Iodide-Catalyzed Carbonylation–Benzylation of Benzyl Chlorides with Potassium Aryltrifluoroborates under Ambient Pressure of Carbon Monoxide

Abstract

Tetra-N-butylammonium iodide (TBAI) catalyzed carbonylation–benzylation of unactivated benzyl chlorides with potassium aryltrifluoroborates using CO gas has been developed. This reaction is transition-metal free, is carried out under ambient pressure, and provides a wide range of 1,2,3-triarylpropan-1-one derivatives in high yields. The novel method represents a significant improvement over the traditional palladium-catalyzed carbonylation.

Key words

benzyl chlorides - carbonylation - potassium aryltrifluoro­borates - radicals - synthetic methods

The three-component coupling of organohalide, organoborane reagent, and carbon monoxide, which is generally referred to as the carbonylative Suzuki reaction,[1] is a powerful synthetic tool for the construction of ketones, which is an important moiety in numerous natural products, pharmaceuticals, photosensitizers, and advanced organic materials.[2] This transformation has become one of the best-known palladium-catalyzed carbonylation reactions over the last few decades owing to its reagent stability, high functional group tolerance, efficiency, and reliability.[3] In contrast to the carbonylative Suzuki coupling of aryl halides that is well-established,[4] the efficient catalytic systems for the carbonylations of alkyl or benzyl halides have been sparsely reported. This is because C(sp³) halides are more reluctant to undergo oxidative addition to Pd⁰ than their unsaturated analogues.[5] Miyaura et al. first reported only two examples of PdCl₂(PPh₃)₂-catalyzed carbonylative coupling of benzyl bromide with arylboronic acid.[6] Subsequently, a more systematic study conducted by the Beller group involved Pd(OAc)₂/PCy₃-catalyzed-catalyzed carbonylative Suzuki reactions of benzyl chlorides with arylboronic acids under 10 bar of CO gas.[7] Recently, α-halomethyl oxime ethers were employed to couple with arylboronic acids at atmospheric pressure of CO gas and in the presence of Pd(PPh₃)₄ (10 mol%) as catalyst.[8] However, these palladium-based catalysts have several limitations: (1) high cost and low abundance of Pd (1 × 10^–6 wt% in the earth crust), (2) sensitive to air and moisture, (3) problems with residual metal in products, and (4) deactivation of Pd catalyst (CO is a strong π-acidic ligand).[9] Therefore, the development of an efficient and practical strategy devoid of metal catalysts is a highly desirable and important challenge.[10]

Although transition-metal-free processes for the carbonylation of organohalides (or pseudo-halides) by utilizing CO gas have been reported,[11] most of them are limited to radical alkoxycarbonylation of aryl halides (or pseudo-halides).[11a] [b] [c] Recently, we reported the first example of transition-metal-free carbonylative Suzuki coupling of aryl halides with potassium aryltrifluoroborates under ambient pressure.[11d] Based on this work, we further developed an unprecedented iodide-mediated radical carbonylation–­benzylation of benzyl chlorides with arylboronic acids.[11e] Unfortunately, this method is ineffective for organotrifluoroborate reagents.

Organotrifluoroborates have been employed in many contexts as an advantageous alternative to boronic acids and boronate esters.[12] These organoboron species are immune to reactions with important classes of reagents (e.g., oxidants, bases, and nucleophiles) commonly utilized in organic synthesis owing to their tetracoordinate nature. Thus, organotrifluoroborates, superior to most other boron species, carry through synthetic operations to build molecular complexity, while leaving the carbon–boron bond intact. This expands the range of synthetic approaches available to target molecules of interest. Although organotrifluoroborates possess these benefits mentioned above, reports of the carbonylation of these species to give the corresponding carbonyl compounds are severely limited.[11d] [13] To our knowledge, only one example of carbonylative Suzuki ­coupling of C(sp³) halides with potassium aryltrifluoroborates has been reported,[13a] albeit air-sensitive palladium catalyst and high pressure of CO gas were required.

Here, we report a transition-metal-free, iodide-­catalyzed radical carbonylation–benzylation of benzyl chlorides with potassium aryltrifluoroborates under ambient pressure of CO gas.[14] [15] This process addresses the problems of Pd-based catalysis as well as the limitation of our previous iodide-mediated carbonylation.[11e] Notably, this transformation provides a step-economic, inexpensive, and operationally simple method for the construction of the 1,2,3-triarylpropan-1-one motif in known pharmaceuticals and biologically active compounds.[16]

Table 1 Optimization of Reaction Conditions^a

Entry	[Cat.] (%)	Base	Solvent	Yield of 3aa (%)b
1	NaI (15)	Na3PO4	PEG-400	31
2	TBAI (15)	Na3PO4	PEG-400	78
3c	THAI (15)	Na3PO4	PEG-400	35
4	KI (15)	Na3PO4	PEG-400	59
5	TBAI (20)	Na3PO4	PEG-400	81
6	–	Na3PO4	PEG-400	–
7	TBAI (20)	Na2CO3	PEG-400	88
8	TBAI (20)	Li2CO3	PEG-400	5
9	TBAI (20)	K2CO3	PEG-400	25
10	TBAI (20)	Cs2CO3	PEG-400	trace
11	TBAI (20)	NaHCO3	PEG-400	86
12	TBAI (20)	KHCO3	PEG-400	5
13	TBAI (20)	K2HPO4	PEG-400	trace
14	TBAI (20)	NaH2PO4	PEG-400	trace
15	TBAI (20)	KF	PEG-400	trace
16d	TBAI (20)	DBU	PEG-400	trace
17	TBAI (20)	Na2CO3	Glycol	trace
18	TBAI (20)	Na2CO3	DMF	35
19	TBAI (20)	Na2CO3	H2O	25

^a Reaction conditions (unless otherwise stated): 1a (0.5 mmol), 2a (0.25 mmol), CO (1 atm), base (1.0 mmol), solvent (2.0 mL), 100 °C, 3 h.

^b Yield of isolated product.

^c THAI (tetraheptylammonium iodide).

Our investigation into this new carbonylative coupling reaction began with exposure of 1,2-dichloro-4-(chloromethyl)benzene (1a) and potassium phenyltrifluoroborate (2a) to a variety of conditions (Table [1]). In initial experiments, we used similar reaction conditions to those under which the iodide-catalyzed carbonylative coupling of benzyl chloride with phenylboronic acid proceeded well (80% yield).[11e] However, the present reaction delivered the product 3aa in just 31% yield (entry 1). This difference may be caused by the increased stability of PhBF₃K.[13a]

Gratifyingly, the use of tetra-N-butylammonium iodide (TBAI) as catalyst increased the yield dramatically (78%; Table [1], entry 2). Raising the amount of catalyst to 20 mol% further improved the yield of 3aa (entry 5). However, replacing TBAI with THAI or KI as the catalyst resulted in much lower yields (entries 3 and 4). The choice of base was crucial for this reaction. Among the tested bases, Na₂CO₃ was optimal, yielding 88% 3aa (entry 7). None of the other solvents tested (glycol, DMF, and H₂O) could replace PEG-400 (entries 17–19).

With the above optimized conditions, the TBAI-catalyzed carbonylation-benzylation of a range of benzyl chlorides was tested.[17] Generally, good to excellent yields and selectivities were observed under ambient pressure of CO (Scheme [1]). The dichloro- and chloro- substituted benzyl chlorides underwent successful coupling with isolated yields of 88, 85, and 87% (3aa–ca), respectively. In a similar manner, benzyl chlorides bearing fluoro or trifluoromethyl substituents were also competent substrates, irrespective of the position of these groups on the phenyl ring (3da–ia). Furthermore, substrates having electron-donating groups such as trifluoromethoxy, methoxy, and methyl worked well with 2a to generate the corresponding products in excellent yields (3ja–ka and 3ma–na). However, 4-methoxylbenzylchloride proved to be unreactive (3oa).

Subsequently, a series of potassium aryltrifluoroborates were investigated. Potassium aryltrifluoroborates bearing electronically neutral, deactivated, and activated cases efficiently coupled with typical benzyl chlorides under normal conditions (Scheme [2]). No significant electronic effects were observed for meta- and para-substituted potassium aryltrifluoroborates. Even the sterically hindered potassium 2-methoxyphenyltrifluoroborate also reacted well in this carbonylation, affording the expected product 3lf in 71% yield. Moreover, the method can move beyond the simple aryl group. Specifically, potassium naphthalen-2-yltrifluoroborate underwent successful coupling with electron-poor, electron-neutral, and electron-rich benzyl chlorides. Potassium aryltrifluoroborate, having a strong electron-withdrawing group, undertook this reaction with lower efficiency (3mi).

To further demonstrate the utility of this method, the present procedure was amenable to a gram-scale reaction. Thus, starting from 10.0 mmol of 1d and 5.0 mmol of 2a, 1.31 g of 3da, which is a potentially useful intermediate for the synthesis of pharmaceuticals and organic materials, was obtained, which corresponds to a yield of 83% (Equation 1).

Equation 1

Scheme 1 TBAI-catalyzed carbonylation–benzylation reactions of 2a with various benzyl chlorides. Reagents and conditions: 1 (1.0 mmol), 2a (0.5 mmol), CO (1 atm), Na₂CO₃ (1.0 mmol), PEG-400 (2.0 mL), 100 °C. Yields of the isolated products are given.

Scheme 2 TBAI-catalyzed carbonylation–benzylation reactions of 1 with 2. Reagents and conditions: 1 (1.0 mmol), 2 (0.5 mmol), CO (balloon), Na₂CO₃ (1.0 mmol), PEG-400 (2.0 mL), 100 °C. Yields of the isolated products are given.

It had been suggested that carbonylative reaction of alkyl halide and arylboronic acid would proceed through a radical mechanism.[14d] [11e] To check the involvement of this process in the present transformation, several control experiments were carried out. A single-electron-acceptor 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) was used, which led to complete inhibition of the model reaction (Scheme [3a]). Furthermore, we also attempted to employ a free-radical scavenger galvinoxyl to trap a radial intermediate; however, the reaction with the galvinoxyl did not result in the generation of product 3aa. Alternatively, benzylated galvinoxyl 3′ was obtained (Scheme [3b]), which suggests the existence of the benzyl radical intermediate in the reaction. The role of iodide was proposed to initiate the reaction through halogen exchange with a benzyl chloride to give a key intermediate benzyl iodide.[11e] Subsequently, the benzyl iodide underwent thermal decomposition to afford benzyl free-radical.[18] A benzyl iodide 1a′ took the place of 1a to react with 2a under normal conditions except that no TBAI was added to give 3aa in 89% yield (Scheme [3c]), a comparable result with the model reaction. In addition, the temperature had a decisive effect on the model reaction: a lower temperature (50 °C) resulted in no conversion of 1a and, under the same conditions, the free-radical scavenger galvinoxyl also did not intercept the benzyl radical. These results support the presence of benzyl iodide intermediate and the process of thermal decomposition of it in the reaction. A carbonylated product benzoic acid (25% yield) was obtained when the model reaction was performed in the absence of 1a (Equation 2), while in the absence of 2a no carbonylated product was observed (Equation 3) under normal conditions, implying that the carbonylation process starting from aryltrifluoroborate is reasonable. Besides these, under the standard conditions, 10–15% of 2-(3,4-dichlorophenyl)-1-phenylethan-1-one (according to GC-MS analysis) was observed throughout the model reaction; its concentration decreasing only toward the end of the carbonylation reaction. This result can be rationalized by a reaction sequence of in situ TBAI-catalyzed conversion of 1a into 2-(3,4-dichlorophenyl)-1-phenylethan-1-one followed by benzylation of the resulting ketone to give the desired product.

Equation 2

Equation 3

Combining all the results mentioned above and previous studies,[14d] [11e] [19] [20] we propose a mechanism for this TBAI-catalyzed carbonylation/benzylation reaction (Scheme [4]). This reaction was initiated by potassium aryltrifluoroborate reacting with carbon monoxide to generate a carbonylated species II, followed by carbonyl 1,2-migration insertion,[19] leading to a key intermediate III. Then, a carbonylated intermediate IV was formed by the insertion of carbon monoxide into III.[19] Benzyl chloride V underwent halogen exchange with iodide to give a benzyl iodide VI, which can thermally decompose into a benzyl free radical VII and an iodine free radical.[18] Subsequently, the active free radical VII was trapped by intermediate IV to provide 1,2-diaryl­ethanone intermediate IX along with [BCOF₃K] radical anion species VIII. Benzylation of IX would give the desired product X and the species VIII undertook single-electron transfer with an iodine free radical to regenerate iodide anion, which would re-enter the catalytic cycle.

Scheme 3 Control experiments conducted to probe the mechanism

Scheme 4 Proposed mechanism for the catalysis

In summary, we have developed the first examples of transition-metal free carbonylations of unactivated benzyl chlorides with stable potassium aryltrifluoroborates under ambient pressure of CO gas. This method represents a significant improvement over the previous palladium-­catalyzed carbonylation: the catalyst TBAI is abundant, inexpensive, and bench stable, the reaction conditions are much milder, and the use of ligands and the costly need to remove residual palladium from the end products are avoided. In consequence, this protocol provides an inexpensive and operationally simple method for the preparation of 1,2,3-triarylpropan-1-ones.

• References and Notes

• 1 Ishiyama T. Kizaki H. Miyaura N. Suzuki A. Tetrahedron Lett. 1993; 34: 7595
• 2a Maeyama K. Yamashita K. Saito H. Aikawa S. Yoshida Y. Polym. J. 2012; 44: 315
• 2b Wen A. Wang Z. Hang T. Jia Y. Zhang T. Wu Y. Gao X. Yang Z. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2007; 856: 348
• 2c Zhao WL. Carreira EM. Org. Lett. 2006; 8: 99
• 2d Ong AL. Kamaruddin AH. Bhatia S. Process Biochem. 2005; 40: 3526
• 2e Furusawa M. Ido Y. Tanaka T. Ito T. Nakaya K. Ibrahim I. Ohyama M. Iinuma M. Shirataka Y. Takahashi Y. Helv. Chim. Acta 2005; 88: 1048
• 2f Bosca F. Miranda MA. J. Photochem. Photobiol., B 1998; 43: 1
• 2g Dorman G. Prestwich GD. Biochemistry 1994; 33: 5661

For some recent reviews on Pd-catalyzed carbonylations of arylhalides, see:
• 3a Wu X.-F. Neumann H. Beller M. Chem. Rev. 2013; 113: 1
• 3b Wu X.-F. Neumann H. Beller M. Chem. Soc. Rev. 2011; 40: 4986
• 3c Grigg R. Mutton SP. Tetrahedron 2010; 66: 5515
• 3d Brennführer A. Neumann H. Beller M. Angew. Chem. Int. Ed. 2009; 48: 4114
• 3e Gadge ST. Bhanage BM. RSC Adv. 2014; 4: 10367
• 3f Fang WW. Zhu HB. Deng QY. Liu SL. Liu XY. Shen YJ. Tu T. Synthesis 2014; 46: 1689
• 4a Wu X.-F. Beller M. Transition Metal Catalyzed Carbonylation Reactions-Carbonylative Activation of C–X Bonds. Springer-Verlag; Berlin, Heidelberg: 2013
• 4b Kollár L. Modern Carbonylation Methods . Wiley-VCH, Verlag GmbH & Co. KgaA; Weinheim: 2008
• 5a Cardenas DJ. Angew. Chem. Int. Ed. 2003; 42: 384
• 5b Nishihara Y. In Applied Cross-Coupling Reactions . Springer; Berlin, Heidelberg: 2013: 1-247

For recent reviews on carbonylation of C(sp³) halides, see:
• 5c Liégault B. Renaud J.-L. Bruneau C. Chem. Soc. Rev. 2008; 37: 290
• 5d Wu LP. Fang XJ. Liu Q. Jackstell R. Beller M. Wu X.-F. ACS Catal. 2014; 4: 2977
• 6 Ishiyama T. Kizaki H. Hayashi T. Suzuki A. Miyaura N. J. Org. Chem. 1998; 63: 4726
• 7 Wu X.-F. Neumann H. Beller M. Tetrahedron Lett. 2010; 51: 6146
• 8 Noverges B. Medio-Simόn M. Asensio G. Adv. Synth. Catal. 2014; 356: 3649
• 9a Zanti G. Peeters D. Eur. J. Inorg. Chem. 2009; 3904
• 9b Welch CJ. Albaneze-Walker J. Leonard WR. Biba M. DaSilva J. Henderson D. Laing B. Mathre DJ. Spencer S. Bu X. Wang T. Org. Process Res. Dev. 2005; 9: 198
• 9c Usluer Ö. Abbas M. Wantz G. Vignau L. Hirsch L. Grana E. Brochon C. Cloutet E. Hadziioannou G. ACS Macro Lett. 2014; 3: 1134

Recent reviews on transition-metal-free processes, see:
• 10a Sun C.-L. Shi Z.-J. Chem. Rev. 2014; 114: 9219
• 10b Yanagisawa S. Itami K. ChemCatChem 2011; 3: 827
• 10c Shirakawa E. Hayashi T. Chem. Lett. 2012; 41: 130
• 10d Mehta VP. Punji B. RSC Adv. 2013; 3: 11957
• 11a Majek M. Jacobi von Wangelin A. Angew. Chem. Int. Ed. 2015; 54: 2270
• 11b Guo W. Lu L.-Q. Wang Y. Wang Y.-N. Chen J.-R. Xiao W.-J. Angew. Chem. Int. Ed. 2015; 54: 2265
• 11c Zhang H. Shi R. Ding A. Lu L. Chen B. Lei A. Angew. Chem. Int. Ed. 2012; 51: 12542
• 11d Jin FL. Han W. Chem. Commun. 2015; 9133
• 11e Jin FL. Zhong YZ. Zhang X. Zhang HC. Zhao Q. Han W. Green Chem. 2016; 18: 2598

For reviews, see:
• 12a Darses S. Genet J.-P. Chem. Rev. 2008; 108: 288
• 12b Doucet H. Eur. J. Org. Chem. 2008; 2013
• 12c Molander GA. Ellis NM. Acc. Chem. Res. 2007; 40: 275
• 12d Stefani HA. Cella R. Vieira AS. Tetrahedron 2007; 63: 3623
• 13a Wu X.-F. Neumann H. Beller M. Adv. Synth. Catal. 2011; 353: 788
• 13b Zhao HY. Han W. Eur. J. Org. Chem. 2016; 4279

For selected examples of radical carbonylative coupling of alkyl halidesand arylboronic acids (or arylboronic esters), see:
• 14a Zhao H.-Y. Feng Z. Luo ZJ. Zhang XG. Angew. Chem. Int. Ed. 2016; 55: 10401
• 14b Pye DR. Cheng L.-J. Mankad NP. Chem. Sci. 2017; 8: 4750
• 14c Roslin S. Odell LR. Chem. Commun. 2017; 6895
• 14d Sumino S. Ui T. Ryu I. Org. Lett. 2013; 15: 3142
• 14e Sumino S. Ui T. Ryu I. Org. Chem. Front. 2015; 2: 1085
• 15a Ryu I. Chem. Soc. Rev. 2001; 30: 16
• 15b Sumino S. Fusano A. Fukuyama T. Ryu I. Acc. Chem. Res. 2014; 47: 1563
• 15c Ryu I. Sonoda N. Angew. Chem. Int. Ed. Engl. 1996; 35: 1050

For recent important examples of radical carbonylations of alkyl halides, see:
• 15d Sargent BT. Alexanian EJ. J. Am. Chem. Soc. 2016; 138: 7520
• 15e Chow SY. Stevens MY. Åkerbladh L. Bergman S. Odell LR. Chem. Eur. J. 2016; 22: 9155
• 15f Yin HF. Skrydstrup T. J. Org. Chem. 2017; 82: 6474
• 15g Kobayashi S. Kawamoto T. Uehara S. Fukuyama T. Ryu I. Org. Lett. 2010; 12: 1548
• 15h Kawamoto T. Okada T. Curran DP. Ryu I. Org. Lett. 2013; 15: 2144
• 15i Fusano A. Fukuyama T. Nishitani S. Inouye T. Ryu I. Org. Lett. 2010; 12: 2410
• 15j McMahon CM. Renn MS. Alexanan EJ. Org. Lett. 2016; 18: 4148
• 15k Sumino S. Ui T. Hamada Y. Fukuyama T. Ryu I. Org. Lett. 2015; 17: 4952
• 15l Domański S. Staszewska-Krajewska O. Chaładaj W. J. Org. Chem. 2017; 82: 7998
• 15m Li YH. Zhu FX. Wang ZC. Rabeah J. Breckner A. Wu X.-F. ChemCatChem 2017; 9: 915
• 16a Zhu X. Wang Y.-F. Ren W. Zhang F.-L. Chiba S. Org. Lett. 2013; 15: 3214
• 16b Anstead GM. Altenbach RJ. Wilson SR. Katzenellenbogen JA. J. Med. Chem. 1988; 31: 1316
• 16c Wang ZY. Dufresne C. Leblanc Y. Li CS. Gauthier JY. Lau CK. Therien M. Roy P. US 6174874, 2001
• 16d Govek SP. Smith ND. ITL 159769, 2011
• 17 Typical Procedure for the Synthesis of 3aa: A 25 mL flask equipped with a magnetic stir bar was charged with potassium phenyltrifluoroborate 2a (0.5 mmol, 94.9 mg), TBAI (0.1 mmol, 37.6 mg), Na₂CO₃ (2.0 mmol, 213.0 mg), and PEG-400 (2 mL) before standard cycles of evacuation and back-filling with anhydrous and pure carbon monoxide. Benzyl chloride 1a (1.0 mmol, 141.4 μL) was added successively. The mixture was then stirred at 100 °C for 3 h. After being allowed to cool to room temperature, the reaction mixture was diluted with water (3 mL) and extracted with diethyl ether (4 × 5 mL). The organic phases were combined, and the volatile components were evaporated in a rotary evaporator. The residue was purified by column chromatography on silica gel (petroleum ether/diethyl ether, 100:1) to afford 3aa (185.7 mg, 88%) as a white solid (mp 119.6–120.0 °C). ¹H NMR (400 MHz, CDCl₃): δ = 7.85–7.83 (m, 2 H), 7.49 (tt, J = 7.4, 1.2 Hz, 1 H), 7.39–7.32 (m, 4 H), 7.24 (d, J = 8.2 Hz, 1 H), 7.20 (d, J = 2.0 Hz, 1 H), 7.04 (dd, J = 8.3, 2.1 Hz, 1 H), 6.87 (dd, J = 8.2, 2.1 Hz, 1 H), 4.70 (t, J = 7.3 Hz, 1 H), 3.45 (dd, J = 13.8, 7.6 Hz, 1 H), 2.96 (dd, J = 13.8, 7.0 Hz, 1 H). ¹³C NMR (100 MHz, CDCl₃): δ = 197.7, 139.2, 138.5, 135.9, 133.5, 133.2, 132.3, 131.8, 131.0, 130.9, 130.6, 130.3, 129.9, 128.8, 128.6, 128.6, 127.6, 54.4, 39.0
• 18 Miiller-Markgraf W. Troe J. J. Phys. Chem. 1988; 92: 4899
• 19 For a review on carbonylation of organoboranes, see: Brown HC. Acc. Chem. Res. 1969; 2: 65
• 20 Chen ZW. Jiang HF. Wang AZ. Yang SR. J. Org. Chem. 2010; 75: 6700