[1,3]Thiazolo[3,2-b][1,2,4]triazol-7-ium salts: synthesis, properties and structural studies

Introduction

Thiazolo[3,2-b][1,2,4]triazoles and other derivatives of 1,2,4-triazoles show diverse biological activity [1], [2], [3], [4], [5], [6], [7]. The most recent approaches toward the synthesis of the thiazolo[3,2-b][1,2,4]triazole system have been described [8], but only the C-5 substitution in this fused system has been studied as regards the chemical properties [9], [10], [11], [12], [13], [14], [15]. Recently there has been significant interest in developing reactions for thiazolo[3,2-b][1,2,4]triazole and [1,2,4]triazole systems’ functionalizations with the aim of expanding the range of potential bio-active compounds [1], [2], [3], [4], [5], [6], [7]. Herein, we discuss the reactivity of the [1,3]thiazolo[3,2-b][1,2,4]triazol-7-ium salts using quantum chemical calculations followed by experimental studies on the reactions of these salts with O,N-nucleophiles, leading to the formation of new polyfunctional derivatives of 1,2,4-triazole.

Results and discussion

Previously, we reported on the synthetic approach to the [1,3]thiazolo[3,2-b][1,2,4]triazol-7-ium salts via electrophilic heterocyclization [16], [17]. The literature also contains a limited number of sources that describe the chemical properties of azolinium cations; however, these data are contradictory [16], [18], [19], [20], [21], [22]. In order to investigate the reactivity of the [1,3]thiazolo[3,2-b][1,2,4]triazol-7-ium cation in more detail, in this work we used a combination of theoretical and experimental approaches.

The bromides 1 and 2 (Figure 1) were used as model salts. The starting geometries of the cations 1 and 2 were pre-optimized using the semiempirical PM7 approach [23], which is one of the best neglect of differential diatomic overlap (NDDO) methods and describes triazole systems well [24]. The re-optimization of the geometries of 1 and 2 with the DFT PBE/6-311G(d,p) level of theory [25] was the next step. This decision was based on our previous studies, where it was found that the PBE functional can model similar C, H, N, S containing organic compounds very well [26], [27]. Table 1 shows the comparison of selected bonds of the optimized geometries of cations 1 and 2 with similar experimental data. Most of the calculated parameters are in good agreement with the experimental parameters and lie within single standard deviations of the experimentally determined values. The only exception is the C33-C34 bond, where the DFT computed length is slightly overestimated as compared with variability range of the experimental bond length (Table 1). This discrepancy can be explained by the fact that none of the presented experimental geometries contain a fragment with the set of bonds chemically and structurally equivalent to the N4⁺-C33-C34 bonds.

Figure 1

The general view of optimized cations 1 and 2 with numbering. R equals H in cation 1, and R is CH₃ in cation 2.

The Fukui function [32] and molecular electrostatic potential [33] are well-established reactivity descriptors that can explain the regioselectivity of the reaction [24], [27]. The wavefunction for the production and analysis of reactivity descriptors was generated at the B3LYP/6-311G(d,p) level. It was demonstrated that the B3LYP functional [34] produces relatively good electron densities [35]. Moreover, the predictive power of reactivity descriptors, obtained by this approach and natural population analysis (NPA) partial charges, [36] is satisfactory [37]. The 6-311G(d,p) basis set [38] was chosen as it generates a reliable electrostatic potential [33]. Table 2 presents the computed reactivity indexes of the condensed Fukui function (CFF) for the nucleophilic attack of cations 1 and 2. As can be seen from Table 2, the most reactive atom for both cations is C6. The reactivity order S2>N5>C7 can also be noted. The rearrangement of 1,2,4-triazole system under the action of bases to 1,3,4-thiadiazole system has been described in the literature [31]. Less plausible is the attack at para-carbon in the C19-H29 phenyl ring. Even less probable is the substitution of Br1 atom by nucleophiles. These data strongly suggest breaking of the S2-C6 thiazoline bond under the action of nucleophiles.

These theoretical results prompted further experimental studies on the behavior of the model salts 1, 2 under the action of nucleophiles (Scheme 1). The reactions between salts 1, 2 and OH⁻ nucleophile in aqueous solutions of NaOH, Na₂CO₃, Na₂SO₃ at room temperature were studied. The cleavage of the thiazoline moiety with the formation of poly-functional symmetric triazoles 3, 4 was observed in all cases. Also, as was shown previously [16], the use of more concentrated solutions of reagents and longer reaction times lead to the formation of the elimination product 5 in the case of salt 1. The reaction is believed to proceed through the cleavage of the S2-C6 thiazoline bond by nucleophile action with the subsequent substitution of bromine. Oxidation of the thiol to a disulfide probably occurs in the presence of atmospheric oxygen and gives the corresponding target triazoles 3, 4. It should be noted that performing these reactions under heating or in different solvents leads to a decrease in selectivity [according to the nuclear magnetic resonance (NMR) spectrum of the reaction mixture].

Scheme 1

The proton nuclear magnetic resonance (¹H NMR), spectrum of compound 3 contains the characteristic signals of exocyclic saturated fragment – the multiplet at 4.79–4.87 ppm for methine proton, the multiplet for oxi-methylene protons at 4.01–4.08 ppm and the doublet of doublets for thiomethylene protons at 3.78 ppm, which (together with the signals of the OH group at 3370 cm⁻¹ and the C=O group at 1695 cm⁻¹ in IR spectrum) confirm the structure. A similar pattern is observed in the case of triazole 4. The ¹H NMR spectrum of triazole 4 contains the multiplet for SCH₂ at 3.10–3.19 ppm, the multiplet for OCH₂ at 4.15–4.21 ppm and a singlet for the methyl group protons at 1.76 ppm. The strong band at 1700 cm⁻¹ (C=O group) is present in the infrared (IR) spectrum of triazole 4; a wide signal at 3400 cm⁻¹ (OH group) is also seen. The different behavior of salts 1 and 2 is observed for the reaction of an N-nucleophile such as morpholine at room temperature with further dilution of the mixture by water (Scheme 1). Specifically, the methyl-substituted salt 2 is transformed into amino-substituted product 6, whereas a similar treatment of the salt 1 gives the elimination product 5, as earlier described by us [16]. In addition to the signals of the CH₃, SCH₂ and NCH₂ groups, the ¹H NMR spectrum of triazole 6 also contains the triplets for protons of methylene groups at 3.05 ppm and 3.45 ppm (morpholine moiety). By contrast, there are only signals of the thiopropenyl moiety in the ¹H NMR spectrum of triazole 5, namely the singlet for the SCH₂ group at 4.20 ppm and the singlets for cis and trans protons of the =CH₂ group at 6.00 ppm and 5.57 ppm, respectively. The IR spectra of triazoles 5 and 6 also show a strong band at 1700 cm⁻¹ (C=O group); whereas the signal of the OH group is absent. In the ¹³C NMR spectra of compounds 3–6, all signals of carbons belong to sp³- and sp²-hybridized carbons. The yields of the triazoles 3–6 are in the range of 64–78%, which indicates regioselectivity of these reactions.

An interesting fact for significantly different stabilities of salts 1 and 2 toward heating can be noted. Thus, salt 1 is stable upon heating in alcohol or acetic acid medium, whereas the heating of salt 2 in ethanol leads to the destruction of the thiazoline ring with the formation of dibromide 7 (Scheme 2). This difference in stability of the N4-C33 bond in cations 1 and 2 is consistent with the following computational results. First, the N4-C33 bond in cation 2 is 0.015 Å longer than that of cation 1 (Table 1). Furthermore, a comparison of the NPA partial charges shows that the N4-C33 bond is more ionic in cation 2 than in 1. Thus, the partial charges of N4/C33 in cation 1 are −0.204/−0.032, whereas in the case of cation 2, the N4/C33 partial charges are −0.216/0.149. The Fukui function is not applicable in the case of Br⁻ attack (Table 2), as it shows the absence of any reactivity of C33 atom. We already faced similar limitations of the CFF [24], [26], [27], and for correct treatment of ionic reactions with the participation of hard nucleophiles the electrostatic forces must be considered. Hence, the electrostatic potential (ESP) is expected to be a useful descriptor for understanding the reaction selectivity between cations 1 and 2 (Figure 2). Thus, in the case of cation 1, the maxima on the ESP surface A, B and C with the corresponding values of 97, 86 and 88 kcal/mol, respectively, are closest to C33. In the case of the S_N2 reaction, the attack of a nucleophile must go opposite to the leaving group (the N4 atom), but in the case of cation 1 there is no maximum in the region opposite to N4, and the angles N4-C33-X, where X can be A, B or C, are 71°, 91° and 105°, respectively. Obviously, the reaction is not favorable for cation 1. By contrast, in the case of cation 2, there is a maximum D (83 kcal/mol) located opposite to N4 at the distance of 2.50 Å from C33. The N4-C33-D angle is 141°, which makes possible the attack of bromide anion with cleavage of thiazoline system possible.

Scheme 2

Figure 2

The electrostatic potential of cations 1 (A) and 2 (B). The maxima in the N4-C33 region are shown as letters with corresponding values in parentheses (kcal/mol).

The formation of linear dibromide 7 was confirmed by analysis of the NMR spectra and the structure was resolved using single-crystal X-ray diffraction analysis. All calculations relevant to crystal structure determination were performed using the SHELX-97 package [39]. The program OLEX2 1.1 [40] was used for the analysis and visualization of the structure and for rendering the illustrations. As salt 2 is a mixture of R- and S-enantiomers, the resulting dibromide 7 is also obtained in the form of a racemic mixture. Figure 3 presents the structure of dibromide 7 that was obtained by superimposition of R- and S-enantiomers.

Figure 3

Crystal structure of the superimposed R- and S-enantiomers of the molecule 7.

The 1,2,4-triazole-3-thiol moiety is almost planar as in similar triazole systems that we have investigated previously [28], [29], [30]. The angles between the plane of the 1,2,4-triazole-3-thiol fragment and C6-C11/C12-C17 phenyl rings are 67.4°/33.0°, respectively. Intermolecular forces in the crystal are formed by the weak interaction of the N1 atom with the H-CHS group, and by the weak interactions between the 2,3-dibromopropyl fragments through the Br···H-C attraction.

Conclusions

The reactivity of [1,3]thiazolo[3,2-b][1,2,4]triazol-7-ium cation toward nucleophiles was analyzed using a combination of theoretical and experimental approaches. As a result, an efficient general synthesis of functional derivatives of 1,2,4-triazole was developed.

Experimental

¹H NMR (400 MHz) and ¹³C NMR (100 MHz) spectra were recorded in hexadeutero dimethyl sulfoxide (DMSO-d₆) as a solvent and with tetramethylsilane (TMS) as an internal standard on a Varian VXR 400 spectrometer. Melting points were determined on a Stuart SMP30 instrument. IR spectra were recorded on a Shimadzu FTIR Prestige 21 instrument with an attenuated total reflection (ATR) accessory. Elemental analyses were performed on an Elementar Vario MICRO cube analyzer. All crystallographic measurements were performed at room temperature [293(2) K] on a single-crystal diffractometer, Oxford Diffraction Xcalibur.

Supplementary material

X-ray crystallographic data.