NaBAr^F‑Mediated Electrophilic Trifluoromethylation of Nonactivated Silyl Enol Ethers

Abstract

In this work, a mild NaBAr^F-mediated electrophilic trifluoromethylation of nonactivated silyl enol ethers is reported, using an Umemoto-type chloride salt thanks to a reactivity-modulation through its counter-anion. Hence, the key to success is the catalytic generation of a highly reactive Umemoto trifluoromethylating agent with the non-coordinative BAr^F ₂₄ anion upon in situ anion exchange initiated by catalytic amounts of the commercially available and simple NaBAr^F ₂₄ salt. This alternative method enables a selective reaction towards α-trifluoromethylated ketones under mild reaction conditions and avoids the use of stoichiometric Sn reagents, offering a practical strategy for embracing further highly demanding substrates in trifluoromethylation reactions.

Fluorinated organic molecules are of increasing importance due to their broad application in the pharmaceutical and agrochemical industry, as well as in material science. − The introduction of a fluorine atom or a fluorinated group often induces significant changes in the chemical, physical, and physiological properties of these compounds, providing new opportunities for drug lead discovery. , Based on the advantage of fluorinated motifs to mimic functional groups in biologically active molecules, , the development of new fluorination reactions became highly appealing. − Among the different transformations, intensive research has been set towards the synthesis of α-trifluoromethylated carbonyls. Thus, various strategies to prepare α-CF₃ ketones have been reported, which include nucleophilic, − electrophilic, ,, radical, , or carbene processes with enolates, alkenes, and alkynes using trifluoromethyl reagents such as Umemoto salts, − Togni’s reagent, , TMSCF₃, , CF₃SO₂Cl, CF₃I, or CF₃Br, , among others. However, the electrophilic trifluoromethylation approaches still present important limitations. Since the introduction of electrophilic sulfonium salts as trifluoromethylating agents by Yagupolskii in 1984 and Umemoto in the early 1990s, further efforts have been made to generalize this chemistry (Scheme ). In one of the early reports of Umemoto, the trifluoromethylation of reactive alkali enolates was achieved using S-(trifluoromethyl)­dibenzothiophenium triflate (e.g., 1-OTf, Umemoto reagent) as CF₃-source. However, a boron-based Lewis acid was required to tune the enolate reactivity. At present, significant progress has been made for the α-trifluoromethylation of activated carbonyls such as β-ketoesters, which can now be considered well established and allow for mild set-ups (Scheme a). − However, the reactions with nonactivated silyl enol ethers still require harsh conditions (DMF at 80–100 °C) or the use of tin reagents (Scheme b). In the latter case, tetrabutylammonium difluorotriphenylstannate (Gingras’ salt) was needed as a mild fluoride source to avoid fast desilylation and subsequent decomposition of the enolate, which represents an intrinsic challenge with this type of substrates.

1. Electrophilic α-Trifluoromethylation of Ketones via Their Enolates with Umemoto-Type Sulfonium Salts for (a) Activated 1,3-dicarbonyl Substrates, (b) Nonactivated Silyl Enol Ethers, and (c) This Approach Using Catalytic NaBAr^F .

Motivated by the current limitations and challenges in this field and following our research programs on organofluorine − and anion chemistry, − we envisioned the feasibility of tuning the reactivity of the trifluoromethylating agent to match the one of the silyl enol ethers by in situ generation of the appropriate reactive species upon counter-anion exchange. Thus, we herein report a NaBAr^F-mediated − selective electrophilic trifluoromethylation of silyl enol ethers with 5-trifluoromethyldibenzothiophenium salts under mild conditions and avoiding the use of stoichiometric tin reagents (Scheme c).

We started our study by choosing the nonactivated cyclic silyl enolate 2a bearing a tert-butyldimethylsilyl (TBS) group as a model substrate and exploring the reactivity of several trifluoromethyl-sulfonium salts with different counter-anions at room temperature (Scheme a). As expected from previous reports, the Umemoto reagent 1a-OTf provided no reactivity in DMF at room temperature, and only a dismissed yield of 11% was obtained in dichloromethane after 72 h. This low reactivity could be attributed to the low solubility of the electrophilic CF₃-reagent in common organic solvents. Therefore, we decided to use a modified 2,8-difluoro-Umemoto sulfonium salt 1b, which was reported to show enhanced solubility and performance. In this case, an interesting counter-anion effect was observed. A clear trend in increasing reactivity to 23% on going from a tight, small anion such as chloride to the less coordinative BF₄-anion was spotted. To ascertain this hypothesis, the non-coordinating BAr^F ₂₄-anion was used, as better results might be expected. However, only a complex mixture of inseparable fluorinated products was formed.

2. Preliminary Reactivity Studies with (a) Different CF₃-Sulfonium Salts and (b) Catalytic 1b-BAr^F with 1b-Cl.

To overcome this issue, the in situ generation of catalytic amounts of 1b-BAr^F ₂₄ salt by anion exchange from a less active species was envisioned, aiming at controlling the selectivity while keeping a good reactivity of the CF₃-reagent. Thus, the reaction was carried out with 1b bearing Cl, OTf, and BF₄ counter-anions in the presence of 10 mol% of NaBAr^F ₂₄. To our delight, the combination with the chloride salt allowed us to boost the yield to 41%, which indicated a more efficient anion exchange by a favored precipitation of NaCl out of the solution. The same experiment with other sulfonium chlorides 1c and 1d was examined, but notably lower performance was observed for which we continued our investigation with the 1b salt. To confirm our anion-exchange pathway hypothesis, a reaction using 0.9 equiv of the 1b-Cl salt in combination with 0.1 equiv of the highly reactive 1b-BAr^F ₂₄ salt was then performed (Scheme b). In this case, a 43% yield and complete selectivity to the desired product 3a was obtained, indicating the crucial catalytic generation of the active species 1b-BAr^F ₂₄.

With these preliminary results in hand, the optimization of the model reaction of 2a with 1b-Cl salt and catalytic NaBAr^F ₂₄ was next carried out (Table ). The use of other BArF salts such as KBAr^F ₂₀ (10 mol%) led to a lower yield of the product (entry 2), whereas the increase of the equivalent of the 1b-Cl reagent did not affect the outcome of the reaction (entry 3).

1. Optimization of the Conditions for the Reaction of 2a .

entry	NaBArF 24 (mol%)	solvent	additive	3a yield (%)
1	10	DCM		41
2	10	DCM		35
3	10	DCM		41
4	10	DCM	4Å MS	49
5	10	DCM	4Å MS	29
6	10	DCM	4Å MS	40
7	20	DCM	4Å MS	32
8	5	DCM	4Å MS	28
9	10	CHCl3	4Å MS	39
10	10	DCE	4Å MS	45
11	10	THF	4Å MS	18
12	10	MeCN	4Å MS	1
13	10	Acetone	4Å MS	--
14	10	DCM	4Å MS	33
15	10	DCM	4Å MS	37
16	10	DCM	4Å MS	28

^aConditions: NaBAr^F ₂₄ (x mol%), 1b-Cl (1.0 equiv), 2a (0.2 mmol, 1.0 equiv), and 4Å MS (60 mg; 300 mg/mmol) as additive (if used) in the corresponding solvent [0.1 M] were reacted at r.t. (∼23 °C) for 72 h under argon atmosphere.

^bIsolated yield.

^cUse of 10 mol% of KBAr^F ₂₀ as the catalyst.

^dUse of 1.5 equiv of 1b-Cl.

^eAddition of 180 mg (900 mg/mmol) of 4Å MS.

^f24 h reaction.

^g6 days reaction time.

^hReaction at 0 °C.

ⁱReaction at 40 °C.

^jReaction at 80 °C.

However, although dry dichloromethane (DCM) was used as solvent, the addition of small amounts of 4Å molecular sieves (300 vs. 900 mg/mmol, entries 4 and 5, respectively) turned out to be beneficial, allowing reaching a 49% yield of 3a (entry 4). Although the 24 h reaction provided a 40% yield (entry 6), we decided to keep 72 h for maximizing the yield. Next, the catalytic loading of NaBAr^F ₂₄ was screened (entries 7 and 8). However, either higher or lower loadings than 10 mol% showed to be less efficient. Next, different solvents were tested (entries 9–13), showing the superiority of DCM, although other chlorinated solvents such as dichloroethane (DCE) were also competitive. Finally, the reaction temperature was varied (entries 14–16), but the highest yield (49%, entry 4) was obtained at room temperature.

After having optimized the conditions for the reaction of 2a, we explored the substrate scope using a variety of silyl enol ethers 2 to give the corresponding 2-trifluoromethylated indanones 3. As shown in Scheme , different substitution on the silyl group was tolerated, whereas labile trimethylsilyl (TMS) or groups containing aromatic residues such as triphenylsilyl (SiPh₃) or tert-butyldiphenylsilyl (TBDPS) led to a notable lower product formation (13–26%). Pleasingly, a 1 g scale synthesis was smoothly achieved starting from 2a and 3a was isolated in 37% yield.

3. Substrate Scope of the Reaction .

^a All reactions were carried out on a 0.2 mmol scale. Isolated yields are given. ⁱ Yield of 3a on a gram-scale synthesis. ⁱⁱ Product 3l was isolated with an inseparable impurity.

Since the best result was obtained with the TBS-enolate, the scope of the substrate substitution on the indanone core was then made with this silyl group. Derivatives substituted by methoxy, alkyl, halogens, and aromatic units were well-tolerated by this catalytic system, while better performance was observed with substrates bearing an electron donor substituent. This could be seen in the series of methoxy substituted enolates 3b–e. High yields were achieved with most of these compounds (42–55%), whereas with the 6-OMe substrate, a significant lower 7% yield (3d) was obtained. Moreover, α-substitution on the enolate unit (3i) and the variation of the ring size from 5 to 6 (3j) were also well-tolerated, as well as the enlargement of the aromatic system (3k). Notably, the method also performed moderately with challenging electron-deficient heteroaromatic groups such as pyridine (3l), while acyclic enolates were not converted to the desired trifluoromethylated products (e.g., 3m), presumably due to a faster competitive hydrolysis.

To gain insights into the reaction mechanism, we conducted several control experiments and kinetic studies (Scheme ). Initially, the role of sodium in the reaction was explored. Thus, the standard reaction was carried out in the presence of 15-crown-5 ether as a Na-scavenger agent (Scheme a). Since the reaction proceeded satisfactorily with only slightly lower yield after 24 h, we concluded that sodium itself is not involved as an activator in the reaction. Next, we investigated the effects of chloride anion and ammonium salts (Scheme b). To this aim, one equivalent of Bu₄NCl was added to the standard reaction as a source of strong coordinating anion, which led to no product formation. Moreover, the replacement of sodium for the more soluble ammonium BAr^F salt also showed a similar negative outcome. These results are in line with the inefficient catalytic generation of the active 1b-BAr^F ₂₄ reagent.

4. Control Experiments and Mechanistic Investigations .

^a (a) Sodium scavenging, (b) ammonium Cl and BAr^F ₂₄ salt effects, (c) use of other chloride-trapping BAr^F salts, (d) addition of extra catalytic amounts of NaBAr^F ₂₄ after 8 h reaction time, and (e) reaction monitoring.

In order to prove the need of abstracting the chloride anion from the trifluoromethylating agent, a chloride-trapping BArF salt was used in catalytic amounts instead of NaBAr^F ₂₄. In particular, the methyl diaryltelluronium salt 4 (10 mol%), known as a halogen abstractor by chalcogen bonding, − led to the product in a similar high 52% yield (Scheme c). Since the hydrolysis of the silyl enol ether to the corresponding ketone is the main competing reaction observed, the addition of an extra 10 mol% of catalyst after 8 h was explored to rule out the hydrolysis of the nucleophile by inefficient final sequestering of chloride as TBSCl (Scheme d). However, no significant improvement in terms of efficiency (30 vs 32% yield) was observed compared with the reaction using 20 mol% of 1b-Cl at once (see Table , entry 7), which is in line with the observed side reactions with higher amounts of the active 1b-BAr^F ₂₄ species and the competing partial decomposition of the nucleophile during the reaction. These observations indicate again the key role of efficient anion exchange and catalytic formation of the reactive Umemoto-type BArF salt for this reaction to take place.

Finally, the reaction progress was monitored by ¹H NMR (Scheme e). Due to the hydrolysis of the silyl enol ethers upon preparation of the samples isolated yields given, only the generation of product 3a could be tracked. Not surprisingly, considering the nonactivated character of the silyl enol ethers used, a relatively slow reaction could be observed at the employed ambient temperature. Although the desired product was formed in 33% after just 6 h, the reaction needs at least 24 h to reach a ∼40% yield and 72 h for ∼50% yield.

Based on our experimental studies and previous reports from the literature, , a proposed reaction mechanism is outlined in Figure . The process is initiated by anion exchange between the inactive 1b-Cl salt and NaBAr^F ₂₄ to form the reactive 1b-BAr^F ₂₄ species in catalytic amounts. The latter can then coordinate to silyl enol ethers 2a to form intermediate I, favoring the CF₃-group transfer by enolate nucleophilic attacks and release of the dibenzothiophene side product (RSR). The formed 2-trifluoromethylated TBS-oxonium species II further reacts with another molecule of 1b-Cl to lead to the final product 3a, along with TBSCl, and regeneration of the active species 1b-BAr^F ₂₄ that can re-enter the catalytic cycle.

1.

Proposed mechanism.

In conclusion, we have developed a novel method for the trifluoromethylation of nonactivated cyclic silyl enol ethers with a CF₃–Umemoto-type chloride salt reagent using commercially available and simple NaBAr^F ₂₄ salt in catalytic amounts under mild conditions. Mechanistic investigations suggest that the reaction proceeds via the in situ formation of a highly active Umemoto-BAr^F ₂₄ species, which should be generated in catalytic amounts to guarantee site selectivity in trifluoromethylation reactions. Thus, a carefully designed anion exchange and chloride-anion scavenging are key for this process. This strategy offers a practical and efficient method for electrophilic trifluoromethylation reactions, opening new possibilities for challenging substrates and enhancing site selectivity.

Experimental Section

General Information

¹H, ¹³C, and ¹⁹F-NMR spectra were recorded in CDCl₃, CD₃CN, or DMSO-d ₆ (reference signal ¹H = 7.26 ppm, ¹³C = 77.17 ppm for CDCl₃; ¹H = 1.94 ppm, ¹³C = 118.26 ppm for CD₃CN; and ¹H = 2.50 ppm, ¹³C = 39.52 ppm for DMSO-d ₆) on a Bruker Neo 400 or Agilent DD2 500. Chemical shifts (δ) are given in ppm, and spin–spin coupling constants (J) are given in Hz. Analytical thin layer chromatography was performed using silica gel 60 F254, and a solution of KMnO₄ or a solution of vanillin served as staining agent. Column chromatography was performed on silica gel 60 (0.040–0.063 mm). Exact masses (HRMS) were performed using electrospray ionization techniques (ESI) or atmospheric-pressure chemical ionization (APCI) and recorded on a Thermo Fisher Scientific Exploris 120 Electrospray Orbitrap, a Thermo Fisher Scientific LTQ Orbitrap XL spectrometer, or a Thermo Fisher Scientific Orbitrap Velos Pro. Gas chromatography with mass analysis (GC-MS) with electron ionization (EI) was performed on a Thermo Fisher Scientific Exactive GC-MS or a Thermo Fisher Scientific ISQ 7000 GC-MS. The employed solvents such as tetrahydrofuran (THF), diethyl ether (Et₂O), DCM, and toluene were distilled in a solvent purification system (SPS) and dried over 3 or 4 Å molecular sieves (MSs). Other solvents, such as DCE and chloroform (CHCl₃) were dried over 4 Å MS. Commercially available reagents were used without further purification. NaBAr^F ₂₄ and 2,8-bis­(trifluoromethoxy)-5-(trifluoromethyl)-5H-dibenzo­[b,d]­thiophen-5-ium chloride (1b-Cl) were thoroughly dried under vacuum and stored under argon. 2,8-Bis­(trifluoromethoxy)-5-(trifluoromethyl)-5H-dibenzo-[b,d]­thiophen-5-ium trifluoromethanesulfonate, silyl enol ether 2r, and telluronium salt 4 were prepared following known literature procedures. Unless otherwise stated, all reactions were conducted under an argon atmosphere. NMR yields in the optimization were obtained using 2,2,2-trifluoracetophenon (0.1 mmol, 0.5 equiv) as the internal standard.

General Procedure for the Trifluoromethylation of Silyl Enol Ethers

To an oven-dried Schlenk-tube, 4 Å MSs (60 mg), 2,8-difluoro-5-(trifluoromethyl)-5H-dibenzo­[b,d]­thiophen-5-ium chloride (64.9 mg, 0.2 mmol, 1.0 equiv), and sodium tetrakis­(3,5-bis­(trifluoromethyl)­phenyl)­borate (17.7 mg, 0.02 mmol, 10 mol%) were added. The reagents were suspended in DCM (2 mL) and stirred for 5 min at r.t. Silyl enol ether (0.2 mmol, 1.0 equiv) was added to the mixture at once, and the reaction mixture was stirred at r.t. for 72 h. The mixture was filtered through a short pad of silica, a small amount of silica was added to the resulting solution, and the solvent was evaporated. The silica mixture was dry-loaded onto a column and eluted with a mixture of pentane/DCM/EtOAc.

Glossary

Abbreviations

BAr^F ₂₄

Tetrakis­(3,5-bis­(trifluoromethyl)­phenyl)­borate

BAr^F ₂₀

Tetrakis­(pentafluorophenyl)­borate

The data underlying this study are available in the published article and its Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsorginorgau.5c00034.

• Additional screening experiments; experimental procedures; characterization data; ¹H, ¹³C, and ¹⁹F NMR collection (PDF)

O.T. has performed all experiments and characterization of the products. T.B. and O.G.M. have supervised the work. All authors have contributed to writing the manuscript and approved the final version. CRediT: Olaf Tjabben investigation, methodology; Tatiana Besset conceptualization, supervision, writing - review & editing; Olga García Mancheño conceptualization, supervision, writing - original draft.

No specific safety precautions are required beyond standard measures.

The authors declare no competing financial interest.

Published as part of ACS Organic & Inorganic Au special issue “Fluorine Chemistry”.