1 Introduction

The success of cuprate superconductors promoted the novel iron-based superconductors as the new challenge both for theorists and applications. Similarly to cuprates, these materials possess a layered structure with superconducting layers sandwiched between charge reservoirs, and the superconductivity is induced by charge doping of an antiferromagnetic parent [1–5]. These materials preserve a short coherence length [6], hence, a very high upper critical field H c2 (over 100 T [7–10]), but also a much smaller anisotropy. In addition, the symmetry of the order parameter is nodeless [11–13]. Therefore, the detrimental effect of the grain boundaries, which is the main drawback of the cuprate superconductors, is substantially reduced in Fe-based superconductors [14]. This is an advantage that compensates the smaller critical temperature T c compared to cuprates. The compounds with the highest critical temperature belong to the fluorine doped LnFeAsO (Ln-1111) family where Ln is a rare earth element. Magnetic critical current density J c was found to be as high as 2×106A cm−2 at 5 K and 14 T [15] in single crystals. However, the wires made of this material can only carry current densities of order 103A cm−2 [16–18], a limitation imposed by the strong granularity of several origins [19–21]. However, it is a very difficult-to-handle compound because of the unavoidable spurious, non-superconducting phases like FeAs [16, 22, 23], which enhances the granularity, and the lability of fluorine, which makes its control difficult during sintering process.

Improvements might be obtained by the addition of several compounds which attenuate the granularity and its effects. The procedure was successfully applied to Sr1−x K x Fe2As2 (Sr-122) compound in which Ag [23–26] or Sn [27] addition proved to considerably improve the grain connectivity. In 1111 compound, the addition of tin provided good results leading to record J c for the Sm-1111 [28]. However, Ag was not used in the 1111 samples yet, though it proved to be useful as sheath for the wires obtained by powder-in-tube technology [29, 30].

Before taking any step forward, it is important to evaluate the intragranular critical current density when silver is added to the Sm-1111 superconductor. Therefore, in this paper we investigate the transport and magnetic properties of a series of Ag-added SmFeAsO1−x F x polycrystalline samples, focusing on the effect on the intragranular critical current density.

2 Experimental

A series of Ag-enriched SmFeAsO1−x F x (Sm-1111) samples were prepared by a two-step solid state reaction method using SmAs as a precursor, together with Fe, Ag, Fe2O3 and FeF3. Details on the preparation were presented in [31]. The nominal composition of the samples is SmFeAsO1−x F x Agy with 0≤y≤0.12. The X-ray diffraction (XRD) (Brucker Advance) and scanning electron microscopy (SEM) were used for the structural and morphological characterization of the samples. The magnetic properties were measured using a SQUID magnetometer (MPMS-Quantum Design) at fields up to 7 T whereas the electrical transport properties were measured in applied magnetic fields up to 14 T with a PPMS equipment (Quantum Design) using the standard four-probe method.

3 Results and Discussions

3.1 Structure and Morphology

The X-ray diffraction data show the presence of the dominant Sm-1111 phase but also a series of additional phases like FeAs, Sm2O3, SmOF, SmFe3, and unreacted silver (Fig. 1). These additional phases are the main drawbacks which prevent the development of Ln-1111 cables, despite the high critical temperature.

Fig. 1
figure 1

X-ray diffraction pattern of SmFeAsO1−x F x Ag y . Indexed peaks belong to the Sm-1111 phase; Silver is marked as Ag. Other phases are marked with symbols: (⧫) SmF3; (□) FeAs; (•) Sm2O3; (∘) SmOF

Full size image

The microstructure of the samples is almost the same for both pristine Sm-1111 and silver-containing samples. We show here only the pristine sample (Fig. 2(a)) and the sample with the highest silver content (Sm-1111Ag0.12) (Fig. 2(b)). Both micrographs show well grown, thick plate-like grains of 2 μm average size and also pores of micrometer size located at the grain boundaries.

Fig. 2
figure 2

SEM micrographs of polycrystalline Sm-1111 samples: (a) pristine Sm-1111; (b) Sm-1111-Ag0.12

Full size image

3.2 Resistive Transition

Figure 3 shows the typical temperature dependence of the resistance up to 300 K for three SmFeAsO1−x F x Ag y samples: y=0.0; 0.06 and 0.12. The addition of silver decreases the critical temperature T c from 54.1 K for y=0 to 43.0 K for y=0.12. Despite this decrease of T c, all samples have similar non-Fermi temperature dependence of the resistance in the normal state. Specifically, it shows a remarkable linearity of the R(T) curves just beyond the critical temperature T c, which extends in a temperature range of almost 100 K followed by a slowing down to a logarithmic dependence R(T)=A(x)log(T/Δ) for temperatures higher than a crossover temperature T cr (Fig. 4). T cr was determined as the point of equal departure of the linear and logarithmic fit from the experimental R(T) curve, i.e., T cr is the temperature for which |R(T cr)−linear fit(T cr)|=|R(T cr)−logarithm fit(T cr)| (see the inset to Fig. 4). The linear-in-T dependence of the resistance at low temperatures is a signature of the quantum criticality [32]. It is typical for all types of superconducting pnictides [33] and marks the suppression of the spin density waves by doping.

Fig. 3
figure 3

Temperature dependence of the normalized electrical resistivity of polycrystalline SmFeAsO1−x F x Ag y

Full size image
Fig. 4
figure 4

Temperature dependence of the electrical resistivity of polycrystalline SmFeAsO1−x F x Ag0.06 and the fit with linear and logarithmic function. Inset shows the determination of the crossover temperature

Full size image

If the low temperature linear dependence can be explained by the scattering on the fluctuations associated to the quantum critical point [32], the logarithmic dependence is still hard to understand. Anyway, the silver content has no effect on the functional form of the R–T curves despite the continuous decrease of the critical temperature. There is no indication on the possible recovery of the SDW state which means that superconductivity is not influenced by the addition of Ag. However, the value of T cr is almost identical to the structural phase transition in the undoped parent SmFeAsO which raises some questions related to the presence of the antiferromagnetic correlations and their role [34]. The dependence of the critical and crossover temperatures, T c and T cr, respectively, on the amount of silver is shown in Fig. 5. It is obvious that for 0≤y≤ 0.06 both temperatures show minor changes with T cr around 160 K, and T c is slightly decreasing from 54 to 53 K. At higher Ag content, both temperatures steeply fall. A similar tendency is shown by the energy scale Δ which appears in the logarithmic dependence of the resistance at high temperatures (Fig. 6). It is noticeable that the high critical temperature requires a high Δ value as the plot of Δ vs T shows (see the inset to Fig. 6). In addition, we find that the normal state is very robust in applied field up to 14 T. Specifically, the functional dependence of the R–T curves, T cr and Δ are field-independent in this field range.

Fig. 5
figure 5

The dependence of the critical temperature T c and crossover temperature T cr as a function of Ag content in SmFeAsO1−x F x Ag y

Full size image
Fig. 6
figure 6

The dependence of the energy scale Δ on the silver content y in polycrystalline SmFeAsO1−x F x Ag y . Inset shows the correlation between Δ and the critical temperature T c

Full size image

3.3 Magnetic Susceptibility

The low field (10−3 T) dc-magnetic susceptibility shows that the sample with y=0 is not single phase in structure but displays high temperature transition at 53.6 K and a large transition at about T c2=30 K (see Fig. 7). The addition of silver makes the transition steeper though the critical temperature decreases. In addition, the magnetization gets positive values as the temperature gets closer to T c even at this low field. That is, the first proof of the strength of the paramagnetic background. The origin of this magnetic background is complex. Any attempt to depict the normal state susceptibility with either the Curie–Weiss relationship or with Brillouin function fails. The samples containing silver start displaying a two step transition only if a magnetic field higher than 0.1 T is applied (Fig. 8) but with the second transition shifted below 20 K. This second transition is not visible in the field cooled regime (see inset to Fig. 7).

Fig. 7
figure 7

Low field (1 mT) temperature dependence of the magnetic moment of the polycrystalline SmFeAsO1−x F x Ag y . Inset shows field cooled (FC) and zero field cooled (ZFC) magnetizations of the pristine sample SmFeAsO1−x F x

Full size image
Fig. 8
figure 8

Temperature dependence of the static magnetic susceptibility of polycrystalline SmFeAsO1−x F x Ag y : (a) samples measured at an applied field of 0.1 T; (b) samples measured at 1 T. Inset shows samples measured at 5 T in the field cooled regime

Full size image

Generally, the sharp decrease below T c2 of the ZFC curves is attributed to the onset of the intergrain superconductivity. Specifically, it marks the flux penetration inside the individual grains.

As the SEM micrographs show (Fig. 2), the degree of granularity is important and exactly the intergrain week links impose the global critical current density. This means that silver substantially reduces the width of transition and makes the week links transparent enough to the Cooper pair but the cost is the slight reduction of the critical temperature (Fig. 7). However, relative modest magnetic fields, about 0.1 T, suppress the intergrain, silver-mediated superconductivity, as the second transition shows up. Above this field, large portions of the ZFC curves are positive in the temperature range used for measurements.

Actually, at high magnetic field, the result is more puzzling. First, in the normal state, above the critical temperature, the silver-free sample shows magnetic susceptibility which decreases with increasing field, whereas below the second transition, in the ZFC regime, the susceptibility shows an opposite trend (see Fig. 9), and the steep decrease of the superconductivity shows that superconductivity becomes robust. The different trends below and above critical temperature are the result of the presence of several phases with different ferromagnetic ordering which reach the saturation around T c and, consequently, their magnetic susceptibility decreases with increasing B. At fields higher than 2 T, a noticeable upturn can be seen for T c2<T<T c. In this temperature range, the magnetization is positive, suggesting a dominant contribution from paramagnetic or magnetically ordered phases, though superconducting phases are present. When the sample is cooled in the field, the trapped field emphasizes this effect but the second transition is no more visible down to 5 K (inset to Fig. 8(b)).

Fig. 9
figure 9

Temperature dependence of the zero-field cooled static magnetic susceptibility of polycrystalline SmFeAsO1−x F x measured for different applied fields. Inset shows the same plot for the field cooled susceptibility

Full size image

The samples containing silver display a smaller magnetic background. As Fig. 8(b) shows, the contribution of the magnetic component in the normal state is five times higher in the Ag-free sample than in the samples containing Ag both at 1 T (main panel) and in 5 T (the inset). Actually, the decrease of magnetic background is not monotonous. For y>0.6 the magnetic background starts to increase, whereas the critical temperature shows a continuous decrease. Since the magnetic behavior does not significantly change below T c, the reduced magnetic background leads to a negative susceptibility even at 1 T, whereas the Ag-free sample already has a positive susceptibility at this field. Overall, the silver within ceramic SmFeAsO1−x F x dilutes the amount of paramagnetic ions but also the fraction of superconducting component.

3.4 Critical Current

The field dependence of the magnetization was used to obtain the critical current density J c with the Bean relationship. Due to the polycrystalline structure of the sample and because of the short coherence length which is typical for superconducting pnictides, the Bean relationship is appropriate to obtain the intragranular critical current density J cg. Figure 10 shows the field dependence for all sample investigated at temperatures between 5 and 30 K. We noticed four regimes of field dependence which are similar to the regimes depicted for single crystals [35]. At very low fields, there is a slow decrease of J cg. As the field increases, J cg crosses to a power law dependence, J cg∼B −α, which is followed an almost B-independent J cg and, finally, by fast decrease. At 2 K, the exponent α is almost constant and takes values around 0.5 but it increases with the temperature to 0.75–0.80 at 20 K. Ta value α=0.5 was proposed by Vinokur et al. to describe pinning by dense point pins in layered superconductors [36] while α=5/8 is consistent with the pinning by sparse large point defects larger than the coherence length ξ [37]. The variation of the exponent α with temperature is usually attributed to the flux creep, hence, is enhanced as the temperature increases. The B-independent plateau at higher fields is attributed to the collective pinning of single vortex arising from the atomic scale fluctuations of the dopant positions which leads to mean free path fluctuations [38, 39]. At even higher fields it crosses to the pinning of small bundles regime which follows an exponential decrease, J cg∼exp[−(B/B 0)3/2] [40]. It is noticeable that a small amount of Ag (y=0.04) leads to an enhancement of J cg which is more consistent at low fields but also reduces the field range of collective pinning regime. As the content of silver is higher than y=0.06, the intragranular critical current density decreases, though all regimes are preserved. The samples with silver display a faster decrease of J cg, and that effect is emphasized at those samples with high content of silver.

Fig. 10
figure 10

Field B dependence of intragranular critical current density J cg of polycrystalline SmFeAsO1−x F x Ag y at different temperatures: (a) 5 K; (b) 15 K; (c) 20 K; (d) 30 K. The fits with the functions typical for the two regimes are shown with dashed lines

Full size image

It is not clear why a small amount of silver enhances J cg whereas for y>0.06 the effect is detrimental. The X-ray diffraction data did not reveal changes of the lattice parameter; hence, silver doping is not expected. However, the onset of the critical temperature shows a continuous decrease when Ag is added both in resistive data (Fig. 5) and magnetization measurements (Fig. 7). In the absence of high resolution microscopy data we can only speculate. A hint might be given by Fig. 7 which clearly shows the enhancement of the transition as the Ag is added. It is clear that silver contributes to the stabilization of the high temperature phase but the price is the decrease of the critical temperature of that phase. Most likely it is a problem of fluorination which seems very inhomogeneous in the pristine sample (y=0), hence, silver improves the homogeneity of the dopant distribution but, to a certain degree, prevents the optimal fluorination for y>0.06. This fact increases the density of the deeper pinning potential wells sites on the account of the more shallow ones. However, as the T c decrease becomes significant, the pinning gets weaker. Therefore, the intragranular critical current density is the result of a subtle balance between the decrease of the critical temperature, hence, of the density of Cooper pairs, and the enhancement of the homogeneity. Surface pinning may also play a role [35]. Silver is detrimental for the glassy FeAs which is found at grain boundaries where it is considered to suppress surface superconductivity.

4 Conclusions

The addition of silver to the polycrystalline Sm-1111 superconductor was investigated by transport and magnetic measurements. The silver addition does not significantly change the superconducting mechanism, keeping the compound out of the influence of SDW. The magnetic data show a very important effect of Ag. Specifically, it enhances the homogeneity of the samples promoting the development of the high temperature phase on the account of lower, more disordered or less fluorinated phases. That effect is, however, accompanied by a decrease of the critical temperature. This is more consistent for high content of Ag, though the X-ray diffraction data do not suggest any Ag diffusion inside the Sm-1111 grains which would change the electronic structure. What is most important, the magnetic intragranular critical current density shows an enhancement for small amount of Ag (y<0.6) in all field regimes we could explore but a depression at higher content. We attribute that behavior to the balance between the enhancement of the homogeneity and depression of the critical temperature when silver is added.