Cadmium Phosphorus Trisulfide (CdPS3)

• Cadmium phosphorus trisulfide (CdPS3) is a layered van der Waals semiconductor characterized by a wide indirect bandgap (~3.4 eV) and an unusually high in-plane refractive index (~3.0), enabling advanced UV–vis applications.
• Precise synthesis methods such as melt growth and femtosecond pulsed laser ablation allow control over phase and defect populations, tailoring its electronic and optical properties.
• Its unique crystallography and tunable band structure facilitate applications in waveguiding, metasurfaces, and photocatalytic nanocomposites, offering breakthroughs in integrated photonics.

Cadmium phosphorus trisulfide (CdPS₃) is a layered van der Waals semiconductor of current interest for ultraviolet–visible (UV–vis) nanophotonics and catalytic heterostructures. It is distinguished among the MPX₃ family by an exceptional combination of a wide indirect bandgap ($E_g \sim 3.4$ eV) and a near-UV in-plane refractive index approaching $n = 3$, breaking the prevailing empirical trade-off described by Moss’s law. This article surveys the crystallographic, electronic, and optical properties of CdPS₃, together with its behavior under external stimuli, phase/defect tunability by advanced synthesis, and key implications for device applications.

1. Crystal Structure and Polymorphism

CdPS₃ exhibits a strongly 2D crystal structure, with electronically decoupled layers held together by van der Waals (vdW) interactions (Povolotskiy et al., 18 Nov 2025, Kuzmin, 2020, Ushkov et al., 9 Dec 2025). At room temperature, the stable phase is monoclinic (space group C2/m), while a low-temperature polymorph is trigonal (space group R3). Lattice parameters determined experimentally and via first-principles optimization agree to within 1%. Typical monoclinic values are

$$a = 6.218 \,\text{Å}, \;\; b = 10.763 \,\text{Å}, \;\; c = 6.867 \,\text{Å}, \;\; \beta = 107.58^\circ.$$

The layer structure consists of hexagonally packed CdS₆ octahedra sharing edges, separated and interconnected by P₂S₆⁴⁻ "ethane-like" bipyramidal units. The resulting 2D motif is a honeycomb network (Figure 1a in Povolotskiy et al. (Povolotskiy et al., 18 Nov 2025)), with each repeating layer approximately 6.8 Å thick. Interlayer spacing is measured as ~3.16 Å for C2/m and ~3.12 Å for R3 (Kuzmin, 2020). Each layer is overall charge-neutral as [Cd²⁺(P₂S₆)²⁻], yielding a stoichiometry Cd:P:S = 1:1:3.

Relevant bond lengths at zero pressure are summarized below:

Bond	Length (C2/m) [Å]	Length (R3) [Å]
Cd–S	2.68–2.60	2.69
P–S	2.07–2.10	2.07
P–P	2.24	2.24
Cd–P	3.73	3.75

High-purity single crystals are obtained via melt growth, and atomically thin flakes down to two monolayers can be prepared by mechanical exfoliation for device studies (Povolotskiy et al., 18 Nov 2025).

2. Electronic Band Structure and Density of States

CdPS₃ is an indirect-gap semiconductor in both polymorphs. First-principles calculations using DFT (PBE+$G_0W_0$), as well as hybrid meta-GGA M06 LCAO methods, consistently yield indirect bandgaps $E_g^{(\text{ind})} = 3.4$ eV (C2/m, room temperature) and $E_g^{(\text{ind})} = 3.3$ eV (R3, low temperature) (Kuzmin, 2020, Povolotskiy et al., 18 Nov 2025). The VBM is located near the $\Gamma$ point, while the CBM is slightly displaced along the $\Gamma$–Y direction. The onset of direct transitions is only ~0.1 eV above the indirect gap, which matches optical absorption spectra (Povolotskiy et al., 18 Nov 2025).

Valence band states are principally S 3p character with minor admixtures of Cd 4d and P 3p, while conduction band manifold is derived from hybridized P 3s/3p, S 3p, and Cd 5s/5p orbitals (Kuzmin, 2020). Application of hydrostatic pressure modulates the gap in a non-monotonic fashion, with $E_g$ peaking at 3.6 eV for C2/m at 8 GPa, and at 4.0 eV for R3 at 30 GPa, indicating a strong structure–property–pressure correlation.

3. Linear Optical Properties and Anomalous Index–Gap Relation

The most distinctive optical property of CdPS₃ is its unusually high in-plane refractive index, which approaches $n_{ab} \approx 3.0$ at $\lambda = 350$ nm and maintains $n_{ab} \approx 2.7$ at 450 nm, with low absorption (extinction $k_{ab} < 10^{-2}$ for $\lambda > 360$ nm) (Povolotskiy et al., 18 Nov 2025). The Sellmeier dispersion in the transparent regime follows

$$n^2(\lambda) = 1 + \sum_{i=1}^2 \frac{B_i \lambda^2}{\lambda^2 - \lambda_i^2}$$

with $B_1 = 1.12$, $\lambda_1 = 150$ nm; $B_2 = 0.85$, $\lambda_2 = 275$ nm.

The out-of-plane index $n_c$ is lower (e.g., $n_c \approx 2.1$ at 500 nm), with modester anisotropy $\Delta n \approx 0.5$. This refractive index–bandgap combination violates the empirical Moss’s law ($E_g n^4 \approx 95$ eV), with $E_g n^4 \approx 275$ eV for CdPS₃, nearly three times higher than conventional semiconductors. This indicates a capacity for simultaneously high transparency and strong optical confinement in the UV–vis, unattainable for other layered dielectrics.

4. Waveguiding, Confinement, and Near-field Optical Characterization

Scattering-type scanning near-field optical microscopy (s-SNOM) has been used to directly visualize tightly confined waveguide modes in mechanically exfoliated CdPS₃ flakes (Povolotskiy et al., 18 Nov 2025). For a 254 nm-thick flake at 700 nm, effective indices are measured as $n_{\text{eff}}^{(\text{TE}_0)} \approx 2.75$ and $n_{\text{eff}}^{(\text{TM}_0)} \approx 2.15$, reducing to 1.85 and 1.60 at 1600 nm, respectively. These results agree with transfer-matrix modeling based on experimentally retrieved $(n, k)$ values, confirming that CdPS₃ guides extreme-subwavelength optical modes.

The high $n$ enables channel waveguides and metasurfaces with dimensions (width $w \approx 94$ nm at 375 nm wavelength for a decay constant $\alpha = 2.49$ µm⁻¹ in SiO₂) that are significantly reduced relative to TiO₂, Si, or MoS₂. Simulations show two 94 nm-wide CdPS₃ guides at 50 nm separation maintain crosstalk lengths on the order of 10 µm, supporting ultra-dense routing architectures. Propagation lengths in the blue/green exceed 100 µm, with mode areas as small as $A_{\text{mode}} \sim (0.1\lambda)^2$.

5. Phase Engineering and Photocatalytic Nanocomposites

CdPS₃ can be processed into nanostructures by femtosecond pulsed laser ablation in liquid (fs-PLAL) (Ushkov et al., 9 Dec 2025). The choice of solvent establishes a route to control phase and defect population. In deionized water, the laser-induced plasma condenses into stoichiometric CdPS₃ nanocrystals preserving the C2/m lattice. In contrast, ablation in isopropanol or acetonitrile induces reductive chemistry, removing P atoms to yield a mixture of CdS quantum dots (QDs) and metallic Cd clusters, registered as mixed-phase or binary/ternary nanocomposites according to Table 1 in (Ushkov et al., 9 Dec 2025):

Solvent	CdPS₃ Content (mol %)	CdS Content (mol %)
DI H₂O	88	12
Acetonitrile	52.6	47.4
IPA	11.3	88.7

Raman spectroscopy, TEM/SAED, and EDX confirm the phase assignments. Band alignment measurements show type-II heterojunctions between CdPS₃ and CdS, as well as Schottky barriers to metallic Cd ($\Phi_B \sim 0.1$ eV). These facilitate spatial charge separation and enhanced photocatalytic activity.

6. Optoelectronic Functionality and Photocatalytic Performance

Optical band positions with respect to the normal hydrogen electrode are $E_{\text{CBM}} \approx -0.8$ V, $E_{\text{VBM}} \approx +2.2$ V for CdPS₃, and $E_{\text{CBM}} \approx -1.0$ V, $E_{\text{VBM}} \approx +1.56$ V for CdS QDs. The hybrid systems, especially CdPS₃/CdS nanocolloids produced via fs-PLAL in isopropanol, exhibit superior charge separation due to the engineered junctions: electrons flow from CdS to metallic Cd, while holes remain in the CdPS₃ valence band.

Photocatalytic degradation of Methylene Blue under 532 nm irradiation demonstrates a kinetic constant $k_{\text{IPA}} \approx 0.077$ min⁻¹, yielding 90% degradation in 30 minutes for CdPS₃/CdS hybrids in IPA, compared to negligible activity in pure water-prepared CdPS₃ (Ushkov et al., 9 Dec 2025). A plausible implication is the critical dependence of visible-light photocatalytic activity on engineered phase composition and defect chemistry, which extends the functionality of wide-bandgap vdW crystals beyond their native UV absorption.

7. Prospects for Ultraviolet–Visible Nanophotonics and Beyond

Owing to its record-high in-plane refractive index and wide gap, CdPS₃ is identified as a benchmark vdW dielectric for deep-UV and visible nanophotonics, metasurfaces, and integrated photonic circuits (Povolotskiy et al., 18 Nov 2025). Material dispersion, transparency, and experimentally validated wave confinement far surpass the limitations imposed by conventional index–gap trade-offs. Sub-100 nm waveguides, deep-UV metasurfaces, and dense on-chip routing are accessible using this material platform.

Solvent-assisted fs-PLAL further offers tunable phase and defect engineering for application to hybrid photochemistry, where precise modulation of optoelectronic structure is enabled by rational solvent choice and pulse protocol (Ushkov et al., 9 Dec 2025). The approach is generalizable to other MPX₃- and ternary-layered systems, but open questions persist regarding quantitative control of defect densities, in situ monitoring of plasma dynamics, long-term operational stability, and integration into functional UV/vis photonic and catalytic devices.

In summary, CdPS₃ is emerging as a multifaceted semiconductor, combining foundational advances in photonic materials science with avenues for tunable optoelectronics and catalysis (Povolotskiy et al., 18 Nov 2025, Kuzmin, 2020, Ushkov et al., 9 Dec 2025).