Carbon chemical shifts steric interactions

Further doubt about the validity of the original Grant-Cheney model was expressed by Seidman and Maciel (185), whose INDO calculations of proximity effects in hydrocarbons revealed that there is no simple correlation between carbon chemical shifts and calculated electron-density increases caused by steric C-H bond polarization they report the conformational relation of interacting bonds and groups to be at least equally important, if not more so (185). 

Olah and Watkins (187) correlated l3C chemical shifts in crowded phenyl-ethanes with bond-electron polarizations brought about by van der Waals interactions. They found that these effects cannot be confined to one single C7-H bond but operate throughout the whole molecule and produce shielding of ortho and deshielding of a- and meta carbon atoms. The para carbon atoms are unaffected, which is taken as evidence that only the o-electron systems of the phenyl groups are involved in these steric interactions (187). 

Certain value judgments not built into the rules need to be applied. Thus, the envelope conformation accepted for the D-ring (43) suggests that a y-interaction should be allowed for the calculation of C-15 and C-18 but not for C-16 since the hydrogens on the latter are tilted away from C-18. Similarly, there is a gauche interaction between C-18 and C-20 in the side chain which influences the chemical shift of both carbons. It is likely that the difficulties at C-17 and C-21 arise from steric interactions which are difficult to account for without assigning a specific conformation to the side chain. The result at C-10 is more difficult to rationalize. 

Although proton chemical shifts are influenced significantly by factors other than electron density, they also reflect the polarization of the enamine framework and the degree of n,n interaction. Thus, the chemical shifts of the vinylic protons are modulated by the same factors discussed for the chemical shifts of the corresponding olefinic carbons, such as amine component, steric and electronic effects of the substituents and ring size effects. In particular, the chemical shift of the proton(s) at C(2) is lowered by increasing njt interaction, in parallel with what has been observed for < C(2). No general correlation exists between the chemical shifts of both nuclei probably as a consequence of their different sensitivity to steric, electronic and, particularly, anisotropic effects of the substituents. Nevertheless, for sets of structurally related compounds, reasonable linear correlations can be found between <5C(2) and <5H(2) (see below). Since the XH-NMR data available for enamines are more abundant than those for 13C and 15N, more complete structural information can be obtained for wider sets of compounds. 

Tables 3-1 through 3-3 contain data on which empirical calculations of chemical shifts can be made. The tables represent a fraction of the data available on the fundamental alkane, alkene, and aromatic structures. Moreover, corrections must be applied in order to avoid nonadditivity caused primarily by steric effects. Thus, three groups on a single carbon atom, two large groups cis to each other on a double bond, or any two ortho groups can cause deviations from the parameters listed in the tables. If sufficient model compounds are available, the corrections shown can be applied. Further empirical calculations are possible for any structural entity, so that the eclipsing strain in cyclobutanes, the variety of steric interactions in cyclopentanones, or the variations in angle strain in norbornanes may be taken into account.

The same intramolecular Lewis base - Lewis acid interaction can be observed when a chlorophosphane is used instead of a fluorophosphane. However, the chloride is less strongly bonded than fluoride, resulting in the displacement of chloride by the phosphane without the use of an auxiliary Lewis acid. The chemical shift of the tricoordinate phosphorus atom is sensitive to the steric bulk of its carbon substituent. Evidently, sterically demanding substituents like tert-butyl hinder the 2T-bonding interaction from nitrogen, resulting in the observed downfield shift. 

We feel that attempts to rationalize chemical shifts for coordinated olefinic and acetylenic carbons in terms of any one factor are not justified. Indeed simple explanations are unlikely to be valid and any true interpretation of shifts will of necessity be nonsimple. Some basic statements do, however, appear reasonable (though unpredictive). For example, the strength of the coordinative interaction will play a role in determining AS (shift on coordination). Factors which affect this interaction, whether they be steric or electronic, will affect the resonances. The nature of the coordinative interaction (the relative orbital energies, the specific metal involved, the electronic... 

However, the data of direct quantum chemical calculations on the PES of this reaction indicate strong steric repulsion in the case of the supra-antara approach I, which makes this mechanism energetically unfavored [2, 3]. But the route II of the parallel approach with the synchronous formation of two bonds C—C is equally energetically unrealizable. The MINDO/3 calculations with precise localization of the transition state by minimization of the gradient norm lead to the structure III. A three-center Cj—C2—C3 interaction, predictable from the perturbation theory [5], takes place in this structure. The form of the transition vector III reflects the character of the carbon atom shifts which determine the reaction path where the processes of breaking and making of the C—C bonds are sharply asynchronous. The ab initio calculations [3, 6] lead to the same conclusion, they indicate a transition state with the structure IV in which two CC bonds lying in parallel planes are spaced 2.237 A apart. 

Table 2 summarizes observed and calculated NMR shifts of a variety of substituted phenylcalcium iodides in THF. Although the additivity of such empirical increments is an oversimplification, neglecting steric and electronic substituent interactions, the calculated values show acceptable deviations from the observed chemical shifts. Larger differences were observed for the ipso carbon of sterically crowded derivatives such as [(Mes)Cal(thf)4] [93] as well as for compounds where quinoid mesomeric forms might contribute to the overall structure like in case of [(p-Me2N-C6H4)Cal(thf)4] [84]. 

---