Alcohol synthesis based on the S_N2 reactions of alkyl halides with the squarate dianion

Abstract

A convenient method has been developed for transforming alkyl halides into the corresponding alcohols via an S_N2 reaction. Treatment of an alkyl halide with the squarate dianion at high temperature produces mono-alkyl squarate, and a one-pot basic hydrolysis of the intermediate affords the alcohol in good yield.

---

Synthesis of alcohols from their corresponding alkyl halides is one of the most important and fundamental transformations in organic chemistry.¹ Spontaneous hydrolysis in wet polar solvents via an S_N1 mechanism is applicable only to highly reactive halides, such as tertiary alkyl halides. Hence, the conversion of primary and secondary alkyl halides into alcohols requires some reactions with an oxygen source.² The S_N2 reaction of an alkyl halide with the hydroxide ion may provide a straightforward method for this purpose, but the “hard” property of the hydroxide ion tends to induce undesirable β-elimination reactions yielding alkenes.³ A dialkyl ether, derived from the desired alcohol and remaining alkyl halide under basic conditions, can also be detected as a side product. Therefore, the use of oxygen nucleophiles with a “soft” property is essential for achieving the S_N2 reaction of an alkyl halide in high yields (Scheme 1).

Scheme 1 (a) Typical conversions of alkyl halides into the corresponding alcohols through S_N2 reactions with oxygen nucleophiles. (b) Conversion mechanism proposed in this work.

Superoxide ions⁴ are considered as representative nucleophiles for converting alkyl halides into alcohols,⁵ but the use of hazardous reagents is a serious drawback, especially in large-scale experiments. In contrast, carboxylate anions undergo S_N2 reactions with alkyl halides to afford the corresponding esters, which can easily be converted into alcohols by saponification.⁶ The delocalization of the anionic charge in a carboxylate anion makes it a “soft” nucleophile. However, the carboxylate ions barely react with the sterically hindered alkyl halides. Similarly, phenolate ions function as soft oxygen nucleophiles in the S_N2 reactions with alkyl halides. p-Methoxyphenolate ions are widely used as reagents because alkyl p-methoxyphenyl ethers can be converted into the corresponding alcohols upon treatment with cerium(VI) ammonium nitrate (CAN) and water.⁷ However, this transformation requires a two-step protocol to obtain alcohols from alkyl halides. Herein, we report that squarate dianions conveniently behave as oxygen nucleophiles that can transform primary and secondary alkyl halides. The dianion species exhibited high reactivity in S_N2 reactions, and the resulting squarate monoesters readily underwent hydrolysis in one pot (Scheme 1b).

Squaric acid (H₂Sq) is a strong acid (pK_a1 = 0.52 and pK_a2 = 3.48), comparable to trifluoroacetic acid, because of the high stability of the corresponding squarate anion.^8,9 The anionic charge of the squarate is completely delocalized across the four-membered ring, and the aromatic character of the conjugated π-system contributes to the exceptionally high stability of the anion.

To elucidate the utility of the soft and non-basic oxygen nucleophile, we explored the S_N2 reactions of the squarate dianion (Sq²⁻) with primary alkyl bromide 1a under various conditions. The screening of suitable bases and solvents for the reaction revealed that the combination of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and DMF was optimal for the in situ generation of Sq²⁻ from H₂Sq (see ESI-1 for details†). Bromide 1a smoothly underwent the substitution reaction upon heating with H₂Sq (1.5 equiv.) and DBU (3.0 equiv.) at 60 °C in DMF, and the reaction mixture containing monoalkyl squarate 2 was treated with a K₂CO₃ aqueous solution to obtain the desired alcohol 3a in 89% yield (Table 1, entry 1). When i-Pr₂NEt was used as the base, 3a was obtained but with a dramatically reduced yield of 27% (entry 2). Investigation of the solvent effects showed that the use of MeCN resulted in good yields of 3a (entry 3), but THF led to a decreased reaction conversion to 3a (43% yield) and a recovery of 1a in 48% yield (entry 4). The absence of H₂Sq prevented the formation of 3, leading to a preference for the elimination of 1a (entry 5), thereby demonstrating the role of 2 as the reaction intermediate in our system. Further examination involved changing the amount of H₂Sq from 1.5 equivalents to 0.5 equivalents because H₂Sq can theoretically react with two equivalents of 1a. However, this reaction resulted in the recovery of 1a in 13% yield and generation of 3a in 56% yield (entry 6), suggesting that the reactivity of 2 was lower than that of Sq²⁻. The decrease in the amounts of DBU to 2.2 equivalents also caused the low conversion of 1a (entry 7). Although the reaction at rt required a longer time than that at 60 °C, the yield of 3a remained high (entry 8). The use of a salt prepared using H₂Sq and 2.0 equivalents of DBU led to a dramatic decrease in the yield of 3a to 30% NMR yield (entry 9). The bis-tetra-n-propylammonium salt of H₂Sq¹⁰ improved the yield of 3a (entry 10), but the yield was lower than that in entry 1.¹¹

Table 1 Change of factors under the optimized reaction conditions^a

Entry	Variation from the standard conditions	Yield of 3a^b (%)	Recovery of 1a^b (%)
a 0.2 mmol scale. b Isolated yield. c 3.6 equivalents of DBU were used. d The reaction time was 1 h. e The reaction temperature ranged from 60 to 150 °C, and the reaction time was 6 h. f NMR yield.
1	None	89	—
2	i-Pr2NEt	27	20
3^c	MeCN	82	—
4^c	THF	43	48
5	Absence of H2Sq	ND	Trace
6	0.5 equiv. of H2Sq	56	13
7	2.2 equiv. of DBU	60	24
8	rt, 6 h	87	—
9^d		30^f	—
10^e		39	—

---

We attempted to isolate mono-alkyl squarate 2; however, its isolation was difficult owing to its high polarity. In contrast, quenching the reaction with an excess amount of methyl iodide produced methyl alkyl ester 4, which was isolated in 73% yield (Scheme 2). Hydrolysis of 4 also occurred when using an aqueous K₂CO₃ solution, but the yield of 3a was lower than that for the conversion of 1a into 3a (entry 1 in Table 1). This result suggests that the reactivity of the dialkyl squaric ester under basic hydrolysis conditions is low, and the oxy-anion of 2 might accelerate the opening of the squarate moiety under the reaction conditions.

Scheme 2 Trapping of monoalkyl squarate 2 with iodomethane and trial hydrolysis of dialkyl squarate 4.

The optimized reaction conditions were applied to the transformation of several alkyl halides into their corresponding alcohols (Table 2). While a prolonged reaction time at a higher temperature was required to achieve the complete consumption of primary chloride 1b, the desired alcohol 3a was obtained in 91% yield after a one-pot hydrolysis (entry 1). Secondary halides 1c and 1d were transformed in a similar manner, giving rise to alcohol 3c in good yields (entries 2 and 3). Note that the reaction of 1c with potassium acetate in DMF at 90 °C did not yield 3c and 1c was recovered in 84% NMR yield. Although the Nozaki group reported that the direct conversion of seven-membered chloride 1e into alcohol 3e failed under various conditions,¹² our method facilitated this conversion when the reaction temperature was elevated to 120 °C, affording 3e in 64% yield (entry 4). These results demonstrate the usefulness of squarate dianions as oxygen-incorporating nucleophiles in S_N2 reactions. The use of benzyl halides (1g–k) enabled the formation of the corresponding alcohols (3g–k) without the influence of the electron-donating or electron-withdrawing groups (entries 5–10). Cinnamyl bromide 1l exhibited good reactivity in the S_N2 reaction of Sq²⁻, and cinnamyl alcohol (3l) was obtained in 57% yield (entry 11). Additionally, allylic chlorides 1m and 1n, derived from isopulegol¹³ and carvone,¹⁴ respectively, also efficiently induced the desired reactions, producing diol 3m and ketoalcohol 3n in 79% and 71% yields, respectively (entries 12 and 13). Propargyl chloride 1o also afforded the corresponding alcohol 3o in 86% yield (entry 14). Next, Sq²⁻-mediated chemoselective reactions were examined using acyclic compounds 1p–r bearing two different halides/pseudohalides. The reaction of 6-iodo-1-chlorohexane (1p) with Sq²⁻ at rt occurred at the alkyl iodide moiety chemoselectively, which afforded 3p in 87% yield (entry 15). The subjection of 7-bromo-1-chloroheptane 1q to similar reaction conditions also induced the chemoselective substitution with the Br group, providing 3q in high yield (entry 16). In contrast, the adoption of 7-bromo-1-heptyl tosylate (1r) to the chemoselective reaction was difficult because the eliminated bromo ion easily underwent substitution with the tosyloxy group. On the other hand, the coincident conversions of the bromo and chloro/tosyloxy groups in 1r/1q proceeded smoothly using 3.0 equivalents of Sq²⁻ at 60–80 °C affording the same product, diol 3r, in 83% yield from 1r and in 85% yield from 1q, respectively (entries 17 and 18). Furthermore, the simultaneous conversion of the 1,2-dibromide moiety of 1s¹⁵ was achieved via the exposure of 2.4 equivalents of H₂Sq and DBU at 120 °C, which afforded 3s in 59% yield after a one-pot hydrolysis (entry 19). The E2 reaction of 1s competed when the proportion of DBU exceeded that of H₂Sq; thus, we used the squarate monoanion, generated from the same equivalents of H₂Sq and DBU, as a nucleophile.

Table 2 Substrate scope of the substitution reaction^a

Entry	Substrate	Temp. (°C)	Product	Yield^b (%)	Entry	Substrate	Temp. (°C)	Product	Yield^b (%)
		time (h)					time (h)
a The reaction was generally conducted on a 0.2 mmol scale. b Isolated yield. c 0.1 mmol scale. d 3.0 equivalents of H2Sq and 6.0 equivalents of DBU were used and one-pot hydrolysis was conducted at 100 °C for 2.5 h. e 2.4 equivalents of H2Sq and 2.4 equivalents of DBU were used and one-pot hydrolysis was conducted at 100 °C for 1 h.
1		80		91	11		60		57
		2					0.7
2		80		76	12		80		79
		2					2
3		90		70	13		80		71
		3					2
4^c		120		64	14		80		86
		6					1
5		60		71	15		rt		87
		0.5					2
6		60		96	16		rt		73
		0.5					6
7		60		69	17^d		60		83
		0.5					1
8		60		74	18^d		80		85
		0.5					2
9		60		87	19^e		120		59
		0.5					2
10		600.5		66

---

To demonstrate another application of squarates, we explored the synthesis of highly functionalized alcohol, namely glycerol (Scheme 3). In the presence of a primary alcohol, triisopropylsilyl ether 5¹⁶ was reacted with N-iodosuccinimide (NIS) to afford iodoether 6 in good yields. The use of methanol, benzyl alcohol (BnOH), and propargyl alcohol resulted in the formation of ethers 6a, 6b, and 6c, respectively. Upon heating with Sq²⁻ generated in situ in DMF at 120 °C for 4 h, followed by a one-pot hydrolysis, alcohols 7a, 7b, and 7c were obtained in high yields (Scheme 3a). On the other hand, the bromoetherification of allyl silyl ether 8¹⁷ using 1,3-dibromo-5,5-dimethylhydantoin and BnOH provided a 1.8 : 1 inseparable mixture of primary bromide 9 and secondary bromide 10. This mixture was subjected to 1.5 equivalents of Sq²⁻ at 100 °C, leading to the desired S_N2 reactions in both compounds. Subsequent hydrolysis resulted in the formation of primary alcohol 11 and secondary alcohol 12 in 47% and 15% yields, respectively. A decrease in the reaction temperature to 60 °C induced a selective reaction of 9, which afforded 11 in 48% yield and 10 was recovered in 30% yield (Scheme 3b). The alcohols 7, 11 and 12 are regarded as analogues of glycerol possessing two different protective groups; therefore, they are expected to be versatile building blocks for the synthesis of various complex lipids.

Scheme 3 (a) Synthesis of 2-methylglycerol derivatives. (b) Synthesis of doubly protected glycerols. (c) Preparation of an azidoalcohol.

We further investigated the synthesis of azidoalcohol 13, a precursor for the synthesis of sphingosine (Scheme 3c).¹⁸ The primary bromo group in dibromide 14, prepared from 8via bromination, underwent a chemoselective reaction upon treatment with 1.5 equivalents of H₂Sq and 2.9 equivalents of DBU. Although this reaction proceeded smoothly to produce 15,¹⁹ the subsequent one-pot hydrolysis of 15 using aqueous K₂CO₃ solution generated bromohydrin 16 and epoxide 17 in a 1 : 1 ratio. In contrast, the treatment of the reaction mixture containing 15 with excess hydrazine monohydrate suppressed the formation of 17, affording 16 in 59% yield. The typical conditions for the replacement of the bromo group with an azido group allowed the conversion of 16 into 13 in 87% yield. The conventional protocol for synthesizing azidoalcohol 13 from olefin 8 involves 5 steps: dihydroxylation, selective protection of the primary alcohol, sulfonylation of the secondary alcohol, azidization, and removal of the protecting group. Our synthetic method of 13 from 8 can occur in 3 steps, indicating the utility of Sq²⁻ as an oxygen nucleophile.

Finally, we demonstrated the Walden inversion of secondary alkyl halides/pseudohalides with a stereogenic center (Scheme 4). The reaction of tosylate 18, which was derived from epiandrosterone, with Sq²⁻ proceeded at 100 °C to provide α-alcohol 19 in 70% yield. Although the application of bicyclic compound 20²⁰ required an excess amount of Sq²⁻ and a higher reaction temperature than that required for 18, β-alcohol 21 was obtained in 50% yield.

Scheme 4 Inversion of the stereochemistry of secondary alkyl bromides and tosylates.

Conclusions

In conclusion, we developed a convenient method for transforming alkyl halides into the corresponding alcohols through an S_N2 reaction using Sq²⁻. The reaction of primary- and secondary-alkyl halides with the dianion species proceeded in DMF at 60–120 °C, and the resulting mono-alkyl squarate was readily hydrolyzed by treatment with an aqueous K₂CO₃ solution. Examples for the inversion of a stereogenic center were also described. Furthermore, methods for synthesizing versatile building blocks of complex lipids have also been developed. We are further exploring the utility of Sq²⁻, a potent and safe oxygen nucleophile, in synthetic organic chemistry.

Data availability

All experimental procedures and spectral data are available in the ESI.†

Author contributions

K. T. conceived the research theme and designed the experiments. K. S., T. F., and T. T. performed the experiments and analyzed the data. K. I. and K. T. commanded this work and wrote the manuscript. T. S. assisted in writing and editing the manuscript. All authors contributed to the discussions.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by JSPS KAKENHI (grant numbers JP20K05485, JP21H01923, and JP21K14616) and the Photo-Excitonix Project of Hokkaido University.

References

1. (a) T. A. Hamlin, M. Swart and F. M. Bickelhaupt, ChemPhysChem, 2018, 19, 1315; (b) V. Lankea and I. Marek, Chem. Sci., 2020, 11, 9378.
2. (a) E. Nakamura, T. Inubushi, S. Aoki and D. Machii, J. Am. Chem. Soc., 1991, 113, 8980; (b) M. Sawamura, Y. Kawaguchi, K. Sato and E. Nakamura, Chem. Lett., 1997, 26, 705; (c) A. Abad, C. Agulló, A. C. Cuñat and I. Navarro, Synlett, 2005, 3355; (d) B. M. Chougala, S. Samundeeswari, M. Holiyachi and L. A. Shastri, ChemistrySelect, 2017, 2, 1290; (e) Y.-M. Cai, Y.-T. Xu, X. Zhang, W.-X. Gao, X.-B. Huang, Y.-B. Zhou, M.-C. Liu and H.-Y. Wu, Org. Lett., 2019, 21, 8479; (f) H. Liu, J. Liu, X. Cheng, X. Jia, L. Yu and Q. Xu, ChemSusChem, 2019, 12, 2994.
3. T. Hansen, J. C. Roozee, F. M. Bickelhaupt and T. A. Hamlin, J. Org. Chem., 2022, 87, 1805.
4. (a) M. Hayyan, M. A. Hashim and I. M. AlNashef, Chem. Rev., 2016, 116, 3029; (b) D. T. Sawyer and J. S. Valentine, Acc. Chem. Res., 1981, 14, 393.
5. (a) R. A. Johnson and E. G. Nidy, J. Org. Chem., 1975, 40, 1680; (b) M. A. Brimble, G. M. Williams, R. Baker and M. James, Tetrahedron Lett., 1990, 31, 3043; (c) A. Matsuzawa, J. Shiraiwa, A. Kasamatsu and K. Sugita, Org. Lett., 2018, 20, 1031.
6. (a) S. Winstein and R. E. Buckles, J. Am. Chem. Soc., 1942, 64, 2780; (b) R. C. Larock, J. Org. Chem., 1974, 39, 3721; (c) N. Ono, T. Yamada, T. Saito, K. Tanaka and A. Kaji, Bull. Chem. Soc. Jpn., 1978, 51, 2401; (d) S. Yamamoto, H. Itani, H. Takahashi, T. Tsuji and W. Nagata, Tetrahedron Lett., 1984, 25, 4545; (e) H. A. Zahalka and Y. Sasson, Synthesis, 1986, 763; (f) J. Alexander, M. L. Renyer and H. Veerapanane, Synth. Commun., 1995, 25, 3875.
7. (a) Y. Masaki, K. Yoshizawa and A. Itoh, Tetrahedron Lett., 1996, 37, 9321; (b) L. F. Tietze and J. Görlitzer, Synthesis, 1998, 873; (c) L. He, H.-S. Byun, J. Smit, J. Wilschut and R. Bittman, J. Am. Chem. Soc., 1999, 121, 3897; (d) T. Toma, Y. Kita and T. Fukuyama, J. Am. Chem. Soc., 2010, 132, 10233; (e) W. Szeja, G. Grynkiewicz, T. Bieg, P. Swierk, A. Byczek, K. Papaj, R. Kitel and A. Rusin, Molecules, 2014, 19, 7072.
8. (a) L. M. Schwartz and L. O. Howard, J. Phys. Chem., 1970, 74, 4374; (b) H. Schmidt, Synthesis, 1980, 961; (c) G. Seltz and P. Imming, Chem. Rev., 1992, 92, 1227.
9. Analogues of squaric acid or squaramide are widely used in various fields. For selected reviews see: (a) M. Ohno, Y. Yamamoto and S. Eguchi, Synlett, 1998, 1167; (b) J. Alemán, A. Parra, H. Jiang and K. A. Jørgensen, Chem. – Eur. J., 2011, 17, 6890; (c) L. A. Marchetti, L. K. Kumawat, N. Mao, J. C. Stephens and R. B. P. Elmes, Chem, 2019, 5, 1398; (d) K. A. Agnew-Francis and C. M. Williams, Chem. Rev., 2020, 120, 11616; (e) J. Chasák, V. Šlachtová, M. Urban and L. Brulíková, Eur. J. Med. Chem., 2021, 209, 112872; (f) T. Grus, H. Lahnif, B. Klasen, E.-S. Moon, L. Greifenstein and F. Roesch, Bioconjugate Chem., 2021, 32, 1223.
10. M. G. Topuzova, S. V. Kotov and T. M. Kolev, Appl. Catal., A, 2005, 281, 157.
11. We also used a di-sodium salt of H₂Sq ( Q. Zhao, J. Wang, Y. Lu, Y. Li, G. Liang and J. Chen, Angew. Chem., Int. Ed., 2016, 55, 1252 (8)) as the reagent, but, the reaction did not occur and 1 was recovered in 80% NMR yield.
12. A. Itoh, T. Saito, K. Oshima and H. Nozaki, Bull. Chem. Soc. Jpn., 1981, 54, 1456.
13. (a) S. G. Hegde, M. K. Vogel, J. Saddler, T. Hrinyo, N. Rockwell, R. Haynes, M. Oliver and J. Wolinsky, Tetrahedron Lett., 1980, 21, 441; (b) S. G. Hegde, D. Beckwith, R. Doti and J. Wolinsky, J. Org. Chem., 1985, 50, 894.
14. M. Xuan, I. Paterson and S. M. Dalby, Org. Lett., 2012, 14, 5492.
15. I. Ryu, H. Matsubara, S. Yasuda, H. Nakamura and D. P. Curran, J. Am. Chem. Soc., 2002, 124, 12946.
16. K. Zhao and R. R. Knowles, J. Am. Chem. Soc., 2022, 144, 137.
17. N. Saygili, R. J. Brown, P. Day, R. Hoelzl, P. Kathirgamanathan, E. R. Mageean, T. Ozturk, M. Pilkington, M. M. B. Qayyum, S. S. Turner, L. Vorwerg and J. D. Wallis, Tetrahedron, 2001, 57, 5015.
18. (a) P. M. Koskinen and A. M. P. Koskinen, Synthesis, 1998, 1075; (b) R. Sridhar, B. Srinivas and K. R. Rao, Tetrahedron, 2009, 65, 10701; (c) Y. Gao, X. He, F. Ding and Y. Zhang, Synthesis, 2016, 4017.
19. Reaction of 14 with H₂Sq and DBU, followed by the addition of excess amounts of MeI, provided methyl squarate 22 and vinyl bromide 23 in 80% and 19% NMR yields, respectively. Purification of the crude product via silica gel chromatography resulted in the decrease of the isolated yield of 22 (ca. 27% yield) owing to the instability of 22.
    .
20. R. M. Carman and B. N. Venzke, Aust. J. Chem., 1973, 26, 2235.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental procedure, compound characterization, and copies of NMR data. See DOI: https://doi.org/10.1039/d3ob01507f

‡ These authors contributed equally to this work.

§ Present address: Graduate School of Pharmaceutical Sciences, Nagoya City University, 467-8603, Japan

---