Nitroalkanes, proton-transfer

Sodium hydroxide has been the most commonly used base in experimental nitroalkane proton transfer reaction studies.However, the computational studies of these reactions have generaUy been with hydroxide ion without the sodium counter ion. Recently a computational study of the proton transfer reactions of the three simple nitroalkanes in the presence of NaOH in water has been carried out and it was found that the presence of Na had an enormous effect on the energetics of the reactions. Double potential energy well diagrams, much like those found for the Sn2 reactions, were recorded for the proton transfer reactions of NM, NE and 2-NP with hydroxide ion in water. The computations included two explicit water molecules in the water cavity. The Gibbs free energies and enthalpies observed for the reactant complex (CPI), the TS and the product complex (CP2) both in the presence and absence of sodium ion and two explicit water molecules are summarized in Table 1.24. 

Superoxide anion formed in situ in a solution exposed to air (i.e. with only a small concentration of O2) has been used as an EGB to generate nitroalkane anions that may add to activated alkenes or to carbonyl compounds [130, 131]. An example is shown in Scheme 33. The reaction is catalytic since the product anion can act as a base toward the nitroalkane. Using the nitroalkane as the solvent favors the proton transfer pathway over the competing addition of the product anion to a second molecule of activated alkene, a pathway that may lead to polymerization [130]. In some cases, better yields of the Michael addition product were obtained if a stoichiometric amount of the anion was formed ex situ (with O2 as the PB), and the activated alkene added subsequently ]130, 132]. 

We will devote the remaining part of this section to a discussion of the proton transfer reaction of nitroalkanes. 

Without a doubt, the proton transfer reaction which has attracted most attention over the past 15 years is the reaction of a series of nitroalkanes with a base, as indicated in (104) (Bordwell et al., 1969, 1970, 1972). The reason... 

The reason why delocalization is not more advanced is that there are constraints imposed on the transition state that prevent extensive delocalization. This was first pointed out by Kresge51 in the context of the deprotonation of nitroalkanes, but it applies to any proton transfer from carbon. The situation is represented in Equation (9) which is a more nuanced version of Equation (3) and allows for a certain degree of charge delocalization into the jt-acceptor (8Y) at the transition... 

Fig. 1 More O Ferrall-Jencks diagram for the deprotonation of a nitroalkane. The curved line shows the reaction coordinate with charge delocalization lagging behind proton transfer.

Acidity constants and rates of reversible deprotonation of triphenylphosphonium ion (54), (55) and pyridinium ions (56), (57) by amines in water, 50 50 v/v DMSO-water, and 90 10 v/v DMSO-water have been determined.157 The intrinsic rate constants for proton transfer were relatively high for all four carbon acids and showed little solvent dependence. This is in contrast with nitroalkanes, which have much lower intrinsic rate constants and show a strong solvent dependence.158... 

The clearest example of the danger in using a as a measure of transition state structure is illustrated in the work of Bordwell et al. (1969, 1970, 1975). In the rate-equilibrium relationship for the deprotonation of a series of nitroalkanes the unprecedented Br nsted slopes of 1 61 for l-aryl-2-nitropropanes and 1-37 for 1-arylnitro-ethanes were obtained. The simple exposition of the mechanistic significance of a disallows values greater than 1. This, coupled with the fact that the transition state for the proton transfer is not product-like (as established by alternative criteria) indicates at best that, in at least some cases, a does not reflect the selectivity of a particular reaction. Several attempts to rationalize these anomalous results have been made. 

Another acid/base anomaly, the anomalous relationship between rates and equilibria for the proton-transfer reactions of nitroalkanes such as H3C—NO2/H3C— CH2—N02/(H3C)2CH—NO2 found in water, does not exist in the gas phase [244]. Only in aqueous solution is the rate of proton abstraction by HO unexpectedly slower for the more acidic nitro compound. 

The analysis as stated does not allow for values of a and j3 outside the limits of zero and unity. Hence when experimental values for a and (3 were recently obtained which were outside this range [76, 77] considerable doubt was thrown upon the general usefulness of a and (3 as an index of the degree of proton transfer. Detailed discussion of these unusual experimental results will be given in Sect. 4.2, but meanwhile an explanation suggested by Kresge [78] will be described here as it illustrates some of the limitations of the foregoing analysis. Results for the reaction of hydroxide ion with thirteen substituted nitroalkanes in 50 % (v/v) methanol—water (63)... 

Figure 7 Proton transfer from nitroalkanes to base. See text for details of the structure at (1,0).

The numerical values of the forward and reverse Bronsted coefficients in a simple proton transfer should sum to unity. Neither a nor p should exceed unity or be less than zero. Bordwell and his co-workers [29] discovered that in the nitroalkane acid-base system (Eqn. 36) the introduction of electron-donating substituents into R lowered the rate constant for reaction but increased the overall equilibrium constant. 

Current thought is that the anomalous parameters reflect an unusual transition state where there is no monotonic change from ground to product states. Bordwell considers that the anomalies arise from the rehybridisation required as well as the proton transfer in passing from nitroalkane to aci-species (Eqn. 37) [29c]. 

---