Solvent-free reduction of carboxylic acids to alcohols with NaBH₄ promoted by 2,4,6-trichloro-1,3,5-triazine and PPh₃ in the presence of K₂CO₃

Abstract

A simple, rapid, and eco-friendly method for NaBH₄ reduction of carboxylic acids to alcohols under solvent-free conditions was developed using a combination of 2,4,6-trichloro-1,3,5-triazine (TCT) with a catalytic amount of triphenylphosphine as an acid activator. With the 1 : 0.2 : 1.5 : 2 mole ratio of TCT : PPh₃ : K₂CO₃ : NaBH₄, carboxylic acids including aromatic acids, aliphatic acids, and N-protected α-amino acids (Fmoc, Z) could readily undergo reduction to give the corresponding alcohols in good to excellent yields within 10 min.

---

Sodium borohydride (NaBH₄) is a versatile reducing agent possessing widespread applications in organic synthesis.¹ Due to its low cost, high stability, and ease of handling, NaBH₄ and its polymer bound analogs are routinely used in the conversion of aldehydes and ketones to the respective alcohols.² Other functional groups such as acyl chlorides, esters, and imines could also undergo reduction under typical room temperature reaction conditions.³ Nevertheless, carboxylic acids are generally inert and cannot be reduced directly without activation.

So far a number of methods have been developed particularly in attempts to convert carboxylic acids into alcohols with NaBH₄.⁴ Carboxylic acids are commonly pre-activated via an in situ formation of active species such as acyl halides, mixed anhydrides, or active esters prior to borohydride reduction. Acid activators such as BF₃·Et₂O,^3a cyanuric fluoride,⁵ BOP reagent,⁶ 2,4,6-trichloro-1,3,5-triazine (TCT)/N-methylmorpholine (NMM),⁷ 1,1′-carbonyldiimidazole,⁸ sulfonylbenzotriazole derivatives,^4c,9 3,4,5-trifluorophenylboronic acids,¹⁰ and 1-propanephosphonic acid cyclic anhydride^4h have been applied in combination with NaBH₄ to achieve the direct reduction of carboxylic acids under mild conditions. Nevertheless, the methods still suffer from some of these limitations including the use of expensive reagents in excessive amount, long reaction times, and difficulty in work-up.

Recently, due to the awareness of environmental problems, considerable efforts have been made toward the synthesis under solvent-free conditions. The methods are not only of interest from ecological point of view, but in many cases also offer several synthetic advantages in terms of yield, selectivity, and simplicity of the reaction procedure. For the reduction with NaBH₄, a number of solvent-free methods have been reported mostly for reduction of carbonyl compounds.^2c,11 However, the protocol for carboxylic acids has yet to be reported.

Our interest in simple, low cost, and low environmental impact protocols has led us to develop a facile and efficient solvent-free method for direct reduction of carboxylic acids to alcohols with NaBH₄ using cheap and readily available TCT as an acid activator. Although TCT has previously been applied in combination with tertiary amines such as NMM^7b,12 and triethylamine,¹³ the acid activation step often requires cooling to 0 °C to avoid decomposition of the formed quaternary N-triazinylammonium chlorides intermediates.¹⁴ Such conditions, however, are not easily amendable under solvent-free conditions.

Since triphenylphosphine (PPh₃) is considered a phosphorus analog of tertiary amine exhibiting high stability and good nucleophilicity.¹⁵ It was envisaged that PPh₃ could be used to activate TCT providing a more stable triazinylphosphonium intermediate which enables the acid activation to be carried out at room temperature in an absence of organic solvent.

In our preliminary studies, solvent-free reduction of benzoic acid with NaBH₄ promoted by TCT-PPh₃ system was investigated as a model reaction. The reduction was carried out by grinding using mortar and pestle at room temperature, while TLC was used to follow the progress of the reaction. The effect of the amount of reagents used was first examined using potassium carbonate which was inert toward TCT as base. The reaction time was kept constant for the ease of comparison. Typically, a specified amount of PPh₃, benzoic acid (1 equiv.), and K₂CO₃ (1.5 equiv.) were added to TCT (1 equiv.). Since the reactant and reagents are all solid, a few drops of CH₂Cl₂ were added to the mixture to facilitate homogeneous mixing and grinding. After grinding for 5 min in which TLC showed completed disappearance of the acid, NaBH₄ (2 equiv.) was then added with continuous grinding for further 5 min. The product was isolated by column chromatography.

According to Table 1, it was found that in the absence of PPh₃ (entry 1), no appreciable amount of the corresponding alcohol was detected. Using 10 mol% of PPh₃ gave the expected alcohol in 59% yield (entry 2). When the amount of PPh₃ was increased to 20 mol% (entry 3), the yield of the reaction improved significantly and this amount was further applied as the optimal value for PPh₃. Since the three chloride atoms on the triazine ring of TCT are known to be reactive toward nucleophiles,¹⁶ it seems possible to perform the acid activation using TCT in less than stoichiometric amount. However, it was found that upon decreasing the amount of TCT from 1 to 0.66 (entry 4), the reaction yield dramatically reduced. This result implied that monoacylated triazine was the key intermediate in reacting with NaBH₄. To determine the effect of the counterion of the carbonate base on the outcome of the reaction, the reduction was carried out using 1 equiv. of TCT and 20 mol% of PPh₃ with cesium, sodium, and calcium carbonate. Since solubility of carbonate bases in organic solvent increases as the ionic radius of the metal within a group increases,¹⁷ Cs₂CO₃ is thus expected to be the most effective base, followed by K₂CO₃, Na₂CO₃, and CaCO₃, respectively. However, as shown in Table 1, using Cs₂CO₃ gave relatively low yield of benzyl alcohol (entry 5), while the uses of Na₂CO₃ and CaCO₃ led to complex mixtures with significant amount of starting materials remained (entries 6–7). It is thus possible that Cs₂CO₃ may be too reactive under the applied condition leading to partial decomposition of the generated intermediates before subsequence reduction.

Table 1 Optimization of benzoic acid reduction with NaBH₄ promoted by TCT–PPh₃^a

Entry	TCT (equiv.)	PPh3 (equiv.)	Base	% Yield
a All reactions were carried out with benzoic acid (0.271 mmol), TCT, PPh3, base (0.406 mmol), and NaBH4 (0.541 mmol). nd = not determined.
1	1	0	K2CO3	—
2	1	0.1	K2CO3	59
3	1	0.2	K2CO3	90
4	0.66	0.2	K2CO3	39
5	1	0.2	Cs2CO3	34
6	1	0.2	Na2CO3	nd
7	1	0.2	CaCO3	nd

---

We next turned our attention to investigate substrate compatibility to obtain the scope and generality of the method. The best reaction condition (Table 1, entry 3) was applied with aromatic acids, aliphatic acids, and N-protected α-amino acids.¹⁸ For direct comparison, the reaction time again was fixed equally for all substrates without further optimization, except for the less reactive substrates. After product isolation, ¹H NMR, ¹³C NMR and GC-MS data were recorded and compared with those reported in literature to confirm product formation.

According to Table 2, aromatic carboxylic acids especially benzoic acid and its electron-rich derivatives reduced readily to provide the expected alcohols in good to excellent yields (entries 1–6). Only the case of 3-(dimethylamino)benzoic acid, the product was obtained in moderate yield (entry 7). Benzoic acid derivatives containing halogen substituents (Cl or I) gave the corresponding products in slightly lower yields (entries 8–10), while poor results were obtained with the less reactive substrates having strong electron-withdrawing NO₂ group (entries 11 and 12). Obviously, longer time is required for activation of the less reactive 4-nitrobenzoic acid since the yield of the alcohol increased with prolonged acid activation step as indicated in parenthesis.

Table 2 Solvent-free reduction of carboxylic acids with NaBH₄ promoted by TCT–PPh₃^a

Entry	Carboxylic acid	% Yield (Ref.)
a Unless otherwise specified, a mixture of TCT (0.271 mmol), PPh3 (0.054 mmol), carboxylic acid (0.271 mmol), and K2CO3 (0.406 mmol) was ground for 5 min before adding NaBH4 (0.541 mmol), followed by grinding for further 5 min. The yields in parenthesis were obtained with 10 min acid activation, followed by 5 min reduction.
1	R = C6H5	90 (ref. 19a)
2	R = 3-CH3C6H4	91 (ref. 19b)
3	R = 4-CH3C6H4	91 (ref. 19c)
4	R = 2-CH3OC6H4	89 (ref. 19d)
5	R = 4-CH3OC6H4	92 (ref. 19e)
6	R = 3,4-(CH3O)2C6H3	95 (ref. 19f)
7	R = 3-(CH3)2NC6H4	75 (ref. 19g)
8	R = 2-ClC6H4	88 (ref. 19h)
9	R = 2-IC6H4	85 (ref. 19i)
10	R = 4-ClC6H4	89 (ref. 19j)
11	R = 3-NO2C6H4	71 (ref. 19k)
12	R = 4-NO2C6H4	45(70) (ref. 19a)
13	R = Cinnamyl	81 (ref. 19l)
14	R = 1-Naphthylacetyl	76 (ref. 19m)
15	R = 5-Phenylvaleryl	74 (ref. 19n)
16	Fmoc-Gly-OH	77(87) (ref. 19o)
17	Fmoc-Ala-OH	64(80) (ref. 19p)
18	Fmoc-Val-OH	66(83) (ref. 19q)
19	Fmoc-Ile-OH	61(78) (ref. 19q)
20	Z-Phe-OH	75(84) (ref. 19r)

---

For conjugated acid, cinnamic acid (entry 13) was reduced smoothly to give the expected allylic alcohol as the major product along with 20% of the saturated alcohol derived from the competitive C=C reduction based on GC-MS analysis. This result was in consistent with the previously reported works on borohydride reduction with the TCT/NMM system.^7b Aliphatic acids including 1-napthylacetic acid and 5-phenylvaleric acid were less reactive than aromatic carboxylic acids and gave the corresponding alcohols in moderate yields (entries 14 and 15).

For N-protected α-amino acids (entries 16–20), the method was less effective possibly due to steric hindrance of these substrates. Amino acids having 9-fluorenylmethyloxycarbonyl (Fmoc) and benzyloxycarbonyl (Z) as the amino protecting groups were reduced to the corresponding amino alcohols in moderate yields without loss of their optical purities indicating no racemization occurred under the applied condition. Upon increasing the time of the first grinding step from 5 min to 10 min, the yields of the corresponding amino alcohols were greatly improved indicating that for the less reactive or steric hindered substrates, the acid activation times need further adjustment to enhance the product yields.

Based on the above results, reaction mechanism for the TCT–PPh₃ mediated carboxylic acid activation prior to borohydride reduction was proposed according to Scheme 1. Since the reaction requires 1 equiv. of TCT to achieve high conversion, the reaction is believed to proceed via nucleophilic displacement of one chloride atom of TCT with PPh₃ to provide a triazinephosphonium chloride I. This highly reactive intermediate then undergoes rapid substitution with a carboxylic acid to give an acylated triazine II prior to reduction with NaBH₄.

Scheme 1 Proposed mechanism for NaBH₄ reduction of carboxylic acid mediated by TCT–PPh₃.

It is important to note that in contrast to the reduction of carboxylic acids activated by the TCT/NMM system where aliphatic acids were more favorable,⁷ our system was more effective with aromatic carboxylic acids. This could presumably due to the π–π stacking interactions between the benzene ring of the aromatic acids with those of phosphonium salt I which accelerates the rate of formation of an active ester II.

In summary, this work reported the first solvent-free method for reduction of carboxylic acids to alcohols with NaBH₄. Using TCT in combination with catalytic amount of PPh₃ as an acid activator, a range of carboxylic acids could be readily reduced to the corresponding alcohols in good to excellent yields within short reaction times. This protocol offers several benefits over the existing methods including the use of inexpensive reagents, reduction of volatile organic solvent with time and energy efficiency. Applications of the developed reagent system on other functional group transformations are currently underway and outcome will be reported shortly.

Acknowledgements

The Center of Excellence for Innovation in Chemistry (PERCH-CIC), the National Research University Project under Thailand's Office of the Higher Education Commission, and the Graduate School, Chiang Mai University are gratefully acknowledged for financial support.

Notes and references

1. L. Zhenjiang, Synlett, 2005, 182–183.
2. (a) D. Setamdideh, Z. Karimi and A. Alipouramjad, J. Chin. Chem. Soc., 2013, 60, 590–596; (b) M. Mohamadi, D. Setamdideh and B. Khezri, Org. Chem. Int., 2013, 127585; (c) H. Mahdavi and E. Haghani, Chin. J. Chem., 2008, 26, 333–337; (d) D. Setamdideh and B. Zeynizadeh, Z. Naturforsch., B: Chem. Sci., 2006, 61, 1275–1281; (e) B. Zeynizadeh and D. Setamdideh, J. Chin. Chem. Soc., 2005, 52, 1179–1184; (f) B. Zeynizadeh and S. Yahyaei, Z. Naturforsch., B: Chem. Sci., 2004, 59, 704–710; (g) B. Zeynizadeh and S. Yahyaei, Bull. Korean Chem. Soc., 2003, 24, 1664–1670; (h) B. Tamami and H. Mahdavi, Tetrahedron, 2003, 59, 821–826.
3. (a) S.-D. Cho, Y.-D. Park, J.-J. Kim, J. R. Falck and Y.-J. Yoon, Bull. Korean Chem. Soc., 2004, 25, 407–409; (b) E. Papavassilopoulou, P. Christofis, D. Terzoglou and P. Moutevelis-Minakakis, Tetrahedron Lett., 2007, 48, 8323–8325; (c) M. M. Lakouraj, M. Tajbakhsh and M. S. Mahalli, Monatsh. Chem., 2008, 139, 117–123.
4. (a) T. Fujisawa, T. Mori and T. Sato, Chem. Lett., 1983, 835–838; (b) B. T. Cho and N. M. Yoon, Synth. Commun., 1985, 15, 917–924; (c) T. Okawara, N. Ikeda, T. Yamasaki and M. Furukawa, Chem. Pharm. Bull., 1988, 36, 3628–3631; (d) Y. Suseela and M. Periasamy, Tetrahedron, 1992, 48, 371–376; (e) J. W. Simek, T. Tuck and K. C. Bush, J. Chem. Educ., 1997, 74, 107–108; (f) S. J. Zhang, W. X. Hu and H. Q. Chen, Chin. Chem. Lett., 2007, 18, 1463–1465; (g) Y. Xu and Y. Wei, Synth. Commun., 2010, 40, 3423–3429; (h) G. Nagendra, C. Madhu, T. M. Vishwanatha and V. V. Sureshbabu, Tetrahedron Lett., 2012, 53, 5059–5063.
5. G. Kokotos and C. Noula, J. Org. Chem., 1996, 61, 6994–6996.
6. R. P. McGeary, Tetrahedron Lett., 1998, 39, 3319–3322.
7. (a) A. D. Sagar, J. S. Pulle, S. M. Reddy and M. V. Yadav, J. Chem. Pharm. Res., 2011, 3, 1103–1108; (b) M. Falorni, A. Porcheddu and M. Taddei, Tetrahedron Lett., 1999, 40, 4395–4396.
8. S.-H. Hwang, M. A. Blaskovich and H.-O. Kim, Open Org. Chem. J., 2008, 2, 107–109.
9. J. A. Morales-Serna, E. Garcia-Rios, J. Bernal, E. Paleo, R. Gavino and J. Cardenas, Synthesis, 2011, 1375–1382.
10. R. H. Tale, K. M. Patil and S. E. Dapurkar, Tetrahedron Lett., 2003, 44, 3427–3428.
11. (a) B. Zeynizadeh and D. Setamdideh, Asian J. Chem., 2009, 21, 3588–3602; (b) B. T. Cho, S. K. Kang, M. S. Kim, S. R. Ryu and D. K. An, Tetrahedron, 2006, 62, 8164–8168; (c) H. Shalbaf, Asian J. Chem., 2010, 22, 6761–6764; (d) B. Zeynizadeh and T. Behyar, J. Braz. Chem. Soc., 2005, 16, 1200–1209; (e) W.-Y. Liu, Q.-H. Xu and Y.-X. Ma, Org. Prep. Proced. Int., 2000, 32, 596–600; (f) D. Setamdideh, Z. Karimi and F. Rahimi, Orient. J. Chem., 2011, 27, 1621–1634.
12. (a) L. De Luca, G. Giacomelli and M. Taddei, J. Org. Chem., 2001, 66, 2534–2537; (b) B. Bandgar and S. Sawant, Synth. Commun., 2006, 36, 859–864; (c) S. Masala and M. Taddei, Org. Lett., 1999, 1, 1355–1357; (d) H. L. Rayle and L. Fellmeth, Org. Process Res. Dev., 1999, 3, 172–176.
13. K. Venkataraman and D. R. Wagle, Tetrahedron Lett., 1979, 3037–3040.
14. B. Kolesinska and Z. J. Kaminski, Tetrahedron, 2009, 65, 3573–3576.
15. L. F. Pedrosa, Synlett, 2008, 1581–1582.
16. G. Blotny, Tetrahedron, 2006, 62, 9507–9522.
17. V. A. Stenger, J. Chem. Eng. Data, 1996, 41, 1111–1113.
18. General procedure; unless otherwise specified, carboxylic acid (0.271 mmol), TCT (0.271 mmol), PPh₃ (0.054 mmol) and K₂CO₃ (0.406 mmol) were mixed and ground for 5 min during which a few drop of CH₂Cl₂ was added to aid the grinding. After addition of NaBH₄ (0.541 mmol), the mixture was ground for further 5 minute. The crude material was purified by short column chromatography using ethyl acetate/hexane as the eluent to afford pure product. All known products were characterized by ¹H- ¹³C-NMR and GC-MS, and their spectroscopic data were consistent with those reported in the literature.
19. (a) L. Ford, F. Atefi, R. D. Singer and P. J. Scammells, Eur. J. Org. Chem., 2011, 942–950; (b) M. M. Mojtahedi, E. Akbarzadeh, R. Sharifi and M. S. Abaee, Org. Lett., 2007, 9, 2791–2793; (c) L. Koren-Selfridge, H. N. Londino, J. K. Vellucci, B. J. Simmons, C. P. Casey and T. B. Clark, Organometallics, 2009, 28, 2085–2090; (d) M. M. Heravi, N. Z. Ahari, H. A. Oskooie and M. Ghassemzadeh, Phosphorus, Sulfur Silicon Relat. Elem., 2005, 180, 1701–1712; (e) H. Quiroz-Florentino, R. I. Hernandez-Benitez, J. A. Avina, E. Burgueno-Tapia and J. Tamariz, Synthesis, 2011, 1106–1112; (f) H. Kanho, S. Yaoya, N. Kawahara, T. Nakane, Y. Takase, K. Masuda and M. Kuroyanagi, Chem. Pharm. Bull., 2005, 53, 361–365; (g) J. Cody and C. J. Fahrni, Tetrahedron, 2004, 60, 11099–11107; (h) N. S. Shaikh, K. Junge and M. Beller, Org. Lett., 2007, 9, 5429–5432; (i) D. Dolenc and B. Plesnicar, J. Org. Chem., 2006, 71, 8028–8036; (j) R. Cano, M. Yus and D. J. Ramon, Tetrahedron, 2011, 67, 8079–8085; (k) B. Basu, B. Mandal, S. Das, P. Das and A. K. Nanda, Beilstein J. Org. Chem., 2008, 4, 53; (l) M. L. Clarke, M. B. Diaz-Valenzuela and A. M. Z. Slawin, Organometallics, 2007, 26, 16–19; (m) L.-C. Li, J.-X. Jiang, J. Ren, Y. Ren, C. U. Pittman Jr and H.-J. Zhu, Eur. J. Org. Chem., 2006, 1981–1990; (n) C. P. Owen, I. Shahid, M. S. Olusanjo, C. H. Patel, S. Dhanani and S. Ahmed, J. Steroid Biochem. Mol. Biol., 2008, 111, 117–127; (o) Y. Harayama, M. Yoshida, D. Kamimura, Y. Wada and Y. Kita, Chem.–Eur. J., 2006, 12, 4893–4899; (p) H. S. Lalithamba and V. V. Sureshbabu, Indian J. Chem., Sect. B: Org. Chem. Incl. Med. Chem., 2010, 49B, 1372–1378; (q) V. V. S. Babu, K. Sudarshan and N. S. Sudarshan, Indian J. Chem., Sect. B: Org. Chem. Incl. Med. Chem., 2006, 45B, 1880–1886; (r) P. S. Kumar, G. D. K. Kumar and S. Baskaran, Eur. J. Org. Chem., 2008, 6063–6067.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental procedure and spectroscopic data. See DOI: 10.1039/c4ra08643k

---