Significant Efficiency Boost in CdS/CdSe Quantum Dot-Sensitized Solar Cells Using (001)-Oriented TiO2 Nanosheet Photoanodes

Background

Quantum dot‑sensitized solar cells (QDSSCs) are emerging as a promising alternative to conventional dye‑sensitized solar cells (DSSCs). Quantum dots (QDs) offer a larger extinction coefficient, tunable band gaps through size and composition control, superior photostability, and the potential for multiple‑exciton generation or hot‑carrier extraction—features that can push the theoretical efficiency beyond the Shockley‑Queisser limit of 32 %.

QDSSCs operate on the same principle as DSSCs but use inorganic nanocrystals in place of organic dyes. The typical configuration consists of a QD‑coated metal‑oxide photoanode, a polysulfide electrolyte, and a platinum counter electrode. Narrow‑bandgap semiconductors such as CdS, CdSe, CdTe, and PbS have been employed as light absorbers, while core/shell combinations (e.g., CdS/CdSe) extend the absorption range and improve charge injection dynamics.

Despite advances, the best‑performing QDSSCs still reach only 6–8 % PCE, largely due to charge recombination and limited QD loading on the photoanode. Mesoporous TiO₂ remains the material of choice for the photoanode because of its low cost, chemical stability, and excellent electron transport. However, TiO₂ performance is highly sensitive to morphology and crystal facet exposure. The thermodynamically stable (101) facet dominates commercial nanoparticles, whereas the high‑energy (001) facet is more reactive, lowers the band‑edge, and suppresses recombination.

Nanostructures exposing the (001) facet—such as nanosheets, hollow spheres, and nanotubes—have shown enhanced performance in photocatalysis, water splitting, and lithium‑ion batteries. Yet, few studies have explored (001)-oriented TiO₂ nanosheets in QDSSC devices. Here, we compare (001)-rich nanosheets with P‑25 nanoparticles for CdS/CdSe QD‑sensitized cells, revealing a substantial PCE improvement.

Methods

Preparation of TiO₂ Photoanodes

TiO₂ nanosheets were synthesized hydrothermally: 2.4 mL HF (48 % wt) was added dropwise to 30 mL titanium butoxide, sealed in a Teflon‑lined autoclave, and heated at 180 °C for 16 h. The resulting precipitate was collected, washed, and dispersed with terpineol and 10 % ethyl cellulose to form a 13 wt % paste. A similar paste was prepared with commercial P‑25. The pastes were screen‑printed onto FTO glass (10 Ω cm⁻²) and annealed at 500 °C for 1 h to achieve conductive films.

QD Deposition

CdS QDs were grown in situ via successive ion‑layer absorption and reaction (SILAR) using 20 mM CdCl₂ and 20 mM Na₂S solutions in a 1:1 methanol/water mixture. The photoanodes were alternately dipped for 1 min in each precursor, repeated over 8 cycles to achieve optimal CdS coverage. CdSe QDs were then deposited by chemical bath deposition: a 70 °C solution containing 2.5 mM Cd(CH₃COO)₂, 2.5 mM Na₂SeSO₃, and 75 mM NH₄OH was applied for 1 h. Two cycles of CdSe deposition provided the best performance.

Device Assembly & Characterization

QDSSCs were assembled in a sandwich configuration: Pt‑coated FTO counter electrode, a 25 µm Surlyn spacer, and the QD‑sensitized TiO₂ photoanode. A polysulfide electrolyte (0.2 M Na₂S/0.2 M S/0.02 M KCl in water) filled the cell. The active area was ~ 0.16 cm². Devices were characterized by FE‑SEM, TEM, GIXRD, ICP‑MS, J‑V curves under AM 1.5 G illumination, IPCE measurements, UV‑VIS absorption, and electrochemical impedance spectroscopy (EIS).

Results and Discussion

Structural analysis confirmed the nanosheets were pure anatase TiO₂ with a dominant (001) orientation, as evidenced by the strong (200) XRD peak and 0.235‑nm lattice spacing in HRTEM. The nanosheets displayed an average side length of 50 nm and a thickness of 5 nm, providing ~ 70 % (001) facet exposure, compared to < 10 % for P‑25.

Nitrogen adsorption–desorption isotherms revealed the nanosheet films are type IV with H3 hysteresis, indicating slit‑like meso‑ and macropores. The BET surface area reached 52.8 cm² g⁻¹, larger than that of P‑25, facilitating greater QD loading.

CdS/CdSe QDs grown by SILAR/CBD showed excitonic absorption peaks at 2.67 eV (CdS) and 1.78 eV (CdSe), confirming quantum confinement. UV‑VIS spectra indicated higher light absorption for the nanosheet electrode, correlating with ICP‑MS results: CdS loading of 5.44 × 10⁻⁹ mol cm⁻² on nanosheets versus 4.59 × 10⁻⁹ mol cm⁻² on P‑25; CdSe loading of 4.57 × 10⁻⁹ mol cm⁻² versus 3.77 × 10⁻⁹ mol cm⁻².

Photovoltaic performance under one‑sun illumination showed the nanosheet device delivering a V_oc of 0.58 V, J_sc of 15.07 mA cm⁻², and η of 4.42 %, compared with 0.52 V, 11.75 mA cm⁻², and 2.86 % for the P‑25 cell—a 54 % efficiency gain. The higher V_oc arises from the lower flat‑band potential of the (001) facet, while the increased J_sc reflects superior QD loading and light harvesting. IPCE spectra showed the nanosheet device achieving ~ 75 % at 600 nm and a 675‑nm cut‑off, confirming enhanced charge collection.

EIS analysis under open‑circuit bias revealed a lower charge‑transfer resistance (R_k) of 28.26 Ω and a shorter electron‑transport resistance (R_w) for the nanosheet device, indicating faster electron diffusion and reduced recombination. The calculated electron diffusion length (L_n) was ~ 21 µm—double that of the P‑25 device—ensuring most photogenerated electrons reach the electrode before recombination.

Conclusions

We have fabricated 2D anatase TiO₂ nanosheets with > 70 % (001) facet exposure via a straightforward hydrothermal method and used them as photoanodes for CdS/CdSe core/shell QDSSCs. The nanosheet‑based cells achieved a PCE of 4.42 %, a 54 % improvement over conventional P‑25 cells, and an IPCE exceeding 70 % between 450–600 nm. The high‑energy (001) surface not only provides more active sites for QD attachment but also reduces surface traps, enhancing electron transport. These findings underscore the potential of (001)-oriented TiO₂ nanosheets as a low‑cost, scalable platform for high‑performance QDSSCs, eliminating the need for additional electron‑transfer linkers.