Synthesis of trifluoromethyl ketones by nucleophilic trifluoromethylation of esters under a fluoroform/KHMDS/triglyme system

A straightforward method that enables the formation of biologically attractive trifluoromethyl ketones from readily available methyl esters using the potent greenhouse gas fluoroform (HCF3, HFC-23) was developed. The combination of fluoroform and KHMDS in triglyme at −40 °C was effective for this transformation, with good yields as high as 92%. Substrate scope of the trifluoromethylation procedure was explored for aromatic, aliphatic, and conjugated methyl esters. This study presents a straightforward trifluoromethylation process of various methyl esters that convert well to the corresponding trifluoromethyl ketones. The tolerance of various pharmacophores under the reaction conditions was also explored.


Introduction
In recent decades, organofluorine molecules have received widespread attention in the field of medicinal chemistry [1][2][3][4]. The introduction of fluorine(s) into organic molecules usually leads to significant changes in the chemical and physicochemical properties of the original compounds [5,6]. Hence, the fluorination and related fluoro-functionalization of drug candi-dates are powerful strategies in drug design to appropriately bias their biological properties, bioavailability, and ADME [7,8]. While tremendous methodologies have been developed for the synthesis of organofluorine compounds [9,10], many of the laboratory methods are not always suitable for industrial production in terms of their synthetic complexity, handling, and Scheme 1: Chemistry of the CF 3 anion generated from HCF 3 . a) Decomposition of the trifluoromethyl anion to difluorocarbene and fluoride. b) A hemiaminaloate adduct of CF 3 anion to DMF. c) Formation of the [P4-t-Bu]H + CF 3 anion salt. d) Encapsulation of K + by glymes. Transformation of esters to trifluoromethyl ketones.
Several useful methods exist for preparing trifluoromethyl ketones [66,67], such as the direct trifluoromethylation of esters by the Ruppert-Prakash reagent (Me 3 SiCF 3 ) [68][69][70][71], but the use of HCF 3 for this transformation reaction is still limited. In 1998, Russel and Roques examined the transformation of methyl benzoate to trifluoromethyl phenyl ketone with HCF 3 in the presence of KHMDS or KH/DMSO in DMF, but the method required DMF and only a single example was indicated (Scheme 2a) [23]. Prakash and co-workers showed the first example of the DMF-free preparation of trifluoromethyl phenyl ketone with HCF 3 in the presence of KHMDS in THF, but they did not examined the scope of the reaction (Scheme 2b) [38]. In 2018, Szymczak and co-workers showed a single example of the preparation of phenyl trifluoromethyl ketone using HCF 3derived borazine CF 3 in 29% yield (Scheme 2c) [43]. Very recently, Han, Lian, and co-workers reported that a protocol using diisopropylaminosodium (NaDA) was useful for the tri- fluoromethylation of esters to trifluoromethyl ketones with HCF 3 at −60 °C (Scheme 2d) [44]. However, the preparation of NaDA was rather complicated and required pre-mixing of diisopropylamine, tetramethylethylenediamine (TMEDA), isoprene, and even more tedious "dispersion sodium" in n-heptane at 25°C for 4 h, before the reaction of esters with HCF 3 at −60 °C. We herein extend our glyme strategy [50] shown in Scheme 1d, the HCF 3 /KHMDS/triglyme system, for the synthesis of trifluoromethyl ketones from esters (Scheme 2e). The combination of HCF 3 and KHMDS in triglyme at −40 °C was found to be effective for this transformation, with good yields as high as 92%. The substrate scope of the trifluoromethylation procedure was explored for aromatic, aliphatic, and conjugated methyl esters. This study presents a straightforward trifluoromethylation process of various methyl esters that convert well to the corresponding trifluoromethyl ketones. The tolerance of various pharmacophores under the reaction conditions was also explored.

Results and Discussion
We first examined the trifluoromethylation reaction of methyl 2-naphthoate (1a) as a model substrate for HCF 3 to optimize the reaction conditions (Table 1). Following our glymes strategy, we initially used t-BuOK as the base in triglyme, and the desired trifluoromethyl ketone 2a was obtained in 29% yield (  Table S1).
We explored the substrate scope of this trifluoromethylation reaction with the optimized conditions in hand (entry 9, Table 1). Various carboxylic esters were investigated in the  presence of 1.1 equiv of HCF 3 and two equiv of KHMDS (Scheme 3). Methyl 2-naphthoate (1a) gave 2a in 75% yield, but sterically demanding methyl 1-naphthoate (1b) gave the desired trifluoromethyl ketone 2b in only lower yield (37%). Functionalities on the benzene ring at the para-position were well-tolerated in the KHMDS/glyme system. Halogen groups, such as chloro (1c), bromo (1d), and reactive iodo (1e) substitutions were also tolerated, resulting in the corresponding trifluoromethyl aryl ketones in moderate yields (56-63%) under basic conditions. The alkyl groups of tert-butyl (1f)-and cyclohexyl (1g)-substituted methyl benzoate derivatives, biphenyl benzoate (1h), and electron-donating 4-methoxybenzoate, were nicely transformed into aryl trifluoromethyl ketones in moderate to high yields (45-92%). Aryl substrates with a halogen attached at the meta-and ortho-positions were also accepted to furnish the desired products (2j-m) in good yields (66-82%). Moreover, di-substituted benzoate (1n), sterically demanding methyl adamantly carboxylate (1o), and conjugated methyl ester (1p) transformed effectively into trifluoromethyl ketones (2o-p) in moderate yields (50-62%). A gram-scale reaction was also Scheme 3: Substrate scope of esters 1 for trifluoromethylation by HCF 3 under the optimized conditions. a Determined by 19 F NMR of the crude 2 with trifluorotoluene as an internal standard. b Isolated yield. c Isolated yield of gram-scale reaction by using 1 g of substrate.
carried out for 1h, 1n, and 1p to furnish 2h, 2p, and 2n in similar isolated yields, 43%, 40%, and 36%, respectively. The double CF 3 addition product 3 was not observed due to the preferential formation of stable tetrahedral species I instead of the CF 3 ketones 2 in the reaction mixture. However, all the yields were moderate to good. This fact could be explained by the appearance of hydrate products 4 in the 19 F NMR spectrum of the crude reaction mixture [72], while the hydrates 4 disappeared completely after purification by silica gel column chromatography [73].
Given the relevance of this trifluoromethylation reaction system for drug discovery, we conducted a robustness screening experiment to gain further information on its tolerance to various pharmacophores ( Table 2). A range of common nitrogen-containing compounds such as pyridine, pyrazine, 1H-pyrazole, 1H-indole, 1-methyl-1H-indole, piperidine, and piperazine were subjected to screening. Pyridine and piperidine slightly hamper the reaction of 1g (Table 2, entries 2 and 7, 80-82%). Other nitrogen-containing compounds have more effect on the yield of the reaction of 1g (Table 2, entries 3-6, 58-72%). Next, a range of common oxygen and sulfur-containing compounds such as furan, tetrahydrofuran, 1,4-dioxane, thiophene, benzo[b]thiophene, dibenzo[b,d]thiophene, and diphenylsulfane were also screened. These substances also have some effect on the reaction ( Table 2, entries 9-15, 63-87%). Besides, silicon-containing compound, trimethyl(phenyl)silane that is more sensitive to fluorine was screened, 79% yield were obtained in this test. To consider the frequency of these motifs in modern pharmaceutical drugs, these tests are necessary, and the resistance of the reaction was also verified from various pharmacophores to be acceptable.

Conclusion
In conclusion, the trifluoromethylation of methyl carboxylates to trifluoromethyl ketones is accomplished under basic conditions with fluoroform in triglyme at −40 °C. An equivalent amount of fluoroform was sufficient for this transformation. A wide variety of medicinally attractive aryl and alkyl trifluoromethyl ketones are obtained in good yields by a relatively simple procedure, although the protocol is not applicable to enolizable esters. Fluoroform is an economical feedstock, and methyl esters are readily available inexpensive precursors. Besides, glymes are versatile solvents for chemical processes in industry [74] would the protocol be useful for the industrial extension, although there are still many points to be overcome such as requirements of low temperature, two equivalents of KHMDS. Further application of this "batch protocol" for a "continuous-flow microreactor" reaction is now ongoing in our laboratory towards industrial collaboration.

Experimental
A test tube containing 1 (0.4 mmol) in triglyme (0.7 mL) was charged with HCF 3 (9.9 mL, 1.1 equiv, measured by a syringe, see the picture in Supporting Information File 1, Figure S1) by cooling in liquid nitrogen under vacuum. KHMDS (160 mg, 2.0 equiv) in triglyme (commercial grade, without drying, 0.3 mL) was added at −40 °C under nitrogen atmosphere, and the reaction mixture was stirred at the same temperature for 4 h. Thereafter, 1 M HCl aq (1.0 mL) was added, and the aqueous layer was extracted with CH 2 Cl 2 (1.0 mL × 3). The combined organic layer was washed with brine, dried over Na 2 SO 4 , concentrated under reduced pressure, and purified by column chromatography on silica gel to give products 2.

Supporting Information
Supporting Information File 1 Optimization of reaction conditions, general procedure and product characterization data.

Funding
This work was supported by JSPS KAKENHI grants JP 18H02553 (KIBAN B, NS).