Spin Signature of Nonlocal Correlation Binding in Metal-Organic Frameworks
Artikel i vetenskaplig tidskrift, 2015

We develop a proper nonempirical spin-density formalism for the van der Waals density functional (vdW-DF) method. We show that this generalization, termed svdW-DF, is firmly rooted in the single-particle nature of exchange and we test it on a range of spin systems. We investigate in detail the role of spin in the nonlocal correlation driven adsorption of H-2 and CO2 in the linear magnets Mn-MOF74, Fe-MOF74, Co-MOF74, and Ni-MOF74. In all cases, we find that spin plays a significant role during the adsorption process despite the general weakness of the molecular-magnetic responses. The case of CO2 adsorption in Ni-MOF74 is particularly interesting, as the inclusion of spin effects results in an increased attraction, opposite to what the diamagnetic nature of CO2 would suggest. We explain this counterintuitive result, tracking the behavior to a coincidental hybridization of the O p states with the Ni d states in the down-spin channel. More generally, by providing insight on nonlocal correlation in concert with spin effects, our nonempirical svdW-DF method opens the door for a deeper understanding of weak nonlocal magnetic interactions.

Författare

T Thonhauser

Wake Forest University

S. Zuluaga

Wake Forest University

Calvin A. Arter

Wake Forest University

Kristian Berland

Chalmers, Mikroteknologi och nanovetenskap (MC2), Elektronikmaterial och system

Elsebeth Schröder

Chalmers, Mikroteknologi och nanovetenskap (MC2), Kvantkomponentfysik

Per Hyldgaard

Chalmers, Teknisk fysik, Elektronikmaterial och system

Physical Review Letters

0031-9007 (ISSN) 1079-7114 (eISSN)

Vol. 115 136402

Drivkrafter

Hållbar utveckling

Innovation och entreprenörskap

Styrkeområden

Nanovetenskap och nanoteknik

Energi

Materialvetenskap

Ämneskategorier

Fysik

Atom- och molekylfysik och optik

Teoretisk kemi

Den kondenserade materiens fysik

Fundament

Grundläggande vetenskaper

Infrastruktur

C3SE (Chalmers Centre for Computational Science and Engineering)

DOI

10.1103/PhysRevLett.115.136402

PubMed

26451571