Transition metal chalcogen phosphates belong to a family of two-dimensional magnetic materials with general formula MPX₃, where M is a transition metal ion and X = {S, Se, or Te}. They exhibit a monoclinic symmetry with transition metal ions arranged on a honeycomb lattice in the ab-plane. While all the members of this family order antiferromagnetically at low temperatures, the nature of magnetic ordering and electronic properties are easily tunable depending on the choice of elements M and X. For example, while FePS₃ is an Ising-type antiferromagnet with T_N = 118 K[1, 2], MnPS₃ is an Heisenberg type antiferromagnet with T_N = 78 K[3, 4], and NiPS₃ is an XY or XXZ type antiferromagnet with T_N = 155 K[5]. The electronic band gap of these materials varies from 1.5 eV (FePS₃) to 3.5 eV (ZnPS₃)[6]. Magnetism in atomically thin limit has been reported for FePS₃[7], Cr₂Ge₂Te₆[8] and CrI₃[9], etc. to name a few.

In NiPS₃, each Ni ion in the honeycomb layer is octahedrally coordinated by six S ions. The P and S ions together form a bipyramidal [P₂S₆]^4- unit. This unit is located at the center of the hexagon formed by the Ni ions. The crystal structure is shown in Fig. 1(a). Below 155 K, NiPS₃ shows a zigzag type antiferromagnetic ordering where the Ni²⁺ spins (S = 1) arrange ferromagnetically along the zigzag chains running parallel to the a-axis and antiferromagnetically along the b-axis, as shown in Fig. 1(a) (Top). In this study, we report single crystal growth of the site-diluted series Ni_1-xZn_xPS₃ using the physical vapor transport technique. The single crystals are rigorously characterized using various experimental probes to confirm their structure and chemical homogeneity.

Figure 1. (Color online) (a) Schematic of crystal structure (b) FESEM images of the single crystals. The inset shows the representatives images of the single crystals.

The high-quality single crystals are investigated using Raman spectroscopy to explore the effect of site-dilution in the honeycomb lattice of Ni ions. We show that a dominant phonon mode near 176 cm^-1, associated with Ni vibrations, shows a Fano-like asymmetry, which is interpreted as arising due to the presence of charge-phonon coupling in NiPS₃.

The morphology of the single crystals was verified using Field Emission Scanning Electron Microscope (FESEM) (Zeiss Ultra Plus) and the chemical homogeneity was confirmed using Energy Dispersive X-ray (EDX) analysis from Oxford instruments. To investigate the phase purity of the single crystals, the powder X-ray diffraction experiment was carried out using Bruker D8 Advance diffractometer by grounding small crystal pieces for each composition. High-purity Si powder was mixed with the sample powder for use as an internal standard. The room temperature Raman measurements were done in a back-scattering geometry using Horiba Jobin-Yvon LabRAM HR spectrometer by employing 633 nm Laser. The incident light on the sample was unpolarized, travelling perpendicular to the ab-plane.

High-quality single crystals of Ni_1-xZn_xPS₃ (x = 0, 0.05, 0.08, 0.13, 0.14) were grown using the physical vapor transport technique without any external transporting agent. The high-purity elemental micro-powders of nickel (Sigma Aldrich, 99.99% trace metal basis), black phosphorous (Sigma Aldrich, 99.99%), sulfur (Sigma Aldrich, 99.98% trace metal basis), and zinc (Sigma Aldrich, 99%) were weighed using a high-precision microbalance in the stoichiometric quantities. The weighed precursors were then transferred into a quartz ampoule. The weighing, mixing, and filling in the quartz tube were done in an argon filled glove box whose O₂ and moisture levels are always maintained at less than 0.1 ppm. The ampoule was flame-sealed under vacuum ( 10^-5 Torr) and placed in a two-zone furnace. For a detailed working of the physical vapor transport method, one can refer to Ref. 10. For the growth of NiPS₃, the source temperature was set at 750 °C and the sink temperature at 730 °C. For the Zn-doped samples, the source temperature, sink temperature, and reaction times are slightly tuned to obtain homogenous single crystals of Ni_1-xZn_xPS₃. Shiny black crystals of dimension 1 cm by 1 cm laterally, with a thickness in some cases reaching up to 1 mm, were obtained at the colder end of the ampoule. A few representative images of the obtained crystals are shown in the insets of Fig. 1(b). The elemental maps for a representative case are shown in the Appendix (Fig. A1). The elemental composition was assessed by collecting 15–20 EDX data sets at various locations of the crystal specimen as shown in Table A1. The average and standard deviations were then computed and used as markers of the crystal quality. The standard deviation for our crystals turned out to be 0.5 to 1 atomic %, which is smaller than the size of the error bar (2–3 atomic %) associated with this technique.

The layered morphology of the single crystals is confirmed from the FESEM images shown in Fig. 1(b), where the EDX composition x of the amount of Zn doping in the crystal is shown. The powder X-ray diffraction pattern, shown in Fig. 2(a), is well-indexed using the monoclinic symmetry (space group C2/m) previously reported. Figure 2(b) shows the variation of lattice parameter with x. The lattice parameters monotonically increase with increasing Zinc doping since Zn²⁺ (0.74 Å) has larger ionic radius compared to Ni²⁺(0.69 Å).

Figure 2. (Color online) (a) Powder X-Ray diffraction on Ni_1-xZn_xPS₃. (b) Trend in the lattice parameter showing a monotonic increase with the increasing Zn concentration.

Raman spectra for NiPS₃ have been recently investigated by Cheng-Tai Kuo et al.[11]. Altogether, eight Raman active phonon modes, three of A1g symmetry and five of E_g symmetry, are predicted at the Brillouin zone center using factor group analysis. The Raman spectrum of our undoped NiPS₃ sample nicely matches with that reported in the literature. The peaks observed in the Raman spectra for Ni_1-xZn_xPS₃ are labeled according to Ref. 11, as shown in Fig. 3(a). The observed red shift in the phonon modes upon Zn-doping is attributed to unit cell expansion upon doping Zn, as evident from the XRD data; also, the Zn²⁺ ion is heavier than Ni²⁺, which may further lead to the shifting of phonon modes towards lower energy.

Figure 3. (Color online) (a) Raman spectra collected at room temperature. (b) Phonon mode Eg(2) fitted using different line shapes, Fano giving the best fit.

Interestingly, the Eg(2) phonon mode at 176 cm^-1, which is attributed to the vibration of Ni ions, cannot be fitted satisfactorily using a Lorentzian line shape. Attempts to fit this phonon using several other line shapes or a combination of two Lorentzians did not yield any satisfactory result, as shown in Fig. 3(b). The line shape that best describes this mode is the Breit-Weigner-Fano (BWF) line shape[12]. The BWF line shape is given by:

The asymmetry in the Fano line shape is quantified using a dimensionless asymmetry parameter q where |1/q| indicates the strength of coupling of the discrete phonon mode with a continuum excitation of likely electronic origin. Here, Γ is the peak width; the peak is centered at ω0, and I₀ is the peak intensity. For NiPS₃, 1/q is -0.13, the negative value indicates antiresonance with the continuum. With Zn doping, the asymmetry increases and the parameter |1/q| increases to values as large as 0.20 for the highest doped sample.

To summarize, we have grown high-quality single crystals of Ni_1-xZn_xPS₃ and studied their morphology, chemical composition, and structural properties using FESEM, EDX, and powder X-ray diffraction. The crystals are further investigated using room temperature Raman spectroscopy. The entire Raman Spectrum exhibits a red shift upon Zn-doping due to lattice expansion, and heavier mass of Zn compared to Ni. One of phonon mode Eg(2) at 176 cm^-1, which originates from the vibration of the Ni atoms, shows a distinct Fano-like asymmetry, indicating the possibility of charge-phonon coupling. The coupling strength |1/q| increases from 0.13 in NiPS₃ to 0.20 in Ni_0.86Zn_0.14PS₃. The charge continuum that we alluded to possibly corresponds to the electronic state ψ=αd8+βd9L_+γd10 L _ 2 (where L_ is the ligand hole) of Ni, where the weight of d9L_ and d10L_2 components increases with initial Zn-doping as revealed by x-ray photoemission spectra to be published elsewhere. Low-temperature Raman experiments will be useful for further understanding.

NP would like to thank CSIR for the fellowship. SS and LH thank QTF, I-HUB at IISER Pune for supporting research on quantum materials. AK acknowledges Prime Minister Research Fellowship by DST, India.