An eccentric rod-like linear connection of two heterocycles: synthesis of pyridine trans-tetrafluoro-λ6-sulfanyl triazoles

An eccentric trans-SF4 unit achieves a rod-like linear connection of two independent N-heterocycles, pyridines and triazoles.


Introduction
Hypervalent sulfur uorides belong to an interesting class of compounds, which are popular among both medicinal and material chemists alike. 1 The pentauoro-l 6 -sulfane (penta-uorosulfanyl, SF 5 ) group has turned out to be the front-runner of this class of compounds, with chemists realizing the varied potential of this moiety. 2 Over the years, a steady increase can be seen in the number of publications related to SF 5 -containing compounds. 2 On the other hand, the tetrauoro-l 6 -sulfane (tetrauorosulfanyl, SF 4 ) moiety, with equally potent functionality, has not been exploited enough and is highly underdeveloped. 3 The SF 4 moiety not only has unique physiochemical properties, which make it suitable as a unit of liquid crystals, 1,3 but also has interesting geometric features that enable it to connect two independent functional groups via the central hypervalent sulfur atom in either the cis or trans conguration of R-SF 4 -R 0 (Fig. 1a). 3,4 Due to the octahedral geometry of the R-SF 4 -R 0 moiety, the trans-SF 4 conguration has the ability to function as a building block in the construction of linear structures via its axial bonds, which makes the trans-SF 4 conguration highly signicant. 3,4 The rod-like connection by the trans-SF 4 unit potentially suggests new approaches for designing novel pharmaceuticals, while also leading to sought-aer uorine-containing drug candidates. As an extension of our research on uorine-containing heterocycles, 5 we were interested in the development of a novel method to connect two heterocyclic rings via a rod-like linear linker. Linear molecules are currently receiving chemists' attention due to their material and biological applications. 6,7 Bicyclo-[1.1.1]pentane (BCP) has gained the most attention in achieving linear connection 7 (Fig. 1b). Due to the intrinsically linear framework and lipophilicity of BCP, it has been proposed as a bioisostere of a psubstituted benzene ring 7e and as an alkenyl group, 7d both of which are found in pharmaceuticals. Our present objective is to propose the trans-SF 4 unit as a novel bio-isosteric unit next to BCP and to design rod-like molecules with two independent (hetero)aromatic rings (Fig. 1c).
While the initial idea of SF 4 -linked diaromatic compounds, Ar-SF 4 -Ar, appeared in 1973, 3a followed by a handful of reports (i.e., 4 papers and 1 patent), 3a-e only circumstantial evidence was reported. Kirsch and co-workers in 1999 obviously isolated trans-SF 4 -linked aromatic building blocks based on direct uorination of the corresponding bis(aryl)sulde (Fig. 1d), 3d but that method has a serious limitation, i.e., the necessity of substrates substituted by p-nitro groups to deactivate the aromatic moiety, preventing the reaction with uorine. Therefore, novel methodologies for trans-SF 4 -linked bis-aromatic compounds are highly desired. Besides, SF 4 -linked heteroaromatic building blocks have never been reported.
Heterocycles, which are common skeletal components of natural products, oen exhibit bioactive properties, and are thus extensively used as pharmaceuticals. 8 Pyridines and triazoles are two of the most commonly occurring N-heterocycles in medicinal chemistry. 9 The use of the trans-SF 4 moiety as a rod-like linear linker for pyridine and triazole groups would provide a fascinating novel set of compounds, which should have a very interesting physiochemical prole due to their linear structure 6,7 and the uorinated N-heterocycles. 10 Herein, we present the rst method for the connection of two independent N-heterocyclic molecules via a rod-like linear trans-SF 4 unit, where the axial bonds of the octahedral disubstituted SF 4 moiety are responsible for the linear connection (Scheme 1). First, the pyridine tetrauorosulfanyl chlorides 1 (Py-SF 4 Cl) were prepared from pyridine disuldes ((Py-S) 2 ) with KF/Cl 2 by oxidative chloro-tetrauorination. It should however be noted that the generation of the tetrauorosulfanyl chloride group in any organic molecule itself is challenging. The chloro-uorination needs to be performed under a completely dry and inert atmosphere in FEP bottles, while the isolation of the product requires special equipment. 2g,h,j Once the pyridine tet-rauorosulfanyl chlorides 1 were synthesized, based on our previous reports 2j,3k their radical addition to alkynes was attempted to furnished pyridine trans-SF 4 -alkenes 2. Then pyridine trans-SF 4 -alkenes 2 were converted to previously unknown pyridine SF 4 -alkynes 3 by treatment with LiOH$H 2 O. Finally, reaction of the pyridine SF 4 -alkynes 3 with azides 4 under thermal Huisgen 1,3-dipolar cycloaddition conditions provided the eccentric, three-dimensionally unique trans-SF 4 linked pyridine and triazole derivatives 5 in high yields (Scheme 1). 11 While a few examples of aryl-SF 4 -aryl were reported, 3 this is the rst example for the synthesis of heteroaryl-SF 4 -heteroaryl systems. An aryl-SF 4 -heteroaryl system was also accessed by the method. Since both heteroaryl and uorinated moieties are sought aer building blocks for drug candidates, our novel, eccentric heteroaromatic molecules should suggest new elds of drug design.

Results and discussion
The key precursors, pyridine SF 4 -alkynes 3, were prepared in good yields from the pyridine SF 4 -alkenes 2 2j,3k by subjecting the latter to dehydrochlorination under basic conditions. Thus, treatment of pyridine SF 4 chloroalkenes 2 with an excess amount of LiOH$H 2 O in DMSO at room temperature furnished the desired pyridine SF 4 alkynes 3. A wide variety of functional groups including halogens (Br, Cl, and F), an electronwithdrawing NO 2 group, and an electron-donating methyl group in the pyridine ring were well tolerated under these strong basic conditions. All the positions of the SF 4 unit on the pyridine ring, namely the ortho-, meta-and para-SF 4 -pyridines 2, were accepted for this transformation. The desired pyridine SF 4 alkynes 3 with an aryl or alkyl group at the terminal position were also obtained in good to excellent yields (Scheme 2).
With the precursor pyridine SF 4 -alkynes 3 in hand, we rst examined the reaction conditions that would allow the azide/ alkyne cycloaddition to take place (see the ESI for details †). Using alkyne 3a and benzyl azide 4a, initially ruthenium catalyst 12 Cp*Ru(PPh 3 ) 2 Cl 2 was used in toluene at 80 C and 110 C to give target product 5a in 24% and 32% yield respectively, as a mixture of 1,4-and 1,5-disubstituted isomers (isomer A and isomer B, entries 1 and 2). When the catalyst loading of Cp*Ru(PPh 3 ) 2 Cl 2 was increased, the yield decreased (entry 3) and a subsequent decrease of the catalyst increased the yield to 56% (entry 4). We then realized that our reaction does not require an Ru catalyst, but undergoes a thermal cycloaddition, where the catalyst initially led to the start of material decomposition. Running the reaction in the absence of a catalyst gave the product satisfactorily in 83% yield (entry 5). Being a thermal reaction, the regioisomers A and B were obtained in a 2 : 1 ratio. 13 The formation of 1,4-disubstituted isomer A was slightly preferred because it avoided the steric repulsions between the bulky SF 4 and the benzyl group (Table 1).
With the optimized reaction conditions in hand, we began the substrate screening by modifying the SF 4 -alkynes 3 (Scheme 3). Changing the halogen on the pyridine ring to Cl Scheme 1 Synthesis of trans-SF 4 linked pyridine and triazole compounds via a cycloaddition reaction.
Scheme 2 Synthesis of pyridine SF 4 -alkynes 3 from alkenes 2. a Reaction of 2 (1.0 equiv.) was performed in the presence of LiOH$H 2 O (10.0 equiv.) in DMSO at rt. gave product 5b in 86% yield with a ratio of 1.6 : 1 (isomers A : B). Having the electron-withdrawing NO 2 group on pyridine gave 5c in an excellent yield of 92%. It is of interest that in this case we observed a reversal in selectivity, with a preference for the 1,5-disubstituted product (isomers A : B ¼ 1 : 2). 4-Me pyridine SF 4 -alkyne 3d and unsubstituted pyridine SF 4 -alkyne 3e both sustained the reaction to give the desired triazole products 5d and 5e in 60% and 71% yield, respectively. Next, we analysed the effect of the electron-withdrawing NO 2 and electrondonating OMe on the phenyl ring of the alkyne moiety (3f and 3g). There did not appear to be any drastic effect of the substituents as both alkynes gave products 5f and 5g in 67% and 70% yield, respectively with a 1.5 : 1 ratio (isomers A : B). Replacing the aromatic ring by an aliphatic straight chain (n-butyl or n-pentyl) also gave products 5h and 5i in 68% and 70% yield, respectively. We further expanded our reaction by employing m-SF 4 pyridine alkyne 3j and p-SF 4 pyridine alkyne 3k as substrates. Both alkynes gave products 5j and 5k in excellent yields of 95% and 91%, respectively. The selectivity of the regioisomers increased slightly to 3 : 1. This reaction could be reproduced at a 1 g scale of alkyne 3a to 5a without any loss of yield.
The substrate scope of azides 4 was further investigated for the cycloaddition (Scheme 4). Changing the substituents on the benzene ring of the benzyl azides 4 gave good results. The presence of an electron-withdrawing group was sustained well to give products 5 having NO 2 (5l), Br (5m), and F (5n) on the benzene ring in good to excellent yields. Even a CN group was borne to give the triazole product 5o in 75% yield. An electron-donating group also gave products 5p (OMe) and 5q (Me) in 67-68% yields. It is noteworthy that, in the case of 5p, selectivity for the regioisomers increased to 6.6 : 1. The single crystal X-ray structures of each of the regioisomers of 5m, isomer A and isomer B, clearly revealed the specic regiochemistry of the octahedral sulfur centre, connecting the pyridine and the triazole parts linearly with its axial bonds and four uorines occupying the equatorial plane. Switching from benzyl to phenyl azide revealed that the 1,4-disubstituted products 5r-t (isomer A) formed exclusively. This was possibly due to the high steric hindrance of the SF 4 moiety and phenyl ring in the 1,5-disubstituted product (isomer B). Aliphatic azides 4 having 6 and 8 carbons were used and the respective products 5u and 5v were obtained (67-66% yields) with almost no regioselectivity. Cyclohexyl azide gave product 5w in moderate (52%) yield, while very bulky adamantyl azide gave exclusively the 1,4-disubstituted product 5x (isomer A) regioselectively in 37% yield. The cycloaddition reaction also proceeded well with two azide derivatives, quinine and epiandrosterone, to give products 5y and 5z respectively, having a druglike structure, in 37% to 67% yields. Isomer A of 5y was in fact positively obtained selectively over isomer B (6.4 : 1).
The cycloaddition reaction was also extended to benzene SF 4alkynes 6, which were synthesized according to a reported procedure 3i (see the ESI for details †). The reactions of these alkynes 6 were completely feasible with the simple benzyl azide 4a (Scheme 5). While the Br and Cl-benzene SF 4 -alkyne 6a and 6b gave products 7a and 7b in good yields of 56% and 70% respectively (isomer A : B ratio 2 : 1 and 2.3 : 1 respectively), the NO 2 -benzene SF 4 -alkyne 6c underwent the reaction to provide
Following the synthesis of the cycloaddition products, we used isomer A of 5a for a further application. Suzuki coupling via the bromo substituent on pyridine was attempted. 14 The use of phenoxyphenyl and benzofuran boronic acids, which are aromatic and heteroaromatic substrates, gave the coupled products 8a and 8b in 54% and 46% yield, respectively (Scheme 6).

Conclusion
In conclusion, we designed and synthesized novel tetrauorol 6 -sulfanes 5 having a linear connection between pyridine and triazole rings via the trans-tetrauoro-l 6 -sulfane (SF 4 ) moiety. The desired compounds 5 were obtained by thermal Huisgen 1,3-dipolar cycloaddition between pyridine SF 4 -alkynes 3 and azides 4. Benzene-SF 4 -triazole product 7 was also synthesized using the same protocol. Further coupling of 5 with boronic acids was also possible. These compounds are eccentric uorinated heterocycles with potential bioactive and surface properties. Further investigations on the application of 5 are underway.

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