arXiv subjects

Santanu Saha

Publications and source records attributed to Santanu Saha.

13 recordsLinked to original sources

arXiv2026

Ordered-to-disordered transfer learning with graph neural networks for formation-energy and HOMO-LUMO gap prediction in high-entropy perovskite oxides

High-entropy perovskite oxides (HEPOs) represent a chemically complex class of materials with promising functional properties, yet their vast compositional space and, chemical/structural disorder pose significant challenge for accurate property prediction. Graph neural networks (GNNs) enable rapid exploration of materials space but are often limited by the availability of representative training data. Here, we investigate ordered-to-disordered transfer learning using GNNs for formation-energy and HOMO-LUMO gap prediction in HEPOs by transferring knowledge learned from chemically ordered perovskites. Four representative GNN models, including CGCNN, GATGNN, ALIGNN and M3GNet are evaluated to understand the role of structural representations, spanning pairwise two-body and angular three-body interactions in transfer performance. We find strong property-dependent transfer behavior: formation-energy prediction transfers effectively to disordered HEPOs, whereas HOMO-LUMO gap prediction shows limited transferability due to its sensitivity to local chemical environments. Incorporating a small HEPO-specific training dataset substantially improves HOMO-LUMO gap prediction. Representation-level analysis using UMAP further highlights the importance of encoding three-body geometric information such as in ALIGNN for capturing complex structure-property relationships and improving transferability.

cond-mat.mtrl-sci↗

arXiv2026

Mapping the influence of symmetry breaking in structure-property relationships of ABO$_3$ perovskites

Perovskite oxides have emerged as an important class of material with promising energy applications owing to their compositional and structural flexibility, which enables stabilization of both low- and high-symmetry phases and gives rise to diverse physical properties. Under ambient conditions, most perovskites adopt low-symmetry structures characterized by octahedral tilting and B-site displacements. Despite their importance, computational studies have largely focused on the ideal cubic phase as modeling these distortions remains challenging. The difficulty stems from the absence of a quantitative framework capable of capturing composition-dependent distortions that can occur through multiple non-equivalent atomic displacement modes, often requiring computationally expensive large supercells to explore the structural landscape. Consequently, the influence of distortions on the stability and properties of low-symmetry perovskites remains insufficiently understood. In this work, we develop an efficient computational framework for the rapid construction and exploration of composition-dependent structural models across both low- and high-symmetry phases. Using $\textit{symmetry constrained templates}$ and $\textit{unconstrained supercell templates}$, we systematically investigate 15 representative compositions to uncover relationships between composition, supercell size and shape, and distortion patterns. Based on these insights, we propose a robust and computationally inexpensive protocol for rapid structural exploration and assess the influence of different distortion modes on key physical properties.

cond-mat.mtrl-sci↗

arXiv2024

HEX: High-pressure Elemental Xstals, a complete Database

This paper introduces the HEX (High-pressure Elemental Xstals) database, a complete database of the ground-state crystal structures of the first 57 elements of the periodic table, from H to La, at 0, 100, 200 and 300 GPa. HEX aims to provide a unified reference for high-pressure research, by compiling all available experimental information on elements at high pressure, and complementing it with the results of accurate evolutionary crystal structure prediction runs based on Density Functional Theory. Besides offering a much-needed reference, our work also serves as a benchmark of the accuracy of current ab-initio methods for crystal structure prediction. We find that, in 98 % of the cases in which experimental information is available, ab-initio crystal structure prediction yields structures which either coincide or are degenerate in enthalpy to within 300 K with experimental ones. The main manuscript contains synthetic tables and figures, while the Crystallographic Information File (cif) for all structures will be available on a figshare online repository when the paper will be published.

cond-mat.mtrl-sci↗

arXiv2022

Mapping Superconductivity in High-Pressure Hydrides: The $Superhydra$ Project

The discovery of high-$T_c$ conventional superconductivity in high-pressure hydrides has helped establish computational methods as a formidable tool to guide material discoveries in a field traditionally dominated by serendipitous experimental search. This paves the way to an ever-increasing use of data-driven approaches to the study and design of superconductors. In this work, we propose a new method to generate meaningful datasets of superconductors, based on element substitution into a small set of representative structural templates, generated by crystal structure prediction methods (MultiTemplate-HighThroughput approach). Our approach realizes an optimal compromise between structural variety and computational efficiency, and can be easily generalized to other elements and compositions. As a first application, we apply it to binary hydrides at high pressure, realizing a database of 880 hypothetical structures, characterized with a set of electronic, vibrational and chemical descriptors. 139 structures of our $Superhydra$ Database are superconducting according to the McMillan-Allen-Dynes approximation. Studying the distribution of $T_c$ and other properties across the database with advanced statistical and visualization techniques, we are able to obtain comprehensive material maps of the phase space of binary hydrides. The $Superhydra$ database can be thought as a first step of a generalized effort to map conventional superconductivity.

cond-mat.supr-con↗

arXiv2021

Importance of feature engineering and database selection in a machine learning model: A case study on carbon crystal structures

Drive towards improved performance of machine learning models has led to the creation of complex features representing a database of condensed matter systems. The complex features, however, do not offer an intuitive explanation on which physical attributes do improve the performance. The effect of the database on the performance of the trained model is often neglected. In this work we seek to understand in depth the effect that the choice of features and the properties of the database have on a machine learning application. In our experiments, we consider the complex phase space of carbon as a test case, for which we use a set of simple, human understandable and cheaply computable features for the aim of predicting the total energy of the crystal structure. Our study shows that (i) the performance of the machine learning model varies depending on the set of features and the database, (ii) is not transferable to every structure in the phase space and (iii) depends on how well structures are represented in the database.

cond-mat.mtrl-sci↗

arXiv2021

Fused borophenes: a new family of superhard materials

The search of new superhard materials has received a strong impulse by industrial demands for low-cost alternatives to diamond and $c$-BN, such as metal borides. In this Letter we introduce a new family of superhard materials, "fused borophenes", containing 2D boron layers which are interlinked to form a 3D network. These materials, identified through a high-throughput scan of BxC1-x structures, exhibit Vicker's hardnesses comparable to those of the best commercial metal borides. Due to their low formation enthalpies, fused borophenes could be synthesized by high-temperature methods, starting from appropriate precursors, or through quenching of high-pressure phases.

cond-mat.mtrl-sci↗

arXiv2021

Potential energy surface study of X@Si$_{32}$X$^-_{44}$(X=Cl, Br) clusters to decipher the stabilization process of Si$_{20}$ fullerene

Efforts toward stabilization of the Si$_{20}$ fullerene through different schemes have failed despite several theoretical predictions. However, recently Tillmann {\it et. al.} succeeded to stabilize the Si$_{20}$ fullerene through exohedral decoration with eight Cl substituents and twelve SiCl$_3$ groups on the surface and enclosing Cl$^-$ ion. A deeper understanding on what factors lead to stabilization will open the path for stabilizing other systems of interest. Here, we employ the minima hopping method within density functional theory to understand the potential energy surface. The study shows that the exo-endo halide decoration of the cage alters the glassy nature of the potential energy surface of pure cage to structure seeker. Further analysis of different properties of the global minima, reveal that the extra electron instead of residing on the central encapsulated atom in the cage, it is distributed on the cage and increases the encapsulation energy; thereby stabilizing the system. We also provide estimates of the stability for different kind of exo-endo halide decorations and their feasible realization in experiments.

cond-mat.mtrl-sci↗

arXiv2020

Comment on "Pentadiamond: A Hard Carbon Allotrope of a Pentagonal Network of sp2 and sp3 C Atoms"

In a recent Letter, Fujii et. al. [Phys. Rev. Lett. 125, 016001] predicted a new carbon allotrope, pentadiamond, with remarkable mechanical properties: elastic moduli comparable or larger than diamond, and negative Poisson's ratio $μ$ = -0.241. The estimated Vicker's hardness is 422 GPa, $\sim$4.6 times higher than in diamond($\sim$92 GPa), the hardest material known to date. Pentadiamond, however was not investigated. Motivated by this, we used different approaches to estimate the elastic properties of pentadiamond with different approximations based on Density Functional Theory. The tests were also carried out on diamond and other known hard carbon allotropes for validation. Through tests, we found that the elastic constants and elastic moduli of pentadiamond reported in Ref. [Phys. Rev. Lett. 125, 016001] are incorrect, and pentadiamond should be considered a non-auxetic soft carbon allotrope.

cond-mat.mtrl-sci↗

arXiv2020

High-Temperature Conventional Superconductivity in the Boron-Carbon system: Material Trends

In this work we probe the possibility of high-temperature conventional superconductivity in the boron-carbon system, using ab-initio screening. A database of 320 metastable structures with fixed composition (50$\%$/50$\%$) is generated with the Minima-Hopping method, and characterized with electronic and vibrational descriptors. Full electron-phonon calculations on sixteen representative structures allow to identify general trends in $T_{\textrm{c}}$ across and within the four families in the energy landscape, and to construct an approximate $T_{\textrm{c}}$ predictor, based on transparently interpretable and easily computable electronic and vibrational descriptors. Based on these, we estimate that around 10$\%$ of all metallic structures should exhibit $T_{\textrm{c}}$'s above 30 $K$. This work is a first step towards ab-initio design of new high-$T_{\textrm{c}}$ superconductors.

cond-mat.supr-con↗

arXiv2018

Comment on the stability of decorated C 48 B 12 heterofullerene

A good hydrogen storage material should adsorb hydrogen in high concentrations and with optimal binding energies. Numerous mixed carbon boron fullerenes which are decorated with metal atoms were previously constructed by hand and proposed as a promising material in this context. We present a fully ab-initio, unbiased structure search in the configurational space of decorated C48B12 and find that most of the hitherto postulated ground state structures are not ground states. We determine the energetically lowest configurations for Be, Ca, Li and Sc decorated C48B12 clusters.

cond-mat.mtrl-sci↗

arXiv2017

The Elephant in the Room of Density Functional Theory Calculations

Using multiwavelets, we have obtained total energies and corresponding atomization energies for the GGA-PBE and hybrid-PBE0 density functionals for a test set of 211 molecules with an unprecedented and guaranteed $μ$Hartree accuracy. These quasi-exact references allow us to quantify the accuracy of standard all-electron basis sets that are believed to be highly accurate for molecules, such as Gaussian-type orbitals (GTOs), all-electron numeric atom-centered orbitals (NAOs) and full-potential augmented plane wave (APW) methods. We show that NAOs are able to achieve the so-called chemical accuracy (1 kcal/mol) for the typical basis set sizes used in applications, for both total and atomization energies. For GTOs, a triple-zeta quality basis has mean errors of ~10kcal/mol in total energies, while chemical accuracy is almost reached for a quintuple-zeta basis. Due to systematic error cancellations, atomization energy errors are reduced by almost an order of magnitude, placing chemical accuracy within reach also for medium to large GTO bases, albeit with significant outliers. In order to check the accuracy of the computed densities, we have also investigated the dipole moments, where in general, only the largest NAO and GTO bases are able to yield errors below 0.01 Debye. The observed errors are similar across the different functionals considered here.

physics.comp-ph↗

arXiv2015

Preparation, structural and magnetic studies on BiFe1-xCrxO3 (x=0.0, 0.05 and 0.1) multiferroic nanoparticles

BiFe1-xCrxO3 (x=0.0, 0.05 and 0.1) nanoparticles are prepared by a combustion method without using any solvent. All the synthesized nanoparticles are single phase in nature, nearly spherical in shape and crystallize in distorted perovskite structure and space group R3c with an average crystallite size of the order of 40 nm. The room temperature magnetization observed in BiFeO3 nanoparticles is larger than that in the bulk. Increasing Cr doping leads to increase in the magnetization and coercivity. Strong superexchange interaction between Fe3+ and Cr3+ atoms is likely to give rise to such increase in magnetization with Cr-doping. Mössbauer data of these nanoparticles show ordered magnetic state in which Fe atoms are in 3+ oxidation states.

cond-mat.mtrl-sci↗

arXiv2015

Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network

Based on an analysis of the short range chemical environment of each atom in a system, standard machine learning based approaches to the construction of interatomic potentials aim at determining directly the central quantity which is the total energy. This prevents for instance an accurate description of the energetics of systems where long range charge transfer is important as well as of ionized systems. We propose therefore not to target directly with machine learning methods the total energy but an intermediate physical quantity namely the charge density, which then in turn allows to determine the total energy. By allowing the electronic charge to distribute itself in an optimal way over the system, we can describe not only neutral but also ionized systems with unprecedented accuracy. We demonstrate the power of our approach for both neutral and ionized NaCl clusters where charge redistribution plays a decisive role for the energetics. We are able to obtain chemical accuracy, i.e. errors of less than a milli Hartree per atom compared to the reference density functional results. The introduction of physically motivated quantities which are determined by the short range atomic environment via a neural network leads also to an increased stability of the machine learning process and transferability of the potential.

cond-mat.mtrl-sci↗