The Role of CHPD and AIMI processing on enhancing JC and transverse connectivity of in-situ MgB2 strand

Research into in-situ MgB2 strand has been focused on improvements in JC through reduction of porosity. Both of cold-high-pressure-densification (CHPD) and advanced-internal-magnesium-infiltration (AIMI) techniques can effectively remove the voids in in-situ MgB2 strands. This study shows the nature of the reduced porosity for in-situ MgB2 strands lies on increases in transverse grain connectivity as well as longitudinal connectivity. The CHPD method bi-axially applying 1.0 GPa and 1.5 GPa yielded 4.2 K JCM∥s of 9.6 × 104 A/cm2 and 8.5 × 104 A/cm2 at 5 T, respectively, with compared with 6.0 × 104 A/cm2 for typical powder-in-tube (PIT) in-situ strand. Moreover, AIMI-processed monofilamentary MgB2 strand obtained even higher JCs and transverse grain connectivity than the CHPD strands.


Introduction
MgB 2 superconducting strands are promising to the practical magnetic application due to its high transition temperature T C (39 K) [1], high coherence length [2,3], and low anisotropy of upper critical fields (B c2 ) [3][4][5][6]. The powder-in-tube (PIT) in-situ MgB 2 strands were fabricated by filling a mixture of Mg and B powder into a non-reactive metallic tube and then being cold-worked into wires or tapes. The PIT strands have large amount of voids elongated along longitudinal strand axis which were left behind by molten Mg powders after heat treatment [7]. The present of voids tends to limit the number of the continuous current path in the in-situ MgB 2 strand and therefore suppress the current-carrying-capacity of the strand.
The CHPD technique can effectively increase the transport properties of PIT in-situ MgB 2 strands by eliminating the pores [8][9][10][11]. The pre-reacted powder-in-tube composite was bi-axially cold-densified at room temperature to increase the mass density of the Mg + B mixture. In this case, the cold-densified MgB 2 strand can obtain higher grain connectivity after heat treatment. Additionally, the Mg reactive -liquid -infiltration (RLI) process, initiated by Giunchi et al [12], also has the ability to eliminate the pores and produce a dense MgB 2 layers in in-situ MgB 2 strands. For RLI process, a Mg rod is inserted axially into a boron-filled metallic tube. After wire drawing the heat treatment (H. T.), MgB 2 layer is formed through the reactive diffusion of Mg into B layer. Since Mg is totally separated with precursor B layer before H. T., the RLI process can totally eliminate the "Mg-site porosity" from MgB 2 layer but induce the formation of a big hole at the central region of the strand [13]. Furthermore, since the molar volume of MgB 2 (17.46 cm 3 /mol) is twice as that of two B atoms (9.18 cm 3 /mol), the volume expansion associated with the reactive transformation from 2B to MgB 2 enables even better connections between MgB 2 grains during heat treatment [13]. Our group named our RLI-processed strands as advancedinternal -magnesium -infiltration (AIMI) strands due to their optimized strand architecture and high J C s [13].
The previous researches mostly focused on investigating the effect of CHPD and AIMI techniques on the transport properties along longitudinal strand axis of in-situ MgB 2 strands, such as transport J C s and longitudinal grain connectivity. However, Shi and Susner pointed out that high anisotropic grain connectivity exists in PIT in-situ strands, which was resulted from the elongated voids between elongated MgB 2 stringers [7]. Therefore longitudinal transport properties are different with the transport properties along transverse strand axis for MgB 2 superconducting wires. In continuing along these lines, the transport properties along transverse strand axis were investigated for the 2.0 mol% C-doped in-situ MgB 2 strands in this study. The anisotropic connectivity of the PIT strands results in the differences between perpendicular magnetic J C (J CM_|_ ) and parallel J C (J CM∥ ). The influence of aspect ratio (S = length/diameter) on transverse and longitudinal J C were investigated for the PIT strand (P00). The CHPD-processed strands were P10 (1.0 GPa cold-pressing) and P15 (1.5 GPa cold-pressing). According to the previous results of our group, the CHPD technique increased the transport J C of the monofilamentary PIT in-situ MgB 2 strand from 3.0 × 10 4 A/cm 2 to 3.6 × 10 4 A/cm 2 at 4.2 K and 10 T due to decreased porosity [11] and AIMI-processed MgB 2 strands attained the 4.2 K, 10 T transport layer J C s of 1.0 ~ 1.5 × 10 5 A/cm 2 [13][14][15][16]. In this study, we compared the J CM∥ s and transverse connectivity of the CHPD-and AIMI-processed strands with those of the P00 strand at 4.2 K and 20 K. The relationship between porosity and transverse flux pinning force density F p∥ , which is J CM∥ × B, for the in-situ MgB 2 strands was also discussed.

Sample preparation
A series of pre-reacted powder-in-tube (PIT) in-situ strands, typically 0.834 mm diameter, with a Nb barrier and a Cu outer sheath were fabricated by Hyper Tech Research, Inc. (HTR). Two PIT strands (designated P10 and P15) were bi-axially densified with 1.0 GPa and 1.5 GPa at room temperature, respectively. The other strand (designed A00) manufactured through AIMI technique were also provided by HTR. The AIMI-processed strand, with 0.55 mm diameter, has a Nb barrier and a Monel outer sheath. The powders used for the present strands were 2 mol% C-doped amorphous B (10 -100 nm) from Specialty Materials Inc. (SMI). The cold-densified PIT strands were heat-treated at 675 °C for 1 h and the AIMI strand was heat -treated at 625 °C for 16 h. The specification and heat treatment (H. T.) conditions of the strands are presented in table 1.

Transport and Magnetic Measurements
The transport I C (I CT ) test was conducted in perpendicular magnetic field up to 13 T in a pool of liquid Helium at 4.2 K on the MgB 2 strands with a total length of 50 mm and a gauge length of 5 mm. The electric criterion used for determining I CT s is 1.0 μV/cm. The magnetizations versus perpendicular and parallel magnetic fields (M -H) loops were measured by a Quantum Design Model 6000 Physical Property Measuring System (PPMS) for all strands with a sample length of 3 -5 mm.

Results
The J CT s of the PIT strands were the transport critical current normalized by MgB 2 core area. As shown in Figure 1(a) the MgB 2 core of the typical PIT in-situ strand is a solid cylinder. Figure1(b)-(d) shows the shape of MgB 2 cores for the CHPD-and AIMI-processed strands are cuboid and hollow cylinder, respectively. The J CT of the AIMI-processed strands, which is also named as transport layer J C , were calculated by dividing the I CT by the area of annulus MgB 2 layer. Values of magnetic J C (J CM_|_ and J CM∥ ) for the MgB 2 strands were extracted from the full M -H loops heights ΔM, using the standard Bean model equations [7,17]: For the PIT in-situ wire P00: Parallel Magnetic J C : J CM = 3ΔM 2R 0 Here R 0 is the radius of the cylinder MgB 2 core in the PIT wire.
For the densified wires P10 and P15: Parallel Magnetic J C : Here a, b are both lengths of the transverse cross -sectional area of cuboid MgB 2 core, a > b.
For the AIMI wire A00: Here R i is the inner diameter of the annulus MgB 2 layer and R 0 is the outer diameter of the annulus MgB 2 layer. Figure 2(a) shows the J CT and J CM_|_ versus B at 4.2 and 20 K for the strand P00. It can be seen that J CM_|_ s agree with J CT at low fields, whereas the bifurcation of J CT and J CM_|_ happened at high fields. Moreover, J CM_|_ s were greatly affected by the aspect ratio S, especially at high magnetic fields. The relationships among J CT , J CM_|_ , and aspect ratio were fully discussed in ref [7] and [18]. Therefore, the values of J CM_|_ are not only determined by the intrinsic properties but also the extrinsic properties of the in-situ MgB 2 strands. As shown in Figure 2  connections between MgB 2 fibers were enhanced. With the formation of high density MgB 2 layer, strand A00 attained 4.2 K, 10 T J CT of 9.4 × 10 4 A/cm 2 , which is 180% higher than those of CHPD-processed strands. On the other hand, the AIMI strand obtained 5 T J CM∥ s of 3.1 × 10 5 A/cm 2 at 4.2 K and 1.7 × 10 4 A/cm 2 at 20 K, which are about 240% higher than those of CHPD strands. Figure 4 shows the F p∥ /F p,max∥ versus B/B irr for all strands, where F p∥ = J CM∥ × B. It can be seen that the peak pinning occurred at b = B/B irr∥ close to 0.2 at 4.2 and 20 K, which is in agreement with the Dew-Hughes/Kramer model [19,20]. In other words, the dominant pinning centers for the all strands are also grain boundaries for the direction along transverse strand axis. According to the previous work [11], the densified wires have decreased porosities, the values of porosity are ~ 50% (p = 0) and ~ 30 % (p = 1.0 or 1.5 GPa). Since the Mg-site porosity was totally eliminated for AIMI-processed strands, the porosity of the AIMI strand is close to 0. Therefore, both CHPD and AIMI processes enable the resulting MgB 2 phases to be denser and more connected. For the discussion, it can be concluded that the enhanced J CT and J CM∥ in densified wires and AIMI wire is correlated with lower porosity (higher grain connectivity).

Transverse Flux Pinning, Porosity and Transverse Grain Connectivity
The connectivity K defined by Rowell [21] can represent the grain connectivity of MgB 2 strands. The connectivity K is calculated by the equation: Wan et al. Page 4 Here Δρ is the difference between the sample's resistivity at 300 K and the sample's resistivity at 40 K and Δρ SC is the resistivity difference for an ideal single crystal. However, transverse connectivity K ∥ is difficult to be determined for the MgB 2 wires. It has been reported that the maximum flux pinning force densities of fully-connected MgB 2 superconductor, where grain connectivity K = 100 %, are estimated to be 90 GN/m 3 at 4.2 K and 22 GN/m 3 at 20 K [22]. Therefore, we can roughly estimate the transverse grain connectivities K ∥ s of the in-situ MgB 2 strands by normalizing the F p,max∥ with the F p,max of fully-connected MgB 2 . Table 2 shows the transverse maximum flux pinning force densities F p,max∥ at 4.2 K and 20 K and estimated transverse grain connectivities for all the strands. The connectivity of 5% is achieved by the 2 mol% C-doped PIT strand, P00. The cold-densification increases the connectivity of PIT strand by 20 %. The AIMI strand A00 obtained the highest transverse grain connectivities of 20 %.

Conclusion
J CM∥ s is merely determined by the intrinsic properties of in-situ MgB 2 strands, so the current-carrying capacity of in-situ MgB 2 strands can be represented by J CM∥ as well as J CT . The CHPD of 1.0 GPa and 1.5 GPa enhanced the 4.2 K, 5 T J CM∥ from 6.0 × 10 4 A/cm 2 to 9.6 × 10 4 A/cm 2 and 8.5 × 10 4 A/cm 2 , respectively. AIMI strand attained the highest J CT and J CM∥ at 4.2 and 20 K due to the formation of a high dense MgB 2 layer. By eliminating the voids in in-situ MgB 2 strands through CHPD and AIMI technique, better connections between MgB 2 grains along transverse strand axis can be obtained in in-situ MgB 2 strands