CQfiFGREefié-EIGNAL muss 0F Di -HETEROA?0M SUBSTRTUTED ALEEEHYEES BY NUCLEAR MAGNET iC ‘RESGNANCE SPECTROSCOPY Thesis for the Degree of Ph. D. MICHEGAN STATE UNIVERSITY, {SAVED JOHN FENOGLIO 1969 THESIS This is to certify that the thesis entitled COIWRI‘MTIONAL ANALYSIS OF a-ILUI‘EROATOM SUBSTITUTED ALDEHYDES BY NUCLEAR MAGNETIC RESOI‘JANCE SPECTROSCOPY presented by David John Fenoglio has been accepted towards fulfillment of the requirements for PhoDo degree in ChemiStrV q/ . 6,. J gem/5550] Mojor professor Datejt/vm' /0,1/?6? 0-169 i. LIBRARY Michig .11 State University IV ..".‘1‘E..§..§9.“§f- w ABSTRACT CWFWMTIGML ANALYSIS OF Cl"HFITEROATO‘I SlBSTITUTED ALIEHYES BY NltLEAR MNEI'IC RESONANCE SPECTROSCCPY by David John Fenoglio Nuclear magnetic resonance spectroscopy has been applied to the cmfomatimal analysis of a—heterostbstituted acetaldehydes . 'Ihe time averaged vicinal spin-spin couplirg constants between the aldehydic and a-protcn(s) of chlor'oacetaldelryde , brmnacetaldenyde , nethozwacetaldehyde , plenoxyacetalderwde , nettwlnercaptoacetaldemde , dichloroaoetalderwde , di- bramacetaldelwde , cyclOpr'opanecarboxaldehyde and glycidaldel'wde were studied at 60-Mc as a Motion of tamerature and solvent. 'Ihe data for the mmosrbstituted acetaldehydes were interpreted in terms of rotamer-s I and II, whereby a single bcnd eclipses the carbonyl group. The data for O o H L R L H H (R)H// H” R Ha?) I II the disubstituted acetaldetwdes were examined in terms of I and II (three- fold barrier to rotation) and I and III (twofold barrier to rotation). -1. David John Fencglio -2. 'Ihese analyses led to the follouirg concltaions. (A) mlomacetaldemde , O R at VL \ H H III branoacetaldelwde , nethmwacetaldemde and phemzwacetaldemde: (l) 'Ihe data are best interpreted in term of a threefold barrier to rotation ammd the 8132-8133 carbon-carbon bard. (2) The mat stable rotmner for these canpcunds is the one with the carbon-hetercatan eclipsirg the carbonyl domile bond. (3) The flee energy and enthalpy differences for I='.-‘II are stmly solvent dependent, being much more negative in solvents of high dielectric cmstant . (B) mtmlnercaptcacetalde- rude: (l) A threefold barrier to rotation about the spa-spB carbon- carbcn bond best fits the data. (2) ‘Ihe most stable rotamer is the one with the carbm-hydrogen bond eclipsing the carbonyl double bond. (3) The flee enery and enthalpy differences for I¢II are not very sensitive to the dielectric constant of the solvent. (C) Dichloroace— taldetwde and dibmncacetaldetwde: (1) lbs data are canistent with a threefold barrier to rotation about the spa-”3 carnal-carbon bond. However, the possibility of a twofold barrier to rotation can not be eliminated. (2) In nonaranatic solvents whose dielectric constant is less than sit, the free energy and enthalpy values for 1:11 are positive, ii... the rotanerwith the carbon-demgenbmd eclipsing the carbonyl ddble bond is more stable. (3) In nmaramtic solvents whose dielectric David John Fenoglio -3- constant is higher than seven, the free energy and enthalpy values are negative, ii” the more polar rotaner is more stable. (D) Cyclopropane— carboxaldetwde and glycidaldehyde: (1) No manhiguous decision can be made as to whether there is a dominant twofold or threefold barrier to rotaticn about the spZ-sp3 carbon—carbon bond. (2) The most stable rotamer is the one with the carbon-twdrogen bond eclipsing the carbonyl dormle bond. (3) 'Ihe cyclopropyl group acts as an electron donor, whereas the oxirane group acts as an electron withdrawer. Nevertheless, the oxirane ring reserrbles the cyclopmpyl ring more than it does the nethoxy m. Chemical shifts of the aldekwdic and nethylene (or nethine) proton(s) were also measured in conJmctim with the coupling cmstants. It was foundthatthe Mealshift resultsareinagreementwithanewmodel for the anisotrcpy of the carbotwl group. This data also reinforce the cmclusions derived fran the coupling cmstant data concerning the stability of the rotaners. WWTIGW. ANALYSIS OF cl'HEI'EROA'I’M SLBSTITUTED AME-MES BY NLCLEAR MAGNETIC W SPECTROSCOPY BY DAVID JON FEWLlO A 'l'tESIS Stbmitted to Michigan State University in.partial fUlfillment of the requirements nor the degree of mam 0F PHILOSG’HY Department of Mstry 1969 U'x TOWANDDAD ANDMYWIFE: JILL 11 ms The author wishes to express his sincere appreciation to Professor G. J. Karabatsos for his help and friendship during the course of this investigatim. He also wishes to thank Professor Karabatscs and the numbers of his research group for the mam stirmllating ccnversatiom outside the world of chemistry . Special thanks are given to Mrs. Celia Miller for typirg the "Hilarity of this manuscript. The author also wishes to emress his appreciation to the National Science Foundation for the financial assistance and the Dow Chemical Carpamr for a Dow Sumner Fellowship (1967). TABLE'OF (IJNI'ENTS IWIW coo0.0000000000000000...oooccoccoooooooocococococoooo mu OOOOOOOOOOOOOOCOOOOOOCOO0.0....0.0....OOOOOOOOOOOOOOOOOOOO A. Coupling Cmstants B. meal- Srfifis 0.0.000000000000000...OOOOOOOOOOOOOOOOOOO DlscmSlm OOOOOOOOOOOOIOOOOOOOOOOOOOOOOOOIOOOOOOOOOOOOO0.0.0.0... A. Cowl-1m Camtmts OOOOOOOOOOOOCOOOOOO0.0.0....0.00.00... I. mmwStituted Awmmdes OOOOOOOOOOOOOOOOOOOOOO. (a) Consideratim. of a 'Nofold Barrier for Chloroacetaldehyde and Brennacetaldehyde........ (b) Relative Stabilities. of. the. mncstbstituted Ammmm 000000OOOOOOOOOOOOOOOO00.00.000.00 'DihalmmW®8'OOIOOOOOOOOO0.0000000000000000000 (a) Consideration ofaTwcfold Barrier Glycidaldemde and Cycloprcpanecarboxaldei‘wde . . . . . . . (a) Twofold or 'Ihreefold Barrier to Rotatim Effect of Solvent Polarity on Rotamer Stabilities . . . (a) Mmosmstituted Acetaldehydes (b) Dihaloacetaldemdes Caxpariscn of Results with Other System (a) ManstbstitutedCase (b) Mermaloacetaldemdes _v_s_ 2-halocyc lohexanones . . (c) DismstitutedCase iv 27 30 141 M ‘45 50 51% 55 57 57 58 59 59 63 65 B. Chemical Shifts EXPERIMENTAL ~- A. Reagents and.Compounds B. Solvents ..... C. Synthesis ... I. all 0.000000...00..0 II a 0.00.... . .... .0.0000 III mmmew& 0000.00.000. ......... Iv. Dib:a'm tawny ¢00000000000... 00000 0............... v Immmwwk 00.0000... 00000 0.0 ............. VI eyelm ............. .............. . klmqv a“ .0.0000..0.0 1D thy ropanecarboxaldehyde . ....... ........ . mM.R. a1 & 000000000000 REFERENCES Page 68 7a 7a 7a 7a 7a 75 76 76 77 78 78 so TABLE I. II. III . VII. VIII. XII. XIII. LIST OF TABLES Vicinal Spin-Spin Couplirg Constants, in c.p.s., of Sammwwta&m¢s 0.0.0....0..000......0.00.0..0. Vicinal Spin-Spin Couplirg Comtants, in c.p.s. , of Some Heteroacetaldetwdes and Cyclopropanecarboxaldemrdel . . . . Temperature Dependence of the Vicinal Spin-Spin Coupling taldem Constants of Chloroacetaldel'wde and Bromoace Tenperature Dependence of the Vicinal Spin-Spin Coupling mmt Of Dimlmmma 00000000000...0.0.00.0000.0 Temperature Dependence of the Vicinal Spin-Spin Coupling Comtant of Dibromoacetaldem'de Tenperature Dependence of the Vicinal Spin-Spin Coupling Constants of menozqracetaldemde, Methoxyacetalderwde , Nbflwhemaptoacetaldelwde , Glycidaldehvde and Cyclo- pmcamwm¢ 00.0..0000000.00.0000000...00000 The Chendcal Shifts of the Hetemacetaldehvdes Solvent Dependence of the Relative Rotaner Pepulatiom of Chloroacetaldelwde and Brancacetaldetwde Solvent Dependence of the Free Emery Difference, so°, Between Rotaners of Chloroacetaldetwde and Bromoace- taltblwde .................................................. Enthalpy Differences , AHO , Between Rotaners of Chloro- amtal&m&mld8mwtalm& 0.0.000000000000000000000 Solvent Dependence of Rotamer Populations of Phenoxyace— taldem'de , Methowacetaldeiwde and Mettwlnercaptoace— taldemrde .................................................. o Solvent Dependence of the Free Energy Difference, AG , Page 8 10 ll! 17 18 21 28 3’4 Between Rotaners of Hemmacetaldetwde , Methoxyacetaldetwde “kWMMtMtal-m 00000000000000000000000000000 Enthalpy Differences , AH° , Between Rotamers of Phenoxyace- taldehyde , Methoxyacetaldel'wde and Mettwlmercaptoace— taldel‘wde .................................................. vi XVII. XVIII. Rotamer Pepulatiom of Blencxyacetaldetwde and mmmawtalmfi 000000.00.00.000..000...000000.000000... Free Enery Differences, AGO, Between Rotamers of Hermacetaldemde and Methmqracetaldehvde Solvent Dependence of the Relative Rotamer Populations of Dichloroacetaldehvde and Dibrunoacetaldehvde Solvent Dependence of the Free fiery Difference , AGO , Between Rotamers of Didlloroacetaldehyde and Dibromo- amede 00000......00000.00000000.00.00000......00.00.. Enthalpy Differences , AH° , Between Rotarers of Dichloro- mtflm¢ m Dibmcetalmfi 00.0....0.0000.00000000 Enery Differences Between Rotaners of Sale Related System Conpariscn of Euthalpies Between Mmdlaloswstituted mmmwStituted mm 0.0000000.00.000.000..000....0 vii P889 '42 ‘43 1&8 149 51 61 67 LIST OF FIGLRES FIGLRE Page 1. Tarperature Dependence of the Cowling Constant for mlmcetalm& 0.000000000000000000000000.00.000.000...o 15 2. Tenperature Dependence of the Cowlirg Coretant for mmmm¢ co00000000000000.000000000000.000.000.000. 16 3. Temperature Dependence of the Cowling Constant for mmmwtalm& 0000000000000000....00000000000000..000 19 it. 'Ienperature Dependence of the Cowling Constant for Dibmxtaldee 000000000000000000000000.000000000000... 20 5. Temperature Dependence of the Cowling Constant for mmwawmma 000000.000000000000000000000000000000000 23 6 . Tenperature Dependence of the Cowling Cmstant for Hemmtalma 000000000000000000000000000000.00.....00 2h 7. Teuperature Dependence of the Cowlirg Constant for mwmmtmmma 000000000.000000000000000.000000. 26 8. Dependence of the Meal Shifts of the Aldehydic and Methylene Protae on Rotamer POpulatims for Chloroace- 9. Dependence of the Chendcal Snifts of the Aldehydic and Methine Protons on Rotaner Populations for Dichloroace— mm¢ am mmwmm& 0000.00.00.0.0000.00000000 70 10. Dependence of the Meal Shifts of the Aldehydic and Dbthylene Protms on Rotaner Populations for Phencxyace- tflm® andmmwawtaldee 0000.000000.00.0..0.00000. 71 ll. Dependence of the Cherxdcal Shifts of the Aldehydic and Natty lene Protons on Rotaner Papulaticns for mmlmmmtalmm 00000000.0000000.000000...00..... 72 V111 INTRGIKITIM Rotational isomer-ism about a carbon-carbon sirgle bond has been a problem investigated by may techniques. In particular, the relative stabilities of 1' and 2‘. have been examined with respect to rotation about the carpal-carbon single bond Joinirg the spa-sp3 hybridized carbon atcms as a nmctim of x, Y and R. For exarrple, a few of these R Y ' Y / l l / R H l 2 ~ ~ studies include Hanan and infrared studies (11 o-haloacetmes (l, 2, 3), haloacetylhalides (u, 5, 6), N—nethylcbloroacetamde (7). etl'wl halo- acetates (8) and 3—ha10propenes (9); ndcrowave studies on acetaldemrde (10), acetone (11), propionaldehyde (12) and olefins (13, 11!) ; electron dif- fractim studies (:1 aliphatic ketmes (15) and aldehydes (l6, l7); and n.m.r. studies an ketones (18), 3-swstituted propenes (19, 20, 21, 22) and aldemdes (23). Certain basic factors have been proposed to explain the results from may of these studies. 'lhese factors include nmbcnded (attractive and repulsive), dipole-dipole and dipole-induced dipole electrostatic inter- actions. For exanple, in additim to electrostatic dipole-dipole inter- actions, nmbmded repulsims between R and Y in retailer 1 of chloro- -1. -2- acetone (1) , chloroacetyl chloride (14) and N—netlwlchlomacetandde (7) have been invoked to explain the different 1/2 ratios of these compounds . In agreenent with these cases is the report (214) that chloroacetaldelwde exists essentially in conformation '1. When this is taken in commotion With the finding that chloroacetone exists in both '1' and 2, this report is certainly consistent with the concept that nonbonded interactions be- tween R and Y significantly affect the relative stabilities of l and ‘2' of chloroacetone. These same factors have been invoked to correlate and interpret a large nurber of data (25) on the relative stabilities of the axial and equatorial conformers of 2-halocyc10hexanones . In contrast, nonbonded repulsions in aldehydic systems (X - O and Y - H) (23) have been shown to have only a minor affect on the stability of the rotaners. For exanple, AH° for 22:3. is -800 and -500 cal/mole when R is methyl or isOpropyl, respectively. Out of the 800 cal/mole mserved when R is method, nabcnded interactions accomt for less than 200 cal/mole. These interactions becone significantly repulsive mly when R is t—butyl, in which case 1 is favored over 3 by 250 cal/mole. Several carpounds containing a single halogen (R - halogen) on the d-carbon to the double bond have been found to exhibit a threefold barrier to rotation about the spa-sp3 carbon—carbon bond. For exanple, AH° for la-tg is -560, -500 and o cal/mole for etlwl fluoroacetate, chloroacetate and bmmacetate (8), respectively; it is -lOOO and ~1900 cal/mole for bromacetyl chloride and branoacetyl bromide (l) , respec— tively; and it is -100, +100 and over +100 cal/mole, respectively, for 3-fluoropropene (19, 26). 3-chlcropropene (20) and 3—bromopropere (27). In nest cases, the data have been interpreted in tame of perfectly eclipsing cmforrnaticns, i_._g_._, dihedral angles of zero between planes -3- HCCandCCXinlandbetweenplanes RCCandCCXin 2. Forchloroaoetyl N N chloride (5) and branoacetyl chloride (1), a dihedral angle of 30° (3) for l was fomd to best agree with the experimentally determined vibra- m H 0 ¢ I 30° H (Br)C Cl 3 Na tional frequencies. It was also pointed out (23) that assignments that were made fran n.m.r. studies could not make such distinctims in the di- hedral angles. Ihe only mondlaloswstituted coupomd that has been described by a twofold barrier to rotation about the sp2-8p3 carbon-carbon bond is fluoroacetyl fluoride (28), whose AHO for 3:22 is -910 cal/mole. o H H\ L F L \ F F l/ H H Analogous dihaloconpomds differfmmthemmhaloccwcmds intwo respects: (a) '32 AH" values for £31 are such more positive than the 3“: correspondirg ones for F2. In all cases, except for dinethoxypro- pene (20), the we stable rotaner is the one with the carbon-hydrogen x H k R k Y Y x, ’ / R R H bond eclipsing the double bond. For exalrple, AH° for .8523 for dichloro— acetyl chloride (6) is +200 cal/mole; it is +500 to +11400 cal/mole for 3,3-d1flu0r0propene; and +800 cal/Incle for 3,3-d1chloropropene (20). (b) 'Ihe etml dihaloacetates, in contrast to the ethyl nmohaloacetates, exhibit a twofold barrier to rotation with MP for gig beirg +25 and 0 cal/mole for ethyl diflucroacetate and ethyl dichloroacetate, respectively (8) . 0 O “ L i L OEt ‘ OEt / x H 9 R e .. Bellamy and William (2‘4), by conparison of the vibrational fie— quencies of acetaldelwde , chloroacetaldehyde , dich loroacetaldehyde and trichloroacetalderwde concluded that in both gaseous and liquid states , dictllor'oacetaldekwde exists in essentially one confomation whose probable structure is 10 . 'Vb Cl Cl H 10 ‘VV There has also been sale question fran the n.m.r. results (23) as to whether cycloprwanecarboxaldetwde in the liquid phase is best described in terns of a twofold barrier to rotation (1,1 and 1A1}, as minimm enery cmfcrmatiom) as found in the gas phase (29), or in terns of a threefold barrier to rotation (11, 11a and 11b as mininun enery cmfomatims) . M W W XL H WL 3i llc 'Vh ‘VW MA. We have investigated dlloroacetaldehvde , brpmacetaldehyde , dichloroace- taldemvde , dibrcrnoacetaldelwde , methoxyacetaldehyde , phenoxyacetaldehyde and netmlrrercaptoacetaldemde , in order to determine the main factor or factors that ccntroll rotaner stabilities . Cyclwrepanecarboxaldemde was also investigated to see if, in the liquid state, there was a dmdnant twofold or threefold rotational barrier went the spa-3p3 carbm-carbm bmd. 'Ihe obvious relation of glycidalde- hyde to cyclopmpanecarboxaldemde also led to the study of this conpound .6. in order to decide whether a twofold (l2 and 13) or threefold (12, Ill and ’VV 'VV 'VV ’Vb 12) rotational barrier is pertinent. O o 0 O \\ H V H H H I / I 0/ H/ H H '1}, 1‘4 53 ... .12 \ 'Ihe effect of the anistropy of the carboml grow has been investigated in recent years (30, 31, 32). A model, l2, described by Jaclanan (33), has been cannonly accepted as best describing the anisotropy of the carbonyl grow. Recently (311), a more refined model, ‘11,- has been suggested. Studies on rigid cage ketones and ketals (35) seem to agree well with this new model. Along with the treasurements of the cowling constants of the 16 17 ‘VM ‘Vb aldehydes mentioned above, the chemical shifts of these aldehydes were treasured in order to determine if this new model, 11, was also in agree- ment with the experimental results. RESLLTS A. Coupling Constants Smmarized in Tables I and II are the Vicinal Spin-spin coupling constants between the aldehydic and the can-protons of chloroaoetalde- hyde, branoacetaldehyde , dichloroaoetaldehyde and dibranoacetaldehyde (Table I) ; phenoxyacetaldehyde, methoxyacetaldehyde, methylmercaptoaoe— taldelwde , glycidaldehyde and cyclopropanecarboxaldehyde (Table II). The coupling constants of these aldehydes were observed in 2.5 - 5% solutions in various solvents . All were averages of seven to ten measurements with a precision of t 0.03 c.p.s. , and were always checked for accuracy and consistency against knovm values (36, 37) of aoetalde- hyde; 2.85, 2.88 and 2.90 c.p.s. at 36, O and -30°, respectively. The cowling constants of chloroaoetaldehyde and bromoaoetaldehyde are smaller than that of aoetaldehyde, as were those of monosubstituted alkylacetaldemdes (23). In contrast, however, to the coupling cmstants of the nmoaJJqlaoetaldehydes that were found to be relatively insensitive to solvent dielectric constant, those of chloroaoetaldehyde and bromaoe— taldehyde decreased sharply with increase in the dielectric constant of the nediun. 'Ihe couplirg constant of bmmacetaldehyde is larger than that of chloroaoetaldehyde when carpared in the same solvent. Similar to the monohaloaoetaldehydes , the coupling constants of the dihaloaoetaldedees decrease sharply with increasing dielectric constant of the solvent . For exanple, for dichloroaoetaldehyde and dibmmaoe- taldemde, the coupling constants are respectively, 14.65 and 5.65 c.p.s. in the luv dielectric constant solvent pentane (a ~ 1.8), and 1.10 and 2.17 c.p.s. in the high dielectric cmstant solvent dinetl‘wlsulfoxide -7- -8- TABLE I. VICINAL.SPIN'SPIN COUPLING ODNSTANTSa: IN C.P.S.: OF SOME HALOACETALDEHYDES f-__—JHH’ c.p.s.-——-] Solventb ClCH2CHO Br'CHzCHO 0113(0112)30H3 2.17 trans-decalin 2.13 2.81 CYCIOhexane 2.11 2.82 001,l 2.05 2.78 011013 1.78 2.62 08281-2 1.59 2.48 CH2012 1.58 2.117 CH3OOCH3 1.23 2.19 (11301: 1.09 2.06 (CH3)2NCHO 0.98 1.97 (083)250 0.83 1. 81 112N080 0. 81 1. 80 06116 1.67 2.55 CGHSCHB 1.68 2.58 068501 1.65 2.514 C6H5CN 1.28 2.25 TABLEI'Ccommuan...) r—-—-—JHH, c.p.s.--—-1 Solvent C12CHCHD Br2CHCHO CH3(CH2)3CH3 “.65 5.65 cycldhexane “.50 5.56 Efggsrdecalin H.h0 5.£7 001“ 9.35 5.36 CH013 3.80 “.82 01121:»2 3.30 11.27 CH2C12 3.35 “.25 CHBOOCHB 1.90 3.08 (CH3)2NCHO 1.35 3.17 CH3CN 1.30 2.89 (083)280 1.10 2.17 06H6 3.00 4.16 C5HSCH3 3.10 “.2“ 05115011 2.10 3.22 C6H5N02 2.30 3.35 Neat 2.90 3.90 aValues at 36 1 2°. b2.5 - 5% solutions. -10- TABLE II. VICINAL SPIN-SPIN 00u>LING mNSTANTsa. IN C.P.S.,OF some HETEROACETALDEHYDES AND CYCLOPROPANECARBOKALDEHYDE F JHH: c.p.s. j Solventb 0611500112080 CH 3OCH2CHO . 01133011ch0 CH3(CH2)3CH3 1.52 1.37 3.63 cyclohexane 1.u9 1.3a 3.63 Eggggrdecalin 1.95 1.29 3.63 001, 1.98 1.27 3.63 CHC13 1.16 0.80 3.5u cnzsr2 1.05 0.78 3.51 CH2C12 0.9a 0.77 3.96 CH3OOCH3 0.73 0.76 3.35 (CH3)2NCHO _c_a 0.149 _c_a 0.1l7 3.25 CH3CN 99,0.uuc g§_0.Hl 3.25 (CH3)230 3.18 HQNCHO 3.10 C6H6 1.15 0.99 3.51 0685083 1.22 0.99 3.51 C6H5C1 1.19 0.92 3.u8 05H50N 0.71 0.69 3.33 051151102 0.71 0.59 3.39 Neat 0.73 3.112 TABLE II (CONTINUED -11. O I ) Solvent CHz-CHCHO CHZ-CHCHO CH3(CHZ)3CH3 h.63 qyclohexane 6.26 “.70 Eggggydecalin 6.28 n.7u 001“ 6.22 n.9u CHéBrQ 6.08 5.75 CH2C12 6.06 5.75 CH3OOCH3 5.90 5.80 (CH3)2NCHO 5.85 5.95 CHBCN 5.9“ 6. 00 (CH3)230 5.79 6.15 H2NCHD 5.98 6.20 C6H5 6.u1 5.30 06H5CH3 6.h0 5.25 06H501 6.23 5.80 CGHSCN 6.06 5.73 C6H§NO2 6 . 014 5.60 aValues at 36 1_2°. b2.5 - 5% solutions. cPoor resolution. (2 ~ 115). Furthenmre, the coupling constants are large in low di- electric constant solvents when carpared to the nmdraloacetaldehydes or nuioallqlacetaldehydes. The only other saturated aldehydes whose coupling constants are large are g-g-butyiaeetamehyde (JHH - 6.0 c.p.s.) and cyclopmpanecarboxaldemde (JHH - 5.75 c.p.s.). The former exist mainly in conformation 6 and the latter exist in over 85% of this same conformation (23). Similarly to the monohaloacetaldelwdes, the coupling constant of dibronnacetaldemde is larger than that of dichloroacetalde— kwde when conpared in the same solvent. The coupling constants of phenoxyacetaldehyde and methoxyacetalde- hyde are the smallest Vicinal coupling constants observed for substituted acetaldelwdes. Their dependence on solvent polarity closely parallels that of the Vicinal coupling constants of the haloacetaldehydes . The Vicinal coupling constant of nethylmercaptoacetaldehyde is considerably larger than either phenoxyacetaldehyde or nethoxyacetaldehyde. It is even larger than that of acetaldehyde (2.85 c.p.s.). It decreased only slightly with increasing dielectric constant of the medium. This would incucate that the polarities of the rotaners of methylmsrcaptoacetaldehyde might be quite similar. Since the differences between the electro- negativities of hydrogen and sulfur are small, this conclusion is reason- able and consistent with l and 2 being the minimum energy conformations of this alwwde. In contrast to the small values of the vicinal coupling constants of phenoxyacetaldehyde , nethoxyacetaldehyde , chloroacetaldehyde , branoacetaldekwde and nethylnercaptoacetaldehyde, the corresponding coupling constants of glycidaldehyde and cyc10pr0panecarboxaldehyde are quite large. The Vicinal coupling constant of glycidaldehwde is the largest observed of am aldelwde, other than those of 0,8-msaturated aldehydes, whose values are about 7.7 c.p.s. (38, 39). InTableIIIis showntheeffect oftenperature onthevicinal spin-spin cowling ccnstants of chloroacetaldehvde and branoacetaldemrde . The couplirg constant of chloroacetaldelwde increases with increasing temperature in all solvents, with the rate of increase being greater in the more polar solvents. From a plot of these trends shown in Fig. l, the couplirg constant seem to becans tenperature independent at about 2.5 c.p.s. The couplirg constant of breunacetaldemae increases in N,N-dinetmlformamide (high dielectric constant), but is constant, or tenperature independent, at 2.75 c.p.s. in Wow (low dielectric mtant). This is sham in Fig. 2. 'me effect of tenperature m the vicinal spin-spin couplirg constants of dichloroacetaldemrde and dibrmmcetaldel'wa are sham in Tables IV and V, respectively. For both aldehydes, increase in tenperature causes a decrease in the cowling cantant in low dielectric cmstant solvents and an increase in the couplirg constant in man dielectric constant solvents. In Fig. 3 and h are plots of teuperatuze 1s. coupling constant for dichloreacetaldehyde and dibrmnacetaldehyde, respectively. The lines converge toward a taIperature independent region of about 3.18 c.p.s. for dichloroacetalcblwde and “.5 c.p.s. for dibrunoacetaldemde. are tenperatme dependence of the viciml coupling cmstants of mflmacetaldelwde , Wasataldetwde , nettwlnercaptoacetaldelwde , cyclopropanecarboxaldetwde and glycidaldemde are sham in Table VI. The couplirg cantants of the femur two alderwdes are observed to increase in all solvents with increasim tenperatme. Fran plots of temperature _v§_ vicinal coupling meant in Fig. 5 and 6, respectively, teIperature imependent regime can be extrapolated to be wont 1.5 c.p.s. for nethonry- acetaldelwde and 1.6 c.p.s. for Macetaldemcb. In cmtrast to the -1l&- . com.” pm @3de .OOMH v.0 9.3m: .gflgdom um I m.mw 8a was. 8n 34 54 as $4 21 eezflmes SUN gum RUN gum mute. EHN mw.m caoalgu oemcoeawuoosggm no ..mdé .325. RA 84 BA $5 8.0 5MB om.H main main .230 OIOZNAMZOV 634 84 $4 the 84 em; Swen mH.N NH.N MH.N OH.N mo.N ggfiomo DMN.N ON.N mH.N MH.N mo.N HO.N figalguu OOHH OOOH com 08 OO\.. 08 00m 00m omH oo omHl 00ml ducgflom F 8303388820 no . .ndd LE. . gag nz< gag ...o mtéhmg wzunrfioo zEmIzEm .3253 ME. mo gag me‘mmA—‘mh .2— 32h -15. 80:. age omH OOH om as o: om Q on. can ggmomgggugguoggg JOE o.o m.o o.m -16- Avove owH 03H oma. OOH om cm 0: ON 0 oml. omozflmmov .~ $8.88» .H qum N Egg—Mug «on 5.23.98 mac—#500 mi... no gag Egg .N or. :.o o...” m.a o.N o.m J“ (039.8.) -17. TABLE IV} TEMPERATURE DEPENDENCE OF THE VICINAL SPIN'SPIN COUPLING CONSTANT 0F DICHLOROACETALDEHYDE r th’ c.p.s. 1 Solventa -30° -15° 0° 15° 36° 50° 70° 90° 110° cycldhexane “.52 “.“6 “.3“ “.33 “.28 trans-decalin “.“1 “.38 “.37 “.3“ “.25 “.18 “.06 “.03 3.99 068501 3.16 3.25 3.29 3.32 3.28 3.37 3.38 3.39 3.38 C6H5CH3 2.7a 2.90 2.98 3.08 3.09 3.25 3.26 3.28 3.29 C6H6 2.89 2.93 3.0a 3.19 3.21 C6H5CN 1.69 1.81 1.98 2.09 2.19 2.29 2.35 2.38 (CH3)2NCHO 0.90 1.0u 1.13 1.19 1.28 1.91 1.50 1.60 1.70 a2.5 - 5% solutions. -18- mmsm mwmmmemmmmaosmewaMmesmnmwum CONSTANT 0F DIBROMDACETALDEHYDE r J HII’ c.p.s. 1 Solventa —30° 0° 15° 36° 60° 70° 100° Eggggrdecalin 5.99 5.73 5.u3 5.32 5.15 CI‘lCl3 5.26 “.98 “.95 “.77 06H5C1 “.78 1.55 “.“6 n.77 u.u9 (“.“8)b 0585083 “.26 “.38 “.“1 “.“8 “.“8 CHZBrz 11.02 11.11 11.11: (CH3)2NCH0 3.15 3.35 3.51 3.59 CH3CN 2.56 2.69 2.9“ 3.09 32.5 - 5% soluticu. bValue at 130° . -19.. g [I]? c \ :1 E ’I'I|\\‘ Eggsé“ 3% g //"\\§%§l8‘°§§°§§ a // II \ 4.6.6526; § II III \ \ :5 IU/ l I \ \\ g // l:| \ \ g // ll \ \ l I III \ ‘\ E I | \ \ 25 I, ll \ I] ll' \ \ /I ll| / I 35 E n “5 “.0 35 5 T(°C) 0 20 “O 60 80 100 120 1‘40 —20 . J' 0. -20- AOOOB OOH OeH OOH OOH Om OO O: ON O ON- zommo .s omozmHmmuv .O Nnmmmo .m m8»%O.= \\ HommOO .m mHomo .m H cflmooelegu . .1: ....1:AH~MM~HVAHH////xu ,1, / .1. H mn>xuanu oz< m§~u>qo . Bax—3585;“; s 3:555; egkwu<>xgyt no 9.72.578 "dz—:50 27515.3 .225; “1:. n5 gun—B SEE .~> Hams. -22. .oow um msz>m .oomH um mSHm>o .ooml um mSHm>o .oOO es .825 .ncoHpsHom an. u .99 mm.z mm.= om.: 05.: :w.: mm.= HH.m :H.m mm.m :HHmomelncmnp oezcmeamsonnmomcmoonaoacad no ..m.a.o .::w .66. Ohm SA. 86. OOA. 86 NH.O mH.O ONO omOstmeeO NHO ONO OHO Rho Owe OO.O 3.6 H9O 5268.88» 856388.48 no ..ndd tan mO.m an.m smO.m OH.m ON.m mm.m 982mm 82 was mm.m Om.m Oi detainee as m...m e; Hem 82 Homes“. m...m om.m Ram 8.». {mm remix. 2a OO.m $5 tum NO.m 5268.88» OOHH OOOH 00m CON. 00m 00m omH OO omHl 00m I HCQHCW P 63833688888282 so .666 .EOIIIL A . . OHOzCzBV H> WES -23. A00vB ova 03H ONH 00H ow cm O: on 0 cm! omozflmmov ... memmOO .m mn>zwn4<fimu<>xcthz mom hz JS’ where Jt is the fledgling cantantanngistheflchg, fieobservedaveragevicinalcomlim cantantwouldbe tenperature independent if 183, 1,80 andeere iso- energetic. Accmdirgly, the vicinal cowling ccnstant would decreme 0 0 0 H H H I l / I R’ / H H a H 8a R .8... $88 .12 wimmmeuingteupemtmeiflfig(w£a)wasmmestablemm$9, and would increasewith increuing temperature iflglgwas less stdale than a. Fran the turpentine dependence of the spin—spin comlirg constants, the following canbe deduced. (a) Chloroaeetaldemde: inboth low and my: dielectric cmstmt solvents, the most stable retailer is 12, 1.3., the mewith (:hlorhie _c_igto the carboml. (b) Brunoacetalchl'wde: in maimelectr'ic mtantsolvents, themst stablerotmts‘lgandin low dielectric cantant solvents, such as m-decalin, all three nature are iseenergetic. (c) Methowacetaldehyde and phenowaeetalde- -3o- -31.. hyde: in both high and low dielectric constant solvents, 12 is the more stable retailer. (d) Methylnercaptoacetaldehyde: in both big: and low dielectric constant solvents, 12 rather than 12 is the more stable rotamer. Rotauer populations and free energy differences , AG° , between individual retainers were calculated fran equations 1 and 2, respectively, where p is the fractional population of~ll8 ($83. + £82) and (1 - p) that of 12. 'me enthalpy differences, AH°, between 18 and 12 were Jobgd - p(Jt + J3)”- + (1 - I>)Jg (1) Obtained fran plots of log Keq 1s l/T, where Keq is given by equation 3. Keq - 2(1 - p)/p ‘ (3) For the wove calculations, the J1; and J8 coupling constants for each aldemIde mt be lawn. Limits and estinates for these parameters have been made frun the experimental data and equation “, which relates the experinental coupling ccnstant to Jt and JS’ either when the three rotaners 1182, 1’82 and 19 are equally populated, or at free rotation about the carbon-carbon bond (usually at very high temperatures). For brmoacetaldelvde, the ten'perature independent value (J obsd in equation ‘0 is 2.75 c.p.s. in _t_ra_n_§_-decalin (Fig. 2). The analogous value for chloroacetaldelwde is greater than 2.23 c.p.s. (highest value within increasirg trend in Fig. l) and is estimated by extrapolation to be -32- about 2.5 c.p.s. These values are certainly reasonable when compared to the corresponding cotplirg constant of acetaldehyde (2.85 c.p.s.). Having established the .10de values for these two aldehydes misting Jt and J8 according to equation “, we can set limits for Jt and J 8' For exanple, the lowest coupling constant of chloroacetaldehyde is 0.6 c.p.s. (in acetonitrile at -30°). In absolute magnitude, therefore, Jg Imst be equal to or smaller than 0.6 c.p.s. for chloro- a«Getaldem'de. If Jt and J8 have the same sign, then, fran equation “, J8 _<_ 0.6 c.p.s. and Jt 3_ 6.3 c.p.s.; if J1; and J8 have opposite signs, J8 1 0.6 c.p.s. and Jt 1 8.7 c.p.s. From the analogous coupling con— stants of acetaldehyde (23), J8 =- 0.5 c.p.s. and Jt = 7.6 c.p.s., a reasmwle estimte of the cowling constants of chloroacetaldehyde would be: J8 =- 0.3 c.p.s. and Jt :- 6.9 c.p.s., both havirg the same s13). For brmnacetaldehyde, the smallest coupling constant observed is 1.“8 c.p.s. (in N,N-dinetmrlfomannde at -30°). 'ihus, J 1 l.“8 c.p.s. 8 and Jt _>_ 5.28 c.p.s., if the sign are the same; and J <__ 1.“8 c.p.s. and 8 Jt _>_ 11.2 c.p.s., if the sign are apposite. Since the couplirg con- Btants of acetalderwde and brmnacetaldemde satisfying equation “ are 2.85 and 2.75-c.p.s., respectively, reasonwle estimates for J8 and Jt of brumacetaldemde are 0.“ c.p.s. and 7.5 c.p.s., respectively. 'Jhe values (Jobsd) which satisfy equation “ for pherIoJIyacetaldem'de, metlmacetalchlwde and netlwlnercaptoacetaldehyde are about 1.6, 1.5 and 3.0 c.p.s., respectively. These values were obtained fran extrap- olations in Figs.5, 6 and 7. With these Jobsd values and equation “, limits and or estimtes can be set for at and J as follows: Since the 8 lowest experinentally Treasured vicinal coupling constant of plenoxy- -33- acetaldehyde is about 0.5 c.p.s., Jg must be equal to or smaller than 0.5 c.p.s. By using equation “ as before: if Jt and J8 have the same sign, then J 5 0.5 c.p.s. and Jt _>_ 3.8 c.p.s. and if they have the 8 opposite sign, Jg _<_ 0.5 c.p.s. and Jt _>_ 5.8 c.p.s. For nethoxyacetalde- mde, the snellest observed coupling constant is about 0.“ c.p.s. Again, Jg must be equal to or sneller than 0.“ c.p.s. Thus, if at and J8 have the sane sign, JB 3 0.“ c.p.s. and at z 3.7 c.p.s.; if J8 and Jt have the opposite sign, J 5 0.“ c.p.s. and Jt _>_ 5.3 c.p.s. 8 The coupling constant of nethylnercaptoacetaldehyde did not vary extensively with solvent polarity and therefore, it is difficult to set reasonable limits for Jt and J g' However, due to the similarity of the J values of acetaldemIde (2.85 c.p.s.), bromoacetaldehyde (2.75 obsd c.p.s.) and mfluyhmmaptoacetalee (3.0 c.p.s.), reasonable values for nethylnercaptoacetaldelwde of 0.6 and 7.8 c.p.s. can be estimated for J8 and Jt’ respectively. By using the previous values for J and JS’ the effects of solvent 1'. polarity on the relative population of the rotamers for chloroacetaldehyde and branoacetaldemde were calculated and are given in Table VIII. The values in columns B and E were calculated using the best estimtes , those in A and D were calculated from coupling constants with the same sign, and those in C and F were calculated from coupling constants of opposite signs. As had been noted previously from the coupling constants, the pOpulation of the more polar rotaner increases as the dielectric con- stant of the solvent increases . This sane effect is seen for chloro- acetaldehyde and bromacetaldemde in Table TX in terms of the free energy differences , AG° , that were calculated from equation 2 . However, this sane trend exists to a smaller extent for bremoacetaldehyde. For exanple, -31- TABLE VIII. SOLVENT IEPENDENCE OF THE RELATIVE ROTAPER POPLLATICNSa 0F CHLOROACETALDEHYDE.AND BRDWDNCETALDEHYDE 0101120110 8101120110 7: 19 7 19 'VV «A. Solvent Ah so Cd De sf FE CH3(CH2)3CH3 us “0 “0 £r_ans_;—decalin “6 Lu “1 30 32 32 cyclohexane ' “7 “5 “2 29 32 32 0011; “9 “7 “3 I 32 33 33 011013 59 55 “9 “0 37 35 CHZBr-z 65 61 53 “7 “1 38 0112012 66 61 53 “8 “2 38 0113000113 78 72 61 63 50 “2 CH3CN 83 76 6“ 69 53 ““ (CH3) ZNCHO 87 79 66 7“ 56 “6 (0113530 92 8“ 69 83 60 “8 HZNCHO 93 85 70 83 61 “8 C6H6 63 58 51 ““ 39 36 0611501 63 59 52 u“ “0 37 CGHSCN 76 70 60 59 “8 “1 aAll values calculated for 36°. bCoupling constants used: J =6.3 and J = 0. 6 (sane 8181). °Coupling constants used: J = 06.9 and =5. 3 (sane sign) dCoupling constants used: Jt = 8. 7and = 0. 6 (gpposite sign). eCouplirg constants used: J = 5. 28 andJ g 8 J5. “8 (sane sign). fCoupling constants used. Jt = .5 andJ =.0 “ (sane sign). 8Coupling constants used: Jt= 11.21 andJg =5. “8 (opposite sign). TABLE IX. 382-135va M OF 1145 FREE ENERGY migrant-ea. 36". mm 01082080 BrCHZCHO :2; 182W? for 0139133 Solvent A B C D E F c113(0112)3CH3 -310 -170 -170 _t__rar_:_s_-deca11n -325 -300 -200 +100 +110 +140 cycld'exare -350 -310 -225 +100 +“0 +“0 CClu -uoo -350 -250 +uo ~o ~o 011013 -650 -560 -1700 -170 -100 -50 0112er -800 -700 -500 -360 -200 -100 0112012 ~835 -710 -500 -375 -230 -100 CH3OOCH3 -1200 -1000 -700 -800 4430 -230 CH3CN -1“oo -1100 -780 -950 -soo -275 (CH3)2NCHO -16oo -1250 -830 -1100 -570 -300 (013)280 -1900 41:50 -920 -1uoo -680 -1400 112mm -2000 -1500 —9uo -11400 -700 -1400 06115 -750 -680 -1150 -300 -150 -80 c685c1 -750 -650 4:75 -300 -180 -100 C6350“ -1130 -950 -675 —650 -380 -200 aCalculated for 36° usirg the correspond}; data in Table VIII. -35- whereas in the least polar solvents 18a is slightly favored, -in the ’VW most polar solvent formamide, $9 is favored by about 700 cal/mole. In Table X are sumerized the enthalpy differences, AH°, between the rotamers of the two haloacetaldehydes . They were calculated from reasonably linear plots of log Keq _v_s_ M. It can be noticed in Table IX and X that while the AG" and AH° values are about equal in solvents of low dielectric constant, AH° is appreciably more negative than 110° in solvents of high dielectric constant. For exanple, 116° and AH° for lays-4:522 of chloroacetaldehvde are both -300 cal/mole in Lrar_rs-decalin, whereas in acetonitrile AG° is -1000 cal/mole and AH" -2500 cal/mole. This sane trend is observed for branoacetaldehyde. In Tables XI, HI and XIII, respectively, are sumnarized the rotaner pOpulations , free energies and enthalpy differences of phenoxyacetal- demyde, methoxyacetaldehyde and methylnercaptoacetalderwde. These values were determined by using the respective Jt and J5 cowling con- stants from above in conjunction with equations 1, 2 and 3. The values given in colums A, A' and C were calculated by using coupling constants Of the sane sign. From these results, the most stable rotamer of phenoxy- acetaldehyde and nethoxyacetaldehyde is 2.2. As was the case for chloro- acetaldemrde and bramacetaldehyde, the stability of this rotaner in- creases with increasing dielectric coretant of the solvent. However, l2 is the more stable rotamer of netmlnercaptoacetaldehvde, but decreases in stability as the dielectric constant of the solvent increases . The stability of $2 of phenowacetaldehyde, methoxyacetaldelwde, chloro- acetaldemrde and bruncacetaldetwde is greater in the aranatic solvents benzene and toluene, than would be expected from their dielectric con- stants (e of 2.3 and 2.“ for benzene and toluene, respectively). This TABLE x. EN'IHALPY DIFFERENCESa. AH“. BETWEEN ROTNVERS 0F WNW AM) WW 010112080 BrCHZCHO AH", cal/mole, AH", cal/mole, for 18:19 for 18:19 M ‘Vb m M Solvent A B C D E F tmm-tbcalin -“00 -300 -250 0 O 0 cyclohexane -500 -“00 -350 C6HSCI -1300 ~900 -700 , (CHBENCHD -2700 -2100 -1000 -38SO -1500 -700 CH3CN -2900 -2500 -1200 amesa values were obtained by plotting the equilibrium constants calculated fran the rotaner papulations in Table VIII _vg l/T. -38- TABLE XI. SOLVENT DEPENIENCE OF ROTMER POPULATIONSa OF PHEIWYACETALDEHYDE, WWW AND PETHYU‘ERCAPTOACETAIDEHYE c6H50cazcuo CH3OCHzCHO CHBSCHZCHO 2 19 z 19 Z 19 M Solvent AP 30 Avd Bye Cf cycldhexane “0 37 “8 39 16 trans-decalin “2 38 51 “1 16 CClu “1 37 52 “l 16 C5013 50 “7 78 58 18 CflzBrQ 67 52 79 59 19 CH2012 73 5“ 80 50 21 013000113 86 61 80 59 214 (CH3)2NCHO >99 69 96 7o 26 GH3CN >99 70 >99 72 26 82mm 31 0636 61 “8 67 51 19 CGHSCHB 56 “5 67 51 19 C6H5C1 61 “8 71 S“ 20 C6HSCN 87 62 8“ 62 2“ Neat 89 62 am values calculated for 36°. bCoupling constants used: Jt - 3.8 and J = 0.5 (sane sign). cCoqalirg mtants used: J = 5. (fippoeite sign). dCoupling constantsused: J = 3. and J = 0. sign). eCoupling constants used Jt = 5.3 ang J3 = 0.“ ( 1te sign). Couplirg cantata used: Jt = 7.8 and J8 = 0.6 (same 8191 . -39- TABLE XII. SG..VENT IEPEMIPCE OF 'n-E FREE Ef‘ERGY DIFFERENCEa, 06°: BETWEEN ROTNMERS OF PHENOXYACETALDEHYDE: METHOXYACETALDEHYDE AND!METHYLMERCAPWOACETALDEHYDE C6H5°CH2CH0 “30°55” CH35"“2CH0 fifié‘flia “Sinai-8115 89.233.223.156 w» 'w w. M. w» m. Solvent A B A' B' c CH3(CHz)3CH3 -130 -70 -380 -120 +600 cycldrexane -180 -90 -380 -150 +600 Egg-deem -2“0 -130 4:80 -200 +600 001,, -200 -100 -1170 -200 +600 011013 -700 -360 -1200 -620 +500 08213::2 «.860 4:70 -1200 -630 +850 (mam2 -960 -530 -1200 -650 +u00 083000113 -1500 -700 -1300 -660 +300 (083)2NCHO -900 .2800 .930 +200 (213014 -950 -3600 -1000 +200 11sz +80 06116 -700 -370 -870 -“60 +500 c6H5d13 -600 -300 -870 .1160 +500 0611501 -700 -380 -1000 -5“0 +1100 0635‘“ -1600 -700 '-lhoo -700 +300 06H5N02 -1600 -700 -1700 -800 +330 Neat -1700 -700 aCalculatedfor36°nunmecorrespondingdatainTableXL -40- TABLE x111. ENTHALPY DIFFERENCESayAH°. BETWEEN ROWAMERS 0F PHENOXY- ACETALDEHYDE, METHGXYACETALDEHYDE ANDtMETHNLMERCAPFO' ACETALIEHYIE 05H5OCH2CI'D CH3OCHZCH) CH3SCI-120HO AH", cal/mole, AH°, cal/mole, AH", cal/mole, for 18 """*‘~—— 12 for 18 ——*‘-— 19 for 18 —"-.— 19 ‘VV 'Vh 'VM M 4v» Solvent A B A' B' C cyclohexane -l“00 -600 tm—decaIin -1200 -500 -l“00 -600 +1000 Cal-150113 -l700 -700 —2200 -800 +900 C5H5€1 -l700 -900 +900 (CH3)2NCHD -1300 -2600 -1200 +500 HZNCID +300 Neat -3600 -1200 8These values were obtained by plotting the equilibrium constants calculated fran the rotaner pmulations in Table XI 1:; VP. -81- is attributed to a type of solute-solvent steriospecific association which will be discussed later. Large discrepancies are particularly obvious between the free energy (Table XII) and enthalpy values (Table XIII) of phenoxyacetal- dem'de and nethoxyaoetaldehyde in the low dielectric constant solvents. Although entrOpy differences may be partly responsible for these dis- crepancies, the choice of J and J 8 probably constitutes the major t source of error. These paraneters were determined on the basis that J satisfying equation “ is 1.6 c.p.s. for phenoxyacetaldehyde and obsd 1.5 c.p.s. for methoxyacetaldehyde. These values are considerably lower than those for acetaldel'wde (2.85 c.p.s.), brcmoacetaldehyde (2.75 c.p.s.) and chloroacetalderwde (2.5 c.p.s.). From electrmegativity coreideratims (“0, “1), they ougrt to be between 2.0 and 2.5 c.p.s. Using values of 2.0 and 2.5 c.p.s. for J in equation “, rotamer obsd papulations , fiee energy and enthalpy differences are sumerized in Tables )CN and XV. The much better correspondence between free energy and enthalpy differences suggest that these values are more reliable and that the correct choice of J and J8 may be critical. 1: a) Consideration of a Wofold Barrier for Chloroacetaldehyge; and Brennacetaldetydé. The results for chloroacetaldehyde and bromoacetaldemtde can be interpreted in terms of a threefold barrier to rotation about the spamsp3 carbon-carbon bond. However, the question of a twofold barrier to rotation must also be examined. Among all structurally relevant mmohalocalpomds studied and reported today, only fluoroacetyl fluoride (28) has been found to have a twofold barrier to rotation about the spa-3p3 carbon—carbon bond. However, if it is assuned that «1A9 and 30 are the equilibriun conformations (twofold -ug- TABLE XIV. ROTANER POPULATImSa OF PHENOXYACETALEHYIE IND WACETAUIEHYDE 0511500820110 CH3OCH2CHO Z 19 % l9 ’VM ‘VV Solvent AP 89 A'd B'e CH3 (CH2 ) 3CH3 55 “6 S3 1‘7 cyclohexane 56 “7 55 “8 trans-decalin 58 “8 56 “9 Jobsd a 1/'3(Jt + 2J3) = 2.5 c.p.s. cn3(o12)3ca3 66 55 6“ 56 cyclohexane 67 56 65 57 trans-decalin. 68 57 67 57 aAll values calculated for 36°. bCouplirg constants used: Jt . 5. 0 =O.ob5(sanesign)andJ =2.0;J=6..5,J=05(sanesign) afid JobsdJ . 2. 5. cCoupling °b§8rgtemts usedé J =- 7. 0, JE g)=agd5J(0pposite sign = 2. 0; J = 5, J = 0 5 JOpposite sign aggupggng mtantg used: th= 5.2, Jg = 0. “ (same Sign) 333d Jobs = 2. 0; Jt = 6. 7, J = 0. “ (same sign) and J = 2. 5. eCoupling cons ants used-. 6.88J = 0. “ (opposite si mgband Jobsd = 2. 0; J =8.H.I.3,Jg=0“ (Oppositgsign) andJobsd=2..5 -143- TABLE xv. FREE ENERGY DIFFERENCEamcfi BETWEEN Romans OF PHENOXY- ACETAwEMJE AND mmvacsmmeE 0585001120110 0H300HZCH0 AG° , cal/mole, AG° , cal/mole, for 18a"-.==l9 for 18a: 19 NM 'Vh W M Solvent A B A' B' Jobsd = l/3(Jt + 2Jg) = 2.0 c.p.s. CH3(CH2)3CH3 -5“0 —320 -510 ~350 cyclohexare -570 ~350 -5“0 -370 tram-decalin -620 -380 -580 -“00 Jamd - 1/3 (Jt + 2.13) .. 2.5 c.p.s. CH3(CH2)3CH3 -830 -550 -780 -570 cyclohexane -880 -570 -810 -590 trans-decalin -900 -580 -850 —610 AH°, cal/mole, for 18 :19 ‘VV ’Vh cyclohexane -800 -500 (11de - 2.0) cycldrexane -300 -500 (J0bad - 2.5) trans-decalin -600 -“00 -600 -“00 tram-decalin -500 -300 -500 -300 (01de . 205) “Calculated for 36° from the data in Table XIV. -““- barrier) of the two mndraloacetaldehydes , then the relevant vicinal Spin-spin coupling constants would be Jg (60°) from 12 and J 1200 from ‘22.” For a twofold barrier to rotation, equation “ becanes equation 0 H ~ L \ H 20 and 'W “'. Since Jg must be equal to or smaller than 0.6 c.p.s. for chloro- 2.5 c.p.s. = l/2(Jg + leoc) (“') acetaldehyde, Jl20° met be equal to or larger than “.“ c.p.s. These results are certainly unreasonable, as J8 and leoo are eXpected to be of similar magnitude (“2). Analogous treatnent for branoacetaldehyde leads to JS 5_ 1.5 c.p.s. and Jlgoo _>_ “.0 c.p.s., which again seems to be measurable. b) Relative Stabilities of the Mmosubstituted Acetaldemges. The relative stabilities of the monosubstituted acetaldehydes (% and £9), as a function of R as Judged Iran the present and previous results (23), are given in rough order belav. These results are valid only in solvents R - CH3>CHZCH3~005H5«00H3>CH(CH3)2>05115w1>13r>c(CH3)3>SCH3 Increased stability: tDecreased stability of lg of ‘12 of la: dielectric constant, such as carbon tetrachloride and saturated 4+5. mdrocarbms. In solvents of high dielectric constant, the nethoxy, phenoxy, drlor'oandbrmrogretxpsmvearreadofthenetmlgr'oupinthe above given order. For the groups preceeding branire in the above order, AH° for £82 $2.2 is negative and for those groups following bromine, it is positive. The position of the more polarizable netmlmercapto group with respect to that of the less polarizable nethozqy, as well as that of bruuine with respect to chlorine, indicates that dipole-induced dipole interactions play minor roles in determining the relative stabili— ties of 128.1 and $9. Nonbonded repulsions are partly responsible for the position of the bulky _t-butyl and nethylnercapto groups. However, their relative positions reinforce the conclusion (23) that nmbonded repulsims are not the overriding factor controlling rotaner stabilities. What this factor is, still renains to be determined. II. Dihaloacetaldel'ydes The data for dichloroacetalderyde and dibremoacetaldehyde in Tables I, IV and V may be interpreted in terms of a threefold barrier to rotation about the sp2-3p3 carbm-carbon single bond with 2,1 and 32 being the equilibrium conformations of these carpomds. As before, by assuming 0 O H L R L L H H H 1’ I / I l R R H H R 22b R 21 223 'VV W W Jt’JS’ the average vicinal coupling constant would be tenperature in- . 1 dependent if 51, 23a and 322 were isoenergetic If the average vic nal coupling constant increases with increasing tenperature, then 522 (or ab) is more stable than 2’1, and if it decreases with increasing tenperature, -u5_ then 2,1 is the more stable rotamer. Fran Tables IV and V, it can be seen that in the non-polar solvents (m-decalin), 2,1 is the more stable rotaner, whereas in polar solvents (acetonitrile), 22 is the more stable rotaner for both dichloroacetaldetwde and dibroneacetaldehyde. Rotaner papulations and free energr differences, 00°, between individual rotamers were calculated, respectively, from equations 5 and 6, where p is the fractional population of 2,1 and (1 - p) that of +1de I th + (l - D)Jg (5) AG0 I -RI‘ln V2011: - JCbSd)/(Jd38d - J8) (6) 3?: + 251’. As before, the enthalpy differences, AH° , were calculated franplots of logKeqygl/I‘. The values ofKeqwere determined from equation 7. Determination of these quantities requires knowledge of Keq - (l - p)/2p (7) Jt and J , as well as equation “. Equation “ relates Jt and J8 either when 2%, 2,22 and 322 are equally populated, or at a state of free rotation about the carbon-carbon single bond (usually at high tenperature). The emerinental coupling constant (Jobsd) satistying equation “ is 3.“ c.p.s. for dichloroacetaldehyde and “.5 c.p.s. for dibrunoacetaldehyde as shown in Figs. 3 and “, respectively. Since the lowest experimentally neasured coupling comtant of dichloroacetaldehyde is 0.9 c.p.s. (Table IV), Jg trust be equal to or smaller than 0.9 c.p.s. From equation “, if Jt and J8 have the sane sign, Jg<_0.9 c.p.s. and Jt Z. 8.“ c.p.s.; if they have 0lipasitze signs, J8 :03 c.p.s. and Jt 1 12.0 c.p.s. Hadever, a reason- -87- able estimate would probably be Jg =- 0.5 c.p.s.. From this and equation “, Jt = 9.1 c.p.s. (sane sign). For dibranoacetaldehyde, the smallest observed coupling constant is 2.17 c.p.s. If Jt and J8 have the sane sign, Jg _<_ 2.17 c.p.s. and Jt _>_ 9.16 c.p.s.; if the signs are opposite, then J 8 ~ 1.6 c.p.s.) of Jt and J8 1 2.17 c.p.s. and Jt 1 17.8“ c.p.s. A reasonable set (estimated Jg would be 10.3 and 1.6 c.p.s., respectively, (sane sign). The effect of solvent polarity on the relative pepulations of the rotaners is shown in Table XVI. These were calculated from equation 5. The values in columns B and B were calculated by using the best estinetes 0f the Jt and J8 coupling constants. Those in A arnd D were calculated fran the coupling constants with the sane sign and those in columns C and F fran coupling constants with Opposite signs. Inspection of these results are not strongly affected by the choice of the coupling constants , Jt and J8. However, the stability of the rotaners is strongly dependent an solvent polarity. The wire polar rotaner 2,216: (or $22.) is more stable in polar solvents, whereas the nonpolar rotaner, $1, is more stable in the mrpolar solvents. In Table XVII are smmardzed the free energ' differences, 00°, between rotaners 2’1 and $22 (or 222) calculated from equation 6. In low dielectric constant solvents, c < 5, ,the free energ' differences are positive for both aldehydes ($1, the less polar rotamer is more stable than 2’22, the more polar retaner). In solvents of dielectric constant geater than 9, the AG°'8 are negative (23% is more stable than 21). The cross-over appears to occur at a dielectric constant of about 6. For exanple, the AG° values are positive in chloroform (e of “.8) and negative in metlwlene bromide (8 of 7.“). The values in armatic solvents are anomalous, as they were -“8— TABLE XVI. SOLVENT DEPENDENCE OF THE RELATIVE ROTAMER POPULATIONSa OF DICHLOROACETALDEHYDE AND DIBROMOACETALDEHYDE 0120:1030 BrZCHCHO z 21 z 21 Solvent ab 39 cd 08 sf 98 CH3(CH2)3CH3 50 “8 “3 50 “7 39 cyclohexane “8 “7 “2 “9 “6 39 Egggsydecalin “7 “5 “1 “7 us 38 001” - “6 “5 “l “6 “3 38 08013 39 38 36 38 37 35 “2&2 32 33 33 30 31 32 0112012 33 33 33 3o 30 32 CH3000H3 13 16 22 13 17 26 (CH3)2NCHO 6 10 18 19 21 28 . CchN 5 9 17 10 15 25 (083)280 3 7 16 . 1 7 22 0585 28 29 3o 29 29 32 C685.083. 29 3o 31 30 3o 32 C6HSCN 16 19 23 15 19 27 CGH'SNO2 19 21 25 17 20 28 Neat 27 27 29 25 26 30 8All values calculated for 36°. bCoupling constants used: Jt = 8.“ and J = 0.9 (sane 8191). cCoupling constants used: Jt = 9.1 and J = 0.5 (game sign). dCoupling constants used: J = 12.0 and.J = 0.9 prposite sisn). eCoupling constants used: J = 9.§6 and.J = 2. (same sign). fConpling constants used: Jt - 10.3 8nd J - 1.6 (§ame sign). SCoupling constants used: Jt - 17.85 and J8 - 2.173(Oppoaite sign). -“9- TABLE xvn. SOLVENT DEPENDENCE OF THE FREE ENERGY 01m“. 116.: BETWEEN ROTNERS W WWW AND DIBROO’ ACETATE-ME 0120110110 BrZCI-ICHD AG°, cal/mole, AG", caMnole. for’215:===££a fer’£1=:===£s: Solvent A. B c D E F CH3(CHZ)3CH3 +“30 +380 +250 +“20 +3110 +150 cyclohexare +380 +350 +220 +390 +320 +1710 m-decalin +3110 +300 +210 +360 +290 +130 001,. +330 +300 +200 +320 +260 +110 03013 um +120 +80 +130 +100 +110 CHZBrnz -“0 ~15 -20 -90 -75 ~30 (112012 -17 -1“ -11 -100 -80 Am 03300CH3 -730 -590 -360 -7“0 -550 -210 (013)2Ndno -1260 -930 -500 -“80 -370 -150 0130M -13“0 -1000 -5“0 .900 —650 -280 (c113)280 -1800 -1200 -620 -1200 -360 061-16 ~16O -120 -90 ~l“0 -120 -50 C6H50H3 -l2O -100 -60 -100 -90 -“0 CGHSCN -600 -3“0 -300 -6“O -“80 -190 °6"5"°2 -“80 -390 -250 -550 4:20 -170 Most -200 ~190 -110 -260 -210 -90 a'n'nese values were calculated from the corresponding data inTable XVI. -50- for the mondnaloacetaldehydes , methoxyacetaldehyde and phenoxyacetal— dehyde, and will be discussed in a separate section. In Table XVIII are smmarized the enthalpy differences between rotamers 2’1 and 22 calculated from reasonably linear plots of log Keq y; M. The enthalpy and free energy differences are about the sane in solvents of low dielectric constant , but the enthalpy differences are appreciably more negative than the free energ' differences in a median of high dielectric comtant. The effect is geater for dichloro- acetaldekyde than for dibranoacetaldehyde. This sane effect is also found in the monohaloacetaldehvdes . a) Consideration of a Twofold Barrier. In view of the finding that the rotational isomerism about the 8p2-8p3 carbon-carbon bond of ethyl dihaloacetates is best described in terms of a twofold barrier to rotation (8) , the experinental data for the dihaloacetaldemdes will be examired in terns of ndnimum energy conformations 2’1 and 2,3. The R O R\‘>/’LH H 6% relevant vicinal spin-spin coupling coretants for a twofold barrier to rotation would be Jt and Jo, where Jt is the trans coupling from 2,1 and Jc is the gig coupling from 22. Equation “ now becomes equation 8. As stated earlier, the values of Jobsd for dichloroacetaldemde and Jua’d - l/2(Jt + JO) (8) -51- TABLE XVIII. ENTHALPY DIFFERENCESa: AH... BETWEEN ROTN‘ERS OF DICHLUIO' ACETAUIHYII AND DIWALDE'ME C12CHCHO BrZCHCHO AHO, cal/mole, AH°, cal/mole, roi- 21..-==:22 for 21:: 22 ‘Vb 'V’b W M Solvent A B C D E F' cycld‘em +300 +300 +200 m'decan“ +300 +300 +200 +600 +500 +230 CHCl3 +600 +500 +230 C6H5Cl 0 0 0 0 0 0 (332312 -200 -100 -100 C6H6 —600 -1450 4:00 C6’15‘3‘3 —600 -500 Am -250 -200 -100 09‘50" . -1200 -1000 -600 (“3’2"“) -2800 -1800 -600 —700 -500 -200 CH3CN -1500 -800 -300 aDnese valnes were obtained by plotting the equilibrium constants calculated fran the rotaner papulatims in Table XVI is; 1/1'. -52- dibroneacetaldemde were found to be 3.“ and “.5 c.p.s., respectively. When these values are compared to those of acetaldelwde (2.85 c.p.s.), Chloroacetaldehyde (2.5 c.p.s.) and branoacetaldehyde (2.75 c.p.s.), which apply to equatiorn “, they are found to be larger than substituent Electron'egativity effects on vicinal proton-proton coupling would have Predicted (“0, “1). For exanple, sustitution of one rydrogen by chlorine or brandne reduces the average vicinal coupling constant of etharne frail 8.0 c.p.s. (“3) to 6.5 c.p.s. for chloroethane and 6.6 c.p.s. for branoethane (“0). Substitution of two hydrogere on the sane carbon by two chlorlnes or bronrines further reduces the coupling to 6.1 c.p.s. (1,1-dichloroethane) and 6.2 c.p.s. (1,1-dibromoethane) (“1). By using this argument of electronegativity, it would be expected that the J obsd values of dichloroacetaldehyde and dibranoacetaldehyde would be sneller than 2.5 and 2.75 c.p.s., respectively, for a threefold barrier to rotation. The experimentally observed higher values of 3.“ and “.5 c.p.s. would seem to contradict a threefold barrier to rotation and would be more aligned with equation 8, i._e_._, a twofold barrier to rotation about the carbon-carbon bond. Havever, these predictions based on substituent electronegativity may be quite false. For exanple, whereas the average Vicinal coupling constant of ethanol (twdroxyl substituted for hydrogen) is about 6.6 c.p.s. and of propionic acid (carboxyl substituted for hydrogen) is 7.“ c.p.s., that of lactic acid (hydroxyl and carboxyl Substituted on the same carbon ) is 7.3 c.p.s., and not about 6.0 c.p.s. as would have been predicted train substituent electronegativity consid- eratiore. Let the assunption of a twofold barrier still be valid. Since the smallest experinentally observed coupling constant of dichloroacetaldehyde ~53- is 0.9 c.p.s. and Jt > Jc, Jc must be equal to or smaller than this value, From this value for Jc and equation 8, Jt must be equal to or greater than 5.9 c.p.s., if Jc and Jt have the sane sign; and equal to or geater than 7.7 c.p.s., if they have Opposite signs. Sinnilar treatment of the data for dibromoacetaldemde gives: Jc 1 2.17 c.p.s. and Jt l 6.83 c.p.s., if they have the sane sign; and JO 1 2.17 c.p.s. and Jt 1 11.17 c.p.s., if they have Opposite signs. The question may “Pi now be asked whether these values for Jt and J c are reasonable in re- lation to ore another. From valence-bond theory (“2), the contact inter- action term describing the dihedral angle dependence of vicinal proton- j proton coupling is approximated by equation 9. The relative nnagnitudes ,_ JHH - A + 80056 + CcosZ¢ (9) of Jt and Jc depends on the values of A, B and C. For ethane (both carborns sp3 and a carbon-carbon distance of 1.51:3), A - “.22, B . —o.5 and c = “-5 c.p.s., the treatnnent predicts Jt = 9.22 c.p.s. and Jc = 8.22 c.p.s. For ethylene (both carbons sp2 and a carbon-carbon distance of 1.35 A), it predicts Jt = 11.9 c.p.s. and Jc = 6.1 c.p.s. Experimentally deter- mined Jt and J c values of ethylenic conpounds agree fairly well, if not in absolute value, at least in the relative magnitudes of the two coup- lirg constants with the predicted values. There are no experimental Jt and Jc values for system with ore carbon atan 8p2 hybridized and the other sp3 hybridized to which our calculated values may be conpared. Sone values are available for systems with both carban atone sp2 hybrid- ized, where the carbon-carbon length is between that of ethane and ethylene. The Jt of 1,3-butadiene (““) and Jc (single bond of 1,3- -511. cyclohexadiene) (“5) are 10.“1 and 5.1“ c.p.s., respectively. For a ,B-unsaturated aldehydes (malondialdehyde and acetylacetaldehyde) , the analogous coupling constants for 2’“ and 22 have been estinnated by n.m.r. to be about 7.7 and 2.8 c.p.s., respectively (39). If it could be con- \___/H \____/H FM 7% 2 2“ 25 m. «A. sidered that the 0.9 c.p.s. value for Jc ($3) of dichloroacetaldehyde is an upper limit, then a threefold rather than a twofold barrier to rotatim best fits the experinental data. However, the above conclusion W be questiorned due to the possibility of the potential well of 22 beirg quite broad, as in fluoroacetyl fluoride (28), in which case con- tributions to J c from torsional oscillations would neke it appear much sneller than it really is. From the above discussion, the question of a twofold or threefold barrier to rotation is still unanswered, but it does illustrate the major weakness of n.m.r. in rendering an unanbiguous conclusion in such cases of rotational isarerism. Irrespective of a twofold or threefold barrier, the conclusion that $1 is the nest stable rotaner in the low dielectric constant solvents would remain valid. III. glycidaldehyde and CyclOQropanecarboxaldehn/de The large vicinal coupling constants of glycidaldehyde and cyclo- pr'opan'necarboxaldekwde- and their decrease with increasing tenperature (Table VII) indicate that 12 and 11, respectively, are the nest stable M M rotamers of these conpounds in solution. From the dependence of their -55- cowling carnstants on solvent dielectric constant (Table II), it can be concluded that the cyclOpropane ring donates electronic charge (22) and the oxirane ring withdraws charge (27). From the per cent change of 'VM 0 H / / 27 “Vb of the two coupling constants with solvent polarity, it appears that the two effects (donation and withdrawl of charge) are of the sane magnitude. If the per cent decrease (_ca 10%) of the coupling constant of glycid— alderwde in going from the least polar solvents to the nest polar solvents is conpared to those of dichloroacetaldehyde (92 75%), dibromoacetalde- hyde (_c_a_ 60%), nethoxyacetaldehyde (.c_a 70%), phenoxyacetaldehyde (ga 70%). chloroacetaldelwde (_c_a_ 60%) and bramacetaldehyde (_c_a_ 35%), it nnay be con- cluded that the oxirane ring acts as a nnuch weaker electron withdrawing group than expected from an alkoxy group. a) Wofold or Threefold Barrier to Rotation. Knowing that the nest stable rotaners of glycidaldehyde and cyclopropanecarboxaldehyde in sol- ution are 13 and “1,1, respectively, it is of interest to determine what the nature of the less stable rotaner is. If the less stable rotaner for cyclopropanecarboxaldehyde is $10, then there would be a twofold barrier to rotation and equation 8 would apply . From the available experinental data, Jobsd’ Jt and Jc can be estimated. The smallest experimentally neasured vicinal coupling constant for cycloprOpanecarboxaldehyde is “.53 c.p.s. (Table VI) in m-decalin at 110°. If it is assuned that rotamers 1’1» and 1,1: are equally populated at this ten'perature, which is a false -55- assunption since the coupling constant is still decreasing with in— creasing tenperature, then Jobsd becones “.53 c.p.s. in equation 8. This value of “.53 c.p.s. would represent an upper limit. The largest experinentally neasured coupling constant of this aldehyde is 6.2 c.p.s . (Table II), in fornemide at 36°. If this is assuned to be the coupling constant observed for 100% of 1,1, then this value would represent Jt' Again, this assumption is incorrect and 6.2 c.p.s. represents a lower limit.of Jt' From these two quantities wt 1 6.2 c.p.s. and Jobsd E. “.53 c.p.s.) and equation 8, Jc is calculated to be equal to or smaller than 2.8“ c.p.s. From the Jt values of aliphatic aldehydes (23) and c,8—unsaturated alderwdes (38, 39), the Jt of cyclOpropane- carboxaldehyde may be estimated to be between 7.0 - 7.7 c.p.s. If this is true, then Jc would be between 1 - 2 c.p.s. As discussed in the section on consideration of a twofold barrier for the dihaloace— taldehydes, the decision of whether such relative values of J t and J c are reasonable or not is a difficult one to make. Since the more accurate and reliable microwave and electron diffraction techniques have shown that cyclOpropanecarboxaldehyde (16), cyclOpropyl nethyl ketane (“6), cyclOprOpanecarbOleic acid chloride (“6) and cyclOprOpane- carboxylic acid fluoride (17) exhibit twofold barriers to rotation in the gas phase, it is reasornable to assume that the sane will be true for cyclOpropanecarboxaldemde in solution. The weakness of the n.m.r. technique to be used as a tool from which to decide such a question is further illustrated by the Opposite conclusions drawn regarding the natme of the barrier to rotation, twofold (“7) _v_s_. threefold (“8, “9), about the analogous spZ-sp3 carbon-carbon bond of vinyl cyclOpropane. From the similarity between the vicinal coupling constants of cyclo— -57.- propanecarboxaldehyde and glycidaldehyde and on the basis of the micro— wave arnd electron diffraction results Just nentioned, it may be also assured that glycidaldehyde exhibits a twofold barrier to rotation. IV. Effect of Solvent Polarity on Rotaner Stabilities a) Monosubstituted Acetaldehydes. The increase in the rotaner ratio Q4122 reflected in the data of Tables VIII, IX, XI and )CIII as the dielectric constant of the nediun increases, is reasonable in view of the higher dipole nement of 22 over that of 2,8. It is also “1, not 28 29 W m understandable that this increase would be nere pronounced for the retailers of chloroacetaldehyde than for those of broneacetaldehyde one to the carbon-chlorine bond being nere polar than the carbon- bronine bond. This large difference in the dipole monents of the two retainers is also responsible for AH° values being much nere negative than the corresponding AG° valnes in solvents of high di- electric constant. Increase in tenperature decreases the dielectric constant of the solvent. This decrease causes a decrease in the ratio 219/58, far nere rapid than would be expected and causes the coupling constants to increase rapidly with increasing tenperature. The result is the calculation of nere negative and hence , inaccurate AH° values. For this reason, in solvents of high dielectric constant, -58- the AG° values reflect the enthalpy differences to a better degree between rotamers whose dipole marents differ greatly, than do the calculated AH° values. The only meaningful AH° values calculated for such rotamers by the terrperature dependence of the spin—spin coupling constants are those in solvents of low dielectric constant. Since in tn:_'_a_ns_-decalin, (511° 2 AG° for monohaloacetaldehydes , the argunent that AS° between rotamers l2 and TX is zero is indeed valid. As expected, whereas methoxyacetaldemIde and phenoxyacetaldehvde behave similarly to the haloacetaldehydes, netrwlmercaptoacetaldehyde does not. The effect of solvent on the enthalpy and free energy differences (Tables XIII and XII, respectively) between the rotamers of nethylnercaptoacetaldelwde is not very pronounced and is probably due to the similar polarities of rotamers lg and R. b) Dihaloacetaldedees. The increase of the relative stability of rotaner 3% over that of gr with increasing dielectric constant of the solvent is sham in Tables XVI, XVII and XVIII. In view of the mgrer dipole moment of 323, this trend is predictably reasonable, and parallels that observed for methoxyacetaldehyde, phenoxyace- taldehyde and the mnmaloacetalderwdes . This large difference in the dipole wrents of the rotaners is also responsible for AH° values being more negative than the corresponding AG° values in solvents of high dielectric constant. The inadequacy of the solvent dielectric constant effect to explain all the changes observed in the AG° values has already been mentioned while disctssing the aromatic solvents benzene and toluene. 0n the basis of the low dielectric constant of these solvents, it would have been eXpected that 2}, would be more stable than 5% rather -59.. than the reverse which was found experimentally. This reverse effect observed in the aromatic solvent is best interpreted in tems of solute—solvent interactions that destabilize a with respect to 3%. Some sort of stereospecific association (50), such as pictured in ‘32 and a, could rationalize the results in terns of stronger non- bonded repulsions between benzene and halogen in 3% than in a. O R s 6 H / R/ H 31 'VD V. Conparison of Results with Other Systems a) Monosubstituted Case. The conclusion from the infrared studies (2“) that chloroacetaldehyde exists essentially in conformation 2.2 is certainly in disagreement with the n.m.r. results here. In fact, $2 is the unjor rotaner, about 55%, only in the low dielectric constant twdrocarbon solvents and in carbon tetrachloride. If the degeneracy factor of two that favors l2 over 2.2 is removed, than 2.2 would be less stable than R by about 300 cal/mole even in these low dielectric constant solvents . The suggested nonbonded repulsions (214) between chlorine and carboryl oxygen cannot be a controlling factor of the relative sta- bilities of l3 and 2.2. It was found (1) that for chloroacetone in the liquid state (e: ~ 30), 32" and 3% are of comparable stability. It was sugested (7) that 32» might have been more stable had it not been -60- O O H C]. L , CH3 Ch3 / , I H’ H H C1 for the norbonded repulsions between the m chlorine and the nethyl. However, if this interpretation were correct, then chloroacetaldehyde would have been expected to exist predominately in 33 rather than l2, as the g_a_u__ch_e_ chlorine-netl'wl interactions in chloroacetone are absent in chloroacetaldehyde. If these nonbonded interactions between the w chlorine and nethyl groups of chloroacetone are affecting the rotaner stabilities, then to account for the results, these inter- actions would have to be attractive rather than repulsive. The enthalpy differences (Table X) for lg $l2 of branoacetalde- hyde are shown to be less negative by 300 cal/mole than the correspond- ing AH° for chloroacetalderwde (m-decalin) . This observation mitigates against the polarizability of the R group (dipole-induced dipole interactions) being very inportant in controlling the ratio I‘d/lg. Also, if one were to consider that in these low dielectric constant solvents (e m 2), the electrostatic dipole-dipole interactions would destabilize TB of chloroacetaldehyde more than R of bromoacetal- dehyde with respect to their other rotamers. Sumnarized in Table XIX are the energy differences (either in low dielectric constant solvents or in the gaseous state) between the rotamers of RCHZCOY compounds, where R is halogen. In all cases, -51- TABLE XIX. EDERGY DIFFERENCES BETWEEN ROTANERS 0F SG‘E RELATED SYSTEMSa 1 System 2 R AH" ,cal/mole Method Reference N N H 0 R O L ——L- K 01 —350 NMR This work *— 11,? H H/ I Br 0 NMR This work R H H o F -560 IR 8 tor—R ( m a Cl - 00 x' t ‘, Et 5 H R Br 0 IR 8 H Cl (r - 01) ca -1200b Raman & IR 5 Iii/KY “Efk Br (Y . Cl) -1ooo Raman & IR 1 Br (Y I Br) -1900 Hanan 8: IR 1 = k -910 Microwave 28 aThese values are either in low dielectric constant solvents, such as pentane and 0011,, or in the gaseous state. bEstimated from ref 5. -6 2... except fluoroacetyl fluoride, the results have been interpreted in terns of a threefold rather than twofold barrier about the sp2-sp3 carbon—carbon bond. In all cases, AH° for 2:422 is negative or zero. The only exception appears to be N-nethylchloroacetamide (7), where AH° is quite positive. It is inpossible to decide (ran the published results (1, 2, 3, it) whether AH° for l¢2 of the monohaloacetones is positive or negative in the gaseous state, although it appears that it is negative in the liquid state. Minor differences of the results may be explained in terms of nonbonded and dipole-dipole electrostatic interactions . For example, substitution of branine for chlorine in the monohaloacetaldehydes and in the etryl acetates (8) causes AH° for l¢2 to become more positive. This same trend has been observed in the 3—halopropenes, where AH° for a:& is -100 cal/mole for 3-fluor0pr0pene (19, 26), +100 cal/mole for 3—chlor0pr'0pene (20) and progressively more positive in 3—bromo- propene and 3-iod0pr0pene (9, 27). The best rationalization of these results is increase in the nonbonded repulsions between halogen and H H H H a I: a E H H / / H’ R H, H 311 35 '\/\a 4/» oxygen (or methylene) in rotamer 2 (or 3%) as the size of the halogen increases . The observation that the AH° values of the haloacetaldehydes are less negative than those of the ethyl acetates (8) and haloacetyl halides -63- (1, u, 5, 28) can be explained in terms of dipole-dipole interactions. The difference between the dipole nonents of l and 2 would be greater for the haloacetaldehydes than the haloesters and haloacetyl halides. The ratio of ye would then be smaller for the haloacetaldehydes than the haloesters and haloacetyl halides. This same argwnent (7) has been used to partly explain the differences in the relative stabilities of the rotamers of chloroacetyl chloride, chloroacetone and N-methyl— chloroacetamide. Nonbonded interactions between the gauchg groups R and Y in rotaner l have been used (28) to explain why fluoroacetyl fluoride exhibits a twofold barrier to rotation, when all other haloacetyl halide studies exhibit a threefold barrier to rotation. As shown in 36;, when both R and Y groups are the small fluorine stars, the repulsion between them is very small and, hence, 6 . 0°. However, when they are the larger atoms chlorine and bromine, the nonbonded repulsions change the equilibriun configuration to e - 30° (1, ll). Even though these arguments show sone success in rationalizing some of the trends observed, they are still inadequate to explain why, in most cases, 2 is so much more stable than I. b) Mmdraloacetaldehydes Ev; 2—halocyclohexanones. Several investi- gations (25) have established that the ratio gas/31$ increases when R is _gu- changed from fluorine to chlorine to bromine. In solvents of low 0 O NH ~12 R H 32 .312 dielectric constant (51) , when R is fluorine, the equatorial rotaner is more stable than the axial (25d, 1') and when it is chlorine or bromine, the axial rotaner is more stable (2‘4a—c, f). In midrocarbon solvents , the free energy difference, AG° , for as :33? was found (25f) to be -—l70 cal/mole, +7lIO cal/mole and +1280 cal/mole for 2- fluorocyclohexanone, 2-chlorocyclohexanone and 2-bromocyclohexanone , respectively. The corresponding values calculated (25f) by taking into account nonbonded, dipole-dipole and dipole-induced dipole interactions were +1130, +1130 and +1100 cal/mole. Irrespective of how the results are interpreted, if the sane criteria are applied to chloroacetaldehyde and bromoacetaldelwde, the AH° values for '12: 12 would turn out to be similar to those of 2-chlorocyclohexanore and 2-brcmocyclohexanone . How- ever, the experinental values are -300 and about +110 cal/mole for chloro- acetaldehyde and bromoacetaldehyde, respectively. It seems therefore, that the basic factor, in addition to all those discussed, controlling rotaner stabilities, might be the sane one restricting the barrier to rotation about carbon-carbon single bonds and also associated with the nature of the axial bonds (52) . The differences between monohaloaldehydes and 2-halocyclohexanones may very well arise from torsional strain. The dihedral angles calculated (25f) for the equatorial (38e) 4: a 16° 17', MA; -55. and axial ($83), a 102° 13', bromccyclohexanores are different than those of the correspondirg acyclic rotaner-s 12 and 1,8. Fran all indications (l, 16, S, 28) the correspmding dihedral angles (¢) for O ’12 and 1,8 are zero and 150 , respectively. H Br ¢¢ = 16° 17' ‘. 0 ¢ 3 102° 13' g c O B. L. kc 38a 33% c) Disubstituted Cases. From infrared studies (24), it was sugested that dichloroacetaldelwde exists in essentially me conforneticn, probably «32' 'mis however, is in disagreement with the n.m.r. results presented here. In low dielectric cmstant solvents, both 51 and 2,2 (or 39;) are 0 Cl H Cl H 39 vs. present in about equal concentrations if the degeneracy of $2 is not taken into account. Cmparison of mmochloroacetaldekwde with dichloroacetalderyde shows that whereas the rotaner with the hydrogen eclipsing the car-bowl grow is more stable for dichloroacetaldehyde in saturated hydrocarbon solvents, -66- o o it is the less stable me in nmochlorcacetaldemde, Lg. , AHI - AH2 (equations 10 and 11) is negative. Sumerized in Table XX are several Z H k as: R |\ y ‘ Y (10) / / / I H H . Z Z H k AH‘Z’ R I L y * (11) j / RI, H’ R o 0 AH; - AHZ values for related system, either in the gas phase or in solvents of low dielectric coretant. In all cases, the AH? - AH: values are regative and of cmparable magnitude. In the case of the halOpropenes, no. _6_ and 1, the differences were attributed (l9 , 20) to less favorable van der Waals attractions between halogenandmdmgenofthenetmleregrowinZthaningduetotre C-R bald being less polar in 1 than in g. This explanation would lead to the opposite results for no. _l_._- 5,, unless it was applied mly to l and .9 A possible explanaticn for _l - i would be electrostatic dipole-dipole interactions that favor 1 over 2 and Q over 1. The dipole mment difference for g and 1 would be nuch larger than between 1 and ‘2' and would lead to a greater energy difference between E and 1 than between 1 and 2 ~. -57. TABLE )0(. COVPARISCN 0F ENTHALPIESa BETWEEN NDNOl'iALDSUBSTI'I'U'I'ED AND DIHALDSlBSTITUTED SYSTEIVS No. Systems AH; -AH§ (cal/"Ole )b Reference 1 CH201CHO _vg cncrzcno _600 This work 2 CH2me X3; CHBI'ZCHO -500 This work 3 CHZF‘COzEt _vg CHF2OOZEt -500 8 ‘8 CHZClCOzEt _Vg CHClzCOzEt -500 8 5 W2CIOOCI E CHCIZCOC]. _c_a_ -1200c 5 ,6 6 CHZFCH-CHZ _V_3_ CHF‘ZCH-CH2 -500 to -1500 19,26 7 CH2010H-CH2 _v§_ CHClZCHI'CHZ -700 20 the enthalpies are either in the gas phase or in low dielectric solvents. bFor AH; and AHE see text, equations 10, and ll. °Ihe enthalpy for the rmmhalo conpound was estimated fran the "data of reference 5. B . Chemical Shifts rlhe chemical shift data for the monosubstituted and disubstituted acetaldehydes best agree with model l1, rather than lg. The latter model would predict that Ha, (:42) in the plane of the carborwl group, would be deshielded. 'Ihe former model would predict the Opposite. 0 Ha L / / Hb no 'Vh From Table VII, it can be seen that in nonaromatic solvents, the chemical shifts of the methylene (or nethine) protons move upfield as the polarity of the solvent decreases. Fran the previous results on the stability of rotamers, it was established that for chloroacetaldehyde, bromoacetalde- hyde, dichloroacetaldehyde, dibranoacetalderwde, methoxyacetaldehyde, phenomacetaldehyde and nethylmercaptoacetaldehyde the stability of w increased with decrease in the dielectric constant. Therefore, Ha is shielded with respect to Hb' 'Ihis shielding effect can be seen graphi— cally in Figs. 8, 9, 10 and 11 for each of the aldehydes. By using the aldehydic and methylene (or methine) chemical shifts in pentane as a reference position, the chemical shifts of these protons in other solvents have been plotted against the per cent population of the rotaner which was obtained from the coupling constant data. The per cent populations plotted for chloroacetaldehyde, bromoacetaldehyde, dichloroacetaldehyde , dibranoacetaldehyde and nethylnercaptoacetaldehyde ~68— O mw om ms OOH an OOH ma om mm o on Suzmm .m .5 o.o $68me .3 \\= er Nozaxpe .m m a a ma «ENE .m /_ m on m m m m . o 04 mo A :8 :0 H x z m P anm \x . m4 M . we“ Ommammov .m .8 a m N @8890 .m \\ W m 3 I m 0.0 {OI NSNmo ... m./_ 0:: U maze .m o z m o z m :8 .m \ m m \ I m 8323; U mmomfimxemmo A a < o; M m 0852 O Egon /_ o 0 m4 0 mm om ma 2: ma OOH ms om mm o zmn4Ihwz_«on mzo~h<43doa mmzzhmz OZ< U_Q>IUQJ< th mo mhufizm 4;_i_<_i_., £33, 5876 (1960). c) J. Allinger and N. L. Allinger, Tetrahedron, 2, 64 (1958). d) A. s. Kende, Tetrahedron Letters, 1}), 13 (1959). e) C. Y. Chen and R. J. W. IeFevre, J. Chem. Soc., 3700 (1965). f) Y. H. Pan and J. B. Stothers, Can. J. Chem., 42, 2943 (1967). H. Hirota, J. Chem. Pth., 4/2, 2071 (1961). A. A. Bothner—By and H. GCmther, Discussions Faraday Soc. , 3’4, 127 (1962). E. Saegebarth and E. B. Wilson, Jr., J. Chem. Phys., 36, 3088 (1967). a) L. s. Bartell, B. L. Carroll and J. P. Guillory, Tetrahedron letters, 1’3, 705 (1964); J. Chem. Phys” 43, 647 (1965). b) R. N. Schwendeman and H. N. Volltrauer, private communication. D. L. Hooper and R. Kaiser, Can. J. Chem., :42, 2363 (1965). A. D. Buckingham, ibid., «3’8, 300 (1960). P. T. Narasimhan and M. T. Rogers, J. Chem. Phys, 31, 1302 (1959). L. M. Jackmen, Applications of Nuclear Magnetic Resonance Spectro- scom in Organic Chemistry, Pergamon Press, New York, N. Y., 1959, pp. 122-1250 G. J. Karabatsos, G. C. Sormichsen, N. Hsi and D. J. Fenoglio, J. Am. Chem. Soc., 89, 5067 (1967). (35) (36) (37) (38) (39) (40) (41) (42) (43) (44) (45) (46) (47) (48) (49) (50) (51) (52) -82- . Stedman and L. D. Davis, Tetrahedron Letters, 12, 1871 (1968). . Abraham and J. A. Pople, Mol. Phys., 3, 609 (1960). N . Pople and T. Schaefer, ibid., 3, 547 (1960). R J R J J. G. Powels and J. H. Strange, ip__i_<_i_., 5, 329 (1962). J A A A. Bothner—By and R. K. Harris, J. Org. Chem., 32’ 254 (1965). s. Ebersole, S. M. Castellano and A. A. Bothner—By, J. Phys. Chem., $3. 3430 (1964). R. J. Abraham and K. G. R. Pachler, Mol. Phys., 7, 165 (1963). ""———-' N a) M. Karplus, J. Chem. Phys., 30, 11 (1959). b) M. Kar'plus, J. Am. Chem. Soc., 85, 2870 (1963). W """n-n-i- ‘- - é a. ' ..' R. Lynden—Bell and N. Shepard, Proc. Roy. Soc., A, 229, 385 (1962). ?’ Hobgood and G. H. Goldstein, J. Mol. Spectry” 12, 76 (1964). E M . Manatt and D. D. Elleman, referred to by J. B. Lambert, . . Durham, P. Lapouter and J. D. Roberts, J. Am. Chem. Soc., 87. 3896 (1965). W L. S. Bartell, J. P. Guillory and A. '1‘. Parks, J. Pth. Chem., 62, 3043 (1965). W. Li'lttke and A. deMeiJere, Arggw. Chem. Interp. Ed. Engl., 5, N 521 (1966). be .,..,33 H. Gilnther and D. Wendisd'l, ibid., 2, 251 (1966). G. R. DeMare and J. S. Martin, J. Am. Chem. Soc., 88, 5033 (1966). 'Vb G. J. Karabatsos and R. A. Taller, Tetrahedron, 34, 3923 (1968) and references cited therein. In contrast to all other studies, it has been reported by K. Kozima and E. Hirota, J. Am. Chem. Soc., 83, 4300 (1961); K. Kozilma and Y. Yamanouchi, 121g. , 81, 4159 (1959), that even in heptane, AE for 393239 e is -0.75 Kcavmole; equatorial conformer is more stable than the axial. E. B. Wilson, Jr., Proc. Natl. Acad. Sci. u. s. A., A}, 816 (1957); Adv. Chem. Phys., 3, 367 (1959). (53) (54) (55) (56) (57) (58) -83— L. F. Fieser, Experiments in Organic Chemistry, D. C. Heath and 00., 3rd ed., Boston, 1957, p. 281. M. N. Schukina, J. Gen. Chem. (131153.), 12, 1653 (1948). L. A. Yanovskaya, A. P. Terentiev and L. I. Belenskiy, ibid., 53. 1594 (1952). A. N. Dey, J. Chem. Soc., 1057 (1937). H. C. Brown and C. P. Garg, J. Am. Chem. Soc., 86, 1085 (1964). E. L. Wick, T. Yamanishi, L. C. Wertheimer, J. E. Hoff, B. E. Proctor and s. A. Goldblith, Food Tech., '1}, 914 (1959). I II: I l I ll