Introducing Deep Eutetic Solvents to Polar Organometallic Chemistry: Chemoselective Addition of Organolithium and Grignard Reagents to Ketones under Air**

Despite their enormous synthetic relevance, the use of polar organolithium and Grignard reagents is greatly limited by their requirements of low temperatures in order to control their reactivity as well as the need of dry organic solvents and inert atmosphere protocols to avoid their fast decomposition. Breaking new ground on the applications of these commodity organometallics in synthesis under more environmentally friendly conditions, this work introduces Deep Eutetic Solvents (DESs) as a green alternative media to carry out chemoselective additions of ketones under air at room temperature. Comparing their reactivities in DES with those observed in pure water suggest that a kinetic activation of the alkylating reagents is taking placing, favouring nucleophilic addition over the competitive hydrolysis, which can be rationalised through formation of halide-rich magnesiate or lithiate species. Grignard and organolithium reagents are exceptionally valuable organometallic reagents in synthesis. Boasting extremely high reactivities, primarily due to the high polarity of their metal-carbon bonds, these reagents are indispensable to any laboratory where synthetic chemistry is carried out. Amongst their numerous applications, their addition reactions to ketones is one of the most versatile and fundamental methodologies to generate new C-C bonds allowing access to tertiary alcohols. However, the chemoselectivity of these processes can be seriously compromised by formation of undesired reduction and/or enolization products, resulting from competing く-hydride elimination and deprotonation reactions respectively. Modern synthetic alternatives to overcome these unwanted side reactions include the use of inorganic salt additives (such as CeCl3, FeCl2 or ZnCl2), as well as the in situ generation of magnesiate (LiMgR3) complexes by mixing Grignard reagents with alkyllithiums. Aiming to boost the nucleophilicity of these organometallic reagents, as well as activating the carbonyl substrates, most such approaches still require the restriction of low temperatures (ranging from 0 to -78 C) to allow chemoselective control of the reaction. Furthermore, the use of dry ethereal solvents and inert atmosphere protocols is mandatory in order to avoid fast degradation of these polar reagents, which can react violently with air or moisture. These experimental constraints can greatly hamper their synthetic usefulness in scale-up industrial processes. Thus development of novel synthetic methodologies to use these reagents under more greener conditions, compatible with the presence of water and air, without having a detrimental effect in performance is the monumental challenge in polar organometallic chemistry. Building new bridges between traditional polar organometallic synthesis and Green Chemistry, by pioneering the use of Grignard and organolithium reagents in Deep Eutectic Solvents (DESs), herein we report remarkable progress towards meeting this challenge which allows chemoselective alkylation of aliphatic and aromatic ketones to be conducted under air at room temperature and without the need of volatile organic solvents (VOCs). DESs have emerged in synthesis as a new family of green solvents which find widespread applications in a variety of areas, spanning electrochemistry, biocatalysis, metal extraction, material chemistry, purification of biodiesel to metal-catalyzed organic reactions. DESs are mostly obtained by mixing a quaternary ammonium salt with a hydrogen-bond donor that can form a complex with the halide anion of the ammonium salt. A popular choice to prepare DESs is the lowcost and readily available ammonium salt choline chloride (ChCl, Fig. 1), which in combination with biorenewable and environmentally benign hydrogen-bond donors [i.e. glycerol (Gly), lactic acid (LA), urea or water, Fig. 1] can form an eutectic mixture. Figure 1. Components used in the synthesis of the DESs employed as solvent in this work. To start this work addition of the Grignard reagent vinylmagnesium bromide to 2’-methoxy-acetophenone (1a) was examined at room temperature, under air, using different stoichiometries, (entries 1-3, [] C. Vidal, Dr. J. Garcia-Alvarez Laboratorio de Compuestos Organometálicos y Catálisis (Unidad Asociada al CSIC) Departamento de Química Orgánica e Inorgánica, Instituto Universitario de Química Organometálica “Enrique Moles”, Facultad de Química, Universidad de Oviedo, E-33071, Oviedo, Spain. E-mail: [email protected] Dr. Hernán-Gómez, Dr A. R. Kennedy, Prof. E. Hevia WestCHEM, Department of Pure and Applied Chemistry University of Strathclyde, Glasgow, UK, G1 1XL E-mail: [email protected] [] Acknowledgements. We are indebted to MICINN of Spain (Projects CTQ2010-14796 and RYC-2011-08451) and the ERC. J. G.-A. thanks the MICINN and the European Social Fund for the award of a “Ramón y Cajal” contract. We also thank the Royal Society (University Research Fellowship to E. H.), and the European Research Council (ERC) for the generous sponsorship of this research. Authors thank Professor R. E. Mulvey for insightful discussions. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201xxxxxx 2 Table 1) in the eutectic mixture 1ChCl/2Gly. Remarkably, the almost instantaneous formation of the relevant tertiary alcohol 2a was observed in all cases, finding a significantly improved yield of 78% when 2 equivalents of the Grignard reagents were employed (entry 3). Replacing Gly by other H-bond donors like ethylene glycol (EG, entry 4) or water (entry 5) led to lower conversions. In all cases the formation of 2a occurs chemoselectively with no side products observed in the crude reaction mixture (only unreacted starting material 1a). Remarkably, when the reaction was carried out using water as a solvent, 2a was obtained in a diminished 21% yield, indicating that the addition reaction to 2a must be significantly slower under these conditions that with the DES, which translates in a greater degree of hydrolysis of the organometallic reagent. Interestingly, using H-donors bearing carbonyl functionalities to generate the DES mixtures [urea (entry 6) or lactic acid (LA, entry 7)] has a negative effect in the overall chemoselectivity of the addition process, yielding 2a in 67 and 30% respectively, along with products generated by addition of Grignard reagent to the C=O bond of these H-donor molecules. Crucially the use of Schlenk techniques (N2 or Ar atmosphere) or low temperatures (0 oC) is not required. In this sense and since DESs have high heat capacities, low temperatures are not needed to cool the reaction (but often mandatory for ethereal solvents). Although attempts to generate the Grignard reagent in the eutectic mixture 1ChCl/2Gly failed, isolated vinylmagnesium bromide also reacts with the aromatic ketone 1a in the pure eutectic mixture (without ethereal co-solvents), although the yield is lower (60%, entry 8) than when a commercial ethereal solution of the Grignard reagent was employed. Table 1. Study of the addition reaction of vinylmagnesium bromide to 2’-methoxyacetophenone (1a) in different Deep Eutectic Solvents. entry Grignard Eutectic Mixture Yield (%) 1 1 mmol 1ChCl/2Gly 20 2 1.5 mmol 1ChCl/2Gly 43 3 2 mmol 1ChCl/2Gly 78 4 2 mmol 1ChCl/2EG 55 5 2 mmol 1ChCl/2H2O 57 6 2 mmol 1ChCl/2Urea 67 7 2 mmol 1ChCl/2LA 30 8 2 mmol 1ChCl/2Gly 60 a Reactions done under air, at room temp. using 1 g of DES. Reaction time 2-3 seconds. 1 mmol of the ketone used. A commercially available 1.0 M solution of vinylmagnesium bromide in THF was used. Determined by GC and 1H NMR. Addition of Grignard reagent to Urea or LA was also seen. e Ethereal solvent removed by exhaustive evacuation. Encouraged by these initial findings, which glimpse the potential that 1ChCl/2Gly and 1ChCl/2H2O have as green solvents for the chemoselective addition of Grignard reagents to ketones under standard bench experimental techniques, we then assessed the scope of this methodology extending our studies to a range of Grignard reagents with both aromatic and aliphatic ketones (Table 2). Similarly to the experiments compiled in Table 1, the addition reactions in both eutectic mixtures were completed in exceptionally short reaction times (2-3 seconds), with in most cases a high degree of selectivity, recovering only unreacted ketone. An exception was the reaction of EtMgCl with benzophenone (1b) (entries 7 and 8, Table 2) where the corresponding reduction product (diphenylmethanol, 3) was also formed (vide infra). Considering first the two different eutectic mixtures employed, although no direct correlation was found between the H-bond donor component of the DESs (Gly or water) and the outcome of the reaction, in general, better yields for the corresponding alcohols 2 were obtained in the Gly-containing eutectic mixture (odd entries in Table 2). Using vinylmagnesium bromide as an alkenylating reagent tertiary alcohols 2a, 2c and 2e can be effectively obtained from reactions with 2’-methoxy-acetophenone (1a) (entries 1 and 2), benzophenone (1b) (entries 5 and 6) and aliphatic ketone 2-pentanone (1c) (entries 9 and 10) respectively. Similarly ethynylmagnesium bromide reacts with 1a, affording 2f in a 77 and a 72% yield depending on the eutectic mixture employed (entries 11 and 12). Confronting the same ketone with EtMgCl, where く-hydride transfer (reduction) is more plausible, led exclusively to addition product 2b in 64 and 73% yields (entries 3 and 4 using 1ChCl/2Gly and 1ChCl/2H2O), although when benzophenone was employed, secondary alcohol 3 was the major product (entries 7 and 8, Table 2), while ethyl adduct 2d was obtained in modest yields, similar to those previously reported for the same reaction using dry THF as solvent at 0 C. Table 2. Addition of various Grignard reagents to ketones 1a-1c in ChCl-based eutectic mixtures. a Reactions performed under air, at room temperature using 1 g of the DES. Reaction time: 2-3 seconds. 1 mmol of ketone used. Commercially available 1.0 M solutions of the relevant Grignard reagents in THF (2 mmol) were employed. Determined by GC and 1H NMR. Formation of diphenylmethanol (3), resulting from the reduction reaction was also observed, with yields displayed in brackets. Theoretical and experimental studies monitoring the addition reactions of carbonyl compounds by Grignard reagents, using neat water as a solvent, have shown that while for allyl Grignard reagents additions take place at a comparable rate to those of the competing hydrolysis processes, alkyl analogues, such as BuMgCl, are much more kinetically retarded (addition reaction is up to 10 times slower), and therefore protonation occurs preferentially, yielding only trace amounts of addition products. This behaviour contrasts sharply with the reactions mentioned above with EtMgCl and 1a where in both eutectic mixtures (1ChCl/2Gly and 1ChCl/2H2O), the tertiary alcohol 2b is the major product (entries 3 and 4 table 2), hinting at some type of kinetic activation of the Grignard reagent may be occurring using the DESs. Furthermore in this case the conversions observed for 2b are greater than that found when 1a is reacted at -78 C in dry THF under strictly inert atmosphere techniques (45%). Germane to this work, Song has recently shown that catalytic amounts of ammonium salt NBu4Cl in THF solutions of Grignard reagents can greatly enhance the chemoselectivity of addition reactions, minimizing formation of enolization and reduction products. The authors proposed that substoichiometric amounts of the salt can shift the position of the Schlenk equilibrium of Grignard reagents to form dinuclear R2Mg.MgX2 species which would favour addition. Since a main component in the eutectic mixtures employed in this work is an ammonium salt, a related activation effect can be operative. To explore this possibility more, we reacted choline chloride (ChCl) with various amounts of the Grignard reagent entry R1 R2 R3 DES Yield (%) of 2 1 o-(MeO)C6H4 Me vinyl 1ChCl/2Gly 2a 78 2 o-(MeO)C6H4 Me vinyl 1ChCl/2H2O 2a 57 3 o-(MeO)C6H4 Me Et 1ChCl/2Gly 2b 64 4 o-(MeO)C6H4 Me Et 1ChCl/2H2O 2b 73 5 Ph Ph vinyl 1ChCl/2Gly 2c 69 6 Ph Ph vinyl 1ChCl/2H2O 2c 44 7c Ph Ph Et 1ChCl/2Gly 2d 24 [46] 8 Ph Ph Et 1ChCl/2H2O 2d 19[46] 9 CH3(CH2)2 Me vinyl 1ChCl/2Gly 2e 79 10 CH3(CH2)2 Me vinyl 1ChCl/2H2O 2e 87 11 o-(MeO)C6H4 Me ethynyl 1ChCl/2Gly 2f 77 12 o-(MeO)C6H4 Me ethynyl 1ChCl/2H2O 2f 72