Acceleration of Acid-Catalyzed Hydrolysis in a Biphasic System by Sodium Tetracyanocyclopentadienides

Abstract

The hydrolysis of tert-butyldimethylsilyl L-menthyl ether (3) in a CH₂Cl₂–1 M HCl biphasic solvent system was accelerated by the addition of sodium tetracyanocyclopentadienides 1. Particularly, the reaction rate was enhanced using sodium salt 1a–c with a lipophilic substituent on the cyclopentadienide ring. From the results obtained by a triphasic experiment, hydrolysis proceeds via the formation of hydronium ion 2 in the aqueous phase by ion exchange, followed by the transfer of 2 to the CH₂Cl₂ phase.

Phase-transfer reactions have garnered increasing attention in organic chemistry because of their mild, sustainable conditions.¹⁾ Phase-transfer catalysts (PTCs) enhance the reactivity of anionic reagents by forming a lipophilic ion pair, mediating a reaction in the organic phase or at the interface between two phases. Enantioselective phase-transfer reactions catalyzed by chiral quaternary ammonium salts are one of the representative methods for synthesizing optically active molecules.^2,3) Anionic PTCs have also been developed for reactions where lipophilic anions enhance the solubility of the reactive cationic species, thereby promoting the reactions. Kobayashi et al. have reported the first anionic phase-transfer reaction using tetrakis(3,5-bis(trifluoromethyl)phenyl)borate (TFPB) anion.^4–6) Recently, asymmetric anionic phase-transfer reactions^7,8) have been reported using chiral phosphates with an aziridinium cation⁹⁾ and N-haloammonium cations.^10–18) However, a limited number of studies have been reported about anionic PTCs. Weakly basic anions are candidates for anionic PTCs because they barely undergo protonation or react with other cationic species. Tetracyanocyclopentadienides (C₅R(CN)₄⁻)^19–21) are weakly basic anions; their conjugate acids exhibit very small pK_a values as superacids, e.g., perchloric acid.^22,23) Recently, we have efficiently synthesized various Na[C₅R(CN)₄] salts 1a–f from tetracyanothiophene and sulfones and achieved direct functionalization at a substitution position R (Fig. 1).²⁴⁾ During our previous study, we noticed that salts 1a–f are miscible in polar organic solvents (MeCN, tetrahydrofuran (THF), EtOAc) and can be extracted with organic solvents by partitioning between EtOAc and H₂O, despite the absence of large hydrophobic functional groups. Such unique lipophilic behavior exhibited by the C₅R(CN)₄ salts 1a–f stimulated us to explore their catalytic activities in a biphasic solvent system.

Fig. 1. Sodium Tetracyanocyclopentadienides Na[C₅R(CN)₄] 1a–f

We envisaged that Na[C₅R(CN)₄] 1 could serve as a carrier of hydronium cations in the organic phase of the CH₂Cl₂–H₂O biphasic system (Fig. 2). In the aqueous phase of the biphasic system, HCl dissociates, forming a hydronium cation (H₃O⁺) and chloride anion (Cl⁻). The addition of 1 to the aqueous phase forms hydronium salt H₃O⁺[C₅R(CN)₄]⁻ 2 by ion exchange, which is a hydronium ion that is “masked” by the lipophilic tetracyanocyclopentadienide group; it can be extracted into the CH₂Cl₂ phase. This working hypothesis could be verified by investigating the hydrolysis of tert-butyldimethylsilyl (TBS) ether 3 to L-menthol in CH₂Cl₂ catalyzed by hydronium salt 2.

Fig. 2. Role of C₅R(CN)₄ Anion in the H₂O–CH₂Cl₂ Biphasic System

The reaction rate for the hydrolysis of the TBS group was significantly affected by acid strength. We examined the effects of sodium salt 1a and acidity of the aqueous phase on the biphasic hydrolysis of 3 (Table 1). Hydrolysis did not proceed in water–CH₂Cl₂ and in a pH 4 buffer solution–CH₂Cl₂ two-phase system (entries 1, 2, respectively). By using 0.03 M HCl, the TBS group was eliminated, albeit very slowly (entry 3). Furthermore, with increasing acid concentration to 1.0 M, the hydrolysis rate increased, and desilylation was completed within 2 h (entry 4). By contrast, in the absence of 1a, the hydrolysis of 3 did not occur even when using 1.0 M HCl as the aqueous phase (entries 5–7). For comparison with representative superacid salts, NaClO₄ and NaBF₄ were examined. However, the desilylation was not induced because they could not be partitioned into the organic phase, attributed to the lack of lipophilicity (entries 5, 6, respectively).

Table 1. Effect of Catalyst and Aqueous Phase Acidity on the Hydrolysis of TBS Ether 3

Entry	Sodium salt	Aqueous phase	Time (h)	L-Menthol (%)
1	Na[C5(CO2Menthyl)(CN)4] 1a	H2O	24	0
2	Na[C5(CO2Menthyl)(CN)4] 1a	pH 4 buffer	24	0
3	Na[C5(CO2Menthyl)(CN)4] 1a	0.03 M HCl	20	89
4	Na[C5(CO2Menthyl)(CN)4] 1a	1.0 M HCl	2	92
5	NaClO4	1.0 M HCl	24	0
6	NaBF4	1.0 M HCl	24	0
7	—	1.0 M HCl	24	0

Next, the effect of substitution of C₅R(CN)₄ salt 1 on the hydrolysis of 3 in a 1 M HCl and CD₂Cl₂ biphasic solvent system was investigated by ¹H-NMR (Fig. 3(A)). Tetracyanocyclopentadienides 1a–c with less polar substituents (R=CO₂menthyl (◆), CH₂CH₂OBn (■), and Ph (▲), respectively) significantly accelerated hydrolysis, while those with polar substituents 1d–f (R=CO₂Et (●), CN (×), and CH₂OH (＊), respectively) resulted in slow reactions. These results suggest that more hydrophobic anions increase the amount of hydronium salt 2 dissolving in CD₂Cl₂. The hydrolysis of the TBS group in a 1 M HCl–CDCl₃ biphasic system was slower than that in 1 M HCl–CD₂Cl₂ (Fig. 3(B)). The decreased reaction rate in CDCl₃ is attributed to a lower solubility of hydronium salt 2 in CDCl₃ than in CD₂Cl₂.²⁵⁾ The electron-withdrawing power of the substituent R of C₅R(CN)₄ anion did not affect the hydrolysis rate, as indicated by the moderate reaction rates of 1d–f.

Fig. 3. NMR Study for the Reaction Rate for the Hydrolysis of 3 Catalyzed by 1 in (A) 1 M HCl–CD₂Cl₂ or (B) 1 M HCl–CDCl₃

Other acid-catalyzed reactions by 1 were examined under the same biphasic system. A hydrolysis of acetal 5 provided diol 6 in only 22% yield after 23 h in the absence of catalyst 1. The addition of catalyst 1c to the biphasic system markedly enhanced the reaction rate, and diol 6 was obtained in 73% yield after 2 h (Eq. 1). As an example of a different type of reaction, we demonstrated an acid-catalyzed rearrangement of cis-stilbene oxide (7) to aldehyde 8, induced by addition of 1e (Eq. 2).

Next, we focused on the phase transfer ability of hydronium cation 2. Initially, we attempted to isolate the hydronium salt 2, an ion pair formed from Na[C₅Ph(CN)₄] (1c) and a hydronium cation, using a separating funnel. However, the CH₂Cl₂ extract of a solution of 1c in 1 M HCl did not afford enough tetracyanocyclopentadienide salt for detection by ¹H-NMR and TLC. Then, another experiment was planned for determining the presence of 2 in CH₂Cl₂ phase using a right-angled U-tube, where two independent aqueous phases were separated by an organic phase (Fig. 4). A U-tube apparatus was equipped with a stir bar and two glass sticks, and another two stir bars were placed on the top of each glass stick. The U-tube was filled with CH₂Cl₂ right up to the upper level of the glass sticks, and an aqueous phase was added over the CH₂Cl₂ phase in both arms (Phases A, B). After adding 1c to phase A, the whole system was stirred for the indicated time.²⁶⁾ The percentages of the amount of the remaining 1c in phase A and the transferred amount to phase B were measured on the basis of the initial amount of 1c.

Fig. 4. Apparatus for the Measurement of the Phase-Transfer Rate

(A) Schematic; (B) Photograph of the apparatus; (C) Photograph of the glass stick.

In the case where both phases A and B were water, the transfer of the C₅Ph(CN)₄ anion from phase A to phase B was not observed regardless of the additive ether in the CH₂Cl₂ phase (Table 2, entries 1, 2, respectively). On the other hand, by changing the phases A and B from water to 1 M HCl, the slow transfer of the C₅Ph(CN)₄ anion was induced from phase A to phase B via the middle organic phase (entry 3). These results supported our hypothesis that H₃O⁺[C₅Ph(CN)₄]⁻ (2), formed from H₃O⁺Cl⁻ and 1, is partitioned between the organic and aqueous phases. The addition of hydrophobic 3 and L-menthol accelerated the rate of transfer from phase A to phase B (entries 5, 6, respectively),²⁷⁾ while with the addition of diethyl ether as an additive into the CH₂Cl₂ phase, the transfer rate of the C₅Ph(CN)₄ anion was slightly affected (entry 4). The acceleration effect of the former additives was attributed to the formation of a hydrophobic oxonium specie such as 4 shown in Fig. 2, which enhances the solublity of the C₅Ph(CN)₄ salt into the CH₂Cl₂ phase.

Table 2. Percentages of the Remaining C₅Ph(CN)₄ Anion 1c in Phase A and the Transferred C₅Ph(CN)₄ Anion to Phase B

Entry	Phase A/Phase B	Additive	Phase A (%)/Phase B (%)a)
			2 h	6 h	24 h	48 h
1	H2O/H2O	—	101/0	100/0	102/0	—
2	H2O/H2O	TBS ether 3	105/0	104/0	104/0	—
3	1 M HCl/1 M HCl	—	88/2	79/3	81/3	77/11
4	1 M HCl/1 M HCl	Diethyl ether	93/1	92/8	76/15	73/21
5	1 M HCl/1 M HCl	TBS ether 3	90/4	68/16	50/32	41/36
6	1 M HCl/1 M HCl	L-Menthol	79/6	65/13	53/25	43/28

a) Each percentage was calculated on the basis of the initial amount in phase A.

The hydronium cation carrier ability of 1 was also confirmed by pH measurements (Fig. 5). A similar experiment as shown in Table 2 was conducted using a U-tube apparatus charged with 1 M HCl for phase A and H₂O for phase B. The change of pH in phase B was measured at the indicated times. In the absence of 1c, the pH of phase B slowly decreased from 7.0 to 5.3 after 24 h (●). The addition of sodium salt 1c accelerated the decrease of pH, attaining pH 2.8 after 24 h (▲). The synergetic effect of an additive and salt 1c on the decrease of pH was also observed for the C₅Ph(CN)₄ anion transfer experiment summarized in Table 2 (entries 4–6). The addition of diethyl ether to the CH₂Cl₂ phase did not change the rate (◆), while hydrophobic 3 or L-menthol significantly accelerated the pH decrease of phase B (× and +). When NaBF₄ was employed instead of 1c, the rate of pH decrease was much slower than when 1c was used (■), consistent with the result of the hydrolysis reactions shown in Table 1, entry 6.

Fig. 5. Change of pH in Phase B When 1 M HCl and H₂O Were Used for Phases A and B, Respectively

In conclusion, we discovered that tetracyanocyclopentadienides 1 promote the hydrolysis of TBS ether 3 in a CH₂Cl₂–1 M HCl biphasic solvent system via the formation of lipophilic hydronium salt 2. The capability of 2 as a phase-transfer catalyst was also demonstrated by experiments conducted using a triphasic system with a right-angled U-tube. Currently, more experiments for verifying the synthetic utility of catalyst 1 is underway in our laboratory.

Experimental

Hydrolysis of TBS Ether 3 in a Biphasic System Using Catalyst 1a (Table 1, Entry 4)

To a solution of TBS ether 3²⁸⁾ (100 mg, 0.37 mmol) in CH₂Cl₂ (1.0 mL) was added 1 M HCl (0.50 mL, 0.50 mmol) and tetracyanocyclopentadienide 1a (7.4 mg, 0.02 mmol) and the reaction mixture was stirred for 2 h. The reaction was quenched with saturated aq NaHCO₃ and the mixture was extracted with EtOAc. The extract was washed with brine, dried over Na₂SO₄, filtered, and concentrated under reduced pressure. Purification by flash chromatography (20% EtOAc in n-hexane) afforded L-menthol (53 mg, 92%). Further elution with EtOAc provided recovered 1a (6.4 mg, 86% recovered).

L-Menthol. ¹H-NMR (CDCl₃, 500 MHz) δ: 3.40 (1H, td, J=10.4, 4.4 Hz), 2.18 (1H, sptd, J=7.0, 2.7 Hz), 1.96 (1H, dtd, J=12.1, 3.9, 2.1 Hz), 1.66 (1H, dqd, J=3.2, 2.5 Hz), 1.61 (1H, dq, J=12.9, 3.2 Hz), 1.56 (1H, br s), 1.42 (1H, tqt, J=11.5, 7.0, 3.4 Hz), 1.11 (1H, ddt, J=12.4, 9.8, 3.1 Hz), 1.01–0.81 (3H, m), 0.93 (3H, d, J=7.0 Hz), 0.91 (3H, d, J=7.0 Hz), 0.81 (3H, d, J=7.0 Hz).²⁹⁾

Hydrolysis of Acetonide 5 in a Biphasic System Using Catalyst 1c (Eq. 1)

To a solution of acetonide 5³⁰⁾ (100 mg, 0.56 mmol) in CH₂Cl₂ (1.5 mL) was added 1 M HCl (0.75 mL) and catalyst 1c (7.4 mg, 0.028 mmol), and the reaction mixture was stirred for 2 h. The reaction was quenched with saturated aq NaHCO₃ and the mixture was extracted with EtOAc. The extract was washed with brine, dried over Na₂SO₄, filtered, and concentrated under reduced pressure. Purification by flash chromatography (80% EtOAc in n-hexane) afforded diol 6 (53 mg, 92%). Further elution with EtOAc provided recovered 1c (6.0 mg, 81% recovered).

Diol 6. ¹H-NMR (CDCl₃, 500 MHz) δ: 7.36–7.26 (5H, m), 4.81 (1H, dd, J=8.1, 3.4 Hz), 3.74 (1H, dd, J=11.3, 3.4 Hz), 3.66 (1H, dd, J=11.3, 8.1 Hz), 2.88 (1H, br s), 2.44 (1H, br s).²⁹⁾

Acid-Catalyzed Rearrangement of cis-Stilbene Oxide (7) to Aldehyde 8 (Eq. 2)

To a solution of cis-stilbene oxide (7) (50.0 mg, 0.25 mmol) in CH₂Cl₂ (1 mL) was added 1 M HCl (0.5 mL) and catalyst 1e (2.9 mg, 0.014 mmol), and the reaction mixture was stirred for 24 h. The reaction was quenched with saturated aq NaHCO₃ and the mixture was extracted with EtOAc. The extract was washed with brine, dried over Na₂SO₄, filtered, and concentrated under reduced pressure. Purification by flash chromatography (10% Et₂O in n-hexane) afforded diol 6 (28.2 mg, 58%). Further elution with EtOAc provided recovered 1e (1.9 mg, 66% recovered).

Aldehyde 8. ¹H-NMR (CDCl₃, 500 MHz) δ: 9.94 (1H, d, J=2.5 Hz), 7.40–7.34 (4H, m), 7.30 (2H, tt, J=7.3, 1.5 Hz), 7.24–7.20 (4H, m), 4.89 (1H, d, J=2.5 Hz).²⁹⁾

General Procedure for the NMR Experiment of the Hydrolysis of TBS Ether 3 in a Biphasic System (Fig. 3)

An NMR tube was charged with a solution of TBS ether 3 (30 mg, 0.11 mmol) and benzene (internal standard, 10 µL) in CD₂Cl₂ (0.6 mL). A solution of tetracyanocyclopentadienide 1 (0.0055 mmol, 0.05 equiv) in 1 M HCl (0.3 mL) was added, followed by vigorous manual shaking. The mixture was allowed to stand for the indicated time. The yields were calculated on the basis of the integration of the ¹H-NMR signals of L-menthol (0.74 ppm for methyl group) and benzene (7.37 ppm).

General Procedure of the U-Tube Experiment for the C₅Ph(CN)₄ Salt Solubility in CH₂Cl₂ (Table 2)

A right-angled U-tube equipped with two glass sticks and three magnetic stir bars (Fig. 4) was charged with a solution of an additive (0.75 mmol) in CH₂Cl₂ (20 mL). A solution of 1c (8 mg, 0.03 mmol) in 1 M HCl or H₂O (3 mL) was added to one side of the U-tube (phase A), and 1 M HCl or H₂O (3 mL) was added to the other side (phase B), and the system was stirred at a rate of approx. 1000 rpm. The amount of the C₅Ph(CN)₄ anion was measured at the indicated time according to the following procedures. A 10 µL sample was collected using a micropipette and diluted with a 0.025% (w/v) phenol aqueous solution in a 5 mL volumetric flask. The resulting solution was analyzed by HPLC (Develosil C30-UG-5, 70% MeOH/25 mM aqueous Na₂HPO₄, 0.8 mL/min, 254 nm, 5.7 min for phenol and 9.5 min for 1c), and the percentages of the remaining C₅Ph(CN)₄ anion in phase A or the transferred C₅Ph(CN)₄ anion to phase B were calculated on the basis of the initial amount of 1c in phase A.

Acknowledgments

This study was partially supported by a Grant-in-Aid for Scientific Research (C) (24590033), Grant-in-Aid for Young Scientists (B) (26860018) from the Japan Society for the Promotion of Science (JSPS), and the Science Research Promotion Fund from the Promotion and Mutual Aid Corporation for Private Schools of Japan.

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) “Phase-Transfer Catalysis Mechanisms and Syntheses,” ed. by Halpern M. E., ACS Symposium Series 659, American Chemical Society, Washington, DC, 1997.
• 2) Hashimoto T., Maruoka K., Chem. Rev., 107, 5656–5682 (2007).
• 3) “Asymmetric Phase Transfer Catalysis,” ed. by Maruoka K., Wiley-VCH, Weinheim, 2008.
• 4) Kobayashi H., Sonoda T., Iwamoto H., Yoshimura M., Chem. Lett., 10, 579–580 (1981).
• 5) Iwamoto H., Kobayashi H., Murer P., Sonoda T., Zollinger H., Bull. Chem. Soc. Jpn., 66, 2590–2602 (1993).
• 6) Ichikawa J., Kobayashi H., Sonoda T., J. Synth. Org. Chem. Jpn., 46, 943–954 (1988).
• 7) Brak K., Jacobsen E. N., Angew. Chem. Int. Ed., 52, 534–561 (2013).
• 8) Mahlau M., List B., Angew. Chem. Int. Ed., 52, 518–533 (2013).
• 9) Hamilton G. L., Kanai T., Toste F. D., J. Am. Chem. Soc., 130, 14984–14986 (2008).
• 10) Rauniyar V., Lackner A. D., Hamilton G. L., Toste F. D., Science, 334, 1681–1684 (2011).
• 11) Phipps R. J., Hiramatsu K., Toste F. D., J. Am. Chem. Soc., 134, 8376–8379 (2012).
• 12) Honjo T., Phipps R. J., Rauniyar V., Toste F. D., Angew. Chem. Int. Ed., 51, 9684–9688 (2012).
• 13) Phipps R. J., Toste F. D., J. Am. Chem. Soc., 135, 1268–1271 (2013).
• 14) Shunatona H. P., Früh N., Wang Y.-M., Rauniyar V., Toste F. D., Angew. Chem. Int. Ed., 52, 7724–7727 (2013).
• 15) Romanov-Michailidis F., Guénée L., Alexakis A., Angew. Chem. Int. Ed., 52, 9266–9270 (2013).
• 16) Xie W., Jiang G., Liu H., Hu J., Pan X., Zhang H., Wan X., Lai Y., Ma D., Angew. Chem. Int. Ed., 52, 12924–12927 (2013).
• 17) Liu H., Jiang G., Pan X., Wan X., Lai Y., Ma D., Xie W., Org. Lett., 16, 1908–1911 (2014).
• 18) Shen Z., Pan X., Lai Y., Hu J., Wan X., Li X., Zhang H., Xie W., Chem. Sci., 6, 6986–6990 (2015).
• 19) Webster O. W., J. Am. Chem. Soc., 87, 1820–1821 (1965).
• 20) Webster O. W., J. Am. Chem. Soc., 88, 4055–4060 (1966).
• 21) Simmons H. E., Vest R. D., Vladuchick S. A., Webster O. W., J. Org. Chem., 45, 5113–5121 (1980).
• 22) Webster O. W., J. Am. Chem. Soc., 88, 3046–3050 (1966).
• 23) Stoyanov E. S., Kim K.-C., Reed C. A., J. Am. Chem. Soc., 128, 8500–8508 (2006).
• 24) Sakai T., Seo S., Matsuoka J., Mori Y., J. Org. Chem., 78, 10978–10985 (2013).
• 25) Octanol–water partition coefficient (log P) of CHCl₃ (1.97) is higher than that of CH₂Cl₂ (1.25). Sangster J., J. Phys. Chem. Ref. Data, 18, 1111–1229 (1989).
• 26) The movie showing the experiment was attached as a supplementary material.
• 27) One referee pointed out that the total concentration of 1c in the combined aqueous phases A and B reduced to ca. 90% based on the initial amount of 1c in every entry. We consider that it is within the limits of experimental error because the total amount of 1c in the combined aqueous phases was with the same after 2 and 48 h for entries 3 and 4. In entries 5 and 6, 1c would be partially dissolved in the CH₂Cl₂ phase after 48 h. A small amount of 1c was detected by ¹H-NMR analysis of a CH₂Cl₂ extract from 1c solution in 1 M HCl, when 20 equiv of L-menthol was used as an additive for the extraction.
• 28) Matsubara H., Maegawa T., Kita Y., Yokoji T., Nomoto A., Org. Biomol. Chem., 12, 5442–5447 (2014).
• 29) Pouchert C. J., Behnke J., “The Aldrich Library of ¹³C and ¹H FT NMR Spectra,” Aldrich Chemical Company, Milwaukee, 1993.
• 30) Kliman L. T., Mlynarski S. N., Morken J. P., J. Am. Chem. Soc., 131, 13210–13211 (2009).