Resonance effect constant

This work already showed that substituent constants of one reaction can only be transferred to another reaction when similar effects are operating and when they are operating to the same extent. In order to find a broader basis for the transfer-ability of substituent constants, they were split into substituent constants for the resonance effect and those for the inductive effect. 

The quatemization reaction of the thiazole nitrogen has been used to evaluate the steric effect of substituents in heterocyclic compounds since thiazole and its alkyl derivatives are good models for such study. In fact, substituents in the 2- and 4-positions of the ring only interact through their steric effects (inductive and resonance effects were constant in the studied series). The thiazole ring is planar, and the geometries of the ground and transition states are identical. Finally, the 2- and 4-positions have been shown to be different (259. 260). 

If resonance effects a/one are considered, it is possible to make some sense of the ranking of various propagation constants. 

Assignment of the signals is completed in Table 30.2. The criteria for assignment are the shift values (resonance effects on the electron density on C and N), multiplicities and coupling constants. Because the difference between them is so small, the assignment of N-8 and N-9 is interchangeable. 

A more ambitious goal is to separate completely resonance effects from polar effects. This involves using separate substituent constants to account for resonance and polar effects. The modified equation, called a dual-substituent-parameter equation, takes... 

In general, the dissection of substituertt effects need not be limited to resonance and polar components, vdiich are of special prominence in reactions of aromatic compounds.. ny type of substituent interaction with a reaction center could be characterized by a substituent constant characteristic of the particular type of interaction and a reaction parameter indicating the sensitivity of the reaction series to that particular type of interactioa For example, it has been suggested that electronegativity and polarizability can be treated as substituent effects separate from polar and resonance effects. This gives rise to the equation... 

If the two interaction mechanisms maintain the same relative importance throughout the reaction series, that is, if the ratio pj/pi is constant, Eq. (7-26) becomes identical in form with Eq. (7-24). This in fact appears to be the usual situation in Hammett correlations, PP- and the two interaction mechanisms are commonly discussed in terms of the inductive and resonance effects of electronic displacement. 

A further complication arises with Ingold s suggestion" that both the inductive and resonance effects are composed of initial state equilibrium displacements that reveal themselves in equilibrium properties like dipole moments and equilibrium constants and of time-dependent displacements produced during reaction by the approach of an attacking reagent, observed rate effects being resultants of both types of electronic effects. Hammett, however, claims that it is not necessary or possible to make this distinction. 

Many other definitions of an ortho substituent constant have been made Shorter has reviewed these. Charton analyzed Oo in terms of Oi and CTr, i.e., = a(Ti -I- fpoR, finding that the distribution of inductive and resonance effects (the ratio a/b) varies widely with the substituent and, therefore, that no general Oo scale is possible. Charton also subjected to analysis according to Eq. (7-47),... 

A more quantitative formulation of the varying resonance effects in electrophilic nuclear substitution reactions has been suggested by Tsuno, who has proposed to use Eq. (2), where Aa+ is a resonance exaltation term, and r is a susceptibility constant. 

Few other reactions of series of substituted pyridines have been investigated extensively. Dondoni, Modena, and Todesco have measured the rate of N-oxidation of a limited series of pyridines and found a good correlation with normal u-values with a p-value of — 2.23. The A-alkylation of pyridines with alkyl iodides in nitrobenzene has been studied by Brown and Cahn and by Clarke and Rothwell. Unfortunately, the only data available are for the parent compound and for alkyl derivatives, and, since the a-values for the various alkyl groups in a given position are substantially constant, this leaves a correlation of only three independent points. However, the rates of A-alkylation of the j8- and y-alkyl derivatives are so nearly equal that it appears as if no correlation existed. Clarke and Rothwell have also studied the alkylation with allyl bromide in nitromethane at various temperatures, and in this case a more extensive series is available. The authors state that no overall Hammett correlation is obtained however, the j8-substituted derivatives fall on one straight line and the y-derivatives on another one with a different slope. The data are shown in Fig. 2. The line for the j8-compounds, p = — 2.53 0.31, r = 0.95, is seen not to be very good the line for the y-derivatives, p = — 1.42 0.06, r = 0.99, is much more satisfactory. It does not seem likely that the discrepancy is due to the intervention of resonance effects, since in this case one would expect the correlation for the y-derivatives to be poorer than that for the j8-analogs. More extensive studies with a wider variety of substituents would seem very desirable. 

The least squares value for the p constant obtained by this procedure is +6.2 it wiU be obviously subject to change as more meta and epi substituents become available. Only the cata-NO group was excluded from the above plot because it causes a strongly enhanced resonance effect in nucleophilic substitution (Section IV,C, l,a) and an anomalous effect of uncertain origin in the dissociation of carboxylic acids. It can be assumed that the reaction constant for 4-chloro-... 

B. The Separation of Inductive and Resonance Effects Substituent Constants from Spectroscopic Studies. 

However, the duality of substituent constants and the attempt to deal with crossconjugation by selecting cr+, a or a in any given case is somewhat artificial. The contribution of the resonance effect of a substituent relative to its inductive effect must in principle vary continuously as the electron-demanding quality of the reaction center is varied, i.e. whether it is electron-rich or electron-poor. A sliding scale of substituent constants would be expected for each substituent having a resonance effect and not just a pair of discrete values a and a for — R, or o and a for + R substituents55. 

Thus, a values for XCH2 groups may be defined free of resonance effects. Values of a for groups other than XCH2 will contain a resonance effect if they are defined from the ionization constants of carboxylic acids. 

The Eo values for 2-substituted 1,4-benzoquinones (sets 45-4 through 45-7, 45-10) show an average value of pr of 59. Thus the resonance effect predominates. For most of these sets, the Op constants are not the best parameters for correlation. By contrast, the electron reduction potentials (set 45-8) show a Pr value of 39, which indicates predominance of the localized effect. The 2,5-disubstituted 1,4-benzoquinones differ distinctly in their behavior from the 2-substituted 1,4-benzoquinones in that they show an average Pr value of 53. The one-electron reduction potentials of these compounds show about the same composition of the electrical effect, with a value of Pr of 50. The only set of Eq values available for the 2,6-disubstituted 1,4-benzoquinones pve a Pr value of 51, comparable to the values observed for the 2,5-disubsti-tuted 1,4-benzoquinones. The 2,3,5,6-tetrasubstituted 1,4-benzoquinones have... 

Finally, in this account of multiparameter extensions of the Hammett equation, we comment briefly on the origins of the a, scale. This had its beginning around 1956 in the a scale of Roberts and Moreland for substituents X in the reactions of 4-X-bicyclo[2.2.2]octane-l derivatives. However, at that time few values of o were available. A more practical basis for a scale of inductive substituent constants lay in the o values for XCHj groups derived from Taft s analysis of the reactivities of aliphatic esters into polar, steric and resonance effects . For the few o values available it was shown that o for X was related to o for XCHj by the equation o = 0.45 <7. Thereafter the factor 0.45 was used to calculate c, values of X from o values of XCH2 . ... 

Because of the scarcity of electronic paramagnetic resonance data, and because of the frequent unreliability of the data from paramagnetism, boiling point elevation, spectrophotometry, and ortho-para hydrogen conversion, most published radical dissociation constants can be accepted only with reservations. An error of 50 % is not at all improbable in many cases. We are therefore not yet in a position to explain, or rather to test our explanations of, small differences in dissociation constants. Table I shows the values of K corresponding to various hexaarylethanes in benzene at 25°. Because of the order of magnitude differences in Table I, however, it is likely that some of the expected large effects, such as steric and resonance effects, exist. 

The effect of a substituent on the aromatic substitution reaction is similar to its effect on electrophilic side chain reactions, but not precisely parallel. Thus the Hammett relationship using the usual sigma or substituent constants gives considerable scatter when applied to aromatic substitution. The scatter is probably due to an increased importance of resonance effects in the nuclear substitution reaction as compared with the side chain reactions. 

---