ENTROPY CONTRIBUTIONS To REMOTE SECONDARY KINETIC HYDROGEN ISOTOPE EFFECTS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY VINCENT FRANCIS SMITH, IR. 197 2 'J) I)ate 0-7639 ENTROPY CONTRIBUTIONS TO REMOTE SECONDARY This is to certify that the thesis entitled KINETIC HYDROGEN ISOTOPE ' EFFECTS presented by Vincent Francis Smith, Jr. has been accepted towards fulfillment of the requirements for Ph.D. degree in Chemistry June 16, 1972 6f (MIR Major professor LIBRARY Michigan State University Y :5 ”HUNG BY "' HUAG & SflNS' 800K BINDERY INC. LIBRARY BINDERS a. warez-9! r. 10.2mm I ‘7' ABSTRACT ENTROPY CONTRIBUTIONS TO REMOTE SECONDARY KINETIC HYDROGEN ISOTOPE EFFECTS BY Vincent Francis Smith, Jr. In an effort to establish the importance of contri- butions to remote secondary kinetic isotope effects from entropy differences between isotopic substrates, the rates of solvolysis of the 2,6,6-trimethyl-gng272-norbornyl p-nitrobenzoates (Ia-c) were measured in 80% ethanol-water over the temperature I Ia: R = R'=CH R ‘IIMII’ CH Ib‘ R 3 10: R = CD 3 (Q0) CH3; R'=CD3 (endo-G-CD 3) ; R'=CH (exo-G-CD ) R' OPNB 3 3 3 range 100° to 150‘ by a spectrophotometric technique. Brown and coworkers1 have suggested that the ioni- zation process in this system is sterically impeded by the Vincent Francis Smith, Jr. interaction which develops between the leaving group and the endo-6-methy1 group. This, in conjunction with Bar- tell's theory2 of steric isotope effects, would lead to the prediction of an inverse effect for both Ib and Ic. However, the observed effects were normal in both cases and were found to vary with temperature as follows: Temp.°C Substrate kH/kD 100.840 endgf6-CD3 0.997 1 0.016 125.217 gigg-G-CD3 1.003 1‘. 0.014 142.244 gngng-CDB 1.027 : 0.013 150.186 flig-G-CDB 1.026 t 0.010 100.840 egng-CD3 1.011 i 0.015 125.217 §§gf6-CD3 1.017 :_0.014 142.244 .e_}_(_o_-6-CD3 1.043 t. 0.013 150.186 eéng-CD3 1.049 : 0.011 From the activation parameters obtained with these data, it was apparent that the enthalpy differences be- tween the isot0pic substrates were consistent with those predicted, and that it was, in fact, entropy differences that gave rise to the surprising normal effects. Vincent Francis Smith, Jr. +_+ +-+ Substrate AH H AH D AS H AS D (cal/mole) (cal/deg/mole) endo-6—CD3 204 1.82 0.53 i 0.20 exo-6-CD3 244 i 96 0.67 i 0.24 Differential solvation and internal rotational differences were considered as possible sources of these entrepy-controlled effects. It was concluded that entropy contributions to effects of a non-bonded nature may indeed be significant and that predictions based upon Bartell's model may be most useful in estimating enthalpy differences alone. REFERENCES 1. S. Ikegami, C. L. Vander Jagt, and H. C. Brown, J. Am. Chem. Soc., 92, 7124 (1968). 2. L. S. Bartell, Ibid., 83, 3567 (1961). ENTROPY CONTRIBUTIONS TO REMOTE SECONDARY KINETIC HYDROGEN ISOTOPE EFFECTS By Vincent Francis Smith, Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1972 To Bonnie, whose patience, tolerance, and hard work has contributed immeasurably to this thesis. ii ACKNOWLEDGMENTS The author wishes to express his appreciation to Professor Gerasimos J. Karabatsos for his interest through- out the course of this investigation. Appreciation is also expressed to Dr. Gerald W. Klein for his confidence and his contribution to the development of the author's practical abilities. The financial support provided by the National Science Foundation and the Department of Chemistry, Michigan State University, is gratefully acknowledged. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . . . EXPERIMENTAL . . . . . . . . . . . . . . . . . . I o syntheSiS o o o o o o o 4 o o o o o o 0 Preparation of 2-methy1-5-norbornene- 2-CarbOXY1iC aCid o o o o o o o o o 0 Separation of epimeric carboxylic acids Preparation of exo-2—methy1-endo-2- hydroxydideuteriomethyl-5-norbornene . Preparation of endo-Z-methyl-exo-Z- hydroxydideuteriomethyl-S-norbornene . Preparation of exo-Z-methyl—endo-Z- tosyloxydideuteriomethyl-5-norbornene Preparation of endo-Z-methyl-exo-Z- tosyloxydideuteriomethyl-S-norbornene Preparation of endo-6-trideuterio- methyl-exo-6-methy1-2-norbornene . . . Preparation of exo-6-trideuteriomethy1- endo-6-methyl-2-norbornene . . . . . . Preparation of exo-2,3-epoxy-endo-6- trideuteriomethyI-exo-6-methy1norbornane Preparation of exo-2,3-epoxy-exo-6- trideuteriomethyI-endo-6-methyInorbornane Preparation of endo-6-trideuteriomethy1- exo-6-methyl-exo-2-norbornanol . . . . . . Preparation of exo-6-trideuteriomethy1- endo-6-methy1- xo-2-norbornanol . . . Preparation of endo-6-trideuteriomethy1- exo-6-methyl-2-norbornanone . . . . . iv Page vii ix 16 16 16 16 18 19 19 20 20 22 22 23 23 24 24 Page Preparation of exo-6-trideuteriomethy1- endo-6-methy1-2-norbornanone . . . . . . . . 26 Preparation of 2,6-dimethy1-endo-6- trideuteriomethyl-endo-2-norBornanol . . . . 26 Preparation of 2,6-dimethy1-exo-6- trideuteriomethyl-endo-2-norEornanol . . . . 27 Preparation of 2,6-dimethy1-endo-6- trideuteriomethyl-endo-2-nor50rny1 p-nitrobenzoate . . . . . . . . . . . . . . 28 Preparation of 2,6-dimethy1-exo-6- trideuteriomethyleendo-2-norBorny1 p-nitrobenzoate . . . . . . . . . . . . . . 29 Preparation of unlabeled compounds . . . . . 29 II. Solvolysis Product Study . . . . . . . . . . . 30 Examination of products generated under solvolytic conditions . . . . . . . . . . . 30 Examination of the stability of 2,6,6- trimethyl-endo-2-norbornanol in the .presence of p-nitrobenzoic acid . . . . . . 31 III. Kinetics O O O O O O I O O O O O O O O O O O O 31 Constant temperature bath . . . . . . . . . 31 Sample ampoules and holder . . . . . . . . . 32 Calibration of Beckmann differential thermometer . . . . . . . . . . . . . . . . 33 Preparation of solvents . . . . . . . . . . 33 Water . . . . . . . . . . . . . . . . . . 33 EthanOl O O O O O 0 O O O O O O O O O O O 33 Mixed solvent . . . . . . . . . . . . . . 33 Determination of Beer's Law . . . . . . . . 33 p-nitrobenzoic acid . . . . . . . . . . . 33 2,6,6-trimethy1-endo-2-norborny p-nitrobenzoate . . . . . . . . . . . . . 34 Kinetics of solvolysis . . . . . . . . . . . 34 RESULTS AND DISCUSSION CONCLUSION . I C O O I REFERENCES . . . . . . vi Table l. 2. 10. LIST OF TABLES y-d3 Isotope effects for the hydrolyses of n-propyl derivatives . . . . . . . . . . . Kinetic isotOpe effects for the reaction of various pyridines with methyl iodide at 5 2 o O O O I O O O O O I O O O O O O O O I Products from the solvolysis of 2,6,6-tri- methyl-endo—Z-norbornyl.p-nitrobenzoate in 80% ethanol-water (V/V) at 150° . . . . Products from the treatment of 2, 6, 6- -tri- methyl- endo-2-norbornanol with p-nitro- benzoic ac1d under solvolytic conditions at 150- 5° for the equivalent of 10- ll half-lives . . . . . . . . . . . . . . . . Beer's Law determination for 2, 6, 6- -tri- methyl- endo-Z-norbornyl p-nitrobenzoate in 95% e——than01 O I O O I O O O O I O I O O Beer's Law determination for 2,6,6-tri- methyl-endo—Z-norbornyl p-nitrobenzoate in cyclohexane . . . . . . . . . . . . . . Beer's Law determination for sodium p- nitrobenzoate in 95% ethanol . . . . . . . Data for the solvolysis of 2, 6, 6- -trimethy1- endo-Z-norbornyl p-nitrobenzoate in run 35 at 149. 790° in 80% ethanol-water (V/V) Data for the solvolysis of 2, 6, 6- -trimethy1- endo- -2-norbornyl p-nitrobenzoate in run 59 at 150.186° in 80% ethanol-water (V/V) . . Rates of solvolysis of 2, 6, 6- -trimethy1- endo-Z-norbornyl p-nitrobenzoates in 80% m5 nol-water (V/V) . . . . . . . . . . . vii Page 39 4O 41 41 42 49 50 52 Table Page 11. Temperature dependence of isotope effects in the solvolysis of 2,6,6—trimethy1-endo-2- norbornyl penitrobenzoates in 80% ethanol- water (V/V) . . . . . . . . . . . . . . . . 55 12. Activation parameters for the solvolysis of 2,6,6-trimethyleendo-2-norbornyl penitro- benzoates in.80% ethanol-water (V/V), . calculated from results obtained at all temperatures . . . . . . . . . . . . . . . . 56 13. Activation parameters for the solvolysis of 2,6,6-trimethyl—endo-2-norbornyl p-nitro- benzoates in 80% ethanol-water (V/V), omitting results obtained at 160.671° . . . 57 14. Comparison of isotOpe effects calculated from activation parameters and those experimen- tally measured 0 O O O O O I I O O O O O O O 60 viii LIST OF FIGURES Figure Page 1. (a) Potential energy curves for decreasing force.constant of isotopic bond . . . . . 3 (b) Potential energy curves for increasing force constant of isotopic bond . . . . . 4 2. Beer's Law plots of 2,6,6-trimethy1-endo-2— norbornyl p—nitrobenzoate and sodium p-nitrobenzoate in 95% ethanol . . . . . . . 43 3. Beer's Law plotof2,6,6-trimethy1-endo-2- norbornyl p-nitrobenzoate in cyclohexane . . 44 ix INTRODUCTION It should be obvious to all but the most naive ob- server, that application of secondary kinetic isotope effects as a mechanistic criterion is useful only to the degree to which the origin of the isotope effect itself is understood. Ironically, it is possible to determine with certainty the factors contributing to a given effect only if the intimate mechanism of the reaction involved is explicitly defined, a situation not often encountered in modern chemistry. One is therefore left with this dichotomous relationship between reaction mechanism and isotope effect, and must bear in mind this interdependence when drawing conclusions from either phenomenon. $Considerable amounts of data have been amassed in efforts directed toward the elucidation of the nature of 1 and 82 secondary hydrogen isotope effects. solvolytic 0 However, effects~arising from deuterium substitution at sites further removed from the reaction center have re-~ ceived relatively little attention. Almost certainly a prime reason for this lack of data is the generally smaller magnitude of these remote effects. The precision and accuracy required for such studies is often attainable only under the most stringent conditions. It is generally accepted that secondary isotope effects arise from force constant changes which occur dur— ing the progression from ground state to transition state during a given reaction. Figure 1 illustrates that if a reaction proceeds such that the force constant associated with the isotopic bond is decreased (the bond weakened) relative to that in the ground state, the activation energy for the deuterated compound will be greater than that for the unlabeled compound; this is due to the greater zero- point energy difference in the stronger bond of the ground state. This leads to the "normal" isotope effect (kH/kD>l). On the other hand, if the reaction involves an increase of force constant (bond strengthening) at the site of isotopic substitution, the "inverse" isotope effect (kH/kD<1) will be exhibited. Presently, the cause of these force constant changes is discussed in terms of inductive effects,2’3 hyperconjugative effects,4 and steric (non-bonded) effects.5 It is still subject to question as to how much each of these factors contributes to the observed isotope effects, especially B-effects. It has been suggested by Karabatsos: §E_213P that, in systems where hyperconjugative stabiliza- tion involving the isotopic bond is possible, less than 10% of the observed isotope effect is due to non-bonded interactions. kH/kD > Potential energy AE+H < AE+D\ / 1 ‘V, i 7(— AE+ H 13+ \V/ Reaction coordinate Figure 1a. Potential energy curves for decreasing force constant of isotopic bond. ‘0 .1 ' In, 'I l . 13+ > 13+ kH/kD < 1 Potential energy 13+ 13+ Reaction coordinate Figure 1b. Potential energy curves for increasing force constant of isotopic bond. In order to assess the relative magnitude of purely steric isotope.effects, a system must be considered wherein hyperconjugative and inductive involvement of the isotopic bond is minimal. Mislow and coworkers7 provided the first unambiguous demonstration of such an effect in their study of the rate of racemization of 9,10-dihydro-4,5-bis(tri- deuteriomethyl)phenanthrene (I); at 42° in benzene, kH/kD = .885. CD3 CD3 Measurement at three other temperatures allowed the deter- mination of activation parameters. Thus AH+H - Afi+ and D AS+H - AS+D were found to be 240 cal/mole and 0.53 e.u. respectively. Consequently, although enthalpy favored the isotOpic substrate by a relatively large amount (corres- 'ponding to a kH/kD = .682), entropy favored the unlabeled compound by an amount such that at 42° AG+H - AG+D was only 80 cal/mole. It must be noted at this point that steric isotope effects, as developed by Bartell,5 result from internal energy changes alone and consequently require activation entropy differences to be zero, i.e., AG+H - AG+D = AH+H - AH+D and AS+H - AS+D = 0. Clearly, the data of Mislow and coworkers do not lend support to this hypothesis. Similar studies of racemization rates in the bi- phenyl (II) and binaphthyl (111)10 systems have been reported. As expected, both systems gave inverse isotope effects. III F— In the biphenyl study, at -19.8° in ethanol kH/kD H was found to be 0.84. Rate measurements over a 20° range yielded isotope effects which were the same within experi- mental error. No meaningful activation parameter differ- ences could be obtained. For III, Carter and Dahlgrenlo found kH/kD = .88 at 64.2° in N,N-dimethylformamide solu- tion. The effect increased to a maximum of 0.83 over a 42° range. Here again activation parameter differences did not provide support for the assumption that Ad+H - AG+D = Afi+H - Afi+D according to Bartell's theory. Afi+H - AH+D was found to be 270 cal/mole and AS+H - AS+D 0.54 e.u. In this as well as the other racemization studies, the activa- tion parameters were shown to be essentially independent of the nature of the solvent, confirming the expectation that little charge development occurs along the reaction coordinate. Earlier, Robertson, gt_al.,ll had reported a solvolytic isotope effect that was assumed to be free of complicating hyperconjugative or inductive contributions. Thus, examination of the hydrolyses of a series of n-propyl derivatives containing a terminal trideuteriomethyl group yielded the results collected in Table 1. It was suggested that steric inhibition of vibrations involving the terminal methyl group was responsible for the observed 5-8% inverse isotope effects. This inhibition could arise from inter- actions between the methyl group and either the leaving group, or the incoming solvent molecule. In the view of the authors, the former was considered more important. Table l. y-d3 Isotope effects for the hydrolyses of n-propyl derivatives. kH/kD Leaving Group T,°C. (+ 006) Benzenesulfonate 54.18 0.947 Methanesulfonate 60.00 .943 Bromide 80.01 .921 Iodide 90.00 .924 More recently, a similar effect was reported by Jewett and Dunlap12 who studied the aqueous ethanolysis of dimethylneopentylcarbinyl chloride (IV) and its 6-nonadeuterated analog. At a CH CH 3 l 3 CH3—:|: —, CH2— : — c1 H3 H3 IV temperature unspecified by the authors kH/kD was 0.983 in 80% aqueous ethanol. This effect might similarly be ex- plained as the result of an increase in non-bonded inter- actions between the y-methyl groups and the (solvated) leaving group, although the authors implied that the ef- fect was inductive in origin. In view of previous success in determining the significance of steric factors by studying heats of re- action of pyridine bases with Lewis acids and the rates of reaction with alkyl halides, Brown and McDonald13 reported the data shown in Table 2. Although the authors favored a purely steric explanation for the effects ob- served, the evidence cited does not appear particularly compelling. An inductive contribution may also be present. However, it was shown in a succeeding paper,14 that the reaction between 2,6-bis(trideuteriomethyl)pyridine and boron trifluoride was more exothermic than the reaction Table 2. Kinetic isotOpe effects for the reaction of various pyridines with methyl iodide at 25°. Pyridine kH/kD 4-Methyl-<_fl_3 0.999 t .003 3-Methyl-g3 .991 i .002 2-Methyl-g3 .971 i .003 2,6-Dimethy1-d6 .913 i .003 Pyridine-4111 .988 i .002 Pyridine-g5 .970 i .006 with the unlabeled substrate by 230 cal/mole; when a sterically less demanding Lewis acid like diborane15 was used, this effect apparently vanished. Since these two Lewis acids are thought to be equally sensitive to polar influences,15 these results may be taken as evidence against the operation of an inductive effect in this equilibrium reaction. Implications that similar inductive effects are operable in the kinetic process are valid only to the extent to which the transition state resembles the product of this equilibrium. It is interesting that the authors also noted that a steric explanation would predict an increasing effect with increasing bulkiness of the alkyl iodide. Thus, the effects of 2-methy1-d3-pyridine with ethyl and isopropyl 10 iodides were determined at 75° and 100° respectively; cal- culations based on the assumption that 138+H - AS+D = 0 were then performed and the isotope effects compared at 25°! A comparison of predicted and experimental effects is in this case of questionable utility for two reasons. Firstly, the trend predicted is reasonable only if it is assumed that as one proceeds from methyl and ethyl to isopropyl, the transition state "tightness" remains un- changed, i.e., that the nucleophile approaches the central carbon to the same extent in each case, thus experiencing greater non-bonded repulsions with any increased substitu- tion about that carbon. In fact, it may be argued16 that as substitution is increased, transition state distances also increase and the attacking nucleophile thus experi- ences lesser non-bonded repulsions, leading to smaller isotope effects. Secondly, without a better understanding of the contribution (if any) of entropy differences to determining non-bonded isotope effects,* values obtained by extrapolations based upon only one temperature should be treated with due caution. In this light it is interesting to note the obser- 6,18 vations of Karabatsos, et al., in their studies of the 8-trideuteriomethy1-l-naphthoyl chlorides (V) and the . *Robertson, et al.,17 have presented certain data wh1ch would 1ndicate that some B-effects are entirely due to entrOpy differences. 11 8-trideuteriomethy1-l-chloromethyl-naphthalenes (VI). Available evidence would indicate that both systems under- go solvolysis by a limiting mechanism.6 Use of Bartell's COCl CD CH C1 CD 3 2 3 VI procedures results in calculated AAE values of about -100 to -300 cal/mole for V and +70 to +49 cal/mole for V1.19 When the acid chloride V was solvolyzed in 95% aqueous acetone at 25° kH/kD = 1.029 : .015 corresponding to AG+H - AG+D==-16.6 : 8.3 cal/mole; and in 75% aqueous acetone over the temperature range -20° to -34° AG+H — AG+D = -60 to -74 cal/mole. The data yielded AH+H - AH+D = -308 i 102 cal/mole, a value in agreement with that cal- culated. Similarly, when the carbinyl chloride VI was solvolyzed in 67% aqueous acetone at 25°, kH/kD = 1.013 1 .022 and AG+H - AG+D = _8 :.13 cal/mole, an effect in direction opposite to that calculated! However, from kinetic measurements at 15° and 35°, it was established that AH+H - AH+D = 140 i 300 cal/mole and AS+h - AS+D = 0.49 i .44 e.u. Again it would appear that the enthalpy of activation agrees with the effect calculated by Bartell's procedure, but the free energy of activation does not! Thus, it appears that entropy differences may indeed play an important role in determining the 12 magnitude of experimental non-bonded isot0pe effects, and this research was initiated to further understand this contribution. In selecting a system with which the steric iso- tOpe effect would be most effectively investigated, it is necessary that, as previously noted, contributions from hyperconjugation or induction be minimized. Consideration should also be given Mislow's conclusion7 "...that steric hydrogen isotope effects are plainly operative only under special conditions of severe overcrowding," and Bartell's observation5b "...that if the number of carbons separating the hydrogen and the leaving group, X, is greater than two or three, the purely steric effect may be very small even if standard molecular models suggest an unusually close H-o-X distance...because the skeletal flexibility in- creases substantially with increasing links in the chain unless unusual constraints are imposed, and the steric effect vanishes rapidly with modest increases of flexi- bility." In addition, the mechanism of the reaction under observation must be understood and defined as completely as possible. Since special conditions of severe overcrowding have been suggested as the cause of the abnormally high solvolytic gxg/gndg rate ratio for the 2,6,6-trimethyl-2- 20 norbornyl-p-nitrobenzoates (VII), it seemed that this 13 system would serve as a convenient probe for further exam~ ination of the steric isotope effect. It has generally been accepted that the rigid norbornane structure consti- tutes an ideal system for the investigation of large steric effects,21 thereby providing the "unusual constraints" deemed necessary by Bartell. Mechanistically, the solvoly- sis of VII is well established in that a wide variety of evidence now indicates that tertiary norbornyl substrates react via classical carbonium ions.22 In particular, the successful trapping of the optically active 1,2-dimethyl- 2-norbornyl cation or ion pair reported by Goering and Humski23 provides convincing support for this mechanism. Brown and coworkers20 have presented kinetic data which indicate that the large repulsive forces between the endng-methyl and the endng-methyl of the gngp-nitroben- zoate substrate are effectively relieved as the transition state geometry is approached. In fact, at 25° in 80% aqueous acetone, gngVII solvolyzed at a rate 726 times greater than endo-2-methy1-exo-2-norbornyl-p-nitrobenzoate. 14 Therefore, based upon steric considerations alone,* it would be expected that substitution of a trideuteriomethyl group in the.gndgf6-position would produce a rather large normal isotope effect. On the other hand, as the transi- tion state is approached by the endgfp-nitrobenzoate sub- strate, the already crowded endng-methyl group becomes even more constrained by the (solvated) leaving group. Thus, when compared at 25° in 80% aqueous acetone, endg- VII is about 6 times less reactive than the gngZ-methyl— gnd2f2-norbornyl derivative. This increased crowding of the endng-methyl group would be expected to lead to an inverse isotope effect with substitution by a trideuterio- methyl group. Similar substitution in the gngG-position of either substrate may be expected to produce an effect reflecting the nature of the interaction (if any) between the geminal methyl groups as well as that between the 3x2: 6-methyl and the 7-hydrogen syn_to it. It would seem reasonable that these effects should be in the same direc- tion as those associated with the endgf6-position but smaller in magnitude (perhaps undetectable). It is interesting to note that Goering and Humski24 also reported on the y-isotope effect for the previously mentioned l,2-dimethyl-gng2-norbornyl-p-nitrobenzoate (VIII). It was found that at 78.5° in 90% aqueous acetone *Both hyperconjugative and inductive effects should be negligible in this 6-isotope effect. 15 VIII kH/kD = 1.02 i .01. In the words of the authors, "The present results show that the y-isotope effect for unas- sisted ionization in an gngnorbornyl system is small, as previously assumed..." It must be pointed out, however, that the previous assumption25 was made for solvolysis at 25°. Again, the need for a better understanding of entropy differences in reactions of isotopic substrates is well illustrated. 1r EXPERIMENTAL I. Synthesis Preparation of 2-methyl-5- norbornene-Z-carboxylic acid To a lOOOml round-bottomed flask equipped with a reflux condenser was added 181g (2.1mol) glacial meth- acrylic acid.(Rohm & Haas) and 168g (2.5mol) freshly dis- tilled cyclopentadiene (dimer from Eastman). The mixture was allowed to react at reflux on a steam bath for 3.5 hrs. Immediate distillation of the product under reduced pressure afforded l70g (1.1mol, 53% theoretical amount) 2-methyl-5-norbornene-2-carboxylic acid, bp. 126-7° at 5mm. Separation of epimeric carboxylic acids" To a 2000ml round-bottomed flask equipped with an addition funnel and magnetic stirring bar was added 1709 (1.1mol) 2-methyl-5-norbornene-2-carboxylic acids and an equimolar amount of sodium bicarbonate in a minimum amount of water such that a homogeneous solution was obtained. To this magnetically stirred solution was added drOpwise a stock solution of iodine and potassium iodide (300g I 2 and 6009 KI in 1800ml water). The addition was continued 16 17 until a yellow color persisted. The precipitate which formed was filtered and washed with bicarbonate solution and dried. Nearly colorless crystals weighing 98.39 (0.355mol), mp. 84-6° without recrystallization, were obtained. The filtrate was carefully acidified by drop- wise addition of dilute hydrochloric acid with stirring; the pH was monitored with pH paper and was taken to ca. 3-4. Extraction with ether followed by washing the com- bined extracts with water and then saturated salt solution, drying over anhydrous magnesium sulfate, and evaporation of the solvent yielded 103g (0.68mol) endgfz- methyl-5-norbornene-gng2-carboxylic acid. The product was twice recrystallized from aqueous acetic acid to mp. 80-3°. The iodolactone obtained previously was reduced 28 Thus, to a lOOOml round-bottomed flask with zinc dust. equipped with a reflux condenser and magnetic stirring bar was added 98.3g (0.354mol) iodolactone and 400ml glacial acetic acid. To this magnetically stirred solu- tion was added 1209 (1.859-at) zinc dust in portions; heat was evolved after a short induction period. After 3hrs. the mixture was filtered and the solid residue washed with 500ml hot water. The filtrates were diluted with 1500ml water and then extracted with ether. The ether extracts were combined and washed with 5% sodium bicarbonate solution; these washingsawere acidified in we the manner described for the exo-acid. This procedure 18 yielded 40.99 solid material which was recrystallized from aqueous ethanol; recovered yield was 32.69 (0.214mol, 61% theoretical amount), mp. 86-98°. Vacuum (12mm) sub- limation of this material yielded 29.89 pure egg-2- methyl-5-norbornene-gndng-carboxylic acid, mp. 107-109°. The identity of each pure epimer was confirmed by its NMR spectrum. Preparation of ngrZ-methyl- gage;2-hydroxydideuteriomethyl- 5-norbornene To a thoroughly dried 500ml three-necked round- bottomed flask equipped with reflux condenser, addition funnel and mechanical stirrer was added 200ml anhydrous ether (dried by distillation from lithium aluminum hyd- ride) and 5.09 (0.12mol) lithium aluminum deuteride (Merck, Sharp and Dohme, 99%). To this stirred suspen- sion was added dropwise a solution of 24.39 (0.16mol) 2x2:2-methy1-5-norbornene-gnggf2-carboxylic acid dis- solved in a minimum amount of anhydrous ether. This mix- ture was then heated at reflux for two days. After cooling, the mixture was treated successively with 100ml water and 100ml dilute hydrochloric acid. The ether layer was separated and the aqueous layer extracted with fresh ether; then the combined ether extracts were washed with 10% sodium hydroxide, water, and saturated sodium chloride, then dried over anhydrous magnesium sulfate. 19 Removal of solvent by distillation left a residue weighing 17.29. On the basis of the amount of unreduced acid re- covered from the basic wash solution above, the product yield was 13.09 (.093mol, 58% theoretical amount). This product was used in the next step without further purifi- cation. Preparation of endo-2-methyl-exgr 2-hydroxydideuteriomethyl-5- nofbornene By the same procedure as described for the epi- meric compound, 18.89 (0.134mol, 84% theoretical amount) endo-2-methy1-exo-2-hydroxydideuteriomethyl-5-norbornene was obtained. Preparation of errZ-methyl-endgr 2-tosyloxydideuteriomethyl-5- norbornene To a 500ml round-bottomed flask fitted with a drying tube was added 17.29 (0.12mol) gngZ-methyl-endgf 2-hydroxydideuteriomethyl-5-norbornene and 100ml dry pyridine. The solution was cooled in an ice bath, then 24.89 (0.13mol) p-toluenesulfonyl chloride was added in two equal portions. The mixture was kept at 0° until pyridine hydrocholoride began to precipitate. Then the reaction mixture was transferred to the refrigerator (ca. 3°) for two days. The product was isolated by the addi- tion of ice chunks to the reaction flask. The white solid which formed was taken up in ether and the ether 20 solution extracted with 10% hydrochloric acid until the odor of pyridine was no longer detectable in the neutralized extract. After a final washing with water and saturated sodium chloride, the product solution was dried over anhydrous magnesium sulfate. Evaporation of the solvent yielded a crystalline solid weighing 28.39. Theoretical yield was 27.39. This material was used in the next step without further purification. Preparation offingggz-methyl-gxg- 2:tosyloxydideuteriomethyl- 5-norbornene A procedure similar to that described for the pre- paration of the epimeric compound yielded 38.39 (0.13 mol, 97% theoretical amount) tosylate from 18.89 (0.134 mol) endo- 2-methy1-exo-2-hydroxydideuteriomethyl-5-norbornene. Preparation of gnfig;§f trideuteriomethyl-gxg-6- methyl-Z-norbornene To a thoroughly dried 250ml round-bottomed flask equipped with reflux condenser, addition funnel, and mecha- nical stirrer was added 100ml dry tetrahydrofuran (distilled from lithium aluminum hydride) and 1.5g (0.036mol) lithium aluminum deuteride (Merck). To this stirred suspension, which was protected from moisture by a drying tube, was added dropwise a solution of 27.39 (o.093mol) gngz- methyl-2392f2-tosyloxydideuteriomethyl-5-norbornene dissolved in 25ml dry tetrahydrofuran. The reaction 21 mixture was then heated at reflux for two days. At this time an additional 2.69 (0.062mol) lithium aluminum deu- teride was added and heating at reflux was continued for another four days. After cooling, the mixture was treated with water; after the initial reaction had subsided, the mixture was added to ca. 2000ml water made slightly acidic by the addition of dilute hydrochloric acid. After the resulting solution was extracted with ether, the combined extracts were washed with fresh water, then saturated sodium chloride, and dried over anhydrous magnesium sul- fate. Slow distillation thru a short Vigreux column yielded 9.39 colorless liquid which was about 60% desired hydrocarbon and 40% l-butanol (identified by mixed injec— tion VPC). Passage of this mixture thru an alumina column (activated, chromatographic grade, 80-200 mesh, Matheson) by using pentane as eluant afforded a clean separation of the hydrocarbon product. Distillation yielded 4.49 (0.035mol, 37% theoretical amount) endng-trideuteriomethyl- gig:6-methyl-2-norbornene, bp. 131-2° at prevailing atmos- pheric pressure. The identity of the product was confirmed by NMR; no contamination by the epimeric hydrocarbon was detectable. 22 Preparation of exo-6-trideuteriomethy1- endo-6-methyl-2-norbornene A procedure similar to that described for the pre- paration of the epimeric compound yielded 6.59 (0.052mol, 40% theoretical amount) 2x276-trideuteriomethy1-gpdgf6- methyl-Z-norbornene, bp. 13l-2°, from 38.39 (0.13mol) 939272-methy1-§x972-tosyloxydideuteriomethyl-5-norbornene r. treated with 5.79 (o.l36mol) lithium aluminum deuteride in tetrahydrofuran at reflux over a period of seven days. 5 Again there was no contamination by the epimeric hydro- carbon as shown by NMR. Preparation of exgr2,3-epoxy-§ndgr 6-trideuteriomethy1-gx9-6- methylnorbornane‘yI To a 250ml three-necked round-bottomed flask equipped with reflux condenser, addition funnel, thermo- meter, and magnetic stirring bar was added 4.49 (0.035mol) 2292f6-trideuteriomethyl-gng6-methy1-2-norbornene and 50ml methylene chloride. Toithis magnetically stirred solution was added dropwise a solution of 7.09 (0.040mol, 8.29 of Aldrich 85% prue) m-chloroperbenzoic acid dis- solved in 100ml methylene cthride. The temperature during the addition was maintained (25°. After stirring for 2hrs., the mixture was treated with 10% sodium sul- fite solution until a test with starch-iodide paper was negative. The precipitated solid m-chlorobenzoic acid was filtered and the filtrate was washed with 5% sodium 23 bicarbonate, water, and then saturated sodium chloride. After drying over anhydrous sodium sulfate and removal of the solvent by evaporation, the theoretical amount of desired epoxide was obtained. This material was used in the next step without further purification. Preparation of exp-2,3-epoxyéexgr ' fr 6-trideuteriometHyl—endo-6- methylnorbornane Treatment, similar to that described for the pre- paration of the epimeric compound, of 6.59 (0.052mol) exo-6-trideuteriomethyl-endo-6-methyl-2-norbornene with 10.49 (0.060mol, 12.29 of Aldrich 85% pure) m-chloroper- benzoic acid again yielded the desired epoxide quantitatively. Preparation of gagg:§f trideuteriomethyl-exo-6- methyl-exg-Z-norbornanol 30 To a 250ml three-necked round-bottomed flask equipped with reflux condenser, addition funnel, and mechanical stirrer was added 100ml N-methylmorpholine (Aldrich, bp. 118-9° and 2.79 (0.070mol) lithium alumi- num hydride; this stirred suspension soon became quite thick and paste-like in appearance. Then, with continued stirring, was added dropwise a solution of 4.99 (0.035mol) corresponding epoxide dissolved in 25ml N-methylmorpholine. The mixture was heated at reflux for 62hrs. After cooling, the mixture was hydrolyzed by the cautious addition of 24 water. This mixture was then transferred to a 1000ml three-necked round-bottomed flask equipped with a dis- tillation head, a steam inlet tube and a glass stopper. Steam distillation was continued until 500ml of distil- late had been collected. This distillate was then extracted with.ether and the combined extracts were washed with water, 10% hydrochloric acid, again with water, and then dried over anhydrous sodium sulfate. Distillation of the solvent afforded 4.49 of residue which contained about 90% desired alcohol as determined by VPC. Therefore the actual yield was ~4.09 (0.028mol, 80% theoretical amount). This material was used in the next step without further purification. Preparation of exo-6- trideuteriomethyl-eggp-6- methyI:exo-2-norbornano Reduction,“similar to that described for the I epimeric compound, of 7.39 (0.052mol) exo-2,3-epoxy-exo- 6-trideuteriomethyl-endo-6-methy1norbornane with 3.89 (0.10mol) lithium aluminum hydride yielded 6.79 (0.047mol, 90% theoretical amount) desired alcohol. Preparation of endg-G- trideuteriomethy -e§g-6- methyl-Z-norbornanone A stock solution was prepared by diluting a mix- ture of 109 (0.0336mol) sodium dichromate dihydrate and 7.50ml (0.134mol) concentrated sulfuric acid to 50ml with 25 water. To a 100ml three-necked round-bottomed flask equipped with reflux condenser, small addition funnel, and thermometer was added 4.49 (0.031mol) of the corres- ponding engZ—norbornanol material (ca. 90% pure) dis- solved in 20ml ether. To this magnetically stirred solution was added 15.0ml (contains 0.010mol sodium dichromate) oxidizing solution at a rate such that the temperature was maintained <30°. After continued stirring for 2hrs. at room temperature, the ether layer was sepa- rated and the aqueous layer extracted with fresh ether. The combined ether extracts were washed with water, saturated sodium bicarbonate, again with water, and then dried over anhydrous sodium sulfate. After removal of the solvent by careful distillation thru a short Vigreux column, the residue was distilled thru a short-path con— denser under reduced pressure. ’A clear colorless liquid weighing 2.019 was collected over the range 75-100° at 12mm. This material was analyzed by VPC and was found to contain about 59% (1.29, 8.5mol, 31% theoretical amount) of the desired ketone. The major impurity was apparently unoxidized alcohol as indicated by VPC re- tention times. The material was used in the next step without further purification. 26 Preparation of exgr6- trideuteriomethyl-endo-6- meEhyIF2—norbornanone Treatment, similar to that described for the pre- paration of the epimeric compound, of 7.49 (0.052mol) corresponding.exefZ-norbornanol material (90% pure) with 25.4m1 (contains 0.017mol sodium dichromate) oxidizing solution yielded 3.279 product which was collected over the range 74v98° at 11mm. This material was about 70% (2.39, 0.016 mol, 34% theoretical amount) desired ketone. Again the major impurity was unoxidized alcohol, and the material was used without further purification. Preparation of 2,6-dimethyl-endgr 6-trideuteriomethyl-endo- 2-norbornanol To a 50ml two-necked pear—shaped flask equipped with reflux condenser, serum cap, and magnetic stirring bar, was added 0.359 (0.014g-at) magnesium metal turn- ings. After the flask had been flame-dried, protected from moisture by a drying tube, and cooled, 15ml of anhydrous ether and 0.9ml (2.049, 0.014mol) methyl iodide (Aldrich) were added and allowed to react. After stirring at room temperature for 0.5hr., the reagent was treated with 0.59 of the corresponding ketone solution (about 0.39, 2.1mmol of egdef6-trideuteriomethyl-eng6-methy1-2- norbornanone); this mixture was heated at reflux over- night. After 18hrs. the mixture was allowed to cool and 27 was hydrolyzed by the addition of saturated ammonium chloride solution. The two clear phases that were ob- tained were separated and the aqueous phase extracted with fresh ether. The combined extracts were washed with 20% sodium thiosulfate, water, and saturated sodium chloride, then dried over anhydrous sodium carbonate. . r: The solvent was removed by slow distillation thru a 16cm. Vigreux column. The residue was transferred to a small sublimation apparatus and heated at 80° under a pressure of 12mm. The solid obtained was recrystallized from pentane and allowed to dry. The yield was 0.229 (1.4mmol, 66% theoretical amount) desired tertiary alcohol, mp. 77-81°. The alcohol was not further purified before conversion to its p-nitrobenzoate. Preparation of 2,6-dimethy1-exgr 6-trideuteriomethyl-e§d9-2-norbornanol Treatment, similar to that described for the pre- paration of the epimeric compound, of 0.59 of correspond- ing ketone material (about 0.359, 2.5mmol of exef6- trideuteriomethyl-epggf6-methy1-2-norbornanone) with 0.014mol methyl Grignard reagent yielded 0.249 (1.5mmol, 60% theoretical amount) desired tertiary alcohol, mp. 69-75°. Again this material was not further purified before conversion to the p-nitrobenzoate. 28 Preparation of 2,6-dimethy1-endo- 6etrideuteriomethyl-endo-2ip norbornyl p-nitrobenzoate 2 To a 50ml two-necked pear-shaped flask equipped with reflux condenser, nitrogen inlet, serum cap, and magnetic stirring bar, was added 0.229 (1.4mmol) tertiary alcohol and 5ml anhydrous ether. Then by syringe was added l.0m1 (1.6mmol) of n-butyllithium (1.6M solution in hexane, Foote Mineral Co.) and the reaction mixture stirred at room temperature under nitrogen for 1.5hrs. Meanwhile, to a 50ml three-necked round-bottomed flask equipped with reflux condenser fitted with a nitrogen inlet, serum cap, glass stepper, and magnetic stirring bar, was added 0.279 (1.4mmol) p-nitrobenzoyl chloride (recrystallized from ligroine immediately prior to use) and 5m1 ether. Then by syringe was added the previously formed lithium alkoxide under an atmosphere of nitrogen. The reaction mixture was allowed to stir at room tempera- ture for 4 hrs.; then the product was isolated by pouring the mixture into 10% sodium carbonate and separating the ether layer. After washing with saturated sodium chloride, the ether solution was dried over anhydrous sodium sul- fate. After removal of the solvent by evaporation, the residue was induced to crystallize by cooling to ca. -25°. The yellow crystals were collected and dried; the product weighed 0.1489 and melted over the range 103-13°. 29 After ten successive recrystallizations from pentane at 0°, 0.0619, mp. ll9-20°, was obtained. The NMR of this compound was obtained and compared to that of the un- labeled material; with the exception of the absence of the signal for the egdng-methyl group, the spectra were identical. Preparation of 2,6-dimethy1-exg- 6-trideuteriomethyl-endgr2— norbornylepénitrobenzoate Treatment, similar to that described for the pre- paration of the epimeric p-nitrobenzoate, of 0.249 (1.5 mmol) of the appropriate tertiary alcohol with 1.1m1 (1.7mmol) n-butyllithium solution, followed with 0.299 (1.53mmol) p-nitrobenzoyl chloride and the usual isola- tion procedure yielded 0.1239 p-nitrobenzoate, mp. 120-2°. Again the NMR of this material showed that the eégf6- methyl group was specifically and completely deuterated. Preparation of unlabeled compounds Instead of the 2-methyl-5-norbornene-2-carboxylic acids, the corresponding methyl esters were prepared in 55% yield by Diels-Alder condensation of cyclopentadiene with methyl methacrylate (Eastman). An excess of lithium aluminum hydride in those steps requiring the deuterated reagent in the preparation of the labeled compounds, allowed the synthesis of the 6,6-dimethyl-2-norbornene 30 in 69% yield from the initially prepared bicyclic methyl esters. Subsequent preparations were performed in the same manner as previously described for the labeled com- pounds; similar yields were also obtained. The end product, 2,6,6—trimethy1-egdef2-norborny1 p-nitrobenzoate, melted at 120-l°. II. Solvolysis Product Study Examination of products generated under solvolytic COHditIOns To a thick-walled glass ampoule was added 0.19 2,6,6—trimethy1—epdgf2-norborny1 p—nitrobenzoate dis- solved in about 5m1 warm ethanol and about 45ml of 80% aqueous ethanol. The ampoule was sealed and the con- tents heated at 150° for 20hrs. After cooling, the ampoule was broken and the contents added to 300ml saturated sodium chloride solution containing 0.029 sodium carbonate. The mixture was extracted with several portions of pentane. The combined extracts were washed with water and then dried over anhydrous sodium carbonate. The solvent was removed by slow fractionation and the residue was examined by VPC (column: 25% SE-52, Chrom. W— AW/DMCS; 70/80 mesh; 6'-7"x 1/4"; temperature: 111°; carrier gas: helium, 25ml/min). 31 Examination of the stability_of 2,6,6-trimethy1- nd eZ—norbornanol in theepresence oIp-nifrobenzoic acid To another ampoule was added about 45ml of 80% aqueous ethanol and 0.059 2,6,6-trimethy1-egdgfz- norbornanol with 0.0559 p-nitrobenzoic acid (recrystal- lized from water). The sealed ampoule was heated at Ff 150-5° for 18hrs. After treatment as described above, the product.mixture was again examined by VPC under con- ditions identical to those described above. III. Kinetics Constant temperature bath A Sgal. glass (Pyrex) jar was placed in a wooden (3/4" plywood) box large enough to accommodate the jar and allow about 1 1/2" between the jar and the box walls on all sides except the top. This space around the jar was filled with vermiculite; the top was covered with a sheet of 1/2" asbestos. The most suitable bath medium was found to be HTF-lOO Ucon Fluid (available from Matheson Scientific) which is a mixture of polyalkylene glycols. This medium was heated to within about 2° of the desired operational temperature by means of a 500W heater. The heater was operated continuously at an appro- priate voltage that was controlled by means of a Variac. The medium was stirred rapidly by a 2 1/2" aluminum 32 propeller with 20 blades which was positioned near the bottom of the bath and was driven by a Talboys, No. 104, 1/18 H.P. electric motor. It was necessary to cool the rheostat with a continuous stream of air so as to prevent excessive heating. Excellent temperature control was maintained using a mercury filled temperature regulator (-35 to 500°F., Micro-Set, Precision No. 62541) in con- junction with a knife heater (125W, Cenco) and an elec- tronic buffer relay (Precision No. 62690). The voltage supplied to the intermittent heater was adjusted so that heating and non-heating periods were of approximately equal duration. Temperature changes were observed with a Beckmann differential thermometer; over the range 100- 160°, bath temperatures were found to vary within the limits of :0.010°. Sample ampoules and holder Individual ampoules were made from 1/4" regular wall Pyrex tubing; each was about 3 1/4" in length and had a neck constriction to facilitate subsequent opening. A wire basket was constructed from wire mesh such that a maximum of 45 of these ampoules could be simultaneously introduced into the bath thru an opening in the asbestos top. The basket was made deep enough to allow the am- poules to set below the surface of the bath top so that a cover could be placed over the opening. 33 Calibration of Beckmann differential‘thermometer The actual temperatures corresponding to each of the settings of the Beckmann differential thermometer were determined with either a Hewlett-Packard Quartz Thermo- meter or a precalibrated platinum resistance thermometer. Preparation of solvents fleEe£.--House-supplied distilled water was re- distilled in an all-glass apparatus. Ethanol.--Absolute ethanol was prepared by treating the commercial material with sodium metal and diethyl succinate as described by Fieser.33. The anhy- drous material was distilled thru a 30cm. Vigreux column. Mixed solvent.--The solvent used for the kinetic runs was 80/20 ethanol-water (V/V). It was prepared in bulk by weighing the appropriate quantities as calculated from densities at 25°. Determination of Beer's Law p-nitrobenzoic acid.--A series of solutions of p-nitrobenzoic acid (recrystallized from water) in 80% ethanol-water was prepared by weighing and by the appro- priate dilution procedures. Then a 1.0 ml sample of each of these was transferred to a 10 ml volumetric flask con- taining 1.0 ml of sodium hydroxide solution (0.1M). Each sample was then diluted to 10 ml with 95% ethanol and 34 the absorbance determined on a Unicam SP.800 spectro- photometer. 2,6,6-trimethyl:e§gg;2-norbornyl p-nitrobenzoate.-- A similar series of solutions of this p-nitrobenzoate (re- crystallized from hexane) in 80% ethanol-water was pre- pared and treated as described above. The absorbances were determined in the same manner. Additionally, a 1.0 ml sample of each of the solutions was added to 1.0 ml of 10% sodium hydroxide solution in a small separatory funnel. 10 m1 of cyclohexane (Spectro-Grade) was added and the mixture was shaken. After settling of the layers, a sample of the cyclohexand portion was withdrawn and its absorbance determined on a Cary 14 spectrOphotometer. Kinetics of solvolysis To each of 10 to 14 ampoules was added about 2 m1 of substrate solution (about 8.0 x 10-4 M, typically 0.0619/250ml) by using a syringe. After all the ampoules had been filled, each was successively cooled in ice water while it was sealed with a gas torch. The ampoules of a given run were placed in the wire basket such that a maximum amount of space was left between them; the basket was then placed in the bath. At this point, the bath temperature decreased rapidly and the 500W heater was ad- justed so as to minimize the temperature drop; after the original temperature had again been attained (after 35 3-5min) the voltage supplied to the heater was returned to its original value. After 40 to 65 min. allowed for equilibration, the first ampoule was withdrawn and im- mediately quenched in an ice-water mixture with rapid shaking. This was taken as time zero and subsequent ampoules were removed and quenched in the same manner at appropriate time intervals. The ampoules were stored at -5° until the run was completed, when they were col- lectively allowed to warm to room temperature in a water bath. Each ampoule was broken Open and 1.00m1 was with- drawn by syringe. This aliquot was treated by one of the following: Simple dilution procedure.--The sample was added to a 10 ml volumetric flask containing 1.0 ml of sodium hydroxide solution (0.1M). After at least 10min, this mixture was diluted to volume with 95% ethanol. The ab- sorbance of this solution was measured on a Cary 14 spectrOphotometer against a solvent blank prepared in the same manner. The measurements were performed at various wavelengths around the absorption maximum for the ester (around 256-257nm). Extraction procedure.--The sample was introduced into a 60ml separatory funnel containing 10ml of 10% sodium hydroxide solution. After having set for about 10min, this mixture was shaken for 1min. with 10ml of cyclohexane (Spectre-Grade) which was added by pipette. 36 After 5 min., the aqueous layer was drawn off and the cyclohexane layer allowed to stand an additional 3 min. Then a sample was withdrawn and transferred to a cuvette wherein the absorbance was measured against a solvent blank prepared in exactly the same manner. The absorb- ance of each sample was determined at four separate wavelengths (250, 255, 260, and 265nm) on a Cary 14 fi_fi spectrophotometer. - Thus from each kinetic run, at least four values 5 for the first order rate constant were determined from the appropriate kinetic expression. The average of these values was taken as the rate constant for that run. RESULTS AND DISCUSS ION In order to probe the origin of kinetic isotope effects, it is first necessary to establish the exact nature of the kinetic process itself. For this reason, an analysis of the solvolysis products from 2,6,6-tri- methyl-egdefZ-norbornyl p-nitrobenzoate in 80% ethanol- water at 150° was performed. The results in Table 3 show that 68% of the products formed are olefinic and 32% are substitution. These compounds were identified by their VPC retention times determined by the mixed injection technique. The production of 32% rearranged alcohol was interesting, although not surprising, in view of the results obtained when 2,6,6-trimethy1-egggf 2-norbornanol was treated with formic acid at room temperature.34 Under these conditions a 90% yield of the l,6,6-trimethy1-exef2-norbornyl formate was obtained presumably through a carbonium ion rearrangement mecha- nism. Therefore, all the solvolysis products may be reasonably assumed to germinate from the 2,6,6-trimethy1- 2-norborny1 cation, or ion pair.* *Bimolecular elimination from tertiary sub- strates under solvolytic conditions have been shown to be negligible.16br35 37 38 However, the possibility that the kinetic pro- cess might have produced the 2,6,6-trimethy1-epg272- norbornanol which subsequently, under the unbuffered conditions of the solvolysis, may have undergone further reaction to yield the observed products could not be dis- missed. Such instances of acyl-oxygen cleavage under solvolytic conditions have been observed previously with rather unreactive p-nitrobenzoates. Thus Bunton, e£_el.,36 found the} a comparison of the ionization rates of iso- bornyl and 2-methy1isoborny1 p-nitrobenzoates in 80% ethanol-water was frustrated by the fact that the iso- bornyl system underwent acyl-oxygen cleavage. Goering and Closson37 have suggested that cyclodecyl p-nitroben- zoate solvolyzes by acyl-oxygen cleavage in 90% acetone- water at 119°. However, from the results collected in Table 4, it may be concluded that if acyl-oxygen cleavage occurs at all in the present system, it does so at a rate which is at least thirty times slower than the rate of ionization at 150°. This is evidenced by the fact that after ten solvolytic half-lives, 83% of the 2,6,6-trimethy1- epdng-norbornanol is recovered unchanged. Therefore any isotope effects observed in this investigation may safely be attributed to the ionization process itself or some other process involving the corresponding ion pair. The suitability of ultraviolet spectrophotometry as the analytical tool for this investigation was 39 Table 3. Products from the solvolysis of 2,6,6-trimethyl- endo-2-norbornyl p-nitrobenzoate in 80% ethanol-water (V/V) at 150°. Product Retention Time(min.) %(norma1ized) A_ 3.5 7.7 E. 5.2 60.4 E 20.4 31.8 I A’= l,6,6-trimethy1-2-norbornene g = 2,6,6-trimethy1-2-norbornene and 6,6-dimethy1-2- methylenenorbornane (roughly a 40:60 mixture) Hill I - l,6,6-trimethy1-exo-Z—norbornanol PC analysis conditions: column- 25% SE-52 (Chrom.W- AW/DMCS), 70/80 mesh, 6'-7"x l/4", 111°; carrier gas- He @ 25m1/min. < confirmed by Beer's Law determinations for the p-nitro- benzoate ester in cyclohexane as well as in 95% ethanol and for the p-nitrobenzoate anion in 95% ethanol. The determinations were made at two different wavelengths in 95% ethanol and at four in cyclohexane. The results of the measurements are collected in Tables 5, 6, and 7; typical plots of the data are shown in Figures 2 and 3. No systematic deviations from Beer's Law were apparent over the concentration range up to 1.2 x 10-4M for the ester in cyclohexane. Similar results were obtained for the p-nitrobenzoate anion in 95% ethanol. However, sub- stantial negative deviations from Beer's Law were 2...; b {LGJDHw-ZDZJJ 40 Table 4. Products from the treatment of 2,6,6- trimethyl-endo-2-norbornanol with p-nitro- benzoate acid under solvolytic condi- tions at 150-5° for the equivalent of 10—11 half-lives. Product Retention Time (min.) %(normalized) 5. 3.5 2.7 B 4.1 1.3 g 5.2 2.0 B 14.6 82.9 E 20.7 11.1 A_= 1,6,6-trimethy1-2-norbornene B = unidentified g = 2,6,6-trimethyl-2-norbornene and 6,6-dimethy1-2- methylenenorbornane g = 2,6,6-trimethy1-endo-2-norbornanol E = 1,6,6-trimethy1-exo-2-norbornanol VPC analysis conditions: column- 25% SE-52(Chrom.W- AW/DMCS), 70/80 mesh, 6'-7"x 1/4", 111°; carrier gas- He @ 25m1/min. Table 5. 41 Beer's Law determination for 2,6,6-trimethy1- endo-Z-norbornyl p-nitrobenzoate in 95% ethanol. Molar Concentration Absorbance at 5 (x 10 ) 255nm 258nm 13.0 1.541 1.575 10.4 1.355 1.379 6.5 0.902 0.922 5.2 0.727 0.741 2.6 0.373 0.372 1.3 0.200 0.201 8255 = 14,000 8258 = 14,300 Table 6. Beer's Law determination for 2,6,6-trimethy1- endo-2-norbornyl p-nitrobenzoate in cyclohexane. Molar Concentration Absorbance at (x 105) 250nm 255nm 260nm 265nm 14.4 1.848 --- --- 1.822 12.0 1.621 1.772 1.756 1.567 9.6 1.283 1.405 1.397 1.244 7.2 0.990 1.082 1.070 0.950 4.8 0.619 0.677 0.673 0.602 2.4 0.334 0.360 0.357 0.318 6250 = 13,500 8255 = 14,600 5260 = 14,600 5265 = 13,000 42 Table 7. Beer's Law determination for sodium p-nitro- benzoate in 95% ethanol. Molar Concentration Absorbance at (x 105) 255nm 258nm 16.8 1.267 1.343 13.4 1.071 1.142 8.4 0.662 0.722 6.7 0.532 0.572 3.4 0.281 0.298 1.7 0.151 0.160 8255 = 7900 8258 = 8900 Absorbance at 255nm 0.5 — Figure 2. 1.0 — 43 / /®’ / / / .6/ / / / / / ., (D = 2,6,6-trimethy1-endo-2-norbornyl p-nitrobenzoate sodium p-nitrobenzoate I I 5.0 10. Molar Concentration x 10 E] n Beer's Law plots of 2,6,6-trimethy1-endo-2- norbornyl p-nitrobenzoate, and sodium p-nitro- benzoate, in 95% ethanol. Absorbance at 255nm Figure 3. 1.5 44 I I 5.0 l .0 Molar Concentration x 10 Beer's Law plot of 2,6,6-trimethyl-endo-2- norbornyl p-nitrobenzoate in cyclohexane. 45 encountered with the p-nitrobenzoate ester in 95% ethanol at concentrations greater than about 8.0 x lo-SM. This may have been a reflection of the limited solubility of the ester at higher concentrations. The calculated molar absorptivities (s) ranged from 13,000 to 14,600 for the ester in either solvent and 7900 to 8900 for the acid anion in 95% ethanol. Thus the kinetics of solvolysis of the 2,6,6-tri- methyl-endo-Z-norbornyl p-nitrobenzoates (Ia-c) in 80% ethanol-water were determined by using the ampoule I . _ ._ Ia. R — R —CH3 (do) R ‘ . CH Ib: R = CH3; R'=CD3 (endo-6-CD3) _ . I: _ - R' OPNB R — CD3, R CH3 (exg 6 CD3) technique and analyzing the contents of each ampoule by measuring its absorbance at four different wavelengths. In this system, an absorbing ester solvolyzes to yield an absorbing acid among the products. The observed de- crease in absorbance with time may be treated by the 38 scheme developed by Roy wherein allowance is made for the generation of an absorbing product. Thus, if 8e = molar absorptivity of ester at given wavelength Ea = molar absorptivity of acid at given wavelength At = absorbance of sample at time t A = absorbance of sample at time = 0 46 Am = absorbance of sample after 10-11 half-lives a = ester concentration at time 0 ce = ester concentration at time t ca = acid concentration at time t then At = case + caea = c 8e + (a-ce)ea c = A -a€a e e -e a kt = 1n g1 e = 1nace-a8a At-aea but as = A e o and as = A a 00 Therefore kt = ln(Ao-Am) - 1n(At-Am) Unfortunately, the absorption spectra of the ester and the acid are similar, such that a direct analysis of the solvolysis mixture is impossible. However, the spectrum of the acid anion is substantially different from that of the ester and thus addition of dilute solutions of sodium hydroxide to each sample would allow the deter- mination of the set of points comprising a kinetic run. The reaction between the unsolvolyzed ester and the added base was shown to be negligibly slow at room temperature, as the absorption spectrum of the pure ester 47 remained unchanged upon standing for 1hr. in the presence of a 10-fold molar excess of base. The average time re- quired for analysis of a particular kinetic sample was 20min. During this study it was noted, however, that some of the later samples taken during a kinetic run did indeed exhibit a small decrease in absorbance with time upon standing with base. The production of small amounts of ethyl p-nitrobenzoate from a reaction between the solvent and the liberated acid was suspected. When a sample of a solution of pure p-nitrobenzoic acid in 80% ethanol-water was exposed to the conditions of the solvoly- sis for the equivalent of twenty half-lives, the initial absorbance of that base-treated sample was indeed observed to decrease to an equilibrium value identical to that of a base-treated sample of the original solution. Therefore, in all subsequent analyses a 10 to 15min. period was allowed for the base treatment of each sample before dilu- tion and measurement. An extensive series of kinetic measurements was initially carried out using the dilution procedure de- scribed in the EXPERIMENTAL section of this thesis. The collected data for each run were fitted to the kinetic equation derived previously: kt = 1n (AC-Am) - 1n (At-Aw) 48 by a non-linear least-squares computer program provided by Professor J. L. Dye of this department (see Ref. 39). The results presented in Table 8 are typical of those obtained using this procedure. Three factors are particularly noteworthy among these results. Firstly, it may be seen that the maximum change in absorbance, i.e., the difference between the absorbances.of the initial and the final kinetic points, is only 0.262 units. Secondly, it is apparent that the errors in the individual rate constants are on the order of 2.5%, a value approaching that of the unacceptable in view of the small differences in rates which are being measured. Finally, it is obvious that the precision with which the infinity point may be measured by this technique leaves much to be desired. Thus, from the same kinetic points two infinity values which differ by only 0.006 absorbance units (see for example the data taken at 265nm) yield two rate constants which differ by 4.4%! Clearly, this technique depends too heavily on the ap- parently unreliable value for the absorbance at infinite time. A dependable alternative to this procedure was developed and carried out as described in the EXPERIMENTAL section of this thesis. This extraction technique afforded data which are typically represented by those of Table 9. Ironically, the three factors mentioned which render the Table 8. 49 Data for the solvolysis of 2,6,6-trimethy1-endo- 2-norborny1 p-nitrobenzoate in run 35 at 149.790° in 80% ethanol-water (V/V). Time (min.) Absorbances at 250nm 255nm 260nm 265nm 0 0.766 0.892 0.978 0.991 35 0.739 0.861 0.948 0.967 70 0.718 0.836 0.922 0.948 111 0.701 0.813 0.902 0.928 155 0.667 0.779 0.867 0.901 200 0.638 0.742 0.830 0.868 251 0.611 0.716 0.804 0.850 310 0.594 0.694 0.783 0.833 372 0.577 0.673 0.762 0.816 440 0.558 0.657 0.748 0.802 515 0.544 0.640 0.731 0.789 582 0.538 0.630 0.719 0.776 3033 0.477 0.561 0.656 0.725 3033 0.480 0.569 0.661 0.731 Wavelength k x 105 Average k x 105 (nm) (sec-1) (sec-1) 250 4.791 i 0.120 255 4.783 i 0.106 260 4.893 1 0.117 from first w value: 265 4.819 1 0.111 4.822 :_0.ll4* 250 4.889 : 0.122 255 5.017 : 0.108 260 5.046 : 0.119 from second w value: 265 5.038 t 0.116 4.998 :_0.ll8* *This error is the average deviation. Table 9. 50 Data for the solvolysis of 2,6,6-trimethy1-endo- 2-norborny1 p-nitrobenzoate in run 59 at 150:186° in 80% ethanol-water (V/V). Time(min.) Absorbances at 250nm 255nm 260nm 265nm 0 0.862 0.955 0.956 0.856 30 0.790 0.876 0.874 0.782 60 0.720 0.793 0.793 0.712 100 0.638 0.707 0.708 0.632 140 .0.560 0.622 0.621 0.557 190 0.472 0.523 0.523 0.467 241 0.409 0.454 0.454 0.407 307 0.321 0.359 0.359 0.322 360 0.277 0.303 0.305 0.273 430 0.223 0.249 0.251 0.225 500 0.176 0.197 0.200 0.182 570 0.140 0.158 0.160 0.144 Wavelength k x 105 Average k x 10 (nm) (sec-l) (sec- 250 5.288 : 0.035 255 5.266 : 0.034 260 5.243 i 0.032 265 5.232 1 0.033 5.257 i_0.034* *This error is the average deviation. 51 former method unreliable, make the new procedure extremely attractive. Thus the maximum change in absorbance is now 0.798 units, more than three times larger than that by the former technique; the errors in the rate constants are only 0.7%, almost four times smaller than previously; and finally, and perhaps most importantly, the determination of infinity absorbance values is eliminated. In control experiments, it was determined that the experimental in- finity absorbances measured after ten or eleven solvoly- tic half—lives were all zero within the limits of the reproducibility of the instrument zero. Therefore, all of the subsequently discussed results are based on the data determined by this latter technique. The rate constants collected in Table 10 were determined from the rate of absorbance decrease over a period of 2 to 2 1/2 half-lives in most cases. However, the runs conducted at 100° were followed for only 0.6 ?; half-life since one half-life at this temperature was E twenty days! The individual rate constants in the table I i are the average of the values determined at four wave- Tm, lengths for each kinetic run. If any one of these four values appeared inconsistent with the other three, the 4d Rule40 was applied. That is, the average of the three unquestioned values was used to calculate the deviation of each of the four original values. The questionable value was retained only if its deviation was within four Table 10. 52 Rates of solvolysis of 2,6,6-trimethy1-endo- 2-norborny1 p-nitrobenzoates in 80% ethanol- water (V/V). 6 Averag k x Run Temp,°C. Substrate k x E? 102: (sec ) (sec ) 78 100.840 d .3878 —-o 80 100.840 do .3920 82 100.840 do .3892 .3897 i .0037 79 100.840 endo-6-CD3 .3857 83 100.840 endo-6-CD3 .3962 .3910 i .0052 81 100.840 exo-6-CD3 .3853 .3853 i .0044 52 125.217 d 5.033 —o 54 125.217 d 4.985 _0 57 125.217 do 5.084 5.034 : .050 58 125.217 endo-6-CD 4.965 5.018 : .052 53 125.217 exo-6-CD3 4.922 55 125.217 exo-6-CD3 4.983 4.952 : .044 71 142.244 d 25.42 _-0 74 142.244 do 25.09 25.26 i 0.22 72 142.244 endo-6-CD3 24.65 75 142.244 endo-G-CD3 24.52 24.58 i 0.23 73 142.244 exo-6-CD3 24.03 76 142.244 exo-6-CD3 24.39 24.21 i 0.21 average, or the average error of the individual rate con- stants of a given set, whichever is larger. *Errors quoted are maximum deviations from the ‘x. ‘4 53 Table 10. (continued) Average k x Run Temp,°C. Substrate k x E? 106:l (sec (sec ) 59 150.186 dO 52.57 62 150.186 d 51.99 52.28 + 0.29 ‘° ‘ E 60 150.186 endo-6-CD3 51.22 63 150.186 endo-6-CD3 50.70 50.96 i 0.38 61 150.186 exo-6-CD3 49.86 64 150.186 exo-6-CD3 49.80 49.83 i 0.44 65 160.671 d 116.8 —0 68 160.671 do 119.4 118.1 1 1.4 66 160.671 endo-6-CD3 117.3 69 160.671 endo-6-CD3 119.6 118.4 : 1.2 67 160.671 exo-6-CD3 115.5 70 160.671 exo-6-CD3 114.4 115.0 : 1.9 times the average deviation of the other three. The errors associated with these individual rate constants of the table were taken as the average standard deviations of the four (or three) rate constants from each run. The errors reported with the average rate constants of Table 10 are of two kinds. The average standard deviation of the individual constants was calculated as well as the maximum deviation of the values comprising a given set 54 from the average of that set. The error reported in the table is the larger of these two. The variation of isotope effect with temperature is shown by the results collected in Table 11. The errors associated with kH/kD values were determined from the formula: 0 = kH/kD(oH2/kH2 + 602/1932)”2 where kH is the average rate constant for the unlabeled compound, kD is the average rate constant for the appro- priate deuterated compound and OH and CD are the errors in the respective rate constants. From the results ob- tained, it is clear that a trend toward increasing normal isotope effect with increasing temperature occurs with both epimerically deuterated compounds, although the results obtained at 160° are inconsistent with this trend. Suspicions of extraneous errors arising from possible de- composition or oxidation of the solvent at this tempera- ture were supported when the thermodynamic activation parameters were calculated. The average rate constants at the different temperatures were fitted to the equation: AH+ l ln(kr/T) = ln(k/h) --—-0-) + 15+ R T R where kr is the rate constant, T is the absolute tempera- ture, R is the gas constant, k is Boltzmann's constant, 55 Table 11. Temperature dependence of isotope effects in the solvolysis of 2,6,6-trimethy1-edddf2- norbornyl p-nitrobenzoates in 80% ethanol- water (V/V). Temp,°C. Substrate kH/kD 100.840 epddf6-CD3 0.997 t 0.016 125.217 endo-6-CD3 1.003 : 0.014 142.244 EBSQTG'CD3 1.027 : 0.013 150.186 229276'C03 1.026 : 0.010 160.671 eddd-G-CD3 0.997 1 0.015 100.840 §§275'CD3 1.011 1 0.015 125.217 exdfG-CD3 1.017 : 0.014 142.244 exdfG-CD3 1.043 t 0.013 150.186 egdfG-CD3 1.049 t 0.011 160.671 EEQTG'CD3 1.027 t 0.021 and h is Planck's constant. The values for AH+ and A S+ listed in Tables 12 and 13 were obtained from the slope and intercept of a plot of ln(kr/T) versus l/T by using a linear least squares computer program (ACTIV) written by Dr. George C. Sonnichsen. The errors reported with the AH+ and AS+ values are two times the standard devia- tion obtained by using the ACTIV program; this allows the error limits to be known with 95% certainty. From the fact that these errors calculated from data collected at Table 12. 56 Activation parameters for the solvolysis of 2,6,6-trimethy1-endo-2-norbornyl p-nitro- benzoates in 80% ethanol—water (V/V), calculated from results obtained at all temperatures. __—_—____-—___-—_———______1F_——_———_—_—____—__—_—_1F'——____ A . AS Substrate (cal/mole) (cal/deg/mole) 5% 30,148 i 342 -7.67 i 0.84 endo-6-CD3 30,065 i 204 -7.89 i 0.50 exo-6-CD3 29,993 i 234 -8.11 i 0.58 +-I'. +...+ Substrate AH H AH D AS H AS D (cal/mole) (cal/deg/mole) endo-6-CD3 83 i 398 0.22 i 0.98 exo-6-CD3 155 i 414 0.44 i 1.02 endo-6-CD3 83 i 168* 0.22 i 0.42* exo-6-CD3 157 i 134* 0.44 i 0.32* *Calculated from a least-squares treatment of a plot of ln(kH/kD) vs. l/T. 57 Table 13. Activation parameters for the solvolysis of 2,6,6-trimethyl-endo-2-norbornyl penitro- benzoates in 80% ethanol-water (V/V), omitting results obtained at 160.67l°. ===¢—'__——=‘_——_——- AH- AS Substrate (cal/mole) (cal/deg/mole) go 30,402 i. 118 -7.02 i 0.30 endo-6-CD3 30,204 i 124 -7.54 i 0.32 ;; exo-6-CD3 30,160 i 116 -7.68 i 0.28 7F. _ + + _ + Substrate A H AH D AS H AS D E (cal/mole) (cal/deg/mole) j N411 endo-6-CD3 198 i 171 0.52 .1: 0.44 exo-6-CD3 242 i 165 0.66 i. 0.41 endo-6-CD3 204 i 82* 0.53 i 0.20* exo-6-CD3 244 _+_ 96* 0.67 i 0.24* *Calculated from a least-squares treatment of a plot of ln(kH/kD) vs. l/T. all five temperatures over a 60° range (Table 12) are two to three times larger than the corresponding errors deter- mined from the same data after deletion of the results obtained at 160° (Table 13), it was concluded that the results obtained at this higher temperature were inaccu- rate; they were consequently discarded. The values obtained for the activation parameters in Table 13 are in good agreement with those reported by 58 20 who studied the solvolysis of 2,6,6- Brown, eE_el., trimethyl-edddf2-norbornyl p-nitrobenzoate in 80% acetone- water. From.kinetic measurements at 125° and 150° they determined AH+ and AS+ to be 31.5 kcal/mole and -6.4 eu. respectively. The values for A +H - AH+D and AS+H - AS+D were determined by two different methods. Subtraction of the appropriate individually determined AH+ and AS+ values gave differences which were in close agreement with those determined from the equation: +- +. +_ + AHH AHDl ASH ASD A linear least squares computer program (HANDS) written by Dr. George C. Sonnichsen was used to calculate AH+H - AH+b and ZB+H - AS+D from the slope and intercept of a plot of ln(kH/kD) versus l/T. The errors listed with these values are again two times the standard deviation obtained by using the HANDS program. It was felt that these limits of error were more realistic than those ob- tained from a statistical combination* of the errors associated with the individual AH+ and AS+ values. *The errors from such a combination were calcu- lated from the formula: 0 = (0A2 + OB2)1/2 where 9A and 03 are the errors in the individual values whose difference has been determined. 59 From these data, it is apparent that the enthalpy contribution to the isotope effects is substantial and in the direction one would predict from consideration of the 20 for steric hindrance to ionization in this system, and of Bartell5 for the conse- arguments of Brown, et al., quential increase in the carbon-hydrogen force constant resulting in an inverse isotope effect. In fact, at 150°, if enthalpy alone were important in determining the iso- tOpe effects, the calculated kH/kD of 0.78 for the 229276“ CD3 compound and 0.74 for the egdf6-CD3 compound are rather impressive. It is somewhat surprising to find that the exo-6-CD not only exhibits an isotope effect, but that it 3 is of about the same magnitude as that of the 223276'CD3° This suggests that there occurs a substantial increase in the repulsive interactions between the egdfG-methyl group and the eydfhydrogen in the 7-position as well as the geminal edddf6-methyl group, as the transition state geometry is approached. It would be interesting to in- vestigate this possibility by measuring the isotope effect in the solvolysis of 2,6,6-trimethy1-7,7-dideuterio- eddde-norbornyl p-nitrobenzoate. It is entropy, however, which is apparently re- sponsible for the unusual effect observed in this system. A comparison of experimental and calculated isotope effects based on the activation parameter differences 60 obtained from the HANDS cOmputer program is presented in Table 14. Table 14. Comparison of isotope effects calculated from activation parameters and those experimentally meaSured. Temp,°C.. Substrate Ad: - AGg(calc) (:E::? t:::? TE? . _.. ,_.,., (cal/mole) 100.340 22g2f6-CD3 6 0.992 0.997 125.217 gaggf6-CD3 -7 1.009 1.003 142.244 22d2f6-CD3 '16 1.020 1.027 150.136 Egggf6-CD3 -20 1.024 1.026 100.840 252f6-CD3 -7 1.008 1.011 125.217 gégf6-CD3 -23 1.029 1.017 142.244 §§2f6-CD3 -34 1.042 1.043 150.186 Ezgf6-CD3 ~40 1.048 1.049 The observed values of 0.53 and 0.67 cal/deg/mole for.AS+H - AS+D of the endo and exo trideuteriomethyl groups respectively are not particularly large for an interaction of this nature. ‘effect observed by Mislow, et al.,7 bis(trideuteriomethyl)phenanthrene It may be recalled that the with 9,10-dihydro-4,5- . . + _ exh1b1ted AS H - AS D—. 0.53 to 0.36 cal/deg/mole. Similarly, the 1,l'-binaphthy1- 2,2'-d’2 system studied by Carter and Dahlgren10 gave values 61 of AS+h - AS+D = 0.54 to 0.60 cal/deg/mole. The important difference between these previous studies and the present investigation is the temperature at which the kinetic measurements were performed. If the 2,6,6-trimethy1-edddf 2-norbornyl p-nitrobenzoates had been studied in the range 22-64° as were these other systems, inverse isotope effects ranging from 0.92 to 0.96 would have been expected. It is interesting to note that in another higher temperature 41 found that the isotope effect study, Heitner and Leffek on the rate of racemization of optically active II d-camphor-lO-sulfonate in water at 100.0° and 119.9° was (CD3)2 II N(CD ) + 3 3 0.996 and 1.02 respectively. As in the present system, enthalpy considerations alone would lead to a predicted large inverse effect. Although suggestions as to the source of these observed entr0py contributions would be purely specula- tive at this point, it may be useful to consider the possibilities. Specific solvation effects have been 42 considered as a source of B-deuterium isotope effects in solvolytic reactions, but considerable amounts of data have been presented which indicate this to be an unlikely cause of these effects. Data collected by Kang43 62 on the solvolyses of the 8-trideuteriomethy1-l-naphthoyl chlorides in.aqueous acetone solvents of varying composi- tions showed.isot0pe effects which were independent of solvent composition, a result which would argue against differential solvation. The fact that in the present study the exo~6-CD compound exhibits an effect similar 3 to that for the edddf6-trideuteriomethy1 compound also argues against specific solvation, since one would expect that a differential involvement of solvent as the transi- tion state is approached would be important only in the immediate vicinity of the ionizing bond. A more likely source of the entropy contribution to the isotope effect depends upon internal rotational differences between deuterated and non-deuterated mole- cules. In their study of the temperature dependence of the B-deuterium isotope effects in the solvolysis of isopropyl substrates, Robertson, e5_e£.,l7 suggested that the major portion of the observed (normal) effects was a result of such differences. Although no clear evidence that CH3 groups possess higher barriers to rotation than CD3 groups in hydrocarbon-like molecules has been experi- mentally obtained, it may be argued that the larger steric requirements of a normal methyl group, as compared to a trideuteriomethyl group, would result in an increased barrier height. This argument seems reasonable in view 44 of the calculations made by Kreevoy and Mason which 63 indicate that as van der Waals repulsions increase, bar- riers to rotation also increase. Application of these arguments to the 2,6,6-trimethy1-eddgf2-norborny1 system allows the rationalization of the observed entropy ef- fects in a strictly qualitative sense. Thus, if it is assumed that as the transition state is approached, the van der Waals repulsions involving the trideuteriomethyl groups increase (as indicated by the enthalpy contribu- tions to the isotope effect); and that once the transi- tion state has been reached, these repulsions have become so large that internal rotational differences due to deuteration have become obscured; then, the lower barrier to internal rotation of the ground state trideuterio- methyl group will manifest itself as a greater decrease in activation entropy for the deuterated molecule. In view of these ideas it is interesting to re- consider the unusual effect reported by Jewett and 12 Dunlap, and mentioned previously in the INTRODUCTION. It may be recalled that in the aqueous ethanolysis of III, CD CH I 3 I 3 CD3— ('3 — CH2 -g -- Cl III CD3 H3 an inverse isotOpe effect of kH/kD = 0,933 was measured at an unspecified temperature (probably around 25°). The rate of solvolysis of the unlabeled compound was found to 64 be greater than that of Eeggfbutyl chloride by a factor of twenty. This rate enhancement was attributed to steric acceleration arising from a severly crowded ground state.45 On this basis, one would have predicted a substantial normal isotOpe.effect from enthalpy considerations alone. If, however,.entr0py contributions are also taken into account in a manner completely analogous to that just described, it is quite reasonable to rationalize the 137 verse effect observed. Thus, with AH+h - AH+b = -300 cal/mole and AS+H - AS+D = -l.06 cal/deg/mole (admittedly convenient values, but nonetheless reasonable), at 25° the isotope effect would be 0.98 corresponding to a AAG+ of +16 cal/mole. It would be extremely interesting to examine the temperature dependence of the isotope effect for this system. Although providing an entertaining and interest- ing rationalization for the effects observed in the pre- 7 sent study as well as the study of Mislow, et al., this approach would appear inapproPriate in explaining the re- sults of Carter and Dahlgrenlo where a substantial entropy contribution was found in the absence of deuterated substituents subject to internal rotational differences. CONCLUSION It would.thus appear that the present study has afforded a.conc1usive demonstration of the existence and the importance of entropy contributions to those isotope effects which arise from interactions between non-bonded atoms or groups of atoms. The utility of isotope effects of this nature in determining reaction mechanisms has been illustrated, while at the same time the futility of conclusions drawn from their measurement at a single temperature has been clearly established. Consideration of the erroneous conclusions which may have been reached regarding Brown's explanation for the decreased reactivity of 2,6,6-trimethy1-e2ddf2-norbornyl p-nitrobenzoate if the normal isotope effect at 150° had been assumed to reflect enthalpy differences alone, makes the importance of these results clear. It would appear that future efforts in this area should be directed toward a better understanding of the underlying source of the entropy contributions that have been observed. 65 LI ST OF REFERENCES 10. 11. 12. a) b) a) b) a) b) G. LIST OF REFERENCES J. M. Harris, R. E. Hall, and P. v. R. Schleyer, J. Am. Chem. Soc., dd, 2551 (1971), and refer- ences cited therein. V. J. Shiner and R. D. Fisher, Ibid., dd, 2553 (1971), and references cited therein. A. Halevi, "Secondary Isotope Effects," Progress in Physical Organic Chemistry, Vol. l, S. G. Cohen, A. Streitweiser, and R. W. Taft, eds., John Wiley and Sons, New York, 1963, and refer- ences cited therein. A. Halevi, Tetrahedron, 174 (1957). E. S. Lewis, Ibid., d, 143 (1959). V. J. Shiner, Jr., Ibid., d, 243 (1959). L. S. Bartell, J. Am. Chem. Soc., dd, 3567 (1961). L. S. Bartell, Iowa State J. Sci., dd, 137 (1961). J. Karabatsos, G. C. Sonnichsen, C. G. Papaioannou, S. E. Scheppele, and R. L. Shone, J. Am. Chem. Soc., dd, 463 (1967). Mislow, R. Graeve, A. J. Gordon, and G. H. Wahl, Jr., Ibid., dd, 1199 (1963); dd, 1733 (1964). Melander and R. E. Carter, Ibid., dd, 295 (1964). Melander and R. E. Carter, Acta Chem. Scand., id, 1138 (1964). E. Carter and L. Dahlgren, Ibid., 23, 504 (1969). T. Leffek, J. A. Llewellyn, and R. E. Robertson, J. Am. Chem. Soc., dd, 6315 (1960). G. Jewett and R. P. Dunlap, Ibid., dd, 809 (1968). 66 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. a) b) a) b) 67 C. Brown and G. J. McDonald, Ibid., dd, 2514 (1966). C. Brown, M. E. Azzaro, J. G. Koelling, and G. J. McDonald, Ibid., dd, 2520 (1966). C. Brown, D. Gintis, and L. Domash, Ibid., _d, 5387 (1956). C. G. Swain and N. D. Hershey, Ibid., _d, 1901 (1972). G. J. Frisone and E. R. Thornton, Ibid., dd, 1211 (1968). T. Leffek, R. E. Robertson, and S. E. Sugamori, Can. J. Chem., dd, 1989 (1961). J. Karabatsos and C. G. Papaioannou, Tet. Letters, 1968, 2629. C. Sonnichsen, Ph.D. Thesis, Michigan State University, 1967. Ikegami, D. L. Vander Jagt, and H. C. Brown, J. Am. Chem. Soc., dd, 7124 (1968). C. Brown and J. Muzzio, Ibid., 8 2811 (1966). .33.: 90, 6216 (1968). C. Brown and M. H. Rei, Ibid., L. Goering and K. Humski, Ibid., dd, 6213 (1968). L. Goering and K. Humski, Ibid., dd, 4594 (1969). L. Murr, A. Nickon, T. D. Swartz, and N. H. Werstiuk, Ibid., dd, 1730 (1967). Beckmann, R. Schaber, and R. Bamberger, Chem. Ber., d1, 997 (1954). K. Alder and W. Gunzl, Chem. Ber., dd, 809 (1960). E. E. van Tameln and M. Shamma, J. Am. Chem. Soc., 1d, 2315 (1954). S. Rondestvedt, Jr. and C. D. Ver Nooy, Ibid., 21, 4878 (1955). C. Brown, J. H. Kawakami, and S. Ikegami, Ibid., dd, 6914 (1970). 'F a -' 1' a . 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. a) b) V. 68 P. v. R. Schleyer, M. M. Donaldson, and W. E. Watts, Ibid., d1, 375 (1965). H. M. Walborsky and D. F. Loncrini, Ibid., 6, 5396 (1954) . "_' _ C. Brown and C. P. Garg, Ibid., dd, 2952 (1961). C. Brown, Ibid., 9, 370 (1967). F. Fieser, Experiments in Organic Chemistry, D. C. Heath and Company, Boston, 1957, p. 286. F. Smith, Jr., unpublished results. S. Winstein, E. Grunwald, and H. W. Jones, d. Am. Chem. Soc., 2d, 2700 (1951). E. Grunwald.and S. Winstein, Ibid., 1d, 846 (1948). A. Bunton, C. O'Connor and D. Whittaker, J. Org. Chem., dd, 2812 (1967). L. Goering and W. D. Closson, J. Am. Chem. Soc., dd, 3511 (1961). S. Roy, Anal. Chem., dd, 1724 (1968). L. Dye and V. A. Nicely, J. Chem. Ed., dd, 443 (1971). A. Skoog and D. M. West, Fundamentals of Analyti- cal Chemistry, Holt, Rinehart, and Winston, New York, 1966, p. 557. Heitner and K. T. Leffek, Can. J. Chem., dd, 2567 (1966). J. Shiner, J. Am. Chem. Soc., 1d, 2925 (1953); Ibid., 1d, 1603 (1954). Uan Gen Kang, Ph.D. Thesis, Michigan State University, E. H. 1972. A. Mason and M. M. Kreevoy, J. Am. Chem. Soc., 11, 5808 (1955). C. Brown and H. L. Berneis, Ibid., 1d, 10 (1953). MICHIGAN STATE UNIVERSITY LIBRARIES IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 31 93 03175 0403