Neutron scattering as a tool to study

quantum magnetism
Radu Coldea
Oxford

Outline
- principles of (magnetic) neutron scattering
- spin waves in a Heisenberg ferromagnet
- spin waves in square-lattice AFM La2CuO4

- quantum renormalization, spinons and method to determine

Hamiltonian triangular AFM S = 1/2 Cs2CuCl4
- quantum phase transition in the Ising chain CoNb2O6 in
transverse field

Neutron reactors
(Institute Laue Langevin)

Three-axis neutron spectrometer

- nuclear fission of Uranium

ki
kf
Q
ki kf = Q
Ei Ef = E

neutron
guide

spectrometers

Clifford
Shull

Bertram
Brockhouse

Spallation neutron sources (ISIS, SNS )

- evaporation when fast protons
hit a heavy nucleus (Ta)
proton synchrotron
accelerator ~800MeV

p+
neutron
guide
time-of-flight
spectrometer

LET@ISIS

~40,000 detector elements

count simultaneously (timestamp each arriving neutron

Magnetic neutron diffraction

Neutrons have
- no charge
- spin-1/2 moment

kf

ki

dipole moment M

Periodic magnetic order

t
=> magnetic Bragg peaks
at Q = t q

magnetic
unit cell

- Intensity ~
Fourier transform of magnetic
moment density (perp to
scattering wavevector Q)

structural

Bragg peak t

magnetic

Inelastic magnetic neutron scattering

kf
ki

sample

excited
state
ground state

matrix element for transition

Fourier transform of magnetic
moment density M a

Fourier transform of
magnetic e- density

polarization factor

Single crystals for inelastic neutron scattering

La2CuO4

Cs2CuCl4
(solution growth)

CoNb2O6
(mirror furnace growth)

4 cm

2.5 cm

7 single-crystal mount ~50 g

(flux growth)

floating-zone mirror furnace

Spin waves in a Heisenberg ferromagnet

H = - Jij SiSj if all Jij > 0 T=0 ground state is
ij
S-

magnon
energy

(q) = -J(q)
(0) +- J(0) + h
1
iqrij
2
Fourier transform
of magnetic couplings

J(q) = _ ijJij e

H = - Jij SiSj - h Siz

ij

= _Jij (Si- Sj+ + Si+Sj- ) +- Jij SizSjz - h Siz

i
ij
ij 2
hopping

coherent propagation of spin-flip states

(if Hamiltonian conservs Sz)
magnon

One spin flip

manifold

S-

exact
result
Zeeman
energy

Wavevector (q)

(q)

Energy

- neutrons flip over one spin

Gap
0

Dispersion images exchange Hamiltonian

|q> =

e iqr |

1
N i

i>

Neutron scattering by ferromagnetic magnons

scattered
incident
neutron
ki

kf

S z = + 1/2

S z = -1/2

Magnon
k

ki - kf = k
Ei - Ef = w(k)
Wavevector k
L (rlu)

Sz = -1

Spin waves in the square-lattice anti-ferromagnet

Ground state has Neel order
(<S> reduced by quantum fluctuations)

J antiferromagnetic

Spin wave excitations

(approximate eigenstates)

La2CuO4
Magnetic Bragg peak (1/2,1/2,0)
- insulating parent of
high-TC cuprates
- square-lattice of CuO2
planes, Cu2+ S=1/2
Keimer at al, PRB (1992)

Neutron scattering experiments on La2CuO4

ki

S=1
magnon
kf

7 single crystal mount, ~50 g

RC et al, PRL 86, 5377 (2001)

Magnetic excitations in La2CuO4

Collect maps of magnetic scattering in the whole
2D Brillouin zone (h,k) at increasing energies E

Energy

325 meV

275 meV

75 meV

Spin-wave dispersion surface

RC et al, PRL 86, 5377 (2001)

Dispersion relations
La2CuO4, T=293 K

Antiferromagnetic
Bragg peak
position
RC et al, PRL 86, 5377 (2001)

Dispersion relation and interactions

- dispersion shape is a direct fingerprint of the magnetic interactions

- wiggle in high-energy dispersion is evidence

for a cyclic-exchange between the 4 spins on a
plaquette in addition to the main exchange J

RC et al, PRL 86, 5377 (2001)

Ring exchange in the Hubbard model

Expand up to 4
electron hops
A.H. MacDonald (1990),
Takahashi (1977)

Hubbard model parameters for La2CuO4

- dispersions and intensities
well described by linear spinwave theory
t = 0.30(2) eV, U = 2.2 (4) eV
U/t = 7.3 1.3 (10 K)
Intensity renormalization factor
Zc = 0.51 +/- 0.13 (predicted 0.61)

RC et al, PRL 86, 5377 (2001)

Magnetic excitations in S =1/2 triangular lattice AFM Cs2CuCl4

- sharp spin-wave mode only very small
weight

- dominant scattering continuum with

strongly-dispersive boundaries
=> Quantum fluctuations very strong, spinwave theory inadequate
1.5

1.25

(0,0.5,l)
Intensity (a.u.)

Intensity (arb. units)

J'

0.1

0.2

E (meV)

0.8

0
0.0

0.2

RC et al, PRB 68, 134424 (2003)

0.4

0.6

Energy (meV)

0.8

1.0

1.2

Experimental method to determine Hamiltonian

via spin waves in the fully-polarized state in high field
RC et al, PRL 88,137203 (2002)

B=0

B >> J

J antiferromagnetic

Fully-polarized
state

State with AF
correlations

BC
Magnetically
ordered, spin
liquid etc.

gapped
magnons

Applied
Field

Exact eigenstate if
Hamiltonian
conserves Sz
Bose condensation
of magnons at Bc

Dispersion is the Fourier

transform of exchange
couplings
(k) = J (k) - J (0) + h
J (k) = 1

S Jd

2 d

exp( i k d)

T=60 mK

1.0
0.8

Bc = 8.42 T
0.6
0.4
0.2

Susceptibility dM/dB (/T)

Magnetization ()

Excitations in the saturated ferromagnetic phase at B=12 T

0.4
0.3
30

0.2
0.1

20

0.0

10

0.0

Lineshapes show sharp

magnon excitations

T=60 mK

10
0

Field // a (Tesla)

10

0.4

Field // a (Tesla)

(021)

2.0

1.5

0.5

0.6

Energy (meV)

(010)

(020)

1.0

0.0

0.5

(021)

[0,2,l]

[0,,-1]
1.5

Energy (meV)

(0,1.447,0)

1.0

1.0

Fourier transform of
couplings J(q)
J=0.374(5) meV
J=0.128(5) meV
_J ~ _1
J
3

J'

0.5

2D Hamiltonian
B = 12 T || a
T < 0.2 K

0.0
1.0

1.5

[0,k,0]
RC et al, PRL 88,137203 (2002)

2.0

J= 0.017(2) meV
interlayer coupling
Da = 0.020(0) meV DM anisotropy _| bc plane

Magnon condensation below critical field induces transverse order

B<Bc

B=Bc

B>Bc

magnon
dispersion

Gap

B (T) || a
4

12

0)
2
0
(

(a)

1.25

1.5

0)
1
(0

1.00

1.0

+
)
1
7,
4
4
.
(0,
)
1
7,
4
4
.
(0,

0.75
0.50
0.25
0.00

Peak intensity
(arb. units)

10

(b) CONE

0.04

FERRO

0.03
0.02

0.5

Bragg peaks appear

where the gap closes

D
150
0.0
1.00 1.25 1.50 1.75 2.00

[0,k,0]

Intensity a.u.)

Energy (meV)

1.50

Energy (meV)

B > Bc
B < Bc

100

50

0.01
0.00
0

B (T) || a

RC et al, PRL 88,137203 (2002)

10

12

0
1.40

1.45

K (rlu)

1.50

Link magnetic order with magnon wavefunctions

Energy (meV)

(000)

Alternating layers
counter-rotate

+D

1.0

-D

0.5

-Q

Sense
selected by
DM couplings

+Q

0.0
-0.5

0.0

0.5

[0 k 0]

D(Si x Sj)

2 condensates at
+Q and -Q
1.6

1.0

1.4
1.2

0.5

1.0
0.0
1.00 1.25 1.50 1.75 2.00

B (T) || a

RC et al, PRL 88,137203 (2002)

[0,k,0]

predicted asymmetry using

magnon wavefunctions

Energy (meV)

Asymmetry <Sb>/<Sc>

1.5

B=Bc

Incommensuration
(2/b)

Quantum renormalization of incommensurate ordering wavevector

Q

B<Bc

mean-field
result

0.05

Bc= 8.44(1)T
c=0.0536(5)

0.04

(0,1.5-,0)
(0,0.5-,1)

0.03
0.02
0

10

Strong quantum
renormalization
20 e0/ecl=0.56
Energy

Collinear
spins
RC et al, PRL 88,137203 (2002)

10

B (T) || a

Wave
vector k
Quantum Classical
Spin-1/2 instability
large-S

12

Magnetic excitations at zero field: spin-waves fractionalize

into pairs of S=1/2 spinons
J'

Si+ Sj- + S-j Si+

J/J ~ 1/3, J=0.37 meV

S=1/2
Spinons

Kohno, Starykh, Balents, Nat. Phys. (07)

Summary on anisotropic triangular AFM Cs2CuCl4

J'

La2CuO4

- anisotropic triangular lattice S=1/2 AFM Cs2CuCl4 has

spiral order coexisting with strong quantum fluctuations
- renormalization of Q-vector and zone-boundary energy
measured by quenching quantum fluctuations via field and
revealing classical behaviour
- dominant continuum scattering (spin-waves fractionalize
into pairs of spinons)
Headings, Hayden, RC , PRL (2010)

Ising magnets and phase transitions

Classical
thermally-driven
continuous phase
transition

- classical Ising model

z

H = - J Si z S j +
i,j

<Sz>

- 2D model Onsager exact solution (1944)

0

<Sz>
quantum fluctuations
add transverse field
- B Sx
quantum tunneling
- B (S++S-) / 2

TC

Temperature

Quantum fluctuations
driven continuous
phase transition

BC

B field

T=0 quantum melting of order

An Ising ferromagnet in transverse field

z

H = - J Si Si+1 +- B S ix
i
T
TC

Quantum
paramagnet

Ferromagnet

0
z

- transverse field
- B Sx = - B (S++S-) / 2
generates quantum fluctuations
that melt the spontaneous
magnetic order at BC ~ J/2

BC
Ordered
moment

Field
x

Magnetization

continuous quantum phase transition

Quantum criticality: experimental challenges

and opportunities

Temperature

Critical
Phase 2

Phase 1

BC

Tuning
parameter

-what is microscopic mechanism of transition, can one observe the

quantum fluctuations that drive transition ?
- how quasiparticles evolve near critical point ?
- what are the fundamental symmetries that govern physics of QCP ?
- what are finite-T properties (interplay of thermal and quantum
fluctuations, under what conditions universal scaling ?

1D Ising chain in transverse field

z

H = - J Si Si+1 + 2 ground states: or

i

domain walls (solitons)

- transverse field B Sx ~ - B (S++S-)

flips spins

Bx
=> propagating solitons
(Jordan Wigner fermions)
classical soliton
(large-S limit)

Gap D

Ferromagnet

solitons

BC

Quantum
paramagnet

spin-flip
quasiparticle

Ising chain at criticality

gapless linear (Dirac) spectrum
w = c|k|
for critical solitons

- w/T scaling expected, special conformal symmetry

- different universality class from Luttinger liquids (1D Heisenberg and XY
AFM chains)

Experimental requirements
1) good 1D character to see solitons
2) low-exchange J ~ 1 meV to access critical field BC ~ J/2 < 10 T
3) strong uniaxial anisotropy (Ising character) but not perfect to still
have transverse g-factor
l LS
Strong Crystal field + Spin Orbit

- best Ising magnets are based on Co2+ 3d7

lowest Kramers doublet effective spin-1/2 Ising-like
2D Ising AF
1D Ising AF

K2CoF4 (Birgeneau 73, Cowley 84)

CsCoCl3 (Goff 95) also CsCoBr3 (Nagler)
J ~ 12 meV BC > 50 T not accessible

Quasi-1D Ising ferromagnet CoNb2O6

Oxygen

zig-zag Co2+ spin chain along c

Ferromagnetic
superexchange
~ 90 bond Co-O-Co
~ 20K ~ 2meV

Ferromagnetic order along chain

Strong easy-axis (Ising) in ac plane

Single crystal
of CoNb2O6
(Oxford
image
furnace)

Co2+

4 cm

Magnetic excitations in 1D phase seen by

T= 5 K, 1D phase above TN
neutron scattering

k1

incident

scattered

- gapped continua characteristic of

2-soliton excitations
RC et al, Science 327, 177 (2010)

k2

k1

k2

Magnetic excitations in zero field

TC

ordered

T= 5 K, 1D phase above TN

1D phase

T= 0.04 K, deep in 3D phase

3D phase

- rich structure : continuum as characteristic of 2-soliton excitations

+ sharp modes (bound states)
RC et al, Science 327, 177 (2010)

BC

Excitations have 1D character

no measurable dispersion chains
|| chain direction

H=-2.5
H=-2.0
H=-1.5

Intensity (arb, units)

0.4
0.3
0.2
0.1
0
1.1

chain direction

1.2

1.3
1.4
Energy (meV)

1.5

Zeeman ladder of bound states in 3D ordered phase

T=5K

Zeeman ladder of bound states

T =40 mK

0.04 K
5K

Intensity (arb units)

1
0.3

L = 0.00(5)
2

0.2

3
4

0.1

0.0
1.0

1.2

1.4

1.6

Energy (meV)

Continuum of
Bound states in
free 2-soliton
confining potential
states
E

2
1

x x x x

Soliton separation
costs energy

V(x) = l x
l ~ J Sz

Longitudinal mean-field -hSz , l=2 h Sz

Zeeman ladder of bound states in 3D ordered phase

T=5K

Zeeman ladder of bound states

T =40 mK

0.04 K
5K

Intensity (arb units)

1
0.3

L = 0.00(5)
2

0.2

3
4

0.1

0.0
1.0

1.2

1.4

1.6

Energy (meV)

Continuum of
Bound states in
free 2-soliton
confining potential
states
E

2
1

x x x x x

Soliton separation
costs energy

V(x) = l x
l ~ J Sz

Longitudinal mean-field -hSz , l=2 h Sz

Zeeman ladder of bound states in 3D ordered phase

T=5K

Zeeman ladder of bound states

T =40 mK

0.04 K
5K

Intensity (arb units)

1
0.3

L = 0.00(5)
2

0.2

3
4

0.1

0.0
1.0

1.2

1.4

1.6

Energy (meV)

Continuum of
Bound states in
free 2-soliton
confining potential
states
E

2
1

x x x x x x

Soliton separation
costs energy

V(x) = l x
l ~ J Sz

Longitudinal mean-field -hSz , l=2 h Sz

Soliton confinement

0.04 K
5K

McCoy&Wu (78)
Schrdingers equation

Intensity (arb units)

1
0.3

L = 0.00(5)
2

0.2

3
4

0.1

0.0

kinetic energy

string tension

=> bound states

Ai(z)

0.5

-0.5
-10

-5

Airy function
zn = 2.33, 4.08, 5.52, 6.78

1.0

1.2

1.4

Energy (meV)

1.6

Phenomenological model of soliton gas

- work perturbatively around the Ising limit
(spin clusters)
J

2-soliton states

=J

-a

-b d n,1

a 2 soliton state
z
gap
from Ising J Siz Si+1

+
i+1

RC et al, Science 327, 177 (2010)

soliton hopping
from XY term

kinetic bound state

i-1

x + Sy Sy
+ S- + S - S+
- Jxy ( Six Si+1
)
~
S
i i+1
i i+1 i i+1

Phenomenological model of soliton gas describes full spectrum

T = 40 mK

Gap:

J ~ 1.94 meV

from Ising zz exchange

Bandwidth

a = 0.12 J domain-wall hopping term

[ microscopic origin SzSx ?]

Kinetic bound state : transverse couplings for nn bond SxSx + SySy , J / J z = 0.24
and 2-nd neighbour AFM along chain J z' = - 0.15 J z
Weak confinement term: hz ~ 0.02 J longitudinal field includes interchain mean-field

numerical calculation agrees with exact analytic solution of effective Hamiltonian

S.B. Rutkevich, J. Stat. Phys (2010)

Experiments in applied transverse field

0T

4T

Energy

3.25 T

Gap decreases
with field
B

- field tunes quasiparticle dispersion

t
torque

wavevector

CoNb2O6
crystal

Field ~ kinetic energy

Place crystal in metallic cage to

prevent movement under high torque

BxSx = (S++S-)/2

Excitations as a function of transverse field

4T

6T

Magnetic 3D LRO
Bragg peak

Two-soliton states
Ordered

Single sharp
mode

Paramagnetic
1000

Excitations change
character above
critical field

800

Intensity (a.u.)

7T

600
400
200
0
0.5

1.0

1.5

Energy(meV)

2.0

Excitations in
transverse field

2-soliton continuum
RC et al, Science 327, 177 (2010)

Summary
- realized experimentally field-tuned quantum phase transition in
quasi 1D Ising magnet CoNb2O6
- observed transmutation of quasiparticles at critical point
ordered
paramagnet
2-soliton
states

spin-flip

Conclusions
- neutron scattering is a very versatile probe of magnetic ordering and
dynamics (meV->eV, magnetic field (15 T->25 T), low T (mK)
- quantitative: probe dispersions of excitations through wellunderstood matrix element (quantitative comparison with theories)
- sample size limited, large crystals needed (advanced crystal growth)
- new neutron sources (ISIS 2nd target station, USA + JAPAN, ESS) will
bring new opportunities: higher flux, higher resolution, wider coverage
50 detectors,
2D data set

40,000 detectors,
4D data set, 4
energies measured
simultaneously,

- complementary to resonant x-ray diffraction and inelastic (RIXS)

(samples 10s of mm, resolution >30 meV, T > few K beam heating)

Collaborators
La2CuO4

CoNb2O6

S.M. Hayden (Bristol),

D.A. Tennant, (HZB Berlin)

G. Aeppli (UCL)

Elisa Wheeler (Oxford)

T.G. Perring(ISIS)
C.D. Frost (ISIS)
T.E. Mason (Oak Ridge)

Ewa Wawrzynska (Bristol)

M. Telling (ISIS)

S.W. Cheong (Rutgers)

K. Habicht, P. Smeibidl, K.Kieffer (HZB)

Z. Fisk (Florida)

D. Prabhakaran (Oxford)

Cs2CuCl4
D. A. Tennant (HZB)
A.M. Tsvelik (Brookhaven)

Ivelisse Cabrera (Oxford)

Jordan Thompson (Oxford)
R. Bewley, T. Guidi (ISIS)

K. Habicht, P. Smeibidl (HZB)

C. Stock, J. Rodriguez-Rivera (NIST)

Z. Tylczynski (Poland)

F. H.L. Essler, N. Robinson (Oxford)

Y. Tokiwa, F. Steglich (Dresden)