Quantum dots (QDs) are semiconductor nanoparticles whose optical and electrical properties can be precisely tailored by adjusting their size and modifying their surface chemistry^[1−3]. QDs synthesized from binary ionic compound semiconductors such as PbS, CdS, PbSe, CdSe, and InAs have gained significant attention in optoelectronics due to their extraordinary properties^[4], including multi-exciton generation^[5], solution processability^[6], tunable absorption spectrum^[7], and narrow emission line width^[8]. The unique characteristics make QDs highly suitable for a wide range of optoelectronic applications, including light-emitting diodes^{[9, 10]}, photodetectors^{[11, 12]}, lasers^[13], and photovoltaic devices^[14−16]. In a typical lead sulfide colloidal quantum dot solar cell (PbS CQDSC), a several-hundred-nanometer-thick n-type CQD absorber layer (CQD AL) is strategically positioned between an electron transport layer (ETL) and a hole transport layer (HTL). The ETL is typically composed of metal oxides such as zinc oxide, tin oxide, or titanium dioxide, while the HTL consists of p-type CQDs or organic materials^[15]. The limited efficiency of photovoltaic devices based on this structure is often attributed to the interfacial defects at the ETL/AL and AL/HTL junctions. These defects promote non- radiative recombination, leading to carrier loss before extraction^[17]. To overcome these challenges and further enhance the performance of CQDSCs, doping technologies have become a key research focus^{[4, 15, 18−20]}. The traditional method for fabricating p-type CQD HTLs relies on the layer-by-layer (LbL) ligand exchange technique. However, this approach is tedious and prone to introducing defects at the AL/HTL interface, which can degrade device performance. Moreover, commonly used ligand solvents, such as acetonitrile (ACN), can damage the underlying AL, further exacerbating defect formation. Additionally, the high reactivity of 1,2-ethanedithiol (EDT), which is normally used during the LbL ligand exchange process, can lead to undesirable surface interactions, contributing to instability and hindering efficient charge extraction. These factors have been identified as key limitations restricting the open-circuit voltage (V_oc) of CQDSCs.

In recent years, p-type CQD inks synthesized via the solution-phase ligand exchange (SPLE) technique have gained significant attention and are considered promising materials for HTL fabrication^{[1, 15−17]}. Traditional p-type PbS CQD inks have primarily been developed using organic thiol ligands, such as EDT, 3-mercaptopropionic acid (MPA), cysteamine (CTA), and 2-mercaptoethanol (ME), which are commonly employed in the LbL technique. These approaches typically follow either a two-step ligand exchange (LbL technique) or a one-step exchange (SPLE technique), where all ligands are introduced simultaneously in solution^{[1, 21−23]}. Nevertheless, organic thiol ligands often suffer from poor solubility and stability, limiting their practical application in p-type CQD inks.

To address these issues, we proposed a novel technique to synthesize p-type PbS CQDs using only inorganic ligands. By incorporating tin (II) iodide (SnI₂) into the traditional lead halide (PbX₂; X = I, Br) ligand solution via the SPLE technique, we successfully obtained stable p-type CQD ink with inorganic ligands. This formulation exhibits excellent colloidal solubility, overcoming the solubility limitations of short-chain organic ligands and enabling the fabrication of high-quality p-type AL or HTL through a one-step deposition process.

Chemicals: Zinc acetate dihydrate (Zn(Ac)₂·2H₂O, Wako, 99.9%), 2-aminoethanol (Wako, 99%), 2-methoxyethanol (Wako, 99%), magnesium acetate tetrahydrate (Mg(Ac)₂·4H₂O, Wako, 99%), lead oxide (PbO, Wako, 99.5%), oleic acid (OA, Sigma−Aldrich, 90%), 1-octadencence (ODE, Sigma−Aldrich, 90%), hexamethyldisilathaine (TMS₂, Sigma-Aldrich, 99%), cadmium chloride (CdCl₂, Wako, 95%), tetradecylphosphonic acid (TDPA, Sigma−Aldrich), oleylamine (OLA, Sigma−Aldrich, 70%), lead iodide (PbI₂, 99%), lead bromine (PbBr₂, Sigma−Aldrich, 95%), N, N-dimethylfomarmamide super dehydrated (DMF, Wako, 99.5%), ammonium acetate (NH₄Ac, Wako, 95%), 1, 2-ethanedithiol (EDT, Sigma−Aldrich, 98%), tin iodide (SnI₂, stream chemicals, 99%), tin shot (Suzu, Wako), butylamine (Sigma−Aldrich, 99.5%), amylamine (Sigma−Aldrich, 99%), hexylamine (Sigma−Aldrich, 99%), chlorobenzene (Sigma−Aldrich, 99.8%), poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS, Sigma−Aldrich).

Synthesis of ETL solution: The ETL solution was synthesized according to a previous report^[23]. Mg-doped ZnO, referred to as ZMO, was prepared by combining 0.296 g of Zn(Ac)₂·2H₂O, 32.17 mg of Mg(Ac)₂·4H₂O, 100.782 mg of 2-aminoethanol, and 5 mL of 2-methoxyethanol. Then, the mixture was stirred at 65 °C for 12 h.

Synthesis of PbS CQDs: PbS CQDs were synthesized and purified following previous research^[15]. Initially, a mixture containing 6 mmol of PbO, 15 mmol of OA, and 50 mL of ODE was prepared in a 100 mL three-neck flask. The solution underwent degassing while stirring in two stages: first at room temperature for 20 min and then at 100 °C for 60 min. After degassing, the mixture gradually cooled down to 84 °C. At this temperature, a TMS₂ sulfide solution (3 mmol TMS₂ sulfide dissolved in 10 mL of pre-degassed ODE) was rapidly injected. The heating source was removed immediately after injection, and the solution was stirred continuously until the temperature cooled to room temperature. A CdCl₂-TDPA-OLA halide passivation solution was introduced when the temperature of the PbS CQD raw solution reached 75 °C. The solution was purified after it had cooled down to room temperature. 13 mL of the raw solution and 20 mL of acetone were added to each of the two tubes. After thorough mixing, the tubes were centrifuged at 5000 rpm for 3 min to separate the QD precipitates. The process was then repeated twice to precipitate all the raw solution. For each tube, the resulting QD precipitates were then redispersed in 5 mL of toluene. For further purification, 10 mL of acetone and 20 mL of methanol were added to the dispersion, followed by centrifugation at 6000 rpm for 3 min. This step was repeated twice to remove any remaining impurities. Finally, the QDs were redispersed in 5 mL of toluene, re-precipitated using 20 mL of acetone and 16 mL of methanol and centrifuged at 8500 rpm for 6 min. The final precipitates were dried under nitrogen flow and dispersed in octane to prepare a colloidal solution with a concentration of 100 mg/mL for further experiments.

Preparation of n-type PbS CQDs and p-type PbS CQDs: The PbX₂ ligand solution was prepared by dissolving 0.1 M of PbI₂, 0.04 M of PbBr₂, and 0.06 M of NH₄Ac in DMF solution, followed by stirring at room temperature for 8 h. For mixing SnI₂ ligand conditions, SnI₂ was incorporated into PbX₂ ligand solution at controlled molar ratio to (PbI₂ + SnI₂) ranging from 1% to 3% (as SnI₂/(PbI₂ + SnI₂) from 1%−3%).

PbS CQD ink was prepared following a reported SPLE method^[15]. A 12 mL octane solution containing OA-coated PbS CQDs (PbS−OA) at a concentration of 20.84 mg/mL was slowly added dropwise into 20 mL of DMF solution under vigorous stirring. The mixture was stirred for an additional 4 min. The octane phase was removed, and the DMF phase was thoroughly washed three times with 10 mL of octane each time to eliminate any residual impurities. PbS CQDs capped with PbX₂ ligands (PbS–PbX₂), and PbS CQDs capped with PbX₂/SnI₂ ligand (PbS–PbX₂/SnI₂) were extracted from the DMF phase by adding 14 mL of toluene, followed by centrifugation at 8500 rpm for 5 min. The PbS–PbX₂ precipitates were vacuum-dried at room temperature for 45 min. Then, they were redispersed in a mixed-amine solution comprising butylamine, amylamine, and hexylamine in a volume ratio of 10 : 3 : 2 at 295 mg/mL for AL ink. To prevent the dissolution of n-type PbS film and p-type PbS film during device assembling, PbS–PbX₂/SnI₂ solid was redispersed in butylamine and chlorobenzene in a volume ratio of 1 : 9 at 70 mg/mL for HTL ink preparation.

Device fabrication: A sequential ultrasonic technique was used to clean the FTO substrate using soap water, distilled water, ethanol, and acetone. The substrate was treated with UV-ozone plasma to enhance its surface qualities after drying with nitrogen gas. The ETL was formed by spin-coating the ZMO solution onto the FTO substrate and heated on a hot plate at 190 °C for 10 min before annealing in a muffle furnace at 290 °C for 30 min. To deposit AL, PbS–PbX₂ CQD ink was spin-coated onto the ETL at 1200 rpm for 30 s. The AL was then heated at 70 °C for 5 min to improve the film quality. For control devices, a LbL technique was used to fabricate the HTL with two EDT-coated PbS CQDs (PbS–EDT) layers. 20 µL of PbS–OA QD solution (50 mg/mL) was spin-coated at 2500 rpm for 18 s for each layer. The film was then submerged in a 0.02 vol.% EDT/ACN solution for 30 s before spin-coating at 2500 rpm for 10 s. To finish the procedure, three rinse-spin cycles with ACN were used to remove excess chemicals and create a homogenous film. For the target devices, HTL was formed using a novel technique that PbS–PbX₂/SnI₂ ink was spin-coated onto the AL surface in one step at 2500 rpm for 30 s. A 100 nm-thick gold electrode was deposited on HTL via thermal evaporation using a custom shadow mask to fabricate PbS CQDSCs with contact sizes of 0.25 cm².

Characterization: PbS CQD film samples were prepared by spin-coating the PbS CQD ink, and PbS–EDT sample was prepared using LbL technique on a glass substrate. The thickness of the PbS CQD film was determined using a Dektak XT surface profiler. X-ray diffraction (XRD) patterns were obtained by Rigaku-D/max-2500 X-ray diffractometer. The size of PbS CQDs was calculated from XRD peaks according to the Debye−Scherrer equation. X-ray photoelectron spectroscopy (XPS) data were collected using a photoelectron spectrometer (S4 JPS-90MX, JEOL Ltd., Japan). The photoelectron yield spectrum (PYS) was measured using an ionization energy measurement system (BIP-KV205, Bunkoukeiki Co., Ltd.). Kelvin Probe (KP) microscopy was measured by FAC-2 instrument (Riken Keiki, Co., Ltd.). The Hall effect and mobility was measured using a Ecopia HMS-3000 Hall measurement system with an applied current ranging from 1 nA to 20 mA, with a compliance voltage of 12 V. Voltage measurements have an input impedance of 2 × 10⁷ Ohms and can accommodate an input voltage range of ±12 V. The Keithley 2400 source was used to measure current density−voltage (J−V) characteristics under AM 1.5 simulated sunlight (100 mW/cm²) and ambient conditions (temperature: 25 °C, relative humidity: 20%−40%) using the solar simulator PEC-L10 (Peccell Technologies, Inc., Japan). The light intensity for J−V characteristics measurements was calibrated using an amorphous silicon photodiode detector (BS-520BK S/N 834, Bunkoukeiki, Japan) and its photogenerated current. A black mask (made by PHOTO PRECISION CO., LTD., Japan) was used to mask the absorber region, resulting in an aperture of 0.16 cm² with an accuracy of ±0.015 mm. The scanning speed was 10 mV/s, with a latency of 0.2 s. The scanning range is −0.1 to 0.7 V (forward) and 0.7 to −0.1 V (reverse). To measure device performance, bias was fixed at the maximum power output point determined from J−V curves.

We developed a reliable and efficient strategy to produce p-type PbS CQD inks using the SPLE method by incorporating SnI₂ into a PbX₂ ligand solution. As the ratio of SnI₂ to (PbI₂ + SnI₂) increases from 1% to 3%, the PbS CQDs capped with PbX₂ and SnI₂ ligands (denoted as PbS–PbX₂/SnI₂ 1%−3% in this work) successfully transition from n-type to p-type while maintaining excellent solubility in a 1 : 9 (v/v) butylamine to chlorobenzene solvent mixture. Photos of PbS–PbX₂ CQD inks and PbS–PbX₂/SnI₂ (1%−3%) inks are shown in Fig. S1.

The successful realization of p-type PbS CQDs is preliminarily verified by Hall effect measurements, which directly assess the dominant charge carrier type based on the Hall coefficient (R_H). Fig. 1(a) presents the R_H values of PbS CQD ink films with varying SnI₂ concentrations. The samples without SnI₂ (0%) and with 1% SnI₂ exhibit n-type characteristics, showing negative R_H values of −8.14 × 10⁶ m³/C and −5.13 × 10⁵ m³/C, respectively. In contrast, samples with 2% and 3% SnI₂ display p-type characteristics, with positive R_H values of 7.01 × 10⁵ m³/C and 2.1 × 10⁷ m³/C, respectively. These results show successful transition of PbS CQDs from n-type to p-type.

We speculate that the transition from n-type to p-type is a gradual process, where the n-type characteristics progressively weaken as the mixing ratio of SnI₂ increases, eventually leading to the enhancement of p-type properties. Interestingly, however, the Hall effect results show different trends depending on the change in R_H (Fig. 1(a)), which is considered to be inversely proportional to the carrier concentration.

To further investigate this behavior, we measured energy levels of PbS QD films with and without SnI₂ mixing to confirm the modifications in energy level structure, as shown in Fig. 1(b). With the increasing amount of SnI₂, the Fermi level (E_F) position of the films remains largely unchanged, while the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) positions shift upward as a whole, resulting in stronger p-type characteristics. This result supports the assumption of a continuous transition.

There are several possible explanations for these seemingly contradictory results. The relationship between R_H and carrier concentration is primarily applicable to bulk semiconductor materials rather than QD semiconductor materials due to the quantum confinement effect and surface effects in QDs. Therefore, the R_H values obtained from the Hall effect measurements should be regarded as a reference to indicate the n-type or p-type characteristics rather than an absolute value of carrier concentration. Moreover, the mixed SnI₂ ligands may play the following roles: (1) passivate defect states; (2) altering the surface dipole state of QDs and shifting the band-edge positions of the valence band and conduction band; and (3) inducing a redox reaction of Sn, which changes the type and concentration of doped carriers.

In general, the transition from n-type to p-type is a complex process resulting from the combined effects of multiple reactions. As the amount of SnI₂ increases, the dominant reaction mechanism may shift. We will explore this mechanism in greater depth in future work. In any case, both the Hall effect measurement results and the changes in the energy level structure confirm that a controlled amount of SnI₂ can effectively convert PbS CQDs from n-type to p-type, demonstrating the flexibility and reliability of our approach.

To investigate the mechanism behind the n-type to p-type transition and the role of SnI₂ incorporation, we first conducted XPS measurements to confirm the presence of Sn and analyze the chemical environment of PbS–PbX₂ films (control) and PbS–PbX₂/SnI₂ 1%−3% films (target). All the XPS spectra were calibrated using the C 1s peak positioned at 284.8 eV as a reference for charge correction^[24].

As depicted in Fig. 1(c), all target samples exhibit distinct Sn signals, confirming the presence of Sn in the surface environment of PbS CQDs. Notably, the Sn 3d peaks shift toward higher binding energies compared to reference values from the NIST database: (1) Sn²⁺ usually appears around 485.5−486.3 eV (Sn 3d_5/2) and 494.0−495.0 eV (Sn 3d_3/2); (2) Sn⁴⁺ is observed at higher binding energies, approximately 486.8−487.5 eV (Sn 3d_5/2) and 495.3−496.5 eV (Sn 3d_3/2). This shift suggests the occurrence of a redox reaction. Additionally, the increasing spatial ratio of Sn²⁺ with higher SnI₂ concentration (as shown in Fig. S2) indicates that Sn⁴⁺ undergoes reduction (Table S1). Since SnI₂ was introduced into the PbX₂ ligand solution, two possible redox pathways are considered: (1) Sn²⁺ binds to PbS QDs and oxidized to Sn⁴⁺ on the QD surface; (2) Sn²⁺ oxidizes to Sn⁴⁺ in the ligand solution before binding to PbS QDs, then Sn⁴⁺ is reduced back to Sn²⁺ upon adsorption. As shown in Fig. 1(d), the Pb 4f signal shifts toward higher binding energy with increasing SnI₂ concentration, indicating electron transfer from Pb atoms to the surface ligand^[25]. The investigation of increasing spatial ratio of Sn²⁺ and the shift of Pb 4f signal supports the second proposed pathway, in which Sn⁴⁺ acts as a strong electron acceptor, extracting electrons from Pb atoms on the QD surface and subsequently reducing itself from Sn⁴⁺ to Sn²⁺. The resulting charge redistribution facilitates the reduction of Sn⁴⁺ to Sn²⁺ and induces hole accumulation, ultimately transforming PbS CQDs from n-type to p-type.

Our findings highlight the critical role of Sn⁴⁺ in the preparation of p-type PbS ink via SnI₂ incorporation into the PbX₂ ligand solution. To further verify this mechanism, we conducted supplementary experiments by substituting SnBr₂, SnCl₂, and SnI₄ for SnI₂ under the 3% mixing condition. All the samples exhibited p-type characteristics, as shown in Fig. S3. Regarding the oxidation process of Sn²⁺ to Sn⁴⁺, we propose that it occurs under ambient conditions through the reaction 2SnI₂ + O₂ → SnO₂ + SnI₄, forming Sn⁴⁺ and SnO₂ precipitates in ligand solution^[26], as shown in Fig. S4. Notwithstanding, we observed that as the SnI₂ concentration increased, the ligand exchange process became increasingly difficult due to excessive SnO₂ precipitation, which hindered the ligand exchange reaction. Among the tested samples, SnI₄ exhibited the best performance. Considering factors such as cost-effectiveness and stability, we opted for an optimal amount of SnI₂ in this study.

In summary, our results demonstrate that incorporating SnI₂ into PbX₂ ligand solution enhances electron withdrawal from Pb atoms to the surface ligand, effectively tuning the electronic properties of PbS CQDs. To further explore the effect of SnI₂ incorporation, we analyzed elemental ratios from the XPS results, summarized in Table S2. The I/S ratio increased from 0.686 (control) to 0.763 (target), and the Br/S ratio increased from 0.217 (control) to 0.269 (target) after SnI₂ incorporation. These data suggest that increasing the SnI₂ concentration during ligand exchange improves surface passivation on PbS CQDs (Fig. S5).

To further study the mechanism and the effect of SnI₂ incorporation in the PbX₂ solution on p-type PbS QD ink preparation, XRD spectroscopy was employed to analyze the crystal structure and estimate the crystallite size. The XRD patterns in Fig. 2 confirm that all diffraction peaks correspond to the cubic phase of PbS (JCPDS #005-0592). The crystallite size of PbS CQDs was determined using the strongest (200) peak, calculated based on the Debye−Scherrer Eq. (1)^[21]:

$D=\frac{K\lambda }{\beta \cos \theta }\mathrm{,}$where K is the Scherrer constant (0.94), λ is the X-ray wavelength (0.15406 nm), β is the full width at half maximum (FWHM) in radians, and θ is the peak position in radians obtained from XRD measurements. Using Rigaku software and Eq. (1), the average crystallite size of PbS CQDs was estimated to be approximately 3.05 ± 0.07 nm for all of the samples.

After confirming the p-type characteristics of PbS CQD inks, we investigated their impact as HTL on the performance of PbS CQDSCs. The schematic structure of the control device (PbS–EDT HTL) and target device (p-type PbS ink HTL) is shown in Fig. S6. Fig. S7 presents the J−V characteristics of champion devices based on PbS–PbX₂/SnI₂ (1%−4%) HTLs in their fresh state. The devices utilizing PbS–PbX₂/SnI₂ 1%, 2%, and 4% SnI₂ HTLs exhibited relatively low efficiency due to two primary factors: (1) an insufficient energy level offset between the HOMO and LUMO of the AL and HTL, leading to increased electron−hole recombination at the AL/HTL interface (Fig. 1(b))^[15]; and (2) an excessive concentration of SnI₂ in the ligand solution, which disrupts the ion equilibrium and surpasses the solubility limit, resulting in precipitation (Fig. S4)^{[27, 28]}. This precipitation is particularly evident in the PbS–PbX₂/SnI₂ 4% sample, where a yellow muddy layer formed during ligand exchange process. Consequently, the optimal device performance was achieved with PbS–PbX₂/SnI₂ 3% as the HTL.

To further assess the effectiveness of PbS–PbX₂/SnI₂ 3% HTL (target), we compared it with conventional PbS–EDT HTLs fabricated via the LbL technique, which served as control devices. Fig. 3 displays the J−V characteristics of both champion devices in their fresh state and after four weeks storage, respectively. The control device achieved an initial power conversion efficiency (PCE) of 10.54%, with a short-circuit current density (J_sc) of 29.44 mA/cm², a V_oc of 0.63 V, and a fill factor (FF) of 0.53 (Fig. 3(a)). However, after four weeks of storage, its PCE declined to 9.83%. In contrast, the target device based on PbS–PbX₂/SnI₂ 3% HTLs initially exhibited a PCE of 10.37%, with a J_sc of 25.15 mA/cm², a V_oc of 0.64 V, and a FF of 0.61. Interestingly, after four weeks, its performance improved slightly to 10.93%, with a J_sc of 29.57 mA/cm², a V_oc of 0.61 V, and a FF of 0.58 (Fig. 3(b)). These results highlight the improved stability and effectiveness of the PbS–PbX₂/SnI₂ 3% HTL, demonstrating its potential for enhancing both efficiency and long-term performance in PbS CQDSCs.

A detailed statistical comparison of device performance between the control and target devices is summarized in Table S3 for the fresh state, after 1-week storage, after 4-weeks storage, after 6-weeks storage, and after 8-weeks storage, respectively. The target device exhibits higher V_oc and FF compared to the control device, as interfacial defects at the AL/HTL interface are effectively mitigated through the one-step ink deposition method. This approach was employed to deposit the PbS–PbX₂/SnI₂ 3% HTL for the target device. The one-step ink deposition technique enables direct and stable HTL formation, significantly reducing interface defects and enhancing overall device stability. As a result, the target device demonstrated a slower degradation in V_oc and FF over time, demonstrating superior operational stability (Fig. 4). These findings highlight the advantages of one-step ink deposition for PbS CQD inks with inorganic ligands as HTLs, providing a promising strategy for the development of stable and efficient CQDSCs.

In summary, we proposed a SPLE method to produce p-type PbS CQD inks using only inorganic ligands for the first time. SnI₂ was introduced into the PbX₂ ligand solution, then Sn²⁺ was oxidized to Sn⁴⁺, which would act as an acceptor on the PbS CQD surface followed by the reduction of Sn⁴⁺ back to Sn²⁺ upon adsorption on QD surface. The resulting electron-withdrawing ligands contribute to stabilizing the QD surface. By adjusting the SnI₂ concentration, we successfully tuned PbS QDs from n-type to p-type, as confirmed by the R_H shift from negative value (n-type) to positive value (p-type). After confirming the p-type characteristics of PbS CQD inks, we evaluated their performance as HTLs in PbS CQDSCs by varying the SnI₂ concentration. The one-step ink deposition technique enables a direct and stable HTL fabrication process, effectively overcoming the limitations of conventional LbL ligand exchange methods.

Compared to control devices using PbS–EDT as HTLs, the target device incorporating PbS–PbX₂/SnI₂ 3% (inorganic p-type PbS CQD inks) exhibits reduced defects at the AL/HTL interface and improved stability, achieving an initial PCE of 10.37% and further increasing to 10.93% after 4 weeks storage. In contrast, the control device showed a decline in performance, decreasing from an initial PCE of 10.54% to 9.83% over the same period.

The innovative p-type CQD ink fabrication approach based on SPLE offers simplicity (one-step process) and flexibility (adjustable HOMO and LUMO position by varying SnI₂ concentration), representing a significant advancement in ligand exchange processes. Beyond its application as a HTL, we are also exploring the potential of p-type CQD ink as an AL for CQDSCs. The proposed strategy provides a promising pathway for the development of efficient and high-performance CQD-based optoelectronic devices.