Copper indium selenide

• CAS: 12018-95-0
• Synonyms: COPPER INDIUM SELENIDE
• Molecular Formula: CuInSe2
• Molecular Weight: 336.3 g/mol
• SMILES: [Cu].[Se].[Se]=[In]
• InChI: InChI=1S/Cu.In.2Se
• InChIKey: KTSFMFGEAAANTF-UHFFFAOYSA-N

Copper indium selenide (CuInSe₂, CIS) is a ternary I-III-VI₂ chalcopyrite semiconductor characterized by a direct band gap of approximately 1.04 eV at room temperature , a tetragonal crystal structure with lattice parameters a = 5.77–5.81 Å and c = 11.56–11.63 Å , and an optical absorption coefficient exceeding 10⁵ cm⁻¹ in the visible-to-near-infrared range. It serves as the foundational narrow-bandgap endpoint of the commercially critical Cu(In,Ga)Se₂ (CIGS) solid solution series and is the prototype material for understanding defect physics in chalcopyrite photovoltaics.

Copper Indium Selenide (CuInSe₂, CAS 12018-95-0): Chalcopyrite Semiconductor Procurement Guide for Photovoltaic and Optoelectronic R&D

Copper indium selenide (CuInSe₂, CIS) is a ternary I-III-VI₂ chalcopyrite semiconductor characterized by a direct band gap of approximately 1.04 eV at room temperature , a tetragonal crystal structure with lattice parameters a = 5.77–5.81 Å and c = 11.56–11.63 Å , and an optical absorption coefficient exceeding 10⁵ cm⁻¹ in the visible-to-near-infrared range . It serves as the foundational narrow-bandgap endpoint of the commercially critical Cu(In,Ga)Se₂ (CIGS) solid solution series and is the prototype material for understanding defect physics in chalcopyrite photovoltaics. Its unique combination of ambipolar dopability, extreme off-stoichiometry tolerance, and the ability to form ordered vacancy compounds distinguishes it from all other I-III-VI₂ family members .

• [1] Soni, A., Gupta, V., Arora, G., & Dashora, A. (2013). Density functional theory of structural, electronic and optical properties of CuXY₂ (X=In, Ga and Y=S, Se) chalcopyrite semiconductors. Physica B: Condensed Matter, 416, 55–63. View Source
• [2] Paszkowicz, W., et al. (1985). The lattice constants of CuInSe₂. Journal of Electronic Materials, 14(6), 677–683. View Source
• [3] Kosyachenko, L. A., Grushko, E. V., & Mikityuk, T. I. (2012). Absorptivity of semiconductors used in the production of solar cell panels. Semiconductors, 46(4), 466–470. View Source
• [4] Zhang, S. B., Wei, S.-H., Zunger, A., & Katayama-Yoshida, H. (1998). Defect physics of the CuInSe₂ chalcopyrite semiconductor. Physical Review B, 57(16), 9642–9656. View Source

Why CuInSe₂ Cannot Be Replaced by Generic Chalcopyrite Analogs: Structural and Electronic Differentiation Evidence

Within the CuXY₂ (X=In,Ga; Y=S,Se) chalcopyrite family, band gap varies by a factor of ~2.3 from 1.04 eV (CuInSe₂) to 2.43 eV (CuGaS₂), and carrier type control differs fundamentally between compounds . CuInSe₂ is the only member of the series that can be reversibly switched between p-type and n-type conduction by simple stoichiometric adjustment or post-growth annealing, whereas CuGaSe₂ resists n-type doping without elaborate co-doping strategies . Furthermore, CuInSe₂ uniquely tolerates massive off-stoichiometry—manifested as a series of ordered vacancy compounds (CuIn₃Se₅, CuIn₅Se₈, Cu₂In₄Se₇, Cu₃In₅Se₉)—without catastrophic degradation of electronic properties, a feature absent in the sulfide analogs and in competing thin-film absorbers such as CdTe . Substituting CuInSe₂ with a wider-bandgap analog or a different thin-film absorber therefore alters not merely the absorption onset but the entire defect-chemical landscape, doping strategy, and device architecture.

• [1] Soni, A., Gupta, V., Arora, G., & Dashora, A. (2013). Density functional theory of structural, electronic and optical properties of CuXY₂ (X=In, Ga and Y=S, Se) chalcopyrite semiconductors. Physica B: Condensed Matter, 416, 55–63. View Source
• [2] Schön, J. H., Schenker, O., Riazi-Nejad, H., & Bucher, E. (1999). Comparison of point defects in CuInSe₂ and CuGaSe₂ single crystals. Solar Energy Materials and Solar Cells, 57(3), 229–245. View Source
• [3] Wei, S.-H., Zhang, S. B., & Zunger, A. (2005). Defect properties of CuInSe₂ and CuGaSe₂. Journal of Physics and Chemistry of Solids, 66(11), 1994–1999. View Source

CuInSe₂ Quantitative Differentiation Evidence: Head-to-Head Data vs. CuGaSe₂, CuInS₂, CdTe, and c-Si

Narrowest Direct Band Gap Among Chalcopyrites Enables Tandem Bottom-Cell and Near-IR Photodetector Applications

CuInSe₂ exhibits the narrowest experimental direct band gap (1.04 eV) among the four endpoint CuXY₂ chalcopyrite compounds, compared to CuGaSe₂ (1.67 eV), CuInS₂ (1.55 eV), and CuGaS₂ (2.43 eV) . This 0.63 eV reduction relative to CuGaSe₂ places CuInSe₂ near the optimal bottom-cell absorption threshold for mechanically stacked and monolithic tandem photovoltaic architectures, while its direct-gap character ensures high oscillator strength for absorption at the band edge, unlike the indirect-gap c-Si .

Evidence Dimension	Experimental direct band gap (eV)
Target Compound Data	1.04 eV (CuInSe₂)
Comparator Or Baseline	CuGaSe₂: 1.67 eV; CuInS₂: 1.55 eV; CuGaS₂: 2.43 eV; c-Si: 1.12 eV (indirect)
Quantified Difference	0.51–1.39 eV narrower than all chalcopyrite analogs; 0.08 eV narrower than c-Si but with direct-gap advantage
Conditions	Experimental room-temperature optical absorption and photoluminescence measurements on single-crystal and thin-film samples; DFT-GGA and LDA theoretical validations

Why This Matters

Procurement of CuInSe₂ is essential when the application requires the narrowest-bandgap chalcopyrite endpoint—any substitution with CuGaSe₂ or CuInS₂ shifts the absorption edge by >0.5 eV, fundamentally altering the spectral response and tandem current-matching design.

• [1] Soni, A., Gupta, V., Arora, G., & Dashora, A. (2013). Density functional theory of structural, electronic and optical properties of CuXY₂ (X=In, Ga and Y=S, Se) chalcopyrite semiconductors. Physica B: Condensed Matter, 416, 55–63. Experimental band gaps: CuGaS₂ 2.43 eV, CuInS₂ 1.55 eV, CuInSe₂ 1.04 eV, CuGaSe₂ 1.67 eV. View Source
• [2] AbuShama, J., Noufi, R., & Johnston, S. (2005). Improved Performance in CuInSe₂ and Surface-Modified CuGaSe₂ Solar Cells. NREL/CP-520-37614. World record CIS cell efficiency 15.0% used as tandem bottom cell. View Source

Reversible Ambipolar Doping by Stoichiometry Control: CuInSe₂ vs. CuGaSe₂

CuInSe₂ can be converted from p-type to n-type and back by annealing in vacuum or under maximum Se pressure, or by controlling the Cu/In ratio during growth . In stark contrast, CuGaSe₂ has resisted n-type doping through native defect control alone; n-type CuGaSe₂ was achieved only through co-doping with Ge and Zn via ion implantation and subsequent Zn-atmosphere annealing, yielding a maximum electron concentration of ~10¹⁶ cm⁻³ at room temperature . The fundamental origin is the higher formation energy of the Ga_Cu antisite donor in CuGaSe₂ compared to In_Cu in CuInSe₂, and the much deeper donor level of Ga_Cu, which acts as an electron trap .

Evidence Dimension	Feasibility of n-type doping via native defects
Target Compound Data	CuInSe₂: Reversible p↔n conversion by vacuum or Se-pressure annealing; n-type carrier concentration up to 10¹⁶–10¹⁷ cm⁻³ achievable via stoichiometry alone
Comparator Or Baseline	CuGaSe₂: n-type impossible via native defects alone; requires Ge+Zn co-doping with complex post-implantation Zn-annealing; maximum n ~10¹⁶ cm⁻³
Quantified Difference	CuInSe₂ achieves n-type conduction with simple thermal processing; CuGaSe₂ requires extrinsic co-dopants and multi-step processing—a qualitative (go/no-go) differentiation in process complexity
Conditions	Single crystals grown by chemical vapor transport; annealing in vacuum or In/Ga atmosphere at elevated temperatures; Hall-effect verification

Why This Matters

For any application requiring both p- and n-type regions of the same base material (homojunctions, bipolar transistors, complementary logic), CuInSe₂ is the only viable chalcopyrite endpoint; CuGaSe₂ cannot serve this function without extrinsic doping that introduces additional complexity and potential contamination.

• [1] Schön, J. H., Schenker, O., Riazi-Nejad, H., & Bucher, E. (1999). Comparison of point defects in CuInSe₂ and CuGaSe₂ single crystals. Solar Energy Materials and Solar Cells, 57(3), 229–245. CuInSe₂ easily transformed p- to n-type by annealing; CuGaSe₂ cannot. View Source
• [2] Schön, J. H., Schenker, O., & Bucher, E. (2000). Transport properties of n-type CuGaSe₂. Thin Solid Films, 361–362, 411–414. n-type CuGaSe₂ achieved only via Ge/Zn co-doping; carrier concentration up to 10¹⁶ cm⁻³. View Source
• [3] Wei, S.-H., Zhang, S. B., & Zunger, A. (2005). Defect properties of CuInSe₂ and CuGaSe₂. Journal of Physics and Chemistry of Solids, 66(11), 1994–1999. Ga_Cu donor level deeper than In_Cu; larger donor formation energy in CuGaSe₂. View Source

Unmatched Off-Stoichiometry Tolerance: Ordered Vacancy Compound Formation Unique to CuInSe₂

CuInSe₂ tolerates extreme Cu-poor off-stoichiometry by spontaneously forming a homologous series of ordered vacancy compounds (OVCs): CuIn₃Se₅, CuIn₅Se₈, Cu₂In₄Se₇, and Cu₃In₅Se₉. These arise from the periodic insertion of the exceptionally low-energy neutral defect pair (2V_Cu⁻ + In_Cu²⁺) into the CuInSe₂ lattice . First-principles calculations show that under Cu-poor conditions, the formation enthalpy of this pair can be near-zero or even exothermic . In contrast, CuGaSe₂ does not support an analogous homologous OVC series, and II-VI or III-V semiconductors (e.g., GaAs, CdTe) require near-perfect stoichiometry to maintain electronic quality .

Evidence Dimension	Off-stoichiometry tolerance range and ordered defect compound formation
Target Compound Data	CuInSe₂: Forms CuIn₃Se₅, CuIn₅Se₈, Cu₂In₄Se₇, Cu₃In₅Se₉; (2V_Cu+In_Cu) pair formation enthalpy near-zero or exothermic under Cu-poor chemical potentials; maintains electrically benign character despite >10²⁰ cm⁻³ native defect concentrations
Comparator Or Baseline	CuGaSe₂: No analogous homologous OVC series reported; GaAs and CdTe: require stoichiometry control within ~0.01% for device-grade electronic quality (class-level baseline)
Quantified Difference	CuInSe₂ accommodates Cu/In ratios deviating from 1.0 by >20% while retaining photovoltaic functionality; III-V and II-VI semiconductors degrade at deviations exceeding ~0.1% (qualitative class-level difference)
Conditions	First-principles defect calculations (LDA/GGA, supercell approach); experimental validation via XRD, TEM, and XAFS on off-stoichiometric bulk crystals and thin films

Why This Matters

For large-area thin-film manufacturing, CuInSe₂'s intrinsic defect tolerance translates to wider process windows, lower rejection rates, and reduced need for precise composition control—directly lowering the cost-per-watt in photovoltaic module production compared to alternatives that demand stringent stoichiometry.

• [1] Zhang, S. B., Wei, S.-H., Zunger, A., & Katayama-Yoshida, H. (1998). Defect physics of the CuInSe₂ chalcopyrite semiconductor. Physical Review B, 57(16), 9642–9656. (2V_Cu+In_Cu) pairs have remarkably low formation enthalpy; explain CuIn₅Se₈, CuIn₃Se₅, Cu₂In₄Se₇, Cu₃In₅Se₉. View Source
• [2] Wei, S.-H., Zhang, S. B., & Zunger, A. (2005). Defect properties of CuInSe₂ and CuGaSe₂. Journal of Physics and Chemistry of Solids, 66(11), 1994–1999. Defect pair formation enthalpy and self-passivation mechanism. View Source
• [3] Noufi, R., et al. (1995). Research on high-efficiency, large-area CuInSe₂-based thin-film modules. NREL Subcontract Report. CIS tolerates large compositional variations while maintaining >15% efficiency. View Source

Absorption Coefficient and Material Utilization: CuInSe₂ Requires ~2000× Thinner Absorber than c-Si for Equivalent Photon Capture

CuInSe₂ achieves near-total absorption of AM1.5 solar photons at a film thickness of only 3–4 µm, compared to 7–8 mm for crystalline silicon (c-Si) and 20–30 µm for CdTe . For 90% photon absorption, CuInSe₂ requires less than 1 µm of material, while c-Si requires approximately 100–200 µm in commercial cells . This difference—a factor of ~2000× in required thickness—stems from CuInSe₂'s direct band gap and absorption coefficient exceeding 10⁵ cm⁻¹ across the visible spectrum .

Evidence Dimension	Absorber thickness required for near-total AM1.5 photon absorption
Target Compound Data	CuInSe₂: 3–4 µm for total absorption; <1 µm for 90% absorption; α > 10⁵ cm⁻¹
Comparator Or Baseline	c-Si: 7–8 mm (total absorption); CdTe: 20–30 µm; a-Si:H: 30–60 µm; CuGaSe₂: 3–4 µm (same class)
Quantified Difference	CuInSe₂ requires ~2000× less material thickness than c-Si and ~7× less than CdTe for equivalent photon absorption
Conditions	AM1.5 global spectrum; 25°C; absorptivity calculations from measured absorption coefficients with consideration for spectral distribution; no reflection losses assumed

Why This Matters

The extreme thinness of CuInSe₂ absorber layers directly reduces raw material consumption, deposition time, and capital equipment throughput requirements in manufacturing—translating to lower bill-of-materials cost and higher production capacity per tool compared to c-Si or CdTe production lines.

• [1] Kosyachenko, L. A., Grushko, E. V., & Mikityuk, T. I. (2012). Absorptivity of semiconductors used in the production of solar cell panels. Semiconductors, 46(4), 466–470. Total absorption at d = 3–4 µm for CIS and CGS; 7–8 mm for c-Si; 20–30 µm for CdTe. View Source
• [2] Noufi, R., et al. (1995). Research on high-efficiency, large-area CuInSe₂-based thin-film modules. NREL Subcontract Report. 90% photon absorption in <1 µm for CIS (direct band gap ~0.95 eV). View Source
• [3] Mickelsen, R. A. & Chen, W. S. (1980). High photocurrent polycrystalline CuInSe₂ thin-film solar cells. Applied Physics Letters, 36(5), 371–373. α > 10⁵ cm⁻¹ for CuInSe₂. View Source

Carrier Mobility Asymmetry: n-Type CuInSe₂ Electron Mobility Exceeds p-Type by >20× and Outperforms CuInS₂

Single-crystal CuInSe₂ exhibits a pronounced mobility asymmetry: n-type electron mobility reaches 660 cm²/V·s, while p-type hole mobility is limited to ~30 cm²/V·s—a factor of 22× difference . This n-type mobility substantially exceeds that of CuInS₂ (~10 cm²/V·s from thin-film Hall measurements ). In p-type Bridgman-grown CuInSe₂ with low Na addition (0–0.2 at.%), the average hole mobility is 17 cm²/V·s, increasing for Se-rich compositions (CuInSe₂.₂) . The high n-type mobility is attributed to the shallowness of the In_Cu donor level and low compensation ratios achievable through stoichiometry control.

Evidence Dimension	Room-temperature carrier mobility (cm²/V·s)
Target Compound Data	CuInSe₂: n-type μₑ = 660 cm²/V·s (single crystal); p-type μₕ = 30 cm²/V·s (single crystal), 17 cm²/V·s (Bridgman polycrystal with Na)
Comparator Or Baseline	CuInS₂: μ = 10.09 cm²/V·s (combined); CIGS thin film: μ = 3–22 cm²/V·s; CuInSe₂ p-type single crystal: 30 cm²/V·s
Quantified Difference	CuInSe₂ n-type mobility is 66× higher than CuInS₂ mobility and >30× higher than typical CIGS thin-film mobility
Conditions	Single-crystal Hall effect measurements at 300 K; Bridgman-grown ingots with controlled Na addition (0–0.2 at.%); thin-film Hall data from compiled literature

Why This Matters

For high-frequency or high-speed optoelectronic devices (photodetectors, modulators), the 660 cm²/V·s n-type mobility of CuInSe₂ enables higher transit frequencies and lower RC delays than any other chalcopyrite endpoint, making it the preferred choice when electron transport governs device speed.

• [1] Parkes, J., Tomlinson, R. D., & Hampshire, M. J. (1973). Electrical properties of CuInSe₂ single crystals. Solid-State Electronics, 16(7), 773–783. n-type μ = 660 cm²/V·s, p-type μ = 30 cm²/V·s. View Source
• [2] Mkawi, E. M., et al. (2024). Table 1: Energy Band Gap and Carrier Mobility of Selected Semiconductor Materials. Nanomaterials, 14, 1783. CuInS₂ carrier mobility = 10.09 cm²/V·s. View Source
• [3] Simchi, H., et al. (2010). Room temperature transport measurements on Bridgman-grown CuInSe₂ with added sodium. Proceedings of SPIE, 7750, 775030. p-type μ_avg = 17 cm²/V·s; higher for CuInSe₂.₂. View Source

Near-Zero Valence Band Offset with CuGaSe₂ Enables Lossless Tandem Heterojunction Integration

The valence band offset between CuInSe₂ and CuGaSe₂ is extrapolated to be only ΔE_V = +0.036 eV, which is effectively flat-band alignment . In contrast, the CuInSe₂/CuInS₂ interface exhibits a substantial valence band offset of ΔE_V = −0.23 eV (with CuInS₂ valence band lying 0.23 eV below that of CuInSe₂) and a conduction band offset of ΔE_C = +0.21 eV . The near-zero ΔE_V between CuInSe₂ and CuGaSe₂ means that hole transport across a CIS/CGS heterojunction encounters negligible energy barrier, enabling efficient tandem and graded-bandgap device architectures without band-offset-induced recombination losses.

Evidence Dimension	Valence band offset ΔE_V and conduction band offset ΔE_C at heterointerface (eV)
Target Compound Data	CuInSe₂/CuGaSe₂: ΔE_V = +0.036 eV, ΔE_C = +0.70 eV
Comparator Or Baseline	CuInSe₂/CuInS₂: ΔE_V = −0.23 eV, ΔE_C = +0.21 eV
Quantified Difference	ΔE_V(CIS/CGS) is 0.266 eV smaller in magnitude than ΔE_V(CIS/CuInS₂); CuInSe₂/CuGaSe₂ interface is ~7× closer to flat-band condition
Conditions	Extrapolated from acceptor energy reference levels in Cu(In,Ga)(Se,S)₂ alloys; polycrystalline thin-film samples; photoelectron spectroscopy and device modeling corroboration

Why This Matters

For tandem solar cell procurement, CuInSe₂ is the natural bottom-cell partner to CuGaSe₂ (or Ga-rich CIGS) top cells because the negligible valence band offset eliminates the need for complex tunnel junctions or interfacial buffer layers that would otherwise be required to overcome band discontinuities—simplifying device fabrication and reducing series resistance losses.

• [1] Turcu, M., Kötschau, I. M., & Rau, U. (2002). Compositional trends of defect energies, band alignments, and recombination mechanisms in the Cu(In,Ga)(Se,S)₂ alloy system. Journal of Applied Physics, 91(3), 1391–1399. ΔE_V(CuInSe₂/CuGaSe₂) = 0.036 eV; ΔE_V(CuInSe₂/CuInS₂) = −0.23 eV. View Source
• [2] Turcu, M. & Rau, U. (2004). Defect energies, band alignments, and charge carrier recombination in polycrystalline Cu(In,Ga)(Se,S)₂ alloys. Journal of Physics and Chemistry of Solids, 64(9–10), 1591–1595. ΔE_C(CuInSe₂/CuInS₂) = 0.21 eV; ΔE_C(CuInSe₂/CuGaSe₂) = 0.70 eV. View Source

CuInSe₂ Procurement-Linked Application Scenarios Derived from Quantitative Differentiation Evidence

Mechanically Stacked and Monolithic Tandem Photovoltaic Bottom Cells

CuInSe₂'s 1.04 eV direct band gap , combined with its near-zero valence band offset of only 0.036 eV relative to CuGaSe₂ , makes it the optimal narrow-gap absorber for tandem solar cells paired with wider-gap chalcopyrite or perovskite top cells. The NREL-demonstrated 15.0% world-record CIS cell served as the bottom component in a mechanical-stacked tandem achieving 15.31% total-area efficiency . The 3–4 µm absorber thickness requirement enables lightweight, flexible tandem modules where c-Si bottom cells (>100 µm) would be impractically thick and rigid.

Near-Infrared Photodetectors and Spectrally Selective Sensors (1.0–1.2 eV Range)

With the narrowest band gap among all Cu-based chalcopyrites (1.04 eV vs. 1.55–2.43 eV for sulfide and Ga-containing analogs) , CuInSe₂ is the only I-III-VI₂ endpoint suitable for photodetectors operating in the 1000–1200 nm near-infrared window without requiring compositional grading. The material's high absorption coefficient (>10⁵ cm⁻¹) ensures efficient photon capture in sub-micron films , enabling fast-response thin-film photodetector architectures.

Homojunction and Bipolar Electronic Devices Requiring Ambipolar Doping

CuInSe₂ is unique among the four CuXY₂ endpoint compounds in its ability to support both p-type and n-type conduction through simple stoichiometric adjustment or post-growth annealing, without extrinsic dopants . This enables fabrication of homojunction p-n diodes, bipolar junction transistors, and complementary logic elements from a single base material—a capability that CuGaSe₂ cannot match without elaborate Ge+Zn co-doping procedures . The n-type electron mobility of 660 cm²/V·s further supports high-frequency device operation.

Defect-Tolerant Large-Area Thin-Film Manufacturing with Wide Process Windows

CuInSe₂'s ability to form electrically benign ordered vacancy compounds (CuIn₃Se₅, CuIn₅Se₈) under Cu-poor growth conditions provides an intrinsic process tolerance that is absent in CdTe, c-Si, and III-V thin-film technologies. Manufacturing lines can operate with wider composition control windows—Cu/In ratios deviating by >20% from unity still yield functional photovoltaic devices —reducing yield loss, metrology burden, and overall production cost per watt relative to stoichiometry-sensitive alternatives.

• [1] Soni, A., Gupta, V., Arora, G., & Dashora, A. (2013). Density functional theory of structural, electronic and optical properties of CuXY₂ (X=In, Ga and Y=S, Se) chalcopyrite semiconductors. Physica B: Condensed Matter, 416, 55–63. View Source
• [2] Turcu, M., Kötschau, I. M., & Rau, U. (2002). Compositional trends of defect energies, band alignments, and recombination mechanisms in the Cu(In,Ga)(Se,S)₂ alloy system. Journal of Applied Physics, 91(3), 1391–1399. View Source
• [3] AbuShama, J., Noufi, R., & Johnston, S. (2005). Improved Performance in CuInSe₂ and Surface-Modified CuGaSe₂ Solar Cells. NREL/CP-520-37614. View Source
• [4] Kosyachenko, L. A., Grushko, E. V., & Mikityuk, T. I. (2012). Absorptivity of semiconductors used in the production of solar cell panels. Semiconductors, 46(4), 466–470. View Source
• [5] Schön, J. H., Schenker, O., Riazi-Nejad, H., & Bucher, E. (1999). Comparison of point defects in CuInSe₂ and CuGaSe₂ single crystals. Solar Energy Materials and Solar Cells, 57(3), 229–245. View Source
• [6] Schön, J. H., Schenker, O., & Bucher, E. (2000). Transport properties of n-type CuGaSe₂. Thin Solid Films, 361–362, 411–414. View Source
• [7] Parkes, J., Tomlinson, R. D., & Hampshire, M. J. (1973). Electrical properties of CuInSe₂ single crystals. Solid-State Electronics, 16(7), 773–783. View Source
• [8] Zhang, S. B., Wei, S.-H., Zunger, A., & Katayama-Yoshida, H. (1998). Defect physics of the CuInSe₂ chalcopyrite semiconductor. Physical Review B, 57(16), 9642–9656. View Source
• [9] Noufi, R., et al. (1995). Research on high-efficiency, large-area CuInSe₂-based thin-film modules. NREL Subcontract Report. View Source