Predicting dipole orientations in spontelectric methyl formate

Eur. Phys. J. D (2021)75:89 https://doi.org/10.1140/epjd/s10053-021-00098-4 THE EUROPEAN PHYSICAL JOURNAL D Regular Article - Atomic and Molecular Collisions Predicting dipole orientations in spontelectric methyl formate Christian Kexel1,2,3,a and Andrey V. Solov’yov3 1 Department of Physics, Goethe University, Max-von-Laue-Str. 1, 60438 Frankfurt, Germany Frankfurt Institute for Advanced Studies, Ruth-Moufang-Str. 1, 60438 Frankfurt, Germany 3 MBN Research Center, Frankfurt Innovation Center Biotechnology, Altenhöferallee 3, 60438 Frankfurt, Germany 2 Received 11 May 2020 / Accepted 18 February 2021 © The Author(s) 2021 Abstract. Capturing intermolecular interactions accurately is essential for describing, e.g., morphology of molecular matter on the nanoscale. When it reveals characteristics which are not directly accessible through experiments or ab initio theories, a model here becomes eminently beneficial. In laboratory astrochemistry, the intense study of ices has led i.a. to the exploration of the spontelectric state of nanofilms. Despite its success in biophysics or biochemistry and despite its predictive power, molecular modeling has however not yet been widely deployed for solid-state astrochemistry. In this article, therefore a pertinent hitherto unaddressed problem is tackled by means of the classical molecular-dynamics method, namely the unknown distribution of relative dipole orientations in spontelectric cis-methyl formate (MF). In doing so, from ab initio data, a molecular model is derived which confirms for the first time the anomalous temperature-dependent polarization of MF. These insights thus represent a further step toward understanding spontelectric behavior. Moreover, unprecedented first-principles predictions are reported regarding the ground-state geometry of the MF trimer and tetramer. In conjunction with the study of the binding to carbonaceous substrates, these additional findings can help to exemplarily elucidate molecular ice formation in astrochemical settings. 1 Introduction and background The theoretical and computational characterization of objects on the nanoscale is a fundamental plus highly-demanding scientific endeavor. Albeit specific systems are dealt with in dedicated and seemingly unlike research disciplines, ranging, e.g., from pharmacology to astrochemistry, their appropriate description bears joint challenges. Common intertwined tasks are (i) the decision on a quantum-mechanical, semi-classical or classical ansatz, (ii) the complications introduced by many-body effects, (iii) the consistent linkage of multiple adjacent scales, defined either by time, characteristic lengths or the number of particles, covering the changeover from individual atoms to the solid state as well as the changeover between inanimate and living systems. There are shared underpinning phenomena, such as thermally driven structural transitions [1], chemical ordering [2], bonding or dissociation events [3,4] as well as cluster formation [5,6]. In spite of the low temperatures and the specific vacuum conditions in the interstellar medium, specOn leave from A. F. Ioffe Physical-Technical Institute, St. Petersburg, Russia. a e-mail: author) kexel@fias.uni-frankfurt.de (corresponding 0123456789().: V,-vol troscopic studies indicate intricate chemical activities, particularly in birth regions of planets and stars, involving diverse organics which reflect the evolution of complexity toward animate matter. Pure gas-phase chemistry is not able to account for the diversity, enhanced abundances and the existence of larger organics, such as MF (see Fig. 1). Hence, explanations additionally imply surface chemistry with molecules persisting in the solid state. Emissions of protostars and stars that move through molecular clouds generate broad infrared absorption bands which originate from condensed phases. Here, the infrared profiles of these ices carry information on temperature, composition or morphology of icy mantles. In the interstellar settings, the underlying processes are similar in nature to the already-mentioned phenomena above concerning nanosystems. They include, e.g., (i) deposition of molecules on dust particles within dense clouds, (ii) creation of ice films on the particles’ surface, (iii) catalyzation of chemical reactions due to the presence of ice, (iv) shocks by protons and swift ions leading to sputtering and to morphological changes, (v) energetic processing by ultraviolet, X-ray and cosmic radiation, (vi) polymerization through thermal processing by adjacent protostars [7–9]. With regard to the theoretical description, in the realm of organic computational chemistry, densityfunctional theory (DFT) conventionally yields a bal- 123 89 Page 2 of 12 anced trade-off between accuracy and numerical efficiency, e.g., for determination of molecular ground-state geometries. DFT is deployable for systems of, say, up to hundreds of atoms. Understandably, this limit is lowered if dynamical descriptions are needed. Moreover, weak intermolecular dispersion forces call for the inclusion of empirical long-range corrections [10] and possibly for more demanding ab initio approaches that treat electronic correlations adequately, such as manybody perturbation theory [11]. Furthermore, numerous problems on the nanoscale involving organics require the simulation of larger systems. Hence, a great variety of classical interaction potentials has been derived [12] which do not consider electronic structure explicitly. The analytic formulation of these force-fields is deduced from quantum-mechanical principles and provides a mapping from geometry of the systems to its potential energy. The accuracy of a molecular model relies largely on the deployed parameters and this model accuracy can be quantified with respect to distinct properties, such as transition temperatures [2,13] or binding energies. Regarding the development of force-fields, some requirements have to be fulfilled. In order to be considered accurate, the interaction potential should (i) exhibit reasonable complexity while being numerically practicable. It should also (ii) be transferable, i.e., accurate for geometries that were not utilized in its derivation [14]. It has become evident that astrochemistry and the study of ices can offer possible research opportunities related to nanoscopic systems. The large-scale production of metal clusters using an ice matrix is, e.g., an instructive example [15]. The scientific investigation of the coupled gas-surface chemistry, motivated by the ongoing discovery of more sophisticated species in the interstellar medium, represents a formidable challenge for ultrahigh-vacuum laboratory experiments and also molecular modeling. However, such models are only infrequently deployed in solid-state astrochemistry with, e.g., Ref. [16] being an exception. Astrochemical theory, e.g., centers around kinetic considerations on the one hand [17] or high-precision quantum chemistry methods for (...truncated)