One-step chemoselective conversion of tetrahydropyranyl ethers to silyl-protected alcohols

Abstract

Aluminium trichloride catalyses the expeditious direct conversion of tetrahydropyranyl ethers to silyl ethers. This one-step transformation is chemoselective versus deprotection of the acetal and hydrosilylation of unsaturated carbon–carbon bonds, and can also be applied to linear acetals. A possible mechanism is tentatively proposed.

---

Introduction

The replacement of one protecting group with another, a common process in the synthesis of polyfunctional molecules, usually requires two separate steps: deprotection and re-protection.¹ One-step conversion of one protecting group to another, when possible, saves time, material, and energy.

Hydroxyl-protecting groups have been extensively explored and are generally classified as giving rise to alkyl ethers, silyl ethers, acetals, or esters; among the most popular are those producing silyl ethers² (–SiR₃) or acetals (tetrahydropyranyl (THP), ethoxyethyl (EE), methoxymethyl (MOM), etc.). A number of methods are now available for direct conversions among the various types,³ yet only a couple concern the formation of silyl ethers from the widely used cyclic acetal (THP) ether.^1,4 Kim et al.⁵ transformed THP ethers into tert-butyldimethylsilyl (TBDMS) ethers by treatment with TBDMSOTf and dimethyl sulfide in dichloromethane. Using Ph₃P instead of Me₂S afforded just slightly decreased yields, but pyridine and Et₃N were ineffective. Primary and secondary alkyl or benzylic THP ethers responded well, yielding the corresponding TBDMS ethers in high yields under very mild conditions (−50 °C), but allylic and tertiary alkyl THP ethers were less responsive. Oriyama⁶ later reported that a mixture of trialkylsilyl trifluoromethanesulfonate and triethylamine converts THP ethers to the corresponding trialkylsilyl ethers at room temperature. Better yields were obtained with phenolic ethers than with aliphatic ethers, conversion of only primary and secondary examples of the latter being reported. The desired conversion also resulted from Sn(OTf)₂-catalysed reduction of THP ethers with a trialkylsilane, at least in the case of simple primary and secondary protected alcohols.

Despite their usefulness, both the above methods suffer from drawbacks (the use of noxious dimethylsulfide, or Lewis acid containing toxic tin, or competitive O-silylation of free hydroxyls by silyltriflate donors) and both afford unsatisfactory yields for sterically demanding aliphatic substrates. There is clearly a need for a “greener” and more generally applicable method.

AlCl₃ is one of the most powerful Lewis acids, and is also probably the most commonly used⁷ in synthetic laboratories and in the chemical industry as a catalyst for Friedel–Crafts reactions, polymerizations, acetal cleavage,⁸ and the hydrosilylation⁹ of unsaturated carbon–carbon bonds. Here we report the use of aluminum trichloride catalyst for the expedient, direct conversion of acetals into silyl ethers. In addition to being effective with primary, secondary, and tertiary alkyl THP ethers, and for a wide range of different silyl protecting groups (including some of the more commonly employed), this reaction is applicable to substrates with unprotected functional groups that are known to be reactive under AlCl₃/R₃SiH conditions, including alkenes and alkynes. It can be also applied to linear acetals (Scheme 1).

Scheme 1 Direct conversion of acetals to silyl ethers.

Results and discussion

Optimization of reaction conditions

With Oriyama's⁶ Sn(OTf)₂-catalysed reaction in mind, we initiated our study by screening a representative set of Lewis acids. We chose the conversion of 1-(2-tetrahydropyranyloxy)octane (1a₁) to 1-(dimethylphenylsilyloxy)octane (2a₁) as the model reaction (Table 1). Silane and catalyst (5 mol%) were mixed in CH₂Cl₂ at 0 °C, and the acetal was then added.¹⁰ As expected, Sn(OTf)₂ worked well for this simple THP-protected substrate, giving a yield of 81% (Table 1, entry 1). The titanium-based Lewis acids CpTiCl₂ and Ti(iOPr)₂ had no effect, while TiCl₄ led to decomposition of the starting material in less than 1 h (entries 2–4). BF₃·Et₂O produced a complex mixture, and InCl₃ afforded but a poor yield, the main product being deprotected octanol (3a) (entries 5 and 6). FeCl₃ gave a better yield (60%, entry 7), though inferior to that of Sn(OTf)₂; and EtAlCl₂ yet a better (74%, entry 8), but required a reaction time of 8 h. Finally, with AlCl₃ an excellent 91% yield was obtained in just half an hour (entry 9), and we proceeded to optimize the experimental conditions for this catalyst.

Table 1 Optimization of Lewis acid and reaction conditions

Entry	Lewis acid	mol%	Time (h)	Solvent	Yield^a (%)
a Isolated yield after column chromatography.b Deprotected octanol (3a) was also obtained.
1	Sn(OTf)2	5	2	CH2Cl2	81
2	CpTiCl2	5	5	CH2Cl2	—
3	Ti(iOPr)2	5	5	CH2Cl2	—
4	TiCl4	5	1	CH2Cl2	Decomp.
5	BF3·Et2O	5	2	CH2Cl2	10
6	InCl3	5	2	CH2Cl2	19^b
7	FeCl3	5	2	CH2Cl2	60
8	EtAlCl2	5	8	CH2Cl2	74
9	AlCl3	5	0.5	CH2Cl2	91
10	AlCl3	2.5	1	CH2Cl2	82
11	AlCl3	10	0.5	CH2Cl2	50^b
12	AlCl3	5	2	Toluene	85
13	AlCl3	5	5	THF	—
14	AlCl3	5	5	DMF	—

---

Decreasing the concentration of AlCl₃ to 2.5 mol% slowed the reaction and lowered the yield (entry 10), while increasing it to 10 mol% favoured deprotection over the desired conversion (entry 11). At this point we also noticed that the absence of water was critical for avoiding THP cleavage, and dried solvent and freshly sublimated AlCl₃ were accordingly used in all subsequent experiments. Trials with alternative solvents identified none better than dichloromethane: the reaction was slightly slower in toluene, and failed to occur to any detectable extent in the coordinating solvents THF and DMF (entries 12–14).

Scope, chemoselectivity and limitations

To evaluate the scope of the reaction on the silyl side we ran the reactions of 1-(2-tetrahydropyranyloxy)octane (1a₁) with an assortment of commercially available silanes (Table 2). Direct conversion proceeded smoothly in all cases, regardless of the steric and/or electronic properties of the silane: although slightly longer reaction times (1 h) were needed for silanes that were bulky (entries 7, 9 and 10) or oxygenated (entries 3 and 8), the yield of the silyl ether 2a_x was always excellent. From among all the silanes tested, PhMe₂SiH was selected for use thereafter in view of its excellent yield, easy visualization by TLC, and low cost.

Table 2 Hydrosilane screening

Entry	R3SiH	Time (h)	Product number	Yield^a (%)
a Isolated yield after column chromatography.
1	PhMe2SiH	0.5	2a1	91
2	BnMe2SiH	0.5	2a2	89
3	(EtO)Me2SiH	1	2a3	80
4	tBuMe2SiH	0.5	2a4	83
5	Ph3SiH	0.5	2a5	79
6	Et3SiH	0.5	2a6	86
7	iPr3SiH	1	2a7	93
8	(EtO)3SiH	1	2a8	78
9	tBu2MeSiH	1	2a9	80
10	tBu3SiH	1	2a10	79

---

To evaluate the scope of the reaction we tested a collection of THP ethers that included different functional groups (Tables 3, 4).

Table 3 Scope of the reaction for THP ethers with no unprotected hydroxyl groups

Entry	R–OTHP	Product number	Yield^a^,^b (%)
a Isolated yield after column chromatography.b Standard conditions: 0.05 eq. of AlCl3, 1.25 eq. of PhMe2SiH, 0.5 h.c 2.00 eq. of PhMe2SiH, 1.25 eq. of PhMe2SiH, 0.5 h.
1		2a1	91
2		2b1	90
3		2c1	89
4		2d1	88
5		2e1	85
6		2f1	80^c
7		2g1	81
8		2h1	80^c
9		2i1	97
10		2j1	81

---

Primary, secondary and even tertiary alkyl acetals (1x₁) were all converted to the corresponding dimethylphenylsilyl ethers 2x₁ in short time and excellent yields, as were allylic, benzylic and propargylic acetals, although an extra equivalent of hydrosilane was required for sterically demanding substrates, entries 6 and 8.

Of particular note, the reaction was compatible with halide, alkene, alkyne and aromatic functional groups, being completely chemoselective for conversion of the protecting group despite these same experimental conditions having been shown to effect the regio- and stereoselective hydrosilylation of alkenes and alkynes.⁹ Substrates with free hydroxyl groups were more problematic under standard conditions (PhMe₂SiH, AlCl₃, CH₂Cl₂, 0 °C), the THP-monoprotected 1,7-heptanediol 1k₁ evolved within minutes to deprotected 1,7-heptanediol (3k) instead of giving the desired 7-(phenyldimethylsilyloxy)-1-heptanol (2k₁); see Table 4, entry 1. The use of the bulky silane iPr₃SiH in toluene allowed useful yields of the corresponding silyl ether to be obtained – 58% in the case of 7-(triisopropylsilyloxy)-1-heptanol (2k₆)¹¹ (entry 2) and 70% in that of the bispropargylic substrate 4-(triisopropylsilyloxy)-but-2-yn-1-ol (2l₆) (entry 3). In this last case the final reaction mixture showed no traces of silane alcoholysis, reduction of the alcohol,¹² hydrosilylation of the alkyne, or cleavage of the acetal.

Table 4 Optimization of chemoselectivity for THP ethers with unprotected hydroxyl groups

Entry	HO–R–OTHP	R3SiH	Solvent	Product^a (%)
a Isolated yield after column chromatography.b 1,7-Heptanediol (3k) was also obtained.
1		PhMe2SiH	CH2Cl2	3k (100)
2		iPr3SiH	Toluene	2k6 (58)^b
3		iPr3SiH	Toluene	2l6 (70)

---

Finally, to explore the possible extension of the method to linear acetals, we subjected the methoxymethyl ether (MOM) 1a₂ and the ethoxyethyl ether (EE) 1a₃ to the standard conditions (Table 5). In these cases the desired product, silyl ether 2a₁, was accompanied by the alkyl ethers 4a_x due to the alternative cleavage of the acetals, the 4a_x : 2a₁ ratio being greater for the α-substituted acetal 1a₃ (27%) than for the α-unsubstituted 1a₂ (16%) (Table 5, entries 2 and 3). In both cases the global yield of 2a₁ and 4a_x exceeded 90%.

Table 5 Extension to linear acetals

Entry	Substrate	2a1 (%)	4ax (%)	Total yield^a (%)
a Isolated yield after column chromatography.
1		91		91
2		78		93
3		68		93

---

Mechanism

On the basis of the above experimental evidence, the tentative mechanism shown in Scheme 2 is proposed. Since pre-mixing of catalyst and silane seems to be critical for the efficiency of the reaction, the activation of the silane by aluminium through hydride abstraction appears to be a key step.¹³ Following that, two pathways are possible (Routes A and B), corresponding to the two ways in which the reactive silyl-aluminium species can coordinate to the acetal oxygen atoms to form the six-membered cyclic transition structure of a concerted mechanism¹⁴ in which charge pushing by one of the oxygens drives cleavage of the other acetal bond. Cleavage releases a silyl ether (2a₁ or 6a_x) and a carbocation (I or II) that subsequently evolves to compound 5a_x or 4a_x.¹⁵ For THP ethers only Route A proceed well, Route B requiring the opening of the pyrane ring; but for the linear acetals both pathways may proceed well, leading to the observed mixtures of compounds 2a₁ and 4a_x.

Scheme 2 Tentative reaction mechanism.

Conclusions

Summing up, we have developed an expedient procedure for the direct transformation of tetrahydropyranyl-protected alcohols into the corresponding silyl ethers by their reaction with hydrosilanes in the presence of catalytic amounts of AlCl₃. The advantages of this protocol – mild reaction conditions, short reaction times, applicability to a variety of substrates (including tertiary alcohols), high yield, and total chemoselectivity even in the presence of free hydroxyls or unsaturated functional groups – make it an attractive and useful addition to the present methodological armamentarium.

Acknowledgements

We thank the University of Santiago de Compostela, the Center for Research in Biological Chemistry and Molecular Materials (CIQUS), The Ministry of Economy and Competitiveness and the Xunta de Galicia for financial support. J. M. is a Juan de la Cierva fellow.

References

1. (a) T. V. Greene and P. G. M. Wuts, Protective Groups in Organic Synthesis, John Wiley and Sons, New York, 3rd edn, 1991, p. 17 and 297; (b) I. Fleming, Protecting groups, Grog Thieme, New York, 1994, p. 21; (c) J. R. Hanson, Protective Groups in Organic Synthesis, Blackwell Science Inc., Malden, MA, 1st edn, 1999, p. 37; (d) G. Sartori, R. Ballini, F. Bigi, G. Bosica, R. Maggi and P. Righi, Chem. Rev., 2004, 104, 199–250.
2. (a) J. W. Gillard, R. Fortin, H. E. Morton, C. Yoakim, C. A. Quesnelle, S. Daignault and Y. Guindon, J. Org. Chem., 1988, 53, 2602–2608; (b) J. S. Davies, C. L. Higginbotham, E. J. Tremeer, C. Brown and R. C. Treadgold, J. Chem. Soc., Perkin Trans. 1, 1992, 3043; (c) The reductive cleavage of aryl ethers by Lewis-base-activated silanes has also been reported: A. Fedorov, A. A. Toutov, N. A. Swisher and R. H. Grubbs, Chem. Sci., 2013, 4, 1640–1645.
3. For a general view see: (a) T. Oriyama, Journal of Synthetic Organic Chemistry, 1996, 490–502. For specific conversions: (b) Ch. S. Reddy, G. Smitha and S. Chandrasekhar, Tetrahedron Lett., 2003, 44, 4693–4695 and (c) B. Das, M. R. Reddy, K. R. Reddy, R. Ramu and P. Thirupathi, J. Mol. Catal. A: Chem., 2006, 248, 185–188 (OTHP to OAc); (d) T. Oriyama, M. Kimura and G. Koga, Bull. Chem. Soc. Jpn., 1994, 67, 885–887 (OPB to OMOM); (e) T. Oriyama, M. Kimura, M. Oda and G. Koga, Synlett, 1993, 437–440 (OBn to OAc); (f) T. Oriyama, K. Yatabe, Y. Kawada and G. Koga, Synlett, 1995, 45–46 (OPMB to OSi); (g) T. Oriyama, T. Watahiki, Y. Kobayashi, H. Hirano and T. Suzuki, Synth. Commun., 1999, 29, 2217–2223 (OAc to OSi); (h) K. Iwanami, K. Yano and T. Oriyama, Synthesis, 2005, 2669–2672 and (i) T. Oriyama, K. Noda and K. Yatabe, Synlett, 1997, 701–703 (OSi to OBn); (j) T. Oriyama, M. Oda, J. Gono and G. Koga, Tetrahedron Lett., 1994, 35, 2027–2030 (OSi to OAc); (k) T. Suzuki and T. Oriyama, Synthesis, 2001, 555–558 (OSi to OTHP).
4. W. E. Parham and E. L. Anderson, J. Am. Chem. Soc., 1948, 70, 4187–4189.
5. S. Kim and I. S. Kee, Tetrahedron Lett., 1990, 31, 2899–2900.
6. (a) T. Oriyama, K. Yatabe, S. Sugawara, Y. Machiguchi and G. Koga, Synlett, 1996, 523–525; (b) T. Suzuki, K. Ohashi and T. Oriyama, Synthesis, 1999, 1561–1563.
7. (a) S. D. Sarma, P. Pahari, S. Hazarika, P. Hazarika and M. Jyoti, ARKIVOC, 2013, 4, 1640–1645; (b) P. H. Groggins, Ind. Eng. Chem., 1931, 23, 152–160.
8. (a) M. T. Crimmins and S. W. Rafferty, Tetrahedron Lett., 1996, 37, 5649–5652. For a model system see: (b) E. D. Moher, P. A. Grieco and J. L. Collins, J. Org. Chem., 1993, 58, 3789–3790.
9. (a) N. Asao, T. Sudo and Y. Yamamoto, J. Org. Chem., 1996, 61, 7654–7655; (b) T. Sudo, N. Asao, V. Gevorgyan and Y. Yamamoto, J. Org. Chem., 1999, 64, 2494–2499; (c) Y.-S. Song, B. R. Yoo, G.-H. Lee and I. N. Jung, Organometallics, 1999, 18, 3109–3115; (d) Y. Liu, S. Yamazaki and S. Yamabe, J. Org. Chem., 2005, 70, 556–561; (e) N. Kato, Y. Tamura, T. Kashiwabara, T. Sani and M. Tanaka, Organometallics, 2010, 29, 5274–5282.
10. Pre-mixing of catalyst and silane was critical for the success of the reaction, since direct mixture of the three reagents led mainly to the free alcohol.
11. A small amount of 1,7-heptanediol (3k) was also obtained.
12. The system HSiR₃/B(C₆F₅)₃ has been used for both the silanolysis and the reduction of alcohols: (a) J. M. Blackwell, K. L. Foster, V. H. Beck and W. E. Piers, J. Org. Chem., 1999, 64, 4887–4892; (b) V. Gevorgyan, M. Rubin, S. Benson, J.-X. Liu and Y. Yamamoto, J. Org. Chem., 2000, 65, 6179–6186.
13. Silyl cation formation has been reviewed in: J. B. Lambert, L. Kania and S. Zhang, Chem. Rev., 1995, 95, 1191–1201.
14. Piers has proposed a sequential mechanism for Lewis-acid-catalysed silanolysis of alcohols (see ref. 12), but in our case AlCl₃ is not able to promote the O-silylation of a free hydroxyl group.
15. The non-detection of dimethyl ether (5a₂) and diethyl ether (5a₃) is attributed to their extreme volatility. The alcoxysilanes 6a_x were possibly eliminated during the working-up of the reaction.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed experimental procedures. See DOI: 10.1039/c4ra00655k

---