Sayan
Kar
a,
Jie
Luo
a,
Michael
Rauch
a,
Yael
Diskin-Posner
b,
Yehoshoa
Ben-David
a and
David
Milstein
*a
aDepartment of Molecular Chemistry and Materials Science, The Weizmann Institute of Science, Rehovot 76100, Israel. E-mail: david.milstein@weizmann.ac.il
bDepartment of Chemical Research Support, The Weizmann Institute of Science, Rehovot 76100, Israel
First published on 4th February 2022
We report the dehydrogenative synthesis of esters from enol ethers using water as the formal oxidant, catalyzed by a newly developed ruthenium acridine-based PNP(Ph)-type complex. Mechanistic experiments and density functional theory (DFT) studies suggest that an inner-sphere stepwise coupled reaction pathway is operational instead of a more intuitive outer-sphere tandem hydration–dehydrogenation pathway.
Using ruthenium and iridium-based pincer complexes, our group and others have reported catalytic primary alcohol oxidation to carboxylate salts under alkaline conditions using water while liberating H2 (Fig. 1a).14–17 Similarly, aqueous reforming of simple alcohols such as methanol,18–23 ethanol,24 glycol,25 glycerol26 is reported by Beller, us, and others. Goldberg and co-workers have reported molecular complexes that can catalyse aldehyde oxidation to carboxylic acids under neutral conditions using water as an oxidant (Fig. 1b).27,28 At the same time, using ruthenium acridine based PNP complexes, our group has demonstrated the oxidation of cyclic amines to lactams (Fig. 1c),29,30 alkenes to ketones (Fig. 1d),31 and amines to carboxylate salts (Fig. 1e).32 Notably, in these reports, the molecular complexes act as dehydrogenation catalysts while the water addition to substrate/intermediate happens spontaneously or is facilitated by additional Lewis acid or base catalysts. No further involvement of the molecular complex during the reaction is observed.
Fig. 1 Previous water-based oxidation reactions catalyzed by molecular complexes (a–e) and this study (f). |
Here we report the dehydrogenative oxidation of vinyl and cyclic enol ethers to the corresponding esters using water as the oxidant. The reaction is selectively catalysed by a newly developed ruthenium acridine PNP(Ph) based complex, which catalyses both hydration and dehydrogenation steps, allowing the reaction to proceed under neutral additive-free conditions. Interestingly, DFT calculations revealed a possible stepwise inner-sphere coupled dehydrogenation–hydration mechanism instead of the intuitive outer-sphere tandem hydration–dehydrogenation route.33 Esters are widespread commodity chemicals with myriad applications, including odorants, lubricants, polymer precursors, and this method provides an alternative to access esters to the traditional Fischer–Speier esterification,34 and other advanced oxidative35–37 and dehydrogenative38–42 methods.
Our investigation started by exploring the reactivity of ethyl vinyl ether with water under neutral conditions in the presence and absence of complex 1, which was previously reported by us for other water-based oxidations (Table 1).29–31 When a mixture comprising 0.30 mmol of ethyl vinyl ether and 0.1 mL of water (5.6 mmol) in 2.0 mL of 1,4-dioxane was heated at 120 °C for 16 h in a 25 mL Schlenk flask, with no added catalyst, no significant reaction was observed by GC and NMR spectroscopy (entry 1). This is in line with previous reports, which showed that hydrolysis of enol ethers to aldehydes and alcohols proceeds only slowly under neutral or basic conditions.43–46 Interestingly, when a similar reaction was carried out in the presence of 0.75 mol% complex of 1, ethyl acetate was observed as a product in 17% yield (entry 2), along with diethyl ether as a minor by-product (4%). The presence of H2 gas in the reactor headspace was detected by GC analysis (Fig. S16†). The formation of ethyl acetate from ethyl vinyl ether presumably proceeds via a hemiacetal-like intermediate upon initial water addition, followed by its dehydrogenation (Fig. S14†). The above results indicate that complex 1 can catalyse both enol ether hydration and dehydrogenation under the reaction conditions, ultimately leading to ester formation. At the same time, the liberated H2 hydrogenates, in the presence of 1, some of the starting material to form the ether side product. Increasing the catalyst loading to 1.5 mol%, water amount to 0.2 mL (11.1 mmol), reaction temperature to 125 °C, and reaction time to 24 h increased the ester yield to 31%, while also increasing the amount of ether by-product (19%; entry 3). Using a larger amount of water (1 mL, 56 mmol) was detrimental to the process, leading to significant hydrolysis of both the enol ether and ester (entry 4). Replacing the substrate with a higher boiling one, namely, butyl vinyl ether, increased conversion to 82%. The desired ester was obtained in 57% yield, along with 24% of the hydrogenated by-product ethyl butyl ether (entry 5). Among different solvents, tetrahydrofuran (THF) was almost equally effective as co-solvent as 1,4-dixoane (entry 6). On the other hand, when toluene was used as co-solvent, the ester yield significantly decreased, likely due to the limited solubility of water in toluene forming a biphasic reaction system (entry 7). Similarly, when only water was used as solvent with no co-solvent, only traces of ester formation were observed due to the sparing solubility of complex 1 in water (entry 8).
Entry | R | Cat. (mol%) | H2O (mL) | T (°C)/t (h) | Conv.a (%) | P1a (%) | P2a (%) |
---|---|---|---|---|---|---|---|
Reaction conditions: Vinyl ether (0.30 mmol), 1,4-dioxane (2.0 mL), water, [Ru], oil bath temperature, and reaction time as specified.a Yields were determined by GC analysis with mesitylene as an internal standard.b The additional products were ethanol (40%), acetaldehyde (20%), and acetic acid (20%).c THF used as co-solvent.d Toluene used as co-solvent.e No 1,4-dioxane was used. | |||||||
1 | Me | — | 0.1 | 120/16 | 1 | 0 | 0 |
2 | Me | 1 (0.75) | 0.1 | 120/16 | 21 | 17 | 4 |
3 | Me | 1 (1.5) | 0.2 | 125/24 | 53 | 31 | 19 |
4 | Me | 1 (1.5) | 1.0 | 125/24 | 100b | 30 | 20 |
5 | n Pr | 1 (1.5) | 0.2 | 125/24 | 82 | 57 | 24 |
6c | n Pr | 1 (1.5) | 0.2 | 125/24 | 75 | 51 | 21 |
7d | n Pr | 1 (1.5) | 0.2 | 125/24 | 35 | 17 | 7 |
8e | n Pr | 1 (1.5) | 2.2 | 125/24 | 86 | 1 | 1 |
9 | n Pr | 2 (1.5) | 0.2 | 125/24 | 7 | 0 | 0 |
10 | n Pr | 3 (1.5) | 0.2 | 125/24 | 2 | 0 | 0 |
11 | n Pr | 4 (1.5) | 0.2 | 125/24 | 3 | 0 | 0 |
12 | n Pr | 5 (1.5) | 0.2 | 125/24 | 12 | 9 | 0 |
13 | n Pr | 6 (1.5) | 0.2 | 125/24 | 67 | 58 | 3 |
14 | n Pr | 6 (1.5) | 0.25 | 125/36 | 91 | 81 | 3 |
Among other complexes, the acridine-based hydrido-chloride complex 2 did not promote the reaction to any significant extent (entry 9). Similarly, the dearomatized bipyridine-based complex 3 and the dearomatized aminopyridine-based (PNNH) complex 4, showed no catalytic activities (entries 10–11).14,47 The acridine-based dicarbonyl complex 5 showed only minor catalytic activity (entry 12), highlighting the importance of a vacant coordination site for the reaction to proceed. Quite interestingly, upon decreasing the electron density at the ruthenium centre, as in the newly synthesized phenyl-substituted analog of complex 1, complex 6 (synthesis and characterization details in the ESI†), similar ester yield (58%) was observed, but with much better selectivity (only 3% of the respective olefin hydrogenation product) (entry 13). The ester yield increased to 81% upon increasing water amount to 0.25 mL and the reaction time to 36 h, without compromising selectivity (entry 14).
Reaction conditions: Vinyl ether (0.3 mmol), 6 (1.5 mol%), H2O (0.25 mL), 1,4-dioxane (2 mL), 125 °C (oil bath temperature), 36 h, closed 25 mL Schlenk flask.a Ester yields were determined by GC or 1H NMR spectroscopy, using mesitylene as an internal standard; yields of the corresponding hydrogenated products (P2) are in parentheses.b Reflux in an open system under argon flow.c 48 h.d 24 h.e H218O, 0.1 mL. |
---|
Compared to the vinyl ether substrates, cyclic enol ethers were less reactive, requiring harsher reaction conditions (1 mL water, 150 °C) to yield lactones via this method (Scheme 1). Under these reaction conditions, dihydrofuran reacted selectively to produce γ-butyrolactone in 89% yield (Scheme 1a), with 90% H2 yield. Dihydropyran showed slightly lower reactivity compared to dihydrofuran, and 72% of the corresponding lactone was observed after similar reaction conditions (Scheme 1b).48 In contrast to the vinyl and cyclic enol ethers, the internal linear enol ether, 1-butenyl ethyl ether, did not react under the reaction conditions in the presence of either complex 6 or 1 as catalyst (Scheme 1c). This is likely due to its increased steric bulk and thermodynamic stability, disfavouring its coordination to the metal centre of the catalysts. A similar reactivity pattern is observed in other organometallic reactions involving alkene coordination, including hydrogenation with Wilkinson's and Vaska's catalysts.49 The addition of a Lewis acid cocatalyst such as LiBF4 resulted in the hydrolysis of the starting material to butanal and ethanol with no significant ester formation.
When a 1,4-dioxane solution containing complex 1, butyl vinyl ether (5 equiv. to 1), and water (10 equiv. to 1) was heated at 100 °C in a J Young NMR tube, formation of ester was observed by 1H NMR, although the 31P NMR revealed only the presence of complex 1a in solution. The reaction was complete after 30 min, forming butyl acetate (70%) and butyl ethyl ether (30%) as products, with the aqua complex 1b present afterward in solution following the consumption of all available vinyl ether (Scheme 2c, Fig. S34†). Thus, complex 1a seemingly acts as the resting state during active catalysis, and other catalytic species formed during reaction are high-energy intermediates, not observed by NMR. Formation of the ester product is irreversible with no observed reactivity of complex 1 with the ester (ethyl acetate), even at elevated temperatures (Scheme 2d).
Similar to butyl vinyl ether, dihydrofuran also coordinated to the vacant site of 1. On the contrary, no interaction was observed when ethyl butenyl ether was added to a solution of 1 (Scheme 2e). This lack of coordination may explain why no reaction is observed for internal enol ethers under the reaction conditions (Scheme 1). The relative binding propensities of different substrates were explored by NMR through exchange experiments at room temperature in C6D6, and among different substrates, butyl vinyl ether coordinated to the metal centre of 1 most strongly, followed by dihydrofuran and water, respectively.
Although no catalytic intermediates were observed in mechanistic investigations, based on previously observed reactivities of the ruthenium acridine system by us and others,30,50–55,63 we surmise that the reaction proceeds via the fac isomer of acridine complex 1/6, with the mer isomers being off-cycle intermediates. (For other selected examples of fac isomers of pincer complexes, see ref. 15, 56–61). Density functional theory (DFT) calculations were carried out to gain further insight regarding the reaction mechanism, with ethyl vinyl ether as the model substrate (Fig. 2). To account for the reaction conditions, the structures were optimized in a 1,4-dioxane continuum following the method reported by Truhlar and co-workers,62 and free energies were computed at 125 °C with standard state corrections. Under the conditions, the fac isomer of 6 was calculated to be close in energy to the mer isomer (5.6 kcal mol−1 uphill) (Fig. 2). However, the outer-sphere water addition to the bound vinyl ether was found quite energy demanding (ΔG 36.1 kcal mol−1), with the relevant transition state associated with simultaneous H2 liberation, akin to a previous report with imines (Fig. 2, in red).29,30 At the same time, outer-sphere water addition to vinyl ether facilitated by the mer isomer of 6 acting as Lewis acid catalyst, similar to a mechanism proposed for alkene hydration with the stronger Lewis acid In(OTf)3,31 is also unlikely, with the relevant Zwitter ionic intermediate not found by computation (ESI sec 6.2†). Interestingly, an inner-sphere stepwise mechanism from the aqua complex 6b-fac involving initial dehydrogenation,30,32 followed by hydroxide insertion to cis-bound vinyl ether, and β-hydride elimination was computed to have an overall low activation barrier (26.0 kcal mol−1 from 6a-fac) (Fig. 2, in blue). In this mechanism, initial dehydrogenation of the aqua complex is the most energy demanding step (6TS1, ΔG: 22.4 kcal mol−1 with respect to 6-mer), with subsequent hydroxide migration (6TS2) and β-H elimination (6TS3) transition states being relatively lower in energy (ΔG: 17.2 and 15.2 kcal mol−1, respectively). The likelihood of the stepwise inner-sphere mechanism is further supported by the observations that – (i) changing the reactant from water to D2O slowed down the ester formation significantly (KIE kH/kD: 2.3), suggesting the involvement of O–H(D) bond in the slowest steps (ESI, sec 3.1†) and (ii) the overall energy barrier of the pathway (26.0 kcal mol−1) is in accordance to the reaction temperature (125 °C). A similar energy profile was found with complex 1 with an overall activation barrier of 28.6 kcal mol−1 (ESI sec 6.3†). Notably, another reaction mechanism involving spontaneous water addition across vinyl ether followed by Ru catalysed dehydrogenation can be eliminated based on experimental observation (Table 1, entry 1) and computational studies (Ea: 51.1 kcal mol−1; ESI sec 6.1†).
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2079903. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/d1gc04574a |
This journal is © The Royal Society of Chemistry 2022 |