Abstract:
The aim of the study was to investigate the stability of pentlandite (Fe,Ni)9S8 and Mn, Co, Cu, Os, Pt, Ir, Ru, Rh and Pd on pentlandite (Fe,Ni)9S8 bulk using density functional theory (DFT) and explore the (FeNi)9S8 surfaces. Furthermore, derive the atomic potentials for (Fe,Ni)9S8 pentlandite and pentlandite–like structures and study their molecular dynamics (MD). This study employed density functional theory (DFT) using the universal cluster expansion (UNCLE) technique to model various structures of (Fe,Ni)₉S₈ pentlandite, including doping with elements such as Mn, Co, Cu, and precious metals (Ru, Rh, Pd, Os, Ir, and Pt) at the Ni and Fe tetrahedral (T) sites. The cluster expansion (CE) identified Fe-rich Fe₅Ni₄S₈ with a tetragonal space group (P42/nmc) as the most stable configuration. Doping Fe₅Ni₄S₈ at the Fe(T) sites with Mn, Cu, Rh, Pd, Ir, or Pt altered the structure to cubic, while doping with Co, Ru, or Os retained the tetragonal structure. Conversely, doping Ni(T) sites with Co transitioned the structure to cubic, while doping with Mn, Cu, Ru, Rh, Pd, Os, Ir, or Pt preserved the tetragonal structure. Among these, Pd doping at the Fe(T) site produced the most thermodynamically stable compound (Fe₃Pd₂Ni₄S₈) with a heat of formation of –130.24 meV. The density of states (DOS) analysis revealed that Pd and Pt were the most favourable substitutions for Fe(T) in pentlandite, while Co preferred Ni(T) sites. Additionally, Mn doping at Fe(T) and Ru doping at Ni(T) induced magnetism in the structure. This research predicted diverse dopant compositions in pentlandite and provided a basis for the discovery of new mineral compounds using the DFT-CE approach. The second part of this study was to investigate the surface stability of Fe5Ni4S8 (P42/nmc) and Fe3Pd2Ni4S8 which was found to be the most stable Fe5Ni4S8 (P42/nmc) doped with Pd on the Fe(T) site. The structural optimization for (400), (220), (111) and (311) Fe5Ni4S8 surfaces and three surfaces for Fe3Pd2Ni4S8 which were (440), (111) and (311) surfaces was performed. All surfaces had distinct terminations and showed the significant changes after optimization. The surface energies for all investigated optimised surfaces were calculated to predict the most favourable Fe5Ni4S8 and Fe3Pd2Ni4S8 surfaces. The calculated surface energies revealed that the (311) surfaces cleaved from both Fe5Ni4S8 and Fe3Pd2Ni4S8 bulk structures was the most stable surface. Additionally, the spin-polarised density of states and charge density for both Fe5Ni4S8 (400), (220), (111) and (311) and Fe3Pd2Ni4S8 (440), (111) and (311) surfaces were calculated. The third part of this study was to study the molecular dynamics (MD) of Fe5Ni4S8 (P42/nmc) pentlandite and pentlandite-like structures both binary Rh9S8, Ir9S8 and also the ternary Ru4Pd5S8 and Ru5Pd4S8. The derived interatomic potentials were used to perform the MD to see the effect of pressure and temperature on the Fe5Ni4S8 (P42/nmc) pentlandite and Rh9S8, Ir9S8, Fe5Ni4S8, Ru5Pd4S8, Ru4Pd5S8, Pn-like bulk structures. The elastic constants, calculated by ab initio density functional theory (DFT) method using Vienna ab–initio simulation package (VASP) code, were used as input data for the development of potentials models. To analyse our molecular dynamics (MD) results, we calculated the MD properties namely, radial distribution functions (RDFs), diffusion coefficient and Mean square displacement (MSD).