The Conversion of 5,5′-Bi(1,2,3-dithiazolylidenes) into Isothiazolo[5,4-d]isothiazoles

Thermolysis of 4,4′-dichloro-, 4,4′-diaryl-, and 4,4′-di(thien-2-yl)-5,5′-bi(1,2,3-dithiazol-ylidenes) affords the respective 3,6-dichloro-, 3,6-diaryl- and 3,6-di(thien-2-yl)isothiazolo[5,4-d]-isothiazoles in low to high yields. The transformation of the 4,4′-diaryl- and 4,4′-di(thien-2-yl)-5,5′-bi(1,2,3-dithiazolylidenes) occurs at lower temperatures in the presence of the thiophiles triphenylphosphine or tetraethylammonium iodide. Optimized reaction conditions and a mechanistic rationale for the thiophile-mediated ring transformation are presented.

In light of the above, and with the aim to develop a milder route to 3,6-di(het)aryl-isothiazolo [5,4-d]isothiazoles 11 we developed and optimized a thiophile-mediated ring transformation for (E)-4,4′-diphenyl-5,5′-bi(1,2,3-dithiazolylidene) (11b), which has no nucleofuges at either C4/4′ and gives a product with no nucleofuges at either C3/6 and therefore was more resistant to thiophile-mediated ring opening reactions ( Table 2, entries 1-10). Interestingly, single crystal X-ray studies were also obtained to support both the (E)-geometry of the diphenylbi(dithiazolylidene) 11b and the structure of the final diphenyl-substituted isothiazoloisothiazole 8b (Figure 3). While both isothiazoloisothiazoles 8a and 8b have planar isothiazoloisothiazole core structures, the phenyl groups in the latter deviate in a conrotatory manner from the isothiazoloisothiazole plane by 8.7°.  Heating a toluene solution of the diphenylbi(dithiazolylidene) 11b at ca. 110 °C for 29 h led to its conversion into diphenylisothiazoloisothiazole 8b in 76% yield together with traces of S8 and 4-phenyl-5H-1,2,3-dithiazole-5-thione (14b) [20] (by TLC) ( Table 2, entry 1). In the presence of thiophiles Et4NI or Ph3P the reaction proceeded significantly faster and in higher yield ( Table 2, entries 2-10). When Ph3P (2-4 equiv) was used as thiophile, diphenylisothiazoloisothiazole 8b was formed in quantitative yield but was accompanied, as expected, by the formation of triphenylphosphine sulfide (Ph3P = S) in 75-78% yield ( Table 2, entries 2 and 3), which required chromatographic separation. Fortunately, the use of Et4NI as thiophile worked equally well, and on a 0.5 mmol scale, in 5 mL of PhMe, the use of Et4NI (0.2 equiv) gave the fastest reaction (2 h) and a quantitative yield of product (Table 2, entry 5) which could be isolated without the need for chromatography. More or less equivalents of Et4NI led to longer reaction times ( Table 2, entries 4 and 6). With these conditions in hand, we then investigated the effect of concentration and temperature. Fortunately, small variations in concentration did not affect the reaction times or yields (data not shown), and carrying out the reaction using 0.1 mmol of ylidene 11b in only 5 mL of PhMe continued to give a near quantitative yield of isothiazoloisothiazole 8b (95%) together with some dithiazolethione 14b (5%) (Table 2, entry 9), but at 0.2 mmol the yield of the desired product 8b dropped significantly (63%) and the amount of undesired dithiazolethione 14b increased (26%) ( Table 2, entry 10). Lowering the reaction temperature with the use of PhH (bp 80 °C) instead of PhMe (bp 110 °C) led to no reaction after 10 h of heating (Table 2, entry 7) while the use of PhCl (bp 132 °C) led to a longer reaction time (8 h), lower yield (72%) and the formation of more dithiazolethione 14b (15%) compared to the PhMe ( Table 2, entry 8). With the thiophile-mediated reaction of the diphenylbi(dithiazolylidene) 11b partially optimized we then carried out a minor investigation into the reactions scope ( Table 2, entries [11][12][13].
Under the optimized reaction conditions, ylidenes bearing aryls containing electron withdrawing para-fluoro and electron releasing para-methoxy substitution worked to give the desired isothiazoloisothiazoles in moderate to excellent yields 69 and 99%, respectively ( Table 2, entries 11 and 12). Furthermore, the important thien-2-yl group, an electron rich hetaryl that is important in organic electronic materials, was tolerated to afford 3,6-di(thien-2-yl)-isothiazolo [5,4-d] isothiazole (8e) in 92% yield ( Table 2, entry 13), potentially opening up this biheterole system for study as a new π spacer for small organic molecules, oligomers or polymers in material sciences. Heating a toluene solution of the diphenylbi(dithiazolylidene) 11b at ca. 110 • C for 29 h led to its conversion into diphenylisothiazoloisothiazole 8b in 76% yield together with traces of S 8 and 4-phenyl-5H-1,2,3-dithiazole-5-thione (14b) [20] (by TLC) ( Table 2, entry 1). In the presence of thiophiles Et 4 NI or Ph 3 P the reaction proceeded significantly faster and in higher yield ( Table 2, entries 2-10). When Ph 3 P (2-4 equiv) was used as thiophile, diphenylisothiazoloisothiazole 8b was formed in quantitative yield but was accompanied, as expected, by the formation of triphenylphosphine sulfide (Ph 3 P = S) in 75-78% yield ( Table 2, entries 2 and 3), which required chromatographic separation. Fortunately, the use of Et 4 NI as thiophile worked equally well, and on a 0.5 mmol scale, in 5 mL of PhMe, the use of Et 4 NI (0.2 equiv) gave the fastest reaction (2 h) and a quantitative yield of product (Table 2, entry 5) which could be isolated without the need for chromatography. More or less equivalents of Et 4 NI led to longer reaction times ( Table 2, entries 4 and 6). With these conditions in hand, we then investigated the effect of concentration and temperature. Fortunately, small variations in concentration did not affect the reaction times or yields (data not shown), and carrying out the reaction using 0.1 mmol of ylidene 11b in only 5 mL of PhMe continued to give a near quantitative yield of isothiazoloisothiazole 8b (95%) together with some dithiazolethione 14b (5%) (Table 2, entry 9), but at 0.2 mmol the yield of the desired product 8b dropped significantly (63%) and the amount of undesired dithiazolethione 14b increased (26%) ( Table 2, entry 10). Lowering the reaction temperature with the use of PhH (bp 80 • C) instead of PhMe (bp 110 • C) led to no reaction after 10 h of heating (Table 2, entry 7) while the use of PhCl (bp 132 • C) led to a longer reaction time (8 h), lower yield (72%) and the formation of more dithiazolethione 14b (15%) compared to the PhMe ( Table 2,

Mechanistic Rationale
Tentatively, we propose the thiophile-mediated reaction ( Table 2) proceeded via an ANRORC style reaction pathway [32][33][34], where the thiophile attacks either the dithiazole S1 or S2 sulfurs to generate a ring opened species that then collapses to give the isothiazole ring system (Scheme 5). At this stage, it is not possible to give an accurate mechanism, as several possibilities exist. Previous studies, nevertheless, reveal the dithiazole S2 atom to be marginally more susceptible to thiophilic attack [49], and based on this we propose the ring opening of both dithiazoles to generate the dianion 17 that then collapses to the 14π 4,8-disubstituted [1][2][3]
Molecules 2018, 23, x FOR PEER REVIEW 7 of 13 (Scheme 5). Attempts to treat 3,6-diphenylisothiazolo [5,4-d]isothiazole 8b with elemental sulfur or active sulfur (DABCO/S8) [50] led to no reaction supporting that the reaction was not in equilibrium and the product was not convertible back to the starting dithiazole 11b.

General Methods and Materials
Powdered anhydrous Na2SO4 was used for drying organic extracts and all volatiles were removed under reduced pressure. Toluene was distilled over CaH2 before use. All reaction mixtures and column eluents were monitored by TLC using commercial glass backed thin layer chromatography (TLC) plates (Merck Kieselgel 60 F254). The plates were observed under UV light at 254 and 365 nm. Elemental analyses were performed on a 2400 Elemental Analyzer (Perkin Elmer Inc., Waltham, MA, USA). Melting points were determined on a Kofler hot-stage apparatus and are uncorrected. Solvents used for recrystallization are indicated after the melting point. 1  High-resolution MS spectra were measured on a Bruker MICROTOF II instrument using electrospray ionization (ESI). The measurement was operated in a positive ion mode (interface capillary voltage-4500 V) or in a negative ion mode (3200 V); mass range was from m/z 50 to 3000 Da; external or internal calibration was done with Electrospray Calibrant Solution (Fluka Chemicals Ltd., Gillingham, UK). A syringe injection was used for solutions in acetonitrile, methanol, or water Scheme 5. Tentative mechanism for the Et 4 NI-mediated ring transformation of dithiazolylidenes 11 into isothiazolo [5,4-d]isothiazoles 8.

General Methods and Materials
Powdered anhydrous Na 2 SO 4 was used for drying organic extracts and all volatiles were removed under reduced pressure. Toluene was distilled over CaH 2 before use. All reaction mixtures and column eluents were monitored by TLC using commercial glass backed thin layer chromatography (TLC) plates (Merck Kieselgel 60 F 254 ). The plates were observed under UV light at 254 and 365 nm. Elemental analyses were performed on a 2400 Elemental Analyzer (Perkin Elmer Inc., Waltham, MA, USA). Melting points were determined on a Kofler hot-stage apparatus and are uncorrected. Solvents used for recrystallization are indicated after the melting point. 1 H and 13 C-NMR spectra were taken with a Bruker AM-300 (at 300.1 and 75.5 MHz) or Bruker DRX500 (at 500.1 and 125.8 MHz) or Bruker AV600 instrument (at 600.1 and 150.9 MHz) (Bruker Ltd., Moscow, Russia) with TMS as the standard. J values are given in Hz. MS spectra (EI, 70 eV) were obtained with a MAT INCOS 50 instrument (Thermo Finnigan LLC, San Jose, CA, USA). High-resolution MS spectra were measured on a Bruker MICROTOF II instrument using electrospray ionization (ESI). The measurement was operated in a positive ion mode (interface capillary voltage-4500 V) or in a negative ion mode (3200 V); mass range was from m/z 50 to 3000 Da; external or internal calibration was done with Electrospray Calibrant Solution (Fluka Chemicals Ltd., Gillingham, UK). A syringe injection was used for solutions in acetonitrile, methanol, or water (flow rate 3 µL/min). Nitrogen was applied as a dry gas; interface temperature was set at 180 • C. IR spectra were measured with a M-80 instrument (Carl Zeiss Jena GmbH, Jena, Germany) in KBr pellets. 4,4 -Disubstituted 5,5 -bi(1,2,3-dithiazolylidenes) 11a-e [23] and (E)-1-(4-bromophenyl)-ethan-1-one oxime [52] were prepared according to the literature.
Data collection for a single crystals 8a, 8b, 8d and 11b ( Figures S1-S4, Supporting Information, SI) was performed at the Center for Molecular Composition Studies of INEOS RAS on a Bruker Smart Apex II CCD diffractometer (Mo Kα radiation, λ = 0.71073 Å, graphite monochromator). Frames were integrated using the Bruker SAINT software package [53] using a narrow-frame algorithm, and a semiempirical absorption correction was applied with the SADABS program [54] using intensity data of the equivalent reflections. All the structures were solved by direct method and refined by the least-squares in anisotropic full-matrix approximation on F 2 hkl . The hydrogen atoms were calculated geometrically and refined in isotropic approximation using the riding model with the SHELX software package [55]. The refinement of the molecules with minor occupancy was performed with the restraints on anisotropic displacement parameters (EADP) and bond lengths and angles (SAME). Detailed crystallographic information is provided in Table 3 and as Supporting Information in CIF format that can be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html or from the Cambridge Crystallographic Data Centre, 12, Union Road, Cambridge CB2 1EZ, UK; Fax: 44-1223-336033 using the reference CCDC numbers (Table 3). To a stirred solution of (E)-1-(4-bromophenyl)ethan-1-one oxime [24] (428 mg, 2 mmol) and sulfur monochloride (0.64 mL, 4 mmol) in acetonitrile (15 mL) at ca. −5 • C under an argon atmosphere was added dropwise pyridine (0.96 mL, 6 mmol). The mixture was stirred at ca. −5 • C for 15 min, then copper powder (192 mg, 3 mmol) was added, the mixture was stirred at room temperature for 1.5 h and then poured into ice water (100 mL). The precipitate was filtered, washed with water, dried and extracted with CH 2 Cl 2 (3 × 15 mL). The combined extracts were dried (CaCl 2 ) and solvents were evaporated under reduced pressure. The residue was rapidly separated by flash chromatography (silica gel Merck 60, n-hexane and then n-hexane/CH 2 Cl 2 mixtures) to afford the title compound 11f (346 mg, 67%) as a black powder, m.p. 181-182 • C (n-hexane); δ H (300 MHz; CD 2 Cl 2 ) 7. 37 16 H 8 Br 2 N 2 S 4 , m/z 515.7910).