The ring-chain transformations observed with heterocyclic structures are a characteristic type of reactivity for this broad class of compounds. Such reactions are well known in the chemistry of hetarenes, but not for carbocyclic arenes. At the same time, they are essential from synthetic point of view, especially in planning the routes of synthesis for the preparation of condensed bi- and polycyclic compounds.1,2,3,4,5,6,7,8,9, 10

The occurrence of ring-chain transformations and understanding of the factors affecting the shifting of equilibrium and stability of the cyclic or open-chain forms are important for performing directed intramolecular cyclization reactions leading to the formation of condensed molecules.1,2,3,4,5,6,7,8,9,10,11,12, 13

Several types of ring-chain transformations exist, which are based on intramolecular cyclization reactions of openchain O-, N-, and С-nucleophiles with electron-deficient carbon atoms of cyclic systems containing strongly polar С=О, С=N, or С=S bonds. Another type of cyclization is associated with intramolecular attack by electron-deficient terminal atoms of the open-chain form on the ring system heteroatoms bearing a lone pair of electrons. Theoretical studies of cyclization reactions involving azide group and the lone pair of electrons belonging to nitrogen atom13 have provided a better theoretical understanding of pericyclic reactions and supported the development of the concept of pseudopericyclic14 and heteroelectrocyclic15 reactions. The latter type also includes the cyclization of azido group located at α-position relative to a heterocyclic nitrogen atom belonging to an azole or azine ring, leading to the formation of tetrazole isomer.16 Such phenomenon is known as azido-tetrazole tautomerism and can be considered to be one of the basic properties of hetaryl azides. This type of ring-chain transformations is reversible and proceeds spontaneously or during the preparation of solutions of such compounds.

A large number of published studies are devoted to the transformations of azide isomer to the tetrazole form in various heterocyclic systems. A substantial contribution to the understanding of these transformations was made by Academician I. Ya. Postovskii, who was one of the first to study these transformations both in the series of tetrazoloazines17,18, 19 and tetrazoloazoles20 by using for this purpose IR spectroscopy in solutions, solid state, and melts. The majority of these pioneering studies and the later works by other authors have been described in review articles.16 , 21,22, 23 In addition, the cyclization of some hetaryl azides to the tetrazole isomer can produce two cyclic forms (Scheme 1). This type of heterocycles includes compounds A derived from 2-azidopyrimidines (Y = CR2, Z = CR3), 2-azido-1,3,5-triazines (Y = N, Z = CR3), and 3-azido-1,2,4-triazines (Y = CR2, Z = N), lacking N-alkyl or N-oxide groups at positions 2 and 4 of the triazine ring. This type of azides A is capable of transformations both to tetrazole T' and to its isomeric cyclic form T (Scheme 1). The characteristic structural difference of 1,2,4-triazine derivatives T' (Y = CR2, Z = N) and T (Y = CR2, Z = N) is evident in the type of ring fusion between the azine and azole moieties. In this case, it can be specified as [5,1-с] and [1,5-b], respectively.

scheme 1

Scheme 1

The unpredictable behavior of azide А requires that researchers working in this field of chemistry not only perform the synthesis of these hetaryl azides and establish the azido-tetrazole equilibrium, but also pay particular attention to proving the structure of tetrazole isomers Т' and Т.

It should be noted that in recent years there is a renewed interest in 3-azido-1,2,4-triazines and their tetrazole analogs, which belong to the group of heterocycles with high nitrogen content. Due to such structure, this class of compounds is considered as a template for the design of new high energy materials.24 , 25 Besides that, the aforementioned series of compounds are also used in search for biologically active compounds.26,27, 28 For these reasons, it is necessary to study the azido-tetrazole equilibrium of 3-azido-1,2,4-triazines A in order to facilitate the future investigations regarding the mechanisms of biological interactions of active structures. An example of practical results from applying tetrazole derivatives of 3-azido-1,2,4-triazines in the synthesis of biologically active compounds is the development of a new method for the preparation of the antiepileptic drug lamotrigine29 (Scheme 2). This approach was based on the reduction of the corresponding tetrazolo[1,5-b][1,2,4]-triazine 1Ta. Obviously, it would be impossible to perform this transformation without the existence of an azidotetrazole equilibrium between the isomeric structures 1Ta and 1Aa. Thus, it is essential to study the ring-chain transformations in the series of 3-azido-1,2,4-triazines in order to enable practical applications of these compounds.

scheme 2

Scheme 2

A literature survey showed that the methods for the preparation of 3-azido-1,2,4-triazines A can be divided into three main synthetic approaches. The first approach involves the treatment of the respective hetarylhydrazines with nitrous acid. The second approach is based on the substitution of good leaving groups at position 3 of 1,2,4-triazine ring with azide groups. The third variant for the synthesis of compounds А involves condensation of the azine moiety with tetrazole ring.

Synthesis from 3-hydrazino-1,2,4-triazines

The reaction of 3-hydrazino-1,2,4-triazines with nitrous acid is one of the traditional routes for the preparation of hetaryl azides. This method was successfully used for converting compounds 2ah to azidotriazines 3Aah, which spontaneously isomerized to tetrazolo[1,5-b][1,2,4]triazines 3Tah (Scheme 3).30 This fact was confirmed by the absence of an absorption band in the range of 2020–2050 cm–1 when examining IR spectra of products 3Tah, which were recorded in KBr and in Nujol. Besides that, the signals of azido group appeared after some time in IR spectra of samples 3Таd that were recorded in СH2Cl2 or CHCl3 solutions. These observations pointed to the reversibility of the conversion of azido species 3Aah to the cyclic isomers 3Tah. The use of CDCl3 as a solvent enabled the monitoring of this rearrangement by 1Н NMR spectroscopy. However, only one isomer existed in DMSO-d 6 solution, which probably was the tetrazole 3Tah. The structure of cyclization products – compounds 3Aah – was conclusively established by using X-ray structural analysis of a sample of compound 3Tf, showing that a type [1,5-b] fusion existed between the tetrazole and 1,2,4-triazine rings.

scheme 3

Scheme 3

It should be also noted that the formation of alternative tetrazole form 3T'ah was not observed in this case, and the authors proposed that the formation of tetrazolo[5,1-c]-[1,2,4]triazines from 3-azido-1,2,4-triazines is unlikely.

An analogous result was obtained by diazotation of 3-hydrazino-5,6-diphenyl-1,2,4-triazine (2i). This reaction led to azide 3Ai, which was converted to tetrazole 3Ti.31 The type of ring fusion between the tetrazole and triazine moieties in this case was established by comparing the absorption spectra of compound 3Ti and selected model structures – pyrazolo[5,1-c][1,2,4]triazine 4 and triazolotriazine 5.

The treatment of hydrazino-1,2,4-triazinones 6ac with nitrous acid led to azides 7Aac (Scheme 4), which spontaneously isomerized to tetrazoles 7Tac.32 Similarly to the case of compounds 3Aah (Scheme 3), the signals of azide group in IR spectrum of product obtained from hetarylhydrazine 6a were observed in the range of 2170–2180 cm–1 both in CHCl3 solution and in Nujol. In the rest of the cases only the signals of cyclic form were observed. The type of ring fusion between the azole and azine moieties in tetrazolotriazines 7Tac was established by comparing 13C NMR chemical shifts of carbon atoms with the analogous signals of model compounds – deaza analogs 1,2,4-triazolo[5,1-c][1,2,4]triazine 8 and 1,2,4-triazolo[1,5-b]-[1,2,4]triazine 9. As shown in Scheme 4, the chemical shifts of 1,2,4-triazine rings in the carbon spectra of compounds 7Tb and 9 were practically identical. This observation led to the conclusion that the products obtained in cyclization reactions of azides 7Aac were tetrazolo[1,5-b]-[1,2,4]triazines 7Тас. It should be emphasized that the isomerization reactions of compounds 7Ab,c have been investigated previously.33 , 34 However, those studies considered only the variant involving the formation of structures 7T'b,c, while data of elemental analysis were used for proving the structure of the tetrazole isomer. Apparently, the conclusions about the structure of the products obtained in cyclization of azides 7Ab,c were erroneous in this case.

Scheme 4.
scheme 4

*

Azido-tetrazole tautomerism was also investigated for product 11A, obtained by treatment of 3-hydrazino-1,2,4-benzotriazine 10 with nitrous acid (Scheme 5).35 , 36 Thus, IR spectrum recorded for CHCl3 solution showed the absorption band of azido group (2150 cm–1), which was absent when recording the spectrum in KBr. These observations led the authors to a conclusion on the transformation of azide 11A to isomer 11T'. At the same time, no solid evidence was presented about the structure of the proposed compound 11T', as well as the possibility of the formation of another isomer 11T was not considered.

Scheme 5.
scheme 5

*

Partial attempts were made in another study37 to establish the direction of isomerization reaction using azide 11A. For this purpose, the UV spectra of the polycyclic structure 11T' and model compounds 12 and 13 were compared. It was found that the spectral characteristics of compounds 11T' and 12 were very close. Further, the authors performed a more detailed analysis of the azidotetrazole tautomerism by using NMR spectroscopy in DMSO-d 6 solution. This method allowed to observe a triple set of signals for compounds 11T', 11A, and 11T in the ratio of 6.5:2.5:1. At the same time, the analysis of chemical shift values for some benzene ring signals pointed to the similarity between 13С NMR spectra of tetrazole 11T' and the model compound – triazolotriazine 14 (Scheme 5). This result provided an additional support to the assumption that the main component of equilibrium mixture, obtained from the reaction product of hetarylhydrazine 10 with nitrous acid, was compound 11T', featuring a [5,1-c] type of ring fusion between the tetrazole and 1,2,4-triazine moieties. Even though this approach is difficult to present as straightforward and ideal, it was possible to establish diagnostic criteria, which allowed to distinguish isomers 11T' and 11T in this series of compounds. Among these characteristics are the chemical shifts of carbon atoms (115.4 and 127.0 ppm, underlined in Scheme 5), which carry hydrogen atoms and are located in the benzene ring, separated by two covalent bonds from the N-4 atom of the triazine ring.

Performing the diazotation reaction of hetarylhydrazine 10 with 15N-labeled NaNO2 (15N, 98%) allowed to selectively introduce this isotopic label into the azido group.38 As a result, compound 11*A was synthesized (Scheme 6). The introduction of label allowed to directly study the azido-tetrazole equilibrium via the analysis of long-range 1H–15N coupling constants (4–6 J HN), without resorting to the use of model compounds, and to observe not only the transitions of cyclic forms to azides, but also to unequivocally prove the structure of the tetrazole isomers. The spin-spin coupling constants between the labeled nitrogen atom and benzene ring protons were measured by using spin echo experiments in 1D 1H NMR spectra. The determined values of the constants were in the range from 0.17 to 0.04 Hz. Similarly to the previous example,37 the spectra acquired for DMSO-d 6 solutions showed a mixture of three isomeric species 11*T', 11*A, and 11*T in 79:12:9 ratio according to the integrated signal intensities. The further analysis of long-range 1H–15N constants showed that all protons in isomer 11*T' participated in spin-spin interaction with the labeled atom, while in the case of tetrazole 11*T the 4–6 J HN constant was measured only for two proton signals (Scheme 6). Long-range 1H–15N coupling constants were not observed for azide 11*A. The values of 4–6 J HN constant allowed to unequivocally confirm the structure of each isomer. The addition of a small amount of trifluoroacetic acid (TFA) to the DMSO-d 6 solution caused insignificant changes in the isomer ratio. The use of pure deuterated TFA (ТFA-d) shifted the equilibrium practically toward azide 11*А. Only small amounts of tetrazole isomer 11*T' (~1%) could be detected in ТFA-d solution. However, despite the very low concentration of cyclic isomer 11*T', the 1H–15N constants for this form were successfully detected. The presence of 15N atom in the structure of these isomers gave an additional opportunity to use 1D 15N NMR spectra for the characterization of the equilibrium. The labeled nitrogen signal for tetrazoles 11*T' and 11*T was observed in these experiments in the range from –36 to –30 ppm, while the chemical shifts of the azido species were observed in the range from –140 to –130 ppm. It is important to note that the measurements of 1H–15N spin-spin coupling constants confirmed the correlation of labeled atom signals in onedimensional 15N NMR spectra.

Scheme 6.
scheme 6

*

Various research groups have investigated the interaction of 3-hydrazino-1,2,4-triazino[5,6-b]indoles 15ad with nitrous acid, generated in phosphoric, acetic, or hydrochloric acid media (Scheme 7).39,40,41, 42 For example, compounds 16Aad were obtained, which spontaneously transformed into the tetrazole form, as confirmed by the absence of absorption bands due to the azido group and the presence of tetrazole ring signals in the range of 1180–1220 cm–1 in IR spectra recorded in KBr.39 , 40 , 42 The ring fusion type between the tetrazole and triazine moieties in compounds 16T'ad was supported by the literature data.37

scheme 7

Scheme 7

Performing the reaction in concentrated H3PO4 allowed to detect the formation of azide 16Aa, which was converted by refluxing in acetic anhydride into heterocycle 16T'e, with the cyclization of azido group also accompanied by an acylation reaction (Scheme 7). The preparation of 3-azido-1,2,4-triazine 16Aa was confirmed by data of IR spectroscopy by the observation of azido group absorption band at 2150 cm–1.41

In order to unequivocally establish the direction of the cyclization reaction of azide 16Aa, experiments were performed with the isotopically labeled compound (Scheme 8).43 The use of K15NO2 in acidic medium in the diazotation reaction of compound 15a allowed to obtain azide 16*Aa, which spontaneously cyclized to tetrazole 16*Ta. The study of labeled sample by 13C NMR spectroscopy in DMSO-d 6 solution revealed the presence of 2-3 J CN constants for two carbon atoms of the 1,2,4-triazine ring of the tetrazole isomer. These spectral properties proved that compound 16*Aa was converted by cyclization into structure 16*Ta, in which the tetrazole and azine moieties were fused according to the type [1,5-b]. It is important to note that the alternative cyclic isomer 16*T'a could not be detected. At the same time, in the case of formation of compound 16*T'a, 13С–15N constants should be observed in one-dimensional 13С NMR spectrum for all carbon atom signals of the 1,2,4-triazine ring.

Scheme 8.
scheme 8

*

The study of azido-tetrazole equilibrium of compound 16*Ta in TFA-d solution allowed to establish that 30 days after dissolution the ratio between structures 16*Аа and 16*Та reached 87:13. The presence of a labeled nitrogen atom provided an opportunity to observe ring-chain interconversion by the method of 15N NMR spectroscopy, while the measurements of 13С–15N spin-spin coupling constants confirmed that the [1,5-b] ring fusion type between the tetrazole and triazine moieties in acidic medium was conserved in the case of isomer 16*Та. Besides that, azide 16*Aа was characterized by a single 3 J CN constant. Thus, the introduction of 15N isotope in the azole ring of tetrazolo[1,5-b][1,2,4]triazines allowed to directly observe not only the transitions of tetrazoles to azides, but also to prove the structure of the cyclic forms.

The reactivity of the 3-azido-1,2,4-triazine naphtho derivatives 18A differed from that of 3-azido-1,2,4-benzotriazines 11A.44 Thus, compound 18A obtained by treatment of heterocycle 17 with nitrous acid was cyclized to tetrazolo[1,5-b]triazine 18T (Scheme 9).

scheme 9

Scheme 9

The presented process was monitored by the method of IR spectroscopy. The spectrum of the crude product featured a strong azido group absorption band at 2120 cm–1 (KBr), which disappeared after recrystallization from dioxane. At the same time, 1Н and 13С NMR spectra acquired in DMSO-d 6 solutions showed the signals of a single tetrazole form 18T. The performed X-ray structural analysis confirmed the hypothesis about cyclization of azide 18A to tetrazolo[1,5-b][1,2,4]triazine 18T.

Furthermore, another case of azido-tetrazole tautomerism for 1,2,4-triazines is discussed in the literature.44 It was found that the product from interaction between compound 19 and НNO2, namely, azide 20A, was isomerized to cyclic form 20T, in which the tetrazole and triazine moieties were joined by type [1,5-b] ring fusion (Scheme 10). The conclusion about the structure of this compound was based on the results of X-ray structural analysis. The conversion of azide 20A to the tetrazole isomer could be established by using IR spectroscopy. Initially, two absorption bands of azido group were observed in IR spectrum of the diazotation product obtained from compound 19 (2150 and 2120 cm–1, КBr). After compound 20А was recrystallized from EtOH, the characteristic azide signals were absent from IR spectrum, confirming the conversion of open-chain form to cyclic isomer 20T.

scheme 10

Scheme 10

Two additional examples have been described for the conversion of 3-azido-1,2,4-triazines 22A and 24A to tetrazolo[1,5-b][1,2,4]triazines 22T and 24Т 45 (Schemes 11 and 12). When NaNO2 was added to solutions of compounds 21 and 23 in aqueous AcOH, respective azides 22A and 24A were formed. The structure of compounds 22A and 24A was supported by data of IR spectroscopy (absorption band at 2140 cm–1 in СНBr3 solution). Refluxing of azidotriazines 22А and 24А in EtOH led to polycyclic structures 22Т and 24T. Alternative structures 22T' and 24T' were excluded from consideration on the basis of results presented in earlier publications.32 , 37

scheme 11

Scheme 11

scheme 12

Scheme 12

The standard procedure for the conversion of hetarylhydrazine 25 to azide 26A was used for obtaining analogs of the antibiotic 2-methylfervenulone46 (Scheme 13). The presence of absorption band in IR spectrum at 2200 cm–1 confirmed the formation of 3-azido-1,2,4-triazine 26A. Heating of azide 26A to 150°C in DMF medium led to tetrazolotriazine 26T. At the same time, the authors recognized that the isomerization of azide can proceed by an alternative route, leading to the formation of compound 26T'. The type of ring fusion between the azole and azine moieties, resulting from the cyclization of azide 26A, was distinguished on the basis of previously published results,30 as well as comparison of IR and UV spectral data with model compound 27T, which is known to show only a single type of fusion between the tetrazole and triazine rings.

scheme 13

Scheme 13

The reaction of 2-hydrazino-1,2,4-triazine 28 with NaNO2 in concentrated H3PO4 medium has been described in the literature47 (Scheme 14). The authors noted that IR spectrum of the obtained compound did not contain an absorption band in the range of 2120–2150 cm–1. For this reason, tricyclic structure 29T' was assigned to the product obtained by diazotation reaction of compound 28. Despite of the two possible directions of cyclization reaction in the case of azide 29A, there were no arguments presented in support of the formation of heterocycle 29T', in which the azole and triazine rings were condensed according to type [5,1-c] ring fusion.

scheme 14

Scheme 14

A different manifestation of azido-tetrazole tautomerism was present in studies devoted to diazotation of 3-hydrazinopyrazolo[4,3-е]triazines 30aс and 32 48,49, 50 (Schemes 15 and 16). It was demonstrated that the reaction resulted in the formation of azides 31Aaс and 33A, which upon crystallization from EtOH or spontaneously cyclized to the corresponding tetrazoles 31Taс and 33T.

scheme 15

Scheme 15

scheme 16

Scheme 16

The structure of azides 31Aa–c and 33A was confirmed by using the data of IR spectroscopy. In the case of compound 31Aа the process of conversion into the cyclic isomer 31Tа was observed by using 1H NMR spectroscopy in CDCl3 solution. The type of ring fusion for compounds 31Та–c was unequivocally established by method of X-ray structural analysis.

In the case of cyclization reaction using azide 33A, the most preferred direction upon initial consideration was the route leading to compound 33T'. However, the performed X-ray structural analysis showed that the condensation of compound 33A led to the formation of betaine structure 33T.

Preparation of 2-azido-1,2,4-triazines by nucleophilic substitution reaction

The second method for the synthesis of 3-azido-1,2,4-triazines and their tetrazole isomers is based on the nucleophilic substitution of a good leaving group at position 3 of 1,2,4-triazine ring with an azido group. Such an approach enabled direct preparation of this class of hetaryl azides.

An alternative method of synthesis was proposed for the previously presented compound 3Ai, including the treatment of chlorotriazine 34 in acetone with aqueous NaN3 solution (Scheme 17).51 In the same work, azidotriazine 36A was obtained by a similar method from chloro derivative 35 (Scheme 18).

Scheme 17.
scheme 17

scheme 18

Scheme 18

The structures of compounds 3Ai and 36A were proved on the basis of IR spectra recorded in KBr, which contained characteristic absorption bands at 2137 and 2135 cm–1, corresponding to the signals of azido groups. Despite the fact that compounds 3Ai and 36A were capable of conversion into isomers 3Ti and 36T or 3T'i and 36T', the authors did not consider the azido-tetrazole tautomerism for the given structures, instead emphasizing the investigation of reactivity toward trialkyl and diethyl phosphonates.

Another use of NaN3 was in the case of chloride substitution in pyridotriazine 37 (Scheme 19).35 , 36 That reaction proceeded in aqueous acetone, and the authors were able to separate two isomeric forms 38A and 38T'. Compound 38A was observed by IR spectroscopy both in solution phase and in solid state, but was found to be less stable than tetrazole form 38T'. Thus, compound 38A underwent gradual spontaneous transformation in DMSO-d 6 solution or in crystalline state, producing tetrazole isomer 38T', which was confirmed by 1H NMR and IR spectral data. The main problem during the study of azido-tetrazole tautomerism of compound 38A was the fact that no consideration was given to the variant of cyclization leading to structure 38T. Besides that, there was clearly insufficient data to support the formation of structure 38T'.

scheme 19

Scheme 19

Another example for the application of NaN3 in the substitution of a halide ion (Scheme 20) has been presented for benzotriazines 39ас,52 resulting in the synthesis of azides 40Aac. The azido-tetrazole equilibrium of triazine 40Aac derivatives was studied by using a combination of methods based on 1H NMR and IR spectroscopy.

Scheme 20.
scheme 20

*

Thus, 1H NMR spectra of compounds 40Aac, which were acquired in DMSO-d 6 and acetone-d 6 solutions, showed the predominance of tetrazole forms, or the content of one of the tetrazole isomers was close to the azide concentration. In the case of heterocyclic compound 40Ab and dimethyl derivative 40Ac, formation of the linear isomers T was observed, while the cyclization of azide 40Aa resulted in the formation of structure 40T'a. When CDCl3 was used as the solvent, the equilibrium was shifted toward the open form, resulting in the domination of azides 40Aac. The assignment of proton signals in azides 40Aас contained in complex mixtures in comparison to the tetrazole isomers was performed by taking into account the data of 1H NMR spectra acquired for bromo derivatives 39ac. It should be noted that the type of ring fusion between tetrazole and triazine moieties was established on the basis of comparing the proton chemical shifts in 1H NMR spectra of the annulated benzene ring of tetrazole forms 40Tac and 40T'ac with those of model compounds, namely, the respective derivatives of imidazo-[1,2-a][1,2,4]triazines. An example of comparison between some 1H NMR spectral signals of isomers 40Ta and 40T'a with compounds 41 and 42 is shown in Scheme 20. These characteristics show that the chemical shifts of benzene ring protons depend on the type of ring fusion between the azole and azine moieties.

The product obtained by reaction of compound 43 with NaN3 was studied by method of IR spectroscopy, which allowed to observe spontaneous cyclization of heterocyclic compound 44A to the tetrazole isomer44 (Scheme 21). In this case, the formation of compounds 44T' or 44T was possible. At the same time, 1Н NMR spectrum acquired for DMSO-d 6 solution of the cyclization product obtained from azido-1,2,4-triazine 44A showed that only one of the two possible tetrazole forms was present. The type of ring fusion between the tetrazole and triazine moieties in this case was established on the basis of 13С NMR spectroscopy data. Following the example of previous work,37 the selected informative feature was the chemical shift of naphthaline ring carbon atom bearing a hydrogen atom and separated by two covalent bonds from the N-4 nitrogen atom of the triazine ring (Scheme 21). When comparing this chemical shift with the analogous signals of benzotriazine 11T' and compound 18T, the product obtained by cyclization of azide 44А was assigned with structure 44T'.

Scheme 21.
scheme 21

*

The reaction of heterocyclic compound 45 with an equimolar amount of NaN3 allowed to obtain tetrazolotriazine 46T by substitution of the sulfonyl group (Scheme 22).53 The tetrazole type of structure for the product obtained by sulfonyl group substitution was determined by IR spectroscopy of crystalline sample, which showed an absence of absorption in the range of 2100–2200 cm–1. It is important to note that the opening of tetrazole ring in compound 46T did not occur even in TFA medium.

scheme 22

Scheme 22

Despite the two possible directions for cyclization of azide 46A (leading to products 46T and 46T'), the aforementioned study did not consider the issue of proving the type of triazine ring fusion.

Nucleophilic substitution by using NaN3 as a reagent was identified as an alternative method for the synthesis of azido and tetrazolo derivatives of azafervenulin, including compound 49Aa.54 Thus, the treatment of compounds 47, 48 with NaN3 led to heterocyclic products 49Aa and 49Tb (Scheme 23). The structures of compounds 49Aa and 49Tb were confirmed on the basis of IR spectral data. The further cyclization of azide 49Aa to tetrazole was not performed, in contrast to another earlier work.46

scheme 23

Scheme 23

The [1,5-b] type of ring fusion between the tetrazole and triazine moieties in compound 49Tb was unequivocally established by X-ray structural analysis, as was described in a later article.55 The study of azido-tetrazole equilibrium of compound 49Tb by 1Н NMR spectroscopy showed that two forms existed in DMSO-d 6 solution – tetrazole form 49Tb and azide form 49Ab. The addition of TFA resulted in shifting of the equilibrium toward the azide form, while the tetrazole isomer was predominant in pyridine. The diagnostic features in this case were the proton signals of N-methyl groups, which did not allow to exclude the possible formation of alternative structure 49T'b.

Synthesis of 3-azido-1,2,4-triazines and their cyclic isomers on the basis of tetrazole derivatives

The diazotation of 5-aminotetrazole and the subsequent coupling of diazoazole 50 with compounds containing an activated CH group served as obe of the approaches to the construction of 1,2,4-triazine ring on the basis of azole moiety. One of the first synthetic studies in this direction was published in 1974 by uzing β-naphthol 51 as the azo component (Scheme 24).56 The coupling reaction was performed under basic conditions with efficient cooling and resulted in the preparation of azo compound 52, which, according to the authors of that study, was converted to tetrazolo[5,1-c][1,2,4]triazine 20T'.

scheme 24

Scheme 24

As established by method of IR spectroscopy, the product from the cyclization of compound 52 existed in crystalline state in its tetrazole form, while in DMSO-d 6 or СDCl3 solutions polycyclic structure 20T' started to isomerize to azide 20A. The type of ring fusion between the tetrazole and triazine rings was established by comparing the UV spectra of compound 20T' and model structures 53 and 54. As a result of this analysis, the isomerization of heterocyclic compound 20T' to tetrazole 20T was excluded.

The conclusions from this work were subsequently revised in an article by the same authors, where 6-bromo-2-naphthol 55 was used as an additional reagent with activated CH group.57 Thus, compound 56 was obtained, which was converted to triazine 57T. The tetrazole character of the obtained isomer was proved by IR spectrum, recorded in KBr pellets. The presence of azidotetrazole tautomerism was also confirmed by studies performed using 1H NMR spectroscopy. The proton spectrum acquired in CDCl3 solution for the condensation product obtained from azo compound 56 contained a double set of signals due to compounds 57A and 57T, showing equal integrated intensity. At the same time, only one isomer 57T was observed in 1H NMR spectra acquired for a DMSO-d 6 solution. The type of fusion between the tetrazole and triazine rings in this case was established by comparing the chemical shifts of proton signals arising from tetracyclic structure 57T and model compounds 58 and 59 (Scheme 24). The main diagnostic features in 1H NMR spectrum, which were used for determining the structure of compound 57T, were the signals of naphthyl ring protons. The introduction of bromine atom in this case significantly simplified the coupling pattern compared to the cyclization products obtained from compound 52. This enabled a more complete comparison of 1Н NMR spectra of tetrazolo[1,5-b][1,2,4]triazine 57T and bromosubstituted azolo[5,1-c][1,2,4]triazines 58 and 59. Thus, the main conclusion of this work showed that the condensation of hydrazones 52 and 56 gave linear structures 20T and 57T, which represented products from the rearrangement of tetrazole isomers 20T' and 57T'. Obviously, the presented process proceeded through respective azides 20А and 57А. Besides that, an alternative method was presented for the preparation of compound 20T, and the conclusions about the structure of this tetrazolotriazine, which were given in the literature,44 , 57 were in good agreement.

The interaction of diazonium salt 50 with various compounds containing an activated CH group provided a convenient method for the synthesis of 2-azido-1,2,4-triazines and their tetrazole analogs. Thus, the reaction of α-formylphenylacetonitriles 60ag and compound 50 led to hydrazones 61ag (Scheme 25).29 At the same time, the coupling process was accompanied by elimination of the formyl group, which was used for the activation of CH acidity of arylacetonitriles, which themselves did not react with diazotetrazole 50. The cyclization of hydrazones 61ag proceeded upon refluxing in DMF and led to structures 1Tag. Similar transformations were also described for hydrazones 63a,b, obtained by using β-dicarbonyl compounds 62a,b.58 It is important to note that the existence of these structures in the cyclic form 64a,b was identified in that study for hydrazones 63а,b, confirming the route of formation for tetrazoles 65T through azides 65A and tetrazolo[5,1-c][1,2,4]triazines 65T'. The conversion of compounds 64a,b to triazines 65T'a,b and their further isomerization to tetrazoles 65Ta,b proceeded upon heating in toluene medium in the presence of p-TsOH (Scheme 25). The structure of compounds 1Tag and 65Та,b was confirmed by data of X-ray structural analysis.

Scheme 25.
scheme 25

*

Another example for the transformation of tetrazolo[5,1-c]-[1,2,4]triazines T' to isomeric tetrazolo[1,5-b][1,2,4]triazines T was presented by using products from the cyclization of hydrazone 67, which was obtained by using nitroacetic ester 66 (Scheme 26).59 The transformation of compound 67 to tetrazolotriazines proceeded upon refluxing with aromatic amines 68ad. Two routes are possible for this transformation (Scheme 26). The route A involves substitution of the nitro group in hydrazone 67 and the preparation of compounds 69ad, which are further cyclized to heterocyclic intermediates 70T'ad, followed by isomerization to tetrazolo[1,5-b][1,2,4]triazines 70Tad via azido-tetrazole rearrangement. An alternative variant of the reaction route (the route B) involves cyclization of hydrazone 67 as the first step. The substitution of nitro group in that case occurs in the already formed triazine ring of tetrazole derivative 71T' or azide 71A. The lack of IR absorption bands of azido and nitro groups in spectra of tetrazolo[1,5-b][1,2,4]triazines 70Tad proved that the formation of tetrazole derivative was accompanied by the substitution of nitro group. The structure of compounds 70Tad was conclusively proved on the basis of X-ray structural analysis of triazine 70Tc.

scheme 26

Scheme 26

The syntheses of 3-azido-1,2,4-triazine 76A and tetrazolotriazine 77T' were reported as examples of azo coupling reactions of diazotetrazole 50 with malonodinitrile 72 and ethyl cyanoacetate 73 (Scheme 27).60 , 61 This process proceeded via the formation of hydrazones 74 and 75, which cyclized spontaneously (compound 74) or upon refluxing in AcOH (compound 75). The initially formed tetrazolo[5,1-c][1,2,4]triazine 76T' spontaneously converted to open-chain form 76A according to this mechanism, while the condensation of compound 75 gave heterocyclic product 77T'. The structure of azide 76А was confirmed by X-ray structural analysis. The azido-tetrazole equilibrium for the cyclization product obtained from hydrazone 74 was not studied by other methods. The bicyclic structure of compound 77T' was confirmed by the absence of absorption band in the range of 2100–2200 cm–1 in the IR spectrum. Besides that, the structure of triazine 77T' was additionally proved by using elemental analysis and UV spectroscopy. The methods described for the structural characterization of compound 77T' did not provide a clear confirmation of the ring fusion type in the bicyclic structure and did not solve the issue about its possible isomerization to tetrazolo[1,5-b]triazine 77T.

scheme 27

Scheme 27

The structure of cyclization product obtained from hydrazone 75 could be conclusively established by performing the coupling reaction with cyanoacetate 73 using 15N-labeled diazotetrazole 50* (Scheme 28).43 Compound 75*, which was thus obtained, exhibited prototropic tautomerism by existing in two forms a and b, which spontaneously cyclized through tetrazoles 77*T'a,b and azides 77*Аa,b to a mixture of isotopomers 77*Ta,b. The reaction with ethyl phenylacetate formyl derivative 78 proceeded analogously.43 , 62 The cyclization of tautomeric forms a and b of hydrazone 79* led to tetrazolotriazines 80*Ta,b with different positions of the 15N isotopic label in the azole ring. The presence of 15N atom in the structures of compounds 77*Ta,b and 80*Ta,b allowed to use the analysis of 13С–15N spin-spin coupling constants for establishing and confirming the structures.43 The carbon spectra of compounds 77*Ta,b and 80*Ta,b showed interaction of the 15N atom with two carbon atoms of the 1,2,4-triazine moiety. These spectral features unequivocally confirmed the [1,5-b] type of ring fusion between the tetrazole and triazine moieties in compounds 77*Ta,b and 80*Ta,b, since the alternative bicyclic structures 77*T'a,b and 80*T'a,b would show 13С–15N coupling constants at each carbon atom of the azine ring, while in the case of azides 77*Aa,b and 80*Aa,b only one carbon atom showed a spin-spin coupling to the labeled atom. Despite the expectations that cyclization of hydrazones 75*a,b and 79*а,b should lead to tetrazolo[5,1-c][1,2,4]triazines, the presence of structures 77*T'a,b and 80*T'a,b was not detected. X-ray structural analysis of unlabeled compound 77T completely confirmed the data obtained on the basis of analysis of 13С–15N coupling constants.

Scheme 28.
scheme 28

*

Besides that, the 13С–15N spin-spin coupling constants and the chemical shift values of labeled atoms in 1D 15N NMR spectra were found to be convenient diagnostic characteristics for studying the azido-tetrazole tautomerismin the given series of compounds (Scheme 28). The significant differences in chemical shifts of the labeled atoms allowed to easily detect ring-chain transformations. By relying on these characteristics it was established that in TFA-d solution the mixture of compounds 77*Ta,b completely rearranged over 12 h to azides 77*Аa,b. At the same time, the product from cyclization of hydrazone 79* in TFA-d solution showed a 60:40 ratio of compounds 80*Аa,b and 80*Тa,b when observed 30 days after dissolution.

The condensation of diaminotetrazole 81 with α-dicarbonyl compounds 82ac provided another example for the construction of tetrazolo[1,5-b][1,2,4]triazines on the basis of azole ring (Scheme 29).63 The reaction of azole 81 with glyoxal 82a and 2,3-butanedione 82b led to the formation of compounds 83Ta,b. On the other hand, the reaction of diamine 81 with methylglyoxal 82с provided a mixture of tetrazolotriazines 83Tc,d.

scheme 29

Scheme 29

The structures of obtained compounds 83Tad were established on the basis of data available in an earlier publication.30 This conclusion was further confirmed by the results of X-ray structural analysis for compound 83Td, which were published in another article.24

The analysis of literature data and the results of studies from our laboratory show that the azido-tetrazole equilibrium in the series of 3-azido-1,2,4-triazines, as a rule, is shifted toward the formation of tetrazolo[1,5-b]-[1,2,4]triazines, and the explanation of this fact requires both additional experimental data and theoretical studies. At the same time, a range of examples are known where the existence of open-chain form and another cyclic isomer (tetrazolo[5,1-c][1,2,4]triazine) has been reported. Therefore, the chemists working in this area need to be very careful with structural characterization and observation of ring-chain equilibria in this series of heterocycles.

As already noted above, X-ray structural analysis is certainly the definitive method for unequivocal determination of the isomeric forms of azidotriazines in crystalline state. At the same time, the study of this type of transformations in solution phase can be effectively accomplished by using modern 1H and 13С NMR experiments. Furthermore, the potential benefit from applying NMR methods significantly increases when labeled nitrogen atoms are introduced into the azido group, leading to the appearance of additinal spectral features, such as 1H–15N and 13С–15N spin-spin coupling constants. Besides that, the introduction of 15N isotopic labels enables the use of 15N NMR spectral data for the characterization of azido-tetrazole equilibrium. Unfortunately, the full potential of this method has not yet been revealed. It can be said with confidence that the further studies using the full range of modern NMR methods with isotopically labeled azaheterocycles will allow to shine the light not only on the nature of azido-tetrazole rearrangements, but also will be useful for the study of other types of ring-chain transformations.