Spontaneous oxidation of bis(heteroaryl)methanes and bis(heteroaryl)carbinols to ketones

Oxidation reactions of bis(heteroaryl)methanes to the corresponding ketones were investigated in the absence of usual oxidation reagents and catalysts. There are strong indications that the reaction pathway involves radical species. The spontaneous oxidation of bis(heteroaryl)carbinols was also investigated by kinetic measures. Strong base catalysis, together with relevant hydrogen/deuteron isotopic effect, were observed. Both reaction pathways, from methane derivatives and from carbinols derivatives, involve the presence of a tautomeric equilibrium of the considered heterocycles


Introduction
The oxidation reaction of hydrocarbons by molecular oxygen is usually carried out by metal catalysis. 1,2Aerobic auto-oxidation of hydrocarbons was reported to be enhanced by the presence of non-ionic and cationic surfactants. 3Some hydrocarbons are oxidised in the presence of Nhydroxyphthalimide and metals or quaternary ammonium bromide. 4Potassium superoxide was reported to initiate autoxidations of some arylmethanes. 5pontaneous oxidation of aryl methanes or of the corresponding carbinols, by atmospheric oxygen, in the absence of common oxidising reagents, is an unusual reaction.
Auto-oxidation of weakly acidic carbon atoms is known to occur in the presence of potassium t-butoxide in apolar solvents (benzene) or in poly(ethylene glycols). 6Under these experimental conditions, fluorene and diphenylmethane are oxidised to fluorenone and benzophenone, respectively.On the contrary, the oxidation of bis(heteroaryl)methanes or bis(heteroaryl)carbinols such as bis(2-benzothiazolyl)methane 7 (1A) or of bis(2benzothiazolyl)carbinol (2A) to bis(2-benzothiazolyl)ketone (3A) (see Scheme 1) is a spontaneous reaction 8 in working up the solutions of 1A and 2A and it may be an undesired side reaction in all studies of 1A and 2A and of related compounds.For example, the studies of the NH/CH tautomerism 8,9 on 1A (and on the related compounds) are complicated by the formation of carbinols 2 or ketones 3. Previously, 10 we reported the oxidation of 1A to 2,2-tetrakis(2-benzothiazolyl)ethane (4) by using a general oxidising reagent.Solutions of 4 are unstable and spontaneously give 5, as illustrated in Scheme 2 In DMSO, the oxidation of 1A to ketone 3A occurs in the presence of the oxygen.Under our experimental conditions, bases such as tertiary amines catalyse the oxidation reaction of Scheme 1: the presence of base is important as well as the presence of an oxidising species, which may be the dimethylsulfoxide.The presence of water smoothly enhances the reactivity.The reactions carried out in the presence of an acid catalyst (CH 3 SO 3 H) produces 3A in very low yields (less than 10%) in 3 days, while addition of an excess of H 2 O 2 (see experimental section) quickly afforded 3A.
Under the same experimental condition of reaction of 1A, diphenyl methane (1D) is not converted into ketone 3D, bis(2-benzoxazolyl)methane (1C) instead, is converted into the correspondent ketone 3C by a rate considerably slower than that of 1A.Compounds 1C and 1F are reported to produce corresponding ketones 3C and 3F by oxidation with Cr 2 O 6 . 11,12Addition of water to the reaction mixture produces moderate increase of the reactivity.As previously described, 10 the oxidation reaction of 1A with the usual oxidizing reagents (KMnO 4 /H 2 SO 4 , or Cr 2 O 6 /H 2 SO 4 ) produces complicated reaction mixtures from which the coupling products 4 and 5 were isolated.
Reaction between 1A and N-bromosuccinimide (in equimolar amount), produces compound 5 in low yields, together with 50% of bromo bis(2-benzothiazolyl)methane (6).In the presence of an excess of N-bromosuccinimide, the dibromo derivative 7 is the major product (see experimental).
No formation of 3A was observed for reactions of 1A carried out in the absence of air or in the presence of radical inhibitors, such as 2,2,6,6-tetramethyl-1-piperidinyloxy radical (TEMPO) or thiophenol.
These facts, probably, indicate that the reaction follows a multi-step pathway with an important step involving formation of a radical species.Kinetic runs, carried out by following the appearance of 3A from 1A, were poorly reproducible and showed a curvilinear plot of log[3A] against time.We tried to obtain EPR evidence for the presence of a radical species.The only signals recorded were those related to the radical species arising from the interaction between the amine (Et 3 N) and small amounts of the ketone 3A, as tested by independent spectrum of 3A and the amine.
A possible intermediate of the reaction from 1A to 3A may be the carbinol 2A.The formation of 2A as an intermediate, is justified by the fact that the oxidation reaction (under the same experimental conditions of 1A) of bis(2-benzothiazolyl)ethane (8) produces the carbinol 9 (see Scheme 3).

Scheme 3
We were not able to obtain evidence for the presence of 2A in the reaction mixtures of 1A, (by TLC analysis and by performing runs directly in the probe of the NMR spectrometer).

Oxidation of bis(heteroaryl)carbinols (2) to bis(heteroaryl)ketones (3)
Different procedures were tried to obtain pure carbinols 2A-C, because ketones 3 were often present in the reaction mixtures.The best way to obtain these compounds involves the reduction of the ketones 3A-C by sodium borohydride, as reported in the experimental section.
The spontaneous oxidation of 2A to ketone 3A, in DMSO, THF, CCl 4 , is faster than that of 1A, and the ketone 3A is obtained in almost quantitative yield.Infact, in the reported solvents, 2A is unstable under ambient conditions, and must be stored as a solid at -25°C.The low stability of carbinols 2 (they convert into ketones 3) necessitates their quick work up.The presence of DMSO is not essential for ketone formation from the carbinol.The key results are presented in Table 2.The behaviour in CCl 4 parallels that in THF.In this case too, addition of water (0.2 mmol, under the experimental conditions of entries 1 and 4 of Table 2) produces a small decrease in the reaction times.
In the reaction mixture of 2A we did not observe the presence of 1A.Compound 1A may be the second product (and it is oxidised to 3A at a lower rate than that of 2A) if a 'dismutationlike' process takes place.The oxidation reaction of carbinols 2A, 2B and 2C show a regular kinetic feature.These reactions are catalysed by bases (DABCO and Et 3 N).At [DABCO] o = 4.9 x 10 -2 mol dm -3 , in THF, the k obs value is not affected by the [2A] o values ( [2] o means the initial concentration values of 2) in the range 3 x 10 -5 to 4 x 10 -4 mol dm -3 (k obs = 6.5±0.2) x 10 -3 s -1 , error is standard deviation).The oxidation of 2A to 3A follows a first order law in the carbinol until high percent of conversion (80%) after which the reaction rate is decreased.The kinetic data obtained by initial (50%) conversion are reported in Table 3.  (10).
The decreased rate probably results from the interactions between the ketone formed during the reaction, and the starting alcohol by an equilibrium that produces an hemi-ketal, this depresses the presence of free alcohol as the ketone concentration becomes high.
Kinetic data are reported in Table 3. k obs Values (in s -1 ) linearly increase with increasing the value of [base].Table 4 summarises data dissection using equation ( 1) where k o (s -1 ) is the oxidation rate in the absence of base and k B is the rate of the catalysed process in s -1 mol -1 dm 3 .
In some cases, intercept values (k o ) show strong uncertainty, and, practically, tend to zero.The same Table 4 reports data concerning the oxidation rate of the deutero carbinol 10.In some cases, the intersection of the straight line of equation 1 agrees well with k obs values independently obtained in the absence of base.The importance of base catalysis with respect to the uncatalysed oxidation is expressed by the k B / k o ratios reported in Table 4.

Oxidation of bis(heteroaryl)methanes (1) to bis(heteroaryl)ketones (3)
The oxidation reaction from CH 2 to CO group most likely occurs in several steps, involving carbinol species 2, but we were not able to obtain evidence for the presence of carbinols 2 in the reaction mixtures from 1A to 3A(including reactions performed directly in the NMR probe).The failure to detect the presence of carbinol 2A is a consequence of the fact that the ketone is obtained by a reaction about 10 3 times faster from 2A than the reaction from 1A and therefore 2A does not accumulate under the experimental conditions.The oxidation of methane derivatives 1 requires the presence of atmospheric oxygen in the solvents (DMSO, THF, CCl 4 ) used.The presence of a radical scavenger did not allow the formation of ketones (entries 6, 11 of Table 1).Clearly, there is an important step involving the formation of a radical species along the oxidation reaction pathway.
We did not observe any difference between the reaction carried out in the dark and those without protection from light.In the literature 1 there are attempts to explain similar oxidation reactions of the C _ H group of substituted methane derivatives to corresponding oxygenated derivatives.
Oxidation of triaryl-methanes (and of 1,1-diphenyl-ethane) in DMSO/t-butyl alcohol mixtures was studied in the presence of potassium t-butoxide. 13This reaction produces mainly carbinols by a radical mechanism (via peroxide derivatives) on the carbanion related to the starting methane (or ethane) derivatives.
A reasonable pathway for these reactions may be a photo-oxidation mechanism, but our evidence does not support this hypothesis.
Another possible reaction pathway involves the anionic species of 1, such as compound 11 in Scheme 4. 11 may be important in explaining the oxidation reaction in the presence of metals.Present data are hardly explained by considering the equilibrium of Scheme 4 to be important in the oxidation pathway, because the heterocyclic derivatives 1C and 1F show a reactivity significantly lower than that of 1A, but the electron-withdrawing powers of the heterocyclic moiety are very similar, as shown by heterocyclic σ values of the Hammett-like treatment (benzothiazole-2-yl σ − = 0.65, 14 benzoxazole-2-yl σ − = 0.68, 14 thiazole-2-yl,  σ -= 0.63, 15 pyridine-2-yl σ − = 1.0 14 ).
The addition of water to the tautomeric form of 1A, may be an alternative for the formation of the carbinol (to the addition of oxygen), probably involving radical species.The inverse addition of water to the double bond should result in a ring opening reaction or return back to 1A.
The last step is an oxidative dehydrogenation: the aromatisation of 12 to 3A should be a simple elimination of hydrogen.All attempts to have evidence of formation of H 2 failed.Even if the data in Table 1 are qualitative, reaction pathway depicted in Scheme 5 (related to the benzothiazole derivative) is a reasonable picture to explain all the reported data.In particular the base is a catalyst to shift the tautomeric equilibrium 16 depicted in Scheme 5.The first, produces the tautomer of 1A which is the species that reacts with water; the second produces the enol preceding the ketone 12.
The last step of Scheme 5 is an oxidative dehydrogenation: 2-benzothiazoline 17 and 2thiazoline 18 are indicated to be reducing agents in mildly experimental conditions.
Under the same experimental conditions reported here, the diphenylmethane is completely unreactive.This reaction pathway agrees with the fact that when the 1,1-bis(2benzothiazolyl)ethane ( 8) is prepared, the corresponding carbinol was recovered from the reaction mixtures.The pathway reported in Scheme 5, as well as that of Scheme 6, is an alternative mechanism to the usual mechanism involving anionic/radical specie. 19hoto-oxygenation of methanes bonding isoquinoline derivatives (similar to that reported here) are indicated to start from the NH tautomer of the heterocyclic moiety. 20idation of bis(heteroaryl)carbinols (2) to bis(heteroaryl)ketones (3) Our evidence supports the idea that the oxidation of carbinols 2 to ketones 3 occurs without intervention of some radical species in the rate limiting step.The presence of a radical scavenger (entry 10 of Table 2) did not affect the yields and the reaction times.Reactions carried out in dark or in sunlight, gave the same results.
With regard to changes in the heterocyclic moiety, the oxidation of 2 parallels the behaviour observed for the oxidation of the CH 2 group of compounds 1.The reactivity order is 2A > 2B > 2C > 2E, 2F > 2D.Also in this case the electron-withdrawing power of the heterocyclic moiety appears to be unimportant.This conclusion arises not only from data presented in Table 2, but also from kinetic data in Table 4.
The rate of both oxidation reactions of compounds 1 and 2 are enhanced by the presence of a base.This fact agrees with a proton departure in a rate limiting step (equilibrium).A possible equilibrium is represented in Scheme 7.
As In the case of the equilibrium depicted in Scheme 4, formation of 11 is a reasonable process.On the contrary, equilibrium of Scheme 7 affording the presence of the anionic specie 13, should compete with the equilibrium of formation of anion 14 from the dissociation of the O-H group, reasonably more acid than the C-H group.As a consequence in the oxidation of carbinols, the pathway involving carbanions like 13, is less likely and base catalysis on the proton departure from the C-H group cannot be considered in the reaction pathway.Probably, the presence of 14 is not relevant, and, in any case it may represent a cul-de-sac.In fact, the relevant importance of a pre-equilibrium affording 14, should remove carbinols 2 from the oxidation reaction, and, consequently, the overall reaction rate should be reduced by the addition of base.A possible reaction pathway parallel to that of Scheme 5, is depicted in Scheme 8.In this scheme, tautomerism of 2 is an important step to obtain the ketone 16, via enol form 15. This statement agrees well with the fact that diphenylmethane is not reactive.In addition, six membered heterocyclic derivatives, such as pyridine derivatives, are less prone to afford N-H tautomers compared to five membered heterocycles. 21Thus, the oxidation of pyridine derivatives is significantly slower than the oxidation of thiazole and benzothiazole derivatives.Regarding five membered ring systems, benzothiazole derivatives are more prone to give 'non-aromatic' tautomeric form than the thiazole derivatives: the oxidation of benzothiazole derivatives is faster than that of thiazole derivatives.From 16, ketones 3 are easily obtained by oxidative dehydrogenation. 18

Conclusions
In conclusion, the oxidation of bis(heteroaryl)methanes occurs by a radical oxidation involving the NH tautomer of 1.The oxidation reactions of carbinols 2 occur via two tautomeric equilibria, which are favored by the presence of base.In both reactions, the partially saturated compounds 12 and 16 are formed in the final step which, probably, is a fast step.

Experimental Section
General Procedures. 1 H and 13 C NMR spectra were recorded on a Varian Gemini spectrometer at 300 and 75.46 MHz, respectively.Chemical shifts are referenced to solvent.J values are given in Hz.Mass spectra were recorded at an ionisation voltage of 70 eV on a VG 7070 E spectrometer.Thin-layer chromatography was performed on Merck Kieselgel 60 F 254 .Melting points were measured with a Büchi apparatus and are uncorrected.THF was distilled from sodium benzophenone ketyl.Air and moisture sensitive solutions and reagents were handled in dried apparatus under an atmosphere of dry nitrogen.Uv/vis spectrophotometric data were recorded with a Perkin Elmer (model Lambda 12) spectrophotometer.Under the reported experimental conditions, uv/vis spectrophotometric analysis (as well as the TLC analysis) did not show evidence for the presence of other oxidation/condensation products other than keto derivatives 3. Kinetic measurements were performed by the usual procedures, by following the appearance of compounds 3, until high percent of conversion, at λ max values here reported: 3A λ max = 343nm (ε = 2.05 x 10 5 ); 3B λ max = 325nm (ε = 2.20 x 10 3 ); 3C λ max = 335nm (ε = 1.82 x 10 5 ) Table 3 reports k obs values (s -1 ).

Table 3 .
Effect of the added bases on the rate of the reaction of bis(2-heteroaryl)carbinols to ketones, at 25°C