high pressure\(\def\hfill{\hskip 5em}\def\hfil{\hskip 3em}\def\eqno#1{\hfil {#1}}\)

Journal logoJOURNAL OF
SYNCHROTRON
RADIATION
ISSN: 1600-5775

Development of an energy-domain 57Fe-Mössbauer spectrometer using synchrotron radiation and its application to ultrahigh-pressure studies with a diamond anvil cell

aJapan Atomic Energy Agency, 1-1-1 Kouto, Sayo-cho, Sayo-gun, Hyogo 679-5148, Japan, bCREST, Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan, cJapan Synchrotron Radiation Research Institute, 1-1-1 Kouto, Sayo-cho, Sayo-gun, Hyogo 679-5198, Japan, dNational Institute of Advanced Industrial Science and Technology, Tsukuba Central 2, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan, and eResearch Reactor Institute, Kyoto University, Kumatori, Sennan-gun, Osaka 590-0494, Japan
*Correspondence e-mail: taka@spring8.or.jp

(Received 1 March 2009; accepted 23 August 2009; online 12 September 2009)

An energy-domain 57Fe-Mössbauer spectrometer using synchrotron radiation (SR) with a diamond anvil cell (DAC) has been developed for ultrahigh-pressure measurements. The main optical system consists of a single-line pure nuclear Bragg reflection from an oscillating 57FeBO3 single crystal near the Néel temperature and an X-ray focusing device. The developed spectrometer can filter the Doppler-shifted single-line 57Fe-Mössbauer radiation with a narrow bandwidth of neV order from a broadband SR source. The focused incident X-rays make it easy to measure a small specimen in the DAC. The present paper introduces the design and performance of the SR 57Fe-Mössbauer spectrometer and its demonstrative applications including the newly discovered result of a pressure-induced magnetic phase transition of polycrystalline 57Fe3BO6 and an unknown high-pressure phase of Gd57Fe2 alloy placed in a DAC under high pressures up to 302 GPa. The achievement of Mössbauer spectroscopy in the multimegabar range is of particular interest to researchers studying the nature of the Earth's core.

1. Introduction

The diamond anvil cell (DAC) is well recognized as a very important device for high-pressure Mössbauer spectroscopy to investigate pressure-induced effects on the magnetic properties, charge states and the local environment of a Mössbauer isotope in materials (Jayaraman, 1983[Jayaraman, A. (1983). Rev. Mod. Phys. 55, 65-108.]). To perform high-pressure Mössbauer spectroscopy using a DAC, we generally use a radioisotope (RI) source or synchrotron radiation (SR) source as a photon source. The SR source, in particular, has many advantages: one advantage is a small beam size, which is important for studies under ultrahigh pressure using a DAC with a small gasket-hole as a sample chamber. Furthermore, the focusing device for SR X-rays allows us to obtain easily a small beam size below 100 µm diameter. In contrast, the size of the point RI source available for conventional Mössbauer spectroscopy is 0.5 mm × 0.5 mm, much larger than the beam size of the SR X-rays.

In high-pressure SR Mössbauer spectroscopy, so-called nuclear forward scattering (NFS) is normally used to obtain the hyperfine splitting (HFS) parameters, including the hyperfine field (Hint), the isomer shift (IS) and the quadrupole splitting (QS). In this method, time-domain interference of the delayed nuclear resonant scattering of SR pulses is measured, and the HFS parameters are determined by numerical analysis of the quantum beats in the obtained NFS time spectrum. The frequencies of the quantum beats reflect the energy splitting of the nuclear states owing to the hyperfine interactions. Experimentally, the first high-pressure NFS spectra were measured by Nasu and co-workers (Nasu, 1994[Nasu, S. (1994). Hyperfine Interact. 90, 59-75.]) and the measurement technique has been developed rapidly with the advent of third-generation SR sources (Lübbers et al., 1999[Lübbers, R., Wortmann, G. & Grünsteudel, H. (1999). Hyperfine Interact. 123/124, 529-559.]; Sturhahn & Jackson, 2007[Sturhahn, W. & Jackson, J. M. (2007). Geol. Soc. Am. Spec. Pap. 421, 157-174.]). Time-domain measurement, however, requires special bunch modes of SR, i.e. a rather long period (T > 100 ns for 57Fe nuclei) between electron bunches in a storage ring; such bunch operation modes are usually limited at SPring-8, the third-generation synchrotron radiation facility in Japan. Currently, the available machine time for high-pressure SR 57Fe-Mössbauer spectroscopy with NFS might be a few weeks per year at SPring-8. Moreover, NFS measurements with short bunch period may make it difficult to analyze the spectrum consisting of complex and/or multi-site hyperfine interactions particularly for users of conventional Mössbauer spectroscopy. Such situations contrast with conventional Mössbauer spectroscopy, which has a long history concerning the data analysis of the energy domain absorption spectrum measured with RI sources.

We have recently developed a new high-pressure SR 57Fe-Mössbauer spectrometer using a pure nuclear Bragg reflection from an oscillating 57FeBO3 single crystal near the Néel temperature, which can filter 14.4 keV single-line 57Fe-Mössbauer radiation from broadband SR (Chumakov et al., 1990[Chumakov, A. I., Zelepukhin, M. V., Smirnov, G. V., Van Bürck, U., Rüffer, R., Hollatz, R., Rüter, H. D. & Gerdau, E. (1990). Phys. Rev. B, 41, 9545-9547.]; Smirnov et al., 1997[Smirnov, G. V., van Bürck, U., Chumakov, A. I., Baron, A. Q. R. & Rüffer, R. (1997). Phys. Rev. B, 55, 5811-5815.]; Smirnov, 2000[Smirnov, G. V. (2000). Hyperfine Interact. 125, 91-112.]; Mitsui et al., 2007a[Mitsui, T., Seto, M., Kikuta, S., Hirao, N., Ohishi, Y., Takei, H., Kobayashi, Y., Kitao, S., Higashitaniguchi, S. & Masuda, R. (2007a). Jpn. J. Appl. Phys. 46, 821-825.],b[Mitsui, T., Seto, M., Masuda, R., Kiriyama, K. & Kobayashi, Y. (2007b). Jpn. J. Appl. Phys. 46, L703-L705.],c[Mitsui, T., Seto, M. & Masuda, R. (2007c). Jpn. J. Appl. Phys. 46, L930-L932.]). As a result, it allows researchers to perform SR-based energy-domain Mössbauer spectroscopy under any bunch-mode operation of SR. Furthermore, the high degree of polarization, full recoillessness, extremely small divergence, small beam size and other excellent beam properties of SR offer the new possibility of various advanced measurement techniques fusing conventional Mössbauer spectroscopy with modern high-precision X-ray optics (Pankhurst et al., 2001[Pankhurst, Q. A., Cohen, N. S., Barquín, L. F., Gibbs, M. R. J. & Smirnov, G. V. (2001). J. Non-Cryst. Solids, 287, 81-87.]; Mitsui et al., 2007d[Mitsui, T., Seto, M., Hirao, N., Ohishi, Y., Kobayashi, Y., Higashitaniguchi, S. & Masuda, R. (2007d). Jpn. J. Appl. Phys. 46, L382-L384.], 2008[Mitsui, T., Seto, M., Masuda, R., Kobayashi, Y. & Kitao, S. (2008). Jpn. J. Appl. Phys. 47, 7136-7139.]; Masuda et al., 2008[Masuda, R., Mitsui, T., Kitao, S., Higashitaniguchi, S., Yoda, Y. & Seto, M. (2008). Jpn. J. Appl. Phys. 47, 8087-8090.]). In the present paper we describe the principle and component devices of the developed spectrometer. As application studies, we present two experimental results concerning high-pressure Mössbauer studies of an iron orthoborate 57Fe3BO6 and an intermetallic compound Gd57Fe2.

2. Principle of SR 57Fe-Mössbauer spectroscopy with nuclear monochromator

Fig. 1[link] shows schematically the principle of energy-domain SR 57Fe-Mössbauer spectroscopy with a variable-frequency nuclear monochromator (VFNM) (Mitsui et al., 2007c[Mitsui, T., Seto, M. & Masuda, R. (2007c). Jpn. J. Appl. Phys. 46, L930-L932.]). In this method, an electronically forbidden pure nuclear Bragg reflection from a 57FeBO3 single crystal near the Néel temperature achieves an extremely high suppression of electronic Thomson scattering and filters 14.4 keV single-line 57Fe-Mössbauer radiation from a SR source. Here, a purely nuclear Bragg reflection is allowed for the (hkl) lattice plane of 57FeBO3 with the conditions h + k + l = odd. The extremely narrow energy bandwidth of neV order is realised by the collapse of the nuclear Bragg reflection near the Néel temperature in the presence of a small magnetic field, Hex ≃ 100 Oe (Smirnov et al., 1986[Smirnov, G. V., Zelepukhin, M. V. & Van Bürck, U. (1986). JETP Lett. 43, 352-355.]).

[Figure 1]
Figure 1
Conceptual diagram of energy-domain SR 57Fe-Mössbauer spectroscopy with a VFNM. (a) Basic principle of the VFNM. (b) Scheme of the energy scan process for obtaining the Mössbauer spectrum.

As shown in Fig. 1(a)[link], the SR 57Fe-Mössbauer radiation is emitted at a fixed beam position by the moving crystal parallel to the reflection plane. Then, the nuclear resonance energy is shifted by the Doppler effect of light according to

[E_{\rm{ref}}=E_0\left[1+(v/c)\cos\theta_{\rm{B}}\right].\eqno(1).]

Here, E0 is the nuclear resonance energy of 57Fe nuclides in the crystal at rest, v is the velocity of the crystal, c is the velocity of light, θB is the Bragg angle and Eref is the energy of the reflected SR 57Fe-Mössbauer radiation in the laboratory system. In this system, if a sample is placed at position A or B in Fig. 1(a)[link], the Mössbauer spectrum in both cases can be measured equally by counting the reflected SR 57Fe-Mössbauer radiation as a function of velocity, as shown in Fig. 1(b)[link]. Here, in the case of the former position (A), the VFNM functions as a nuclear energy analyzer.

3. Nuclear monochromator system installed at BL11XU of SPring-8

To develop materials science using energy-domain SR 57Fe-Mössbauer spectroscopy, a nuclear monochromator system has been designed and installed at the JAEA beamline (BL11XU) of SPring-8. The system consists of a VFNM unit and a precision diffractometer. Their external views are shown in Figs. 2(a) and 2(b)[link], respectively.

[Figure 2]
Figure 2
The nuclear monochromator system installed at BL11XU of SPring-8. (a) Light oven for heating an oscillating 57FeBO3 single crystal. (b) External view of the high-precision diffractometer for nuclear Bragg scattering.

The VFNM utilizes a large, very high quality 95%-57Fe-enriched 57FeBO3 single crystal to filter high-throughput 14.4 keV 57Fe-Mössbauer radiation from SR (see Fig. 2a[link]). The crystal is grown by the flux method (Mitsui et al., 2005[Mitsui, T., Takei, H., Kitao, S., Seto, M., Harami, T., Zhang, X. W., Yoda, Y. & Kikuta, S. (2005). Trans. Mater. Res. Soc. Jpn, 30, 7-10.]). It is formed as a thin platelet with its surface parallel to the (111) easy magnetization plane. In that work, plane-wave X-ray topography proves that there is no curvature or growth boundaries throughout the whole crystal surface beyond 1.0 arcsec. In order to heat the crystal to the Néel temperature, a light oven, of weight 95 g, is used. The crystal is placed on a copper heating stage and is covered with a compact aluminium case, which is partially shielded by a Kapton polyimide film for a beam access area. The temperature of a 57FeBO3 crystal is precisely stabilized with a precision within 0.01 K near the Néel temperature. Then, built-in small adjustable magnetic poles apply the external magnetic field within the range 100–150 Oe to the crystal surface and normal to the scattering plane. Under these conditions the heated crystal is oscillated parallel to the basal (111) plane using a Mössbauer velocity transducer (WissEL MVT-1000).

In order to perform high-resolution measurements of nuclear Bragg reflections from the 57FeBO3 crystal, a special diffractometer has also been installed at BL11XU. As shown in Fig. 2(b)[link], it consists of two high-precision goniometers and some motorized swivel and translation stages. The device parameters are summarized in Table 1[link].

Table 1
Parameters of the high-precision diffractometer of the VFNM

  Range Resolution Capacity
First goniometer
θ ±90° 0.00002° 500 kg
2θ arm −45° to 150° 0.0001° 250 kg
Swivel (Rx1) ±10° 0.000765° 5 kg
Z-axis (Z1) ±10 mm 0.25 mm 10 kg
 
Horizontal travel
Coarse ±100 mm (manual)  
Fine ±15 mm 0.16 mm  
 
Second goniometer
θ ±90° 0.0001° 25 kg
2θ arm ±90° 0.0002° 10 kg
Swivel (Rx2) ±10° 0.000765° 5 kg
Z-axis (Z2) ±10 mm 0.25 mm 10 kg
 
Stage size 650 × 1150 × 1040 (−200 to +50); vertical axis, driven by stepping motor

In normal mode, the VFNM unit is attached to the first goniometer, and the Bragg angle of a 57FeBO3 single crystal is adjusted with a high angular resolution below 1.0 arcsec. As an advantage of this apparatus, the first goniometer is equipped with a special load compensation mechanism on the 2θ axis. In future application studies, it will support the use of additional optical components and heavy experimental devices downstream of the VFNM.

4. Energy-domain SR Mössbauer spectrometer for high-pressure measurement

The application of ultrahigh-pressure Mössbauer spectroscopy requires measurements on the extremely small sample enclosed in the DAC. Therefore, we have developed a new energy-domain high-pressure SR 57Fe-Mössbauer spectrometer (HP-SRMS), which makes it possible to perform small-target research with a high signal-to-noise ratio. The main optical components are a high-resolution monochromator, an X-ray focusing mirror and a VFNM system. The external view and optics of the spectrometer are shown in Figs. 3(a) and 3(b)[link], respectively.

[Figure 3]
Figure 3
SR 57Fe-Mössbauer spectrometer for ultrahigh-pressure measurements. (a) External view, (b) nuclear diffraction optics. SR: synchrotron radiation from the undulator of BL11XU; DM: double-crystal monochromator, Si 111 reflections; HRM: high-energy resolution monochromator, nested-type channel-cut Si 511 × Si 975 reflections; MXFM: multilayer X-ray focusing mirror; F: focusing position; VFNM: variable-frequency nuclear monochromator; 57FeBO3 crystal near the Néel temperature; H: magnetic field; S: 0.4 mm × 5.0 mm slit; D: NaI detector.

The experimental procedure is as follows. The electron current of the SPring-8 storage ring is kept at 100 mA at 8.0 GeV by its `top-up' operation (Tanaka et al., 2006[Tanaka, H. et al. (2006). J. Synchrotron Rad. 13, 378-391.]). A liquid-nitrogen-cooled Si(111) double-crystal monochromator is used to handle the high heat load of undulator radiation (Shiwaku et al., 2004[Shiwaku, H., Mitsui, T., Tozawa, K., Kiriyama, K., Harami, T. & Mochizuki, T. (2004). AIP Conf. Proc. 705, 659-662.]), and σ-polarized incident X-rays with an energy width of 2.5 meV at 14.4 keV nuclear resonance of 57Fe are produced by a nested high-resolution monochromator (HRM) consisting of asymmetric Si(511) and asymmetric Si(975) channel-cut crystals (Mitsui et al., 2001[Mitsui, T., Kitao, S., Zhang, X. W., Marushita, M. & Seto, M. (2001). Nucl. Instrum. Methods Phys. Res. A, 467-468, 1105-1108.]). The beam size is 0.4 mm × 1.8 mm and the total flux is 1.0 × 1010 counts s−1, typically. A small-size probe beam is obtained using a bent elliptical multilayer X-ray focusing mirror (MXFM), which is coated with 50 layers of W(13 Å)/Si(39.5 Å) on a high-quality quartz SiO2 base (Mitsui et al., 2004[Mitsui, T., Kobayashi, Y. & Seto, M. (2004). Jpn. J. Appl. Phys. 43, 389-393.]). Here, the incident X-rays are reflected at a Bragg peak angle of 8.6 mrad and are horizontally focused with a size of 400 µm × 20 µm at the focus position (F) 600 mm downstream from the centre position of the MXFM. The total flux is about 4.5 × 109 counts s−1. The focused X-rays are ultrafinely monochromated to a bandwidth of about 15.4 neV by a VFNM using a single-line pure nuclear Bragg reflection from a 57FeBO3 single crystal near the Néel temperature (348.8 K) in a 150 Oe external field. In order to change the nuclear resonance energy, the crystal is oscillated in the sinusoidal velocity mode with a frequency range of 8.0 ≤ f ≤ 10.0 Hz. Behind a slit (S), the nuclear Bragg diffracted X-rays are detected by a NaI scintillation detector. The typical peak photon-counting rates of 57FeBO3 (111) and (333) reflections are about 7.5 × 103 and 3.5 × 103 counts s−1, respectively. Their noise levels are below 4.0%. In this optics, if the sample enclosed in the DAC is placed at the beam focus position (F), the Mössbauer absorption spectrum can be measured by counting the intensity of a single-line nuclear Bragg reflection as a function of velocity as described in §2[link].

5. Application experiments

To evaluate the experimental performance of the HP-SRMS, we have carried out high-pressure studies of polycrystalline 57Fe3BO6 and Gd57Fe2. The latter is, to our knowledge, the first report on Mössbauer spectroscopy under multimegabar pressures (up to P ≃ 300 GPa).

5.1. Pressure-induced phase transition from the antiferromagnetic to non-magnetic state in 57Fe3BO6

Under ambient pressure, iron orthoborate Fe3BO6 belongs to the space group D2h16 (Pnma) and below TN = 508 K represents an antiferromagnetic structure with weak ferromagnetism. The elemental cell contains 12 Fe3+ ions occupying non-equivalent 4c and 8d sites and the 57Fe-Mössbauer spectrum shows a typical profile with multi-site hyperfine interactions (Wolfe et al., 1969[Wolfe, R., Pierce, R. D., Eibschütz, M. & Nielsen, J. W. (1969). Solid. State Commun. 7, 949-952.]). Although there are many papers dealing with the magnetic properties of Fe3BO6, the high-pressure phase is still unknown. Therefore, we have carried out SR Mössbauer spectroscopy of Fe3BO6 under high pressure as the first performance-test experiment of the HP-SRMS.

Isotopically enriched polycrystalline samples were prepared by grinding the flux-grown 57Fe3BO6 (57Fe, 95%) crystal in an agate mortar. The sample and small ruby chips were enclosed in a hole of diameter 59 µm in a rhenium gasket between the bevelled-diamond anvils of 150 µm culet size. NaCl was used as the pressure-transmitting medium. The pressure was estimated using the shift of the ruby R1 fluorescence line. The pressure inhomogeneity was estimated to be less than 2.0 GPa from ruby chips in the sample chamber.

SR 57Fe-Mössbauer spectra in the pressure range 0 ≤ P ≤ 59 GPa were measured at room temperature by a VFNM using the 57FeBO3 (333) reflection plane. Statistically sufficient spectra were obtained with a short measurement time of 2.0 h. This shows the high potential of the HP-SRMS in high-pressure studies using a DAC. Note that typical data collection times for such high-pressure measurements using a RI source is several days or more.

Fig. 4[link] presents typical absorption spectra for different pressures. From the results of data fitting, one can recognize that all spectra consist of two hyperfine subspectra with an intensity ratio of 1:2. This is clear evidence that the iron atoms are in two non-equivalent crystallographic sites over the whole pressure range up to 59 GPa. The calculated HFS parameters are plotted in Fig. 5[link].

[Figure 4]
Figure 4
Room-temperature SR 57Fe-Mössbauer spectra of polycrystalline 57Fe3BO6 for pressures below and above the magnetic phase transition. Solid and dashed lines correspond to calculations.
[Figure 5]
Figure 5
Experimental results for the hyperfine splitting parameters obtained from the room-temperature SR Mössbauer spectroscopy of polycrystalline 57Fe3BO6. Dashed lines act as guides to the eye.

At ambient pressure, the spectrum shows well resolved Zeeman sextets owing to 4c and 8d iron sites. The hyperfine fields were 45.3 and 42.2 T, respectively. These values were in good agreement with previous results (Wolfe et al., 1969[Wolfe, R., Pierce, R. D., Eibschütz, M. & Nielsen, J. W. (1969). Solid. State Commun. 7, 949-952.]). Here, the parameters IS and QS are typical of the Fe3+ high-spin (HS) 3d5 state (S = 5/2) in an octahedral environment (Wolfe et al., 1969[Wolfe, R., Pierce, R. D., Eibschütz, M. & Nielsen, J. W. (1969). Solid. State Commun. 7, 949-952.]; Greenwood & Gibb, 1971[Greenwood, N. N. & Gibb, T. C. (1971). Mössbauer Spectroscopy, p. 91. London: Chapman and Hall.]).

In the pressure range 0 < P < 49 GPa, the hyperfine fields at the 4c and 8d sites show a gradual increase up to 48.1 and 46.4 T, respectively. This is likely to be due to the increase in Néel temperature caused by the strengthening of the exchange interaction with the lattice volume decrease. In contrast, the IS and QS parameters show no pronounced pressure effect up to 49 GPa. In Fig. 5[link], the negative slope in IS (P) reflects only the increase in s-electron density at the 57Fe nucleus with pressure.

At P ≥ 49 GPa, new non-magnetic quadrupole doublets appear at the centre position of the spectrum and its abundance clearly increases with pressure. Then, all HFS parameters change abruptly at around P = 49 GPa as shown in Fig. 5[link]. These results give clear evidence of the magnetic phase transition of the antiferromagnet Fe3BO6 state to the non-magnetic state. The coexistence of a magnetic and a non-magnetic phase is presumably due to the effect of pressure gradients. Actually, at P = 59 GPa, all of the iron transforms into the non-magnetic phase and the spectrum shows an asymmetric doublet which consists of the two quadrupole subspectra with an intensity ratio of 1:2. Their IS are almost the same value, 0.1 mm s−1, which is the typical value of the Fe3+ ions in the low-spin (LS) state (S = 1/2) (Greenwood & Gibb, 1971[Greenwood, N. N. & Gibb, T. C. (1971). Mössbauer Spectroscopy, p. 91. London: Chapman and Hall.]). On the other hand, the QS show different values of 1.25 mm s−1 and 0.79 mm s−1, respectively, which implies that two different iron sites exist in 57Fe3BO6 after the transition. These results suggest that the pressure-induced transition of Fe3BO6 is strongly related to an isostructural HS → LS transition. Recently, similar phase transition behaviour has been observed in 57FeBO3, whose magnetic moment collapses with the spin crossover HS → LS at a pressure of about 46 GPa. The change of the magnetic and electronic structures is explained by Mott's transition with rupturing of strong dd-electron correlations (Troyan et al., 2001[Troyan, I. A., Gavrilyuk, A. G., Sarkisyan, V. A., Lyubutin, I. S., Rüffer, R., Leupold, O., Barla, A., Doyle, B. & Chumakov, A. I. (2001). JETP Lett. 74, 26-29.]; Sarkisyan et al., 2002[Sarkisyan, V. A., Troyan, I. A., Lyubutin, I. S. Gavrilyuk, A. G. & Kashuba, A. F. (2002). JETP Lett. 76, 788-793.]).

In the present preliminary study using a HP-SRMS, at least, we could observe a new pressure-induced magnetic phase transition of Fe3BO6, accompanying drastic changes of the electronic structure of iron. However, to elucidate the more detailed mechanism of the high-pressure-induced transition in Fe3BO6, we require additional measurements involving Mössbauer spectroscopy at low temperature, X-ray diffraction and electric conductivity in a DAC.

5.2. High-pressure SR Mössbauer spectroscopy of GdFe2 in the multimegabar range

The intermetallic compound GdFe2 under ambient conditions is a ferrimagnet with high Tc (∼790 K) and belongs to the cubic C15 Laves phase structure with space group [Fd\bar3m]. Although the structural and magnetic properties are generally well understood, there are few reports on its high-pressure study. Nevertheless, one can expect that the magnetism of GdFe2 exhibits some unusual behaviour in the high-pressure phase. This is because the magnetism in GdFe2 arises not only from the itinerant 3d electrons of Fe atoms but also from the localized 4f electrons of Gd atoms; the former is strongly influenced by pressure compared with the latter. As a typical phenomenon, the high-pressure studies using NFS demonstrate that the magnetic order of GdFe2 is maintained in the whole pressure region up to 105 GPa (Lübbers et al., 1999[Lübbers, R., Wortmann, G. & Grünsteudel, H. (1999). Hyperfine Interact. 123/124, 529-559.]). In addition, it has also been suggested that a pressure-induced (C15 → C14) structural phase transition of GdFe2 occurs in the high-pressure phase above 50 GPa (Lübbers et al., 1999[Lübbers, R., Wortmann, G. & Grünsteudel, H. (1999). Hyperfine Interact. 123/124, 529-559.]; Reiß, 2000[Reiß, G. (2000). Dissertation, Paderborn University, Germany (downloadable from http://deposit.ddb.de/cgi-bin/dokserv?idn=961546808&dok_var=d1&dok_ext=pdf&filename=961546808.pdf ).]). To understand such phenomena in detail, it is essentially important to perform a high-pressure study of GdFe2 in the multimegabar range. Therefore, as a feasibility study in the multimegabar pressure range with HP-SRMS, we have carried out SR 57Fe-Mössbauer spectroscopy of GdFe2 under high pressures up to 302 GPa.

A 57Fe-95%-enriched Gd57Fe2 alloy was prepared by arc melting in an argon atmosphere. In order to obtain a single phase of GdFe2, the ingot was annealed at 1073 K for 200 h in a vacuum chamber. Polycrystalline samples were obtained by grinding the ingot sample in an agate mortar. A symmetrical diamond anvil cell with double-bevelled diamond anvils with 30 µm culet was used to generate multimegabar pressures. The sample was loaded into a hole of diameter 18 µm in a rhenium gasket without a pressure medium. The pressure gradients in the sample chamber were estimated to be less than 10 GPa at a maximum pressure of 302 GPa. Normally, we use the equations of state of elemental metals (e.g. Pt, Au etc.) for determination of the sample pressure in the multimegabar pressure range. Consequently, synchrotron radiation X-ray diffraction (SR-XRD) experiments are required for measuring the volume compression data of them. However, in the present experiment we could not carry out SR-XRD and SR-Mössbauer spectroscopy during the same SR machine time. Therefore, a diamond anvil Raman gauge was used as a practical pressure-determination method in the multimegabar pressure range without using SR-XRD (Akahama & Kawamura, 2004[Akahama, Y. & Kawamura, H. (2004). J. Appl. Phys. 96, 3748-3751.], 2007[Akahama, Y. & Kawamura, H. (2007). High Press. Res. 27, 473-482.]). The room-temperature SR 57Fe-Mössbauer spectra of Gd57Fe2 under high pressures up to 302 GPa were measured by a VFNM using the 57FeBO3 (111) reflection plane.

Fig. 6[link] presents the Mössbauer spectra of Gd57Fe2 for different pressures. Each spectrum has been measured with a considerably short data collection time of 3.0 h. In spite of the ultrahigh-pressure measurements up to 302 GPa, all spectra are obtained with good statistical quality. These results prove that the HP-SRMS easily realises even ultrahigh-pressure SR 57Fe-Mössbauer spectroscopy in the multimegabar region. One should note that such measurements have not been achieved by conventional RI-Mössbauer spectroscopy.

[Figure 6]
Figure 6
Room-temperature SR 57Fe-Mössbauer spectra of polycrystalline Gd57Fe2 under multimegabar pressures. Solid and dashed lines correspond to calculations.

At ambient pressure the spectrum is fitted by a superposition of four Zeeman sextets, which agrees with the well known model of a random orientation of magnetic easy axis relative to the crystallographic directions. The estimated HFS values are consistent with previous studies (Atzmony & Dariel, 1974[Atzmony, U. & Dariel, M. P. (1974). Phys. Rev. B, 10, 2060-2067.]; Genin et al., 1981[Genin, J. M., Bauer, Ph. & Besnus, M. J. (1981). Phys. Status Solidi A, 64, 325-333.]; Mori et al., 1998[Mori, K., Onodera, H., Aoki, K. & Masumoto, T. (1998). J. Alloys Compd. 270, 35-41.]).

When the pressure rises to 51 GPa, the splitting of resonance lines decreases because of a decrease in the hyperfine field at the iron nuclei. Moreover, at a glance, one can observe an additional broadened absorption dip at the centre portion of the spectrum. The new subspectral component indicates a principle change of structural and magnetic properties of GdFe2. In the pressure range 51 < P ≤ 302 GPa, the hyperfine field decreases gradually with the build up of pressure and finally disappears at about 194 GPa. The strong magnetic conservation in a high-pressure phase originates in the effect of the localized 4f electrons of Gd (Lübbers et al., 1999[Lübbers, R., Wortmann, G. & Grünsteudel, H. (1999). Hyperfine Interact. 123/124, 529-559.]). On the other hand, at P ≥ 248 GPa, the spectra of non-magnetic GdFe2 clearly show the two well resolved quadrupole doublets, corresponding to the two crystallographically non-equivalent iron sites. The intensity ratio of the corresponding lines of two components is approximately in the ratio 3:1 (see the calculated dashed lines in Fig. 6[link]). This result strongly suggests that the high-pressure phase of GdFe2 is a hexagonal C14 structure, which is likely formed by a pressure-induced structural phase transition at about 50 GPa. This is because, in the non-magnetic C14 structure, the iron atoms are generally located on the two crystallographically non-equivalent sites (6h and 2a) with an abundance ratio of 3:1. In contrast, GdFe2 is in the C15 structure; all the iron sites are located in equivalent sites in the absence of a hyperfine magnetic field (Atzmony & Dariel, 1974[Atzmony, U. & Dariel, M. P. (1974). Phys. Rev. B, 10, 2060-2067.]). Therefore, the nuclear quadrupole interaction is axially symmetric, and only the doublet of one component should be observed. However, this is contradictory to the present experimental results. Actually, the spectra above 51 GPa have been fitted well by assuming GdFe2 in the C14 structure. The calculated results for the averaged hyperfine structure parameters (Havg, ISavg and QSavg) are plotted in Fig. 7[link]. As was expected, at P = 51 GPa, ISavg and QSavg show significant changes compared with the corresponding values at ambient pressure. In contrast, at pressures above 51 GPa they show no pronounced pressure dependence up to 302 GPa. These results also provide clear evidence of the pressure-induced structural transition of GdFe2 from the C15 to the C14 phase at about 50 GPa (Lübbers et al., 1999[Lübbers, R., Wortmann, G. & Grünsteudel, H. (1999). Hyperfine Interact. 123/124, 529-559.]). However, the details of the mechanism are not explained completely. Currently, in order to elucidate the structural and magnetic properties on the high-pressure phase in GdFe2 in more detail, we are preparing further high-pressure measurements in the multimegabar range, including SR-XRD and SR-Mössbauer spectroscopy at low temperature.

[Figure 7]
Figure 7
Experimental results for the averaged hyperfine splitting parameters obtained from the room-temperature SR 57Fe-Mössbauer spectroscopy of polycrystalline Gd57Fe2.

6. Conclusion

A new SR-based 57Fe-Mössbauer spectrometer, consisting of a HRM, a MXFM and a VFNM, has been developed for performing high-pressure studies. It allows us to perform energy-domain high-pressure SR 57Fe-Mössbauer spectroscopy using a DAC of micrometre-scale gasket-hole size.

In application studies, this spectrometer could observe clearly a pressure-induced magnetic phase transition of 57Fe3BO6 at about 50 GPa and a new high-pressure phase of non-magnetic Gd57Fe2 with C14 hexagonal structure in the multimegabar region. In particular, the latter study demonstrates that the spectrometer can probe different solid-state properties at multimegabar pressures. It promises that the spectrometer can become a powerful tool for the research field of the interior of the Earth and planets in future studies; this is because the pressure in the iron-alloy core of the Earth ranges from 135 GPa to 365 GPa.

Acknowledgements

The authors would like to thank Professor S. Kikuta, Professor S. Nasu and Dr K. Aoki for various fruitful discussions, valuable comments and continuous encouragement, respectively. The authors are also grateful to the accelerator groups of SPring-8 for the excellent top-up operation. This work was partly supported by the New Energy and Industrial Technology Development Organization (NEDO) under the Advanced Fundamental Research Project on Hydrogen Storage Materials. NH was partially supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Science, Sport and Technology of the Japanese Government.

References

First citationAkahama, Y. & Kawamura, H. (2004). J. Appl. Phys. 96, 3748–3751.  Web of Science CrossRef CAS Google Scholar
First citationAkahama, Y. & Kawamura, H. (2007). High Press. Res. 27, 473–482.  CrossRef CAS Google Scholar
First citationAtzmony, U. & Dariel, M. P. (1974). Phys. Rev. B, 10, 2060–2067.  CrossRef CAS Web of Science Google Scholar
First citationChumakov, A. I., Zelepukhin, M. V., Smirnov, G. V., Van Bürck, U., Rüffer, R., Hollatz, R., Rüter, H. D. & Gerdau, E. (1990). Phys. Rev. B, 41, 9545–9547.  CrossRef CAS Web of Science Google Scholar
First citationGenin, J. M., Bauer, Ph. & Besnus, M. J. (1981). Phys. Status Solidi A, 64, 325–333.  CrossRef CAS Google Scholar
First citationGreenwood, N. N. & Gibb, T. C. (1971). Mössbauer Spectroscopy, p. 91. London: Chapman and Hall.  Google Scholar
First citationJayaraman, A. (1983). Rev. Mod. Phys. 55, 65–108.  CrossRef CAS Web of Science Google Scholar
First citationLübbers, R., Wortmann, G. & Grünsteudel, H. (1999). Hyperfine Interact. 123/124, 529–559.  Google Scholar
First citationMasuda, R., Mitsui, T., Kitao, S., Higashitaniguchi, S., Yoda, Y. & Seto, M. (2008). Jpn. J. Appl. Phys. 47, 8087–8090.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Kitao, S., Zhang, X. W., Marushita, M. & Seto, M. (2001). Nucl. Instrum. Methods Phys. Res. A, 467468, 1105–1108.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Kobayashi, Y. & Seto, M. (2004). Jpn. J. Appl. Phys. 43, 389–393.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Seto, M., Hirao, N., Ohishi, Y., Kobayashi, Y., Higashitaniguchi, S. & Masuda, R. (2007d). Jpn. J. Appl. Phys. 46, L382–L384.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Seto, M., Kikuta, S., Hirao, N., Ohishi, Y., Takei, H., Kobayashi, Y., Kitao, S., Higashitaniguchi, S. & Masuda, R. (2007a). Jpn. J. Appl. Phys. 46, 821–825.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Seto, M. & Masuda, R. (2007c). Jpn. J. Appl. Phys. 46, L930–L932.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Seto, M., Masuda, R., Kiriyama, K. & Kobayashi, Y. (2007b). Jpn. J. Appl. Phys. 46, L703–L705.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Seto, M., Masuda, R., Kobayashi, Y. & Kitao, S. (2008). Jpn. J. Appl. Phys. 47, 7136–7139.  Web of Science CrossRef CAS Google Scholar
First citationMitsui, T., Takei, H., Kitao, S., Seto, M., Harami, T., Zhang, X. W., Yoda, Y. & Kikuta, S. (2005). Trans. Mater. Res. Soc. Jpn, 30, 7–10.  CAS Google Scholar
First citationMori, K., Onodera, H., Aoki, K. & Masumoto, T. (1998). J. Alloys Compd. 270, 35–41.  Web of Science CrossRef CAS Google Scholar
First citationNasu, S. (1994). Hyperfine Interact. 90, 59–75.  CrossRef CAS Web of Science Google Scholar
First citationPankhurst, Q. A., Cohen, N. S., Barquín, L. F., Gibbs, M. R. J. & Smirnov, G. V. (2001). J. Non-Cryst. Solids, 287, 81–87.  Web of Science CrossRef CAS Google Scholar
First citationReiß, G. (2000). Dissertation, Paderborn University, Germany (downloadable from http://deposit.ddb.de/cgi-bin/dokserv?idn=961546808&dok_var=d1&dok_ext=pdf&filename=961546808.pdf ).  Google Scholar
First citationSarkisyan, V. A., Troyan, I. A., Lyubutin, I. S. Gavrilyuk, A. G. & Kashuba, A. F. (2002). JETP Lett. 76, 788–793.  Web of Science CrossRef Google Scholar
First citationShiwaku, H., Mitsui, T., Tozawa, K., Kiriyama, K., Harami, T. & Mochizuki, T. (2004). AIP Conf. Proc. 705, 659–662.  CrossRef CAS Google Scholar
First citationSmirnov, G. V. (2000). Hyperfine Interact. 125, 91–112.  Web of Science CrossRef CAS Google Scholar
First citationSmirnov, G. V., van Bürck, U., Chumakov, A. I., Baron, A. Q. R. & Rüffer, R. (1997). Phys. Rev. B, 55, 5811–5815.  CrossRef CAS Web of Science Google Scholar
First citationSmirnov, G. V., Zelepukhin, M. V. & Van Bürck, U. (1986). JETP Lett. 43, 352–355.  Google Scholar
First citationSturhahn, W. & Jackson, J. M. (2007). Geol. Soc. Am. Spec. Pap. 421, 157–174.  Google Scholar
First citationTanaka, H. et al. (2006). J. Synchrotron Rad. 13, 378–391.  Web of Science CrossRef IUCr Journals Google Scholar
First citationTroyan, I. A., Gavrilyuk, A. G., Sarkisyan, V. A., Lyubutin, I. S., Rüffer, R., Leupold, O., Barla, A., Doyle, B. & Chumakov, A. I. (2001). JETP Lett. 74, 26–29.  Web of Science CrossRef Google Scholar
First citationWolfe, R., Pierce, R. D., Eibschütz, M. & Nielsen, J. W. (1969). Solid. State Commun. 7, 949–952.  CrossRef CAS Web of Science Google Scholar

© International Union of Crystallography. Prior permission is not required to reproduce short quotations, tables and figures from this article, provided the original authors and source are cited. For more information, click here.

Journal logoJOURNAL OF
SYNCHROTRON
RADIATION
ISSN: 1600-5775
Follow J. Synchrotron Rad.
Sign up for e-alerts
Follow J. Synchrotron Rad. on Twitter
Follow us on facebook
Sign up for RSS feeds