Malononitrile proton transfer with

Fig. 5. Rates of proton transfer for cyanocarbon acids. Open and closed points represent forward (log10 k, ) and reverse (logI0 -i ) rate coefficients, respectively, and ApK is the statistically corrected difference in acidities between the cyanocarbon acid and base (B). Points are o and for f-butylmalononitrile reacting with carboxylate ions and H20 a and for malononitrile with formate ion and H20 and for bromomalononitrile with phosphate ion and morpholine x for 1,4-dicyano-2-butene with phenolate ions. Redrawn with permission from F. Hibbert and F. A. Long, J. Am. Chem. Soc., 94 (1972) 2647. Copyright by the American Chemical Society.

For the malononitriles and l,4-dicyano-2-butene, very low primary isotope effects are observed (Table 3). For 2-methyl-3-phenylpropionitrile reacting with methoxide ion in pure methanol, fcH/fcD = 1.15 [119]. Isotope effects [121] on proton transfer from ketones [89] and nitro-paraffins [5] are much higher than these values. It will now been shown that the low primary isotope effects observed for cyanocarbons support [113] the conclusion [64] reached from the high Bronsted exponents that proton transfer occurs through a transition state in which the proton is almost fully transferred. The equilibrium isotope effect (KH /KD = Kr /l) on the ionization of malononitriles (80)... 

The importance of the low isotope effects which have been observed for proton transfer from cyanocarbons has been emphasized [125]. Until recently it has been assumed that a rate-determining proton transfer would give rise to an appreciable isotope effect. The results for malononitriles show that this is not always the case and small isotope effects can be expected when the transition state in a proton transfer reaction is strongly reactant-like or product-like and this has been confirmed by theoretical calculation [126]. Inverse primary isotope effects (kH/kD < 1) may even be possible with a product-like transition state when the equilibrium isotope effect is in this direction [126]. The observation of an appreciable kinetic isotope effect for a reaction indicates that a hydrogen transfer is involved, but a small isotope effect does not prove the contrary and this may have led, in the past, to errors in the interpretation of... 

A number of other proton transfer reactions from carbon which have been studied using this approach are shown in Table 8. The results should be treated with reserve as it has not yet been established fully that the derived Bronsted exponents correspond exactly with those determined in the conventional way. One problem concerns the assumption that the activity coefficient ratios cancel, but doubts have also been raised by one of the originators of the method that, unless solvent effects on the transition state are intermediate between those on the reactants and products, anomalous Bronsted exponents will be obtained [172(c)]. The Bronsted exponents determined for menthone and the other ketones in Table 8 are roughly those expected by comparison with the values obtained for ketones using the conventional procedure (Table 2). For nitroethane the two values j3 = 0.72 and 0.65 which are shown in Table 8 result from the use of different H functions determined with amine and carbon acid indicators, respectively. Both values are roughly similar to the values (0.50 [103], 0.65 [104]), obtained by varying the base catalyst in aqueous solution. The result for 2-methyl-3-phenylpropionitrile fits in well with the exponents determined for malononitriles by general base catalysis but differs from the value j3 0.71 shown for l,4-dicyano-2-butene in Table 8. This latter result is also different from the values j3 = 0.94 and 0.98 determined for l,4-dicyano-2-butene in aqueous solution with phenolate ions and amines, respectively. However, the different results for l,4-dicyano-2-butene are to be expected, since hydroxide ion is the base catalyst used in the acidity function procedure and this does not fit the Bronsted plot observed for phenolate ions and amines. The primary kinetic isotope effects [114] also show that there are differences between the hydroxide ion catalysed reaction (feH/feD = 3.5) and the reaction catalysed by phenolate ions (kH /kP = 1.4). The result for chloroform, (3 = 0.98 shown in Table 8, fits in satisfactorily with the most recent results for amine catalysed detritiation [171(a)] from which a value 3 = 1.15 0.07 was obtained. 

The small isotope effects observed in proton transfer from cyanocarbon acids to various bases shown in Table 3 (for example feH/feD = 1.46 for proton transfer from malononitrile to water) are compatible with an extremely product-like transition state in which the proton is almost fully transferred [113] (Sect. 4.3). Similar conclusions may be reached from the small isotope effects observed for chloroform (feH/feD = 1.41 0.01 [114] and 1.11 0.05 [171]) and phenylacetylene (kH/kD = 0.95 0.09 [143]) for reaction with hydroxide ion, and for reaction of disulphones with water (feH/feD = 2.2 0.1 [65]). In all these cases the magnitude of the Bronsted exponent is close to the limiting value of unity as expected for a product-like transition state. 

A mechanism for the piperazine-catalyzed formation of 4//-chromenes is complex cascade of reactions, starting with piperazine acting as a base which activates malononitrile, promoting Knoevenagel condensation, and also formation of an enamine, followed by Michael condensation, proton transfer, intermolecular cycliza-tion via a nucleophilic addition of the enolate oxygen to the nitrile group (hetero-Thorpe-Ziegler), and finally hydrolysis and tautomerization. 

---