A Selective Single Step Amidation of Polyfluoroarenes

Abstract

This chemistry establishes a method for the synthesis of per- and poly-fluoroaryl acid amides, utilizing nucleophilic aromatic substitution. Traditionally, such amides are constructed in a two-step process, namely, ammonolysis and then N-acylation. Herein, good yields of N-polyfluoroaryl acid amides were achieved in a single step under mild reaction conditions. Key to achieving optimal yields is the use of two equivalents of the nucleophile. In addition, the mechanism of the reaction is discussed which has implications for other related nucleophilic substitutions.

Graphical Abstract

1. Introduction

Per- and polyfluoroarenes are important targets because they are active components in many agrichemicals, pharmaceuticals, and industrial manufacturing products [1]. More specifically, polyfluoroaryl amides are a common subclass found within polyfluoroarenes as seen in the insecticide Teflubenzuron[2] and the cystic fibrosis drug Tezacaftor [3] (Figure 1). Many of the current methods of synthesizing these polyfluoroarenes involve selective fluorine installation. Traditional procedures for direct fluorine installation such as halex, Balz-Scheiman, and C–H fluorination have significant drawbacks including harsh reaction conditions and/or low selectivity which often result in poor yields [4][5][6]. Perhaps more concerning, these methods often only allow limited exploration of fluorinated space which pertains to the number, location and directionality of fluorine atoms within a molecule. Furthermore, due to the functional group incompatibility of these reactions, the exploration of the fluorinated space must be accomplished at an early stage before the molecules become too structurally complex and possesses incompatible functionality. Arguably, as a result of these difficulties discovery programs investigating polyfluoroaryl chemical space are often left with incomplete understanding. Consequently, methods that enable exploration of fluorinated space at a late-stage are valuable.

Figure 1.

Bioactive compounds containing polyfluoroaryl amides or derivatives.

An alternative approach to exploration of organofluorine space involves the selective defluorination [7][8][9][10][11] and substitution of highly fluorinated organofluorines, a process akin to sculpting. The advantage being two-fold. Namely, within perfluoroarenes all the difficult to install C–F bonds are already in place, and are readily available as building blocks. Novel substrates can then be reached through selective hydrodefluorination or functionalization [12]. Thus, our lab and many others have explored the development of such reactions in hopes of developing sufficient tools that enable discovery programs to more fully explore organofluorine space, which is expected to positively impact the discovery of molecules with enhanced properties.

Nucleophilic aromatic substitution using nitrogen-based nucleophiles has been previously investigated on highly fluorinated pyrimidines and pyridines [13][14] with azides. More recently, azides of highly fluorinated arenes have been shown to be readily coupled with aldehydes to make the corresponding acid amide [15][16]. Further, the reaction of hexafluorobenzene with amine based nucleophiles has also been studied [17], as well as N-silyl amides [18]. Indeed, previous methods for the synthesis of N-polyfluoroaryl acid amides build from this methodology. Typically, the synthesis involves two steps starting from the commercially available perfluoroarene (Scheme 1A). The first step usually involves the addition of ammonium hydroxide to make the C–N bond, followed by an isolation step, and then formation of the N-amide in a second step via treatment with the acid anhydride and an acid [12]. In particular, perchloric acid has been used, and is corrosive and can produce explosive salts as side products or on metal surfaces with which it comes in contact. Thus, there are significant handling concerns, and as a result, it has become a less accessible reagent in many labs.

Scheme 1.

A comparison of the proposed work with the traditional route.

We speculated that the amide motif could be synthesized more directly, by adding it as an entire nucleophilic unit. Further, doing so had the potential to take place under milder reaction conditions which could allow a greater degree of functional group tolerance (Scheme 1B). Thus, we set out to develop reaction conditions for the direct C–F substitution of polyfluoroarenes with deprotonated acid amide nucleophiles.

2. Results and Discussion

Neutral amides are not generally strong enough nucleophiles to undergo S_NAr with the polyfluoroarenes, but upon deprotonation their nucleophilicity is significantly enhanced [19][20]. Therefore, our synthetic plan involved the use of a base to deprotonate the amide N–H, followed by the addition of the polyfluorinated arene to yield the desired N-polyfluoroaryl amide in a single step. We first observed that the use of just one equivalent of strong base (vide infra), or alternatively the use of weaker bases such as tertiary amines, led to nitrogen di-arylation or very low conversions altogether in the case of amine bases (Scheme 2). The di-arylation result can be understood to arise from the NH acidity of the product. Because the highly electron withdrawing fluoroarene is expected to stabilize the charge of the incipient aza-anion formed upon deprotonation, the polyfluoroarylation results in a significant acidification of the N–H bond. In the presence of weak bases, the product N–H is therefore selectively deprotonated. In the presence of one equivalent of strong base, K₁ may be adequately shifted towards the aza-anion, but the formation of product serves to protonate the initial amide. In both cases, the result is formation of a second N-arylation event. We reasoned that this could be avoided by using two equivalents of strong base that would fully deprotonate both the initial amide and the product amide. With both amides present during the reaction, the initial amide would outcompete the product amide for bond formation.

Scheme 2.

Mechanistic rational for the formation of N-monoaryl- and N-diaryl amides.

Based on our predictions, we attempted to synthesize N-(4-cyano-2,3,5,6-tetrafluorophenyl)-2,2,2-trifluoroacetamide (3a) from pentafluorobenzonitrile using 2.15 equivalents of NaH and 2.15 equivalents of trifluoroacetamide (Scheme 3), and we found that it indeed selectively produces the mono-arylated amide product (Table 1, entry 1). We then reduced the NaH loading such that only an equimolar amount of nucleophile would be generated. As anticipated, we observed a 2:1 diarylation to intended product ratio (entry 2). This result confirmed that adding sufficient base to account for the formation of the acidic product was key to avoiding the N,N-diaryl amide product (3a’). We next explored the effect of temperature on the reaction. When we ran the reaction at −10 °C, we only achieved 61% conversion (entry 3). However, as we increased the temperature the conversion increased (entries 2, 3, and 4). Unfortunately, as the temperature increased we observed the formation of a side product that appeared to arise from hydroxide addition [21]. This assignment was supported by GC-MS. Suspecting that this was due to adventitious water, we began using a modified Schlenk technique in which more rigorously anhydrous conditions were achieved. As a consequence, the hydroxide addition product was completely suppressed and resulted in a 99% assay yield of 3a as determined by ¹⁹F NMR.

Scheme 3.

Optimization of reaction conditions.

Table 1.

Pentafluorobenzonitrile optimization

Entry	X equiv	T °C	% Conversion	Scale (mg)	3a/3a’
1a	2.1	10	99	500	>20/1
2	1.0	27	99	50	0.33
3	2.1	−10	61	50	>20/1
4	2.1	−5	73	50	>20/1
5	2.1	0	85	50	>20/1
6a	2.1	−5	98	50	>20/1

^aRigorously anhydrous and air-free

After achieving excellent NMR assay yields we turned our attention to developing workup and purification conditions. Since we expected this reaction to be used in a preparative manner, we sought to avoid the need for column chromatography altogether. After quenching the reactions with a dilute HCl solution, we expected the reaction mixture to contain our intended product, 3a, a minimal amount of the di-arylated product, 3a’, excess trifluoroacetamide, and fluoride salts. We believed that the product amide, 3a, along with any undesired 3a’ could be extracted into 10:1 hexanes:ethyl acetate mixture. Meanwhile, excess trifluoroacetamide and fluoride salts could be rejected by aqueous washes. 3a treated in this way was a solid, and recrystallization from hexanes allowed removal of any trace diarylated product, 3a’. Indeed, by this method, we were able to achieve an 86% isolated yield with better than 99% purity on a 500 mg scale without the aid of column chromatography.

2.3. Scope

We next sought to evaluate the scope. Prior to campus closure [22], we assessed five additional compounds both in terms of the fluoroarene and the nature of the amide (Scheme 4). While the optimal equivalents of the base and the amide varied subtly (among the trifluoroacetamide products), the optimal reaction temperature varied significantly among the substrates. We found that the highly activated pentafluorobenzonitrile (3a and 4b) went to >99% conversion at 10 °C, but the less activated compounds octafluoronapthalene (4e) and perfluorobenzene (4f) required much higher temperatures of 70 °C and 140 °C, respectively. Moving from the trifluoroacetamide to the simple acetamide nucleophile (compound 4b), we found that the optimal temperature was the same as the reaction with trifluoroacetamide (4a). However, the NaH and the acetamide equivalents had to be quadrupled. This may have reflected the increased tendency of this amide to carry water into the reaction mixture. In previous reports of nucleophilic addition to 2,3,4-trifluoronitrobenzene (4d) with nitromethonate and Meldrum’s acid anion the ortho substitution product was the primary or exclusive product formed [23][24]. Instead, we observed complete conversion within 1 h and amidation exclusively at the 4-position, as evidenced in the fluorine and proton NMR spectra. Importantly, reaction of the trifluoronitrobenzene shows that perfluorination is not a hard requirement, though it clearly facilitates the reaction. Attempts to utilize the procedure to directly amidate minimally fluorinated 2-fluoropyridine resulted in 80% conversion in 1 h at 160 °C. Evidence for the product was seen in the crude 1H NMR and GCMS. The crude GCMS gave 2 well resolved signals consistent with the expected molecular ion (M+ = 136). They are consistent with N- and O-substitution products, suggesting the reaction may be possible even with monofluorinated pyridines. However, isolation of the products proved challenging. Pentafluoropyridine similarly underwent smooth reaction at room temperature to give selective amidation at the C4 position. Lastly, perfluoronapthalene, devoid of other activating functional groups, cleanly afforded product 4c, but required an increased temperature (70 °C) and elongated reaction time.

Scheme 4.

Evaluation of the scope of the 1 -step amidation. Assay yields were determined using 19F NMR and were based on the observed product distributions. Isolated yields are based off the mass of the isolated material.

The relatively low isolated yield is believed to be due to the incomplete development of workup conditions [22]. Similarly, N-(perfluorophenyl)acetamide (4f) could be synthesized from hexafluorobenzene, but required even higher temperatures (140 °C) due to the absence of additional electron withdrawing groups in the substrate. We accomplished this by use of microwave vial that could maintain the volatile hexafluorobenzene even at the elevated temperatures.

3. Conclusions

In conclusion, we have explored the preparative single step synthesis of polyfluoroarylated amides. This method allows the expedited synthesis of N-polyfluoroaryl amides in a single step in high yields and purities, which should enable faster production of these compounds and inherently increase accessibility of fluorinated compounds. Importantly, we found that product inhibition was a key complicating factor that could be circumvented by using sufficient amount of nucleophile such that the acidic product did not interfere with the desired reaction. This is a general strategy that is common to polyfluoroarylation owing to the aryl group’s tendency to acidify nearby protons.

4. Experimental

General Experimental:

All reagents were obtained from commercial suppliers (Aldrich, VWR, TCI Chemicals, and Oakwood Chemicals) and used without further purification unless otherwise noted. Dry tetrahydrofuran (THF) was distilled after refluxing over sodium until dry as indicated by the benzophenone ketyl radical. Reactions were monitored by a combination of thin layer chromatography (TLC), (obtained from sorbent technologies Silica XHL TLC Plates, w/UV254, glass backed, 250 μm, 20 × 20 cm) and were visualized with ultraviolet light, potassium permanganate stain, GC-MS (QP 2010S, Shimadzu equipped with auto sampler), ¹⁹F NMR and ¹H NMR (vide infra). Isolations were carried out using Teledyne Isco Combiflash Rf 200i flash chromatograph with Redisep Rf normal phase silica (4 g, 12 g, 24 g, 40 g) with product detection at 254 and 288 nm and by ELSD (evaporative light scattering detection). NMR spectra were obtained on a 400 MHz Bruker Avance III spectrometer and Neo 600 MHz. ¹H, ¹⁹F and ¹³C NMR chemical shifts are reported in ppm relative to the residual proteo solvent peak (¹H, ¹⁹F, ¹³C). Reactions were set up in oil baths at various temperatures that are indicated below. Alternatively, reactions performed above their atmospheric boiling points were performed in a Biotage Initiator Microwave Synthesizer using 0.5– 2 mL vials with penetrable septa at the indicated temperatures.

General procedure for the one-step amidation of fluoroarene:

NaH (stored in a glovebox) was added into a flame-dried round bottom flask with a stir bar. The flask was capped and the atmosphere exchanged with Ar (3x). Then trifluoroacetamide was dissolved in dry THF (previously dried by refluxing over Na) in a flame-dried flask with a rubber septum. The amide solution was transferred using a syringe to the flask containing the NaH which had been placed in an ice bath to cool under positive Ar pressure which allowed venting of the H₂ pressure formed upon addition of the acetamide. The mixture was allowed to stir for 20 minutes before the temperature was adjusted to the reaction temperature. Finally, the fluoroarene was added to the flask and the temperature was maintained. After 3 hours, the reaction was quenched by the addition of 0.1 M HCl. Workup: The reaction mixture was extracted 3 times with a hexanes: ethyl acetate mixture (3x reaction volume). The organic layer was washed with H₂O (3x) with 10x reaction volume. Finally, the organic layer was dried with MgSO₄, concentrated, and purified via recrystallization or column chromatography.

4a. N-(4-cyano-2,3,5,6-tetrafluorophenyl)-2,2,2-trifluoroacetamide was prepared via the general procedure described above, using 1.0 equiv of pentafluorobenzonitrile (500 mg), 2.1 equiv trifluoroacetamide, and 2.15 equiv of NaH. The reaction temperature was maintained at 10 °C for 3 h. Mass isolated 636.6 mg, 85.83% isolated yield, mp 95–95 °C, yellow solid. Yellow solid; mp 95–96 °C. ¹⁹F NMR (376 MHz, Chloroform-d) δ −74.8, −130.8 – −130.9 (m), −140.0 – −140.2 (m). ¹H NMR (400 MHz, Chloroform-d) δ 8.26 (s, 1H). ¹³C NMR (151 MHz, Chloroform-d) δ 154.8 (q, J = 40.1 Hz), 147.5 (ddt, J = 263.5, 14.7, 4.3 Hz), 142.1 (dddd, J = 256.8, 13.6, 4.5, 3.0 Hz), 119.4 (tt, J = 13.7, 2.7 Hz), 115.1 (q, J = 287.9 Hz), 106.7 (t, J = 3.5 Hz), 93.7 (t, J = 17.2 Hz).

4b N-(4-cyano-2,3,5,6-tetrafluorophenyl)acetamide was prepared via the general procedure described above, using 1.0 equiv of pentafluorobenzonitrile (50 mg), 8.4 equiv acetamide, and 8.6 equiv of NaH. The reaction temperature was maintained at 10 °C for 4.5 h. The crude product was purified via column chromatography. Mass isolated: 30.0 mg, 60% isolated yield, white solid; mp 148–152 °C. The compound matches the previously reported spectra. [26]

4c. N-(2,3-difluoro-4-nitrophenyl)-2,2,2-trifluoroacetamide was prepared via the general procedure described above, using 1.0 equiv of pentafluorobenzonitrile (500 mg), 1.3 equiv trifluoroacetamide, and 2.4 equiv of NaH. The reaction temperature was maintained at 60 °C for 1 h. Mass isolated: 652 mg; 85.6% yield, mp 123–126 °C. ¹⁹F NMR (376 MHz, Chloroform-d) δ −75.4 (d, J = 1.1 Hz), −121.3 (ddd, J = 19.9, 8.4, 4.8 Hz), −130.9 (ddd, J = 19.9, 7.1, 2.2 Hz). ¹H NMR (400 MHz, Chloroform-d) δ 9.31 (s, 1H), 8.09 (ddd, J = 9.4, 4.8, 2.2 Hz, 1H), 7.37 (ddd, J = 9.4, 8.4, 7.1 Hz, 1H). ¹³C NMR (101 MHz, DMSO) δ 155.8 (q, J = 38.5 Hz), 153.4 (dd, J = 257.0, 11.9 Hz), 145.9 (dd, J = 254.5, 14.9 Hz), 142.0 (d, J = 3.4 Hz), 122.3 (dd, J = 9.1, 3.8 Hz), 120.2 (dd, J = 13.5, 2.5 Hz), 117.2 (d, J = 19.0 Hz), δ 116.0 (d, J = 287.9 Hz).

4d. 2,2,2-trifluoro-N-(perfluoropyridin-4-yl)acetamide was prepared via the general procedure described above, using 1.0 equiv of pentafluorobenzonitrile (500 mg), 3.784 equiv trifluoroacetamide, and 2.31 equiv of NaH. The reaction temperature was maintained at 27 °C for 3.5 h. Mass isolated: 133.8 mg; percent yield 74.7%. ¹⁹F NMR (376 MHz, CDCl₃) δ −74.9, −87.1 – −87.5 (m), −143.9 – −144.1 (m). ¹H NMR (400 MHz, Chloroform-d) δ 7.99 (s, 1H). ¹³C NMR (101 MHz, CDCl3) δ 154.4 (q, J = 40.2 Hz), 143.8 (apparent d, J = 245.9 Hz), 136.9 (dd, J = 265.3, 37.7 Hz), 125.4 – 125.0 (m), 115.2 (q, J = 287.9 Hz).

4e. 2,2,2-trifluoro-N-(perfluoronaphthalen-2-yl)acetamide was prepared via the general procedure described above, using 1.0 equiv of pentafluorobenzonitrile (50 mg), 2.1 equiv trifluoroacetamide, and 2.15 equiv of NaH. The reaction temperature was maintained at 70 °C for 48 h. The crude product was purified via column chromatography. Mass isolated: 10.5 mg; 21% isolated yield, white solid. ¹⁹F NMR (376 MHz, Chloroform-d) δ −74.9, −123.1 (dd, J = 65.5, 17.7 Hz), −140.3 (dq, J = 12.9, 4.3 Hz), −143.1 (dt, J = 64.6, 16.7 Hz), −143.9 – −145.2 (m), −145.4 – −147.0 (m), 151.0 (t, J = 18.7 Hz), −153.3. ¹H NMR (400 MHz, Chloroform-d) 7.69 (s, 1H).

4f. N-(perfluorophenyl)acetamide was prepared via the general procedure described above, using 1.0 equiv of pentafluorobenzonitrile (56 mg), 2.4 equiv acetamide, and 2.3 equiv of NaH. The reaction temperature was maintained at 140 °C for 2 h by sealing the reaction mixture in a microwave vial. An NMR assay yield 97% was obtained. White solid. The spectra match literature values. [25]

4g. N-(pyridin-2-yl)acetamide was prepared via the general procedure described above, using 1.0 equiv of 2-fluoropyridine (97.1 mg), 2.0 equiv acetamide, and 2.0 equiv of NaH. The reaction temperature was maintained at 160 °C for 1 h by sealing the reaction mixture in a microwave vial. An 19F NMR NMR yield indicated 80% conversion based on remaining 2-fluoropyridine. After workup with 0.1 M HCl and extracted with CDCl₃. GCMS showed two major peaks consistent with expected mass of the molecular ion. Calculated mass: C7H8N2O m/z: 136.06 (100.0%), 137.07 (7.7%). Observed Mass: (M+, 136.2), (M+1, 137.2), (M-Me, 121.1).

Highlights.

• Novel method allows for selective polyfluoroarylation of amides in a single step

• Good yields and mild reaction conditions