March 14, 2005
Time:
02:40pm
ps6120.tex
Physica Scripta. Vol. T115, 632–634, 2005
Mn K-Edge Magnetic EXAFS for Ni-Mn Alloys
T. Miyanaga*1 , T. Okazaki1 , R. Maruko1 , S. Nagamatsu2 , T. Fujikawa2 and Y. Sakisaka1
1 Department
2 Graduate
of Materials Science and Technology, Faculty of Science and Technology, Hirosaki University, Hirosaki, Aomori 036-8561, Japan
School of Science, Chiba University, Yayoi-cho 1-33, Inage, Chiba 263-8522, Japan
Received June 26, 2003; accepted in revised form January 19, 2004
pacs number: 03.30+p
Abstract
2. Experimental and Theory
We have measured Mn K-edge magnetic EXAFS for Ni0.76 Mn0.24 alloy. The 2nd
peaks in the Fourier transform attributed to Mn-Mn pairs in the ordered phase
are enhanced in comparison with those in non-magnetic EXAFS. This result
suggests that some Ni atoms in the 2nd shell are replaced by Mn atoms, which
have large magnetic moment in comparison with Ni atoms, due to heat-treatment
induced atomic ordering. Semi-relativistic theoretical calculations well explain
the observed magnetic EXAFS. We estimate the ratio of the magnetic moments
of Ni to Mn atom in Ni-Mn alloy, within a single scattering limit.
A 21 m foil sample of Ni0.76 Mn0.24 was prepared by polishing
carefully and annealing for 100 hours at 693 K under Ar
atmosphere in order to make certain an ordered state. The
reason why we study this concentration is that the ordered
phase of Ni0.76 Mn0.24 alloy is easily obtained by the shorter
heat treatment time than that of Ni0.75 Mn0.25 alloy, in which
the residual oxygen can oxidize the sample. The saturation
magnetization of this sample is 56.4 emu/g and the degree of
order S is deduced to about 0.74. Magnetic EXAFS of Mn k-edge
(6.4–7.2 keV) for Ni0.76 Mn0.24 foil was measured at BL28B of
Photon Factory in KEK (2.5 GeV) with circularly polarized Xrays that were generated by an ellipsoid multi-pole wiggler and
monochromatized using a Si(111) double-crystal monochromator
[13, 14]. The absorption spectra were measured in transmission
mode at 30 K. Instead of switching helicity of the incident photons,
the applied magnetic field, about 0.6 T which is still enough to
saturate the magnetization of the sample, was reserved every 4
seconds to obtain MCD effect.
We perform semi-relativistic theoretical calculation of the
magnetic EXAFS for binary alloy as the ordered model of Ni3 Mn
including up to 4th shell atoms (∼5A). In the case of the binary
system, it is important to determine the electronic state for each
atom. To calculate the phase shift for magnetic EXAFS, we refer
to the electronic structures obtained from FLAPW method [15].
The details of the present theoretical approach are discussed
elsewhere [15, 16]. In this calculation, multiple scattering effect,
the amplitude reduction factor by Debye-Waller factor and manybody effects are not taken into account.
1. Introduction
The correlation between the structure and magnetism of 3d
transition metal alloys has long been an important topic in
condensed matter physics. Among the 3d transition metal
alloys, the Mn based alloys have recently attracted special
attention because of their potential applications in magnetic
recording technology. Ni3 Mn alloy has been an interesting
example of such alloys and can undergo a structural phase
transition where the magnetic behavior is sensitive to the atomic
arrangement [1]. Cable and Child studied the magnetic moment
distributions in disordered Nix Mn1−x (x = 0.80–0.95) alloys by
polarized-neutron diffuse-scattering [2]. They showed that the
magnetic moment of isolated Mn atom in Ni0.95 Mn0.05 alloy is
3.50 B , which decreases with Mn concentration. Two possible
explanations are proposed: (1) The ferromagnetic component
of Mn moment decreases by becoming randomly oriented or
antiferromagneically coupled. In this case the local magnetic
moment of Mn atom is conserved even in the disorder state. (2)
Averaging over up-spin and down-spin magnetic configurations
on Mn atoms. In this case, the net local moment of Mn atom
decreases. They concluded that the magnetic distribution is
complicated and not specifically defined by the neutron-crosssection measurement, since the neutron scattering technique is
based on long-range order.
On the other hand, magnetic XAFS is a powerful tool for
the investigation of local magnetic structures with short-range
order in alloys, and uses circular polarized X-ray beam from a
synchrotron radiation source [3–7]. The interpretation of magnetic
EXAFS is clear and direct for the study of local magnetic
structures, although the signal is much smaller than that of
magnetic XANES. Recently, some theoretical approaches have
been developed for magnetic EXAFS analyses [8–10]. Recent
measurements of magnetic XAFS have been reported for alloys
[11] and theoretical analyses have been applied to FeCo and FePt
alloys [9] and disordered GdCo alloys [12].
In this paper, we experimentally and theoretically study
magnetic EXAFS for ordered Ni0.76 Mn0.24 alloys prepared by
heat treatments.
* e-mail: [email protected]
Physica Scripta T115
3. Results and Discussion
First we discuss the result for non-magnetic EXAFS. Figure 1
(a) shows the non-magnetic EXAFS oscillation function, k(k),
for experimental Mn K-edge for Ni0.76 Mn0.24 foil sample (solid
line) and theoretical calculation by a model of ordered fcc Ni3 Mn
(dashed line). For experimental spectra the magnetic EXAFS
signal is normalized by the edge jump of the non-magnetic X-ray
absorption spectrum to be unity. Similar experimental spectrum
was observed in a previous paper [15]. However, the present result
is much better than the previous one. As the figure shows, the main
oscillating parts for experimental and theoretical calculation are
coinciding. On the other hand, a more fine oscillation is found
in the experimental spectrum. Figure 1 (b) shows a comparison
between experimental and theoretical Fourier transforms (FT) of
the non-magnetic EXAFS k(k) shown in Fig. 1 (a). For the
data processing we use the XANADU code [17]. The k-range
of the Fourier transform is 2.0 ∼ 10.8 Å−1 . Solid line (dashed
line) shows the result for the experimental (theoretical) EXAFS.
Vertical arrows in Fig. 1 (b) denote the crystallographic positions
C Physica Scripta 2005
March 14, 2005
Time:
02:40pm
ps6120.tex
Mn K-Edge Magnetic EXAFS for Ni-Mn Alloys
633
Fig. 1. (a) Non-magnetic EXAFS k(k) spectra for the Mn K-edge of experimental
data for Ni0.76 Mn0.24 alloy (solid line) and theoretical calculation as a model of
ordered Ni3 Mn (dashed line). (b) Fourier transforms of non-magnetic EXAFS
k(k) shown in (a).
Fig. 2. (a) Magnetic EXAFS k(k) spectra for the Mn K-edge of experimental data
for Ni0.76 Mn0.24 alloy (solid line) and theoretical calculation as a model of ordered
Ni3 Mn (dashed line). (b) Fourier transforms of the magnetic EXAFS k(k) shown
in (a).
of nearest neighbor atoms in Ni-Mn fcc lattice from the central Mn
atom. First nearest neighbor atoms are located at 2.54 Å, second
at 3.59 Å, third at 4.5 Å and fourth at 5.08 Å. In Fig. 1 (b), the
phase shift correction is not considered, so the peak position is
not equal to these crystallographic data. The amplitude of the
calculated EXAFS is larger than that for the experimental one,
since we neglect thermal vibration, electron mean-free and manybody effects in the present analyses (the factor is about 0.7). The
FT of the first and second peak for the theoretical calculation
almost overlapp the experimental one. However, deviation is
found between theoretical calculation and experimental spectrum
for the third and fourth peaks (experimental peaks is enhanced in
comparison to theoretical). This phenomenon is considered as a
multiple scattering effect; especially a focusing effect is expected
in the fourth shell atoms. In the present theoretical calculation,
multiple scattering is not taken into account. Since our present
calculation cannot reproduce such an effect, we discuss mainly
the results for the first and second shells.
Next we discuss the result for magnetic EXAFS. Figure 2
(a) shows a comparison between experimental magnetic EXAFS
k(k) spectrum and theoretical one. The main oscillating part
almost coincides with that of non-magnetic EXAFS. On the other
hand, the fine oscillating structure is more pronounced than in the
case of non-magnetic EXAFS. Figure 2 (b) presents a comparison
between experimental and theoretical calculation for the FT of
magnetic EXAFS shown in Fig. 2 (a). Deviation is found in the
third and fourth peaks between experimental and theoretical. This
is considered as multiple scattering or focusing effect as for nonmagnetic EXAFS. For the first and second peaks, we can find
clear enhancement of the second peak in magnetic EXAFS in
comparison to non-magnetic one (Fig. 1 (b)). This result can
be interpreted as follows. After appropriate heat treatment, the
Ni-Mn alloys are ordered. In the ordered phase, Mn atoms with
large magnetic moment (∼3.5 B from neutron diffraction [2]) are
exchanged to Ni atoms that are located in the second shell around
an X-ray absorbing Mn atom. Therefore the second nearest Mn
atoms strongly contribute to the magnetic EXAFS.
Another important effect we should consider is a strong
multiple scattering effect, which was reported in the higher shell
of magnetic EXAFS by Ahlers & Shutz [6] and Wende et al.
[7]. Wende et al. have shown the quite interesting result that the
multiple scattering effect is enhanced in magnetic EXAFS and
the FT peak for the second shell overlaps the triangle multiple
scattering path for bcc structure [7]. In their case, if the first nearest
distance is 2.49 Å (as a typical model for Fe), the distance for the
third shell is 4.06 Å and the corresponding triangle path is 3.92 Å,
in which it is difficult to distinguish them with the FT method [7].
On the other hand, in the present fcc Ni3 Mn case, the distance for
the second shell is 3.59 Å and the corresponding triangle path
is 3.81 Å (the difference is 0.22 Å). We carefully checked the
simulation of FT (k-range is 2–10.8 Å−1 ) for various situations
and certify that the peaks of 3.59 Å and 3.81 Å do not overlapped
and can be distinguished from each other. The peak of the triangle
path (3.81 Å) should be included at longer distance than the
second peak. However, since the second peak position is slightly
shifted in Fig. 2 (b), a small portion of the triangle path effect
may be mixed into the second shell contribution. Unfortunately
our present calculation cannot take the multiple scattering
process into account. So we have to consider a single scattering
limit.
ptIn the theoretical calculation for magnetic EXAFS, we can
handle the electronic configuration and also magnetic moments of
each Ni and Mn atom. For the ordered Ni3 Mn model, the 1st and
C Physica Scripta 2005
Physica Scripta T115
March 14, 2005
634
Time:
02:40pm
ps6120.tex
T. Miyanaga et al.
4. Conclusion
We measure the Mn K-edge magnetic EXAFS for Ni0.76 Mn0.24 ,
where the 2nd peaks in the Fourier transform are enhanced
in comparison with those in non-magnetic EXAFS. This result
suggests that some Ni atoms in the 2nd shell are replaced by Mn
atoms with larger magnetic moment than Ni atom, due to heattreatment induced atomic ordering. Semi-relativistic theoretical
calculation well explains the observed magnetic EXAFS. We
propose a simple method to evaluate the ratio of the magnetic
moments for each atom in Ni-Mn alloy from magnetic EXAFS,
which is applicable to disordered systems, even within a single
scattering approximation.
Acknowledgements
Fig. 3. The ratio of the calculated Fourier peak intensity Mn/Ni as a function of
the ratio of magnetic moments Mn/Ni atom used for the theoretical calculation.
2nd peak in the Fourier transform are attributed to the Ni and Mn
atom. The ratio of the peak intensities of the 1st to 2nd peak in the
FT of the magnetic EXAFS depends on the magnetic moment of
Ni and Mn atoms. Figure 3 shows calculated Fourier peak intensity
Mn (2nd peak)/Ni (1st peak) as a function of the ratio of magnetic
moments of Mn and Ni atom used for the theoretical calculation.
From the FT in Fig. 2 (b), we obtain the ratio of 2nd/1st
peaks 0.41. Using the theoretical curve shown in Fig. 3, we
estimated the ratio of the magnetic moment Mn/Ni ∼ 4.0 (as
shown in Fig. 3). This value is much smaller than that obtained
from neutron study (Mn/Ni ∼ 10) [18], however it is fairly close
to the result derived from a recent band calculation using FLAPW
method (Mn/Ni ∼ 4.8) [15]. As a reason for the small value of
the ratio of magnetic moments Mn/Ni, the ordered Ni3 Mn model
is not adequate for the present Ni0.76 Mn0.24 sample (S ∼ 0.74).
The electronic states used in the theoretical calculation are still
inaccurate or the multiple scattering effect cannot be neglected
in the contribution of second shell peak. While there are still
problems we should solve in this simple method to derive the
ratio of the magnetic moments of each atom, this method has
a potential for disordered magnetic systems in which neutron
diffraction cannot be applied. To improved this method, more
accurate analyses of magnetic EXAFS, e.g. multiple scattering
process [7], thermal vibration and spin disorder [5] and many
body effect, are necessary to take into account.
Physica Scripta T115
The authors thank Dr. T. Iwazumi, Dr.Y. Watanabe and Dr. T. Nakamura for kindly
supporting the measurement of magnetic XAFS. One of the authors (T.M.) thanks
Prof. Yu. A. Babanov and Prof. A. Z. Menshikov for a fruitful discussion and to
Prof. E. D. Crozier for encouragement of this work. This work was performed
under the approval of proposal No. 98G236 and No. 2001G232 of the Photon
Factory (PF) at KEK.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Okazaki, T., Jpn. J. Appl. Phys. 34, 1537 (1995).
Cable, J. W. and Child, H. R., Phys. Rev. B 10, 4607 (1974).
Schutz, G., et al., Phys. Rev. Lett. 58, 737 (1987).
Lovesey, S. W. and Collins, S. P., “X-ray Scattering and Absorption by
Magnetic Materials,” (Oxford Science Publisher, 1996), pp 151.
Wende, H., et al., J. Appl. Phys. 83, 7028 (1998).
Ahlers, D. and Shutz, G., Phys. Rev. B 57, 3466 (1998).
Wend, H., et al., J. Synchrotron Rad. 6, 696 (1999).
Ankdinov, A. L. and Rehr, J. J., Phys. Rev. B 52, 10214 (1995).
Popescu, V., Ebert, H. and Jenkins, C., J. Synchrotron Rad. 6, 711 (1999).
Nagamatsu, S. and Fujikawa, T., J. Synchrotron Rad. 8, 416 (2001).
Pizzini, S., et al., Phys. Rev. B 50, 3779 (1994).
Yamamoto, I., Nagamatsu, S., Nakamura, T., Fujikawa, T. and Nagao, S.,
J. Elect. Spect. 125, 89 (2002).
Iwazumi, T., Koyama, A. and Sakurai, Y., Rev. Sci. Instrum. 66, 1691
(1995).
Kobayashi, K., Maruyama, H., Iwazumi, T. and Yamazaki, H., Solid State
Commun. 97, 491 (1997).
Miyanaga, T., et al., J. Synchrotron Rad. 10, 113 (2003).
Fujikawa, T. and Nagamatsu, S., Jpn. J. Appl. Phys. 41, 2005 (2001).
Fujikawa, T. and Nagamatsu, S., J. Electr. Spectr. 129, 55 (2003).
Sakane, H., et al., Jpn. J. Appl. Phys. 32, 4641 (1993).
Paoletti, A. and Ricci, F. P., J. Appl. Phys. 34, 1571 (1963).
C Physica Scripta 2005