Letter pubs.acs.org/OrgLett
Direct Palladium-Catalyzed Carbonylative Transformation of Allylic Alcohols and Related Derivatives Fu-Peng Wu,† Jin-Bao Peng,*,† Lu-Yang Fu,† Xinxin Qi,† and Xiao-Feng Wu*,†,‡ †
Department of Chemistry, Zhejiang Sci-Tech University, Hangzhou, Zhejiang 310018, People’s Republic of China Leibniz-Institut für Katalyse e.V. an der, Institution Universität Rostock, Albert-Einstein-Straße 29a, Rostock 18059, Germany
‡
S Supporting Information *
ABSTRACT: A direct, palladium-catalyzed, carbonylative transformation of allylic alcohols for the synthesis of β,γunsaturated carboxylic acids has been developed. With formic acid as the CO source, various allylic alcohols were conveniently transformed into the corresponding β,γ-unsaturated carboxylic acids with excellent linear and (E)-selectivity. The reaction was performed under mild conditions; toxic CO gas manipulation and high-pressure equipment were avoided in this procedure.
T
alcohols in the presence of phenols to afford the corresponding β,γ-unsaturated esters.5a Soon after, Alper and Xiao reported a novel thiocarbonylation reaction with thiols as nucleophiles to afford β,γ-unsaturated thioesters.5b Recently, Beller and coworkers developed an elegant carbonylation method of allylic alcohols with aliphatic alcohols and aromatic amines to produce β,γ-unsaturated esters5c and amides. 5d However, these procedures require the use of toxic carbon monoxide gas, and special equipment is usually needed for handling high-pressure gas, which make them not ideal for benchtop synthesis. To circumvent the problem of carbon monoxide manipulation and develop user-friendly carbonylation methods, several research groups, including ours, have developed many efficient carbonylation reactions based on CO surrogates.6 Recently, we have reported a general carboxylation reaction of aryl halides using formic acid as the CO source. Based on our continuous interest in palladium-catalyzed carbonylation reactions with formic acid as a green CO source, we recently became interested in developing a direct carbonylation of allylic compounds with formic acid. We assume that formic acid would act in dual roles, both as a CO source and as a source of the hydroxyl group. Initially, cinnamyl alcohol was selected as a model substrate. To our delight, after a solution of cinnamyl alcohol, formic acid, and DCC was stirred in the presence of Pd(OAc)2 and Xantphos in 1,4-dioxane at 50 °C, the hydroxylation reaction proceeded successfully and the desired product (E)-4-phenylbut-3-enoic acid was isolated in 46% yield (Table 1, entry 1). Notably, no (Z)-isomer was detected in this transformation. Mechanistically, base was not required for this reaction, but the acidity of the solution might have some important influence on catalytic cycle. Thus, we also tested the effect of bases.
he transition-metal-catalyzed carbonylation reaction has emerged as a versatile and powerful tool for the formation of various carbonyl compounds.1 A variety of efficient carbonylation reactions have been developed during the past two decades. The majority of the previous works were focused on the carbonylation of organic (pseudo)halides, which are of great interest for both the academic and the industrial community. However, the development of carbonylation of alcohols, which are much more readily available, has attracted less attention and remains a challenging goal. In this respect, allylic alcohol has shown great potential due to its specialized structure and reactivity. Compared to the flourishing development of transition-metal-catalyzed nucleophilic substitution reactions,2 there are only a few investigations of the carbonylation reaction of allylic compounds. In addition, carbonylation of allylic compounds represents a direct and general method for the synthesis of the valuable synthetic building blocks, β,γ-unsaturated carboxylic acids.3 Typically, an activated allylic substrate with a good leaving group (i.e., allylic sulfonates, acetates, carbonates, or halogens) is used in these reactions.4 In spite of their good reactivity, these substrates usually suffer two main drawbacks: first, multistep preinstallation of the leaving group was usually required for the preparation of these substrates, thus reducing the general efficiency. Second, stoichiometric amounts of wastes were generated with the release of the leaving group, which also might influence the subsequent reactions. The direct use of unprotected allylic alcohol as the substrates turns out to be a very economical and environmentally friendly method since it gives water as the sole byproduct. Unfortunately, the high activation barrier required for C−OH bond-breaking and the nucleophilicity of the hydroxyl group have hindered allylic alcohols from their application in carbonylation reactions. As a result, there are very few reports on this subject.5 In 1997, the Miura group demonstrated the direct carbonylation of allyl © 2017 American Chemical Society
Received: September 8, 2017 Published: September 18, 2017 5474
DOI: 10.1021/acs.orglett.7b02801 Org. Lett. 2017, 19, 5474−5477
Letter
Organic Letters Table 1. Optimization of the Reaction Conditionsa
Table 2. Substrate Scope of the Carboxylation Reaction: Primary Allylic Alcoholsa
entry
ligand
base
temp (°C)
yieldb (%)
1 2 3 4 5c 6c 7 8 9d 10e 11f
Xantphos Xantphos Xantphos Xantphos PPh3 P(o-tolyl)3 Xantphos Xantphos Xantphos Xantphos Xantphos
none NaHCO3 HCO2Na Et3N NaHCO3 NaHCO3 NaHCO3 NaHCO3 NaHCO3 NaHCO3 NaHCO3
50 50 50 50 50 50 60 70 60 60 60
46 53 35 trace 8 trace 67 63 59 76 58
a
Reaction conditions: cinnamyl alcohol (0.5 mmol), formic acid (3.5 mmol), Pd(OAc)2 (3 mol %), Xantphos (3 mol %), base (1.0 mmol), 4 Å MS (80 mg). DCC (1.0 mmol), 1,4-dioxane (2 mL), 20 h. b Isolated yield. c6 mol %. d0.5 mmol of NaHCO3 was used. e1.5 mmol of NaHCO3 was used. fXantphos (4.5 mol %).
Surprisingly, the addition of sodium bicarbonate increased the yield to 53% (Table 1, entry 2), but sodium formate resulted in lower yield (Table 1, entry 3). Organic base nearly stopped the reaction, and only trace amount of the product was detected (Table 1, entry 4). Subsequently, various phosphine ligands were investigated. On the basis of our previous work on the ligand-controlled carbonylation of aryl halides, bidentate phosphine ligand with a large bite angle facilitated the hydroxycarbonylation to give the acid product.6i Indeed, triphenylphosphine showed much lower reactivity and gave a poor yield of acid (Table 1, entry 5). Monodentate ligands such as substituted triphenylphosphines and other bidentate ligands with small bite angles like DPPP and DPPF were all ineffective for this transformation (Table 1, entry 6 and also see details in Supporting Information). It should be mentioned that when monodentate phosphines were used as the ligand, formylation6i product was not detected; instead, cinnamyl formate was obtained in 10−30% yields, which indicates that the nature of ligands has a significant influence in the oxidative addition step of the palladium to the C−O bond. The reaction temperature also played an important role. When the reaction was performed at 60 °C, the yield was improved to 67% (Table 1, entry 7). However, increasing the temperature to 70 °C resulted in a lower yield (Table 1, entry 8). Further screening of the additives revealed that acid additives decreased the yield (see details in the SI). The best result was obtained when 3 equiv of sodium bicarbonate was added, and the desired product was isolated in 76% yield (Table 1, entry 10). Finally, when we tested the influence of the ratio of palladium and phosphine ligand, higher ligand loading reduced the yield (Table 1, entry 11). With the optimized reaction conditions in hand (Table 1, entry 10), we investigated the substrates scope of the reaction with various substituted allylic alcohols. First, as summarized in Table 2, a series of primary allylic alcohols was successfully applied to the carboxylation reaction. The steric and electronic property of the substitution on the benzene ring does not affect the yield and selectivity of the reaction. Both electron-donating group (Table 2, entries 2−5) and electron-withdrawing group (Table 2, entries 6−9) substituted cinnamyl alcohols were all well tolerated and delivered the corresponding acids in 64−76%
a
Reaction conditions: allylic alcohol (0.5 mmol), formic acid (3.5 mmol), Pd(OAc)2 (3 mol %), Xantphos (3 mol %), NaHCO3 (1.5 mmol), 4 Å MS (80 mg). DCC (1.0 mmol), 1,4-dioxane (2 mL), 60 °C, 20 h, isolated yield.
yields. All of the reactions proceeded regio- and stereoselectively. No branched or (Z)-isomers were detected in these reactions. In addition, heteroaryl- and 2-naphthyl-substituted allylic alcohols were also subjected to the carboxylation reaction, and the corresponding products were conveniently generated in good yields with excellent selectivity (Table 2, entries 10−12). However, when alkyl substituted allylic alcohol was used in this reaction, the desired product β,γ-unsaturated carboxylic acid was obtained in moderate yield with a decreased stereoselectivity (E/Z = 8:1, Table 2, entry 13). We then turned our attention to test the generality of the carboxylation reaction with different secondary and tertiary allylic alcohols. As demonstrated in Table 3, various aryl- and heteroaryl-substituted secondary allylic alcohols were subjected to the optimized reaction conditions, and the carboxylation 5475
DOI: 10.1021/acs.orglett.7b02801 Org. Lett. 2017, 19, 5474−5477
Letter
Organic Letters
were delivered in moderate yield with lower E/Z selectivity (Table 3, entries 13 and 14). It should be mentioned that the secondary and tertiary allylic alcohols can be easily prepared from the nucleophilic addition of alkenyl metal reagent to aldehydes and ketones. Thus, this procedure provides a twostep alternative method for the synthesis of β,γ-unsaturated carboxylic acids from aldehydes and ketones by introducing a C3 unit. We further investigated the carboxylation of other allylic substrates, as shown in Table 4. When activated cinnamyl
Table 3. Substrate Scope of the Carboxylation Reaction: Secondary and Tertiary Allylic Alcoholsa
Table 4. Investigation of Other Leaving Groupsa
entry
X
yield (%)
E/Z
1 2 3 4 5
OAc OBoc OMe Cl NMe2
75 66 18 79 14
>99:1 >99:1 >99:1 >99:1 >99:1
a
Reaction conditions: allylic compounds (0.5 mmol), formic acid (3.5 mmol), Pd(OAc)2 (3 mol %), Xantphos (3 mol %), NaHCO3 (1.5 mmol), 4 Å MS (80 mg). DCC (1.0 mmol), 1,4-dioxane (2 mL), 60 °C, 20 h, isolated yield.
acetate and carbonate were employed as the substrate, the carboxylation reaction proceeded effectively and gave yields and selectivity similar to those of the unprotected alcohol (Table 4, entries 1 and 2). However, when cinnamyl methyl ether was applied in this reaction, owing to the poor leaving ability of the methoxyl group, only 18% yield of product was generated (Table 4, entry 3). This also indicates that formic acid did play an activating rule by converting the hydroxyl group to its formate. Although cinnamyl chloride gave a slightly higher yield than cinnamyl alcohol, as stated above, its availability might be a problem which restricts the substrate scope. Interestingly, when cinnamyl amine was subjected to the optimized condition, the carboxylation reaction proceeded as well, albeit with a low yield, and no aminocarbonylation product was observed. On the basis of these results and previous reports, a possible reaction mechanism is proposed in Scheme 1. First, allylic Scheme 1. Plausible Reaction Mechanism a
Reaction conditions: allylic alcohol (0.5 mmol), formic acid (3.5 mmol), Pd(OAc)2 (3 mol %), Xantphos (3 mol %), NaHCO3 (1.5 mmol), 4 Å MS (80 mg). DCC (1.0 mmol), 1,4-dioxane (2 mL), 60 °C, 20 h, isolated yield.
reaction proceeded smoothly. Interestingly, in all of these reactions, the double bond had migrated and linear (E)-β,γunsaturated carboxylic acids were isolated in moderate to good yields (Table 3, entries 1−11). No branched isomer was observed. This might be caused by the stereoisomerization of the allylic palladium complex after the oxidative addition of Pd0 to the C−O bond. Similar to the primary allylic alcohol, cyclohexyl-substituted allylic alcohol gave a lower yield and decreased E/Z selectivity (Table 3, entry 12). In addition, tertiary allylic alcohols were also tolerated in this transformation, and trisubstituted β,γ-unsaturated carboxylic acids 5476
DOI: 10.1021/acs.orglett.7b02801 Org. Lett. 2017, 19, 5474−5477
Letter
Organic Letters alcohol and relative derivatives 1 or 1′ reacted with Pd0 to give an allylic palladium complex 4, which thermodynamically favors the linear configuration 5 owing to the tremendous steric effect of the bisphosphine ligated palladium structure. Subsequently, coordination and insertion of 5 with CO, which was generated in situ from the reaction of formic acid with DCC, and X ligand exchange of Pd−X with formic acid afforded the acylpalladium formic acid complex 7. Reductive elimination delivered the formic anhydride 8 and regenerated Pd0 for the next catalyst cycle. The formic anhydride 8 then decomposed to generate the product 2 and release CO at the same time. In conclusion, a palladium-catalyzed direct carbonylative transformation of unactivated allylic alcohol with formic acid has been established. With formic acid as the CO source, a wide range of primary, secondary, and tertiary allylic alcohols were transformed into the corresponding β,γ-unsaturated carboxylic acids in good yields. The reaction was conducted in a userfriendly manner in which no gas manipulations were required, and excellent regio- and stereoselectivity was obtained.
■
(4) (a) Loh, T.-P.; Cao, G.-Q.; Yin, Z. Tetrahedron Lett. 1999, 40, 2649−2652. (b) Miyazawa, M.; Yamamoto, K. Chem. Lett. 1994, 23, 491−494. (c) Troisi, L.; Granito, C.; Rosato, F.; Videtta, V. Tetrahedron Lett. 2010, 51, 371−373. (d) Imada, Y.; Shibata, O.; Murahashi, S. I. J. Organomet. Chem. 1993, 451, 183−194. (e) Tommasi, S.; Perrone, S.; Rosato, F.; Salomone, A.; Troisi, L. Synthesis 2012, 44, 423−430. (5) (a) Satoh, T.; Ikeda, M.; Kushino, Y.; Miura, M.; Nomura, M. J. Org. Chem. 1997, 62, 2662−2664. (b) Xiao, W.-J.; Alper, H. J. Org. Chem. 1998, 63, 7939−7944. (c) Liu, Q.; Wu, L.; Jiao, H.; Fang, X.; Jackstell, R.; Beller, M. Angew. Chem., Int. Ed. 2013, 52, 8064−8068. (d) Li, H.; Neumann, H.; Beller, M. Chem. - Eur. J. 2016, 22, 10050− 10056. (e) Fu, M.-C.; Shang, R.; Cheng, W.-M.; Fu, Y. Chem. - Eur. J. 2017, 23, 8818−8822. (6) (a) Cacchi, S.; Fabrizi, G.; Goggiamani, A. Org. Lett. 2003, 5, 4269−4272. (b) Korsager, S.; Taaning, R. H.; Skrydstrup, T. J. Am. Chem. Soc. 2013, 135, 2891−2894. (c) Morimoto, T.; Kakiuchi, K. Angew. Chem., Int. Ed. 2004, 43, 5580−5588. (d) Wu, L.; Liu, Q.; Jackstell, R.; Beller, M. Angew. Chem., Int. Ed. 2014, 53, 6310−6320. (e) Gautam, P.; Bhanage, B. M. Catal. Sci. Technol. 2015, 5, 4663− 4702. (f) Qi, X.; Jiang, L.-B.; Li, C.-L.; Li, R.; Wu, X.-F. Chem. - Asian J. 2015, 10, 1870−1873. (g) Qi, X.; Li, C.-L.; Jiang, L.-B.; Zhang, W.-Q.; Wu, X.-F. Catal. Sci. Technol. 2016, 6, 3099−3107. (h) Peng, J.-B.; Wu, F.-P.; Li, C.-L.; Qi, X.; Wu, X.-F. Eur. J. Org. Chem. 2017, 2017, 1434− 1437. (i) Wu, F.-P.; Peng, J.-B.; Meng, L.-S.; Qi, X.; Wu, X.-F. ChemCatChem 2017, 9, 3121−3124. (j) Seo, Y.-S.; Kim, D.-S.; Jun, C.H. Chem. - Asian J. 2016, 11, 3508−3512.
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.7b02801. Experimental details and characterization data (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. ORCID
Xiao-Feng Wu: 0000-0001-6622-3328 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are thankful for financial support from NSFC (21472174, 21602201, 21602204), the Education Department of Zhejiang Province (Y201636555), Zhejiang Sci-Tech University (16062095-Y), and Zhejiang Natural Science Fund for Distinguished Young Scholars (LR16B020002)
■
REFERENCES
(1) (a) Beller, M. Applied Homogeneous Catalysis with Organometallic Compounds, 2nd ed.; Wiley-VCH: Weinheim, 2002; Vol. 1, pp 145− 156. (b) Brennführer, A.; Neumann, H.; Beller, M. Angew. Chem., Int. Ed. 2009, 48, 4114−4133. (c) Gabriele, B.; Mancuso, R.; Salerno, G. Eur. J. Org. Chem. 2012, 2012, 6825−6839. (d) Peng, J.-B.; Qi, X.; Wu, X.-F. Synlett 2017, 28, 175−194. (e) Wu, X.-F. RSC Adv. 2016, 6, 83831−83837. (f) Peng, J.-B.; Qi, X.; Wu, X.-F. ChemSusChem 2016, 9, 2279−2283. (2) For selected reviews, see: (a) Sundararaju, B.; Achard, M.; Bruneau, C. Chem. Soc. Rev. 2012, 41, 4467−4483. (b) Lumbroso, A.; Cooke, M. L.; Breit, B. Angew. Chem., Int. Ed. 2013, 52, 1890−1932. (c) Butt, N. A.; Zhang, W. Chem. Soc. Rev. 2015, 44, 7929−7967. (3) (a) Houk, K. N. Chem. Rev. 1976, 76, 1−74. (b) Liu, J.; Liu, Q.; Franke, R.; Jackstell, R.; Beller, M. J. Am. Chem. Soc. 2015, 137, 8556− 8563. (c) Michigami, K.; Mita, T.; Sato, Y. J. Am. Chem. Soc. 2017, 139, 6094−6097. (d) van Gemmeren, M.; Börjesson, M.; Tortajada, A.; Sun, S.-Z.; Okura, K.; Martin, R. Angew. Chem., Int. Ed. 2017, 56, 6558−6562. 5477
DOI: 10.1021/acs.orglett.7b02801 Org. Lett. 2017, 19, 5474−5477
