Low-temperature high magnetic field powder x-ray

Low-temperature high magnetic field powder x-ray diffraction setup for field-induced structural phase transition studies from 2 to 300 K and at 0 to 8-T field Aga Shahee, Shivani Sharma, Dhirendra Kumar, Poonam Yadav, Preeti Bhardwaj, Nandkishor Ghodke, Kiran Singh, N. P. Lalla, and P. Chaddah Citation: Review of Scientific Instruments 87, 105110 (2016); doi: 10.1063/1.4963843 View online: http://dx.doi.org/10.1063/1.4963843 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/87/10?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Magnetic and charge ordering properties of Bi0.6-x(RE)xCa0.4MnO3 (0.0 ≤ x ≤ 0.6) perovskite manganites J. Appl. Phys. 111, 07E128 (2012); 10.1063/1.3675982 Antiferromagnet-ferromagnet phase transition in lightly doped manganites Low Temp. Phys. 31, 819 (2005); 10.1063/1.2008144 X-ray powder diffractometer for in situ structural studies in magnetic fields from 0 to 35 kOe between 2.2 and 315 K Rev. Sci. Instrum. 75, 1081 (2004); 10.1063/1.1667253 High magnetic field studies of 3d and 4f magnetism in (R 0.7 A 0.3 )MnO 3 : R=La 3+ , Pr 3+ , Nd 3+ , A=Ca 2+ , Sr 2+ , Ba 2+ , Pb 2+ J. Appl. Phys. 85, 5384 (1999); 10.1063/1.369985 Cation disorder and size effects upon magnetic transitions in Ln 0.5 A 0.5 MnO 3 manganites J. Appl. Phys. 82, 6181 (1997); 10.1063/1.366543

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 14.139.62.174 On: Fri, 07 Oct 2016 04:55:41

REVIEW OF SCIENTIFIC INSTRUMENTS 87, 105110 (2016)

Low-temperature high magnetic field powder x-ray diffraction setup for field-induced structural phase transition studies from 2 to 300 K and at 0 to 8-T field Aga Shahee, Shivani Sharma, Dhirendra Kumar, Poonam Yadav, Preeti Bhardwaj, Nandkishor Ghodke, Kiran Singh, N. P. Lalla,a) and P. Chaddah UGC-DAE Consortium for Scientific Research, University Campus, Khandawa Road, Indore, MP 452001, India

(Received 22 July 2016; accepted 17 September 2016; published online 6 October 2016) A low-temperature and high magnetic field powder x-ray diffractometer (XRD) has been developed at UGC-DAE CSR (UGC: University Grant Commission, DAE: Department of Atomic Energy, and CSR: Consortium for scientific research), Indore, India. The setup has been developed around an 18 kW rotating anode x-ray source delivering Cu-Kα x-rays coming from a vertical line source. It works in a symmetric θ-2θ parallel beam geometry. It consists of a liquid helium cryostat with an 8 T split-pair Nb-Ti superconducting magnet comprising two x-ray windows each covering an angular range of 65◦. This is mounted on a non-magnetic type heavy duty goniometer equipped with all necessary motions along with data collection accessories. The incident x-ray beam has been made parallel using a parabolic multilayer mirror. The scattered x-ray is detected using a NaI detector through a 0.1◦ acceptance solar collimator. To control the motions of the goniometer, a computer programme has been developed. The wide-angle scattering data can be collected in a range of 2◦–115◦ of 2θ with a resolution of ∼0.1◦. The whole setup is tightly shielded for the scattered x-rays using a lead hutch. The functioning of the goniometer and the artifacts arising possibly due to the effect of stray magnetic field on the goniometer motions, on the x-ray source, and on the detector have been characterized by collecting powder XRD data of a National Institute of Standards and Technology certified standard reference material LaB6 (SRM-660b) and Si powder in zero-field and in-field conditions. Occurrence of field induced structural-phase transitions has been demonstrated on various samples like Pr0.5Sr0.5MnO3, Nd0.49Sr0.51MnO3−δ and La0.175Pr0.45Ca0.375MnO3 by collecting data in zero field cool and field cool conditions. Published by AIP Publishing. [http://dx.doi.org/10.1063/1.4963843]

INTRODUCTION

Ever since its advent about a century back, the technique of x-ray diffraction (XRD) has directly or indirectly enriched the knowledge base of a variety of basic and applied disciplines in the arena of science. It has been immensely used in the area of materials science for crystal structure determination and phase transition studies. Using XRD the structural phase diagrams of materials were initially investigated mainly against temperature (above room temperature) and composition. But with the advent of brighter x-ray sources, cryogenics, and efficient high-pressure techniques, the phase diagrams of materials have also been studied against low-temperatures and high-pressure as well. Like composition, temperature, and pressure, magnetic field is also a thermodynamic variable. Materials which pose magneto-structural coupling will show changes in their structure as a function of applied field. Magnetostriction has been known since long as direct effect of magnetic field on the atomic structure of magnetic materials. But it is only since last decade that researchers1–4 have developed infield XRD setups for studying the effects of magnetic field on the crystal structure of materials. The current progress of materials science regarding multifunctional materials (MFMs) has put forward further new challenges to understand the microa)Author to whom correspondence should be addressed. Electronic mail:

[email protected]

scopic origin of their physical properties. Most of such MFMs have magneto-elastic coupling and are therefore expected to show field induced structural changes. In manganites,3 hustler alloys,4 magnetocaloric materials,5,6 and 1-D spin-quantum magnets,7,8 field induced structural phase transitions have been seen. To understand the underlying structure property correlations, it is necessary to investigate their structural evolution as a function of magnetic field. Depending on the requirement, the in-field XRD facilities have been developed around laboratory based sources1–6 as well as synchrotrons.7–17 Initial efforts in this direction were made by Ohsumi et al.,1,2 Shimomura et al.,3 and Ma et al.;4 however, the designs were mostly limited to scanning just few peak-profiles of single crystals. The availability of several orders of higher intense beams at synchrotron sources permits the design flexibility to develop LTHM-XRD setups to study materials’ structure under high dc-fields in the range of 1315 T.7,8 Various pulsed field based systems9–17 have also been developed at synchrotrons to study materials at fields as high as 30-40 T. While the pulsed high magnetic-field setups are inevitable for certain type of studies and have benefits of moderate power consumption, however, the extremely high rate of flux-change in such setups causes heating of the samples.18 Keeping in view the studies requiring frequent need at once own ease and discretion, the importance of laboratory x-ray based powder XRD setups, which are capable of providing reasonably high-resolution refinable data, remains intact.

0034-6748/2016/87(10)/105110/9/$30.00 87, 105110-1 Published by AIP Publishing. Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 14.139.62.174 On: Fri, 07 Oct 2016 04:55:41

105110-2

Shahee et al.

In this regard perhaps the first attempt was made by Pecharsky et al.5 and Holm et al.,6 at the Ames Laboratory, USA. Their system comprises a rotating anode Mo x-ray source, a symmetric Brag-Brentano θ-2θ focusing diffractometer, and a moderate 3.5 T magnet. They were the first to quantify the field induced changes in the crystal structure in a magneto-caloric material Gd5Ge4 and thus show the effectiveness of the laboratory based LTHM-XRD setups. Debye-Scherrer camera based systems have also been developed for 5 T room-temperature and 10 T high-temperature studies by Mitsui et al.19,20 Very recently Wang et al.,18 have developed a unique setup, which is based on a fixed tube Mo source and an image plate (IP) detector system and uses a 25 T dc-field. This dc-field value for LTHM-XRD setup is perhaps the highest. Das et al.21 have developed a LTHM-XRD setup which is designed around a new type of magnet concept, trapped field magnet (TFM). This setup provides a very large range of assessable scattering directions and is suitable for polarized x-ray scattering experiments. There are materials, e.g., phase separated manganites and magnetic shape memory alloys, where very high field is not needed, and a moderate high-field ∼8 T can serve the purpose very well. Robust split-pair magnets in an 8 T field range are readily available as per the required x-ray optical window. Infield powder XRD setup developed for moderate high dc field and working in symmetric diffraction geometry to produce refinable powder XRD data is still fewer.6 Keeping in view the necessity of such a setup to corroborate with the current scenario in materials science, a moderate high-field (8 T) LTHM powder x-ray diffraction setup has been developed at UGC-DAE CSR (UGC: University Grant Commission, DAE: Department of Atomic Energy, CSR: Consortium for scientific research), Indore, India. The beauty of this setup is that it assures constant field in the sample during the XRD scan. A varying field may have some significant effects, which may be difficult to account. The setup is working very nicely since 2014.

INSTRUMENTATION AND DESIGN

Looking into the requirement of an 8 T field and thereby large stray magnetic field (SMF) of the readily available splitpair magnets, the design of an angle dispersive XRD setup on a fixed rotating anode source is a bit complicated task. Constrained by the various design considerations arising due to SMF present around the magnet, about 400 G at ∼30 cm, one cannot simply go with readily available diffractometers. Large SMF present around the goniometer is likely to hinder or even totally block the motion of the stepper motors used in the goniometers. Therefore the active components of the goniometer, like worm/gear and ball bearings, were chosen to be non-magnetic type so that they do not get stuck during their operation under high-field. The other serious problem caused by SMF is the Lorentz force arising due to the interaction between SMF and the free electrons moving in vacuum in some of the main components of the diffractometer kept around the high field magnet. These are the thermionic electrons emitted from the filament of the x-ray tube and the photo-electrons in the photomultiplier tube (PMT) in a NaI-scintillation detector.

Rev. Sci. Instrum. 87, 105110 (2016)

The SMF will change the path of the thermionic electrons in an x-ray tube and displace the source position thereby changing the source intensity and increasing the zero-shift error of the 2θ zero position aligned under zero-field condition. Since one cannot keep the x-ray tube too far from the sample, i.e., the magnet center, therefore, to minimize the odd effects due to SMF, the x-ray tube needs to be magnetically shielded. Holm et al.6 have shown a clear positive effect of magnetic shielding of the source on the 2θ zero-shift value. A detailed study of the effect of SMF on the x-ray tube output intensity has been done by Brzozowski et al.22 during MRI development studies. In our design, the distance of closest approach between the magnet center and the source was limited by the overall mechanical dimensions of the goniometer components and the x-ray tube mounting table. The closest distance of approach was ∼100 cm. At this distance, the SMF is ∼40 G. As described in the following, a magnetic shield comprising two layers of 0.3 mm thick µ-metal sheet was found to be sufficient. Since the sample has to lie in the center of the magnet, ∼100 cm large distance between the magnet center and the source was to adversely affect the intensity of the x-ray incident on the sample. To compensate the odd effect due to large source to sample distance, a parabolic multi-layer mirror (Fox 1D-infin, Xenocs) has been used to sagittally focus the divergent x-rays to make it an intense and parallel beam incident on the sample. The mirror assembly is housed in a chamber, which is evacuated to minimize the x-ray absorption. It is designed to collect ∼0.6◦ of horizontal divergence from the source and make it almost parallel (beam divergence after mirror ∼0.025◦) in the scattering plane. To prevent the incident x-rays from being absorbed along the ∼60 cm long air path, an evacuated beam pipe has been used, which is sealed with ∼40 µm thick Mylar windows at both the ends. The vertical opening of the final aperture at the exit port of the beam pipe (∼11 mm) is matched with the vertical dimension of the sample (18 mm), so that it does not expose anything other than the sample giving unwanted spurious peaks. This allows an axial (vertical) divergence of ∼0.8◦ on the sample, which is much lesser than the typical value of ∼2◦, used in reasonably high resolution commercial diffractometers. The setup works in a parallel beam symmetric θ-2θ diffraction geometry. Compared to symmetric θ-2θ focusing geometry, a parallel beam symmetric θ-2θ geometry makes the measured scattering angle (2θ) rather invariant against the shift and minor tilt of the sample surface. Despite the cylindrical symmetry of the variable temperature insert (VTI) and the sample fixing insert, about the sample, some minor drift of the sample surface may always be present due to the inevitable anisotropy of the thermal stress generated during cooling or heating cycles. Thus a parallel beam symmetric θ-2θ geometry is better suited for such setups. A vertical view of the necessary x-ray optics and ray diagram is shown in Fig. 1. The setup has been developed around an 8 T Nb-Ti splitpair superconducting magnet (Oxford UK), which provides vertical field. The cryostat comprises two x-ray windows each covering an angular range of 65◦ and is designed to accept vertical loading of sample, mounted in an insert. The flat rectangular sample is vertically placed in the uniform (0.5%) field region in the centre of the magnet poles and remains


105110-3

Shahee et al.

FIG. 1. A vertical view of the main components of the LTHM XRD setup. RA: Rotating anode x-ray source, GM: parabolic focusing mirror, BL: evacuated beam-line, θ: θ-Circle 2θ: 2θ-Circle, LSS: long Soller slit, SC: scintillation counter, and S-Sample.

rigidly fixed with respect to it. The sample insert has a specially designed notch, which fits in a groove on the VTI of the cryostat. The magnet is mounted on rotation/translation (Rx/Tx2) stages with an aluminum spacer in between so that the center of the sample surface lies exactly at the center of rotation of the Rx stage. The rotation stage Rx, which is on the top of the translation stage Tx-2, allows the sample to be tilted (±5◦) in the plane perpendicular to the incident beam direction and the translation stage Tx-2 translates (±50 mm) the sample horizontally and normal to the incident beam. Thus the combination of Rx/Tx-2 stages allow us to properly align the flat surface of the sample with the coaxial axes of θ and 2θ circles of the goniometer, on which the two stages Rx/Tx-2 are finally mounted. The Rx/Tx-2 stages and the θ2θ circles are of heavy duty type (Huber GmbH, Germany). All these components are in turn mounted on another heavy duty translation stage Tx-1. Tx-1 stage allows the horizontal motion (±50 mm) normal to the incident beam direction so that the θ-2θ axes of the goniometer can be aligned with the incident x-ray beam path. The factory (Rigaku, Japan) supplied rotating anode x-ray generator machines which usually comprise heavy, ∼400 Kg of miled-steel table, which is basically a large ferromagnetic mass. Therefore, keeping in view the precautionary prohibitive instruction from the magnet manufacturer (Oxford, UK) that not to keep the magnet near (closer than 5-m) any large ferromagnetic mass, the x-ray tube part of the rotating anode generator was demounted from its usual miled-steel table and reinstalled on an equally heavy aluminum table. The aluminum table was designed to be capable of movement required during its horizontal alignment and stable fixing on the laboratory floor. The essential alignment related to the beam-direction and beam-positioning with respect to the θ-2θ axes of the goniometer was done before mounting the cryostat/magnet system on the goniometer. The parallel x-ray beam (divergence 0.025◦, size ∼0.5 mm at the sample place) of Cu-Kα coming from the parabolic mirror optics falls on the sample. The scattered x-rays are detected using a NaI scintillation detector through a 24 cm long Soller slit (LSS) with 0.1◦ horizontal acceptance. The solar collimator with 0.1◦ horizontal acceptance provides reasonably high resolution and also prevents unwanted diffraction signals coming from scattering by various windows (Beryllium and Kapton) used in the cryostat.

Rev. Sci. Instrum. 87, 105110 (2016)

Both are mounted on the diffraction arm of the goniometer. The vertical opening of the solar collimator is ∼11 mm, which allows ∼2.5◦ axial angular opening at the active (scintillator) surface of the NaI detector, kept at ∼48 cm from the sample. The hardware of goniometer and the detector is interfaced with a computer, and to control the required motions of the goniometer, during alignment as well as data collection, a computer programme has been developed. An automatic control/interface for the main x-ray exit shutter of the rotating anode x-ray source has also been designed and fabricated. It automatically opens about a fraction of seconds just before the XRD scan starts. A schematic view of the sample position and all the possible motions of the goniometer is shown in Fig. 2(a). The 2θ zero is aligned by directly scanning the 2θ circle across the incident parallel beam coming from the mirror. The 2θ zero is fixed in the center of the profile as shown in Fig. 2(b). The θ zero is aligned by cutting the direct beam intensity to half by translating the sample surface in the beam while the detector is kept at 2θ = 0 and then scanning the θ circle. This process is iterated at least twice to achieve the best possible zero θ alignment. The θ = 0 position is fixed at the vertex of the triangular intensity profile Fig. 2(c). The whole setup is housed in a tightly shielded lead hutch to prevent the users around from getting exposed from the x-rays. The setup consists of a staircase for easy top loading of the sample. Camera photographs of the interior and the shielded exterior views of the setup are shown in Figs. 3(a) and 3(b). The setup needs ∼0.3 cc of powder sample which is firmly glued on a copper sample holder with GE-varnish. Fig. 3(c) shows a view of the sample holder insert. The sample holder is then clamped at a prefixed position on the sample insert. The sample insert is then mounted vertically between the pole pieces of the split-pair magnet such that the sample surface remains vertically aligned, parallel, with the field direction. The notch/groove arrangement of the insert provides a prefixed orientation to the sample surface with respect to the incident x-ray beam path. The sample surface is further properly aligned to the θ-axis through θ zero alignment. A Cernox temperature sensor is mounted on the sample insert. The sample temperature follows the VTI, which uses a separate Cernox temperature sensor. The sample temperature can be varied in the range of 2–300 K with an average stability better than 20 mK and with absolute accuracy of around ±0.1 K. The available scattering angle ranges from 2◦ to 115◦ of 2θ. Statistically good quality XRD data in 2θ range of 10◦ to 110◦ can be collected within an hour at ∼10 kW working power of the rotating anode x-ray source. As pointed out above, intensity problems were faced during the in-filed XRD experiments using the NaI scintillation detector. In the very first trial attempt at a field as low as 0.5 T (and PMT unshielded), the intensity count was found to become almost zero. This was all due to the quenching of the efficiency of PMT, caused by the SMF induced trajectory change of the photoelectrons relative to the di-anodes in PMT. Therefore the rotation of the PMT about its own cylinder axis brought drastic improvement (∼90%) in the intensity count at 0.5 T. The PMT of the detector is situated at ∼50 cm from the magnet center. At this point, a vertical SMF of ∼120 G is present perpendicular to the cylinder axis of the PMT. The


105110-4

Shahee et al.

Rev. Sci. Instrum. 87, 105110 (2016)

FIG. 2. (a) A schematic view showing the sample position, field direction, and all the possible motions to manipulate the diffractometer during alignment and diffraction measurement. (b) The profile of the direct beam during 2θ scan showing 2θ = zero alignment, and (c) profile of the direct beam intensity variation during θ-scan of the sample surface while 2θ = 0, showing θ = zero alignment.

rotation of the PMT about this axis will change the angle between the trajectory of the photoelectrons and the SMF direction (vertical). Minimizing this angle will minimize the Lorentz force and the displacement of the photoelectrons relative to the di-anodes. This will in turn maximize the intensity counts. In this way, we could optimize an orientation of the PMT for the existing SMF at 0.5 T working field. But at higher working fields, just the orientation-optimization was not sufficient and therefore we had to magnetically seal the

PMT with cylindrical tube made out of 0.3 mm thick µ-metal sheets. With the trial and error, a tube comprising 7 layers of such µ-metal sheet was found to provide a reasonably efficient shielding against the SMF of ∼120 G at 8 T working field of the magnet. Once the detector shield got reasonably optimized, attempt was made to seal the x-ray tube as well, which was in a rather lower, ∼40 G, SMF region. Two layers of the same µmetal sheet were found to be sufficient. This was fixed directly on the x-ray tube using glue tape.

FIG. 3. Photographic view (a) of the interior of the LTHM-XRD diffractometer showing all its components, (b) view after the lead hutch shielding, and (c) view of the sample holder insert. Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 14.139.62.174 On: Fri, 07 Oct

2016 04:55:41

105110-5

Shahee et al.

FIG. 4. Variation of the diffracted intensity of Si (111) peak as a function of applied field ramped from 0 to 8 T. This intensity monitoring has been done to characterize the net adverse effect of SMF on the x-ray source as well as the NaI detector after its magnetic shielding using µ-metal. The measured intensity remains almost constant up to 6 T.

The effectiveness of the total shielding, including both the x-ray tube and the PMT, was characterized by monitoring the intensity count of Si-(111) reflection while the magnet was set to ramp from 0 to 8 T. Fig. 4 shows the result of this field ramping. It can be seen that the intensity remains almost constant upto 6 T; however, from 6 to 8 T the intensity slowly decreases by 6%-8% of its zero-field value. This indicates room for further improvement in the shielding using better quality µ-metal sheet. However this will not cast much effect on the experimental results. It should be noted that since the applied field remains fixed during the 2θ scan and SMF is cylindrically symmetric, this small relative decrease of the intensity remains constant throughout the angular range of the XRD pattern and therefore does not cast any adverse effect on the determination of the crystallographic parameters and relative phase fraction determination, say during a phase transition study. The result in Fig. 4 shows that the setup is reasonably efficiently shielded against the odd effects of the SMF and now be used for structural studies at non-ambient conditions of lowtemperatures (2-300 K) and high magnetic field up to 8 T. The existing literature on high field XRD development shows that people have tried optimizing various detectors for such setups and have found the image plate (IP) detector to be quite suitable against the problems related to SMF. It works without any artifacts even in filed as high as 10 T.20 Therefore, keeping in view the above described results obtained using NaI scintillation detector, it appears imperative to compare these two detectors. The insensitivity of an IP detector against high fields makes it ideal to be used with setups having SMF around it. IP based diffraction setups are mostly designed to work in Debye-Scherrer transmission geometry.18,20 This diffraction geometry does not require movement of any detector or goniometer for data collection and therefore is ideal for diffraction setups using high field. Unlike the case of Mitsui et al.,20 where a cylindrically curved IP has been used, the constrains arising from the use of cryogenics in LTHM XRD setups demand using a flat and reasonably large IP detector. In such

Rev. Sci. Instrum. 87, 105110 (2016)

setups, the positioning reproducibility of the IP is extremely important. This is essential for the accuracy of the measured lattice parameters and also for the best possible resolution. For obtaining the best possible resolution in Debye-Scherrer geometry, one needs to expose the tiniest possible sample volume and the narrowest possible parallel beam. A small sample volume may result in spottiness and weak intensity of the Debye-Scherrer rings. This problem with intensity is overcome by 2π integration along the Debye-Scherrer rigs. The technique to calibrate an IP is very well discussed by Rajiv et al.23 Although the available benefits of an IP detector, i.e., its capability of collecting quick diffraction data, are fully enjoyed with a synchrotron source only, as discussed above, even for lab based LTHM-XRD setup, IP appears to be the best suited because one does not need to bother about the problems related to SMF. On the other hand, NaI scintillation detector has been widely used in laboratory based diffractometers for last several decades. It mostly uses line source of Cu-Kα x-rays and works in reflection geometry. It is a well proven setup providing high resolution, refinable data from flat samples. The intensity problem related to SMF and the PMT in the scintillation detector is totally solvable. The successfully developed and working setups as described in Refs. 6 and 19 and also in the present report do prove this fact. Of course this requires magnetic shielding using µ-metals, which is not at all a tedious task and requires not even 1% of the total effort in developing the complete setup. Depending on the permeability value of the µmetal available, the effective shielding is very much achievable for the SMF value as high as 150 G, i.e., a magnet energized to field as high as 8-10 T. The act of shielding is anyway required even for an IP based setup too. The presence of SMF in the nearby vicinity of the x-ray tube of the IP based setup also needs to be shielded. The above described comparison clearly tells that as far as the maximum applied field in the LTHM XRD is limited to values like 8-10 T, both the IP based setups and the NaI based setups are equally promising.

SETUP CHARACTERIZATION

The proper functioning of the goniometer against the possible artifacts caused by the adverse effects of SMF on the goniometer, source, and the detector has been characterized by collecting powder XRD data of a NIST (National Institute of Standards and Technology) certified standard reference material LaB6 (SRM 660b), as a function of temperature, and of Si powder, as a function of applied field. To demonstrate the field induced structural changes, LTHM XRD of Pr0.5Sr0.5MnO3 (PSMO)24–26 was carried out in zero field cool (ZFC) and field cool (FC) conditions. Figure 5 shows Rietveld refined XRD profile of NIST (SRM 660b) LaB6 powder collected at room temperature. The refined room temperature lattice parameter of LaB6 is found to be 4.1566 Å, which differs only at the 4th place from the reported value of 4.1569 Å. Inset (a) of Fig. 5 shows the expanded view of a peak profile showing well resolved Kα1 and Kα2 lines. The accuracy of the refined lattice parameter and the quality of fit with goodness parameter 1.09 assure high


105110-6

Shahee et al.

Rev. Sci. Instrum. 87, 105110 (2016)

FIG. 5. Rietveld refined room temperature XRD profile of NIST LaB6 (SRM 660b) sample. The inset (a) depicts the instrument resolution function (IRF) and the inset (b) shows the temperature dependence of the LaB6 lattice parameter obtained after Rietveld refinement of the XRD data collected at various temperatures down to 2 K.

quality and accuracy of the data obtained using the developed LTHM-XRD. Inset (b) of Fig. 5 shows the variation of the full width at half maxima (FWHM) of the LaB6 XRD peak profiles as a function of scattering angle. The FWHM has been obtained from the output file of the Rietveld refinement. Since SRM 660b comprises strain free LaB6 crystalline particles of average size of ∼10 µm, no size or strain broadening is expected to contribute to the observed total peak width. Thus the angular variation of FWHM, as shown in the inset (b) of Fig. 5, basically represents the peak broadening contribution arising from the instrumental factor only. The curve in the inset (b) of Fig. 5 thus represents the instrumental resolution function (IRF) of the LTHM-XRD setup. The FWHM remains nearly constant in between 20o and 70◦ of 2θ with an average

FIG. 7. (a) 4.2 K ZFC-XRD data of Pr0.5Sr0.5MnO3 (PSMO) taken at zero and 8 T applied fields. It clearly shows field induced structural phase change of the AFM insulating Fmmm phase of PSMO into FM metallic I4/mcm phase. (b) Comparison of the 4.2 K 8 T FC-XRD and 4.2 K 8 T ZFC-XRD clearly shows the enhancement in the phase fraction of I4/mcm phase of PSMO.

value ∼0.0956◦ with a very shallow minima of 0.096◦, which is very close to the designed value of horizontal acceptance of the long Soller slit (LSS). Slightly lower (by 0.004◦) value of the observed minima of 0.096◦ of the IRF, as compared to the designed value of 0.1◦, may be due to some inevitable minor error in the Soller slit alignment on the diffraction arm or due to some error in its fabrication in the factory itself. Anyway the closeness of these values directly implies that the angular resolution of the setup is mainly governed by the acceptance angle of the Soller slit. However the shallow minima in IRF appears to be a combined effect of small but finite beam divergence (0.025◦) and the sample transparency. The rise of IRF above 70◦ (∼0.125◦ at 110◦) is attributed to the axial divergence of the incident beam. The XRD profiles of LaB6 taken at low-temperatures, down to 2 K, were also refined to obtain accurate lattice parameters. Inset (c) of Fig. 5 shows the temperature variation of the refined lattice parameter of NIST LaB6 (SRM 660b) sample. The above presented results approve the proper functioning of the goniometer. Keeping in view the non-magnetic nature of Si, to assess the possible artifacts caused by the effect of SMF on the diffractometer, in-field diffraction measurements were carried out on Si-powder sample. Figure 6 shows the in-filed XRD data of Si-powder taken at various fields from zero to

FIG. 6. Low-temperature (96 K) XRD data of Silicon powder collected at various magnetic field values. The enlarged view of the peak profiles presented in the insets clearly shows that up to 6 T field, all the XRD profiles exactly superimpose, that is, effectively no cumulative artifact of stray magnetic field is present up to 6 T field. However above 6 T field about 6%-8% decrease in the peak intensity of the profiles was noticed. Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 14.139.62.174 On: Fri, 07 Oct

2016 04:55:41

105110-7

Shahee et al.

Rev. Sci. Instrum. 87, 105110 (2016)

8 T at 96 K. It can be noticed that up to 6 T, the collected patterns superimpose but above 6 T field, ∼6%-8% decrease in the peak intensity is observed. However the angular positions of the peaks remain the same in the whole range of the field variation. The decrease in the peak intensity is due to the effect of SMF on the detector. SETUP APPLICATION

Figure 7(a) shows 4.2 K ZFC XRD profiles of Pr0.5Sr0.5 MnO3 (PSMO) taken at zero and 8 T fields. The peaks in the ZFC XRD correspond to (400), (040), and (004) of Fmmm antiferromagnetic (AFM) insulating phase coexisting with (220) and (004) reflections of about 5% of the high-temperature ferromagnetic (FM) I4/mcm phase. It can be noticed that on application of 8 T field, the peaks corresponding to Fmmm phase get suppressed and that of I4/mcm phase get enhanced to a relative phase fraction of 55%-45%, respectively. This fraction was found to further increase to 80%-20% in the FC XRD data. This observation clearly shows the occurrence of field induced structural phase transition of Fmmm AFM insulating phase into I4/mcm FM metallic phase in PSMO. The increasing phase fraction of the I4/mcm FM metallic phase

basically gives rise to the large magnetoresistance effect in PSMO. To further check the capability of the setup, we carried out LTHM-XRD studies on Nd0.49Sr0.51MnO3−δ (NSMO)27,28 sample. Figure 8(a) shows the ZFC XRD data at 150 K taken during increasing-decreasing field cycles from 0 to 8 T. It can be seen that as field increases, the (220) and (004) reflections of the low-temperature phase of I4/mcm slowly get suppressed and new (220) and (004) reflections, which correspond to high temperature phase of I4/mcm, slowly emerge. This transformation was fully reversible. It was analyzed that at 150 K NSMO undergoes a reversible isostructural phase transition from a strong Jahn-Teller (JT) distorted I4/mcm phase to a weak JT-distorted I4/mcm phase on the application of field. Similar reversibility was seen also in the magnetoresistance of the NSMO sample. However at temperatures below 100 K, NSMO gets hardened against the applied field and no significant phase change was seen; see Figs. 8(b) and 8(c). Details of this study will be published elsewhere. To check the capability of the developed LTHM-XRD setup to deliver quantitative structural phase information as a function of applied filed, we carried out field dependent low-temperature diffraction studies on charge-ordered (CO)

FIG. 8. Magnetic field dependent low-temperature structural phase transition studies on Nd0.49Sr0.51MnO3−δ (NSMO). At ∼150 K, NSMO shows a reversible structural phase transition from a strong JT-distorted tetragonal phase to a weak JT-distorted tetragonal phase. At lower temperatures of 100 K and 5 K, it becomes hardened against the applied field. Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 14.139.62.174 On: Fri, 07 Oct

2016 04:55:41

105110-8

Shahee et al.

Rev. Sci. Instrum. 87, 105110 (2016)

FIG. 9. Magnetic field dependent low-temperature structural phase transition studies on La0.175Pr0.45Ca0.375MnO3 sample. Increase in the minimum required field value for the P21/m CO AFM to Pnma FM metallic phase transformation can be seen. Nearly complete irreversibility of this transformation at 2 K shows the field induced devitrification of the kinetically arrested glass-like phase of P21/m CO AFM phase.

La0.175Pr0.45Ca0.375MnO3 (LPCMO). LPCMO is a celebrated electronic as well as magnetic phase separated manganite,29–31 which has undergone a variety of studies.32–35 It is reported to show dynamic to frozen transition36,37 of its phase separated state below ∼39 K and has been probed by various techniques. However no such study involving field induced structural phase change has been done till date. Figs. 9(a) and 9(b) show field-induced irreversible phase transition of LPCMO from P21/m CO AFM to Pnma FM metallic phase. This is quite clear from the field dependent integrated intensity plot of (202) and (040) reflections of Pnma phase. At 2 K, the transformed Pnma phase remains stable because it corresponds to the ground state phase. The irreversibility slowly vanishes at higher temperatures. Nearly complete irreversibility of the field induced transformation at 2 K shows the field induced devitrification of the kinetically frozen glasslike state of the phase-separated P21/m CO AFM/FM phase to stable Pnma FM phase. Based on the field dependent lowtemperature structural transition studies carried out at various temperatures and fields (during increasing field cycle), an HT phase diagram is constructed for LPCMO. This is shown in Fig. 10. It can be clearly seen that the required minimum field for P21/m to Pnma phase transformation undergoes nonmonotonous change. It shows a minimum at ∼39 K. The increase of the minimum required field value for the P21/m to Pnma transformation shows initiation of the freezing of the P21/m to Pnma phase transformation, resulting in a frozen phase separated phase. The freezing causes magnetic stiffness of the P21/m CO AFM to Pnma FM metallic phase against applied field. The H-T phase diagram constructed purely based

FIG. 10. H-T phase diagram constructed based on low-temperature structural phase transition studies on La0.175Pr0.45Ca0.375MnO3 done during increasing field. This H-T phase diagram constructed based on structural measurements exactly matches with the H-T phase diagram constructed based on magnetic measurements.

on our structural measurements exactly matches with the H-T phase diagram constructed based on magnetic measurements reported by Ghivelder et al.36,37 CONCLUSION

A low-temperature (2-300 K) and high-magnetic field (8 T), powder x-ray diffraction setup (LTHM-XRD) using Cu-Kα x-rays has been developed around an 18 kW rotating


105110-9

Shahee et al.

anode source at UGC-DAE CSR, Indore, India. The instrumental resolution of the setup is about 0.1◦. The powder XRD data collected using this setup gets very well refined using Rietveld refinement and reproduces the reported lattice parameters. To monitor its capability towards field induced phase-transitions studies, LTHM-XRD experiments have been performed on few manganites with temperature stability better than 20 mK and absolute accuracy of ±0.1 K. Field induced transition from AFM insulating Fmmm phase to FM metallic I4/mcm phase has been observed in Pr0.5Sr0.5MnO3. Field induced suppression of JT-distortion has been observed in I4/mcm phase of Nd0.49Sr0.51MnO3−δ. Dynamic to frozen phase separated state has been structurally probed in La0.175Pr0.45Ca0.375MnO3 and based on that, a H-T phase diagram has been constructed, which exactly matches with the reported H-T phase diagram constructed based on magnetic measurements. ACKNOWLEDGMENTS

Authors thank the Director, Dr. A. K. Sinha, the Centre Director, Dr. V. Ganesan, and the former Centre Director, Professor A. Gupta for their constant support and encouragement during the development of LTHM-XRD facility. The authors also like to thank Dr. A. Banerjee for useful discussion regarding stray magnetic field shielding of the detector and the x-ray tube. Thanks to Dr. A. K. Sinha, Dr. S. K. Rai, and Dr. Archana Sagdeo, RRCAT, Indore for discussion and providing help during IRF measurement. Thanks are also due to Dr. R. Rawat and Er. P. Sarvanan and cryogenics team for their help during magnet cooling. 1H.

Ohsumi, K. Tajima, N. Wakabayashi, Y. Shimodo, K. Kamishima, and T. Goto, J. Phys. Soc. Jpn., Part 1 66, 1896 (1997). 2H. Ohsumi and K. Tajima, J. Phys. Soc. Jpn., Part 1 67, 1883 (1998). 3S. Shimomura, K. Tajima, N. Wakabayashi, S. Kobayashi, H. Kuwahara, and Y. Tokura, J. Phys. Soc. Jpn., Part 1 68, 1943 (1999). 4Y. Ma, S. Awaji, K. Watanabe, M. Matsumoto, and N. Kobayashi, Solid State Commun. 113, 671 (2000). 5V. K. Pecharsky, A. P. Holm, K. A. Gschneidner, and R. Rink, Phys. Rev. Lett. 91, 197204 (2003). 6A. P. Holm, V. K. Pecharsky, K. A. Gschneidner, R. Rink, and M. N. Jirmanus, Rev. Sci. Instrum. 75, 1081 (2004). 7V. Kiryukhin, B. Keimer, J. P. Hill, S. M. Coad, and D. McK. Paul, Phys. Rev. B 54, 7269 (1996). 8K. Katsumata, Phys. Scr. 71, CC7–CC13 (2005). 9Y. H. Matsuda, Y. Ueda, H. Nojiri, T. Takahashi, T. Inami, K. Ohwada, Y. Murakami, and T. Arima, Physica B 346, 519 (2004). 10T. Inami, K. Ohwada, Y. H. Matsuda, Y. Ueda, H. Nojiri, Y. Murakami, T. Arima, H. Ohta, W. Zhang, and K. Yoshimura, Nucl. Instrum. Methods Phys. Res., Sect. B 238, 233 (2005).

Rev. Sci. Instrum. 87, 105110 (2016) 11Y.

H. Matsuda, T. Inami, K. Ohwada, Y. Murata, H. Nojiri, Y. Murakami, H. Ohta, W. Zhang, and K. Yoshimura, J. Phys. Soc. Jpn., Part 1 75, 024710 (2006). 12P. Frings, J. Vanacken, C. Detlefs, F. Duc, J. E. Lorenzo, M. Nardone, J. Billette, A. Zitouni, W. Bras, and G. L. J. A. Rikken, Rev. Sci. Instrum. 77, 063903 (2006). 13Z. Islam, J. P. C. Ruff, H. Nojiri, Y. H. Matsuda, K. A. Ross, B. D. Gaulin, Z. Qu, and J. C. Lang, Rev. Sci. Instrum. 80, 113902 (2009). 14J. Billette, F. Duc, P. Frings, M. Nardone, A. Zitouni, C. Detlefs, T. Roth, W. Crichton, J. E. Lorenzo, and G. L. J. A. Rikken, Rev. Sci. Instrum. 83, 043904 (2012). 15Z. Islam, D. Capatina, J. P. C. Ruff, R. K. Das, E. Trakhtenberg, H. Nojiri, Y. Narumi, U. Welp, and P. C. Canfield, Rev. Sci. Instrum. 83, 035101 (2012). 16Y. H. Matsuda and T. Inami, J. Phys. Soc. Jpn., Part 1 82, 021009 (2013). 17F. Duc, X. Fabrèges, T. Roth, C. Detlefs, P. Frings, M. Nardone, J. Billette, M. Lesourd, L. Zhang, A. Zitouni, P. Delescluse, J. Béard, J. P. Nicolin, and G. L. J. A. Rikken, Rev. Sci. Instrum. 85, 053905 (2014). 18S. Wang, A. E. Kovalev, A. V. Suslov, and T. Siegrist, Rev. Sci. Instrum. 86, 123902 (2015). 19Y. Mitsui, K. Koyama, and K. Watanabe, Sci. Technol. Adv. Mater. 10, 014612 (2009). 20Y. Mitsui, K. Koyama, K. Takahashi, and K. Watanabe, Rev. Sci. Instrum. 82, 125104 (2011). 21R. K. Das, Z. Islam, J. P. C. Ruff, R. P. Sawh, R. Weinstein, P. C. Canfield, J.-W. Kim, and J. C. Lang, Rev. Sci. Instrum. 83, 065103 (2012). 22L. Brzozowski, A. Ganguly, M. Pop, Z. Wen, R. Bennett, R. Fahrig, and J. A. Rowlands, Med. Phys. 33, 3033 (2006). 23P. Rajiv, B. Hinrichsen, R. Dinnebier, M. Jansen, and M. Joswig, Powder Diffr. 22, 3 (2007). 24Y. Tomioka, A. Asamitsu, Y. Moritomo, H. Kuwahara, and Y. Tokura, Phys. Rev. Lett. 74, 25 (1995). 25A. Banerjee, A. K. Pramanik, K. Kumar, and P. Chaddah, J. Phys.: Condens. Matter 18, L605–L611 (2006). 26R. Mahendiran, M. R. Ibarra, C. Marquina, B. Garcia-Landa, L. Morellon, A. Maignan, B. Raveau, A. Arulraj, and C. N. R. Rao, Appl. Phys. Lett. 82, 242 (2003). 27H. Kuwahara, T. Okuda, Y. Tomioka, A. Asamitsu, and Y. Tokura, Phys. Rev. Lett. 82, 21 (1999). 28R. Rawat, K. Mukherjee, K. Kumar, A. Banerjee, and P. Chaddah, J. Phys.: Condens. Matter 19, 256211 (2007). 29M. Uehara, S. Mori, C. H. Che, and S.-W. Cheong, Nature 399, 560 (1999). 30V. Kiryukhin, B. G. Kim, V. Podzorov, S.-W. Cheong, T. Y. Koo, J. P. Hill, I. Moon, and Y. H. Jeong, Phys. Rev. B 63, 024420 (2000). 31T. Z. Ward, J. D. Budai, Z. Gai, J. Z. Tischler, L. Yin, and J. Shen, Nat. Phys. 5, 885 (2009). 32J. Q. He, V. Volkov, T. Asaka, S. Chaudhuri, R. C. Budhani, and Y. Zhu, Phys. Rev. B 82, 224404 (2010). 33D. D. Sarma, D. Topwal, U. Manju, S. R. Krishnakumar, M. Bertolo, S. La Rosa, G. Cautero, T. Y. Koo, P. A. Sharma, S.-W. Cheong, and A. Fujimori, Phys. Rev. Lett. 93, 097202 (2004). 34Mark H. Burkhardt, M. A. Hossain, S. Sarkar, Y.-D. Chuang, A. G. Cruz Gonzalez, A. Doran, A. Scholl, A. T. Young, N. Tahir, Y. J. Choi, S.-W. Cheong, H. A. Du¨rr, and J. Stohr, Phys. Rev. Lett. 108, 237202 (2012). 35C. Israel, W. Wu, and A. de Lozanne, Appl. Phys. Lett. 89, 032502 (2006). 36L. Ghivelder, R. S. Freitas, M. G. das Virgens, M. A. Continentino, H. Martinho, L. Granja, M. Quintero, G. Leyva, P. Levy, and F. Parisi, Phys. Rev. B 69, 214414 (2004). 37L. Ghivelder and F. Parisi, Phys. Rev. B 71, 184425 (2005).
