Resonant x-ray diffraction studies of Ca3Ru2O7

Resonant enhancement of magnetic scattering (electric dipole transitions) ... 110 K. 260 K. Novel phase. E = 2.9685 keV (L. 2. ) • Magnetic transition at T. N.
1MB taille 1 téléchargements 554 vues
Resonant scattering at the Ru L-edges Ph. Leininger, B. Bohnenbuck, I. Zegkinoglou, J. Strempfer, C. S. Nelson and B. Keimer

Ruthanate materials Outline Rich phase diagram: x=0: TMI (357 K) and AFM below 110 K (small FM component) x=2 metallic and superconducting below 1.5 K (unconv. spin triplet) - T and doping dependent Structural distortions - Strong e- correlations - Spin-orbit coupling O. Friedt et al. Phys. Rev. B 63, 174432 (2001)

Interplay between Orbital, Magnetic and Structural degree of freedom? • Ca2RuO4 • RuSr2GdCu2O8 (Ru1212) • Ca3Ru2O7

Ca2RuO4 (x=0) - Single-layered - a ~ b ~ c/2 - K2NiF4-type structure strongly distorted (s. g. : Pbca) Ru-Oapical bond rotated around the long c Octahedron tilt around an axis lying in the RuO2 plane

c

b

a

- 44Ru4+: 4d4 configuration

- Antiferromagnetic ordering: 2 magnetic modes: ‘A-centered’ mode [(100) (La2CuO4 type),(011)]: TN = 110K ‘B-centered’ mode [(010) (La2NiO4 type)]: TN = 150K - S = 1 (low spin state) - m along b (m = 1.3μB < 2μB )

Braden et al., PRB 58, 847 (1998)

Resonant X-Ray Diffraction (RXD) How to investigate the structural and magnetic properties? X- ray scattering incoming x-ray coupled to : - electronic charge - magnetic moment (m) orbital magnetization density ≠ spin density: Not the case for neutron incoming x-ray ~ absorption edge Æ Resonant enhancement of magnetic scattering (electric dipole transitions) Ru L edges: - Intermediate energy range - Probe directly the d bands

RXD : Ru L2 edge APS 4ID-D (100)

L2 edge L2’ edge

(011)

Temperature Dependence at (100) (100) L2 edge

Novel phase

E = 2.9685 keV (L2) • Magnetic transition at TN=110 K • Second phase transition at ~260 K

L2’ edge

E = 2.9725 keV (2nd resonant pic) • No anomalies up to 320 K -> excitations to unoccupied states (eg) I. Zegkinoglou et al.,PRL 95, 136401 (2005)

110 K

260 K

Temperature Dependence at (011)

• Identical temperature dependences at (011) and (100) • New phase is characterized by the same

propagation vector as the low-temperature antiferromagnetic state Origin of this new phase?

Origin of the new phase

Polarization dependence at (100) only σ-π’ contribution Æ no charge scattering

260 K Muon spin rotation: Octahedral tilt angles: no anomaly at 260 K Æ no tilt order origin

260 K

Braden et al., PRB 58, 847 (1998)

no magnetic moment above TN Æ no magnetic origin

Orbital Order is responsible for phase transition at 260 K at (100) and (011)

Orbital Order: Theoretical Predictions

•Hotta et al. [3-orbital Hubbard model, el-ph interactions] -> Antiferro-orbital-ordering (AFO) : explain XAS experiments Doubling periodicity in a b (½ ½ 0): not observed experimentally • Anisimov et al., Jung et al. [first principle cal.: LDA+U, σ and XAS] : -> 4dxy Ferro-orbital (FO) ordering • Kubota et al. (RXD: interference technique) -> Ferro-O ordering • Lee et al.: (optical spectro. and multi-orbital Hubbard model) -> Phase coexistence of AF-O and F-O ordered states

Still an contradiction between theory and experiment

zx yz xy Ru: 4d4 (degenerate)

T and azimuthal dependence at (110) No charge and magn. peak in Pbca

Ψ 2θ

No anomalies up to Troom

σ-π‘ and σ - σ ≠ 0

- T and polarization dependence ≠ from (100 & 011) - Azimuthal dependence well reproduced by tilt order model (symmetry consideration)

Resonant scattering at (110) due to tilt order

RuSr2GdCu2O8 (Ru1212) • Tetragonal structure (P4/mmm)

Cu

a=b=3.836 Å, c=11.563 Å • Octahedra rotated about c and

to c

• Alternating - superconducting Cu-O layers (TC≅ 35 K) • Ru ions order AF below TN≅ 136 K

O Ru O Cu

• Gd: AF, TN = 2.5 K

||m|| = 1.18 µB predicted along c Magnetic ordering ?

Temperature Lynn et al., PRB 61, R14964 (2000)

Gd

Ru

Energy dependence at (½ ½ ½)

L2 edge

L2’ edge

calculated absorption coef. from fluorescence

Temperature dependence at (½ ½ ½)

- T > 50 K x-rays and neutrons: identical TN~ 138 K

Below 50 K (superconducting state): Integrated Int. x-rays ≠ neutron Magnetic – superconducting interplay? Æ Difficult to obtain further info.

Azimuthal dependence at (½ ½ ½)

Azimuthal dependence resonant dipole magentic scattering simulation

I σ −σ ' ( 1 1 1 ) ∝ F σ −σ ' ( 1 2 2 2

I

σ −π '

( 12 12 12 )

∝F

σ −π '

2 1 1 2 2 2

)

2

( 12 12 12 )

=0

= A cos 2 Ψ − B cos Ψ + C

Hill and McMorrow, Acta Cryst. A52, 236 (1996)

Direction of magnetic moment RXD : Intensity max. when m in the diffraction plane Intensity max. when m is ~55º off from c axis

≠ powder neutron diffraction (m||c)

(111)

μ

• The direction of the magnetic moment is ~55º off the c-direction

α μp 77º

53º c cp

• From fitting of azimuthal dependence: angle between m and (111)= α = 45.6º • Calculated direction of m : (102)

Ca3Ru2O7 • Orthorhombic structure (rotation and tilting distortion) • Double layer • Paramagnetic metal at high T • AFM ordering temperature below TN = 56K • Structural phase transition at TS = 48K - Large changes of lattice constants - Increase of resistivity below TS Æ Metal insulator-like transition

Yoshida et al., PRB 72, 054412 (2005)

Æ Quasi 2D metal below 30K

- Ordering propogation vector?

Energy Dependence Bessy: KMC-1 • Large resonance enhanchement at reflections (001) and (110) • Calculated structure factor: A-type AFM structure: Ferromagnetic bilayers Antiferromagnetic coupling along c-axis

Azimuthal Dependence Simulation

I

π −σ '

I

π −π '

∝F

π −σ ' 2

= m1cosθ + m3 sinθ

∝F

π −π ' 2

= − m2 sin2θ

2

2

⎛ m1 ⎞ ⎛ cos ψ sin α ⎞ ⎜ ⎟ ⎜ ⎟ m = ⎜ m2 ⎟ = m⎜ sin ψ sin α ⎟ ⎜m ⎟ ⎜ − cos α ⎟ ⎝ 3⎠ ⎝ ⎠ Hill and McMorrow, Acta Cryst. A52, 236 (1996)

Æ Reorientation of magnetic moment at TS - T < TS: m || b - T > TS: m || a

Temperature Dependence

• Similar T-dependence at (001) and (110) • Intensity constant below TS • Drastic intensity loss at TS due to reorientation of magnetic moment • Decreasing magnetic moment in AFM-metallic phase

TS TN

Orbital Order? ƒ Raman scattering: - Large structural changes at TS and orbital-ordering ƒ RXD: No Anti-Ferro-Orbital (AFO) Order at (100), (010), (1/2 0 0),… Reason… Magnetic and Orbital Order at the same reciprocal space positions ƒ Outlook: Search for FO-Ordering: Resonant x-ray interference technique

I(90°+ΔФ) - I(90°-ΔФ)∞ 2Re[Fσσ´F*σπ]×sin2θAsin2ΔФ]

Ferro-orbital ordering => Interference term Next beamtime (few weeks)

Takashi Kiyama et al., J. Phys. Soc. Jpn., Vol. 72, 785 (2003)

Conclusions

ƒ New orbitally ordered paramagnetic phase discovered T∞ = 260 K (x=0) ƒ Tilt ordering above 260 K (x = 0) ƒ m along (102) in Ru1212 ƒ Reorientation of m at TS, No AFO-order (Ca3+) Æ Orbital order weakly coupled to crystal lattice (no structural phase transition at TOO) Æ Magnetic structure sensitive to small lattice distortions caused by doping

Collaborators C. T. Lin

MPI, Stuttgart

J.C. Lang, G. Srajer

APS, Argonne, USA

E. Schierle

HMI at BESSY, Berlin

C. Schüßler-Langeheine

Universität Köln

N. Kikugawa

National Institute for Material Science, Tsukuba, Japan

Y. Yoshida, S. I. Ikeda

National Institute of Advanced Industrial Science and Technoloy, Tsukuba, Japan

H. Fukazawa, S. Nakatsuji, Y. Maeno

Department of Physics, Kyoto University, Kyoto, Japan