Palladium-catalyzed difluoroalkylative carbonylation of styrenes toward difluoropentanedioates

The introduction of fluorine atoms into organic molecules is an attractive but challenging topic. In this work, an interesting palladium-catalyzed difluoroalkylative carbonylation of aryl olefins has been developed. A wide range of aryl olefins were transformed into the corresponding difluoropentanedioate compounds with good functional-group tolerance and excellent regioselectivity. Inexpensive ethyl bromodifluoroacetate acts both as a difluoroalkyl precursor and a nucleophile here. Additionally, a scale–up reaction was also performed successfully, and further transformations of the obtained product were shown as well.


Introduction
Organic uorides play an important role in organic synthesis, medicinal chemistry, and materials science due to their special physical and chemical properties. 1 The introduction of a uorine atom into an organic molecule can oen change the biological activity and physical properties of the compound. Among the uorine containing moieties, the diuoromethylene group has good metabolic stability, and its electronwithdrawing character can affect the electronic properties, chemical properties, and biological reactivity of the adjacent functional groups among the uorine-containing compounds, thus it exists in diverse drug molecules. 2 For example, Maraviroc is a CCR5 co-receptor antagonist used for treating CCR5-tropic HIV-1 infection together with other antiretroviral medications. 3 Gemcitabine, a nucleoside metabolic inhibitor, is used as an adjunct therapy in the treatment of certain types of ovarian cancer, non-small cell lung carcinoma, metastatic breast cancer, and as a single agent for pancreatic cancer. 4 Lumacaor is a protein chaperone, used for the treatment of cystic brosis in patients who are homozygous for the F508del mutation in the CFTR gene by combining with ivacaor. 5 Tauprost, an ophthalmic prostaglandin analogue, has been used to lower intraocular pressure in patients with ocular hypertension or open-angle glaucoma (Scheme 1). 6 Due to the double bond of aryl alkenes conjugated with an aromatic ring, they are quite activated and will usually lead to selectivity and reactivity issues in organic transformations. On the other hand, it represents an attractive route to construct uorinated compounds by using aryl olens and commercially available uoroalkyl halides as the starting materials. 7 Various transition-metal-catalyzed-, 8 photoinduced-9 or N-heterocyclic carbene (NHC) catalyzed-10 uoroalkylation reactions of aryl alkenes with uoroalkyl halides have been developed for the preparation of difunctional saturated uorine-containing compounds (Scheme 2a).
Transition metal-catalyzed carbonylation is an effective strategy for preparing various functionalized carbonylcontaining compounds. 11 In recent years, more and more procedures for uoroalkylative carbonylations of olens have been developed for the construction of uorine-containing carbonylated compounds. For example, Liu's group reported a novel cooperative strategy based on palladium-catalyzed and iodine(III)-mediated b-uorocarboxylation of alkenes; 12 palladium-catalyzed multi-component peruoroalkylative carbonylation for the synthesis of b-peruoroalkyl esters and amides was also realized; 13 and a copper-catalyzed 1,2-tri-uoromethylation carbonylation of unactivated alkenes to get b-triuoromethylated aliphatic carboxylic acid derivatives has been reported recently. 14 However, most of them are based on unactivated aliphatic olens, and the transformation of aryl olens in the uoroalkylative carbonylation reaction to give the corresponding ester product is still not reported.
Under all the above discussed backgrounds, we became interested in developing a new uoroalkylative carbonylation procedure for aryl olens to construct uorine-containing carbonylated compounds. However, this strategy faces several challenges. As depicted in Scheme 2b, palladium-mediated single-electron reduction of uoroalkyl halide forms radical A, which subsequently adds to the aryl olens to generate a new benzylic carbon radical B. The intermediate B recombines with palladium to form a new active benzylic intermediate C.
Meanwhile, B may dimerize to form D, which may even continue to polymerize with other aryl olens to form product E. Intermediate C can give product F aer b-hydride elimination. 7a Intermediate C can also continue to add to another olen to form G, and then b-hydride elimination takes place to give product H. 15 Theoretically, uoroalkyl halide can also react directly with a nucleophile to give compound I, and C may be quenched by a nucleophile to form J. Hence, a selective and efficient procedure for uoroalkylative carbonylation of aryl olens to construct uorine-containing carbonylated compounds is a challenging topic.
Aer systematic optimizations, herein, we developed an efficient palladium-catalyzed diuoroalkylative carbonylation reaction for aryl olens. Ethyl bromodiuoroacetate acts both as a diuoroalkyl precursor and a nucleophile in this system. A wide range of aryl olens were transformed into the corresponding diuoropentanedioate compounds in good yields with broad functional group tolerance and excellent regioselectivity (Scheme 2c).

Results and discussion
Initially, styrene 1a and cheap ethyl bromodiuoroacetate 2a were chosen as the model substrates to evaluate the feasibility of this diuoroalkylation carbonylation reaction. To our delight, with DiPEA as the base and assisted by B(OH) 3 in dioxane at 80 C, the desired product 3aa was obtained in 57% yield in the presence of PdCl 2 and using Xantphos as the ligand (Table 1, entry 1). Subsequently, various bases were studied, and a lower yield was observed with Na 2 CO 3 (Table 1, entry 2). The desired product 3aa was not detected when using NaO t Bu as the base (Table 1, entry 3). We then studied the effect of palladium precatalysts, but unfortunately, reduced yields were obtained when Pd(OAc) 2 , Pd(PPh 3 ) 4 , or Pd(TFA) 2 were tested (Table 1, entries 4-6). Further screening showed that 0.4 mmol of boric acid was proven to be the best for the target transformation ( Then we studied the effect of ligands, but reduced yields were obtained with the tested ligands (Table 1, entries 10-13). Interestingly, the yield of 3aa was almost not changed when we decreased the pressure of CO to 5 bar (Table 1, entry 14). To our satisfaction, 3aa was obtained in 81% yield when CsF was used as an additive (Table 1, entry 15).
With the best reaction conditions in hand, we conducted our investigation into substrate scope, and a variety of aryl olens were tested (Scheme 3). Aryl olens with electron-donating groups, such as methyl, tert-butyl, methoxy, benzeneoxy, and 3, 4-dimethoxy groups were tolerated well to give the desired diuoropentanedioate products in moderate to high isolated yields (3ba-3ha), Notably, the yield can reach up to 88% when the olen bears a methoxy group at the para-position. It should be mentioned that ortho-substituted styrene provided the corresponding product in lower yield compared with parasubstituted styrene probably due to the steric hindrance (3ca vs 3ba). Aryl olens bearing electron-withdrawing groups, such as triuoromethyl, acetate, can afford the target products in moderate to good yields as well (3ia-3ja). For those substrates with halogen groups, including uoro, chloro, and bromo substituents, the desired products were isolated in good yields (3ka-3oa). To our delight, 3-vinylbenzo thiophene and 3-vinylquinoline were also tolerable under our standard conditions (3pa-3qa). Moreover, substrates with 1-biphenyl and 2-naphthalene moieties could also work well to give the corresponding products in good yields (3ra-3sa). 1-Vinylnaphthalene gave the corresponding product in lower yield compared with 2-vinylnaphthalene probably because of the steric hindrance (3ta vs. 3sa). Moreover, six additional examples of bromodi-uoroacetates were examined and the desired products 3ab-3ag were all isolated in moderate to good yields. However, when ethyl 2-bromo-2-uoroacetate and ethyl 2-bromoacetate were tested, very low or no yield of the desired product could be detected (3ah, 3ai).
To demonstrate the scalability and utility of this method, we conducted a scale-up reaction and further transformations of the obtained product 3aa. The desired diethyl 2, 2-diuoro-4phenylpentanedioate 3aa can still be obtained in 63% yield when we expanded the reaction by 10 times (Scheme 4a). Then 2, 2-diuoro-4-phenylpentanedioic acid was obtained with 92% yield by alkaline hydrolysis of the product 3aa in THF/H 2 O with LiOH as the base (Scheme 4b). Subsequently, 98% yield of 2,2-diuoro-4-phenylpentane-1,5-diol was achieved from the  product 3aa by using lithium aluminum hydride as the reductant (Scheme 4c).
In order to gain some insight into the reaction mechanism, several control experiments were performed. Firstly, the target product 3aa was not observed in the absence of carbon monoxide gas under the standard conditions (Scheme 5a). Secondly, only a trace amount of the target product 3aa was detected when the radical inhibitor BHT (2,3 equiv.) or 1, 1-DPE (1,1-diphenylethylene, 3 equiv.) was added to our model reaction under the standard conditions (Scheme 5b and 5c). Similarly, the yield of 3aa was decreased to 23% when the radical inhibitor TEMPO (3 equiv.) was added (Scheme 5d). Furthermore, radical inhibitors 1, 1-DPE and TEMPO both trapped the diuoroacetate radical and detected it in GC-MS (see the ESI †), which indicates that this reaction involves radical intermediates. Moreover, the possible intermediate 6 was prepared and then reacted with ethyl bro-modiuoroacetate 2a under standard conditions, and the desired product 3aa was obtained in 56% yield (Scheme 5e). The radical nature of this reaction was also proven by the ringopening radical clock reaction and 55% of the corresponding product 8 was obtained under our standard conditions (Scheme 5f).
Based on the above results and literature studies, 13,16 a plausible reaction mechanism is proposed (Scheme 6). The catalytic cycle starts from the active catalyst Pd 0 Ln species, which was generated from the PdCl 2 pre-catalyst. Then, the Pd 0 Ln complex induced a SET (single-electron transfer) process of bromodi-uoroacetate to give the corresponding diuoroacetate radical and a Pd I LnX species, followed by the addition of the diuoroacetate radical to aryl olen to give a new secondary benzylic radical I. Subsequently, the Pd I LnX species was reincorporated with the carbon radical I to afford the key intermediate II. It is important to mention that the complex II can be converted into III through reductive elimination. However, the reaction is reversible and compound III can react with the reactive Pd 0 Ln species and be reconverted into II. Aer the insertion of carbon monoxide, complex II will be transformed into intermediate IV.
Finally, intermediate IV reacts with another molecule of bro-modiuoroacetate and gives the desired nal product aer the reductive elimination procedure. Meanwhile, in the presence of DiPEA, Pd 0 Ln will be regenerated for the next catalytic cycle. Although their roles are not very clear, we believe that B(OH) 3 can promote the decomposition of bromodiuoroacetate for alcohol release and CsF is more like a buffer here.

Conclusions
In summary, a novel palladium-catalyzed procedure for diuoroalkylative carbonylation of aryl olens has been developed. We have overcome the previous inability to take advantage of aryl olens in diuoroalkylative carbonylation. A variety of aryl olens were transformed into the corresponding diuoropentanedioate compounds in good yields with broad functional group tolerance and excellent selectivity. Additionally, the scaled-up reaction to a 3 mmol scale can be performed smoothly with a similar yield. Furthermore, the produced diuoropentanedioate product can be efficiently converted to the corresponding diacid and diol in a facile manner.

Author contributions
XFW surprised this project and revised the manuscirpt. ZPB and YZ performed all the experiments. ZPB prepared the rst version of this manuscript.

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