Massimo Delle Piane

A new massively parallel version of CRYSTAL for large systems on high performance computing architectures

Fully ab initio treatment of complex solid systems needs computational software which is able to efficiently take advantage of the growing power of high performance computing (HPC) architectures. Recent improvements in CRYSTAL, a periodic ab initio code that uses a Gaussian basis set, allows treatment of very large unit cells for crystalline systems on HPC architectures with high parallel efficiency in terms of running time and memory requirements. The latter is a crucial point, due to the trend toward architectures relying on a very high number of cores with associated relatively low memory availability. An exhaustive performance analysis shows that density functional calculations, based on a hybrid functional, of low‐symmetry systems containing up to 100,000 atomic orbitals and 8000 atoms are feasible on the most advanced HPC architectures available to European researchers today, using thousands of processors.

Silica-Based Materials as Drug Adsorbents: First Principle Investigation on the Role of Water Microsolvation on Ibuprofen Adsorption

Silica-based materials find applications as excipients and, particularly for those of mesoporous nature, as drug delivery agents for pharmaceutical formulations. Their performance can be crucially affected by water moisture, as it can modify the behavior of these formulations, by limiting their shelf life. Here we describe the role of water microsolvation on the features of ibuprofen adsorbed on a model of amorphous silica surface by means of density functional theory (DFT) simulations. Starting from the results of the simulation of ibuprofen in interaction with a dry hydrophobic amorphous silica surface, a limited number of water molecules has been added to study the configurational landscape of the microsolvated system. Structural and energetics properties, as well as the role of dispersive forces, have been investigated. Our simulations have revealed that the silica surface exhibits a higher affinity for water than for ibuprofen, even if several structures coexist at room temperature, with an active competition of ibuprofen and water for the exposed surface silanols. Dispersive interactions play a key role in this system, as pure DFT fails to correctly describe its potential energy surface. Indeed, van der Waals forces are the leading contribution to adsorption, independently of whether the drug is hydrogen-bonded directly to the surface or via water molecules.

Propionic acid derivatives confined in mesoporous silica: monomers or dimers? The case of ibuprofen investigated by static and dynamic ab initio simulations

Confinement in mesoporous silica can greatly increase the solubility of pharmaceutical compounds. Propionic acid derivatives (a very popular class of drugs that include ibuprofen and ketoprofen) would greatly benefit from such technology, given their common apolar character. However, it is still debated whether, after confinement, these drugs are adsorbed on the pore walls as individual molecules or they keep the H-bonded dimeric structure that exists in their crystalline form. Their physical state inside the mesopores could have important consequences on the final performances of the drug delivery system. We employed accurate periodic density functional theory simulations, both static and dynamic, to investigate the issue. We simulated ibuprofen, as a model for all propionic acid derivatives, adsorbed both as a monomer and as a dimer inside a realistic model for the MCM-41 mesoporous silica. We found that adsorption is energetically favored in both cases, driven by both vdW and H-bond interactions. However, through ab initio molecular dynamics, we observed a continuous forming, breaking and reforming of these interactions. In the end, by comparing computed energetics, vibrational spectra and mobility, we were able to provide some important clues on the physical state of this class of drugs inside mesoporous silica, helping to define which drug family (monomer or dimer) is more probable after confinement.

Water at hydroxyapatite surfaces: the effect of coverage and surface termination as investigated by all-electron B3LYP-D* simulations

Hydroxyapatite [HA, Ca10(PO4)6(OH)2], the main constituent of bones and teeth enamels, is a widely studied and employed biomaterial. Its applications span from dental to orthopedic implants, including bone tissue engineering scaffolds, coating, filler and many others. Previous theoretical and experimental studies have already characterized the physical–chemical foundations of water adsorption on a number of HA surfaces, an essential step in the mechanism of biomaterial integration. Here, we extend such knowledge by simulating, at a hybrid DFT level of theory, different HA surface terminations, both stoichiometric and non-stoichiometric, as free and in interaction with water. Such a goal is achieved at an unprecedented accuracy, with a large all-electron basis set and including dispersion forces contributions. The calculated results are then compared with experimental micro-calorimetric data, showing a good agreement in the loading trend of the (010) surfaces. More generally, this theoretical approach is confirmed to be an efficient tool to analyze these biomaterials, giving the possibility to investigate the HA behavior toward more complex molecules, from amino acids to collagen, at the here-presented level of theory, to shed some light on the complex biomineralization process of human bones and teeth.

Computational Study of Acidic and Basic Functionalized Crystalline Silica Surfaces as a Model for Biomaterial Interfaces.

In silico modeling of acidic (CH2COOH) or basic (CH2NH2) functionalized silica surfaces has been carried out by means of a density functional approach based on a gradient-corrected functional to provide insight into the characterization of experimentally functionalized surfaces via a plasma method. Hydroxylated surfaces of crystalline cristobalite (sporting 4.8 OH/nm(2)) mimic an amorphous silica interface as unsubstituted material. To functionalize the silica surface we transformed the surface Si-OH groups into Si-CH2COOH and Si-CH2NH2 moieties to represent acidic/basic chemical character for the substitution. Structures, energetics, electronic, and vibrational properties were computed and compared as a function of the increasing loading of the functional groups (from 1 to 4 per surface unit cell). Classical molecular dynamics simulations of selected cases have been performed through a Reax-FF reactive force field to assess the mobility of the surface added chains. Both DFT and force field calculations identify the CH2NH2 moderate surface loading (1 group per unit cell) as the most stable functionalization, at variance with the case of the CH2COOH group, where higher loadings are preferred (2 groups per unit cell). The vibrational fingerprints of the surface functionalities, which are the ν(C═O) stretching and δ(NH2) bending modes for acidic/basic cases, have been characterized as a function of substitution percentage in order to guide the assignment of the experimental data. The final results highlighted the different behavior of the two types of functionalization. On the one hand, the frequency associated with the ν(C═O) mode shifts to lower wavenumbers as a function of the H-bond strength between the surface functionalities (both COOH and SiOH groups), and on the other hand, the δ(NH2) frequency shift seems to be caused by a subtle balance between the H-bond donor and acceptor abilities of the NH2 moiety. Both sets of data are in general agreement with experimental measurements on the corresponding silica-functionalized materials and provide finer details for a deeper interpretation of experimental spectra.

Ab Initio Modeling of Hydrogen Bond Interaction at Silica Surfaces With Focus on Silica/Drugs Systems

Abstract Among materials usually employed for drug delivery, silica plays a key role, particularly in its mesoporous form. Although much research has been performed on the topic of silica drug delivery, the understanding of the interactions occurring between the material surface and drug molecules is still scarce, despite this knowledge is essential for determining the final performance of a drug-delivery system (DDS). Molecular modeling can give a precious insight on this issue, acting as a virtual microscope to study the processes occurring inside the drug carrier. Ab initio simulations, in particular, can accurately predict geometries and enthalpies of adsorption, infrared and nuclear magnetic resonance spectra, and other experimental observables that can help pharmaceutical researchers to better predict the features of novel DDSs.

Structural and computational assessment of the influence of wet-chemical post-processing of the Al-substituted cubic Li7La3Zr2O12

Li7La3Zr2O12 (LLZO) and related compounds are considered as promising candidates for future all-solid-state Li-ion battery applications. Still, the processing of those materials into thin membranes with the right stoichiometry and crystal structure is difficult and laborious. The sensitivity of the Li-ion conductive garnets against moisture and the associated Li+/H+ cation exchange makes their processing even more difficult. Formulation of suitable polymer/ceramic hybrid solid state electrolytes could be a prosperous way to reach the future large scale production of solid state Li-ion batteries. In fact, solvent mediated and/or slurry based wet-processing of the LLZO, e.g., tape-casting, could result in irreversible Li-ion loss of the pristine material due to Li+/H+ cation exchange. The concomitant structural changes and loss in functionality in terms of Li-ion conductivity are the results of the above process. Therefore, in the present work a systematic study on the chemical stability and structural retention of Al-substituted LLZO in different solvents is reported. It was found that Li+/H+ exchange in LLZO occurs upon solvent immersion, and its magnitude is dependent on the availability of -OH functional groups of the solvent molecules. As a result, a larger degree of Li+/H+ exchange causes higher increase of the lattice parameter of the LLZO, determined by synchrotron diffraction analyses. The expansion of the cubic unit cell was ascertained, when Li+ was replaced by H+ in the host lattice, by ab initio computational studies. The application of the most common solvent as dispersion medium, i.e., high purity water, causes the most significant Li+/H+ exchange and, therefore, structural change, while acetonitrile was proven to be the best suitable solvent for wet postprocessing of LLZO. Finally, computational calculations suggested that the Li+/H+ exchange could result in diminished ionic, i.e., mixed Li+-H+, conductivity due to the insertion of protons with lower mobility than that of Li-ions.