Harnessing Alkyl Amines as Electrophiles for Nickel-Catalyzed Cross Couplings via C–N Bond Activation

Abstract

We developed a strategy to harness alkyl amines as alkylating agents via C–N bond activation. This Suzuki–Miyaura cross coupling of alkyl pyridinium salts, readily formed from primary amines, is the first example of a metal-catalyzed cross coupling via C–N bond activation of an amine with an unactivated alkyl group. This reaction enjoys broad scope and functional group tolerance. Primary and secondary alkyl groups can be installed. Preliminary studies suggest a Ni^I/Ni^III catalytic cycle.

Graphical Abstract

Primary amines are prevalent across a wide range of molecules, from simple building blocks and synthetic intermediates, to biomolecules, drugs, and natural products (Scheme 1A).¹ The amino (NH₂) group is easily installed, is amenable to late-stage functionalization, and offers advantages such as purification via acid/base extraction. Although these benefits are well appreciated in the synthesis of nitrogen-containing products, alkyl amines have yet to be broadly recognized as alkylating agents. We envisioned that this underutilized reactivity of alkyl amines could be unlocked via metal-catalyzed C–N bond activation in a cross coupling. However, few cross couplings employ amine derivatives (Scheme 1B).² Cross couplings have been achieved via cleavage of various C_sp2–N bonds.³ For C_sp3–N bonds, electronically activated (benzylic and allylic) and strain-activated (aziridinyl) C–N bonds have been employed.⁴ However, there are no cross couplings of an alkyl amine derivative with an unactivated alkyl group.

Scheme 1.

Alkyl amines and their potential in the context of cross-coupling reactions

In contrast, intense efforts have identified other reagents to install alkyl groups lacking activation (Scheme 1C).⁵ Following pioneering developments with alkyl halides,⁶ pseudohalides,^{6n, 7} and organometallic nucleophiles,⁸ dual photoredox/nickel catalysis has enabled use of oxalates,⁹ carboxylic acids,¹⁰ 1,4-dihydropyridines,¹¹ organoboronates,¹² and organosilicates.¹³ Redox-active esters are also potent alkylating agents without need for a photocatalyst.¹⁴

Cross coupling an alkyl amine derivative would offer exciting complimentary opportunities in synthesis and late-stage functionalization. Our previous efforts toward C–N activation relied on benzylic trimethylammonium salts, in which chemoselectivity for the benzylic C–N bond is achieved via electronic activation.^4a–c Due to the similarity of the alkyl groups and diminished reactivity of non-benzylic C–N bonds, activation of nonbenzylic alkyl amines required a new strategy. Toward this goal, we were drawn to Katritzky pyridinium salts (3, Scheme 1D).^{3b, 15} These air- and moisture-stable solids are easily prepared in a single step via condensation of a primary amine with commercially available 2,4,6-triphenylpyrylium tetrafluoroborate.¹⁶ Unlike pyridinium cations lacking 2,6-substitution, which undergo addition to the pyridinium ring,¹⁷ these pyridinium salts have been employed as alkyl electrophiles in non-metal-catalyzed transformations via S_N2¹⁸ or radical mechanisms.^{15a, 19} We envisioned that they would also serve as alkyl electrophiles in metal-catalyzed cross couplings, particularly Suzuki–Miyaura reactions with commercially available and functional group tolerant aryl boronic acids.²⁰ Herein, we report the first example of a metal-catalyzed cross coupling via C–N bond activation of an amine derivative with an unactivated alkyl group. This reaction enjoys broad scope and functional group tolerance in the alkyl pyridinium and (hetero)aryl boronic acid. Both primary and secondary alkyl groups can be installed.

Optimization began with the cross coupling of pyridinium 3a and p-tolylboronic acid. Conditions similar to those for benzyl trimethylammonium triflates gave only 6% yield (Table 1, entry 1).^4a However, with bathophenanthroline (BPhen) as ligand, the yield increased (entry 2). KO^tBu as base also led to improvement (entry 3). The use of inexpensive, air- and moisture-stable Ni(OAc)₂·4H₂O gave increased yield, enabling set-up without an inert atmosphere glovebox (entry 4). Further improvement was realized by premixing Ni(OAc)₂·4H₂O and BPhen before the reaction (entry 5). Suspecting solubility was important, EtOH was added, giving 81% yield (entry 6). Not surprisingly, KOEt can be used (entry 7). Ni(OAc)₂·4H₂O, BPhen, and KO^tBu are necessary (entries 8, 9, and 11). Replacing BPhen with bipy decreases yield (entry 10), as does K₃PO₄ instead of KO^tBu (entry 12).

Table 1.

Optimization^a

entry	[Ni]	ligand	base	yield (%)b
1	Ni(cod)2	PPh2Cy	K3PO4	6
2	Ni(cod)2	BPhen	K3PO4	21
3	Ni(cod)2	BPhen	KOtBu	24
4	Ni(OAc)2·4H2O	BPhen	KOtBu	39
5c	Ni(OAc)2·4H2O	BPhen	KOtBu	52
6c,d	Ni(OAc)2·4H2O	BPhen	KOtBu	81
7c,d	Ni(OAc)2·4H2O	BPhen	KOEt	68
8c,d	–	BPhen	KOtBu	0
9c,d	Ni(OAc)2·4H2O	–	KOtBu	0
10c,d	Ni(OAc)2·4H2O	bipy	KOtBu	54
11c,d	Ni(OAc)2·4H2O	BPhen	–	0
12c,d	Ni(OAc)2·4H2O	BPhen	K3PO4	3

^aConditions: pyridinium salt 3a (0.1 mmol), [Ni] (10 mol %), ligand (24 mol %), p-TolB(OH)₂ (3.0 equiv), base (3.4 equiv), dioxane (0.1 M), 60 °C, 24 h.

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

^cTwo mixtures (Vial 1: [Ni], BPhen, dioxane. Vial 2: p-TolB(OH)₂, KO^tBu, EtOH, dioxane. 3a in either vial.) were stirred for 1 h before combining.

^dEtOH (5 equiv) added.

Under optimized conditions, we observed broad scope of alkyl pyridinium salts (Scheme 2). The reaction is somewhat tolerant of moisture and can be set up without oven-dried glassware, but low yield was observed when only minimal precautions were taken against air and moisture (6).¹⁶ Both primary and secondary (cyclic and acyclic) alkyl groups work. β-Substituted alkyl groups can be installed, including enantioenriched examples with β-stereocenters (17). Many functional groups, including ethers, silyl ethers, acetals, and esters, were well tolerated. Excitingly, unlike trimethylammonium substrates, this strategy is selective for primary amines; pyridinium formation and cross coupling does not affect tertiary or Boc-protected amines. Heterocycles, including piperidines, piperazines, and morpholine, can be used.

Scheme 2.

Reaction scope

^aConditions: pyridinium salt 3 (1.0 mmol), Ni(OAc)₂·4H₂O (10 mol %), BPhen (24 mol %), ArB(OH)₂ (3.0 equiv), KOt-Bu (3.4 equiv), EtOH (5 equiv), dioxane (0.1 M), 60 °C, 24 h. Average isolated yields (±6%) from duplicate experiments. ^bSingle experiment. ^cGlassware not oven-dried before use. ^dMinimal precautions to protect from air and moisture (see Supporting Information). ^e0.5-mmol scale. ^f0.05-mmol scale. p-(CF₃)C₆H₄OH (2 equiv) added. Yield determined by 1H NMR using 1,3,5-(OMe)₃C₆H₃ as internal standard. ^g12 mol % BPhen, dioxane (0.025 M).

Several products demonstrate the utility of this chemistry to create novel, potentially bioactive compounds from natural or synthetic molecules (Scheme 2). Products 16 and 17 are derived from proline and isoleucine, respectively. The synthesis of 17, which required only 4 steps from N-Boc isoleucine, is representative of the ease of synthesis enabled by this cross coupling. Cross coupling of the amino side chain of N-Boc lysine also proceeded in good yield, albeit poor conservation of ee (18). However, a much higher ee, but lower yield, was observed with the use of an acidic additive, suggesting conditions can be identified to solve this problem.²¹ Products 19 and 20 are derived from an amine intermediate in the synthesis of Mosapride, a treatment for gastrointestinal disorders.²² Product 29 derives from an amine intermediate used in the synthesis of Lipitor^®, an anti-cholesterol drug.²³

Broad scope was also achieved with the aryl boronic acid. Various functionalities were tolerated, including aryl chlorides (7) and fluorides (14, 16), methyl ketones (8, 12), esters (9), amides (10), ethers (11, 17), alkenes (13), silyl-protected alkynes (18), acetals (19), and nitriles (20). Given the prevalence of heterocycles in bioactive molecules, we investigated heteroarylboronic acids. N-Methyl indole was easily installed (15). Pyridyl boronic acids can also be used under slightly altered conditions (21–29). Both 3- and 4-pyridyl groups work, including those with fluoride, ether, and morpholino substituents. Notably, 2-fluoropyridines 21, 25, and 29 are primed for elaboration via S_NAr chemistry. Unsubstituted pyridyl was also successful (23).

This cross coupling could proceed via a Ni^0/II or Ni^I/III cycle. A Ni^0/II mechanism would involve two-electron oxidative addition (S_N1 or S_N2), whereas a Ni^I/III cycle would proceed via single-electron transfer (SET) from a Ni^I intermediate to the pyridinium. Although pyridinium salts undergo S_N2 reactions,¹⁸ they are also single-electron acceptors,^15a and have been exploited as photosensitizers²⁴ and sources of nitrogen radicals.²⁵ Also, cross coupling of 3p, prepared from(S)-2-aminooctane, resulted in racemic 30; cross coupling of cyclopropane 3q gave ring-opened 31; and addition of TEMPO provided trapped product 32 (Scheme 3). These results, the superiority of bipyridyl ligands, which are often employed in Ni^I/III catalysis,⁵ and the fact that Ni^II precursors outcompete Ni⁰ suggest a SET mechanism (Scheme 4). Similarly to redox-active esters, pyridinium 3 undergoes SET with a Ni(I) intermediate, triggering fragmentation to give alkyl radical B, which recombines with an arylnickel(II) intermediate to give Ni^III species C. Reductive elimination provides 4. Both monometallic “transmetallation-first” and radical chain bimetallic SET oxidative addition are known; we cannot currently distinguish these possibilities.^{6w, 26}

Scheme 3.

Mechanistic experiments

Scheme 4.

Mechanistic proposal

In sum, we developed a nickel-catalyzed cross coupling of alkyl pyridinium salts with aryl boronic acids. When combined with efficient formation of pyridinium salts from primary amines, this method enables transformation of primary alkyl amines to alkyl arenes. This reaction is the first cross coupling to install unactivated alkyl groups via C–N bond activation. Additional highlights include selectivity for primary amines, broad scope for primary and secondary alkyl groups, and wide tolerance of functional groups and heterocycles. Mechanistic experiments suggest a Ni^I/III cycle. Current efforts are underway to expand the scope and utility of this chemistry, which we hope will find broad use in synthesis and late-stage functionalization of alkyl amines.