Process Induced Texturing in Ex-situ Magnesium Diboride Based Wires

1 ASC2024-3MPo2C-04 Process Induced Texturing in Ex-situ Magnesium Diboride Based Wires. Alberto Ansaldo, Sergio Marras, Paola Mauceri, Tiziana Spina, Andrea Tumino Abstract—Texturing plays a key role in MgB2 wires performance. One of the limits of powder in tube wires production with respect to other techniques, e.g. pulsed lased deposition, is the lack of a proper control of the grain orientation with respect to the wire main axis. In this paper we will discuss the influence of the pipe mechanical deformation on the powder texturing and the impact of the powder preparation and pre-processing on the result. Index Terms— Crystal growth, Ex-situ Powder-In-Tube, Magnesium boride wire, Magnesium compounds, Materials science and technology, Physics, Powders, Superconducting materials, Technology, Texturing, micro-XRD. I. INTRODUCTION URRENT in magnesium diboride (MgB2) based superconducting wires is affected by a number of parameters including grain size, connectivity and texturing [1], due to the inherent anisotropy of the MgB2 crystals [2]. The most common approach to evaluate MgB2 texturing relies on measuring the magnetic anisotropy of the current with respect to the external magnetic field [3]. Unfortunately, transport-based methods allow to evaluate the crystal orientation anisotropy along the wire cross-section only, while a proper alignment of the MgB2 crystals along the wire main axes should lead to a significant current boost in symmetrical section wires (i.e., round, square, hexagonal and so on) as well. In this perspective, a direct measurement of the average orientation of the crystallite in the cable would be highly desirable. In the past, studies on the crystallites alignment in tapes were made by means of synchrotron light XRD [4, 5, 6], but this method is not directly appliable when the metallic sleeves exceed certain thicknesses and it requires access to a synchrotron facility. In this paper we show how, by means of micro-XRD on the cross-sections, we are able to evaluate the crystallite alignment of different samples of ex-situ MgB2 powder in tube (PIT) wires. Moreover, we show how different powders and preparation methods lead to various degrees of alignment due to the mechanical forces applied during rolling. This work was supported in part by EU’s SCARLET project (“Superconducting cables for sustainable energy transition”), Horizon Europe research and innovation programme under grant agreement No. 101075602. (Corresponding author: Alberto Ansaldo, [email protected]). Alberto Ansaldo, Paola Mauceri, Tiziana Spina, and Andrea Tumino are with ASG Superconductors S.p.A., 16133, Genova, Italy (e-mail: ans[email protected]t, [email protected], [email protected], [email protected]). II. EXPERIMENTAL MgB2 powders were synthesized starting from commercial amorphous boron and magnesium powder. For “small” crystallite powder (SC, Fig. 1a), boron and magnesium were mixed in an open stainless-steel crucible in stoichiometric ratio. The crucible was then placed in a tubular furnace and heated at about 900°C for 1 hour. For the “large” crystallite powder (LC, Fig. 1b), precursors were mixed in stoichiometric ratio and put in an iron tube about 12 cm long, 15 mm in diameter, sealed by two soldered iron stoppers. The iron tubes were then placed in a tube furnace and heated at about 1000°C for 6 days. All the preparations, reactions and handling of the powders were carried out in an inert argon atmosphere. A third powder was prepared by wet milling the “large” crystallite powder (MLC), to partially separate out polycrystalline aggregates in single crystallites (Fig. 1c). Powders were characterized by XRD, SEM and laser diffraction. Finally, the powders were pressed in Sergio Marras is with the Istituto Italiano di Tecnologia, 16163, Genova, Italy (e-mail: [email protected]). Color versions of one or more of the Fig.s in this article are available online at http://ieeexplore.ieee.org C Fig. 1. Examples of the three as prepared powders: a) "small crystallites", b) "large crystallites", and c) wet milled large crystallites. Fig. 2. Size and cross section of the samples used in this paper: Billet, rolled at 2.5 mm (R2.5), rolled at 1.1 mm (R1.1), flat ribbon 2.9 mm by 0.25 mm (F0.25). Author Accepted Manuscript version of the article by Alberto Ansaldo et al. in IEEE Transactions on Applied Superconductivity, 35 (2025), 6200204 https://doi.org/10.1109/TASC.2024.3512526 Distributed under the terms of the Creative Commons Attribution License (CC BY 4.0) 2 ASC2024-3MPo2C-04 nickel tubes (12 mm external diameter, 6 mm internal bore) sealed by tin stoppers. The resulting billets were then cold rolled up to the final shape. We took samples at three different sizes and shapes i.e., 2.5 mm (R2.5), and 1.1 mm (R1.1) false square, and 2.9 mm by 0.2 5 mm flat ribbon (F0.25, see Fig. 2). The samples were sintered in a linear oven at about 800°C, crosscut and lapped. SEM images of the surfaces after lapping are reported in Fig. 3. Micro-XRD was performed by a 3rd generation Empyrean (Malvern-PANalytical, Fig. 4(a) equipped with a Cu source and GaliPIX3D area detector Fig. 4(b)). The focused X-ray spot on the sample was about 50 µm in diameter. It is worth noting that, due to the very high transparency of the MgB2 to X-ray, even if the spot size is very small, the sample volume involved in the measurement is not negligible and the signal from the outer nickel sheath can be very strong. For this reason, the measurements were carried out in the very centre of the samples to minimize the contribution from the nickel sheath. Rietveld refinement is performed by PowderCell 2.4 software applying the March-Dollase model for the preferred orientation. III. RESULTS AND DISCUSSIONS A. Determination of the measurement range Due to the presence of different phases in the wires (e.g., MgB2, MgO, Ni) a proper measurement range must be selected where the overlap between MgB2 peaks and the other phases is negligible. Looking at the theoretical diffractogram (Fig. 5), we select the 2Θ range 23°-46° as there are at least 3 discernible MgB2 peaks, namely directions <0 0 1> (~25°), <1 0 0> (~34°), and <1 0 1> (~42°). The partial overlap of MgB2 1 0 1 peak with main MgO peak is not an issue due to the very low content of MgO of the samples (less than 1% in volume). A sketch of MgB2 main crystal planes is reported in Fig. 6. B. March-Dollase distribution and texturing To estimate the samples texturing, a Rietveld refinement [7] is performed, using the March-Dollase approach [8, 9]. The relative intensities of the simulated peaks, while varying the sample alignment with respect to the <1 0 0> direction, are reported in Fig. 7. Data are plotted as a function of the o1 parameter of the March-Dollase distribution. In short, if o1=1, the sample is random; values smaller than 1 correspond to a preferential alignment in the reference direction (<1 0 0> in this case); values larger than 1 model a preferential alignment in the perpendicular direction, i.e., <0 0 1> in our case. As a rule of thumb, if the preferred orientation is along the sample surface normal, its intensity is enhanced, and vice versa. Examples of computed normalized March-Dollase distributions for different values of o1 are reported in Fig. 8. The intensity of the vector in the polar representation describes the relative abundance of crystallites oriented in that direction. To represent the degree of orientation/texturing, different metrics have been proposed in the past [10], but for our aim we think that the median direction Fig. 3. Polished cross sections of three different samples of 2.9 mm by 0.25 mm tape made with a) "small crystallite", b) "large crystallite”, and c) milled large crystallite powders respectively. Fig. 5. Theoretical diffractogram for the main compounds. Fig. 7. Computed peak intensities while varying the texturing level (o1 parameter). Fig. 6. MgB2 main crystal planes and their normal directions. Fig. 4. Detail of a) the Malvern-PANalytical Empyran, and b) the focusing X-ray lens used for this paper. Author Accepted Manuscript version of the article by Alberto Ansaldo et al. in IEEE Transactions on Applied Superconductivity, 35 (2025), 6200204 https://doi.org/10.1109/TASC.2024.3512526 Distributed under the terms of the Creative Commons Attribution License (CC BY 4.0) 3 ASC2024-3MPo2C-04 of the distribution (i.e. the angle that corresponds to the 50% of the distribution) is a more significant value. The median angle for the different distributions is superimposed in Fig. 8 as well. A plot of the median value as a function of o1 is reported in Fig. 9. It is worth noting that due to the symmetry of the system, median values are computed in the 0-90° range. For a given direction a median value equal to 45° implies a random distribution (o1=1), values lower than 45° imply that the chosen orientation is the preferential one (o1<1) and, on the contrary, if the values are larger than 45° the preferred orientation of the sample is the other main crystallographic direction (o1>1). C. Texturing measurements in differently shaped wires. In the case of MgB2, the best performance, in terms of critical current and critical field, is along the {0 0 1} plane [11]. While in a ribbon or a flat wire this condition corresponds, usually, to having the <0 0 1> direction facing the flat side of the specimen, making quite easy to evaluate it, this is not true for an arbitrary shaped wire. If we consider, as an example, a perfectly round wire (Fig. 10) with all the crystallites facing the surface (i.e., the {0 0 1} planes along the wire main axis) a measurement of a lateral cross-section would result in a spread of the values nearly indistinguishable from a random sample due to the low MgB2 extinction coefficient for X-rays. An alternative and more reliable approach is to perform the measurement by crosscutting the wire and performing the measurement on that surface by looking at the <1 0 0> direction. In this condition, a perfectly textured sample, in fact, will have no signal from the <0 0 1> while the signal from the <1 0 0> direction will be maximized. In other words, knowing the degree of alignment of the crystals in the <1 0 0> direction on the cross-section, due to the geometry of the system, provide a robust measurement of the orthogonality of the <0 0 1> direction with respect to the wire main axes. Please not that <1 0 0> and <0 1 0> directions are indistinguishable in a hexagonal crystal. This means that in all the paper any reference to the <1 0 0> refers to all the equivalent directions. Fig. 9. Computed values of the median angle of the MarchDollase distribution in the 0-90° range. Fig. 10. Hypothetical perfectly texturized round MgB2 wire and side measurement set-up schematics. Fig. 8. Examples of March-Dollase polar distribution for different values of o1, and the relative median angles with respect to the chosen reference crystallographic direction. Fig. 11. MicroXRD diffractograms (solid lines) and fits (dashed lines) of the following samples: a) billet, filled with LC; R2.5 samples filled with b) SC, c) LC, and d) MLC powders; R1.1 samples filled with e) SC, f) LC, and g) MLC powders; F0.25 samples filled with h) SC, i) LC, and j) MLC powders. Author Accepted Manuscript version of the article by Alberto Ansaldo et al. in IEEE Transactions on Applied Superconductivity, 35 (2025), 6200204 https://doi.org/10.1109/TASC.2024.3512526 Distributed under the terms of the Creative Commons Attribution License (CC BY 4.0) 4 ASC2024-3MPo2C-04 D. Experimental results. Measured micro-diffractograms for the three different kinds of samples, and corresponding Rietveld refinements are reported in Fig. 11. The fits (dashed lines) are in good agreement with the measurements (solid lines). The difference between measurement and fit that can sometimes be appreciated for the weaker peak (Mg 0 0 1) is mostly due to the background noise and do not influence particularly orientation estimation. As the fit is based overall diffractogram, the two other peaks allow to univocally identify the crystal alignment. By applying the March-Dollase approach to the Rietveld refinement, we get for each sample the o1 parameter from which we can compute the corresponding polar distribution (Fig. 12). The resulting median orientations of the crystallites facing cross section are reported in TABLE I and plotted in Fig. 13. Micro-XRD measurements make clear the impact of the mechanical forces occurring during ex-situ PIT process. Small crystals are less prone to texturing during rolling. The texturing increases with deformation. The values are consistent, ranging from 60° to 64°. This suggests that most of the texturing should occur during the first deformations. Large crystals are strongly affected by the deformation process. The texturing increases with the deformation process and peaks with a median direction of 79°. Notably, the starting billet, due to the filling process, shows a texturing in the "wrong" direction (median 20°, less than 45° i.e., a random sample). In the case of the milled large crystals, the effect of separating out the single large crystallites is quite evident: these samples get immediately textured with the first deformations (median direction 76°) but then the value saturates at 78°. Notably after the last square deformation step (R1.1) the texturing in both LC and MLC samples is nearly identical, suggesting the external forces we apply during the mechanical deformation are more adequate to separate out the crystallites. CONCLUSIONS The deformation process occurring in PIT ex-situ wire production can cause a significant texturing of the MgB2 powder. This mechanical texturing is strongly influenced by crystallites size, shape and aggregation. The ability of controlling those aspects in wire production can lead to a significant improvement in MgB2 wires performance. ACKNOWLEDGMENT This work was partially funded by the European Union’s project SCARLET (“Superconducting cables for sustainable energy transition”), Horizon Europe research and innovation programme under grant agreement No. 101075602. REFERENCES [1] M. Eisterer e H. W. Weber, «The influence of weak texture on the critical currents in polycrystalline MgB2,» Superconductor Science and Technology, vol. 23, n. 3, p. 034006, 2010. [2] M. Eisterer, M. Zehetmayer e H. W. Weber, «Current Percolation and Anisotropy in Polycrystalline MgB2,» Physical Review Letters, vol. 90, n. 24, pp. 247002-1-4, 2003. [3] P. Kováč, «Critical current anisotropy of MgB2 tapes,» J. Phys.: Conf. Ser., vol. 153, pp. 012019 1-6, 2009. [4] M. Eisterer, W. Häßler e P. Kováč, «Critical currents in weakly textured MgB2: Nonlinear transport,» Physical Review B, n. 80, p. 174516, 2009. [5] W. Häßler, P. Kováč, M. Eisterer, A. B. Abrahamsen, M. Herrmann, C. Rodig1, K. Nenkov, B. Holzapfel, T. Melisek, M. Kulich, M. Zehetmayer, J. Bednarcik e J.-C. Grivel, «Anisotropy of the critical current in MgB2 tapes made of high energy milled precursor powder,» Supercond. Sci. Technol., n. 23, pp. 065011 1-6, 2010. [6] M. Vignolo, G. Romano, D. Nardelli, E. Bellingeri, A. Martinelli, A. Bitchkov, C. Bernini, A. Malagoli, V. Braccini e C. Ferdeghini, «In situ high-energy synchrotron x-ray diffraction investigation of phase formation and sintering in MgB2 tapes,» Superconductor Science and Technology, n. 24, p. 065014, 2011. [7] H. M. Rietveld, «A profile refinement method for nuclear and magnetic structures,» Journal of Applied Crystallography, n. 2, pp. 65-71, 1969. TABLE I MEDIAN ORIENTATION OF THE CRYSTALLITES FACING THE WIRE CROSS SECTION, WITH RESPECT TO <0 0 1> Sample Powder Billet R2.5 R1.1 F0.25 SC 60° 63° 64° LC 20° 64° 78° 79° MLC 76° 78° 78° Fig. 12. Polar representation of the March-Dollase distributions computed for the different wire samples analysed. Please note that <1 0 0> and <0 1 0> directions are indistinguishable (equivalent) in a hexagonal crystal. Fig. 13. Angular distribution median value [0°-90° range]. Author Accepted Manuscript version of the article by Alberto Ansaldo et al. in IEEE Transactions on Applied Superconductivity, 35 (2025), 6200204 https://doi.org/10.1109/TASC.2024.3512526 Distributed under the terms of the Creative Commons Attribution License (CC BY 4.0) 5 ASC2024-3MPo2C-04 [8] A. March, «Mathematische Theorie der Regelung nach der Korngestah bei Affiner Deformation. Kristallographie,» Zeitschrift für Kristallographie - Crystalline Materials, n. 81, pp. 285-297, 1932. [9] W. A. Dollase, «Correction of intensities for preferred orientation in powder diffractometry: application of the March model,» Journal of Applied Crystallography, n. 19, pp. 267-272, 1986. [10] E. Zolotoyabko, «Determination of the degree of preferred orientation within the March–Dollase approach,» Journal of Applied Crystallography, n. 42, p. 513–518, 2009. [11] C. Buzea e T. Yamashita, «Review of superconducting properties of MgB2,» Superconductors, Science & Technology, vol. 14 , n. 11, p. R115, 2001. Author Accepted Manuscript version of the article by Alberto Ansaldo et al. in IEEE Transactions on Applied Superconductivity, 35 (2025), 6200204 https://doi.org/10.1109/TASC.2024.3512526 Distributed under the terms of the Creative Commons Attribution License (CC BY 4.0)