arXiv subjects

Edan T. Marcial

Publications and source records attributed to Edan T. Marcial.

3 recordsLinked to original sources

arXiv2026

Performance of universal machine learning potentials in global optimization of inorganic crystal structures

Rapid development of universal machine learning potentials (uMLPs) and expansion of training data sets are reshaping the state of the art in atomistic simulation, highlighting the need for concurrent systematic benchmarking of their capabilities. Global optimization is among the most demanding uMLP applications because unconstrained exploration includes probing motifs not present in reference sets. We examined twelve pretrained uMLPs in unconstrained evolutionary searches to assess whether these models can consistently predict complex nonmagnetic crystal structure ground states under ambient pressure across diverse inorganic systems. Our findings demonstrate that the considered M3GNet, MACE-MATPES, MACE-mh-1, SevenNet, EquiformerV2, EquiformerV3, MatterSim, GRACE, eSEN, Orb-v3, PET-MAD, and PET-OAM models span a wide performance range, from near {\it ab initio} to essentially non-predictive, in their ability to resolve competing phases within low-energy basins. Additional tests on hcp-Zn, MB$_4$ (M = Cr, Mn, and Fe), and LiB$_{y}$ ($y\approx 0.9$) ground states reveal that several uMLPs capture fine energy differences arising from subtle electronic structure features.

cond-mat.mtrl-sci↗

arXiv2024

Stability-superconductivity map for compressed Na-intercalated graphite

A recent ab initio investigation of Na-C binary compounds under moderate pressures has uncovered a possible stable NaC$_4$ superconductor with an estimated critical temperature up to 41K. We revisit this promising binary system by performing a more focused exploration of Na-intercalated graphite configurations, assessing the sensitivity of their thermodynamic stability to density functional approximations at different (T,P) conditions, and examining their superconducting properties with the anisotropic Migdal-Eliashberg formalism. The combinatorial screening of possible Na arrangements reveals additional stable stoichiometries, i.e., Na$_3$C$_{10}$, NaC$_8$, NaC$_{10}$, and NaC$_{12}$, that redefine the previously proposed convex hulls for pressures up to 10 GPa. The evaluation of formation enthalpies with different van der Waals functionals indicates that the proposed compounds might not be thermodynamically stable at zero temperature but some of them could stabilize due to the vibrational entropy or form via cold compression if graphite is used as a starting material. Our more rigorous modeling of the electron-phonon coupling in NaC$_4$ confirms the material's potential for high-temperature superconductivity, with a critical temperature reaching 48 K at 10 GPa, and reveals a well-defined two-gap structure unusual for an electron-doped compound. By tracking the position of the intercalant nearly free electron states with respect to the Fermi level in viable Na-C compounds, we map out the range of pressures and compositions needed for strong electron-phonon coupling and identify Na$_3$C$_{10}$ as an equally promising superconductor.

cond-mat.supr-con↗

arXiv2024

Prospect of high-temperature superconductivity in layered metal borocarbides

Delithiation of the known layered LiBC compound was predicted to induce conventional superconductivity at liquid nitrogen temperatures but extensive experimental work over the past two decades has detected no signs of the expected superconducting transition. Using a combination of first-principles stability analysis and anisotropic Migdal-Eliashberg formalism, we have investigated possible Li$_x$BC morphologies and established what particular transformations of the planar honeycomb BC layers are detrimental to the material's superconductivity. We propose that Li$_x$BC reintercalation with select alkali and alkaline earth metals could lead to synthesis of otherwise inaccessible metastable Li$_x$M$_y$BC superconductors with critical temperatures ($T_{c}$) up to 73 K. The large-scale exploration of metal borocarbides has revealed that NaBC and Li$_{1/2}$Na$_y$BC layered phases are likely true ground states at low temperatures. The findings indicate that this compositional space may host overlooked synthesizable compounds with potential to break the $T_{c}$ record for conventional superconductivity in ambient-pressure materials.

cond-mat.supr-con↗