Introduction

As isoelectronic analogs of III–V zincblende materials, zinc-based chalcopyrites ZnXPn2 (X = Si, Ge, and Sn; Pn = P, As, and Sb) have recently attracted great attention because of their potential technological applications in nonlinear optics, optoelectronics, and photovoltaics1,2,3,4,5,6,7,8,9,10. For instance, ZnGeP2 can be applied to nonlinear optical devices based on its large nonlinear coefficient, birefringence, and large-area growth availability3. ZnSnP2 is an absorber material for solar cells, whose bandgap can be effectively engineered by tuning the cation disorder6. Furthermore, a recent first-principles calculation predicted that ZnSnP2 displays large shift-current conductivity, a bulk photovoltaic phenomenon correlated with the Berry connection between the valence and conduction bands10. As for ZnSiP2, it possesses a direct bandgap of Eg ~2.01 eV9 and a typically cation-ordered tetragonal structure with intrinsic stability and defect tolerance2. Apart from its inexpensive, earth-abundant, and nontoxic elemental constituents, ZnSiP2 has become a promising candidate for transitional tandem solar cells owing to its small lattice mismatch and good refraction index matching with Si, little parasitic below-bandgap absorption, excellent photoresponse, and high open-circuit voltage2,11,12,13. After disorder is introduced into the cation sublattice, ZnSiP2 can be further used as a high-performance anode material for next-generation Li-ion batteries1.

As one of the fundamental state parameters, pressure is an effective and clean way to tune the lattice constant, crystal structure, and electronic state, thus varying the fundamental physical properties of materials. Regarding the pressure engineering of photovoltaic and optoelectronic materials, various exotic phenomena have been revealed recently, including photoluminescence (PL) emission enhancement, prolonged carrier lifetime, bandgap optimization, and superconductivity14,15,16,17,18,19,20,21. For the photovoltaic material ZnSiP2, Bhadram et al.22 reported that it undergoes a phase transition from tetragonal to cubic rock-salt structure between 27 and 30 GPa, in agreement with the ab initio investigations23. However, a systematic investigation of the electronic and optical properties of ZnSiP2 under pressure is still lacking to date.

Here, we systematically investigate the pressure effect on the structural, optical, and electronic properties of the chalcopyrite semiconductor ZnSiP2 through various experimental measurements, including synchrotron X-ray diffraction (XRD), Raman spectroscopy, PL spectroscopy, optical microscopy, and electrical transport measurements. We show that pressure-induced modulations in optical and electronic properties correlate well with the structural evolution. For the optical properties, the peak energy of the PL spectrum displays a plateau between 1.4 and 8.7 GPa due to the presence of disorder in the cation sublattice. Moreover, along with the structural phase transition from tetragonal to cubic phase, a V-shaped superconducting behavior is observed.

Materials and methods

Sample synthesis and characterization at ambient pressure

Single crystals of ZnSiP2 were grown via a flux method24. Room-temperature X-ray diffraction (XRD) patterns of single crystals were obtained by using a Rigaku X-ray diffractometer with Cu Kα radiation (λ = 1.5406 Å). The atomic proportions of the crystals were characterized by energy dispersive X-ray spectroscopy (EDXS). Absorption spectra were collected by using a UV/Vis/NIR spectrometer (CRAIC 20/30PV).

High-pressure PL spectra, Raman spectra, and X-ray diffraction measurements

High-pressure PL emission, Raman scattering, and angle-dispersive synchrotron XRD experiments were conducted in symmetric diamond anvil cells (DACs) with rhenium (Re) as the gasket and silicone oil as the pressure-transmitting medium (PTM). The culet of the diamond was 300 μm. The PL and Raman spectra were recorded in a Renishaw spectrometer (λ = 532 nm) at powers of 0.25 mW and 2.5 mW, respectively. The dimensions of cleaved single-crystal flakes were ~50 × 50 × 10 µm3. Powder angle-dispersive XRD experiments were carried out at beamline BL15U1 of Shanghai Synchrotron Radiation Facility (SSRF). The wavelength of the monochromatic X-ray beam was 0.6199 Å. The Dioptas25 and Rietica26 programs were employed for image integrations and Le Bail refinements, respectively.

High-pressure electrical transport measurements

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 × 40 × 10 μm3 was loaded together with NaCl fine powder and ruby powder. A four-probe configuration was utilized to measure the 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 experiments27.

Results and discussion

The synthesized ZnSiP2 single crystals were characterized via various experimental techniques under ambient conditions. Fig. S1 displays a single-crystal XRD pattern collected from a flake freshly cleaved from a bulk crystal, which shows a (101) orientation of the cleavage plane. The EDXS measurement reveals that the cleaved flake is off-stoichiometric with a real composition of Zn0.92±0.02SiP2.07±0.03. The absorption spectrum shown in Fig. S2a yields a bandgap of ~2.06 eV, consistent with that reported in recent literature2. The PL spectrum at room-temperature (see Fig. S2b) features a broad emission band peaking at ~919 nm (equivalent to 1.36 eV), which can be ascribed to the donor–acceptor pair transitions via defects (vacancies and antisites)2,6. In ref. 2 Martinez et al. showed that the PL peak energy of ZnSiP2 varies from 1.8 eV to 1.67 eV as the temperature is increased from 5 K to 100 K, illustrating a temperature effect on the PL peak position. Based on the data of ref. 2 a value of ~1.4 eV at 300 K is obtained by extrapolation according to the approximate model for temperature-dependent PL spectra28,29. These results consistently confirm the high quality of our samples.

High-pressure synchrotron XRD experiments were conducted on powdered single-crystal ZnSiP2 to explore the structural stability at high pressures. Fig. 1a shows the representative XRD patterns at room-temperature. We note that the peaks at ~15° and ~17° arise from the gasket (Re) (see Fig. S3), which persist in the entire pressure region. In addition, all the other peaks progressively shift to larger angles without the appearance of new peaks under compression up to 23.3 GPa, indicating the stability of the pristine tetragonal structure. A structural phase transition to the cubic rock-salt type (Fm-3m, No. 225) is detected upon further compression, in agreement with the results of Bhadram et al.22 Above 23.3 GPa, one can see that the intensity of the peak at ~15° shows a subtle enhancement, and a broad peak develops at ~22° in the meantime. The structural transition is complete ca. 36.7 GPa, where the high-pressure cubic phase possesses cation disorder, as evidenced by the broad features in the XRD patterns. Although peak broadening in the XRD pattern can be caused by nonhydrostatic compression associated with the PTM, similar broadening behavior was also observed by Bhadram et al.22 They used Ar and He as the PTMs, which provide better hydrostatic conditions than the silicone oil used in our case22. The extracted lattice parameters a and c are displayed in the upper part of Fig. 1c. The unit-cell volume as a function of pressure can be fitted by the third-order Birch-Murnaghan equation of state30; see the solid lines in the lower panel of Fig. 1c. The fittings yield ambient pressure volume V0 = 302.0(4) Å3, bulk modulus B0 = 97.5(6) GPa, and first-order derivative of the bulk modulus at zero pressure B0’ = 6.3(0) for the pristine tetragonal phase and 124.9(8) Å3, 110.0(1) GPa, and 3.4(5) for the high-pressure cubic phase. Note that the errors caused by Le Bail refinements and/or nonhydrostatic conditions are not included for the equation of state fitting. The structural phase transition yields a unit-cell volume contraction ΔV/V ~ 19.1% at 27.6 GPa, similar to the case of a previous report22.

Fig. 1: Structural phase transition of ZnSiP2 under pressure.
figure 1

a Typical powder synchrotron X-ray diffraction patterns at room-temperature (λ = 0.6199 Å). The asterisks (*) denote the peaks of the rhenium gasket. b Raman spectra of ZnSiP2 single crystals at room-temperature (λ = 532 nm) with a semilogarithmic scale. c Upper: Pressure-dependent lattice parameters a and c for the pristine tetragonal (I-42d, Z = 4) and high-pressure cubic (Fm-3m, Z = 2) phases. Lower: Unit-cell volume as a function of pressure. The solid lines are the fittings according to the third-order Birch-Murnaghan equation of state. d Pressure-dependent Raman modes of ZnSiP2.

Full size image

Figure 1b depicts selective room-temperature Raman spectra of ZnSiP2 at various pressures. At 0.2 GPa, the Raman vibrational modes centered at 102.0, 129.9, 185.3, 264.8, 334.4, 338.2, 344.6, 464.8, 494.8, and 519.7 cm−1 can be assigned to E, B1, E, ET, A1, B1, B2T, EL, B2T-ET, and B2L-EL, respectively31. The evolutions of these modes under pressure are qualitatively consistent with those reported in ref. 22, which were believed to accord with a scenario of a two-stage transition. Based on the model proposed by Bernard and Zunger32, it was suggested that Zn and Si cations substitute each other in the first stage, leading to a partially cation-disordered sublattice in the low-pressure tetragonal phase. The second stage involves the structural transition from the tetragonal phase to the high-pressure cubic phase22. The strain energy, set up by the atomic size mismatch between the Zn-P and Si-P bond lengths, could be the reason to control the nature of the state of order in chalcopyrite ZnSiP222,33. Owing to the lattice instability caused by progressive development of cation disorder, the modes below 200 cm−1 that correspond to acoustic zone-center phonon modes display abnormal evolutions in both frequency and full width at half maximum (FWHM). As shown in Fig. S4, in contrast to the continuous blueshift of optical phonon modes, B1 and E demonstrate a crossover from blueshift to redshift ca. 8 and 15 GPa, respectively. Meanwhile, the FWHM of each mode shows corresponding changes. Along with the structural transition to the cubic phase upon further compression, these modes show an abrupt reduction in intensity above 27.8 GPa and completely disappear when the structural transition completes at 36.7 GPa. A similar pressure-induced two-stage order-disorder transition was also claimed in defect chalcopyrite CdAl2S434, nevertheless, direct experimental evidence of the disorder in ZnSiP2 is lacking.

We further investigated the pressure effect on the optical properties of ZnSiP2. Figure 2a displays the PL spectra of ZnSiP2 at different pressures. At 0.7 GPa, both the profile and PL peak position are analogous to those at ambient pressure (see Fig. S2b). As the pressure increases up to 8.7 GPa, the relative intensity of the PL peak gradually decreases, but the profile remains nearly unchanged. The PL intensity undergoes a strong suppression at 11.7 GPa and becomes almost undetectable at 16.9 GPa, as shown in Fig. 2a. In contrast to the common expectation that the peak energy should decrease monotonically with increasing pressure due to the enhancement of orbital overlapping caused by lattice shrinkage, the PL peak energy of ZnSiP2 displays a plateau between 1.4 and 8.7 GPa followed by an abrupt decrease at higher pressures (see Fig. 2b). Note that anomalous evolution of the PL energy under pressure has been observed in organolead perovskites17,19. For example, the blueshift of the PL energy in perovskite (MA)PbBr3 was attributed to pressure-induced amorphization breaking certain bonds17. In (MA)PbI3, the abrupt blueshift of the PL energy was ascribed to octahedral tilting, which decreases orbital overlapping19. For chalcopyrite ZnSiP2, Martinez et al.2 showed that SiZn2+ and/or PSi1+ antisite defects contribute to the donor level, while ZnSi2- and/or SiP1- antisite defects form the acceptor level. The abnormal plateau could be attributed to the progressive development of cation disorder because donor–acceptor pair transitions are the primary PL mechanism in ZnSiP2. The optical micrographs in Fig. 2c demonstrate a piezochromic transition in compressed ZnSiP2. During the compression process, ZnSiP2 changes from its original transparent red to translucent dark red at 7.5 GPa and eventually turns to opaque black at 16.7 GPa.

Fig. 2: Pressure evolution of the luminescent properties of ZnSiP2 single crystals.
figure 2

a Room-temperature PL spectra under compression. b Pressure dependence of the PL peak. The red star denotes the result at ambient pressure. The dashed line is a guide to the eyes. c Optical micrographs of ZnSiP2 in a DAC upon compression, which demonstrates a piezochromic phenomenon.

Full size image

Because of the insulating nature of ZnSiP2, the resistance at 300 K is beyond our instrumental limit (107 Ω) and could not be detected below 14.0 GPa. Figure 3a shows the temperature dependence of the resistance R(T) of ZnSiP2 at various pressures up to 55.5 GPa. Starting from 14.0 to 21.5 GPa, ZnSiP2 displays semiconducting behavior, as evidenced by the increase in resistance upon cooling. A further increase in pressure leads to a semiconducting-metallic transition at 24.6 GPa. Strikingly, the occurrence of metallization is accompanied by a resistance drop below ∼8.2 K (see Fig. 3b). The drop in the R(T) curve becomes increasingly sharper, and zero resistance is finally observed at 37.1 GPa, signaling pressure-induced superconductivity in ZnSiP2. Meanwhile, the superconducting critical temperature Tc decreases with increasing applied pressure, reaching a minimum at 37.1 GPa, followed by a continuous increase up to 55.5 GPa, the highest pressure applied in this study. We further measured the R(T) curves under various magnetic fields to determine the upper critical field at 44.4 GPa, as shown in Fig. 3c. By defining Tc as the onset temperature of the superconducting transition, we constructed the temperature-magnetic field phase diagram in the inset of Fig. 3c. According to the Werthamer–Helfand–Hohenberg (WHH) model35, the yielded upper critical field μ0Hc(0) is ~3.0 T. Note that the upper critical field is much lower than the resultant Pauli limiting field of μ0HP(0) = 1.84Tc, which suggests the absence of Pauli pair breaking.

Fig. 3: Experimental evidence of pressure-induced superconductivity in ZnSiP2.
figure 3

a Selective temperature-resistance curves R(T) of ZnSiP2 single crystals under compression. Inset: R(T) curves in the range of 10–100 K. b Low-temperature R(T) curves around the superconducting transition. The curve at 24.6 GPa is multiplied by 0.5 for comparison. c R(T) curves under various magnetic fields at 44.4 GPa. Inset: Temperature-dependent upper critical field μ0Hc at 44.4 GPa. The solid line represents the fitting based on the WHH model. Here, the Tc value was defined as the onset temperature of the superconducting transition.

Full size image

To obtain a comprehensive understanding of the evolution of the PL, electrical conductivity, and crystal structure in pressurized ZnSiP2, we constructed a phase diagram, as shown in Fig. 4. It is clear that the appearance of superconductivity can be directly attributed to the structural transition from the tetragonal to cubic phase ca. 23 GPa. In the tetragonal phase, ZnSiP2 is situated in the semiconducting regime but possesses the characteristic of cation disorder. As a prelude of the structural transition, the disorder in the cation sublattice gradually develops during compression, leading to the plateau present from 1.4 to 8.7 GPa in the pressure evolution of the PL peak energy. Along with the structural transition above ca. 23 GPa, ZnSiP2 evolves into the superconducting regime. In agreement with the coexistence of tetragonal and cubic phases between 23 and 37 GPa, a measurable resistance is still observed at temperatures below the sharp resistance drop. Compared with the XRD results of Bhadram et al., who used Ar and He as PTMs, the coexisting pressure range is much larger in our case, which indicates that the coexistence of the two structures can be attributed not only to the incompleteness of the phase transition itself but also to the nonhydrostatic conditions. In the cubic phase, the superconducting phase diagram features an abnormal V-shaped evolution of Tc. We note that a similar V-shaped Tc(P) behavior was previously reported in some superconducting compounds, such as AFe2As2 (A = K, Rb, Cs)36,37, PbTaSe238, and TlInTe239, while the underlying mechanism is still under debate. On the one hand, the origin of V-shaped Tc behavior has been discussed in terms of a change in the superconducting pairing symmetry or a Lifshitz transition across the critical pressure36,38. On the other hand, Yesudhas et al.39 observed unusual giant phonon softening (Ag mode) concomitant with the V-shaped Tc(P) behavior in TlInTe2. In our case, one can see that the structural phase transition from the tetragonal to cubic phase is almost complete around the valley pressure of 36.7 GPa, which implies that the V-shaped behavior might be associated with abnormal evolution of the electron-phonon coupling due to the incompleteness of the phase transition in that pressure regime and/or to the nonhydrostatic conditions.

Fig. 4: Pressure evolution of the structural, optical, and electronic properties of ZnSiP2.
figure 4

Pressure evolution of the superconducting transition onset temperature Tc and PL peak energy from PL analyses. The colored areas are guides to the eyes, indicating the distinct conducting states, i.e., semiconductor, metal, and superconductor. The black and red horizontal arrows indicate the low-pressure tetragonal phase and high-pressure cubic phase, respectively.

Full size image

Conclusions

In summary, by systematically investigating the pressure effect on the optical and electronic properties, we demonstrate the pressure-induced abnormal evolution of the PL spectrum as well as superconductivity in the chalcopyrite semiconductor ZnSiP2. The abnormal evolution of the PL peak energy is accompanied by a piezochromic transition and is attributed to the gradual development of disorder in the cation sublattice. The superconductivity that shows a V-shaped Tc(P) phase diagram can be directly correlated with the structural phase transition from the tetragonal to cubic phase ca. 23 GPa. Based on the fact that a material with optoelectronic and photovoltaic applications is transformed into a superconductor, these findings provide crucial insight into the structure–property relationships in chalcopyrite semiconductors.