Meinwald-type rearrangement of monosubstituted epoxides to methyl ketones using an [Al porphyrin]⁺[Co(CO)₄]⁻ catalyst

Abstract

A Meinwald-type rearrangement of monosubstituted epoxides to methyl ketones using a well-defined aluminum porphyrin catalyst is reported. This reaction shows good functional group tolerance under mild reaction conditions to selectively give methyl ketones in excellent yields.

---

Epoxides are versatile intermediates in organic chemistry due to their inherent high reactivity and synthetic availability. The oxirane polarity and ring strain enable these moieties to undergo a variety of useful reactions, including nucleophilic ring opening, deoxygenation, reduction, and a variety of rearrangements.¹ One reaction that has attracted a great deal of attention due to its high efficiency and synthetic potential is the Lewis-acid induced isomerization of epoxides to carbonyl compounds, known as the Meinwald rearrangement (Fig. 1).^1,2

Fig. 1 Mechanism of the Meinwald rearrangement of epoxides activated by a Lewis acid.

Despite previous work to make this transformation synthetically useful, multiple products are commonly observed due to unselective ring opening and substituent migration.³ While the migratory aptitude of various substituents can often be tuned by the Lewis acid and conditions employed,⁴ selective ring opening using Lewis acids generally relies on substrate bias to form a single, stable tertiary (Fig. 1, R³ and R⁴ ≠ H) or benzylic (R³ or R⁴ = aryl) carbocation.⁵ Similarly, terminal epoxides generally yield aldehyde products because the instability of primary carbocations (R³ = R⁴ = H) disfavors ketone formation.⁶

Methyl ketones have been observed as the major product of the isomerization of monosubstituted epoxides when lithium iodide⁷ or transition metals—such as Fe,⁸ Ru,⁹ Co,¹⁰ Rh,¹¹ or Pd¹²—are used to alter the mechanism of ring opening. While important progress has been made in this field, existing systems still have drawbacks, including limited substrate scope, modest yields, high reaction temperature, and high catalyst loading.

During our group's investigation of carbonylative ring expansion of epoxides to β-lactones using catalysts of the form [Lewis Acid]⁺[Co(CO)₄]⁻,¹³ we observed concurrent ketone formation, especially at low pressures of carbon monoxide (Fig. 2).¹⁴ We decided to explore this reaction in more detail and selected the porphyrin-based catalyst¹⁵ [pClTPPAl(THF)₂]⁺[Co(CO)₄]⁻ (1) because of its straightforward synthesis and known functional group tolerance.¹⁶ Herein we report a mild, high yielding, and chemo- and regioselective rearrangement of monosubstituted epoxides to methyl ketones.

Fig. 2 Catalytic (A) carbonylation and (B) isomerization of mono-substituted epoxides.

Optimization of the reaction conditions with 1-hexene oxide (2a, Table 1) showed that the best solvent for this transformation was THF (entries 1–6). A concentration screen revealed 1.0 M as the most effective, with both higher and lower concentrations resulting in lower conversions (entries 6–11). We initially hypothesized that at higher concentrations the polarity of the ketone product might affect the rate of the reaction; however, adding 1–2 equivalents of acetone at the beginning of the reaction did not appreciably affect the conversion.¹⁷ Increasing the catalyst loading from 1 to 2 mol% resulted in full conversion of epoxide at room temperature (entry 12).

Table 1 Evaluation of solvents and epoxide concentration in the isomerization of epoxide 2a by [pClTPPAl(THF)₂]⁺[Co(CO)₄]⁻ (1)

Entry	Mol% catalyst 1	Solvent	Concentration 2a (M)	Conv.^a (%)
a Determined by ^1H NMR spectroscopy of the crude reaction mixture.
1	1	1,4-Dioxane	0.5	21
2	1	Benzene	0.5	27
3	1	Toluene	0.5	31
4	1	Ether	0.5	36
5	1	Hexanes	0.5	28
6	1	THF	0.5	45
7	1	THF	1.0	82
8	1	THF	1.5	72
9	1	THF	2.0	67
10	1	THF	4.0	71
11	1	Neat	8.7	38
12	2	THF	1.0	>99

---

Once viable reaction conditions were found, the substrate scope of the isomerization was explored (Chart 1). Aliphatic epoxides, including those containing substantial α-branching, were rearranged in excellent yields (≥89%, 3a–i), and no aldehyde or alcohol side products were observed. Capillary GC analysis of the crude reaction mixtures revealed trace amounts of β-lactone formation, which would result in partial deactivation of 1via loss of a CO ligand bound to the cobaltate counterion.¹⁷

Chart 1 Isolated yields and substrate scope for the rearrangement of monosubstituted epoxides catalyzed by 1. ^a Quantitative GC yield versus dodecane internal standard. ^b 5 mol% 1 used. ^c 2 mol% 1 per epoxide moiety used based on the molecular weight of the repeat unit.

As previously reported,¹⁶ the aluminum porphyrin catalyst 1 tolerates a variety of functionality under the same mild reaction conditions while maintaining high isolated yields (3j–p). Notably, this system tolerates a terminal olefin (3q), which demonstrates its potential over other ketone-forming reactions such as the Wacker oxidation. Even though this substrate required 5 mol% catalyst to achieve full conversion of epoxide, we were able to access the methyl ketone in the presence of a terminal olefin by utilizing the difference in reactivity of an epoxide and a double bond. Conversely, the Wacker oxidation would result in a diketone from the corresponding diene.¹⁸

In addition to small molecule substrates, this transformation works cleanly on polycarbonate 2r using the same reaction conditions with 2 mol% 1 per epoxide moiety. This substrate is prepared via a Zn-catalyzed copolymerization of vinylcyclohexene dioxide and CO₂.¹⁹ As expected, the molecular weight and molecular weight distribution of the polymer did not change during this transformation, but a clear change in the NMR and IR spectra indicate clean conversion to the ketone-functionalized polymer (3r).¹⁷

A limitation of this method was revealed during our attempt to rearrange terminal epoxides with aromatic substituents. Styrene oxide resulted in a 1 : 1.6 mixture of acetophenone and phenylacetaldehyde due to competing S_N2 attack at the sterically unhindered methylene and the adjacent benzylic methine (Table 2, entry 1). While a variety of catalysts are known to selectively isomerize styrene oxide to either acetophenone²⁰ or phenylacetaldehyde,^5c,d,21 cobaltate-based catalysts have been reported to give a mixture of products with this substrate.²² Attempts to rearrange 2,3-disubstituted epoxides with 1 led to low conversions under standard reaction conditions (entry 2).

Table 2 Substrate limitations of epoxide rearrangement using 1^a

Entry	Substrate	Product(s)	Conv.^b (%)
a Conditions: [epoxide] = 1.0 M in THF, 2 mol% catalyst 1, 22 °C, 18 h. b Conversion determined by GC or ^1H NMR analysis of the crude reaction mixture. c Determined by GC analysis.
1			90 (3s : 4s 1 : 1.6)^c
2			31

---

We propose that 1 reacts via a similar mechanism as that reported by Eisenmann^10a and Kagan,^10c in which Co₂(CO)₈ was combined with MeOH to form [Co(MeOH)₆]⁺²[Co(CO)₄]₂⁻in situ. Kagan's proposed catalytic cycle involves ring opening via a Lewis acid-assisted S_N2 attack of the cobaltate at the least substituted carbon followed by classical β-hydride elimination from cobalt (Fig. 2B). This proposal was based on the observed epoxide reactivities and product distributions, but the full mechanism has yet to be thoroughly studied.

We utilized our well-defined complexes to probe the synergistic nature of this two-component catalyst system by replacing either the aluminum porphyrin with the non-Lewis acidic sodium cation or the cobaltate with the non-nucleophilic tetraphenyl borate anion. Neither of these variants was able to promote the isomerization of 2a (Table 3, entries 2 and 3). Similarly, pClTPPAlCl, the precursor to catalyst 1, does not lead to any conversion to ketone (entry 4). This conclusively shows that both the Lewis acid cation and nucleophilic anion are necessary for this reaction, which supports the first step in Kagan's proposed mechanism and parallels the ring-opening step of carbonylation (Fig. 2A),²³ which catalyst 1 is also known to do. The fact that cobalt is required to effect β-hydride elimination clearly distinguishes this mechanism from the Lewis acid pathway depicted in Fig. 1.

Table 3 Catalyst control experiments for the isomerization of 2a

Entry	Catalyst	Conv.^a (%)
a Determined by ^1H NMR spectroscopy of the crude reaction mixture.
1	[pClTPPAl(THF)2]^+[Co(CO)4]^− (1)	>99
2	[pClTPPAl(THF)2]^+[BPh4]^−	<1
3	Na[Co(CO)4]	<1
4	pClTPPAlCl	<1

---

The β-hydride elimination step was investigated by subjecting isobutylene oxide to standard reaction conditions.¹⁷ This 2,2-disubstituted epoxide results in a tertiary alkoxide upon S_N2 ring opening that cannot undergo β-hydride elimination. As expected, no methyl shift to form a ketone was observed. Less than 2% conversion to isobutyraldehyde occurred, which could arise from either an S_N1-type attack of cobaltate²⁴ or the classic Lewis-acid mediated mechanism (Fig. 1).²⁵

We have presented a new application of a readily available, well-defined catalyst for the isomerization of monosubstituted epoxides to methyl ketones in excellent yields. Compared to previous catalyst systems, 1 is effective under mild conditions, low catalyst loading, and has good functional group tolerance. Current work focuses on applying more recent, selective catalysts to the rearrangement of 2,3-disubstituted epoxides.

We are grateful to the Department of Energy (DE-FG02-05ER15687) and the National Science Foundation (DGE-1144153, fellowship to J.R.L.) for funding.

Notes and references

1. For reviews see: (a) J. G. Smith, Synthesis, 1984, 629; (b) C.-Y. Huang and A. G. Doyle, Chem. Rev., 2014, 114, 8153; (c) S. C. Bergmeier and D. J. Lapinsky, Prog. Heterocycl. Chem., 2013, 25, 47.
2. (a) J. Meinwald, S. S. Labana and M. S. Chadha, J. Am. Chem. Soc., 1963, 85, 582; For a review see: (b) B. Rickborn, in Comprehensive Organic Synthesis, ed. B. M. Trost, I. Fleming and G. Pattenden, Pergamon, Oxford, 1991, vol. 3, 733.
3. (a) H. O. House, J. Am. Chem. Soc., 1955, 77, 3070; (b) H. O. House, J. Am. Chem. Soc., 1955, 77, 5083; (c) B. Rickborn and R. M. Gerkin, J. Am. Chem. Soc., 1971, 93, 1693; (d) A. Miyashita, T. Shimada, A. Sugawara and H. Nohira, Chem. Lett., 1986, 1323.
4. (a) K. Maruoka, N. Murase, R. Bureau, T. Ooi and H. Yamamoto, Tetrahedron, 1994, 50, 3663; (b) F. Martínez, C. del Campo and E. F. Llama, J. Chem. Soc., Perkin Trans. 1, 2000, 1749.
5. For selected examples see: (a) R. Sudha, K. M. Narasimhan, V. Saraswathy and S. Sankararaman, J. Org. Chem., 1996, 61, 1877; (b) K. Suda, S. Nakajima, Y. Satoh and T. Takanami, Chem. Commun., 2009, 1255; (c) E. Ertürk, M. Göllü and A. S. Demir, Tetrahedron, 2010, 66, 2373; (d) V. Gudla and R. Balamurugan, Tetrahedron Lett., 2012, 53, 5243; (e) D. J. Vyas, E. Larionov, C. Besnard, L. Guénée and C. Mazet, J. Am. Chem. Soc., 2013, 135, 6177; (f) N. Humbert, D. J. Vyas, C. Besnard and C. Mazet, Chem. Commun., 2014, 50, 10592.
6. (a) A. Yanagisawa, K. Yasue and H. Yamamoto, J. Chem. Soc., Chem. Commun., 1994, 2103; (b) K. Suda, K. Baba, S. Nakajima and T. Takanami, Tetrahedron Lett., 1999, 40, 7243; (c) A. Procopio, R. Dalpozzo, A. De Nino, M. Nardi, G. Sindona and A. Tagarelli, Synlett, 2004, 2633.
7. Z. An, R. D'Aloisio and C. Venturello, Synthesis, 1992, 1229.
8. T. Kauffmann, C. Neiteler and G. Neiteler, Chem. Ber., 1994, 127, 659.
9. C.-L. Chang, M. P. Kumar and R.-S. Liu, J. Org. Chem., 2004, 69, 2793.
10. (a) J. L. Eisenmann, J. Org. Chem., 1962, 27, 2706; (b) J.-Y. Kim and T. Kwan, Chem. Pharm. Bull., 1970, 18, 1040; (c) J. Prandi, J. L. Namy, G. Menoret and H. B. Kagan, J. Organomet. Chem., 1985, 285, 449.
11. D. Milstein, J. Am. Chem. Soc., 1982, 104, 5227.
12. (a) S. Kulasegaram and R. J. Kulawiec, J. Org. Chem., 1994, 59, 7195; (b) S. Kulasegaram and R. J. Kulawiec, J. Org. Chem., 1997, 62, 6547 (mechanism).
13. T. L. Church, Y. D. Y. L. Getzler, C. M. Byrne and G. W. Coates, Chem. Commun., 2007, 657.
14. J. W. Kramer, E. B. Lobkovsky and G. W. Coates, Org. Lett., 2006, 8, 3709.
15. Iron, ruthenium, and chromium porphyrin catalysts have precedence in the Meinwald rearrangement literature, but they generally operate via a Lewis acidic mechanism which results in aldehyde products from terminal epoxides (Fig. 1). For Fe see ref. 6b and (a) T. Takanami, R. Hirabe, M. Ueno, F. Hino and K. Suda, Chem. Lett., 1996, 1031; For Ru see: (b) J. Chen and C.-M. Che, Angew. Chem., Int. Ed., 2004, 43, 4950; (c) G. Jiang, J. Chen, H.-Y. Thu, J.-S. Huang, N. Zhu and C.-M. Che, Angew. Chem., Int. Ed., 2008, 120, 6740; For Cr see: (d) K. Suda, T. Kikkawa, S. Nakajima and T. Takanami, J. Am. Chem. Soc., 2004, 126, 9554.
16. (a) J. M. Rowley, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2007, 129, 4948; (b) M. Mulzer, B. J. Tiegs, Y. Wang, G. W. Coates and G. A. O'Doherty, J. Am. Chem. Soc., 2014, 136, 10814.
17. See ESI† for additional information.
18. For a Wacker oxidation on a similar diene substrate see: T. Mitsudome, T. Umetani, N. Nosaka, K. Mori, T. Mizugaki, K. Ebitani and K. Kaneda, Angew. Chem., Int. Ed., 2005, 45, 481.
19. W. C. Ellis, Y. Jung, M. Mulzer, R. DiGirolamo, E. B. Lobkovsky and G. W. Coates, Chem. Sci., 2014, 5, 4004.
20. E. C. H. Sykes, M. S. Tikhov and R. M. Lambert, Catal. Lett., 2002, 78, 7.
21. (a) I. Karamé, M. L. Tommasino and M. Lemaire, Tetrahedron Lett., 2003, 44, 7687; (b) M. W. C. Robinson, K. S. Pillinger and A. E. Graham, Tetrahedron Lett., 2006, 47, 5919; (c) B. C. Ranu and U. Jana, J. Org. Chem., 1998, 63, 8212.
22. A. Cabrera, F. Mathé, Y. Castanet, A. Mortreux and F. Petit, J. Mol. Catal., 1991, 64, L11.
23. T. L. Church, Y. D. Y. L. Getzler and G. W. Coates, J. Am. Chem. Soc., 2006, 128, 10125.
24. Both S_N2 and S_N1-type attacks were observed during the carbonylation of isobutylene oxide: (a) V. Mahadevan, Y. D. Y. L. Getzler and G. W. Coates, Angew. Chem., Int. Ed., 2002, 41, 2781; (b) Y. D. Y. L. Getzler, V. Mahadevan, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2002, 124, 1174.
25. This alternative pathway is favored when using a non-nucleophilic anion so that conversion to isobutyraldehyde increased to 62% with [pClTPPAl(THF₂)]⁺[BPh₄]⁻.

---

Footnote

† Electronic supplementary information (ESI) available: Expanded Tables S1 and S2, detailed experimental procedures, characterization data. See DOI: 10.1039/c4qo00324a

---