AgInS₂/CdSe type-II core/shell quantum dot-sensitized solar cells with an efficiency of 11.75% under 0.1 sun

Abstract

We report the fabrication and photovoltaic performance of new type-II AgInS₂/CdSe core/shell quantum dot-sensitized solar cells (QDSSCs). AgInS₂ core and CdSe shell QDs are grown using the successive ionic layer adsorption and reaction (SILAR) method. X-ray diffraction reveals the orthorhombic and cubic structures for AgInS₂ and CdSe, respectively. Optical spectra reveal the bandgaps of 1.91, 1.78, and 1.57 eV for AgInS₂, CdSe, and AgInS₂/CdSe, respectively. Cyclic voltammetry measurements show a type-II band alignment for the AgInS₂/CdSe QDs. Liquid-junction AgInS₂/CdSe QDSSCs are fabricated using a polysulfide electrolyte. Single-layered AgInS₂ and CdSe QDSSCs are prepared simultaneously for comparison. The AgInS₂/CdSe cell yields a PCE of 8.39% (V_oc = 0.64 V, FF = 56.0%) under 1 sun, significantly higher than the single-layered AgInS₂ (3.04%) and CdSe (3.67%) QDSSCs. Moreover, the J_sc increases from 7.72 in single-layered AgInS₂ to 23.28 mA cm⁻² in core/shell AgInS₂/CdSe. The PCE of AgInS₂/CdSe further increases to 11.75% under 0.1 sun. The PCE of 8.39% is comparable to the best PCE (8.26%) of Dy-doped CdTe/CdS for all type-II core/shell QDSSCs reported. The improved performance of the core/shell QDSSCs is attributed to the spatial separation of electron/hole, cascade energy structure, broader absorption band, and a possible indirect interfacial transition.

---

1 Introduction

Quantum dot-sensitized solar cells (QDSSCs) are a potential low-cost alternative to Si-based photovoltaics. A QDSSC adopts the basic architecture of a dye-sensitized solar cell in which a mesoporous TiO₂ photoanode is coated with a layer of light-absorbing semiconductor QDs. The utilization of semiconductor QDs as sensitizers has the advantages of a high absorption coefficient,¹ tunable bandgap due to the quantum-size effect,² solution processability, and the possibility of multiexciton generation by a single photon.³ The most widely studied sensitizer materials for QDSSCs have been binary metal sulfides such as CdS, CdSe, PbS, PbSe, Ag₂S, etc.^4–9 Due to the narrow absorption ranges, the power conversion efficiencies (PCEs) of QDSSCs based on binary metal sulfides are low, typically ∼3–5%.^10–12 Recently, ternary QDSSCs, such as CuInSe₂, CuInS₂, and CdSe_xTe_1−x, have shown improved PCEs of ∼7–9%.^13–15 The highest efficiency for QDSSCs was obtained in the quaternary system Zn–Cu–In–Se (ZCISe) with PCEs of ∼13–15%.^16,17 However, these best PCEs are still lower than those of other solar materials, such as Si and perovskite, which exceed 20%. The lower efficiency can be attributed to inefficient light harvesting due to the narrow absorption ranges of the QDs and high carrier recombination occurring at the QD surface. The absorption range can be expanded by lowering the bandgap through several approaches, such as alloying, co-sensitizing, or the core/shell structure. In the alloying strategy for metal sulfides, the band gap can be tuned by changing the stoichiometries of the constituent cation or anion elements via cation alloying, anion alloying, or cation/anion alloying. For example, the bandgap of CdSe_xTe_1−x can be reduced from 700 to 900 nm by controlling the ratio of the anion S with Se, leading to significantly improved performance.¹⁸ In the second strategy of co-sensitization, two QDs with different bandgaps are deposited simultaneously on the photoanode. The synergetic absorption allows the co-sensitized QDs to extend the absorption range and improve efficiency. Many co-sensitized QD systems, such as CdS/CdSe, CdS/CdTe, CdSe/CdTe, and CuInS/CdS, have been investigated.^19–22

Core/shell QDs are another effective approach to expanding the light absorption range of solar cells. A core/shell QD is formed by combining two different QDs into a single QD in the core/shell geometry. The core/shell QD systems can be classified into four types: type-I, reverse type-I, type-II, and quasi-type-II, depending on the bandgap and the relative band alignment between the conduction band (CB) and valence band (VB) edges of the core and shell materials. In type-I core–shell QDs, the bandgap of the core material is lower than that of the shell material. The CB edge of the shell is higher than that of the core, and the VB edge of the shell is lower than that of the core. Therefore, the electrons and holes are both localized in the core. Many type-I core/shell QD systems have been studied, including CdSeTe/CdS, PbS/CdS, CuInS₂/ZnS, etc., with the best PCE of ∼9%.^23–25 The reverse type-I core/shell QD system is opposite to the type-I QD, in which the bandgap of the core material is larger than that of the shell material. Both electrons and holes are now delocalized in the shell region, which allows fast injection of the photoelectrons from the shell to the electrolyte and the TiO₂ electron transporting layer. CdS/CdSe is the most extensively studied reverse type-I core/shell QD system, with high PCEs of 5.32% by Pan et al. and 7.24% by Cao et al., respectively.^26,27 In a type-II core/shell QD system, the shell's CB edge (or the VB edge) is situated between the CB and VB edges of the core. The electrons are located in the CB of the shell, and the holes are located in the VB of the core, forming separated electron and hole wavefunctions. The spatial separation of electrons and holes reduces carrier recombination, which prolongs the electron lifetime, a beneficial effect for photovoltaic devices. An additional advantage of type-II core/shell QDs is the notable extension of the absorption edge towards the red spectrum, which occurs due to the exciplex state, providing an extra pathway for enhancing spectral response through effective bandgap reduction.²⁸ The energy bandgap of the exciplex state extends the absorption ranges of the type-II core/sell QDs to a longer wavelength than the constituent core and shell, resulting in improved light harvesting ability. The first type-II core/shell QD system was ZnSe/CdS, reported by Ning et al.²⁹ Other type-II core/shell QD systems, including CdTe/CdSe, ZnTe/CdSe, ZnTe/ZnSe, CdSe_xTe_1−x/CdS, etc., have subsequently been reported.^30–33 The best PCEs among all type-II core/shell QDSSCs reported to date are ZnTe/CdSe (7.17%)²⁸ and Dy-doped CdTe/CdS (8.26%).³⁴

Currently, the materials of core/shell QDs are primarily limited to binary sulfides such as CdS, CdSe, ZnSe, ZnTe, etc. In contrast, ternary semiconductors provide a broader selection of materials because there are many more ternary semiconductors in nature. Amongst the many ternary metal sulfides, the I–III–VI semiconductor AgInS₂ is an attractive candidate for solar absorbers. AgInS₂ has several advantageous properties: (a) a high absorption coefficient (>10⁴ cm⁻¹) in the visible region,^35,36 (b) direct bandgap, (c) suitable bandgap:³⁷ 1.86 eV and 2.03 eV ³⁸ for tetragonal structure, and 1.98 eV ^37,38 and 2.09 eV ³⁹ for orthorhombic structure, close to the ideal Shockley–Queisser gap of 1.4 eV.⁴⁰ These properties make AgInS₂ a potential solar absorber. AgInS₂ QDSSCs have been reported to achieve a PCE of 2.91%.⁴¹

Here, AgInS₂/CdSe type-II core/shell QDs were fabricated by employing AgInS₂ as the core material and CdSe as the shell material. The binary metal sulfide CdSe was chosen as the shell because it has a suitable bandgap of 1.7 eV for the solar absorber. The band alignment of AgInS₂ and CdSe forms a type-II core/shell structure. The AgInS₂ core and CdSe shell QDs were synthesized using the successive ionic layer adsorption and reaction (SILAR) method.^42,43 Single-layered AgInS₂ and CdSe QDSSCs were fabricated simultaneously for comparison. The AgInS₂/CdSe core/shell QDSSCs yielded a PCE of 8.39% under 1 sun, significantly higher than the 3.04% of the single-layered AgInS₂ core and 3.67% of the CdSe shell QDSSCs. Meanwhile, the J_sc increased from 7.72 mA cm⁻² in AgInS₂ to 23.28 mA cm⁻² in AgInS₂/CdSe. The PCE of the AgInS₂/CdSe QDSSC further increased to 11.7% at the reduced light intensity of 0.1 sun. To our best knowledge, the PCE of 8.39% (1 sun) here is higher than the best PCE of 8.26% in Dy-doped CdTe/CdS among all type-II core/shell QDSSCs reported so far. We discuss the possible reasons for the enhanced photovoltaic performance in the core/shell structure.

2 Materials and methods

Fig. 1 illustrates a schematic diagram of a liquid-junction AgInS₂/CdSe core/shell QDSSC consisting of three main components: a photoelectrode, a counter electrode, and a polysulfide electrolyte. Details of preparation for each component are described as follows.

Fig. 1 Schematic diagram of an AgInS₂/CdSe core/shell QDSSC.

2.1 Materials

Fluorine-doped tin oxide (FTO, 7 Ω per square) glass was purchased from Ruilong (Taiwan). TiO₂ 30 NR-D paste and TiO₂ WER2-0 Titania Paste were obtained from GreatCell Solar. Silver nitrate (Ag(NO)₃) and copper(II) sulfate pentahydrate (CuSO₄·5H₂O) were purchased from Honeywell Fluka. Sodium sulfide nonahydrate (Na₂S·9H₂O), selenium powder, and sodium thiosulfate anhydrous (Na₂S₂O₃) were obtained from Thermo Fisher Scientific. Zinc acetate dihydrate (Zn(CH₃COO)₂·2H₂O) and potassium chloride (KCl) were from JT Baker Chemical Company. Cadmium acetate (Cd(CH₃COO)₂·2H₂O), indium chloride (In(Cl)₃), and sodium borohydride (NBrH₄) were from Alfa Aesar. Urea (CH₄N₂O) was from Sigma Aldrich. The sulfur powder was from Showa Chemical. All materials were used directly without any further purification.

2.2 Preparation of TiO₂ electrodes

The TiO₂ electrode was prepared using the following steps. First, a TiO₂ blocking layer, 50–70 nm in thickness, was coated on the precleaned FTO glass by spin-coating 50 μl of a 0.247 M titanium(IV) isopropoxide solution. Then, a mesoporous TiO₂ (mp-TiO₂) layer with a thickness of 10–12 μm, particle size ∼30 nm, and 0.3 cm in diameter was coated on top of the blocking layer by doctor-blading TiO₂ paste (Greatcell DSL 30NR-T), then heated at 120 °C for 5 min. Thirdly, a TiO₂ scattering layer, particle size 350 nm, thickness ∼5 μm, and diameter 0.5 cm, was coated on top of the mp-TiO₂ layer by doctor blading a prepared solution (Greatcell WER4-0). The prepared TiO₂ electrode was finally annealed at 500 °C for 90 min.⁴⁴

2.3 Preparation of AgInS₂ core QDs

AgInS₂ QDs were grown onto an mp-TiO₂ electrode using a two-stage successive ionic layer adsorption and reaction (SILAR) method.^45–48 First, binary Ag–S QDs seeds were grown onto an mp-TiO₂ electrode. Second, binary In–S QD seeds were deposited on the Ag–S QD seeds. Post-annealing at an elevated temperature transformed the Ag–S/In–S double-layered structure to the AgInS₂ phase. In detail, the mp-TiO₂ electrode was first dipped into a 25 °C, 0.05 M Ag(NO)₃ ethanol solution for 30 s, then rinsed using ethanol and dried on a hotplate at 40 °C. Then, the sample was immersed into a 25 °C, 0.1 M Na₂S·9H₂O methanol/DI water (1 : 1, v/v) solution for 120 s, rinsed with methanol, and dried on a hotplate at 40 °C. These two steps were denoted as one SILAR cycle of Ag–S. The number of cycles was fixed into two to obtain the suitable amount of Ag–S seeds. For the In–S seeds deposition, the Ag–S-coated TiO₂ electrode was immersed into a 25 °C, 0.1 M In(Cl)₃ ethanol solution for 120 s, rinsed with ethanol, and dried on the hotplate at 40 °C. Then, the sample was dipped into a 25 °C, 0.1 M Na₂S·9H₂O methanol/DI water (1 : 1, v/v) solution for 120 s and rinsed with methanol and dried on a hotplate at 40 °C. The two-dipping process was referred to as one SILAR cycle of In–S seeds. The number of cycles was fixed at seven to obtain the suitable amount of AgInS₂ QDs. Finally, the sample was annealed under the nitrogen atmosphere at 300 °C for 30 min, yielding the AgInS₂ core QDs.

2.4 Preparation of CdSe shell QDs

The CdSe shell was grown onto the AgInS₂-coated mp-TiO₂ electrode using the SILAR method. In detail, a Se precursor solution was prepared by dissolving 0.36 g of selenium powder into 125 ml of ethanol solution containing 0.3 g of NBrH₄. The AgInS₂-coated TiO₂ electrode was dipped into a 0.1 M Cd(CH₃COO)₂·2H₂O methanol solution for 60 s, rinsed with methanol, and dried on a hotplate at 40 °C. The sample was then immersed in the Se precursor solution for 90 s, rinsed using ethanol, and dried on a hotplate at 40 °C. This two-step dipping process was called one SILAR cycle of CdSe. The cycle was repeated six times to obtain the desired amount of CdSe shell QDs.

2.5 Preparation of a ZnSe passivation layer

The synthesized AgInS₂/CdSe core/shell QDs were passivated with a ZnSe layer using the SILAR method to reduce carrier recombination. Briefly, the prepared sample was dipped into a 0.1 Zn(CH₃COO)₂·2H₂O DI-water solution for 120 s, rinsed with DI-water, and dried on a hotplate at 40 °C. Then, the sample was immersed in the Se solution mentioned above for 90 s, rinsed with ethanol, and dried on a hotplate at 40 °C. This two-step process was repeated four times, forming a ZnSe passivation layer on the AgInS₂/CdSe core/shell QDs.

2.6 Assembly of solar cells and characterization

The AgInS₂/CdSe core/shell QDs were assembled into the liquid-junction QD-sensitized solar cell structure by sandwiching the prepared AgInS₂/CdSe electrode with a CuS counter electrode using a parafilm spacer of thickness 190 μm. The CuS counter electrode was prepared using the chemical bath deposition method.⁴⁷ In brief, three to six precleaned FTO glass was immersed horizontally in a 40 ml solution containing 1.564 g of CuSO₄·5H₂O, 3.968 g of Na₂S₂O_3, and 0.969 g of CH₄N₂O, for 70 min on a hotplate at 55 °C, rinsed with DI-water and dried at 60 °C. The polysulfide electrolyte, consisting of 1 M Na₂S, 2 M S, and 0.2 M KCl in methanol/water (7 : 3 by volume), was injected into the space between the TiO₂ electrode and the CuS counter electrode before J–V measurements.

The photocurrent–voltage (J–V) characteristics were measured using a Keithley 2400 source meter under 100 mW cm⁻² light illumination from a 150 W Xe lamp. A batch of three to six samples was measured parallel to ensure data consistency. A metal mask of the size 3 mm × 3 mm was used during measurements to avoid overestimated current densities. The X-ray diffractometer (XRD) patterns were obtained using a Bruker XRD D2 Phaser. Transmission electron microscopy (TEM) was conducted using JEOL JEM-2010. Energy dispersive spectroscopy (EDS) was analyzed using a JEOL JSM-6700F scanning electron microscope (SEM). Optical characteristics were obtained from a Hitachi U-2800A UV-vis spectrophotometer. External quantum efficiency (EQE) spectra were collected from an Acton monochromator with a 250 W tungsten halogen lamp source (without white-light biasing).

2.7 Electrochemical measurements

Cyclic voltammograms (CV) were retrieved from Bio-Logic SAS SP-150. The reference electrode employed in this study was an Ag/AgCl electrode, while the counter electrode utilized platinum (Pt). The working electrode was prepared through the following steps: (i) a mesoporous titanium dioxide (mp-TiO₂) layer, possessing a thickness ranging from 18 to 20 μm, particle size of 30 nm, and dimensions of 1.2 × 1 cm, was deposited onto the blocking layer using the doctor-blading technique with TiO₂ paste; (ii) the AgInS₂ and CdSe quantum dots (QDs) were synthesized within the mp-TiO₂ layer following established procedures. Electrochemical experiments were executed within a sample vial equipped with a specialized cap. The reference electrode was strategically positioned between the working electrode, connected to the equipment via an electric clamp, and the Pt counter electrode. The measurement comprised several cycles within a potential range spanning from −1.0 to 2.0 V, employing a scan rate of 100 mV s⁻¹. A 0.1 M KCl solution served as the supporting electrolyte during the electrochemical analysis. In a setup where AgInS₂ and CdSe QDs are grown on mp-TiO₂ nanoparticles, the cyclic voltammetry (CV) signal predominantly reflects the redox reactions at the electrode surface. Despite TiO₂ being present, the electrochemical behavior of AgInS₂ and CdSe QDs dominates the observed signal. Utilizing the anodic and cathodic peaks of AgInS₂ and CdSe QDs grown onto mp-TiO₂ nanoparticles is a reasonable approach to estimating the position of the valence band (VB) and conduction band (CB) position. Therefore, in this study, the valence band level (E_VB) and conduction band level (E_CB) positions were estimated by analyzing oxidation and reduction peak potentials.⁴⁹ After completing each round of experiments, the potentials were adjusted in reference to the normal hydrogen electrode (NHE) and the potential of the Ag/AgCl electrode. The electrochemical bandgap was calculated as the difference between these peak potentials.⁵⁰

3 Results and discussion

3.1 Structural analysis

The crystalline structures of the synthesized QDs were analyzed using XRD. As depicted in Fig. 2(a), all peaks from AgInS₂ QDs match correctly with the orthorhombic AgInS₂ crystal structure (JCPDS No. 00-025-1328) without impurities; some peaks overlap with the TiO₂ peaks (JCPDS No. 00-002-0387), namely the (1 2 0) peak at 24.99°, (2 0 2) at 37.12°, and (1 2 3) at 48.05°, while the AgInS₂ peaks of (0 0 2), (1 2 1), (2 0 1), (0 4 0), (3 2 0), (2 4 0), (0 4 2), and (3 2 2) are well separated with the TiO₂ background. The CdSe QDs show the cubic structure (JCPDS No. 00-019-0191), consisting of three prominent peaks: (1 1 1) at 25.35°, (2 2 0) at 42.01°, and (3 1 1) at 49.67°. In the XRD pattern of AgInS₂/CdSe core/shell QDs, some AgInS₂ peaks, such as (0 0 2) at 26.58°, (1 2 1) at 28.38°, (2 0 1) at 28.75°, (0 4 0) at 43.69°, and (3 2 0) at 44.53°, become less apparent due to partial overlap with the CdSe peaks of (1 1 1) at 25.35°, and (2 2 0) at 42.01°. The deconvoluted analysis (ESI Fig. S1†) shows that the patterns contain all corresponding peaks of AgInS₂ and CdSe, indicating the successful formation of AgInS₂ and CdSe crystal structures. The SAED peaks of AgInS₂ QDs (Fig. 2(b)) and AgInS₂/CdSe QDs (Fig. 2(c)) are in agreement with the XRD analysis.

Fig. 2 (a) XRD pattern of AgInS₂ QDs, CdSe QDs, and AgInS₂/CdSe core/shell QDs, (b) selected area (electron) diffraction (SAED) of AgInS₂ QDs, and (c) SAED of AgInS₂/CdSe core/shell QDs.

In this study, lattice constants of the synthesized QDs were determined by analyzing the XRD peak shifts of AgInS₂ and CdSe concerning the standard position within the XRD pattern of related materials. These measurements were compared with the lattice parameters of ideal single crystals of AgInS₂ and CdSe obtained from a reputable XRD database. The findings, showcased in Table 1, present the calculated results of these comparisons.

Table 1 Lattice constant and volume of the cell of ideal single crystal AgInS₂ (JCPDS No. 00-025-1328), synthesized AgInS₂, ideal single crystal CdSe (JCPDS No. 00-019-0191), and synthesized CdSe

Sample	Lattice constant (Å)			Volume of cell (pm^3)
	a	b	c
Ideal single crystal AgInS2 (JCPDS No. 00-025-1328)	7.001	8.278	6.698	388.18 × 10^6
Synthesized AgInS2 QDs	6.989	8.204	6.682	383.13 × 10^6
Ideal single crystal CdSe (JCPDS No. 00-019-0191)	6.077	6.077	6.077	224.42 × 10^6
Synthesized CdSe QDs	6.008	6.008	6.008	216.86 × 10^6

---

Table 1 shows that the determined lattice constants of the synthesized QDs differ from the database references for ideal crystal structures of AgInS₂ (JCPDS No. 00-025-1328) and CdSe (JCPDS No. 00-019-0191), implying that the synthesized QDs might undergo the lattice distortion phenomenon.⁵¹ Furthermore, Table 2 includes calculations of the strain resulting from varied d-spacings obtained using the Bragg equation and eqn (1) compared to the ideal d-spacings listed in the same XRD database.

where ε is the microstrain, Δd is the difference in d-spacing, and d_reference is the reference d-spacing.

Table 2 Calculated microstrain resulting from the difference between the measured d-spacing and the ideal single crystal d-spacing

Sample	Miller indices	Measured d-spacing (Å)	Ideal d-spacing (Å)	Calculated microstrain (× 10^−6)
AgInS2	002	3.341	3.350	−2.687
	040	2.051	2.070	−9.178
	320	2.026	2.033	−3.445
CdSe	220	2.124	2.149	−11.633

---

Table 2 reveals that all calculated microstrain values exhibit a negative sign, indicating compressive strain and suggesting compression or contraction within the crystal lattice. Specifically, CdSe QDs demonstrate a larger microstrain value than AgInS₂ QDs. This difference can be attributed to the growth process utilized for CdSe in this study, which omitted the annealing step after SILAR. The absence of annealing can cause various effects, including incomplete relaxation of internal stresses, limited structural reorganization, and accumulation of defects, resulting in larger strain.

The preceding outcomes suggest that these distortions and strains emerged during the QD growth process on TiO₂ using the SILAR method. Several contributing factors may have led to these lattice distortions and strains, such as (i) SILAR involves sequential adsorption of precursor ions or molecules followed by reaction or deposition, leading to nanoscale structure growth. Irregularities in nucleation, growth kinetics, or layer thickness variations contribute to lattice strain; (ii) interface mismatches between layers or substrate in SILAR-deposited materials introduce strain due to differences in lattice parameters or surface energies; (iii) epitaxial or quasi-epitaxial growth in SILAR may cause strain if deposited layers deviate from substrate orientation; (iv) surface defects, variations in layer quality, or surface states influence material strain; (v) the deposition process induces mechanical stress from factors like island growth or lattice mismatch, (vi) quantum dot boundary defects with differing orientations can cause lattice distortion. Identifying specific strain-inducing factors often necessitates detailed analysis using techniques such as transmission electron microscopy (TEM). The advanced characterization of the lattice mismatch using high-resolution transmission electron microscopy (HR-TEM) was carried out to enhance the analysis, as detailed in the subsequent explanation. While the above analysis requires careful consideration of instrumental effects influencing discrepancies between the synthesized QDs and the ideal single crystal of related materials, combining XRD pattern analysis with TEM findings enables the identification of defects in AgInS₂ and CdSe QDs grown on TiO₂ nanoparticle surfaces. Lattice distortion, strain, and lattice mismatch contribute to the formation and propagation of defects in the crystal lattice. These defects may account for unfavorable charge carrier recombination, a factor further detailed in the subsequent photovoltaic performance analysis.

3.2 Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) analysis

The surface morphology of TiO₂/AgInS₂ core and TiO₂/AgInS₂/CdSe core/shell QDs was analyzed using SEM, as depicted in Fig. 3(a) and (b). The TiO₂/AgInS₂ sample shows a porous structure. The AgInS₂/CdSe QDs exhibit a slightly larger particle size. The elemental composition of the QDs was analyzed using SEM-EDS. As shown in Fig. 3(c), the atomic percentage of Ag, In, and S were 1.57%, 1.15%, and 2.00%, respectively, yielding an elemental ratio of 1 : 0.73 : 1.27, which can be approximated to a 1 : 1 : 2 ratio; this result is in good accordance with the XRD pattern. In the AgInS₂/CdSe core/shell QDs (Fig. 3(d)), the Cd and Se atomic percentages were 5.10% and 6.12%, giving an approximate ratio of 1 : 1 for Cd : Se. Meanwhile, the atomic percentage of Ag, In, and S slightly changed to 1.37%, 1.37%, and 2.32%, respectively, maintaining the 1 : 1 : 2 ratio for Ag : In : S.

Fig. 3 Top view of SEM pictures of (a) TiO₂/AgInS₂ and (b) TiO₂/AgInS₂/CdSe core/shell with a magnification of 100 000 times; EDS spectra of (c) TiO₂/AgInS₂ and (d) TiO₂/AgInS₂/CdSe core/shell.

3.3 Transmission electron microscopy (TEM) analysis

The morphology of the synthesized QDs was investigated using TEM, as shown in Fig. 4. The bare mp-TiO₂ nanoparticles have a hexagonal shape with round edges and an average particle size of 30 nm (Fig. 4(a)). The AgInS₂ QDs, denoted with red arrows in Fig. 4(b), were distributed randomly on TiO₂ nanoparticles. The AgInS₂ particle sizes were in the range of 4–12 nm with an average size of 6.3 nm (the normal distribution of particle sizes is depicted in Fig. 4(g)). Fig. 4(d) depicts the lattice fringes of three individual AgInS₂ QDs with different planes. The d-spacings of the lattice fringes for the (1 2 3) plane were 0.185 nm and 0.188 nm for two individual QDs, showing reasonable consistency. The lattice fringes with the (0 4 0) plane have a d-spacing of 0.208 nm. These results agree with the orthorhombic AgInS₂ crystal structure (JCPDS No. 00-025-1328). From Fig. 4(d), it is evident that two neighboring QDs possess different orientations, resulting in a calculated lattice mismatch of 9.26% and a detected grain boundary. The mismatch between their lattices leads to strain accumulation at the interface. This accumulated strain, aiming for relief, instigates the emergence of defects or irregularities near the boundary. Additionally, the lattice distortion caused by this mismatch induces defect formation to accommodate the strain. These accumulated defects may contribute to the development of specific structural anomalies near the boundary. These findings suggest the existence of such imperfections within the AgInS₂ QDs grown on TiO₂ nanoparticles. The subsequent section analyzing photovoltaic performance further explores how these imperfections impact the overall efficiency of the photovoltaic system.

Fig. 4 TEM images of (a) mp-TiO₂ nanoparticles, (b) AgInS₂ QDs grown onto mp-TiO₂ nanoparticles (the red arrows display the AgInS₂ QDs), (c) AgInS₂/CdSe QDs grown onto mp-TiO₂ nanoparticles (the red arrows display the AgInS₂/CdSe QDs), (d) lattice fringes of AgInS₂ QDs, (e) lattice fringes of AgInS₂/CdSe core/shell QDs (the red arrow indicates the diagram shown in (f)), (f) diagram that shows the boundary of the TiO₂, AgInS₂, and CdSe QD images shown in (e), (g) distribution of particle size of AgInS₂, and (h) distribution of particle size of AgInS₂/CdSe.

The formation of AgInS₂/CdSe core/shell QDs enlarged the overall particle sizes, as given in Fig. 4(c); the AgInS₂/CdSe core/shell QDs were distributed densely on the mp-TiO₂; the particle sizes in the range of 7–15 nm with an average particle size of 11.7 nm (Fig. 4(h) depicts the normal distribution of the particle size). As illustrated in Fig. 4(e), the AgInS₂/CdSe core/shell QD is grown onto mp-TiO₂, which has three different planes: (2 0 0), (2 2 0), and (3 1 2). The CdSe QD with the plane of (3 1 1) is attached to the mp-TiO₂ ((2 0 0) plane) and covers the AgInS₂ core QD with the plane of (0 4 0), while the CdSe plane of (2 2 0) is adhered with another side of AgInS₂ core QD ((0 4 0) plane). All the corresponding lattice fringes agree with the crystal structures of each material. Fig. 4(f) illustrates a schematic diagram that shows the boundary of the respective CdSe QD, AgInS₂ QD, and a TiO₂ nanoparticle shown in Fig. 4(e). As seen, the AgInS₂ QD is surrounded by the CdSe QD. Meanwhile, the CdSe QD is attached to the TiO₂ nanoparticle. The diagram further elucidates the formation of the core/shell QD structure. Based on the observations in Fig. 4(e), the CdSe shell manifests diverse orientations, resulting in a lattice mismatch of 13.27%. Notably, the calculated lattice mismatch between the CdSe shell and TiO₂ nanoparticles varies significantly, ranging from 36.02% to 45.02%, depending on the nanoparticle orientation. However, the lattice mismatches between the AgInS₂ core and the CdSe shell are comparatively lower at 6.64% and 7.65%. Essentially, the strain accumulation primarily responsible for inducing defects at the interface predominantly stems from the lattice mismatch between the CdSe shell and TiO₂ nanoparticles. These results align with the XRD pattern analysis related to the strain analysis.

3.4 Optical spectra analysis

Fig. 5 compares the optical spectra of AgInS₂, CdSe, and AgInS₂/CdSe QDs. The transmission spectra T(λ) of each sample in Fig. 5(a) were measured as I_QDs/I_TiO2, where I_QDs and I_TiO2 are the light intensity passing through TiO₂/QDs sample and TiO₂ sample, respectively. The transmission spectra shifted to the longer wavelength region in the order of AgInS₂, CdSe, and AgInS₂/CdSe core/shell, indicating that the bandgap decreases with the order of AgInS₂, CdSe, and AgInS₂/CdSe core/shell. The addition of the ZnSe passivation layer appears to exert negligible influence on the optical spectra, as depicted in Fig. 5(a). Fig. 5(b) shows the absorbance spectra A(λ) = −log₁₀T(λ). The AgInS₂/CdSe core/shell had the largest absorbance value at a given wavelength, CdSe the second, and AgInS₂ the smallest. For example, at λ = 700 nm, A = 0.504, 0.114, and 0.086 for AgInS₂/CdSe, CdSe, and AgInS₂, respectively. Fig. 5(c) shows the Tauc plots of the samples. The optical energy bandgap (E_g,op) was estimated by finding the intercept of the extrapolated line to the x-axis. The Tauc plots yielded the E_g of 1.91 eV (AgInS₂), 1.78 eV (CdSe), and 1.57 eV (AgInS₂/CdSe), respectively. The 1.91 eV of AgInS₂ is consistent with the bandgap of 1.98 eV for orthorhombic AgInS₂.^37,38 The 1.78 eV of CdSe also agrees with the reported CdSe bandgap.^52–54 Intriguingly, the core/shell QDs had an E_g of 1.57 eV, lower than the constituent core AgInS₂ (1.91 eV) and CdSe shell (1.78 eV). Two possible explanations are: (i) the core/shell QDs were synthesized using a number of SILAR cycles larger than the core and shell. As reported by Robinovich and Hodes, the E_g of QD samples prepared by SILAR would decrease with an increasing number of SILAR cycles.⁵⁵ The lower E_g of the core/shell QDs could be a consequence of the larger number of SILAR cycles used, and (ii) the type-II core/shell structure induces a narrow effective bandgap owing to the presence of the exciplex state. This will be explained in detail in the section on cyclic voltammetry below. The ZnSe passivation layer has minimal impact, with the energy bandgap remaining approximately consistent with that of the AgInS₂/CdSe core–shell QDs at 1.57 eV.

Fig. 5 Optical spectra of AgInS₂, CdSe, AgInS₂/CdSe: (a) transmission spectra T(λ), (b) absorbance A(λ), and (c) Tauc plot (Ahυ)²vs. hυ.

3.5 Cyclic voltammetry analysis

The determination of the core/shell type of AgInS₂/CdSe QDs relies significantly on the positions of the valence band level (E_VB) and conduction band level (E_CB) of AgInS₂ and CdSe QDs grown onto TiO₂ nanoparticles. The estimation of band alignment was achieved through CV measurements, as exemplified in Fig. 6. In Fig. 6(a) and (b), during the potential scan from 0 V to a more positive value, an anodic peak (A) emerges, and conversely, upon reversing the cycle, a cathodic peak (C) is observed. These anodic and cathodic peaks are attributed to the electron transfer between the QDs and the working electrode through the E_VB and E_CB, respectively. Consequently, the E_VB and E_CB were calculated from the onset of anodic (A) and cathodic (C) peaks of the QDs. The onset of anodic peak (E^ox_onset) was calculated by determining the intersection of two tangents drawn at the point where the oxidation current begins to rise and the background current in the cyclic voltammograms. Similarly, the onset of the cathodic peak (E^red_onset) was computed by identifying the point where the reduction current starts to increase and determining the intersection of two tangents drawn at that position and the background current in the cyclic voltammograms; the numerical values enclosed in brackets within Fig. 6(a) and (b) indicate the onset potential. The energy levels were calculated using formulas (2) and (3).⁵⁰

E_CB = −e(E^red_onset + 4.4) eV (2)

E_VB = −e(E^ox_onset + 4.4) eV (3)

Fig. 6 Cyclic voltammetry (CV) curves of (a) AgInS₂ QDs, (b) CdSe QDs grown onto TiO₂ nanoparticles; the letters A and C in the figures express the anodic and cathodic peak, respectively; the number in the bracket shows the onset potential, (c) band alignment of AgInS₂ and CdSe QDs determined from CV curves, and (d) schematic diagram of AgInS₂/CdSe type-II core/shell band alignment with an energy bandgap from exciplex state.

Based on Fig. 6(a), the AgInS₂ QDs showed the oxidation peak of Ag at 0.14 V; the E_VB and E_CB of AgInS₂ QDs are −5.85 eV and −4.09 eV, respectively, resulting in a bandgap (E_g,CV) of 1.76 eV. The bandgap derived from CV measurements may not invariably align with the optical bandgap, as it hinges on the specific semiconductor material under investigation. The observed peak shift in CV could potentially arise from trapping sites enabling numerous electrons and holes to engage in the reaction. This distinction underscores that the CV-derived bandgap might not be directly synonymous with the optical bandgap due to the influence of trapping mechanisms within the semiconductor material.⁵⁶ The outcomes observed in this study align with prior research investigating the energy bandgap using CV. Mohan and Renuga documented the E_VB and E_CB of AgInS₂ nanocrystals as −5.61 eV and −3.87 eV, respectively.⁵⁷ Similarly, Jagadeeswararao et al. reported the E_VB and E_CB of AgInS₂ nanocrystals as −6.0 eV and −3.94 eV.⁵⁸ The CdSe QDs in Fig. 6(b) yielded respective E_VB of −5.64 eV and E_CB of −3.89 eV, resulting in an E_g,CV of 1.75 eV. These results are comparable with many other studies measuring the electrochemical bandgap of CdSe QDs. For instance, the reported values for the E_VB and E_CB of CdSe QDs were −5.86 eV and −3.66 eV,⁵⁶ respectively. Additionally, another study documented the E_CB of CdSe QDs as −3.71 eV.⁵⁹ Furthermore, a range of values was observed for the E_VB and E_CB of CdSe QDs, varying in size from 1.9 nm to 3.4 nm. These values spanned from −5.68 eV and −2.83 eV to −3.26 eV and −5.47 eV, respectively.⁶⁰ It is noteworthy to consider potential experimental uncertainties and methodological variations between measurement techniques when comparing these results. Considering this comparative analysis, the inferred E_VB and E_CB values in our study appear logically justifiable, potentially indicating a core–shell type configuration. Consequently, based on this analysis, it is evident that the E_VB and E_CB levels of CdSe QDs are higher compared to the corresponding levels of AgInS₂ QDs, indicating the formation of a type-II core/shell structure, as depicted in Fig. 6(c). In this type-II core/shell QD structure, the electrons are confined to the core while the holes are confined to the shell, leading to effective charge separation.⁵⁴ This structure produces an exciplex state, which offers an additional route (indirect interfacial transition) to improve spectral response, consequently lowering the effective bandgap (E_g (Exciplex)), as shown in Fig. 6(d).²⁸

3.6 Photovoltaic performance

The photovoltaic characteristics of AgInS₂/CdSe core/shell QDSSCs were measured under 1 sun (100 mW cm⁻²) light intensity. Single-layered AgInS₂ core and CdSe shell QDSSCs were studied simultaneously for comparison. Table 3 lists the photovoltaic parameters, including the short-circuit current density (J_sc), open-circuit voltage (V_oc), fill factor (FF), and power conversion efficiency (PCE). Fig. 7(a) depicts the J–V curves for each sample. The single-layered AgInS₂ core QDSSC achieved an average PCE of 3.04% (J_sc = 7.72 mA cm⁻², V_oc = 0.66 V, FF = 60.0%). The PCE of 3.04% achieved in this work is slightly higher than the best PCE of 2.91% for AgInS₂ QDSSCs reported recently.⁴¹ Building upon the theory mentioned in the literature,⁶¹ the maximum attainable J_sc for an ideal solar absorber material, with an absorption onset at a wavelength λ of 713 nm (as per EQE spectra), is estimated to be 20.4 mA cm⁻². Nevertheless, the generated J_sc in this study amounts to only 7.72 mA cm⁻². This outcome could be elucidated by considering the XRD pattern and TEM analysis, which suggest that the reduced J_sc observed in AgInS₂ QDSSCs might stem from increased carrier recombination due to defects linked to imperfections within the AgInS₂ QDs grown on TiO₂. Meanwhile, the single-layered CdSe shell QDSSCs achieved an average PCE of 3.67% (J_sc = 12.9 mA cm⁻², V_oc = 0.52 V, FF = 55.1%). The low PCE observed in CdSe shell QDSSCs can also be ascribed to carrier recombination due to imperfections within the CdSe QDs cultivated on TiO₂. Nevertheless, the AgInS₂/CdSe core/shell QDSSCs (sample no.5) achieved an average PCE of 6.49% (J_sc = 16.95 mA cm⁻², V_oc = 0.59 V, FF = 64.9%), a significant increase that nearly doubles the PCEs of the single-layered core and shell QDSSCs. Higher J_sc is assigned to the extension of the light-absorption range (explained in the optical spectra) and the multiplication of electron injections from the core or the shell. However, the V_oc decreases by 11%, which can be explained based on the standard diode eqn (4)where T is the temperature, k_B is the Boltzmann constant, q is the electron charge, and J₀ is the saturation current density. The J₀ depends on the recombination in the solar cells; therefore, based on eqn (4), lower V_oc is ascribed to a higher amount of recombination in the cells. As mentioned in the XRD pattern and TEM analysis, the CdSe generated through the SILAR method without annealing harbors defects arising from lattice mismatches and crystal imperfections, subsequently functioning as recombination centers. The AgInS₂/CdSe core/shell QDSSCs were then coated with a ZnSe passivation layer to suppress carrier recombination. The ZnSe-coated AgInS₂/CdSe/ZnSe QDSSC (sample no. 7) yielded an average PCE of 8.39% (J_sc = 23.28 mA cm⁻², V_oc = 0.64 V, FF = 56.0%), a 29% improvement over that of the uncoated core/shell cell (sample no. 5). As depicted in Fig. 7(c), the presence of the ZnSe passivation layer effectively hinders the recombination of electrons originating from the conduction bands of CdSe, AgInS₂, or TiO₂ with S_n²⁻ in the electrolyte. This preventive effect contributes to an enhancement in the overall performance of the solar cells. The PCE of 8.39% achieved here is higher than the best PCEs among all type-II core/shell QDSSCs reported to date, such as ZnTe/CdSe (7.17%) and Dy-doped CdTe/CdS (8.26%).^31,34

Table 3 Photovoltaic performance of single-layered AgInS₂, CdSe and core/shell AgInS₂/CdSe QDSSCs

No.	Sample	J sc (mA cm^−2)	V oc (V)	FF (%)	PCE (%)
a Average value from three cells. b Champion cell.
1	AgInS2^a	7.72 ± 0.75	0.66 ± 0.02	60.0 ± 0.3	3.04 ± 0.29
2	AgInS2^b	8.58	0.66	59.7	3.38
3	CdSe^a	12.9 ± 1.18	0.52 ± 0.00	55.1 ± 6.4	3.67 ± 0.08
4	CdSe^b	11.55	0.52	62.42	3.75
5	AgInS2/CdSe^a	16.95 ± 0.43	0.59 ± 0.02	64.9 ± 1.6	6.49 ± 0.04
6	AgInS2/CdSe^b	16.73	0.61	63.9	6.53
7	AgInS2/CdSe/ZnSe^a	23.28 ± 1.09	0.64 ± 0.02	56.0 ± 0.7	8.39 ± 0.18
8	AgInS2/CdSe/ZnSe^b	23.85	0.65	55.2	8.57

---

Fig. 7 (a) J–V curves of AgInS₂, CdSe, AgInS₂/CdSe core/shell, and AgInS₂/CdSe core/shell/ZnSe under 1 sun, (b) J–V curves of AgInS₂/CdSe core/shell/ZnSe under different sun intensities, and (c) schematic diagram of electron transfer from an AgInS₂/CdSe core/shell QD into a TiO₂ nanoparticle with a ZnSe passivation layer.

Semiconductor nanoparticles prepared by SILAR contain a large number of defects due to the low-temperature synthesis, as evident from the XRD pattern and TEM analysis. These defects act as recombination centers. When illuminated under 1 sun, representing a higher photon influx, there is an increased generation of electron–hole pairs. However, a substantial portion of these pairs undergo recombination. Consequently, scrutinizing the current–voltage (J–V) curves at low light intensities proves valuable in alleviating the adverse effects of recombination from these defects, thereby improving the overall performance. Fig. 7(b) shows the I–V curves of the best AgInS₂/CdSe core/shell QDSSCs under various reduced sun intensities. Table 4 lists the photovoltaic parameters. The PCE enhanced from 8.57% (1 sun) to 11.15% (0.25 sun) and further to 11.75% (0.1 sun), a 36.5% improvement over the 1 sun data. The enhanced PCE arose from both J_sc and FF, J_sc boosted from 23.85 mA cm⁻² (1 sun) to 3.07 mA cm⁻² (0.1 sun), which could be normalized to 30.7 mA cm⁻² (1 sun). In addition, the FF increased from 55.26% (1 sun) to 64.80 (0.1 sun). The enhanced J_sc and FF indicate improved photocarrier collection. The improvement can be explained in terms of the multiple trapping model for carrier recombination proposed by Tachiya and Seki.⁶² The model assumes carriers could be trapped and de-trapped repeatedly from trap surface sites. The theory predicts a sublinear J_scvs. light intensity I₀ power law: J_sc ∝ I₀^α (α < 1). Analysis of the J_scvs. I₀ data in Table 3 yielded α = 0.89 (ESI Fig. S2†), consistent with the multiple trapping model. Similar sublinear relations n_e ∝ I₀^0.6 and n_e ∝ I₀^0.45 have been reported in dye-sensitized solar cells,^63,64 where n_e is the electron density. In addition, the power laws of J_sc ∝ I₀^0.72 and J_sc ∝ I₀^0.52 have been reported in Pb₅Sb₈S₁₇ and NaSbS₂ QDSSCs, respectively.^65,66 According to the multiple trapping model, J_sc decreases at a lower rate than that of the light intensity I₀ because α < 1. For example, if α = 0.5 and the light intensity I₀ decreases by four times, then J_sc decreases only by two times. The number of photocarriers is reduced under low light intensities, which reduces the photocarrier trapping/de-trapping activity and the occurrence of carrier recombination, leading to more efficient carrier collection and higher efficiency.

Table 4 Photovoltaic performance of AgInS₂/CdSe/ZnSe core/shell QDSSCs under different sun intensities

Sun intensity	J sc (mA cm^−2)	V oc (V)	FF (%)	PCE (%)
1 sun	23.85	0.65	55.26	8.57
0.25 sun	7.34	0.62	61.27	11.15
0.10 sun	3.07	0.59	64.80	11.75

---

3.7 External quantum efficiency (EQE)

EQE is essential for evaluating the efficiency of photoelectron generation. Fig. 8 depicts the EQE spectra and EQE-deduced integrated current density of three QDSSCs: AgInS₂, CdSe, and AgInS₂/CdSe/ZnSe. The AgInS₂ core QDSSCs covered the EQE spectra range of 350–700 nm with a nearly constant EQE value of ∼65% over the range of 350–600 nm. The EQE onset, corresponding to the bandgap E_g,pv, of the AgInS₂ QDSSCs, was 712 nm (E_g,pv = 1.74 eV). The CdSe shell QDSSCs covered the spectra range of 350–720 nm with a nearly constant EQE value of ∼68% over the range of 400–650 nm. The EQE onset was 733 nm (E_g,pv = 1.69 eV). Moreover, the CdSe shell exhibited significantly higher EQE values than that of the AgInS₂ core over the long-wavelength region of 600–700 nm. For the AgInS₂/CdSe core/shell QDSSCs, the EQE spectra exhibited a slightly broader range of 350–725 nm with significantly higher EQE values of 80–84% over the spectral range of 350–625 nm. Meanwhile, the EQE onset decreased slightly to E_g,pv = 1.62 eV (765 nm). The addition of the ZnSe passivation layer improved the EQE response and reduced the bandgap tails, particularly evident beyond 800 nm, accompanied by substantially elevated EQE values ranging from 82% to 85% over the interval of 350–700 nm. Concurrently, the EQE onset is similar to the AgInS₂/CdSe core/shell QDSSCs with E_g,pv = 1.61 eV (770 nm). Two notable results of the EQE spectra are: (a) the core/shell QDSSCs yielded higher EQE values (84%) than that of the core and shell (∼65%), (b) the core/shell QDSSCs exhibited slightly broader spectral range (up to 765 nm) compared to the core (710 nm) and CdSe shell (730 nm). Lastly, some weak EQE signals (EQE < 15%) appeared in the long-wavelength region of λ > 800 nm. These weak EQE signals are attributed to the localized states within the bandgap that arise from defects in QDs, similar to a semiconductor's Urbach tail.

Fig. 8 EQE spectra of AgInS₂, CdSe, and AgInS₂/CdSe/ZnSe QDSSCs and their integrated current density.

The photocurrent generated by a solar cell was calculated by integrating the area under an EQE curve. The integrated J_scs were 14.41, 16.75, and 23.26 mA cm⁻² for the AgInS₂ core, CdSe shell, and AgInS₂/CdSe core/shell, respectively, as shown in the right axis of Fig. 8. The integrated J_sc of 23.26 mA cm⁻² for the AgInS₂/CdSe core/shell is consistent with the J_sc of 23.28 mA cm⁻² obtained from I–V measurements shown in Table 2.

The AgInS₂/CdSe core/shell solar cell exhibited a distinctly higher PCE (8.39%, 1 sun) than that of both the single-layered AgInS₂ core (3.04%) and the CdSe shell cells (3.67%). As Table 1 indicates, the improved PCE arises mainly from J_sc, which is 23.28 mA cm⁻² for AgInS₂/CdSe core/shell, compared to 7.72 mA cm⁻² (AgInS₂) and 12.9 mA cm⁻² (CdSe). The significantly larger improvement in J_sc is attributed to five features of the core/shell structure: (1) spatial separation of electron and holes: in the AgInS₂/CdSe core/shell QD, the photogenerated electrons are located in the CB of AgInS₂ core while holes in the VB of CdSe shell. The spatially separated electrons and holes reduce the overlap of their wavefunctions and, hence, carrier recombination, (2) shell layer as an energy barrier for recombination: the CB of the CdSe shell is higher than the CB of AgInS₂ core. The CdSe CB, acting as an energy barrier, prevents photoelectrons in the CB of AgInS₂ from recombining with holes in the electrolyte, (3) cascade energy structure: the CB and VB of AgInS₂/CdSe core/shell form a cascade energy structure, as shown in Fig. 6(c). Electrons can jump from the CB of the CdSe shell to the CB of the AgInS₂ core, then into the CB of the TiO₂ electron transport layer. Similarly, holes can jump from the VB of the AgInS₂ core and then to the VB of CdSe. Cascade energy structures can facilitate the electron injection and collection in a solar cell, as has been shown in many types of solar cells such as Zn–Cu–In–S QDs^67–70 and dye-sensitized solar cells,⁷¹ (4) broader absorption spectrum: the AgInS₂/CdSe core/shell QD has an absorption onset of 765 nm (1.62 eV), which is slightly broader than the AgInS₂ core (712 nm, 1.74 eV) and CdSe shell (733 nm, 1.69 eV), as revealed in the EQE spectra. According to the literature,⁶¹ the maximum J_sc for a perfect solar absorber material (i.e., EQE = 100%) with an absorption onset of wavelength λ is ∼24.0 mA cm⁻² for λ = 765 nm, 21.0 mA cm⁻² for λ = 733 nm and 19.5 mA cm⁻² for λ = 712 nm. The broader absorption range of the core/shell can produce a J_sc 15% more than that of the core and shell, (5) the decrease in the effective bandgap resulting from the exciplex state represents a distinctive characteristic of the type-II core/shell structure. As illustrated in the band energy diagram shown in Fig. 6(d), the energy bandgap originating from the exciplex state is lower than that of both AgInS₂ (1.76 eV) and CdSe (1.75 eV), thereby broadening the absorption spectrum of the QDs. In other words, the exciplex state may have the potential for an indirect interfacial transition. To summarize, the reduced carrier recombination through spatially separated carriers, improved electron injection through the cascade energy structure, better light harvesting through the broader absorption band, and a possible indirect interfacial transition lead to a J_sc in the core/shell cell twice larger than that of the core and shell.

4 Conclusions

We demonstrated type-II AgInS₂/CdSe core/shell QDSSCs prepared on mesoporous TiO₂ using SILAR. The AgInS₂/CdSe core/shell QDSSCs achieve a PCE of 8.39% under 1 sun and 11.75% under 0.1 sun. The PCE and J_sc of the core/shell cells are nearly three times larger than those of the single-layered AgInS₂ cells. The PCE of the AgInS₂/CdSe cells ranks as one of the top efficiency among all type-II core/shell QDSSCs reported. The cascade energy structure and spatial separation of electron/hole are the primary reasons that lead to the improved performance in the AgInS₂/CdSe core/shell structure.

Author contributions

Siti Utari Rahayu: conceptualization, data curation, formal analysis, investigation, methodology, validation, visualization, and writing – original draft. Yu-Rou Wang: data curation, investigation. Jen-Bin Shi: formal analysis. Ming-Way Lee: conceptualization, formal analysis, funding acquisition, methodology, project administration, resources, supervision, validation, and writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors are grateful for the financial support from the Ministry of Science and Technology (MOST) in Taiwan under grant no. MOST 111-2112-M-005-017. We are grateful to Prof. Ja-an Annie Ho and Dr Thirupatthi Murugan of the Department of Biochemical Science and Technology at National Taiwan University for their support and discussions on the CV measurements.

References

1. S. Gorer and G. Hodes, J. Phys. Chem., 1994, 98, 5338–5346.
2. I. Moreels, K. Lambert, D. De Muynck, F. Vanhaecke, D. Poelman, J. C. Martins, G. Allan and Z. Hens, Chem. Mater., 2007, 19, 6101–6106.
3. R. D. Schaller and V. I. Klimov, Phys. Rev. Lett., 2004, 92, 186601.
4. P. Sudhagar, E. J. Juárez-Pérez, Y. S. Kang and I. Mora-Seró, in Low-cost Nanomaterials, ed. Z. Lin and J. Wang, Springer London, London, 2014, pp. 89–136.
5. G. Larramona, C. Choné, A. Jacob, D. Sakakura, B. Delatouche, D. Péré, X. Cieren, M. Nagino and R. Bayón, Chem. Mater., 2006, 18, 1688–1696.
6. N. Guijarro, T. Lana-Villarreal, I. Mora-Seró, J. Bisquert and R. Gómez, J. Phys. Chem. C, 2009, 113, 4208–4214.
7. Y. Itzhaik, O. Niitsoo, M. Page and G. Hodes, J. Phys. Chem. C, 2009, 113, 4254–4256.
8. V. González-Pedro, C. Sima, G. Marzari, P. P. Boix, S. Giménez, Q. Shen, T. Dittrich and I. Mora-Seró, Phys. Chem. Chem. Phys., 2013, 15, 13835.
9. A. Tubtimtae, K.-L. Wu, H.-Y. Tung, M.-W. Lee and G. J. Wang, Electrochem. Commun., 2010, 12, 1158–1160.
10. H. Zhang, K. Cheng, Y. M. Hou, Z. Fang, Z. X. Pan, W. J. Wu, J. L. Hua and X. H. Zhong, Chem. Commun., 2012, 48, 11235.
11. M. Ye, X. Gao, X. Hong, Q. Liu, C. He, X. Liu and C. Lin, Sustainable Energy Fuels, 2017, 1, 1217–1231.
12. D. Sharma, R. Jha and S. Kumar, Sol. Energy Mater. Sol. Cells, 2016, 155, 294–322.
13. W. Li, Z. Pan and X. Zhong, J. Mater. Chem. A, 2015, 3, 1649–1655.
14. Z. Pan, I. Mora-Seró, Q. Shen, H. Zhang, Y. Li, K. Zhao, J. Wang, X. Zhong and J. Bisquert, J. Am. Chem. Soc., 2014, 136, 9203–9210.
15. Z. Ren, Z. Wang, R. Wang, Z. Pan, X. Gong and X. Zhong, Chem. Mater., 2016, 28, 2323–2330.
16. H. Song, Y. Lin, M. Zhou, H. Rao, Z. Pan and X. Zhong, Angew. Chem., Int. Ed., 2021, 60, 6137–6144.
17. J. Du, R. Singh, I. Fedin, A. S. Fuhr and V. I. Klimov, Nat. Energy, 2020, 5, 409–417.
18. Z. Ren, J. Wang, Z. Pan, K. Zhao, H. Zhang, Y. Li, Y. Zhao, I. Mora-Sero, J. Bisquert and X. Zhong, Chem. Mater., 2015, 27, 8398–8405.
19. Md. A. Hossain, J. R. Jennings, Z. Y. Koh and Q. Wang, ACS Nano, 2011, 5, 3172–3181.
20. Q. Wang, S. Li, J. Qiao, R. Jin, Y. Yu and S. Gao, Sol. Energy Mater. Sol. Cells, 2015, 132, 650–654.
21. J. H. Bang and P. V. Kamat, ACS Nano, 2009, 3, 1467–1476.
22. T.-L. Li, Y.-L. Lee and H. Teng, Energy Environ. Sci., 2012, 5, 5315–5324.
23. J. Yang, J. Wang, K. Zhao, T. Izuishi, Y. Li, Q. Shen and X. Zhong, J. Phys. Chem. C, 2015, 119, 28800–28808.
24. D. C. J. Neo, C. Cheng, S. D. Stranks, S. M. Fairclough, J. S. Kim, A. I. Kirkland, J. M. Smith, H. J. Snaith, H. E. Assender and A. A. R. Watt, Chem. Mater., 2014, 26, 4004–4013.
25. L. Liu, H. Li, Z. Liu and Y.-H. Xie, J. Colloid Interface Sci., 2019, 546, 276–284.
26. Z. Pan, H. Zhang, K. Cheng, Y. Hou, J. Hua and X. Zhong, ACS Nano, 2012, 6, 3982–3991.
27. F. Huang, L. Zhang, Q. Zhang, J. Hou, H. Wang, H. Wang, S. Peng, J. Liu and G. Cao, ACS Appl. Mater. Interfaces, 2016, 8, 34482–34489.
28. J. Wang, I. Mora-Seró, Z. Pan, K. Zhao, H. Zhang, Y. Feng, G. Yang, X. Zhong and J. Bisquert, J. Am. Chem. Soc., 2013, 135, 15913–15922.
29. Z. Ning, H. Tian, C. Yuan, Y. Fu, H. Qin, L. Sun and H. Ågren, Chem. Commun., 2011, 47, 1536–1538.
30. S. Kim, B. Fisher, H.-J. Eisler and M. Bawendi, J. Am. Chem. Soc., 2003, 125, 11466–11467.
31. S. Jiao, Q. Shen, I. Mora-Seró, J. Wang, Z. Pan, K. Zhao, Y. Kuga, X. Zhong and J. Bisquert, ACS Nano, 2015, 9, 908–915.
32. J. Bang, J. Park, J. H. Lee, N. Won, J. Nam, J. Lim, B. Y. Chang, H. J. Lee, B. Chon, J. Shin, J. B. Park, J. H. Choi, K. Cho, S. M. Park, T. Joo and S. Kim, Chem. Mater., 2010, 22, 233–240.
33. J. Luo, H. Wei, F. Li, Q. Huang, D. Li, Y. Luo and Q. Meng, Chem. Commun., 2014, 50, 3464.
34. A. R. Amani-Ghadim, M. Mousavi and F. Bayat, J. Power Sources, 2022, 539, 231624.
35. M. A. Sunil and P. Kodali, AIP Conf. Proc., 2018, 1992(1), 040035.
36. M. Calixto-Rodriguez, H. Martínez, M. E. Calixto, Y. Peña, D. Martínez-Escobar, A. Tiburcio-Silver and A. Sanchez-Juarez, Mater. Sci. Eng. B, 2010, 174, 253–256.
37. J. L. Shay, B. Tell, L. M. Schiavone, H. M. Kasper and F. Thiel, Phys. Rev. B: Condens. Matter Mater. Phys., 1974, 9, 1719–1723.
38. M. L. A. Aguilera, M. Ortega-López, V. M. S. Resendiz, J. A. Hernández and M. A. G. Trujillo, Mater. Sci. Eng. B, 2003, 102, 380–384.
39. J. Liu, S. Chen, Q. Liu, Y. Zhu and Y. Lu, Comput. Mater. Sci., 2014, 91, 159–164.
40. W. Shockley and H. J. Queisser, J. Appl. Phys., 1961, 32, 510–519.
41. C. Cai, L. Zhai, Y. Ma, C. Zou, L. Zhang, Y. Yang and S. Huang, J. Power Sources, 2017, 341, 11–18.
42. H. M. Pathan, B. R. Sankapal, J. D. Desai and C. D. Lokhande, Mater. Chem. Phys., 2003, 78, 11–14.
43. K. B. Chaudhari, N. M. Gosavi, N. G. Deshpande and S. R. Gosavi, J. Sci.: Adv. Mater. Devices, 2016, 1, 476–481.
44. P. Wang, S. M. Zakeeruddin, P. Comte, R. Charvet, R. Humphry-Baker and M. Grätzel, J. Phys. Chem. B, 2003, 107, 14336–14341.
45. N. M. Noer, S. U. Rahayu, K. Sebayang, H. A. Sianturi and M.-W. Lee, J. Phys.: Conf. Ser., 2020, 1542, 012041.
46. K. Sebayang, S. U. Rahayu, M.-W. Lee, S. Gea, H. Ginting and A. Warman, Rasayan J. Chem., 2021, 14, 88–93.
47. E. N. Chiristina, S. U. Rahayu, A. Tubtimtae, J.-B. Shi and M.-W. Lee, RSC Adv., 2022, 12, 31093–31101.
48. S. U. Rahayu, Y.-L. Tai, P. Boon-On, Y.-R. Wang and M.-W. Lee, Opt. Mater., 2023, 145, 114416.
49. S. K. Haram, B. M. Quinn and A. J. Bard, J. Am. Chem. Soc., 2001, 123, 8860–8861.
50. L. Micaroni, F. C. Nart and I. A. Hümmelgen, J. Solid State Electrochem., 2002, 7, 55–59.
51. W. Qin, T. Nagase, Y. Umakoshi and J. A. Szpunar, Philos. Mag. Lett., 2008, 88, 169–179.
52. M. Kuno, Phys. Chem. Chem. Phys., 2008, 10, 620–639.
53. M. Böhmler, Z. Wang, A. Myalitsin, A. Mews and A. Hartschuh, Angew. Chem., Int. Ed., 2011, 50, 11536–11538.
54. L. Zhao, L. Hu and X. Fang, Adv. Funct. Mater., 2012, 22, 1551–1566.
55. E. Rabinovich and G. Hodes, J. Phys. Chem. C, 2013, 117, 1611–1620.
56. S. Lee, K. Lee, W. D. Kim, S. Lee, D. J. Shin and D. C. Lee, J. Phys. Chem. C, 2014, 118, 23627–23634.
57. C. Neela Mohan and V. Renuga, J. Alloys Compd., 2019, 787, 972–981.
58. M. Jagadeeswararao, A. Swarnkar, G. B. Markad and A. Nag, J. Phys. Chem. C, 2016, 120, 19461–19469.
59. X. Ma, A. Mews and T. Kipp, J. Phys. Chem. C, 2013, 117, 16698–16708.
60. S. N. Inamdar, P. P. Ingole and S. K. Haram, ChemPhysChem, 2008, 9, 2574–2579.
61. G. P. Smestad, F. C. Krebs, C. M. Lampert, C. G. Granqvist, K. L. Chopra, X. Mathew and H. Takakura, Sol. Energy Mater. Sol. Cells, 2008, 92, 371–373.
62. K. Seki, K. Marumoto and M. Tachiya, Appl. Phys. Express, 2013, 6, 051603.
63. G. Schlichthörl, N. G. Park and A. J. Frank, J. Phys. Chem. B, 1999, 103, 782–791.
64. G. Franco, J. Gehring, L. M. Peter, E. A. Ponomarev and I. Uhlendorf, J. Phys. Chem. B, 1999, 103, 692–698.
65. P. Boon-on, C.-R. Chiang, R. Rajendran, N. Suriyawong, J.-B. Shi and M.-W. Lee, J. Power Sources, 2021, 485, 229324.
66. W.-C. Sun, S. U. Rahayu and M.-W. Lee, IEEE J. Photovolt., 2018, 8, 1011–1016.
67. Z. Pan, L. Yue, H. Rao, J. Zhang, X. Zhong, Z. Zhu and A. K. -Y. Jen, Adv. Mater., 2019, 31, 1903696.
68. L. Yan, Q. Xue, M. Liu, Z. Zhu, J. Tian, Z. Li, Z. Chen, Z. Chen, H. Yan, H. Yip and Y. Cao, Adv. Mater., 2018, 30, 1802509.
69. J. Ding, J. Duan, C. Guo and Q. Tang, J. Mater. Chem. A, 2018, 6, 21999–22004.
70. P. Liang, C. Liao, C. Chueh, F. Zuo, S. T. Williams, X. Xin, J. Lin and A. K. -Y. Jen, Adv. Mater., 2014, 26, 3748–3754.
71. K. Kakiage, Y. Aoyama, T. Yano, K. Oya, J. Fujisawa and M. Hanaya, Chem. Commun., 2015, 51, 15894–15897.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3se01249b

---