- Home
- Users & Science
- Scientific Documentation
- ESRF Highlights
- ESRF Highlights 2008
- X-ray absorption and magnetic scattering
- Au and Fe magnetic moments in disordered AuFe alloys
Au and Fe magnetic moments in disordered AuFe alloys
Recent developments in third generation synchrotron radiation facilities has made possible the element-specific determination of magnetic moments in multi-element materials via the X-ray magnetic circular dichroism (XMCD) technique. The limit of detection is now well below 0.01 µB/atom even for thin films. Moreover, the ability to separate spin and orbital magnetic moment contributions has rendered XMCD as the most powerful tool for the complete magnetic characterisation of materials and for clarifying subjects that had remained controversial in the past.
In this work, we have determined the 5d induced magnetic moment of Au via XMCD, for disordered fcc and bcc AuFe alloys in the concentration range 25 - 97 at. % Fe. The exotic fcc phase is not predicted in the binary alloy phase diagram at room temperature. However, one may produce single phase fcc disordered AuFe alloys via rapid quenching of the molten alloy for concentrations up to 53 at. % Fe. Iron is in the regular low-spin state in the bcc Fe-rich alloys, but in the high-spin state in the ferromagnetic fcc alloys. This fact may further influence the Au induced magnetic moment and it makes the AuFe alloys extremely interesting from the point of view of magnetism; however, they could not be thoroughly studied in the past and controversial results were published, due to the lack of experimental techniques with element specificity and shell selectivity.
Fig. 100: Normalised XANES and XMCD spectra at the L3,2 edges of Au in AuxFe1–x alloys, x as indicated. The spectra reveal a straightforward relation between the 5d induced magnetic moments of Au and the number of Fe nearest neighbours. |
Figure 100 presents the normalised XANES and XMCD spectra recorded at the L3,2 edges of Au in AuxFe1–x disordered alloys. The data were recorded at the beamline ID12 using the fluorescence detection scheme. The intense photon flux of the Apple II undulator with a very high degree of circular polarisation was necessary in order for high quality data to be accumulated for the Au signal. The sizeable XMCD signal reveals that Au has acquired an induced magnetic moment. Knowing the direction of the magnetic field and the helicity of the beam we were able to conclude that Au is polarised parallel to the Fe magnetic moment. Our analysis followed the magneto-optical sum rules. The outcome was the precise determination of the 5d spin and orbital magnetic moment of Au in the disordered AuFe alloys. The results show that the total 5d induced magnetic moment of Au in the fcc alloys scales with the number N of nearest Fe neighbours, from MAu = 0.1 µB/atom, for N = 3, to MAu = 0.2 µB/atom, for N = 6. Iron in these alloys is in a high spin state carrying a total magnetic moment of about 3 µB/atom as our superconducting quantum interference device (SQUID) magnetometry measurements have revealed. The maximum value of 5d induced magnetic moment of Au is 0.33 µB/atom and is exhibited when Au is placed as an impurity in a bcc Fe environment, i.e. when all its first neighbours are Fe atoms. In the bcc environment the total magnetic moment of Fe is 2.2 µB/atom. Thus, one may conclude that the total 5d induced Au moment seem to follow a simple scheme of the type MAu (N/Ntotal)*MFe regardless of the crystallographic structure. Our experimental results for the Au and Fe magnetic moments compare well with first principle calculations, as Figure 101 reveals.
Fig. 101: Au and Fe magnetic moments by experiment and theory as a function of the Fe at. % concentration in disordered AuFe alloys. The lines are guides to the eye. |
In conclusion, the relationship between the induced magnetism of Au and the Fe magnetic moments for disordered bcc and fcc AuFe alloys have been found and a simple formula for the Au induced magnetic moments was deduced. The experimental results, recorded via a combination of XMCD and SQUID resolve previous controversial experimental data and are in good agreement with our first principle calculations [1].
Authors
F. Wilhelm (a), P. Poulopoulos (b), V. Kapaklis (b), J.-P. Kappler (c), G. Schmerber (c), A. Derory (c), N. Jaouen (a), A. Rogalev (a), A.N. Yaresko (d), C. Politis (b,e).
(a) ESRF
(b) University of Patras (Greece)
(c) IPCMS Strasbourg (France)
(d) MPI Dresden (Germany)
(e) INT Karlsruhe (Germany)
References
[1] F. Wilhelm, P. Poulopoulos, V. Kapaklis, J.-P. Kappler, N. Jaouen, A. Rogalev, A.N. Yaresko, and C. Politis, Phys. Rev. B. 77, 224414 (2008).