Palladium-Catalyzed Aminocarbonylation of Benzyl Chlorides using Ammonia

Palladium-catalyzed coupling reactions have become one of the most popular methodologies for CC bond formation in organic synthesis.1, 2 In this area, both academic and industrial chemists constantly develop new applications and improve known procedures. Among the different coupling reactions, also palladium-catalyzed carbonylation reactions of aryl (pseudo)halides have experienced impressive improvements during the last three decades.3 On the other hand, related palladium-catalyzed carbonylation of allyl or benzyl halides are rarely reported.4, 5 One of the more recent examples constitutes the work of Troisi and coworkers, who disclosed the synthesis of secondary amides through the palladium-catalyzed aminocarbonylation of benzyl halides and amines.5d In the last two years, our group also developed the carbonylative coupling of benzyl chlorides with aryl boronic acids and aryl trifluoroborates to give 1,2-diarylethanones in moderate to good yields.5h, i

Primary benzyl amides represent interesting pharmaceutical compounds.6 Traditionally, these compounds are synthesized by hydration of the corresponding benzyl cyanides7 or condensation of 2-arylacetic acids with ammonia.8 In addition, carbonylation of inexpensive benzyl chlorides with ammonia offers a straightforward access to this class of compounds. Clearly, ammonia is the most inexpensive source for the preparation of primary amides. It avoids the use of less atom-economical ammonia synthons, such as hexamethyldisilazane (HMDS), N-tert-butylamides, and hydroxylamine.9 However, a potential problem of this transformation is the direct reaction of ammonia with benzyl chloride to give benzylamines. Hence, to the best of our knowledge, the carbonylation of benzyl halides to primary 2-arylacetamides has not been described.

More recently, we reported the aminocarbonylation of aryl halides using gaseous ammonia10 to synthesize various primary benzamides in high yields. However, by using benzyl chloride as substrate, only low yields (<10 %) of the desired product were obtained. Taking the advantages of ammonia11 and the importance of primary benzyl amides6 into account, we wish to report herein a general and efficient methodology for the synthesis of primary benzyl amides through a palladium-catalyzed aminocarbonylation of benzyl chlorides using ammonia as the amine source.

Based on our experience with palladium-catalyzed carbonylation reactions,12 we initially investigated the aminocarbonylation of benzyl chloride in the presence of 1 mol % [(cinnamyl)PdCl]₂ and varying ligands. As shown in Table 1, the model reaction was performed at low pressure [2 bar (1 bar=0.1 MPa) NH₃ and 2 bar CO] in dioxane at 100 °C.

Table 1. Aminocarbonylation of benzyl chloride: Testing of ligands.^[a]

Entry	Ligand	Concentration [mol %]	Conversion^[b] [%]	Yield^[b] [%]
[a] Reaction conditions: Benzyl chloride (1 mmol), CO (2 bar), NH3 (2 bar), [(cinnamyl)PdCl]2 (0.01 mmol), ligand, dioxane (2 mL), 100 °C, 16 h. [b] Conversion and yield were determined by using GC; dodecane was used as internal standard. [c] DPPP: 1,3-bis(diphenylphosphino)propane. [d] Xantphos: 4,5-bis(diphenylphosphino)-9,9-dimethylxanthene. [e] TFP: tri(2-furyl)phosphine.
1	DPPP^[c]	2	100	74
2	DPPB	2	100	80
3	DPPF	2	100	64
4	DtBPF	2	83	42
5	Xantphos^[d]	2	100	67
6	DPEphos	2	100	79
7	PCy3	4	100	29
8	P(tBu)3	4	10	0
9	TFP^[e]	4	67	24
10	P(o-tolyl)3	4	15	0
11	BuPAd2	4	100	63
12	PPh3	4	100	79

All bidentate ligands tested gave the primary amide in moderate to good yields (42–80 %; Table 1, entries 1–6). The best result was obtained by using 1,4-bis(diphenylphosphino)-butane (DPPB, 80 %; Table 1, entry 2). Interestingly, 1,1′-bis(di-tert-butylphosphino)ferrocene (DtBPF), a more electron-rich ligand compared to 1,1′-bis(diphenylphosphino)ferrocene (DPPF), gave a lower yield and conversion (Table 1, entries 3 and 4). Furthermore, six monophosphine ligands were tested. Tricyclohexylphosphine (PCy₃) resulted in 29 % yield of 2-phenylacetamide, whereas tri-tert-butylphosphine (P(tBu)₃) gave no desired product at all (Table 1, entries 7 and 8). On the other hand, di(1-adamantyl)-n-butylphosphine (BuPAd₂), a versatile ligand in palladium-catalyzed coupling reactions,13 led to 63 % of the desired product with total conversion of benzyl chloride (Table 1, entry 11). Fortunately, the most simple and inexpensive phosphine ligand PPh₃ resulted in 79 % yield of the primary amide (Table 1, entry 12). Considering the price of DPPB, 1,1′-[(oxydi-2,1-phenylene)]bis[1,1-di-phenylphosphine] (DPEphos), and PPh₃, we decided to use PPh₃ for further optimization. In this coupling reaction, ammonia is not only the amination reagent, but also acts as base. Moreover, benzylamine, a reasonable by-product, was not detected during the optimization processes.

Next, different palladium precursors were tested (Table 2). Pd(OAc)₂ gave 89 % yield with 100 % selectivity (Table 2, entry 1). Fortunately, palladium trifluoroacetate (Pd(TFA)₂) resulted in 99 % yield and full conversion of the starting material (Table 2, entry 2). Palladium halide complexes, except for PdI₂, also worked well in our system: PdCl₂ gave 95 % yield, whereas PdBr₂ and Na₂PdCl₄ gave 99 % yield (Table 2, entries 3–5). Because of the lower price of PdBr₂, we choose this palladium precursor for the following range of substrates.

Table 2. Aminocarbonylation of benzyl chloride: Testing of palladium precursors (2 mol %).^[a]

Entry	Palladium source	Conversion^[b] [%]	Yield^[b] [%]
[a] Reaction conditions: Benzyl chloride (1 mmol), CO (2 bar), NH3 (2 bar), [Pd] (0.02 mmol), PPh3 (0.04 mmol), dioxane (2 mL), 100 °C, 16 h. [b] Conversion and yield were determined by using GC; dodecane was used as internal standard. [c] 1 mol %
1	Pd(OAc)2	89	89
2	Pd(TFA)2	100	99
3	PdCl2	98	95
4	PdBr2	100	99
5	PdI2	91	47
6	Na2PdCl4	100	99
7	PdBr2	45	40

A possible reaction mechanism based on previous mechanistic investigations of palladium-catalyzed carbonylations of aryl halides is proposed in Scheme 1.14 The reaction starts with oxidative addition of benzyl chloride to a Pd⁰ species to form the benzyl Pd^II complex. After coordination and insertion of CO, the respective acyl Pd^II complexes are produced. Subsequent nucleophilic attack of ammonia gave the primary amide as the last product. Finally, the reaction of HPd^IIX with ammonia regenerates the active Pd⁰ species, which starts the next reaction cycle.

Scheme 1. Proposed reaction mechanism.

As shown in Table 3, the value of the methodology is proven by the aminocarbonylation of various commercially available benzyl chlorides. In general, all primary 2-arylacetamides were synthesized in good to excellent yields. Simple para- and meta-alkyl-substituted benzyl chlorides gave the corresponding amides in 74–89 % yield (Table 3, entries 2–4). 3-Methoxybenzyl chloride gave 3-methoxyphenylacetamide in 98 % yield (Table 3, entry 5). Benzyl chlorides with electron-withdrawing groups also successfully gave the corresponding products in 83–94 % yields (Table 3, entries 6–8). Not surprisingly, aminocarbonylation of the chloride-substituted substrate proceeded chemoselectively (Table 3, entry 8). In addition to benzyl chlorides, cinnamyl chloride, an example of an allyl chloride, can be used to give the respective primary amide in 75 % isolated yield (Table 3, entry 9).

Table 3. Palladium-catalyzed aminocarbonylation of benzyl chlorides using PdBr₂ as precursor.^[a]

Entry	Benzyl chloride	Amide	Yield [%]^[b]
[a] Reaction conditions: Benzyl Chlorides (1 mmol), CO (2 bar), NH3 (2 bar), PdBr2 (0.02 mmol), PPh3 (0.04 mmol), dioxane (2 mL), 100 °C, 16 h. [b] Isolated yield.
1			97
2			74
3			80
4			89
5			98
6			94
7			92
8			83
9			75

In conclusion, a palladium-catalyzed aminocarbonylation of benzyl chlorides using ammonia has been developed. Various primary 2-arylacetamides have been synthesized in good to excellent yields (up to 98 %). Notably, an inexpensive palladium catalyst system is successfully applied for this atom-efficient methodology.

Experimental Section

A typical reaction procedure for the aminocarbonylation of benzyl chloride to 2-phenylacetamide was performed as follows. A 4 mL vial was charged with PdBr₂ (2 mol %), PPh₃ (4 mol %), and a stirring bar. Then, dioxane (2 mL) and benzyl chloride (1 mmol) were injected by using syringe. The vial (or several vials) was placed in an alloy plate, which was transferred into a 300 mL autoclave of the 4560 series from Parr Instruments under an argon atmosphere. After flushing the autoclave three times by using ammonia, the pressure of ammonia and bar CO was adjusted to 2 bar (1 bar=0.1 MPa) at ambient temperature. Then, the reaction was performed for 16 h at 100 °C. After the reaction was finished, the autoclave was cooled down to room temperature and the pressure was released carefully. The solution was extracted 3–5 times by using ethyl acetate (2–3 mL). After evaporation of the organic solvent, the residue was adsorbed on silica gel, and the crude product was purified by using column chromatography using n-heptane/AcOEt (1:1–0:1) as eluent. ¹H NMR (300 MHz, [D₆]DMSO): δ=7.56 (bs, 1 H), 7.24–7.29 (m, 5 H), 6.92 (bs, 1 H), 3.38 ppm (s, 2 H); ¹³C NMR (75 MHz, [D₆]DMSO): δ=172.3, 136.6, 129.1, 128.2, 126.3, 42.3 ppm.

Acknowledgements

The authors thank the state of Mecklenburg-Vorpommern and the Bundesministerium für Bildung und Forschung (BMBF) for financial support. We also thank Mrs. S. Leiminger, Dr. W. Baumann, and Dr. C. Fischer (all LIKAT) for analytical support.