The Supporting Material contains the cartesian coordinates of optimized minima and transition state for the reaction studied in the paper, in .xyz format, computed using ORCA code. The folders are relative to: structures of the first benchmark, optimized at 𝜔B97M-V/def2-TZVP level of theory; structures of the second benchmark, optimized at B3LYP-D3(BJ), 𝜔B97M-V and r2SCAN-3c levels of theory, plus NEB structures at 𝜔B97M-V level; adsorption of H2O, H2S, CO and CS on forsterite, cluster A and cluster B at 𝜔B97M-V/def2-TZVP level of theory: adsorption of the 9 S-bearing species on cluster C at 𝜔B97M-V/def2-TZVP level of theory.
Formamide is abundant in the interstellar medium and was also present during the formation of the Solar system through the accretion process of interstellar dust. Under the physicochemical conditions of primordial Earth, formamide could have undergone decomposition, either via dehydration (HCN + H2O) or via decarbonylation (CO + NH3). The first reactive channel provides HCN, which is an essential molecular building block for the formation of RNA/DNA bases, crucial for the emergence of life on Earth. In this work, we studied, at the CCSD(T)/cc-pVTZ level, the two competitive routes of formamide decomposition, i.e. dehydration and decarbonylation, either in liquid formamide (by using the polarization continuum model technique) or at the interface between liquid formamide and amorphous silica. Amorphous silica was adopted as a convenient model of the crystalline silica phases ubiquitously present in the primordial (and actual) Earth's crust, and also due to its relevance in catalysis, adsorption and chromatography. Results show that: (i) silica surface sites catalyse both decomposition channels by reducing the activation barriers by about 100 kJ mol-1 with respect to the reactions in homogeneous medium, and (ii) the dehydration channel, giving rise to HCN, is strongly favoured from a kinetic standpoint over decarbonylation, the latter being, instead, slightly favoured from a thermodynamic point of view.
Abstract Water is one of the most important and abundant molecules in star-forming regions. In protoplanetary disks, where planets and comets form, H 2 O is in a gas or solid form, depending on the dust temperature, i.e., the distance from the center and its binding energy (BE). Not surprisingly, several experimental and theoretical studies of the H 2 O BE have been published. We report new ab initio calculations carried out on a large model of interstellar ice, where we identified 144 different adsorption sites. The BE associated with those sites ranges between 14.2 kJ mol −1 (1705 K) and 61.6 kJ mol −1 (7390 K). The distribution of the computed BEs as a function of BE follows a Gaussian peaked at 35.4 kJ mol −1 (4230 K) with a standard deviation of 9.7 kJ mol −1 (1160 K). The computed pre-exponential factor ( ν ) ranges between 9 × 10 12 and 6 × 10 14 s −1 . We evaluated the impact of the newly calculated BE and ν distributions on the snowline of a generic protoplanetary disk. We found that the region where water is frozen onto the ice is much smaller (a factor of 10 smaller radius) than that computed with the single BE (5600 K) and ν (2 × 10 12 s −1 ) values commonly adopted by astrochemical models. Besides, ∼10% of water remains frozen in relatively warm (∼150 K) regions, where the single BE and ν model would predict a full release of the ice in the gas phase. This last aspect may have an impact on the quantity trapped in the planetesimals eventually forming rocky planets.
ABSTRACT Despite hydrogen sulphide (H2S) has been predicted to be the major reservoir of S-bearing species on the icy mantles of interstellar grains, no solid H2S has been detected so far. A crucial parameter that governs whether or not a species remains frozen on to the grain mantles is its binding energy (BE). We present a new computational study of the H2S BE on a large amorphous water ice surface, constituted by 200 water molecules. The resulting H2S BE distribution ranges from 57 K (0.5 kJ mol−1) to 2406 K (20.0 kJ mol−1), with the average μ = 984 K (8.2 kJ mol−1). We discuss the reasons why the low bound of the newly computed BE distribution, which testifies to the very weak interaction of H2S with the ice surface, has never been found by previous theoretical or experimental works before. In addition, the low H2S BEs may also explain why frozen H2S is not detected in interstellar ices. Following previous molecular dynamics studies that show that the energy of reactions occurring on ice surfaces is quickly absorbed by the water molecules of the ice and conservatively assuming that 10 per cent of the HS + H → H2S formation energy (−369.5 kJ mol−1) is left to the newly formed H2S, its energy is more than twice the largest BE and five times the average BE and, hence, H2S will most likely leave the water surface.
Computational modeling of protein/surface systems is challenging since the conformational variations of the protein and its interactions with the surface need to be considered at once. Adoption of first-principles methods to this purpose is overwhelming and computationally extremely expensive so that, in many cases, dramatically simplified systems (e.g., small peptides or amino acids) are used at the expenses of modeling nonrealistic systems. In this work, we propose a cost-effective strategy for the modeling of peptide/surface interactions at a full quantum mechanical level, taking the adsorption of polyglycine on the TiO2 (101) anatase surface as a test case. Our approach is based on applying the periodic boundary conditions for both the surface model and the polyglycine peptide, giving rise to full periodic polyglycine/TiO2 surface systems. By proceeding this way, the considered complexes are modeled with a drastically reduced number of atoms compared with the finite-analogous systems, modeling the polypeptide structures at the same time in a realistic way. Within our modeling approach, full periodic density functional theory calculations (including implicit solvation effects) and ab initio molecular dynamics (AIMD) simulations at the PBE-D2* theory level have been carried out to investigate the adsorption and relative stability of the different polyglycine structures (i.e., extended primary, β-sheet, and α-helix) on the TiO2 surface. It has been found that, upon adsorption, secondary structures become partially denatured because the peptide C═O groups form Ti–O═C dative bonds. AIMD simulations have been fundamental to identify these phenomena because thermal and entropic effects are of paramount importance. Irrespective of the simulated environments (gas phase and implicit solvent), adsorption of the α-helix is more favorable than that of the β-sheet because in the former, more Ti–O═C bonds are formed and the adsorbed secondary structure results less distorted with respect to the isolated state. Under the implicit water solvent, additionally, adsorbed β-sheet structures weaken with respect to their isolated states as the H-bonds between the strands are longer due to solvation effects. Accordingly, the results indicate that the preferred conformation upon adsorption is the α-helix over the β-sheet.
Abstract Interstellar Grains (IGs) spread in the Interstellar Medium (ISM) host a multitude of chemical reactions that could lead to the production of interstellar Complex Organic Molecules (iCOMs), relevant in the context of prebiotic chemistry. These IGs are composed of a silicate-based core covered by several layers of amorphous water ice, known as a grain mantle. Molecules from the ISM gas-phase can be adsorbed at the grain surfaces, diffuse and react to give iCOMs and ultimately desorbed back to the gas phase. Thus, the study of the Binding Energy (BE) of these molecules at the water ice grain surface is important to understand the molecular composition of the ISM and its evolution in time. In this paper, we propose to use a recently developed semiempirical quantum approach, named GFN-xTB, and more precisely the GFN2 method, to compute the BE of several molecular species at the crystalline water ice slab model. This method is very cheap in term of computing power and time and was already showed in a previous work to be very accurate with small water clusters. To support our proposition, we decided to use, as a benchmark, the recent work published by some of us in which a crystalline model of proton-ordered water ice (P-ice) was adopted to predict the BEs of 21 molecules relevant in the ISM. The relatively good results obtained confirm GFN2 as the method of choice to model adsorption processes occurring at the icy grains in the ISM. The only notable exception was for the CO molecule, in which both structure and BE are badly predicted by GFN2, a real pity due to the relevance of CO in astrochemistry.
