Approach to Pyrido[2,1-b][1,3]benzothiazol-1-ones via In Situ Generation of Acyl(1,3-benzothiazol-2-yl)ketenes by Thermolysis of Pyrrolo[2,1-c][1,4]benzothiazine-1,2,4-triones

Acyl(imidoyl)ketenes are highly reactive heterocumulenes that enable diversity-oriented synthesis of various drug-like heterocycles. Such ketenes, bearing heterocyclic substituents, afford angularly fused pyridin-2(1H)-ones in their [4+2]-cyclodimerization reactions. We have utilized this property for the development of a new synthetic approach to pharmaceutically interesting pyrido[2,1-b][1,3]benzothiazol-1-ones via the [4+2]-cyclodimerization of acyl(1,3-benzothiazol-2-yl)ketenes generated in situ. The thermal behaviors of 3-aroylpyrrolo[2,1-c][1,4]benzothiazine-1,2,4-triones and 3-benzoylpyrrolo[2,1-b][1,3]benzothiazole-1,2-dione (two new types of [e]-fused 1H-pyrrole-2,3-diones reported by us recently) have been studied by thermal analysis and HPLC to elucidate their capability to be a source of acyl(1,3-benzothiazol-2-yl)ketenes. As a result, we have found that only 3-aroylpyrrolo[2,1-c][1,4]benzothiazine-1,2,4-triones are suitable for this. The experimental results are supplemented with computational studies that demonstrate that thermolysis of 3-aroylpyrrolo[2,1-c][1,4]benzothiazine-1,2,4-triones proceeds through an unprecedented cascade of two thermal decarbonylations. Based on these studies, we discovered a novel mode of thermal transformation of [e]-fused 1H-pyrrole-2,3-diones and developed a new pot, atom, and step economic synthetic approach to pyrido[2,1-b][1,3]benzothiazol-1-ones. The synthesized drug-like pyrido[2,1-b][1,3]benzothiazol-1-ones are of interest to pharmaceutics, since their close analogs show significant antiviral activity.

Acyl(imidoyl)ketenes are versatile synthetic platforms that enable diversity-oriented synthesis (DOS) of various drug-like heterocyclic systems on their basis [9]. These compounds are highly reactive chemical species and can only be used in organic synthesis if generated in situ [9]. Nevertheless, their reactions tend to have high yields, short reaction times, high selectivity, pot, atom and step economy (PASE), and simple purification procedures [9].

Results and Discussion
For the development of the approach to pyrido [2,1-b] [1,3]benzothiazol-1-ones 2, we considered two types of potent sources of acyl(1,3-benzothiazol-2-yl)ketenes 3, APBTTs 1 [10,11] and 3-benzoylpyrrolo [2,1-b] [1,3]benzothiazole-1,2-dione 4 (Scheme 3) [11]. Compound 4 was assumed to undergo a classical pattern for FPDs of thermal decarbonylation [9], cheletropic elimination of CO from C 1 =O position, affording the desired ketene 3a. Compounds 1 were hypothesized to undergo elimination of two molecules of CO from C 1 =O (as a classical pattern of thermal decarbonylation for FPDs [9]) and C 4 =O positions (as a reactivity feature of APBTTs 1, observed by us earlier in nucleophilic reactions [11]), affording the desired ketenes 3. To check our assumptions, we studied the thermal decomposition of compounds 1ag, 4 by simultaneous thermal analysis (STA) ( Table 1, Figures 2 and 3, Supplementary Materials). According to the data obtained, APBTTs 1a-g underwent thermal decomposition with a weight loss accompanied by an exothermic effect ( Figure 2). The values of the weight loss corresponded to the elimination of two CO molecules from APBTTs 1a-g (Table 1), which was in a good accordance with our assumptions about the thermolysis pattern of these compounds. In the case of compound 4, it underwent thermal decomposition with a weight loss accompanied by an exothermic effect consisting of two peaks ( Figure  3). The value of the weight loss was one and a half times more than the calculated one for the elimination of CO molecules from compound 4 (Table 1). Apparently, compound 4 took a different pathway of thermolysis from that proposed by us.
Molecules 2023, 28, x FOR PEER REVIEW 3 of 12 yl)ketenes 3 generated in situ from APBTTs 1 (Scheme 2). The experimental results were supplemented with computational studies to elucidate the reaction mechanism. Scheme 2.
To check our assumptions, we studied the thermal decomposition of compounds 1ag, 4 by simultaneous thermal analysis (STA) (Table 1, Figure 2 and Figure 3, Supplementary Materials). According to the data obtained, APBTTs 1a-g underwent thermal decomposition with a weight loss accompanied by an exothermic effect ( Figure  2). The values of the weight loss corresponded to the elimination of two CO molecules from APBTTs 1a-g (Table 1), which was in a good accordance with our assumptions about the thermolysis pattern of these compounds. In the case of compound 4, it underwent thermal decomposition with a weight loss accompanied by an exothermic effect consisting of two peaks ( Figure 3). The value of the weight loss was one and a half times more than the calculated one for the elimination of CO molecules from compound 4 (Table 1). Apparently, compound 4 took a different pathway of thermolysis from that proposed by us. To check our assumptions, we studied the thermal decomposition of compounds 1a-g, 4 by simultaneous thermal analysis (STA) ( Table 1, Figures 2

and 3, Supplementary Materials).
According to the data obtained, APBTTs 1a-g underwent thermal decomposition with a weight loss accompanied by an exothermic effect ( Figure 2). The values of the weight loss corresponded to the elimination of two CO molecules from APBTTs 1a-g (Table 1), which was in a good accordance with our assumptions about the thermolysis pattern of these compounds. In the case of compound 4, it underwent thermal decomposition with a weight loss accompanied by an exothermic effect consisting of two peaks ( Figure 3). The value of the weight loss was one and a half times more than the calculated one for the elimination of CO molecules from compound 4 (Table 1). Apparently, compound 4 took a different pathway of thermolysis from that proposed by us.
Examination by HPLC-UV (Supplementary Materials) of the products of thermolysis of compounds 1a-g, 4, obtained by measuring their melting points in a capillary, revealed that thermolysis of APBTTs 1a-g resulted in the formation of a single product, and thermolysis of compound 4 resulted in a complex mixture of products, among which only trace amounts of thermolysis product corresponding to APBTT 1a were observed. In addition, compounds 1a-d decomposed without melting (visually dark violet solid turned to a yellow solid), and compounds 1e-g, 4 melted with decomposition.
Thus, the results of STA and HPLC-UV (Supplementary Materials) clearly indicated that APBTTs 1a-g could be suitable candidates for the synthesis of compounds 2a-g, while compound 4 was not.
Then, we successfully scaled up the thermolysis of APBTTs 1a-g under solvent-free conditions to 0.3 mmol (about 100 mg) and isolated products 2a-g in good yields (58-91%) by simple recrystallization of the reaction mixtures. The structures of compounds 2a,f were unequivocally proved by single crystal X-ray analyses (CCDC 2277018 (2a), 2277017 (2f), Supplementary Materials). It should be mentioned that compounds 2a-g had very poor solubility, and there were problems with the acquisition of their NMR spectra; therefore, their 13 C NMR spectra were obtained involving solid-state (ssNMR) and cryoprobe NMR techniques (Supplementary Materials).
Since the structures of compounds 2a,f were proved by single crystal X-ray analyses, they can be considered as reference structures for the establishment of the structures of other compounds 2 by comparison of their spectral characteristics.       Figure 4. Characteristic signals in NMR spectra (CDCl3) (left structure: red is for 1 H NMR signals; blue is for 13 C NMR signals in the downfield; and green is for 13 C NMR signals in the upfield) and atom numbering (right structure) of compound 2a, whose structure was proved by single crystal Xray analysis.  . Characteristic signals in NMR spectra (CDCl 3 ) (left structure: red is for 1 H NMR signals; blue is for 13 C NMR signals in the downfield; and green is for 13 C NMR signals in the upfield) and atom numbering (right structure) of compound 2a, whose structure was proved by single crystal X-ray analysis. We were most interested in the first stage of the process under study, since experimentally we observed that thermolysis of compound 4 did not afford the target compound 2a (results of STA and HPLC-UV studies).
Results of DFT calculations of total electronic energies, enthalpies, and Gibbs free energies of reaction for elementary stages of different pathways for 1a → 2a transformation (  We were most interested in the first stage of the process under study, since experimentally we observed that thermolysis of compound 4 did not afford the target compound 2a (results of STA and HPLC-UV studies).
Results of DFT calculations of total electronic energies, enthalpies, and Gibbs free energies of reaction for elementary stages of different pathways for 1a → 2a transformation (  Note that results of DFT calculations of total electronic energies, enthalpies, and Gibbs free energies of activation for elementary stages of different pathways for 1a → 3a transformation ( Figure 5, Note that results of DFT calculations of total electronic energies, enthalpies, and Gibbs free energies of activation for elementary stages of different pathways for 1a → 3a transformation ( Figure 5, Table 3) also revealed that transformation 1a → 3a occurred via intermediate 4, whereas the formation of alternative intermediate I1 was found to be highly thermodynamically and kinetically unfavorable.  Thus, we observed a discrepancy between theoretical calculations and experimental observations. So, keeping the obtained results in mind, we suppose that when individual compound 4 was heated slowly and gradually from room temperature in our experiments, it underwent another way of decomposition from that we assumed above (Scheme 3). This makes thermolytic transformations of 3-benzoylpyrrolo [2,1b] [1,3]benzothiazole-1,2-dione 4 an intriguing object for further studies.  Table 3. Calculated values of total electronic energies, enthalpies, and Gibbs free energies of activation (∆E = , ∆H = , and ∆G = ) for elementary stages of different pathways for 1a → 3a transformation 1 . Thus, we observed a discrepancy between theoretical calculations and experimental observations. So, keeping the obtained results in mind, we suppose that when individual compound 4 was heated slowly and gradually from room temperature in our experiments, it underwent another way of decomposition from that we assumed above (Scheme 3). This makes thermolytic transformations of 3-benzoylpyrrolo [2,1-b] [1,3]benzothiazole-1,2-dione 4 an intriguing object for further studies.

General Procedure for Compounds 2a-g
The corresponding APBTT 1 (0.3 mmol) was put into an oven-dried test tube and pressed slightly. Then, it was heated at 220 • C on a metal bath for 3 min. The reaction mixture was cooled to room temperature and scrubbed with toluene (5 mL). The resulting precipitate was filtered off and recrystallized from toluene to afford an appropriate pyrido [2,1-b] [1,3]benzothiazol-1-one 2.

Computational Details
The DFT calculations for all model structures were carried out at the M06-2X/6-31G* level of theory with the help of the Gaussian-09 program package [17]. No symmetry restrictions have been applied during the geometry optimization procedure. The Hessian matrices were calculated analytically for all optimized model structures to prove the location of the correct minimum or saddle point (transition state) on the potential energy surface. The Cartesian atomic coordinates for all model structures are presented in attached xyz-files (Supplementary Materials).
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/molecules28145495/s1, copies of NMR spectra for new compounds 2a-g, STA plots for compounds 1a-g, 4, details of DFT calculations, Cartesian atomic coordinates for all model structures, HPLC-UV chromatograms, and ORTEP images of X-ray crystal structures.