It is virtually impossible to evaluate the magnetic properties of large anisotropic magnetic molecules numerically exactly due to the huge Hilbert space dimensions as well as due to the absence of symmetries. Here we propose to advance the Finite-Tem
perature Lanczos Method (FTLM) to the case of single-ion anisotropy. The main obstacle, namely the loss of the spin rotational symmetry about the field axis, can be overcome by choosing symmetry related random vectors for the approximate evaluation of the partition function. We demonstrate that now thermodynamic functions for anisotropic magnetic molecules of unprecedented size can be evaluated.
The characterization and manipulation of deposited magnetic clusters or molecules on surfaces is a prerequisite for their future utilization. In recent years techniques like spin-flip inelastic electron tunneling spectroscopy using a scanning tunneli
ng microscope proved to be very precise in determining e.g. exchange constants in deposited finite spin chains in the meV range. In this article we tackle the problem numerically by investigating the transition from where a pure spin Hamiltonian is sufficient to the point where the interaction with the surface significantly alters the magnetic properties. To this end we study the static, i.e. equilibrium impurity magnetization of antiferromagnetic chains for varying couplings to a conduction electron band of a metal substrate. We show under which circumstances the screening of a part of the system enables one to deduce molecular parameters of the remainder from level crossings in an applied field.
Free-standing carbon nanomembranes (CNM) with molecular thickness and macroscopic size are fascinating objects both for fundamental reasons and for applications in nanotechnology. Although being made from simple and identical precursors their interna
l structure is not fully known and hard to simulate due to the large system size that is necessary to draw definite conclusions. We performed large-scale classical molecular dynamics investigations of biphenyl-based carbon nanomembranes. We show that one-dimensional graphene-like stripes constitute a highly symmetric quasi one-dimensional ground state. This state does not crosslink. Instead crosslinked structures are formed from highly excited precursors with a sufficient amount of broken phenyls. The internal structure of CNM is very likely a disordered metastable state which is formed in the process of cooling.
We discuss the magnetocaloric properties of gadolinium containing magnetic molecules which potentially could be used for sub-Kelvin cooling. We show that a degeneracy of a singlet ground state could be advantageous in order to support adiabatic proce
sses to low temperatures and simultaneously minimize disturbing dipolar interactions. Since the Hilbert spaces of such spin systems assume very large dimensions we evaluate the necessary thermodynamic observables by means of the Finite-Temperature Lanczos Method.
In this Letter we report how thermodynamic properties of a giant frustrated magnetic Keplerate molecule of N=30 spins s=1/2 can be evaluated with the help of the highly accurate finite-temperature Lanczos method. The comparison to experimental data s
hows excellent agreement. Since this molecule is structurally related to the archetypical kagome lattice antiferromagnet we expect new detailed insight into properties of this important class of frustrated materials.
The very interesting magnetic properties of frustrated magnetic molecules are often hardly accessible due to the prohibitive size of the related Hilbert spaces. The finite-temperature Lanczos method is able to treat spin systems for Hilbert space siz
es up to 10^9. Here we first demonstrate for exactly solvable systems that the method is indeed accurate. Then we discuss the thermal properties of one of the biggest magnetic molecules synthesized to date, the icosidodecahedron with antiferromagnetically coupled spins of s=1/2. We show how genuine quantum features such as the magnetization plateau behave as a function of temperature.
The determination of the energy spectra of small spin systems as for instance given by magnetic molecules is a demanding numerical problem. In this work we review numerical approaches to diagonalize the Heisenberg Hamiltonian that employ symmetries;
in particular we focus on the spin-rotational symmetry SU(2) in combination with point-group symmetries. With these methods one is able to block-diagonalize the Hamiltonian and thus to treat spin systems of unprecedented size. In addition it provides a spectroscopic labeling by irreducible representations that is helpful when interpreting transitions induced by Electron Paramagnetic Resonance (EPR), Nuclear Magnetic Resonance (NMR) or Inelastic Neutron Scattering (INS). It is our aim to provide the reader with detailed knowledge on how to set up such a diagonalization scheme.
