Choosing the right molecular machine learning potential

Quantum-chemistry simulations based on potential energy surfaces of molecules provide invaluable insight into the physicochemical processes at the atomistic level and yield such important observables as reaction rates and spectra. Machine learning potentials promise to significantly reduce the computational cost and hence enable otherwise unfeasible simulations. However, the surging number of such potentials begs the question of which one to choose or whether we still need to develop yet another one. Here, we address this question by evaluating the performance of popular machine learning potentials in terms of accuracy and computational cost. In addition, we deliver structured information for non-specialists in machine learning to guide them through the maze of acronyms, recognize each potential''s main features, and judge what they could expect from each one.

This article provides a lifeline for those lost in the sea of the molecular machine learning potentials by providing a balanced overview and evaluation of popular potentials.     

Fermi Arcs and DC Transport in Nanowires of Dirac and Weyl Semimetals

The transport properties and electron states in cylinder nanowires of Dirac and Weyl semimetals are studied paying special attention to the structure and properties of the surface Fermi arcs. The latter make the electric charge and current density distributions in nanowires strongly nonuniform as the majority of the charge density is accumulated at the surface. It is found that a Weyl semimetal wire also supports a magnetization current localized mainly at the surface because of the Fermi arcs contribution. By using the Kubo linear response approach, the direct current (DC) conductivity is calculated and it is found that its spatial profile is nontrivial. By explicitly separating the contributions of the surface and bulk states, it is shown that when the electric chemical potential and/or the radius of the wire is small, the electron transport is determined primarily by the Fermi arcs and the electrical conductivity is much higher at the surface than in the bulk. Due to the rise of the surface-bulk transition rate, the relative contribution of the surface states to the total conductivity gradually diminishes as the chemical potential increases. In addition, the DC conductivity at the surface demonstrates noticeable peaks when the Fermi level crosses energies of the surface states.     

Accessing the First nido-Carborane-Substituted Diphosphetane: A Ligand and Synthon for nido-Carboranylphosphanes

Deboronation of a carborane-substituted diphosphetane 2 in toluene yielded the first nido-carboranyldiphosphetane 1 . The P−P bond in 1 can be broken via dismutation reactions with diaryl dichalcogenides yielding nido-carboranyl bis-phosphanes that were not accessible via established synthetic protocols. Additionally, transition metal complexes of 1 could be isolated including one coordination polymer. Notably, when the deboronation of 2 is carried out in ethanol, unprecedented nido-carborane-substituted secondary bis-phosphane monoxides ( 3 , 4 ) are obtained. These compounds are interesting starting materials for further reactivity studies due to their P−H bonds. Experimental findings are supported by DFT calculations including the calculation of reaction mechanisms and NMR spectroscopic parameters.     

On the directed gas phase synthesis of the imidoborane molecule (HNBH)--an isoelectronic molecule of acetylene (HCCH)

The elementary reaction of ground state boron atoms, (B((2)P(j))), with ammonia (NH(3)(X(1)A(1))) was conducted under single collision conditions at a collision energy of 20.5 ± 0.4 kJ mol(-1) in a crossed molecular beams machine. Combined with electronic structure calculations, our experimental results suggested that the reaction was initiated by a barrier-less addition of the boron atom to the nonbonding electron pair of the nitrogen atom forming a weakly bound BNH(3) collision complex. This intermediate underwent a hydrogen shift to a doublet HBNH(2) radical that decomposed via atomic hydrogen loss to at least the imidoborane (HBNH(X(1)Σ(+)) molecule, an isoelectronic species of acetylene (HCCH(X(1)Σ(g)(+))). Our studies are also discussed in light of the isoelectronic C(2)H(3) potential energy surface accessed via the isoelectronic carbon-methyl system.     

On the Hamming distance in combinatorial optimization problems on hypergraph matchings

In this note we consider the properties of the Hamming distance in combinatorial optimization problems on hypergraph matchings, also known as multidimensional assignment problems. It is shown that the Hamming distance between feasible solutions of hypergraph matching problems can be computed as an optimal value of linear assignment problem. For random hypergraph matching problems, an upper bound on the expected Hamming distance to the optimal solution is derived, and an exact expression is obtained in the special case of multidimensional assignment problems with 2 elements in each dimension.     

Risk optimization with p-order conic constraints: A linear programming approach

The paper considers solving of linear programming problems with p-order conic constraints that are related to a certain class of stochastic optimization models with risk objective or constraints. The proposed approach is based on construction of polyhedral approximations for p-order cones, and then invoking a Benders decomposition scheme that allows for efficient solving of the approximating problems. The conducted case study of portfolio optimization with p-order conic constraints demonstrates that the developed computational techniques compare favorably against a number of benchmark methods, including second-order conic programming methods.     

Facile access to a series of large polycondensed pyridazines and their utility for the supramolecular synthesis of coordination polymers

Domino cyclization of ketoenols and hydrazine leads to a series of polycondensed pyridazines, which reveal potential as rigid N-donor multidentate ligands for supramolecular synthesis of open coordination polymers.     

Cadmium(II) chloride,bromide and iodide complexes with 4,4′‐bipyridazine: when are diazine and halide bridges (in)compatible?

In poly[di‐μ‐chlorido‐μ‐(4,4′‐bipyridazine)‐κ²N¹:N^1′‐cadmium(II)], [CdCl₂(C₈H₆N₄)]_n, (I), and its isomorphous bromide analogue, [CdBr₂(C₈H₆N₄)]_n, (II), the halide atom lies on a mirror plane and the Cd^II ion resides at the intersection of two perpendicular mirror planes with m2m site symmetry. The pyridazine rings of the ligand lie in a mirror plane and are related to each other by a second mirror plane perpendicular to the first. The compounds adopt the characteristic structure of the [M^IIX₂(bipy)] type (bipy is bipyridine) based on crosslinking of [Cd(μ‐X)₂]_n chains [Cd—Cl = 2.5955 (9) and 2.6688 (9) Å; Cd—Br = 2.7089 (4) and 2.8041 (3) Å] by bitopic rod‐like organic ligands [Cd—N = 2.368 (3)–2.380 (3) Å]. This feature is discussed in terms of supramolecular stabilization, implying that the periodicity of the inorganic chain [Cd...Cd = 3.7802 (4) Å in (I) and 3.9432 (3) Å in (II)] is favourable for extensive parallel π–π stacking of monodentate pyridazine rings, with centroid–centroid distances of 3.7751 (4) Å in (I) and 3.9359 (4) Å in (II). This is not the case for the longer iodide bridges, which cannot stabilize such a pattern. In poly[tetra‐μ‐iodido‐μ₄‐(4,4′‐bipyridazine)‐κ⁴N¹:N²:N^1′:N^2′‐dicadmium(II)], [Cd₂I₄(C₈H₆N₄)]_n, (III), the ligands are situated across a centre of inversion; they are tetradentate [Cd—N = 2.488 (2) and 2.516 (2) Å] and link successive [Cd(μ‐I)₂]_n chains [Cd—I = 2.8816 (3)–3.0069 (4) Å] into corrugated layers.