Acceptorless, reversible dehydrogenation and hydrogenation reactions involving N-heterocycles are reported with a well-defined cobalt complex supported by an aminobis(phosphine) [PN(H)P] pincer ligand. Several N-heterocycle substrates have been evaluated under dehydrogenation and hydrogenation conditions. The cobalt-catalyzed amine dehydrogenation step, a key step in the dehydrogenation process, has been independently verified. Control studies with related cycloalkanes suggest that a direct acceptorless alkane dehydrogenation pathway is unlikely. The metal−ligand cooperativity is probed with the related [PN(Me)P] derivative of the cobalt catalyst. These results suggest a bifunctional dehydrogenation pathway and a nonbifunctional hydrogenation mechanism.

Hydrogenation of alkenes containing polarized CC double bonds has been achieved with iron-based homogeneous catalysts bearing a bis­(phosphino)­amine pincer ligand. Under standard catalytic conditions (5 mol % of (PNHPiPr)­Fe­(H)2(CO) (PNHPiPr = NH­(CH2CH2PiPr2)2), 23 °C, 1 atm of H2), styrene derivatives containing electron-withdrawing para substituents reacted much more quickly than both the parent styrene and substituted styrenes with an electron-donating group. Selective hydrogenation of CC double bonds occurs in the presence of other reducible functionalities such as −CO2Me, −CN, and N-heterocycles. For the α,β-unsaturated ketone benzalacetone, both CC and CO bonds have been reduced in the final product, but NMR analysis at the initial stage of catalysis demonstrates that the CO bond is reduced much more rapidly than the CC bond. Although Hanson and co-workers have proposed a nonbifunctional alkene hydrogenation mechanism for related nickel and cobalt catalysts, the iron system described here operates via a stepwise metal–ligand cooperative pathway of Fe–H hydride transfer, resulting in an ionic intermediate, followed by N–H proton transfer from the pincer ligand to form the hydrogenated product. Experimental and computational studies indicate that the polarization of the CC bond is imperative for hydrogenation with this iron catalyst.

Over the past few decades, molecular dynamics simulations and first-principles calculations have become two major approaches to predict the lattice thermal conductivity (κL), which are however limited by insufficient accuracy and high computational cost, respectively. To overcome such inherent disadvantages, machine learning (ML) has been successfully used to accurately predict κL in a high-throughput style. In this review, we give some introductions of recent ML works on the direct and indirect prediction of κL, where the derivations and applications of data-driven models are discussed in details. A brief summary of current works and future perspectives are given in the end.