Protonation radical anions

Electron transfer reduction of pyridines in both acid and alkaline solution generates the protonated radical-anion. This rapidly accepts a further electron and a proton to give a mixture of dihydropyridines. Enamine structures in these dihydro-pyridines can tautomerise to the imine, which is more readily reduced than the original pyridine molecule. Further reaction of the 1,4-dihydropyridine leads to piperidine while reduction of the t, 2-dihydropyridine leads to a tetrahydropyridine in which the alkene group cannot tautomerise to the imine and which is not therefore reduced to the piperidine stage. The reaction sequence is illustrated for 2,6-dimethyl-pyridine 18 which yields the thermodynamically favoured cis-2,6-dimethylpiperidine in which the two alkyl substituents occupy equatorial conformations. 

There has been some discussion about the detailed mechanism of the reduction of aromatic nitro compounds. In acid solution the slow step in the four-electron reduction has been found [96,98] to be the uptake of the second electron, the reduction of ArN02H to the dihydroxylamine, ArN(OH)2. According to some authors, ArN02H [96,98,99] is formed on protonation of the radical anion, but there are also strong proponents for its formation by electron transfer to a preprotonated nitro compound [100,101]. In water the pKa of the protonated radical anion, ArN02H, for most aromatic nitro compounds, is between 2 and 4 [102]. 

Nitrones have been used to indicate the presence of short-lived radicals, as the addition of a radical to a nitrone forms a radical which is stable enough for an esr-investigation [96]. Reduction of a nitrone as A-/-butylphenylnitrone in DMF yields the Schiff base, which is further reduced. In the presence of a proton donor the protonated radical anion from the Schiff base may add to the nitrone to a stable radical, so it is recommended not to have a proton donor present when using nitrones for investigation of the formation of radicals during a reduction. 

We consider here the detailed consequences of the reduction in aprotic solvents, two sequential steps, of azobenzene and analogous aromatic azo compounds [12-15]. The dianion is considerably more basic than the radical anion, and the dianion is only long-lived in very dry nonacidic solvents [14-16]. Both the dianion and the radical anion derived from azobenzene and substituted azobenzenes have been used as EGBs. The anion resulting from protonation of the dianion is less basic (by several pATa units) than the dianion itself but more basic than the radical anion [15]. Using the dianion as EGB may therefore result in mono- or diprotonation, depending on the strength of the acid. The radical anion leads directly to hydrazobenzene due to the further reduction of the protonated radical anion (Scheme 3) and fast protonation of the more basic anion [pATa(Ph-NH-NH-Ph) =26.1, pATa(Ph-NH-N -Ph) 31, pAa(Ph-NH-N -Ph) < 20]. 

Thymine and uracil behave similar to typical carbonyl compounds, i.e. the first intermediate is a radical anion [reaction (55)] which is in equilibrium with its oxygen-protonated conjugated acid [reaction (57)]. The other functional groups, especially the second carbonyl function, withdraw electron density, and hence the p a values of these oxygen-protonated radical anions are much lower (thymine p a = 6.9 [52]) than those of the corresponding radicals from simple carbonyl compounds. However, the C(6)-protonated isomers [cf. reaction (58)]... 

In the presence of a proton source, the radical anion is protonated and further reduction occurs (the Birch reduction Part B, Section 5.5.1). In general, when no proton source is present, it is relatively difficult to add a second electron. Solutions of the radical anions of aromatic hydrocarbons can be maintained for relatively long periods in the absence of oxygen or protons. 

Two classes of charged radicals derived from ketones have been well studied. Ketyls are radical anions formed by one-electron reduction of carbonyl compounds. The formation of the benzophenone radical anion by reduction with sodium metal is an example. This radical anion is deep blue in color and is veiy reactive toward both oxygen and protons. Many detailed studies on the structure and spectral properties of this and related radical anions have been carried out. A common chemical reaction of the ketyl radicals is coupling to form a diamagnetic dianion. This occurs reversibly for simple aromatic ketyls. The dimerization is promoted by protonation of one or both of the ketyls because the electrostatic repulsion is then removed. The coupling process leads to reductive dimerization of carbonyl compounds, a reaction that will be discussed in detail in Section 5.5.3 of Part B. 

One-electron reduction of a-dicarbonyl compounds gives radical anions known as setnidiones. Closely related are the products of one-electron reduction of aromatic quinones, the semiquinones. Both semidiones and semiquinones can be protonated to give neutral radicals which are relatively stable. 

Nonetheless, the reduction clearly proceeds by initial addition of one electroi to the aj omatic ring and the resulting radical-anion must be protonated ii some way before the second electron addition can occur. 

Obviously the structures and yields of Birch reduction products are determined at the two protonation stages. The ring positions at which both protonations occur are determined kinetically the first protonation or 7t-complex collapse is rate determining and irreversible, and the second protonation normally is irreversible under the reaction conditions. In theory, the radical-anion could protonate at any one of the six carbon atoms of the ring and each of the possible cyclohexadienyl carbanions formed subsequently could protonate at any one of three positions. Undoubtedly the steric and electronic factors discussed above determine the kinetically favored positions of protonation, but at present it is difficult to evaluate the importance of each factor in specific cases. A brief summary of some empirical and theoretical data regarding the favored positions of protonation follows. 

In a recent publication, D. A. Burnham [Tetrahedron, 25, 897 (1969)] presents other calculations which indicate that protonation of the radical-anion of anisole is more likely to occur meta to the methoxyl group rather than ortho as Zimmerman proposes. 

Reduction of a conjugated enone to a saturated ketone requires the addition of two electrons and two protons. As in the case of the Birch reduction of aromatic compounds, the exact order of these additions has been the subject of study and speculation. Barton proposed that two electrons add initially giving a dicarbanion of the structure (49) which then is protonated rapidly at the / -position by ammonia, forming the enolate salt (50) of the saturated ketone. Stork later suggested that the radical-anion (51), a one electron... 

Reversible electron addition to the enone forms the radical anion. Rate determining protonation of the radical anion occurs on oxygen to afford an allylic free radical [Eq. (4b) which undergoes rapid reduction to an allylic carbanion [Eq. (4c)]. Rapid protonation of this ion is followed by proton removal from the oxygen of the neutral enol to afford the enolate ion [Eq. (4c)]. 

When saturated steroidal ketones are reduced in ammonia, an alcohol is usually present to act as a proton donor and high yields of steroidal alcohols are obtained. Under these conditions, reduction probably proceeds by protonation of the radical-anion (or ketyl) (61), which results from a one electron addition to the carbonyl group, followed by addition of a second electron and proton. Barton has proposed that reduction proceeds via protonation of the dianion (62) arising from addition of two electrons to the carbonyl group. This proposal implies that the ketyl (61) undergoes addition of a second electron in preference to undergoing protonation by the... 

Flavin coenzymes can exist in any of three different redox states. Fully oxidized flavin is converted to a semiqulnone by a one-electron transfer, as shown in Figure 18.22. At physiological pH, the semiqulnone is a neutral radical, blue in color, with a A ax of 570 nm. The semiqulnone possesses a pAl of about 8.4. When it loses a proton at higher pH values, it becomes a radical anion, displaying a red color with a A ax of 490 nm. The semiqulnone radical is particularly stable, owing to extensive delocalization of the unpaired electron across the 77-electron system of the isoalloxazine. A second one-electron transfer converts the semiqulnone to the completely reduced dihydroflavin as shown in Figure 18.22. 

The proposed mechanism for the conversion of the furanone 118 to the spiro-cyclic lactones 119 and 120 involves electron transfer to the a -unsaturated methyl ester electrophore to generate an anion radical 118 which cyclizes on the /3-carbon of the furanone. The resulting radical anion 121 acquires a proton, giving rise to the neutral radical 122, which undergoes successive electron transfer and protonation to afford the lactones 119 and 120 (Scheme 38) (91T383). 

Evidence for the radical anion 3 came from esr spectroscopic experiments, thus supporting this mechanism. The radical anion is protonated by the alcohol to give... 

The ESR spectrum of the thioxanthene S, S-dioxide radical anion itself shows that the two possible conformers coexist, since the two methylene protons are not equivalent. In the case of the 9-monoalkyl derivatives, the large coupling constant observed for the 9-proton leads to the conclusion that the 9-substituent is in the boat equatorial position as in II1 F Thus the radical anions and the neutral molecule display different conformations. The protons in the 9-position of the radical anions of cis-9-methylthioxanthene S-oxides (2, n — 1, R1 = H, R2 = CH3) have an appreciable coupling constant10 which suggests that these radical anions have the substituent in the pseudo-axial position. Furthermore, in the radical anions the S—O bond is pseudo-axial. These situations are exactly the opposite of that observed for the neutral compound. 

A particular case of a [3C+2S] cycloaddition is that described by Sierra et al. related to the tail-to-tail dimerisation of alkynylcarbenes by reaction of these complexes with C8K (potassium graphite) at low temperature and further acid hydrolysis [69] (Scheme 24). In fact, this process should be considered as a [3C+2C] cycloaddition as two molecules of the carbene complex are involved in the reaction. Remarkable features of this reaction are (i) the formation of radical anion complexes by one-electron transfer from the potassium to the carbene complex, (ii) the tail-to-tail dimerisation to form a biscarbene anion intermediate and finally (iii) the protonation with a strong acid to produce the... 

The excited triplet states of quinones can be fairly readily populated by irradiation and nuclear polarization observed (Cocivera, 1968). Hydrogen atom abstraction leads to the relatively stable semiquinone radicals and, in alkaline media, radical anions. Recombination of radical pairs formed in this way can give rise to CIDNP signals, as found on irradiation of phenanthraquinone (20) in the presence of donors such as fluorene, xanthene and diphenylmethane (Maruyama et al., 1971a, c Shindo et al., 1971 see also Maruyama et al., 1972). The adducts are believed to have the 1,2-structure (21) with the methine proton appearing in absorption in the polarized spectrum, as expected for a triplet precursor. Consistently, thermal decomposition of 21 as shown in equation (61) leads to polarization of the reactant but now in emission (Maruyama... 

Thus a single two-electron wave is observed and only one product, the alcohol, can be isolated. Finally, at high pH neither the ketone nor the radical anion are protonated by this basic medium and it is not until the dianion, formed by successive electron transfers, that protonation occurs. 

---