Nilufa Khatun,
Arghya Banerjee,
Sourav Kumar Santra,
Wajid Ali and
Bhisma K. Patel*
Department of Chemistry, Indian Institute of Technology Guwahati, 781 039, Assam, India. E-mail: patel@iitg.ernet.in; Fax: +91-3612690762
First published on 16th April 2015
A copper catalysed ortho-benzoxylation of 2-arylpyridines has been accomplished using benzylic ethers as the alternative arylcarboxy sources (ArCOO–) via sp2 C–H bond activation. The use of the Pd/TBHP catalytic system is reported to install an o-aroyl (ArCO–) moiety at the 2-arylpyridine while the Cu/TBHP combination fixes a benzoxy (ArCOO–) group at the ortho site.
Benzylic ethers are commonly used as protecting groups for alcohols that can be easily cleaved under suitable oxidising or reducing conditions.5 It has also served as the dormant synthetic equivalents of aldehydes, carboxylic acids or esters depending upon the reaction conditions.6 Benzyl ether served as aroyl (ArCO–) equivalent both under palladium(II)7a or copper(I/II)7b catalysed reactions utilising TBHP as the oxidant (Scheme 1(a) and (b)). But under Cu catalysed reaction conditions, it resulted in O-aroylation and not C-aroylation.7b Catalyst dependent selectivity is not uncommon in literature. In our earlier works, during the synthesis of 2-aminobenzothiazoles from 2-halothioureas, catalyst CuI followed C–X (X = halogen) bond breaking path while PdII-preferred the C–H activation path.8 Further, divergent reactivity was observed using alkyl benzenes,2b terminal alkenes/alkynes,3c and benzyl amines.9 They all serve as aroyl (ArCO–) surrogates for substrate-directed ortho-aroylation when the catalyst used was PdII, while the use of CuII catalyst preferred installing aryl carboxy (ArCOO–) groups at the ortho site of directing arenes.3a,4,10
| Entry | Catalyst (mol%) | Oxidant (equiv.) | Solvent | Yield (%) |
|---|---|---|---|---|
| a Reaction conditions: 2-phenylpyridine (1), (0.5 mmol), dibenzyl ether (a), (0.75 mmol), PhCl (0.5 mL), 22 h.b Isolated yield.c Reaction carried out at 100 °C. | ||||
| 1 | CuI (10) | TBHP (3) | DCE | 23 |
| 2 | CuCl (10) | TBHP (3) | DCE | 17 |
| 3 | CuBr (10) | TBHP (3) | DCE | 18 |
| 4 | Cu(OAc)2 (10) | TBHP (3) | DCE | 26 |
| 5 | Cu(OTf)2 (10) | TBHP (3) | DCE | 18 |
| 6 | CuCl2 (10) | TBHP (3) | DCE | 11 |
| 7 | CuBr2 (10) | TBHP (3) | DCE | 9 |
| 8 | CuO (10) | TBHP (3) | DCE | 11 |
| 9 | CuSO4·5H2O (10) | TBHP (3) | DCE | 8 |
| 10 | Cu(OAc)2 (20) | TBHP (3) | DCE | 37 |
| 11 | Cu(OAc)2 (30) | TBHP (3) | DCE | 41 |
| 12 | Cu(OAc)2 (20) | TBHP (3) | DMSO | 6 |
| 13 | Cu(OAc)2 (20) | TBHP (3) | DMF | 5 |
| 14 | Cu(OAc)2 (20) | TBHP (3) | CH3CN | 13 |
| 15 | Cu(OAc)2 (20) | TBHP (3) | PhCl | 41 |
| 16 | Cu(OAc)2 (20) | Aq. TBHP (3) | PhCl | 47 |
| 17 | Cu(OAc)2 (20) | Aq. TBHP (6) | PhCl | 58 |
| 18 | Cu(OAc)2 (20) | H2O2 (6) | PhCl | 00 |
| 19 | Cu(OAc)2 (20) | DTBP (6) | PhCl | 00 |
| 20 | Cu(OAc)2 (20) | m-CPBA (6) | PhCl | 00 |
| 21 | Cu(OAc)2 (20) | Oxone (6) | PhCl | 00 |
| 22 | Cu(OAc)2 (20) | K2S2O8 (6) | PhCl | 00 |
| 23 | — | Aq. TBHP (6) | PhCl | 00 |
| 24 | Cu(OAc)2 (20) | — | PhCl | 00 |
| 25 | Cu(OAc)2 (20) | Aq. TBHP (6) | PhCl | 47c |
Encouraged by the finding of benzyl ethers serving as the surrogate of ArCOO–, a series of reactions were carried out by varying catalysts, oxidants and solvents to arrive at the best possible yield. At first, the efficacies of various copper salts were screened keeping all other parameters constant. Among the catalysts tested (Table 1, entries 2–9) such as CuCl (17%), CuBr (18%), Cu(OTf)2 (18%), CuCl2 (11%), CuBr2 (9%), CuO (11%) and CuSO4·5H2O (8%) in DCE solvent, Cu(OAc)2 (26%) (Table 1, entry 4) was found to be the ideal. The yield of the desired product (1a) marginally improved (37%) when the amount of Cu(OAc)2 was increased to 20 mol% (Table 1, entry 10). Furthermore, the use of excess Cu(OAc)2 (upto 30 mol%) did not improve the yield (41%) significantly (Table 1, entry 11). Polar aprotic solvents such as DMSO, DMF and CH3CN were found to be less effective for this transformation as illustrated in Table 1, (entries 12–14). The use of chlorobenzene as the solvent provided better yield (41%) with lesser side products (Table 1, entry 15) than DCE (Table 1, entry 10). The use of 70% aqueous solution of TBHP in a lieu of decane TBHP was found to be better (47%) for this transformation as shown in Table 1, entry 16. A further improvement in the yield (upto 58%) of (1a) was observed when the aq. TBHP quantity was increased to two fold (6 equiv.) (Table 1, entry 17). Instead of TBHP, other oxidants such as H2O2, di-tert-butyl peroxide (DTBP), m-chloroperbenzoic acid (m-CPBA), oxone and K2S2O8 were also tested during the screening of the reaction. As shown in Table 1, entries 18–22, all these oxidants are found to be ineffective for this transformation. Notably, either in the absence of copper salt or TBHP, the reaction failed to yield the desired product (1a). The yield of product (1a) dropped to 47% when the reaction temperature was decreased from 120 °C to 100 °C. Finally, the optimised reaction condition was, the use of 2-phenylpyridine (1) (0.5 mmol), dibenzyl ethers (a) (0.75 mmol), Cu(OAc)2 (20 mol%), TBHP (aq. 70%) (6 equiv.) in chlorobenzene (0.5 mL) at 120 °C.
With the above optimised conditions in hand, the scope of this strategy was then implemented to the reaction between 2-phenylpyridine (1) and various substituted dibenzyl ethers and the results are summarised in Scheme 3. Dibenzyl ethers having electron neutral –H (a) and electron donating groups such as o-Me (b), m-Me (c), p-Me (d) and p-OMe (e) as well as electron withdrawing p-Br (f) and p-Cl (g) substituents were all found to serve as ArCOO– sources and gave good to moderate yields of corresponding products of (1a–1g). The presence of electron donating substituents in the aryl ring of dibenzyl ethers irrespective of their position of attachments (b–e) provided better yields than those possessing electron-withdrawing substituents (f and g) as shown in Scheme 3. The efficacy of this coupling reaction was further executed with substituted 2-phenylpyridines such as 2-p-tolylpyridine (2). Reaction of (2) with various substituted dibenzyl ethers (a–g) were then carried out and all provided good to moderate yields of their respective products (2a–2g) as shown in Scheme 3. Similarly, 2-m-tolylpyridine (3) when treated with dibenzyl ether (a) under the reaction conditions gave a good yield of the desired product (3a).
Further 2-(4-methoxyphenyl)pyridine (4), another activated substrate when treated with dibenzyl ethers possessing electron neutral –H (a) and electron withdrawing p-Cl (g) under the present reaction conditions, provided (4a) and (4g) in 75% and 54% yields respectively as shown in Scheme 3. Dibenzyl ether (a) also served as ArCOO– surrogate with other 2-phenylpyridine derivatives possessing electron-withdrawing substituents such as p-Br (5) and p-Cl (6), giving o-benzoxylated products (5a) and (6a) respectively in moderate yields. However, directed arenes bearing activated substituents such as p-Me (2), m-Me (3) and p-OMe (4) gave better yields compare to those possessing electron withdrawing substituents such as p-Br (5) and p-Cl (6) (Scheme 3). This is because of the better electrophilic metallation of Cu(II) catalyst with activated 2-aryl rings of 2-arylpyridine. To check whether both the halves of a dibenzyl ether acted as the arylcarboxy source or not, an unsymmetrical dibenzyl ether (eg) was treated with 2-phenylpyridine (1). Under the present reaction conditions, the unsymmetrical dibenzyl ether (eg) provided a separable mixture of (1e) and (1g) in a ratio of 1.5
:
1 as shown in Scheme 4(i). This result reconfirms that both the halves served as aryl carboxy source but the activated aryl ring acts as a better o-benzoxy source compare to its deactivated counterpart (Scheme 4(i)).
A set of controlled reactions were performed to gather the evidences to support the postulated reaction mechanism for this transformation. Analysis of the crude reaction products of dibenzyl ether (a) under the optimised conditions in the absence of directing substrate (1) revealed the presence of benzyl alcohol (A), benzaldehyde (B), benzoic acid (C) and benzylbenzoate (D) as detected by HRMS, which is consistent with previous observations.7b Thus, to find out the possible active intermediates, reaction of benzyl alcohol (A), benzaldehyde (B), benzoic acid (C) and benzylbenzoate (D) with (1) were carried out separately under otherwise identical conditions. Interestingly benzyl alcohol (A) and benzaldehyde (B) when reacted with (1), provided the corresponding product (1a) in 61% and 65% yields respectively (Scheme 4(ii), path I and path II). These results strongly support their intermediacy during this transformation. While benzoic acid (C) and benzylbenzoate (D) yielded only 7% and 10% of (1a), suggesting those species may not be the active coupling partners (Scheme 4(ii), path III and path IV). The use of benzoic acid in lieu of dibenzyl ether provided only a trace of o-benzoxylated product supporting the presence of active benzoxy radical and not the benzoate anion in the medium. Thus, the possibility of oxidation of the ligand chelated CuII species to CuIII species via disproportionation reaction21 is less feasible as compared to its oxidation via active benzoxy radical. Furthermore, to support the radical nature of the coupling, a reaction was conducted in the presence of a radical quencher TEMPO under standard conditions. A substantial drop in the yield of (1a) (13%) along with the formation of TEMPO-ester (X) confirms the radical nature of the reaction as illustrated in Scheme 4(iii). Results obtained from controlled experiments and from our recent reports3a,4 a plausible reaction mechanism has been postulated as shown in Scheme 5(i). Presumably, TBHP in the presence of copper catalyst produces species (a′) via an initial SET mechanism. The intermediate species (a′) undergo proton abstraction of α-sp3 C–H bond to give an oxonium species (a′′). However alternative path involving α-sp3 C–H proton abstraction followed by a SET mechanism to form oxonium species (a′′) cannot be ruled out.20 A nucleophilic attack of water on oxonium species leads to the formation of an unstable hemi-acetal species (benzyloxy)(phenyl)methanol (a′′′). This hemi-acetal species easily cleaved to give an equimolar mixture of benzyl alcohol (A) and benzaldehyde (B). Thus formed benzyl alcohol (A) generated is further oxidised to the corresponding benzaldehyde (B). Due to the presence of an excess of TBHP, the in situ generated (B′) obtained by the proton abstraction of benzaldehyde (B), forms a perester species (E). Homolytic cleavage of this perester (E) forms carboxy radical (C′). Further oxidative addition of this carboxy radical (C′) with cyclometallated Cu complex (F) lead to the formation of a unstable CuIII intermediate (G). Finally, a reductive elimination of (G) installs a benzoxy moiety at the ortho site of (1) forming CuI species. The generated CuI catalyst is oxidised to CuII for subsequent catalytic cycle as shown in Scheme 5(i). Unsymmetrical dibenzyl ether (eg) provided (1e) as the major product. This is due to the formation of a more stable oxonium species by the α-sp3 C–H proton abstraction from activated ring side (eg′′). The in situ generated hemi-acetal intermediate (eg′′′) cleaved to equimolar mixture of 4-methoxybenzaldehyde and 4-chlorobenzylalcohol as shown in Scheme 5(ii). The in situ generated 4-methoxy benzaldehyde then undergo preferential coupling with 2-phenylpyridine (1). This favoured coupling of electron-donating substituent is true even when an equimolar mixture of 4-methoxy benzaldehyde and 4-chlorobenzaldehyde were reacted with (1) under the optimised conditions. The ratio of corresponding o-benzoxylated products (1e) and (1g) obtained were 7
:
3 confirming our assumption. The higher propensity of formation of o-benzoxylated product derived from electron-donating part of unsymmetrical dibenzylether (eg) has been ascertain even when the reaction was performed with 0.5 mmol of unsymmetrical ether (eg). The ratio of products (1e) and (1g) (1.45
:
1) obtained were almost identical (1.50
:
1) using 0.75 mmol (1.5 equiv.) of (eg) supporting our presumption.
:
ethylacetate: 9
:
1) as the eluent to give pure compound (2-(pyridin-2-yl)phenyl benzoate) (1a, 0.08 g, 58%) as a brownish oil material.
Footnote |
| † Electronic supplementary information (ESI) available: 1H and 13C NMR spectra. See DOI: 10.1039/c5ra03462k |
| This journal is © The Royal Society of Chemistry 2015 |