Oxidative Rearrangement of Tertiary Propargylic Alcohols

Abstract

An oxidative rearrangement of tertiary alcohols mediated by m-CPBA is described that generates tetrasubstituted alkenes with a carboxylic acid substituent. The mechanism of the reaction is proposed to proceed through epoxidation of the alkyne to form an oxirene that undergoes a 1,2-aryl shift.

Key words

oxidation - rearrangement - propargylic alcohols - carboxylic acids - peroxy acid

The oxidation of acetylenes with peroxy acids is capricious and can lead to the formation of a variety of products resulting from either oxidation, rearrangement, or addition depending on the reaction conditions used. In addition, only low yields of products have been reported in the scant number of examples on this topic. McDonald and Schwab reported that phenylacetylene was oxidised to five different compounds by perbenzoic acid: ethyl phenylacetate, methyl phenylacetate, benzaldehyde, benzoic acid, and methyl benzoate (Scheme [1]).[ ¹ ] The overall yield reported was greater than 100%, as the benzoic acid could also have come from reduction of the perbenzoic acid.

Scheme 1 McDonald and Schwab’s oxidation reaction

Stille and Whitehurst reported the oxidation of diphenylacetylene by perbenzoic acid and m-CBPA and obtained a mixture of products in both cases including benzil and benzoic peroxyanhydride with ratios which were found to be solvent dependent (Scheme [2]).[ ² ] The mechanism for the formation of these products was proposed to proceed through epoxidation of the alkyne followed by either ring opening or rearrangement to the ketene and subsequent addition or elimination processes.[ ³ ]

As part of a wider programme concerned with the reactivity of polyvalent iodine species, we became interested in the rearrangement of tertiary propargyl alcohols. Tertiary propargyl alcohols are useful synthetic intermediates that have found widespread use in organic synthesis.[ ⁴ ] In particular, the rearrangement of these compounds has garnered significant attention.[5] [6] During these studies we discovered a hypoiodite-induced rearrangement to α-iodoenones (Scheme [3])[ ⁷ ] as well as the reaction detailed herein.

Scheme 2 Stille and Whitehurst’s oxidation reaction

Scheme 3 Our group’s hypoiodite-initiated rearrangement reaction

With the aim of extending our work utilising in situ generated iodine(III) species, we aimed to effect an oxidative rearrangement of propargyl alcohols 1 into diketones 2 through a semipinacol-type process. Accordingly, 1a was treated with 20 mol% of iodobenzene, 1.5 equivalents of m-CPBA, and 1.5 equivalents of 4-toluenesulfonic acid monohydrate in acetonitrile at room temperature (Table [1]).[ ⁸ ] However, instead of the anticipated oxidative rearrangement, elimination of water occurred to generate enyne 4 (Table [1], entry 1). Replacement of the strong acid with acetic acid led to no conversion whatsoever (Table [1], entry 2). However, treatment with trifluoroacetic acid also led to formation of enyne 4 (Table [1], entry 3). At this point, it was decided that an acid with a pK _a in between that of trifluoroacetic acid and acetic acid was necessary. Accordingly, trichloroacetic acid was tested, and a new product was formed (Table [1], entry 4). Analysis suggested that this was not the anticipated product 2a but that an alternative oxidative rearrangement had taken place to generate the enoic acid 3a. This rearrangement is completely unprecedented for this substrate and involves a 1,2-aryl migration instead of the anticipated 1,2-alkyl migration. Only a low yield of 3a was obtained, however the remainder of the mass balance was unreacted starting material. Attempts to improve the conversion, and hence the yield of the product, were unsuccessful. Using either 2- or 4-iodoanisole as the catalyst led to complete reaction inhibition (Table [1], entries 5 and 6), whereas 2-iodonitrobenzene led to complete conversion into enyne 4 (Table [1], entry 7). During these studies it became evident that iodobenzene was not required and that the reaction was mediated by m-CPBA and not a hypervalent iodine species (Table [1], entry 8). Changing the solvent from acetonitrile to dichloromethane, TFE, or nitromethane afforded the dehydrated product 4 whilst methanol gave a mixture of unidentifiable compounds. Using Oxone^® instead of m-CPBA led to complete conversion into enyne 4. Epoxidation of enyne 4 was not observed in any reactions.

Table 1 Optimization of the Reaction Conditions

Entry	Acid	Catalyst	Yield (%)a
1	p-toluenesulfonic acid monohydrate	iodobenzene	4 83
2	acetic acid	iodobenzene	0
3	trifluoroacetic acid	iodobenzene	4 80
4	trichloroacetic acid	iodobenzene	3a 20
5	trichloroacetic acid	2-iodoanisole	0
6	trichloroacetic acid	4-iodoanisole	0
7	trichloroacetic acid	2-iodonitrobenzene	4 78
8	trichloroacetic acid	–	3a 20

^a Yield of isolated product.

The rearrangement of a few derivatives with these conditions was investigated in turn (Scheme [4]).[ ⁹ ] With a 4-methyl substituent, substrate 1b rearranged to provide 42% yield of isolated product 3b. With a 4-ethyl group, 32% yield of 3c was obtained and with a 4-tert-butyl group 40% yield of 3d was obtained. Substrates 1e and 1f containing more strongly electron-donating and electron-withdrawing aromatic substituents, respectively, where synthesised and submitted to the same reaction conditions but no desired products could be isolated. The rearrangement of the cyclohexyl substrate 1g proceeded with moderate efficiency furnishing 37% of the acid 3g. However, cycloheptyl substrate 1h failed to rearrange, and a mixture of unidentified products was obtained.

Scheme 4 Scope of the rearrangement reaction

The structure of the products was determined with the aid of NOESY and HMBC NMR experiments (Figure [1]). The NOESY spectrum showed an interaction between the arene CH and a CH₂ in the aliphatic ring. The HMBC showed interactions between the alkene C and all four of the CH₂ groups in the ring. Subsequently, the methyl ester of compound 3a was synthesised and the analytical data matched those in the literature.[ ¹⁰ ]

Figure 1 Structure determination by NMR experiments

Methyl ether 5 was prepared and subjected to the reaction conditions, however, no reaction took place (Scheme [5]),[ ¹¹ ] thus demonstrating that the presence of the hydroxyl group is crucial for the reaction to occur. Presumably, the hydroxyl group preorganises and directs the epoxidation of the alkyne through a hydrogen-bonding interaction.[ ¹² ]

Scheme 5 Mechanistic investigation

The mechanism of this rearrangement is proposed to proceed through hydroxyl-directed epoxidation of the alkyne to form an oxirene that undergoes a 1,2-aryl shift to form a ketene intermediate (Scheme [6]). Addition of water followed by elimination of water generates the final product. An acid is required for the reaction to occur, however, a fine balance exists between formation of products 3 and 4.

Scheme 6 Postulated rearrangement mechanism

In conclusion, an unprecedented oxidative rearrangement of tertiary propargylic alcohols to enoic acids is presented. This process converts readily accessible compounds into tetrasubstituted alkenes with a carboxylic acid substituent. The reaction likely proceeds through a hydrogen-bond directed alkyne epoxidation followed by a 1,2-aryl migration.

Acknowledgment

We thank the University of Huddersfield for funding.

• References and Notes

• 1 McDonald RN, Schwab PA. J. Am. Chem. Soc. 1964; 86: 4866
• 2 Stille JK, Whitehurst DD. J. Am. Chem. Soc. 1964; 86: 4871
• 3a Ortiz de Montellano PR, Kunze KL. J. Am. Chem. Soc. 1980; 102: 7373
• 3b Crandall JK, Conover WW. II. J. Org. Chem. 1978; 43: 1323
• 3c Csizmadia IG, Font J, Strausz OP. J. Am. Chem. Soc. 1968; 90: 7360
• 4 For a review, see: Bauer EB. Synthesis 2012; 44: 1131

For example, see:
• 5a Angara GJ, Bovonsombat P, McNelis E. Tetrahedron Lett. 1992; 33: 2285
• 5b Bovonsombat P, McNelis E. Tetrahedron 1993; 49: 1525
• 5c Chen J.-M, Huang X. Synthesis 2004; 2459
• 5d Bovonsombat P, McNelis E. Synth. Commun. 1995; 25: 1223
• 5e Djuardi E, Bovonsombat P, McNelis E. Tetrahedron 1994; 50: 11793
• 5f Bovonsombat P, McNelis E. Tetrahedron Lett. 1993; 34: 4277
• 5g Bovonsombat P, McNelis E. Tetrahedron Lett. 1993; 34: 8205
• 6 For a review of the Meyer–Schuster rearrangement, see: Engel DA, Dudley GB. Org. Biomol. Chem. 2009; 7: 4149
• 7 Moran WJ, Rodríguez A. Org. Biomol. Chem. 2012; 10: 8590
• 8 Rodríguez A, Moran WJ. Org. Lett. 2011; 13: 2220
• 9 Typical Experimental Procedure 1-[(4-tert-Butylphenyl)ethynyl]cyclopentanol (1a, 50 mg, 0.21 mmol), m-CPBA (54 mg, 0.31 mmol), and trichloroacetic acid (51 mg, 0.31 mmol) were dissolved in MeCN (1 mL) at r.t. under a nitrogen atmosphere. The mixture was stirred overnight until solid precipitation was evident (typically less than 15 h). The reaction mixture was quenched with sat. aq Na₂S₂O₃ solution and extracted with CH₂Cl₂. The organic layer was washed with sat. aq NaHCO₃ solution, dried over MgSO₄, filtered, and concentrated under reduced pressure. The residue was purified by flash chromatography (silica gel; 20:1 PE–EtOAc) to afford 2-(4-tert-butylphenyl)-2-cyclopentylideneacetic acid (3a) as a pale yellow solid (22 mg, 40%); mp 193–195 °C. IR (neat): 1252 (s), 1284 (s), 1618 (m), 1674 (s), 2955 (w) cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 1.32 (9 H, s), 1.58 (2 H, quin, J = 6.8 Hz), 1.76 (2 H, quin, J = 6.9 Hz), 2.23 (2 H, t, J = 7.2 Hz), 2.89 (2 H, t, J = 7.3 Hz), 7.10 (2 H, d, J = 8.4 Hz), 7.35 (2 H, d, J = 8.4 Hz), 11.35 (1 H, br). ¹³C NMR (100 MHz, CDCl₃): δ = 26.1, 27.1, 31.7 (3 C), 34.8, 34.9, 36.5, 125.1, 125.4 (2 C), 129.2 (2 C), 135.9, 149.9, 168.1, 173.0. MS: m/z = 281.2 [M + 23]. HRMS: m/z calcd for C₁₇H₂₂NaO₂: 281.1512; found: 281.1518.
• 10 Martín-Vaca B, Rudler H. J. Chem. Soc., Perkin Trans. 1 1997; 3119
• 11 Zhang H, Fu X, Chen J, Wang E, Liu Y, Li Y. J. Org. Chem. 2009; 74: 9351
• 12 For a review, see: Hoveyda AH, Evans DA, Fu GC. Chem. Rev. 1993; 93: 1307