Tuning Magnetic and Transport Properties in Quasi-2D (Mn_1−xNi_x)₂P₂S₆ Single Crystals

Abstract

We report an optimized chemical vapor transport method to grow single crystals of (Mn_1−xNi_x)₂P₂S₆ where x = 0, 0.3, 0.5, 0.7, and 1. Single crystals up to 4 mm × 3 mm × 200 μm were obtained by this method. As-grown crystals are characterized by means of scanning electron microscopy and powder X-ray diffraction measurements. The structural characterization shows that all crystals crystallize in monoclinic symmetry with the space group C2/m (No. 12). We have further investigated the magnetic properties of this series of single crystals. The magnetic measurements of the all as-grown single crystals show long-range antiferromagnetic order along all principal crystallographic axes. Overall, the Néel temperature T_N is non-monotonous; with increasing Ni²⁺ doping, the temperature of the antiferromagnetic phase transition first decreases from 80 K for pristine Mn₂P₂S₆ (x = 0) up to x = 0.5 and then increases again to 155 K for pure Ni₂P₂S₆ (x = 1). The magnetic anisotropy switches from out-of-plane to in-plane as a function of composition in (Mn_1−xNi_x)₂P₂S₆ series. Transport studies under hydrostatic pressure on the parent compound Mn₂P₂S₆ evidence an insulator-metal transition at an applied critical pressure of ~22 GPa.

1. Introduction

In recent years, research of functional two-dimensional (2D) materials has stimulated activities aimed at the synthesis of new materials and studies of their functionalities. This class has drawn a great deal of interest and attention chiefly because of its unique electronic [1,2,3], magnetic [4,5,6,7,8], and optical properties [9,10,11,12] to the bulk counterpart. Layered van der Waals (vdW) materials dominate among the current 2D materials. The vdW interaction only weakly couples the atomic layers together along the c axis, even in a bulk vdW material, and as a result, the electron confinement in the 2D lattice often leads to some specific properties, such as anisotropic magnetic behavior [13,14] and an anisotropic conductivity [15,16,17]. Fabricating heterostructures and optoelectronic nanodevices using these materials have resulted in impressive findings in different fields [1,18,19,20,21].

The transition metal phosphorus trichalcogenides (TM₂P₂Ch₆ TM = transition metal; Ch = chalcogen) family belongs to the layered van der Waals materials class [22]. All of these materials have very similar crystal structures, a monoclinic unit cell with the space group of C2/m but different magnetic properties depending on the underlying 3d transition metal [23]. The TM₂P₂Ch₆ compounds are all semiconductors at ambient pressure [22], with band gaps for most of them larger than 1 eV as determined by optical measurements [24] with extremely high room-temperature resistivities.

Band structure calculations [25,26] lead to the conclusions that TM₂P₂Ch₆ are Mott insulators, and by applying external pressure, these materials could be driven to an insulator-metal or Mott transition [27]. These results further stimulated substantial interest to tune the parameters of the system to an intermediate state to access the physics that is not yet fully understood. Additionally, many superconducting materials are low dimensional and lie close to antiferromagnetic Mott-insulator phases in phase diagrams. Theoretical calculations [28] show that these states increase the possibility of introducing superconductivity. Thus, tunable 2D antiferromagnetic Mott insulators provide a promising playground to investigate the basic mechanisms of several unsolved problems in condensed matter physics.

Field-effect transistors based on bulk and few-layer Mn₂P₂S₆ have been fabricated [29]. These devices show the potential for good ultraviolet photodetectors. Bulk and few-layer Mn₂P₂S₆ on top of indium titanium oxide (ITO) coated Si substrate also shows tunneling transport phenomena [30]. The estimated barrier height of thin Mn₂P₂S₆ flakes is 1.31 eV (±0.01) [30]. These investigations bring the magnetic van der Waals material Mn₂P₂S₆ very close to various applications such as field-effect transistors and magnetic tunnel junction fabrication.

Mn₂P₂S₆ and Ni₂P₂S₆ are well-known members of the transition metal phosphorus trichalcogenides family. They present interesting anisotropic antiferromagnetism below the Néel temperatures of 80 K for Mn₂P₂S₆ and 155 K for Ni₂P₂S₆ [31,32], which is not well understood. The main difference between these two pristine compounds and the key ingredient in the TM₂P₂Ch₆ family is the underlying anisotropy that dictates the long-range magnetic order down to the monolayer, which eventually depends on the 3d transition metal ions [33,34,35,36]. This anisotropy of the system can be tuned systematically by several methods, for example, with chemical substitution or with applied pressure. Motivated by this, here we present two different ways of tuning the ground state of Mn₂P₂S₆ single crystals.

In this work, we report the optimized synthesis and crystal growth conditions of (Mn_1−xNi_x)₂P₂S₆. We have investigated the crystal structure and magnetic properties of the series of single crystals. We observe two different effects on the magnetic properties, namely a shift in the long-range order temperature T_N and a broad anomaly at higher temperatures above T_N, which becomes very sharp for selected compositions. Additional transport studies under hydrostatic pressure evidence an insulator-metal transition at a critical pressure of ~22 GPa.

The present article is organized in the following order. In the first section, we present the optimized methods for crystal growth for our series of crystals, (Mn_1−xNi_x)₂P₂S₆. In the later sections, results of scanning electron microscopy, energy-dispersive X-ray spectroscopy (SEM-EDX), and X-ray diffraction (XRD) are shown. Finally, detailed magnetic and electrical transport measurements are presented.

2. Experimental Methods and Crystal Growth

The composition of our crystals was determined using energy-dispersive X-ray spectroscopy (EDX), with an accelerating voltage of 30 kV. A scanning electron microscope was used to obtain electron microscopic images using two types of signals: the backscattered electrons (BSE) for chemical contrast and the secondary electrons (SE) for topographic contrast. X-ray powder diffraction data were collected using a STOE powder laboratory diffractometer in transmission geometry with Cu-$K_{\alpha 1}$ radiation (the wavelength (λ) is 1.540560 Å) from a curved Ge (111) single crystal monochromator and detected by a MYTHEN 1 K 12.5° linear position-sensitive detector manufactured by DECTRIS (Baden, Switzerland). An XRD pattern of a polycrystalline sample was obtained by grinding as-grown single crystals.

Temperature and field-dependent magnetization studies were performed using a Quantum Design Superconducting Quantum Interference Device Vibrating Sample Magnetometer (SQUID-VSM, from Quantum Design, San Diego, CA, USA). The measurements were performed for field-cooled (FC) conditions and for two crystallographic directions, i.e., parallel and perpendicular to the growth direction of the crystal.

High-pressure transport experiments were performed in a screw-pressure-type DAC made of CuBe alloy. A pair of anvil culets of 300 μm was used. A mixture of epoxy and fine cubic boron nitride (c-BN) powder was compressed firmly to insulate the electrodes from the steel gasket. A single-crystal flake with dimensions of ~120 × 45 × 10 μm³ was loaded together with NaCl fine powder and ruby powder. A five-probe configuration was utilized to measure the resistance and Hall resistance of the flake, where the external magnetic field was perpendicular to the surface of the flake. The ruby fluorescence shift was used to calibrate the pressure at room temperature in all experiments [37].

Single crystals of (Mn_1−xNi_x)₂P₂S₆ were grown by the chemical vapor transport technique. All preparation steps were performed in a glovebox before sealing the ampule. The proper weights of the starting materials manganese (powder, Alfa Aesar, 99.9998%), nickel (powder-100 mesh, Sigma Aldrich, Kandel, Germany, 99.99%), red phosphorus (lumps, Alfa Aesar, Haverhill, MA, 99.999%), and sulfur (pieces, Alfa Aesar, 99.999%) with a molar ratio of (Mn_1−xNi_x):P:S = 1:1:3 (for x = 0; 0.3; 0.5; 0.7; 1) and a total charge mass of 1 g were thoroughly mixed with iodine (0.05 g). The starting material was loaded in a quartz ampoule (6 mm inner diameter, 2 mm wall thickness) and then was cooled by liquid nitrogen to avoid significant losses of the transport agent during the evacuation process to a residual pressure of 10⁻⁸ bar. The ampoule was sealed under static pressure at a length of approximately 12 cm by the oxy-hydrogen flame. Then, the closed ampules were heated in a two-zone furnace. A similar approach was used by us for the crystal growth of the closely related sister compounds such as (Fe_1−xNi_x)₂P₂S₆ and AgCrP₂S_6. [38,39].

The optimized temperatures for Ni₂P₂S₆ [35,38] were chosen as 750 °C and 700 °C for the hot and the cold zone, respectively, and for Mn₂P₂S₆ [40,41] as 680 °C and 630 °C. For the substituted (Mn_1−xNi_x)₂P₂S₆ samples, due to the absence of published information, the following temperature profile has been optimized by us. Initially, the furnace was heated homogeneously to 300 °C with 100 °C/h and dwelled for 24 h to provide a pre-reaction of P and S with the transition elements. After that, an inverse transport gradient is applied to transport the particles adhering to the walls to one side of the ampoule, which is the charge region, to avoid the formation of random nucleation centers. Later, the charge region was heated to 720 °C in 4 h with a dwelling time of 336 h, whereas the sink region was initially heated up to 770 °C in 4 h and dwelled at this temperature for 24 h. Later, the temperature in the growth zone was gradually reduced during one day to 670 °C to slowly form the transport temperature gradient for controlling nucleation and was held at this temperature for 289 h. As a result, the temperature gradient was set for vapor transport between 720 °C (charge) and 670 °C (sink) for 12 days. In the final stage, the charge region was cooled to the sink temperature in 1 h before both regions were furnace-cooled to room temperature.

Thin lustrous black and green in the Mn₂P₂S₆ case plate-like crystals perpendicular to the c axis in the size of up to 4 mm × 3 mm × 200 μm were obtained (as shown in Figure 1. All as-grown crystals show a layered morphology, and they are easily exfoliated by scotch tape.

3. Characterization: Compositional and Structural Analysis

All as-grown single crystals of the series exhibit the typical features of a layered van der Waals structure, which is in line with the pristine compounds (Figure 1). As an example, crystals are shown in Figure 1a–e with clear hexagonal layers and a well-defined flat facade with 120° angles, which clearly indicates that the crystals grew along the symmetry axes. Backscattered electron (BSE) images of our samples have no change of chemical contrast on the surface of the crystals, as shown in Figure 1f–h. This indicates a homogeneous elemental composition on the respective area of the crystal.

To gain further information about the distribution of elements for our (Mn_1−xNi_x)₂P₂S₆ samples, an elemental mapping was performed on an area of the crystal with dimensions of about 150 × 100 μm² as shown in Figure 2a–f. The results indicate a homogeneous distribution of Mn, Ni, P, and S.

The composition of the as-grown single crystals was determined by energy-dispersive X-ray spectroscopy (EDX). The result of EDX measurements highly depends on the sample topography [42]. EDX was mapped out via measuring different areas and points on the surface of several crystals representing all doping concentrations (see Figure 2 for representative results on (Mn_0.7Ni_0.3)₂P₂S₆). As shown in

Table 1. Compositional analysis of as-grown single crystals of (Mn_1−xNi_x)₂P₂S₆.

xnom	Expec. Comp.	Mean Comp. (EDX)	xexp
0	Mn2P2S6	Mn2.02(5)P2.04(1)S5.93(5)	0
0.3	Mn1.4Ni0.6P2S6	Mn1.45(1)Ni0.55(1)P2.03(6)S5.96(2)	0.28
0.5	MnNiP2S6	Mn0.80(6)Ni0.91(7)P2.07(9)S6.19(8)	0.45
0.7	Mn0.6Ni1.4P2S6	Mn0.62(4)Ni1.33(3)P2.04(1)S6.00(2)	0.67
1	Ni2P2S6	Ni2.04(6)P2.03(2)S5.92(3)	1

, all (Mn_1−xNi_x)₂P₂S₆ compounds show the expected composition. The experimental value x_exp is found in the range of the nominal value x_nom, considering a systematic uncertainty of this ratio of approximately 5%.

The structural characterization and phase purity were confirmed by powder X-ray diffraction using a STOE powder diffractometer (STOE, Darmstadt, Germany). All compositions show the same monoclinic space group C2/m (No. 12). There were no additional impurity phases in the pXRD patterns. Figure 3 shows the pXRD patterns for the full series of samples (Mn_1−xNi_x)₂P₂S₆. There is a systematic shift of all the Bragg peaks toward lower angles with increasing manganese content. Significant changes in relative intensities are observed due to a strongly preferred orientation of the plate-like powder crystallites. The interlayer spacing d is calculated from the peak position using Bragg’s law $n\lambda =2dsin\theta$ and plotted in Figure 4 as a function of Ni substitution for (Mn_1−xNi_x)₂P₂S₆. The interlayer spacing and, therefore, the van der Waals gap decrease linearly with x due to the smaller atomic radius of Ni compared to Mn.

All obtained pXRD patterns were described by the atomic model proposed by Klingen et al. for Fe₂P₂S₆ [43]. In this model, the transition metal ions create hexagons that are octahedrally surrounded by Sulfur. They form a honeycomb structure in the ab plane. P-P pairs are located in the void of each honeycomb. These layers are poorly bonded to each other by van der Waals bonds, which giving rise to a layered structure. Crystal structures projecting different crystallographic axes of (Mn_1−xNi_x)₂P₂S₆ are shown in Figure 5a–c.

The Rietveld refinement of the X-ray data was carried out using the FULLPROF software for a 30% substituted sample, i.e., (Mn_0.7Ni_0.3)₂P₂S₆. The measured pattern (red circles) is shown in Figure 6 together with the calculated pattern based on the Rietveld analysis (black line), the difference between measured and calculated pattern (blue line), and the calculated Bragg positions for a monoclinic unit cell with the space group C2/m. The calculated peak positions and intensities are in satisfactory agreement with the experimental data.

Table 2. Structural parameters and residual factors of Rietveld refinement.

pXRD
Composition	(Mn0.7Ni0.3)2P2S6
Space group	C2/m (No. 12)
Wavelength (Å)	1.540560
2θ range (◦)	10–100
Step Size (◦)	0.015
Temperature (K)	293
a (Å)	6.009 (7)
b (Å)	10.405 (3)
c (Å)	6.779 (3)
β (◦)	107.45 (6)
V (Å3)	404.40 (6)
Goodness-of-Fit	2.06
Bragg R-factor	7.61
RF-factor	22.23

summarizes the results of the structural refinement and the lattice parameters. Note that the presence of a high concentration of stacking faults is a well-known problem in layered vdW compounds and was observed for Ni₂P₂S₆ in Refs. [45,46]. Moreover, stacking faults in the samples manifest by an asymmetry of the 00l reflections in the pXRD patterns. This might explain the high value of our R-factor from our X-ray analysis.

4. Magnetization and Transport

4.1. Magnetic Properties

Figure 7a–d shows the magnetization measurements as a function of temperature for the entire series of (Mn_1−xNi_x)₂P₂S₆ single crystals with a magnetic field of µ₀H = 1 T applied parallel and perpendicular to the ab plane. The overall behavior of M/H is very similar for all compositions with two prominent features in the magnetization curve, i.e., a broad maximum at elevated temperatures, T_max, related to (low-dimensional) magnetic spin correlations in these quasi-two-dimensional van der Waals materials [34,35,36], together with an inflection point at a somewhat lower temperature, T_N, signaling the phase transition into a long-range ordered antiferromagnetic state. The Néel temperature T_N is defined from the sharp peak in the derivative dχ/dT for H along the crystallographic c axis. Overall, the doping dependence of T_N is non-monotonous (see also Figure 8): While the end member Mn₂P₂S₆ has T_N ~ 77 K, with increasing Ni²⁺ content, the temperature of the antiferromagnetic phase transition first decreases up to x = 0.5 and then increases up to T_N = 155 K for Ni₂P₂S₆ (x = 1).

The single-crystalline nature of our samples allows us to further comment on the orientation dependence of the normalized magnetization with respect to the alignment in a magnetic field, viz., the magnetic anisotropy and its evolution as a function of Ni substitution. Below T_N, for Mn₂P₂S₆, we find M/H_{|| c} < M/H_||ab with magnetic moments oriented along the c direction in agreement with the literature [34]. Upon Ni substitution, this anisotropy is gradually suppressed up to about x ~ 0.7 before it gets reversed for even higher substitution levels up to x = 1 (i.e., for Ni₂P₂S_6, the ordered magnetic moments are oriented in the ab plane). A clear anisotropic behavior of T_max is observed for all intermediate compounds (Mn_1−xNi_x)₂P₂S₆ (x_Ni = 0.3, 0.5, 0.7). It begins to appear already for x = 0.3 and reaches the maximum anisotropy at x = 0.7, then decreases again.

The observed shift of T_N to lower temperature compared to the parent compound is similar to the behavior of bimetallic (Mn_1−xM_x)₂P₂S₆ substituted with diamagnetic ions Mg²⁺ or Zn²⁺. In that series, however, for x > 0.3, a percolation threshold was observed in Refs. [46,47,48,49]. In our (Fe_1-xNi_x)₂P₂S₆ substitution series [38], we also observed a gradual evolution of the magnetic transition temperature up to x = 1. Note that while T_N shifts in both series, (Mn_1−xNi_x)₂P₂S₆ and (Fe_1−xNi_x)₂P₂S₆, a minimum in the T_N(x) dependence was only observed for (Mn_1−xNi_x)₂P₂S₆. Moreover, the magnetic anisotropy, as well as the broad maximum at T_max, show two distinctly different behaviors for (Fe_1−xNi_x)₂P₂S₆, one up to x_Ni = 0.9, which is similar to the magnetic anisotropy of Fe₂P₂S₆ and another one for Ni₂P₂S₆, whereas a gradual but non-monotonous behavior is observed for T_max(x) in our (Mn_1−xNi_x)₂P₂S₆ series.

Note that the parent compound Ni₂P₂S₆ has a broad maximum well above T_N, which gets more pronounced for Mn substitution and becomes the dominant feature for Mn₂P₂S₆. The shape of the maximum seems to be determined by the difference between T_max and the inflection point. The smaller the difference, the sharper the maximum becomes. Ni₂P₂S₆ shows the highest T_max as well as the highest T_N, whereas (Mn_0.5Ni_0.5)₂P₂S₆ shows the lowest T_max and T_N and they appear very close to each other in contrast to both parent compounds.

Figure 9a–e shows the magnetization as a function of the applied field for (Mn_1−xNi_x)₂P₂S₆ at T = 1.8 K. The field was applied parallel and perpendicular to the ab plane. A linear dependence of M vs. H is observed for fields in the ab plane for all compositions, except for x = 1. The largest difference between the curves for H_||ab and H_||c is observed for higher nickel contents x = 1 and 0.7 (Figure 9,e). For x = 0.5, the curves overlap, and then, clear spin-flop transitions are observed at µ₀H_c = 1.9 T and 3.7 T for x = 0.3 and 0, respectively, for magnetic fields applied parallel to the c axis (see Figure 9a,b). The spin-flop is also seen for x = 0.5, but it is very weak and broadened. Figure 9f shows the difference $\Delta M=M\left(H_{\left|\right|ab}\right)-M\left(H_{\left|\right|c}\right)$ as a function of the magnetic field at T = 1.8 K. (Mn_1−xNi_x)₂P₂S₆ samples with x < 0.5 show a positive pronounced $\Delta M$ in the positive field direction in line with an easy-axis anisotropy. In samples with x > 0.5, this behavior is inversed, showing an easy-plane anisotropy. The change of the anisotropy is mainly determined via the interplay between the single-ion anisotropy and the crystal field effect (CEF) schemes together with the effects of the spin orbit coupling (SOC) [34]. While the single-ion anisotropy and the trigonal distortion are small/negligible for Mn, they are positive (D > 0) and large for Ni. Such a compositional-dependent switch of the magnetic anisotropy has also been observed in other van der Waals systems such as mixed halides [50]. The observed compositional dependence of the switch of magnetic anisotropy is very appealing for possible applications in the future.

4.2. Transport Properties

Electrical resistivity, Hall, and magnetotransport measurements were performed on single crystals of Mn₂P₂S_6. Figure 10 shows the temperature-dependent resistance R(T) of our Mn₂P₂S₆ single crystal at various pressures up to 59.0 GPa. As shown in Figure 10a, Mn₂P₂S₆ displays an insulating behavior at 18.8 GPa, similar to the early reports from Wang et al. [51]. With increasing pressure up to 22.2 GPa, the resistance in the whole temperature range decreases remarkably, and a metal-insulator transition appears at T_MIT~250 K for 22.2 GPa (see Figure 10b). One can see that T_MIT initially decreases up to 25 GPa and then increases in the pressure range 25.0–30.9 GPa. Notably, the sample exhibits a metallic behavior in the whole temperature range up to room temperature at 34.1 GPa. As shown in Figure 10c, the metallic behavior maintains with increasing pressure up to 59.0 GPa, and no traces of superconductivity are detected down to 1.8 K.

High-pressure Hall resistance and magnetoresistance measurements were further performed to detect the evolution of charge carriers in pressurized Mn₂P₂S₆. Figure 11a shows the selective Hall resistivity curves R_xy(B) measured at 10 K under pressure, where the magnetic field is applied perpendicular to the ab-plane. Above 18.8 GPa, the R_xy(B) curves exhibit a quasi-linear feature with a positive slope, indicating hole-dominated charge carriers. According to the single-band model, the Hall coefficient R_H, extracted from the slope of R_xy(B) around zero fields, decreases initially with increasing pressure and then increases above 30.9 GPa (see Figure 11b). Figure 10c displays the magnetoresistance MR = [(ρ(B) − ρ(0)]/ρ(0) × 100% at 10 K at various pressures. Note that the MR is very small with negative values. Similar to the case of the Hall coefficient, the absolute values of the specific MR at 5 T and 10 K initially decrease and then increase above 30.9 GPa.

5. Conclusions

In summary, we have successfully grown single crystals of (Mn_1−xNi_x)₂P₂S₆ via the chemical vapor transport technique using Iodine as a transport agent. The optimized crystal growth conditions yielded a series of (Mn_1−xNi_x)₂P₂S₆ with (x = 0, 0.3, 0.5, 0.7, 1) single crystals up to 4 mm × 3 mm × 200 μm. As-grown crystals obtained by this method were characterized by means of SEM\EDX for compositional analysis and powder X-ray measurements for the structural analysis. Compositional analysis confirms the chemical homogeneity for both parent compounds as well as for the substituted series. The structural characterization shows that all crystals crystallize in the monoclinic space group C2/m (No. 12). The magnetic measurements as a function of temperature were performed for applied fields parallel and perpendicular to the crystallographic ab plane. All as-grown single crystals show the presence of long-range antiferromagnetic order. The end member Ni₂P₂S₆ has the highest T_N ~155 K and non-monotonously shifts to lower temperature as a function of Mn substitution. The shift in the long-range magnetic order and the gradual change in the anisotropy show some similarity to the sister compounds like (Fe_1−xNi_x)₂P₂S_6, and yet there are some significant differences, such as the lowest T_N and T_max for the (Mn_0.5Ni_0.5)₂P₂S₆ composition. The magnetic anisotropy switches from out-of-plane to in-plane as a function of composition in the (Mn_1−xNi_x)₂P₂S₆ series. Our electrical transport measurements on the end member Mn₂P₂S₆ clearly indicate an insulator-to-metal transition at a critical pressure of ~22 GPa. For pressures higher than 40 GPa, Mn₂P₂S₆ becomes metallic at room temperature. The magnetoresistance measured at 10 K shows a negative magnetoresistance. The successful growth of high-quality single crystals of our series (Mn_1−xNi_x)₂P₂S₆ opens an opportunity for further anisotropic investigations in the future.