Nickel-catalysed sequential hydrodefluorination of pyridines: mechanistic insights led to the discovery of bench-stable precatalysts

Abstract

The nickel(0) complex [Ni(^iPrPN)(COD)] (^iPrPN = 2-[(N-diisopropylphosphino)methylamino]pyridine, COD = 1,5-cyclooctadiene) was an efficient precatalyst for the hydrodefluorination of pyridines employing pinacolborane (HBPin). 2-fluoro and 2,6-difluoropyridines were hydrodefluorinated at the 2- and 6-positions at room temperature in 3 h 30 min. The impact of the number of fluorine atoms and their position at the pyridyl ring in the efficiency of the catalyst was explored. Mechanistic experiments for the hydrodefluorination of 2,6-difluoropyridine allowed to identify COD decoordination followed by C–F oxidative addition as the catalyst entry pathway to the cycle and the [Ni(^iPrPN)(COD)] complex as the catalyst resting-state. The Ni(II) fluoride complexes, [NiF(^iPrPN)(6-Fpy)] (6-Fpy = 6-fluoropyrid-2-yl) and [NiF(^iPrPN)(py)] (py = 2-pyridyl) were independently synthesized and identified as intermediates in the two subsequent hydrodefluorination cycles operative through single-turnover experiments. Both Ni(II) fluoride complexes were found to be bench-stable precatalysts for the process with a comparable efficiency to [Ni(^iPrPN)(COD)] in the presence of a substoichiometric amount of COD to prevent catalyst deactivation.

---

Introduction

The hydrodefluorination (HDF) of small molecules finds applications in the synthesis of drugs, agrochemicals and materials, and is key for the remediation of perfluoroalkyl substances (PFAs).¹ Due to the strong nature of the C–F bonds, their activation² and catalytic transformation into C–H bonds by transition-metal complexes³ has attracted a lot of attention. Among small molecules, pyridines and low-degree fluorinated pyridines are a key motif in biologically active molecules,⁴ therefore, methods for the hydrodefluorination of pyridines hold promise for fluorinated drug synthesis and discovery, as well as for studying how fluorine affects the physicochemical and metabolic properties of these drugs.⁵ Additionally, HDF serves as a model process for deuterodefluorination (DDF),⁶ which enables the selective introduction of deuterium at specific positions on the pyridyl ring, constituting a relevant strategy for the synthesis of deuterated drugs.^6c,f Although notable advances have been achieved in the catalytic hydrodefluorination (HDF) of arenes,^3,7 including the discovery of main-group catalysts and of systems exhibiting broad substrate scope, the HDF of pyridines⁷ remains predominantly reliant on precious metal catalysts that require harsh conditions and the use of alkylsilanes as H sources to be efficient.^7b–e,8 More importantly, their efficiency is limited to the HDF of 2,3,4,5,6-pentafluoropyridine at the 4-position (Scheme 1a, top),^7a–c,e with the stoichiometric C–F activation or the catalytic HDF of more pharmaceutically-relevant partially fluorinated pyridines (containing 1–4 F atoms) being rare (Scheme 1a, bottom). Additionally, the limited substrate scope of the reported systems has precluded the understanding of the role that the number of F and their position or the presence of other functional groups at the pyridyl ring play on the HDF process, limiting further synthetic applications. In fact, only a Ru^7e and a Ti^9a,b and catalyst have been described efficient for the HDF of partially fluorinated pyridines, with the [TiCp₂F₂] (Cp = cyclopentadienyl) complex being found efficient for the HDF of a 4-substituted 2,6-difluoropyridine derivative (Scheme 1a, bottom).^9a Remarkably this HDF reaction found a direct application in synthesis, allowing to develop a synthetic route for ligand starting from the fluorinated starting material that cannot be synthetized otherwise. However, the system required 110 °C, 18 h, 30 mol% of catalyst loading and a 1.5-fold excess of H₂SiPh₂ to be efficient and employed a highly air- and moisture-sensitive Ti complex as precatalyst.

Scheme 1 State-of-the-art for the transition-metal catalysed HDF and mediated C–F activation of pyridines. (a) Catalytic HDF of pentafluoropyridine (top) and of a 4-substituted 2,6-difluoropyridine catalysed by a Ti complex (bottom) (ref. 9); (b) stoichiometric C–F bond activation of pentafluoropyridine by an in situ generated Ni(0) complex (ref. 10a); (c) this work: Ni-catalysed 2,6-regioselective HDF of partially fluorinated pyridines with pinacolborane.

In this context, nickel(0) complexes emerge as promising catalysts, as they have been found to show a marked preference for C–F bond activation over C–H,^10,2a,d and to be selective for the 2-position. Seminal work by Perutz and coworkers supported that the in situ generated 14-electron Ni(0) species, [Ni(PEt₃)₂], was efficient for the stoichiometric C–F activation at the 2-position of pentafluoropyridine and 2,3,5,6-tetrafluoropyridine under mild conditions (Scheme 1b).¹⁰ Although this stoichiometric C–F bond activation was leveraged to catalytic C–C bond formation employing SnBu₃(CH₂=CH₂),¹¹ the efficiency of this system for catalytic HDF has not been reported.

Attaining chemoselectivity for C–F activation in the presence of weaker C–H bonds^2d as well as achieving a high efficiency in pyridines with a decreased number of F¹² are among the main challenges for catalysts for the HDF of partially fluorinated pyridines.^7e,9 As an example, the presence of C–H bonds has been found to slow down the activation of C–F bonds,^2d with C–F bond activation of pentafluoropyridine proceeding five times faster than that of 2,3,5,6-tetrafluoropyridine with the [Ni(PEt₃)₂] species.^10c Additionally, air- and moisture-stable precatalysts are virtually absent in this field, limiting the practical applications of the process. Although Ni complexes for the HDF and C–F functionalization of fluorinated arenes have been reported,¹³ a Ni catalyst for the HDF of partially fluorinated pyridines remains unknown to date.

In this work we report the first regioselective HDF of partially fluorinated pyridines with pinacolborane (HBPin) employing a Ni(0) precatalyst (Scheme 1c). The efficiency of the system was found to depend on the number of F in the substrate as well as on their position. Mechanistic studies provided insights into the precatalyst entry pathway to the cycle, a catalyst deactivation pathway, the identity of the catalyst resting-state and enabled the discovery of two bench-stable Ni(II) precatalysts that were identified as key intermediates in the catalytic cycle.

Results and discussion

Catalyst screening and optimization of the catalytic conditions

Our research commenced with the assessment of the catalytic efficiency of the novel Ni(0) complexes [Ni(^RPN)(COD)] (^RPN = 2-[(N-dialkylphosphino)methylamino]pyridine, R = iPr (Ni1), R = Cy (Ni2), R = Ph (Ni3), COD = 1,5-cyclooctadiene, see Scheme 2a and ESI† for synthetic details) for the HDF of 2,6-difluoropyridine (1) with HBPin at room temperature for 30 min. In the presence of NEt₃ as FBPin scavenger,‡Ni1 (see Scheme 2a for the solid-state structure) afforded an 80% conversion of the starting material yielding a mixture of 2-fluoropyridine (1a, 25% yield) and pyridine (1b, 47% yield) from the single and double hydrodefluorination of 1 respectively at the 2- and 6-positions (entry 1 in Table in Scheme 2a). The Ni2 complex containing a cyclohexyl PN ligand afforded comparable conversion of 1 and yields of the products (entry 2 in Table in Scheme 2a), however, when phenyl groups were present at the P (Ni3) the catalytic efficiency was precluded (entry 3 in Table in Scheme 2a). These results support that the steric profile at the P plays a minor role in the transformation and that a strong σ-donating and poor π-accepting phosphine at the PN ligand is key to enable turnover. Additionally, the presence of a P donor at the ligand was also relevant to enable efficient catalysis, since when the Ni4 complex, containing 4,4′-ditertbutyl-2,2′-bipyridine, was employed as precatalyst, the starting material was fully recovered (entry 4 in Table in Scheme 2a). While [Ni(COD)₂] has been reported efficient for the directed HDF of fluorinated arenes,^13h the presence of the PN ligand in Ni1 was key for the HDF of pyridines to proceed, and when [Ni(COD)₂] was employed as the precatalyst, the starting material was recovered even when the reaction was conducted at 80 °C (see pp. S19 and S21 in the ESI†). In line with this observation, equimolar mixtures of [Ni(COD)₂] and the ^iPrPN or ^CyPN ligands were catalytically competent (see entries 1 and 2 in Table S2 in the ESI†), although afforded poorer conversions and yields (e.g. 43% conversion and 19% and <5% yields for 1a and 1b respectively for the Ni(COD)₂/^iPrPN system) than the well-defined precatalysts Ni1 and Ni2, presumably as a result of incomplete ligand coordination before catalytic turnover. The use of other ligands, including PEt₃, which has been reported to in situ afford the [Ni(PEt₃)₂] species efficient for the stoichiometric C–F activation of pyridines (Scheme 1b),¹⁰ led to full recovery of the starting material (entries 3–6 in Table S2 in the ESI†). The identity of the H source was also key enable efficient catalysis with HBPin being the most efficient when Ni1 was employed as the precatalyst, whereas HSi(OEt)₃ yielded lower conversion (62%) and yields of the products (46% and 6% for 1a and 1b respectively) and HBCat (Cat = catecholate), H₂SiPh₂ and HSiEt₃ afforded negligible yields of the products (see Table S1 in the ESI†). Since Ni1 and Ni2 showed a comparable efficiency for the process and because (a) the synthesis, purification, and crystallization of Ni1 was found to be more straightforward than that of Ni2, and (b) PCliPr₂, employed for the synthesis of the ^iPrPN ligand, was more affordable than PClCy₂, all the studies reported below were conducted employing Ni1 as precatalyst.

Scheme 2 Catalyst discovery and method optimization for the HDF of 2,6-difluoropyridine (1). (a) Catalyst screening and solid-state structure of Ni1 (right) with thermal ellipsoids at 30% probability; (b) optimization of the reaction conditions employing Ni1 as precatalyst. All the reactions were run employing 5 mol% of Ni1 as precatalyst in a 0.66 M THF solution and 3 equiv. of NEt₃. The conversions of 1 and 1a were calculated by ¹⁹F NMR spectroscopy integration employing fluorobenzene as internal standard. Yields of 1b were calculated by ¹H NMR spectroscopy employing mesitylene as internal standard.

An optimization of the reaction conditions employing Ni1 as the precatalyst and 1 as the substrate (see Scheme 2b for selected entries and see Table S6 in the ESI† for the full optimization) identified 2.5 equiv. HBPin, 25 °C, 3.5 h and THF as the solvent as the optimum, affording 1b as the exclusive product in 99% yield (entry 1 of the Table in Scheme 2b). When 2-fluoropyridine (1a) was employed as the starting material, only 1.5 equiv. of HBPin and 3 h were required for the single hydrodefluorination, affording 1b in 97% yield (see Table S7 in the ESI† for the optimization of conditions). Shorter times as well as lower or higher equiv. of HBPin resulted in decreased yields of 1b (entries 2 and 3, Table Scheme 2b), whereas the reaction showed a comparable yield of 1b when it was conducted at 40 °C (entry 6, Table Scheme 2b). However, raising up the temperature to 60 or 80 °C resulted in decreased conversions of 1 and yields of 1b, suggesting that the catalyst could be thermally unstable or that there could be a deactivation pathway operative at those temperatures (see below for mechanistic rationale). The system showed good efficiency in the green solvent 2-methyltetrahydrofuran (2-MeTHF, entry 8, Table Scheme 2b) as well as in THP, Et₂O (see entries 18 and 20 in Table S6 in the ESI†) and toluene (entry 9, Table Scheme 2b), however the yield of 1b decreased significantly when the reaction was conducted in pentane (entry 10, Table Scheme 2b). Control experiments under the optimized conditions resulted in the recovery of the starting materials (pp. S13–15 in the ESI†), supporting Ni1 as responsible for the HDF of 1 and 1a. Because aliphatic phosphines are well-known catalysts for the HDF of perfluorinated pyridines under harsher conditions,^3f,7f,g the catalytic HDF of 1 and of 1a were conducted employing 5 mol% of ^iPrPN as the precatalyst under the optimized conditions (page S21 in the ESI†) and at 80 °C (page S20 in the ESI†), however, in both cases the starting material was recovered, ruling out phosphine catalysis from decoordination of ^iPrPN in Ni1 as responsible for the catalytic transformation.

Effect of pyridine substitution on catalyst efficiency

With a set of optimized conditions in hand, the impact of the number of fluorine atoms and their position on the efficiency of Ni1 was explored. 2,6-Difluoropyridines (Scheme 3a, left) and 2-fluoro-6-H-pyridines (Scheme 3a, right) containing F at the 3-, 4- or 5-positions were selectively hydrodefluorinated at the 2- and 6-positions. The presence of F at the 3- and 5-positions (5, 7 and 8) resulted in the formation of the HDF products in excellent yields (e.g. >99% yield of 5b from 8), in contrast, when a F was present at the 4-position (4 and 4a), the efficiency of Ni1 significantly decreased (e.g. 16% yield of 4b from 4a), however, formation of 2,6-difluoropyridine, 2-fluoropyridine or pyridine from single, double or triple HDF including that at the 4-position of 4, was not observed, supporting that the catalyst retained its regioselectivity for these substrates.

Scheme 3 Substrate scope for the Ni1-catalyzed HDF of pyridines. (a) Evaluation of Ni1 efficiency for pyridines with different number of F at different positions. Conversions are listed below each substrate and were calculated by ¹⁹F NMR spectroscopy integration employing fluorobenzene as internal standard. Yields are listed below each product. ¹Yields were calculated by ¹⁹F NMR spectroscopy integration employing fluorobenzene as internal standard. ²Yields were calculated by ¹H NMR spectroscopy employing mesitylene as internal standard. *The reaction was run employing 1.5 equiv. of HBPin for 3 h in a 1 M THF solution. (b) Deactivating effect of F or BPin substituents at the 3-position for the sequential HDF of 2,6-difluoropyridine. All the reactions were run under the conditions described in part a.

For 2,3,6-trifluoropyridine (2), the formation of a mixture of products, 2a and 2b, in a 1 : 1.2 2a : 2b ratio from single and double HDF respectively, was observed, suggesting a decreased efficiency of Ni1 than for the HDF 1. The presence of 2a in the reaction crude suggests an incomplete HDF of this product and hints that the catalyst targets the F at the 2-position in the first HDF cycle, and that at the 6-position in the second cycle. An optimization of the reaction time allowed to obtain 2a (13 min) or 2b (16 h) as the major products (see Scheme 3a) with the reaction reaching >99% conversion of 2 in 13 min. Aiming to understand the role of the F atom for each of the substrates in the first and second cycles (2 and 2a respectively), the HDF of 2,3-difluoropyridine (7) and 2,5-difluoropyridine (2a) were conducted. For 7, the system was more efficient than for 2-fluoropyridine (1a) affording 3-fluoropyridine in 93% yield (vs. 68% yield of pyridine when 1a was the starting material). In contrast, for 2a the system was slightly less efficient than for 1a, affording 3-fluoropyridine in 57% yield. These results point toward the F at the 3-position facilitating the HDF at the 2-position but hindering it at the 6-position, allowing to rationalize why Ni1 targets the F at the 2-position in the first HDF cycle and that at the 6-position in the second for substrate 2. Additionally, they suggest that the incomplete double HDF of 2 is due to a more challenging second HDF cycle for which 2,5-difluoropyridine is the starting material.

The presence of a BPin substituent at the 3-position of the 2,6-difluoropyridine 3 afforded exclusively the product from a single HDF at the 6-position (3a, Scheme 3b right). This result suggests that the steric hindrance imposed by the BPin prevented the activation of the C–F bond at the 2-position, therefore precluding a double HDF. Additionally, product 3a was formed in lower yield (87%) than when 1 was employed as the substrate (99% yield of 1b) supporting that the BPin substituent at the 3-position also hindered the HDF at the 6-position. These results indicate that Ni1 is highly sensitive to the presence of substituents (X = H, F, BPin) at the 3-position of the pyridyl ring, which control whether a double HDF (for X = H, Scheme 3b, left), a single HDF (for X = BPin, Scheme 3b, right) or mixtures of products from double and single HDF (for X = F, Scheme 3b, bottom) are accessed.

Both, Ni1 and Ni2, were inefficient for the HDF of pentafluoropyridine (6) as well as for pyridines containing F only in the 3- and 5-positions (2b and 5b) for which the starting materials were recovered, further supporting the inability of the Ni intermediates to activate C–F bonds at the 3 and 5-positions.

Mechanistic insights

To gain insights into the intermediates involved in the catalytic cycle, the reaction of Ni1 with 1.1 equiv. of 1 in THF-d₈ at room temperature was monitored by ¹H, ¹⁹F and ³¹P{¹H} NMR spectroscopy (Scheme 4a, top). The NMR spectra supported that the reaction reached completion in 8 h yielding COD and a single Ni complex, Ni5, in quantitative yield formed at the expense of Ni1 and 1 (pp. S72–78 in the ESI†). The full characterization of independently synthesized Ni5 (see p. S8 in the ESI† for synthetic details) by NMR spectroscopy and single-crystal X-ray diffraction (Scheme 4b, top) supported its identity as the Ni(II) complex [NiF(^iPrPN)(6-Fpy)] (6-Fpy = 6-fluoropyrid-2-yl).¹⁴ A pathway to rationalize the formation of Ni5 from Ni1 and 1 involves COD decoordination in Ni1 to yield the unsaturated 14-electron Ni(0) species [Ni(^iPrPN)], followed by the C–F oxidative addition of 1. The [Ni(^iPrPN)] species was not detected by NMR spectroscopy, supporting a fast C–F oxidative addition in the NMR timescale.

Scheme 4 Mechanistic insights into the Ni1-catalyzed HDF of 2,6-difluoropyridine and discovery of bench-stable precatalysts. (a) Stoichiometric reaction of Ni1 with 1 or 1a followed by addition of HBPin (top) and with HBPin followed by addition of 1 (bottom); (b) solid-state structures of Ni5 (top) and Ni6 (bottom) (ellipsoids drawn at the 30% probability level); (c) competition experiment employing 1, 1a and Ni1 and (d) HDF of 1 catalysed by 5 mol% of Ni5 and of Ni5 or Ni6 after being exposed to atmospheric air and moisture for 24 h.

Addition of 1 equiv. of HBPin to the reaction mixture at room temperature resulted in the immediate disappearance of the signals attributed to Ni5 in the ¹⁹F NMR spectrum and the appearance of a new signal attributed to 2-fluoropyridine (1a) formed in quantitative yield as well as on the quantitative regeneration of Ni1 as confirmed by ³¹P{¹H} and ¹H NMR spectroscopy (Scheme 4a, top and pp. S72–78 in the ESI†). The immediate formation of Ni1 from Ni5 upon addition of HBPin supports a fast reaction of Ni5 with HBPin in the NMR timescale followed by a fast COD recoordination to the putative [Ni(^iPrPN)]. Formation of 1a supports Ni5 as a plausible intermediate in the cycle and the reaction of Ni1 with 1 as the catalyst entry pathway to the cycle. In line with this observation, the ¹H, ¹¹B and ³¹P{¹H} NMR monitoring of the reaction of Ni1 with 1.0 equiv. of HBPin in THF-d₈ at room temperature for 16 hours supported the presence of Ni1 and HBPin as the major species, hinting that they did not react at rt. Heating up to 80 °C for 1 h resulted in consumption of Ni1 and afforded a mixture of unidentified P-containing species that did not yield any product upon addition of 1 (Scheme 4a bottom and pp. S102–106 in the ESI†). Additionally, heating a solution of Ni1 in THF-d₈ at 80 °C for 1 h resulted in a minimum degree of thermal decomposition (see pp. S107 and 108 in the ESI†), with the species formed not being the same as those present in the ³¹P{¹H} NMR spectra from the reaction with HBPin, supporting that catalyst deactivation at 80 °C is due to reaction with HBPin rather than to thermal decomposition of Ni1. These results suggest that the reaction of Ni1 with HBPin is a catalyst deactivation pathway that could account for the decreased catalyst efficiency observed at 80 °C (see entry 25 in Table S6 in the ESI†).

An analogous result was obtained when the reaction of Ni1 with 2-fluoropyridine (1a) was monitored by ¹H, ¹⁹F and ³¹P{¹H} NMR spectroscopy, supporting the formation of the Ni(II) complex [NiF(^iPrPN)(py)] (Ni6, py = 2-pyridyl) as the exclusive product (see pp. S77–83 in the ESI†). Addition of HBPin resulted in the formation of pyridine (1b) as the exclusive product, supporting Ni6 as an intermediate in the second HDF cycle that yields 1b from 1a. Ni6 was independently synthesized (see page S9 in the ESI† for synthetic details) and fully characterized by NMR spectroscopy and single crystal X-ray diffraction (see Scheme 4b bottom for the solid-state structure) and showed a significantly decreased solubility in THF compared to that of Ni5. When an equimolar mixture of 1 and 1a was added to 1 equiv. of Ni1 the reaction reached completion in 2 h, yielding the complexes Ni5 and Ni6 in 80% and 20% yield respectively (Scheme 4c), supporting a faster C–F oxidative addition of 1 by Ni1 than of 1a. This observation is in line with previous reports for the C–F oxidative addition of pentafluoropyridine and 2,3,5,6-tetrafluoropyridine by [Ni(PEt₃)₂] which pointed toward a decreased rate of C–F oxidative addition for the latter, suggesting that a decreased number of fluorine atoms in the pyridine led to a slower C–F activation.^10c

When the HDF of 1 was conducted employing 5 mol% of Ni5 under the conditions in Scheme 4d, product 1b was obtained in 84% yield, supporting Ni5 as an efficient precatalyst for the process. Additionally, when 5 mol% of COD were added to the reaction mixture, the yield of 1b was increased to 97% (Scheme 4d) and was comparable to that when Ni1 was used as precatalyst (99%), hinting a relevant role of COD in the catalytic cycle (see below for a mechanistic rationale). More interestingly, when Ni5 or Ni6 that had been exposed to atmospheric air and moisture for 24 h were independently employed as precatalysts, comparable yields of 1b were obtained (Scheme 4d), supporting both complexes as a bench-stable precatalysts for the process.

Intrigued by the poor efficiency of the catalyst when a F at the 4-position was present in the substrate, the stoichiometric reactions of Ni1 with 2,4,6-trifluoropyridine (4) or pentafluoropyridine (6) followed by the addition of HBPin were monitored by NMR spectroscopy (Scheme 5a and b respectively and pages S90–S101 in the ESI†). When Ni1 was reacted with 1.1 equiv. of 4, the Ni–F complex, Ni7, was formed as the exclusive product from C–F oxidative addition at the 2-position of 4. The fact that Ni7 was formed as the exclusive product is consistent with the observed 2,6-regioselectivity for the catalytic HDF of 4. Additionally, the reaction reached completion in 25 min, supporting that the C–F oxidative addition of 4 is faster than that of 1 (8 h) and of 1a (41 h) and suggesting that the origin of the inefficiency of Ni1 toward this substrate is not due to a slow C–F oxidative addition. However, whereas Ni5 formed in quantitative yield from the C–F oxidative addition of 1, Ni7 formed only in 70% yield, hinting that the C–F oxidative addition of 4 is less efficient than that of 1 presumably due to partial decomposition of Ni7. Addition of 1 equiv. of HBPin to the crude mixture resulted in the immediate formation of 2,4-difluoropyridine (4a) in 18% yield and 4-fluoropyridine (4b) in <5% yield, supporting that the reaction of Ni7 with HBPin was not efficient, affording the HDF products in low yields (Scheme 5a). Additionally, as judged by ³¹P NMR spectroscopy, Ni1 was only partially regenerated along with the formation of a mixture of unidentified P-containing species, presumably from catalyst deactivation. Taken together all these observations suggest that the presence of a F at the 4-position of the substrate leads to a less efficient C–F oxidative addition and metathesis steps and to the partial decomposition of the catalytically active Ni species after the turnover, which allow to rationalize the observed low efficiency of Ni1 in the catalytic HDF of 4 and 4a.

Scheme 5 Stoichiometric reactions of Ni1 with (a) 2,4,6-trifluoropyridine and (b) pentafluoropyridine followed by the addition of HBPin providing insights into catalyst inefficiency for substrates with F at the 4-position (a) and catalyst loss of regioselectivity in the HDF of pentafluoropyridine (b).

A low efficiency of the C–F oxidative addition and the subsequent reaction with HBPin was also observed when pentafluoropyridine (6) was employed as the pyridine, which could be attributed to the presence of a F at the 4-position. In this instance, the reaction between Ni1 and 6 reached completion in <20 minutes, supporting a faster C–F oxidative addition than for 4, 1 and 1a and in good agreement with the trends previously reported^2d,10c whereby an increase in the fluorination degree of the pyridine results in a faster rate of the C–F oxidative addition (from fastest C–F oxidative addition to slowest: 6 > 4 > 1 > 1a). Remarkably, a mixture of two Ni–F complexes, Ni8 and Ni9, were formed as the products in 27% and 58% yields respectively (1 : 2.1 Ni8 : Ni9 ratio). These complexes resulted from the C–F oxidative addition of 6 at the 2- (minor) and 4- (major) positions respectively, suggesting that Ni1 was not regioselective for the C–F oxidative addition of substrate 6. Since substrate 4, with a F at the 4-position, retains the 2,6-regioselectivity for the C–F oxidative addition, the origin of the loss of regioselectivity for 6 could be traced back to the presence of F atoms at the 3- and 5-positions. Addition of HBPin to the crude mixture afforded a mixture of 3,4,5,6-tetrafluoropyridine (from Ni8, single HDF at the 2-position), 2,3,5,6-tetrafluoropyridine (5, from Ni9, single HDF at the 4-position), and 2,3,5-trifluoropyridine (8, from sequential HDF at the 4 and 2-positions) in a 1 : 3.6 : 2 ratio. The fact that 5 was formed in this reaction allows to understand why this product was observed in trace amount in the catalytic HDF of 6 (see page S151 in the ESI†). All the yields for the HDF products were lower (<20%, see Scheme 5b) than those of the corresponding parent Ni8 and Ni9 complexes, again pointing toward an inefficient reaction of the Ni–F complexes with HBPin. Additionally, the catalytically active species Ni1 was not regenerated after turnover, the ³¹P NMR spectrum after addition of HBPin showing signals consistent with the presence of a mixture of unidentified catalyst deactivation products.

Taken together, these results shed some light onto the origins of the inefficiency of Ni1 for the HDF of substrates containing F at the 4-position, whereby a slow C–F oxidative addition is not the origin of this inefficiency, but an inefficient metathesis with HBPin and the inability of the intermediates to regenerate the catalytically active species after turnover. Additionally, they put into relevance the role that the presence of F at the 3 and 5-positions play on the regioselectivity of the reaction, with F at these positions resulting in the loss of catalyst regioselectivity. Further efforts are undergoing in our laboratory to fully understand the origin of these effects.

To gain insights into the identity of the catalyst resting-state, the HDF of 1 with HBPin employing 10 mol% of Ni1 as precatalyst was monitored by ¹H, ¹⁹F, ¹¹B and ³¹P{¹H} spectroscopy in THF-d₈ at room temperature for 8 hours (pp. S26–33 in the ESI†, see Scheme 6a for the reaction profile). The ¹⁹F NMR spectra supported two catalytic cycles operative simultaneously, one for the transformation of 1 into 1a and another for that of 1a into 1b-BFPin. As a result of FBPin decomposition into BF₃ and B₂Pin₃ in the reaction medium,¹⁵1b-BFPin was found to partially convert into 1b-BF₃ over time. The independently NMR-monitored HDF of 1a (pp. S34–41 in the ESI†) further supported the identity of the second cycle in the HDF of 1 and revealed a comparable efficiency of Ni1 for the HDF of 1 and of 1a (94% and 90% conversion for 1 and 1a respectively after 1 h). This observation, along with the preference of Ni1 for 1 in the competition experiment above, allow to propose that the C–F oxidative addition step is not the rate-determining in the catalytic cycles. The ³¹P{¹H} NMR spectra showed one singlet at 120.8 ppm during the catalytic turnover, which was attributed to Ni1, allowing to identify this species as the catalyst resting state for both catalytic cycles. The appearance of 5 unidentified peaks in the ³¹P{¹H} NMR spectra and signals attributable to free COD in the ¹H NMR spectra, suggested that catalyst deactivation started being operative at high-substrate conversion (94% of 1 and 85% of 1a for the HDF of 1, 1 h of reaction). None of the signals attributed to catalyst deactivation products in the ³¹P{¹H} NMR spectra were present in that of the stochiometric reaction of Ni1 with HBPin at 80 °C, suggesting that deactivation of Ni1 by reaction with HBPin is not operative in the catalytic reaction at room temperature. When the Ni1-catalyzed HDF of 1 was conducted in the presence of 5 mol% of COD and monitored by NMR spectroscopy, the reaction only reached 39% conversion of 1 in 1 h (see pp. S45–53 in the ESI†), lower than in the absence of COD (94%, see Scheme 6b for the reaction profile). This observation supports that COD decoordination is necessary for C–F bond activation, which would take place on the unsaturated 14-electron species [Ni(^iPrPN)]. Additionally, COD decoordination should be reversible, the presence of excess COD shifting the equilibrium toward Ni1 which hindered the C–F oxidative addition step and the overall catalytic cycle. The unsaturated [Ni(^iPrPN)] species could be stabilized after its formation by THF (solvent) coordination, however, the use of non-coordinating solvents such as pentane did not prevent catalytic efficiency (see entry 22 in Table S6 in the ESI†), suggesting that solvent coordination to stabilize the C–F activating species is not a requirement for this system.

Scheme 6 Reaction profiles and proposed catalytic cycle. Reaction profiles for the HDF of 1 catalysed by (a) 10 mol% of Ni1 and (b) by 10 mol% of Ni1, Ni5 or Ni1 in the presence of 5 mol% of COD and (c) proposed catalytic cycle for the Ni1-catalyzed sequential HDF of 1.

NMR monitoring of the HDF of 1 employing 10 mol% of Ni5 as precatalyst, supported an efficiency of Ni5 comparable to Ni1 in the initial 30 min of the reaction (86% conversion of 1 with Ni5vs. 83% conversion with Ni1, see Scheme 6b and pp. S62–71 in the ESI†). However, as evidenced by the presence of a mixture of unidentified species in the ³¹P{¹H} NMR spectra, the deactivation of Ni5 started being operative after 10 minutes of reaction (63% conversion of 1 and 62% of 1a), earlier than for Ni1 (see above). As a result, it took significantly longer for Ni5 to reach 97% conversion of 1a than for Ni1 (8 h for Ni5vs. 2 h 45 min for Ni1). Early catalyst deactivation in Ni5 could be attributed to partial decomposition of the unsaturated C–F activating species [Ni(^iPrPN)] in the absence of COD. This observation allows to rationalize why the Ni5-catalyzed reaction in the presence of 5 mol% of COD afforded 1b in higher yield than in the absence of COD (see above), suggesting that COD stabilizes the [Ni(^iPrPN)] species by coordination, in turn delaying catalyst deactivation.

Based on the experimental evidence presented above, a plausible catalytic cycle for the formation of pyridine (1b) by the Ni1-catalyzed double HDF of 1 is depicted in Scheme 6c. The Ni1 precatalyst generates the 14-electron active species [Ni(^iPrPN)] by COD decoordination (rate-determining step) which promotes the C–F oxidative addition (regioselectivity-determining step) of 1 affording Ni5. Subsequent reaction with HBPin generates FBPin and the [NiH(^iPrPN)(2-Fpy)] intermediate. This intermediate has not been detected by NMR spectroscopy in the stoichiometric reaction of Ni5 with HBPin, unlike the analogous intermediate for the Pd-catalyzed HDF of pentafluoropyridine,^7c suggesting that it undergoes a faster C–H reductive elimination to afford 1a and regenerate the [Ni(^iPrPN)] active species. Because Ni1 is the catalyst resting state, the [Ni(^iPrPN)] active species should be in equilibrium with Ni1 by COD coordination/decoordination. An analogous catalytic cycle could be proposed for the transformation of 1a into 1b, that coordinates to the Lewis acidic FBPin byproduct affording 1b-BFPin,§ which partially converts into 1b-BF₃ over time (see p. S158 in the ESI† for the full catalytic cycle). Reaction of 1b-BFPin and 1b-BF₃ with NEt₃ allows to obtain 1b. Additional experiments are ongoing in our laboratory to gain further details into the mechanism operative in the transformation.

Conclusions

In conclusion, the first catalyst for the single and sequential 2,6-regioselective hydrodefluorination of partially fluorinated pyridines with HBPin has been discovered. The Ni(0) complex, [Ni(^iPrPN)(COD)], was an efficient precatalyst under mild conditions, with the number of F substituents at the pyridine ring and their position impacting the efficiency of the system. The presence of a BPin substituent at the 3-position of a 2,6-difluoropyridine allowed to conduct a single hydrodefluorination while retaining the F at the 2-position in the product. Mechanistic studies for the HDF of 2,6-difluoropyridine supported two simultaneous HDF cycles operative, with the [Ni(^iPrPN)(COD)] complex as the resting-state for both cycles. A competition experiment supported the catalyst preference for the C–F oxidative addition of 2,6-difluoropyridine rather than that of 2-fluoropyridine. Additionally, stoichiometric reactions allowed to rationalize the impact of pyridine substitution in catalyst efficiency and regioselectivity, with F atoms at the 4-position resulting in the loss of catalyst efficiency as a result of an inefficient metathesis with HBPin and the inability of the intermediates to regenerate the catalytically active species, and at the 3- and 5-positions resulting in the loss of catalyst regioselectivity as a result of a non-selective C–F oxidative addition. Single-turnover experiments and stoichiometric reactions supported that the [Ni(^iPrPN)(COD)] precatalyst entered the cycle by reversible COD decoordination followed by C–F oxidative addition to yield the [NiF(^iPrPN)(6-Fpy)] or [NiF(^iPrPN)(py)] complexes, which afforded 2-fluoropyridine or pyridine, respectively, upon reaction with HBPin and were found to be bench-stable efficient precatalysts for the transformation.

Author contributions

RA conceived the project, designed the experiments, supervised the experimental work, refined, and prepared for publication the X-ray diffraction data of Ni1, Ni5 and Ni6 and drafted the manuscript. VDA conducted the synthesis and characterization of Ni5 and Ni6, the substrate scope and the mechanistic studies, including the NMR monitoring of catalytic and stoichiometric reactions, and collected the X-ray diffraction data of Ni1, Ni5 and Ni6. RN discovered the HDF reaction, conducted the optimization of the reaction conditions, the control experiments, the in situ catalyst screening and the synthesis and characterization of Ni1–Ni4. VDA and RN drafted the ESI†. The manuscript was written through contributions of all authors/All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the ESI† including complete experimental details, characterization data and NMR spectroscopic data (PDF).

Crystallographic data for Ni1 (2420313), Ni5 (2420312) and Ni6 (2429444) has been deposited at the CCDC database under the cited accession numbers.

Acknowledgements

VDA thanks the Achievement Rewards for College Scientists (ARCS) Foundation for a fellowship. RA thanks Prof. Julio Perez for insightful discussions. This work was supported by funds from the NSF LEAPS-MPS-CHE program (grant # 2316526 to R. A.) and the NSF MRI program (CHE-2216471). University of California, Merced, start-up funds for Rebeca Arevalo.

References

1. (a) M. F. Kuehnel, D. Lentz and T. Braun, Synthesis of Fluorinated Building Blocks by Transition-Metal-Mediated Hydrodefluorination Reactions, Angew. Chem., Int. Ed., 2013, 52, 3328–3348; (b) J. B. Sap, N. J. Straathof, T. Knauber, C. F. Meyer, M. Médebielle, L. Buglioni, C. Genicot, A. A. Trabanco, T. Noël, C. W. am Ende and V. Gouverneur, Organophotoredox Hydrodefluorination of Trifluoromethylarenes with Translational Applicability to Drug Discovery, J. Am. Chem. Soc., 2020, 142, 9181–9187; (c) H. Amii and K. Uneyama, C−F Bond Activation in Organic Synthesis, Chem. Rev., 2009, 109, 2119–2183; (d) L. V. Hooker and J. S. Bandar, Synthetic Advantages of Defluorinative C−F Bond Functionalization, Angew. Chem., Int. Ed., 2023, 62, e20230880; (e) T. Ahrens, J. Kohlmann, M. Ahrens and T. Braun, Functionalization of Fluorinated Molecules by Transition-Metal-Mediated C–F Bond Activation To Access Fluorinated Building Blocks, Chem. Rev., 2014, 115, 931–972.
2. (a) S. A. Johnson, J. A. Hatnean and M. E. Doster, Functionalization of Fluorinated Aromatics by Nickel-Mediated C–H and C–F Bond Oxidative Addition: Prospects for the Synthesis of Fluorine-Containing Pharmaceuticals, Progress in Inorg. Chem, ed. K. D. Karlin, Wiley, Hoboken, NJ, 2011, vol. 57, pp. 255–352; (b) R. N. Perutz, T. Braun and D. Michael, Transition Metal-Mediated C–F Bond Activation, Comprehensive Organometallic Chemistry III, ed. P. Mingos and R. H. Crabtree, Elsevier Science, 2007, vol. 1, pp. 725–758; (c) J. L. Kiplinger, T. G. Richmond and C. E. Osterberg, Activation of Carbon-Fluorine Bonds by Metal Complexes, Chem. Rev., 1994, 94, 373–431; (d) E. Clot, O. Eisenstein, N. Jasim, S. A. Macgregor, J. E. McGrady and R. N. Perutz, C−F and C−H Bond Activation of Fluorobenzenes and Fluoropyridines at Transition Metal Centers: How Fluorine Tips the Scales, Acc. Chem. Res., 2011, 44, 333–348; (e) J. Weaver and S. Senaweera, C–F Activation and Functionalization of Perfluoro- and Polyfluoroarenes, Tetrahedron, 2014, 70, 7413–7428; (f) J.-Y. Hu and J.-L. Zhang, Hydrodefluorination Reactions Catalyzed by Transition-Metal Complexes, Organometallic Fluorine Chemistry, ed. T. Braun and R. P. Hughes, Springer, 1st edn, 2015, pp. 143–196; (g) N. A. Jasim, R. N. Perutz, A. C. Whitwood, T. Braun, J. Izundu, B. Neumann, S. Rothfeld and H.-G. Stammler, Contrasting Reactivity of Fluoropyridines at Palladium and Platinum: C−F Oxidative Addition at Palladium, P−C and C−F Activation at Platinum, Organometallics, 2004, 23, 6140–6149; (h) T. Braun and R. N. Perutz, Routes to Fluorinated Organic Derivatives by Nickel Mediated C–F Activation of Heteroaromatics, Chem. Commun., 2002, 2749–2757; (i) S. Erhardt and S. A. Macgregor, Computational Study of the Reaction of C₆F₆ with [IrMe(PEt₃)₃]: Identification of a Phosphine-Assisted C−F Activation Pathway via a Metallophosphorane Intermediate, J. Am. Chem. Soc., 2008, 130, 15490–15498; (j) B. L. Edelbach and W. D. Jones, Mechanism of Carbon−Fluorine Bond Activation by (C₅Me₅)Rh(PMe₃)H₂, J. Am. Chem. Soc., 1997, 119, 7734–7742; (k) M. Reinhold, J. E. McGrady and R. N. Perutz, A Comparison of C−F and C−H Bond Activation by Zerovalent Ni and Pt: a Density Functional Study, J. Am. Chem. Soc., 2004, 126, 5268–5276; (l) A. Das and N. Chatani, The Directing Group: a Tool for Efficient and Selective C–F Bond Activation, ACS Catal., 2021, 11, 12915–12930; (m) T. Schaub and U. Radius, Efficient C-F and C-C Activation by a Novel N-Heterocyclic Carbene–Nickel(0) Complex, Chem. – Eur. J., 2005, 11, 5024–5030; (n) D. Ertler, M. W. Kuntze-Fechner, S. Dürr, K. Lubitz and U. Radius, C–F Bond Activation of Perfluorinated Arenes Using NHC-Stabilized Cobalt Half-Sandwich Complexes, New J. Chem., 2021, 45, 14999–15016; (o) T. R. Dugan, J. M. Goldberg, W. W. Brennessel and P. L. Holland, Low-Coordinate Cobalt Fluoride Complexes: Synthesis, Reactions, and Production From C–F Activation Reactions, Organometallics, 2012, 31, 1349–1360; (p) S. A. Johnson, E. T. Taylor and S. J. Cruise, A Combined Experimental and Computational Study of Unexpected C−F Bond Activation Intermediates and Selectivity in the Reaction of Pentafluorobenzene with a (PEt₃)₂Ni Synthon, Organometallics, 2009, 28, 3842–3855; (q) J. A. Garduño, A. Arévalo and J. J. García, Bond and Small Molecule Activation with Low-Valent Nickel Complexes, Dalton Trans., 2015, 44, 13419–13438; (r) M. E. Doster and S. A. Johnson, Selective C–F Bond Activation of Tetrafluorobenzenes by Nickel(0) with a Nitrogen Donor Analogous to N-Heterocyclic Carbenes, Angew. Chem., Int. Ed., 2009, 48, 2185–2187; (s) T. Schaub, P. Fischer, A. Steffen, T. Braun, U. Radius and A. Mix, C–F Activation of Fluorinated Arenes using NHC-Stabilized Nickel(0) Complexes: Selectivity and Mechanistic Investigations, J. Am. Chem. Soc., 2008, 130, 9304–9317; (t) M. W. Kuntze-Fechner, H. Verplancke, L. Tendera, M. Diefenbach, I. Krummenacher, H. Braunschweig, T. B. Marder, M. C. Holthausen and U. Radius, Coligand role in the NHC nickel catalyzed C–F bond activation: investigations on the insertion of bis(N2(HC) nickel into the C–F bond of hexafluorobenzene, Chem. Sci., 2020, 11, 11009–11023.
3. (a) M. K. Whittlesey and E. Peris, Catalytic Hydrodefluorination with Late Transition Metal Complexes, ACS Catal., 2014, 4, 3152–3159; (b) J. T. Moore and C. C. Lu, Catalytic Hydrogenolysis of Aryl C–F Bonds Using a Bimetallic Rhodium–Indium Complex, J. Am. Chem. Soc., 2020, 142, 11641–11646; (c) S. S. Chitnis, F. Krischer and D. W. Stephan, Catalytic Hydrodefluorination of C−F Bonds by an Air-Stable P(III) Lewis Acid, Chem. – Eur. J., 2018, 24, 6543–6546; (d) N. O. Andrella, N. Xu, B. M. Gabidullin, C. Ehm and R. T. Baker, Selective Copper Complex-Catalyzed Hydrodefluorination of Fluoroalkenes and Allyl Fluorides: a Tale of Two Mechanisms, J. Am. Chem. Soc., 2019, 141, 11506–11521; (e) Z. Chen, C. He, Z. Yin, L. Chen, Y. He and X. Zhang, Palladium-Catalyzed Ortho-Selective C-F Activation of Polyfluoroarenes with Triethylsilane: a Facile Access to Partially Fluorinated Aromatics, Angew. Chem., Int. Ed., 2013, 52, 5813–5817; (f) A. Arévalo, A. Tlahuext-Aca, M. Flores-Alamo and J. J. García, On the Catalytic Hydrodefluorination of Fluoroaromatics Using Nickel Complexes: the True Role of the Phosphine, J. Am. Chem. Soc., 2014, 136, 4634–4639; (g) S. Sabater, J. A. Mata and E. Peris, Hydrodefluorination of Carbon–Fluorine Bonds by the Synergistic Action of a Ruthenium–Palladium Catalyst, Nat. Commun., 2013, 4, 2553; (h) J. Vela, J. M. Smith, Y. Yu, N. A. Ketterer, C. J. Flaschenriem, R. J. Lachicotte and P. L. Holland, Synthesis and Reactivity of Low-Coordinate Iron(II) Fluoride Complexes and Their Use in the Catalytic Hydrodefluorination of Fluorocarbons, J. Am. Chem. Soc., 2005, 127, 7857–7870; (i) N. M. Hein, F. S. Pick and M. D. Fryzuk, Synthesis and Reactivity of a Low-Coordinate Iron(II) Hydride Complex: Applications in Catalytic Hydrodefluorination, Inorg. Chem., 2017, 56, 14513–14523; (j) B. Zhu and S. Sakaki, C(Sp³)–F Bond Activation and Hydrodefluorination of the CF₃ Group Catalyzed by a Nickel(II) Hydride Complex: Theoretical Insight Into the Mechanism with a Spin-State Change and Two Ion-Pair Intermediates, ACS Catal., 2021, 11, 10681–10693; (k) O. Ekkert, S. D. Strudley, A. Rozenfeld, A. J. White and M. R. Crimmin, Rhodium Catalyzed, Carbon–Hydrogen Bond Directed Hydrodefluorination of Fluoroarenes, Organometallics, 2014, 33, 7027–7030; (l) A. Matsunami, S. Kuwata and Y. Kayaki, Hydrodefluorination of Fluoroarenes Using Hydrogen Transfer Catalysts with a Bifunctional Iridium/NH Moiety, ACS Catal., 2016, 6, 5181–5185; (m) H. Lv, J.-H. Zhan, Y.-B. Cai, Y. Yu, B. Wang and J.-L. Zhang, Π–Π Interaction Assisted Hydrodefluorination of Perfluoroarenes by Gold Hydride: a Case of Synergistic Effect on C–F Bond Activation, J. Am. Chem. Soc., 2012, 134, 16216–16227; (n) V. H. Mai and G. I. Nikonov, Hydrodefluorination of Fluoroaromatics by Isopropyl Alcohol Catalyzed by a Ruthenium NHC Complex an Unusual Role of the Carbene Ligand, ACS Catal., 2016, 6, 7956–7961; (o) S. P. Reade, M. F. Mahon and M. K. Whittlesey, Catalytic Hydrodefluorination of Aromatic Fluorocarbons by Ruthenium N-Heterocyclic Carbene Complexes, J. Am. Chem. Soc., 2009, 131, 1847–1861; (p) L. Schwartsburd, M. F. Mahon, R. C. Poulten, M. R. Warren and M. K. Whittlesey, Mechanistic Studies of the Rhodium NHC Catalyzed Hydrodefluorination of Polyfluorotoluenes, Organometallics, 2014, 33, 6165–6170; (q) M. K. Cybulski, D. McKay, S. A. Macgregor, M. F. Mahon and M. K. Whittlesey, Room Temperature Regioselective Catalytic Hydrodefluorination of Fluoroarenes with trans–[Ru(NHC)₄H₂ ] through a Concerted Nucleophilic Ru−H Attack Pathway, Angew. Chem., Int. Ed., 2017, 56, 1515–1519; (r) M. K. Cybulski, J. E. Nicholls, J. P. Lowe, M. F. Mahon and M. K. Whittlesey, Catalytic Hydrodefluorination of Fluoroarenes Using Ru(IMe₄)₂L₂H₂ (IMe₄ = 1,3,4,5-Tetramethylimidazol-2-ylidene; L₂ = (PPh₃)₂, dppe, dppp, dppm) Complexes, Organometallics, 2017, 36, 2308–2316; (s) R. Doi, M. Yasuda, N. Kajita, K. Koh and S. Ogoshi, Nickel-Catalyzed Exhaustive Hydrodefluorination of Perfluoroalkyl Arenes, J. Am. Chem. Soc., 2023, 145, 11449–11456.
4. (a) E. Vitaku, D. T. Smith and J. T. Njardarson, Analysis of the Structural Diversity, Substitution Patterns, and Frequency of Nitrogen Heterocycles among U.S. FDA Approved Pharmaceuticals, J. Med. Chem., 2014, 57, 10257–10274; (b) R. A. Ward and J. G. Kettle, Systematic Enumeration of Heteroaromatic Ring Systems as Reagents for Use in Medicinal Chemistry, J. Med. Chem., 2011, 54, 4670–4677; (c) R. D. Taylor, M. MacCoss and A. D. Lawson, Rings in Drugs, J. Med. Chem., 2014, 57, 5845–5859.
5. (a) M. Inoue, Y. Sumii and N. Shibata, Contribution of Organofluorine Compounds of Pharmaceuticals, ACS Omega, 2020, 5, 10633–10640; (b) M. Hird, Fluorinated Liquid Crystals – Properties and Applications, Chem. Soc. Rev., 2007, 36, 2070–2095; (c) S. Purser, P. R. Moore, S. Swallow and V. Gouverneur, Fluorine in Medicinal Chemistry, Chem. Soc. Rev., 2008, 37, 320–330; (d) V. Gouverneur and K. Seppelt, Introduction: Fluorine Chemistry, Chem. Rev., 2015, 115, 563–565; (e) K. Müller, C. Faeh and F. Diederich, Fluorin in Pharmaceuticals: Looking Beyond Intuition, Science, 2007, 317, 1881–1886; (f) D. O'Hagan, Fluorine in Health Care: Organofluorine Containing Blockbuster Drugs, J. Fluorine Chem., 2010, 131, 1071–1081; (g) J. He, Z. Li, G. Dhawan, W. Zhang, A. E. Sorochinsky, G. Butler, V. A. Soloshonok and J. Han, Fluorine-Containing Drugs Approved by the FDA in 2021, Chin. Chem. Lett., 2023, 34, 107578; (h) G. Chandra, D. V. Singh, G. K. Mahato and S. Patel, Fluorine-a small magic bullet atom in the drug development: perspective to FDA approved and COVID-19 recommended drugs, Chem. Zvesti, 2023, 1, 1–22; (i) G. Shabir, A. Saeed, W. Zahid, F. Naseer, Z. Riaz, N. Khalil, Muneeba and F. Albericio, Chemistry and Pharmacology of Fluorinated Drugs Approved by the FDA (2016–2022), Pharmaceuticals, 2023, 16, 1162; (j) D. O'Hagan and R. J. Young, Future Challenges and opportunities with flourine in drugs, Med. Chem. Res., 2023, 32, 1231–1234.
6. (a) J. Atzrodt, V. Derdau, W. J. Kerr and M. Reid, Deuterium- and Tritium-Labelled Compounds: Applications in the Life Sciences, Angew. Chem., Int. Ed., 2018, 57, 1758–1784; (b) N. Li, Y. Li, X. Wu, C. Zhu and J. Xie, Radical Deuteration, Chem. Soc. Rev., 2022, 51, 6291–6306; (c) R. M. Di Martino, B. D. Maxwell and T. Pirali, Deuterium in Drug Discovery: Progress, Opportunities and Challenges, Nat. Rev. Drug Discovery, 2023, 22, 562–584; (d) Z.-H. Yuan, H. Xin, Y. Gao, G.-W. Sun, R.-Y. Zhao, X.-H. Duan and L.-N. Guo, Photoredox-catalyzed Selective Deuterodefluorination of α,α-difluoroarylacetic Esters and Amides, Org. Chem. Front., 2024, 11, 4479–4486; (e) P. Yang, H. Yu, R. Zhai, J. S. Zhou and B. Tang, Nickel-catalyzed hydrodefluorination/deuterodefluorination of CF₃-Alkenes with Formic Acid, Chem. Commun., 2024, 60, 6548–6551; (f) R. Sun, G. Li, F. Xie, Q. Zhang, Y.-J. Sun and W. Dai, Exhaustive Hydrodefluorination or Deuterodefluorination of Trifluoromethylarenes via Metal-Free Photoredox Catalysis, Org. Lett., 2024, 26, 928–932; (g) Y. Chen and Y. Du, The Application of Deuteration Strategy in Drug Design, ChemMedChem, 2025, 20, e202400836.
7. (a) Y. Pang, M. Leutzsch, N. Nöthling, F. Katzenburg and J. Cornella, Catalytic Hydrodefluorination via Oxidative Addition, Ligand Metathesis, and Reductive Elimination at Bi(I)/Bi(III) Centers, J. Am. Chem. Soc., 2021, 143, 12487–12493; (b) J. J. Gair, R. L. Grey, S. Giroux and M. A. Brodney, Palladium Catalyzed Hydrodefluorination of Fluoro-(Hetero)Arenes, Org. Lett., 2019, 21, 2482–2487; (c) D. Breyer, T. Braun and P. Kläring, Synthesis and Reactivity of the Fluoro Complex Trans-[Pd(F)(4-C₅NF₄)(iPr₂PCH₂CH₂OCH₃)₂]: C–F Bond Formation and Catalytic C–F Bond Activation Reactions, Organometallics, 2012, 31, 1417–1424; (d) D. Breyer, T. Braun and A. Penner, Isolation and Reactivity of Palladium Hydrido Complexes: Intermediates in the Hydrodefluorination of Pentafluoropyridine, Dalton Trans., 2010, 39, 7513; (e) D. McKay, I. M. Riddlestone, S. A. Macgregor, M. F. Mahon and M. K. Whittlesey, Mechanistic Study of Ru-NHC-Catalyzed Hydrodefluorination of Fluoropyridines: the Influence of the NHC on the Regioselectivity of C–F Activation and Chemoselectivity of C–F Versus C–H Bond Cleavage, ACS Catal., 2015, 5, 776–787; (f) S. Bonfante, C. Lorber, J. M. Lynam, A. Simonneau and J. M. Slattery, Metallomimetic C–F Activation Catalysis by Simple Phosphines, J. Am. Chem. Soc., 2024, 146, 2005–2014; (g) K. Chulsky, I. Malahov, D. Bawari and R. Dobrovetsky, Metallomimetic Chemistry of a Cationic, Geometrically Constrained Phosphine in the Catalytic Hydrodefluorination and Amination of Ar–F Bonds, J. Am. Chem. Soc., 2023, 145, 3786–3794; (h) U. Jäger-Fiedler, M. Klahn, P. Arndt, W. Baumann, A. Spannenberg, V. V. Burlakov and U. Rosenthal, Room-temperature catalytic hydrodefluorination of pentafluoro-pyridine by zirconocene fluoro complexes and diisobutylaluminumhydride, J. Mol. Catal. A: Chem., 2007, 261, 184–189; (i) S. Yow, S. J. Gates, A. J. White and M. R. Crimmin, Zirconocene Dichloride Catalyzed Hydrodefluorination of C(sp²)-F Bonds, Angew. Chem., Int. Ed., 2012, 51, 12559–12563.
8. For the Co and Cu-catalyzed HDF of pentafluoropyridine at the 4-position, see: (a) J. Li, T. Zheng, H. Sun and X. Li, Selectively Catalytic Hydrodefluorination of Perfluoroarenes by Co(PMe₃)₄ with Sodium Formate as Reducing Agent and Mechanism Study, Dalton Trans., 2013, 42, 13048; (b) H. Lv, Y. Cai and J. Zhang, Copper-Catalyzed Hydrodefluorination of Fluoroarenes by Copper Hydride Intermediates, Angew. Chem., Int. Ed., 2013, 52, 3203–3207.
9. For a Ti catalyst efficient for a substituted 2,6-difluoropyridine motif in a ligand, see: (a) G. Podolan, D. Lentz and H. Reissig, Selective Catalytic Hydrodefluorination as a Key Step for the Synthesis of Hitherto Inaccessible Aminopyridine Derivatives, Angew. Chem., Int. Ed., 2013, 52, 9491–9494; (b) G. Podolan, P. Jungk, D. Lentz, R. Zimmer and H. Reissig, Studies on the Synthesis of Specifically Fluorinated 4-Amino- pyridine Derivatives by Regioselective Nucleophilic Aromatic Substitution and Catalytic Hydrodefluorination, Adv. Synth. Catal., 2015, 357, 3215–3228.
10. (a) L. Cronin, C. L. Higgitt, R. Karch and R. N. Perutz, Rapid Intermolecular Carbon−Fluorine Bond Activation of Pentafluoropyridine at Nickel(0): Comparative Reactivity of Fluorinated Arene and Fluorinated Pyridine Derivatives, Organometallics, 1997, 16, 4920–4928; (b) J. A. Hatnean and S. A. Johnson, Experimental Study of the Reaction of a Ni(PEt₃)₂ Synthon with Polyfluorinated Pyridines: Concerted, Phosphine-Assisted, or Radical C–F Bond Activation Mechanisms?, Organometallics, 2012, 31, 1361–1373; (c) S. Burling, P. I. Elliott, N. A. Jasim, R. J. Lindup, J. McKenna, R. N. Perutz, S. J. Archibald and A. C. Whitwood, C-F Bond Activation at Ni(0) and Simple Reactions of Square Planar Ni(II) Fluoride Complexes, Dalton Trans., 2005, 3686; (d) T. Braun, S. Parsons, R. N. Perutz and M. Voith, Reactivity of a Nickel Fluoride Complex: Preparation of New Tetrafluoropyridyl Derivatives, Organometallics, 1999, 18, 1710–1716; (e) M. I. Sladek, T. Braun, B. Neumann and H.-G. Stammler, Aromatic C–F activation at Ni in the presence of a carbon–chlorine bond: the nickel mediated synthesis of new pyrimidines, J. Chem. Soc., Dalton Trans., 2002, 3, 297–299.
11. T. Braun, R. N. Perutz and M. I. Sladek, Catalytic C–F Activation of Polyfluorinated Pyridines by Nickel-Mediated Cross-Coupling Reactions, Chem. Commun., 2001, 2254–2255.
12. For a recently published Zr complex capable of mediating the functionalization of 2,6-difluoro and 2,4,6-trifluoropyridine at the 2-position see: S. Bonfante, T. F. Tanner, C. Lorber, J. M. Lynam, A. Simonneau and J. M. Slattery, Zirconium-mediated carbon–fluorine bond functionalisation through cyclohexyne “umpolung”, Chem. Sci., 2025, 16, 3552–3559.
13. (a) J. Xiao, J. Wu, W. Zhao and S. Cao, NiCl₂(PCy₃)₂-Catalyzed Hydrodefluorination of Fluoroarenes with LiAl(O-T-Bu)₃H, J. Fluor. Chem., 2013, 146, 76–79; (b) P. Fischer, K. Götz, A. Eichhorn and U. Radius, Decisive Steps of the Hydrodefluorination of Fluoroaromatics Using [Ni(NHC)₂], Organometallics, 2012, 31, 1374–1383; (c) Y. He, Z. Chen, C. He and X. Zhang, Nickel-Catalyzed Ortho-Selective Hydrodefluorination of N-Containing Heterocycle-Polyfluoroarenes, Chin. J. Chem., 2013, 31, 873–877; (d) J. Zhou, M. W. Kuntze-Fechner, R. Bertermann, U. S. Paul, J. H. Berthel, A. Friedrich, Z. Du, T. B. Marder and U. Radius, Preparing (Multi)Fluoroarenes as Building Blocks for Synthesis: Nickel-Catalyzed Borylation of Polyfluoroarenes via C–F Bond Cleavage, J. Am. Chem. Soc., 2016, 138, 5250–5253; (e) Y.-M. Tian, X.-N. Guo, M. W. Kuntze-Fechner, I. Krummenacher, H. Braunschweig, U. Radius, A. Steffen and T. B. Marder, Selective Photocatalytic C–F Borylation of Polyfluoroarenes by Rh/Ni Dual Catalysis Providing Valuable Fluorinated Arylboronate Esters, J. Am. Chem. Soc., 2018, 140, 17612–17623; (f) A. Steffen, M. I. Sladek, T. Braun, B. Neumann and H.-G. Stammler, Catalytic C−C Coupling Reactions at Nickel by C−F Activation of a Pyrimidine in the Presence of a C−Cl Bond: the Crucial Role of Highly Reactive Fluoro Complexes, Organometallics, 2005, 24, 4057–4064; (g) X.-W. Liu, J. Echavarren, C. Zarate and R. Martin, Ni-Catalyzed Borylation of Aryl Fluorides via C–F Cleavage, J. Am. Chem. Soc., 2015, 137, 12470–12473; (h) A. Morishige, A. Matsuura and N. Chatani, Hydrodefluorination of ortho-Fluoro Aromatic Amides with 2-Propanol, Chem. Lett., 2023, 52, 63–66; (i) J. Liu and M. J. Robins, Fluoro, Alkylsulfanyl, and Alkylsulfonyl Leaving Groups in Suzuki Cross-Coupling Reactions of Purine 2′-Deoxynucleosides and Nucleosides, Org. Lett., 2005, 7, 1149–1151; (j) M. J. O'Neill, T. Riesebeck and J. Cornella, Thorpe–Ingold Effect in Branch-Selective Alkylation of Unactivated Aryl Fluorides, Angew. Chem., Int. Ed., 2018, 57, 9103–9107; (k) K. Wang and W. Kong, Synthesis of Fluorinated Compounds by Nickel-Catalyzed Defluorinative Cross-Coupling Reactions, ACS Catal., 2023, 13, 12238–12268.
14. For analogous Ni(II) complexes from C-F oxidative addition of pentafluoropyridine and tetrafluoropyridine see: D. Breyer, J. Berger, T. Braun and S. Mebs, Nickel fluoro complexes as intermediates in catalytic cross-coupling reactions, J. Fluor. Chem., 2012, 143, 263–271.
15. L. Kuehn, M. Stang, S. Würtemberger-Pietsch, A. Friedrich, H. Schneider, U. Radius and T. B. Marder, FBpin and its adducts and their role in catalytic borylations, Faraday Discuss., 2019, 220, 350–363.

---

Footnotes

† Electronic supplementary information (ESI) available: Complete experimental details, characterization data, NMR spectroscopic data (PDF). CCDC 2420312 (Ni5), 2420313 (Ni1) and 2429444 (Ni6). For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d5qi01257k

‡ NEt₃ has been previously employed as an additive in the HDF of 2,3,4,5,6-pentafluoropyridine with HBPin catalyzed by a Pd complex in ref. 7c, however, the role of the additive was not described.

§ The 1a-BFPin adduct has not been detected in the reaction medium by NMR spectroscopy. The formation of this adduct is presumably less favourable than that of 1b-BFPin as a result of the weaker Lewis basicity of 1a compared to 1b.

---