Copper(ii)-catalyzed tandem cyclization for the synthesis of benzo[d][1,3]thiazin-2-yl phosphonates involving C–P and C–S bond formation

A copper(ii)-catalyzed, high-efficiency and atom-economical synthesis of valuable organophosphorus compounds via tandem cyclization of o-alkynylphenyl isothiocyanates with phosphites is described. This protocol, having a good functional-group compatibility, provides a simple and direct pathway to organophosphorus heterocycles in good yields under mild conditions. The method could be efficiently scaled up to gram scale, thus providing a potential application of this cascade cyclization strategy in synthesis.

As the valuable precursors of many biologically active molecules, 1 organophosphorus compounds have wide applications in the eld of materials science, 2 medicinal chemistry, 3 organic synthesis, 4 natural products, 5 and ligand chemistry. 6 For example, a-amino and a-hydroxy phosphonic acids have been found to act as antibiotics, 7 antitumor agents, 8 and enzyme inhibitors. 9 Therefore, the synthesis of these organophosphorus compounds is still appealing. The construction of a C(sp 2 )-P bond on heterocycles is one of the most fundamental methods to synthesize the organophosphorus compounds. Through the efforts of many chemists, several extensively valuable methods have been established and developed. 10 For instance, the Duan group and the Ackermann group developed an Ag-mediated C-H/P-H functionalization method to construct a Csp 2 -P bond by using arylphosphine oxides and internal alkynes as the substrates (Scheme 1a). 11 At the same time, Studer and coworkers reported a novel Ag-catalyzed radical cascade reaction for the synthesis of 6-phosphorylated phenanthridines from 2isocyanobiphenyls and diphenylphosphine oxides (Scheme 1b). 12 Recently, Liang and co-workers also developed two cases of cascade functionalization of N-(p-methoxyaryl)propiolamides and alkynol substrates with diphenylphosphine oxides to construct phosphorylated heterocycles (Scheme 1c and d). 13 Although various utilized methods for the construction of Csp 2 -P bond on heterocycles have been established, the development of a new synthetic strategy from easily prepared starting materials is still a challenging task.
Recently, transition-metal-catalyzed cascade cyclization of oalkynylphenyl isothiocyanates with various nucleophiles provides a new and powerful synthetic strategy to synthesize different heterocycles. o-Alkynylphenyl isothiocyanates have extensively been used as versatile organic synthons for the construction of different compounds such as indoles, 14 quinoline, 15 thiazine 16 due to its high reactivity and easy preparation. Encouraged by this fascinating research and our continuing interest in the transformation of o-alkynylphenyl isothiocyanates, 17 we herein report an efficient copper-catalyzed cyclization of o-alkynylphenyl isothiocyanates with phosphites for the synthesis of phosphorylated heterocycles and related derivatives (Scheme 1e).
According to the literature procedure, 14c,18 the starting oalkynylphenyl isothiocyanates were prepared via the Sonogashira coupling of 2-iodoanilines with terminal alkynes, 19 followed by reacting with thiophosgene. At the outset, we used o-phenylethynylphenyl isothiocyanate 1a and diethyl phosphonate 2 as the substrates in a model reaction to optimize the conditions, and the results are summarized in Table 1. Firstly, different copper salts (20 mol%) were screened in the presence of Cs 2 CO 3 (2.0 equiv.) used as the base in MeCN (2 mL) at 80 C for 18 h (Table 1, entries 1-8). CuCl 2 was the best choice, leading to the desired product 3a in 41% yield. It is worth noting that trace amounts of the products were obtained in the absence of metal salts (Table 1, entry 9) and no product was obtained in the absence of base (Table 1, entry 10). These results indicated that the combination of a Lewis acid catalyst and a base is indispensable to afford the target product. Subsequently, we examined the base effect on the reaction (Table 1, entries 11-15). Lower yields were observed when other bases such as K 2 CO 3 , K 3 PO 4 , t-BuOK, NaOH, and Et 3 N were employed, whereas DBU gave the best yield (Table 1, entry 15). We next examined the solvent effect (Table 1, entries 16-21). When DCM was employed as the solvent, the highest yield of 60% was obtained (Table 1, entry 20). Finally, we examined the effect of temperature on the reaction. When the reaction temperature was reduced to 45 C, the reaction was completed with a yield of 75% (Table 1, entry 23). Increasing the reaction temperature to 100 C or reducing the reaction temperature to 25 C resulted in a diminished yield (Table 1, entries 22 and 24).
In order to further demonstrate the substrate scope, different o-alkynylphenyl isothiocyanates and phosphites were then explored; the results are summarized in Table 2. All reactions proceeded smoothly, leading to the desired 4H-benzo[d] [1,3] thiazin-2-yl phosphonate in moderate to good yields. For example, the substituents on the R 2 position of substrates 1 showed obvious electronic effects on the reaction. Compared with the substrates 1 with an electron-rich aryl group such as p-MeOC 6 H 4 and p-MeC 6 H 4 at the R 2 position, the reaction of the R 2 group in the substrates 1 bearing an electron-decient aryl, such as p-FC 6 H 4 , p-ClC 6 H 4 , and p-BrC 6 H 4 , could lead to the desired products (3b-3d) in lower yields. Surprisingly, no desired products were obtained when the R 2 group was an alkyl group, such as methyl, ethyl, n-butyl, t-butyl, and n-hexyl. However, when the R 2 group in the substrate 1 was the cyclopropyl group, the desired 4H-benzo[d] [1,3]thiazin-2-yl phosphonate was obtained in 52% yield. Electronic properties and substitution position on the benzene ring of substrate 1 did not hamper the reaction process. With both of electron withdrawing groups, such as F-, Cl-, Br-, CF 3 -, and electron-donating group Me-on the benzene ring, the reactions could afford desired products 3h-3o in moderate to good yields. The other two substitution products 3p and 3q were also obtained successfully under standard conditions. Diphenylphosphine oxide was also a suitable substrate for this cyclization, by reacting with 1 under the standard conditions, the corresponding products (Z)-(4-benzylidene-4H-benzo[d] [1,3] thiazin-2-yl)diphenylphosphine oxide 3s and (Z)-(4-benzylidene-6-bromo-4H-benzo[d] [1,3]thiazin-2-yl)diphenylphosphine oxide 3t were obtained in 60% yield and 52% yield, respectively. The structure of 3t was further conrmed using X-ray diffraction analysis (see Fig. S2 in the ESI †). Due to a kinetic effect according to Baldwin's rules and a smaller steric effect compared to the Eisomer, all products were uniformly formed as the Z-isomer. 20 It is worth mentioning that all these reactions could be efficiently scaled up to gram scale under the optimal conditions, providing a potential application in the synthesis industry.
In order to insight into the reaction mechanism more clearly, two radical control experiments were carried out. The reaction proceeded smoothly by using the radical scavenger 2,2,6,6tetramethyl-1-piperidinyl-oxy (TEMPO) or butylated hydroxytoluene (BHT) probably suggesting that the reaction may not undergo a radical pathway (Scheme 3). Taking the experimental results into account, a possible mechanism was proposed, which is shown in Scheme 4. Firstly, in the presence of a base, isothiocyanate moiety in compound 1 was attacked by phosphite to produce the intermediate A.
Intermediate A could then undergo isomerization to afford intermediate B.
This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 32211-32215 | 32213 organophosphorus heterocycles could be synthesized in good yields involving C-P and C-S bond formation in one pot. This present cascade cyclization strategy represented an effective way to construct phosphorus-containing small molecular N,Sheterocycles.

Conflicts of interest
There are no conicts to declare.