Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
&
'
Overview
introduction + motivation numerical methods results The HeisenbergStar The Sierpinskigasket conclusion
&
'
Introduction
quantum spin systems = quantum many-body systems basic element = magnetic moment of particles = spin spin - quantum object with ,,uctuation
interaction = exchange between spins i j s s isotropic Heisenberg exchange J ss s i sj s i sj s i sj J (ij ) = J (xx + yy + zz ) & %
ferromagnet
antiferromagnet
f ... frustration, T ... temperatur d ... dimension,z ... coordination number, S ... spin quantum number & %
disorder
f*T z*S*d
'
Motivation
temperatur
discovery by Bednorz & Mller 1986 on u doped La2 CuO4 with Tc =35K (recent progress:
max Tc =125K)
400 antiferromagnetism
BCS theory not able to describe this phenomenon current research magnetic correlations are essential for describing HTSC magnetic properties of the undoped cuprates are well described by Heisenberg quantum spin systems magnetic order destroyed by doping doping frustration
0.1
0.2
doping
0.3
&
'
&
' What is known about the magnetic order of Heisenberg quantum spin systems with spin 1/2 ?
Mermin-Wagner theorem for Heisenberg spin systems with d = 1 or 2
T > 0 longrange magnetic order (LRO) impossible T=0 (ground state) LRO possible
&
'
Aim
analytic solution are rare 1. search for models with analytic solution HeisenbergStar
d=1: no LRO
3. What about the transition between d=1 and d=2? & Sierpinskigasket
8
'
Numerical methods
1. full diagonalization of small systems (N15) calculation of thermodynamic properties 2. exact diagonalization with Lanczos for systems up to N=36 calculation of the ground state and some excitations 3. Decoupled-Cell quantum Monte-Carlo method for systems up to N1000 importance sampling method for calculating thermodynamical properties single spin ip due to a local surrounding (decoupled cell) avoiding ,,sign problem 4. variational wave function calculation for systems up to N1000 improving variational parameters for an ansatz wave function calculation of the ground state and rst excitation
&
'
Results
The HeisenbergStar
s0 si + J2 HR {si }
i=1
Si
J2
Sj
S0
&
'
surrounding HR {si }
1 N
N
LK HR =
si si+1
i=1
&
LM HR
4 = 2 N
si sj
i,j=1 iA,jB
11
'
linear chain: quasi-analytic solution with Bethe-Ansatz square lattice: exact diagonalization with N26 and extrapolation N Lieb-Mattis model: analytic solution
E0 -0.25 linear chain square lattice Lieb-Mattis model -0.5 0 0.25 0.5 J2/J1 0.75 1
&
12
'
P =
| si sj |.
i,j=1
rst transition analytically soluble (independent of surrounding and system size) surrounding ferromagnetic until 1 J2 /J1 = 4 minimum P maximum E0 strong frustration rapid change of P N : destruction of LRO for linear chain and square lattice by frustration
&
J2 >> J1 : central spin decouples magnetic order is known: linear chain - no LRO square lattice and Lieb-Mattis LRO present
13
order-from-disorder eect strong frustration high degeneracy of the ground state thermal uctuations lifting of degeneracy the state is more symmetric spin correlations are increasing T : paramagnet
kT
0.5
&
14
'
HeisenbergStar analytic solution frustration reduces or destroys the magnetic order frustration + thermal uctuations order-from-disorder eect
&
15
'
The Sierpinskigasket
H =J
i,j=1
si sj
&
'
13
14
6 12 3 5 8
11
10
classical longrange order with 3sublattice structure 120 angle between the classical spins
Are the quantum uctuations able to destroy the classical longrange order?
consideration of spin correlation, ground state energy, energy spectra, lowtemperature specic heat of signatures for magnetic ordering
17
&
'
linear chain: power law decay (analytical solution) square lattice: slight decay to a constant (not seen due to small lattice) Sierpinskigasket: strong (exponential) decay
'
if e0 is independent of N hint for a disordered ground state exact diagonalization and variational calculations
square lattice -0.65 ED VMC -0.35 Sierpinski-gasket ED VMC
e0
0
-0.67 8 eED()=-0.671
-0.38
1/N3/2
N=16
N=1095
N=15
1/N
N=6
square lattice: eo scales as expected with N 3/2 Sierpinskigasket: no scaling of e0 with system size
'
energy spectrum
for lattices with LRO: Pisatower structure of the low lying energy levels Emin (S) = E0 + AS(S + 1) ; A 1/N
N : degeneracy of the low lying states symmetry broken ground state with LRO
-0.1 Es -0.3 -0.3 -0.5 square lattice N=16 2 3 S(S+1) 4 -0.4 S=1/2 3/2 Sierpinski-gasket N=15 5/2 7/2 S(S+1) -0.2 Es
-0.7 S= 0 1
no Pisatower for Sierpinskigasket additional states wit S=1/2 below the rst S=3/2 excitation
&
'
full diagonalization for N=15 Decoupled-Cell quantum Monte-Carlo method for N1095
C 0.2
ordered systems one broad maximum Sierpinski-gasket: additional maxima for kT < 0.5 very similar to the d=2 kagome lattice (disordered lattice?)
'
Conclusion
investigation of ground state and thermodynamics of antiferromagnetic Heisenberg quantum spin systems with s=1/2 ??? How does frustration or dimension inuence the magnetic order ???
Heisenberg-Star: analytic solution frustration weakening or destruction of longrange magnetic order frustration + thermal uctuations order-from-disorder eect increase of dimension or coordination number stabilization of magnetic order Sierpinskigasket with dimension df < 2 + frustration destruction of magnetic order
&
22