Reaction condition recommendation is an essential element for the realization of computer-assisted synthetic planning. Accurate suggestions of reaction conditions are required for experimental validation and can have a significant effect on the success or failure of an attempted transformation. However, de novo condition recommendation remains a challenging and under-explored problem and relies heavily on chemists’ knowledge and experience. In this work, we develop a neural-network model to predict the chemical context (catalyst(s), solvent(s), reagent(s)), as well as the temperature most suitable for any particular organic reaction. Trained on ∼10 million examples from Reaxys, the model is able to propose conditions where a close match to the recorded catalyst, solvent, and reagent is found within the top-10 predictions 69.6% of the time, with top-10 accuracies for individual species reaching 80–90%. Temperature is accurately predicted within ±20 °C from the recorded temperature in 60–70% of test cases, with higher accuracy for cases with correct chemical context predictions. The utility of the model is illustrated through several examples spanning a range of common reaction classes. We also demonstrate that the model implicitly learns a continuous numerical embedding of solvent and reagent species that captures their functional similarity.

Carbohydrate epimerization is an essential technology for the widespread production of rare sugars. In contrast to other enzymes, most epimerases are only active on sugars substituted with phosphate or nucleotide groups, thus drastically restricting their use. Here we show that Sn-Beta zeolite in the presence of sodium tetraborate catalyses the selective epimerization of aldoses in aqueous media. Specifically, a 5 wt% aldose (for example, glucose, xylose or arabinose) solution with a 4:1 aldose:sodium tetraborate molar ratio reacted with catalytic amounts of Sn-Beta yields near-equilibrium epimerization product distributions. The reaction proceeds by way of a 1,2 carbon shift wherein the bond between C-2 and C-3 is cleaved and a new bond between C-1 and C-3 is formed, with C-1 moving to the C-2 position with an inverted configuration. This work provides a general method of performing carbohydrate epimerizations that surmounts the main disadvantages of current enzymatic and inorganic processes.

No abstract

No abstract

Lewis acidic zeolites are used to synthesize unsaturated dicarboxylic acid esters via aldol condensation of keto esters. Hafnium-containing BEA (Hf-BEA) zeolites catalyze the condensation of ethyl pyruvate into diethyl 2-methyl-4-oxopent-2-enedioate and diethyl 2-methylene-4-oxopentanedioate (an itaconic acid ester analogue) with a selectivity of ca. 80% at ca. 60% conversion in a packed-bed reactor. The catalyst is stable for 132 h on stream, reaching a turnover number of 5110 molEP molHf –1. Analysis of the dynamic behavior of Hf-BEA under flow conditions and studies with Na-exchanged zeolites suggest that Hf­(IV) open sites possess dual functionality for Lewis and Brønsted acid catalysis.

No abstract

Levulinate derivatives are an attractive platform for the production of renewable chemicals. Here we report on the oxidation of methyl levulinate into dimethyl succinate with peroxides under mild conditions using Brønsted and Lewis acid catalysts. Selectivities to succinate and acetate derivatives of approximately 60 and 40 %, respectively, were obtained with strong Brønsted acids in methanol. Although the molecular structure (i.e., carbon-chain length and branching around the C=O group) and the oxidant type affect the product distribution, solvent choice has the strongest impact on changing the location of oxygen insertion into the carbon backbone. Specifically, switching the solvent from methanol to heptane resulted in a decrease in the succinate/acetate ratio from 1.6 to 0.3. In contrast to Brønsted acids, we demonstrate that the nature of the metal cation is responsible for changing the reaction selectivity of water-tolerant Lewis acidic triflate salts.The conversion of renewable resources into commodity and specialty chemicals through chemical or biological routes has attracted considerable attention. [1] Thermocatalytic routes offer an attractive alternative to biological routes, with less stringent requirements for temperature and pH control, as well as potentially less energy-intensive product separation and purification processes. [1a, 2] Succinic acid (SA) has been identified as one of the top 12 building blocks from biomass by the US Department of Energy. [3] The market for SA is expected to exceed $ 1.1 billion in revenue by 2020. [4] Succinates, formed from the esterification of SA with monoalcohols, are important plasticizers, lubricants, and chemical intermediates. [5] A thermocatalytic route to produce SA and other succinate derivatives from renewable resources is highly desirable.Levulinic acid (LA) is a key platform molecule that can be readily produced from lignocellulosic carbohydrates. [6] Upon scale up, the estimated LA price is expected to decrease to less than $ 1 kg À1 from the current commercial price of approximately $ 3.5 kg À1 . [7] The carbonyl group in LA and its esters can be oxidized to produce SA and its esters, respectively; however, linear aliphatic ketones are generally difficult to oxidize selectively. Consequently, only a few studies, mostly using vanadium-, ruthenium-, and manganese-based catalysts, have been reported on the oxidation of LA derivatives. [8] These catalytic systems suffer from several challenges, including necessitating high reaction temperatures, using toxic reagents, or releasing a stoichiometric amount of CO 2 . Recently, Podolean et al. demonstrated that Ru-based magnetic nanoparticles are efficient catalysts for the oxidation of LA into SA under 10 bar of O 2 at 150 8C. [1b] The authors hypothesized that strong Brønsted acid sites are responsible for catalyzing the oxidation of the LA carbon backbone via a Baeyer-Villiger (BV) mechanism. Unfortunately, no systematic studies on the role of Brønsted acids in the BV oxidation of LA into SA were report...

No abstract