Temperature-dependence of magnetism of free Fe clusters

Similar documents
Electronic, magnetic and spectroscopic properties of free Fe clusters

Size- and site-dependence of XMCD spectra of iron clusters from ab-initio calculations

Induced magnetic moments make MAE calculation fun or nightmare

EFFECTIVE MAGNETIC HAMILTONIANS: ab initio determination

The Munich SPRKKR-program package A spin polarised relativistic Korringa-Kohn-Rostoker code for calculating solid state properties

Tuning magnetic anisotropy, Kondo screening and Dzyaloshinskii-Moriya interaction in pairs of Fe adatoms

Pressure dependence of Curie temperature and resistivity of complex Heusler alloys

shows the difference between observed (black) and calculated patterns (red). Vertical ticks indicate

Ab Initio Calculation of Exchange Interactions, Adiabatic Spin-Waves, and Curie Temperature of Itinerant Ferromagnets

MAGNETIC ANISOTROPY IN TIGHT-BINDING

Teoría del Funcional de la Densidad (Density Functional Theory)

COLLEGE PHYSICS. Chapter 30 ATOMIC PHYSICS

Magnetic properties of spherical fcc clusters with radial surface anisotropy

Skyrmions à la carte

Density Functional Theory. Martin Lüders Daresbury Laboratory

Neutron and x-ray spectroscopy

Electronic and Magnetic properties of pure and doped manganese clusters

Quantum Theory of Many-Particle Systems, Phys. 540

Magnetic structure of free iron clusters compared to iron crystal surfaces

Impact of magnetism upon chemical interactions in Fe alloys

Band calculations: Theory and Applications

One-dimensional magnetism of one-dimensional metallic chains in bulk MnB 4.

Phase transitions and finite-size scaling

Electronic correlations in models and materials. Jan Kuneš

2 B B D (E) Paramagnetic Susceptibility. m s probability. A) Bound Electrons in Atoms

Coupled-cluster theory for medium-mass nuclei

Finite-temperature magnetism in bcc Fe under compression. Xianwei Sha* and R. E. Cohen

Surface effects in frustrated magnetic materials: phase transition and spin resistivity

What did you learn in the last lecture?

Chapter 3. The (L)APW+lo Method. 3.1 Choosing A Basis Set

The Gutzwiller Density Functional Theory

Electron spins in nonmagnetic semiconductors

Electron transport simulations from first principles

Skyrmion à la carte. Bertrand Dupé. Skyrmion à la carte Bertrand Dupé. Institute of Physics, Johannes Gutenberg University of Mainz, Germany

Nuclear Structure and Reactions using Lattice Effective Field Theory

Correlation Between Magnetism and Structure in Fe alloys: the case of Fe-Cr and Fe-Pt

ELECTRONIC STRUCTURE OF DISORDERED ALLOYS, SURFACES AND INTERFACES

M04M.1 Particles on a Line

Anisotropic Magnetic Structures in Iron-Based Superconductors

Ni 8 Cu n Ni 9. Lectue 4 Trilayers a prototype of multilayers. for FM1 and FM2 interlayer exchange coupling IEC, J inter

1. A ring of radius α has a charge distribution on it that varies as λ(θ) = λ 0 sin(θ), where λ 0 > 0, as shown in the figure.

Chapter 2. Theoretical background. 2.1 Itinerant ferromagnets and antiferromagnets

Spins and spin-orbit coupling in semiconductors, metals, and nanostructures

复习题. 2 Calculate the intensity of magnetic field in the air gap of the magnetic circuit shown in the figure. Use the values N=200,

Finite-temperature magnetism of ultrathin lms and nanoclusters PhD Thesis Booklet. Levente Rózsa Supervisor: László Udvardi

Mean-field theory. Alessandro Vindigni. ETH October 29, Laboratorium für Festkörperphysik, ETH Zürich

Feynman diagrams in nuclear physics at low and intermediate energies

Exchange bias in core/shell magnetic nanoparticles: experimental results and numerical simulations

Magnetism. Eric Bousquet. University of Liège. Abinit School, Lyon, France 16/05/2014

Proximity-induced magnetization dynamics, interaction effects, and phase transitions on a topological surface

tunneling theory of few interacting atoms in a trap

Introduction to spin and spin-dependent phenomenon

University of New Mexico

Nuclear Structure for the Crust of Neutron Stars

J 12 J 23 J 34. Driving forces in the nano-magnetism world. Intra-atomic exchange, electron correlation effects: Inter-atomic exchange: MAGNETIC ORDER

A Monte Carlo Implementation of the Ising Model in Python

Damping of magnetization dynamics

First principle calculations of plutonium and plutonium compounds: part 1

DFT EXERCISES. FELIPE CERVANTES SODI January 2006

1. Electricity and Magnetism (Fall 1995, Part 1) A metal sphere has a radius R and a charge Q.

Shell model description of dipole strength at low energy

Department of Physics PRELIMINARY EXAMINATION 2015 Part II. Long Questions

Intermediate valence in Yb Intermetallic compounds

Coupled Cluster Method for Quantum Spin Systems

COLLECTIVE SPIN STATES IN THE ELECTRON GAS IN DIFFERENT DIMENSIONS AND GEOMETRIES*)

DYNAMICS of a QUANTUM VORTEX

Lectures on magnetism at the Fudan University, Shanghai October 2005

Magnetic anisotropy in frustrated clusters and monolayers: Cr on triangular Au(111) surface

ICCP Project 2 - Advanced Monte Carlo Methods Choose one of the three options below

Block Iterative Eigensolvers for Sequences of Dense Correlated Eigenvalue Problems

Quantum spin systems - models and computational methods

Metropolis Monte Carlo simulation of the Ising Model

Introduction to Density Functional Theory

Nuclear structure from chiral-perturbation-theory two- plus three-nucleon interactions

1. Nuclear Size. A typical atom radius is a few!10 "10 m (Angstroms). The nuclear radius is a few!10 "15 m (Fermi).

Markov Chain Monte Carlo The Metropolis-Hastings Algorithm

Introduction to MRI. Spin & Magnetic Moments. Relaxation (T1, T2) Spin Echoes. 2DFT Imaging. K-space & Spatial Resolution.

New T=1 effective interactions for the f 5/2 p 3/2 p 1/2 g 9/2 model space: Implications for valence-mirror symmetry and seniority isomers

The Overhauser Instability

w2dynamics : operation and applications

Ferromagnetism and Metal-Insulator Transition in Hubbard Model with Alloy Disorder

Skyrmions and Anomalous Hall Effect in a Dzyaloshinskii-Moriya Magnet

E. not enough information given to decide

All-Electron Full-Potential Calculations at O(ASA) Speed A Fata Morgana?

Single Spin Qubits, Qubit Gates and Qubit Transfer with Quantum Dots

Modeling cold collisions Atoms Molecules

Multi-Scale Modeling from First Principles

Decoherence in molecular magnets: Fe 8 and Mn 12

Quantum and classical annealing in spin glasses and quantum computing. Anders W Sandvik, Boston University

Electronic structure of correlated electron systems. Lecture 2

Theory of magnetic interactions in real materials. Mikhail Katsnelson

Computational Nanoscience: Do It Yourself!

NCTS One-Day Workshop. Yasutami Takada. cf. YT, Ryo Maezono, & Kanako Yoshizawa, Phys. Rev. B 92, (2015)

The Magnetic Properties of Superparamagnetic Particles by a Monte Carlo Method

Neutrino-Nucleus Reactions Based on Recent Progress of Shell Model Calculations

Chapter 23 Term083 Term082

First-Principles Calculation of Exchange Interactions

Chapter 3. Magnetic Model. 3.1 Magnetic interactions

Cluster Models for Light Nuclei

The EOS of neutron matter, and the effect of Λ hyperons to neutron star structure

Transcription:

Temperature-dependence of magnetism of free Fe clusters O. Šipr 1, S. Bornemann 2, J. Minár 2, S. Polesya 2, H. Ebert 2 1 Institute of Physics, Academy of Sciences CR, Prague, Czech Republic 2 Universität München, Department Chemie, München, Germany p.1/19

Outline Introducing finite-temperature magnetism (a bit of theory) p.2/19

Outline Introducing finite-temperature magnetism (a bit of theory) Exchange coupling in clusters p.2/19

Outline Introducing finite-temperature magnetism (a bit of theory) Exchange coupling in clusters Dependence of mean magnetization of clusters on temperature and on cluster size p.2/19

Outline Introducing finite-temperature magnetism (a bit of theory) Exchange coupling in clusters Dependence of mean magnetization of clusters on temperature and on cluster size Critical ( Curie ) temperature T c of clusters p.2/19

Outline Introducing finite-temperature magnetism (a bit of theory) Exchange coupling in clusters Dependence of mean magnetization of clusters on temperature and on cluster size Critical ( Curie ) temperature T c of clusters Temperature-dependence of magnetic profiles of clusters p.2/19

Motivation Magnetism is about ordering of magnetic moments p.3/19

Motivation Magnetism is about ordering of magnetic moments Rule of thumb for studying different geometries: follow the coordination number! p.3/19

Motivation Magnetism is about ordering of magnetic moments Rule of thumb for studying different geometries: follow the coordination number! Lowering the coordination number increase of magnetic moments enhanced magnetization decrease of coupling reduced magnetization p.3/19

Motivation Magnetism is about ordering of magnetic moments Rule of thumb for studying different geometries: follow the coordination number! Lowering the coordination number increase of magnetic moments enhanced magnetization decrease of coupling reduced magnetization When comparing surfaces with bulk: competition between enhancement of magnetic moments and reduction of their coupling p.3/19

Bulk, surfaces, clusters,... Clusters contain a large portion of surface atoms p.4/19

Bulk, surfaces, clusters,... Clusters contain a large portion of surface atoms However, their properties cannot be expressed as a mere linear combination of surface and bulk contributions p.4/19

Bulk, surfaces, clusters,... Clusters contain a large portion of surface atoms However, their properties cannot be expressed as a mere linear combination of surface and bulk contributions Finite cluster dimensions and associated quantum size effects are significant factors p.4/19

Bulk, surfaces, clusters,... Clusters contain a large portion of surface atoms However, their properties cannot be expressed as a mere linear combination of surface and bulk contributions Finite cluster dimensions and associated quantum size effects are significant factors intuition has a limited role p.4/19

Bulk, surfaces, clusters,... Clusters contain a large portion of surface atoms However, their properties cannot be expressed as a mere linear combination of surface and bulk contributions Finite cluster dimensions and associated quantum size effects are significant factors intuition has a limited role one has to perform calculations for the real stuff p.4/19

Systems to be studied Free spherical Fe clusters, geometry taken as if cut from a bcc Fe crystal Cluster sizes: between 9 atoms (1 coordination shell) and 89 atoms (7 coordinations shells) shells atoms radius [a.u.] Fe cluster 27 atoms view in Z direction 3rd shell 2nd shell 1st shell center 1 9 4.70 2 15 5.42 3 27 7.67 4 51 8.99 5 59 9.39 6 65 10.85 7 89 11.82 p.5/19

Calculations for T =0 Ab-initio within LDA framework (material specific) Scalar-relativistic real-space spin-polarized multiple-scattering formalism Atomic sphere approximation (ASA) Using SPRKKR code http://olymp.cup.uni-muenchen.de/ak/ebert/sprkkr p.6/19

Calculations for T 0 For localized moments, finite-temperature magnetism can be described by a classical Heisenberg hamiltonian H eff = i j J ij e i e j e i e j p.7/19

Mapping DFT onto Heisenberg Coupling constants J ij can be obtained from ground-state electronic properties: J ij = 1 EF 4π Im de Tr [ (t 1 i t 1 i ) τ ij (t 1 j t 1 j ) τ ji ] [Liechtenstein et al. JMMM 67, 65 (1987)] p.8/19

Mapping DFT onto Heisenberg Coupling constants J ij can be obtained from ground-state electronic properties: J ij = 1 EF 4π Im de Tr [ (t 1 i t 1 i ) τ ij (t 1 j t 1 j ) τ ji ] [Liechtenstein et al. JMMM 67, 65 (1987)] Valid only if magnetism can be described by localized magnetic moments (fine for Fe) p.8/19

From J ij to M(T ) Mean magnetization M(T ) of a system described by a classical Heisenberg hamiltonian is M(T ) = k M k exp( E k k exp( E k k B T ) k B T ) M k is the magnetization of the system for a particular configuration k of the directions of spins E k is the energy of configuration k p.9/19

From J ij to M(T ) Mean magnetization M(T ) of a system described by a classical Heisenberg hamiltonian is M(T ) = k M k exp( E k k exp( E k k B T ) k B T ) M k is the magnetization of the system for a particular configuration k of the directions of spins E k is the energy of configuration k Practical evaluation: Monte Carlo method with the importance sampling Metropolis algorithm p.9/19

From J ij to M(T ) Mean magnetization M(T ) of a system described by a classical Heisenberg hamiltonian is M(T ) = k M k exp( E k k exp( E k k B T ) k B T ) M k is the magnetization of the system for a particular configuration k of the directions of spins E k is the energy of configuration k Practical evaluation: Monte Carlo method with the importance sampling Metropolis algorithm For bulk Fe, this procedure yields finite-temperature results that are in a good agreement with experiment [Pajda et al. PRB 64, 174402 (2001)] p.9/19

Fe clusters at T =0 mag. moment per atom [ B ] 3.2 3.0 2.8 2.6 2.4 2.2 Average mag. moments bulk 2.0 0 20 40 60 80 100 cluster size (number of atoms) Average magnetic moment of clusters oscillates with cluster size Further reading: O. Šipr et al. Phys. Rev. B 70, 174423 (2004) p.10/19

Fe clusters at T =0 mag. moment per atom [ B ] 3.2 3.0 2.8 2.6 2.4 2.2 Average mag. moments bulk 2.0 0 20 40 60 80 100 cluster size (number of atoms) mag. moment at site [ B ] 3.2 3.0 2.8 2.6 2.4 2.2 2.0 89-atom cluster 0 2 4 6 8 10 12 distance from center [a.u.] Average magnetic moment of clusters oscillates with cluster size Local magnetic moments increase when going from the center outwards in an oscillatory way Further reading: O. Šipr et al. Phys. Rev. B 70, 174423 (2004) p.10/19

J ij in bulk and in clusters 50 Exchange coupling between atoms 40 coupling parameter J ij [mev] 30 20 10 bulk crystal bulk 0-10 4 6 8 10 12 distance between atoms i and j [a.u.] Atom i is fixed, atom j scans coordination shells around i p.11/19

J ij in bulk and in clusters 50 Exchange coupling between atoms 40 coupling parameter J ij [mev] 30 20 10 bulk crystal 89-atom cluster, central atom bulk central 0 central -10 4 6 8 10 12 distance between atoms i and j [a.u.] Atom i is fixed, atom j scans coordination shells around i p.11/19

J ij in bulk and in clusters 50 Exchange coupling between atoms 40 coupling parameter J ij [mev] 30 20 10 bulk outermost crystal 89-atom cluster, central atom 89-atom cluster, outermost atom bulk central 0 central -10 4 6 8 10 12 distance between atoms i and j [a.u.] outermost Atom i is fixed, atom j scans coordination shells around i Oscillatory decay of J ij with distance J ij for a given distance differs a lot between clusters and crystals p.11/19

Site-dependence of j J ij Total strength with which one spin (at site i) is held in its direction: Energy needed to flip the spin of atom i while keeping all the remaining spins collinear: J i = j i J ij p.12/19

Site-dependence of j J ij sum of all coupling parameters J i = j J ij [mev] 300 200 100 Total coupling of a particular atom 15 cluster of 15 atoms 0 2 4 6 8 10 12 distance of atom i from the center of the cluster [a.u.] p.12/19

Site-dependence of j J ij sum of all coupling parameters J i = j J ij [mev] 300 200 100 51 Total coupling of a particular atom 15 cluster of 51 atoms cluster of 15 atoms 0 2 4 6 8 10 12 distance of atom i from the center of the cluster [a.u.] p.12/19

Site-dependence of j J ij sum of all coupling parameters J i = j J ij [mev] 300 200 100 51 Total coupling of a particular atom 15 cluster of 89 atoms cluster of 51 atoms cluster of 15 atoms 0 2 4 6 8 10 12 distance of atom i from the center of the cluster [a.u.] 89 p.12/19

Site-dependence of j J ij sum of all coupling parameters J i = j J ij [mev] 300 200 100 51 Total coupling of a particular atom 15 bulk crystal cluster of 89 atoms cluster of 51 atoms cluster of 15 atoms 0 2 4 6 8 10 12 distance of atom i from the center of the cluster [a.u.] 89 Mild differences in J ij translate into large differences in j J ij No systematics in cluster size, no systematics in the position of atom within a cluster p.12/19

Decay of magnetization with T 3.5 3.0 Temperature-dependence of magnetization mean magnetization M(T) [ B ] 2.5 2.0 1.5 1.0 0.5 bulk cluster of 89 atoms crystal 89 [Bulk M(T ) curve was extrapolated to calculated T C ] 0.0 0 200 400 600 800 1000 1200 temperature T [K] M(T ) curves are more shallow in clusters than in bulk p.13/19

Decay of magnetization with T 3.5 3.0 Temperature-dependence of magnetization mean magnetization M(T) [ B ] 2.5 2.0 1.5 1.0 0.5 bulk cluster of 59 atoms cluster of 89 atoms crystal 89 59 [Bulk M(T ) curve was extrapolated to calculated T C ] 0.0 0 200 400 600 800 1000 1200 temperature T [K] M(T ) curves are more shallow in clusters than in bulk Small clusters more shallow M(T ) curves than large clusters p.13/19

Decay of magnetization with T 3.5 3.0 Temperature-dependence of magnetization mean magnetization M(T) [ B ] 2.5 2.0 1.5 1.0 0.5 bulk cluster of 27 atoms cluster of 59 atoms cluster of 89 atoms crystal 27 89 59 [Bulk M(T ) curve was extrapolated to calculated T C ] 0.0 0 200 400 600 800 1000 1200 temperature T [K] M(T ) curves are more shallow in clusters than in bulk Small clusters more shallow M(T ) curves than large clusters p.13/19

Decay of magnetization with T 3.5 3.0 Temperature-dependence of magnetization mean magnetization M(T) [ B ] 2.5 2.0 1.5 1.0 0.5 bulk cluster of 15 atoms cluster of 27 atoms cluster of 59 atoms cluster of 89 atoms crystal 15 27 89 59 [Bulk M(T ) curve was extrapolated to calculated T C ] 0.0 0 200 400 600 800 1000 1200 temperature T [K] M(T ) curves are more shallow in clusters than in bulk Small clusters more shallow M(T ) curves than large clusters p.13/19

Experimental and theoretical M(T ) 3.5 mean magnetization M(T) [ B ] 3.0 2.5 2.0 1.5 1.0 0.5 50-60 82-92 M(T) by experiment and by theory 50-60 atoms, experiment 82-92 atoms, experiment 59 atoms, theory 89 atoms, theory 89 59 Experiment from Billas et al. PRL 71, 4067 (1993) 0.0 0 200 400 600 800 1000 1200 temperature T [K] Caution: Each experimental curve corresponds to a range of cluster sizes (it represents an average over several M(T ) curves which may differ quite a lot one from another) p.14/19

M as function of cluster size mean magnetization M(T) [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Magnetization as function of cluster size 0 K 0.0 0 20 40 60 80 100 cluster size (number of atoms) p.15/19

M as function of cluster size mean magnetization M(T) [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Magnetization as function of cluster size 0 K 90 K 290 K 0.0 0 20 40 60 80 100 cluster size (number of atoms) Dependence of M on cluster size does not really vary with T for low ( experimental ) temperatures p.15/19

M as function of cluster size mean magnetization M(T) [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Magnetization as function of cluster size 0 K 90 K 290 K 610 K 810 K 1130 K 0.0 0 20 40 60 80 100 cluster size (number of atoms) Dependence of M on cluster size does not really vary with T for low ( experimental ) temperatures For large T, magnetization of large clusters is significantly reduced p.15/19

Dependence of T c on cluster size 3.5 3.0 M(T) and its derivative mean magnetization M(T) [ B ] 2.5 2.0 1.5 1.0 0.5 59 atoms, theory 89 atoms, theory 89 59 0.0 0 200 400 600 800 1000 1200 temperature T [K] Critical temperature T c defined as the inflection point of M(T ) curves (no phase transition for finite systems) p.16/19

Dependence of T c on cluster size 1200 1000 bulk critical temperature T c [K] 800 600 400 200 clusters crystal clusters, calculated with proper J ij mean magnetization M(T) [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 59 atoms, theory 89 atoms, theory M(T) and its derivative 0.0 0 200 400 600 800 1000 1200 temperature T [K] 89 59 0 0 50 100 150 200 cluster size (number of atoms) Critical temperature T c defined as the inflection point of M(T ) curves (no phase transition for finite systems) T c oscillates with cluster size p.16/19

Dependence of T c on cluster size 1200 1000 bulk critical temperature T c [K] 800 600 400 200 proper J ij bulk-like J ij crystal clusters, calculated with proper J ij clusters, calculated with bulk-like J ij mean magnetization M(T) [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 59 atoms, theory 89 atoms, theory M(T) and its derivative 0.0 0 200 400 600 800 1000 1200 temperature T [K] 89 59 0 0 50 100 150 200 cluster size (number of atoms) Critical temperature T c defined as the inflection point of M(T ) curves (no phase transition for finite systems) T c oscillates with cluster size Proper cluster-adjusted J ij have to be taken into account p.16/19

Shell-resolved magnetization Expectations: outer shells have smaller coordination numbers than inner shells M in outer shells should decay more quickly with T than M of inner shells p.17/19

Shell-resolved magnetization 1.0 normalized projection of magnetization 0.8 0.6 0.4 0.2 Shell-resolved projection of M(T) for a 89-atom cluster 1 st shell 7 th shell whole cluster Projecting M of a shell onto the direction of the total M of the 89-atom cluster (Projections are normalized to T =0) 0.0 0 200 400 600 800 1000 1200 temperature T [K] p.17/19

Shell-resolved magnetization 1.0 normalized projection of magnetization 0.8 0.6 0.4 0.2 Shell-resolved projection of M(T) for a 89-atom cluster 1 st shell 2 nd shell 7 th shell whole cluster Projecting M of a shell onto the direction of the total M of the 89-atom cluster (Projections are normalized to T =0) 0.0 0 200 400 600 800 1000 1200 temperature T [K] Not monotonous in order of shells p.17/19

Shell-resolved magnetization 1.0 normalized projection of magnetization 0.8 0.6 0.4 0.2 Shell-resolved projection of M(T) for a 89-atom cluster 1 st shell 2 nd shell 3 rd shell 7 th shell whole cluster Projecting M of a shell onto the direction of the total M of the 89-atom cluster (Projections are normalized to T =0) 0.0 0 200 400 600 800 1000 1200 temperature T [K] Not monotonous in order of shells Although M of outer shells usually decays faster than M of inner shells, no systematics can be found. p.17/19

Magnetic profile for T 0 projection of shell-resolved magnetization [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 4 6 8 10 12 distance from the center of the cluster [a.u.] 0 K Cluster of 89 atoms (7 coordination shells) p.18/19

Magnetic profile for T 0 projection of shell-resolved magnetization [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Magnetic profile of 89-atom cluster depending on temperature 0.0 4 6 8 10 12 distance from the center of the cluster [a.u.] 0 K 90 K 290 K 610 K 810 K 1130 K Cluster of 89 atoms (7 coordination shells) Magnetic profile gets more flat if temperature increases p.18/19

Magnetic profile for T 0 projection of shell-resolved magnetization [ B ] 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Magnetic profile of 89-atom cluster depending on temperature 0.0 4 6 8 10 12 distance from the center of the cluster [a.u.] 0 K 90 K 290 K 610 K 810 K 1130 K Cluster of 89 atoms (7 coordination shells) Magnetic profile gets more flat if temperature increases M of outermost layers is similar in magnitude to M of inner layers even for large T (i.e., no drastic decrease of surface M) p.18/19

Summary Exchange coupling constants J ij in clusters differ from the bulk, with no obvious systematics p.19/19

Summary Exchange coupling constants J ij in clusters differ from the bulk, with no obvious systematics Magnetization M(T ) curves are more shallow for small clusters than for large clusters p.19/19

Summary Exchange coupling constants J ij in clusters differ from the bulk, with no obvious systematics Magnetization M(T ) curves are more shallow for small clusters than for large clusters For large T, magnetization of large clusters is significantly reduced p.19/19

Summary Exchange coupling constants J ij in clusters differ from the bulk, with no obvious systematics Magnetization M(T ) curves are more shallow for small clusters than for large clusters For large T, magnetization of large clusters is significantly reduced Critical temperature T c oscillates with cluster size p.19/19

Summary Exchange coupling constants J ij in clusters differ from the bulk, with no obvious systematics Magnetization M(T ) curves are more shallow for small clusters than for large clusters For large T, magnetization of large clusters is significantly reduced Critical temperature T c oscillates with cluster size (Normalized) M of the outer shells usually decreases with T more quickly than M of inner shells p.19/19

Summary Exchange coupling constants J ij in clusters differ from the bulk, with no obvious systematics Magnetization M(T ) curves are more shallow for small clusters than for large clusters For large T, magnetization of large clusters is significantly reduced Critical temperature T c oscillates with cluster size (Normalized) M of the outer shells usually decreases with T more quickly than M of inner shells Simple models (such as taking J ij from bulk) do not work, systematical trends cannot be guessed beforehand one really has to calculate all the quantities of interest p.19/19