduced to the alcohol with sodium borohydride, mp 83.5-85 "C. Anal. Calcd for C16H1704N: C, 66.87; H , 5.97; N, 4.88. Found: C,66.14;H, 5.50;N,4.71. l-Phenyl-3-p-nitrophenyl-2-propanol.' l9 1-Phenyl-3-p-nitrophen y l - l - p r ~ p e n e ' was ~ ~ converted to l-phenyl-3-p-nitrophenyl-2pr0panone.l3~Reduction with sodium borohydride122gave l-phenyl-3-p-nitrophenyl-2-propanol, mp 109.6-1 11 "C. l-Phenyl-3-p-anisyl-2-propanoI.' l 9 The precursor, I-phenyl-3p-ani~yl-2-propanone'~~ was reduced with sodium borohydride to the alcohol, mp 53.4-54.6 "C. Anal. Calcd for C16H1802:c , 79.29; H, 7.49. Found: C, 79.06; H, 6.86. 1,3-Di-p-ani~yl-2-propanol.'~~ The alcohol, mp 81 -85 "C, was prepared by lithium aluminum hydride reduction of the diary1 acetone, 132.1 3 3 1-Aryl-2-propyl and 1,3-Diaryl-2-propyl Tosylates. Tosylates of the alcohols described above were preparedlI9 in the usual manner.115b Spectra were consistent with the structures proposed. Physical constants and analytical data are summarized in Table XVIII. Kinetic Procedures. Anhydrous acetic acid was distilled from a small amount of acetic anhydride.134Absolute ethanol was purified by the method of Lund and B j e r r ~ m .Aqueous '~~ ethanol (50%v/v) was prepared by mixing solvolytic grade ethanol with an equal volume of distilled water. Formic acid was stored over barium oxide for 2 weeks, decanted onto fresh barium oxide, and distilled under reduced pressure, bp 29.6-30 "C (46 mm).2iKarl Fischer analyses indicated 40%64,72 are difficult to obtain, and are especially sensitive to medium composition. Comparison with the limited available literature (Table XX) indicates relatively good agreement, considering the variations in the three analytical methods among four different research groups. 1-Aryl-2-propyl and 1,3-Diaryl-2-propyl Tosylates. The acetolysis rate constants reported in ref 13a-c and here were detemined for the most part by the usual ampule technique,134titrating the liberated p-toluenesulfonic acid with a glacial acetic acid solution of sodium acetate. Some rate constants were also determined by a conductometric method,119which is shown to agree well with the titrimetric procedure for 1-@-tolyl)-2-propyl tosylate. Formolysis rate constants were determined by conductance using special glass cells (25-mL volume) with bright platinum electrodes, and a recording Wheatstone bridge.119Some formolysis rates were checked potentiometrically, using the method of Winstein and Marshall.*' A Heath Model EU-20-1 l recording pH meter was used for this titration. The formolysis solutions were diluted with 50 mL of glacial acetic acid and titrated to the potentiometric inflection point with an acetic acid solution of sodium acetate. The formic acid solvent was purified by storage for at least 3 days at room temperature over solid boric anhydride, followed by fractional distillation under vacuum, bp 30-33 "C (50 ").*I Acknowledgment. We are grateful to the National Science

/ 1OO:l / January 4 , 1978

Foundation, the donors of the Petroleum Research Fund, administered by the American Chemical Society, the National Institutes of Health for a fellowship to F.L.S. and Grant GM-19134, and Hoffmann-La Roche, Nutley, N.J., for their support of this work. We wish to thank G. A. Olah, D. FgrcaSiu, and T. W. Bentley for valuable discussions and W. F. Sliwinski for generous assistance in writing the computer programs used in this work. References and Notes Preliminary accounts of this work have been presented: (a) J. M. Harris, F. L. Schadt, P. v. R. Schleyer, and C. J. Lancelot, 7th Annual Middle Atlantic Regional Meeting of the American Chemical Society, Philadelphia, Pa.. Feb 15. 1972;Ib) F. L. Schadt and P. v. R. Schlever. J. Am. Chem. Soc., 95, 7860 (1973). (a) National Institutes of Health Predoctoral Fellow, 1989-1974.(b) FMC Corporation, Princeton, N.J. (c) Address correspondence to the lnstitut fur Organische Chemie, Universitat Erlangen-Nurnberg, Henkestr. 42, 8520 Erlangen, West Germany. For a review, see C. J. Lancelot, D. J. Cram, and P. v. R. Schleyer in "Carbonium Ions", Vol. Ill, G. A. Olah and P. v. R. Schleyer, Ed., Interscience, New York, N.Y., 1972,Chapter 27,p 1347. (a)D. J. Cram, J. Am. Chem. SOC.,71,3863(1949);(b) ibid., 74,2129,

2137 (1952). (a) D. J. Cram, J. Am. Chem. SOC.,86,3767 (1964);(b) D. J. Cram and J. A. Thompson, ibid., 69,6766 (1967);(c) ibid., 91, 1778 (1969). (a) This term designates solvolyses which proceed without direct resonance stabilization or enhanced carbon-carbon hyperconjugative or neighboring-group participation. (b) See ref 17, 19,20 and 26. S.Winstein, B. Morse, E. Grunwald, K. Schreiber. and J. Corse, J. Am. Chem. Soc, 74, 1113(1952). Reference 14a,footnote 11. S.Winstein and K. C. Schreiber, J. Am. Chem. SOC.,74, 2165 (1952). S.Winstein, C. R. Lindegren, H. Marshall, and L. L. Ingraham, J. Am. Chem. SOC.,75, 147 (1953). (a) H. C. Brown, Chem. Soc., Spec. Pub/.. No. 18 (1962)."The Transition State"; (b) H. C. Brown, K. J. Morgan, and F. J. Chloupek, J. Am. Chem. SOC., 87, 2137 (1965);(c) C. J. Kim and H. C Brown, ibid., 90, 2082 IlPfiAl ~

H. C. Brown and C. J. Kim, J. Am. Chem. Soc.,91,4286 (1969);(b) ibid., 91,4287 (1969);(c) bid., 91,4289 (1969). la) C.J. Lancelot and P.v. R. Schiever. J. Am. Chem. Soc.. 91. 4291 i1969);(biibid., 91,4296 (1969);(cfC.J. Lancelot, J. J. Harper,and P. v. R. Schleyer, ibid., 91,4294(1969);(d) P. v. R. Schleyer and C. J. Lancelot, ibid., 91,4297(1969);(e)J. M. Harris, F. L. Schadt, C. J. Lancelot. and P. v. R. Schleyer, ibid., 91,7508 (1969);(f) D. J. Raber, J. M. Harris, and P. v. R. Schleyer, ibid., 93,4829 (1971). (a) H. C. Brown, C. J. Kim, C. J. Lancelot. and P. v. R. Schleyer, J. Am. Chem. SOC.,92, 5244 (1970);(b) H. C. Brown and C. J. Kim, ibid., 93,

5765 (1971).

H. C. Brown and C. J. Kim, J. Am. Chem. SOC., 94, 5043, 5051 (1972).

f is the fraction of intimate phenonium ion pair which does not return. (a) J. L. Fry, C. J. Lancelot, L. K. M. Lam, J. M. Harris, R. C. Bingham, D. J. Raber, R. E. Hall, and P. v. R. Schleyer, J. Am. Chem. SOC.,92,2538 (1970);(b) J.L. Fry. J. M. Harris, R. C. Bin@", and P. v. R. Schleyer, ibid., 92,2540 (1970);(c) P. v. R. Schleyer, J. L. Fry, L. K. M. Lam, and C. J. Lancelot. ibid., 92,2542 (1970);(d) S.H. Liggero, J. J. Harper, P. v. R. Schleyer, A. P. Krapcho, and D. E. Horn, ibid., 92,3789(1970);(e)J. M. Hawis, D. J. Raber, R. E. Hall, and P. v. R. Schleyer, ibid., 92,5729 (1970); (f) J. M. Harris, R. E. Hall, and P. v. R. Schleyer, ibid., 93,2551 (1971); (g) D. J. Raber, J. M. Harris, R. E. Hall, and P. v. R. Schleyer, ibid., 93,4821 (1971). F. L. Schadt, P. v, R. Schleyer, and T. W. Bentley, Tetrahedron Lett., 2335 I,.-. 1 9741. .,-

T. W. Bentley, F. L. Schadt, and P. v. R. Schleyer, J. Am. Chem. SO~.,94, 992 11972k ibid.. 98.7667 119761. T. W.Bentley and P. v. R.' Schleyer, J. Am. Chem. SOC., 98, 7658

(1976).

S.Winstein and H. Marshall, J. Am. Chem. SOC.,74, 1120 (1952). H. C. Brown, R. Bernheimer. C. J. Kim, and S. E. Scheppele, J. Am. Chem. SOC.,89, 370 (1967). (a) V. J. Shiner, Jr.. in "Isotope Effects in Chemical Reactions", C. J. Collins and N. S. Bowman, Ed., Van Nostrand-Reinhold, New York, N.Y., 1970,pp 90-159;(b) V. J. Shiner, Jr.. and W. Dowd, J. Am. Chem. SOC., 91,6528(1969);(c) V. J. Shiner, Jr., R. D. Fisher, and W. Dowd, ibid., 91, 7748 (1969). (a) W. M. Schuberl and P.H. LeFevre, J. Am. Chem. Soc.,91,7746(1969); (b) W. M. Schubert and W. L. Henson, ibid., 93,6299 (1971);(c) W. M. Schubert and P. H. LeFevre, ibid., 94, 1639 (1972). A. F. Diaz, I. Lazdins, and S. Winstein, J. Am. Chem. SOC.,90, 1904

(1968). For recent critical reviews of the mechanism of solvolysis and ion pair formation, see (a) D. J. Raber and J. M. Harris, J. Chem. Educ., 49,60 (1972);(b) J. M. Harris, Prog. Phys. Org. Chem., 11, 89-173 (1974);(c) D. J. Raber, J. M. Harris, and P. v. R. Schleyer in "Ions and Ion Pairs in Organic Reactions", Vol. 2,M. Szwarc, Ed., Wiley, New Yo&, N.Y., 1974, p 247;(d) D. J. McLennan, Acc. Chem. Res., 9, 281 (1976);(e) T. W. Bentley and P. v. R. Schleyer. Adv. Phys. Org. Chem., 14, 1 (1977);(1) P. v. R. Schleyer in "Reaction Transition States", J. E. Dubois, Ed.. Gordon and Breach, New York, N.Y., 1972,p 197;see also, ref 20 and T. W.

Schadt, Lancelot, Schleyer

Effect of Solvent on @-ArylalkylSolcolysis

Bentley. S.H. Liggero, M. A. Imhoff, and P. v. R. Schieyer, J. Am. Chem. SOC., 96, 1970 (1974). (27) (a) R. A. Sneen, Acc. Chem. Res., 6,46 (1973); (b) R. A. Sneen and J. W. Larsen. J. Am. Chem. SOC., 91,362 (1969); (c) R. A. Sneen and J. W. Larsen; ibid., 91, 6031 (1969). H. Weiner and R. A. Sneen, J. Am. Chem. SOC., 87, 292 (1965). B. G. Ramsey and N. K. Das, J. Am. Chem. Soc., 94,4233 (1972). J. A. Cramer and J. G. Jewett, J. Am. Chem. SOC., 94, 1377 (1972). Since F16 is relatively constant for all participating P-arylethyl substrates in a given solvent,32plots of log kAor log FkA will give essentially the same slope. M. G. Jones and J. L. Coke, J. Am. Chem. SOC.,91,4284 (1969). R. Heck and S. Winstein, J. Am. Chem. SOC.,79, 3432 (1957). Y. Yukawa and Y. Tsuno, Bull. Chem. SOC.Jpn., 32, 971 (1959). S.Winstein and A. F. Diaz, J. Am. Chem. SOC., 91, 4300 (1969). H. Tanida, T. Tsuji, H. Ishitobi, and T. Irie. J. Org. Chem., 34, 1086 (1969). Rate constants for the p N 0 2and p C N neophyl derivative@ were corrected to 0.7kt in view of 25% methyl migrated and 2% nonrearranged product for the pNOp compound. Tanida et al.36reported slightly different values for the Yukawa-Tsuno correlation using uncorrected values: p = -3.70, r = 0.55, correin coeff 0.936 (see ref 34). 95, 1247 (1973); (a) D. S.Noyce and R. L. Castenson, J. Am. Chem. SOC., (b) D. S. Noyce, R. L. Castenson, and D. A. Meyers, J. Org. Chem., 37, 4222 (1972); (c) R. L. Castenson, Ph.D. Thesis, University of California, Berkeley, Calif., 1971. (a) C. C. Lee, G. P. Slater, and J. W. T. Spinks, Can. J. Chem., 35, 1417 (19571: (bl C. C. Lee. R . Tkachuk, and G. P. Slater, Tetrahedron, 7, 206 i1959j: (40) E. F. Jenny and S.Winstein, Helv. Chim. Acta. 41, 807 (1958). (41) A. Diaz, I. Lazdins, and S. Winstein, J. Am. Chem. SOC., 90, 6546 (1968). (42) I. L. Reich, A. F. Diaz, and S. Winstein, J. Am. Chem. SOC.,94, 2256 (1972). (43) T. Ando, N. Shimizu, S.-G. Kim, Y. Tsuno, and Y. Yukawa. Tetrahedron Lett., 117 (1973). (44) (a) P. E. Peterson, D. M. Chevli. and K. A. Sipp, J. Org. Chem., 33, 972 (1968): (b) G. L. Nelson, G. C. Levy, and J. D. Cargioli, J. Am. Chem. SOC., 94,3089 (1972); (c) R. Baker, R. W. Bott, C. Eaborn, and P.M. Greasiey, J. Chem. SOC.,627 (1964); (d) U.Svanholm and V. D. Parker, J. Chem. SOC.,Chem. Commun. 645 (1972); (e) S.Searles, Jr., and M. Tamres in "The Chemistry of the Ether Linkage", S.Patai, Ed., interscience, New York, N.Y., 1967, pp 291, 294. (45) R. W. Tan, E. Price, I. R. Fox, I. C. Lewis, K. K. Andersen, and G. T. Davis, J. Am. Chem. Soc.,85,709,3146 (1963). (46) M. J. Kamlet, E. G. Kayser, J. W. Eastes, and W. H. Gilligan, J. Am. Chem. SOC., 95, 5210 (1973). (47) R. M. Guidry and R. S.Drago, J. Am. Chem. SOC.,95, 759 (1973). (48) M. D. Bentley and M. J. S. Dewar, J. Am. Chem. SOC., 92, 3996 (1970). (49) Y. Yukawa, Y. Tsuno. and M. Sawada, Bull. Chem. SOC.Jpn.. 45, 1206 (1972). (50) B. G. van Leuwen and R. J. Oueilette, J. Am. Chem. SOC., 90, 7056 (1968). (51) A. Streitwieser, Jr., H. A. Hammond, R. H. Jagow. R. M. Williams, R. G. Jesaitis, C. J. Chang, and R. Wolf, J. Am. Chem. SOC., 92, 5141 (1970). (52) R. J. Ouellette and B. G. van Leuwen, J. Am. Chem. SOC., 90, 7061 (1968). (53) H. C. Brown and T. Inukai, J. Am. Chem. Soc., 83, 4825 (1961). (54) I. C. Lewis and L. S.Singer, J. Chem. fhys., 43, 2712 (1965). (55) H. C. Brown and J. D. Cleveland, J. Org. Chem., 41, 1792 (1976). (56) Better agreement of % FkJk, from kinetics and product data is obtained if the latter is measured by percent retention of configuration in ester product: 35 (AcOH) and 85 (HC02H).'3dHowever, since other processes such as olefin formation and hydride shift can occur and are usually included in the rate constant, k,, percent yield of ester with retained configuration is taken as the more appropriate measure. (57) S.Winstein, E. Grunwald, and H. W. Jones, J. Am. Chem. SOC.,73, 2700 (1951). (58) S.Winstein, A. H. Fainberg, and€. Grunwald, J. Am. Chem. SOC.,79, 4146 (1957). (59) E. Grunwaid and S.Winstein, J. Am. Chem. SOC.,70, 846 (1948). (60) A. H. Fainberg and S.Winstein, J. Am. Chem. SOC.,78, 2770 (1956). (61) (a) A. H. Fainberg and S.Winstein, J. Am. Chem. SOC.,79, 1608 (1957): (b) S.G. Smith, A. H. Fainberg, and S. Winstein, ibid., 83, 618 (1961). (62) (a) V. J. Shiner, Jr., W. Dowd, R. 0. Fisher, S. R. Hartshorn, M. A. Kessick. L. Milakofsky. and M. W. Rapp, J. Am. Chem. SOC.,91, 4838 (1969); (b) D. J. Raber, R. C. Bingham, J. M. Harris, J. L. Fry, and P. v. R. Schleyer, ibid., 92,5977 (1970);(c) D. E. Sunko. I. Szele, and M. Tomic, Tetrahedron Lett., 1827 (1972); (d) D. E. Sunko and I. Szele, ibid., 3617 (1972); (e) J. M. Harris, D. J. Raber, W. C. Neal, Jr., and M. D. Dukes, ibid., 2331 (1974). (63) (a) 2 . Rappoport and J. Kaspi, J. Am. Chem. SOC.,96, 586 (1974); (b) 2 . Rappoport and J. Kaspi, ibid.. S6, 4518 (1974); (c) D. J. Raber, M. D. Dukes, and J. Gregory, Tetrahedron Lett., 667 (1974). (64) P. Haake and P. S. Ossip, J. Am. Chem. SOC.,93, 6924 (1971). (65) P. E. Peterson and F. J. Waller, J. Am. Chem. SOC., 94, 991 (1972). (66) D. A. da Roza, L. J. Andrews, and R. M. Keefer, J. Am. Chem. SOC.,95, 7003 (1973). (67) Values of kSo(2.5 X and ~ A O(3.5 X IO-'), the rates for 80% aqueous ethanol extrapdated from plots4' of log k, vs. log k,(EtOTs) and log k:, vs. log k,(neophyCOTs), agree well with the intercepts of eq 5 and and 6.0 X respectively. 6, 2.45 X (68) Fvaries with solvent (Tables ill and Vi), and therefore log k:, instead of log F k l must be plotted. '

'

245

(69) Inclusion of a solvent nucleophilicity term yields the correlation, log ( k : , l k ~ O= ) 0.64Y 0.05N, with only slightly better correlation: correin coeff 0.998. (70) Inclusion of the solvent nucleophilicity terms yielded the equation log ( k J k i o ) = 0.65Y - 0.23N (correin coeff 0.996). The observation of a negative sign for I:, is analogous to the situation for the primary P-phew ethyl system ( / A = -0.05),although the magnitude in the present case is substantially greater. Since the degree of correlation is not greatly enhanced by including I, and because a negative value of l is somewhat difficult to interpret, we will use eq 8 in the following analysis. (7 1) R. N. McDonald, N. L. Wolte, and H. E. Petty, J. Org. Chem., 38, 1106 (1973). (72) J. E. Nordlander and W. G. Deadman, J. Am. Chem. SOC., 90, 1590 (1968). (73) (a) G. A. Olah, M. B. Comisarow, E. Namanworth, and B. Ramsey, J. Am. Chem. SOC.,89, 5259 (1967). (b) G. A. Olah and R. D. Porter, ibid., 92, 7627 (1970); G. A. Oiah, R. D. Porter, and D. P. Kelly, ibid., 93, 464 (1971). (74) J. E. Nordlander and W. J. Kelly, J. Am. Chem. SOC.,91, 996 (1969). (75) C. Yamagami and A. Sera, Chem. Lett., 741 (1972). (76) P. E. Peterson and R. J. Kamat, J. Am. Chem. Soc., 91, 4521 (1969). (77) klo(lV) and kp(V) were extrapolated from the appropriate plots. it was assumed kso(lV)= kto(V). (78) F. R . Jensen and R. J. Ouellette. J. Am. Chem. SOC.,85, 367 (1963). (79) J. L. Coke, F. E. McFarlane, M. C. Mourning, and M. G. Jones, J. Am. Chem. Soc., 91, 1154 (1969). (80) R . J. Jablonski and E. I. Snyder, J. Am. Chem. SOC.,91, 4445 (1969). (81) S. A. Sherrod and R. G. Bergman, J. Am. Chem. SOC., 93, 1925 (1971). (82) See Table X and footnote d therein. (83) (a) J. E. Leffler and E. Grunwald, "Rates and Equilibria of Organic Reactions", Wiley, New York, N.Y., 1963. pp 312-314; (b) Table iI, footnote a; (c) Y. Okamoto, T. inukai, and H. C. Brown, J. Am. Chem. SOC.,80, 4972 (1958); (d) N. B. Chapman and J. Shorter, Ed., "Advances in Linear Free Energy Relationships", Plenum Press, New York, N.Y., 1972. (84) K. Yates, G. H. Schmid. T. W. Regulski, D. G. Garratt, H.-W. Leung, and R. McDonald, J. Am. Chem. SOC.,95, 160 (1973). (85) (a) H. C. Brown and J. D. Brady, J. Am. Chem. SOC., 74,3570 (1952); (b) M. Tamres, ibid., 74, 3375 (1952); (c) P. v. R. Schleyer, D. S.Trifan and R. Bacskai, ibid., 80, 6691 (1958), and references therein; (d) D. Cook, Y. Lupien, and W. G. Schneider, Can. J, Chem., 34, 964 (1956). (86) (a) W. M. Schubert and D. F. Gurka, J. Am. Chem. SOC.,91, 1443 (1969): (b) A. Himoe and L. M. Stock, ibid., 91, 1452 (1969); (c) L. M. Stock and M. R. Wasielewski, J. Org. Chem., 36, 1002 (1971): (d) H. C. Brown and R. A. Wirkkala, J. Am. Chem. SOC.,88, 1447, 1453, 1456 (1966). (87) (a) C:D. Johnson and K. Schofield. J. Am. Chem. SOC.,95, 270 (1973); (b) D. FHrcagiu. J. Chem. Educ., 52,76 (1975); (c) C. D. Johnson, Chem. Rev., 75,755 (1975); (d) 1. Roberts, C. D. Johnson, and P. G. Taylor Tetrahedron, 2123 (1977); (e) R. N. McDonald, J. M. Richmond,J. R . Curtis, H. E. Petty, and R. J. Mobley, J. Am. Chem. SOC., 99, 5739 (1977). (88) G. S.Hammond, J. Am. Chem. SOC.,77, 334 (1955). (89) D. S.Kemp and M. L. Casey, J. Am. Chem. SOC., 95, 6870 (1973). (90) C. D. Ritchie, Acc. Chem. Res., 5, 348 (1972). (91) V. J. Shiner, Jr., and R. C. Seib, J. Am. Chem. SOC., 98, 862 (1976). (92) D. E. Sunko and S. BorEiC in "isotope Effects in Chemical Reactions", C. J. Collins and N. S.Bowman, Ed., Van Nostrand-Reinhoid. New York, N.Y., 1970, p 176. (93) A. Streitwieser. Jr., "Solvoiytic Displacement Reactions", McGraw-Hill, New York, N.Y., 1962, pp 174, 175. (94) S. Winstein, J. Am. Chem. SOC.,87, 381 (1965). (95) (a) W. H. Saunders, Jr., S.Asperger, and D. H. Edison, J. Am. Chem. SOC., 80, 2421 (1958); (b) W. H.Saunders, Jr., and R . Giaser, ibid., 82, 3586 (1960). (96) (a) V. J. Shiner, Jr., M. W. Rapp, E. A. Haievi. and M. Wolfsberg, J. Am. Chem. Soc., 90, 7171 (1968): (b) V. J. Shiner, Jr., and W. Dowd, ibid., 93, 1029 (1971); (c) V. J. Shiner, Jr., and R. D. Fisher, ibid., 93, 2553 (1971). (97) K. Humski, V. Sendijarevic, and V. J. Shiner, Jr., J. Am. Chem. Soc., 95, 7722 11973). - -, (98) (a) YT Yukawa, T. Ando, K. Token, M. Kawada, and S . 4 . Kim, Tetrahedron Lett.. 2367 (1969): fb) ibid.. 847 119711. (99) (a) Y.'Yukawa, S . 4 . Kkn, and H. Yaktaka, Tetrahedron Lett., 373 (1973); (b) H. Yamataka, S.-G. Kim, T. Ando, and Y. Yukawa, ibid., 4767 (1973). (100) (a) J. M. W. Scott. Can. J. Chem., 48,3807 (1970); (b) J. Koskikailio, Acta Chem. Scand., 26, 1201 (1972). (101) The R+ group in ion pair RfX- in Scheme iI is considered to possess no bonding character to neighboring group or solvent. (102) M. H. Abraham, J. Chem. SOC.,Chem. Commun., 51 (1973). (103) F. P. Lossing and G. P. Semeluk, Can. J, Chem., 48, 955 (1970). (104) (a)M. J. S.Dewar, lnt. Congr. Pure Appl. Chem., 23rd, 1, 1 (1971), Special Lectures presented at Boston, Mass.; (b) N. Bodor, M. J. S.Dewar, and 94, 5303 (1972); (c) L. Radom, J. A. Pople, D. H. Lo, J. Am. Chem. SOC., V. Buss, and P. v. R. Schleyer, ibid., 94, 31 1 (1972); (d) M. Saunders, P. Vogel, E.L.Hagan, and J. Rosenfeid, Acc. Chem. Res., 6, 53 (1973). (105) (a) G. A. Olah, J. R. DeMember, R. H. Schlosberg. and Y. Haipern, J. Am. Chem. Soc., 94, 156 (1972); (b) G. A. Olah, J. R. DeMember, and J. Shen, ibid., 95, 4952 (1973). (106) (a) R. Baird and S.Winstein, J. Am. Chem. SOC., 85, 567 (1963); (b) A. H. Fainberg and S.Winstein, ibid., 79, 1602 (1957). (107) C. C. Lee, D. Unger, and S.Vassie, Can. J. Chem., 50, 1371 (1972). (1081 Evaluated in the same four solvents as eo 7 for k,(PhCH&H(OTs)CH,). ". . correln coeff 0.995. (109) S.L. Loukas, M. R. Velkou,andG. A. Gregoriou. Chem. Commun., 1199 (1969). (1 IO) S.L. Loukas, M. R. Velkou, and G. A. Gregoriou, Chem. Commun., 251

-

\

-

246

Journal of the American Chemical Society

(1970). (111) Elimination and hydride shiR processes are usually Included in ks. See ref 3, 1%. 14b, and 20 and Perspectives and Conclusions. (112) F.G.Bordwell, P. F. Wi1ey.andT.G. Mecca, J. Am. Chem. Soc., 87, 132 (1975). (113) S.Winstein and L. L. Ingraham, J. Am. Chem. Soc.,77, 1738 (1955). (1 14) S. Winstein, E. Allred, R. Heck, and R. Glick, Tetrahedron, 3, 1 (1958). (1 15) L. F. Fieser and M. Fieser, “Reagents for Organic Synthesis”, Wiley, New York, N.Y.. 1967: (a) p 584; (b) p 1180. (1 16) F. G. Mann and J. Watson, J. Chem. Soc., 505 (1947). (1 17) W. A. Wisansky and S. Ansbacher, Org. Synth., 28,46 (1948). (118) W. E. Bachmann and J. C. Sheehan, J. Am. Chem. SOC., 82, 2687 (1940). ( 119) C. J. Lancelot, Ph.D. Thesis, Princeton University, Princeton, N.J. (1971). (120) Table IV, footnote a. (121) S. E. Coan &E. I.Becker, “Organic Syntheses”, Collect. Vol. IV, Wiley, New York, N.Y.. 1963, pp 174, 176. (122) Typical procedure: G. Berti and A. Marsili, Ann. Chim. (Rome), 51, 675 (1961). Cf. Chem. Abstr., 58, 2363a (1962). (123) C, H. DePuy, D. L. Storm, J. T. Frey, and C. G. Naylor, J. Org. Chem., 35, 2746 (1970). (124) S. Chiavarelli, G. Settimj, and H. M. Alves, Gazz. Chim. /tal., 87, 109 (1957). (125) C. G. Overberger and H.Biletch, J. Am. Chem. SOC., 73,4880 (1951). (126) H. G. Walker and C. R. Hauser, J. Am. Chem. SOC., 88, 1386 (1946). (127) A. Buzas and C. Dufour, Bull. SOC.Chim. Fr., 139 (1950). (128) V. R. Likhterov and V. S. Etlis, Zh.Obshch. Khim., 27, 2887 (1957). (129) H. Woodburn and C. Stuntz, J. Am. Chsm. Soc., 72, 1361 (1950). (130) H. 0. Houseand W. F. Berkowitz, J. Org. Chem., 28, 2271 (1963).

1 1OO:l 1 January 4, 1978

(131) C. D. Hurd and W. W. Jenkins, J. Org. Chem., 22, 1418 (1957). (132) J. J. Harper, Ph.D. Thesis, Princeton University, Prlnceton. N.J., (1988). (133) J. C. Sauer in “Organic Syntheses”, Collect. Vol. IV, N. Rabjohn, Ed., Wiley, New York, N.Y., 1963, p 560. (134) S. Winstein, C. Hanson, and E. Grunwald, J. Am. Chem. Soc.,70, 812 (1948). (135) H. Lunda! J. Bjerrum, Chem. Ber., 84, 210(1931); see L. Fieser, “Experiments in Organic Chemistry”. 3rd ed, D. C. Heath, Boston, Mass., 1957, p 285. (136) D. F. Detar in “Computer Programs for Chemistry”, Vol. I, D. F. Detar. Ed., W. A. Benjamin, New York, N.Y., 1968, pp 128-173. (137) W. F. Sliwinski, Ph.D. Thesis, Princeton University, Prlnceton, N.J. (1972). (138) C. G. Swain and C. R. Morgan, J. Org. Chem., 29,2097 (1964). (139) P. E. Peterson, R. E. Kelley, Jr., R. Belloli, and K. A. Sipp, J. Am. Chem. Soc., 87, 5169 (1965). (140) J. S.Fritz, Anal. Chem., 26, 1701 (1954). (141) J. E. Nordlander, R. R. Gruetzmacher, W. J. Kelly, and S. P. Jindal, J. Am. Chem. Soc., 96, 181 (1974). (142) D. D. Perrin, W. L. F. Armarego, and D. R. Perrin, “Purification of Laboratory Chemicals”, Pergamon Press, Oxford, 1966, p 268. (143) R. E. Banks, “Flourocarbons and Their Derivatives”, 2nd, Ed., American Elsevier, New York, N.Y., 1970, pp, 72, 77. (144) Results showed that solvolysis grade CF,COOH, stored In a screw top bottle on the laboratory shelf with occasional opening, absorbed moisture to give the same effect. (145) A. Streitwieser, Jr., A. Lewis, I. Schwager, R. W. Fish, and S. Labana, J. Am. Chem. SOC.,82,6525 (1970). (146) C. Eaborn, P. M. Jackson, and R. Taylor, J. Chem. SOC.8, 613 (1986).

The Solution Conformation of Nicotine. A ‘H and 2H Nuclear Magnetic Resonance Investigation T. Phil Pitner,* William B. Edwards, III,* Ronald L. Bassfield,* and Jerry F. Whidby* Contribution from the Philip Morris U.S.A.Research Center, P.O.Box 26583, Richmond, Virginia 23261. Received June 29, 1977

Abstract: The ‘H NMR spectrum of nicotine (1) is analyzed in detail. Selectively deuterated nicotine analogues afford considerable simplification of the pyrrolidine ‘H resonances by allowing partition of this seven-spin system into three-, four- and fivespin systems. In addition, the *H NMR chemical shifts of these analogues provide a means of assigning ‘ H NMR chemical shifts to specific protons unambiguously. The vicinal coupling constants of the pyrrolidine ring protons suggest an envelope conformation for the five-membered ring, in which the methyl and pyridine moieties assume equatorial positions. A perpendicular spatial orientation of the pyridine and pyrrolidine rings is supported by two observations: a small long-range coupling constant (