Magnetic structure of GdCu6

A. Devishvili1,2, M. Rotter2,3, M. Doerr4, B. Beuneu5, G. Behr6.1 Insitut Laue Langevin, Grenoble, France. 2 Institut fur Physikalische Chemie, University of Vienna, Austria 3 University of Oxford, Dept. of Physics, Clarendon Lab., Oxford, UK

4 Institut fur Festkorperphysik, TU Dresden, Germany 5 Labaratoire Leon Brillouin, CEA-CNRS, Saclay, France 6 Institut fur Festkorper und Werkstoffforschung, Dresden, Germany

References:[1] J. Jensen, A. R. Mackintosh, Rare Earth Magnetism, Clarendon Press Oxford (1991)[2] M. Rotter, M. Loewenhaupt, S. Kramp, T. Reif, N. M. Pyka, W. Schmidt, R. v. d. Kamp, Europ. Phys. J. B 14 (2000) 29.[3] M. Loewenhaupt, M. Rotter, S. Kramp, Physica B 276-278 (2000) 602.[4] M. Rotter, M. Loewenhaupt, M. Doerr, A. Lindbaum, H. Sassik, K. Ziebeck and B. Beuneu, Phys. Rev. B 68 (2003) 144418.[5] S. Takayanagi, Y. Onuki, K. Ina, T. Komatsubara, N. Wada, T. Watanabe, T. Sakakibara, T. Goto, Journal of Physical Society of Japan 58 (1989) 1031Acnowledgements: We acknowledge support by the Austrian Science Foundation (FWF) Project No. P16778-N02 and the French-Austria Bilateral Scientific Techni-cal Exchange Program “Amadee” project 17/2003. We are grateful to P. Ambroise for the helpful assistance at LLB, Saclay. Part of this work was performed within the DFG-funded program of the Sonderforschungsbereich 463.

- The rare earth systems are of particular interest due to the well localized 4f shell. - The Gd3+ 4f shell is half filled with spherically symmetric charge density and has no orbital moment (L=0). - There are no first order crystal field effects present in this type of systems. - Since the crystal field is claimed to be responsible for magnetic anisotropy [1-3] in rear earth compounds the investigation of anisotropic GdCu

6 system may reveal the complexity of exchange interaction.

- Most of the experimentally observed magnetic features have been reproduced by calculations. - GdCu

6 can be considered as well described within standard model of rare earth magnetism.

- The magnetic structure observed by neutrons is in agreement with expectations from the dipolar model [4]. - Comparing the results of our model with the published single crystal bulk magnetisation and susceptibility data we infer that some source of anisotropy other than the dipolar interaction must be present in GdCu

6.

ab

c

Motivation Conclusion

Nuclear and magnetic structure investigation

Method: hot neutron powder diffractionInstrument: 7C2 (LLB, Saclay).Wavelength: 0.57 Å (Ge (311) mono)�Patterns collected: 2 K and 30 K Counting time: 9 hours.Background I

bkg estimated with:

Ibkg

= Icd

+ k(Iempty

- Icd

) where k = 0.3807, I

cd is fully absorbing (cadmium

foil) sample and Iempty

is an empty sample holder.

30 KSpacegroup: P n m aLattice parameters: a=8.023 Å, b=5.027 Å, c=10.075 Å.Atomic positions:Atom X Y Z B

iso Occup.

Gd 0.258 0.250 0.565 0.126 4.0 Cu1 0.047 0.474 0.313 0.229 8.0 Cu2 0.054 0.250 0.073 0.327 4.0 Cu3 0.144 0.250 0.849 0.327 4.0 Cu4 0.286 0.250 0.273 0.327 4.0 Cu5 0.381 0.250 0.004 0.327 4.0

2 KThe magnetic propagation vector: (2/9 0 0) � (0.22 0 0)Magnetic structure: helical structure with moments in bc - plane or a collinear struc-ture with equal moments in bc - plane ([021] direction e. g. � 45˚ from b and c). The helical structure is presented in the fit.

Red open circles represent experimental data measured above the ordering tempera-ture. The line represents the Rietveld refinement of the nuclear structure.

Dots represent the magnitude of the exchange constants J(R) where R is the distance to equivalent neighbors.

Data (dashed lines) was taken from [5]. Calcu-lated magnetic susceptibility is presented as solid lines.

Magnetic field pulse applied along [001] direction. Static magnetic field up to 14 T was used to obtain magnetization for main crystallographic directions. Data taken from [5] (dashed lines) is confronted by results of numerical simulation (bold lines).

The data (black open circles) was obtained by subtracting the specific heat of non mag-netic analog (e. g. C

m = C

GdCu6 - C

LaCu6). Data taken from [5]. Solid blue line represents the cal-

culated magnetic contribution to specific heat by mean field approximation described in simulation section.

GdCu6 hot neutron diffraction data obtained at 7C2, Saclay. Black open circles represent experimental data measured below the

ordering temperature. Lines represent fit of magnetic and nuclear structures. Black dots below represent the difference be-tween the high and low temperature diffraction pattern.

Magnetic structure ith propaga-tion vector (2/9 0 0) at T=2 K fitted from neutron powder dif-fraction.

Calculated magnetic struc-ture with propagation vector (1/6 0 0) at T=2 K using the mean field model.

Mean Field Model

Hamiltonian:

RKKY Classical Dipolar Zeeman term

Classical dipolar:

RKKY:

RKKY parametersfitted from bulk/neutron data:A = -100meVk =(0.85, 1.8125, 1.4)Å-1

Calculation method: Monte-Carlo self consistancy energy minimization. McPhase program package*

Propagation vector: ( 1/6 0 0 )

Magnetic structure: helix,magnetic moments in bc-plane.

* - www.mcphas.de

Magnetic susceptibility

RKKY exchange

Magnetization

Magnetic contribution to the specific heat

Neutron Scattering

Mean field model