and computed CH acidity of 1,2,3- and 1,2,4-triazoles. Application to the synthesis of resveratrol analogues

: 1-Aryl- and 2-aryl-1,2,3-triazoles were synthesized by N -arylation of the corresponding azoles using aryl iodides. The deprotometalations of 1-phenyl-1,2,3-triazole and -1,2,4-triazole were performed using a 2,2,6,6-tetramethylpiperidino-based mixed lithium-zinc combination and occurred at the most acidic site, affording by iodolysis the 5-substituted derivatives. Dideprotonation was noted from 1-(2-thienyl)-1,2,4-triazole by increasing the amount of base. From 2-phenyl-1,2,3-triazoles, and in particular from 2-(4-trifluoromethoxy)phenyl-1,2,3-triazole, reactions at the 4 position of the triazolyl, but also ortho to the triazolyl on the phenyl group, were observed. The results were analyzed with the help of the CH acidities of the substrates, determined in THF solution using the DFT B3LYP method. 4-Iodo-2-phenyl-1,2,3-triazole and 4-iodo-2-(2-iodophenyl)-1,2,3-triazole were next involved in Suzuki coupling reactions to furnish the corresponding 4-arylated and 4,2’ diarylated derivatives. When evaluated for biological activities, the latter (which are resveratrol analogues) showed moderate antibacterial activity and promising antiproliferative effect against MDA-MB-231 cell line.


Introduction
Di-and triazoles are key elements present in compounds of biological interest 1 or in materials for a wide range of applications.
We herein describe our attempts to use the lithium-zinc combination above mentioned for the deprotometalation (followed by iodolysis) of the azole substrates 1 and 2 shown in Scheme 1. Earlier we have shown that the regioselectivity of the same reaction for the related substrates 3 6f and 4 6h is partly determined by the acidity of the different hydrogens in their molecules. As a consequence, we similarly tried to rationalize the reaction results using the CH acidities in THF of the heteroaromatic substrates calculated by using the homodesmic reaction approach within the density functional theory (DFT) framework. Finally, iodides generated by deprotometalation-iodolysis were involved in palladium-catalyzed Suzuki crosscoupling reactions, and the resulting arylated triazoles (which are resveratrol analogues) were evaluated for their biological activity. Scheme 1. Substrates for which the deprotometalation has been studied.
By replacing the phenyl group connected to the 1,2,4-triazole by a 2-thienyl group (substrate 2b), a reaction at the 5 position of the aza-heterocycle was still observed, as demonstrated by the isolation of the corresponding iodide 6b in 95% yield. By increasing the amount of base (0.75 equiv of ZnCl 2 ·TMEDA and 2.25 equiv of LiTMP instead of 0.5 equiv of ZnCl 2 ·TMEDA and 1.5 equiv of LiTMP), the iodide 6b became the minor product formed (8% yield) due to competitive dideprotometalation, as previously noted in the other azole series. 6f,6h Indeed, the diiodides 7b and 7b' were obtained in 59 and 28% yield, respectively (Scheme 4). The iodides 5a and 6a, as well as the major isomer 7b, were identified unequivocally by X-ray structure analysis ( Figure 1).  The triazole moiety is a well-known bioisostere of amide and ester groups and a linker replacement of double bonds. 10 In addition, resveratrol is a compound with a versatile biological activity. 11 In order to progress towards triazole-modified resveratrol analogues (Scheme 5, A), 2-(4-trifluoromethoxyphenyl)-1,2,3-triazole (8) was identified as a good precursor due to the impact of the trifluoromethoxy group on the biological activity of the derivatives. 12 It was synthesized by regioselective palladium-catalysed N-arylation of 1,2,3triazole at its N2 position using 1-bromo-4-(trifluoromethoxy)benzene under conditions described by Buchwald and co-workers. 13 Unfortunately, treatment of 8 by the base (0.5 equiv of ZnCl 2 ·TMEDA and 1.5 equiv of LiTMP) as before led, after addition of iodine, to a mixture from which only pure 2-(2-iodo-4-trifluoromethoxyphenyl)-1,2,3-triazole (9a) could be isolated. Analyzing the crude showed, besides the monoiodide 9a (estimated 25% yield), the formation of the diiodide 10 and the monoiodide 9b (in estimated 21 and 6% yield, respectively) and the recovery of substrate (about 33% yield) (Scheme 6).  This disappointing result led us to rather consider the synthesis of targets without trifluoromethoxy group. We thus chose 2-phenyl-1,2,3-triazole (4a), and synthesized the monoiodide 4b and diiodide 4b' as described previously. 6h The monoiodide 4b was first reacted with arylboronic acids (stoichiometric amount) in a Suzuki-type cross-coupling procedure 14 using catalytic Pd(dba) 2 (dba = dibenzylidene acetone) and triphenylphosphine, dioxane as solvent and CsF instead of a base to afford the compounds 4c,d in 54-70% yields.
For easier purification, we turned to a dba-free reported procedure 15 using catalytic PdCl 2 and triphenylphosphine, a biphasic medium and Na 2 CO 3 as a base, and isolated the arylated compounds 4e,f in 87-98% yields (Scheme 7). The structures of the compounds 4d and 4e were confirmed by X-ray diffraction ( Figure 2).  When involved in the reaction, the diiodide 4b' led to a complex mixture due to unselective mono-coupling. We decided to employ the arylboronic acid in excess (4 equiv) in order to synthesize the bis-coupling products. Thus, the bis-arylated derivatives were obtained in yields ranging from 73 to 92% (Scheme 8).

Computational aspects
The studies on CH acidity of triazoles and their derivatives are not numerous. A brief review of papers devoted to experimental and theoretical investigation of CH acidity of azoles is presented in our previous publication. 16 Fraser reported the pK a value (26.2) of 1-propyl-1H-1,2,4-triazole in THF solution. 17 Gas-phase deprotonation energies for several triazoles were estimated by means of semi-empirical calculations. 18 We also recently contributed to this field by DFT computed values for substituted triazoles, 19 N-aryl triazoles, 6h N-aryl pyrazoles 6f and N-aryl benzotriazoles. 6l In the present paper, the DFT calculated CH acidities of several N-aryl triazoles, both in gas phase (see Supplementary data) and in THF solution (Scheme 9), are presented. These D acid G and pK a values were obtained by using the theoretical protocol described thoroughly  All the calculations were performed by using the DFT B3LYP method. The geometries were optimized using the 6-31G(d) basis set. No symmetry constraints were applied. In order to perform stationary points characterization and to calculate zero-point vibrational energies (ZPVE) and thermal corrections, vibrational frequencies were calculated at the same level of theory. The single point energy calculations were performed using the 6-311+G(d,p) basis set and tight convergence criteria. The gas phase Gibbs energies (G 0 298 ) were calculated for each isolated species using the following equation: The gas phase acidities D acid G were determined as the Gibbs energies of deprotonation of ) by the following formula: Whereas the solvent influence was treated by using the polarized continuum model (PCM) with the default parameters for THF, 20 the PCM energies E PCM were calculated at the B3LYP/6-311+G(d,p) level using geometries optimized for isolated structures. The Gibbs energies in solution G s were calculated for each species by the equation: The pK a values were calculated by means of the following homodesmic reaction: where Het-H is an appropriate heterocycle with experimentally known pK a value. In this study, 1-propylpyrazole was chosen as reference compound since its pK a value in THF found by Fraser et al, 17 35.9, was supposed to be close to those for the investigated substrates.
Within this approach the Gibbs energy of the homodesmic reaction (D r G s ) and the pK a value are linked together by the following equation: It is obvious that the compounds 2b and 8 exist in form of several rotamers due to sterical interaction between adjacent hydrogens or/and heteroatom lone pairs. In such cases, the data on Scheme 9 (and Supplementary data) refer to the most stable ones.
There are several potential deprotonation sites in the investigated substrates. When comparing the CH acidity in gas-phase (see Supplementary data) and in THF solution (Scheme 9), the correlation can be easily seen. Investigation of gas-phase CH acidity (see Supplementary data) is of a great importance because these values are free of solvent influence and can be used for acidity scale development. The calculated values of gas-phase acidity of the investigated compounds lie within the range of 360 to 381 kcal mol -1 , which is typical for weak CH acids.
When analyzing the pK a values distribution for the substrates, one can notice that the 5 position of the triazole ring is clearly the most acidic for 1a, 2a and 2b. According to our previous experience, these absolute pK a values below 30 should hint the regioselectivities unequivocally. In contrast, for 8, CH acidities of the same magnitude should lead to poor selectivity.

Discussion
The calculations of the CH acidities in THF (Scheme 9) allowed us to comment the regioselectivities observed in the course of the reactions.
When treated with the lithium-zinc base, the substrates 1a, 2a and 2b were first attacked at the 5 position of the triazole ring. This could be easily rationalized since the 5 position is clearly the most acidic of these substrates, with pK a values of 27.7, 28.4 and 27.1, respectively. Using the base in excess with the substrate 2b led to the formation of the diiodides 7b and 7b', a result that could be due to rather low pK a values of 32.6 and 30.2, respectively at the 3 and 5 position of the 2-thienyl group.
To rationalize the formation of a mixture from 8 using the base prepared from ZnCl 2 ·TMEDA (0.5 equiv) and LiTMP (1.5 equiv), the pK a values in THF solution of 8 were compared with those of 2-phenyl-1,2,3-triazole (4a), which was deprotometalated at its 4 position under the same reaction conditions (Scheme 10). 6h When 4-substituted by a strongly electron-withdrawing trifluoromethoxy group, which is known to exhibit a long range effect, 21 the phenyl group becomes more prone to deprotonation and can compete with the triazolyl ring. Such an effect could be at the origin of the observed formation of a mixture of iodides from 8.

Biological evaluation
Due to their structural similarity with the resveratrol skeleton, most of the synthesized derivatives were biologically evaluated. As a preliminary screening, ten of these triazoles (the compounds 4c-f, 4c'-g' and 8) were assessed for their antibacterial activities against a representative sample of the bacterial species, the most frequently encountered at the Hospital ). For all the bacteria tested, both Gram-positive and Gram-negative, we were able to determine antibacterial activities (i.e. determination of the Minimum Inhibitory Concentration). More interesting, even if we observe globally a weak antibacterial activity for these compounds, all of them have an antibacterial activity, and to our knowledge, this is the first time that this kind of antibacterial activity is described against these pathogens. On particular interest, the 4,2'diaryl derivatives 4e' and 4f' as well as the 4-aryl derivatives 4e and 4f showed higher antibacterial activities (with MICs = 64 μg.mL -1 ). Nevertheless, at this stage it is difficult to speculate on the possibility to have promising antibacterial candidates. But we believe that this triazole derivatives could be of interest, and it is necessary to have more data (for example, synthesize other derivatives, or to conduct Structure Activity Relationships studies) to be able to obtain or identify potent antibacterial candidates.  (Table 3). 52.0 ± 5.7 40.0 ± 2.8 17.5 ± 0.7 45.5 ± 2.1 67.0 ± 1.4 129.9 ± 3.6

Conclusions
The different 1-aryl triazoles involved in the deprotometalation-iodolysis sequence were functionalized at their most acidic site, which is the 5 position of the aza-ring. By increasing the amount of base, it proved possible to also deprotonate the aryl group connected to the azole.
In the case of 2-phenyl-1,2,3-triazole, the 4-iodo and 4,2'-diiodo derivatives were involved in Suzuki coupling to afford 4-aryl and 4,2'-diaryl compounds, which were evaluated for biological activities. This synthesis of triazole derivatives is promising for the generation of potent antiproliferative compounds.

General
Metalation reactions were performed under an argon atmosphere. THF was distilled over sodium/benzophenone. Column chromatography separations were achieved on silica gel (40-63 μm). Melting points were measured on a Kofler apparatus. IR spectra were taken on a Perkin-Elmer Spectrum 100 spectrometer. 1 H and 13 C Nuclear Magnetic Resonance (NMR) spectra were recorded on a Bruker Avance III spectrometer at 300 and 75 MHz, respectively.
1 H chemical shifts (δ) are given in ppm relative to the solvent residual peak, 13 C chemical shifts are relative to the central peak of the solvent signal. 23 Mass spectra (HRMS) measurements were performed at the CRMPO (Centre Régional de Mesures Physiques de l'Ouest) of Rennes using a Waters Q-TOF 2 instrument in positive electrospray CI mode.

Crystallography
The samples were studied with graphite monochromatized Mo-Ka radiation (l = 0.71073 Å). X-ray diffraction data were collected at T = 150(2) K (compounds 4e, 5a, 6a, 7b) or 294(2) K (compound 4d) using APEXII Bruker-AXS diffractometer. The structure was solved by direct methods using the SIR97 program, 24 and then refined with full-matrix leastsquare methods based on F 2 (SHELX-97) 25 with the aid of the WINGX program. 26 All nonhydrogen atoms were refined with anisotropic atomic displacement parameters. Except oxygen linked hydrogen atom that was introduced in the structural model through Fourier difference maps analysis (4e), H atoms were finally included in their calculated positions.