Deaminative Reductive Methylation of Alkylpyridinium Salts

Abstract

Methyl groups can imbue valuable properties in organic molecules, often leading to enhanced bioactivity. To enable efficient installation of methyl groups on simple building blocks and in late-stage functionalization, a nickel-catalyzed reductive coupling of secondary Katritzky alkylpyridinium salts with methyl iodide was developed. When coupled with formation of the pyridinium salt from an alkyl amine, this method allows amino groups to be readily transformed into methyl groups with broad functional group and heterocycle tolerance.

Graphical Abstract

Even though methyl groups are the simplest alkyl group, they provide important structural effects in organic molecules.¹ Methyl groups often influence the properties of bioactive compounds by increasing their selectivity, binding affinity, metabolic stability, or lipophilicity,² leading to the so-called “the magic methyl effect.”¹ Due to their importance, methods to install a methyl group on an sp² carbon have been widely developed,³ but mild and general methods to forge a C(sp³)–methyl bond are not as well known.^{3e, 4} Classically, these methylations have involved the addition of strong methyl nucleophiles to alkyl halides or pseudohalides, but these methyl nucleophiles often present functional group limitations.⁵ To overcome these restrictions, recent efforts have focused on reductive cross-electrophile couplings with methyl electrophiles (Scheme 1B). Both photoredox and nickel catalysis have been used to cross-couple alkyl bromides with methyl tosylate.^{3e, 4}

Scheme 1.

C(sp³) Methylation

We envisioned that alkylpyridinium salts could also be attractive substrates for a reductive methylation (Scheme 1C). When coupled with efficient formation of the Katritzky pyridinium salt from an alkyl amine (1),⁶ this would allow amino groups to be readily transformed into methyl groups. In addition to providing a new, non-canonical, functional group transformation, this advance would be advantageous because alkyl amines are widely available, easily prepared via various methods, and can be carried through multiple synthetic steps with suitable protection.⁷ They are also abundant in a multitude of bioactive and natural products, offering opportunities for late-stage functionalization.⁸ Because of their abundance and diversity, we and others have developed deaminative methods for a variety of reactions including arylations,⁹ vinylations,^{9d, 10} alkylations,^{9g, 9j, 11} borylations,¹² and others.¹³ Despite the wide breadth of groups that can now be installed in these deaminative reactions, methods to simply convert amino groups into methyl substituents remain limited. In fact, the only deaminative methylation of alkylpyridinium salts is our Negishi cross-coupling, which requires basic conditions. The exceptional functional group tolerance seen in our and others’ reductive arylation^{9g–i, 14} and alkylation^{9g, 9j, 11f, 11g} methods encouraged us to develop a reductive methylation in order to avoid the harsh conditions of the Negishi coupling. Herein, we report the conditions for the nickel-catalyzed, reductive coupling of alkylpyridinium salts with methyl iodide (Scheme 1C). Using Mn⁰ as reductant,¹⁵ this reaction proceeds under remarkably straightforward and mild conditions, even in comparison to other reductive methylations, which typically require a base (see Scheme 1B).

We selected the reductive coupling of pyridinium salt 2a and methyl iodide to develop this method. We began by investigating the catalyst systems that had been used in our Negishi alkylation;^11b 4,4’,4”-tri-tert-butyl-2,2’:6’,2”-terpyridine (ttbtpy) had proven optimal for primary alkylpyridinium salts, and 2,6-bis(pyrazol-1-yl)pyridine (1-bpp) for secondary alkylpyridinium salts. Using ttbtpy as ligand, we observed a promising 37% yield (Table 1, entry 1). The use of 1-bpp increased the yield to 61% (entry 2). The use of bidentate ligands, such as 4,4’-di-tert-butyl-2,2’-bipyridine, resulted in lower yields (see Supporting Information), similar to what we have observed in other alkyl–alkyl cross-couplings.^{11a, 11b} With 1-bpp, a further increase in yield to 67% was achieved with the use of NiBr₂·DME (entry 3). In these reactions, we observe high conversion to triphenylpyridine, consistent with efficient C–N bond cleavage. The major observed by-product is hydrodeamination (or reduction) of pyridinium salt 2a (see Supporting Information). Decreasing the reaction temperature to room temperature resulted in 75% yield (entry 4). Utilizing the simple one-component catalyst, (1-bpp)NiBr₂, gave similar results to the two-component catalyst system (entry 5). Control experiments showed that both nickel and reductant are needed (entries 6 and 7). Somewhat surprisingly, Zn was ineffective as reductant at room temperature (entry 8), could be used at 80 °C (entry 9). Tetrakis(dimethylamino)ethylene (TDAE) was also effective (entry 10), suggesting that the reaction does not proceed via an organomanganese methyl or alkyl intermediate. Other methylating agents provided low yields (14% with MeOTf, 12% with MeOTs).

Table 1.

Reaction Optimization^a

entry	[Ni]	ligand	reductant	yield (%)b
1c	NiCl2·DME	ttbtpy	Mn	37
2	NiCl2·DME	1-bpp	Mn	61
3	NiBr2·DME	1-bpp	Mn	67
4d	NiBr2·DME	1-bpp	Mn	75
5d	(1-bpp)NiBr2	None	Mn	75
6d	None	None	Mn	0
7d	NiBr2·DME	1-bpp	None	0
8d,e	NiBr2·DME	1-bpp	Zn	16
9f	NiCl2·DME	1-bpp	Zn	63
10d,e	NiBr2·DME	1-bpp	TDAE	63

^aConditions: alkylpyridinium salt 2a (0.10 mmol), [Ni] (10 mol %), ligand (12 mol %), reductant (2.0 equiv), CH₃I (1.2 equiv), DMA (0.2 M), 80 °C, 16–22 h, unless otherwise noted.

^bDetermined by ¹H NMR spectroscopic analysis using 1,3,5-trimethoxybenzene as internal standard.

^cCH₃I (1.0 equiv), pyridinium salt 2a (1.2 equiv), NBu₄I (2.0 equiv).

^dRoom temperature.

^eDMF (0.2 M).

^fCH₃I (1.0 equiv), pyridinium salt 2a (1.2 equiv), NBu₄I (1.0 equiv).

Under these optimized conditions (Table 1, entry 5), both cyclic and acyclic secondary alkylpyridinium salts underwent methylation in high yield (Scheme 2). A variety of saturated and aromatic heterocycles were tolerated, including piperidine (4–6), pyrimidine (4, 5), pyrazole (7), pyrone (12), indole (18), and oxepane (19). The methylation was also amenable to the use of diverse functional groups, such as trifluoromethyls (4), carbamates (6, 8–9), ethers (10, 14–18), esters (11–14, 18–20), ketones (13, 19–20), and aryl chlorides (18). Protic functional groups, such as secondary carbamates, can also be used (8–9). The advantage of using a reductive coupling approach is particularly clear with arylboronate ester 5, whose BPin handle can be used for further functionalization. For alkylpyridinium substrates containing an additional stereocenter, we observed up to 5:1 dr in the methylation of cyclohexylpyridinium salts (9), but this is dependent on both the position and identity of the substituent (see 8 and 10). With respect to limitations in substrate scope, primary alkylpyridinium salts resulted in low yield, likely due to the reactivity of the methyl iodide outcompeting the slower reacting primary alkylpyridinium substrate. We also observed low yields with pyridinium salts derived from 2-aminopyrrolidines, 2-aminoazetidines, and α-amino acids (see Supporting Information).

Scheme 2.

Scope of Methylation^a

^a Conditions: alkylpyridinium salt (1.0 mmol, 1.0 equiv), NiBr₂(1-bpp) (10 mol %), Mn (2.0 equiv), CH₃I (1.2 equiv), DMA (5 mL), rt, 1 h. Average yield of duplicate experiments (±8%), unless noted otherwise. ^b Single experiment. ^c 24 h. ^d Diastereomeric ratio (dr) of crude reaction mixture determined by ¹H NMR. ^e Glassware was not dried. ^f Minimal precautions to protect from air and moisture. See Supporting Information.

In addition to these studies of functional group and heterocycle tolerance, this method is applicable to pyridinium salts derived from pharmaceuticals. Methylation of the pyridinium salt of the anti-arrhythmic mexiletine proceeds in 80% yield (15). When this reaction is run in glassware that has not been oven-dried or without precaution against moisture or air, product is still observed, albeit in reduced yield. Notably, this reaction can also be used to install isotopically labelled methyl groups (16, 17), offering possibilities for application in isotope effect studies for mechanistic analysis and the preparation of deuterated drug molecules.¹⁶ To future demonstrate the use of this method on molecules of the complexity found in pharmaceuticals, we prepared pyridinium derivatives of the anti-inflammatory carboxylic acids indomethacin, isoxepac, and ketoprofen via esterification with alcohol 2p (Scheme 3).^11e Subsequent methylation of these substrates proceeded in 45–69% yield (18–20). These examples highlight the utility of this method in late-stage functionalization of advanced intermediates. The non-canonical transformation of amino groups (NH₂) into methyl (CH₃) represents an intriguing possibility in structure-activity relationship (SAR) investigation.

Scheme 3.

Late-stage Methylation of Pharmaceutically Relevant Compounds

We hypothesize that this reaction proceeds via similar mechanistic steps to those proposed in previous reductive couplings of alkylpyridinium salts, in which the alkylpyridinium undergoes single electron transfer (SET) from a Ni(I) intermediate or Mn to generate a neutral dihydropyridyl radical.^{9f, 9g, 9i, 14, 17} This dihydropyridyl radical then fragments to give triphenylpyridine and an alkyl radical, which can recombine with a Ni(II) intermediate and proceed to product. Consistent with this mechanistic framework, we observed TEMPO-trapped adduct 21 when TEMPO is added to the methylation conditions, providing support that an alkyl radical is generated from the pyridinium salt (Scheme 4). It is not clear if the methyl iodide undergoes activation via SET or a two-electron process, although it should be noted that the methyl-TEMPO adduct was not observed. In addition, the use of an organic reductant (TDAE) resulted in 63% yield of the desired methylated product (4), eliminating the likelihood of a methylmanganese intermediate. Further studies are needed to elucidate additional mechanistic details.

Scheme 4.

Mechanistic Experiments

To conclude, we have developed a nickel-catalyzed methylation of alkylpyridinium salts using methyliodide. This method utilizes a single-component nickel catalyst to efficiently transform amino groups to methyls via Katritzky pyridinium salt intermediates. This transformation boasts broad functional group and heterocycle tolerance on the secondary alkylpyridinium intermediates.