Graphene oxide-catalyzed trifluoromethylation of alkynes with quinoxalinones and Langlois' reagent

The direct C–H trifluoromethylation of alkynes and quinoxalinones has been achieved using a graphene oxide/Langlois' reagent system. This multi-component tandem reaction using graphene oxide as the catalyst and Langlois' reagent as the robust CF3 radical source results in the formation of olefinic C–CF3 to access a series of 3-trifluoroalkylated quinoxalin-2(1H)-ones.


Introduction
Graphene oxide (GO), exfoliated from oxidized graphite, has a large p-network and specic surface area, abundant defects and a wide range of oxygen functional groups involving alcohols, epoxides, carboxylic acids and bridging 1,3-ethers. 1 This ultrathin 2D carbocatalyst, using its inherent reactivity of oxidation, acidity and catalytic activity, has been successfully applied to oxidation, acid-or base-catalyzed transformations, oxidative coupling of amines, hydration of alkynes, aza-Michael addition reactions of amines and a,b-unsaturated compounds, ring opening of epoxy compounds, and dehydrogenationoxidation of nitrogen heterocycles. 2 However, GO as a catalyst in organic transformation still faces serious challenges, compared with its applications in electronics, optical materials, solar cells and biosensors. 3 Quinoxalin-2(1H)-ones are important compounds that exist in various bioactive natural products and pharmaceutical compounds, especially CF 3 -containing quinoxalin-2(1H)-ones showing potent anticancer activities (Fig. 1). 4 A tremendous amount of triuoromethylation strategies have been developed and are widespread in the pharmaceutical, agrochemical and material science elds ever since the incorporation of the CF 3 group was introduced into different organic molecules. 5 Various triuoroalkylated regents, such as Langlois (sodium triuoromethanesulnate, NaSO 2 CF 3 ), 6 Togni, 7 Umemoto, 8 and Ruppert-Prakash, 9 have been developed and successfully employed in triuoromethylation reactions. In this regard, many research groups have explored diverse synthetic approaches involving inorganic/organic oxidants, photocatalysis, or electrocatalysis, which synthesize 3-tri-uoromethylquinoxalin-2(1H)-ones using quinoxalin-2(1H)ones and CF 3 SO 2 Na, TMSCF 3 , or Zn(SO 2 CF 3 ) 2 as the tri-uoromethylated agent (Scheme 1, eqn (1)). 10 Within this arena, Langlois' reagent, as an electrophilic triuoromethyl radical, has been frequently used in the addition of electron-rich olens and alkynes (Scheme 1, eqn (2)). 11 In particular, a few threecomponent protocols of "CF 3 "-radical addition to terminal or unactivated alkenes with Langlois' reagent have delivered various 3-triuoroalkylated quinoxalin-2(1H)-ones under metal-free conditions (Scheme 1, eqn (3)). 12 These tri-uoroalkylated reactions exhibit mild reaction conditions, low catalyst loading, ubiquity and versatile functionality of directing C-CF 3 activation. Inspired by these successful discoveries, we turned our attention to the multi-component triuoromethylation of alkynes with quinoxalinones and Langlois' reagent. We envision that GO will facilitate CF 3 SO 2 -Na to generate the electrophilic CF 3 radical intermediate, which enables an intermolecular addition with alkynes to provide nucleophilic alkyl radicals that combine with the C3 position of the quinoxalin-2(1H)-ones.

Results and discussion
The GO material used in this investigation was prepared by Hummers oxidation of graphite and subsequent exfoliation, as reported. 13 The obtained GO material has been characterized by X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), and infrared spectrum (IR) analysis (see the ESI †).
Based on our hypothesis, our investigation commenced with the reaction of N-methylquinoxalin-2(1H)-one (1a), phenylacetylene (2a) and the CF 3 -transfer regent (Langlois' reagent) in the presence of GO (80 wt%) and a mixture of acetonitrile and EA (1 : 1 v/v) at 120 C. It was observed that the desired product, triuoroalkylated quinoxalinone (3a) was obtained as an E/Z isomeric mixture (2.8 : 1) with an isolated yield of 85% (Table 1, entry 1; additional details in Tables S1 and S2 in ESI †). The GO loading-screening results showed that reducing the amount of GO from 80 wt% to 50 wt% dramatically decreased the yield from 85% to 22%, but increasing it further to 100 wt% did not show any negative effect on the yield, suggesting that GO was required only in an 80 wt% loading (entries 2 and 3). Screening of other temperatures including 60 C (yield 10%) and 140 C (yield 79%) did not afford any advantage over 120 C (entries 4 and 5). The use of a single solvent, such as MeCN or EA, led to a very low yield (entries 6 and 7). While changing to other solvents mixed in MeCN, the triuoromethylation proceeded adversely and the yields diminished (entries 8 and 9). When Ar was used instead of air, the reaction was signicantly inhibited (entry 10). The absence of GO failed to afford 3a (entry 11), indicating that this carbocatalyst was essential in the promotion of triuoromethylation. Ultimately, we found that 1a, 2a and CF 3 SO 2 Na in the presence of 80 wt% GO at 120 C for 6 hours afforded the desired product 3a with 85% yield (E/Z ¼ 2.8 : 1, which was determined by the proton NMR of the mixture of E/Z isomers).
Aer exploring the scope of the quinoxalin-2(1H)-one derivatives, we turned our attention to the variation of the system of different alkynes under the optimized conditions (Table 3). Phenylacetylene substrates bearing methyl, methoxy and uoro groups successfully carried out the Csp-CF 3 bond addition and afforded the desired products (3y-3aa). When aliphatic alkynes with cyclohexane or cyclopropane substituents were assessed using the present method, moderate to excellent yields were obtained (3ab-3ac). The reaction of other alkyl alkynes could work equally well to produce the desired products with good yields (3ad and 3ae). Regarding alkyl alkynes, such as those with cyano, chloro, hydroxyl, and protected hydroxyl groups, lower reaction efficiencies were generally observed, supplying the triuoromethylquinoxalin-2(1H)-ones with yields ranging from 32% to 63% (3af to 3ai). Besides terminal alkynes, to our delight, this strategy also obeyed the triuoromethylation of propargylic Csp-CF 3 bond addition to provide 3aj, ak with 53% and 51% yields, respectively.
In order to demonstrate the effectiveness of this new strategy, a gram scale reaction was performed under the standard conditions. N-Methylquinoxalin-2(1H)-one (1a) (10 mmol), phenylacetylene (2a) (20 mmol) and Langlois' reagent (4 mmol) were subjected to the reaction in the presence of GO (1.28 g, 80 wt%) in a mixture of acetonitrile and EA (1 : 1 v/v) (60 mL) at 120 C. Aer 6 h, the desired product 3a was obtained with 72% yield, which demonstrated the practical application of this protocol to prepare 3-triuoroalkylated quinoxalin-2(1H)-ones on a gram-scale (Scheme 2).
To gain more insight into these multi-component tandem reactions of nucleophilic addition, we evaluated the effect of the radical scavengers butylated hydroxytoluene (BHT) and 2,2,6,6tetramethyl-1-piperidinyloxy (TEMPO). As shown in Scheme 3, the reactions were signicantly inhibited, indicating that radicals might be involved. In accordance with the literature, we conceived the generation of the triuoromethyl radical intermediate from Langlois' reagent. 5b,c,14 TEMPO was introduced to the optimized reaction conditions and we observed the formation of TEMPO adducts (I and II), which were conrmed by MS analysis (Scheme 3b). GO with high-oxygen content is usually labile at higher temperatures. X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared (FT-IR) analysis showed that a large number of epoxide groups and carbonyl functional groups are lost when GO participates in the reaction (Fig. S2-S4 †). GO and recovered-GO (GO-R) were analyzed by XPS in order to study On the basis of these observations, and the established catalytic activity of graphene oxide, 2,15,16 a plausible pathway for Csp-CF 3 bond addition has been outlined in Scheme 4. The edge sites with unpaired electrons and oxygen functional groups of the porous GO shell constitute the active sites, which enhance the formation of the triuoromethyl radical and the superoxide radical by a tandem single-electron transfer (SET). The initial reaction of NaSO 2 CF 3 generates the triuoromethyl radical (detected by MS) in the GO catalytic system by a SET process, 17 and is subsequently added to the alkyne to afford the trans-alkenyl complex A (detected by MS). The N-methylquinoxalin-2(1H)-one (1a) is immediately trapped into the generated alkenyl radical, resulting in the intermediate Nalkylquinoxalinone radical B. The unpaired electrons of GO reduce O 2 to form O 2 c À , 2g,15b which can abstract hydrogen atoms Table 2 Conversion of the prepared quinoxalin-2(1H)-ones, phenylacetylene and Langlois' reagent into 3-trifluoromethylquinoxalin-2(1H)ones a,b a Reaction conditions: 1 (0.2 mmol, 1 equiv.), 2a (0.4 mmol, 2 equiv.), Langlois' reagent (0.4 mmol, 2 equiv.), GO (80 wt%), air, 120 C, and 6 h. b Isolated yield. from radical B to furnish the adduct product 3a. At the same time, the active sites of porous GO are released to maintain the catalytic cycle.

Conclusions
We have developed a tandem radical addition strategy of an O 2assisted one-pot process to synthesize 3-triuoroalkylated quinoxalin-2(1H)-ones under ambient atmosphere and 80 wt% GO content. Mechanistic studies reveal that GO affected the formation of triuoromethyl and superoxide radicals and was subsequently capable of performing an auto-tandem radical addition. This unique multi-component tandem reaction facilitated access to CF 3 -substituted quinoxalinones by the catalysis of a readily available carbon material in the absence of metal catalysts. In a broader perspective, these transformations also emphasize the potential of GO in synthetic chemistry, particularly in metal-free synthesis, rather than simply using a precursor to graphene-based materials.

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