• azole
• direct arylation

Our own interest in the development of new and convenient protocols for the highly regioselective synthesis of (hetero)arylazoles via palladium-catalyzed intermolecular direct (hetero)arylation reactions of azoles with (hetero)aryl halides prompted us to design general and selective procedures for the synthesis of unsymmetrically 4,5-diaryl substituted 1-methyl-1H-imidazoles 1, 4(5)-aryl-1H-imidazoles 2 and unsymmetrically 4,5-diaryl substituted 1H-imidazoles 3, in which selective direct C-5 arylation reactions of azole nuclei had to be key-steps. In particular, we demonstrated that a variety of 4,5-diaryl-1-methyl-1H-imidazoles 1 can be efficiently prepared by a highly regioselective three-step procedure in which the first step involves the synthesis of 5-aryl-1-methyl-1H-imidazoles 7 by a highly selective Pd(OAc)2/P(2-furyl)3-catalyzed direct C-5 (hetero)arylation of commercially available 1-methyl-1H-imidazole (4) with activated, unactivated and deactivated aryl bromides 6 in the presence of K2CO3 as the base in DMF at 110 °C. So obtained (hetero)arylimidazoles 7 were then converted into the required compounds 1 via selective C-4 bromination followed by a Suzuki-type reaction of the resulting 5-aryl-4-bromo-1-methyl-1H-imidazoles 8 with arylboronic acids 9 in the presence of a catalytic amount of PdCl2(dppf) under phase-transfer conditions.
Next, our attention was focused on the design and development of practical and efficient procedures for the synthesis of 4(5)-aryl-1H-imidazoles 2 and, in this context we showed that commercially available 1-benzyl-1H-imidazole (19) can undergo highly regioselective Pd(OAc)2/P(2-furyl)3-catalyzed direct C-5 (hetero)arylation with electron-neutral or electron-poor aryl bromides 6 in the presence of K2CO3 as base in DMF at 140 °C. N-Debenzylation of the resulting 5-aryl-1-benzyl-1H-imidazoles 21 via Pd-catalyzed transfer hydrogenation furnished compounds 2. Remarkably, all our attempts to prepare these compounds by Pd-catalyzed direct C-5 arylation of (NH)-free imidazole (18) with aryl iodides 5 were unsuccessful. 4(5)-Aryl-1H-imidazoles 2 were also alternatively efficiently synthesized on a multigram scale by a concise one-step procedure involving a PdCl2(dppf)-catalyzed Suzuki-Miyaura type reaction between commercially available 4(5)-bromo-1H-imidazole (26) and arylboronic acids 9 under phase-transfer conditions.
5-Aryl-1-benzyl-1H-imidazoles 21 were also employed as convenient starting materials for the regioselective synthesis of 4,5-diaryl-1H-imidazoles 3 by a protocol in which 5-aryl-4-bromo-1-benzyl-1H-imidazoles 27, prepared by selective C-4 bromination of compounds 21, underwent a PdCl2(dppf)-catalyzed Suzuki-type reaction with arylboronic acids 9 under phase-transfer conditions. N-Debenzylation of the resulting 1-benzyl-4,5-diaryl-1H-imidazoles 28 gave heteroarenes 3. Unfortunately this approach suffered from the occasionally difficult N-debenzylation of compounds 28 either by Pd-catalyzed transfer hydrogenation or by Pd-catalyzed hydrogenolysis with H2.
The synthesis of two naturally-occurring 2-alkyl-5-aryl-1H-imidazoles, Catharsitoxins D (31a) and E (31b), from cheap commercially available starting materials was also tackled. However, due to the problems sometimes encountered in the N-debenzylation of 1-benzyl-1H-imidazole derivatives, the synthesis of these natural products was performed via highly regioselective C-5 arylation of the corresponding 2-alkyl-1-methoxymethyl-1H-imidazoles, 34a and 34b, respectively, with bromobenzene in the presence of K2CO3 as the base and a Pd(OAc)2/P(2-furyl)3 catalyst system. N-Deprotection of the resulting 5-phenylimidazole derivatives in acidic conditions then provided the required natural products.
Our attention was also turned to the development of a general and efficient method for the direct and highly regioselective C-2 arylation of 1,3-azoles with aryl iodides 5 and we found that a large variety of azoles, including, 1-methyl-1H-imidazole (4), 1-benzyl-1H-imidazole (19), 5-aryl-1-benzyl-1H-imidazoles 21, 1-aryl-1H-imidazoles 36, oxazole (40), thiazole (41), benzothiazole (53) and 1-methyl-1H-benzimidazole (55), are able to undergo Pd(OAc)2-catalyzed and CuI-mediated C-2 arylation reaction with deactivated, unactivated, and activated aryl iodides 5 in DMF at 140 °C under base-free and ligandless conditions. The required 2-arylheterocycles were so obtained in high yields and chemical purity.
This unprecedented procedure was also successfully employed for the highly selective C-2 arylation of substrates containing base-sensitive groups such as the NH groups in 1H-imidazole (18), benzimidazole (44) and 1H-indole (47a) without a preliminary N-protection. Remarkably, N-arylated byproducts were never observed under the above mentioned experimental conditions. We also demonstrated that when the arylation reactions were performed in a solvent with mildly basic characteristics such as DMF or DMA, this base-free and ligandless procedure could involve the use of heterocycle substrates that do not contain a basic pyridine-like nitrogen atom, such as 1H-indole (47a). In fact, by using these experimental conditions we were able to synthesize 2-aryl-1H-indoles 48a from 47a with complete C-2 selectivity, albeit in modest yields.
Additionally, we showed that 2-aryl-1-phenyl-1H-imidazoles 37 can be prepared in satisfactory yields by a one-pot domino HALEX and Pd-catalyzed and Cu-mediated arylation reactions of commercially available 1-phenyl-1H-imidazole (36a) with activated and unactivated aryl bromides 6 under base-free and ligandless conditions. However, this protocol proved to be unsuitable for the C-2 arylation of 36a with a deactivated aryl bromide such as 4-bromoanisole (6c). Nevertheless, it was subsequently discovered that more severe reaction conditions than those used with aryl iodides could allow us the C-2 arylation of 36a with aryl bromide 6c under base-free and ligandless conditions without resorting to HALEX reactions.
Our new arylation protocol was then successfully employed for the preparation of a large variety of 2-arylazoles that include pharmacologically active compounds and their synthetic precursors such as compounds 52, 54a or 54c and 46c and 56a.
In the course of our study on the regioselective direct C-2 arylation reactions, we also developed two new protocols for the highly selective synthesis of 2,4(5)-diaryl-1H-imidazoles 57 in which Pd-catalyzed and Cu-mediated direct C-2 arylation reactions under base-free and ligandless conditions were used as key-steps. In the first protocol, 4(5)-aryl-1H-imidazoles 2, which were prepared as mentioned above, were shown to be able to undergo highly selective Pd-catalyzed and Cu-mediated direct C-2 arylation with unactivated and activated aryl iodides 5 and bromides 6 to give the required 2,4(5)-diaryl-1H-imidazoles 57 in modest to good yields. On the other hand, the second protocol involved the synthesis of compounds 57 via functional group tolerant Pd-catalyzed and Cu-mediated C-2 arylation reactions of 5-aryl-1-benzyl-1H-imidazoles 21 with aryl iodides 5 or aryl bromides 6 under base-free and ligandless conditions, followed by Pd-catalyzed N-debenzylation of the resulting 1-benzyl-2,5-diaryl-1H-imidazoles 25. Interestingly, the overall yields of the two-step synthesis of diarylimidazoles 57 via direct C-2 arylation of 4(5)-aryl-1H-imidazoles 2, prepared by Suzuki-Miyaura reaction of 4(5)-bromo-1H-imidazole (26) with arylboronic acids 9, proved to be comparable to those achieved by the three-step method involving the preparation of 5-aryl-1-benzyl-1H-imidazoles 21 from 1-benzyl-1H-imidazole (19) and the C-2 arylation of compounds 21, followed by N-debenzylation.
The data obtained in our study on the Pd-catalyzed and Cu-promoted base-free direct C-2 arylations of azoles allowed us to hypothesize a plausible reaction mechanism for these reactions that involves the formation of organocopper(I) derivatives and their transmetalation reaction with arylpalladium(II) halide species, followed by reductive elimination. However, a reaction mechanism involving the presence of organopalladium(II), organopalladium(IV) species, and organocopper(I) derivatives can not be excluded.
In an attempt to improve the Pd(OAc)2-catalyzed and CuI-mediated reaction of 1H-indole (47a) with aryl iodides 5 in DMF at 140 °C or in DMA at 160 °C under base-free and ligandless conditions, which selectively provided 2-aryl-1H-indoles 48a in modest yields, we developed a new and inexpensive version of the Ullmann reaction for the selective, efficient and functional group-tolerant N-arylation of 1H-indoles 47 and 9H-carbazole (59) with aryl iodides 5. In fact, we discovered that compounds 47 and 59 undergo highly selective reaction with electron-rich and electron-poor aryl iodides 5 in DMA at 160 °C under base-free and ligandless conditions to give the corresponding N-arylated derivatives in high yields. The experimental conditions for this reaction allowed an unprecedented tolerance of functional groups in the 1H-indoles 47 and facilitated the workup of the reaction mixtures and isolation of the required chemically pure compounds.
Finally, we applied our synthetic protocols to the preparation of new 1,2- and 1,5-diaryl-1H-imidazoles, 37 and 39, respectively, which can be considered Z-restricted analogues of combretastatin A-4 (CA-4) (61), and we evaluated their antitumor activity against the NCI 60 tumor cell lines panel. Interestingly, two 1,5-diaryl-1H-imidazoles, compounds 39h and 39i, were found to be more cytotoxic than CA-4 (61). Moreover, docking experiments showed a good correlation between the mean Log molar drug concentration (MG–MID Log GI50) values of all tested compounds 37 and 39 and their calculated interaction energies with the colchicine binding site of αβ-tubulin. It is also worth noting that some selected imidazoles, including 5-(4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)-1H-imidazole (39g), 5-(3-fluoro-4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)-1H-imidazole (39i), 5-(naphthalen-2-yl)-1-(3,4,5-trimethoxyphenyl)-1H-imidazole (39h), 5-(4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)-1H-imidazol-3-ium chloride (67g), 5-(naphthalen-2-yl)-1-(3,4,5-trimethoxyphenyl)-1H-imidazol-3-ium chloride (67h) and 5-(3-fluoro-4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)-1H-imidazol-3-ium chloride (67i), proved to exhibit vascular disrupting activity in vitro on human umbilical vein endothelial cells (HUVECs), and in vivo on experimental tumors and that compound 39g was found to be able to inhibit significantly the tumor growth at a 150 mg/Kg/day dose (52 % inhibition after 25 days) in an in vivo test performed on MD-MBA-435 breast cancer cells xenotransplanted in immunodeficient mice.