Spark plasma sintering of MgB2 in the two-temperature route

https://doi.org/10.1016/j.physc.2012.01.023Get rights and content

Abstract

Superconducting bulks of MgB2 were obtained by an ex-situ two-temperature route applied to spark plasma sintering (SPS). Processing of samples was performed at lower temperatures than previously reported. Samples produced by the two-temperature route show a higher morphological uniformity, a higher density (above 98%), a higher Vickers hardness, and undesirable stronger microscale flux jumps, as indicated by magnetic relaxation measurements when compared to a sample obtained by the one-temperature route (95.3% relative density). At the same time, all sintered samples show approximately constant crystallite size, critical current density, irreversibility field, critical temperature, weight fraction of impurity phases (MgB4 and MgO), and the amount of carbon accidentally introduced during SPS processing.

Highlights

► MgB2 was obtained by ex-situ SPS in the two-temperature (A) route (98% density). ► Tc, Jc, Hirr are similar to a one-temperature (B) SPS sample (95.3% density). ► Crystallite size, amount of impurity phases, accidental O2 and C intake are similar. ► Sample A shows better morphological uniformity. ► A higher Vickers hardness and stronger microscale flux jumps are for sample A.

Introduction

MgB2 is a promising material for the next generation of superconductivity applications. However, owing to the high volatility of Mg, a severe problem is that the porosity in the bulks is usually high, especially when prepared by the in situ reaction between Mg and B. Low density has dramatic consequences on supercurrent-carrying area leading to a low critical current density, Jc [1], [2]. It is essential to prepare MgB2 materials with high densities and improved grain connectivity. In reality, in order to successfully compete with practical Nb-based superconductors, the challenge is to simultaneously achieve a high level densification and to control grain boundaries, disorder/morphology and defects for the enhancement of current-carrying ability of MgB2 in the absence of additions or substitutions.

Sintering is a complicated process of microstructure evolution, with the main outcome being porosity elimination. However, in polycrystals, accelerated grain growth always accompanies the final-stage sintering. On the other hand, a nanostructured MgB2 ceramic is of high interest because it is well established that grain boundaries have a strong vortex pinning effect [1]. Therefore, a higher density of boundaries in the volume unit as for a nanostructured ceramic is a powerful strategy to enhance Jc. MgB2 bulks with high density have been produced mostly by pressure-assisted methods such as hot isostatic pressing (HIP) [3], [4], hot pressing (HP) [5], [6], and spark plasma sintering (SPS) [7], [8], [9], [10], [11], [12], [13], [14], [15], [16]. In general, to obtain a nanostructured ceramic SPS is suitable because it uses high heating and cooling rates impossible to achieve with the other methods. It should be also noted that SPS has some particular features related to pulsed current application that is leading to benefic non thermal effects. Although still not well understood and controversial, literature indicates on formation of hot spots, enhanced surface electro diffusion producing a grain boundaries “cleaning” effect, spark or spark plasma phenomena [17]. Consequences are accelerated sintering, sintering temperature may be lower, and defects possibly useful for vortex pinning in the particular case of superconductors may occur. All these features of SPS allow production of MgB2 bulks of top quality (Table 1). We note that it is still unclear if thermal conditions are optimum and one can observe that especially maximum temperature in the SPS processing is within a large range. On the one hand, this is due to the processing route, namely in situ reactive SPS applied on mixtures of Mg-based powders and B or ex-situ applied on MgB2 powders. For similar pulsed current features, the differences can be due to equipment specifics (most experiments were conducted on German or Japanese SPS machines, but very often they are not specified, see Table 1), size of the graphite die/punches unit, use of a thermocouple or a pyrometer to monitor temperature, the amount of powder to be consolidated. Comparative analysis of the results is also difficult because of different raw materials and due to measurements interpretation. In such circumstances it is of interest to carefully conduct systematic experiments and apply the same equipment and procedures. This may suggest fine optimization processes and in this work we follow this idea.

Dancer et al. [13] found that above a density of 90% in the ex-situ SPS processed samples at 1250 °C there is no need for full elimination of porosity since Jc values from magnetization measurements are not influenced anymore and suggested that microstructural processes begin to be important. To investigate this statement we conducted experiments of two-temperature SPS processing. First, we applied a higher temperature T1 for a short time and, then, at a lower temperature T2, dwell time was longer. This approach was tested with excellent results on several materials for conventional sintering [18], [19], [20]. Wang et al. [18] explained the suppression of the grain growth, while the porosity is eliminated at densities above 90%, because of the higher activation energy of a grain boundary network pinned by triple points and affecting the grain growth than for the grain boundaries that are leading to porosity elimination. Usually the difference ΔT between T1 and T2 is about 100–150 °C for a non-reactive route. To the authors knowledge only one paper approaches the two-temperature route for MgB2 synthesis [21]. The paper is however for the in situ conventional two-temperature route and ΔT was 440 °C.

In this work, we applied two-temperature route on reacted MgB2, i.e., the ex-situ route. At the same time, we used SPS and the maximum temperature T1 was about 100 °C lower than the temperature used by Dancer et al. [13] for a one-temperature SPS route. Apparently, it is possible to compare our results with those from Ref. [13] since SPS machines are of the same type (FCT, Germany) and raw powders are from the same supplier (Alfa Aesar). Lowering the processing temperature is important not only for the suppression of the grain growth, but, perhaps, it is useful to control and virtually to decrease evaporation of Mg during SPS processing. Boiling point of Mg is 1090 °C [21] and, paying attention to the above uncertainties about temperature determination, strong evaporation was noted for SPS processing of MgB2 above 1050 °C [10] (machine type and temperature measurement device, i.e. thermocouple or pyrometer are not specified). Furthermore, Schmidt et al. [7] reported that peritectic decomposition of MgB2 into Mg and MgB4 occurs at 897 °C. Again, a shorter processing time, a lower maximum temperature and a two-temperature sintering might be a useful process to minimize decomposition of MgB2 and evaporation or oxidation of the free Mg.

Section snippets

Experimental

MgB2 powder (Alfa Aesar, 98% purity), with an average particle size of ∼2.3 μm given by the supplier (our SEM images show particles of 0.15–0.3 μm and aggregates of 1–2.5 μm [22], Fig. 1) was compacted using a FCT Systeme GmbH – HP D 5 SPS furnace. The MgB2 powder (3 g) was wrapped into C-paper and loaded into a graphite die and sintered with different regimes (Table 2, Fig. 2), in vacuum, and under a mechanical uniaxial pressure of 95 MPa. The heating rate was 160 °C/min, and the temperature was

Results and discussions

Samples have shown high apparent density (Table 2). The highest value of the relative density, RSPS, is for the sample TW (99.2%). Comparative to sample TWL, this sample was processed for a shorter time, but at a higher second dwell temperature. The height displacement of the samples was measured in situ, and the increase of the relative density during processing is shown in Fig. 2 left.

Curves of relative density from Fig. 2 were plotted based on displacement, d (mm), versus time, t (sec.), as

Conclusion

We obtained dense MgB2 ceramic by different SPS routes at temperatures lower than previously reported. Two-temperature route is useful to preserve the initial particle size while applying a longer processing time and attaining a high relative density. This route is also shown not to influence significantly the critical temperature, the amount of impurity phases, and the accidental carbon and oxygen intake into the samples. All these results suggest that Mg evaporation is not a problem despite

Acknowledgments

Work at NIMP was performed with financial support from PCCE 9/2010, Romania. PB acknowledges financial support from MANA, Japan and DB from the European Social Fund through POSDRU/89/1.5/S/54785 project. Authors thank Dr. M. Enculescu for technical assistance with SEM.

References (39)

  • C.F. Liu et al.

    Physica C

    (2003)
  • M.B. Maple et al.

    Physica C

    (2002)
  • A. Tampieri et al.

    Physica C

    (2004)
  • C.H. Hsieh et al.

    Solid State Commun.

    (2006)
  • J. Schmidt et al.

    Solid State Sci.

    (2003)
  • K.J. Song et al.

    Physica C

    (2005)
  • H.-L. Suo et al.

    Physica C

    (2007)
  • L. Lutterotti

    Nuclear Inst. Methods Phys. Res. B

    (2010)
  • G.K. Williamson et al.

    Acta Metall.

    (1953)
  • S. Lee et al.

    Physica C

    (2004)
  • L. Miu et al.

    Physica C

    (2008)
  • A.P. Malozemoff

    Physica C

    (1991)
  • D.C. Larbalestier

    Nature

    (2001)
  • Y. Takano et al.

    Appl. Phys. Lett.

    (2001)
  • S.Y. Lee et al.

    J. Am. Ceram. Soc.

    (2003)
  • S.H. Shim et al.

    J. Am. Ceram. Soc.

    (2005)
  • D.K. Kang et al.

    Met. Mater. Int.

    (2009)
  • C.E.J. Dancer et al.

    Supercond. Sci. Technol.

    (2009)
  • G. Aldica et al.

    J. Optoelectron. Adv. Mater.

    (2007)
  • Cited by (0)

    View full text