INTRODUCTION

The emergence of a method for Csp2–N bond formation, proposed independently by Chan [1] and Lam [2] and based on the copper(II)-catalyzed reaction of amines with arylboronic acids, is an important achievement of modern chemistry. In the last decade, there has been an extensive development of this approach to N-(hetero)arylation of amines, as follows from the three reviews published in the recent few years [3–5]. The Chan–Lam coupling utilizes available copper(II) acetate or other copper(II) salts, weak bases, and cheap solvents, proceeds at room temperature, and requires neither inert atmosphere nor additional ligands. Owing to its simplicity, the Chan–Lam reac­tion have found wide application in organic synthesis, and efforts of many research teams resulted in finding conditions for effective synthesis of compounds with various Csp2–Nu bonds in the presence of copper catalysts. Despite the diversity of organoboron com­pounds and amines that can be involved in the Chan–Lam reactions, these reactions are most commonly performed between arylboronic acids or their analogues (arylboron pinacolates, arylboron difluorides) and primary aromatic amines, such as aniline and its derivatives and NH-heterocyclic amines. Accordingly, there are numerous procedures whose successful application depends on the reactant nature [6–8].

Unlike other methods of Csp2–N bond formation based on catalytic reaction between a nucleophile and an electrophile, both reactants in the Chan–Lam reac­tion are nucleophiles. Therefore, its mechanism still remains a matter of discussion. For example, Stahl et al. [9] presumed the formation of copper(III) com­pounds via disproportionation of organic copper salt, however without strong experimental proofs. Watson et al. [10] have made a significant progress in explaining the mechanism of this reaction. The authors proposed a catalytic cycle for the model reaction of [1,1′-biphenyl]-4-ylboronic acid with morpholine and piperidine in the presence of a stoichiometric amount of copper(II) acetate and triethylamine as a base (Scheme 1). However, even this version involves a questionable stage, namely the reaction of two copper(II) compounds with the formation of Cu(I) and Cu(III) derivatives.

Scheme
scheme 1

1.

Full size image

In addition, Schaper et al. [11, 12] largely con­tributed to the study of the reaction mechanism. Catalytic cycles with the participation of sulfonato diketimine and iminoarylsulfonate copper(II) com­plexes were proposed on the basis of the results of careful kinetic experiments. Since the amine nature can significantly influence the efficiency of the process, the simplicity of the method is thus faced not only with the complexity of studying and understanding the reaction mechanism but also with the strong dependence of the reaction outcome on the reactant nature, as well as on the nature of copper catalyst and its concentration, oxidant, solvent, base, and additives, the presence of water, and even temperature. The overwhelming majority of reported examples of the Chan–Lam reactions involved aromatic and heteroaromatic amines, as well as NH-heterocycles with a weakly basic nitrogen atom. The Chan–Lam reaction was used to obtain many drugs such as inhibitors of succinate–cytochrome C reductase [13] and retinoic acid 4-hydroxylase (CYP26) [14], β3-adrenergic receptor agonists [15], inhibitors of MEK kinase [16], and selective FPR agonists [17]. The product yields were sometimes quite low, down to a few percent, and the reactions were often carried out in the stoichiometric version [18]. Examples of the Chan–Lam reaction with primary aliphatic amines are less common, and these reactions were mainly used for the creation of drugs as well. For instance, Judd et al. [19] synthesized anti­cancer agents that are inhibitors of isoprenylcysteine carboxyl methyltransferase by the arylation of 3-amino­propyl substituent in tetrahydropyran [19]. Barlaam et al. [20] introduced a 3,4,5-trifluorophenyl substituent into a more sterically hindered amino group in functionally substituted chromen-4-one via the Chan–Lam reaction with the goal of obtaining an analogue of PI3Kβ/δ inhibitor. A series of BET bromodomain inhibitors were synthesized by the arylation of 4-aminotetrahydroquinoline derivatives with substituted phenylboronic acids [21]. In all these cases, stoichiometric amounts of copper(II) acetate (1–1.5 equiv) were used, and the reactions were carried out in dichloromethane or dichloroethane in the pres­ence of an organic base such as triethylamine, diiso­propylamine, and pyridine. It should be noted that the yields of the target products were generally quite modest (not exceeding 50%). These data indicate the necessity of further studying the Chan–Lam reaction with primary aliphatic amines.

The choice of adamantane-containing primary amines as subjects for our study was determined, on the one hand, by their high and diverse biological activity, especially of (hetero)aromatic compounds possessing an adamantane fragment [22–31]. On the other hand, it was possible to vary over a wide range the reactivity of the amino group by changing its steric environment, which is important for studying success­ful application of particular catalytic procedures. In this work, we studied the Chan–Lam reaction between p-tolylboronic acid and a series of adamantane-containing amines and diamines with different steric hindrances at the primary amino group.

RESULTS AND DISCUSSION

We have recently studied N,N′-diarylation of linear diamines and oxadiamines [32]. Optimization of the conditions showed that, when a catalytic amount of copper(II) acetate was used, the best solvent was acetonitrile (slightly worse results were obtained in DMF and DMSO with comparable dielectric con­stants), and the best base was 1,8-diazabicyclo[5.4.0]­undec-7-ene (DBU); triethylamine and diisopropyl­amine were less efficient. Based on these findings, we studied the conditions for the model reaction of 2-(adamantan-1-yl)ethanamine (1) with p-tolylboronic acid in the presence of copper(II) acetate (Scheme 2, Table 1). The reactions were carried out at room tempera­ture for 24 h in the presence of DBU or triethylamine in closed flasks with a sufficient volume (50 mL per 0.1 mmol of amine) to provide a required amount of atmospheric oxygen.

Scheme
scheme 2

2.

Full size image
Table 1. Reaction of 2-(adamantan-1-yl)ethanamine (1) with p-tolylboronic acid in the presence of Cu(OAc)2·H2O (20 mol %) in acetonitrile (c = 0.1 M) at 25°C
Full size table

It was found that in the presence of 20 mol % of Cu(OAc)2 at an amine concentration of 0.1 M the yield of 2 increased in parallel with the amounts of p-tolyl­boronic acid and DBU (Table 1; entry nos. 1–7). The maximum yield (64%) in the presence of 2 equiv of the base was achieved when 4 equiv of the acid was used (Table. 1, entry no. 5). The yield appreciably decreased when the reaction mixture was diluted to an amine concentration of 0.05 M (Table 1, entry no. 4). By raising the amount of DBU to 4 equiv we succeeded in improving the yield of the arylation product to 75% in the presence of 1.2 equiv of p-tolylboronic acid (Table 1, entry no. 6) and to 86% with 2 equiv of the acid (Table 1, run no. 7). When triethylamine was used as the base, the best yield (65%) was achieved with 4 equiv of the acid and 2 equiv of triethylamine (Table 1, entry no. 10); this result is comparable with that obtained using DBU (Table. 1, entry no. 5). Increase of the amount of triethylamine to 4 equiv in the reaction with 4 equiv of p-tolylboronic acid did not improve the yield of 2 (Table 1, entry no. 11).

Our recent study of copper-catalyzed amination of (hetero)aryl halides demonstrated surprisingly high efficiency of free copper nanoparticles (CuNPs) with a sufficiently large size [33–35]. Therefore, copper nanoparticles of different average sizes (25, 70, and 85 nm), as well as powdered copper(I) and copper(II) oxides, were tested as catalysts in the Chan–Lam reaction. We found that 25-nm copper nanoparticles provided a 59% yield of 2 using 2 equiv of p-tolyl­boronic acid and 4 equiv DBU (Table 2, entry no. 1) and that the yield decreased at other acid-to-base ratios (Table 2; entry nos. 2, 3); no reaction was observed in the presence of triethylamine (Table 2, entry no. 4). Larger copper nanoparticles are also capable of cata­lyzing the reaction (Table 2; entry nos. 5, 6). Obviously, the possibility to catalyze the Chan–Lam reaction by copper(0) nanoparticles is related to their fairly easy oxidation in solution with atmospheric oxygen in the presence of a base, though solid CuNPs are much more stable to oxidation in air. Presumably, DBU acts as a ligand toward dissolved Cu(II) particles, which facilitates the transition of CuNPs to solution. The yields of 2 significantly decreased in the presence of powdered copper(I) and copper(II) oxides (Table 2; entry nos. 7, 8), which may be due to their poor solubility in comparison with CuNPs under the given conditions. In many cases, the reaction was accom­panied by side formation of p-cresol (oxidation product of p-tolylboronic acid), which can be separated by chromatography.

Table 2. Reaction of 2-(adamantan-1-yl)ethanamine (1) with p-tolylboronic acid in the presence of copper compounds (20 mol %) in acetonitrile (c = 0.1 M) at 25°C
Full size table

Adamantane-containing amines 3–9 with different steric hindrances at the amino group were subjected to arylation with p-tolylboronic acid (Scheme 3, Table 3) under the conditions corresponding to the formation of 60% of compound 2 (2 equiv of p-tolylboronic acid and 2 equiv of DBU). The results obtained with amines 3–6 were fairly good, and the yields of arylation products 12–15 were 59–74% (Table 3, entry nos. 1–4) which, on the whole, were even slightly better than in the reaction with amine 1. The yield of 16 from adamantan-2-amine (7), in which the amino group is directly linked to the adamantane skeleton, was lower (Table 3, entry no. 5). The yields of the arylation products further decreased with increasing steric hindrances at the amino group in compounds 8 and 9 (Table 3, run nos. 6–9); in these cases, the use of 4 equiv of p-tolylboronic acid (against standard 2 equiv) did not improve the yield of 17 and 18 (Table 3; entry nos. 7, 9).

Scheme
scheme 3

3.

Full size image
Table 3. Reaction of adamantane-containing amines 3–9 and diamines 10 and 11 with p-tolylboronic acid in the presence of Cu(OAc)2·H2O (20 mol %) in acetonitrile (c = 0.1 M) at 25°C
Full size table

The arylation of adamantane-containing diamines 10 and 11, in which the amino groups are separated from the bridgehead positions of the adamantane core by spacers of different length, was carried out using 4 equiv of p-tolylboronic acid and 2.5 equiv of DBU. In the presence of 20 mol % of copper(II) acetate, the yields of N,N′-diarylation products 19 and 20 were 66 and 42%, respectively (Table 3; entry nos. 10, 12). However, an increase in the catalyst loading to 40 mol % resulted in reduced yields of the products, 49 and 39%, respectively (Table 3; entry nos. 11, 13). These findings correlate with the previously observed dependence of the yield of N,N′-diarylation products in the Chan–Lam reactions of diamines and oxadiamines on the amount of copper(II) acetate [32].

Since we failed to obtain compounds 17 and 18 in acceptable yields by the Chan–Lam reaction, the corresponding sterically hindered amines 8 and 9 were subjected to Pd(0)-catalyzed arylation with p-bromo­toluene (Scheme 4). For comparison, similar reactions were performed with some other amines in which the amino group is sterically more accessible, namely compounds 4–7. The reactions were carried out with equimolar amounts of the reactants in the presence of the standard catalytic system Pd(dba)2/BINAP [4/4.5 mol %; dba = dibenzylideneacetone, BINAP = 2,2′-bis(diphenylphosphino)-1,1′-binaphthalene] and sodium tert-butoxide as a base in boiling dioxane.

Scheme
scheme 4

4.

Full size image

In almost all cases, the yields ranged from 50 to 65%, and only compound 16 was obtained in 85% yield; in the latter case, the reaction was performed with amine 7 hydrochloride using a double amount of sodium tert-butoxide. The moderate yields in the reactions with sterically unhindered amines 4–6 can be attributed to the side N,N-diarylation of the primary amino group, which was not observed in the copper-catalyzed reactions where the yields of the corre­sponding products 13–15 were higher. However, the yield of N,N-diarylation products did not exceed 5–9%. On the other hand, no N,N-diarylation was observed in the reactions with sterically hindered amines 8 and 9, and insufficiently high yields were associated only with the lower reactivity of these compounds.

EXPERIMENTAL

The 1H and 13C NMR spectra were recorded on a Bruker Avance-400 spectrometer at 400 and 100.6 MHz, respectively, using CDCl3 as solvent; the chemical shifts were measured relative to the residual proton and carbon signals of the solvent (δ 7.25, δC 77.00 ppm). The mass spectra (MALDI-TOF, positive ion detection) were recorded on a Bruker Daltonics Autoflex II mass spectrometer using 1,8,9-tri­hydroxyanthracene as matrix and poly(ethylene glycols) PEG-200 and PEG-300 as calibration standards. Silica gel (Merck, 40/60 μ) was used for preparative column chromatography. Commercially available p-tolylboronic acid (1), adamantan-2-amine (7, hydrochloride), triethylamine, 1,8-diazabicyclo­[5.4.0]­undec-7-ene (DBU), copper(II) acetate mono­hydrate, copper nanoparticles, copper(I) oxide, copper(II) oxide, rac-2,2′-bis(diphenylphosphino)-1,1′-binaphthalene (BINAP), and sodium tert-butoxide were used without further purification. Dioxane was dried by successive distillation over alkali and metallic sodium. Adamantane-containing amines 1, 3–6, 8, and 9 and diamines 10 and 11 were synthesized according to the reported procedures [36–41], and Pd(dba)2 was prepared as described in [42]. The spectral charac­teristics of compounds 2 [43], 12 [34], 19, and 20 [44] were reported previously.

N-Arylation of adamantane-containing amines 3–9 and diamines 10 and 11 (general procedures). a. Chan–Lam arylation. A 50–100-mL one-neck flask equipped with a magnetic stirrer was charged with 1 mL of acetonitrile, of amine 1 or 3–9 or diamine 10 or 11 (0.1 mmol), p-tolylboronic acid [0.2 mmol (27 mg) in the reactions with monoamines or 0.4 mmol (54 mg) in the reactions with diamines], Cu(OAc)2·H2O (4 mg, 20 mol %), and DBU [0.2 mmol (30 mg) or 0.25 mmol (38 mg) for mono- and diamines, respec­tively]. The flask was capped, and the mixture was stirred at room temperature for 24 h. A 30-μL sample of the mixture was taken, transferred into an NMR ampule, and dissolved in 0.6 mL of CDCl3, and the 1H NMR spectrum was recorded to confirm the reaction completeness. To isolate the products, the precipitate was separated and washed with dichloro­methane (5 mL), the combined organic fractions were evaporated under reduced pressure, and the residue was subjected to silica gel chromatography using successively dichloromethane–petroleum ether (1:10 to 1:1) and dichloromethane as eluents.

b. Palladium-catalyzed amination. A screw-capped vial containing a magnetic stir bar was charged with 0.2 mmol (34 mg) of p-bromotoluene, 4.5 mg (4 mol %) of Pd(dba)2, 5.5 mg (4.5 mol %) of BINAP, 1 mL of anhydrous dioxane, 0.2 mmol of amine 4–9, and 0.3 mmol (29 mg) of sodium tert-butoxide. The mixture was heated with stirring for 8 h, and the products were isolated as described above in a.

N-(Adamantan-1-ylmethyl)-4-methylaniline (13) was synthesized from 16.5 mg (0.1 mol) (a) or 33 mg (0.2 mmol) (b) of amine 4. Yield 19 mg (74%, a), 28 mg (54%, b). 1H NMR spectrum, δ, ppm: 1.57– 1.58 m (6H, CH2Ad), 1.64–1.79 m (6H, CH2Ad), 1.99 br.s (3H, CHAd), 2.22 s (3H, CH3), 2.77 s (2H, CH2N), 3.52 br.s (1H, NH), 6.53–6.55 m (2H, o-H), 6.95–6.97 m (2H, m-H). 13C NMR spectrum, δC, ppm: 20.3 (CH3), 28.3 (3C, CHAd), 33.9 (CAd), 37.1 (3C, CH2Ad), 40.7 (3C, CH2Ad), 56.6 (CH2N), 112.7 (2C, Co), 125.8 (Cp), 129.6 (2C, Cm), 147.0 (2C, Ci). Mass spectrum (MALDI-TOF): m/z 256.203 [M + H]+. C18H26N. Calculated: M + H 256.207.

N-[2-(Adamantan-2-yl)propyl]-4-methylaniline (14) was synthesized from 19 mg (0.1 mol) (a) or 39 mg (0.2 mmol) (b) of amine 5. Yield 19 mg (66%, a), 34 mg (60%, b). 1H NMR spectrum, δ, ppm: 0.96 d (3H, CH3CH, 3J = 6.6 Hz), 1.49–1.53 m (2H, HAd), 1.68–1.95 m (14H, HAd, CH3CH), 2.23 s (3H, 4-CH3), 2.76 d.d (1H, CH2N, 2J = 12.2, 3J = 8.2 Hz), 3.27 d.d (1H, CH2N, 2J = 12.2, 3J = 3.2 Hz), 6.52– 6.54 m (2H, o-H), 6.97–6.99 m (2H, m-H); the NH signal was not identified. 13C NMR spectrum, δC, ppm: 16.1 (CH3CH), 20.3 (4-CH3), 27.7 (CHAd), 28.8 (CHAd), 29.1 (CHAd), 29.3 (CHAd), 31.7 (CH2Ad), 32.0 (CH2Ad), 32.2 (CHCH3), 38.2 (CH2Ad), 39.2 (CH2Ad), 39.3 (CH2Ad), 47.6 (CHAd), 48.3 (CH2N), 112.7 (2C, Co), 126.0 (Cp), 129.7 (2C, Cm), 146.5 (Ci). Mass spectrum (MALDI-TOF): m/z 284.240 [M + H]+. C20H30N. Calculated: M + H 284.237.

N-[2-(Adamantan-2-yl)butyl]-4-methylaniline (15) was synthesized from 21 mg (0.1 mol) (a) or 41 mg (0.2 mmol) (b) of amine 6. Yield 16 mg (59%, a), 30 mg (50%, b). 1H NMR spectrum, δ, ppm: 0.86 t (3H, CH3CH2, 3J = 7.5 Hz), 1.32–1.40 m (1H, CH2CH3), 1.50–1.60 m (3H, HAd, CH2CH3), 1.68– 1.97 m (16H, HAd), 2.23 s (3H, 4-CH3), 2.96 d.d (1H, CH2N, 2J = 12.2, 3J = 6.7 Hz), 3.17 d.d (1H, CH2N, 2J = 12.2, 3J = 3.6 Hz), 6.53–6.55 m (2H, o-H), 6.97–6.99 m (2H, m-H); the NH signal was not identified. 13C NMR spectrum, δC, ppm: 9.8 (CH3CH2), 20.3 (CH3CH2), 20.6 (4-CH3), 27.7 (CHAd), 28.0 (CHAd), 28.8 (CHAd), 29.8 (CHAd), 31.8 (CH2Ad), 32.1 (CH2Ad), 36.9 (AdCH), 38.2 (CH2Ad), 39.3 (2C, CH2Ad), 43.0 (CHAd), 44.5 (CH2N), 112.7 (2C, Co), 126.0 (Cp), 129.7 (2C, Cm), 146.5 (Ci). Mass spectrum (MALDI-TOF): m/z 298.248 [M + H]+. C21H32N. Calculated: M + H 298.253.

N-(4-Methylphenyl)adamantan-2-amine (16) was synthesized from 15 mg (0.1 mol) of amine 7 (a) or 38 mg (0.2 mmol) of amine 7 hydrochloride (b). Yield 12 mg (49%, a), 41 mg (50%, b). 1H NMR spectrum, δ, ppm: 1.56–1.59 m (2H, HAd), 1.74 br.s (2H, HAd), 1.78–1.93 m (8H, HAd), 2.01 br.s (2H, HAd), 2.22 s (3H, 4-CH3), 3.51 br.s (1H, NH), 6.52–6.54 m (2H, o-H), 6.96–6.98 m (2H, m-H). 13C NMR spectrum, δC, ppm: 20.3 (CH3), 27.3 (CHAd), 27.5 (CHAd), 31.5 (4C, CH2Ad), 37.4 (2C, CHAd), 37.7 (CH2Ad), 57.0 (CHN), 113.3 (2C, Co), 125.9 (Cp), 129.7 (2C, Cm), 145.1 (Ci). Mass spectrum (MALDI-TOF): m/z 242.188 [M + H]+. C17H24N. Calculated: M + H 242.191.

N-[1-(Adamantan-1-yl)ethyl]-4-methylaniline (17) was synthesized from 18 mg (0.1 mol) (a) or 36 mg (0.2 mmol) (b) of amine 8. Yield 9.5 mg (35%, a), 35 mg (65%, b). 1H NMR spectrum, δ, ppm: 1.04 d (3H, CH3CH, 3J = 6.6 Hz), 1.51–1.54 m (3H, CH2Ad), 1.63–1.76 m (9H, CH2Ad), 1.99 br.s (2H, CHAd), 2.22 s (3H, 4-CH3), 3.00 q (1H, CHN, 3J = 6.7 Hz), 6.50–6.52 m (2H, o-H), 6.94–6.96 m (2H, m-H); the NH signal was not identified. 13C NMR spectrum, δC, ppm: 14.4 (CH3CH), 20.3 (4-CH3), 28.5 (3C, CHAd), 36.5 (CAd), 37.2 (3C, CH2Ad), 38.8 (3C, CH2Ad), 57.9 (CHN), 113.1 (2C, Co), 125.5 (Cp), 129.7 (2C, Cm), 146.4 (Ci). Mass spectrum (MALDI-TOF): m/z 270.218 [M + H]+. C19H28N. Calculated: M + H 270.222.

N-[(Adamantan-1-yl)(phenyl)methyl]-4-methyl­aniline (18) was synthesized from 24 mg (0.1 mol) (a) or 48 mg (0.2 mmol) (b) of amine 9. Yield 6 mg (19%, a), 37 mg (56%, b). 1H NMR spectrum, δ, ppm: 1.46–1.51 m (3H, CH2Ad), 1.57–1.90 m (9H, CH2Ad), 1.98 br.s (2H, CHAd), 2.14 s (3H, CH3), 3.84 s (1H, NH), 6.40–6.42 m (2H, o-H), 6.84–6.86 m (2H, m-H), 7.18–7.34 m (5H, Ph); the NH signal was not iden­tified. 13C NMR spectrum, δC, ppm: 20.3 (CH3), 28.4 (3C, CHAd), 36.9 (3C, CH2Ad), 39.2 (3C, CH2Ad), 68.2 (CHN), 113.1 (2C, Co), 125.8 (Cp), 126.6 (C4), 127.5 (2C, CHPh), 128.6 (2C, CHPh), 129.5 (2C, Cm), 140.4 (C1), 145.6 (C1); quaternary C(Ad) signal was not identified. Mass spectrum (MALDI-TOF): m/z 332.240 [M + H]+. C24H30N. Calculated: M + H 332.238.

CONCLUSIONS

The Chan–Lam reaction can be used to obtain N-aryl derivatives of adamantane-containing amines with a sterically unhindered amino group. The optimal conditions include 2 equiv of p-tolylboronic acid, 2 equiv of DBU, 0.1 M amine solution in acetonitrile, and 20 mol % of copper(II) acetate; under these con­ditions, the yields of the target products range up to 74% from monoamines and 66% from diamines. The possibility of using medium-size free copper nano­particles to catalyze the Chan–Lam reaction has been demonstrated. In the case of sterically hindered ada­mantane-containing amines, such as 8 and 9, Pd(0)-catalyzed amination provides better results, whereas copper- and palladium-catalyzed reactions with the other amines give comparable results.