Iodine-DMSO mediated conversion of N-arylcyanothioformamides to N-arylcyanoformamides and the unexpected formation of 2-cyanobenzothiazoles

Cyanoformamides are ubiquitous as useful components for assembling key intermediates and bioactive molecules. The development of an efficient and simple approach to this motif is a challenge. Herein, we demonstrate the effectiveness of the I2-DMSO oxidative system in the preparation of N-arylcyanoformamides from N-arylcyanothioformamides. The synthetic method features mild conditions, broad substrate scope, and high reaction efficiency. Furthermore, this method provides an excellent entry to exclusively afford 2-cyanobenzothiazoles which are useful substrates to access new luciferin analogs. The structures of all new products were elucidated by multinuclear NMR spectroscopy and high accuracy mass spectral analysis. Crystal-structure determination by means of single-crystal X-ray diffraction was carried out on (4-bromophenyl)carbamoyl cyanide, 5,6-dimethoxybenzo[d]thiazole-2-carbonitrile, 5-(benzyloxy)benzo[d]oxazole-2-carbonitrile, 4,7-dimethoxybenzo[d]thiazole-2-carbonitrile, and (5-iodo-2,4-dimethoxyphenyl)carbamoyl cyanide, a key intermediate with mechanistic implications.


Introduction
Cyanoformamides are valuable and versatile building blocks used for constructing synthetically useful intermediates and many bioactive compounds. Cyanoformamides bearing an alkynyl tether undergo intramolecular cyanoamidation to produce veto seven-membered ring a-alkylidene lactams whereas those possessing a 1,1-disubstituted alkenyl groups afford 3,3-disubstituted oxindoles having a quaternary carbon center. 1 An enantioselective 2 and a diastereoselective 3 asymmetric version of the latter reaction has also been developed. The related di(cyanoformamide) precursors are also synthetically useful and their key role in the cascade cyanoamidation route to synthesize the madangamine core is noteworthy (Scheme 1). 4 Cyanoformamides also add across alkynes by nickel/BPh 3 -catalyzed cyanocarbamoylation to give b-cyanosubstituted acrylamides (Scheme 1). 5 The synthesis of carbamoyl amidoximes from cyanoformamides 6 and formation of bketo Weinreb amides and unsymmetrical ketones has also been reported (Scheme 1). 7 Likewise, cyanoformamides have been utilized in the preparation of 1,8-dihydroindeno[2,1-b]pyrrole-2carboxamide and the carboxylate derivatives. 8 Upon treatment with aluminum azide, cyanoformamides convert to the corresponding bioactive antiallergic tetrazole-5-carboxamides. 9 Transformation of the cyanoformamide function into the tetrazol-5-carboxamide has also been achieved with Me 3 SiN 3 -Bu 2 SnO and was used to prepare 5-aryl-1,3,4-oxadiazoles used for glycogen phosphorylase b (RMGPb) inhibition. 10 Interestingly, cyanoformamide is the nitrile derivative of formamide, a species responsible for the synthesis of nucleic acid precursors under prebiotic conditions in interstellar space. 11 Furthermore, such a ubiquitous motif is also present in several natural products like ceratinamine, 12 and its 7-hydroxyceratinamine derivative, 13 subereamide A, 14 and 12-hydroxysubereamide C (Scheme 1). 14 Substantial efforts have been directed toward the development of synthetic methodologies to prepare cyanoformamides. One early strategy described reacting primary and secondary amines with carbonyl cyanide. However, this method was deemed unsuitable for large scale preparation due to the production of toxic hydrogen cyanide. 15 As an alternative, reacting the amines with triphosgene followed by substitution reaction of the resulting chlorocarbamates with cyanide ion provided acceptable yields. 15 Several other earlier reports described the formation of cyanoformamides. [16][17][18][19][20][21] For instance, hydration of cyanogen under high pressure using excess water gave 1-cyanoformamide, 16 whereas reaction of 5-hydroxyimino-1,3-dioxine-4,6-dione (isonitroso Meldrum's acid) with carbodiimides (N,N 0 -dicyclohexylcarbodiimide and N,N 0 -diisopropylcarbodiimide) gave N-cyclohexylcyanoformamide and Nisopropylcyanoformamide, respectively. 17 Some other reagents like tetracyanoethylene, 18 5-tosyloxyimino-2,2-dimethyl-1,3dioxane-4,6-dione, 19 4-chloro-5H-1,2,3-dithiazol-5-one, 20 tetraalkyl-cyanoformamidinium salt, 21 or dichlorosulfenyl chlorides 22 have been employed in the synthesis of similar types of synthetic compounds. Unfortunately, the structural complexity and toxicity associated with these reagents hampered their use. Recently, other more direct synthetic methods have been developed (Scheme 2). [23][24][25][26][27][28][29] For instance, Muñoz reported that the reaction of primary amines with tetramethylphenylguanidine and cyanophosphonates at À10 C under an atmosphere of CO 2 furnishes cyanoformamides in good yields. 23 Dong and co-workers 24 employed phosphoryltrichloride (POCl 3 ) to convert 1-acyl-1-carbamoyl oximes to cyanoformamides, while concurrently Wu and co-workers 25 reported an eco-friendly method for the conversion of 2-oxoaldehydes into cyanoformamides using iodosobenzene diacetate (IBD) as oxidant. Zhang and co-workers described the transformation of triuoropropanamide precursors into cyanoformamides via a sequence of C-CF 3 bond breaking process and subsequent nitrogenation using tert-butyl nitrite as the source of nitrogen. 26 At the same time, Schwartz's group described solvent-free access to secondary and tertiary cyanoformamides from TMSCN and carbamoyl imidazoles. 27 Recently, cyanoformamides were prepared from N,N-disubstituted aminomalononitriles 28 with CsF as the promoter and in another study, 4,5-dioxo-imidazolinium cation activation of 1-acyl-1carbamoyl oximes was used. 29 Electrochemical synthesis of cyanoformamides was also reported starting from trichloroacetonitrile and secondary amines mediated by heptamethyl cobyrinate, a B 12 derivative. 30 Therefore, efficient and convenient methods for the synthesis of cyanoformamides are still highly desirable.

Results and discussion
At the outset of our work, we were interested in preparing 5imino-1,3-diphenyl-2-thioxoimidazolidin-4-ones and 5-imino-1,3-diphenyl-2-selenoxoimidazolidin-4-ones as an extension to our previous work. 31,32 We attempted a previously reported procedure where Papadopoulos prepared N-phenylcyanoformamide by reacting phenyl isocyanate with potassium cyanide in water (Scheme 3). 33 The arylcyanoformamide product was only characterized by melting point. The author noted the slow precipitation of N,N-diphenylurea upon standing of the alkaline reaction mixture due to the dissociation of the anion of N-phenylcyanoformamide to form phenyl isocyanate. However, in our hands, and aer multiple attempts to duplicate the above method, phenyl isocyanate reacted competitively with water to produce phenylcarbamic acid (see ESI section ‡). Attempts to run the same reaction using ethanol-water mixture (87 : 13) produced ethyl phenylcarbamate as the major product (Scheme 3). Clearly, the reactivity of the isocyanate group renders the preceding strategy impractical. On the contrary, isothiocyanates are less reactive and comprise more convenient precursors to prepare the cyanoformamide. Hence, we envisaged that cyanoformamides could be obtained directly from cyanothioformamides by converting the thione to the carbonyl function. Herein, we present an efficient method for the synthesis of N-arylcyanoformamides from N-arylcyanothioformamides using simple iodine-DMSO oxidative system and report unexpected formation of cyanobenzothiazoles. Following the design in Scheme 3, several N-arylcyanothioformamides were prepared on large scale (20 mmol) from commercially available isothiocyanates and potassium cyanide in water-ethanol in good yields (see ESI S2-S253 ‡). Initially, N-4tolylcyanothioformamide (1a) was selected as the model substrate (Table 1) to examine its reaction with I 2 . Ketcham and Schaumann reported that the oxidation of 1a at 80 C using 16 mol% I 2 produced cyanoformamide 2a in 86% yield, although they abandoned the method and opted to employ other more convenient procedures to prepare 2a. 34 Indeed, the initial test to reproduce the formation of 2a at 80 C using 16 mol% of the I 2 /DMSO oxidant system only resulted in partial conversion (30%) of 1a to the expected product 2a as indicated by 1 HNMR (Table 1, entry 1). The product 2a could not be separated and puried by column chromatography from 1a as both exhibit the same R f value. This was not surprising and presented a purication challenge for all N-arylcyanoformamide substrates as, they too, would likely have similar R f values to their corresponding starting materials. Therefore, for this methodology to be useful, complete, and clean conversion of all starting material 1 to product 2 is required. Thus, further variation in the amount of iodine conrmed that 1.1 equivalent is optimal at 80 C to completely transform 1a to 2a (Table 1, entry 4). Next, the reaction was carried out on different substrates using the optimal conditions (1.1 equiv. I 2 , 80 C, 6 h) (Table 1, entries 5-10). Unfortunately, many substrates (1b-e) furnished the desired products 2b-e in low yields (entries 5-8), while others like 3-(uorophenyl)carbamothioyl cyanide (1f) and 4-(nitrophenyl)carbamothioyl cyanide (1g) afforded complex mixtures (entries 9 and 10). N-Arylcyanothioformamides and the N-arylcyanoformamides products are temperature sensitive and may extrude HCN and undergo a reversible reaction to form the isothiocyanates and isocyanates, respectively, at elevated temperatures. Thus, from yield and safety perspectives, ambient conditions are better suited for both, the substrate, and product.
Thus, with the above in mind, the best reaction conditions resulting in complete conversion of 1a at 80 C (1.1 equiv. I 2 ) were applied to N-4-tolylcyanothioformamide (1a) at ambient temperature (20 C), resulting in a disappointing 10% conversion to 2a (  [13][14][15][16] established that 2.75 equiv. I 2 was required for complete conversion of 1a to 2a. With the enhanced reaction conditions in hand (2.75 equiv. I 2 , 20 C, 19 h) (Table 2, entry 16), these were rst applied to (4-methoxyphenyl)carbamothioyl cyanide (1i), resulting in 25% conversion to the target product 2i. Further increase in the amount of I 2 to 3 and 3.5 equivalents resulted in 36% and 44% conversion to 2i, respectively. However, testing the latest conditions (3.5 equiv. I 2 , 20 C, 19 h) on the related (4-ethoxyphenyl)carbamothioyl cyanide (1j) resulted in only 27% conversion to 2j. Potassium iodide (KI) was Table 1 Reaction condition studies at various temperatures using variations in the amount of I 2 and several different N-arylcyanothioformamides  also explored as an alternative source of iodine in the optimization of conditions. Thus, treatment of 1j using the optimized conditions using KI (3.5 equiv. KI, 38 C, 19 h) failed to give product 2j and the starting material was recovered unchanged. Potassium iodide (1 equimolar) was also used as a co-reagent with iodine (3.5 equiv. KI, RT, 19 h) in the preceding reaction but no enhancement in conversion was detected, suggesting that KI was not a suitable replacement for iodine. At this point, it became clear that investigating the impact of slight elevation in temperature was warranted to establish the optimal value suitable for a wide range of substrates. Thus, when 1j was treated with 3.5 equiv. I 2 and heated at various low temperatures (29.5-38 C) for 19 h (Table 1, entries [23][24][25][26], complete and clean conversion to 2j was observed at 37-38 C (>95% based on the minimum detection limit of 1 HNMR).
All new cyanothioformamides and cyanoformamides were characterized by standard spectroscopic and analytical techniques (mp, IR, 1D and 2D NMR, and HRMS). The physical and spectral data of known compounds matched those reported (see Experimental and ESI sections ‡). The most distinctive signal to distinguish the formamide product from the cyanothioformamide starting material is that of the carbonyl (C]O) group which appears around 1700 cm À1 in the IR region and resonates around 140-144 ppm in the 13 C NMR compared to approximately 160-165 ppm for the thiocarbonyl (C]S) group. Structural verication of (4-bromophenyl)carbamoyl cyanide (2g 0 ) by single crystal X-ray crystallography, as a representative example of the cyanoformamide products, is shown in Fig. 1. Clearly, the nitrile function remained intact (exhibiting a typical linear bond angle ¼ 176. 3(3) for N(11)-C(11)-C(12) and C(11)-N(11) bond length ¼ 1.141(4)Å) while the thiocarbonyl has clearly been converted to the carbonyl where the bond length of O(1)-C(2) ¼ 1.227(3)Å. Unlike their thiocarbonyl counterparts, the cyanoformamide products generally appear as one tautomer possibly due to strong intermolecular hydrogen bonding between O 11 and N 2 -H (Fig. 1). The two molecules in the unit cell are arranged tail to tail to accommodate hydrogen bonding as shown in Fig. 1. The bond length between O 11 and N 2 -H is 2.135Å, indicating strong interaction.
An unexpected number of 2-cyanobenzothiazoles were formed exclusively and in good to very high yield (64-98%) when the N-arylcyanothioformamides 1l 0 -q 0 (see ESI section ‡) Fig. 1 Thermal ellipsoid plots of (4-bromophenyl)carbamoyl cyanide (2g 0 )with ellipsoids drawn at 50% probability level. Selected bond distances (Å) and angles (deg) for compound 2g 0 : Br (11) were treated in the usual way with iodine in DMSO using the optimized conditions (3.5 equiv. I 2 , 38 C, 19 h) (Table 3). The resulting light yellow/orange/brown products could be isolated cleanly without the need for ash chromatography and are very stable at room temperature. High-resolution mass spectrometry (HRMS) and NMR measurements corroborated the suggested structures 3a-f (Table 3). Very few synthetic methods are available for the synthesis of 2-cyanobenzothiazoles which are themselves scarce in the literature. Thus, the iodine-DMSO system comprises a novel approach to access 3a-f. While 3c and 3d are unreported, the remaining analogues in the 3a-f series appear in the literature with partial (only 1 H NMR) or even no reported NMR or physical properties data. Thus, 3a-f were extensively characterized (vide infra and see ESI ‡).
The benzothiazole nucleus has a wide prole of biological activities. 35 In particular, the 2-cyanobenzothiazole derivatives have been recently used in self-uorescent hyaluronic acidbased gel for dermal applications 36 and as linkers in the development of single-molecule strategy to characterize the folded state of individual proteins during membrane translocation. 37 Recently, 2-cyanobenzothiazole was incorporated into gold nanoparticles to enhance imaging and treatment of breast cancer 38 and was also used in site-specic immobilization of biomolecules by reaction with terminal cysteine. 39 Perhaps the most intriguing application of 2-cyanobenzothiazoles entails their use as precursors to access new luciferin analogs for bioluminescence imaging applications. [40][41][42] On the other hand, the iodine/DMSO oxidation system has truly revolutionized synthetic practices in a plethora of reactions involving oxidation processes. [43][44][45][46][47][48][49][50][51][52][53][54][55][56] This oxidant has been particularly used in C-N bond chemistry as a greener solution to existing conventional synthetic methodologies and to avoid employing harsh, toxic, and expensive metals and reagents. The wide and abundant availability of iodine and DMSO, ease of preparation, moisture and air stability, atom and step economy, as well as its environmentally benign nature render such system very convenient. Mechanistically, the I 2 /DMSO oxidant system has been largely described to involve prior iodination of substrates. Iodine in catalytic amount is oen regenerated in the reaction from the oxidation of HI with DMSO with concurrent production of dimethyl sulde (DMS) and is mainly reachable at higher temperatures. Notable biologically potent molecules that have been constructed through key C-N bond formation using I 2 /DMSO include a-ketoamides 46 and a-ketoimides, 47 imidazoles, 48 quinoxalines, 49 pyrazines, 49 quinazolinones, 50 isatin, 51 amides, 52 thioamides, 52 thiazoles, 53 triazoles, 54 oxindoles, 55 oxadiazoles 56 and oxazoles. 56 Thus, to fully conrm the chemical structures of heterocycles 3a-f (Table 2) and prove the formation of the new C-S quaternary center, extensive one-dimensional (1D) ( 1 H-, 13 C-, 13 C-CRAPT NMR) and two-dimensional (2D) homonuclear ( 1 H-1 H-gDQCOSY) and heteronuclear ( 1 H-13 C-gHSQC, 1 H-13 C-gHMBC) correlation NMR spectrometry experiments were initially performed on all compounds (see ESI section ‡). Hence, using 5,6dimethoxybenzo[d]thiazole-2-carbonitrile (3e) as a representative model for the remaining structurally related cyanobenzothiazoles, the relevant NMR spectra that were used for structural proof and chemical shi assignment are shown in Fig. 2.
Analysis of the 13 C-CRAPT NMR spectrum (Fig. 2, spectrum b) of 3e conrmed the presence of the expected 10 signals (2 Table 3 Unexpected formation of 2-cyanobenzothiazoles 3a-f from various N-arylcyanothioformamides and (3-iodo-4,6-dimethoxyphenyl) carbamoyl cyanide 3g aromatic CH's, 5 aromatic quaternary carbons, 1 cyano carbon, and 2 methoxy groups) which is consistent with all carbons being magnetically nonequivalent. The most striking feature of the 13 C-CRAPT NMR of 3e, compared to the precursor (3,4dimethoxyphenyl)carbamothioyl cyanide (1p 0 ) (see ESI ‡), is the presence of only 2 aromatic CH's (d 104.9 & 102.9 ppm) as indicated by their negative phase and an additional quaternary carbon in the former (3e), suggesting that one proton has been removed from 1p 0 and replaced with a quaternary center in product 3e. Evidence supporting the suggested regiochemistry of 3e cyclization at C 6 (IUPAC numbering) rather than C 2 is based on the presence of two singlets (d 7.82 & 7.69 ppm) for the two aromatic CH's in the 1 HNMR spectrum of 3e (Fig. 2, spectrum a). Further, the two CH's of 3e do not show any correlation in the 1 H-1 H-gDQCOSY NMR (see ESI ‡), clearly indicating that they are isolated spin systems and are not coupled. These protons are attached to carbon atoms and could not be stemming from a NH group as indicated by the strong correlation contours with the carbons at d 102.9 & 104.9 ppm in the 1 H-13 C-gHSQC NMR spectrum (Fig. 2, spectrum c). In fact, the 1 HNMR spectrum of 3e is lacking the typical NH signal observed in products 2a-k 0 . Conclusive evidence supporting cyclization and the creation of the new ArC-S bond in 3e stems from the 1 H- 13 C-gHMBC NMR spectrum (Fig. 2, spectrum d). The nitrile and C]N chemical shis were easily identied at d 113.9 (CN) and 133.0 (C]N) ppm, respectively, since they did not show any long-range 1 H- 13 C heteronuclear multiple bond correlations with the CH protons. On the contrary, the two ArC-OMe quaternary carbons were identied as the signals at d 151.3 (C-O) and 150.7 (C-O) ppm due to strong 1 H-13 C long-range correlation cross peaks with the two methoxy groups at d 3.88/ 3.87 ppm. The two remaining quaternary centers of the fused heterocycle, the C-S (d 128.8) and C-N (d 146.5), were instrumental proof of heterocyclization. Clearly, both protons at d 7.82 & 7.69 ppm are totally correlated with the two adjacent carbon atoms of the fused ring (d 128.8 & 146.5), as well as with the other two adjacent ArC-OMe quaternary carbon atoms (d 151.3 & 150.7) in the 1 H-13 C-gHMBC NMR spectrum (6 contour correlation squares in the aromatic region) (Fig. 2, spectrum d). Pleasingly, we were able to grow crystals suitable for X-ray diffraction analysis. Thus, structural verication of 3e was also carried out by single crystal X-ray crystallography (Fig. 3). Clearly, 3e comprises a 5-membered heterocyclic ring containing sulfur and C]N (characterized by short bond length of N(7)-C(6) ¼ 1.303(3)Å and typical trigonal planar geometry where N(7)-C(8)-C(4) ¼ 115.1(2) ), indicating that heterocyclization of the cyanothioformamide precursor 1p 0 is faster than desulfurization. The nitrile group is also intact, displaying the typical linear bond angle ¼ 178. 7(3) for N(15)-C(14)-C (6).
The structures of cyanobenzothiazoles 3c and 3f were also proven by single crystal X-ray crystallography, highlighting the role of the alkoxy group in cyclization. The single crystal X-ray structure of 5-(benzyloxy)benzo[d]oxazole-2-carbonitrile (3c) is shown in Fig. 4. The benzyloxy group clearly directs cyclization to the less hindered para position (C 7 atom; Fig. 4 X-ray numbering).  13 C-CRAPT NMR spectrum; (c) 1 H- (d) 1H-  The single crystal X-ray structure of 4,7-dimethoxybenzo[d] thiazole-2-carbonitrile (3f) is shown in Fig. 5. The C 1 methoxy group directs cyclization to the more hindered C 9 position since the para position (C 4 atom in Fig. 5) is substituted.
Interestingly, the cyclization reaction in all cases was completely regioselective, exclusively producing cyclized products 3b-e in which the CH para to the alkoxy or thiomethyl groups was the site of oxidative cyclization (Scheme 4). However, in case of product 3f, the more hindered ortho-CH was involved in the cyclization reaction due to the absence of a para-CH. Clearly, the alkoxy and thiomethyl groups direct the cyclization reaction and offer mechanistic implications. Though, it seems the presence of a substituent is not mandatory for annulation as suggested by the cyclization of naphthalen-1ylcarbamothioyl cyanide (1l 0 ) to the naphthyl derivative 3a.
Using (3,4-dimethoxyphenyl)carbamothioyl cyanide (1p 0 ) as a representative example, the proposed mechanism of heterocyclization is shown in Scheme 4. Mechanistically, it is conceivable that the cyanothioformanilide precursor 1 undergoes fast iodination and subsequent rearomatization to generate intermediate I. Indirect evidence for this mechanism is based on the isolation of (3-iodo-4,6-dimethoxyphenyl) carbamoyl cyanide (3g) ( Table 3) from the reaction of its precursor (2,4-dimethoxyphenyl)carbamothioyl cyanide (1r 0 ) (see ESI section ‡) with I 2 -DMSO. Next, intramolecular nucleophilic attack by the thioformamide sulfur atom, followed by elimination of a HI molecule produces the cyanobenzothiazole 3e. To shed further light on the mechanism, free radical trapping control experiments were performed. Thus, the reaction of 1p 0 with I 2 -DMSO was conducted in the presence of equimolar amounts of TEMPO (2,2,6,6-tetramethyl-1-piperidinyloxy) and BHT (2,6-ditertbutyl-4-methylphenol) as the radical inhibitors. We observed that product 3e was obtained in 63% and 62% yields, respectively, which suggested that a free radical pathway leading to free radical intermediates was not involved in the transformation process and formation of 3e.
The structure of (5-iodo-2,4-dimethoxyphenyl)carbamoyl cyanide (3g) could not be fully established based on 1D and 2D NMR, especially the position of iodine on the aromatic ring (position 2 (X-ray numbering C 6 ) vs. 3 (X-ray numbering C 1 )). Thus, structural verication of 3g by single crystal X-ray crystallography was carried out as shown in Fig. 6. As expected, the   iodination has been directed to the o/p position by the two methoxy groups, rendering 3g unsuitable species for cyclization. On the contrary, products 3b-f were all possible since iodination presumably occurs next to the cyanothioformamide group as directed by the alkoxy or thiomethyl groups. Finally, the preparation of cyanobenzothiazoles was amenable to scaleup to gram quantities as shown by the synthesis of 3a, 3b, 3d on a large scale from their precursors (10 mmol scale) 90%, 95%, and 87% isolated yields, respectively.
In conclusion, the I 2 -DMSO mediated desulfurization of 1 for the synthesis of cyanoformamides 2 at 38 C has been successfully demonstrated. The reaction tolerated a range of functional groups including various halides, alkoxides, esters, cyano, nitro, thiomethyl, and triuoromethyl functions and afforded a broad scope of products. It is expected that the current synthetic technique could become candidate for the synthesis of cyanoformamides because it is practical, scalable, uses a simple reagent system, and offers mild reaction conditions. The I 2 -DMSO oxidative system has also proven useful to access 2-cyanobenzothiazoles which may serve as useful precursors to access new luciferin analogs.

General information
Reactions were conducted with magnetic stirring in air-dried glassware. All reagents and reaction solvents were used as received without any further purication. Analytical thin-layer chromatography (TLC) was used to follow the progress of reactions and was carried out on precoated silica gel plates (HSGF 254) and visualized under UV irradiation (254 nm). Flash column chromatography was performed using silica gel (200À300 mesh) Scheme 4 Proposed mechanism for the formation of 2-cyanobenzothiazoles from N-arylcyanothioformamide. in cases where pure analytical samples were required. 1 H and  or CDCl 3 on a Bruker DPX 300 and 75 MHz NMR spectrometer and on a Varian 400 and 100 MHz NMR spectrometer. The NMR chemical shis (d) are reported in parts per million (ppm) relative to the residual solvent peak ( 1 H-NMR d 7.26 for CDCl 3 , d 2.50 for DMSO-d 6 ; . The following abbreviations were used to explain NMR peak multiplicities: br s ¼ broad signal, s ¼ singlet, d ¼ doublet, t ¼ triplet, q ¼ quartet, p ¼ pentet, sept ¼ septet, app ¼ apparent, and m ¼ multiplet. IR spectra were recorded using a Bruker FT-IR spectrometer and a Thermo Nicolet Nexus 470 FT-IR. High-resolution mass analyses (HRMS) were obtained using a Waters Q-TOF Premier mass spectrometer [electrospray ionization (ESI)]. Melting points were measured using a capillary melting point apparatus (MEL-TEMP) in degrees Celsius ( C).

Procedures
The N-Arylcyanothioformamide 1 (0.5 mmol) was heated at 38 C for 19 h in 2 mL of DMSO with 444 mg (1.75 mmol, 3.5 equiv.) of iodine. The reaction mixture was treated with 4 mL of sodium thiosulfate (1 M) and was extraction with ether (10 mL). The colorless or faint yellow ether extract was washed with a brine solution (2 Â 10 mL), dried (Na 2 SO 4 ), and concentrated in vacuo to afford the corresponding N-arylcyanoformamide 2 or cyanbenzoxazole 3 product in the specied chemical yield.