Part I. The Structure of 1,2-Dimethylnorbornyl Cation. Part II. The Relative Stabilities of Some Tertiary Aliphatic Cations Citation McFarland, James Thomas (1969) Part I. The Structure of 1,2-Dimethylnorbornyl Cation. Part II. The Relative Stabilities of Some Tertiary Aliphatic Cations. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/NJST-QK11. https://resolver.caltech.edu/CaltechTHESIS:03292017-092122515 Abstract Part I. The pmr spectrum of 1,2-dimethylnorbornyl cation indicates that the Wagner-Meerwein rearrangement is rapid on the nmr time scale at -100° C. Furthermore the protons at C-6 are deshielded by 1.3 ppm with respect to the protons at C-5. This is consistent with the nonclassical structure for this cation. Part II. Mixtures of organic chlorides have been forced to compete for less than one equivalent of antimony penta-fluoride in order to judge relative cation stabilities. The results indicate no stabilization of 2,3,3-trimethyl- 2-butyl cation by methyl bridging. However methylene bridging seems to stabilize 1,2-dimethylnorbornyl cation. The qualitative order of cation stabilities measured by this method is not always the same as that measured by solvolysis experiments. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Richards, John H. Thesis Committee: Unknown, Unknown Defense Date: 8 August 1968 Funders: Funding Agency Grant Number Caltech UNSPECIFIED NSF UNSPECIFIED Record Number: CaltechTHESIS:03292017-092122515 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:03292017-092122515 DOI: 10.7907/NJST-QK11 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 10115 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 29 Mar 2017 17:35 Last Modified: 03 May 2024 20:45 Thesis Files Preview PDF - Final Version See Usage Policy. 27MB Repository Staff Only: item control page

Part I. The Structure of 1,2-Dimethylnorbornyl Cation Part II. The Relative Stabilities of Some Tertiary Aliphatic Cations Thesis by James Thomas McFarland In Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1969 (Submitted August 8, 1968) ii ACKNOWLEDGMENTS I would like to thank Professor John H. Richards for his help in the preparation of this thesis. This thesis could never have been completed without the advice, encouragement, and friendship of the graduate students and postdoctoral research fellows of the Richards Group. I would especially like to thank Mr. P. C. Schaefer, Mr. F. W. Dahlquist, and Mr. D. M. Mog. I owe special thanks to several people who have contributed much throughout my years of education: to my parents for their sacrifices and encouragement, to Professor T. R. Williams for his enthusiasm and assistance, and to my wife for her patience and understanding. I would like to thank the California Institute of Technology and the National Science Foundation for financial assistance in the form of a teaching assistantship, an NSF summer fellowship, and an NSF trainee fellowship. iii Abstract Part I The pmr spectrum of 1,2-dimethylnorbornyl cation indicates that the Wagner-Meerwein rearrangement is rapid on the nmr time scale at -100° C. Furthermore the protons at C-6 are deshielded by 1.3 ppm with respect to the protons at C-5. This is consistent with the nonclassical structure for this cation. Part II Mixtures of organic chlorides have been forced to compete for less than one equivalent of antimony pentafluoride in order to judge relative cation stabilities. The results indicate no stabilization of 2,3,3-trimethyl2-butyl cation by methyl bridging . However methylene bridging seems to stabilize 1,2-dimethylnorbornyl cation. The qualitative order of cation stabilities measured by this method is not always the same as that measured by solvolysis experiments~ iv TABLE OF CONTENTS Part I page I. Introduction A. The Question of Nonclassical Tertiary Cations 1 B. Stable Carbonium Ions 10 II . Results A. 1,2-Dimethylnorbornyl Cation 15 B. 2-Methylnorbornyl Cation 20 C. Methyl-d3-cyclopentyl Cation 25 III. Discussion A. Spectral Assignment for 1,2-Dimethylnorbornyl Cation 26 B. Comparison of Ionization Shifts for Tertiary Aliphatic Ions 28 C. Spectral Assignment for 2-Methylnorbornyl Cation 29 D. Discussion of Deshielding of Protons at C-6 34 IV. Experimental A. Syntheses 38 B. Epimerization of 1,2-Dimethylnorbornyl Alcohol 45 C. Preparation of Cations 46 Part II I . Introduction II . Results A. Cation-Chloride Exchange Reactions in Antimony Pentafluoride v B. III. IV. 1) !-Butyl Cation 52 2) !-Amyl Cation 57 3) 2,3,3-Trimethyl-2-butyl Cation 59 4) Methylcyclopentyl Cation 61 5) 1,2-Dimethylnorbornyl Cation 61 6) 2-Methylnorbornyl Cation 61 Cation Competition Experiments 1) Qualitative Comparisons with !-Butyl Cation 64 2) Quantitative Comparisons with !-Butyl Cation 67 3) Comparison of Methylcyclopentyl Cation and 1,2-Dimethylnorbornyl Cation 71 4) Comparison of 1,2-Dimethylnorbornyl Cation and 2-Methylnorbornyl Cation 73 Discussion A. Comparison of Cation Stabilities from Direct Competition and from Solvolysis Experiments B. Discussion of Methyl Bridging in 2,3,3- 81 Trimethyl-2-butyl Cation c. The Role of Methylene Bridging in 1,2- . 83 Dimethylnorbornyl Ca¥ion 79 Experimental A. Synthesis 88 B. Preparation of Antimony Pentafluoride Solutions 88 c. Cation-Chloride Equilibrations 89 vi D. Cation Competition Experiments 90 References 91 Propositions 94 vii To Judy and Kevin INTRODUCTION In 1952 Winstein proposed that the intermediate in the solvolysis of 2-exo-norbornyl arylsulfonates is a symmetrical nonclassical ion, I. I This proposal was based on the accelerated solvolysis rate of the exo-norbornyl arylsulfonate compared to cyclohexyl arylsulfonate and on the stereospecificity of the attack on the cation (1,2,3). Presumably the accelera- tion in rate of reaction reflects increased cation stability, and the specificity of attack by the nucleophile is a consequence of carbon bridging. H. C. Brown suggested three kinds of carbonium ions: classical static ions, classical equilibrating ions, and nonclassical bridged ions. -R-...~ ,,,.,' ; ' 1' E "..- ')( • •• • • '' ' ' 'RY l I classical static (a) ~ ... ,'1' • t E ' i I \ ' '' ' RY classical equilibrating (b) 2 R--Y ,-... 'I ~'. R--X 1' ,I I \ t E nonclassical bridged (c) Figure 1 In the case of the static ions (a), the energy barrier for equilibration of two symmetrical cations is higher than that for conversion to products. Equilibration between ions takes place in case (b) because the energy barrier for this process is lower than the barrier leading to products . In the nonclassical bridged ion, the barrier has disappeared, and the ion is stabilized by charge delocalization. Brown has urged a re-examination of all ions consid~ ered to be nonclassical to see if they can be assigned to the classical equilibrating group (4). A nonclassical structure was considered for the 2,3, 3-trimethyl-2-butyl cation (II), but labeling work proved that this type of methyl bridging was not important. 3 Treatment of 2,3,3-trimethyl-2-butanol-l- 14 C with concen- trated hydrochloric acid gives only 12% rearrangement after one minute. Nearly 58% rearrangement is observed under reversible conditions where 100% of the chloride product has exchanged with hydrochloric acid. These results indi- cate that methyl bridging is not important under irreversible conditions (5). This result has been confirmed in deuterium labeling studies on III. Very little scrambling of the deuterium label occurs, CD ' 3 ~ - CD 3 x III indicating only a small amount of equilibration relative to solvent capture. reaction III-OH +HCl . iii=g~ ~g~§g~ H+ ) % scrambling III-Cl+ ~O __. . . .,._.) iii=gg~~ + H2 0 12 26 19 Figure 2 These reactions are usually considered to involve carbonium ion intermediates; therefore, the conclusion that bridging is unimportant seems valid (6). The question of methylene bridging in a more sterically favorable bicyclic tertiary cation has been extensively studied. The 1,2-diarylnorbornyl cation seems 4 to be a classical ion with the positive charge localized on one carbon. The ultraviolet spectrum, chemical reactivity, and thermodynamic stability of the monoaryl- and 1,2-diarylnorbornyl cations are quite similar. The pmr spectrum at high temperature shows a single aryl peak which loses detail and broadens at -20° C., indicating impending nonequivalence (7). H. C. Brown has examined 2-methyl- and 1,2-dimethyl-· norbornyl derivatives to see if these cation systems are classical. He criticized the use of _t-butyl or cyclohexyl derivatives as model compounds in those anchimeric assistance studies which show rate increases for solvolysis of norbornyl derivatives (1,2,3). Instead, he prefers methyl- cyclopentyl compounds as models. The rate of solvolysis of 2-methyl-2-exo-chloronorbornane is only five times as great as that of the model compound, l-chloro-1,2-dimethylcyclopentane. Such a small rate increase indicates no significant participation by the 1,6 sigma bond (8). Another argument favoring a classical 2-methylnorbornyl cation is that introduction of a methyl group has a large effect on solvolysis rates in the norbornyl system but a much smaller effect on a resonance delocalized system (Figure 3). Brown's conclusion is that norbornyl cation does not behave like a resonance stabilized cation(9i Equilibration of 2-methyl-~-norbornyl alcohol, 3 /\ R c ,, X cp' ' R CH 3 , ~~R o~ CH CH3, C,,, x Figure 3 0.216 39L~ 1,800 55,000 30.0 24.4 5.42x10- 4 5.55 5.62 175,000 55,000 kH kJVJe 3.21Xlo- 5 Me Cl 1.66 Et OH 0.377 H Cl o.o36 HOAc 1.57x10- 6 H OTS 0.081 Et OH 0.0184 R=H X=OTS V1 6 2-methyl-endo-norbornyl alcohol, and l-methyl-2-exonorbornyl alcohol in perchloric acid-water-dioxane gives an equilibrium product distribution of 1.3:1.0:3.6. The same experiment on 1,2-dimethyl exo- and endo-norbornyl alcohol gives an exo to endo ratio of 2.6 :1.0. Furthermore, the rate of formation of the 2-methyl-2-exo-norbornyl alcohol is twice that for formation of l-methyl-2-exo~norbornyl alcohol. Brown takes these facts as evidence that the rate of capture of 2-methylnorbornyl cation is much faster than the Wagner-Meerwein rearrangement to the secondary 1-methylnorbornyl cation. This is more evidence that 2-methyl- norbornyl cation is a classical tertiary ion in equilibrium with a small amount of secondary ion (10). Hydrochlorination of 1-methylnorbornene gives a 45:55 distribution between 3-chloro-1-methylnorbornane and 2chloro-2-methylnorbornane. Brown argues that these data are inconsistent with a delocalized cation in which the methyl group would have a much larger directive effect in -the reaction (11). Brown argues that since the 2-methyl-2-norbornyl cation is classical, the introduction of a methyl group at C-1 should have a very large effect if the 1,2-dimethylnorbornyl cation is nonclassical . Furthermore, this large substituent effect should be observed in the exo but not the endo derivatives. The results are shown in Figure 4. 7 (50°)rate consgant x 10 sec-1 Me/H rel .rates 0.50 1.00 boPHB 1.16 2.3 tH_.boPHS 1.67 3.4 183 2.2 1.00 165 9.4 4.3 0.012 1.00 0.057 4.8 Compound boPNB CH3 exo/endo CH3 t.H 3 .K C v, 3 c1::J._ OPN B CH3 · ~OPNB 3 Ctt3 cQe113 OPN6 ct/CH~ CH3 OPN 8 Figure 4 8 The data indicate that 1,2-dimethylnorbornyl cation is classical (12). Finally Brown has studied the hydrochlorination of l-methyl-d3 -2-methylenenorbornane to see if C-1 and C-2 become equivalent, as demanded for a nonclassical ion. Using neat olefins and very short reaction times, scrambled products amount to 35% of total. In ether or methylene chloride, scrambling as low as 52% is observed. The experi- ment gives equivalent results with l-methyl-2-methylenect2-norbornane. Equilibration is not complete, so 1,2-di- methylnorbornyl cation is in the classical equilibrating category of cations (13). Even though 2-methyl- and 1,2-dimethylnorbornyl cations are classical, they show high specificity for ~­ substitution (14) (Figure 5) and high exo and endo rates of solvolysis (Figure 4). Brown points out that exo substitu- tion and high exo:endo rate ratios are therefore not good criteria for formation of nonclassical ions. Generation of norbornyl, 2-methylnorbornyl, and 1,2dimethylnorbornyl cations through the 11 route from 2( ~3-cyclopentenyl) ethyl £-nitrobenzoates (IV) seems to indicate a delocalization of charge over both C-1 and C-2. The effect of the first methyl substitution, V, in this system is a 7.0-fold increase in rate over the unsubstituted compound. Addition of another methyl group, VI, gives a rate increase of 5.5. The products from the solvolyses are ct:iOT5 CU3 ~ec H 56 57 (NaBH 0 0 Figure 5 42 35 o.6 0.5 4 added as trapping agent) o.8 3 o.8 3 cQCH3 cl:4u. ~U3 3 3 ~ c!:t ~H % products \.0 10 2-methyl-2-exo-norbornyl acetate from V and 1,2-dimethylexo-norbornyl acetate from VI. )=<z ~ HOAc'> ~i IV R =R =H v R 1 =H; R2 =cH 3 1 2 VI C.Hz - X These data seem to be inconsistent with Brown's classical 1,2-dimethylnorbornyl cation which has the charge localized on only one carbon. If this were the case, the second methyl substitution should be much less effective in stabilizLng the transition state (15). G. A. Olah has had much success observing alkyl cations in sulfur dioxide solutions of antimony pentafluoride. The cations are generated from alkyl halides in this highly ionizing solution. In this manner, !-butyl cation can be generated from !-butyl fluoride. The methyl groups appear as one singlet at -4.07 ppm from TMS. In the same manner, !-amyl cation can be generated from~-amyl fluoride; the pmr spectrum of cation VII consists of a triplet at @ CH 3 a - c - c~b- CH3c ' CH 3 a VII -1.72 ppm, a triplet at -3.97 ppm, and a septet at -4.30 11 ppm from TMS. ~' The upfield triplet is assigned to protons the low field triplet to protons ~' and the septet to b. The long range coupling between the methyl and methylerE groups through the sp 2 carbon is about 7 cps. The spectrum methyl doublet at -5.05 of isopropyl cation consists of a and a methine septet at -13.5 ppm (neat SbF , no so 2 5 added) (16). Methylcyclopentyl cation (VIII) is also of interest to this study. The spectrum is quite complex, consisting of an upfield quintet at -2 .• 47 ppm which is assigned to the {?>-methylene protons. K H c,I.. ff VIII The low field multiplet is assigned by analogy to the methyl and methylene protons in !_-amyl cation. The methyl protons are at -3.98 ppm and the c< -methylene protons at -4.20 ppm. A long range coupling constant of 4.o cps is a ssigned for coupling between the methyl and ~ -methylene protons. The methylcyclopentyl cation is generated from e ither methylcyclopentyl or cyclohexyl halides or alcohols ( 17) . 12 Another interesting tertiary cation is the 2,2,3trimethyl-2-butyl cation (IX). The spectrum of IX consists of a singlet at -2.86 ppm. IX All the methyl groups are equivalent as low as -180° C. -10° C. the ion cleaves to give i-butyl cation. At Olah has assigned a classical equilibrating structure to the cation because he feels the methyl resonance of a protonated tetramethylcyclopropane would occur at high field due to anisotropy from the cyclopropyl ring and delocalization of the positive charge. The model he uses for the nonclassi- cal bridged system is tetramethylethylene halonium ion, where the methyl shift is 0.95-0.75 ppm. He assigns the barrier to methyl rearrangement as less than 2-3 kcal/ mole ( 18). The tricyclic adamantyl and bicyclic norbornyl cations will also be of interest in later discussion. adamantyl cation (X) The 13 spectrum consists of a lowfield l'-proton resonance at -5.42 ppm, a ~ -proton resonance at -4.50 ppm, and a proton resonance at -2.67 ppm. b- The interesting point is that the "t· -protons are deshielded with respect to those~ to the positive charge. Olah considered the possibility that an intermolecular hydride shift by the Y.. protons might explain the deshielding but discounted this on the basis of temperature independence of the spectrum from +40 to -90° C. This leaves charge delocalization by hypercon- jugation and backside overlap between the C-H sp3 orbitals and the empty £ orbital to explain the deshielding. Olah prefers delocalization by backside overlap (19). The pmr spectrum of norbornyl cation consists of a single peak at -3.1 ppm. On cooling to -60° C. this peak splits into three peaks at -5.35 ppm (area 4), -3.15 ppm (area 1), and -2.20 ppm (area 6). This is explained by a slow 3,2 hydride shift, which is frozen out in the low temperature spectrum. Cooling as low as -143° C. produced no further change in the spectrum. On this basis, the maximum barrier to Wagner-Meerwein rearrangement and 1,6 hydride shifts is assigned as 5.5 kcal/mole (20). It seemed desirable to study the pmr spectra of the 1,2-dimethylnorbornyl and 2-methylnorbornyl cations. The pmr spectrum of the 1,2-dimethyl ion should allow the study of Wagner-Meerwein rearrangement in the absence of significant amounts of 1,6 and 2,3 hydride shift to less stable 14 secondary cations . The spectra should also shed some light on on the nature of the charge distribution in tertiary norbornyl cations. This information has bearing on the question of whether tertiary norbornyl cations are classical . 15 RESULTS The dimethylnorbornyl cation was prepared either from a mixture of 1,2-dimethyl-2-~- and endo-norbornyl chlorid2s or from the exo or endo alcohols. The 1,2-dimethyl-2-endo- norbornyl alcohol and 1,2-dimethyl-2-exo-norbornyl alcohol were prepared and the spectra recorded (21,22). The pmr spectrumof the exo isomer consisted of methyl resonances at 63 and 69 cps from TMS, while these resonances for the endo isomer occurred at 62 and 70 cps from TMS. The two isomers gave the same mixture in sulfur dioxide, indicating epimerization in thisrolvent. The pmr data and assignment of the exo and endo alcohols and chlorides used in the cation experiments are shown in Figure 6. Careful epimerization studies on the alcoholsshowed the equilibrium mixture in sulfur dioxide without added catalyst at 65° C. to be 53:46 in favor of the _..._ exo alcohol. The data compare favorably with 2.6:1.0 for the exo:endo ratio in water-dioxane (10). All indications are that the ground state free energy differences between exo and endo are small. chloridesand alcohols The 1,2-dimethyl-2-norbornyl cation spectrum (Figure 7) w~s temperature independent from o° C. to -105° C. broadening At the lowest temperatures there wasslight (~ 4 cps) of the peak at 149 cps, but this is not large enough to be assigned to impendi!lg nonequivalence of the two methyl groups. CH3(!) M3@ OH from soc1 2 & alcohol CCl4 91 Figure 6 72 19 24 17 55 62 86 82 so 2 11 8 8 6 cps 22 19 ___ 1 £ CH3a) f, CH3 74 80 22 24 62 63 CH 3@ b(TMS) ____ __cp s ___ 96 89 33 32 70 69 CH-=> Q) b( TMS) __ _..) _ c_ps_ . __ cc1 4 so 2 cc1 4 CClL~ solvent & alcohol flf..~© from HCl C.l-la © fl:l-~ cfllt3 ~~ t;/;:{.._ c H:/D H ~CH:/S> @ (;fH3 c;QoH compound ® exo 100% exo ~67% = 33% endo --67% exo ~ 33% endo 30% endo ~ 70% exo % 100% endo 100% exo ~% isomer 0\ }-l z.so ISO 100 50 Figure 7 pmr spectrum of 1,2-dimethylnorbornyl cation zco ._.., -..J 18 An attempt was made to prepare the fully deuterated l,2-dimethyl-d6-2-norbornyl alcohol, but some of the label at the C-1 methyl group was lost during reduction of the nitrate ester of 1-methyl-2-norbornyl alcohol. lfJ.ass spec- tral analysis of 1-methyl-2-norcamphor indicated 1.9 hydrogens in the partially deuterated methyl group (Figure 8) . peak intensity M-1 0.062 M 1.000 0.331 M+l 1.086 0.355 M-t2 o.724 0.218 M+3 0.331 0.097 M+4 0.059 specie abundance total H 1.9 Figure 8 The pmr spectrum of the dimethyl-d4. 1 -norbornyl cation showed the resonance at 149 cps to be that of the methyl groups and four other protons . (Figure 9). The cation solution was quenched with water to give a 47% yield of a compound whose pmr and ir spectra were identical with those of 1,2-dimethylnorbornyl alcohol. The pmr spectrum did not indicate any secondary alcohols. Gas- liquid partition chromatography showed the product to be a mixture of exo and endo alcohols. The percentage of endo b 17 cat scans a 16 cat scans CD 3 CH 1 • 9 Dl.l dimethyl cation dimethyl cation 2.2 b 9.6 2.1 a Figure 9 5.8 -1.9 methyl protons 3.9 3:0 -6.o methyl protons 149 100 cps Integrated Areas 2.9 3.3 181 1-1 \.0 20 alcohol varied in four experiments from 14% to 30%. 2-Methylnorbornyl cation was prepared from either 2-methyl-2-norbornyl alcohol or 2-methyl-2-chloronorbornane. The methyl resonances of the alcohols and chlorides are shown in Figure 10. The pmr spectrum of 2-methylnorbornyl cation at -67° C. is shown in Figure 11. temperature independent from -58 The spectrum was C. to -100 C. The cation generated from 2-methyl-d3-2-endo norbornyl alcohol gave the integrated results in Figure 12. clearly the resonance at 196 cps. The methyl group is When the cation is quenched with water, an alcohol is obtained whose pmr and ir spectra are the same as those of 2-methyl-2-norbornyl alcohol. Gas-liquid partition chromatography indicates that the alcohol is a mixture of 97% 2-methyl-2-norbornyl alcohol and 3% l-methyl-2-norbornyl alcohol. The cation spectrum shows temperature dependent changes on warming to -18° C. or -29° C. These changes do not represent thermal decomposition; on cooling to -67° C. ·the original spectrum is obtained. spectrum is shown in Figure 13. The high temperature The integrals for this spectrum and for 2-methyl-d3-2-norbornyl cation are shewn in figure 14. Dll ~Ua OH C...3 cl:tei 50C.h.> Figure 10 98 62 CJ;;_~ cq_l!H3 HCJ. '> ~OH 76 exo ~ndo-Grignard stereochemistry exo multiplet 221 cps , ..... c1 ,,,..,H Jmethyl(TMS) in CCl4 cps 64 OH ChCH3 Compound f\) I-' 22 ._,...--......----------------------......................-a o l ~ ........... 0 0 t'-\0 I ~ '--' c 0 •rl ~ ro 0 r-1 a ~ ~ c H 0 H M M c © ~ 0 0 ~ r-1 ~ ~ ~ © sI N ~ 0 s~ H ~ 0 © ~ ro H 8 ~ s ~ H M ~ ~ ~ b b 36 cat scans a 10 cat scans cn 3 o.8 o.8 b Figure 12 3.0 b 7.2 a 1.0 a CH 3 206 cps, 196 cps, 185 cps 299 cps (-67°c) Integrated Areas 5.2 b 4 .8 a 117 & higher cps I\) w Ko 1.51> zoo ISO 100 Figure 13 pmr spectrum of 2-methylnorbor nyl cation (-18 ° C) ~00 So ·~ .i::- I\) 25 Integrated Areas 210 cps 193 cps 180 cps 104 cps 4.9 8.1 8.5 4.5 6.o 4.o Figure 14 (integrals are from single -scan spectra) The methylcyclopentyl cation pmr spectrum consists of an upfield quintet at 140 cps and a downfield multiplet. The lower part of the multiplet at 246 cps has been assigned to the methylene protons 0\ to the methyl group at 230 cps (17). Olah did not confirm his assignment by deuterium studies, so 1-methyl-d -l-chlorocyclopentane was 3 synthesized and the carbonium ion prepared. The deuteromethylcyclopentyl cation pmr spectrum consisted of two quintets: one upfield at 140 cps and a broad downfield quintet at 2L~6 cps. Therefore,Olah 1 s assignment of the spectrum is confirmed. The breadth of the downfield quintet compared to the upfield quintet is consistent with ·1ong range coupling between D and H through the sp2 carbon. The long range H-H coupling constan~ between the methyl and methylene protons is reported by Olah as 4 cps (17) . 26 DISCUSSION The following assignment of the 1,2-dimethyl-2norbornyl cation (XI) is consistent with chemical shifts, integration of the spectrum, deuteration, and observed coupling. ,' ,' ' ' , , , , ,_ - - - - - -' ' -~ XI Deuteration of the methyl groups establishes that the methyl protons, a, occur in the sharp peak at 149 cps. Integration of this peak indicates 3.6-3.9 protons in the shoulder of the methyl peak. These protons are assigned to the four methylene protonso< to the positive charge. This is consistent with the relative shift of the methylene protons in other tertiary carbonium ions. These protons occur just downfield (13 cps and 20 cps) of the methyl resonance in methylcyclopentyl cation and in .:!?_-amyl cation. This leaves the protons at C-5 and C-6 as the high field and low field resonance peaks. It seems reasonable to assign the C-6 protons, ~' to the low field peak because of their proximity to the positive charge. The high field triplet 27 and the low field broadened triplet are apparently coupled to each ·other JH-H ~ 6.o cps. This is consistent with the large coupling constant predicted for two vicinal protons with a dihedral angle <#> =O such as the protons at C-5 and C-6 (23). The total width of the overlapped pattern at 149 cps is not enough to explain coupling to either the low or high field resonances . This leaves the bridgehead proton, £, at the low field resonance and may account for the poor resolution of this signal. The bridgehead proton's resonance appears to be downfield farther than expected. However, the bridgehead proton in 1-methylnorbornane is downfield 0.7 ppm from the ring methylene protons (24). It therefore seems possible that the bridgehead proton could occur 1.3 ppm downfield from the methylene proton farthest from the positive charge. The interesting point in the 1,2-dimethylnorbornyl cation spectrum is that the methylene protons at C-6, while not directly bonded to the carbonium ion center, are farther downfield than the methylene protons at C-3 and C-7 which are (;( to the positive charge. The lack of broadening of the spectrum to -105° C. deserves more comment. Assuming that 1,6 and 3,2 hydride shifts play a small role in the observed spectrum because of the small amount of' secondary cation present, the only observable rate process would be equilibration of the methyl groups by Wagner-Meerwein rearrangements . We can 28 estimate such a rate process from estimated shifts of the nonequivalent methyl groups using the shape function equation for intermediate rate of exchange between two ~ 7C.1 = 1: J_ Figure 15 sites with equal populations. It would have been possible to see broadening of 2.'tf'((,/A-11:8) =l (25). The chemical shift can be estimated from the shift of the methyl group in 2-methylnorbornyl cation at 195 cps and the single methyl group in ~-amyl cation at 105 cps. Thus-flA-~=90/sec, and k1 minimum:2.82xlo~ 2 sec- 1 ; this corresponds to a maximum AG*= 7. 5 kcal/mole. estimate of .OG*' This compares with Olah' s = 5. 5 kcal/mole for the Wagner-Meerwein rearrangement in norbornyl cation (20). Some comparisons of the dimethylnorbornyl cation with other tertiary cations are possible. Figure 16 shows the chemical shift differences of the methyl groups in a few tertiary cations and their chlorides (i.e.chemical shifts due to ionization). The effect of charge delocalization in the last two cations in Figure 16 is striking. The dimethylnorbornyl cation shows no observable long range coupling between the methyl group and the methylene o<_ to the 29 positive charge unlike !-amyl and methylcyclopentyl cations. cation cationchloride cation chloride methyl cyclopentyl 134 232 98 !_-amyl 142 238 96 !-butyl 138 238 100 2-methylnorbornyl 110 196 86 1,2-dimethylnorbornyl 65 149 2,3,3-trimethylbutyl 64 164 86 exo 82 endo 100 Figure 16 The pmr data are not good enough to justify an unambiguous assignment for the 2-methylnorbornyl cation. How- ever, it is possible to make an assignment consistent with the observed spectrum. The low field peak of one proton at 300 ppm, the secondary alcohols recovered in quenching of the ion, and the value of the chemical shift difference from the chloride to the cation all indicate significant charge delocalization from the tertiary cation. We have ·assigned the low field single proton peak to the proton at C-1. H u XII 30 The ion is assigned the delocalized structure XII because the proposed Wagner-Meerwein shift which places positive charge at C-1 is still fast at the lowest temperatures studied. The protons at C-3, being o( to the positive charge, are assigned to the resonance at 180 cps. The C-4 single proton is also assigned to the 180 cps resonance by analogy to 1,2-dimethylnorbornyl and norbornyl cations, where the chemical shift of this proton is 180 and 190 cps respectively (20). The four protons at C-5 and C-7 and one of the protons at C-6 are assigned to the high field resonance at 117 cps. This leaves the other proton at C-6 at 200 cps. The location of the three methyl protons at 196 cps can be seen from the spectrum of the deuteromethylnorbornyl cation. This assignment is consistent with the spectrum and integrals at -67° C. for both the deuterated and undeuterated cation. This assignment is also consistent with thehigh temperature spectra if 1,6 hydride shifts are allowed to cause fast equilibration of ions A and B (p.32). Since the tertiary ion is present in far larger concentration than the secondary, the presence of such secondary ions will facilitate proton equilibration. Their presence may also cause small downfield shifts due to a small population of secondary ions in rapid equilibrium with the tertiary ion. The Wagner-Meerwein rearrangements and 1,6 hydride shifts 31 between secondary cations equilibrate the hydrogens at C-6 and C-1, at C-5 and C-7, and the ~and endo hydrogens at C-3 as seen in Figure 17. This explains the sharpening of the resonance at 185 due to exchange of the exo and endo protons at C-3. It also explains the upfield shift of the proton at C-1 and the downfield shift of one of the protons at C-6. The upfield resonance at 117 cps sharpens into a syrmnetrical pattern because of equilibration between C-5 and C-7. The low field broad multiplet in the high temperature spectrum is assigned to the protons at C-1 and C-6 which seem to be exchanging at intermediate rate. The upfield resonance and the resonance at 185 cps sharpen substantially while that at ~200 cps remains broad e~en though equilibration occurs by the same rate process because the 'C:·~ term in the nmr line shape equati01 remains larger for the protons at C-1 and C-6 (200 cps) than for the protons at C-7, C-5 and C-3. This is due to the much larger chemical shift difference between the protons at C-1 and C-6 than between the C-5 and C-7 or the C-3 endo and exo protons. The explanation for the relatively slower 1,6 hydride shift and Wagner-Meerwein rearrangement in the 2-methylnorbornyl cation relative to that in norbornyl cation is that the lifetime for equilibration between A and B is dete:rnDne:i by the amount of secondary cation present. This is pre- sumed to be low in comparison to total 2-methylnorbornyl 32 1,~ w.-M. ct'l.Cll3 l 11,c. hydride 11 @Cb..CNs hydride . Figure 17 33 cation. The Wagner-Meerwein rearrangement of tertiary to secondary cation seems to be occurring even at -105° C. This can be seen in the large deshielding of the proton at C-1, indicating some positive charge at this position. If we take its normal shift as 180 cps (proton at C-4), the observed shift due to the charge at C-1 and C-2 is 120 cps. A comparison with the shift difference between the protons at C-3 and those at C-5 and C-7 in 2-methylnorbornyl cation indicates that a deshielding of 80-90 cps is reasonable for protons o( to the positive charge. This indicates that the resonance for the proton at C-1 is about 30 cps farther down field than expected. In light of the secondary products isolated on quenching and the small shift in the methyl resonance between 2-methylnorbornyl cation and chloride (Figure 16), it seems possible that this extra deshielding at C-1 is a result of charge delocalization to the . 1-methylnorbornyl cation formed by a rapid Wagner-Meerwein rearrangement. These spectral observations are not inconsistent with nonclassical formulation for 1,2-dimethylnorbornyl and 2-methylnorbornyl cations. Furthermore, the secondary methylnorbornyl cations seem to be unusually stable with respect to the tertiary 2-methylnorbornyl cation. This extra stability is consistent with the nonclassical formulation for secondary norbornyl cations. 34 The C-6 protons of 1,2-dimethylnorbornyl cation are deshielded 80 cps with respect to the C-5 protons. In 2- methylnorbornyl cation one of the C-6 protons is deshielded by~ 100 cps with respect to the other C-6 proton and the protons at C-5. deshielding. Several explanations can account for this A 1,6 hydride shift would account for de- shielding of the C-6 exo proton in 2-methylnorbornyl cation and of both equivalent protons in the 1,2-dimethylnorbornyl cation. This cannot account for the total 80 cps or 100 cps shift observed, however. If we take the maximum de- shielding of the proton at C-1 d ue to Wagner-Me erwein rearrangement of 2-methylnorbornyl cation as 30 cps (p.331 then the C-6 proton should be deshielded by about 30 cps due to a 1,6 hydride shift. This expectation arises because the stability of the secondary cations formed by a 1,6 hydride shift or Wagner-Meerwein rearrangement should be nearly the same. Another indication that 1,6 hydride shifts are unimportant in 1,2-dimethylnorbornyl cation is · the equality of the chemical shifts due to ionization for this cation and for the 2,3,3-trimethyl-2-butyl cation (Fi gure 16). Since charge delocalization by hydride shift is h ighly unlikely in 2,3,3-trimethyl-2-butyl cation, this observation indicates the absence of such delocalization in 1,2-<iimethylnorbornyl cation. Another effect that could account for the stereospecific deshielding is overlap between the empty £ orbital and the backside lobe of the sp3 35 C-H orbital for the exo proton (Figure 13) (19). This would account for some deshielding of the C-6 ~ proton XIII in 2-methylnorbornyl cation and both C-6 protons in the 1,2-dimethylnorbornyl cation. Another possibility for charge delocalization is through resonance form XIV which has been proposed for nonclassical norbornyl cations. XIV This type of delocalization would deshield both protons at C-6 . Finally, a simple electric field effect on the C-H chemical shift due to a positive charge in the molecule could cause the deshielding. The equations used for relating chemical shift of a proton to the electric field due to excess charge is Arr= - aE 2 - bE2 where Ez is the component of the electric field along the C-H bond. A more convenient form of this equation is 36 fl<:s = 12. 5 x lo-6 L;_, [ CAq cos g 1 ) /Ri ~ -17. o x lo-6 1 ( LAq /R 2)2 . i l i where Ri is the distance from the charge _.6qi to the C-H bond and gi is the angle between the electric field vector and the C-H bond (27). For these calculations Gi and Ri were·evaluated from molecular models fur the classical equilibrating 1,2-dimethylnorbornyl cation and the nonclassical cation where one-half the positive charge is at C-1 and,C-2. The bond distance from C-1 and C-2 to C-6 for the nonclassical case was taken as the average of the C-H bond distance from C-6 to C-2 and the nonbonded distance from C-6 to C-1. The effect of the counter ion was ignored in these calculations. The results of the calculations for · 1,2-dimethylnorbornyl cation are shown in Figure 18. Chemical Shift Difference 22m classical eguilibrating nonclassical observed H5-H5 - 0.07 - 1.14 - 1.3 H5-H4 0.76 0.01 0.06 H5-H3,7 exo 1.04 o.4o - 0.50 H6-H3,7 endo 1.38 0.15 - 0.50 -- Figure 18 The nonclassical model seems to predict the downfield shift of the protons at C-6 while the classical model does not. The calculation does not account for the relative 37 position of the C-6 proton resonance to that of the protons o( to the positive charge. A small amount of positive charge character at C-6 would, however, move these C-6 protons downfield farther. A suggestion for determining · whether part of the deshielding is due to charge delocalization is discussed in the propositions. The unexpected deshielding of the C-6 protons in 1,2-dimethylnorbornyl cation may be explained in two ways. It may be the result of the electric field associated with the charge distribution in the nonclassical cation plus any charge delocalization to C-6. The other explanation is that the cation is classical, and deshielding occurs by overlap of the back lobe of the C-H orbital with the empty cation £orbital. Models indicate that the 1,2-dimethyl- norbornyl cation is not nearly as favorable for such overlap as is admantyl cation where this effect is claimed to cause a deshielding of 0.92 ppm for the '('- protons relative to the tion. f-> protons. Therefore, we favor the first explana- 38 EXPERIMENTAL A. Syntheses 1) 2-Methyl-2-endo-norbornyl alcohol t::l::Lo + ClhMsI cq_CH3 OH Methyl magnesium iodide (0.557 moles) was added with stirring under nitrogen to 40.8 g (0.372 moles) norcamphor (Aldrich). The alcohol product was isolated and distilled to give 39.6 g (0.314 moles, 84%) 2-methyl-2endo-norbornyl alcohol (21). 2) l-Methyl-2-exo-norbornyi nitrate ) 2-Methyl-2-endo-norbornyl alcohol (67 g, 0.532 moles) was added to 370 ml concentrated nitric acid (new bottle, N2 o4 free). This mixture was stirred at room temperature for 10 hours. It was then poured into 1 liter of ice water, and the organic product was extracted with ligroin (30-60). The ligroin was washed with sodium bicarbonate and dried with magnesium sulfate. The crude yield of nitrate ester was 84 g (0.492 moles, 92%) (21). 39 3) 1-Methyl-2-exo-norbornyl alcohol ~ (a) Al\oy Jkva.rda.'s L-f;;J-0ND1. OH e The nitrate ester (84 g, 0 . 492 moles) was dissolved in 1 liter 85% ethanol. To this mixture, portions of 50 g Devarda 1 s alloy and 46 g potassium hydroxide in 46 ml water, were added at 6, 10, 25 and 53 hr. During addition the solution was cooled to keep the stirred mixture at room temperature. The ethanol was removed , and the product steam-distilled to give 35 g (0 . 278 moles, 57%) l-methyl-2exo-norbornyl alcohol: bp 83-88 (20 mm) (21). -Pd, c. Et OH (b) '1;1oH The nitrate ester (8 . 0 g, 0.047 moles) was dissolved in ethanol, and this solution was placed in a Parr reducingapparatus. Catalytic reduction was carried out at 30 psi with 20 g 5% Pd on charcoal. The reduction was com- plete in 2 hr and gave 3.2 g (0 . 025 moles, 54%) alcohol . (c) c/;loH The nitrate ester (12 . 4 g, 0.073 moles) was refluxed overnight with 9.1 g (0.240 moles) lithium aluminum hydride 40 in ether. The excess lithium aluminum hydride was destroyed with ethyl acetate, and the reaction mixture was poured into 200 ml 3 N hydrochloric acid. tracted with ether. The aqueous layer was ex- The ether layer contained 7.4 g (0.058 moles, 81%) crude alcohol product. 4) 1-Methylnorcamphor C/:loH 1-Methyl-2-exo-norbornyl alcohol (35 g, 0.278 moles) - was dissolved in 78 ml acetone, and Jones reagent /:!7 g chromium trioxide (0.270 moles), 18 ml water, 23 ml sulfuric aci~ was added to the alcohol during the course of one hour. was kept at o° During addition of the reagent, the solution C~ The ice bath was removed, and the mix- ture was stirred at room temperature for 30 minutes. The reaction mixture was quenched with 750 ml water, and the product wa s ex tracted with ligroin (30-60). The ligroin solution was washed with sodium bicarbonate and dried. The product, 1-methylnorcamphor : bp 73-80 (19-21 mm), weighed 18.5 g (0.149 moles, 54%). 5) 1,2-dimethyl-2-endo-norbornyl alcohol 41 The ketone (18.5 g, 0.149 moles) was treated with methyl magnesium iodide as before to give 20.7 g (0.148 moles, 99%) 1,2-dimethyl-2-endo-norbornyl alcohol (22). 6) 1,2-Dimethyl-2-chloronorbornane (a) Me OH 1,2-Dimethyl-2-endo-norbornyl alcohol (9.7 g, 0.069 moles) was dissolved in 6 ml methanol. This solution was stirred overnight with 80 ml concentrated hydrochloric acid. The mixture was extracted with ligroin (30-60), and the chloride was isolated and sublimed to give 7.5 g (0.047 moles, 68%) (22). The chloride was dissolved in 40-60% ethanol, and 10 ml 0.0883 N sodium hydroxide was added. The mixture was allowed to stand overnight and was back-titrated with 0.0530 N potassium acid phthalate. 1,2-Dimethyl-2-chloro- norbornane showed 98% purity for a tertiary chloride with molecular weight 158.5. (b) SOCJ.z p~ridirte 1,2-Dimethyl-2-endo-norbornyl alcohol (1.40 g, 0.010 moles) was dissolved in 15 ml dry ether. To this solution 0.790 g pyridine (0.010 moles) and 0.595 g (0.005 moles) 42 thionyl chloride were added at -78° C. The solution was warmed to room temperature, and the pyridine hydrochloride was removed by filtration. The ether was removed with a rotary evaporator, and more thionyl chloride (0.595 g, 0.005 moles) was added at room temperature. This mixture was stirred at room temperature overnight. The product was extracted with ligroin, dried, and the ligroin removed to give 1,2-dimethyl-2-exo-norbornyl chloride. 7) 2-Methyl-2-chloronorbornane 50Uz 2-Methyl-2-endo-norbornyl alcohol (7.18 g, 0.057 moles) was treated with pyridine-thionyl chloride as before. The product was 6.65 g (0.046 moles, 80%) 2-methyl-2chloronorbornane. Titration as before showed the product to be 97% tertiary chloride of molecular weight 144.5. 8) l-Methyl-2-chloronorbornane 2-Methyl-2-endo-norbornyl alcohol, when treated with hydrochloric acid as before, gave predominantly (~ 80%) 1-methyl-2-chloronorbornane. 43 9) 1,2-Dimethyl-2-exo-norbornyl alcohol r--J'1 OHS ~H~ rf"1 ~ ~c_°~ 1,2-Dimethyl-2-chloronorbornane (1.7 g, 0.011 moles) was dissolved in 5.2 ml acetone, 1.7 ml ether, and 3 ml water. To the solution 12.1 ml 1 N sodium hydroxide was added. The solution was stirred overnight and extracted with ligroin (30-60). The product was chromatographed on alumina to give 432 mg (0.003 moles, 29%) 1,2-dimethyl-2exo-norbornyl alcohol (22). 10) 1-Methyl-1-chlorocyclopentane HO f_M3 6 HCl Methylcyclopentyl alcohol (9.0 g, 0.090 moles, Aldrich) was placed in a test tube, and dry hydrochloric acid was bubbled through the alcohol for 30 minutes. After initial addition of hydrochloric acid, the reaction vessel was cooled in an i ce bath . The product was extracted with ether, and this ether solution was washed with sodium bicarbonate and dried. The chloride: bp 105-115, was dis- tilled to give 5.0 g (0 . 042 moles, 47%) l-methyl-1-chlorocyclopentane (28). Titration as before indicated the product to be 98% tertiary chloride of molecular weight 118.5. 44 B. Deuterated Compounds 1) 2-Methyl-d 3 -2-endo-norbornyl alcohol Deuteromethyl magnesium iodide (0.012 moles) was prepared from 16.8 g (0.012 moles) deuteromethyl iodide (99% d 3 , Merck of Canada) and 2.84 g (0.012 moles) magnesium. Norcamphor (11.87 g, 0.011 moles, Aldrich) was treated as before with the Grignard reagent. The product was isolated as 8.9 g (0.007 moles, 64%) 2-methyl-d 3 -2-endo-norbornyl alcohol: bp 83 (20 mm). 2) 1,2-Dimethyl-d 4 • 1 -2-endo-norbornyl alcohol 1-Methyl-d 1 . 1 -2-norbornyl ketone (3.0 g, 0.024 moles) · in 10 ml ether was treated with Grignard reagent prepared from 3.85 g (0.029 moles) deuteromethyl iodide and 647 mg (0.027 moles) magnesium. The product was isolated and chromatographed on alumina to give 900 mg (0.006 moles, 26%) 1;2-dimethyl-d4 . 1 -2-endo-norbornyl alcohol. 3) l-Methyl-d 3 -cyclopentyl alcohol Cyclopentanone (bp 30°;Matheson, Coleman, Bell) was treated with 10% excess deuteromethyl magnesium iodide as before. The product was 300 mg deuteromethylcyclopentyl alcohol. C. Epimerization Study 1,2-Dimethyl-2-exo-norbornyl alcohol (14.7 mg) and 1,2-dimethyl-2-endo-norbornyl alcohol were dissolved in 2.5 ml sulfur dioxide. The solutions were sealed in glass tubes and heated to 60-65° C. Tubes were opened and quenched in cold saturated sodium bicarbonate solution. The alcohols were extracted with ether and analyzed by gasliquid partition chromatography. The column used in analysis was 12 feet of 10% carbowax 20M on chromasorb G. The epimerization was carried on until there was no change in the exo:endo ratio. The results were as follows: 46 starting alcohol % exo % endo from endo 53 47 from exo 52 48 from endo 55 45 from exo 54 46 54 46 average 6Geq =.92.4 cal/mole Figure 19 D. Cation Formation Cations were produced from either alcohol (exo or endo) or chlorides. is described. A typical preparation from an alcohol 1,2-Dimethyl-2-endo-norbornyl alcohol (227 mg) was dissolved in 1.7 ml sulfurylchlorofluoride (Matheson, Coleman, Bell) at -78° C. The alcohol was dissolved by warming, and the solution was cooled to -78° C. sulfonic acid (360 p-1, K. & K., redistilled) pentafluoride (450 Fluoro- and antimony 141, Allied Chemical, distilled at atmospheric pressure under argon in a quartz distilling apparatuaj were placed in the reaction vessel (Figure 20). Sulfur dioxide was condensed in the reaction vessel until the antimony pentafluoride-sulfur dioxide complex dissolved. The alcohol dissolved in the sulfurylchlorofluoride was removed and transferred to the reaction vessel in a 0.5 ml glass syringe. The syringe was fitted with a small gauge #20 needle and was precooled in liquid nitrogen. Stirring 47 Figure 20 was provided by bubbling nitrogen through the sulfur dioxide solution of acids. The addition of alcohol was carried out slowly with vigorous stirring. The entire reac- tion vessel and nmr tube were immersed in a -78° C.. bath during cation preparation. The cation was sucked into the nmr tube; the tube was removed and capped or sealed. capillary tube was added to each tube as a pmr A TMS reference. All spectra were taken on the Varian A56-6o spectrometer at low temperature. Many of the spectra were accumulated on a time-averaging device, CAT, using 1024 memory cells to accumulate the spectra. All cation quenching experiments were carried out by pouring the cation into ice water, extracting with ether, and washing the ether layer with sodium bicarbonate. These 48 quenched solutions were analyzed by gas-liquid partition chromatography on a carbowax column as previously described. The pmr and ir spectra of the products were also recorded. 49 INTRODUCTION A recent article has reported the exchange between trityl halides and trityl carboniurn ions as observed by pmr (29). By studying the concentration dependence of exchange induced line broadening, the authors determined that the reaction is an SN Ar3C ...Cl - 1 process (Figure 21). + Ar~c"*" y- ~ The Ar3C .... y- +Ar~C +c1 - Figure 21 authors extended their studies to give relative cation stabilities by carrying out exchange reactions between two cations (Figure 22). By integrating the signals in the K~ ~ + ') RX + R 0 y- Figure 22 slow exchange case or observing the average signal of a sharp methyl or methoxy group, it proved possible to evaluate Keq (30). The technique has been used only for those compounds whose cations have been crystalized. Several other methods of evaluating relative cation stabilities have been investigated. One of these is the study of the carbinol-carbonium ion equilibrium in strongly acidic solution (usually sulfuric acid). The resulting 50 carbonium ion concentration is measured by ultraviolet spectroscopy (31). This uv method is limited to those aromatic cations which are sufficiently ionized in sulfuric acid. This method also relies on the validity of the HR acidity function. More recently relative cation stabilities have been studied by comparing E0 measurements for the half cell sha-.1n in Figure 23. The comparison of the half cell reaction ..... R +e ~ ! R-R Figure 23 for two cations gives their relative stability. These stability differences are close to those measured by the carbinol-carbonium ion equilibrium in strong acid (32). Again this method requires an easily prepared cation and has been used only for trityl cations. The heats of formation (calorimetric) of several carbonium ions have been measured in antimony pentafluoridefluorosulfonic acid mixtures. been studied in this way. Two aliphatic compounds have Cyclohexene and methylcyclopen- tene both produce methylcyclopentyl cation with heats of -16.8 and -18.3 kcal/mole, respec- formation tively, The difference of 1.5 kcal/mole corresponds to the difference of 1.4 kcal/mole in the heats of formation of the alkenes (33). It is our intention to extend some of these cation 51 stability measurements to aliphatic cations in sulfur dioxide solutions of antimony pentafluoride. We feel that such relative stability measurements will have bearing on the nonclassical formulation for certain aliphatic cations. 52 RESULTS Several cation-chloride mixtures were studied in solutions containing less than one equivalent of antimony pentafluoride. We needed to know first if the cation- chloride exchange was fast or slow on the nmr time scale and second if the reaction of a tertiary chloride with antimony pentafluoride was quantitative. The most care- fully studied cation was ~-butyl. A saturated solution of antimony pentafluoride in sulfur dioxide was prepared, and 0.5 ml aliquots were withdrawn in a cold syringe fitted with a teflon needle. Each aliquot was sealed in a glass tube and kept in a freezer. These aliquots were then titrated with ~-butyl chloride by adding weighed amounts of the chloride and observing the average chloride-cation resonance (Figure 24). mg ~-BuCl 1 2 3 4 5 6 7 8 9 10 11 12 33.2 51. 6 52.8 64 .2 66.1 75.4 75.4 76.3 101.5 120.9 152.7 250.8 chem shift (TMS) mmoles t-BuCl fraction cation Aobs<:i - A el. A cod-. - A<?Jl 238 cps 242 148 231 194 140 154 159 167 145 137 115 105 100 cps 0.358 0.557 0.570 o.692 0.714 0.814 0.814 0.824 1.095 1.302 1.648 2.705 100% 100% 35% 96% 69% 29% 40% 43% 46% 33% 28% 11% 4% 0% Figure 24 53 It can be seen from Figure 24 that the results are only semi-quantitative, probably because of the difficulty of keeping water out of the antimony pentafluoride solutions. It is felt that it is still possible to get a reasonable estimate of Keq and the average concentration of the acid solutions. A graph of % cation against chloride concentration is shown in Figure 25. The general trend is toward lower shifts at larger ~-butyl chloride concentrations. To simultaneously determine Keq and the acid concentration the following model was chosen. chloride + SbF 5 !~ >- cation+ +- SbF 5 c1 - 6 :.fraction cation= .bobs- ..6c1 tfcat. - $Cl X HA = rnmoles t-BuCl =original SbF5 - rnmoles AX =cation - rnmoles Keq - ~cation)+ · (SbFP c1-) Keq -- (AX }2 (X- AX) (HA-AX) Keq - ( AX) 2 HAX-HA AX-X AX+ ( AX) 2 Keq - chloride) · 1 HA &X. 1 Keq -- (Sb 5) HA - HA AX ;:... 1 - +l o · (t:3~x ·~ ~x) + ( 1 - i) I I I l> .2. I I e4 M -~ .s 6 c.a:t. - ~Cl ~ - 6Ci 1.0!1 .4 ~0 z0 30 50 40 <l0 ~1..fo 310 Figure 25 s-Bu.. chlor-.1de 100 i -2:, "f'Y\-n"\O \ e S .s "'fl '°0 s z.o f..'f 0 1z. l .8 +=- Vl 55 K~q ~ ~ - .· 1)~(1 -_ ~) ( A plot of l/Keq against HA will be a straight line for each value of 6. and X. The intersection of all these lines will give the values of l/Keq and HA. The lines are con- structed from the values of ~ and X at the intercepts. for the l/Keq = 0 intercept HA =6X for the HA - 0 intercept 1~ 1/Keq -::.. l - The evaluation of these intercepts is shown in Figure 26. 12 3 4 r '.) r 0 7 8 9 10 11 12 J_ A x A 1.03 0.35 0.96 o. 69 0.29 o.4o o.43 o.49 0.33 0.28 0.11 0.04 0.358 0.557 0.570 0.692 0.714 0.814 0.814 0.824 1.095 1.302 1.648 2.708 0.97 2.86 1.04 1.45 3.45 2.50 2.32 2.17 3.03 3.57 9.09 25.0 ' L::::.Y. l - b.. - 0.369 0.195 0.548 o.477 0.207 0.325 0.350 o.403 0.362 0.370 0.180 0.110 0.03 -1.86 -0.04 -0 .l~5 -2.45 -1.50 -1.32 -1.17 -2.03 -2.57 -8.09 -24.o Figure 26 All the lines on the graph should intersect at one common point (Figure 27). That they do not is a conse- quence of varying acid concentrations. Most of the points Of intersection do fall in the area marked with a circle in Figure 27. We feel this is consistent with the qualitative observations that Keq is large (i.e. almost 1.00 ,, . -:1.CTO - t .oo HA- mmole.s Figure 27 , 57 quantitative reaction of the antimony pentafluoride). The acid concentrations seem to fall between 0.3 and 0.5 mmoles per half-milliliter tube. We feel points which represent large amounts of !_-butyl chloride do not converge near the others because the large amounts of chloride used change the character of the solvent. If we now choose the points which do seem to come close to convergence (i.e. all but 2, 5, 11, and 12) and impose the condition l/Keq~O, iA =,6..x Therefore, a plot of slope HA. 1 - 6. vs X should be a straight line of The least squares plot shown in Figure 28 gives HA:o.297::!:0.0Lf 4. The plot shows straight line behavior (variance 0.0759) indicating that our model with ~Kl eq .-0 is not too far afield. Solutions of !_-butyl chloride and !_-butyl cation gave mixtures of !_-butyl chloride and tbutyl alcohol when quenched with water. Other cation-chloride solutions were observed in less than one equivalent of antimony pentafluoride to see if they were stable. These fell into two groups at -78° C; those that showed fast exchange (averaged resonances) and those which showed slow chloride cation exchange. !_-Amyl cation (XIII) in less than one equivalent of acid showed fast exchange. are shown in Figure 29. The results of the experiments The acid concentration was 58 4.o 3.5 3.o 2.5 2.0 1.5 -D. j_ i· O ~mmm.o----'-----------------------------01 x .4 .. (o .g Figure 28 1.2. 59 estimated by titration with !-butyl chloride. The data · x I - c I CH x = ® a X .:: Cl 3 XIII in Figure 29 indicate that in both cases !-amyl chloride is equilibrating between chloride and cation at fast exchange. This is reflected in the averaged chemical shifts and coupling constants for the two equivalent methyl groups. Assigning the percentage of cation in the two cases using the low field methyl resonance gives 56% and 61% . This also is consistent with JHa-Hb =. 3.5 cps, indicating nearly half chloride and half cation. The singlet account- ing for about 30% of the total integral is assigned to ~­ butyl chloride and ~-butyl cation. This is confirmed by a quenching experiment in which ~-butyl alcohol was produced as well as ~-amyl chloride and alcohol. Cleavage of ~-amyl cation to give small amounts of t-butyl cation has been previously reported (34). Another carbonium ion for which the exchange reaction is fast is 2,3,3-trimethyl butyl cation (XIV). for the exchange are shown in Figure 30. The data The singlet in the spectrum of 2,3,3-trimethylbutyl chloride in less than one equivalent of antimony pentafluoride indicates that Figure 29 182 cps (triplet JHA-Hb:::3.5 cps % 1 nunole ~-amyl 87 cps 86 cps 64 cps (triplet) 102 cps (quartet - two peaks under singlet) 0.77 nunoles t-amyl chloride 0.45-0.55 mmoles SbF 5 96 cps (singlet) JHA-HB=.. O 105 cps (triplet c 265 (multiplet) 175 cps 200 cps (triplet) (broad multiplet) JHA- HB= 3. Ocps chloride ~-amyl 238 cp s (triplet) JHA-HB =7cps b chloride o.45-0.55 nunoles SbF 5 cation ~-amyl a °' 141 cps singlet singlet total integral 138 cps 0 ~1/3 other 61 a b )H3 1H3 - c - c I X x=@ J CH 3 x=c1 b XIV exchange between chloride and cation is fast on the nmr time scale. One of the cations studied, methylcyclopentyl, decomposed at -78° C., when less than one equivalent of antimony pentafluoride was added. The spectrum was recorded at -70° C., -50° C., and -30° C. to verify that the broad resonances observed at -105 cps and a sharper peak at -50 cps were not the result of intermediate exchange rate between cation and chloride. independent. The spectrum was temperature Quenching experiments on the cation revealed no methylcyclopentyl alcohol. Both products formed had gas- liquid partition chromatography retention times longer than the alcohol. Two of the cations showed slow exchange between cation and chloride. The results (for 1,2-dimethylnorbor- nyl cation) are shown in Figure 31. The results of the 2-methylnorbornyl cation experiment in less than one equivalent of acid are shown in Figure 32. These data show slow exchange between the 2-methylnorbornyl cation and 1-methyl- 2,3,3-trimethyl butyl chloride (.96 mmoles) antimony pentafluoride ( .5-.6 mmoles) 2,3,3-trimethyl butyl chloride 2,3,3-trimethyl butyl cation 164 cps (singlet ) b Figure 30 11% cation from linear interpolation 90 cps (broad singlet) 100 cps 68 cps (average weighted for number of protons) 81 cps a 0\ [\) methyl resonance 1,2 dimethylnorbornyl cation 149 cps 1,2 dimethylnorbornyl chloride 86, 62 82, 55 1,2-dimeth¥1norbornyl chloride ( L 9 moles) antimony pentaf luoride ( 1.1 moles) 145 ~33% endo ~67% exo 90, 65 85, 59 ~40% ~60% endo exo ~35% integration gives cation Figure 31 methyl resonance 2-methylnorbornyl cation 194 cps (singlet) 180 cps methylene singlet 2-methyl-2-chloronorbornane 86 cps 1-methyl-2-chloronorbornane 65 cps ' ct .,..C.~H 220 cps = 2-methyl-2-chloronorbornane .0.5 equiv. antimony pentafluoride 194 cps 70 cps 228 cps (180 cps) ' Cl /C:_H Figure 32 2-chloronorbornane. There is no evidence of the tertiary isomer in the spectrum, indicating that the 1-methyl-2chloronorbornane is several kcal more stable solution. in this 64 Several cations were equilibrated with ~-butyl cation to obtain semi-quantitative data on their relative stabilities. For each of these experiments the sealed tutes of antimony pentafluoride containing 0.3-0.5 mmoles The data are summarized in Figure 33. were used. The isopropyl doublet has a chemical shift of 88 cps in sulfur dioxide. A few comments can be made about these data. It is clear that methylcyclopentyl cation not only decomposes, but that it also removes the antimony pentafluoride from solution. Otherwise the ~-butyl resonance would be much f arther downfield in the second equilibration experiment in Figure 33. It is also clear that the chloride present in the third experiment is the rearranged 1-methyl-2-chloronorbornane. It is possible to estimate the minimum free energy differences from these experiments if we assign the maximum amount of less stable cation and minimum amount of more stable cation. These estimates are shown in Figure 34 and are based on estimated chemical shift errors or integration errors. These equilibration experiments were per- formed at equilibrium. methylnorbornyl ~-butyl This was checked with the 1,2-di- cation-~-butyl cation equilibration. The chloride was added to the acid first, and the results were the same as for simultaneous addition. All the ions in Figure 34 are too far apart in energy to measUIB their relative stabilities accurately, indicating that the 0 .347 0 .356 0.500 o .476 is opropyl c hloride methylcyclopentyl chloride 2-methylnorbornyl chloride 1,2-d imethylnorbornyl chloride - mmoles t-BuCl o.492 0.500 0.310 0. 350 mmoles chloride II none obsvd. 71 c ation decomposed CHl ch oride II b2 1 CAT scans % % 150 202 97 >85%b 70'{! large smal l 0% 0% ~ 8% 100% ca~ion II cationII t-B1.J±) CH in 0.3-0.5 mmoles SbF 5 a 16 CAT scans Figure 33 100 102 111 240 3 t-Bu-CH Equil~pration of t-Bu-t and R+ 380 t-butyl _!-propyl t-butyl dimethylnorbornyl t-butyl methylnorbornyl 136 36 -- cations t-butyl methylcyclopentyl Keq min. 85% 60% ? 96% minimum more stable Figure 34 kcal/mole 2.0 kcal/mole 1.4 kcal/mole 2.3 min. AG 4% 4% 10% 6% maximum less stable 0\ 0\ 67 cyclic cations are much more stable than !-butyl cation. · This is especially reflected by 2-methylnorbornyl cation, where the difference in ground state free energy between the tertiary and secondary chlorides should partially offset the difference in cation stability (Figure 35). In spite of this, the equilibrium constant is s,till too large to be measured by this method. MN© t E t-BuCl 30 MN-Cl 20 MN-Cl Figure 35 Two of the cations equilibrated with !-butyl cation were close enough in energy to make quantitative estimates of relative stabilities. The data for the !-amyl and 2,3, 3-trimethylbutyl cation equilibrations with !-butyl cation are shown in Figure 36. Again these experiments were performed with sealed tubes, judged to contain about 0.45- 0.55 moles antimony pentafluoride. It is clear from inte- gration of the peaks for the !-butyl-!-amyl equilibration that no appreciable decomposition of !-amyl cation to !-butyl has occurred; so that the molar ratios of total t-amyl 151 cps 183 cps CH3 ~-butyl 193 cps (triplet J H-¥ :::. 5 cps (mu tiplet) 233 cps 211 cps T~3)2-C t-amyl cont'd page 69 *evaluated by tracing spectra and weighing - (o.49 mmoles) t-amyl - (0.47 mmoles) t-butyl - '(o . 47 mmoles) calc: t-butyl 0.96 t-amyl obsvd: * t-butyl 1.08 t-amyl t-butyl - '(o.45 mmoles) area cation 105 cps 90 cps CH2~3 36% 59% t-Bu+ - % 68% 81% t-amyl + % 0\ co 0.51 0.55 t-butyl 2,3,3-trimethylbutyl 1.53 1. 77 1. 65-x- 2. 08 * 131 106 151 115 TMB CH 3 22% 4% Figure 36 ~ 8lJ % Lt1% ~MB+ width t-butyl - % + t-Bu *integral average of tracing and weighing and pk x singlet (9 proton s ) : TMB singlet (15 protons) 0.58 0.52 mmoles t-butyl 2,3,3-trimethy lbutyl cation r.e la ti ve area obsvd: calc: t-Butyl TMB TMB -CH !_- butyl !_- butyl 3 CJ'\ \0 70 !-butyl and !_-amyl species is about the same. Equilibrium constants and free energy differences are shown in Figure 37. Keq .AG (in kcal /mole) 1) !_-butyl-!_-amyl 3.0 o.435 2) !_-butyl-!_-amyl 3.8 0.527 1) t-butyl-TMB 16.7 1.11 2) !_-butyl-TMB 18.6 1.16 av.0.48 kcal/mole av.1.14 kcal/mole Figure 37 The mixtures of E_-amyl and E_-butyl cation were quenched, and the major products were !_-butyl alcohol, !_-amyl chloride, and !-amyl alcohol. The !-butyl chloride could not be separated from ether on the carbowax column. Integrated areas showed the amounts of !_-butyl alcohol and !-amyl alcohol to be about the same (no internal standard was used, however). Quenching of mixtures of E_-butyl cation and 2,3,3trimethylbutyl cation with water gave !-butyl alcohol, 2,3,3-trimethyl-2-chlorobutane, and 2,3,3-trimethyl-2-butyl alcohol. The ratios of integrated areas were b.37, 0.13, and 0.50. Even though methylcyclopentyl cation is not stable in less than one equivalent of acid, the original cation formed did not release the antimony pentafluoride complexed 71 to it as the cation decomposed. It appeared that the decomposition occurred from the cation and was essentially irreversible (no methylcyclopentyl alcohol in the products of quenching). Therefore, it seemed possible to gain some idea about the original cation equilibrium by using the ~0.3-0.5 moles of antimony pentafluoride to carry out equilibrations with methylcyclopentyl cation. Several qualitative experiments indicated that the dimethylnorbornyl cation is more stable than the methylcyclopentyl cation. There seems to be little enough methylcyclopentyl cation present in the solution containing appreciable amounts of 1,2-dimethylnorbornyl cation that the methylcyclopentyl chloride does not decompose at an observable rate. The results of one cation experiment with methylcyclopentyl chloride and 1,2-dimethylnorbornyl chloride in antimony pentafluoride are given in Figure 38. Methyl - cyclopentyl chloride in sulfur dioxide has a methyl resonance at 98 cps. The species observed are 1,2-dimethyl- norbornyl cation, 1,2-dimethyl-2-exo- and endo-norbornyl chloride, and methylcyclopentyl chloride. All the re s onances are farther downfield than normal, but the assi gnment is still clear. The ratio of methylcyclopentyl ch loride to the dimethylnorbornyl species is 1.23, which compares to a molar ratio of 1.05. This is an indication that none of the methylcyclopentyl chloride has been lost 3.6 157 7.5 106 13.7 113 methylcyclo·pentyl chloride 3 protons methylcyclopentyl _ dimethylnorbornyl -1. 23 84 dimethylnorbornyl chloride - 6 ptns shift Figure 38 integrals by tracing and weighing from CAT spectra (20 scans) 32% dimethyl cation/total methylcyclopentyl chloride 0.516 mmoles dimethylnorbornyl chloride 0.490 rmnoles 3 dimethylnorbornyl cation - 10 protons CH f\) --.J 73 by decomposition through the cation which in turn is a qualitative indication of greater stability of the di§futhylnorbornyl cation Direct equilibration of 1,2-dimethylnorbornyl and 2-methylnorbornyl cations was attempted. In two experiments on this system, the only species detected in the spectrum were the 1,2-dimethylnorbornyl cation and l-methyl-2-chloronorbornane. In the first experiment there was still some 1,2-dimethyl-2-chloronorbornane left, as can be seen by the values of the integrals. This could not be integrated because it was superimposed on peaks from the 2-methyl-2chloronorbornane. In the second experiment the relative integrals confirm that most of the dimethylnorbornyl species are present as the cation, and most of the methylnorbornyl species as the l-methyl-2-chloronorbornane. These data indicate that there is a large free energy difference for the exchange reaction. An estimate for the minimum free energy difference for this equilibration is - ~1.4 kcal/mole. This comes from a maximum estimate of 20% for 1,2-dimethyl-2-chloronorbornane which gives a total integral for dimethylnorbornyl species of 3.9. Using the integrals in equilibration (2) of Figure 39, this leaves a maximum for methylnorbornyl cation of 10% and Keq minimum =36. An independent equilibration of the tertiary 2methyl-2-chloronorbornane and the secondary l-methyl-2- 5 (240 pl; ~3.3 'mmoles ) SbF5 ( 215 µl; ~3. 0 1nmoles ) 1,2-dimethyl-2-chloronorbornane (356 mg; 2.25 mmoles) (2) 2-methyl-2-chloronorbornane (326 mg; 2.25 mmoles) SbF 1,2-dimethyl-2-chloronorbornane (476 mg; 3.0 mmoles ) (1) 2-methyl-2-chloronorbornane (434 mg; 3.0 mmoles ) 3.3 4.1 integral proton 63 68 cps methyl chloride CH re sonance 3 3.6 8.5 integral proton 'c 'Cl /H 227 224 R/ R Figure 39 integrals were evaluated by tracing and weighing (av. of 2 determinations) 142 cps 146 c ps d imethyl cation-CH resonance 3 +::- --..:] 75 chloronorbornane indicated an equilibrium (i.e. no change with time) concentration of )90% tertiary isomer. There- fore, the free energy difference between the tertiary and secondary isomers is~ 1. O kcal/mole. It is impossible to tell if the 1,2-dimethylnorbornyl cation is significantly more stable than the methylnorbornyl cation because of the isomerization of 2-methyl-2-chloronorbornane. When less acid is introduced, the rate of isomerization of 2-methyl-2-chloronorbornane is slowed to an observable rate. Dimethylnorbornyl chloride (0.515 mmoles) and methylnorbornyl chloride (0.502 mmoles) were dissolved in sulfurylchlorofluoride. The chloride solution was added to 20 pl antimony pentafluoride in sulfur dioxide. The pmr spectrum of the solution showed no dimethylnorbornyl cation, so another 20p1 of antimony pentafluoride was added. The results of the experiment are shown in Figure 40. These data are consistent with the scheme shown Figure 41. peak height - methyl time after mixing t::::. 0 t-= 2 min. t=30 min. t=40 min. t > 40 min. 153 cps 106 cps 102 cps 97 cps 79 cps 72 cps 14 14 14 14 26 21 15 15 15 24 18 19 19 19 17 16 15 16 16 20 22 29 28 28 13 9 9 8 8 Figure 40 in 76 Figure 41 The methyl resonances were assigned as follows: 153 cps - 1,2-dimethylnorbornyl cation; 106 cps - 2-methylnorbornyl chloride; 102, 97, 79, and 72 cps - dimethylnorbornyl c.hloride, and 79 cps - 1-methylnorbornyl chloride. was no methylnorbornyl cation visible. There The integrals from the time-averaged spectrum are shown in Figure 42. The ratio of total dimethylnorbornyl species and methylnorbornyl species (0.83) is slightly below the expected molar ratio of 0.97. The percentage of dimethylnorbornyl cation is 24%, and no methylnorbornyl cation is seen in the spectrum. This indicated that the dimethylnorbornyl cation is the more stable. If we assign the difference between stabilities of 1,2-dimethylnorbornyl cation and 2-methylnorbornyl cation to an electronic inductive effect, 77 integral integral corrected for # protons 153 106 102 & 97 79 & 72 40.6 40.9 41.0 z+o.13" 27.2 26.8 26.0 26.7 39.5 40.6 40.5 40.2 56.8 56.7 57.4 57.0 12.3 26.7 40.2 57.0 DMN DMN-Cl MN-Cl 12.3 40.2 26.7+(57.o 40.2)=43.5 corrected integral 24% DMN~ cation total DMN/MNN • 83 integrals from CAT spectrum are by tracing and weighing. Figure 42 we predict that the equilibrium constant for the equilibration will be the same order as that for the !_-butyl - ~-amyl system (Keq::::.3.4). This means that 8% of the methylnorbor- nyl species will be cation. After isomerization of the tertiary methylnorbornyl chloride, the energy difference will be at least 1.4 kcal/mole (see p. 73). Thus the equilibrium constant will change from Keq::3.4 to at 1 least Keq.=35. The net result of this would be the release of the antimony pentafluoride associated with methylnorbornyl cation. The integral of the methyl groups of dimethylnorbornyl cation should increase by ::::;: 30% relative to the integral for the methyl groups of all the chloride species. Figure 43. That this is not the case is seen in 78 integral time 153 cps 106,102,97,79,and 72 cps 2 min. 27 73 30 min. 22 78 40 min. 24 76 >4o min. 25 75 24 !: 5. 9 (95% confidence) 76 average Integrals were evaluated by tracing and weighing; numbers are the average of two determinations and are normalized. Figure 43 79 DISCUSSION It is interesting to compare results of cation equilibrations to solvolysis experiments in order to see whether the cation stabilities are reflected in solvolysis rate experiments. The expectation that the difference in cation stability and solvolysis rate ratios should correJate is borne out qualitatively in all cases but 5 and 7 in Fi gure 44. The solvolysis work on the substituted !_-butyl system was done in 1949 by H. C. Brown . Instead of find- ing a simple addition of methyl inductive effects in substitution of methyl groups onto !_-butyl chloride, Brown discovered that this system was more complicated. On sub s titution of the first methyl (!_-amyl chloride) the rate of solvolysis increased. The introduction of a second methyl depressed the rate below that for !_-butyl chloride, while the third methyl substitution caused a rate increase. This behavior was described in terms of changing relative ma ~ nitudes effe cts. of the inductiveJ hyperconjugative, and steric On the substitution of one methyl group, the inductive effect predominates, increasing the rate. The second methyl group decreases the rate because the hyperconjugat i ve effect predominatesJ and the 1-propyl group is not as stabilizing as an ethyl group. The rate goes up with the third substitution because steric stress facilitates the solvolysis (35). 1.9 1.88 x io 1 (c) 8) dimethylnorbornyl methylcyclopentyl 6G large 1.14 o.48 >2. 0 (min) >1.4 (min) large ">2.3 (min) (equil.) ! Figure 44 4.8Cc) 9) dimethylnorbornyl 1.0 ')>0.5 methylnorbornyl al ref. 9 25° C chlorides b ref. 35 25° C chlorides c ref. 12 50° C p-nitrobenzoates d estimated from methylnorbornyl chloride-!.-butylchloride relative rate and methylnorbornyl (PNB) and dimethylnorbornyl (PNB). (exo isomer) 0.11 1.21(b) - 7) 2,3,3-trimethyl-t-butyl - !.-butyl o.88(b) 6) 2,3-dimethy l-t-butyl !.-butyl -0.08 0.30 1.67 (b) 5) !.-amyl - !.-butyl 3.5 3.50 x 102 (a) 3) methylnorborny l - !.-butyl 4.4 2.5 6.53 x 10 1 (a) 2) methylcyclope ntyl - !.-butyl 4) dimethylnorbornyl - !.-butyl ~. 7 x 103 (d) 6.4 AG kcal/mole ~rom rel. rates) 5.50 x 104 (a) relative rates 1) !.-buty l - i-propyl cations CP 0 81 Our cation equilibrations do not follow the order observed for the solvolyses in this particular series. The first methyl group is observed to stabilize the cation, and the second and third methyl substitutions are also stabilizing in a slightly less than additive fashion. We do not feel inclined to invoke this unusual juggling of hyperconjugative and steric effects to explain our data. presently studying the 2,3-dimethyl-~-butyl We are cation to see if it fits in our proposed progression or is less stable than ~-butyl and ~-amyl cation as the solvolysis experiments would predict. It is possible that the hyperconjugative and steric arguments are not necessary to explain the effect of methyl substitution on t-butyl cation stability. The results obtained so far can be correlated on the basis of increasing cation stability due to greater inductive election release by methyl groups compared to hydrogen. The question concerning cation stabilization by methyl or methylene bridging is our main concern. The charge in 2,3,3-trimethylbutyl cation is distributed over two tertiary carbon atoms (see Figure 16). Furthermore the methyl shift which causes this charge distribution is associated with a maximum free energy barrier of 2-3 kcal/mole (18). These data suggest the possibility of stabilization of the cation through methyl bridging (Figure 45). 82 CH ,'' . it> '' ' '\ CH3 - c - c - CH 3 I CH3 CH3 ' Figure 45 Such bridging would be very ineffective in !-amyl cation, because the methyl shift would generate the much less stable primary cation. Therefore the substitution of the second and third methyl group in !-butyl cation should be much more stabilizing than the substitution of the first if the nonclassical structure (Figure 45) were important. In fact, the 2,3,3-trimethylbutyl cation shows no extra stabilization over that predicted from the effect of the first methyl substitution. If the second and third methyl groupsshowed the same effect as the first, the free energy difference between !-butyl cation and 2,3,3-trimethyl-2butyl cation would be 1.44 kcal/mole. is 6 6 =1.14 kcal/mole. The measured value It is apparent that methyl bridg- ing adds no extra stability to this cation even though the positive charge of the carbonium ion is distributed over t wo carbon atoms. This observation is in agreement with that made earlier by J. D. Roberts that the non- classical structure does not contribute significantly in reaction of 2,3,3-trimethyl-2-butanol with hydrochloric acid (5). The same conclusion has been reached by 83 H. C. Brown for solvolysis and hydrochlorination reactions of 2,3,3-trimethylbutyl derivatives (6). It is more difficult to reach a conclusion on the role of methylene bridging in the tertiary bycyclic cations. The two relative stability measurements which have application to this question are those between 1,2-dimethylnorbornyl cation and methylcyclopentyl cation or 2-methylnorbornyl cation. The most conclusive observation we can make from these measurements is that they are not consistent with our expectations for a classical 1,2-dimethylnorbornyl cation. H. C. Brown has advocated the use of methylcyclopentyl cations as a model system for classical bicyclic cations (8). Chemical shift data on ionization . of methylcyclopentyl cation in antimony pentafluoride indicate that the positive charge is localized on one carbon (Figure 16). Our expectation is, therefore, that the classical 1,2-dimethylnorbornyl cation should be more stable than methylcyclopentyl cation only by virtue of a sm.a ll methyl inductive effect. ~0.48 Such an inductive effect is kcal/mole in ~-amyl cation. Therefore, we believe the observation that methylcyclopentyl chloride competes poorly with 1,2-dimethylnorbornyl chloride for antimony pentafluoride is inconsistent with our expectations for a classical 1,2-dimethylnorbornyl cation. The direct equilibration of 2-methylnorbornyl cation and 1,2-dimethylnorbornyl cation is more convincing. In 84 the absence of steric strain on introduction of a methyl group at C-1 the effect of such a substitution should be to stabilize 1,2-dimethylnorbornyl cation over 2-methylnorbornyl cation by about 0.48 kcal/mole. Models indicate little steric strain on introduction of the methyl group at C-1. Furthermore other authors have also observed that they would expect l-methyl-2-norbornyl alcohol to be free of substituent-induced strains (10). Finally, there are some experimental data that qualitatively indicate little steric strain on introduction of the methyl group. The methyl group is larger in terms of steric interaction than either a chloro or an hydroxyl group. This difference is reflected in the following A values: methyl; 1.5-1.9; chloro, 0.3-0.5; and hydroxyl, o.4-0.9 (36). These values are the energy differences in kcal/mole between the axial and equatorial substituents in cyclohexane. The steric difference is also reflected in an axial-axial methyl-hydroxyl interaction of 2.2-2.4 kcal/mole and a methyl-methyl interaction of 3.7 kcal/mole (36). The most serious steric effect of a methyl substitution at C-1 would be the interaction with the substituent in the exo position at C-2. Therefore, if this interaction were large, an exo chloro or hydroxyl group would be favored over an exo methyl group. The net result is that in comparison to the exo-endo equilibria in other norbornyl compounds, 1,2-dimethylnorbornyl compounds should show a large exo:endo ratio. case (Figure 46). Our data indicate this is not the In the absence of further evidence we will assume the ground state free energies of 2-methyl2-chloronorbornane and 1,2-dimethyl-2-chloronorbornane are about the same. If we assume that both 2-methylnorbornyl cation and 1,2-dimethylnorbornyl cation .are classical, the equilibrium constant for their exchange should be ~ 3.8. This would mean that in the equilibration experiment just after addition of acid we should have seen 8% methylnorbornyl cation and 25% 1,2-diemthylnorbornyl cation. Furthermore as 2-methyl-2-chloronorbornane isomerized to l-methyl-2chloronorbornane, the antimony pentafluoride associated with the 2-methylnorbornyl cation would be released, and the 1,2-dimethylnorbornyl cation would increase by onethird relative .,t o the chloride present. Neither of these expectations was confirmed by the experiment which did not ·reveal the presence of any methylnorbornyl cation and did not show a significant increase in the amount of 1,2dime thylnorbornyl cation with time. We feel that the inte gration procedure shown in Fig ure 43 is sufficiently reproducible to assure that the amount of 1,2-dimethylnorbornyl cation increased by much less than one-third during the isomerization . Therefore the addition of the 25°c. 2.la 4b 2-phenyl-2-norbornyl alcohol 2-norbornyl alcohol bref. 37 Figure 46 25°c. l.3a 2-methyl-2-norbornyl alcohol aref. 10 25°c. 2.6a 1,2-dimethyl-2-norbornyl alcohol 100°C. 65°c. 1.1 1,2-dimethyl-2-norbornyl alcohol 25°c. temp. 2 exo/endo 1,2-dimethyl -2-chloronorbornane compound isopropyl alcohol 60% aq. dioxane 60% aq. dioxane 60% aq. dioxane so 2 HCl-MeOH sol vent CP O'I 87 second methyl group seems to give the cation a degree of stability greater than would be anticipated on the basis of inductive stabilization. In Figure 16 there are three cations which show significant charge delocalization in the ~nr spectra. Cation stability measurements indicate that methyl bridging in 2,3,3-trimethylbutyl cation is not important. Similar measurements on 1,2-dimethylnorbornyl cation indicate that methylene bridging does stabilize the cation, but the extent of this extra stabilization is unknown. Our experi- ments allow no decision to be made on the role of methylene bridging in 2-methylnorbornyl cation. 88 EXPERIMENTAL 2,3,3-Trimethyl-2-butyl alcohol 2,3,3-Trimethyl-2-chlorobutane (Aldrich, sublimed) was dissolved in a mixture of acetone and water. A few drops of sulfuric acid were added as a catalyst, and the solution was heated at 80° C. for two hours. The solution was extracted with ether, and the product isolated. The product was a white solid, identified by its pmr and ir spectra as 2,3,3-trimethyl-2-butyl alcohol. Preparation of saturated antimony pentafluoride solution Antimony pentafluoride was placed in the reaction vessel (Figure 20). Sulfur dioxide was added to the cold vessel (-78°c.), and the vessel was heated until the solution was homogeneous. The solution was recooled to -78°c., and some of the antimony pentafluoride precipitated. This solution was allowed to stand under nitrogen for 30 minutes. Aliquots (0.5 ml) were withdrawn in syringes cooled in liquid nitrogen and fitted with teflon needles. The ali- quots were placed in glass tubes, and the tubes were sealed. 0 These tubes were stored in the freezer (-28 C.) until needed. 89 Equilibration in less than one equivalent of acid The appropriate chloride was weighed into a small glass round bottom flask (2 ml) fitted with a serum cap. The flask was cooled and sulfurylchlorofluoride condensed into the vessel until the solution was homogeneous. An attempt was made to keep the volume of solution ~0.6 ml. The 0.5 ml tube of antimony pentafluoride was placed in the reaction vessel, and the vessel was immersed in an acetone-dry ice bath. The cold solution of chloride was then added to the antimony pentafluoride solution which was vigorously stirred with nitrogen. The cation solution was withdrawn into an nmr tube which was either sealed with a torch or with a plastic cap and parafilm. Those tubes sealed with the plastic cap did not seem to pick up moisture over several days (i.e. the percentage of cation remained the same). Quenching experiments were performed by pouring the contents of the nmr tubes into ice water and extracting the solution with ether or petroleum ether. The organic layer was washed with sodium bicarbonate and dried. The organic solution was then analyzed on a 12 foot column of 10% carbowax 20M on chrom G using a Beckman GC-5 flame ionizaa tion gas chromatograph. In the case of 2-methylnorbornyl chloride and 1,2dimethylnorbornyl chloride the equilibrations in less than one equivalent of acid were not carried out with the 90 standardized tubes of antimony pentafluoride. The pro- cedure was the same except that an approximate amount of antimony pentafluoride was pipetted into the reaction vessel, and the sulfur dioxide solution was prepared. Equilibration of two cations The procedure was the same as above except that both chlorides were weighed into the 2 ml round bottom flask, dissolved in sulfurylchlorofluoride, and added simultaneousJy to the antimony pentafluoride solution. Where one chloride was a solid and the other a liquid, the solid was weighed into the flask, the flask was capped, the liquid introduced by microsyringe, and the flask was weighed again. Equilibration of 2-methylnorbornyl cation and 1,2-dimethylnorbornyl cation The experiments were carried out as before except that the tubes of standardized antimony pentafluoride were not used. Antimony pentafluoride ( ~ 20 p 1) was added to the reaction vessel, and the sulfur dioxide solution was prepared. The spectrum indicated no cation species, so another 20 pl of antimony pentafluoride was added and the solution withdrawn and sealed as before. This solution was immediately placed in the spectrometer, and the nmr spectrum was recorded. 91 References 1) S. Winstein, et al.; J. Amer. Chem. Soc., 7_4, ,... 1127, 2) s. Winstein and D. s. Trifan, J. Amer. Chem. Soc., (1952). 74, 1147, (1952). t,/l>I 3) S. Winstein and D. 74, 1154, (1952) . s. Tri fan, J. Amer. Chem. Soc., ~ 4) H. C. Brown, et al., J. Amer. Chem. Soc., 87, 2137, 5) J. D. Roberts and J. A. Yancey, J. Amer. Chem. Soc., """" ( 1965). 77, 5558, (1955). ~ - 6) H. C. Brown and C. J. Kim, J. Amer. Chem. Soc., 90, 7) P. vonR. Schleyer, et al., J. Amer. Chem. Soc., 85, 8) H. 9) H. 10) H. C. Brown and M. H. Rei, J. Amer. Chem. Soc., 88, 11) H. 12) H. C. Brown and Min-Hon Rei, J. Amer. Chem. Soc., 86, 2082, ( 1968). 479, ( 1963) . V\Ao c. Brown and F. J. Chloupek, J. Amer. Chem. Soc., 85, 2322, (1963). """"c. Brown and M. H. Rei, J. Amer. Chem. Soc., 86, u.... 5008, (1964). 5335, (1966). ""'" -- c. Brown and K. T. Liu, J. Amer. Chem. Soc., 89, 3900, (1967). 5004 J ( 1964) v~ 0 13) H. C. Brown and K. T. Liu, J. Amer. Chem. Soc., 89, 14) H. C. Brown and H. M. Bell, J. Amer. Chem. Soc., 86, 15) P. D. Barlett and S. D. Sargent, J. Amer. Chem. Soc., _... 466, (1967). .-. 5006, ( 1964) . 97, 1297, (1965). ~ 16) G. A. Olah, et al., J. Amer. Chem. Soc., 86, 1360 (1964). '-' 92 17) G. A. Olah, et al., J. Amer. Chem. Soc., 89, 2692, ( 1967). - 18) G. A. Olah and J. Lukas, J. Amer. Chem. Soc., 89, 4739, (1967). .,_ 19) G. A. Olah, et al., J. Amer. Chem. Soc., 86, 4195, ( 1964). ~ 20) G. A. Olah, et al., J. Amer. Chem. Soc., 86, 5680, (1964) . '-" 21) H. Toivonen, Anal. Acad. Sci. Fennicae, Ser. A, 72, (1956) . ~ 22) H. Toivonen, Suomen Kemistilehti, 33B, 66-9, (1960). 23) M. Karplus, J. Chem. Phys., 30, 11, (1959). 24) G. Tiers in "N.M.R. Data Tables for Organic Compounds," F. Bover, Ed., Interscience Pub., New York, N.Y., 1967, p. 228. 25) J. A. Pople, W. G. Schneider, and H. J. Bernsteen, 26) A. D. Buckingham, Can~ J. Chem., 38, 301, (1960). 27) M. P. Schweizer, et al., J. Amer. Chem. Soc., 86, 696, ( 1964) . ""'"" 28) H. Meerwein and M. Muhlendyk, Ann. Chem. 405, 171, (1914). --. 29) H. H. Freedman, A. E. Young, and V. R. Sandel, J. Amer. Chem. Soc.,~, 4722, (1964) . 30) A. E. Young, V. R. Sandel, and H. H. Freedman, J. Amer. Chem. Soc., 88, 4532, (1966) . ~ ....- "High-resolution Nuclear Magnetic Resonance," McGraw-Hill Book Company, New York, N.Y., 1959, pp. 218-223. ~ ""'- 31) a) N. C. Deno, et al., J. Org. Chem., 19, 155, ( 1954 ). b) N. C. Deno, et al., J. Amer. Chem. Soc., 77, 3 04 4 , ( 19 5 5 ) • """'" c) N. C. Deno, et al., J. Amer. Chem. Soc., 77, 3051, ( 1955) . ~ ~ 93 d) N. C. Deno, et al., J. Amer. Chem. Soc., 79, 5804, (1957). 32) ""'"" a) E. D. Jenson and R. W. Taft, J. Amer. Chem. Soc., b) R. W. Taft and T. D. McKeever, J. Amer. Chem. Soc., 86, 116, (1964). ""'"" 87, 2489, (1965). 33) E. M. Arnett and J. W. Loesen, J. Amer. Chem. Soc., 34) M. Saunders and E. L. Hagen, J. Amer. Chem. Soc., 90, 35) H. C. Brown and R. S. Fletcher, J . Amer. Chem. Soc., 36) Ernest L. Eliel, 11 Stereochemistry of Carbon Compounds," McGraw-Hill Book Company, New York, N.Y., 1962, pp. 90, 791, (1968). """' 2436, (1968). - Jl, 1845, (1949). 236-237. 37) F. Wilcox, Jr., M. Sexton, and M. F. Wilcox, J. Org. Chem., 28, 1079, (1963). -- """"""' 94 PROPOSITION I Bridged halonium ions have been proposed for each of the halogens except fluorine. Anchimeric assistance by bromine and iodine has been established by solvolysis of 2-substituted-cyclohexyl brosylates. The trans:cis rate ratio for solvolysis of 2-bromo-cyclohexyl brosylate is 800 while the trans:cis rate ratio is only 4 for the chlorosubstituted compound (1). The participation by iodine is even more striking and is estimated to provide 8.5 kcal/ mole stabilization by bridging (2). There is little indica- tion of anchimeric assistance by the chlorine substituent (1). Further evidence for the bromonium ion comes from the observation that threo-3-bromo-2-butanol gives d,1-2,3dibromobutane on bromination while the erythro isomer gives meso-dibromobutane (3). There is also evidence that the reaction of 3-chloro-2-methylpropene with hypochlorous acid proceeds % 38% through a symmetrical chloronium ion ( 4) • More recently G. A. Olah has examined the halosubstituted stable carbonium ions generated from 2-methyl2-fluoro-3-methyl-3-halo butane in antimony pentafluoride. He argues that all but the fluoro-substituted cations are nonclassical halonium ions. This conclusion is drawn from the chemical shift difference on ionization shown below: 95 @ - c a CH 3 x 6 on ionization ppm. fluoro chloro bromo iodo -1. 76 -1.15 -1.08 -1.05 This is the same order expected on the basis of electronegativi ty, but the authors feel the large chemical shift of the protons in the fluoro compound is indicative of much smaller amount of positive charge on halogen (5). It is proposed that the direct equilibration of these halogen-substituted cations be carried out to determine the free energy difference between them. The observed energy difference would indicate the amount of cation stabilization from halogen bridging superimposed on the normal substituent effect. If all the ions except the fluoro-substituted ion are stabilized by halogen bridging, there should be a much larger decrease in free energy on going from chlorine to fluorine than iodine to bromine or bromine to chlorine. It would also be interesting to prepare halogensubsti tuted ions\I). This would permit estimation of a normal substituent effect in the absence of halogen bridging. These two experiments would allow quantitative esti- mation of the amount of cation stabilization due to 96 halogen bridging. CH 3 X I I CH - C - CH 2 3 @ I References 1) S. Winstein, et al., ( 1951). 2) S. Winstein, et al., J. Amer. Chem. Soc., 70, 821, 3) S. Winstein and H. J. Lucas, J. Amer. Chem. Soc., 61, 1576, 2845, (1939). ~ 4) P. B. D. dela Mare, J. Chem. Soc., 1481,(1957). 5) G. A. Olah and J. M. Bollinger, J. Amer. Chem. Soc., 89, 4 744, (1967). (194~. ~ J. Amer. Chem. Soc., 73, 5458, ~ ~ 97 PROPOSITION II It would be interesting to have some measure of the degree of ionization in solvolysis experiments relative to that of a 11 stable 11 cation in antimony pentafluoride . This might be evaluated by the degree to which a substituent is stabilizing or destabiliz.ing in solvolyses and in equilibration of two cations. However, our experiments seem to indicate that the qualitative direction of substituent effects is not the same in the two experiments. It is possible that this is a consequence of different degrees of ionization in the solvolysis of two substituted compounds. It is proposed that the deuterium isotope effect on solvolysis and cation equilibration of the ~-butyl system be studied. It is hoped that the degree of ionization will be the same in the transition states for solvolysis of ~~butyl chloride and its deuterated analogs. The isotope effects measured in solvolysis experiments on l-butyl chloride are shown below (1): compound (CH3 ) 2 CC1CH rel. rate. 3 1.00 (CH ) 2 CC1CD 3 3 (CH3)CCl(CD3) 2 1.33 (cD 3 ) CCl 3 2.32 1.71 If the equilibrium constants are of the same order as the relative rates, these differences would be easily measured 98 by cation equilibration. Hyperconjugation (i.e. delocalizations of C-H electron into the vacant £orbital) is the generally accepted mechanism used to explain f!. deuterium isotope effects. The most convincing evidence that these isotope effects are hyperconjugative in nature comes from their steric dependence. The data for solvolysis of IIa, IIb, and Ile indicate that a C-D bond directed perpendicular to an incipient£ orbital does not give a ~ deuterium isotope effect. This is in agreement with the theory of hyper- conjugative origin of these effects (2). H IIa Ilb D Ile kH/kD Ilb IIc 1.14 0. 986 The isotope effect arises from stretching the C-H bond by overlap with the £orbital . The magnitude of this 99 bond weakening and therefore the isotope effect should depend on the amount of positive charge and the extent of overlap (3). We feel therefore that the magnitude of the deuterium substituent effect on cation stability should be a good indication of the amount of positive charge in stable cations relative to solvolysis transition states. References 1) v. J. Shirer, et al., J. Amer. Chem. Soc., 85, 2413, 2) """"' V. J. Shirer, et al., J. Amer. Chem. Soc., 85, 2416, '-(1963). 3) A. Streitweiser, et al., J. Amer. Chem. Soc., 80, 0.963). 2326' (19581 "\A. 100 PROPOSITION III G. A. Olah has shown that 1 3c-H coupling constants for carbonium ions are related to the hybridization at a pa rticular carbon atom . bonium ions The coupling constant for two car- is ~ 126 and 128 cps, while for the carbons J 1 3c-H=.168 and 164 cps. pre cursor hydro- These coupling constants g ive calculated s character values of 25.2 and 25.6% and 33 .6 and 33 .8% respectively in good agreement with the predicted sp 2 and sp3 hybridization (1). Recently it has become possible to measure l3c chemical shifts and 1 3c-H coupling constants from natural abundance l3c nmr spectra of neat solutions (1) . The resonance assigned to the C-6 protons of dimethylnorbornyl cation is 80 cps downfield of that for the C-5 proton. It is possible that this is a result of charge delocalization to C-6 by a 1,6 hydride shift or by resonan::e de localization. It is proposed that 1 3c enriched ethylene be used to prepare 1,2-dimethyl-2-norbornyl alcohol, and that the 101 13 c chemical shifts and 13c-H coupling constants at C-5 and C-6 be compared in the alcohol and cation. If our original proton assignments were correct, we would expect the C-6 resonance to occur downfield of C-5 after ionization due to positive charge at the C-6 position. Since 1 3c chemical shifts are much larger than proton shifts, the downfield shift for 1 3c should be more striking than for protons. It also should be possible to tell if there is any appreciable change in hybridization of C-6 (relative to C-5) on ionization of the alcohol from the l3c-H coupling constant. This would give an indication of any appreciable change toward sp 2 hybridization at C-6. If on ionization the re is a change to sp 2 hybridization, it would be interestr ing to do a careful product study on the cation quench solution, looking for the following compounds: o~ This would indicate whether delocalization by 1,6 hydride shift or resonance delocalization were more important. 102 By proton decoupling of the 13c nmr spectrum it also would be possible to confirm the proton chemical shifts for the cation. References 1) G. A. Olah and M. B. Comisarow, J. Amer. Chem. Soc., 88, 1818, (1966). ~ 2) F. J. Weigert and J. D. Roberts, J. Amer. Chem. Soc., 89, 2967, (1967). ~ 103 PROPOSITION IV Carbonic anhydrase is a metal containing enzyme which catalyzes the hydration of carbon dioxide. It has been shown recently by differential infrared analysis that carbon dioxide bound to the enzyme is not coordinated to the zinc (II). Ionic inhibitors of the enzyme reaction such as N - and Hco - seem to be bound to the zinc (II), 3 3 and both displace carbon dioxide from the binding cavity of the enzyme (1). This enzyme also catalyzes the hydra- tion of several aliphatic aldehydes and is inhibited by aliphatic alcohols, SNc-, N -, and acetazolamide (2). The 3 zinc can be replaced by cobalt (II) and enzyme activity is still preserved (1,2). Since the zinc is not involved in substrate binding, its proposed role is the coordination of the water molecule which attacks the substrate (2). It is proposed that study of the cobalt (II) enzyme be carried out to determine if the inhibitors of this enzyme displace water in the first coordination sphere of the cobalt, and if these inhibitors are bound directly to the cobalt. This will not unequivically establish the role of the metal ion in the reaction, but it will indicate whether it is likely that its role is coordination of the water used in hydration of carbon dioxide or aldehydes. The alcohol inhibitors are subject to investigation of broadening of the nmr proton resonances due to coordina- 104 tion to paramagnetic cobalt (II) ion. In the fast exchange limit for enzyme substrate interaction in the nmr, it is possible to calculate an upper limit for separation of the inhibitor protons from the manganese (II) in manganesecarboxypeptidase. From the observed broadening in the nmr spectrum of methoxyacetic acid, it was determined that this molecule is directly coordinated to the metal, i.e. in the first coordination shell (3). By applying this procedure to the alcohol inhibitors of carbonic anhydrase, it may be possible to show whether these are coordinated to cobalt. Furthermore, investigation of the water protons by nmr in the presence of the ionic and organic inhibitors may reveal whether these are displacing water from cobalt. In manganese (II)-CPD, the water signal is very broad when a mixture of Mn•(H2 o) 6 and Mn-CPD, is examined. This is due to binding of water to the paramagnetic manganese ion. When inhibitor is added, the signal sharpens because water is displaced from the Mn-CPD. From the temperature and concentration independence of T1 and T2 and the magnitude of broadening, the authors were able to say that a water mo·lecule is displaced from the primary coordination shell of manganese in Mn-CPD (4). A similar experiment is proposed for inhibitors of cobalt carbonic anhydrase. These experiments would give further indication of the 105 role of the metal ion in the binding of water and inhibitors to carbonic anhydrase. References 1) M. R. Riepe and J. H. Wong, J. Amer. Chem. Soc., 89, 2) Y. Pocker and D. G. Dickerson, Biochemistry, 7, 3) G. Novan, et al., PNAS, 60, 86, (1968). 4) """' 4229, (1967). ~ 1995, (1968). - - """"'" R. G. Shulman, et al., PNAS, 56, 39, (1966). -~ 106 PROPOSITION V An interesting feature of substrate binding to acetylcholine est.erase is the requirement for a positively charged group on the substrate. Such positively charged substituents are thought to induce conformational changes upon binding to ACE. In the phosphorylation of ACE by phosphates with charged and uncharged leaving groups, Wilson, et al. have concluded that the positively charged leaving groups may complex with the enzyme causing a conformation change in ACE (1). Catalysis of sulfonation of ACE by substituted ammonium ions also suggests a conformational change induced by these ammonium activators (2). Changeux found that at low ionic strength flaxedil, a "reversible 11 inhibitor of ACE, is only partially competitive with substrate. This suggests another binding site on ACE which might promote allosteric conformational changes (3). Derivatives of aminonaphthalenesulfonic acids do not fluoresce in water but do so when bound to hydrophobic sites on enzymes (4,5). It is proposed that such compounds be used as fluorescent indicators of substrate and inhibitor induced conformational changes in ACE . It would be of interest to bind charged ammonium compounds in order to determine if the binding of such compounds is competitive with fluor binding and if binding of these inhibitors or activators quenches fluorescence. 107 Such quenching is taken as an indication of conformational changes induced by the binding of inhibitors (5). Similar studies would be carried out using flaxedil under conditions of low ionic strength to test Changeux 1 s suggestion that such binding induces allosteric transition. Enzyme acylation should also be studied to see if acylation is competitive with fluor binding and if there is a conformational change occurring during acylation. Such a change has been postulated for o( -chymotrypsin (6). Finally the pH dependence of the fluorescence of the fluor-enzyme complex can be determined to probe the envircnment of the enzyme-fluor binding site. The development of a fluorescent substrate such as ?. fluor-C-0 choline might permit formation of fluorescent labelled acyl-enzyme. This would permit studies of the environment of the active site. References 1) R. J. Kitz, S. Ginsber ~, and I. B. Wilson, Mol. Pharmacol., 3, 225, tl957). ~ 2) R. J. Kitz and I. B. Wilson, J. Biol. Chem., 238, 745, (1963). "'-'"" 3) J.P . Changeux, Mol. Pharmacol., 2, 369, (1966). 4) W. O. McClure and G. M. Edelman, Biochemistry, 5, 1908, (1966). --. """' 108 5) W. O. McClure and G. M. Edelman, Biochemistry, 6, 559, ( 1967) . \-. 6) S. A. Bernhard, B. F. See, and F. H. Tashjian, J. Molec. Biol., 18, 405, (1966). """" .