6715
O n the Conformation of /+Substituted Ethyl Radicals’ D. Griller2 and K. U. Ingold*
Contribution from the Division of Chemistry, National Research Council of Canada, Ottawa, Canada. Received October 5 , 1973 Abstract: A large number of l,l-di-rert-butyl-2-substitutedethyl radicals, B2CCH2MR,, have been prepared, principally by radical addition to 1,I-di-terr-butylethylene. The epr spectral parameters for these sterically hindered radicals are compared with those of the analogous, but unhindered, 2-substituted ethyl radicals, CH2CH2MR,. In B2CCH2MR,all MR, groups eclipse the orbital occupied by the unpaired electron (i.e.,the C, p. orbital). This same conformation (2) is adopted by CH2CH2MR,radicals when M is from rows 2 , 3, or 4 of the periodic table, but when M is from row 1the preferred conformation (1) is that with the MR, group lying in the nodal plane of the C, pl orbital. Since the calculated extents of hyperconjugative delocalization of unpaired spin density into the CB-M bond are about the same in B2CCH2MR,whether M is carbon or silicon, it is concluded that the difference in conformation produced by these two elements in CH,CH2MR, radicals cannot be due to hyperconjugation. Some other interaction, such as pn-dr bonding, must be invoked to account for the preference for conformation 2 with row 2 elements. Although P-hydrogen splittings serve to distinguish between conformation 1 and 2, it is suggested that they may not be a reliable guide to possible small distortions of the 0carbon from its normal tetrahedral geometry. Other conclusions are that B2CCH2C6H5 and analogous radicals have the phenyl ring arranged with one edge towards the unpaired electron and that all BrCCH?MR,radicals have planar ligand geometry at the a carbon.
T
he conformations adopted by 0-substituted ethyl radicals, R,MCH2CH2, are determined by measurement of the isotropic epr hyperfine splittings due to the p hydrogens, aHB. These splittings depend on the angle 0 between the principal axis of the p orbital containing the unpaired electron and the C-H bond on the pcarbon atom and can be represented by the empirical relationship
a%
=
Bo -k B cos2 0
a”, N 15 G. For example, with MR, 17.68 G a t -70”.6
=
SiH3, uHp =
MR,,
*‘I\
I
(1)
where Bo and B are constants (Boe 3 f 2 G, B N 48 i 5 G ) . 3 Under conditions where rotation about the C,-CB bond is rapid the average value of cos2 6 is 0.5 and aHos 27 G . For many alkyl radicals the most stable conformation (i.e., that approached at low temperatures) is 1, with 0 tending toward 30” and u H b toward 39 G as the temperature is lowered. For example, for the propyl radical (MR, = CH3) aHo N 29 G at -20” and shows a steady increase with decreasing temperature with a value of =33 G a t - 140”.4 However, when M is from rows 2, 3, or 4 of the periodic table conformation 2 is favored5-18 with 6 60” and
-
(1) Issued as N R C C No. 14257. (2) N R C C Postdoctoral Fellow, 1973-1975. (3) (a) P. B. Ayscough, “Electron Spin Resonance in Chemistry,” Methuen, London, 1967; (b) H. Fischer i n “Free Radicals,” Vol. 11, J. I' is the observed coupling and aOalis the coupling constant for M with unit unpaired electron density in its valence shell s atomic orbital. 3 4 , 3 6 With BsCCH2CF3, BsCCHsCCl3, and B2CCH2SiMe3 (all of which adopt conformation 2) we calculate that pu 0.108, 0.122, and 0.106, respectively. Thus, in conformation 2 the extents of hyperconjugative delocalization towards CF3, CCI3, and SiMe3 are very similar.:IG However, when the steric constraint of the =T
(33) K.S. Chen and J. K.Koclii, private communication. (34) Although approximate sp' hybridization at M is assumed in this equation, fairly large changes in the s : p ratio \vould not substantially alter our conclusions. (35) Values of mJ1 (ignoring differences in g factor) were taken from J. E. Wcrtz and J. R . Bolton, "Electron Spin Resonance," McGrawHill, New 'fork, N. Y., 1972. (36) Unfortunately, the most obvious comparison between carbon and silicon was not obtained since the B&CH?B radical was not formed when B . \\as generatt'd in the presence of B2C=CH2 (presumably either for steric reasons or because B . is insufficiently electrophilic). However, \\e feel that the SiMes adduct should, in any case, be compared with the CFi or C C h adducts since a t least some of the factors that are believed to promote hyperconjugation to a n SiMea group should also enhance hyperconjugation to CF? or CC1s relative to CMe3. That is, when R is more electronegative than M, the M-R bonds will adopt more ?r character and the Cp-M bond more u character. As a consequence, the Cp-H bonds have increased ?r character and the hyperconjugative canonical structure B?C=CH? ' M R , becomes more favored.37 This suggests that comparisons between different M should invol~ e groups M here the electronegativity difference between M and R (X\I - XR) is similar. Values of (x31 - X R ) 3 n are: SiMes = -0.7, CMe' = 0, CCli = -0.55, and CFa = - 1 . 3 . That is, in our opinion,xg the p o value found for BZCCHzSiMe3 should, indeed be compared with the values found for B?CCH?CFs and B?CCH;.CCh. However, the fact that electronegativity differences do not wholly determine p o is indicated by BpCCH,SiCI? for \vhich p u = 0.255 and ( x u - XR) = - 1.25. (37) As a corollary, hyperconjugation in B&H?MR, and CH2CH2MR, will be favored for nonplanar . M R , radicals (see introductory
Journal of the Ailrerican Chemical Societ!, / 96:21
tert-butyl groups is removed, the a H p splittings indicate that for CH2CH2CF3free rotation occurs about the C,-CB bond at ambient temperatures and the slight increase in the splitting with decreasing temperature (see Results) indicates that conformation 1 is preferred. This conformation (or one close to it)41is also preferred by .CH2CH2CCI3 (aHa = 22.3 G from -20 to - 160"). In contrast, . CH2CHySiMe310 and . CH2CH2SiEt3j adopt conformation 2. Hence, hyperconjugation cannot be responsible for the different conformational preferences when M = carbon or silicon4?in unhindered P-substituted ethyl radical^^^*^^ and some other interaction (possibly pT-dT b ~ n d i n g ) ~must , ~ ~ be invoked to explain the preference for conformation 2 when M is from row 2, 3, or 4 of the periodic table. In radicals such as .CH2CH2CIand .CH2CH2PR2(in which M has lone pairs of electrons) pn-pn homoconjugative bonding seems to be fairly well a c ~ e p t e d , ' " ~ (see, ~ . ~ ~however, following section). a% Values and Conformational Assignments. The common interpretation of aF1p simply in terms of the angle 0 between the p orbital containing the unpaired electron and the Cp-H bonds is extremely valuable in distinguishing between conformations 1 and 2. However, it is possible that more subtle conformational effects are not so easily identified. If the ligands on the /3 carbon of a 8-substituted ethyl radical are arranged tetrahedrally, then the minimum value of a H 8 expected on the basis of eq 1 is ca. 15 G (see introductory statements). ** Although a slightly smaller minimum v-alue may obtain for B2cCH2MR,,33 the actual values of aHo for many of these radicals are considerably less than 15 G. Unexpectedly low values of 0% have normally been interpreted in terms of a bridging structure5' 1 5 , 4 7 involving movement of M towards the 2p, orbital of the CY carbon. As a consequence, the P hydrogens are displaced away from their tetrahedral position towards the nodal plane of the unpaired electron orbital so that their interaction with this electron decreases. statements) and . M R , are nonplanar when R is morc electronegative than M.4O That is, hyperconjugation will he favored for MR, = CFJ, CC13, and SiMea (relative to CMe3) and the hybridization of M in the adduct radicals34 should be similar for these three groups. (38) M. L. Huggins,J. Amer. Cheni. Soc., 75,4123 (1953). (39) A diEerent interpretation of our results has been ofrered by Professor M.C. R. Symons (private coinniunication). (40) For leading references see rsf 3b; and also, J. Cooper, A . Hud, 209 (1972); R. V. Lloyd and son, and R . A. Jackson, Mol. P h ~ s . 23, M. T , Rogers, J . Aixer. Chem. Soc., 95, 1512, 2459 (1973); L. I
