Magnetic circular dichroism studies. Part 55. The aliphatic nitro chromophore

摘要

1979 907 Magnetic Circular Dichroism Studies. Part 55.l The Aliphatic Nitro Chromophore By Gunter Barth, Nada Waespe-Sarcevic, Robert E. Linder, Edward Bunnenberg, and Carl Djerassi," Department of Chemistry, Stanford University, Stanford, California 94305, U.S.A. Lloyd Seamans and Albert Moscowitz, Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455, U.S.A. The magnetic circular dichroism (m.c.d.) of twelve nitroalkanes is reported. For several of these compounds two distinct m.c.d. bands are observed in the 240-350 nm region. The presence of these bands is interpreted in terms of two electronic transitions, viz. a u+r* transition at ca. 320 nm and an n+x* transition at ca. 270-280 nm. The magneto-optical B values for the a+n" transition appear to be small in magnitude throughout the series of com- pounds ; the B values for the n+x" transition are generally much larger in magnitude and are sensitive to alkyl substitution and to solvent effects.A qualitative interpretation of the spectra is given in terms of the effects of static and structural-vibrational perturbations on the m.c.d. associated with symmetry-forbidden transitions. Solvent effects observed in the m.c.d. spectra of the nitroalkanes are also discussed. SOMEtime ago we reported the magnetic circular dichroism (m.c.d.) spectra of a number of aliphatic ket~nes,~Jand we noted then that for saturated ketones there appeared to be a variation with geometrical structure of the sign and magnitude of the magneto- optical B term associated with the n+x* transition.Later,4-6 group theoretical arguments were employed to develop a set of sector rules useful for analysing the magneto-optical activity associated with a locally symmetry-forbidden transition when that transition gains intensity through static perturbations. This analytical protocol was subsequently extended to include vibrational-structural (vibructional) and it has proven quite useful for correlating the aforementioned m.c.d. data on saturated ketones with their structures.*-ll It is of interest now to extend the investigations to include other chromophores which exhibit forbidden electronic transitions of n+x* character. Nitroalkanes show a weak absorption band at ca. 270-280 nm, which has been assigned by several workers l2 to an n+x* transition, and we report here the m.c.d.spectrum associated with this transition for a number of such compounds. A detailed interpretation of the observed magneto-optical activity of the nitro- alkanes along the lines of that presented for ketones is not possible at this time. Amongst other reasons, this is due to special circumstances introduced by internal X A rotation about the C,-N bond. However, an inter-pretation based upon vibructional and gross structural effects provides a qualitative understanding of the data. EXPERIMENTAL The m.c.d. spectra were obtained with JASCO model J-5 and J-40circular dichrometers equipped with 49.5 kG superconducting and 15.0 kG electromagnets, re~pective1y.l~ The U.V.absorption spectra were recorded on a CaIy 14M spectrophotometer. The magneto-optical B values and the dipole strengths were obtained by planimetric in-tegration of the normalized spectra using the relationships (1) and (2)l4 where the molar magnetic ellipticity el, is B(debye2 bohr magnetons/cm-l) w D(debye2)w (1O8.9hmaJ1 I,,,, amp;(A)dh (2) expressed in units of deg cm2 dmol-l gauss-l and the ex- tinction coefficient E has units of 1000 cm2 mol-l. In those instances where amplitudes of both positive and negative sign were observed in the m.c.d. spectrum, the integration was carried out over both bands and the reported B values are those appropriate to the entire band. For an absorption band which is partially overlapped on the short wavelength side by another band of greater intensity, it was assumed that the band shape is symmetrical with respect to the absorption maximum. The compounds used in this study were obtained from the following sources : nitromethane, spectroquality from MC/B (Norwood) ; nitroethane and 2-nitropropane from Aldrich; 1-nitrobutane, 2-nitro-2,4,4-trimethylpentane, nitrocyclohexane, and nitrocyclopentane were samples used in previous mass spectrometric work l6 and were provided by Professor H.Feuer (Purdue University) ; 2-nitro-2-methylpropane was prepared from the corresponding amine by the procedure of Kornblum et al. ; 16a 1-nitropropane and 1-nitropentane from the corresponding bromides by the Victor-Meyer reaction ; lsb 1-nitroadamantane by nitration of adamantane; l7 and 2-nitroadamantane by oxidation of the amine (Aldrich Chemical Co.) with perbenzoic acid in benzene solution. The purities of all samples were checked by g.1.c.The solvents used were methanol (Photrex re- agent from Baker Chemical Co.), iso-octane (Spectro-quality, from MC/B), and 1,1,1,3,3,3-hexafluoropropan-2-o1 (Sequanalgrade, from Pierce Chemical Co.) . RESULTS The m.c.d. and absorption data are compiled in Table 1 and the spectra, in iso-octane, of a few representative nitro- alkanes which differ in the degree of methyl substitution at the a-carbon are reproduced in Figure 1. There nitro- J.C.S. Perkin I1 TABLE1 M.c.d. and absorption data for aliphatic and alicyclic nitro compounds Absorption M.c.d./nmC Compound Solvent a A,a,./nm (E) A)imax.(1O4x ~IM) 10'B 102Da Nitromethane (1) Neat 271 (15.1) 275 (-4.7) 19 3.1 I0 278 (17.9) 278 (-5.8) 22.9 2.7 M 273 (15.3) 277 (-5.4) 22.8 2.7 HFIP 265 (15.7) 268 (-4.2) 15 2.7 Nitroethane (2) Neat 275 (19.0) 278 (-4.5) 17.0 3.7 I0 278 (20.6) 278 (-5.1) 21.7 3.3 M 274 (18.2) 278 (-4.4) 17.7 3.2 1-Nitropropane(3) Neat 277 (22.9) 277 (-3.2) 12.6 4.3 I0 279 (23.6) 280 (-4.0) 16.8 4.1 M 275 (22.3) 278 (-3.1) 12.4 4.0 HFIP 269 (30.0) SO (0.5) -2.4 5.5 l-Nit robutane (4) Neat 277 (24.7) 277 (-2.5) 10.0 5.0 I0 278 (24.2) 280 (-3.5) 14.2 3.0 M 273 (25.0) 275 (-3.3) 13.5 4.1 HFIP 267 (37.8) 310 (0.5) -2 6.2 l-Nitropentane (5) Neat 277 (27.5) 277 (-3.1) 13.0 5.1 I0 278 (25.3) 279 (-3.3) 14.1 4.2 M 274 (24.8) 277 (-2.6) 9.4 4.5 HFIP 269 (38.0) 280 (1.2) -5.6 6.7 2-Nitropropane(6) I0 279 (22.0) 275 (-0.7), 0.5 3.4 310 (0.4)M 278 (21.1) 300 (0.8) -4.6 3.4 HFIP 270 (28.1) 270 (2.0) -10 4.4 Nitrocyclohexane (7) Neat 278 (33.3) 270 (-1.7), 3.3 5.9 310 (0.6) I0 280 (28.6) 270 (-1.6), 6.4 4.4 320 (0.5)M 277 (30.0) 265 (-l.l), -0.6 4.9 305 (1.0)HFIP 270 (58.0) 274 (5.9) -25 9.0 Nitrocyclopentane(8) Neat 278 (25.9) 279 (-1.8), 6.8 4.6 I0 280 (24.1) 280 (-2.9) 11 3.8 M 277 (22.5) 275 (-l.O), 2.4 3.8 320 (0.4)HFIP 269 (31.0) 272 (3.9) -12 5.1 2-Nitroadamantane (9) I0 282 (38.4) 282 (-7.9) 36 6.3 M 279 (32.2) 280 (-6.3) 25 5.6 2-Methyl-2-nitropropane (10) I0 280 (24.0) 280 (1.3) -9.7 3.5 M 279 (22.2) 285 (3.8) -20.3 3.4 HFIP 271 (31.3) 273 (6.5) -25 4.9 l-Nitroadamantane (11) I0 281 (30.5) 282 (-1.3), 3.1 5.9 320 (0.6)M 280 (38.6) 285 (6.2) -20.6 5.9 HFIP 270 S (69.7) 278 (13.4) -50 Z-Nitro-2,4,4-trimethylpentane(12) I0 280 (24.5) 285 (6.9) -32 3.7 M 280 (25.6) 285 (8.1) -35 3.7 HFIP 271 (42.3) 272 (13.3) -55 6.9 (I I0 = iso-octane,M = methanol, HFIP = 1,1,1,3,3,3-hexafluoropropan-2-o1.Molar extinction coefficient in 1 000 cm2mol-l S = shoulder. Molar magnetic ellipticityh deg cm2 dmol-1 G-l.B Value in units of debye2 bohr magnetons/cm-l. 0 Dipolestrength in debye2 methane (1)exhibits the largest negative m.c.d. amplitude, highly dependent upon the degree of methyl substitution with a B value of 22.9 x whereas the spectrum of 2-at the cc-carbon. The latter band is negative for nitro- methyl-2-nitropropane (10) shows a positive m.c.d.ampli- methane, negative but less intense for the secondary nitro- tude with a B value of -9.7 x The m.c.d. spectrum alkanes, and is positive for the tertiary nitro compounds. of l-nitrobutane (4)has a single band of negative sign which The changes in the m.c.d. sign pattern are therefore very is of lower intensity (B 14.2 x lo-') than that of nitro- similar to those observed in the series of aliphatic ketones methane. The m.c.d. spectrum of 2-nitropropane (6)shows wherein the sign of the m.c.d. band associated with the bands of opposite sign at 275 (negative) and 310 nm carbonyl n-m* transition changes progressively from (positive).Bisignate curves are also observed in the negative to positive with an increase in the number of m.c.d. spectra of other compounds in which the nitro group cc-methyl substituents. It can also be seen from a com-is attached to a secondary carbon atom, e.g., nitrocyclo-parison of the B values given in Table 1 that in addition hexane (7)and nitrocyclopentane (8). This situation also to the influence of a-alkyl substituents, there is a further obtains for l-nitroadamantane (1l), which has a tertiary effect associated with non-cc-substitution. For example, a-carbon atom. the B values for the series of secondary nitroalkanes (6)-The experimental data therefore suggest that the m.c.d. (9) vary from 0.5 x to 36 x lo-'. spectra of the nitroalkanes in the wavelength region under Spectral changes associated with changes in solvent consideration are composed of two overlapping bands.It polarity or hydrogen-bonding ability have been widely appears that the longer wavelength band is of low intensity used to assign the n+r* character of an absorption band. throughout the series of compounds investigated, whereas In fact, the assignment of the weak absorption band at the sign and amplitude of the shorter wavelength band are ca. 270-280 nm in nitro compounds rests mainly upon the small blue shift wllich this band undergoes in polar sol- vents.lE We have therefore obtained the m.c.d. spectra of the nitroalkanes in three different solvents, uiz.,iso-octane, methanol, and 1, 1 ,1,3,3,3-hexafluoropropan-2-o1(HFIP).The latter solvent is known to undergo particularly strong 36.0 27.0 m 18.0!~~ , -~ 9.0 TABLE2 Nitromethane electronic states State Energy (eV) N (ground) 0.0 a+x* 3.4 nojx* 3.9 x,-+n* 6.4 xoja* 9.6 Data from rcf. 21. TABLE3 Calculated and experimental dipole strengths for nitro- methane Chroniophoric vibrations Dipole strength x lo2(debye2) n+x* a+x* Out-of-plane bend (b,) 1.02 0.48 In-plane bend (b,) 0.02 0.06 Antisymmetric stretch (b,) 0.06 0.12 Equilibrium (no vibration) 0.002 0.06 Experimental b 2.7 a All excitations from the highest n or cr orbitals to virtual orbitals and all excitations from bonding orbitals to the lowest x* orbitals were included in the CI basis set.6 Solvent: iso-octane. symmetry forbidden n+x* transition and a transition of o+n* character at lower energy. Although it is form- ally electric dipole allowed, the CNDO/S-CI calculations of Harris 21 yield a dipole strength of zero for the latter tran- sition. Some CNDOIB-CI calculations which we performed 2:-. (Table 3) support Harrisrsquo;s results. We therefore conclude --*--that the a+x* band should lie to the red of the n+n* 250 2rsquo;70 290 310 330 350 x /nm FIGURE M.c.d. (upper curves) and absorption (lower curves) 1 spectra (in iso-octane) of a series of a-substituted nitroalkanes in the near U.V. spectral region hydrogen bonding due to the electron-withdrawing effect of the fluorine atoms,lB The spectral data are listed in Table 1, and Figurci 2 shows the m.c.d.and absorption spectra of nitrocyclohexane (7) in these three solvents as an example of the changes observed. In going from iso- octane to methanol as the solvent the bisignate nature of the m.c.d. spectrum of (7) is preserved, but there is a con- siderable blue shift in the crossover point; the absorption spectrum is nearly constant. Dramatic increases in both the m.c.d. and absorption intensities are observed in HFIP solution. Here the m.c.d. is all positive and the absorption maximum is shifted to the blue by ca. 10 nm relative to the maximum in iso-octane solution. As can be readily seen from tht: data in Table 1, the solvent-induced changes in B values are consistent throughout the entire series; a relatively small decrease in B is observed upon going from iso-octane to methanol, and a considerably larger decrease is observed upon going to HFIP solution.Furthermore, the B values of the neat liquids, which can be regarded as solutions of the nitroalkanes in weakly polar solvents (themselves), are consistently less positive than the B values obtainetl in iso-octane solution. Spectral A ssignwzsnts.-The two detailed molecular orbital calculations which have been carried out for nitro- methane 20~2~are in essential agreement with respect to the assignment of the lower energy transitions. The assign- ments obtained by 13arris 21 are given in Table 2. Two electronic transitions are predicted to lie in the region of the spectrum which concerns us here.These are the locally band, and that both bands gain the major part of their 0.1, 0.2 f -a350.0 c -0.2 -0.4 260 280 300 320 340 360 h /nm FIGURE M.c.d. (upper curves) and absorption (lower curves) 2 spectra of nitrocyclohexane in iso-octane (IO), methanol (amp;I),and lfl,1,3,3,3-hexafluoropropan-2-o1(HFIP) intensities through vibrational coupling with higher-lying states. When such vibrational perturbations are included in the CNDOI2-CI calculations,6 the n+x* band is pre- dicted to be the more intense of the two. The above theoretical predictions are supported by the solvent effect work previously mentioned and by studies of the natural optical activity of a number of nitrosteroids 22,23 which exhibit, in addition to the 280 nm band, a Cotton effect at ca.330 nm. As was pointed out previously, the m.c.d. spectra of several of the nitroalkanes exhibit two bands of opposite sign, a situation which is superficially similar to that observed in the n1.c.d. spectra of some saturated ketone^.^*^*^^^-" In the case of the ketones, the oppositely signed bands were shown5y6 to arise from different vibra- tional modes through which the carbonyl n+x* transition gains intensity. In principle, this interpretation of the m.c.d. spectra could also be applied to the nitroalkanes. However, in view of the evidence outlined in the preceding paragraph, we favour the assignment of the positive m.c.d. amplitude at longer wavelength to the G+T* transition, with this band being at least partially overlapped by the stronger n+x* transition which appears at 270-280 nm in the absorption spectrum.In further support of this assignment it is worthwhile to note that the U.V. spectra of virtually all of the nitroalkanes investigated in this study are decidedly unsymmetrical about the 270-280 nni absorption maximum and show a long tail which extends to the red of 350 nm (see, for example, Figure 1). It further appears from the data that the magnitude of the B term associated with the G+X* transition is small for all the compounds. Hence the differences in the B values reported in Table 1, which values were obtained by integration over both bands, will be interpreted as reflecting predominantly the changes in the contributions which the n+x* transition makes to the total observed spectrum.Before concluding this section it is necessary to point out that deMaine et aZ.24325have obtained some evidence from U.V. spectra that nitromethane is capable of forming dimers in nonpolar solvents. It might therefore appear possible to attribute the two observed m.c.d. bands to different (monomeric and dimeric) species. We have been unable to reproduce the results of deMaine and his co- workers, but in order to further eliminate this possibility we investigated the concentration dependence of the m.c.d. spectrum of nitrocyclopentane (8), a molecule in which the two oppositely signed B terms are of approximately the same magnitude.The normalized intensities of the two m.c.d. bands remained unchanged, within experimental error, over a concentration range of from 8 (neat liquid) to 0.06h1, a result which is at variance with the hypothesis that the oppositely signed B terms reflect the presence of a monomer-dimer equilibrium. DISCUSSION A locally symmetry-forbidden transition gains in-tensity through static (structural) and vibrational perturbations of the chromophoric eigenstates. A recent theoretical treatment 8 has shown that the B value of such a transition can be written as the sum (3) of three B(A+J) =BV(A+J) +BS(A+J) +BVS(A-+J) (3) partial B values. Here BS(A+J) represents the contribution of the static perturbation, BV(A+J) that of the vibrational ones, and the term BVS(A+J) arises from the interaction of the vibrational perturbations with the totally symmetric part of the static one.J.C.S. Perkin I1 The partial B values may be further decomposed: equations (4) where BrS(A+J) arises from that part (4) of the static perturbation which forms a basis for the rth irreducible representation of the chromophoric point group, and (5) and (6) where the summations are over BV(A+J) = 2 BrV(A+J) (5) r BVS(A+J) = 2B,""A+J) (6) r the normal vibrational modes. Thus, BrV(A+J) is the purely vibrational contribution of the rth normal mode to the total B value, and BrV'(A+j) rsults from the interaction of the rth normal mode with the totally symmetric part of the static perturbation.In our earlier work on saturated ketone^,^ we found ' vibructional ' effects, associated with the BV + BVS terms of equation (3), to be a principal factor governing the observed variations in the m.c.d. intensity. Hence, we began our analysis of the nitroalkane spectra by focusing upon these same terms. The electronic states of the nitro chromophore form bases for the irreducible representations of the Czz.point group and, to the order of approximation implied by equations (3)-(6) , these states are mixed among them- selves by vibrational perturbations and by a static perturbation U which is associated with the substituent atoms. If one neglects interactions among them, the perturbing atoms contribute additively to U equation (7) where Ui depends only upon the nature of the ith substituents (7) a perturber and its geometric disposition vis-2-vis the nitro chromophore.If we consider only the static perturbation and use first-order perturbation theory, we may write the electronic states of a nitroalkane as equation (8) where i-Ko#Aa IAO) and KO)are unperturbed chromophoric states with energies EAoand EKO,respectively. It is shown in ref. 8 equation (19) et seq. that B,VS((A+J) depends additively -upon first-order changes wrought by the totally symmetric component of U in symmetry-allowed matrix elements. We wish therefore to consider quantities such as those in equation (9) where U(Al) is the totally symmetric part of Ui, 0is an arbitrary operator which forms a basis for one of the irreducible representations of Czv, and where the matrix element (Ao(alJo;,is symmetry-allowed.Specifically, equation (9) represents a change in the magnitude, but not the polarization, of an electric or magnetic dipole transition moment from which m.c.d. intensity is vibrationally borrowed, or a change in the magnitude of a vibrational coupling term which effects the intensity borrowing. If we assume free rotation about the Ca-N bond (see Appendix), we may average equation (9) over the rotation angle to obtain an expression of the form (10) substituents 40lJ-A0161J0= 2 fi (10)i wherefi depends only upon the nature of the ith extra- chomophoric atom, its distance from the origin (which lies on the Ca-N axis), its perpendicular distance from the Ca-N axis, and the form of the operator 8.BrvS(A-,J) is a weighted sum of such changes in symmetry-allowed matrix elements, and so we may write equation (11)where g,i is a function of the same properties of substituent i as isfi. If we now restrict ourselves to carbon or hydrogen substituents bonded directly to the cr-carbon, and assume that all a-substituents of a given type have the same Ca-substituent bond length and are equidistant from the C,-N axis, equation (11) becomes (12) where nH and nc == nHg?'H + ncgamp; (12) are the number of hydrogen and carbon substituents, respectively. We now further assume that the differ- ences in normal modes among structurally similar nitro- alkanes are unimportant for purposes of estimating the RrVS terms.The vibructional contributions to B(nx*), denoted by B(nx*),for such nitroalkanes may then be written as (13)where B, represents the purely vibrational B(nn*) = BV(nx*) -+ ~vs(nx*) = Bo + nHgH + nCgC (13) contribution, which is assumed to be constant, and gH and gc depend only upon the type of substituent atom. Since nH + nc = 3 for the nitroalkanes, we may simply write (14)where Po and PI are constants and the B, term B(nn*)= Po + ncP1 (14) in equation (13) has been absorbed into Po. Equation (14) predicts that, subject to the assump- tions stated above, the vibructional contribution to the B value associated with the n+x* transition of a nitroalkane should be a linear function of the number of 8-carbons in the molecule.If these vibructional con- tributions are important factors in determining the total B values, one would thus expect to observe a general correlation between the degree of a-substitution in a 911 nitroalkane and the total B value associated with its n+x* transition. When the data are arranged so that the compounds are divided into classes on the basis of their degree of a-substitution (Figure 3), it is evident that such a correlation exists. For example, the B values (in iso-octane) of nitromethane (1), nitroethane (2), 2-nitropropane (6), and 2-methyl-2-nitropropane (10)are 22.9,21.7,9.5, and -9.7, all x respectively.This trend towards more negative values of B(nx*)with increasing substitution at the a-carbon appears to be quite general; the only apparent exceptions are the nitroadamantanes (9) and (11). In these instances the rigidity of the adamantane skeleton probably renders + (9) 0 1 2 3 No.of d carbon substltuents FIGURE3 B Values of a series of nitroalkanes plotted as a function of the degree of a-substitution. The solid line gives the Bvalues predicted by equation (14) when the values for (1)and (10) are used in evaluating the constants Po and PI invalid the assumption of constant vibrational contri- butions to B(nx*). Once the classification of the nitroalkanes on the basis of their degree of a-substitution has been accom- plished, our model assumes that the variance in B values within a class is accounted for by non-a-vibructional effects and by the purely static terms in equation (3).In the analysis of the m.c.d. of saturated ketone^,^,^.^ we could further refine the model by using information obtained from the spectrum of formaldehyde about the purely vibrational contributions to B(nx*). Unfortun-ately, the prototypical nitroalkane, HNO,, is unobtain- able, and so no such analysis is possible here. A further complication appears when we attempt to quantify the effects of static (structural) perturbations. A protocol for doing this was developed for saturated ketone^,^ but again such an analysis is not possible for the nitro- alkanes. This difficulty arises from the fact that, due to the assumed free internal rotation about the C,-N bond, no structural perturbations of a pure symmetry type can be isolated.All the nitroalkanes thus exhibit static perturbations (whether electrostatic or mesomeric in nature) belonging to all three nontotally symmetric represent ations of the chromophoric point group. Furthermore, any non-a-vibructional effects, which are associated with the totally symmetric parts of the static perturbations, are for the same reason inseparable from the purely structural effects. About all that can be said then is that, all other things being equal, there should be a general correlation of m.c.d. intensity within a class with the total number of extrachromophoric atoms, and that the contribution of an extrachromophoric atomic perturber should decrease with increasing distance of the atom from the nitro chromophore.An example of this type of trend can be seen in the data for the primary nitroalkanes in iso-octane solution : the B values for nitroethane (2), l-nitropropane (3), l-nitrobutane (a), and 1-nitropentane (5)are 21.7, 16.8, 14.2, and 14.1, all x lo-', respectively. When we consider the secondary and tertiary nitro- alkanes, the assumption of constancy of BV within a class becomes considerably more tenuous. This has already been alluded to above with regard to the nitro- adamantanes relative to the other compounds. More-over, there may be a significant degree of conformational mobility in (7) and (8),*and there is some evidence for steric hinderance of free internal rotation of the nitro chromophore in (12) (see Appendix).In view of these considerations, further attempts at analysis of the m.c.d. spectra of the secondary and tertiary nitroalkanes are unwarranted outside of a considerably more detailed framework than is used in this work. Conclusions.-The near U.V. m.c.d. spectra of nitro- alkanes are most likely due to two electronic transitions. These give rise to a a+x* band at ca. 320 nm and an n+n* band at ca. 270-280 nm. The B values of the o+x* band are small; those of the n+n* band are usually larger in magnitude and vary with molecular structure in a manner that is at least qualitatively explicable. In particular, the observed trend toward more negative B values with increasing degree of sub-stitution at the a-carbon can be attributed to vibruc-tional effects, in which static (structural) perturbations interact with vibrational perturbations so as to alter the efficacy of the vibronic coupling through which the n+n* transition gains m.c.d.intensity. APPENDIX Internal Rotational Barriers in Nitroa2kanes.-The in-ternal rotational barriers in nitromethane and nitroethane have been measured by microwave spectroscopy and are ca. 0.01 kcal m01-1.27y28 These molecules can therefore be considered to exhibit free rotation about the C-N bond * The population ratio of axial to equatorial conformers in (7)is 20 : the populations for (8) are not known J.C.S. Perkin I1 at room temperature.Similarly low barriers of from 0.07 to 0.10 kcal mol-1 have been calculated for trifl~oro-,~~,~~ tribrorno-,29 and trichloro-nitr~rnethane,~~and it seems reasonable to conclude that an analogous situation exists for those nitroalkanes in which the nitro group is bonded to a tertiary carbon atom, e.g. compounds (10)-(12). Com-pound (12) could well be an exception however, since the Dreiding model shows that the t-butyl group in the 2- position interferes sterically with the rotation of the nitro group. No experimental data are available on the internal rotational barriers in compounds in which the nitro chromo- pliore is attached to a secondary carbon atom but, in view of the data on nitroethane,28 which indicate that no increase in the barrier results from the replacement of one hydrogen atom of nitromethane by a methyl group, we assume that the secondary nitro compounds can also be considered to exhibit essentially free rotation about the C-N bond.We thank Professor H. Feuer, Purdue University, for a gift of several nitroalkanes. This work was supported in part by the National Science Foundation, the National Institutes of Health, and the ' Stiftung fur Stipendien auf dem Gebiete der Chemie, Switzerland. ti11162 Received, 23rd June, 19781 REFERENCES 1 Part 54, R. E. Linder, R. Records, G. Barth, E. Bunnen-berg, C. Djerassi, B. E. Hedlund, A. Rosenberg, E. S. Benson, L. Seamans, and A. Moscowitz, Analyt. Biochem., 1978, 90, 474. G.Barth, E. Bunnenberg, and C. Djerassi, Chem. Comm., 1969, 1276. 3 G. Barth, E. Bunnenberg, C. Djerassi, D. Elder, and R. Records, Symp. Faraday SOL, 1969, 3, 40. L. Seamans and A. Moscowitz, J. Chem. Phys., 1972, 56, 1099. 5 L. Seamans, A. Moscowitz, G. Barth, E. Bunnenberg, and C. Djerassi, J. Amer. Chem. SOC.,1972, 94, 6464. 6 R. E. Linder, E. Bunnenberg, L. Seamans, and A. Mos-cowitz, J. Chem. Phys., 1974, 60, 1943. R. N. Totty and J. Hudec, Chem. Comm., 1971, 1533. L. Seamans, A. Moscowitz, R. E. Linder, K. Morrill, J. S. Dixon, G. Barth, E. Bunnenberg, and C. Djerassi, J. Amer. Chern. SOC.,1977, 99, 724.* R. E. Linder, K. Morrill, J. S. Dixon, G. Barth, E. Bunnen-berg, C. Djerassi, L. Seamans, and A. Moscowitz, J. Amer. Chem.SOC.,1977, 99, 727. lo K. Morrill, R. E. Linder, E. M. Bruckmann, G. Barth, E. Bunnenberg, C. Djerassi, L. Seamans, and A. Moscowitz, Tetra-hedvon, 1977, 83, 907. l1 J. S. Dixon, R. E. Linder, G. Barth, E. Bunnenberg, C. Djerassi, L. Seamans, and A. Moscowitz, Spectroscopy Letters, 1977, 9, 777. l2 For a review see G. H. Wagniere and C. N.,R. Rao in ' The Chemistry of the Nitro and Nitroso Groups, ed. H. Feuer, Interscience, New York, 1969. l3 G. Barth, J. H. Dawson, P. M. Dolinger, R. E. Linder, E. Bunnenberg, and C. Djerassi, Analyt. Biochem., 1975, 65, 100. l4 P. J. Stephens, R. L. Mowery, and P. N. Schatz, J. Chem. Phys., 1971, 55, 224. l5 R. T. Alpin, M. Fischer, D. Becher, H. Budzikiewicz, and C. Djerassi, I. Amer. Chem. SOC.,1965, 87, 4888.l6 (a) N. Kornblum, R. J. Clutter, and W. J. Jones, J. Amer. Chem. SOC.,1956, 68,4003; (b) N. Kornblum, B. Taub, and H. Ungnade, ibid., 1954, '96, 3209. l7 G. W. Smith and H. D. Williams, J. Org. Chem., 1961, 26, 2207. A. Balasubramanian and C. N. R. Rao, Spectrochim. Ada, 1962,18, 1337. l9 A. Fratiello, R. E. Schuster, G. A. Vidulich, J. Bragin, and D. Liu, J. Amer. Chem. SOC.,1973,95, 631. 2o K. L. McEwen, J. Chem. Phys., 1960, 32, 1801. 21 L. E. Harris, J. Chem. Phys.,1973, 58, 5615. 1979 22 C. Djerassi, H. Wolf, and E. Bunnenberg, J. Amer. Chew SOC.,1963, 85, 2835. 23 G. Snatzkc, J. Chern. SOC., 1965, 5002. 24 P. A. D. deMaine, M. M. deMaine, and A. G. Goble, Trans. Faraday SOC.,1957, 53, 427. 25 P. A. D. deMaine, M. M. deMaine, A. A. Briggs, and G. E. McAlonie,J. Mol. Spectroscopy, 1960, 4, 398. 26 W. F. Trager and A. C. Huitric, J. Org. Chem., 1965, 30, 3257. 27 E. Tannenbaum, R. D. Johnson, R. J. Myers, and W. D. Gwinn, J. Chem. Phys., 1954, 22, 949. 28 J. Ekkers, A. Bauder, and H. H. Giinthard, Chem. Phys. Letters, 1973, 22, 249. 29 M. I. Dakhis, A. A. Levin, and V. A. Shlyapochnikov, J. Mol. Structure, 1972, 14, 321. 30 W. M. Tolles, E. Tannenbaum-Handelman, and W. D. Gwinn, J. Chern. Phys., 1965, 43, 3019.