A dilute magnetic semiconductor (DMS), which has potential to control charge and spin in a single material, is very applicable to spintronic devices.^[1–3] Since the discoveries of (Ga,Mn)As and (In,Mn)As, the III–V based DMSs have received much attention as prototypical DMS materials.^[4–6] However, in either (Ga,Mn)As or (In,Mn)As, heterovalent (Ga³⁺, Mn²⁺) or (In³⁺, Mn²⁺) substitution leads to difficulties in individual control of carrier and spin doping and seriously limited chemical solubility. These two obstacles prevent further improving the Curie temperature (T_C) of the III–V based DMS.

Recently, a series of new DMS materials with independent doping of carrier and spin have been discovered, such as Li(Zn,Mn)As and (Ba,K)(Zn,Mn)₂As₂ (BZA).^[7,8] Many progresses have been made in these new DMSs on both fundamental studies and potential applications.^[9–28] Among the new DMS materials, BZA has a maximum Curie temperature (T_C) of 230 K, which is a reliable record of carrier-mediated ferromagnetic DMS.^[29] Large single crystals of BZA have been grown.^[15] Taking advantage of single crystal-based Andreev reflection junction, the spin polarization rate of 66% is obtained for BZA, which is comparable to prototypical III–V based DMS.^[15,30]

However, the lack of reliable room temperature ferromagnetism is still one of the open questions for all DMS materials. Recently, a series of studies have concentrated on physical pressure-effect on BZA to gain an insight into the mechanism of the ferromagnetism in BZA and in turn to search for a possible way to increase T_C. Although these works showed suppression of T_C with increasing external pressure, one may wonder how lattice expansion will affect the ferromagnetism in BZA. In this article, we substitute As with Sb and P to study the negative and positive chemical pressure-effect on BZA.

Two teams of samples have been synthesized: the first is based on K-under-doped (Ba_0.9K_0.1)(Zn_0.85Mn_0.15)₂(As_1−yPn_y)₂ (Pn = Sb and P; y = 0.1 and 0.2), the second is based on K-optimal-doped (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂(As_1−yPn_y)₂ (Pn = Sb and P; y = 0.1 and 0.2). The polycrystallines of all the samples were synthesized with conventional solid reaction under the protection of high-purity argon as described in Ref. [8]. Powder x-ray diffraction (PXRD) was performed using Cu radiation with a Philips X’pert diffractometer at room temperature. The DC magnetic susceptibility was characterized by a superconducting quantum interference device (SQUID) magnetometer. A physical property measurement system (PPMS) was used for electricity transport and Hall effect measurements.

All the samples crystalize into tetrahedral BaZn₂As₂ phase (β-phase, space group , as shown in Fig. 1(a)). Figure 1(b) shows the PXRD patterns of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂(As_1−yPn_y)₂ (Pn = Sb and P; y = 0.1 and 0.2) series samples as typical examples. As expected, except some minor nonmagnetic or antiferromagnetic impurities, all the diffraction peaks move to lower degrees with higher Sb-doping level due to the larger ionic radius of Sb than As. For the same reason, P-doping moves the diffraction peaks to higher degrees. As shown in Fig. 1(b), the calculated lattice parameters further demonstrate the expansion and compression of lattice by Sb- and P-doping, respectively. The lattice parameters monotonically change with doping levels, indicating successful chemical solutions of Sb- and P-dopants.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (a) The crystal structure of (Ba,K)(Zn,Mn)2(As1−yPny)2. (b) PXRD patterns of (Ba0.7K0.3)(Zn0.85Mn0.15)2(As1−yPny)2 (Pn = Sb and P; y = 0.1 and 0.2). The mark star stands for antiferromagnetic impurity MnAs, the solid diamond for nonmagnetic As, and the open diamond for nonmagnetic Zn3As2. (c) Lattice constants and cell volumes of the samples in panel (b).

Figures 2(a) and 2(b) show the DC magnetization versus temperature (M(T)) of (Ba_1−xK_x)(Zn_0.85Mn_0.15)₂(As_1−yPn_y)₂ with x = 0.1, 0.3; y = 0.1, 0.2; and Pn = Sb, P. Herein the T_C of (Ba_0.9K_0.1)(Zn_0.85Mn_0.15)₂As₂ and (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂ are 70 K and 185 K, respectively, consistent with our previous report.^[8] For both K-under-doped and K-optimal-doped samples, Sb and P dopings steadily reduce T_C (Fig. 2(c)). The Curie–Weiss law, , where is the temperature-independent term, has been employed to fit the magnetic data above their T_C. In Fig. 2(c), the obtained paramagnetic temperatures (θ) share the same alteration tendency with T_C for varying Sb- and P-doping levels. Meanwhile, coercivity field H_C and saturation moment M_S show some non-linear relationship upon doping, particularly in K-optimal-doped (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂(As_1−yPn_y)₂ (Pn = Sb and P; y = 0.1 and 0.2). In fig 2(d), we subtract the small H-linear component in the field-dependent magnetization loops (M(H)), which is presumably due to remaining paramagnetic spins and/or field-induced polarization,^[6] to obtain the saturation moment at 2 K. H_C and M_S of 10% Sb are almost identical with those of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂, even though the former’s T_C is significantly lower than that of the latter. For the P-doped samples, H_C of 10% P is close to that of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂, but its M_S is only half of the latter. With higher P level, both H_C and M_S are apparently reduced.

Fig. 2.

Figure Option ViewDownloadNew Window

Fig. 2. The temperature dependent magnetization of (a) (Ba0.7K0.3)(Zn0.85Mn0.15)2(As1−yPny)2 (Pn = Sb and P; y = 0.1 and 0.2) and (b) (Ba0.9K0.1)(Zn0.85Mn0.15)2(As1−yPny)2 (y = 0.1 and 0.2 for Pn = Sb; y = 0.1 for Pn = P). The inset is the Curie–Weiss fit of (Ba0.7K0.3)(Zn0.85Mn0.15)2(As0.8Pn0.2)2 (c) Curie temperature and paramagnetic temperature of the samples in panel (a). (d) Field dependent magnetization of the samples in panel (a) at 2 K.

Figure 3(a) shows the temperature dependence of resistivity, ρ (T), under H = 0 T and 5 T for (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂(As_0.9Pn_0.1)₂ with Pn = Sb, As, and P. For all the three samples, the resistivity increases with decreasing temperature, indicating a semiconducting behavior. In the entire temperature range, the resistivity of the Sb-doped sample is much higher than that of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂. In contrast, the resistivity of the P-doped sample is lower than that of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂ in most range of temperature. The tendency of resistivity changing with chemical pressure is consistent with our previous physical pressure studies which exhibited pressure induced semiconductor–metal transition in BZA.^[8,10] In this work, (As,P) substitution compresses the cell volume, acting as a positive chemical pressure to increase the conductivity of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂. In contrary, (As, Sb) substitution, which extends the cell volume, produces a negative chemical pressure and thus decreases the conductivity.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (a) (Ba0.7K0.3)(Zn0.85Mn0.15)2(As1−yPny)2 (Pn = Sb and P; y = 0.1) samples’ temperature dependent resistivity. (b) Field dependent resistivity of the samples in panel (a) at 5 K.

All three samples show negative magnetoresistance (MR) at low temperature regions. As shown in Fig. 3(b), the hystereses (i.e., H_C) observed from the ρ (H) curves of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂ and the Sb-doped sample are consistent with the M(H) loops, while H_C of the P-doped one on ρ (H) is somewhat smaller than that of the corresponding magnetic loop. MR_2K (MR defined as ( )/ of (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂ is about −7% at H = 5 T, consistent with the previous report.^[8,9] In many ferromagnetic materials, negative magnetoresistance results from the reduction of spin-dependent scattering by aligning the spins in the applied field. Within this scenario, compared to (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂, the decrease of MR in the Sb-doped sample is reasonable as the Sb-doped one has weaker ferromagnetic ordering. Unexpectedly, the P-doped sample shows much larger MR than (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂, even T_C of the former sample is smaller than that of the latter. Magnetoresistance effect is one of the most attractive functions of spintronic materials or devices. This modification on MR significantly extends the potential application of BZA.

In DMS, ferromagnetic ordering is mediated by carriers. Thus Hall effect was measured to probe possible change upon Sb- or P- doping. Figure 4 shows the Hall resistivity versus field ( for (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂As₂ at 2 K and (Ba_0.7K_0.3)(Zn_0.85Mn_0.15)₂(As_0.9Pn_0.1)₂ (Pn = Sb, As, and P) at 2 K, 50 K, and 100 K. The results demonstrate hole as the major carrier for all the three samples. At 2 K, anomalous Hall effect dominates the low field region for each sample. The shape of loops is consistent with M(H) loops at 2 K for the three samples, because the anomalous Hall part is proportional to magnetization according to the phenomenological expression of the Hall resistivity, , where R₀ is the ordinary Hall coefficient, R_s is the anomalous Hall coefficient, and M is the magnetization. With increasing temperature, the anomalous Hall part significantly decreases due to the weakened magnetization. Hole concentrations (n_p) calculated with the data in the high field region are tabulated in Table 1. Considering interference from remaining anomalous Hall part even in the high field region, we argue that n_p from Hall resistivity of 100 K is closer to the authentic value than that form the low-temperature. The n_p does not decrease but even slightly increases by Sb- and P-doping. Thus, the cause of the considerable reduction of T_C with Sb- and P-doping is still an open question for future studies.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. (Ba0.7K0.3)(Zn0.85Mn0.15)2(As1−yPny)2 (Pn = Sb, As, and P; y = 0.1) samples’ field dependent Hall resistivity at (a) 2 K, (b) 50 K, and (c) 100 K.

Table 1.

Table 1.

Table 1. The hole concentrations (in units of 1020 cm−3) of (Ba0.7K0.3)(Zn0.85Mn0.15)2(As1−yPny)2 for y=0.1 . . Temperature/K Pn = Sb Pn = P Pn = As 2 1.3 1.0 1.1 50 1.2 1.3 0.9 100 1.0 1.5 1.0

Table 1. The hole concentrations (in units of 1020 cm−3) of (Ba0.7K0.3)(Zn0.85Mn0.15)2(As1−yPny)2 for y=0.1 . .

Studies of external pressure-effects on BZA revealed suppression of Curie temperature upon compression. Iso-valent substitution of Sb or P into As in BZA to produce chemical pressure is an attractive attempt to tune the magnetic properties because previous studies on physical pressure-effects on BZA have revealed the suppression of Curie temperature upon compression. Although lattice extension does not increase T_C as expected, the change of resistivity with modification of cell volume has been reproduced. Nevertheless, the large MR effect induced by P-doping offers a new boulevard to improve magnetoelectric properties for spintronic materials.

Authors are grateful for the discussion with Maekawa S.