Suzuki–Miyaura Cross-Coupling of Aryl Carbamates and Sulfamates: Experimental and Computational Studies

Abstract

The first Suzuki–Miyaura cross-coupling reactions of the synthetically versatile O-aryl carbamate and O-sulfamate groups is described. The transformations utilize the inexpensive, bench-stable catalyst NiCl₂(PCy₃)₂ to furnish biaryls in good to excellent yields. A broad scope for this methodology has been demonstrated. Substrates with electron-donating and electron-withdrawing groups (EDGs, EWGs) are tolerated, in addition to those that possess ortho substitutents. Furthermore, heteroaryl substrates may be employed as coupling partners. A computational study providing the full catalytic cycles for these cross-coupling reactions is described. The oxidative additions with carbamates and sulfamates occur via a five-centered transition state, resulting in the exclusive cleavage of the Ar–O bond. Water is found to stabilize the Ni–carbamate catalyst resting state, and thus provides rationalization of the relative decreased rate of coupling of carbamates. Several synthetic applications are presented to showcase the utility of the methodology in the synthesis of polysubstituted aromatic compounds of natural product and bioactive molecule interest.

Introduction

Transition metal-catalyzed cross-coupling reactions provide a powerful means to assemble carbon–carbon (C–C) and carbon–heteroatom (C–X) bonds.¹ Although halides are most commonly employed as the electrophilic partner,^1,2 phenolic derivatives (Figure 1), or ‘pseudohalides’, offer a valuable alternative given that phenols are typically inexpensive and readily available materials.³ Cross-couplings of aryl sulfonates have been most widely studied and a range of C–C and C–X bond forming reactions are now established.^1,4,5 Recent studies have focused on the development of less common phenol-based electrophiles,⁶ such as ethers,⁷ esters,⁸ carbamates,⁹ and sulfamates^9b,10 since they are commonly more robust, typically unreactive toward Pd-catalysis, and show synthetic advantage for the regioselective construction of aromatics by C–H activation and directed ortho metalation (DoM) chemistry.^11,12,13,14

Figure 1.

Phenol-Based Cross-Coupling Partners.

Inspired by the reasons outlined above, our laboratories have pursued the development of cross-coupling reactions involving phenol-derived carbamates and sulfamates (Scheme 1). Previous studies have demonstrated the utility of these reaction partners in nickel-catalyzed Kumada couplings.^9d,9e,10 However, the corresponding Suzuki–Miyaura couplings of these substrates have remained elusive, despite the numerous benefits of organoboronate coupling methodologies. Such advantages include the low toxicity, wide availability, and pronounced stability of organoboronates, in addition to their broad functional group tolerance.^1,15 In this article, we report a) the development of the Ni-catalyzed Suzuki–Miyaura cross-coupling reactions of aryl O-carbamates and O-sulfamates, b) the broad scope of these transformations, which includes the cross-coupling of heterocyclic substrates, c) computational studies that elucidate the complete catalytic cycle of these couplings, and d) a variety of synthetic applications, including DoM – linked tactics and a concise synthesis of the anti-inflammatory drug flurbiprofen.¹⁶

Scheme 1.

Results and Discussion

Suzuki–Miyaura Cross-coupling Reactions of Aryl O-Carbamates

A key challenge in achieving the Suzuki–Miyaura cross-coupling of aryl carbamates lies in activating the fairly inert aryl carbon–oxygen bond of these substrates. A similar obstacle had been overcome in our previously reported Suzuki–Miyaura coupling of aryl pivalates.^8a Encouraged by our prior success, we explored NiCl₂(PCy₃)₂-promoted conditions to effect the desired Suzuki–Miyaura coupling of aryl carbamates (Table 1). Of note, NiCl₂(PCy₃)₂ is readily available, considerably stable to air and water, and can be used on the bench-top rather than in a glovebox.^17,18,19 Initial studies were directed toward the coupling of fused-aromatic systems, which are typically superior substrates in Ni-catalyzed couplings of phenolic derivatives.^7,8,9 Unfortunately, applying our optimal conditions for pivalate coupling (i.e., NiCl₂(PCy₃)₂ (5 mol%) and K₃PO₄ (4.5 equiv) in toluene at 80 °C) to a 1-naphthyl carbamate substrate led only to trace amounts of cross-coupled product. By raising the temperature to 110 °C, however, the desired biaryl was obtained in 51% yield (entry 1). Further optimization ultimately established more forcing conditions that delivered the targeted product in 86% yield (entry 2).

Table 1.

Cross-coupling of aryl carbamates with arylboronic acids^a

entry	Ar–OR	(HO)2B–Ar	product	yieldc
1b		1a		51%
2				86%
3		1a		47%
4		2a		54%
5		2a		77%
6		1a		52%
7		2a		41%

^aConditions: NiCl₂(PCy₃)₂ (10 mol%), ArB(OH)₂ (4 equiv), K₃PO₄ (7.2 equiv), toluene (0.3 M), 130 °C for 24 h.

^bConditions: NiCl₂(PCy₃)₂ (5 mol%), ArB(OH)₂ (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 110 °C, 24 h.

^cYields of isolated products.

Additional carbamate substrates were examined under our Ni-catalyzed reaction conditions (Table 1).²⁰ 2-Naphthyl carbamates gave products in lower yields (entries 3 and 4). The reaction proved tolerant of an EWG (–CO₂Me, entry 4) and an EDG (–OMe, entry 5) on the naphthyl ring. The corresponding reactions of aryl carbamates proved more challenging. Nonetheless, carbamates derived from phenol and p-methoxyphenol were converted to the corresponding cross-coupled products in 52% and 41% yield, respectively (entries 6 and 7).

Further studies were undertaken to uncover higher yielding and more generally useful reaction conditions. The N,N-diethyl carbamate of 2-naphthol was subjected to NiCl₂(PCy₃)₂ with variations in temperature, solvent, ligand additive, and organoboron species (Table 2). When employing o-xylenes at 150 °C, cross-coupling with boroxine 3b proceeded sluggishly but nevertheless furnished the desired biaryl in 61% yield (entry 1). Mixtures of boronic acids and boroxines were also examined. Using a 1:1 mixture of 3a:3b, the desired biaryl was obtained in only 26% yield (entry 2). These results, coupled with the observation that 3a liberates excessive water in organic solvents, led us to hypothesize that, although some water is necessary to generate the catalytically active boronate species, excessive water can be detrimental to the carbamate cross-coupling reaction.²¹ Furthermore, Shi had previously reported the critical role of water in the Suzuki–Miyaura coupling of aryl pivalate esters.^8b By using a 1:10 ratio of 3a:3b,²² and thereby minimizing the water content, a quantitative yield of cross-coupled product was obtained (entry 3). Conducting the reaction at 120 °C, with toluene as solvent and 10 mol% Ni catalyst, gave a lower yield of the biaryl adduct (entry 4).²³

Table 2.

Optimization studies for naphthyl 2-O-carbamate coupling

entryb	solvent	temp	PCy3HBF4	K3PO4	ArB(OR)2	yielda
1	o-xylene	150 °C	10 mol%	5 equiv	3b (2.5 equiv)	61%
2	o-xylene	150 °C	10 mol%	5 equiv	3a:3b (1:1, 2.5 equiv)	26%
3	o-xylene	150 °C	10 mol%	5 equiv	3a:3b (1:10, 2 equiv)	100% (84%)
4c	toluene	120 °C	-	7.2 equiv	3a:3b (1:10, 2.5 equiv)	62%

^aYield by GC/MS analysis (yield of isolated product).

^bAll reactions were run for 20 h with the exception of entry 4, which was run for 5 h.

^c10 mol% NiCl₂(PCy₃)₂.

Having found optimal and reproducible conditions, we turned our attention to defining the scope and functional group tolerance of the carbamate cross-coupling reaction (Table 3).²⁰ Substrates derived from 2-naphthol, 1-naphthol, and phenol underwent smooth coupling (entries 1–3). Furthermore, although a substrate bearing the electron-withdrawing fluoro substituent was tolerated (entry 4), coupling in the presence of a cyano derivative was less fruitful (entry 5). The latter result can be explained by competitive cross-coupling at the cyano group, a transformation reported recently by Shi.²⁴ Finally, an electron-rich substrate gave only low yields of product (entry 6).

Table 3.

Cross-coupling of aryl carbamates under improved conditions^a

entry	Ar–OC(O)NEt2	(RO)2B–Ar	product	yieldb
1		3		100% (84%)
2		2		100% (82%)
3		1		64% (58%)
4		1		80% (69%)
5		2		36% (28%)
6		2		23% (31%)

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), PCy₃HBF₄ (10 mol%), ArB(OR)₂ (2.5 equiv), ratio of Ar₃B₃O₃:ArB(OH)₂ = 10:1 (4 equiv), K₃PO₄ (5 equiv).

^bYield by GC/MS analysis (yield of isolated product).

Several ortho-substituted aryl carbamates were tested in the Suzuki–Miyaura coupling (Table 4). Substrates of this type can be readily synthesized by DoM¹² or transition metal-catalyzed C–H functionalization.^14b–e Substrates with o-benzyl, -alkenyl, and -phenyl groups were all tolerated (entries 1–3). Coupling of the o-methoxy substrate proceeded in modest yield (entry 4), whereas coupling of a 2,4-dimethylated substrate was unsuccessful (entry 5). In view of the coupling of other ortho-substituted systems to afford good to excellent yields of products (entries 1 – 3), rationalization of these results based on steric effects is premature.

Table 4.

Cross-coupling of ortho-substituted aryl O-carbamates^a

entry	Ar–OC(O)NEt2	product	yieldb
1			60% (70%)
2			99% (93%)
3			69% (50%)
4			40% (36%)
5			8%

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), PCy₃HBF₄ (10 mol%), 2 (2.5 equiv), ratio of Ph₃B₃O₃:PhB(OH)₂ = 10:1 (4 equiv), K₃PO₄ (5 equiv).

^bYield by GC/MS analysis (yield of isolated product).

As shown in Table 5, the carbamate cross-coupling methodology is also applicable to heterocyclic substrates. Thus, the 3-pyridyl-carbamate was efficiently cross-coupled with a variety of organoboron species (entries 1–5). In addition, a quinoline-derived substrate was tolerated (entry 6) and a carbazole-containing substrate underwent conversion to the desired biaryl under our optimal reaction conditions albeit in modest yield (entry 7).

Table 5.

Cross-coupling of heterocyclic aryl O-carbamates and scope of aryl boronates^a

entry	HetAr–OC(O)NEt2	(RO)2B–Ar	product	yieldb
1		1		(84%)
2		2		100% (85%)
3		3		(87%)
4		4		(70%)
5		5		(65%)
6		2		100% (51%)
7		2		45% (36%)

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), PCy₃HBF₄ (10 mol%), ArB(OR)₂ (2.5 equiv), ratio of Ar₃B₃O₃:ArB(OH)₂ = 10:1 (4 equiv), K₃PO₄ (5 equiv).

^bYield by GC/MS analysis (yield of isolated product).

Suzuki–Miyaura Cross-coupling Reactions of Aryl Sulfamates

Concurrent with the above studies on aryl carbamates, aryl sulfamates were targeted as substrates for the Ni-catalyzed Suzuki–Miyaura cross-coupling reactions. Although initial efforts to effect this transformation with dppp as ligand gave initial encouragement,^10a employing the tricyclohexylphosphine ligand led to improved results, ultimately rendering aryl sulfamates superior Suzuki–Miyaura coupling partners to the corresponding carbamates (Table 6).²⁰ Naphthyl substrates were smoothly converted to biaryl products,²⁵ even in the presence of an EWG or EDG (entries 1–3). Most strikingly, the reaction proceeded comparably well when operating on aryl derivatives (entries 4–9). Methyl and the electron-withdrawing CF₃ substituents are tolerated (entries 5–7). Substrates bearing electron-rich methoxy or amino substituents also afforded very good yields of coupled products (entries 8 and 9). Moreover, an enone sulfamate participated in the Suzuki–Miyaura cross-coupling reaction (entry 10).

Table 6.

Cross-coupling of aryl sulfamates^a

entry	Ar–OSO2NMe2	(HO)2B–Ar	product	yieldb
1		1a		95%
2		2a		72%
3		2a		92%
4		1a		87%
5		1a		89%
6		1a		91%
7		2a		81%
8		2a		80%
9		2a		76%
10		2a		75%

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), ArB(OH)₂ (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 110 °C for 24 h.

^bYields of isolated products.

In view of the availability of many ortho-substituted aryl sulfamates by DoM chemistry, such derivatives were also evaluated in the Suzuki–Miyaura cross-coupling reaction (Table 7).^20,26 The transformation was found to be tolerant of an ortho-cresol derived substrate, in addition to the sterically burdened sulfamate prepared from 2,6-dimethylphenol (entries 1 and 2). Furthermore, substrates bearing ortho-trimethylsilyl, -phenyl, and -methoxy substituents underwent cross-coupling to give the corresponding products in excellent yields (entries 3–5). Interestingly, a substrate possessing a bulky ortho-t-butylketone substitutent could also be utilized in this methodology (entry 6).

Table 7.

Cross-coupling of o-substituted aryl sulfamates^a

entry	Ar–OSO2NMe2	(HO)2B–Ar	product	yieldb
1		1a		92%
2		1a		63%
3c		1a		92%
4		1a		93%
5		2a		90%
6c		2a		85%

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), ArB(OH)₂ (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 110 °C for 24 h.

^bYields of isolated products.

^cConditions: NiCl₂(PCy₃)₂ (10 mol%), ArB(OH)₂ (4 equiv), K₃PO₄ (7.2 equiv), toluene (0.3 M), 130 °C for 24 h.

Although the DoM chemistry of aryl sulfamates was initially reported using N,N-diethyl substrates,^10a the corresponding N,N-dimethyl aryl sulfamates were found to undergo metalation under the identical reported reaction conditions. Scheme 2 highlights syntheses of substrates 8–11 beginning from phenyl sulfamate 6, which, in turn, is easily prepared from phenol and commercially available dimethylsulfamoyl chloride²⁷ in quantitative yield. Compounds 9 and 10 were obtained by lithiation of phenyl sulfamate 6, followed by quenching with TMSCl and PivCl, respectively. Similarly, the boronate 7 was derived by quenching the intermediate lithio species with B(OMe)₃, followed by treatment with pinacol.^10a Boronate 7 served as the common precursor to substituted sulfamates 8 and 11. Whereas methoxysulfamate 11 was prepared by a straightforward oxidation²⁸/methylation sequence, ortho-phenyl sulfamate 8 was accessed by a Pd-catalyzed Suzuki–Miyaura cross-coupling. It is notable that the sulfamate remains undisturbed under the Pd-mediated reaction conditions.

Scheme 2.

The scope of the sulfamate cross-coupling reaction was also found to be broad with respect to the boronic acid component (Table 8). A methyl substituent was tolerated at the para, meta, and ortho positions (entries 1–3), as was a 4-methoxymethyl group (entry 4). Cross-coupling of a boronic acid bearing the electron-donating methoxy group proceeded in 95% yield (entry 5). Finally, electron-withdrawing trifluoromethyl, fluoro, and acetyl substituents were compatible with the sulfamate coupling methodology (entries 6–8).

Table 8.

Scope of aryl boronic acid in the Suzuki–Miyaura cross-coupling of aryl O-sulfamates^a

entry	(HO)2B–Ar	product	yieldb
1			68%
2			77%
3c			74%
4			80%
5			95%
6d			86%
7			93%
8			62%

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), ArB(OH)₂ (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 110 °C for 24 h.

^bYields of isolated products.

^cConditions: NiCl₂(PCy₃)₂ (10 mol%), ArB(OH)₂ (4 equiv), K₃PO₄ (7.2 equiv), toluene (0.3 M), 130 °C for 24 h.

^dConditions: NiCl₂(PCy₃)₂ (5 mol%), ArB(OH)₂ (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 120 °C for 24 h.

Suzuki–Miyaura Cross-coupling Reactions of Heterocyclic O-Sulfamates

Given the importance of heterocycles in medicinal agents, we probed the use of heterocyclic partners in the sulfamate Suzuki–Miyaura cross-coupling process.²⁹ As shown in Table 9, a variety of heterocyclic aryl sulfamates were suitable for this methodology, although more forcing reaction conditions were often required to achieve synthetically useful yields. Coupling of a dihydrobenzofuran-derived substrate afforded the desired biaryl in 88% yield (entry 1). Similar success was observed in the coupling of nitrogen-containing heteroaryl sulfamates (entries 2–6). In addition to indole and carbazole (entries 2 and 3), the pyridine and quinoline heterocycles, each possessing basic amine functionality, were also tolerated (entries 4–6).

Table 9.

Cross-coupling of heterocyclic aryl O-sulfamates with phenyl boronic acid^a

entry	HetAr–OSO2NMe2	product	yieldb
1c			88%
2			75%
3c			89%
4c			72%
5c			89%
6c			74%

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), 2a (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 110 °C, 24 h.

^bYields of isolated products.

^cConditions: NiCl₂(PCy₃)₂ (10 mol%), 2a (4 equiv), K₃PO₄ (7.2 equiv), toluene (0.3 M), 130 °C for 24 h.

The scope of the sulfamate cross-coupling reaction with respect to heteroaryl boronic acids is summarized in Table 10. Benzofuran- and furan-containing substrates underwent smooth cross-coupling under our standard reaction conditions (entries 1 and 2). Furthermore, a sulfur-containing heterocyclic boronic acid could be employed (entry 3). A pyridine 3-boronic acid derivative was also tolerated in our Suzuki–Miyaura coupling methodology (entry 4).

Table 10.

Cross-coupling of naphthyl 1-O-sulfamates with heterocyclic aryl boronic acids^a

entry	(HO)2B–HetAr	product	yieldb
1			67%
2			79%
3			80%
4			80%

^aConditions: NiCl₂(PCy₃)₂ (5 mol%), HetArB(OH)₂ (2.5 equiv), K₃PO₄ (4.5 equiv), toluene (0.3 M), 110 °C, 24 h.

^bYields of isolated products.

As an additional important test of the sulfamate coupling methodology, we attempted a Suzuki–Miyaura reaction wherein both coupling partners were heterocyclic substrates (Scheme 3).²⁹ We were delighted to find that the desired cross-coupling between quinoline-derived sulfamate 12 and pyridinyl boronic acid 13 proceeded smoothly to furnish biaryl 14 in 97% yield. This result underscores the critical tolerance of the sulfamate cross-coupling process to basic nitrogen substituents.

Scheme 3.

Mechanistic Studies

Pd-catalyzed Suzuki-Miyaura cross-couplings have been studied computationally by various groups.^30,31 The three key steps in the catalytic cycle, oxidative addition,³² transmetallation,³¹ and reductive elimination,³³ have been studied carefully for reactions involving a variety of substrates. The mechanism of Ni-catalyzed Suzuki-Miyaura cross-coupling with aryl acetates has been recently investigated theoretically by Li et al.³⁴ Here we report the first theoretical study of the catalytic cycles of the Ni-catalyzed Suzuki–Miyaura cross-coupling with O-carbamates and O-sulfamates using density functional theory (DFT). The selectivities between couplings with the Ar–O bond and the O–carbonyl/sulfonyl bond and the effects of water on the reactivities are also described.

Figures 3 and 4 depict the catalytic cycles for the Suzuki–Miyaura cross-coupling of aryl carbamates and sulfamates, respectively, as determined by DFT calculations. The geometries of important transition structures are shown in Figure 5 for the coupling of N,N-dimethyl phenyl carbamate and N,N-dimethyl phenyl sulfamate with phenyl boronic acid. The PCy₃ ligand used in the experiments was also used in the calculations. Geometry optimizations and frequency calculations were performed using B3LYP³⁵ and a mixed basis set employing LANL2DZ for metal and 6–31G(d) for other atoms. Conformational searches of the PCy₃ ligand were performed. The initial geometry of PCy₃ was taken from the crystal structure of Ni(PCy₃)(C₂H₄)₂.³⁶ Several rotamers of the PCy₃ ligands in the Ni complexes were tested as the initial geometry in the optimizations. Energies reported are Gibbs free energies in solution, which involve zero-point vibrational energy corrections, thermal corrections to Gibbs free energy at 298 K, and solvation free energy corrections computed by singlet point CPCM³⁷ calculations on gas-phase optimized geometries (toluene was used as solvent). The molecular cavities were built up using the United Atom Topological Model (UAHF). Vibrational frequencies were calculated for all optimized structures to confirm the nature of the stationary points. All calculations were performed using Gaussian 03.³⁸

Figure 3.

Gibbs free energy profile of Ni-catalyzed Suzuki-Miyaura cross-coupling reaction of phenyl N,N-dimethyl O-carbamate 15 with phenylboronic acid. PCy₃ was used as ligand in the calculations. For clarity, the cyclohexyl groups on the ligand are not shown.

Figure 4.

Gibbs free energy profile of the Ni-catalyzed Suzuki-Miyaura cross-coupling reaction of N,N-dimethyl phenyl O-sulfamate with phenylboronic acid. PCy₃ was used as ligand in the calculations. For clarity, the cyclohexyl groups on the ligand are not shown.

Figure 5.

Transition state structures of Ni-catalyzed oxidative additions of (a) N,N-dimethyl phenyl O-carbamate and (b) N,N-dimethyl phenyl O-sulfamate. PCy₃ was used as ligand in the calculations. For clarity, the cyclohexyl groups on the ligand are not shown.

The oxidative addition of aryl carbamates may occur via several different pathways: the Ni may be mono- or bis-ligated; the oxidative addition may occur at the Ph–O bond or the O–carbonyl bond of the carbamates. Previous theoretical studies suggested that the oxidative addition of aryl halides to Pd(0) catalysts involves formation of an η² LPd(ArX) pre-reaction complex.³² The η² LPd(ArX) complex may be generated through ligand dissociation from PdL₂ followed by coordination with aryl halide or through a concerted or stepwise associative displacement pathway.³⁹ Recent density functional calculations by Li et al. suggested that the oxidative addition of phenylacetates in Ni-catalyzed Suzuki–Miyaura couplings also involves an η² LNi(ArX) pre-reaction complex.^8c Upon dissociation of a PCy₃ ligand, the Ni catalyst coordinates with the substrate to generate an η² complex 16, which is slightly less stable than Ni(PCy₃)₂.⁴⁰ Li et al. suggested that the oxidative addition of phenylacetates occurs via a three-centered transition state, and the weaker PhO–carbonyl bond is more reactive compared to the Ph–O bond.^8c We investigated the possible pathways in the oxidative additions with phenyl carbamates (Figures 3 and 5) and found that the preferred pathway involves oxidative addition at the Ph–O bond via a five-centered transition state (TS17, Figures 3 and 5) in which Ni is mono-ligated and coordinated with the carbonyl oxygen.^41,42 The corresponding three-centered transition state (TS30) that uses a single oxygen of the carboxylate to bridge requires 7.4 kcal/mol higher activation energy. In contrast to previous theoretical studies by Li et al., the Ph–O bond in carbamates is more reactive in oxidative addition than the PhO–carbonyl bond, although the former is a stronger bond in terms of bond dissociation energies.⁴³ Oxidative addition at the O–carbonyl bond can only occur via a three-centered transition state (TS31) and requires 3.9 kcal/mol higher energy than the oxidative addition at the Ph–O bond (TS17). Thus, the oxidative addition occurs exclusively at the Ph–O bond due to a favorable five-centered transition state. The carbonyl group is acting as a directing group to activate the Ph–O bond in the oxidative addition. It is conceivable that such activating effects by adjacent oxygenation is also present in oxidative additions with carbonates, sulfonates, and sulfamates, etc.

A stable phenyl Ni(II)-carbamate complex 18 is formed after the oxidative addition (Figure 3). The carbamate is κ₂-coordinated with Ni. Subsequent ligand exchange of the carbamate complex 18 with phenylboronate leads to phenyl Ni(II) boronate complex 20, which is 5.5 kcal/mol less stable than the Ni-carbamate complex. The detailed mechanism of this ligand exchange step has been suggested to be stepwise.^15a,44 Previous theoretical studies suggested that the ligand exchange does not have large barrier.^8c,31b,g Thus, we assume the transformation from 18 to 20 is facile.

The following transmetallation (TS21) is the rate-determining step of the catalytic cycle, and requires an activation energy of 30.2 kcal/mol from the catalyst resting complex 18. The transmetallation transition state (TS21) is consistant with the four-center transition state proposed in previous theoretical studies of Pd- and Ni-catalyzed Suzuki-Miyaura couplings.^{31, 8c} The two aryl groups are cis to each other in the transmetallation transition state. The trans transition state is 3.5 kcal/mol less stable, presumably due to greater steric repulsions between the ligand and the aryl groups. Subsequent reductive elimination of the diphenyl Ni(II) complex (23 → TS24) is facile, requiring only a 2.9 kcal/mol activation energy.

Similarly, the oxidative addition of N,N-dimethyl phenyl sulfamate occurs via a mono-ligated five-membered transition state (TS27).⁴¹ Three-center transition states TS32 and TS33 are both much higher in energy (Figure 5). The activation barrier of oxidative additions of sulfamate 6 is 10.4 kcal/mol with respect to the Ni(PCy₃)₂ complex, which is 3.1 kcal/mol lower than that of the oxidative addition of N,N-dimethyl phenyl carbamate 15. The higher reactivity of the sulfamate in oxidative addition is due to the weaker Ph–O bond in sulfamate than the corresponding Ph–O bond in the carbamate group. Nonetheless, the oxidative addition with aryl carbamates and aryl sulfamates are both predicted to be very facile. The differences of their reactivities are attributed to the different activation barriers in the rate-determining transmetalation step. After the oxidative addition of the carbamate, a stable phenyl Ni(II)-carbamate complex 18 is formed. Subsequent ligand exchange with phenyl boronate requires 5.5 kcal/mol energy to form the Ni(II) boronate complex 20. In contrast, since sulfamate is a better leaving group, the Ni(II) boronate complex 20 is formed spontaneously from the Ni(II)-sulfamate complex 28. The activation barrier of the rate-determining transmetalation step for the cross-coupling with the sulfamate (ΔG^‡ = 24.7 kcal/mol) is much lower than that for the corresponding carbamate (ΔG^‡ = 30.2 kcal/mol). The subsequent steps after transmetalation are identical for the coupling reactions for carbamates and sulfamates.

To support the computational finding that transmetallation is the rate-determining step in the cross-coupling reactions described above, a series of experiments were carried out. Sulfamate 34 was independently subjected to reactions with boronic acids 1a, 2a, and 4a in the presence of NiCl₂(PCy₃)₂ and K₃PO₄ in toluene at 80 °C (Figure 6¹H NMR analysis using hexamethylbenzene as internal standard. The relative rate of cross-coupling was found to be dependent on the individual boronic acid employed, with a direct correlation between electron-richness of the boronic acid and reaction rate (i.e., relative rate of conversion: 1a>2a>3a).⁴⁵ These findings are consistent with a rate-determing transmetalation step for the sulfamate cross-coupling process.⁴⁶

Figure 6.

Qualitative relative rates of cross-coupling depending on boronic acid

Similar to the reports by Shi in related Ni-catalyzed Suzuki–Miyaura cross-couplings,^8b we have observed that water can play a critical role in the success or failure of a coupling reaction (vida supra). To better understand these experimental findings, we examined the role of water computationally. In the coupling with phenyl carbamate, water can coordinate with Ni and stabilize the catalyst resting state, the Ni-carbamate complex 18. A six-membered cyclic Ni(II)-water-carbamate complex 19 is formed and is 1.1 kcal/mol more stable than 18 (Figure 3).⁴⁷ Coordination with water increases the barrier of transmetalation to 31.3 kcal/mol (19 → TS21), and thus decreases the reactivity of the coupling for the carbamate. In the coupling with phenyl sulfamate, the catalyst resting state is the Ni(II)-boronate complex 20. Upon coordination with a water molecule, a similar six-membered cyclic complex 29 is formed in equilibrium. However, 29 is 7.7 kcal/mol less stable than 20. This suggests that coordination with water does not affect the barrier of transmetalation in the coupling reaction of the sulfamate. This agrees with the experimental observation that Suzuki–Miyaura cross-couplings of aryl sulfamates are less sensitive to water than couplings of aryl carbamates.

In contrast to the high reactivity of the Ni catalyst in oxidative addition with carbamates and sulfamates, Pd catalysts are much less reactive in the oxidative addition step. Oxidative addition of phenyl N,N-dimethylcarbamate with Pd also prefers a five-membered mono-ligated transition state. The activation barrier is 42.2 kcal/mol with respect to the Pd(PCy₃)₂ complex, much higher than that of the corresponding Ni catalyst. Similarly, oxidative addition of phenyl N,N-dimethylsulfamate also requires a very high activation barrier of 39.7 kcal/mol. These observations are in agreement with previous mechanistic and theoretical studies that the oxidative addition step to Ni(0) is more facile than that to Pd(0).^32,40 Therefore, due to the extremely high activation barriers for oxidative addition, Pd catalysts are not effective in couplings with carbamates and sulfamates.

Synthetic applications

The scope and limitations of our sulfamate and carbamate coupling methodologies were further examined by way of a variety of synthetic applications. In each of the studies undertaken, the synthetically useful capability of carbamates and sulfamates to function as directed metalation groups (DMGs) and Suzuki–Miyaura coupling partners was exploited. These studies showcase the utility of our methodology in the synthesis of polysubstituted aromatic compounds, with relevance to natural product and bioactive molecule synthesis.

Scheme 4 depicts a concise synthesis of 5-phenyl-2H-chromene 38 beginning from bis(carbamate) precursor 36. Thus, in a one-pot procedure involving sequential treatment with t-BuLi, 3-methylbut-2-enal, and AcOH, compound 36 was converted into 2H-chromene carbamate 37.^48,49 Subsequent Suzuki–Miyaura cross-coupling provided biaryl 38 in 56% yield. Biaryl 38 possesses a heterocyclic framework of bioactivity⁵⁰ and natural product interest.⁵¹

Scheme 4.

In another application, the unique heterotriaryl 43 was constructed using carbamate DoM and cross-coupling methodology (Scheme 5). ortho-Methoxy carbamate 39 was first transformed into boronic acid 40 in 88% yield using a standard lithiation/borylation protocol. Subsequent Pd-catalyzed Suzuki–Miyaura cross-coupling with iodobenzofuran 41 delivered arylated product 42 without disturbance of the aryl carbamate under these conditions.⁵² The subsequent Ni-catalyzed carbamate cross-coupling with phenyl boronic acid provided the targeted heterotriaryl 43.⁵³

Scheme 5.

An illustration of the DoM/cross-coupling protocol beginning from a heteroaryl carbamate is presented in Scheme 6. Lithiation/borylation of 44 afforded the boropinacolate 45, which was subjected to standard Suzuki-Miyaura cross-coupling conditions with bromide 46 using Pd catalysis to furnish heterobiaryl 47.⁵⁴ Heteroaryl carbamate 47 was found to be an excellent substrate for the Ni-catalyzed cross-coupling using our standard conditions, to afford the heterotriaryl product 48 in 91% yield. Compound 48 represents a class of pyridines with nonidentical diaryl substitution for which only two synthetic methods are available.⁵⁵

Scheme 6.

As an application to bioactive molecule synthesis, the anti-inflammatory drug flurbiprofen⁵⁶ was prepared using sulfamate methodology (Scheme 7). Boronic acid 49, derived from o-lithiation/borylation of N,N-dimethyl phenyl sulfamate, was fluorinated using the conditions described by Furuya and Ritter⁵⁷ to provide fluorosulfamate 50. Alternative routes to generate 50 by direct lithiation/fluorination of N,N-dimethyl phenyl sulfamate were unsuccessful despite numerous attempts.⁵⁸ Nonetheless, para-selective electrophilic iodination of 50 furnished 51 in 64% yield. With the aryl sulfamate being inert to Pd catalysis, we carried out a site-selective enolate coupling to install the necessary propionate side chain. Whereas enolate coupling of aryl iodide 51 under Buchwald’s Pd-based conditions was feasible,⁵⁹ higher yields of 52 were obtained using a Ni-catalyzed variant.⁶⁰ With the sulfamate remaining undisturbed, exposure of 52 to our Ni-catalyzed conditions facilitated the key sulfamate cross-coupling and delivered the biaryl 53. Acid-mediated hydrolysis furnished flurbiprofen (54) in 84% yield over the two steps. It should be emphasized that the aryl fluoride of 52 was chemically inert under our Ni-catalyzed cross-coupling conditions.⁶¹

Scheme 7.

We have observed that aryl carbamates amd sulfamates are unreactive toward Pd(0) catalysis (vida supra) and related processes.⁶² This feature allows the sequential cross-coupling of an aryl halide, followed by either an aryl sulfamate or carbamate coupling process (see Scheme 7). To further probe related issues of orthogonality, we questioned if it would be possible to couple aryl sulfamates in the presence of aryl carbamates. Although aryl sulfamates generally provide higher yields of cross-coupled products compared to aryl carbamates, the relative reactivity of these substrates had not been determined previously. As shown in Figure 7, an equimolar mixture of phenyl carbamate 55 and phenyl sulfamate 6 was treated with an excess of boronic acid 3a under Ni-catalyzed cross-coupling conditions. Although significant selectivity was observed at elevated temperatures, complete selectivity for sulfamate coupling was readily achieved at 50 °C, as determined by ¹H NMR analysis with hexamethylbenzene as internal standard.⁴⁵ The selectivity for sulfamate coupling is attributed to the lower oxidative addition barrier than that in the corresponding step for carbamates (see above computational studies). Analogous experiments were conducted on the naphthyl-based substrates, carbamate 56 and sulfamate 57, and a high selectivity for naphthyl sulfamate over carbamate cross-coupling was observed at 40 °C.⁴⁵ We expect that these observations will be of synthetic value.

Figure 7.

Intermolecular competition experiments of aryl sulfamates and aryl carbamates.

Conclusion

In summary, we have discovered the first Suzuki–Miyaura cross-coupling reactions of the synthetically versatile O-aryl carbamate and O-sulfamate groups. The transformations utilize the inexpensive, bench-stable catalyst NiCl₂(PCy₃)₂ to deliver biaryls in good to excellent yields. The methodology is tolerant of substrates bearing electron-donating and electron-withdrawing groups, in addition to those that possess ortho substitutents and heterocyclic frameworks. Furthermore, a computational study has revealed the full catalytic cycles for these cross-coupling reactions, thus shedding light on various mechanistic details, rationalizing sulfamate over carbamate higher reactivity, and indicating the role of water in the transition state. As demonstrated by the given synthetic applications, the methodology provides an efficient means to access polysubstituted aromatic compounds, with relevance to both natural product and bioactive molecule synthesis. The orthogonal use of the sulfamate or carbamate reactivities, in combination with directed ortho metalation (DoM) and other aryl O-based cross-coupling reactions in arene and heteroarene synthesis, may be anticipated.