Boron-Catalyzed Carboxylic Acid-Selective Aldol Reaction with Trifluoromethyl Ketones

Abstract

A catalytic carboxylic acid-selective aldol reaction with trifluoromethyl ketones was developed. Reversible and selective covalent bond formation between a boron catalyst and a carboxylic acid is key to realizing the unprecedented catalytic aldol reaction of simple carboxylic acids. The reaction proceeded chemoselectively at the α-position of carboxylic acid even in the presence of ketone, ester, or amide functional groups in the donor substrates. The chemoselectivity is beneficial for late-stage derivatizations of biologically relevant compounds, as demonstrated by the conversion of indomethacin and triacetylcholic acid.

Fluorine-containing organic molecules are attractive for pharmaceutical development.^1–6) The introduction of fluorine atoms in place of hydrogen atoms considerably alter the physicochemical properties of compounds, mainly due to the high electronegativity and hydrophobic nature of fluorine atoms. Compounds containing fluorine exhibit improved metabolic stability, increased lipophilicity, and other unique properties favorable for pharmacokinetics and drug potencies. A trifluoromethyl group (CF₃) is a particularly attractive fluorinated functional group, and many drugs containing the CF₃ group have been developed, such as efavirenz, celecoxib, and sitagliptin. Incorporation of a CF₃ group has been intensively pursued over the past few decades through nucleophilic, electrophilic, and radical CF₃ active species.^7–11)

As an alternative approach to producing CF₃-containing value-added functional molecules, transformations of readily available CF₃-containing substrates are simple and practical.¹²⁾ The aldol reaction, among other approaches, exploiting trifluoromethyl ketones as electrophiles is an attractive transformation that provides biologically relevant α-CF₃ tertiary alcohols.^13–15) In this context, organocatalysts are highly effective catalysts for the introduction of ketones to trifluoromethyl ketones in asymmetric aldol reactions.^16–25) In contrast, the use of nucleophiles with a carboxylic acid oxidation state is scarcely developed, even for racemic aldol reactions with trifluoromethyl ketones. Moreover, almost all preceding studies relied on active methylene compounds (Chart 1a), such as isocyanoacetic acid derivatives,^26,27) ethyl diazoacetate,²⁸⁾ and malonic acid half esters,²⁹⁾ or pre-activated ketene silyl acetals^30–32) (Chart 1b). Because of the high pK_a values of the α-proton of simple carboxylic acid derivatives, a direct catalytic aldol reaction of non-activated carboxylic acid derivatives remained to be explored. A notable exception was reported by Kobayashi: a barium-catalyzed aldol reaction of an N-tert-butoxycarbonyl (Boc) acetoanilide with α,α,α-trifluoroacetophenone supplied the desired aldol product in high yield, although only one example is described.³³⁾ Herein, we report the first catalytic direct aldol reaction of carboxylic acids promoted by a boron catalyst (Chart 1c). The mild and chemoselective reaction conditions markedly expand the scope of the aldol reaction of carboxylic acid derivatives with trifluoromethyl ketones.

Chart 1. Catalytic Aldol Reaction of Carboxylic Acid Derivatives with Trifluoromethyl Ketones

We recently developed a carboxylic acid-selective asymmetric Mannich reaction and diastereoselective aldol reaction using BH₃·SMe₂ as a promoter.^34,35) The boron compound selectively recognizes and activates carboxylic acids via reversible covalent bond formation (Chart 1c: acyloxyborane formation),^36–42) and thus enables enolate formation from the carboxylic acid by mild organic base. Based on the Mannich reaction of carboxylic acids,³⁴⁾ we selected toluene as solvent and applied 10 mol% BH₃·SMe₂ for a reaction between propionic acid (1a) and α,α,α-trifluoroacetophenone (2a) (Table 1). The aldol reaction proceeded smoothly with 1 eq of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) to give the corresponding product 3aa in 86% yield (entry 2). Increasing the amount of DBU to 3 eq improved the reactivity to afford 3aa in 99% yield (entry 3), albeit with slightly decreased diastereoselectivity. When isolated syn-3aa was subjected to the same reaction conditions as entry 3 in the absence of 1a and 2a, syn-3aa was recovered in quantitative yield without changing the diastereomeric ratio. The result indicated that the observed diastereomeric ratio was kinetically-determined, and epimerization after C–C bond formation did not proceed.⁴³⁾ Less basic organic bases and inorganic bases were totally ineffective (entries 4–7). BCl₃ was found to be a similarly effective catalyst (entry 9) to BH₃·SMe₂, while BF₃·Et₂O did not promote the reaction at all (entry 8). Considering the relative strength of B–F, B–Cl, and B–O bonds,⁴⁴⁾ formation of the key acyloxyborane is not possible with BF₃·Et₂O, while BCl₃ would smoothly produce the active acyloxyborane species. Thus, the results suggest the importance of acyloxyborane formation for the facile generation of the carboxylic acid enolate. Similarly, B(OMe)₃ was not a suitable catalyst, probably due to the difficult substitution of the methoxy group with carboxylate (entry 10). Because neither pinacol borane (entry 11), catechol borane (entry 12), nor phenylboronic acid (entry 13) supplied the product, substituents of acyloxyborane strongly affected the reactivity of the aldol reaction. Since the reactivity is the most important factor for broadening the substrate scope, we determined that the conditions in entry 3 (10 mol% BH₃·SMe₂ and 3 eq DBU) as the optimal.⁴⁵⁾

Table 1. Optimization of Aldol Reaction

Entry	B source	Base	Yield (%)a)	syn/anti
1	—	DBU	0	—
2b)	BH3·SMe2	DBU	86	4.0/1
3	BH3·SMe2	DBU	99	2.1/1
4	BH3·SMe2	NEt3	0	—
5	BH3·SMe2	iPr2NEt	0	—
6	BH3·SMe2	K2CO3	0	—
7	BH3·SMe2	Cs2CO3	0	—
8	BF3·Et2O	DBU	0	—
9	BCl3	DBU	96	1.6/1
10	B(OMe)3	DBU	0	—
11	pinBH	DBU	0	—
12	catBH	DBU	0	—
13	PhB(OH)2	DBU	0	—

a) Yield was determined by ¹H-NMR analysis using tBuOMe as an internal standard. b) One equivalent of DBU was used.

With the optimized conditions in hand, the scope of applicable trifluoromethyl ketones was examined (Table 2). The reaction proceeded efficiently with electron-donating as well as electron-withdrawing groups at the para position of the aryl substituent (3ba, 3ca). Although the reaction rate was slower with electron-rich 1b, a prolonged reaction time (12 h) afforded product 3ba in 92% yield. A substituent at the ortho position slightly disrupted the reaction and product 3da was obtained in 62% yield. A thiophene group tolerated the conditions and 3ea was obtained in 98% yield. The reaction with highly electrophilic methyl trifluoropyruvate gave 3fa in moderate yield (36% yield) along with unidentified byproducts. Because of the existence of highly acidic α-proton in aliphatic trifluoromethyl ketones, an attempt to apply the optimized condition to 1,1,1-trifluoro-2-butanone (1g) failed.

Table 2. Substrate Scope of Carboxylic Acid Aldol Reaction with Trifluoromethyl Ketones

Isolated yield was determined after conversion of the aldol products into methyl esters by TMSCHN₂. a) Reaction time was 12 h. b) Thirty-three molecule percent of BH₃·SMe₂ was used. c) Two equivalents of DBU was used. d) Diastereomeric ratio was determined by HPLC analysis.

The generality of carboxylic acids was broad. Acetic acid (2b) reacted smoothly to produce 3ab in 69% yield. Carboxylic acids 2c and d containing C–C double and triple bonds, respectively, which are potentially susceptible to hydroboration of the multiple bonds, were also competent under the standard conditions to give the desired aldol products in good yield (3ac: 74% yield, 3ad: 66% yield). Because acyloxyborane formation is a carboxylic acid-selective process, the aldol reaction proceeded exclusively at the α-position of a carboxy group in the presence of amide and ester groups (3ae, 3af). When a ketone containing 2g was used under the standard conditions, a reaction at the α-position of ketone competed with the desired carboxylic acid aldol reaction. Complete carboxy group selectivity was attained by increasing the catalyst loading to 33 mol% and decreasing the amount of DBU to 2 eq to afford desired product 3ag in 80% yield. Because BH₃·SMe₂ contains three reactive hydrides, all the carboxylic acid was activated as the acyloxyborane using 33 mol% of BH₃·SMe₂, which led to high carboxylic acid-selectivity. The carboxy group-selective reaction enabled the transformation of an acidic amino acid ester, benzyloxycarbonyl (Cbz)-glutamic acid (Glu)-OMe (2h), at the side-chain in good yield. Furthermore, the method could be applied to biologically relevant compounds. Indomethacin (2i), an anti-inflammatory drug, produced corresponding product 3ai in 63% yield and syn/anti=5.3/1. A steroidal triacetylcholic acid was also successfully transformed to aldol product 3aj with complete chemoselectivity in 81% yield. Thus, the method can be used for late-stage derivatization of drugs and drug-like compounds.

In summary, we developed the first boron-catalyzed carboxylic acid aldol reaction with trifluoromethyl ketones as electrophiles. The reversible covalent bond formation between the carboxy group and boron catalyst is key to the high carboxylic acid-selectivity: the reaction proceeded exclusively at the α-position of the carboxy group, even in the presence of an amide, ester, or ketone group. Further investigation for catalytic asymmetric aldol reactions of aldehydes and mechanistic studies are ongoing in our laboratory.

Acknowledgments

This work was partially supported by a Grant-in-Aid for Scientific Research (A) from the Japan Society for the Promotion of Science (JSPS) (MK), a Grant-in-Aid for Scientific Research (C) from the JSPS (YS), and a TORAY Award in Synthetic Organic Chemistry, Japan (YS). We thank K. Aoki for his contribution to preliminary experiments of the carboxylic acid aldol reaction.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

The online version of this article contains supplementary materials.

References and Notes

• 1) Purser S., Moore P. R., Swallow S., Gouverneur V., Chem. Soc. Rev., 37, 320–330 (2008).
• 2) Kirk K. L., J. Fluor. Chem., 127, 1013–1029 (2006).
• 3) Zhou Y., Wang J., Gu Z., Wang S., Zhu W., Aceña J. L., Soloshonok V. A., Izawa K., Liu H., Chem. Rev., 116, 422–518 (2016).
• 4) Gillis E. P., Eastman K. J., Hill M. D., Donnelly D. J., Meanwell N. A., J. Med. Chem., 58, 8315–8359 (2015).
• 5) Zanda M., New J. Chem., 28, 1401–1411 (2004).
• 6) Jeschke P., ChemBioChem, 5, 570–589 (2004).
• 7) Kirk K. L., Org. Process Res. Dev., 12, 305–321 (2008).
• 8) Yang X., Wu T., Phipps R. J., Toste F. D., Chem. Rev., 115, 826–870 (2015).
• 9) Liu X., Xu C., Wang M., Liu Q., Chem. Rev., 115, 683–730 (2015).
• 10) Ma J. A., Cahard D., J. Fluor. Chem., 128, 975–996 (2007).
• 11) Alonso C., Marigorta E. M., Rubiales G., Palacios F., Chem. Rev., 115, 1847–1935 (2015).
• 12) Nie J., Guo H.-C., Cahard D., Ma J.-A., Chem. Rev., 111, 455–529 (2011).
• 13) Schäcke H., Schottelius A., Döcke W.-D., Strehlke P., Jaroch S., Schmees N., Rehwinkel H., Hennekes H., Asadullah K., Proc. Natl. Acad. Sci. U.S.A., 101, 227–232 (2004).
• 14) Young S. D., Britcher S. F., Tran L. O., Payne L. S., Lumma W. C., Lyle T. A., Huff J. R., Anderson P. S., Olsen D. B., Carroll S. S., Pettibone D. J., O’Brien J. A., Ball R. G., Balani S. K., Lin J. H., Chen I.-W., Schleif W. A., Sardana V. V., Long W. J., Byrnes V. W., Emini E. A., Antimicrob. Agents Chemother., 39, 2602–2605 (1995).
• 15) Sani M., Belotti D., Giavazzi R., Panzeri W., Volonterio A., Zanda M., Tetrahedron Lett., 45, 1611–1615 (2004).
• 16) Wang X.-J., Zhao Y., Liu J.-T., Org. Lett., 9, 1343–1345 (2007).
• 17) Hara N., Tamura R., Funabhashi Y., Nakamura S., J. Org. Chem., 13, 1662–1665 (2011).
• 18) Duangdee N., Harnying W., Rulli G., Neudörfl J.-M., Gröger H., Berkessel A., J. Am. Chem. Soc., 134, 11196–11205 (2012).
• 19) Zheng Y., Xiong H.-Y., Nie J., Hua M.-Q., Ma J. A., Chem. Commun., 48, 4308–4310 (2012).
• 20) Kokotos C. G., J. Org. Chem., 77, 1131–1135 (2012).
• 21) Lin J., Kang T., Liu Q., He L., Tetrahedron Asymmetry, 25, 949–955 (2014).
• 22) Zong H., Huang H., Bian G., Song L., J. Org. Chem., 79, 11768–11773 (2014).
• 23) Yang W., Cui Y.-M., Zhou W., Li L., Yang K.-F., Zheng Z.-J., Lu Y., Xu L.-W., Synlett, 25, 1461–1465 (2014).
• 24) Lutete L. M., Miyamoto T., Ikemoto T., Tetrahedron Lett., 57, 1220–1223 (2016).
• 25) Jing Z., Bai X., Chen W., Zhang G., Zhu B., Jiang Z., Org. Lett., 18, 260–263 (2016).
• 26) Soloshonok V. A., Hayashi T., Ishikawa K., Nagashima N., Tetrahedron Lett., 35, 1055–1058 (1994).
• 27) Soloshonok V. A., Kacharov A. D., Avilov D. V., Ishikawa K., Nagashima N., Hayashi T., J. Org. Chem., 62, 3470–3479 (1997).
• 28) Cui H.-F., Wang L., Yang L.-J., Nie J., Zheng Y., Ma J.-A., Tetrahedron, 67, 8470–8476 (2011).
• 29) Li X.-J., Xiong H.-Y., Hua M.-Q., Nie J., Zheng Y., Ma J.-A., Tetrahedron Lett., 53, 2117–2120 (2012).
• 30) Song J. J., Tan Z., Reeves J. T., Yee N. K., Senanayake C. H., Org. Lett., 9, 1013–1016 (2007).
• 31) Song J. J., Tan Z., Reeves J. T., Fandrick D. R., Yee N. K., Senanayake C. H., Org. Lett., 10, 877–880 (2008).
• 32) Chintareddy V. R., Wadhwa K., Verkade J. G., J. Org. Chem., 74, 8118–8132 (2009).
• 33) Saito S., Kobayashi S., J. Am. Chem. Soc., 128, 8704–8705 (2006).
• 34) Nagai H., Morita Y., Shimizu Y., Kanai M., Org. Lett., 18, 2276–2279 (2016).
• 35) Morita Y., Yamamoto T., Nagai H., Shimizu Y., Kanai M., J. Am. Chem. Soc., 137, 7075–7078 (2015).
• 36) Furuta K., Miwa Y., Iwanaga K., Yamamoto H., J. Am. Chem. Soc., 110, 6254–6255 (1988).
• 37) Maki T., Ishihara K., Yamamoto H., Tetrahedron, 63, 8645–8657 (2007).
• 38) Al-Zoubi R. M., Marion O., Hall D. G., Angew. Chem. Int. Ed., 47, 2876–2879 (2008).
• 39) Zheng H., McDonald R., Hall D. G., Chem. Eur. J., 16, 5454–5460 (2010).
• 40) Azuma T., Murata A., Kobayashi Y., Inokuma T., Takemoto Y., Org. Lett., 16, 4256–4259 (2014).
• 41) Hayama N., Azuma T., Kobayashi Y., Takemoto Y., Chem. Pharm. Bull., 64, 704–717 (2016).
• 42) Ishihara K., Lu Y., Chem. Sci., 7, 1276–1280 (2016).
• 43) See Supplementary Materials for details.
• 44) Bond enthalpies of B–F (BF₃), B–Cl (BCl₃), and B–O (B(OMe)₃) were 644.3, 442.3, and 519.3 kJ/mol, respectively. Holbrook J. B., Smith B. C., Housecroft C. E., Wade K., Polyhedron, 1, 701–706 (1982).
• 45) Addition of ligands did not improve the diastereoselectivity. Ts-L-Val (10 mol%): 79% yield, syn-3aa : anti-3aa=1.2 : 1; (S)-3,3′-I₂-BINOL (10 mol%): 36% yield, syn-3aa : anti-3aa=2.3 : 1.