Report

Primary Alcohols from Terminal Olefins: Formal Anti-Markovnikov Hydration via Triple Relay Catalysis

See allHide authors and affiliations

Science  16 Sep 2011:
Vol. 333, Issue 6049, pp. 1609-1612
DOI: 10.1126/science.1208685

Abstract

Alcohol synthesis is critical to the chemical and pharmaceutical industries. The addition of water across olefins to form primary alcohols (anti-Markovnikov olefin hydration) would be a broadly useful reaction but has largely proven elusive; an indirect hydroboration/oxidation sequence requiring stoichiometric borane and oxidant is currently the most practical methodology. Here, we report a more direct approach with the use of a triple relay catalysis system that couples palladium-catalyzed oxidation, acid-catalyzed hydrolysis, and ruthenium-catalyzed reduction cycles. Aryl-substituted terminal olefins are converted to primary alcohols by net reaction with water in good yield and excellent regioselectivity.

Olefin hydration, the addition of water across a carbon–carbon double bond, is an important industrial process for the synthesis of alcohols (1) and can be readily catalyzed by acids, metal oxides, zeolites, and clays (2). However, in accord with Markovnikov’s rule, the proton bonds to the less substituted carbon in these processes (3), and thus, primary alcohols (except ethanol) are difficult to obtain (Fig. 1, equation 1). Given the broad usefulness of primary alcohols in bulk/fine chemical and pharmaceutical industries (1), there is a compelling need to develop selective catalysts for direct anti-Markovnikov hydration of alkenes (4, 5). Currently, a popular indirect protocol involving hydroboration/oxidation (3) affords hydration products with anti-Markovnikov regioselectivity (Fig. 1, equation 2). This two-step process requires a stoichiometric amount of borane reagents and generates boron waste that is difficult to recycle. Moreover, the peroxides used in the oxidation step raise safety concerns for large-scale production. A hydroformylation/reduction sequence can also produce primary alcohols albeit through a homologation process (Fig. 1, equation 3) (6, 7). Campbell et al. recently developed an interesting strategy using a Pd-catalyzed allylic oxidation/ester hydrolysis/olefin reduction sequence to achieve the transformation of terminal olefins to primary alcohols (8). Although effective, this approach requires a three-step operation.

Fig. 1

Synthesis of alcohols from olefins. R, alkyl or aryl groups; R′, hydrogen or alkyl groups.

In contrast to anti-Markovnikov olefin hydroamination (9) and alkyne hydration (10), two closely related reactions, very limited success has been achieved toward anti-Markovnikov olefin hydration (Fig. 1, equation 4) (5, 9). In 1986, it was reported that trans-PtHCl(PMe3)2 (Me, methyl) was able to catalyze hydration of 1-hexene to 1-hexanol (11); unfortunately, this work was difficult to reproduce (5, 12), as was the method in a later report (13). Reliable one-step catalytic protocols are currently limited to certain activated classes of olefin (14, 15). Here, we describe an effective method for the direct synthesis of primary alcohols from nonactivated terminal olefins (defined as non-Michael reaction acceptors) using a triple relay catalysis (16) system, as progress toward an ideal system for catalytic anti-Markovnikov olefin hydration.

Our strategy for anti-Markovnikov olefin hydration is based on a two-catalyst cooperative system involving an oxidation cycle followed by a reduction cycle (Fig. 2). Studies of the Wacker oxidation (17) have established that PdII salts such as PdCl2 (X = Cl) can oxidize olefins in the presence of water to produce a carbonyl compound, an acid (HCl), and a palladium hydride (H-Pd-Cl). Precedents also exist for certain metal hydrides (M-H; M = Ru, Ir, Fe, Ni, Pt, etc.) to readily reduce carbonyl compounds, while in the process forming metal alkoxides that can be subsequently protonated by an acid (HX) to give an alcohol and metal salts (M-X) (18). In principle, these two known processes could be combined through a hydride transfer from Pd to M, in which the carbonyl compound and the acid generated from the oxidation cycle constitute the reactants in the reduction cycle, to provide a facile catalytic methodology for the hydration of olefins. The success of this strategy relies on three criteria: First, the oxidation of olefins must be selective for aldehyde products, as normal Wacker oxidation favors methyl ketones (Markovnikov addition) (17). Second, the oxidation cycle must be compatible with the reduction cycle. Third, migration of the hydride from Pd to M should be facile. In this work, we focus on addressing the challenges of anti-Markovnikov selectivity and compatibility of the two cycles; thus, we developed a modified catalytic system by adding both oxidant and reductant to turn over the two catalysts individually (Fig. 2). Success in this modified system provides critical information about the feasibility of the ideal system.

Fig. 2

Proposed cooperative catalytic system for alcohol synthesis from olefins and water. Asterisk: In the ideal system, either the hydride would be directly transferred from Pd to M or the oxidant [O] and the reductant [H] would be coupled with each other. X, anionic ligands, such as chlorides and acetates.

We chose styrene as the initial substrate because under the acid-catalyzed hydration conditions, the secondary alcohol, 1-phenylethanol, is essentially the only product formed due to the generation of a stable benzyl cation (Fig. 1, equation 1) (19). Furthermore, under Brown’s classic hydroboration conditions, styrene is a challenging substrate to functionalize with high anti-Markovnikov selectivity (the selectivity with diborane is only 4.2:1, whereas catechol borane gives 11.5:1 selectivity) (20). In addition, the product, 2-phenylethanol, is an important ingredient in artificial flavors, perfumery, and soaps (21). Few successful aldehyde-selective Wacker reactions with styrene derivatives have been reported [for a review, see (22)]. When t-BuOH is used as a solvent, high aldehyde selectivities in Wacker oxidations have been well established, albeit in low yields. In the case of styrene, only a 9% yield of phenylacetaldehyde was obtained (2225). Recently, Wright et al. reported better aldehyde selectivity (6.4:1) with styrene, using PdCl2 as the catalyst and DMF/H2O (DMF, dimethylformamide) as the mixed-solvent under an inert atmosphere; the chloride ligands were proposed to play an integral role in the regioselectivity (26). We envisioned that by combining both the t-BuOH solvent effect and the chloride-ligand factor, we would further enhance the anti-Markovnikov selectivity. For the reduction cycle, we selected a combination of i-PrOH and Shvo’s catalyst (27), because i-PrOH can serve as an inexpensive, clean, and safe reductant via metal-catalyzed transfer hydrogenation [for a review series, see (28)], and Shvo’s complex is commonly used as a catalyst for transfer hydrogenation of carbonyl compounds and is also known to tolerate aqueous conditions (29).

To our delight, after initial optimization of the reaction conditions, we obtained 2-phenylethanol (2a) in 77% yield (Fig. 3, equation 5) with exceptionally high anti-Markovnikov selectivity (38:1) (Fig. 3, equation 6). Subsequently, we investigated the role of each reactant through a series of control experiments (see table S1). Under the standard conditions, the reaction proceeded with an excellent product selectivity, and by-products (3a to 6a) were all formed in less than 2% yield. The absence of the Pd catalyst shut down the production of oxygenated products completely, although the over-reduction product (ethylbenzene) still formed in 26% yield. Without Shvo’s catalyst, we observed no alcohol products, and aldehyde 5a was the major product. CuCl2 was originally intended as a co-oxidant and later appeared to play a critical role in slowing down the over-reduction, as the absence of CuCl2 led to substantially increased yields of ethylbenzene. 1,4-Benzoquinone (BQ) is widely used as a hydrogen acceptor and two-electron oxidant in PdII-catalyzed reactions (30) and was found to be the best co-oxidant for this transformation; omission of this component resulted in no alcohol formation. The role of i-PrOH as the reductant was highlighted by formation of aldehyde 5a (57% yield) almost exclusively in its absence. t-BuOH proved to be responsible for enhanced reactivity and selectivity, whereas without t-BuOH, we obtained lower yields and regioselectivity of the primary alcohol. As expected, removal of water from the reaction mixture (using 4 Å molecular sieves) is detrimental: No oxygenated product was observed under anhydrous conditions.

Fig. 3

General reaction scheme for styrene hydration to produce 2-phenylethanol (2a), ethylbenzene (3a), 1-phenylethanol (4a), phenylacetaldehyde (5a), and acetophenone (6a). Pr, propyl; Bu, butyl; Ph, phenyl.

We next examined the substrate scope on a preparative scale (0.4 mmol, Table 1). The primary alcohol products were isolated, purified with silica-gel flash column chromatography, and characterized via nuclear magnetic resonance (NMR) spectroscopy and high-resolution mass spectrometry or identified by comparison of the NMR and gas chromatography–mass spectrometry (GC-MS) data with the authentic samples. In general, aryl-substituted terminal olefins provide good yields of primary alcohols with excellent anti-Markovnikov selectivity (≥20:1). A number of functional groups are tolerated under these reaction conditions, such as alkyl, naphthyl, trifluormethyl, and nitro groups, as well as various halides. Aliphatic olefins also provided hydration products, despite the challenging nature of these substrates [entries 11 and 12 (31)]. Although obtaining high regioselectivity for aliphatic substrates is more difficult, these results are promising because under previous conditions, only a Markovnikov product was observed for aliphatic substrates (26). One key merit of this method is that the major stoichiometric by-product, 1,4-hydroquinone (HBQ), can be easily recovered (see supporting online material) and converted to BQ in an excellent yield via a facile aerobic oxidation (32).

Table 1

[Pd]/[Ru]-catalyzed hydration of functionalized styrenes, 1-octene, and allylbenzene.

View this table:

A triple relay catalysis system is proposed for this formal anti-Markovnikov olefin hydration process (Fig. 4A). We postulate that in the presence of t-BuOH, the olefin (a) would first undergo Pd-catalyzed oxidation to generate a t-butyl vinyl ether (b). Due to the bulkiness of t-BuOH, the linear vinyl ether is preferred, which constitutes a key factor underlying the high anti-Markovnikov selectivity (22, 24). During such a Wacker-like process, we expect the generation of acids (HCl and hydroquinone). Subsequently, in the presence of water, ether b would be converted to aldehyde c through acid-catalyzed hydrolysis. Finally, aldehyde c would be reduced to primary alcohol d via Ru-catalyzed transfer-hydrogenation reaction.

Fig. 4

(A and B) Proposed mechanism and initial mechanistic studies.

We then conducted a number of experiments to examine this hypothesis. In the absence of water, i-PrOH, and Shvo’s catalyst, we observed a mixture of vinyl t-butyl ether geometric isomer 7 and aldehyde 5a by 1H-NMR spectroscopy and GC-MS, providing evidence for the proposed t-butyloxypalladation pathway (Fig. 4B, equation 7). When i-C3H7OD and t-C4H9OD were used, we observed mono- and di-deuterium incorporation at the β position, supporting a proton-mediated enol ether hydrolysis pathway, although deuteration via an aldehyde-enol tautomerization after the aldehyde formation cannot be ruled out (Fig. 4, equation 8) (33). We used regular H2O because it only constitutes 0.14% by volume and undergoes rapid H/D exchange with deuterated alcohols. When i-C3D7OD and t-C4H9OD were used instead, we witnessed 87% deuterium incorporation at the α position as well, strongly supporting an i-PrOH–mediated transfer hydrogenation mechanism (Fig. 4, equation 9).

Compared to the classic hydroboration/oxidation sequence, our approach is still far from perfect, with its relatively high catalyst loadings and use of stoichiometric BQ. However, we are strongly encouraged by the excellent selectivity with aryl-substituted olefins, initial promising results with aliphatic alkenes, and the facile recovery of BQ to reduce the overall expense. Despite being in its infancy, this methodology has demonstrated great potential and will stimulate ongoing research in the field of olefin hydration.

Supporting Online Material

www.sciencemag.org/cgi/content/full/333/6049/1609/DC1

Materials and Methods

SOM Text

Table S1

NMR Spectra

References (3442)

References and Notes

  1. When allylbenzene was used as the substrate, the major by-product was β-methylstyrene arising from olefin isomerization.
  2. Acknowledgments: We gratefully acknowledge financial support from the King Abdullah University of Science and Technology Center in Development, King Fahd University of Petroleum and Minerals, and the NSF. G.D. thanks the Camille and Henry Dreyfus Foundation for a postdoctoral fellowship. P.T. thanks A*STAR (Agency for Science, Technology and Research) for a postdoctoral fellowship (2009 to 2011). P.T. and Z.K.W. contributed equally to this paper. We also thank V. Lavallo and B. K. Keitz for proofreading the manuscript. A provisional patent was filed for work described in this Report.
View Abstract

Navigate This Article