For thin-film photovoltaic power generation, the highest-leverage sputtering targets are molybdenum for the back contact, copper indium gallium selenide (CIGS) or cadmium telluride (CdTe) family compounds for the absorber, and indium tin oxide (ITO) or aluminum-doped zinc oxide (AZO) for the transparent window layer, with 3N5 to 5N purity as the working spec [S1][S4].
Argon is the working gas in essentially every PVD line that deposits these layers because its atomic mass efficiently transfers kinetic energy to the target on ion impact [S2]. Power-generation PVD is one of the four main downstream markets for the global sputtering target industry, alongside semiconductors, flat-panel displays, and data storage, with photovoltaic growth driven by renewable-energy build-out [S3].
Material families that actually show up on a PV production line
CIGS targets in the CuInGaSe and CuInSe composition set are deposited at 3N5 to 5N purity and are the absorber layer in flexible and rigid CIGS modules [S4]. CdTe and CdS targets at 4N to 5N purity cover the absorber and buffer layers in CdTe module production, while GaAs targets at 3N serve III-V concentrator and space cells [S4].
Transparent conductive oxide (TCO) layers are deposited from ITO targets in the standard 1:9 In2O3:SnO2 weight ratio at 4N, or from AZO/ZAO targets with ZnO/Al2O3 ratios from 0.5 to 2 wt%, both at 3N [S1][S4]. Mo targets at 3N to 3N5 are the workhorse back contact, with intrinsic ZnO, ZnO:MgO (15 to 30 at% MgO), and ZnO:Ga (0.5 to 2.5 wt% Ga2O3) targets covering the high-resistivity window stack [S4]. CuIn, CuGa, and CuSe precursor targets are also used for two-step CIGS selenization processes [S4].
Selection criteria engineers actually use
Purity is the first hard gate: photovoltaic-grade sputtering targets are typically 3N (99.9%) minimum, with 4N (99.99%) or 5N (99.999%) required for absorber and TCO layers where trace metal contamination degrades cell efficiency [S4][S7]. High density, generally above 95% of theoretical, is needed to suppress arcing and particle generation during long production campaigns [S7].
Target geometry is the second decision. Planar targets are the default for small-area and R&D chambers; rotary (cylindrical) targets are specified for large-area architectural glass and high-throughput float-line coaters because they give higher material utilization, often above 70%, and longer campaign life between target changes [S3]. Sputter-ready surface treatments, including precision machining, wet polishing, and plasma etch conditioning, can cut the burn-in window from 1 to 6 hours (10 to 35 kWh per target) on a semiconductor-style PVD chamber, which translates directly into more substrates per shift [S2].
Bonding method comes third. Indium bonding, elastomeric bonding, and OFHC copper backing plates are the three commercial options, and the choice is driven by the target material's thermal expansion, the power supply pulse profile, and cooling-water flow at the cathode [S5]. For brittle ceramic TCO targets like ITO and AZO, elastomeric bonding is common because it tolerates differential thermal expansion without cracking the target. Vendor lists consistently call out Al, Cu, and Au alongside alloys and ceramics as the everyday material set for PVD work, with purity cited as the dominant procurement criterion [S9].
Direct comparison of the leading PV target options
Molybdenum (Mo, 3N to 3N5) is the back contact of choice because of its work function, stability against Se during CIGS selenization, and good adhesion to soda-lime glass; pure Mo is the baseline, and no widely adopted thin-film alternative has displaced it for CIGS [S4].
CIGS (CuInGaSe, 3N5 to 5N) holds the highest lab-cell efficiency among polycrystalline thin-film absorber materials, with CuInSe2 (CIS, 3N5 to 4N) used where Ga grading is not required. CdTe (4N to 5N) is the dominant absorber in commercial utility-scale thin-film modules because of its near-ideal bandgap of about 1.45 eV and high absorption coefficient. GaAs (3N) is reserved for space and concentrator cells where efficiency justifies the cost of the III-V target. For the window layer, ITO (1:9 In2O3:SnO2, 4N) gives the lowest sheet resistance below 10 ohm/sq at typical thicknesses, while AZO (ZnO/Al2O3, 3N) is the lower-cost workhorse for large-area amorphous-silicon and CdTe production lines [S1][S4].
On cost, AZO and Mo are typically 1/3 to 1/5 the price per kilogram of ITO and CIGS. On lead time, Mo and AZO are commonly ex-stock in standard planar formats, while rotary CIGS and rotary ITO carry longer lead times of 8 to 16 weeks from specialist suppliers [S3][S4]. On toxicity and regulatory exposure, CdTe and CdS carry cadmium handling requirements under RoHS and equivalent regional rules, which is a real procurement constraint in the EU and Japan.
Standards, sourcing, and the supply landscape
There is no single international standard that defines a "PV-grade" sputtering target. The relevant manufacturing practice norms sit under the ISO 9001 quality system for target producers, the IEC 61215 and IEC 61646 module qualification frameworks that indirectly drive target purity, and the RoHS hazardous-substance regime that governs CdTe and CdS chemistry inside the EU [S3]. Sourcing decisions are driven by raw-material traceability, especially for In, Te, and Ga, where single-country supply concentration has been a recurring procurement risk.
The global market is fragmented across specialty metals suppliers and dedicated PVD materials companies. Materion Corporation expanded its semiconductor PVD footprint in Asia in April 2025 through a strategic acquisition, which directly affects CIGS and Mo target availability for module makers in the region [S3]. Able Target Limited publishes a public PV target reference list with the 30+ material and purity combinations above [S4]. Kurt J. Lesker's materials division stocks precious-metal targets, including Au, Ag, Pt, Pd, Ir, Rh, and Ru, for grid and bus-bar metallization layers, with backing plates and target-bonding services [S5]. The broader market also lists Cu, Al, Ti, Ta, Mo, and ceramic targets such as Al2O3, SiO2, ZnO, and ITO as the high-volume material set for energy and power applications [S3].
Burn-in, yield, and the cost-of-ownership angle
Burn-in is a non-productive step that removes adsorbed moisture, surface oxides, and the plastically deformed surface layer introduced by machining. On a semiconductor-style PVD chamber, burn-in can consume 1 to 6 hours and 10 to 35 kWh per target, with sputter-ready surface treatment cutting that window sharply and stabilizing sheet-resistance uniformity earlier in target life [S2]. The same logic applies to large-area PV lines: a treated Ti target showed lower in-film particle counts at ≥0.3 µm, a quality lever that translates into higher cell yield [S2].
Sputtering yield reference tables for Ti, Al, Cu, Mo, and the ITO/AZO family of compounds let a process engineer back-calculate the kW power and target thickness needed to hit a given film thickness per substrate pass, given the power supply pulse profile and the power distribution capacity of the tool [S8]. Planar target utilization of 30 to 50% versus rotary target utilization above 70% is the single largest consumable-cost lever on a high-throughput PV line.
Limits, failure modes, and what to avoid
Pure ITO targets are increasingly substituted by AZO on cost grounds, with ITO retained for the highest-efficiency heterojunction and perovskite-on-silicon tandem cells where sheet resistance below 10 ohm/sq is non-negotiable [S1][S4]. Indium supply concentration is a real risk for ITO-heavy lines, and many producers now run dual-qualified AZO and ITO recipes on the same rotary PVD platform.
CdTe and CdS targets bring a regulatory and waste-stream burden that does not apply to CIGS or silicon thin-film stacks, which is the main reason CdTe lines are concentrated in a small number of specialized module producers. CIGS targets are also notoriously sensitive to Cu/III stoichiometry drift during deposition; that is why the CuIn, CuGa, and CuInGa precursor split is used in selenization flows, rather than a single quaternary sputter target [S4]. For all of these, the target-bonding interface and the power supply arcing envelope are the two failure modes that drive unplanned downtime, and both are listed as key technical risks in vendor selection guides [S5][S9].
Two trackable signals for the next six months: vendor announcements around rotary AZO and rotary ITO capacity additions for architectural and tandem PV lines, and any expansion of regional Mo and CIGS target-bonding services to shorten the 8 to 16 week lead time for large-area rotary formats. Process engineers specifying targets for new power-generation PVD lines should also review the sputtering target reference page for the full material and bonding matrix before locking in a bill of materials.
This topic is covered further in US Magnet Supplier Map 2026: Flexible, NdFeB, SmCo, and Alnico Shortlist for Spec.