Am SUBSTITUTED POLYCYCLIC omens DEPENDENCE or c 13 NUCLEAR MAGNETIC RESONANCE PARAMETERS .oN ELECTRON DEMAND - Dissertation for the Degree of Ph. D. MICHIGAN; STATE UNIVERSITY ROBERT ERNEST BOTTD 19.2.5 1511mm s'w ZWT DEPARTMENT a)? x :1 "'1 EAST LANSING‘ Mxthvwm 45:14: 8! NC, 1 N” U: 9 HM BUCK B!NDEP.Y I LIBQ El E , | gnmnflyymw i‘ 'Lraéaev Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c:/ClRC/DateDue.p65—p. 15 ABSTRACT Aryl Substituted Polycyclic Cations. Dependence of C-13 Nuclear Magnetic Resonance Parameters on Electron Demand BY Robert Ernest Botto The charge dependence of both B-carbon chemical shifts for a series of bicyclo [3.2.1] is very closely studied. Plots cations against pentyl) exhibit 2-aryl-2—bicyclo [2.2.2]- or 6-aryl-6- octyl cations, egg-g or géa-g, respectively, the same throughout the range of aryl groups of the carbocation chemical shifts of these those of classical models (3.3; arylcyclo— ideal linear behavior. By contrast similar plots of chemical shifts of C(l) 2&- C(3) for a series of Robert Ernest Botto 2-aryl-2-norbornyl cations EEETQ and 2-aryl-gng5,6-tri- methylene-Z-norbornyl cations géa-g show substantial de- viation from linearity for substituents (X) on the aryl group more electron demanding than mechloro. A plot of the a-carbon chemical shifts of either series versus those of our bicyclo [3.2.1] octyl model reveals a dramatic reversal in slope in the same region. The results are consistent with the onset of C(l)-C(6) bond participation in those norbornyl or 252:5,6-trimethylenenorbornyl cations with substituents more electron demanding than mechlorophenyl. A plot of the olefinic cmr parameters of a series of 2-arylnorbornen-2-yl cations gga-g against Brown 0+ sub- stituent constants is consistent with the onset of w-par- ticipation in those cations more electron demanding than 2-phenylnorbornenyl cation. The impressive turnabout in the a-carbon chemical shifts implicates a rehybridization of the carbocation center on electron demand. Free-energy relationships of the C-13 chemical shifts for a series of 3-aryl-3-nortricylyl cations gga-g suggest that there is extensive charge delocalization into the cyclopropyl moiety without extensive rehybridization at the carbocation center. Aryl Substituted Polycyclic Cations. Dependence of C-13 Nuclear Magnetic Resonance Parameters on Electron Demand BY Robert Ernest Botto A Dissertation Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1975 .‘I' FQVMUW migespec} W (YT/E3 Cowmeajes _ best 4 jock 1'1 T; Bout SJQHVQ gov Sam tk“: a! *5 fi 30 u... set S oWfiEkTW 3 iii to Wake»: 1th 1V TABLE OF CONTENTS Chapter Page LIST OF TABLES . . . . . . . . . . . . . . . . . . vi PERSPECTIVE. . . . . . . . . . . . . . . . . . . . 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . 6 The Norbornyl Cation . . . . . . . . . . . . . 7 Cyclopr0pylcarbiny1 Cations. . . . . . . . . . 14 Allyl Carbocations . . . . . . . . . . . . . . 21 RESULTS. . . . . . . . . . . . . . . . . . . . . . 27 DISCUSSION . . . . . . . . . . . . . . . . . . . . 41 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . 73 Carbonium Ion Formation. . . . . . . . . . . . 74 Carbonium Ion Precursors . . . . . . . . . . . 74 IR and NMR Spectra . . . . . . . . . . . . . . 75 REFERENCES 0 O O O O I O O O O O O O O O O O O O O 83 LIST OF TABLES Table ‘ Page 1 CMR Average Chemical Shift (C1,C2) in Substituted Norbornyl Carbocations. . . 13 2 13C Chemical Shifts for 2-Aryl-2- Bicyclo [2.2.2] Octyl Cations (egg-g). . . 29 3 Extent of Rearrangement to 6-Aryl-6- Bicyclo [3.2.1] Octyl Cations %%€'%° . . . 31 4 Selected l3C Chemical Shifts for 6- Aryl-G-Bicyclo [3.2.1] Octyl Cations (ggg-g). . . . . . . . . . . . . . . . . . 32 5 13C Chemical Shifts for 2-Aryl-2- Bicyclo [2.2.1] Heptyl Cations (gég-g) . . 34 6 Selected l3C Chemical Shifts for 2-Ary1- exo-S,6-Trimethylene-2-Norbornyl Cations (aée—g) O O O O O O O O O O O O O O O O O O 35 7 13C Chemical Shifts for 2-Arylnor- bornen-Z-yl Cations (egg-g). . . . . . . . 37 8 13C Chemical Shifts for 3-Ary1-3- Nortricyclyl Cations (age-g) . . . . . . . 39 vi Table Page 9 Typical Aryl Carbon Chemical Shifts in Aryl-Substituted Polycyclic cations O O O O O O O O O O O O O O O O O 40 10 Preparation of 2-Aryl-exo-5,6-Tri- methylene-endo-2-Norbornanols. . . . . . 76 11 Preparation of 2-Ary1norbornen-2-ols . . 77 12 Preparation of 3-Aryl-3-Tricyclo- [2.2.1.02’6] Heptanols . . . . . . . . . 78 13 Preparation of 6-Ary1-6-Bicyclo- [3.2.1] Octanols . . . . . . . . . . . . 79 vii PERSPECTIVE The existence of carbocationic species has been pro- 1,2 posed since the turn of the century when Norris and Kehrman3 independently observed that colorless derivatives of triphenylmethane give striking yellow solutions in con- centrated sulfuric acid and form orange complexes with aluminum and tin halides. The intense colors of these extraordinary compounds in solution and their sensitivity to hydrolysis were the only properties remarked upon in these early papers. Less than one year after these per- plexing observations von Baeyer recognized4 the salt-like character of the compounds formed in the solutions of tri- phenylcarbinol and sulfuric acid. He supposed a correla- tion between the formation of salt and the appearance of color-termed "halochromy". Such salts were named carbonium salts. In 1902 Gomberg pointed out the inconsistency of 5 "In my opinion, the von Baeyer's nomenclature. He wrote: name 'carbonium' should be applied to salts in which an increase of the number of valences of the carbon takes place, just as in the case of the ammonium, sulfonium, iodonium, and oxonium salts." Gomberg coined the term "carbyl salt" which never quite caught on. Later Dilthey 6 for and Wizinger suggested the name of carbenium salts such species in accord with their theory of coordinately unsaturated chromophores. Interestingly this controversy still remains with us. With the discovery of pentavalent cations of carbon in highly protic media, Olah7 suggested replacement of the term "carbonium ion" with "carbenium ion" for the trivalent cationic species. Farnuma, arguing that the "onium" suffix has not been exclusively reserved for the highest valence state, feels such change inappropriate. Both authors sur- prisingly agree on the perfectly adequate term "carbocation", which (this author feels) can be justified by its analogy with "carbanion" and for the sake of simplicity, could easily replace both incongruous nomenclatures for positive charged species of carbon currently in use. The existence of carbocationic salts was finally con- 9 firmed in 1909 by independent discoveries of Hofmann and Gomberg10 when both men isolated crystalline, colored anhydrous perchlorates from triarylcarbinols. The identity of these colored species in solution was later confirmed by their very characteristic absorption spectrum by Hantzsch11 Conductimetric measurements12 as well as molal freezing point depression studies13 further substantiated the ionic nature of these species - an electron deficient, trivalent carbon species had become reality. During the period that followed, 1920-1940, the nature of the chemical bond had become better understood and electronic theory matured from a sapling to bear many con- ceptual fruits. It was an age of exploration and an age of refinement - physical-organic chemistry was born. One of the most daring as well as important concepts visualized during this period concerned the intermediacy of carbocation species in the course of reactions involving non-ionic reactants and leading to non-ionic products. Most of the research provided evidence in favor or against the inter- vention of transient cationic intermediates in specific chemical reactions. Carbocations were considered intermediates in skeletal rearrangements. Wagner's discovered rearrangement of camphene hydrochloride in 1899 was rediscovered by Meer- wein in 192214 . His kinetic study of these reactions led him to conclude that: ". . . the rearrangement takes place only after a preceding ionization" and involves the rearrange- ment of the cation. Meerwein's rearrangement theory was later generalized by Whitmore in 1948 in a then comprehensive review on carbocationsls. The kinetic approach proved to be invaluable in the years to come. Substituent effects were noted and reaction types were classified by OliverlG. Hughes and Ingold17 undertook an extensive research effort during 1933-5 which placed reaction types in one of two categories: unimolecular or bimolecular. Their theory on the duality of reaction mechanism in solvolytic reactions dissipated the quandry concerning Walden inversion.18 The increasing stability of carbocations with branching was realized as well as the importance of solvent interactions during the course of a solvolytic reaction. Although many of the experimental techniques remained the same after 1940, a higher level of sophistication was attained partially by the development of new instru- mental and analytical methods. The advancement of theoretical chemistry had great impact on the understanding of reaction 19 allowed the con- processes. The transition state theory struction of a graphical representation of reacting systems - a reaction coordinate. From the fields of thermodynamics and statistical mechanics emerged extrathermodynamic rela- tionships which permitted investigation of the actual mechanisms of substituent or medium effects and which pro- vided a view of the ordering effects present in the transi- tion statezo. Increasingly accurate representations of the structures of the fundamental states of the reactants also allowed a better understanding of their transition states. Stable carbocations have been prepared in strongly acidic media and their properties determined spectroscopically. Indeed, after 1940 investigators aspired to far more am- bitious experimentation than ever before. Perhaps a climax was reached when it was observed that some unimolecular processes proceed with participation by neighboring carbon atoms. Winstein21 suggested the term "anchimeric assistance" while "synartetic acceleration" was proposed by Ingold22 for this general phenomenon. Electronic delocalization of a saturated carbon-carbon bond was first suggested in a paper by Nevell, deSalas, 23 and Wilson in 1939. Supporting evidence did not appear in the literature until some ten years later. Winstein 24 revealed their studies of the chemical behavior and Trifan of the norbornyl cation and postulated a symmetrically- bridged, delocalized cationic intermediate. Bridged cationic intermediates were considered in the solvolyses of un- saturated systems and small ring compounds (i.e. three and four membered rings) as well. The dialectic which followed concerning nonclassical behavior stimulated an enormous quantity of chemical research spanning the last twenty- five years. During this period a number of thorough reviews on the subjects of g- and lfparticipation representative of both sides of the controversy have appeared in the litera— turezs. Thus dawned the era of nonclassical cations only two decades after the concept that classical cations are dis- crete reaction intermediates was placed on a firm experi- mental foundation. Perhaps an age of enlightenment, it brought us closer to physical reality than ever before. The ideological confrontations found in the voluminous litera- ture on the subject serve to illustrate that we ourselves limit our interpretations, yet that all interpretations serve us. Max Planck writes: "The ideal aim before the mind (of the scientist) is to understand the external world of reality. But the means . . . to attain this end are what are known in physical science as measurements, and these give no direct information about external phenomena. As such they contain no explicit information and have to be interpreted." INTRODUCTION Within the past twenty-five years an extensive research effort has been undertaken to evaluate the viability of the concept of bridged, nonclassical ions. Until the early sixties the nonclassical formulation was based upon evidence derived from kinetic, stereochemical and isotope tracer studies. Yet all the experimental observations supporting the intervention of bridged ions could be explained in terms of rapidly equilibrating, classical ions. Subsequent methods to generate stable, long-lived carbocations in low nucleo- philicity solvents have since allowed direct spectroscopic measurements to be made. These spectroscopic studies provide the chemist with an invaluable tool for the structure elu- cidation of carbocationic species. If one assumes that cations formed in poorly solvating media are structurally equivalent with those cationic intermediates produced along the reaction coordinate during solvolysis, then the conclu- sions drawn from the study of cations in super acid are relevant to the nonclassical ion problem. A nonclassical cation has been defined by Sargent as "a positively charged organic moiety, representing a free energy minimum with respect to internuclear distortions, which is capable of delocalizing positive charge density by means of a multicentre molecular orbital formed, at least in part, by sigma overlap of atomic orbitals." The charge dispersal seemingly apparent in the 2-norbornyl, cyclopropyl- carbinyl and unsymmetrical homoallylic cations has been ascribed to this phenomenon: either inferred from the solvolysis data of the parent systems or concluded from the spectral parameters obtained for stable cations in highly acidic media. We shall discuss each purported nonclassical system in- dividually; and, in view of the voluminous literature already available on this subject we shall select those experiments and hypotheses which seem relevant to our results. The Norbornyl Cation In 1949 Winstein and Trifan26 proposed the symmetrically bridged cation 1 (suggesting the contribution from three canonical forms (1a, lb, lg) to account for their observa- tions found in the solvolysis reactions of 2527 and 2292-2- norbornyl p-bromobenzenesulphonates. Since that time the Eh, «45+» [5 «13 «UR k8 proposal that nonclassical ions are involved in the solvolysis of 2-norbornyl derivatives has rested on three major founda- tions: (1) unusually fast rates of solvolysis for the egg derivatives, (2) high egg/gndg rate ratios, (3) and stereo- selective formation of solvolysis products. Whereas proponents of nonclassical participation explain the rate ratio of the 2-norbornyl epimers in terms of stereoelectronic factors, Brown argues that the 2292f norbornyl transition state is destabilized on the basis of steric grounds27. Schleyer suggests28 that tortional and nonbonded interactions in the 2292 derivatives leading to the transition state may well account for the observed gxg/ggdg rate ratio. Brown also argues that one need not invoke non- classical participation to account for the high stereoselec- tivity of the solvolysis products. He has demonstrated the strong preference for egg orientation in reactions of U- shaped systems, including norbornyl, not involving cationic intermediateszg. Other studies which lend support to Brown's views are: (l) the enhanced rates of solvolysis of highly branched tertiary derivatives with increasing steric strain3o. As model compounds, they strongly support the contention that relief of steric strain is the major contributing factor in the solvolysis of tertiary norbornyl derivatives, (2) the failure of appropriate substitution to enhance solvolysis rates of norbornyl derivatives31, (3) considerable retention of optical activity in the deamination of optically active norbornyl amine832, (4) the incomplete deuterium scrambling observed in the ionic addition of DCI to norbornene33. One of the difficulties in the interpretation of data in the area of norbornyl cation chemistry has been the unavailability of suitable model compounds for direct com- parison. However, the 2-tricyclo [4.3.0.03’7] nonyl deriva- tive % studied by Nickon and Swartz34 is unique in that the substituent at C-2 occupies both an egg and an 2292 bonding relationship to a norbornyl ring system. The im- portance of steric and torsional effects in the presence OBs of a favorable stereoelectronic environment for participation can be evaluated. Solvolysis of compound % is 225 times faster than ggdng-norbornyl brosylate, nearly two-thirds the rate of solvolysis of 25952-norbornyl brosylate. It is apparent that steric hindrance to ionization and tor- sional effects play a minor role. The solvolysis of g is the most convincing evidence to date asserting the intervention of o-participation in the solvolysis of secon- dary norbornyl derivatives. 10 35 of a rapidly equilibrating Schleyer's observation pair of classical 1,2 dianisylnorbornyl cations 3a and 3b support Brown's model. The interconversion of 3% and ab is rapid on a pmr time scale even at -70°C. The absence ‘—-A V—— O . 0CH3 OCH3 of bridging was rationalized in terms of unfavorable non- bonding interactions which may exist in the bridged species ll 95 when one or both the aryl groups are oriented to allow overlap with the electron deficient center. In addition, Winstein contends that the cationic center is substantially stabilized by the p—anisyl moiety and need not require ad- ditional stabilization.36 The 1,2-dimethoxy-2-norbornyl cation studied by Nickon37 (in which the nonbonded repulsions should be absent or at least far less severe than in cation 9) similarly exhibits a temperature dependent pmr spectrum consistent with rapidly equilibrating, classical structures 4a and 4b. The cmr spectrum of the 2-methoxy-2-norbornyl cation has been discussed by Olah38. That the positive charge rests pri- marily on the oxygen atom suggests Nickon's disubstituted cation to be a poor model for comparison with the norbornyl cation, in which charge is concentrated on the carbons directly. Olah has demonstrated that the 1,2 diphenyl-Z-norbornyl cation is c1assica1,undergoing rapid dengenate 1,2 Wagner- 12' Meerwein shift and with very little charge delocalization into the phenyl rings39. Out of plane n-p distortion of the phenyl substituents implicate unfavorable nonbonded interactions between them as is the case in the dianisyl derivative (vide supra). In contrast the 2-methyl-2-norbornyl cation 3 appears to be a partially o-delocalized iondo. A nearly 70 ppm C-13 chemical shift difference of the carbocation centers is observed when S is compared with classical models such as methylcyclopentyl cation é. Olah later employs average cmr chemical shifts to distinguish among rapidly equilibrat- ing, symmetrically delocalized and partially delocalized structures (Table l)38 of substituted norbornyl cations. Cmr and pmr studies of the parent norbornyl cation are best accommodated by a nonclassical model41. Its Raman spectrum at -78°C is consistent with nortricyclene-like rather than norbornane-like skeletal symmetry41. The photoelectron spectrum (ESCA) of the norbornyl cation further supports 42 this conclusion although the quantitative accuracy in- volved is recently viewed with suspicion43. 13 ON m0 Hmfihwfixm OH 0>fl¥MHOH MHMHSm HMOflEflSU % 0N+ om+ ha: m~+ Hm+ wouwamoono no madmauumm 05+ omNHHMOOHmo no maamowuuoEEhm HH- maflumunflaanvm savanna mm+ comaamoono no adamwunmm mcflumunAHHsvm Savanna coaumu HacnonuocamnumZIN cofiumo Hmcuonuozlm coaumu ammoumamuomIMnm coaumu Hmcuonuocaanumaflonm.a unflnm Hmoasmno omnum>4 omumoncm umwnm doomsmno :oflumowmwmmmao momum>¢ UmH fiOH mCOflHMOOQHMU t Hacuonuoz cmusuaumnsm an A~0.Huv unfigm Hmoflamgu omnum>¢ mzu .H «Hams II‘ 'llll“|lll|l|'l‘ I llll 14 44 find a linear Lastly, although Brown and Takeuchi dependence of log k for the solvolysis of a number of 2- aryl-Z-norbornyl derivatives with 0+ values, Farnum and Wolf45 observe nonlinear free energy behavior of 2-ary1-2- norbornyl cations in FSOBH. A plot of the pmr chemical shifts of H(l) versus H(3) shows nonlinearity for substit- uents on the aryl group more electron withdrawing than hydrogen and which is not consistent with a pair of rapidly equilibrating, classical ions. Cyclopropylcarbinyl Cations The ability of cyclopropane rings to conjugate with adjacent n-orbitals has been realized for some time. In 1917 Kohler and Conant46 concluded that a cyclopropyl moiety in the proper position can form a conjugated system similar in properties with those found for conjugated olefinic compounds. Their conclusions were based on chem- ical reactivity. Later, as evidence amassed from solvolysis studies a true appreciation of the stability of cyclopropyl- carbinyl cations was realized. Roberts and Mazur have demonstrated the unusually fast reactivity of cyclopropylcarbinyl halides in solvolysis reactions. In fact they are considerably more reactive than analogously constituted allyic halides47. Although the intense conjugation in cyclopropylcarbinyl cations is widely accepted, the mode of such interaction is 15 still the subject of serious debate. Roberts initially proposed an attractive "nonclassical" (coined by Roberts48) cationic intermediate of structure 148'49 to account for the abnormally large solvolytic r' - + /C H O-~ -..__, 2.“ "‘an \CH" 2 L. -L reactivities of cyclopropylcarbinyl, cyclobutyl and allyl- carbinyl derivatives and the striking ease of interconversion among them. Subsequent isotopic scrambling studies revealed that 14C-distribution in the products from cyclopropyl- carbinyl solvolysis is extensive but not completely randomso. (The symmetry of structure 2 is incompatible with the data; nor does it receive theoretical support51.) Roberts posed the intermediacy of o-participating, unsymmetrical bicyclo- butyl cations ga, Sb and gg, which equilibrate rapidly, to explain the results. 16 Indeed many structures for the cyc10propy1carbinyl cation have been considered (z-l§)47-59. + $3 17 The profusion of these dashed structures tends to compli- cate matters. Simply two modes of stabilization of strained cyclo- propyl bonds are worthy of consideration on the basis of experimental evidence and theoretical considerations: c-bond participation and hyperconjugative interaction. The first requires atomic movement so as to allow maximum o-p overlap of atomic orbitals (Figure 2) irrespective of which dashed representation (0, 2, lg, 19) is chosen. — £::> - +, .0 . 0""0 Lo 0 Figure 2 The hyperconjugative mode of interaction is available with- out changing the spatial arrangement of atoms. It can be represented by partial p-n overlap and partial p-c overlap of orbitals without distortion of the carbon framework (Figure 3, structures 10 and ll)53 or as a vertical process in the Frank-Condon sense (Figure 4)59. Hyperconjugative interaction requires one of the bisected structures indi- cated in Figure 4. 18 Figure 3 Figure 4 The bisected structure has been strongly indicated by the discovery that cyclopropanecarboxaldehyde6o 61 and cyclo- propyl methyl ketone exist in a gig and trans isomer, both of which have bisected conformations. The bisected conformation presumably permits maximum overlap of the p-orbital of the adjacent carbonyl carbon with the cyclo- propane "bent bonds". Hart and coworkers have shown that successive replace- ment of isopropyl by cyclopropyl groups lead to large solvolysis rate enhancements by nearly constant increments per cyclopropylmoiety’introducedsz. That similar modes of conjugation are present in the mono-, di- and tricyclo- propyl carbocation implicates the bisected structure. Methyl substituent effects on the rates of cyclopropyl- carbinyl solvolysis are evidence for symmetrical transition 63 states. The rigid conformation of the adamantane nucleus allows 19 quantitative evaluation of the stereoelectronic require- ments of cyclopropylcarbinyl cations. Adamantyl derivatives 1% which incorporate the cyclopropylcarbinyl system in the perpendicular conformation exhibit rates of solvolysis 64 retarded by inductive interaction . Solvolysis rates for those derivatives lg which incorporate a bisected cyclo- prOpyl interaction are predictably fast. Pittman and Olah66 have shown that the two methyl groups in the 2-cyclopropyl-2-propyl cation 19 are not equivalent. The result is consistent with the bisected structure 0 20 Although nmr examination of methyl- and dimethylcyclo- propyl carbocations has established their preferred bi- 66-68 sected geometry , Olah suggests a bridged structure for the parent system67’68 based on a discrepancy between observed and calculated (estimated from a classical model) 13C chemical shifts. The methylene carbon shift calculated for a set of rapidly equilibrating classical ions is 115 ppm (TMS) as compared to the observed value of 57.6 ppm (TMS). 13 Brown and Kelly have measured the C-‘H coupling con- stants for cyclopropyl, methylcycloprOpyl and dimethyl- cyclopropyl carbocations. The methine coupling constant for the parent compound g0 is in accord with the formulation 130 , 190 H £8 £1 %% 187 £3 £3 21 as an equilibrating set of ions with the bisected arrange- 13 1 ment. An increase in the C- 70 H coupling constant postu- lated to accompany an increase in angle strain (arising from.c-bridging) is not observed. Examination of the pmr and cmr parameters of 8,9-dehydro- 2-adamantyl cations 3&2 and 36b by Olah and coworkers70 1“ ll :11 3362‘ R m: R CH3 suggests the absence of nonclassical participation in these compounds. Allyl Carbocations In 1946 Shoppee71 had demonstrated that cholesteryl chloride undergoes nucleophilic substitution with complete retention of configuration while ch01estanyl chloride re- acted with inversion as expected. To explain these results 22 he invoked the participation of the adjacent double bond. 72 Later Winstein and Adams found that solvolysis of choles- teryl tosylate 36 in pure methanol leads to the formation of methyl ether 21 with retention of configuration while ‘\ W H.. \\ \\ '/' CH3O H ..... llllllll ___9’ llllllll £1 TsO ‘M . ;. 4°33” ‘\ Q 90 OCH3 2.2 the same reaction buffered with potassium acetate yields cyclopropylcarbinyl ether 30. The formation of the two products and the stereochemical outcome was explained in terms of nonclassical ion 32 as an intermediate in the re- action. Solvolysis of gxg or endng-norbornenyl halides or sulfonates and deamination S-norbornenyl-amines with nitrous acid yield nortricyclic derivatives73. Roberts postulates a hyperconjugative, homoallylic interaction to account for the products and/or the rates of solvolysis. 14 The relatively incomplete scrambling of C- labeled dehydronorbornyl derivatives during solvolysis leads Roberts 23 to pose the intermediacy of a pair of slowly equilibrating enantiomorphic cations 30 and 2174. However, Roberts explains the formolysis of allylcarbinyl tosylate (in 1964), which formylizes 3.7 times faster than n-butyl tosylate, in terms of the formation of "bicyclo- butonium ion“ intermediates75. The most dramatic example of homoallylic assistance is found in the acetolysis of anti-7-norbornenyl tosylate. This compound reacts faster than its saturated analog by a factor of 1011 and reacts with complete retention of configuration76’77. The occurence of a sharp break in the p-o+ plot for a series of 7-aryl-7-norbornenyl‘p-nitrobenzoates indicates a dramatic change in the mechanism of solvolysis78. A similar plot of aryl-substituted saturated analogues is linear over the same range of 0+ values. The results are consistent with the onset of neighboring group participation by the n-electrons of the norbornenyl double bond. Further- more it provides evidence that participation can be a linear function of electron demand. 24 A plot of the pmr parameters reported by Winstein79 versus 0+ constants (Figure 5) shows similar results for 7-aryl-7-norbornenyl cations with the exception that the break occurs at a different value of 0+. Thus solvolysis rates and nmr parameters of carbocations in super acid cor- relate qualitatively with Brown 0+ constants although not quantitatively. 3.4- '\ \ L \ \ \\ 3,0_ \ \ \\ p. \\ 112,3(1') 2 6- \ 0 ¥“ ‘”v..~.‘_ ‘~~‘.‘ ‘--, 2.2% 1 1 I n 1 1 1 1 1 l .4 0 -.4 -.8 -l.2 + c Figure 5. Graph of H(Z),(3) Chemical Shift in 7-Aryl- 7-Norbornenyl Cations Xi: 0+ Constants78. 25 The nature of participation in a homoallyl system is much like that apparent in cyclopropylcarbinyl. In fact ab initio molecular orbital calculations indicate that all forms of the homoallyl cation collapse without activation energy to a bisected cyclopropylcarbinyl cationeo. Much effort has been directed toward structure elucida- tion of various carbocations by means of nmr spectroscopy. Many of the structural conclusions drawn from these studies have rested on chemical shift comparisons to known models. The recent development of Fourier transform (FT) nuclear magnetic resonance spectroscopy has made cmr studies both practical and routine. Extension of this spectrosc0pic method to the area of stable carbocation chemistry has led to a better understanding of the structural geometry and electronic characteristics or organic ions. Spiesecke and Schneider81 were the first to show the validity of a charge density to cmr chemical shift relationship which is contingent upon related structures having the same geometry. Cmr spectros- 82 also reveal copic studies of classical benzylic cations that cmr parameters reflect the positive charge density. Hammett-type relationships of these cmr parameters (to Brown 0+ constants) roughly parallel those observed in solvolytic rate studies of the same systems. Thus, it seems reasonable that our extension of the extrathermodynamic relationship of cmr parameters of stable carbocations to the area of nonclassical cations may reveal 26 changes in geometry and charge distribution in these ions with increasing electronic demand. Moreover, comparing our data to those observed in solvolytic studies of the same systems could provide information concerning the nature of those transition states involving stabilization from distant w— and 0- electrons. We have prepared several series of aryl-substituted polycyclic cations to explore these pos- sibilities. RESULTS Experimental procedures for the preparation of the cations for this study are described in the experimental section. The carbon-13 nmr spectra were obtained by Fourier transform (FT) spectroscopy. Coupled spectra were obtained by the off-resonance decoupling technique. That 13 C parameters are invariable with temperature is essential to the interpretation of our data. In temperature studies involving equilibria among aryllbicyclooctyl cations (vide infra), we find that the cmr chemical shifts of cations 332. k and 333, b are constant over an 85 degree temperature range (-60° to +25°), and that those of cations gag-g and egg-e are constant over 60 degrees (-60° to 0°). A series of 2-aryl-2-bicyclo [2.2.2] octyl cations gag-g were generated from the precursor 2-aryl-2-bicyclo- [2.2.2] octanols in FSO3H at -78°. ééé‘é 27 28 The C-13 chemical shifts for these cations are reported at -30°. Cation egg (X = CF3) was prepared from the correspond- ing alcohol in FSO H-SbFS-SOZClF at -110° and its spectrum 3 determined at -90°. Viscosity line broadening and coinci- dental interference from the 13 CD3 resonance lines of the d6-acetone lock preclude accurate evaluation of the C6,8 chemical shift of 32g. The cmr data are summarized in Table 2. The five carbon-l3 resonance lines assigned to the bicyclo [2.2.2] octyl skeletal carbons not including C2 reflect the symmetry present in these cations. We have been unsuccessful in our attempts to prepare 2-(gigrB,S-trifluoromethyl) phenyl-Z—bicyclo [2.2.2] octyl cation 32g. Careful ionization of alcohol 34 in FSO3H- SbFS-SOZCIF at -120° provided a spectrum, recorded at -100°, which is consistent with skeletal rearrangement to the 6-aryl-6-bicyclo [3.2.1] octyl cation 33g. Wolf has CF LA) .susoflamsuasz in .mze Hmcumuxm on m>flumamu coHHHflE mom muumm Am 29 awe s.v~ m.mm m.Hm m.~m v.mm~ mmUVmsoum Am o.m~ m.q~ m.am H.ms «.mv ~.os~ “mammo-m Am m.m~ s.a~ m.Hm s.sa m.ms m.~s~ mmmu in m.m~ m.e~ m.om o.sv «.5v s.so~ mvmsoum Au ~.m~ m.v~ m.om H.mv H.mv s.mm~ «Ammuvmmmo-s.m in m.m~ h.v~ m.m~ o.Hq m.~v A.sm~ mmu0¢smu-m Am luvm.su luvb.mu lasso lucmu lasso almc+u m>sum>auma alwawwmv mconumo assoc fi~.~.~H oaososmumuasuauw sou mumsnm Hmossmsu u .N manna ma 30 previously reported that careful ionization of 3% or its olefin analog gives a mixture of cations at -100°, egg being the minor component82. Perhaps the length of time required to perform the cmr experiment coupled with poor resolution of the spectra at these low temperatures prevented our detecting cation 32g. The propensity for rearrangement of the bicyclo [2.2.2]- octyl skeleton allowed spectral determination of 6-aryl-6 bicyclo [3.2.1] octyl cations 338-3 to be made. The ease of this skeletal reorganization was heavily dependent on the electron withdrawing ability of the aryl substituent. The extent of rearrangement is summarized in Table 3. Warming cations 33% and 32b in the nmr probe to +25° failed to induce rearrangement. At temperatures above +25° decom- position precluded rearrangement. Consequently, cations 333 and age were prepared from their alcohol precursors in FSO3H at -78°. Their spectra were recorded at -30°. The 13 C nmr parameters of the cationic center, C5 and C7 for these ions are presented in Table 4. 13C chemical shifts for all the bicyclo- We cannot assign [3.2.1] octyl skeletal carbons with assurance since appropri- ate models are not available for comparison. Nonequivalence of the orthg and meta carbons of the aryl group for ions ég-EfOCHB and 33-di-CH3 suggests that rotation about the C-C+ bond is slow on the cmr time scale. This is true for the other p—OCH3 and di-CH3 phenyl, and some aryl substituted 31 Table 3. Extent of Rearrangement to 6-Aryl-6-Bicyclo [3.2.1] octyl Cations gag-g. Derivative Percent Conversion Temperature (°C) a) pr6H4OCH3 0% +25° b) 3,4-C6H3(CH3)2 0% +25° c) gecsn4r 50% 0° d) C6H5 60% 0° e) EfC6H4Br 70% 0° f) E-C6H4CF3 ~100% -61° g) 3,5-C6H3(CF3)2 ~100% -9o° 32 Table 4. Selected 13c Chemical Shifts for 6-Aryl-6-Bicyclo [3.2.1] octyl Cations (33a-g)a. Derivative C+ C C 5 7 a) p§C6H4OCH3 235.5 49.8 44.5 b) 3,4-C6H3(CH3)2 256.4 53.1 47.5 c) pr6H4F 26243 55.2 49.1 d) CGH5 268.6 55.7 49.6 e) pr6H4Br 265.4 55.9 49.6 f) EfC6H4CF3 279.5 59.1 52.5 g) 3,5-C6H3(CF3)2 282.2 60.8 53.6 a) Parts per million relative to external TMS. 33 polycyclic cations in this study as well. 2-Ary1norbornyl cations egg-g were prepared from their corresponding alcohols in FSO3H at -78°. Their spectra were taken at -30°. Pmr spectra for these ions egg-g have pre- viously been reported by Wolf and Farnum45. The cmr spectrum for 34g has been obtained by Lam83. The 13 C parameters for the skeletal norbornyl carbons are listed in Table 5. Assign- ments of chemical shifts were made by comparison with those found for 2-norboranone and with the aid of coupled spectra. A series of 8-aryl-8-tricyclo [5.2.1.02'6] decyl cations éék-g were generated from their alcohol precursors in FSO3H I388? at -78°. Cation age was prepared under identical conditions via ionization of its precursor olefin. It is not possible 13 to assign C chemical shifts to all the skeletal carbons with assurance (particularly C3, C4, C5 on the trimethylene 13C parameters bridge) in these cations. However, those which are pertinent to this study are unambiguously assigned (Table 6). Solutions of 2-arylnorbornen-2-yl cations age-e, gener- ated by dissolving "Freon-ll" solutions of the corresponding 34 .muflmum>wco oumum :mmanowz .mwmmne .m.z .qu moan Scum sumo An .mza Hmcumuxo on o>wumamu cowaawe mom uumm Am e.m~ m.~e m.mm m.~e ~.mm «.5» m.mem «Ammuvmmsu-m.m Am e.m~ H.~e H.5m m.He s.~m e.em m.mo~ mmoemmo-m Am m.m~ m.av ¢.vm H.o¢ m.om m.mm m.mm~ HmvmmUIm Am e.m~ e.ae e.em m.mm m.om e.mm m.om~ mmmo is m.m~ m.He m.mn m.em e.om m.mm m.mm~ semen-m .0 o.m~ e.He H.mm o.mm H.me m.mm m.am~ «immuvmmmu-e.m in -- -- -- -- ~.se m.mm mmm nmmUOesmo-m in Auvmo Auveu xuveo leveo luvmo lasso Amv+o m>eum>euma Maw-wwwv mcoaumo Haumon HH.N.~H cacaowmlmnamudlm How muwfinm HMOflEwco U .m manna MA 35 Table 6. Selected 13 C Chemical Shifts for 2-Aryl-exo- - - - - -:—'a 5,6 Trimethylene 2 Norbornyl Cations (323 g) . Derivative C+(8) C1(d) C3(t) C6(d) pr6H4OCH3 231.9 56.3 45.4 48.8 3,4-cen4(ca3)2 250.4 60.2 48.6 50.9 pr6H4F ' 255.0 62.4 49.9 52.0 0685 260.4 63.2 50.6 52.7 pr6H4Br 258.2 63.6 50.9 52.9 2:0684083 268.0 67.6 53.2 55.8 3,5-C6H3(CF3)2 268.4 69.5 54.1 57.2 a) Parts per million relative to external TMS. 36 alcohols in FSO3H at -78°, are stable at low temperatures. The spectra of cations 368, 36d and 36% were recorded at -30°, -60° and -80°,respectively. However, solutions prepared from alcohol precursors (QQ-EfOCH3) and (36-di-CH3) under identical conditions furnished spectra with a different pattern of absorptions. The spectra are devoid of any 13C olefinic carbon atoms. An absorption appearing at 6 88.0 in both spectra, and which appears as a doublet in the proton-coupled cmr spectra, is consistent with addition of FSO3H across the double bond. Perhaps conjugation of the olefinic moiety in cations 688-5 is sufficient to preclude addition of FSOBH, where it is not in 333'?“ Cations 36a and 368 were instead prepared at -78° employ- ing a limited amount of FSOBH diluted with SOZC1F. Their spectra are reported at -30°. The cmr data for cations egg-e are summarized in Table 7. Chemical shift assign- ments were made by comparison with those found for the parent ketone 3186 and with the aid of off-resonance FT spectra. 37 .mze Hmcuwuxw ou m>wumamu cowaafla Mom muumm Am mammovemmo-m.m H.me o.mHH e.mme H.6e m.mm m.ee H.mH~ m.me m.sHH e.oee o.ee e.em m.me 6.6mm lmeuvemso-m o.me m.e~a m.mse s.me m.mm ~.~e e.ee~ memo e.se m.e~H m.sma m.me H.mm e.mm s.~e~ «Ammuvmmeo-e.m e.me m.m~e m.omH m.~e ~.em m.sm m.e- mmUOemso-m luvmo lasso Russo levee “peso onflu lmv+o m>eum>euma a 888 - .mAm-Mmmv mcowumo Hmumnnmauonuocahudam How unmanm HmowEmnu UmH .b manna 38 3-Aryl-3-nortricyclyl cations egg-f were prepared in the usual manner in FSO3H at -78°. Cations 38a and 38b are reported at -30° while cations ggg-g are reported at -50°. The cmr data for these ions are compiled in Table 8. Relatively minor differences in like-aryl carbon chemical shifts indicate that the extent of charge delocalization into the aromatic systems are similar within the series of poly- cyclic cations investigated. Typical aryl carbon chemical shifts are given in Table 9. In the case of the Effluoro- 130-19 phenyl derivatives long range F coupling (4.3 Hz) to the cationic center is observed. 39 Table 8. 13C Chemical Shifts for 3-Aryl-3-Nortricyclyl Cations (egg-f)“. . . + Derivative C (s) C1’6(d) C2(d) C4(d) C5’7(t) p—C6H4OCH3 234.8 43.2 33.7 38.8 38.5 pr6H4CH3 250.7 54.3 41.8 40.0 40.0 C6H5 257.1 61.3 47.0 40.5 40.8 m§C6H4Cl 258.3 68.1 52.1 41.2 41.7 pr6H4CF3 260.0 73.5 56.2 41.3 42.3 3,5-C6H3(CF3)2 258.5 80.5 61.2 41.9 43.1 a) Parts per million relative to external TMS. 40 Table 9. Typical Aryl Carbon Chemical Shifts in Aryl- Substituted Polycyclic Cationsa. x Cpara Cortho Cmeta CB g-ocn3 181.9 143.7, 143.4 120.2, 118.8 131.8 2-083 167.6 140.4, 140.2 134.0, 133.2 132.4 pr 152.4 141.9, 141.1 132.6, 132.5 133.8 Ear 177.6 146.9, 146.2 121.6, 120.5 131.1 (288 Hz)b (14.9 Hz) (4.2 Hz) prr 152.9 141.8, 141.1 136.5 132.5 p—CFB 146.6 140.9 (br) 134.7 128.8 (br) a) Parts per million relative to external TMS. 13 b) C-F Coupling. DISCUSSION Graphs (Figures 6-11) correlating cmr chemical shifts are constructed from selected parameters found in Tables 2, 4-8. The following symbols will represent the various polycyclic cations in the figures below: BCO [2.2.2] 2-aryl-2-bicyclo [2.2.2] octyl cations. BCO [3.2.1] 6-aryl-6-bicyclo [3.2.1] octyl cations. N = 2-aryl-2-norbornyl cations = 2-aryl-2-bicyclo [2.2.1] heptyl cations. XTN = 2-aryl-5,6-gngtrimethylene-2-norbornyl cations. NE = 2-arylnorbornen-2-yl cations. NTC = 3-ary1-3-nortricyclyl cations = 3-aryl-3-tricyclo [2.2.1.02'6] heptyl cations. In their extensive pmr study revealing the nonlinear free energy behavior of 2—aryl-2-norbornyl cations, Wolf and Farnum45 chose as their classical model the bicyclo- [2.2.2] octyl system. Indeed this system was known to have less tendency to exhibit those properties usually associated with o-participation. Their study, however, was limited by rearrangement of the more electron demanding aryl-sub- stituted cations (prF3, _,mf-(CF3)2) in superacid media gig successive Wagner-Meerwein and 7,2 hydride shifts. Our cmr study of 2-aryl-2-bicyclo [2.2.2] octyl cations is limited by this same factor. The equilibrium among 41 $C|(ffwfi sififfifisasxfi Figure 6. 42 ‘I‘ "5 SC; ("m3 Graph of C(l) vs. C(3) Chemical Shifts in 2- Aryl—2-Bicyclof2.2.2]octy1 Cations 48876- {Mil II [I l l I I l {fimql‘dfllflfl Ell Hill II" n mam-mum- :— lfllfll H mum-3391mm: R I 4.8- Email-1311mm i 1*, lfliigigfi! 3%... IJ . lair 554:1 __- .. as: Effig== '23:” E QEE 553%? $33: - fl : fiwfi§ n .... fiasggg 3% fish 12% “1 mm: flUiflliHl 3.5, ms fllflfiiflhmh *8 mail! '7“- NEWS. . mama-81141;: . mama-11mm man WWW“ IH III 139% . ,lflfiflfllfliflfllflflm ' B’sflfififlfijflm .4... E '-<- Efi Eififlfllii‘amfii-‘fliihfi-‘Hfl-HH 91.72? I i" «u u s c. mm Figure 7. Graph of C(S) vs. C(7) Chemical Shifts in 6-Aryl— 6-Bicyclo[3.2.IToctyl Cations gag-g. 44 I .muuasm anode—o u :o a onuao Haucunoaohuabd can Havoc—”H.N.nHoHoaon-o-Heflzuw mo :ouuuauwwow 6...: no .3 r3 .6 656: 8N 0““) [I251 039 +33 6| '0" ‘9' Figure 9 . 45 ‘ c; ( Pf“) Graph of C(l) vs. C(3) Chemical Shiftl in 2-Ary1- 2—Norborny1 ‘CaHons m—g. SC- mm) Md) (hm-Jon .33 sc° acouzu mm) -Aryl-6-Bicyclo[3.2.1]octyl Correlation of 2-Aryl-2-Norbornyl and 6 Carbocation Chemical Shifts. Figure 10. 47 (1061.11.- l'll- 7|: Sc; ( PP“) ions Graph of C(l) vs. C(3) Chemical Shifts in 2- Aryl-exo-S,6-TEImethylene-2-Norbornyl Cat [248%- Figure 11. «“1 um: .03. SC’ stoma-J (pr-53 Correlation of 2-Aryl-exo-5,6-Trimethylene-2-Norbornyl and 6-Ary1- Figure 12. 6-Bicyclo[3.2.l]octyl Carbocation Chemical Shifts. 49 (uddy’as ggEEEEQEV' EEEEEEEEEEEE . EEEEEE Ea EEEEEEEEE gtasgggg a EEEE‘E giaaagrg EEEEEEE EfiEfigng-_ EEEEEEE EEEE fig": EEEEEEE EEEE§§%E*; EEEEEEE- Eamw Eafig '53EEEEE EEEEEE EEE’ Eaaaggg EEEEEEEEF EEEEE a - - EEEEEE§_ : J" EEEEEEE' EEEEEEee EEEEEEE EE EEEEET“ EE. .EEEEE, EE .EEEEE=- EE EEEEE. EE EELEI EE -lm& ; EE- 1% 115mmmmmmmmmsmmmi- ufiEW$fl%wmmmmmmwmmmmwm E3 §E_:m' EE EEEE. 'E -T§;§§* .“F‘ EEEEE-=; , aw' EEEEEEE & x aEEgEEi“.“ E. EEEEEEEEEEEEo $52-23 ..::... ='- -EE%$:;-§§== : 'u .1 " :1. . “agar “"a. .... xi “-‘EEEEEEEPE §E7¢ié§8 . 52 EEqamfiw 231.42.: ' 2 EEE m _EEEEEEEFEEEEEEEEFEEEEEEEEEE“_ -§§E fififififlfififimfigmfiEEEEEEEEEEEE EMEEEEWEE_EEIEEE EEEEEE EEEEEE -EEEEEEEEWQEEEEEEQEQEEEEEVE a 2 9 9' Mmm -.: . E:- ' $5 15 fig -$E§:§§g- I EEEEES “:3 8tfifigggg Egg gEEEwE :g=. 'EEEE"EEE ‘E_ 'EEE§ Effi _§ .EEEEEEEE E EEEEEEEE 'E .EEEEEELW= L EE §~ IE 3% 855 EE a?” Graph of C(S) vs. C(6) Chemieal Shifts in 2-Ary1norbornen-2-yl Cations 363-3. Figure 13. 50 {‘4‘ ’ SN .33 =fi gmgafi EEEEEEE EEEEEEE EEEE g fl MW MIEWEMWE H Eaggmdrggggg g E_‘ .EEEEE ESEEEEEEEEEEE 5 EEE== EEEEEEEEEEEE. EEEEEEEEE§E_ EEEEEEEEEEE“ EE E‘EgEE §§E_E- : fig; g;- ' E‘ Wumm‘i Wu WERE!!! %%WMWMWWW%WWIM mwmmwmm “@lp "mgmmimmmgmggfiignan, Q E‘Hflfllfl mmmmmMMwm EEEWMMMMM_ EahggfiwfiaéE EEEEEEEEEWEEIIEEEEEEE EEEEEEEE 3 mmmng SC’IflXfiJJJ¢fiFn 2—y1 and 6-Ary1-6-Bicyclo[3.2.1] Correlation of 2-Ary1norbornene Carbocation Chemical Shifts. Figure 14. .WIme meowumw Hmummnn .No.H.~.NH O UhUHHBIMIHhHfiIm Ga mUMflfim HMOflEGfiU AVVU .m> Amy .AHWU mo nmwuo .ma musmflm 9“”) DLN 93 3 52 “‘ ”[32.” (m3 icyc10[2.2.1.02'6] heptyl and 6-Ary1- 6-Bicyclo[3.2.1]octyl Carbocation Chemical Shifts. Correlation of 3-Aryl-3-Tr Figure 16. 53 arylbicyclooctyl cations lies strongly in favor of the more stable 6-ary1-6-bicyclo [3.2.1] octyl cations. Rearrange- ment precludes our observing the m,m'-(CF3)2-phenyl substi- tuted cation even at extremely low temperature. Even more disheartening, we find the cmr spectrum of cation 3&3 (pf CF3) poorly resolved. Although a plot of C(l) versus C(3) chemical shifts is very closely linear throughout the limited range of aryl groups studied (Figure 6), our confi- dence in these data leaves much to be desired. A study of 6-aryl-6-bicyclo [3.2.1] octyl cations gag-g has distinct advantages: (1) They are by far the most stable among the equilibrating arylbicyclooctyl cations in 83 and thus are easily obtained. (2) Geom- superacid media etry about the carbocation center is similar to that of norbornyl. (3) Participation or 1,2-alky1 migration in the parent cation is degenerate as in the norbornyl cate ion. 54 A plot of the cmr chemical shifts of C(S) versus C(7) for cations gee-g is linear throughout the range of aryl substituents studied (Figure 7). The ordering of points corresponds qualitatively to Brown 0+ constants which re- flect the ability of a particular aromatic substituent to lower the energy of the cation by inductive and resonance stabilization. Specifically, a decrease in the electron releasing capacity of an aryl substituent should result in a corresponding increase in charge density at C(6). Thus a deshielding of the C(5) and C(7) resonances derives from the increase in positive charge at C(6). For classical ions in which only a-inductive effects are transmitted along the chemical bonds one would expect to obtain a linear expression from a plot of the B-carbon parameters with in- creasing charge at Ca' Indeed this is precisely what we observe. In Figure 8 the carbon—l3 shifts for the carbocation center in the bicyclo [3.2.1] octyl cations are plotted against those in a series of arylcyclopentyl cations as determined by Chambers88. That a reasonably linear cor- relation is observed with such a well-established classical model confirms the classical nature of these bicyclic cations throughout the free-energy range studied. In contrast a nonlinear behavior in a graph of C(l) versus C(3) chemical shifts is observed in a correspond- ing series of 2-aryl-2-norbornyl cations (Figure 9). This plot shows marked deviation from linearity between the 55 mechloro- and p—trifluoromethylphenyl cations, i;2; 0+ = 0.373-0.61. It has been our assumption that an increase in positive charge at C(l) relative to C(3) would lead to a break in the plot. Thus we would expect the change in slope to arise from additional deshielding of C(l) with little change in C(3) chemical shifts. A comparison with the plot for the arylbicyclooctyl cations verifies that the deviation is caused primarily by a displacement of the chemical shift of C(l) rather than that of C(3). Assuming that the C(3) chemical shifts of cations 33$ and gég are very similar to those expected if no break oc- curred, then the C(l) chemical shifts are off the line by 1.4 and 3.2 ppm, respectively. The abrupt deviation in our plot suggests that a change in the mechanism for charge transmission to C(l) has taken place. Furthermore, a dramatic reversal of slope to higher field is encountered when one plots the arylnorbornyl a-carbon chemical shifts against those of arylbicyclooctyl (Figure 10). The chem- ical shift of cation Qég is shielded by 10 ppm relative to that assumed for linear behavior of the plot. As we have already stated in the Introduction, the ex- perimental observations supporting the intervention of a bridged-norbornyl cation have also been explained in terms of rapidly equilibrating, classical ions. Do rapidly equilibrating, classical norbornyl cations account for our results? 56 For systems undergoing rapid interconversion such as egg 2 3334' the central carbon ('C) will give rise to a single signal whose position is some average of the shield- ings for the nonequivalent carbons in the individual ions. The value of this average signal will depend on the relative concentrations of the ions present at equilibrium. A o- bridged species will exhibit a single signal for these carbons also, but its position may be expected to differ from that of the corresponding equilibrating system since c138. the hybridization of these carbons will be altere We can determine the expected value of the average signal for g4g (°C = 6274)_ gégé ('C W674) 'C for equilibrating ions 33% and 343$ by simply estimating their relative concentrations at equilibrium. Assuming that the aryl group would not significantly alter the rate of a 3,2 hydride shift in agg; relative to the known rate for norbornyl cation, WOlf was able to esti- mate that the ratio of ééai/éég at +70° could not possibly be larger than 10-4. Alternatively, a similar value can 57 be derived from the solvolysis rate data obtained for the §,E"(CF3)2-phenyl- and secondary-enggfnorbornyl-pfnitro- benzoates, gg and 38, respectively44. One must, however, assume: (1) that the electronic demands in a developing cationic center during solvolysis are less than those in a fully developed cation, (2) that both compounds solvolyze | CF3 OPNB F C 3 OPNB .253 49 relative rates: 1 510-4 xig classical transition states, (3) that g2 and 4Q experience similar steric factors leading to their respective transi- tion states during solvolysis, and (4) that introduction of an aryl substituent at C(l) does not drastically alter the stability (enerQY) of a classical 2-norbornyl cation. Considering these factors, then it seems intuitively reason- able that the difference in log k for solvolysis of com- pounds 32 and 4g is proportional to the energy difference between their classical transitionostates, which should define the minimum energy between their respective classical 58 ions. As models they provide a reasonable estimate (consider assumption (4)) for the maximum ratio of gég&/gég present at equilibrium, 3:3; 2 10". Returning to Figure lO,we can estimate a 'C(2) chemical shift of 6274 for classical ion 33g by merely extending the line which includes those well-behaved, aryl-substituted cations. We have intentionally underestimated the value for the 'C(1) chemical shift expected for classical ion 33%;, m574. Thus a 10'4 molar concentration present at equilibrium would lead to an upfield shift of approximately 0.02 ppm (200 x 10.4) for the a-carbon in cation 333' We, however, observe a 10 ppm shift to higher field. We therefore con- clude that equilibrating classical cations cannot account for our data. Our results seem consistent with the onset of C(l)-C(6) bond participation in those norbornyl cations more electron demanding than Z-mfchlorophenyl-2-norbornyl cation. The pmr study of the arylnorbornyl and arylbicyclo- [2.2.2] octyl series revealed that an unusual effect was present in the Erhalogen substituted cations causing their points to fall off the line. Farnum posed an equilibrium mixture of monomeric and dimeric cationic species (Figure 17) to account for their anomalous behavior, since large changes in the anisotropic environment associated with the proximity of an additional aryl moiety in dimeric species 33d should have a profound effect on the H(l) and H(3) chemical shifts. However, our dilution study45 over a 59 SO-fold change in concentration did not reveal any signi- ficant changes in the pmr spectrum (specifically in the H(l) and H(3) chemical shifts) of the pfiodophenylnorbornyl X 2 .___ 9© 9 Q. G :3 éé Figure 17 derivative. Inspection of Figures 6, 7 and 9 reveals that the 13C chemical shifts of the Ethalogen derivatives are remark- ably well-behaved. These results support the original hypothesis that a dimeric species is present, since magnetic anisotropy in the dimer would be expected to affect the 13C 1 chemical shifts proportionately much less than the H shifts. Thus, although the relative magnitude of the 13 magnetic effects is the same, the total range of C chem- ical shifts caused by electronic factors is 30 times greater than that of protonsag. Therefore large changes in anisotropy associated with the presence of gag is expected to have relatively little effect on the chemical shifts of the B-carbons to the cation center. In addition to o-participation there is another mode 60 of stabilization that is worthy of consideration. Traylor has proposed a hyperconjugative interaction of the C(1)— C(6) bond to account for the solvolytic behavior of the "59 of norbornyl system. Such "vertical stabilization neighboring o-bonds is available without changing the reactant geometry. In an attempt to demonstrate hyper- conjugation Brown90 observed a normal solvolysis rate for the p-exo-Z-norbornylcumyl derivative 3% compared with those for appropriate models, i.e. Erisopropyl, etc. fig 4% However, the strained C(l) orbital which is predisposed to hyperconjugation with a vacant pforbital at C(2) is in- sulated from the aryl pforbitals by C(2) in compound 3*. Brown's study seems poorly designed. Nonclassical participation results in a rehybridization about C(2) to allow a-overlap with the C(6) orbital. The increase in C(2)-C(6) overlap should occur only at the expense of C(2) overlap with the aryl substituent. We feel that obstruction of aryl conjugation by o-participation of the C(l)-C(6) bond is the source of the break in our plot of the arylnorbornyl B-carbon chemical shifts. On 61 the other hand, hyperconjugation of the C(1)-C(6) bond does not require rehybridization at C(2) to be effective and should not interfere with aryl conjugation. If two inde- pendently stabilizing influences within the molecule operate harmoniously with one another, then we are con- vinced that a plot of C(l) versus C(3) chemical shifts would be a sum of two independent linear plots and thus would be linear over the entire free-energy range. To explore the possibility of vertical stabilization we have prepared a series of 2-aryl-gng5,6-trimethylene- 2-norbornyl cations 333-3. In this tricyclic norbornyl analog the structural reorganization which accompanies nonclassical participation would force the trimethylene bridge into a pseudo-gndg bonding relationship to the norbornyl skeleton. The five-membered ring formed by the trimethylene bridge must become distorted causing an in- crease in the free-energy of the system. The ratio of 2x97 and fingers,6-trimethylene-Z-norbornyl formate esters pro- duced by the addition of formic acid to dicyclopentadiene under equilibrating conditions suggests that the 252:5,6- trimethylene derivative is ~4.2 kcal more stableloo. Hyper- conjugative stabilization should not suffer from severe steric interactions, and consequently, should be a lower energy process. Figures 11 and 12 contain in graphic form the pertinent information for the 2-ary1-S,6-trimethylene-2-norbornyl cations géa-g. The nonlinear behavior of the C(1) and 62 C(3) chemical shifts with increasing electron demand clearly demonstrates that we are observing the onset of charge leakage to C(l). Appreciable deviation from linearity occurs at a 0+ value similar to that observed in a similar plot for the arylnorbornyl cations. Moreover, a change of slope to higher field occurs in the Ca chemical shift cor- relation with our bicyclooctyl standard. The qualitative similarity of both plots to those found for the arylnor— bornyl cations leaves little room for doubt that we are indeed observing the onset of nonclassical participation, although quantitatively the change in slope is not nearly so dramatic in this case as in the norbornyl system. In the aryltricyclic cations the free-energy decrease assoc- iated with electron delocalization more than compensates for the small free-energy increase associated with minimal atomic movement causing steric interactions, although the overall free-energy associated with c-participation is some- what greater than that of the parent system. Our data suggest that with sufficient electron demand norbornyl- like systems require participation with atomic movement even in those cases in which an unfavorable energy change is associated with the movement of atoms. Figures 13 and 14 provide in graphic form chemical shift correlations of C(S) versus C(6) and Ca versus those of our bicyclooctyl model, respectively, for a series of 2- arylnorbornen-Z-yl cations QéQ-g. It is evident that the 63 shielding of the C(6) resonance relates linearly to the de- shielding effect found for C(S). The correlation implies that charge distribution placing charge at C(S) involves 2 toward sp3. Furthermore, rehybridization of C(6) from sp the sharp break in the C(S) chemical shift-o+ plot (Figure 18) indicates that a dramatic change in the mechanism of charge transmission to C(S) occurs at the phenyl derivative (0+ N 0). The impressive turnabout in the Ca chemical shifts to higher field reflect this change, and at the same time, implicate rehybridization of the carbocation center with increasing electron demand. The results are consistent with the formulation that minimal charge delocalization involving little structural and hybridizational change is followed by substantial electron supply involving dramatic changes in structure and hybridization under the increasing demand of the cationic center. These data, however, do not distinguish between nonclassical n-participation and rapidly equilibrating. classical ions 4g and 38’ since it is most difficult to “ti 4___ A Ar 4% 64 determine the relative energies for these structurally different ions with certainty. Brown points to the need for caution in extrapolating data from superacid media to solvolytic mediagl. we agree. However, to conclude that qualitative comparisons are there- fore invalid seems unreasonable. A graph of the 25232292 rate ratio for a series of 2-arylnorbornen-2-yl penitro- 92 stands benzoates as determined by Brown (Figure 19) beside our plot of the C(S) chemical shifts of the 2-aryl- norbornen-Z-yl cations versus 0+ constants (Figure 18) . A plot of the olefinic pmr parameters of the 7-arylnorbornen- 79 7-yl cations reported by Gassman and Ritchie versus 0+ constants is superimposed on a Hammett plot for a 78 in Figure series of 7-arylnorbornen-7-yl penitrobenzoates 20. At first glance our plot appears significantly dif- . ferent from that for the 7-norbornenyl cations. Let us construct a hypothetical curve which depicts the change in the nmr parameters of a n-participating moiety over an extremely broad free-energy range (Figure 21). From left to right, initially the change in chemical shift is some linear function of electron demand. That a drastic change in slope occurs indicates the onset of participation involving the n-electrons causing a deshield- ing of the olefinic resonances. This portion of the curve should again be linear with respect to free-energy. Even- tually electron donation saturates. Whereas Figure 20 is 65 II . mucmumsoo +0 .m> mcowumo H>Imlcmcuonuo:H>H¢|~ mmumoucmnouuwz a“ uuflnm Hmoflsmgo Amvo «0 sauna .ma museum um H>mm5cmGHOQH0cH>H¢ IN“ :1 owumm mumm ovum 6me co panama couuomam +5 mawmmwnocH mo uowmmm .mH musmfim $6... 0.0 0.0- . g . . . . J . . . o . c . 4 .1 1| II .xfll. .1 oils, 0 . J. I .II. ,5 . ll' . 1.10 . \AfI'.AIIPI Vl'q.(i||tillllliliillwli‘illi . .25fi¥8£.8_ V;M.1 m.::.---.,_g- ye; ,W m.%:a. 4-x.” 9- QT o..- .0..- 0.. on o- v- i . /m H. .1 . n .. _ : _ . . . _ . .,. M . d < a q . ab . f O c “T l .....*._ i 4 i : ! l , , ' , g . Taff’wifl f I i. l I I 1 Al 1 I .1 w - . - .1... ,.. w», _:. 1h 3»... g 1. 4 . , . ,_ .1 1.: a... ... _. .p :. Or. .......: . . _...- ”.1 ...fl. .H H -Hh--... -E. . ,. (add) $3 3 .............. fi—oH—4o—4 -1 . .. L . . . I » I +4 I ................ M 1--.... J (I . . O . . . 1 1 .TI .,.. . .. .' | . 1) l ‘1 L p.- a-o-~-.-—y—o—o- O « A l . . . . q . , c ¢ 9 L 115.... . I If. .n -tu‘ll.f.l’v .. 66 .mucmumcoo 0 .MM mummum0ucmnouuwz LN Hasnncocuonuocamumnh mo mmfiumm w you mucmumcoo mummg won can mausofluuo Hans-cucuonuocflsuauh an uuflnm Huoflsmno 1M1.1~1m mo sauna .o~ musmwm +|b v!.. - 7715* L! ' 1,111 ”#1.. j)Bo' 1* . ». ”a--. 1w. ..w-»- me.- _. w , ,»_ -1..;1.: 9.. «u .1a. OJu “‘4 33" I." "qr—-W-4 .- u.) “H ..,.. ..... . H. . . ... 1.1 .. .4 , . ..: ., , :1 . [Mull I!» ll.l|.0'. Tull-I: QI'LIII .Ovl‘l II...M‘ II..JWIIX.H.'I,I». r . 11% 1| I|I| ... . ’4] {.11- III...) . . i. . ._ 1 . _ ._ 4 . . 1 i ..1 I-\ r. . : .1. . .W 1 r . W.w.- _ ., .rr 1 I l , . W V . , f, ; . _, .+ . 1 1 ,1 1 . I . 1 . 1-. mt; M . .. . H. . 1 g 31.. .. ..... 1 1 . . - . W. .. I- ... ...-.1. ...fi . .. . j 1 vi -111 1 h .r +0. ibill.qt o i V a A .291 I’Ln....ilv)-In¢ .141:- .|..il I01 .1- -..Ivll-W. ..r: .M .1. 1 ...-fl.“ 1. 1.11 . 1 :1 ,1. .1 .11 1....W1: ..1 . L 1.... W 1. ....W.H-.1-.1. .1 .- .1.A....- ...-. ,. . 11. ,; o.” ......WH- . n 1 .. ”- W... :1 . . . .1 . .. ./ ... - .1: 5;. ”Hm-.. - . ..- 1 ..... . . H-JW .1... v. .1.- .fi... -1, .....W ..... . 1 Inn--131:- - .-. Fr). - . 1- i -.¢mw.v ....... W1- u-r vyvbrt ... A H w /r n c u r :,. - J 1 L1 ., .. .. . 1 -W» Jl... . , 1 #:La%gj-ru-u1 . 1 a. 1 .flfix_ .. 1 r -1 .VI 1 1 1 -. . . .. - _. . . . ... . . .. ...... .. . . Wm.n.....WWJ..:-:.¢-.1-.m-,...Jw-www.1- -.- W-_r- .....1......-..-....:-..:} ..IO --_+----.z. -.--...W 1.1.11.1- -.:, -../i -.- - H 0.9. - I . . . . 1- 1 , .. s. . tn Wt... .-.-.8... ..m--..-.-.--JI...H... . . . . 1. ...1 1 ...... W .-. .1 1 _ . . .1. -......W.-. .- ....W -1 _ ., 1 . 1:: l - ..:-J “H4113 ‘ “W1“; ‘ ”i..- ” --.}. 67 Onset of n-participa- tion Deshielding of olefinic resonances n-participa- Saturation of If tion Figure 21. Dependence of nmr Parameters in the Presence of n-Participation With Increasing Electron Demand. best described by the lower half of our hypothetical graph where saturation of w-participation occurs, Figure 18 is best represented by the upper half which reflects the onset of participation. Both figures, however, reveal the same general phenomenon, £12; w-participation. Compare Figures 18 and 19. The onset of participation is observed in the cations generated in FSO3H before it is observed in solvolysis (0+ 0 vs. + 1.04). It seems reasonable that the more electron demanding cations would benefit more from electron donation from nearby n- and o-bonds than their corresponding solvated transition states. 68 From the data presented in Figure 20, Gassman and Richey conclude that the 7-aryl group eliminates double bond in- volvement somewhat more readily in the cations than in the transition states for solvolysis, iLththe onset of n- participation occurs much earlier in the solvolysis of the 7-norbornenyl derivatives than it does in the corresponding cations in superacid. However, if we now replot their data as a function of n-participation (Figure 22), it then becomes evident that their interpretation is incorrect. While the solvolysis results are indeed consistent with the onset of n—participation, their nmr observations are consistent with our aforementioned formulation that the break in the plot reflects a saturation phenomenon. The small changes in the H(2) and H(3) chemical shifts to aryl substitution in the region 0+ 0.0 - 1.04 are consistent with a pair of classical cyclopropyl carbocation structures or their resonance hybrid having most of the positive charge concentrated at C(2) and C(3) and with little charge at C(7), since in these classical structures the pmr parameters should be affected insignificantly by long range inductive effects of the 7-aryl substituent. Further- more, the pmr Spectrum of the pfmethoxyphenyl derivative suggests that it approximates a classical structure; there- fore, the onset of w-participation must occur before (at a 0+ value more negative than -O.78) the 7-pfmethoxypheny1- norbornenyl cation. In view of these findings let us re-examine both the 69 95 J Orig; or: .cowummaowuummnp MO cowuocnm m mm cwuuoamwm om musmflm N wudmwm 70 solvolysis and our carbocation studies of the arylnorbornyl derivatives. From his investigation of the solvolysis of arylnorbornyl derivatives, Brown has concluded that there is no evidence for nonclassical participation in the tran- sition statesgl. Thus, a plot of the 95232292 rate ratio delineates a linear free-energy relationship over a range of aryl substituents from pfmethoxyphenyl to m,mf—trifluoro- methylphenyl, approximately 75% of the difference in activa- tion energy between 2-pfmethoxypheny1norbornyl and norbornyl derivatives. In our plot of C(l) versus C(3) chemical shifts we find a deviation from linearity for the arylnor- bornyl cations: The break in our plot occurs between the mechloro- and p—trifluoromethylphenyl derivatives. We have shown that for the cations we have studied the onset of participation in solvolysis lags behind that ob- served in the study of the cations in powerful acids. Indeed, we have shown that this observation finds a reason- able interpretation in the relative electron demands of the solvolysis transition state and the cation in weakly solvating media. Comparing the solvolysis and stable carbo- cation studies for the norbornyl and norbornenyl derivatives, we would expect the break in Brown's plot of the norbornyl derivatives to occur for those aryl substituents more electron withdrawing than the m,mf-trifluoromethylphenyl derivative. Therefore, we conclude that Brown's investi- gation is incomplete. Our investigation of a series of 3-ary1-3-nortricyclyl 71 cations gggaf reveals that there is extensive charge de- localization into the cyclopropyl ring (Figure 15). While there is very little deshielding of the C(4) resonance, the C(1) and C(6) chemical shifts concurrently move to lower field with increasing electron demand. A break in the C(l), (6) chemical shift — 0+ relation- ship (Figure 23) occurs for those derivatives more electron demanding than the 3-phenyl-3-nortricyclyl cation. The change in slope reflects the inability of this cyclopropyl moiety to stabilize the more electron demanding cations as efficiently. Furthermore, the reversal in slope to higher field in a plot of the Ca chemical shifts versus 80 0‘ VI SC .,. ("“0 SO q-+ Figure 23. Graph of C(1),(6) Chemical Shift in 3-Aryl-3- Nortricyclyl Cations 23. 0+ Constants. 72 those of our bicyclooctyl model (Figure 16) implies a changing geometry about the carbocation center. Are these results consistent with the presence of nonclassical par- ticipation involving a cycloprOpyl ring in which atomic movement is repressed by the rigid nortricyclic skeleton? Comparison of the relative rates of solvolysis for compounds 3% and gé reveals a major contribution of the cyclopropylcarbinyl moiety93. However, the ability of a cycloprOpyl ring to stabilize an incipient secondary carbo- cation by neighboring group participation can be demonstrated X X j 4% a Relative Rate 1 ~108 by a rate difference of up to 1014.65'94-96 Again, can this discrepancy in enhanced rates of solvolysis reflect the inability of the constrained cyclopropylcarbinyl moiety to participate with atomic movement? The answer to this question must await further investigation. EXPERIMENTAL Melting points were taken on a Thomas Hoover capillary melting point apparatus and are uncorrected. Infrared spectra were recorded on a Perkin-Elmer Grating Infrared Spectro- photometer, Model 3278. They were calibrated with the 6.23 p band of a polystyrene film reference. Mass spectral analysis were performed by Mrs. R. L. Guile at Michigan State University using a Hitachi Mass Spectrometer, Model RMU-6. Nuclear Magnetic Resonance (nmr) spectra were ob- tained using a Varian T-60 Spectrometer. The nmr data are presented in the following manner: 6 6.00 (2H, d, J=4). All spectra are recorded in delta (6) units relative to tetra- methylsilane (TMS). The multiplicities are s=singlet, d= doublet, t=triplet, q=quartet and m=multiplet; br=broad. Carbon-13 Magnetic Resonance (cmr) spectra were obtained using a Varian CFT-ZO Spectrometer equipped with a Varian V-6040 N-M-R Variable Temperature Controller. All chemical shifts are expressed in parts per million (ppm) downfield from external capillary TMS and d6-acetone. Primary carbon- hydrogen coupling multiplicities were determined by the off- resonance decoupling (spin-tickling) procedure which does not allow determination of the size of coupling. The temperatures at which the various cmr spectra were recorded are accurate within i3°C. Temperatures were calibrated for the probe, not for the sample. 73 74 Carbonium Ion Formation The acidic medium was chosen to try to ensure complete ionization of the carbonium ion precursor yet to avoid un- wanted side reactions. The carbonium ions were formed using either of the following methods: (1) When FSO3H was used the carbonium ion precursor was dissolved in CFCl3 (Freon 11) and slowly added drop- wise to the rapidly stirred acid at -78°C and maintained under a blanket of nitrogen. (2) When a limited amount of FSO3H was necessary, SbF5 was added, or low temperatures were desired, SOZCIF (50- 85%) was employed as a cosolvent. The carbonium ion was prepared directly in the sample tube. Neat carbonium ion precursor was introduced to the ionizing medium (maintained at -78° or -110°) and carefully mixed with a glass rod until a uniform mush was produced. The sample was allowed to warm to the desired temperature. Additional 802C1F was added and the sample vibro-stirred until a homogeneous mixture resulted. Carbonium Ion Precursors The alcohols used in this study were prepared from the appropriate Grignard reagent and ketone; i.e. 2-norborn- anone97, bicyclo [2.2.2]-octan-2-one98, tricyclo [2.2.1.02'6] hepta-3-one99, tricyclo [5.2.1.02’6] deca-8-oneloo, 2- 97 norbornenone . In every case a 10% molar excess of aryl 75 bromide and a 20% molar excess of Magnesium metal was used with respect to ketone. Yields of aryl alcohols ranged from 60% to 95% based on ketone. Tables 10-13 summarize the important physical and spec- tral data obtained for the compounds prepared. IR and NMR Spectra A. 2-Aryl-exo-5,6-Trimethylene-endo-2-Norbornanols. pr6H4OCH3 (olefin): nmr (CC14) 6 6.70 (4H, AA'BB', Av = 31 Hz, J = 8.5 Hz) 5.91 (1H, d, J=3 Hz) 3.61 (3H, s) 2.77 (1H, m) 2.48 (1H, m) 2.20-.68 (10H, m): ir (neat) u 6.20, 9.62. (15H, m) 2.12 (6H, br s); ir (neat) u 3.04, 6.17, 9.59. pr6H4F: nmr (CC14) 6 7.17 (2H, m) 6.77 (2H, distorted t, J=9 Hz) 2.62 (2H, m) 2.10 (2H, d, J=4.5 Hz) 0.60-2.0 (11H, m): ir (nujol) u 2.98, 6.17, 9.58. p—C6H4Br: nmr (CC14) 6 7.24 (4H, AA'BB', Av m0 Hz, J=9 Hz) 0.60-2.86 (15H, m); ir (neat) u 3.04, 6.24, 9.89. ir (KBr) u 2.98, 6.14, 8.92. 3,5-C6H4(CF3)2: nmr (10% CDC13/CC14) 6 7.80 (2H, br s) 7.61 (1H br s) 0.60-3.0 (15H, m): ir (KBr) u 3.02, 6.14, 8.88. 76 caumflo.. unaoa conumEMHnsm. “mango mam .maumc wen «mm «on .mma-m.a~a ---- cmmmflmmflo Nimmocmmmoum.m Amaumv sum Amaumc mum . mmm mmm om.omum.ms ---- cmmmamsau mmoemmuum Amanmv can as N as mNN mmm oosumm amp cowmeflo memo AmH-ac om~.mm~ as ”.0 pm mom.oom mom.mom ---- cone oumaammau umemmo-m Amauac m- . as H.o um ma ma v m mew mew oomumm .oma on m o m a cum Amaumv mmm . Ama-mc Hem as H c an «N ma N m m m mmm mmm ---- .omH o m o A moo m o-¢.m as H.o um oqw cam ---- coca coumsao ..mmUOQmmoum Aocsomv m\E A.0Hmuv m\E me am no an maneuom w>Hum>mea « unadomaoz muuowmm was: .mHocmsuonHOZINlocumlmcmahzumEHualw.mlomeH>H¢|~ mo coquHQOnm .oa manna 77 .ucflom cowumeflanam t. Amawmv mmm mmm oomnmm ssommo um commammao mammovmmmonm.m Ammumc WNW smm osmumm eeowmo um cmmmameao mmoemou.m Assume WMM mmH nnnnn ssommo um. ovammao mmmo Amfl-ac “MM «HN ..... as.mmw um cmammao Nimmuvmmmu-¢.m Amanac WWW mam ammumm eaowmm um memameao mm004mmuum Accsomv m\s “.0Hmov m\E as am no an waseuom m>wum>aumo c unasomaoz muuommm mum: .mHouwncwcuonuocH>H¢|~ mo coaumummonm .HH manna 78 .ucwom coflumeflansm .4 Amanmv mom mmm mmm om.moaumoa as H.o um .HMH commammao «Ammovmmwoum.m as m an n ma «a m v m «mm «mm ammuqm 0-H o m m u mu a cum Asmumv mma.mma 1mm-mc sma.mma as H.o an MA ma 4 m I -~.o- o-.o- .mm-m.vm .om oHo m o Ho m one Aaumv mma SE H.o um mma mma ..... coca csammao mmmo Amaumv mma BE H.o um oo~ oom .m.~o-am .HHH cmamvao mmoemmoum “Humv mam ES H.o um mam mam .vm-mm .OHH Nemameao mmUquooum Awesomv m\E A.oamov m\E me am Mo on maneuom m>flum>wumn a umasomaoz manommm mum: .maocmummn nw.~o.H.N.NH oaoaofluelmnawumlm mo coaumuwmwum .NH manna 79 .ucwom cowumeHndm fl .ma-mc mHN as H.o um omm omm ..... .oma cmmmmflo Nammucmmoo-v.m ES H.o um mmm mmm .ov-mm .mma Ncowmflo mmUOvmmoum Awesomv m\& A.onov o\E me am no on masauom m>aum>flumo t muuommm mum: amasomaoz .maocmuoo mH.N.MH odoaowmsmlamu