- Email: [email protected]
In Situ X-ray Diffraction Study of the Formation of Fe(Se,Te) from Various Precursors
Available online at www.sciencedirect.com
Physics Procedia 36 (2012) 600 – 605
Superconductivity Centennial Conference
In situ X-ray diffraction study of the formation of Fe(Se,Te) from various precursors J.-C. Grivela*, Y. Zhaoa, A.C. Wulffa, P.G.A.P. Pallewattaa, J. Bednarčíkb and M.v. Zimmermannb a
Materials Research Division, Risø National Laboratory for Sustainable Energy, Technical University of Denmark, Frederiksborgvej 399, DK – 4000 Roskilde, Denmark b Hamburger Synchrotronstrahlungslabor, HASYLAB at Deutsches Elektronen-Synchrotron DESY, Notkestrasse 85, D – 22603 Hamburg, Germany
Abstract The formation of the FeSe0.5Te0.5 phase was studied by means of high energy synchrotron x-ray diffraction. The precursors consisted of Fe, Se and Te or Se0.5Te0.5 powder mixtures and were encased in a metal (Cu/Nb) composite sheath to prevent evaporation of Se and Te during high temperature processing. In all cases (Fe – Se – Te ternary mixture; Fe - Se0.5Te0.5 binary mixtures with two different Fe particle sizes) the ternary alloy forms via Fe(Se,Te)2 and Fe3(Se,Te)4 intermediate phases. When unreacted Se and Te powders are used in the precursor, partial Se1-xTex alloying takes place during heating prior and during the formation of the intermediate phases. As the alloying is incomplete, the resulting Fe(Se,Te) phase is not homogeneous. Using pre-alloyed Se0.5Te0.5 allows a better control of the reaction although homogeneisation also occurs in the Fe(Se,Te) phase as a consequence of the phase equilibria of the Se – Te system. The grain size of the starting Fe powder has no influence on the reaction path for the grain sizes used in the present study. However, the reaction rate for Fe(Se,Te) formation is clearly sensitive to this parameter.
© 2012 2011 Published Published by by Elsevier Elsevier B.V. Ltd. Selection Rogalla and © Selection and/or and/or peer-review peer-review under under responsibility responsibility of of Horst the Guest Editors. Peter Kes. Keywords: Fe(Se,Te), phase formation, wires, synchrotron, in-situ x-ray diffraction
1. Introduction Among the recently discovered Fe-based superconductors, β-FeSe not only possesses the simplest structure but has as well the advantage of not containing arsenic, which is interesting in view of safety if this class of superconducting materials happens to compete with other types of superconductors for the large scale manufacture and use of wires in power applications. The critical temperature (T c) of β-FeSe is not especially high (about 10K) but it can be enhanced to 15K by doping with Te [1]. Furthermore, the
1875-3892 © 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of the Guest Editors. doi:10.1016/j.phpro.2012.06.175
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
upper critical field Hc2(0) of the Fe(Se1-xTex) phase can reach values in excess of 50T [2]. A few attempts at preparing FeSe wires have already been reported but the critical current densities remain low [3,4]. Besides possible intrinsic limitations related for example to grain boundary coupling, it appears that improving the microstructure of such wires might play a role in enhancing their superconducting performance. A detailed description of the phase and microstructure evolution taking place inside the core of the wires can play a significant role in optimizing some processing parameters and obtaining a superconducting core of better quality. Recently, we reported two phase formation studies performed in-situ by means of hard x-ray synchrotron diffraction in the FeSe and the related FeTe systems respectively [5,6]. It was found that the FeTe phase forms according to a relatively simple reaction scheme involving only the FeTe2 phase as an intermediate. In contrast, the FeSe reaction path involves three intermediate products, which appear in the sequence FeSe2, Fe3Se4 and Fe7Se8. In this contribution we present new investigations on the phase formation of the FeSe0.5Te0.5 compound starting from different precursors. Like for our previous studies, the precursor powder mixtures were packed into a composite Cu/Nb metallic sheath to prevent losses of Se or Te by evaporation or sublimation during the heat treatment. Using a high-energy synchrotron beam allows following the reactions occurring inside the metal sheath in transmission geometry.
2. Experimental details The starting reagents consisted of metal powders: Fe (99.9% purity, <10μm), Se (99.9%, -200 mesh) and Te (99.99%, -325 mesh ). In one case (see section 3.3) another Fe powder batch was used with 99+% purity and -200 mesh (<74μm) grain size in order to study the effect of Fe particle size on the reaction mechanism and kinetics. In all other samples the <10μm Fe powder was used. For one sample, these powders were directly mixed together with a FeSe0.5Te0.5 nominal composition, whereas for the other samples Se and Te were first reacted together at 200°C for 20h, followed by 250°C for 20h in order to obtain a SeTe alloy prior to mixing with Fe powder with an overall FeSe 0.5Te0.5 composition. Powder mixing was performed by low-energy ball milling in 4 cm diameter polyethylene jars for 24h with NiO balls. The mixed powders were packed into a Nb tube, which was then itself inserted into a Cu tube with larger diameter. The two composite samples were mechanically deformed by means of groove-rolling down to a wire with about 3 mm2 cross section area. The in-situ diffraction measurements were conducted at the high energy x-ray beamline BW5, located at the storage ring DORIS III at DESY. The photon energy of the incident beam was 80 keV. Short pieces (≈ 4cm length) cut from longer wires were clamped in a high-temperature steel holder inserted in a quartz tube. The sample holder assembly was placed in a dedicated high-temperature furnace equipped with Kapton windows and a stainless steel heat shield with holes for beam entrance and exit. The samples were maintained in a flow of Ar (d 0.5 ppm residual O2) during the runs. A heating rate of 2°C/min was used to reach the maximum temperature. A thermometer was situated close to the samples and in contact with the steel holder. The acquisition time was 2s for each diffraction pattern. A beam cross-section of 1x1 mm2 was chosen to probe the ceramic core throughout the diameter of the wires. Absorption scans were used to position the samples in the beam. Diffraction patterns were recorded on a two-dimensional image plate and evaluated using the fit2d software package [7]. The intensity of the signal was normalised to the synchrotron positron beam current value, which varies with time during the experiment.
601
602
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
3. Results and discussion 3.1. Fe – Se0.5Te0.5 mixture In the as deformed wire, only the Se0.5Te0.5 and Fe phases are detected in the diffraction patterns. The appearance of other compounds is evidenced from 248°C, i.e. about 60°C below the partial melting temperature of Se0.5Te0.5. According to the Se – Te phase diagramme determined by Ghosh et al. [8], the composition used in our sample crosses the solidus at 310°C. As the temperature further increases, a Serich liquid is expected to form, whereas the composition of the solid slowly becomes richer in Te. This feature is illustrated in Fig. 1, where the (101) reflection of the Se0.5Te0.5 phase (now better described as Se0.5-xTe0.5+x) clearly shifts towards higher d-spacing values in the temperature interval separating the solidus and liquidus (about 360°C) of the system. The new phases forming in the powder core of the wire are Fe(Se,Te)2 and Fe3(Se,Te)4 as well as a minor amount of Fe(Se,Te). In a Fe-Se mixture, the corresponding FeSe2 and Fe3Se4 phases are also forming as soon as Se melts [5]. However, no equivalent to the Fe7Se8 phase, which was observed at increasing temperatures in the FeSe phase formation study, is evidenced in the present sample. The intensity of the Fe(Se,Te)2 phase increases up to 365°C before decreasing, whereas the amount of Fe3(Se,Te)4 appears to increase up to 388°C. The intensity of the reflections of the latter phase then starts to decrease. In contrast the amount of Fe(Se,Te) increases steadily up to 635°C and is the only phase detected in the samples at the end of the heating ramp (690°C). Whereas a structural phase transition was observed in FeSe between 450°C and 500°C, the tetragonal Fe(Se,Te) phase formed in the present sample does not show any evidence for a transformation.
F
Q
o
179 C o 202 C o 225 C o 249 C o 272 C o 295 C o 318 C o 341 C
Nb
Intensity [a.u.]
Cu
Q
| T
2,4
T
|
|
J
2,2
(101)
2,6
2,8
3,0
3,2
3,4
d - spacing [Angstrom]
Fig. 1. Radially integrated diffraction patterns recorded at different temperatures on a wire containing a starting mixture of Fe and Se0.5Te0.5 powders. F Fe, z SeTe, Fe(Se,Te), | Fe(Se,Te)2,d Fe3(Se,Te)4.
603
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
d-spacing [Angstrom]
As shown in Fig.2 for the (111) reflection of Fe(Se,Te), the peaks of this phase first shift rapidly towards higher d-spacing values at a fast and non linear rate. However, from 450°C, this shift takes place at a slower rate and becomes linear versus temperature. This can be interpreted as the consequence of a variation of composition. Due to the fact that the liquid phase formed at the beginning of the fusion of the the Se0.5Te0.5 phase is Se-rich, the Fe(Se,Te) formed during the initial stage is expected to be Te-deficient and thus have shorter lattice parameters than the target FeSe0.5Te0.5 composition. However, as the temperature increases, the liquid becomes enriched in Te and induces in turn a Te enrichment of the Fe(Se,Te) phase until the overall composition of the phase reaches the nominal atomic ratio. At this point, the temperature variation of the d-spacing becomes linear and only reflects normal thermal expansion. The full width at half maximum (FWHM) of the Fe(Se,Te) reflections increases during this reaction step up to 450°C. This indicates that the Se and Te distribution in the newly formed phase is inhomogeneous. Shortly before the phase enters the linear thermal expansion regime, the FWHM decreases, suggesting that solid-state diffusion allows some homogenization, which may continue after the reaction is finished. o
412 C o 447 C o 470 C o 494 C o 518 C o 542 C o 566 C o 589 C
Intensity [a.u.]
Fe(Se,Te) (111)
Fe3(Se,Te)4
FWHM
Fe(Se,Te)2
2,5
2,6
d-spacing [Angstrom]
2,7
2,57 2,56 2,55 2,54 2,53 2,52 2,51 0,055 0,050 0,045 0,040 0,035 0,030 0,025 0,020 0,015 0,010
350
400
450
500
550
Temperature [oC]
Fig. 2. (left) Radially integrated diffraction patterns recorded at different temperatures on a wire containing a starting mixture of Fe and Se0.5Te0.5 powders. (right) Evolution of the (111) peak position of Fe(Se,Te) as well as its FWHM as a function of temperature. The broken red line is a guide to the eye.
3.2. Fe – Se - Te mixture In the wire containing a mixture of Fe, Se and Te powders in Fe:Se:Te = 1:0.5:0.5 nominal atomic ratio, the situation is more complex. From about 200°C, i.e. prior to the formation of new compounds, there is evidence for Se-Te inter-diffusion as illustrated in Fig.3a, where the background level between the respective (101) peaks of Se and Te is clearly increasing with temperature. However, complete alloying is by far not complete before the start of the reaction of the chalcogen elements with Fe. At 250°C, Fe(Se,Te)2 and Fe3(Se,Te)4 have appeared as in the case of the Fe – Se0.5Te0.5 powder mixture but in the present case the reflections are systematically shifted to lower d-spacing values, which indicates that these new phases are richer in Se than it was the case at the beginning of the reaction in the other sample. The temperature for the appearance of these phases is approximately the same in both cases. The variation of the cell parameters of Te as a function of temperature is non monotonic as illustrated in Fig. 3b by the behavior of the (101) reflection of this phase. Starting from 215°C, the expansion is no more linear and the d-spacing even decreases between 240°C and 280°C before increasing sharply again. This departure from linear behavior coincides with a significant acceleration of the lattice expansion rate in Se (see peak shift in Fig. 3b) that is probably related to Te diffusion into the Se, which has not yet molten and has for effect to increase the melting point of the remaining phase. The melting temperature of pure Se is of 221°C and increases upon alloying with Te [8].
604
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
The fast consumption of the Se that has not been incorporated into the Se1-xTex alloy results in the formation of a Se-rich Fe(Se,Te) phase. As shown in Fig.3c in the case of the (111) reflection of this phase, a shoulder appears on the high d-spacing side after about 100 minutes at 350°C. A similar feature is observed for all other Fe(Se,Te) peaks and suggests that a Te-rich phase is formed at that stage from the Se-Te alloy that formed prior to the melting of the un-alloyed Se. This phenomenon is not quite similar to that observed in the Fe + Se0.5Te0.5 sample (section 3.1). Whereas in the latter case the compositional change was progressive, in the present sample it appears that two separate phase formation reactions occur. The pairs of reflection did not merge during the course of the experiment, showing that homogeneisation of the composition is not straightforward at 350°C. 3,250
Fe3(Se,Te)4
Te
3,245 3,240
d-spacing [Angstrom]
Intensity [a.u.]
a
Se
Fe(Se,Te)2
c
b
72 min 101 min 144 min 180 min 223 min 286 min 330 min
Te
3,235 3,230
Intensity [a.u.]
o
151 C o 202 C o 228 C o 254 C o 280 C o 292 C o 305 C o 331 C
3,225 3,220 3,12 3,10 3,08 3,06
Se
3,04 3,02 3,00
2,6
2,8
3,0
3,2
50
100
d - spacing [Angstrom]
150
200
250
300
350
2,40
2,45
2,50
2,55
d-spacing [Angstrom]
Temperature [oC]
Fig. 3. (a) Radially integrated diffraction patterns recorded at different temperatures on a wire containing a starting mixture of Fe, Se and Te powders. (b) Evolution of the (101) peak position of Se and Te as a function of temperature. (c) Detail of the diffraction pattern showing the coexistence of Se-rich and Te-rich Fe(Se,Te) at various sintering times at 350°C.
3.3. Effect of Fe particle size
Integrated intensity [a.u.]
The effect of Fe particle size has been studied in a wire containing a mixture of Fe (< 10 μm or < 74 μm particle sizes) and Se0.5Te0.5 powders. No significant variations could be evidenced in the reaction path, i.e. the same intermediate phases were formed in the same sequence, in both samples. However, a clear difference was observed in the kinetics of Fe(Se,Te) phase formation. Fig. 4 shows a clear shift to higher
250
300
350
400
450
500
550
o
Temperature [ C] Fig. 4. Integrated intensity of the reflections (101) of (Se,Te) - diamonds, (111) of Fe(Se,Te)2 – triangles, and (111) of Fe(Se,Te) – squares, as a function of temperature for the samples prepared from Se0.5Te0.5 and Fe mixtures with Fe particle size of <10μm(filled symbols) and <74μm (open symbols).
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
temperatures in the sample with larger Fe particles for the Se0.5Te0.5 phase disappearance; the formation and consumption of the Fe(Se,Te)2 compound and the formation of Fe(Se,Te). The Fe3(Se,Te)4 phase (not shown in Fig.4 for clarity) follows a trend similar to that of Fe(Se,Te) 2. 4. Conclusion The formation process of the FeSe0.5Te0.5 phase is affected by the precursor materials used for its preparation, although the reaction sequence appears to be unchanged and characterized by the formation of Fe(Se,Te)2 and Fe3(Se,Te)4 intermediates in all cases studied. When Fe is mixed with pre-alloyed Se0.5Te0.5 powders, the reaction is initiated by the partial melting of the Se0.5Te0.5 phase. The formation of a Se-rich liquid results in Se-rich phases including Fe(Se,Te), which become enriched in Te as the melting of Se0.5Te0.5 proceeds. In such kind of precursor mixtures, the effect of the particle size of the starting Fe powder was studied and found to strongly influence the reaction kinetics as larger particles shift the reaction sequence to higher temperatures. When unreacted Se and Te powders are used in the precursor, partial Se1-xTex alloying takes place during heating prior to and during the formation of the intermediate phases. Since the alloying is incomplete, Se-rich Fe(Se,Te) forms at relatively low temperature prior to the melting of the Te-rich Se-Te alloy. This results in an inhomogeneous Fe(Se,Te) phase. Acknowledgements The authors gratefully acknowledge the technical assistance of M. Wichmann and L. Lorentzen as well as financial support from DANSCATT. References [1] Yeh K-W, Huang T-W, Huang I-L, Chen T-K, Hau F-C, Hsu F-C, Wu PM, Lee Y-C, Chu YY, Chen C-L, Luo J-Y, Yan DC, Wu M-K. Tellurium substitution effect on superconductivity of the alpha-phase iron selenide. Europhys Lett 2008;84:37002 (4pp). [2] Tsurkan V, Deisenhofer J, Günther A, Kant Ch, Klemm M, Krug von Nidda HA, Schrettle F, Loidl A. Physical properties of FeSe0.5Te0.5 single crystals grown under different conditions. Eur Phys J B 2011;79:289-299. [3] Mizuguchi Y, Deguchi K, Tsuda S, Yamaguchi T, Takeya H, Kumakura H, Takano Y. Fabrication of the iron-based superconducting wire using Fe(Se,Te). Appl Phys Exp 2009;2:083004 (3pp). [4] Gao Z, Qi Y, Wang L, Wang D, Zhang X, Yao C, Ma Y. Superconducting properties of FeSe wires and tapes prepared by a gas diffusion technique. Supercond Sci Technol 2011;24:065022 (4pp). [5] Grivel J-C, Wulff AC, Zhao Y, Andersen NH, Bednarčík J, Zimmermann Mv. In-situ observation of the formation of FeSe. Supercond Sci Technol 2011;24:015007 (4p) [6] Grivel J-C, Wulff AC, Zhao Y, Bednarčík J, Zimmermann Mv. In-situ study on the formation of FeTe. J Mater Sci 2011;46:4540-4544 [7] Hammersley AP, Svensson SO, Hanfland M, Fitch AN, Häusermann D. Two-dimensional detector software: From real detector to idealised image or two-thete scan. High Pressure Research 1996;14:235-248 [8] Ghosh G, Lukas HL, Delaey L. A thermodynamic assessment of the Se-Te system. CALPHAD 1988;12;295-299
605
Physics Procedia 36 (2012) 600 – 605
Superconductivity Centennial Conference
In situ X-ray diffraction study of the formation of Fe(Se,Te) from various precursors J.-C. Grivela*, Y. Zhaoa, A.C. Wulffa, P.G.A.P. Pallewattaa, J. Bednarčíkb and M.v. Zimmermannb a
Materials Research Division, Risø National Laboratory for Sustainable Energy, Technical University of Denmark, Frederiksborgvej 399, DK – 4000 Roskilde, Denmark b Hamburger Synchrotronstrahlungslabor, HASYLAB at Deutsches Elektronen-Synchrotron DESY, Notkestrasse 85, D – 22603 Hamburg, Germany
Abstract The formation of the FeSe0.5Te0.5 phase was studied by means of high energy synchrotron x-ray diffraction. The precursors consisted of Fe, Se and Te or Se0.5Te0.5 powder mixtures and were encased in a metal (Cu/Nb) composite sheath to prevent evaporation of Se and Te during high temperature processing. In all cases (Fe – Se – Te ternary mixture; Fe - Se0.5Te0.5 binary mixtures with two different Fe particle sizes) the ternary alloy forms via Fe(Se,Te)2 and Fe3(Se,Te)4 intermediate phases. When unreacted Se and Te powders are used in the precursor, partial Se1-xTex alloying takes place during heating prior and during the formation of the intermediate phases. As the alloying is incomplete, the resulting Fe(Se,Te) phase is not homogeneous. Using pre-alloyed Se0.5Te0.5 allows a better control of the reaction although homogeneisation also occurs in the Fe(Se,Te) phase as a consequence of the phase equilibria of the Se – Te system. The grain size of the starting Fe powder has no influence on the reaction path for the grain sizes used in the present study. However, the reaction rate for Fe(Se,Te) formation is clearly sensitive to this parameter.
© 2012 2011 Published Published by by Elsevier Elsevier B.V. Ltd. Selection Rogalla and © Selection and/or and/or peer-review peer-review under under responsibility responsibility of of Horst the Guest Editors. Peter Kes. Keywords: Fe(Se,Te), phase formation, wires, synchrotron, in-situ x-ray diffraction
1. Introduction Among the recently discovered Fe-based superconductors, β-FeSe not only possesses the simplest structure but has as well the advantage of not containing arsenic, which is interesting in view of safety if this class of superconducting materials happens to compete with other types of superconductors for the large scale manufacture and use of wires in power applications. The critical temperature (T c) of β-FeSe is not especially high (about 10K) but it can be enhanced to 15K by doping with Te [1]. Furthermore, the
1875-3892 © 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of the Guest Editors. doi:10.1016/j.phpro.2012.06.175
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
upper critical field Hc2(0) of the Fe(Se1-xTex) phase can reach values in excess of 50T [2]. A few attempts at preparing FeSe wires have already been reported but the critical current densities remain low [3,4]. Besides possible intrinsic limitations related for example to grain boundary coupling, it appears that improving the microstructure of such wires might play a role in enhancing their superconducting performance. A detailed description of the phase and microstructure evolution taking place inside the core of the wires can play a significant role in optimizing some processing parameters and obtaining a superconducting core of better quality. Recently, we reported two phase formation studies performed in-situ by means of hard x-ray synchrotron diffraction in the FeSe and the related FeTe systems respectively [5,6]. It was found that the FeTe phase forms according to a relatively simple reaction scheme involving only the FeTe2 phase as an intermediate. In contrast, the FeSe reaction path involves three intermediate products, which appear in the sequence FeSe2, Fe3Se4 and Fe7Se8. In this contribution we present new investigations on the phase formation of the FeSe0.5Te0.5 compound starting from different precursors. Like for our previous studies, the precursor powder mixtures were packed into a composite Cu/Nb metallic sheath to prevent losses of Se or Te by evaporation or sublimation during the heat treatment. Using a high-energy synchrotron beam allows following the reactions occurring inside the metal sheath in transmission geometry.
2. Experimental details The starting reagents consisted of metal powders: Fe (99.9% purity, <10μm), Se (99.9%, -200 mesh) and Te (99.99%, -325 mesh ). In one case (see section 3.3) another Fe powder batch was used with 99+% purity and -200 mesh (<74μm) grain size in order to study the effect of Fe particle size on the reaction mechanism and kinetics. In all other samples the <10μm Fe powder was used. For one sample, these powders were directly mixed together with a FeSe0.5Te0.5 nominal composition, whereas for the other samples Se and Te were first reacted together at 200°C for 20h, followed by 250°C for 20h in order to obtain a SeTe alloy prior to mixing with Fe powder with an overall FeSe 0.5Te0.5 composition. Powder mixing was performed by low-energy ball milling in 4 cm diameter polyethylene jars for 24h with NiO balls. The mixed powders were packed into a Nb tube, which was then itself inserted into a Cu tube with larger diameter. The two composite samples were mechanically deformed by means of groove-rolling down to a wire with about 3 mm2 cross section area. The in-situ diffraction measurements were conducted at the high energy x-ray beamline BW5, located at the storage ring DORIS III at DESY. The photon energy of the incident beam was 80 keV. Short pieces (≈ 4cm length) cut from longer wires were clamped in a high-temperature steel holder inserted in a quartz tube. The sample holder assembly was placed in a dedicated high-temperature furnace equipped with Kapton windows and a stainless steel heat shield with holes for beam entrance and exit. The samples were maintained in a flow of Ar (d 0.5 ppm residual O2) during the runs. A heating rate of 2°C/min was used to reach the maximum temperature. A thermometer was situated close to the samples and in contact with the steel holder. The acquisition time was 2s for each diffraction pattern. A beam cross-section of 1x1 mm2 was chosen to probe the ceramic core throughout the diameter of the wires. Absorption scans were used to position the samples in the beam. Diffraction patterns were recorded on a two-dimensional image plate and evaluated using the fit2d software package [7]. The intensity of the signal was normalised to the synchrotron positron beam current value, which varies with time during the experiment.
601
602
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
3. Results and discussion 3.1. Fe – Se0.5Te0.5 mixture In the as deformed wire, only the Se0.5Te0.5 and Fe phases are detected in the diffraction patterns. The appearance of other compounds is evidenced from 248°C, i.e. about 60°C below the partial melting temperature of Se0.5Te0.5. According to the Se – Te phase diagramme determined by Ghosh et al. [8], the composition used in our sample crosses the solidus at 310°C. As the temperature further increases, a Serich liquid is expected to form, whereas the composition of the solid slowly becomes richer in Te. This feature is illustrated in Fig. 1, where the (101) reflection of the Se0.5Te0.5 phase (now better described as Se0.5-xTe0.5+x) clearly shifts towards higher d-spacing values in the temperature interval separating the solidus and liquidus (about 360°C) of the system. The new phases forming in the powder core of the wire are Fe(Se,Te)2 and Fe3(Se,Te)4 as well as a minor amount of Fe(Se,Te). In a Fe-Se mixture, the corresponding FeSe2 and Fe3Se4 phases are also forming as soon as Se melts [5]. However, no equivalent to the Fe7Se8 phase, which was observed at increasing temperatures in the FeSe phase formation study, is evidenced in the present sample. The intensity of the Fe(Se,Te)2 phase increases up to 365°C before decreasing, whereas the amount of Fe3(Se,Te)4 appears to increase up to 388°C. The intensity of the reflections of the latter phase then starts to decrease. In contrast the amount of Fe(Se,Te) increases steadily up to 635°C and is the only phase detected in the samples at the end of the heating ramp (690°C). Whereas a structural phase transition was observed in FeSe between 450°C and 500°C, the tetragonal Fe(Se,Te) phase formed in the present sample does not show any evidence for a transformation.
F
Q
o
179 C o 202 C o 225 C o 249 C o 272 C o 295 C o 318 C o 341 C
Nb
Intensity [a.u.]
Cu
Q
| T
2,4
T
|
|
J
2,2
(101)
2,6
2,8
3,0
3,2
3,4
d - spacing [Angstrom]
Fig. 1. Radially integrated diffraction patterns recorded at different temperatures on a wire containing a starting mixture of Fe and Se0.5Te0.5 powders. F Fe, z SeTe, Fe(Se,Te), | Fe(Se,Te)2,d Fe3(Se,Te)4.
603
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
d-spacing [Angstrom]
As shown in Fig.2 for the (111) reflection of Fe(Se,Te), the peaks of this phase first shift rapidly towards higher d-spacing values at a fast and non linear rate. However, from 450°C, this shift takes place at a slower rate and becomes linear versus temperature. This can be interpreted as the consequence of a variation of composition. Due to the fact that the liquid phase formed at the beginning of the fusion of the the Se0.5Te0.5 phase is Se-rich, the Fe(Se,Te) formed during the initial stage is expected to be Te-deficient and thus have shorter lattice parameters than the target FeSe0.5Te0.5 composition. However, as the temperature increases, the liquid becomes enriched in Te and induces in turn a Te enrichment of the Fe(Se,Te) phase until the overall composition of the phase reaches the nominal atomic ratio. At this point, the temperature variation of the d-spacing becomes linear and only reflects normal thermal expansion. The full width at half maximum (FWHM) of the Fe(Se,Te) reflections increases during this reaction step up to 450°C. This indicates that the Se and Te distribution in the newly formed phase is inhomogeneous. Shortly before the phase enters the linear thermal expansion regime, the FWHM decreases, suggesting that solid-state diffusion allows some homogenization, which may continue after the reaction is finished. o
412 C o 447 C o 470 C o 494 C o 518 C o 542 C o 566 C o 589 C
Intensity [a.u.]
Fe(Se,Te) (111)
Fe3(Se,Te)4
FWHM
Fe(Se,Te)2
2,5
2,6
d-spacing [Angstrom]
2,7
2,57 2,56 2,55 2,54 2,53 2,52 2,51 0,055 0,050 0,045 0,040 0,035 0,030 0,025 0,020 0,015 0,010
350
400
450
500
550
Temperature [oC]
Fig. 2. (left) Radially integrated diffraction patterns recorded at different temperatures on a wire containing a starting mixture of Fe and Se0.5Te0.5 powders. (right) Evolution of the (111) peak position of Fe(Se,Te) as well as its FWHM as a function of temperature. The broken red line is a guide to the eye.
3.2. Fe – Se - Te mixture In the wire containing a mixture of Fe, Se and Te powders in Fe:Se:Te = 1:0.5:0.5 nominal atomic ratio, the situation is more complex. From about 200°C, i.e. prior to the formation of new compounds, there is evidence for Se-Te inter-diffusion as illustrated in Fig.3a, where the background level between the respective (101) peaks of Se and Te is clearly increasing with temperature. However, complete alloying is by far not complete before the start of the reaction of the chalcogen elements with Fe. At 250°C, Fe(Se,Te)2 and Fe3(Se,Te)4 have appeared as in the case of the Fe – Se0.5Te0.5 powder mixture but in the present case the reflections are systematically shifted to lower d-spacing values, which indicates that these new phases are richer in Se than it was the case at the beginning of the reaction in the other sample. The temperature for the appearance of these phases is approximately the same in both cases. The variation of the cell parameters of Te as a function of temperature is non monotonic as illustrated in Fig. 3b by the behavior of the (101) reflection of this phase. Starting from 215°C, the expansion is no more linear and the d-spacing even decreases between 240°C and 280°C before increasing sharply again. This departure from linear behavior coincides with a significant acceleration of the lattice expansion rate in Se (see peak shift in Fig. 3b) that is probably related to Te diffusion into the Se, which has not yet molten and has for effect to increase the melting point of the remaining phase. The melting temperature of pure Se is of 221°C and increases upon alloying with Te [8].
604
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
The fast consumption of the Se that has not been incorporated into the Se1-xTex alloy results in the formation of a Se-rich Fe(Se,Te) phase. As shown in Fig.3c in the case of the (111) reflection of this phase, a shoulder appears on the high d-spacing side after about 100 minutes at 350°C. A similar feature is observed for all other Fe(Se,Te) peaks and suggests that a Te-rich phase is formed at that stage from the Se-Te alloy that formed prior to the melting of the un-alloyed Se. This phenomenon is not quite similar to that observed in the Fe + Se0.5Te0.5 sample (section 3.1). Whereas in the latter case the compositional change was progressive, in the present sample it appears that two separate phase formation reactions occur. The pairs of reflection did not merge during the course of the experiment, showing that homogeneisation of the composition is not straightforward at 350°C. 3,250
Fe3(Se,Te)4
Te
3,245 3,240
d-spacing [Angstrom]
Intensity [a.u.]
a
Se
Fe(Se,Te)2
c
b
72 min 101 min 144 min 180 min 223 min 286 min 330 min
Te
3,235 3,230
Intensity [a.u.]
o
151 C o 202 C o 228 C o 254 C o 280 C o 292 C o 305 C o 331 C
3,225 3,220 3,12 3,10 3,08 3,06
Se
3,04 3,02 3,00
2,6
2,8
3,0
3,2
50
100
d - spacing [Angstrom]
150
200
250
300
350
2,40
2,45
2,50
2,55
d-spacing [Angstrom]
Temperature [oC]
Fig. 3. (a) Radially integrated diffraction patterns recorded at different temperatures on a wire containing a starting mixture of Fe, Se and Te powders. (b) Evolution of the (101) peak position of Se and Te as a function of temperature. (c) Detail of the diffraction pattern showing the coexistence of Se-rich and Te-rich Fe(Se,Te) at various sintering times at 350°C.
3.3. Effect of Fe particle size
Integrated intensity [a.u.]
The effect of Fe particle size has been studied in a wire containing a mixture of Fe (< 10 μm or < 74 μm particle sizes) and Se0.5Te0.5 powders. No significant variations could be evidenced in the reaction path, i.e. the same intermediate phases were formed in the same sequence, in both samples. However, a clear difference was observed in the kinetics of Fe(Se,Te) phase formation. Fig. 4 shows a clear shift to higher
250
300
350
400
450
500
550
o
Temperature [ C] Fig. 4. Integrated intensity of the reflections (101) of (Se,Te) - diamonds, (111) of Fe(Se,Te)2 – triangles, and (111) of Fe(Se,Te) – squares, as a function of temperature for the samples prepared from Se0.5Te0.5 and Fe mixtures with Fe particle size of <10μm(filled symbols) and <74μm (open symbols).
J.C. Grivel et al. / Physics Procedia 36 (2012) 600 – 605
temperatures in the sample with larger Fe particles for the Se0.5Te0.5 phase disappearance; the formation and consumption of the Fe(Se,Te)2 compound and the formation of Fe(Se,Te). The Fe3(Se,Te)4 phase (not shown in Fig.4 for clarity) follows a trend similar to that of Fe(Se,Te) 2. 4. Conclusion The formation process of the FeSe0.5Te0.5 phase is affected by the precursor materials used for its preparation, although the reaction sequence appears to be unchanged and characterized by the formation of Fe(Se,Te)2 and Fe3(Se,Te)4 intermediates in all cases studied. When Fe is mixed with pre-alloyed Se0.5Te0.5 powders, the reaction is initiated by the partial melting of the Se0.5Te0.5 phase. The formation of a Se-rich liquid results in Se-rich phases including Fe(Se,Te), which become enriched in Te as the melting of Se0.5Te0.5 proceeds. In such kind of precursor mixtures, the effect of the particle size of the starting Fe powder was studied and found to strongly influence the reaction kinetics as larger particles shift the reaction sequence to higher temperatures. When unreacted Se and Te powders are used in the precursor, partial Se1-xTex alloying takes place during heating prior to and during the formation of the intermediate phases. Since the alloying is incomplete, Se-rich Fe(Se,Te) forms at relatively low temperature prior to the melting of the Te-rich Se-Te alloy. This results in an inhomogeneous Fe(Se,Te) phase. Acknowledgements The authors gratefully acknowledge the technical assistance of M. Wichmann and L. Lorentzen as well as financial support from DANSCATT. References [1] Yeh K-W, Huang T-W, Huang I-L, Chen T-K, Hau F-C, Hsu F-C, Wu PM, Lee Y-C, Chu YY, Chen C-L, Luo J-Y, Yan DC, Wu M-K. Tellurium substitution effect on superconductivity of the alpha-phase iron selenide. Europhys Lett 2008;84:37002 (4pp). [2] Tsurkan V, Deisenhofer J, Günther A, Kant Ch, Klemm M, Krug von Nidda HA, Schrettle F, Loidl A. Physical properties of FeSe0.5Te0.5 single crystals grown under different conditions. Eur Phys J B 2011;79:289-299. [3] Mizuguchi Y, Deguchi K, Tsuda S, Yamaguchi T, Takeya H, Kumakura H, Takano Y. Fabrication of the iron-based superconducting wire using Fe(Se,Te). Appl Phys Exp 2009;2:083004 (3pp). [4] Gao Z, Qi Y, Wang L, Wang D, Zhang X, Yao C, Ma Y. Superconducting properties of FeSe wires and tapes prepared by a gas diffusion technique. Supercond Sci Technol 2011;24:065022 (4pp). [5] Grivel J-C, Wulff AC, Zhao Y, Andersen NH, Bednarčík J, Zimmermann Mv. In-situ observation of the formation of FeSe. Supercond Sci Technol 2011;24:015007 (4p) [6] Grivel J-C, Wulff AC, Zhao Y, Bednarčík J, Zimmermann Mv. In-situ study on the formation of FeTe. J Mater Sci 2011;46:4540-4544 [7] Hammersley AP, Svensson SO, Hanfland M, Fitch AN, Häusermann D. Two-dimensional detector software: From real detector to idealised image or two-thete scan. High Pressure Research 1996;14:235-248 [8] Ghosh G, Lukas HL, Delaey L. A thermodynamic assessment of the Se-Te system. CALPHAD 1988;12;295-299
605