Facile Rh(III)-Catalyzed Synthesis of Fluorinated Pyridines

Abstract

A Rh(III)-catalyzed C–H functionalization approach was developed for the preparation of multi-substituted 3-fluoropyridines from α-fluoro-α,β-unsaturated oximes and alkynes. Oximes substituted with aryl, heteroaryl and alkyl β-substituents were effective coupling partners, as were symmetrical and unsymmetrical alkynes with aryl and alkyl substituents. The first examples of coupling α,β-unsaturated oximes with terminal alkynes was also demonstrated and proceeded with uniformly high regioselectivity to provide single 3-fluoropyridine regioisomers. Reactions were also conveniently set up in air on the bench top.

Nitrogen heterocycles and their fluorinated analogues are ubiquitous and highly desirable motifs in pharmaceutical compounds.^1-3 While facile new syntheses of fluorinated pyri-dines have emerged in recent years,⁴ current methods of constructing pyridines with fluorine substitution at the 3-position require either functional group transformations upon preinstalled functionality at this site on the pyridine ring^5–9 or rely on heavily functionalized building blocks.^10–13 Herein we describe a new Rh(III)-catalyzed C–H functionalization approach to prepare 3-fluoropyridines bearing multiple substituents from α-fluoro-α,β-unsaturated oximes and alkynes.

Chiba¹⁴ and Rovis¹⁵ have established the utility of [Cp*RhCl₂]₂/metal acetate salt catalyst systems for the synthesis of multi-substituted pyridines from α,β-unsaturated oximes and internal alkynes.^16-17 However, we found that the nucleophilic alcoholic solvents utilized in their protocols, MeOH or 2,2,2-trifluoroethanol (TFE), posed a problem for the construction of fluorinated analogues due to alcohol displacement of the fluorine under the basic reaction conditions (Table 1, entries 1–2). To avoid fluoride displacement we examined a range of nonhydroxylic solvents, and while most proved to be ineffective (see Table S1 in the SI), ethyl acetate resulted in complete conversion to fluoropyridine 3a with minimal byproduct formation as determined by ¹⁹F NMR (entry 3). Unfortunately, very low conversion to fluoropyridine 3b was observed when diphenylacetylene (2b) was used as the alkyne partner, both with CsOPiv (entry 4) and the more soluble Bu₄NOAc as the acetate base (entry 5) even at a higher reaction temperature (entry 6). The sterically hindered alcohol solvents i-PrOH (entry 7) and t-BuOH (entry 8) were explored with the goal of improving reaction rate while minimizing fluoride displacement. t-BuOH proved to be the most effective in providing complete conversion with minimal byproduct formation (entry 8). Additionally, the loading of CsOPiv was evaluated, and 20 mol % was determined to be optimal (see Table S2 in the SI).

Table 1.

Solvent screen for Rh(III)-catalyzed fluoro-pyridine formation^a

entry	R	solvent	temp (°C)	oxime 1ab	pyridine 3b	byprodb
1	Et	MeOH	60	0%	48%	52%b
2	Ph	TFE	60	36%	29%	35%b
3	Et	EtOAc	60	0%	88%	12%
4	Ph	EtOAc	60	95%	2%	3%
5	Ph	EtOAcc	60	91%	4%	5%
6	Ph	EtOAcc	80	59%	22%	19%
7	Ph	i-PrOH	80	18%	57%	25%
8	Ph	t-BuOH	80	0%	89%	11%

^aAll reaction were set up in the glove box and run under nitrogen.

^bPercentages determined by ¹⁹F NMR with pen-tafluorobenzaldehyde as an external standard. Byproduct isolated by column chromatography and determined to be the product of fluoride displacement by the methoxy group.

^cRun with 20 mol % of Bu4NOAc instead of CsOPiv.

CsOPiv is highly hygroscopic as are the other carboxylate salts that have been used with Rh(III) catalysts in pyridine synthesis. For bench-top reactions we therefore envisaged that it would be important to determine the tolerance of the reaction to moisture. This was investigated by evaluating the effect of increasing amounts of water upon the reaction of oxime 1b and alkyne 2b, which are two of the more challenging coupling partners (Table 2). Significantly, up to stoichiometric amounts of water had minimal effect on either the yield of 3c or the formation of byproducts (entries 1-5). Furthermore, at 10 or more equivalents of added water, the reaction conversion was actually higher and was accompanied by only a small increase in byproduct formation (entries 6 and 7). Finally, increasing the reaction concentration from 0.1 M to 0.5 M, which is desirable for preparative reactions, resulted in a modest increase in conversion and yield (entry 8).

Table 2.

Concentration and added water screen for Rh(III)-catalyzed fluoropyridine formation^a

entry	[1a]	H2O	oxime 1bb	pyridine 3cb	byprodb
1	0.10 M	none	50%	27%	23%
2	0.10 M	10%	43%	32%	25%
3	0.10 M	20%	43%	33%	26%
4	0.10 M	40%	43%	32%	25%
5	0.10 M	100%	33%	42%	25%
6	0.10 M	1000%	7%	62%	31%
7	0.10 M	2000%	9%	61%	28%
8	0.50 M	1000%	3%	68%	29%

^aAll reactions were set up in a glove box and run under nitrogen.

^bPercentages were determined by ¹⁹F NMR with pen-tafluorobenzaldehyde as an external standard.

Because the synthesis protocol uses water and a high oxidation state catalyst, we also investigated the feasibility of pyridine synthesis with the reaction set up on the benchtop in air (Table 3). For the coupling of oxime 1a to alkyne 2a, no detrimental effect on the reaction rate or selectivity was observed when the reaction was set up in air (see entry 1 vs 2).

Table 3.

Comparison of Rh(III)-catalyzed fluoropyridine formations run under nitrogen and air^a

entry	time	oxime 1ac	pyridine 3ac	byprodc
1a	15 min	28%	63%	9%
	30 min	9%	81%	10%
2b	15 min	18%	74%	8%
	30 min	1%	91%	8%

^aSet up in a glove box and run under nitrogen.

^bSet up on bench-top in air.

^cPercentages were determined by ¹⁹F NMR with pentafluorobenzaldehyde as an external standard.

With optimized bench-top conditions established, we next explored the scope and generality of fluoropyridine synthesis (Scheme 1). Oximes 1 substituted with phenyl (3a, 3b, 3d-3g), alkyl (3c, 3h, 3i) and the electron-rich furyl (3j-3l) at the β-position each provided 3-fluoropyridines in moderate to excellent yields (Scheme 1). Symmetrical dialkyl and diaryl alkynes coupled in comparable yields for the different oxime coupling partners, as exemplified by 3-fluoropyridine 3a versus 3b, 3h versus 3c, and 3j versus 3k. Unsymmetrical internal alkynes also provided 3-fluoropyridines 3f, 3g, 3i and 3lin good yields, but with variable regioselectivities as has been previously reported for the preparation of non-fluorinated pyridines.^14-15 Attempts to incorporate internal alkynes with bulky t-butyl or TMS substituents were not successful.

Scheme 1.

Scope of Rh(III)-catalyzed fluoropyridine formation from oximes and internal alkynes^a

^a All reactions were set up in air on bench-top. Yields are based upon the mass balance of pure material after column chromatography.

To the best of our knowledge, terminal alkynes have not previously been demonstrated to be viable partners for Rh(III)-catalyzed pyridine formation.¹⁸ Moreover, while we had reported conditions for the Rh(I)-catalyzed synthesis of pyridines from α,β-unsaturated oximes and terminal alkynes, the regioselectivities were generally modest.¹⁹ In the current study, a range of terminal alkynes 2 coupled efficiently with oximes 1 to give single regioisomers of the 3-fluoropyridines 3(Scheme 2). Straight-chain alkyl (3m, 3r-3s) and branched alkyl (3n, 3p-3q) terminal alkynes, and even neohexyne (3o), were effective coupling partners. The complete selectivity for the formation of the 5-substituted 3-fluoropyridines 3 is consistent with the regioselectivity observed by Fagnou et al. for terminal alkyne insertion in their Rh(III)-catalyzed synthesis of isoquinolones.²⁰ Coupling of phenylacetylene with oxime 1a was also attempted, but did not yield any of the desired 3-fluoropyridine.

Scheme 2.

Scope of Rh(III)-catalyzed fluoropyridine formation from oximes and terminal alkynes^a

^a All reactions were set up in air on the bench-top. Yields are based upon the mass balance of pure material after column chromatography.

In conclusion, we have developed a one-step method for the preparation of 3-fluoropyridines from α-fluoro-α,β-unsaturated oximes and alkynes by Rh(III)-catalyzed C–H functionalization. The method is straightforward with reaction set up on the bench-top. α-Fluoro-α,β-unsaturated oximes and alkynes with a variety of alkyl, aryl and heteroaryl substituents are effective coupling partners. Moreover, the first examples of coupling terminal alkynes with α,β-unsaturated oximes with uniformly high selectivity provides an efficient approach to obtain single isomers of the 3-fluoropyridine products with predictable regioselectivity.