Upload
oscar-ramsey
View
215
Download
1
Tags:
Embed Size (px)
Citation preview
Probing Ferroelectricity, Charge Density Wave Dynamics, and
Magnetism with Submicron X-ray DiffractionPaul G. Evans
Department of Materials Science and Engineering
University of Wisconsin, Madison
APS SACMicrobeam ReviewJanuary 21, 2004
Outline• Brief Introduction to Microbeam Experiments at
Sector 7 of the APS• Overview: Physical Phenomena and Motivation• In Depth: Polarization Switching and Fatigue in
PZT Thin Films, Magnetic Domain Evolution in Antiferromagnetic Chromium
• Conclusions, Future Directions
MHATT-CAT’s mission:To develop a productive, open-access center for world-class research in x-ray science exploiting the unique characteristics of the APS, especially timing
and brilliance.
Unique Capabilities:• Ultrafast Laser Facility• Staff Expertise in Ultrafast Optics and X-ray Science, Diffraction, Scattering, and Spectroscopy• White Beam Diffraction with Submicron Focusing with G. Ice-style Mirrors• Small Angle Scattering/XPCS with Fast CCD Direct Detection
MHATT-CAT Sector 7
Center for Real-time X-ray
Studies www.mhatt.aps.anl.gov
• Relaxation dynamics of magnetic polymers using XPCS (collaboration with Ford Research, Dearborn, MI)
• Micro-fluorescence mapping of mineral intrusions (Dan Core, Geosciences Dept., Michigan)
• Development of Li-metal x-ray lenses (Nino Pereia, Ecopulse Inc., Fairfax, VA)
• Studies of core excitations in Kr microjets (Linda Young, ANL)
• Doped magnetic semiconductors (Frank Tsui, U. of North Carolina, Y. Chu, APS)
Microbeam General User
Research at Sector 7
DYNAMICS IN MAGNETORHEOLOGICAL ELASTOMERS
William F. Schlotter, Codrin Cionca, Sirinivas S. Paruchuri, Jevne B. Cunningham, Eric Dufresne, Steve B. Dierker, Dohn Arms and Roy ClarkeUniversity of Michigan, Ann Arbor John M. Ginder and Mark E. NicholsFord Motor Company, Research Laboratory, Dearborn, MI
CCD
Scintillation counter
electromagnet
sample
detectors
Coherent small angle scattering.
MR sample is rigidly mounted while direction of magnetic field is alternated
W. F. Schlotter, et al., Int. J. Mod. Phys. B 16 2426-32 (2002).
focused
unfocused
LITHIUM REFRACTIVE LENSES
Sawtooth Li Lens
Recent results: flux gain ~ 40spot size ~ 20 m
E. M. Dufresne, et al., Appl. Phys. Lett. 79, 4085 (2001).
N. R. Pereira, et al., Rev. Sci. Instrum. 75 37 (2004).
Sawtooth and Parabolic Lithium Lenses
X-ray Microdiffraction as a Tool for Physics and Materials Science
• 6 to 30 keV
• 1010 ph/s/0.01%BW
• Minimum spot (APS, sector 2)- 0.15 x 0.15 µm2
Contrast from: Diffraction, Composition, Ferroelectric Polarization, Magnetization.Problems with existing techniques: time resolution, electrodes, quantification.
Submicron Science with X-ray Diffraction
Strain in MEMS-inspired substrates for SiGe epitaxial growth. (Sector 2)
With B. Lai, Z. Cai (APS),P. Rugheimer, and M. Lagally (U. Wisc.)(P. G. Evans et al., submitted, Jan. 2004)
Charge density wave dynamics in NbSe3 (Sector 2)
With R. Thorne (Cornell U.)
-0.4
-0.2
0
0.2
0.4
-300 -200 -100 0 100 200 300
(220) reflection(004) reflectionsubstrate (220)
rock
ing
cu
rve
pe
ak
shift
(d
eg
.)
position (m)
Si3N
4Si
3N
4bridge
V=0 V=131.3 mV
Submicron Science
Spin density wave antiferromagnetism in Cr(Sector 2)
With E. Isaacs, B. Lai, Z. Cai (ANL)(P. G. Evans et al., Science 2002)
Polarization reversal and piezoelectric distortion in ferroelectric PZT thin films.(Sector 7)
With C.-B. Eom, (U. Wisc) and E. Dufresne (U. Mich/Sector 7)(D.-H. Do et al., submitted, Nov. 2003)
Outline• Brief Introduction to Microbeam Experiments at
Sector 7 of the APS• Overview: Physical Phenomena and Motivation• In Depth: Polarization Switching and Fatigue in
PZT Thin Films, Magnetic Domain Evolution in Antiferromagnetic Chromium
• Conclusions, Future Directions
Epitaxial PZT Thin Film Capacitors
2 Pr 95 C/cm2
angle relative to STO (002) (sec.)
inte
nsity
tetragonal PbZr1-xTixO3
x=0.5580 nm or 160 nm PZT thickness
Time Resolved Microdiffraction with Zone Plate Optics at Sector 7
Piezoelectric xyz stage
Electrical connection to sample
Avalanche photodiode detector (up to ~3 M cps).
Incident beam
Use avalanche photodiode with multichannel scaler to time-resolve the diffraction signal during voltage pulses.
Spatial Resolution: Knife EdgeFluorescence Scan
Aside: Spatial Resolution at Sector 7
0.6 m FWHM
Limitations at Sector 7: Vibrations, Be Windows (?)
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
Diamond Demonstration Version Diamond Demonstration Version Diamond Demonstration Version
(002) x-ray reflection
x-ray reflection
)200(
Intensities are different because PZT lacks inversion symmetry!
Probing Ferroelectric Polarization Switching with X-ray Microdiffraction
X-ray photon energy energy 10 keV.
Maps of the intensity of PZT (002) reflection vs. the position of the beam on the sample. Intensity following –10 V pulse is 25% higher than
Imaging PZT Polarization Switching
0.8V 1.2V 1.6V
2.0V 2.8V
-10V
20 m
1. Apply negative voltage pulse to produce uniform polarization.
Switching with E Ec
2. Apply a positive pulse near Ec.
Device switches in well defined areas.
500 s triangle pulses
-10 -5 0 5 10-20
0
20
40
60
80
100
120
140
Applied Voltage (V)
Δ Pr
SwitchedPolarization
Quantitative Relationship of Polarization to Switched Area
Switched Area Measured from X-ray Microdiffraction Maps is Proportional to the Total Switched Polarization.
Time Resolved Diffraction During Switching- 10 V
+ 10 V
20 m
time (ms)0 1
0 20 40 60 80 1000
10
20
30
40
50
20 40 60 80 100
1750
2000
2250
2500
2750
3000
inte
nsit
ytime (ms)
0 1
time (ms)0 1
20 40 60 80 100
2000
2500
3000
inte
nsit
y
0 20 40 60 80 1000
10
20
30
40
50 +10V
0V
-10V
Applied
Voltage
time (ms)0 1
two
thet
a (d
eg.)
34.52
35.02
Switching of 200 m diameter device limited by RC time constant.
+10V
0V
-10V
Applied
Voltage
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.timems
34.62
34.72
34.82
34.92
35.02
owt
atehtged.
Positive pulses
Negative pulses
Quantitative Measurement of Piezoelectric Distortion
Quantitative measurement: Requires No Standards.
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.timems
34.62
34.72
34.82
34.92
35.02
owt
atehtged.
+10V
0V
-10V
Applied
Voltage
Both polarizations haved33=60 pm V-1.
257-5
An Electromechanical Puzzle
We think: Difference between top and bottom contacts.
During switching experiments the piezoelectric coefficients for positive and negative polarizations are different.
-10 -5 0 5 10-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
Str
ain
(%)
Voltage(V)
257-5 257-7
0 20 40 60 80 1000
10
20
30
40
50
time (ms)
0 1
two
thet
a (d
eg.)
34.52
35.02
Polarization Fatigue
Polarization Fatigue: 160 nm, ± 10 V pulses
Higher Electric Field Pulses Partially Restore the Switchable
Polarization
Polarization Fatigue at Higher Electric Fields
Polarization Fatigue at Higher Electric Fields
Structural Signature of Fatigue at High Electric Fields
Compare Structures at Points Outside the Device With Areas Fatigued at High Electric Fields
Theta-Two Theta Scans at Two Locations
80 nm thickness,105 cycles, ± 10 V
Conclusions: Ferroelectrics
1. Fatigue in PZT devices with Pt electrodes pins the polarization in the state that would normally follow a positive voltage pulse to the bottom electrode.
2. Fatigue at higher electric fields in PZT thin films is accompanied by the gradual spread of a structurally distinct area.
Fatigue
1. X-ray microdiffraction provides a new avenue to studying polarization switching in ferroelectric devices.
2. Quantitative agreement of x-ray observations with electrical measurements of switching.
3. Potential to take advantage of the flexibility of x-ray scattering techniques: resonant scattering, time resolved diffraction…
Imaging
Spin density wave domains in chromium
• Domains are responsible for macroscopically observable magnetic, mechanical, electrical phenomena.
• Previous domain imaging experiments are at the 1 mm scale.
Cr is a spin density wave (SDW) antiferromagnet.
Antiferromagnetic Domains: •Modulation direction Q any <001>
Three possible Q domains.•Spin polarizations S, also <001>
Two S domains in transverse phase. Just one S (||Q) in longitudinal phase.
Cr unit cell
S || [100]
Q || [001]
•SDW leads to strain wave and charge density wave (CDW).
•Spins are transverse T=123 to 311 K, longitudinal for T<123 K.
Fermi surface nesting in Cr(q) susceptibility, response of hypothetical non-magnetic
system to magnetic perturbation with wavevector q
Difficult to calculate (q) directly, but common feature is
K KqK
qEE
1)(
where EK+q and EK are pairs of filled and empty states differing in wavevector by q.
QSDW
Band structure of Cr: Fermi surfaces nest with Q=(0,0,1-) , incommensurate with lattice.
Force Radiation
-eE electric dipole “Thomson scattering”
-eEmagnetic quadrupole
electric dipole)( H
H magnetic dipole
After F. de Bergevin and M. Brunel, Acta Cryst. A 37 314 (1981).
EE
EH
H E
H H
-e
-e
-e
-e
(Non-Resonant) Magnetic X-ray Scattering: Classical Picture
Cr in reciprocal space
CDWnear (0 0 2)
SDWnear (0 0 1)
L
K
Magnetic scattering appears near forbidden lattice reflections.
Also: Strain wave (CDW) reflections near allowed lattice reflections.
Form images using either type of reflection.
H
incident beam k
diffracted beam k’
ST
SL
ST
Non-resonant magnetic x-ray diffraction from Cr
Most important term of cross section scales as:
2)ˆˆ( 'kkS
Polar plot of cross section as a function of spin direction for a Q || (001) domain in our geometry.
k’-k || Cr (0 0 1-)
APS Station 2ID-D
All three Q domains are present
Room temperature laboratory diffractometer scans with large mm-scale beam.
h scan near (2 0 0)
k scan near (0 2 0)
l scan near (0 0 2)0.4
0.6
0.8
1
1.2
0.04 0.06 0.08 0.1
inte
nsity
(a.u
.)
(2+,0,0)
(0,2+,0)
(0,0,2+)
Domain populationsH : K : L4.65 : 1.35 : 1 Visit one CDW
reflection from each family.
B C D W 1 0 0 m
0
4
8
0 . 9 5 0 . 9 5 3 0 . 9 5 6L
C
1
2
SD
W
inte
nsity
co
unts
s-1
A S D W 1 0 0 m
2 1
*
SDW Domains at 130 KSDW magnetic reflection
CDWcharged reflection
incident beam h=5.8 keV
incident beam h=11.6 keV
Spin-flip transitionTransverse SDW phase Longitudinal SDW phase
TSF=123 Kin bulk Cr
Image SDW reflection as a function of temperature.
Magnetic reflection disappears!
10 m 2
4 3
1
10 m 10 m 10 m
10 m 10 m 10 m
140 K 130 K 125 K 120 K
110 K 119 K 115 K 1
4
Q || (001) and SQ
Q (001) or S||Q
Repeat with charged CDW reflectionSDW Magnetic reflection CDW Charged reflection
T=130 K T=130 K
T=110 K T=110 K
Spin flip transition begins at Q domain edges
Nominally first order transition is broadened by several degrees, even at micron scale.
8.6
8.7
0
5
10
110 120 130 140
T (K)
2
3
4
1
mea
n in
tens
ity
(cou
nts
s-1
)
mag
net
iza
tion
(em
u/cm
3 )
H=2 kG
H=0
T (K)
10 m 2
4 3
1
10 m 10 m 10 m
10 m 10 m 10 m
140 K 130 K 125 K 120 K
110 K 119 K 115 K 1
4
Q || (001) and SQ
Q (001) or S||Q
Learning more about domain walls
CDWnear (0 0 2)
SDWnear (0 0 1)
L
K
H
red Q || [100] green Q || [010] blue Q || [001]
1) So far we’ve looked at Q || [001]. What happens in neighboring Q domains?
2) Two spin polarizations within transverse phase.
transverse phase
S || [010]
transverse phase
S || [001]
longitudinal phase
S || [100]
Magnetic cross sections in a Q || [100] domain
incident beam k
diffracted beam k’
Cross section with S along [100] or [010] than along [001].
Magnetic imaging of three domains
Conclusions: Chromium• Self organized or artificial domains at small scales are
key to macroscopic properties. Imaging is important.• Spin flip transition in Cr begins at domain walls upon
cooling.• Future work in Cr:
– Control of domain walls
– Separation of bulk and interface effects
Conclusion
• Today: Cr antiferromagnetism and PZT ferroelectricity
• Future directions:– Coupling of strain between layers in multilayer
films– Direct measurements of ferroelectric
polarization domain wall velocity– Size effects in ferroelectric materials