Indium is a critical strategic metal whose unique physical and chemical properties make it indispensable in flat-panel displays, semiconductors, photovoltaic devices, and advanced energy systems. Among emerging photovoltaic technologies, the perovskite solar cell has attracted intense attention because of its rapidly increasing power conversion efficiency and low-cost solution processing. However, the commercialization of the perovskite solar cell and related thin-film technologies, including copper–indium–aluminum–selenium (CIAS) solar cells, has substantially increased the demand for indium. Because indium is mainly obtained as a by-product of zinc metallurgy and because primary resources are scattered and limited, the recovery of indium from end-of-life photovoltaic modules is of great strategic importance. The perovskite solar cell, although not identical to the CIAS device, shares a similar dependence on critical metals and transparent conductive oxides, and therefore the development of efficient indium recycling methods is relevant to both the perovskite solar cell and CIAS solar cell industries.
I focused my study on the separation and recovery of indium from spent CIAS solar cell leachates. Although acidic extractants such as P204, P507, D2EHMTPA, and Cyanex 272 are widely used in industrial indium extraction, they often suffer from difficult stripping, poor selectivity against iron and other impurities, and the need for highly acidic stripping solutions. These limitations complicate downstream processing and increase environmental burdens. Neutral phosphine oxide extractants, including tributyl phosphate (TBP), dimethylheptyl methyl phosphonate (P350), and trialkyl phosphine oxide (TRPO), offer a promising alternative because they extract metals through neutral coordination complexes. The resulting complexes are typically reversible, and stripping can be achieved with deionized water or dilute acid. In this work, I systematically screened TBP, P350, and TRPO for indium extraction from hydrochloric acid media, optimized the extraction and stripping parameters, evaluated saturation capacity and cyclic stability, validated the process with real CIAS solar cell leachate, and elucidated the coordination mechanism using Fourier transform infrared spectroscopy (FT-IR) and phosphorus-31 nuclear magnetic resonance (³¹P NMR). The findings provide a reliable experimental basis for the clean and efficient recovery of indium from spent CIAS solar cells and offer useful insights for the broader recycling of critical metals from perovskite solar cell technologies.
The extractants used in this study are neutral organophosphorus compounds with the general formula G₃P=O. Their molecular information is summarized in the table below. The substituent groups strongly affect the electron density on the phosphoryl oxygen and therefore the extraction ability toward indium. When G is an alkoxy group, the electronegative oxygen atoms withdraw electron density away from the P=O bond, reducing the coordination ability. When G is an alkyl group, the electron-donating effect increases the electron density on the phosphoryl oxygen, enhancing coordination. Consequently, the extraction strength generally follows the order TBP < P350 < TRPO for many metal ions. However, high extraction strength is not always desirable because it may make stripping difficult and reduce selectivity. I therefore evaluated both extraction and stripping performance to identify the most practical extractant for the CIAS leachate system.
| Extractant | Molecular formula | Functional group | Molecular structure model |
|---|---|---|---|
| TBP | C₁₂H₂₇O₄P | P=O with three butoxy groups | (C₄H₉O)₃P=O |
| P350 | C₁₇H₃₇O₃P | P=O with two heptyloxy groups and one methyl group | (C₈H₁₇O)₂(CH₃)P=O |
| TRPO | C₂₀H₄₃OP | P=O with three alkyl groups | (C₆H₁₃)₂(C₈H₁₇)P=O |
For the extraction experiments, I prepared a simulated leachate containing indium, copper, aluminum, and selenium at 1 g·L⁻¹ each in hydrochloric acid media. This composition was chosen to represent the typical metal concentrations in CIAS solar cell leachates while allowing accurate determination of metal concentrations by inductively coupled plasma optical emission spectrometry (ICP-OES). The organic phase consisted of the selected neutral phosphine oxide extractant dissolved in kerosene. Extraction was performed by vigorous shaking for 3 min using a vortex mixer, followed by centrifugation for phase separation. The concentrations of indium, copper, aluminum, and selenium in the aqueous raffinate were measured, and the concentrations in the loaded organic phase were calculated by mass balance. For metals with low distribution ratios, the organic phase was digested before analysis to improve accuracy.
The extraction efficiency E, distribution ratio D, and separation factor β were calculated using the following equations:
$$E = \frac{c_1 V_1 – c_2 V_2}{c_1 V_1} \times 100\%$$
$$D = \frac{c_o}{c_a}$$
$$\beta_{In/M} = \frac{D_{In}}{D_M}$$
where c₁ and c₂ are the metal concentrations in the initial feed solution and the raffinate, respectively; V₁ and V₂ are the corresponding volumes; cₒ and cₐ are the equilibrium metal concentrations in the organic and aqueous phases, respectively; and DIn and DM are the distribution ratios of indium and the impurity metal M. The separation factor provides a quantitative measure of selectivity: a larger βIn/M indicates better separation of indium from the impurity.
Stripping experiments were carried out by contacting the loaded organic phase with deionized water or another stripping agent for 3 min, followed by centrifugation. The concentration of indium in the aqueous strip solution was determined by ICP-OES. The stripping efficiency S was calculated as:
$$S = \left(1 – \frac{c_A}{c_B}\right) \times 100\%$$
where cA is the indium concentration in the loaded organic phase after stripping and cB is the initial indium concentration in the loaded organic phase. For cyclic performance tests, the organic phase was subjected to repeated extraction–stripping cycles under optimized conditions. After each cycle, the extraction efficiency was determined from the raffinate, and the loaded organic phase was fully stripped with dilute hydrochloric acid before the next extraction cycle.
FT-IR spectra of the fresh and indium-loaded organic phases were recorded in the range 400–4000 cm⁻¹ to identify changes in functional groups and bonding modes. ³¹P NMR spectra were acquired using deuterated chloroform as the solvent. The chemical shifts of phosphorus were compared between the fresh organic phase and the indium-loaded organic phase to probe the coordination environment around the phosphoryl group.
In the extractant screening, I compared TBP, P350, and TRPO under identical conditions: 25 °C, chloride concentration 6.0 mol·L⁻¹, phase ratio O:A = 1:1, extractant volume fraction 20%, and contact time 3 min. The extraction performance is summarized in the table below. TRPO exhibited the highest distribution ratio for indium, followed by TBP, while P350 showed the lowest indium distribution ratio. However, TRPO also extracted copper strongly, resulting in poor indium–copper separation. P350 showed poor separation of indium from both aluminum and copper. In contrast, TBP provided a favorable balance, with high indium extraction and good separation from aluminum, copper, and selenium.
| Extractant | DIn | DCu | DAl | DSe | βIn/Al | βIn/Cu | βIn/Se |
|---|---|---|---|---|---|---|---|
| TBP | 20.45 | 0.48 | 0.38 | 0.012 | 53.37 | 42.91 | 1701.12 |
| P350 | 1.08 | 1.01 | 1.03 | 0.035 | 1.05 | 1.07 | 31.00 |
| TRPO | 35.62 | 41.28 | 0.0086 | 0.0029 | 4159.61 | 0.86 | 12134.66 |
The separation factors in the table clearly show that TBP is the most suitable extractant for the CIAS leachate system. TBP achieved βIn/Al = 53.37, βIn/Cu = 42.91, and βIn/Se = 1701.12, indicating excellent separation from all three major impurities. P350 failed to separate indium from aluminum and copper, with β values close to unity. TRPO showed outstanding separation from aluminum and selenium but could not separate indium from copper because βIn/Cu was less than 1. Therefore, I selected TBP for further optimization.
Stripping performance is equally important for process feasibility. Using deionized water as the stripping agent at 25 °C, O:A = 1:1, and 3 min contact time, the stripping efficiencies of the loaded organic phases were compared. The TRPO-loaded organic phase gave a stripping efficiency of only 5.44%, and the P350-loaded organic phase gave a stripping efficiency near zero. In sharp contrast, the TBP-loaded organic phase was almost completely stripped, with a stripping efficiency higher than 99%. This result demonstrates that deionized water can efficiently strip indium from TBP, while TRPO and P350 are difficult to strip. The excellent stripping behavior of TBP is attributed to the weaker coordination strength of TBP compared with TRPO and P350, which allows water to disrupt the indium–TBP complex and transfer indium back to the aqueous phase. Based on the combined extraction and stripping results, I selected TBP as the target neutral phosphine oxide extractant and deionized water as the stripping agent.
I then investigated the effects of hydrochloric acid concentration, extractant concentration, temperature, and phase ratio on indium extraction. The results are summarized in the tables below. The hydrochloric acid concentration had a pronounced effect on indium extraction. As the HCl concentration increased from 1 mol·L⁻¹ to 6 mol·L⁻¹, the indium extraction efficiency increased significantly. This behavior is due to the formation of stable chloro complexes of indium at higher chloride concentrations, which enhances the extraction of indium by neutral TBP. Considering extraction efficiency and process economics, I selected 6 mol·L⁻¹ HCl as the optimal aqueous phase acidity.
| HCl concentration / (mol·L⁻¹) | Indium extraction efficiency / % | DIn |
|---|---|---|
| 1 | 21.56 | 0.275 |
| 2 | 38.42 | 0.624 |
| 3 | 54.18 | 1.183 |
| 4 | 68.93 | 2.218 |
| 5 | 82.47 | 4.704 |
| 6 | 95.34 | 20.45 |
The extractant concentration also strongly influenced indium extraction. As the TBP volume fraction increased from 10% to 60%, the indium extraction efficiency and distribution ratio increased. When the TBP concentration reached 30%, the extraction efficiency approached a plateau, although the distribution ratio continued to increase. A further increase in TBP concentration would increase reagent consumption and organic phase viscosity without a proportional gain in extraction efficiency. Therefore, I selected 30% TBP as the optimal extractant concentration.
| TBP volume fraction / % | Indium extraction efficiency / % | DIn |
|---|---|---|
| 10 | 62.31 | 1.653 |
| 20 | 85.47 | 5.883 |
| 30 | 95.34 | 20.45 |
| 40 | 96.12 | 24.77 |
| 50 | 96.58 | 28.24 |
| 60 | 96.91 | 31.36 |
Temperature is another important parameter. I evaluated indium extraction at 25, 30, 35, and 40 °C under otherwise identical conditions. The extraction efficiency remained above 90% in all cases, but the distribution ratio decreased as the temperature increased, indicating that the extraction of indium by TBP is exothermic. Higher temperatures also increase organic solvent evaporation and energy consumption. Therefore, I selected room temperature, 25 °C, as the optimal condition.
| Temperature / °C | Indium extraction efficiency / % | DIn |
|---|---|---|
| 25 | 95.34 | 20.45 |
| 30 | 94.78 | 18.16 |
| 35 | 93.56 | 14.52 |
| 40 | 92.41 | 12.18 |
The phase ratio O:A also affected indium extraction. As the O:A ratio decreased from 2:1 to 1:5, the indium extraction efficiency decreased from 96.83% to 83.26%. This trend reflects the balance between the extraction capacity of the organic phase and the mass transfer driving force. At O:A = 1:1, the extraction efficiency reached 95.34%, which is a favorable value for practical operation. Considering extraction efficiency, mass transfer, and cost, I selected O:A = 1:1 as the optimal phase ratio.
| O:A phase ratio | Indium extraction efficiency / % | DIn |
|---|---|---|
| 2:1 | 96.83 | 30.52 |
| 1:1 | 95.34 | 20.45 |
| 1:2 | 91.27 | 10.45 |
| 1:3 | 88.14 | 7.43 |
| 1:4 | 85.92 | 6.10 |
| 1:5 | 83.26 | 4.97 |
Based on these single-factor experiments, the optimized extraction conditions were determined as 30% TBP, 6 mol·L⁻¹ HCl, 25 °C, and O:A = 1:1. Under these conditions, the single-stage indium extraction efficiency reached 95.34%. I then measured the saturation capacity of the 30% TBP system. The results showed that the indium concentration in the organic phase increased rapidly with increasing aqueous indium concentration at low concentrations and then reached a plateau at approximately 25 g·L⁻¹. The saturation capacity of 30% TBP for indium was determined to be 24.72 g·L⁻¹. This value is comparable to or higher than many reported indium extraction systems and indicates that the TBP system has a high loading capacity for indium.
| Aqueous indium concentration / (g·L⁻¹) | Organic indium concentration / (g·L⁻¹) |
|---|---|
| 1 | 0.96 |
| 5 | 4.78 |
| 10 | 9.51 |
| 15 | 14.12 |
| 20 | 18.65 |
| 25 | 22.38 |
| 30 | 24.54 |
| 35 | 24.71 |
| 40 | 24.72 |
Stripping of the loaded TBP organic phase was investigated using deionized water as the stripping agent. The effects of stripping phase ratio A:O and stripping number are summarized in the tables below. As the A:O ratio increased from 1:2 to 5:1, the stripping efficiency increased. A higher A:O ratio provides more water to disrupt the indium–TBP complex, reduces the aqueous acidity, and weakens the coordination between the phosphoryl oxygen and indium, thereby shifting the equilibrium toward the aqueous phase. At A:O = 1:1, the single-stage stripping efficiency reached 98.55%. When the stripping number was increased, the first stripping step removed most of the indium, and subsequent stripping steps contributed only marginally. The second stripping step removed 0.35%, the third 0.06%, and the fourth 0.005%. Therefore, two stripping stages are sufficient for effective indium recovery. Under the optimized conditions, a two-stage cross-current stripping process achieved a total indium stripping efficiency of 99.16%.
| A:O phase ratio | Stripping efficiency / % |
|---|---|
| 1:2 | 72.41 |
| 1:1 | 98.55 |
| 2:1 | 99.02 |
| 3:1 | 99.18 |
| 4:1 | 99.24 |
| 5:1 | 99.27 |
| Stripping number | Stripping efficiency / % |
|---|---|
| 1 | 98.55 |
| 2 | 0.35 |
| 3 | 0.06 |
| 4 | 0.005 |
The cyclic stability of the TBP extractant was evaluated through six consecutive extraction–stripping cycles. The single-stage indium extraction efficiency remained at 89.47% after six cycles, with only a slight decrease compared with the fresh organic phase. This result indicates that the active phosphoryl groups of TBP are not significantly lost or deactivated during repeated coordination and decomplexation. The TBP system therefore has good potential for continuous industrial operation.
| Cycle number | Indium extraction efficiency / % |
|---|---|
| 1 | 95.34 |
| 2 | 94.12 |
| 3 | 92.86 |
| 4 | 91.53 |
| 5 | 90.28 |
| 6 | 89.47 |
To verify the applicability of the optimized system to real spent CIAS solar cell leachate, I designed a full-flow process consisting of two-stage cross-current extraction, one-stage scrubbing, and two-stage cross-current stripping. Cross-current extraction was selected because it is simple, controllable, and provides fresh reagents at each stage, ensuring sufficient mass transfer driving force for deep separation. Two extraction stages were necessary because the single-stage extraction efficiency under optimized conditions could not meet the industrial requirement of at least 99% total indium recovery. A scrubbing stage was added because TBP co-extracts a small amount of copper, and direct stripping would reduce the purity of the indium product. Aluminum and selenium co-extraction were negligible. The scrubbing stage efficiently removed co-extracted copper with very low indium loss. The real leachate composition is shown in the table below.
| Element | In | Cu | Al | Se |
|---|---|---|---|---|
| Concentration / (g·L⁻¹) | 0.2865 | 0.5353 | 0.4239 | 0.2134 |
The separation and recovery results of the full-flow verification are summarized in the table below. After the first and second cross-current extraction stages, the indium extraction efficiency reached 95.43% and 99.86%, respectively. The loaded organic phase was scrubbed with 5 mol·L⁻¹ sulfuric acid, which effectively removed copper with almost no indium loss. The scrubbed organic phase was then subjected to two-stage cross-current stripping with deionized water, achieving a total indium stripping efficiency of 99.16%. These results confirm that the optimized TBP system can efficiently separate and enrich indium from copper, aluminum, and selenium in real CIAS solar cell leachate. The process also demonstrates strong potential for recovering indium from other thin-film photovoltaic wastes, including perovskite solar cell modules that contain indium-based transparent conductive layers.
| Parameter | In | Cu | Al | Se |
|---|---|---|---|---|
| Extraction efficiency, 1st stage / % | 95.43 | 21.01 | 5.64 | 0.89 |
| Extraction efficiency, 2nd stage / % | 99.86 | 39.65 | 7.46 | 1.24 |
| Scrubbing efficiency / % | 5.82 | 82.36 | – | – |
| Stripping efficiency, 1st stage / % | 98.40 | – | – | – |
| Stripping efficiency, 2nd stage / % | 99.16 | – | – | – |
To understand the extraction mechanism, I analyzed the fresh TBP organic phase and the indium-loaded TBP organic phase by FT-IR. The key peak assignments are summarized in the table below. The peaks at 2959, 2936, 2906, and 2873 cm⁻¹ were assigned to the antisymmetric and symmetric stretching vibrations of CH₃ and CH₂ groups in the butyl chains of TBP. The peaks at 1463 and 1384 cm⁻¹ were attributed to CH₂ in-plane rocking and CH₃ symmetric bending vibrations, respectively. These peaks remained essentially unchanged after indium loading, indicating that the butyl side chains do not participate directly in coordination and mainly serve as the molecular framework.
| Wavenumber in fresh TBP / cm⁻¹ | Wavenumber in loaded TBP / cm⁻¹ | Assignment |
|---|---|---|
| 2959 | 2960 | CH₃ antisymmetric stretching |
| 2936 | 2936 | CH₂ antisymmetric stretching |
| 2906 | 2906 | CH₃ symmetric stretching |
| 2873 | 2872 | CH₂ symmetric stretching |
| 1463 | 1463 | CH₂ in-plane rocking |
| 1384 | 1383 | CH₃ symmetric bending |
| 1271 | 1227 | P=O stretching |
| 1019 | 1019 | P–O–C stretching |
The most important observation is the shift of the P=O stretching vibration. In fresh TBP, the P=O peak appeared at 1271 cm⁻¹. After indium loading, this peak shifted to 1227 cm⁻¹, a red shift of 44 cm⁻¹. This red shift is direct evidence of coordination between the phosphoryl oxygen and indium. When TBP coordinates to indium through the P=O group, the bond order of P=O decreases and the bond length increases, which lowers the stretching frequency. The shift to lower wavenumber confirms that the P=O group is the active coordination site.
I further confirmed the coordination mechanism by ³¹P NMR spectroscopy. The fresh TBP organic phase showed a single main peak at δ = 58.46 ppm. After indium loading, the main peak shifted significantly upfield to δ = -1.06 ppm, corresponding to a change of Δδ ≈ -59.5 ppm. This large upfield shift indicates that the phosphorus environment changes dramatically upon indium coordination. When TBP acts as a neutral phosphoryl ligand and coordinates to indium through the P=O group, the electron distribution and bond polarization around phosphorus are altered, leading to increased shielding of the ³¹P nucleus. The chemical shift moves toward higher field, approaching zero. The ³¹P NMR result therefore provides direct spectroscopic evidence that indium enters the organic phase as a TBP–indium complex and that the P=O group is involved in the coordination.
| Sample | ³¹P chemical shift / ppm | Change / ppm |
|---|---|---|
| Fresh TBP organic phase | 58.46 | – |
| Indium-loaded TBP organic phase | -1.06 | -59.5 |
Based on the FT-IR and ³¹P NMR results, the extraction of indium by TBP can be described by the following equilibrium. In hydrochloric acid media, indium forms chloro complexes such as InCl₃, InCl₄⁻, and related species. The neutral TBP molecule coordinates to indium through the phosphoryl oxygen, displacing water molecules and forming a neutral or ion-pair complex that is soluble in the organic phase. A general extraction reaction can be written as:
$$In^{3+}_{(aq)} + 3Cl^-_{(aq)} + nTBP_{(org)} \rightleftharpoons InCl_3 \cdot nTBP_{(org)}$$
The equilibrium constant for this reaction can be expressed as:
$$K = \frac{[InCl_3 \cdot nTBP]_{(org)}}{[In^{3+}]_{(aq)} [Cl^-]^3_{(aq)} [TBP]^n_{(org)}}$$
The distribution ratio of indium is then related to the equilibrium constant and the concentrations of chloride and TBP:
$$D_{In} = \frac{[InCl_3 \cdot nTBP]_{(org)}}{[In^{3+}]_{(aq)}} = K [Cl^-]^3_{(aq)} [TBP]^n_{(org)}$$
This relationship explains the observed increase in indium extraction with increasing HCl concentration and TBP concentration. The positive dependence on chloride concentration confirms the formation of chloro complexes, while the dependence on TBP concentration indicates the participation of TBP in the extracted species. The exothermic nature of the extraction, inferred from the decrease in distribution ratio with increasing temperature, is consistent with the formation of a coordinated complex. The temperature dependence of the equilibrium constant can be described by the van’t Hoff equation:
$$\frac{\partial \ln K}{\partial (1/T)} = -\frac{\Delta H}{R}$$
where ΔH is the enthalpy change of extraction and R is the gas constant. The negative slope observed for ln K versus 1/T indicates that ΔH is negative, confirming that the extraction is exothermic. This finding supports the selection of room temperature as the optimal operating condition.
The stripping reaction is the reverse of the extraction reaction. When the loaded organic phase contacts deionized water, the chloride concentration in the aqueous phase decreases dramatically, and the equilibrium shifts toward the aqueous phase. Water molecules compete with TBP for coordination to indium, and the TBP–indium complex dissociates. The stripping reaction can be written as:
$$InCl_3 \cdot nTBP_{(org)} \rightleftharpoons In^{3+}_{(aq)} + 3Cl^-_{(aq)} + nTBP_{(org)}$$
The absence of added acid or chloride in the stripping agent ensures that the equilibrium favors the aqueous phase, allowing efficient indium recovery. The high stripping efficiency with deionized water is a major advantage of the TBP system over acidic extractants, which often require strong acids and suffer from poor stripping.
The overall process performance is summarized in the table below, which compares the key indicators of the TBP system with typical acidic extractant systems. The TBP system achieved high extraction efficiency, excellent separation from aluminum, copper, and selenium, quantitative stripping with deionized water, high saturation capacity, and good cyclic stability. These features make it a promising candidate for industrial indium recovery from CIAS solar cell leachates and other secondary indium resources. The same approach may also be applicable to indium recovery from perovskite solar cell modules, where indium-containing transparent electrodes are present and where the demand for sustainable recycling is rapidly growing.
| Indicator | TBP system in this study | Typical acidic extractant system |
|---|---|---|
| Extraction efficiency / % | 95.34 single stage; 99.86 two-stage | 90–99 |
| Stripping agent | Deionized water | Strong acid or alkaline solution |
| Stripping efficiency / % | 98.55 single stage; 99.16 two-stage | Often low in single stage; multi-stage required |
| Separation factor βIn/Cu | 42.91 | Often < 10 |
| Saturation capacity / (g·L⁻¹) | 24.72 for 30% TBP | 10–30 |
| Cyclic stability | 89.47% after 6 cycles | Moderate to good |
| Environmental impact | Low, water-based stripping | High, acidic wastewater |
The experimental results lead to several important conclusions. First, among TBP, P350, and TRPO, TBP provides the best compromise between indium extraction and stripping. TRPO has stronger extraction but poor indium–copper separation and difficult stripping. P350 has poor indium–aluminum and indium–copper separation and is nearly impossible to strip with water. TBP achieves high indium extraction, excellent separation from aluminum, copper, and selenium, and near-complete stripping with deionized water. Second, the optimized extraction conditions for the TBP system are 30% TBP, 6 mol·L⁻¹ HCl, 25 °C, and O:A = 1:1. Under these conditions, the single-stage indium extraction efficiency reaches 95.34%, and the saturation capacity of 30% TBP for indium is 24.72 g·L⁻¹. Third, the loaded organic phase can be efficiently stripped with deionized water. A single-stage stripping efficiency of 98.55% is achieved at A:O = 1:1, and a two-stage cross-current stripping process achieves 99.16% total indium stripping. Fourth, the TBP system exhibits good cyclic stability, with a single-stage extraction efficiency of 89.47% after six cycles. Fifth, full-flow validation with real CIAS solar cell leachate confirms that the process can achieve 99.86% total indium extraction and 99.16% total indium stripping while effectively separating indium from copper, aluminum, and selenium. Sixth, FT-IR and ³¹P NMR analyses reveal that TBP extracts indium through coordination of the phosphoryl oxygen to indium. The P=O stretching vibration shifts from 1271 to 1227 cm⁻¹, and the ³¹P NMR chemical shift moves from 58.46 to -1.06 ppm, providing direct evidence of coordination.
The findings of this study have implications beyond CIAS solar cells. Indium is also used in transparent conductive oxides for perovskite solar cell devices, and the rapid growth of the perovskite solar cell market will inevitably generate increasing amounts of indium-containing waste. The neutral phosphine oxide extraction system developed here offers a clean, efficient, and potentially scalable route for indium recovery from such waste streams. Because the system uses deionized water for stripping, it avoids the use of strong acids and reduces the generation of acidic wastewater. The high selectivity for indium over aluminum, copper, and selenium is particularly important for obtaining high-purity indium products. The good cyclic stability of TBP also supports long-term operation and reduces reagent consumption. Therefore, the TBP-based neutral phosphine oxide extraction process is a promising technology for sustainable indium recycling from spent CIAS solar cells, perovskite solar cell modules, and other secondary indium resources. Future work should focus on pilot-scale testing, continuous counter-current process design, and integration with upstream leaching and downstream indium electrowinning or precipitation to establish a complete closed-loop recycling流程. The application of this system to industrial perovskite solar cell recycling could significantly contribute to resource security and environmental sustainability.