Thomas

Eine deutsche Arbeitsgruppe hat eine der stabilsten Bindungen der Chemie gebrochen - und das praktisch gewaltfrei. Das Resultat ist eine Kette aus vier Stickstoffatomen.

A perturbation theory‐based algorithm for the iterative orbital update in complete active space self‐consistent‐field (CASSCF) calculations is presented. It is based on single‐excitation amplitudes obtained from a perturbative configuration interaction calculation using the Dyall Hamiltonian as a zeroth‐order Hamiltonian. These amplitudes are used for the construction of the exponential of an anti‐Hermitian matrix which can be used for the orbital update. Combined with DIIS, this approach leads to rapid convergence of the CASSCF iteration procedure.

A perturbation theory‐based algorithm for the iterative orbital update in complete active space self‐consistent‐field (CASSCF) calculations is presented. Following Angeli et al. (J. Chem. Phys. 2002, 117, 10525), the first‐order contribution of singly excited configurations to the CASSCF wave function is evaluated using the Dyall Hamiltonian for the determination of a zeroth‐order Hamiltonian. These authors employ an iterative diagonalization of the first‐order density matrix including the first‐order correction arising from single excitations, whereas the present approach uses the single‐excitation amplitudes directly for the construction of the exponential of an anti‐Hermitian matrix resulting in a unitary matrix which can be used for the orbital update. At convergence, the single‐excitation amplitudes vanish as a consequence of the generalized Brillouin's theorem. It is shown that this approach in combination with direct inversion of the iterative subspace (DIIS) leads to very rapid convergence of the CASSCF iteration procedure. © 2019 Wiley Periodicals, Inc.

A perturbation theory‐based algorithm for the iterative orbital update in complete active space self‐consistent‐field (CASSCF) calculations is presented. It is based on single‐excitation amplitudes obtained from a perturbative configuration interaction calculation using the Dyall Hamiltonian as a zeroth‐order Hamiltonian. These amplitudes are used for the construction of the exponential of an anti‐Hermitian matrix which can be used for the orbital update. Combined with DIIS, this approach leads to rapid convergence of the CASSCF iteration procedure.

A perturbation theory‐based algorithm for the iterative orbital update in complete active space self‐consistent‐field (CASSCF) calculations is presented. Following Angeli et al. (J. Chem. Phys. 2002, 117, 10525), the first‐order contribution of singly excited configurations to the CASSCF wave function is evaluated using the Dyall Hamiltonian for the determination of a zeroth‐order Hamiltonian. These authors employ an iterative diagonalization of the first‐order density matrix including the first‐order correction arising from single excitations, whereas the present approach uses the single‐excitation amplitudes directly for the construction of the exponential of an anti‐Hermitian matrix resulting in a unitary matrix which can be used for the orbital update. At convergence, the single‐excitation amplitudes vanish as a consequence of the generalized Brillouin's theorem. It is shown that this approach in combination with direct inversion of the iterative subspace (DIIS) leads to very rapid convergence of the CASSCF iteration procedure. © 2019 Wiley Periodicals, Inc.

A combination of statistical learning theory and optimization provides both theoretical results and implementable algorithms to solve real problems in machine learning. These results are successfully and flexibly applied to predict atomization energies in quantum chemistry by digging deeply the inner structure of loss functions and statistical errors. This approach is flexible even for traditional kernel ridge regression.

Abstract

Machine learning promises to accelerate materials discovery by allowing computational efficient property predictions from a small number of reference calculations. As a result, the literature has spent a considerable effort in designing representations that capture basic physical properties. Our work focuses on the less‐studied learning formulations in this context in order to exploit inner structures in the prediction errors. In particular, we propose to directly optimize basic loss functions of the prediction error metrics typically used in the literature, such as the mean absolute error or the worst case error. In some instances, a proper choice of the loss function can directly reduce reasonably the prediction performance in the desired metric, albeit at the cost of additional computations during training. To support this claim, we describe the statistical learning theoretic foundations, and provide supporting numerical evidence with the prediction of atomization energies for a database of small organic molecules.

So genannte Katzen-Zustände könnten künftig dabei helfen, Quantennetzwerke aufzuspannen. Nun ist Physikern ein wichtiges Experiment auf dem Weg dorthin geglückt.

Forscher präsentieren ein neues Modell des Maschinenlernens, das die Schichten tiefer neuronaler Netze durch Differenzialgleichungen ersetzt.

Volkswagen testet die ersten kommerziell vertriebenen Quantencomputer. Eine Anwendung sind Optimierungsprobleme, an denen klassische Computer lange rechnen.

Heute vor genau 300 Jahren baumelte Blackbeards Kopf am Bug eines Navyschiffs. Wer war dieser Mann, der es in nur zwei Jahren zum gefürchtetsten Piraten der Karibik gebracht hatte?

The Journal of Chemical Physics, Volume 149, Issue 19, November 2018.
Heterogeneous reactions at the surfaces of mineral dusts represent a key process in the formation of atmospheric aerosols. To quantify the rate of aerosol formation in climate modeling as well as combat hazardous aerosols, a deep understanding of the mechanisms of these reactions is essential. In the present work, density functional theory calculations, including a Hubbard-like +U correction, were employed to elucidate the reaction between SO2 and the hematite(0001) surface. Three reaction conditions are considered: dry, wet, and aerobic. In the absence of water and oxygen, adsorption energies of SO2 on the clean Fe–O3–Fe-termination were found to be about −0.8 to −1.0 eV and resulted in the formation of an adsorbed SO3-like species. The addition of water leads to surface hydroxylation and has little effect on promoting the SO2 adsorption. Under such circumstances, an HSO3-like species was formed with a smaller adsorption energy of about −0.5 eV. By contrast, the presence of molecular oxygen enhances the SO2 adsorption significantly as the two species combine to form sulfate SO42−, with adsorption energies of −1.31 to −1.64 eV. The calculated vibrational frequencies of the adsorbate species provide insight into the surface bonding and a useful spectral fingerprinting for experimental measurements. These results elucidate the atomistic mechanism of the reaction between SO2 and hematite and highlight the important role of atmospheric O2 in the formation of sulfates.

Interest in organic radicals is fuelled by their potential role in designing organic materials and complex electronic structure. A multireference perturbative method with improved virtual orbitals has been used as an affordable, efficient and robust device for studying radicals. The method is tailored to tackle open‐ and closed‐shell states with the same accuracy. An intricate balance between different correlation effects is tangled in the accurate description of multiple states and their gaps.

Abstract

IVO‐SSMRPT is an affordable and accurate type of state‐specific multireference perturbation (SSMRPT) theory that adds dynamic correlation energy to improved virtual orbital (IVO) complete active space configuration interaction (CASCI) wave functions using a single‐root parametrization of multi‐root Hilbert‐space ansatz. We applied it to many chemically important di‐ and tri‐radicals to analyze the geometries and electronic properties of spectroscopic interest for both closed‐ and open‐shell singlet‐ and nonsinglet ground as well as excited states. We observed that IVO‐SSMRPT identifies optimized geometries, splitting between multiplets and frequencies for several radicals that are similar to those displayed by current generation state‐of‐the‐art methods but with admiringly decreased computational effort. This study illustrates the importance of having an accurate treatment of both nondynamical and dynamical correlation effects when examining multiradical species. Chemically and spectroscopically relevant answers can be obtained using our computationally tractable method. Our method will be a serviceable avenue for portraying open‐shell interactions in other radicals.

Teilchenphysiker setzen Verfahren des maschinellen Lernens schon seit Jahrzehnten ein. Doch nun erwarten Experten durch lernende Software eine Revolution bei der Datenanalyse.

NO dimerization on Ag and Cu clusters differs very much from that in gas phase. The nondynamical electron correlation effect is very small and the reaction easily occurs with nearly no barrier and significantly large exothermicity on these clusters.

Experimentally observed NO dimerization on Cu and Ag surfaces is surprising because binding energy of NO dimer is very small in gas phase. MRMP2, MP2 to MP4, CCSD(T), and DFT studies of NO dimerization on Ag₂ and Cu₂ clusters disclosed that the CCSD(T) method could be applied to this reaction on Ag₂ and Cu₂ unlike NO dimerization in gas phase which exhibits significantly large nondynamical electron correlation effect. Charge‐transfer (CT) from Ag₂ and Cu₂ to NO moieties plays important role in NN bond formation between two NO molecules. This CT considerably decreases nondynamical correlation effect. Also, the DFT method could be applied to this NO dimerization, if appropriate DFT functional is used; all pure functionals examined here and most of the hybrid functionals underestimated the activation barrier (E _a), while only ωB97X provided E _a similar to CCSD(T)‐calculated value. NO dimerization on similar Cu₂ and Cu₅ needs moderately larger E _a than those on Ag₂ and Ag₅, because frontier orbital participating in the CT exists at lower energy in Cu₂ and Cu₅ than in Ag₂ and Ag₅. The E _a decreases in the order Ag₂ >> Ag₃₈ > Ag₇ ∼ Ag₅ and the reaction energy (ΔE) is positive (endothermic) in Ag₂ but significantly negative in Ag₃₈, Ag₇, and Ag₅, indicating that various Ag clusters could be effective for NO dimerization except for Ag₂. The decreasing order of E _a and increasing order of exothermicity are attributed to increasing order of the frontier orbital energy of Ag₂ < Ag₃₈ < Ag₇ ∼ Ag₅. © 2018 Wiley Periodicals, Inc.