EFFECTS OF STRAIN AND ANNEALING TEMPERATURE ON CRITICAL CURRENT DENSITIES IN Cu/MgB2 and Ni/MgB2 SUPERCONDUCTORS
V.Beilin*, M. Roth*, E. Dul’kin*, E. Yashchin*, Y. Lapides**, J.Greenberg**,
E. Mojaev*, M. Tsindlekht***, E. Galstyan***,Y. Felner***
* Department of Applied Physics
** Institute of Inorganic Chemistry
*** The Racah Institute of Physics
The Hebrew University of Jerusalem,Jerusalem91904, Israel
ABSTRACT
Critical current density, Jc, ac susceptibility (real part, χ’) and X-ray diffraction of Cu/MgB2 and Ni/MgB2 wires and tapes were studied in their as-deformed and thermally processed (annealed) states. The Jc increase and degradation throughout the wires/tapes deformation (drawing, rolling) routes was observed. Annealing of Cu/MgB2 tapes in the temperature range of 250°C to 850°C caused Jc degradation (down to a negligible level at elevated temperatures). Sintering of Ni/MgB2 tapes in the 850°C - 950°C range resulted in achieving high Jc > 1.3x105A/cm2 at 4.2K, while thermal processing at the lower temperatures of 750-800°C caused the degradation of core connectivity and, thus, of Jc that dropped below its value in non-treated tapes. Two-step annealing consisting of low- and high-temperature steps did not lead to Jc degradation (JD). JD in Ni/MgB2 tapes was explained by the formation of thin MgB4 layers around MgB2 granules leading to their decoupling. Another possible origin of this effect might be stress relaxation in the MgB2 core during Ni sheath softening.
1. INTRODUCTION
Since the recent discovery of superconductivity in MgB2 [1], extensive research activities have been focused on the superconductivity mechanism as well as on the methods of manufacturing practically applicable conductors of this material. It has become quite evident that some specific characteristics of this compound, such as its low cost, high critical current density, relatively high critical temperature, Tc, of 39K (the highest among low-temperature superconductors), etc. make MgB2 a very attractive material for high power applications. Significant advantage of MgB2 over higher-Tc superconductors, in particular over Bi-2223, is the absence of weak links at grain boundaries [2], at least in the thermally-processed state. The particularly important feature of granular MgB2 (e.g. in the form of mechanically processed powder within a metallic sheath) is its ability to carry high superconducting currents in an unsintered state [3]. At the same time, as-deformed wires, although exhibiting Jc as high as about 105 A/cm2 at 4.2K, demonstrate weak-link-like behavior [4]that may depend on their mechanical prehistory. Evolution of deformation-induced intergranular weak links in course of subsequent thermal treatment is an issue of high practical significance. Accordingly, the choice of the deformation path and thermal processing required for the core sintering must be addressed in developing metal-clad MgB2 wires and tapes. Meanwhile, the influence of deformation prehistory in combination with subsequent heat treatment on the wire or tape performance still has not been studied adequately. In the present work, the Jc, ac susceptibility of Cu/MgB2 and Ni/MgB2 wires and tapes have been studied in as-deformed states as well as after heat treatment at various temperatures. We have also performed the X-ray diffraction (XRD) and the first acoustic emission characterization of as-sintered metal-clad MgB2 tapes.
2. EXPERIMENTAL
Cu/MgB2 and Ni/MgB2 wires and tapes were fabricated by the powder-in-tube method using MgB2 precursor powder of Alfa Aesar (ex-situ route) The details of the wires and tapes preparation and processing could be found elsewhere [4,5]. Powder-packed Cu and Ni tubes were drawn to various diameters down to 1.2mm and then rolled into the tapes of various thicknessdown to 0.18 mm. Cu/MgB2 wires and tapes were fabricated as well by the in-situ route, where MgB2 phase was synthesized in the process of wires and tape thermal processing. Mixture of Mg and B powders with an atomic ratio close to MgB2 stoichiometry was used as a precursor in this case. Deformation degree, , in the both drawing and rolling processes were calculated as the ratio of initial cross section area at the beginning of the process, Fin, to that at the given processing step, Fps : = Fin/ Fps. Thermal operations were carried out in flowing high-purity Ar in the temperature range of 250°C to 950°C. XRD characterization used for the core phase analysis was carried out with Philips PW-1710 diffractometer using Kradiation. Jc measurements were performed by the four-point probe method with current and voltage contacts formed by soldering at the surface of metallic sheaths. Maximal current we could measure using an available current source did not exceed 240 and, sometimes, 120A. Therefore, in the cases where we did not reach the critical state in the samples due to these current limitations or due to excessive heating of current contacts, only the “lower-limit” Jc values were presented below. The ac susceptibility was measured in the range of 5-50K at the frequency of 770 Hz in the driving magnetic field of about 0.08Oe. In this work, the acoustical emission (AE) method [6] was for the first time applied to probe the sintering of MgB2 superconducting core. AE measurements were carried out in the process of tape bending at the room temperature around a cylindrical rod of 6-8 mm in diameter with bending angles varying from 10o to 90o. Generation of AE signals was considered as an indication of nucleation and growth of stress-induced cracks which would only be observed in the sintered (solid) powdered core.
3. RESULTS AND DISCUSSION
Jc values of as-deformed wires and tapes vs. cross-section area reduction = Fin/ Fps are shown in Table1 along with the corresponding data for tapes thermally treated at various temperatures, Tann. It can be seen that the Jc growth during the drawing process is followed by severe Jc degradation down to about zero value along with the growth. An interesting result has been that rolling the as-drawn Ni/MgB2 wire with negligible Jc has resulted in Jc recovery followed by its significant growth with further increase. Jc of Cu/MgB2 tapes at high values has decreased with growth.
Figs. 1,2 show the real part of the ac susceptibility, ’, of Cu/MgB2 and Nu/MgB2 wires and tapes. = (T) traces in as-deformed state demonstrate a two-step superconducting transition with two transition temperatures (Fig.1 and curves 1,4 in Fig. 2). The higher transition temperature, Tonset, is 39K (in accordance with the published data for MgB2) in all cases, and it features the strong-link intragranular behavior. The lower transition temperature, TL, corresponds to the weak-link-like intergrnular coupling in as-deformed wires and tapes.
Fig.1. ’ vs. T dependence for as-deformed with different cross section area reductions δ and annealed Cu/MgB2 wires and tapes: 1- round wire, δ = 13.3; 2-round wire, δ = 15.2; 3- tape, δ = 35.8; 4- tape, δ = 50.3; 5- tape 3 subjected to annealing at 600°C; arrows indicate TL temperatures.
Fig.2. ’ vs. T dependence for as-deformed and thermally treated Ni/MgB2 wires and tapes 1- as-rolled tape, δ = 15.7; 2- tape treated at 750°C; 3 - tape as-sintered at 950°C; 4- as-drawn wire, δ = 7.9; arrows indicate TL temperatures.
Jc reduction after initial growth along with growth during drawing (Table 1) implies that the powder density (PD) exceeds its limiting value (critical packing density [7]) under the given deformation conditions. Therefore, further wire straining and thus MgB2 powder movement would proceed along with the core disintegration resulting in the observed Jc reduction.
Table 1. Experimentally determined parameters of Cu/MgB2 and NiMgB2 wires and tapes
Material / Shape / State / Straindegree / Jc, A/cm2
at 4.2K / TL, K
Cu/MgB2 / Wire / as-drawn / 13.3 / 30
Cu/MgB2 / Wire / as-drawn / 15.2 / 28
Cu/MgB2 / Tape / as-rolled / 35.8 / 10,000 / 33
Cu/MgB2 / Tape / as-rolled / 50.3 / 210 / 28
Cu/MgB2 / Wire / Tanna- 650 oC, 0.5h / <25
Cu/MgB2 / Wire / Tanna- 750 oC, 0.5h / <25
Cu/MgB2 / Wire / Tanna- 850 oC, 0.5h / <25
Cu/MgB2 / Tape / Tannb- 600 oC, 0.5h / <25 / 22
Cu/MgB2 / Tape / Tannb- 250 oC, 1h / 3300
Cu/MgB2 / Tape / Tannb- 400 oC, 1h / 300
Cu/MgB2 / Wire / in-situ, 700oC / 115,000
Cu/MgB2 / Tape / in-situ, 700oC / 195,000
Ni/MgB2 / Wire / as-drawn / 3.3 / 450
Ni/MgB2 / Wire / as-drawn / 6.25 / 1,100
Ni/MgB2 / Wire / as-drawn / 7.9 / 28
Ni/MgB2 / Wire / as-drawn / 10.2 / 270
Ni/MgB2 / Wire / as-drawn / 14.3 / <100
Ni/MgB2 / Tape / as-rolledc / 15.7 / 32
Ni/MgB2 / Tape / as-rolledc / 17.1 / 4,350
Ni/MgB2 / Tape / as-rolledc / 19.7 / 14,800
Ni/MgB2 / Tape / as-rolledc / 26.9 / 22,900
Ni/MgB2 / Tape / Tannd- 650 oC, 1h / - / 15,700 / -
Ni/MgB2 / Tape / Tannd- 750 oC, 1h / - / 520 / -
Ni/MgB2 / Tape / Tann d -800 oC, 1h / - / 1,300 / -
Ni/MgB2 / Tape / Tann d -850 oC, 0.5h / - / >32,700 / -
Ni/MgB2 / Tape / Tannd -900 oC, 0.5h / - / >32,700 / -
Ni/MgB2 / Tape / Tannd -950 oC, 0.5h / - / >32,700 / -
Ni/MgB2 / Tape / Tann -915 oC, 1h / 24.7 / >130,000
a- round wires with the strain degree of 13.9
b- tapes with the strain degree of 35.8
c- tapes rolled out of the round wire with the strain degree of 14.3 prior to rolling.
d- tapes with the strain degree of 17.3
TL decrease as well as Jc drop with increase for as-deformed Cu/MgB2 tapes (Fig.1, Table 1) were supposedly caused by the same mechanism.
The Jc and TL growth under subsequent rolling of as-drawn wires (Table 1, curves 1,3 in Fig. 1) are the indications of recovery and further enhancement in the core connectivity by powder redistribution. Different behavior of the core connectivity along the drawing and rolling paths of the PIT procedure originates supposedly from different stress states in the deformation zone for both processes.
The main goals of thermal processing (annealing) the MgB2 tapes are sintering the core to provide high current-carrying ability of the tapes, as well as sheath softening to make possible further operations e.g. the tape bending for coil preparation. In order to probe the MgB2 core sintering, we have employed the AE characterization. In the AE experiments, Ni/MgB2 tapes subjected to annealing in the temperature range of 650°C to 950°C have been studied. It is found that the tapes treated at Tann 800°C did not generate AE signals under bending tests. AE signals only appear starting from TT of 800°C. At Tann850°C to 950°C AE signals was observed even at minimal bending angles. Accordingly 850°C has been accepted as the temperature where the core sintering is well-established (sintering threshold, Tst ) in the studied samples.
It is shown in Table 1 that annealing of Cu/MgB2 tapes fabricated by the ex-situ route, results in Jc degradation down to a negligible level in the temperature range of 250°C to 850°C. Accordingly, we have observed a broadened superconductive transition with lowered TL temperature in the ’ vs. T curve in Fig.1 for the Cu/MgB2 tape as-annealed at 600°C (curve 5). Further increase in annealing temperature would not result in Jc enhancement due to strong Cu-Mg interaction. Contrary to that, the in-situ route at 700°C provides the high current-carrying ability of Cu/MgB2 tapes (Table 1). Hereafter we describe just the results for in-situ tapes. Annealing the as-deformed Ni/MgB2 tapes at 900°C-950°C has resulted in a sharp one-step superconducting transition and drastic Jc enhancement (see Table 1 and Fig. 2, curve 3). These data indicate the suppression of the weak-link behavior and the establishment of a long-range strong-link connectivity over the tape core. It means that thermal processing at this temperature provides the necessary core sintering and may be used, in principle, as a finishing operation in the process of Ni/MgB2 tapes fabrication. At the same time, this temperature is too high as leading to strong core-sheath interaction with the formation of a reactive layer at the core-sheath interface. Therefore, we have studied the tape annealing at lower temperatures and found that high core connectivity as well as the corresponding high Jc are maintained down to Tann of 850°C (Table 1) in accordance with the AE data. Meanwhile, further reduction in Tann has resulted in severe Jc degradation as compared to the as-rolled tapes.
The χ(T) trace for the tape treated at 750°C demonstrates a one-step behavior (without the intergranular step) with a broadened transition (curve 2 in Fig. 2) indicating the deterioration of the core connectivity. The samples of two other batches subjected to 1h annealing at 650°C to 900°C also show. Jc degradation, but with the differences in temperature windows between all three batches (see Table 1 for batch 1 and Fig. 3 for all three batches).
There are three possible processes which may be responsible for the observed Jc degradation in Ni/MgB2 tapes at the intermediate Tann, namely: a) chemical interaction between the MgB2 core and Ni sheath resulting in the formation of low-temperature Ni-containing phases, b) MgB2 decomposition during annealing and c) mechanical interaction between the sheath and core. With regard to the first factor,XRD analysis reveals indeed some differences in phase compositions (not presented here) at the Ni/core interface between the tapes treated at 750°C and 950°C. When studying the “a” scenario we have used samples subjected to thermal processing consisting of two steps, I and II, at different temperatures: i) 750°C (I) + 950°C (II), ii) 950°C (I)+ 750°C (II). If the low-temperature (750°C) annealing virtually causes the formation of the phases (no matter-stable or nonstable ones) responsible for the Jc degradationfound, we are prone to observe such effect in one of these samples. However both samples have demonstrated high Jc similarly to that in the sample subjected to one-step annealing at 950°C.
In order to check the possibility of the “b” scenario, the equilibrium composition of MgB2 core in Ar atmosphere has been calculated for a closed system in a wide temperature range using ASTRA software [8] and the IVTANTERMO Thermodynamic Database [9]. The results of the calculations are presented in Table 2. Apparently, the main equilibrium species formed due to MgB2 decomposition are gaseous Mg and solid MgB4.It may be anticipated that in case of an unsintered tape (for Tann < Tst), Mg evaporation from the tape core will be dominated by evaporation from the surfaces of granules (rather than from their volumes) throughout the core volume.
Table 2. Equilibrium composition of MgB2 in 1 atm Ar*.
Temperature, KPhase / Species / 873 / 973 / 1073 / 1173 / 1273
Gaseous / B2 / 0 / 0 / 0 / 0 / 0
B / 0 / 0 / 6.81x10-21 / 1.68x10-18 / 1.74x10-16
Mg / 0.0000346 / 0.0004461 / 0.0035404 / 0.0195775 / 0.0821073
Mg2 / 1.46x10-12 / 1.84x10-10 / 8.997x10-9 / 2.184x10-7 / 3.112x10-6
Condensed / MgB2 / 0.909084 / 0.908998 / 0.908348 / 0.904917 / 0.890392
MgB4 / 5.31x10-6 / 0.0000686 / 0.0005461 / 0.0030694 / 0.0137508
MgB12 / 0 / 0 / 0 / 0 / 0
B / 0 / 0 / 0 / 0 / 0
Mg / 0 / 0 / 0 / 0 / 0
* Partial pressures of the gases are given in atm, concentrations of the condensed phases - in mass fractions.
Therefore, the MgB4 phase may be nucleated inhomogeneously over the granule volume, preferably within a near-surface volume. In order to explain the obtained results, a simplified picture has been considered where all MgB4 is located inside a thin single-phase surface layer. It is possible to expect then the core decoupling into separate MgB2 granules surrounded by insulating MgB4 shells and thus Jc degradation, if the thickness of this intergranular barrier layer,Δbl, fulfills the following condition: Δblλ, where λ is a coherence length. A similar approach has been used in [10] for MgO layers around powder particles in unsintered Cu/MgB2 tapes. The characteristic particle size, Sbl, corresponding to the equality of Δbl ≈ λ can be roughly estimated using the data on MgB4 concentrations, CMgB4, as follows:
Sbl= k*λ / CMgB4 (T), (1)
where for spherical particles Sbl is a sphere of radius R and k = 3, while for plate-like particles Sblis a plate of thickness t, and k = 1. Taking into account the λ anisotropy and the scatter of published data [11] as well as moderate c-axis core texturing in MgB2 tapes [5], we have adopted for our estimations λ ≈ 5 nm.
Based on the calculated data of Table 2, the estimated R value for T = 800°C is about 20 μm and t ≈ 7 μm. In the above estimations, we have taken the mass density ratio between MgB2 and MgB4 phases as 1.5:1 based on their atomic compositions. As already notedabove, the data of Table 2 are calculated for a closed system, albeit in real MgB2 tapes some Mg losses have been observed [12]. Meanwhile, as it has been found in [13], in an open system the MgB4 concentration just below 900°C may be as high as 80%, i.e. more than three orders of magnitude higher than in the closed system. Therefore, even minor deviations from closeness will give rise to a significant CMgB4 enhancement and thus to Sbl reduction and/or to Δbl growth. If we assume that the Ceq value increases up to 1% due to minor Mg losses, it results in Δbl of about 20λ.
High-temperature processing, e.g. at 950oC, facilitates particle sintering (fusing) in the solid phase and, consequently, a drastic reduction of the core internal surface. In this case Mg will evaporate mainly from the outercore surface that would not cause the deterioration of intergrain connectivity and thus the Jc degradation. In addition, the decrease of evaporation surface area would result in the reduction of evaporation rate. Results of the experiments with two-step annealing can be easily understood in the frameworks of the above model. MgB4 barrier layers formed at the first step at 750°C in “i” experiment, will be fast eliminated at the sintering step by diffusion-induced core homogenezation at the higher temperature of 950°C. In “ii” experiment this layer would not be formed at all, because the core will be sintered already at the first high-temperature (950°C) step. The second step at 750°C will be applied to the as-sintered core, where Mg evaporates predominantly from outer core surface not giving rise to the formation of intergranular barrier layers.
The third possible scenario of Jc degradation may be connected with the softening of metallic sheath in the process of annealing at temperatures Tann, which are sufficiently high to provide this softening, e.g. by recrystallization, and they are still below Tst. In this case, stresses accumulated inside the MgB2 core during the drawing and rolling processes will be relaxing that would result in the looseness of as-densified core followed by degradation of its connectivity. This effect is responsible for Jc degradation in Cu/MgB2 tapes (see [10,14]) that starts at temperatures as low as 250°C (see Tabl.1) where no MgB2 decomposition can take place.
At higher temperatures, the contribution of Cu-Mg interaction becomes dominating. As to the Ni/MgB2 tapes, the observation of two Jc-degradation windows at “low” and “high” temperatures in batch 3 (Fig.3) may indicate the synergy of both Mg decomposition and stress relaxation mechanisms. Accordingly, differences between Jc vs. Tann data between the Ni/MgB2 samples of different batches may originate from minor differences in the tape fabrication paths influencing the granule sizes as well as the sheath strength and core density.
Fig.3. Normalized critical current densities of Ni/MgB2 tapes subjected to annealing at various temperatures; Jmax - maximal current density applied to a sample at maximal Tann.
4. SUMMARY
Core densification and subsequent disintegration of MgB2 wires leads to growth and a following drastic drop in the critical current, Jc, along the drawing path. Rolling of the as-drawn wire causes the increase of core connectivity and, thus, of Jc due to powder redistribution.
Annealing Cu/MgB2 tapes in the temperature range of 250oC to 850oC results in Jc degradation (down to a negligible level at elevated temperatures) caused by stress relaxation in the core due to Cu softening at low temperatures and presumably by Cu-Mg interaction as well at high temperatures.
Sintering of as-deformed Ni/MgB2 tapes at 900-950°C suppresses the weak links formed during tape rolling and results in establishing long-range strong connectivity throughout the tape core and high Jc. Severe Jc degradation upon annealing of Ni/MgB2 tapes at 750oC - 800oC has been revealed. The Jc degradation effect presumably originates from the formation (at intermediate temperatures) of thin insulating MgB4 layers around MgB2 granules which cause their decoupling. Another possible reason of this effect is stress relaxation in the MgB2 core due Ni sheath softening, that may also cause granules decoupling and thus connectivity reduction.