MENU
Advanced Materials Intelligence
  1. Home
  2. Transparent Conductive Oxides Explained | Perovskite Solar Cells

Transparent Conductive Oxides Explained | Perovskite Solar Cells

Back to Technology, Institution & Company Guides

TECHNOLOGY EXPLAINER

Transparent Conductive Oxides (ITO, IZO, FTO)
— letting light through and carrying current, and what that balance costs

These thin films serve as both the entrance for light and the exit for current in a perovskite solar cell. They are used as the bottom electrode on glass, as the top electrode in tandems, and within the recombination layer. There are three issues: the tug of war between transparency and conductivity, damage to the layers below from sputtering, and indium.

Built from primary sources: US Geological Survey (USGS) Mineral Commodity Summaries 2026 and peer-reviewed papers (Nature Energy, Nature Communications, Advanced Materials, ACS Applied Materials & Interfaces, Energy & Environmental Science and others) / Last updated September 2026

Conceptual image of soft light passing at an angle through a plain, clear glass plate with a faint blue-grey tint, against a dark background
AI-generated concept image. An impression of a thin film that carries current while letting light through. It does not represent the real colour or transmittance of any film, or a product.
What this article covers
  1. What a transparent conductive oxide is, in three points
  2. Where they are used in the cell
  3. How ITO, IZO, FTO and others differ
  4. A materials engineer's view (1): easing the transparency-conductivity trade-off through mobility
  5. Sputter damage — particles, UV light and heat
  6. Buffer layers — one extra layer to prevent damage
  7. A materials engineer's view (2): protective layers are also chosen for speed
  8. Indium — concentrated supply, with our calculation
  9. Strengths and open problems
  10. Glossary / References / Claim-to-source audit
How claims are labelled in this article

Sourced = stated in published material (link given)
Our calculation = a value this article derived from assumptions it states
Not yet confirmed = a plan, target or outlook with no confirmed track record
Structural readings and materials-design interpretations are marked separately as Commentary. Efficiency figures are given with whether they are certified, and their area.

1. What a transparent conductive oxide is, in three points

A transparent conductive oxide (TCO) is a thin oxide film that transmits visible light well and yet conducts electricity like a metal. The main examples are tin-doped indium oxide (ITO), fluorine-doped tin oxide (FTO), indium zinc oxide (IZO) and aluminium-doped zinc oxide (AZO).

  • What is used: a 2025 paper in ACS Applied Materials & Interfaces (Magliano et al.) says ITO, AZO and IZO are the most common transparent electrodes in semi-transparent perovskite cells because of their high transparency and electron mobilitySourced
  • How they are made: the same paper says TCOs are usually deposited by sputtering, the most favoured method for its film quality and compatibility with large-scale productionSourced
  • What the problem is: the 2026 edition of the US Geological Survey (USGS) Mineral Commodity Summaries says ITO production continues to account for most of global indium consumption, and puts world refinery production (2025 estimate) at about 1,100 t, of which China accounts for 70%Sourced
The single most important line in this article

Putting a transparent electrode on top of a perovskite means exposing the perovskite to a plasma. Magliano et al. summarise the damage: high-energy particles break chemical bonds, UV light from the plasma degrades organic materials, and the rise in substrate temperature imposes thermal stressSourced. That is why many cells have an extra layer whose only job is protection before the electrode.

2. Where they are used in the cell

Where transparent conductive oxides sit: three uses (concept) Light blue bands are the TCO. Layer thicknesses are not to scale (this article's framing) (1) Bottom electrode (2) Top electrode (3) In the recombination layer Perovskite Glass Perovskite Lower cell Perovskite Silicon Made first, on glass or film; high temperatures are allowed Made later, on top of the perovskite. Yellow = buffer layer Connects the upper and lower cells; must be thin and clear Note: the buffer in (2) follows Magliano et al. [Ref. 2] and Bush et al. [Ref. 4]; (3) follows Hooijer et al. [Ref. 7]. Note: the high-temperature step on FTO in (1) follows Chen et al. [Ref. 9]. Thicknesses and colours are schematic. Note: the split into three is this article's own. A single cell can include more than one of these uses.
Fig. 1 Concept diagram (vector drawing). The use at each position follows Magliano et al. [Ref. 2], Bush et al. [Ref. 4], Hooijer et al. [Ref. 7] and Chen et al. [Ref. 9]. The split into three is this article's own framing, and layer thicknesses and colours are schematic.

The properties required differ considerably between the three positions (this article's summary).

  • (1) Bottom electrode: made first, on glass or film, so there is no risk of damaging layers underneath. In 2025 Chen et al. made NiO on FTO glass by spray pyrolysis at 500 °CSourced
  • (2) Top electrode: sputtered afterwards, on top of the finished perovskite. This is where damage occurs (Section 5)
  • (3) Recombination layer: Hooijer et al. note that the TCO sits at the centre of high-performance monolithic structures because it handles lateral conduction, protection from solvents, binding sites for SAMs, and recombinationSourced

3. How ITO, IZO, FTO and others differ

MaterialDeposition temperature (as reported)What the primary sources say
ITO (tin-doped indium oxide)Can be made at room temperatureThe most widely used transparent electrode in flexible perovskite cells, because of low-temperature processing, chemical stability towards perovskites, high transparency, matching band structure and mature mass-production processes. On the other hand, it is noted to crack when bent, to have low near-infrared transmission, and to have chemical-stability problems with PEDOT:PSS and with acids and bases. By room-temperature DC sputtering it is amorphous, with higher sheet resistance and lower transmittance
IZO (indium zinc oxide)Can be made at 100 °COne comparison finds IZO has lower sheet resistance and higher fill factor than ITO, while ITO has higher transmittance and larger short-circuit current. Like ITO, it cracks readily when bent
FTO (fluorine-doped tin oxide)Above 350 °CThe deposition temperature is too high for polymer films. Used as the bottom electrode on glass substrates
AZO (Al-doped zinc oxide)Can be made at room temperatureEvidence has been reported that it interacts chemically with perovskite films and lowers performance
In2O3:H (hydrogenated indium oxide)Room-temperature sputteringA high-mobility TCO. Deposited at room temperature without post-annealing, a 149 nm amorphous film had a mobility of 51.3 cm²/(V·s) and a sheet resistance of 25.7 Ω/□ (Fu et al.)

All Sourced (ITO, IZO, FTO, AZO: the review by Xu et al. [Ref. 6]; In2O3:H: Fu et al. [Ref. 5]). Deposition temperatures are literature values cited in the review by Xu et al. and vary with equipment and conditions.

The review by Chavan et al. (2023) cites an example of ITO made by DC and RF magnetron sputtering at 27 °C with 80 to 85% transmittance and a sheet resistance of 20 to 25 Ω/□ at about 300 nm thicknessSourced. The same review lists the drawbacks of ITO as high cost, insufficient transmission from blue to near-ultraviolet, chemical instability, a weak barrier against ions, and mechanical brittleness, and cites reports that the sheet resistance of ITO rose above 180 °C on PET and at 235 °C on PENSourced.

4. A materials engineer's view (1): easing the transparency-conductivity trade-off through mobility

Why this matters for materials engineers: more carriers, or faster carriers

The sheet resistance of a transparent conductive film is the film's resistivity divided by its thickness. To lower it, either make the film thicker or lower the resistivity. But a thicker film absorbs more light, and uses more indium. Lowering resistivity means either more carriers (free electrons) or higher mobility, and more carriers make the film absorb more near-infrared light: this is the well-known tug of war in designing transparent conductive films (our commentary).

The hydrogenated indium oxide (In2O3:H) used by Fu et al. is described as a high-mobility TCO with excellent visible and near-infrared transmissionSourced. Here is a numerical comparisonOur calculation.

  • ITO (Chavan et al. example): 20 to 25 Ω/□ × 300 nm → resistivity of about 6.0 to 7.5×10⁻⁴ Ω·cm
  • In2O3:H (Fu et al. example): 25.7 Ω/□ × 149 nm → resistivity of about 3.8×10⁻⁴ Ω·cm
  • So In2O3:H achieves almost the same sheet resistance at about half the thickness (149 ÷ 300 = about 0.50)

Half the thickness means roughly half the indium for the same area, and less light absorbed (our commentary). Assumptions and limits: the two examples differ in fabrication conditions and in who measured them, and the ITO value is a single example cited in a review. Resistivity was converted by this article as sheet resistance × thickness.

Crystalline or amorphous also matters. The review by Xu et al. says ITO made by room-temperature DC sputtering is amorphous, with high sheet resistance, low transmittance, a rough surface and insufficient flexibility, and cites an example in which annealing up to 300 °C on colourless polyimide activated the Sn dopant and drove crystallisation, giving lower sheet resistance than on PETSourced. You want to crystallise it, but the substrate cannot take the temperature: on top of a perovskite or on a plastic film, this constraint is always there (our commentary).

5. Sputter damage — particles, UV light and heat

Three kinds of damage the layers below suffer during sputtering (concept) Based on the summary by Magliano et al. (2025). Arrows and symbols are schematic, not energies or amounts No protection With a buffer layer Plasma (sputter source) Plasma (sputter source) Particles UV light Heat Perovskite and organic transport layers Buffer layer (SnO2, AZO etc.) Perovskite and organic transport layers Bond breaking, photo-oxidation, thermal decomposition Blocks direct impacts and UV light Note: the three kinds of damage and the role of the buffer layer follow Magliano et al. [Ref. 2]. Note: heat travels through the film, so a buffer layer may not block it completely. The figure is a schematic. Note: the left/right split is this article's own framing.
Fig. 2 Concept diagram (vector drawing). The three damage pathways and the role of the buffer layer follow Magliano et al. [Ref. 2]. Arrows, layer thicknesses and colours are schematic and do not show energies or amounts of damage.

According to Magliano et al., in sputtering, high-energy ions in a plasma strike a target and the ejected atoms settle on the substrate. In the process, damage arises by three routesSourced:

  • High-energy particles hit the substrate and, if their energy exceeds the bond dissociation energy, break chemical bonds
  • UV light emitted by the plasma causes photo-oxidation and photochemical degradation of the perovskite and of exposed transport layers
  • Energy delivered during deposition raises the substrate temperature, and the thermal stress decomposes organic materials and alters crystal structures and interfaces

One countermeasure is “soft” sputtering that controls pressure, deposition rate, deposition time, substrate temperature and target-to-substrate distance, but the paper notes that equipment geometry and constraints mean this is not always possible, and that it can make deposition impractically slow or undermine reproducibilitySourced.

Hooijer et al. also list soft sputtering, which lowers particle energy with RF power, low power and high pressure, and reactive plasma deposition (RPD) as lower-damage alternatives, while judging sputter damage to be a known problem but manageable at production scale, with sputtering widely establishedSourced.

6. Buffer layers — one extra layer to prevent damage

ReportBuffer layerWhat was published
Bush et al. (Advanced Materials, 2016)Solution-processed oxide nanoparticlesA nanoparticle buffer layer protected the layers below, giving 12.3% for a semi-transparent cell with sputtered ITO and 18.0% in a mechanically stacked tandem with silicon. The T80 lifetime when operated at the maximum power point at 100 °C in air without encapsulation was 124 hours
Bush et al. (Nature Energy, 2017)ALD tin oxideAn ALD tin oxide buffer layer prevented shunts with negligible parasitic absorption and made sputtering of a transparent top electrode possible. 23.6% for a two-terminal tandem on 1 cm². The window layer doubled as a diffusion barrier and withstood a damp-heat test of 1,000 hours at 85 °C and 85% RH
Magliano et al. (ACS AMI, 2025)Solution-processed AZO nanoparticlesCompared ZnO, AZO and SnO2 nanoparticle dispersions; AZO was best for resistance to sputter damage, crystallinity and UV shielding. Up to 18.1% for semi-transparent cells and up to 25.3% for two-terminal tandems with silicon

All Sourced (Bush et al. [Refs. 3 and 4], Magliano et al. [Ref. 2]). All are values reported in the papers, with differing measurement conditions and areas.

The review by Xu et al. gives examples of buffer layers in flexible cells, ALD VOx films, evaporated MoOx films and solution-processed AZO films, and names facing-target sputtering, which confines the plasma energy, as another effective methodSourced.

7. A materials engineer's view (2): protective layers are also chosen for speed

Conceptual image of a very thin clear film over a dark flat surface against a dark background, with a single smooth reflection picked out by light from an angle
Fig. 3 AI-generated concept image. An impression of a very thin protective film that shields the layer below. It does not represent a real film thickness, material, colour or product.
Why this matters for materials engineers: ALD wins on quality, coating competes on speed and cost

The most common buffer layer in p-i-n cells is ALD SnO2Sourced. It is dense and covers steps well. Even so, Magliano et al. point out that the extremely low deposition rate of ALD is a major obstacle to industrialisation, compounded by expensive precursors and poor material utilisation, and that careful tuning of conditions is also needed to avoid degrading the perovskiteSourced.

This is where coating commercial dispersions of metal-oxide nanoparticles comes in. Magliano et al. compared commercial ZnO, AZO and SnO2 dispersions and found AZO bestSourced. Hooijer et al. likewise say solution-processed metal-oxide nanoparticles are a scalable way to prevent sputter damage, but are limited to flat surfaces onlySourced.

Here is an opening for materials suppliers. What is wanted is a nanoparticle dispersion in a solvent that does not dissolve the perovskite below, that forms a dense film at low temperature, blocks UV, conducts electrons and survives sputtering from above. Dispersion stability, particle size, surface modification, solvent selection: this is ink and pigment technology (our commentary). Whether such layers can also follow textured surfaces such as those of silicon tandems remains the next challenge (as Hooijer et al. point out).

8. Indium — concentrated supply, with our calculation

ItemWhat USGS Mineral Commodity Summaries 2026 says
World refinery production1,090 t in 2024 and 1,100 t in 2025 (estimated), both rounded. Refinery capacity (2025 estimate) 1,700 t
By countryChina 760 t, Republic of Korea 180 t, Japan 65 t, Canada 40 t and others (2025 estimates). China accounts for 70% of the world total
Export controlsIn February 2025 China's Ministry of Commerce imposed new export controls on critical minerals including indium
End usesITO production continues to account for most of global indium consumption, mainly as conductive coatings on flat panels such as liquid crystal displays
ResourcesRecovered mainly from zinc sulfide ores (sphalerite). The indium content of zinc deposits from which it is recovered ranges from less than 1 ppm to 100 ppm
SubstitutesCarbon nanotubes and graphene as alternatives to ITO electrodes in solar cells, PEDOT for flexible applications, and copper or silver nanowires for touch panels, among others, are being developed or considered

All Sourced (USGS Mineral Commodity Summaries 2026, “Indium” [Ref. 1]). Country figures include estimates, and USGS itself notes that data are limited for many countries.

Our calculation: how many grams of indium stay in the film per square metre?
  • Assumptions: take the ITO film density as 7.1 g/cm³ and the composition as 90 wt% In2O3 and 10 wt% SnO2. Both are general assumptions set by this article, not values from its references
  • Mass fraction of indium in In2O3: 2×114.82 ÷ (2×114.82 + 3×16.00) = about 0.827Our calculation
  • Volume of a 100 nm film over 1 m²: 1 m² × 100 nm = 0.1 cm³ → mass 0.71 g
  • Indium: 0.71 × 0.90 × 0.827 = about 0.53 g per m² (per 100 nm of thickness)Our calculation
  • Per GW: at an assumed module efficiency of 25%, 1 m² = 250 W, so 1 GW = 4 million m² → 0.53 g × 4 million = about 2.1 t (per 100 nm of thickness)Our calculation
  • Against world refinery production: 2.1 ÷ 1,100 = about 0.19% (per GW, per 100 nm of thickness)Our calculation

Assumptions and limits: this is only the indium that stays in the film. In sputtering, only part of the target ends up in the film and some is deposited on chamber walls and elsewhere (that share is not covered in this article). The amount also depends on how many transparent conductive layers a cell has and how thick each one is (Hooijer et al. give about 40 to 70 nm for the TCO in a recombination layerSourced). There may be little indium in the film, but 70% of the supply is concentrated in one country and export controls have been imposed: quantity and procurement need to be considered as separate problems (our commentary).

Indium left in an ITO film (our calculation) Values per 100 nm of thickness. Density, composition and module efficiency are our assumptions about 0.53 g Indium per 1 m² Assumes 7.1 g/cm³, In2O3 90 wt% Our calculation about 2.1 t Per 1 GW Assumes 25% module efficiency Our calculation about 0.19% Of world refinery production Divided by 1,100 t (2025 estimate) Our calculation Small in quantity, yet 70% of supply is in China, with export controls since February 2025 Quantity and procurement need to be considered as separate problems Note: 1,100 t, 70% and the export controls follow USGS Mineral Commodity Summaries 2026 [Ref. 1]. Note: 0.53 g, 2.1 t and 0.19% are this article's calculations, not published figures. They cover only what stays in the film, not material lost during sputtering. Values change with the number and thickness of layers. Note: the density of 7.1 g/cm³, 90 wt% In2O3 and 25% efficiency are our assumptions.
Fig. 4 Drawing that includes our calculation (vector drawing). World refinery production, China's share and the export controls follow USGS Mineral Commodity Summaries 2026 [Ref. 1]. The figures of 0.53 g per m², 2.1 t per GW and 0.19% were calculated by this article and are not published figures. The ITO density, composition and module efficiency are our own assumptions, and material lost during sputtering is not included.

There is also research aimed at doing without indium. Hooijer et al. say zinc tin oxide (ZTO) has been demonstrated as an indium-free recombination layer, and that aluminium-doped zinc oxide and doped titanium oxide are also worth consideringSourced. However, whether these will be used in mass production with performance and durability equal to indium-based materials could not be confirmed at the time of our researchNot yet confirmed.

9. Strengths and open problems

Strengths and challenges of transparent conductive oxides (this article's framing) Strengths Challenges Sputtering: good film quality, suits volume ITO can be deposited even at room temperature ITO-coated film is already a volume product High-mobility materials allow thinner films Buffer layers can prevent sputter damage Indium left in the film: about 0.5 g per m² Sputter particles, UV light and heat Room-temperature films: amorphous, high resistance ITO and IZO crack when bent ALD buffer layers are slow and costly Low near-infrared transmission 70% of indium supply is in China Note: items follow the sources [Refs. 1 to 9] and our calculations (Sections 4 and 8). Note: the split into two groups is this article's own and is not an established industry assessment.
Fig. 5 Concept diagram (vector drawing). Each item follows the sources [Refs. 1 to 9] and this article's calculations (Sections 4 and 8). Sorting them into strengths and challenges is this article's own framing, not an established industry assessment. The figure of about 0.5 g per m² is a calculated value for a 100 nm film, based on our own assumptions.

(1) Can the TCO in the recombination layer be made thinner?

Hooijer et al. point out that the TCO in the recombination layer (around 40 to 70 nm) causes optical loss, cracking, shunts, sputter damage and weaker adhesion, problems that grow with scale-up, while also citing a report that optimising reflection with ITO under 20 nm thick can recover more than 1 mA/cm² of currentSourced. Assuming an operating voltage of 1.7 V, 1 mA/cm² corresponds to about 1.7 points of efficiencyOur calculation. How thin it can be made while staying stable in production cannot yet be confirmedNot yet confirmed.

(2) Long-term durability

Bush et al. (2017) reported that, with the window layer doubling as a diffusion barrier, cells withstood a test of 1,000 hours at 85 °C and 85% RHSourced. How many years cells last outdoors, however, could not be confirmed in the primary sources within the scope of this articleNot yet confirmed.

The article in summary
  • Transparent conductive oxides are used in three positions: bottom electrode, top electrode and recombination layer (this article's framing)
  • ITO can be deposited at room temperature and its production processes are mature, but it cracks easily, and room-temperature films are amorphous with higher resistanceSourced
  • In one example, high-mobility In2O3:H matched the sheet resistance of ITO at about half the thicknessOur calculation
  • Sputtering damages the perovskite through particles, UV light and heat; buffer layers of ALD SnO2 or AZO nanoparticles protect itSourced
  • Indium left in an ITO film is about 0.53 g per m² at 100 nm, or about 2.1 t per GW (with assumptions)Our calculation. 70% of the supply is in ChinaSourced

10. Glossary

Transparent conductive oxide (TCO)
A thin oxide film that transmits visible light and conducts electricity, such as ITO, FTO, IZO and AZO.
ITO
Tin-doped indium oxide. The most widely used transparent conductive film.
IZO
Indium zinc oxide. A transparent conductive film that can be deposited at low temperature.
FTO
Fluorine-doped tin oxide. Made at high temperature and used on heat-resistant glass substrates.
AZO
Aluminium-doped zinc oxide. Uses no indium.
Sheet resistance (Ω/□)
The in-plane resistance of a thin film: resistivity divided by thickness.
Mobility
How easily charge moves under an electric field. The higher it is, the fewer carriers are needed to conduct.
Sputtering
A deposition method in which ions in a plasma strike a target and the ejected atoms settle on the substrate.
Sputter damage
Degradation of the layers below by particles, UV light and heat during sputtering.
Buffer layer
A thin layer inserted to protect the layers below from later steps such as sputtering.
ALD (atomic layer deposition)
A deposition method that builds a film one atomic layer at a time by alternating precursor gases. Dense but slow.
Reactive plasma deposition (RPD)
A method that evaporates and activates the source material with a plasma. Regarded as causing less damage.
Amorphous
A state without the regular atomic arrangement of a crystal.

11. References (primary sources)

  1. U.S. Geological Survey “Mineral Commodity Summaries 2026: Indium”, February 2026 (PDF) https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-indium.pdf
  2. Magliano, E. et al. “Solution-Processed Metal-Oxide Nanoparticles to Prevent The Sputtering Damage in Perovskite/Silicon Tandem Solar Cells”, ACS Applied Materials & Interfaces 17, 17599–17610 (2025, open access) https://doi.org/10.1021/acsami.5c00090
  3. Bush, K.A. et al. “Thermal and Environmental Stability of Semi-Transparent Perovskite Solar Cells for Tandems Enabled by a Solution-Processed Nanoparticle Buffer Layer and Sputtered ITO Electrode”, Advanced Materials 28, 3937–3943 (2016) https://doi.org/10.1002/adma.201505279
  4. Bush, K.A. et al. “23.6%-efficient monolithic perovskite/silicon tandem solar cells with improved stability”, Nature Energy 2, 17009 (2017) https://doi.org/10.1038/nenergy.2017.9
  5. Fu, F. et al. “Low-temperature-processed efficient semi-transparent planar perovskite solar cells for bifacial and tandem applications”, Nature Communications 6, 8932 (2015, open access) https://doi.org/10.1038/ncomms9932
  6. Xu, Y. et al. “Recent Progress of Electrode Materials for Flexible Perovskite Solar Cells”, Nano-Micro Letters 14, 117 (2022, open access) https://doi.org/10.1007/s40820-022-00859-9
  7. Hooijer, R. et al. “Why more junctions do not yet deliver: interconnection challenges in perovskite multijunction solar cells”, Energy & Environmental Science 19, 4582–4596 (2026, open access) https://doi.org/10.1039/d6ee01631f
  8. Chavan, G.T. et al. “A Brief Review of Transparent Conducting Oxides (TCO): The Influence of Different Deposition Techniques on the Efficiency of Solar Cells”, Nanomaterials 13, 1226 (2023, open access) https://doi.org/10.3390/nano13071226
  9. Chen, M. et al. “Low-temperature sequential deposition for efficient inverted perovskite solar cells”, Nature Communications 16, 5746 (2025, open access) https://doi.org/10.1038/s41467-025-61144-y

12. Claim-to-source audit

Claim in the textBasisLabel
That ITO production continues to account for most of global indium consumption (mainly conductive coatings on flat panels). World refinery production of 1,090 t in 2024 and an estimated 1,100 t in 2025 (rounded), with refinery capacity of 1,700 t. China 760 t, Republic of Korea 180 t, Japan 65 t, Canada 40 t and others, with China at 70% of the world total. China's export controls of February 2025. Recovery from sphalerite and indium contents from less than 1 ppm to 100 ppm. Substitute materials (carbon nanotubes and graphene as alternatives to ITO in solar cells, PEDOT, copper and silver nanowires). That data are limited for many countriesUSGS Mineral Commodity Summaries 2026, “Indium”. Reference 1 https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-indium.pdfSourced
That ITO, AZO and IZO are the most common transparent electrodes in semi-transparent cells because of their high transparency and mobility. That TCOs are usually deposited by sputtering, the most favoured method for its film quality and compatibility with large-scale production. The three sputter-damage pathways (bond breaking by high-energy particles, photo-oxidation and photochemical degradation by plasma UV, thermal stress from substrate heating). Soft sputtering by control of pressure, rate, time, substrate temperature and distance, and its limits. That ALD SnO2 is the most common buffer layer in p-i-n cells. The slowness of ALD, expensive precursors, poor material utilisation and the need for careful tuning of conditions. That AZO was best in a comparison of commercial ZnO, AZO and SnO2 dispersions. 18.1% for semi-transparent cells and 25.3% for two-terminal tandemsMagliano et al. (ACS Applied Materials & Interfaces, 2025). Reference 2 https://doi.org/10.1021/acsami.5c00090Sourced
That a solution-processed oxide nanoparticle buffer layer enabled sputtered ITO, giving 12.3% for semi-transparent cells and 18.0% for mechanically stacked tandems, with a T80 lifetime of 124 hours at the maximum power point at 100 °C in air without encapsulationBush et al. (Advanced Materials, 2016). Reference 3 https://doi.org/10.1002/adma.201505279Sourced
That an ALD tin oxide buffer layer prevented shunts with negligible parasitic absorption and enabled sputtering of a transparent top electrode. 23.6% for a two-terminal tandem on 1 cm². That the window layer doubled as a diffusion barrier and withstood 1,000 hours at 85 °C and 85% RHBush et al. (Nature Energy, 2017). Reference 4 https://doi.org/10.1038/nenergy.2017.9Sourced
That In2O3:H is a high-mobility TCO with excellent visible and near-infrared transmission. That a 149 nm amorphous film sputtered at room temperature without post-annealing had a mobility of 51.3 cm²/(V·s) and a sheet resistance of 25.7 Ω/□Fu et al. (Nature Communications, 2015). Reference 5 https://doi.org/10.1038/ncomms9932Sourced
That FTO is deposited above 350 °C and cannot be used on polymer films, that IZO can be made at 100 °C, and ITO and AZO at room temperature. That AZO interacts chemically with perovskites. That IZO has lower sheet resistance and higher FF while ITO has higher transmittance and larger Jsc. That IZO also cracks when bent. The reasons ITO is the most widely used in flexible cells (low-temperature deposition, chemical stability, transparency, band alignment, mature mass-production processes). ITO cracking under bending, low near-infrared transmission, and chemical-stability issues with PEDOT:PSS, acids and bases. That room-temperature DC-sputtered ITO is amorphous, with high resistance, low transmission, roughness and insufficient flexibility. The example of annealing up to 300 °C on colourless polyimide activating the Sn dopant and driving crystallisation, with lower sheet resistance than on PET. Buffer layers of ALD VOx, evaporated MoOx and solution-processed AZO, and facing-target sputteringXu et al. (Nano-Micro Letters, 2022). Reference 6 https://doi.org/10.1007/s40820-022-00859-9Sourced
That the TCO in the recombination layer handles lateral conduction, protection from solvents, SAM binding sites and recombination. Soft sputtering (RF, low power, high pressure) and reactive plasma deposition. That sputter damage is a known problem but regarded as manageable at production scale. That solution-processed metal-oxide nanoparticles are scalable but limited to flat surfaces. That ZTO has been demonstrated as an indium-free recombination layer, and that AZO and doped titanium oxide are also worth considering. Optical loss, cracking, shunts, sputter damage and weaker adhesion from the recombination-layer TCO (around 40 to 70 nm). The report that optimising reflection with ITO under 20 nm can recover more than 1 mA/cm²Hooijer et al. (Energy & Environmental Science, 2026). Reference 7 https://doi.org/10.1039/d6ee01631fSourced
ITO by DC and RF magnetron sputtering at 27 °C with 80 to 85% transmittance and 20 to 25 Ω/□ at about 300 nm. The drawbacks of ITO (high cost, insufficient blue to near-UV transmission, chemical instability, weak ion barrier, brittleness). Reports of sheet resistance rising above 180 °C for ITO on PET and at 235 °C on PENChavan et al. (Nanomaterials, 2023). Reference 8 https://doi.org/10.3390/nano13071226Sourced
That NiO was made on FTO glass by spray pyrolysis at 500 °CChen et al. (Nature Communications, 2025). Reference 9 https://doi.org/10.1038/s41467-025-61144-ySourced
The resistivities of ITO and In2O3:H (about 6.0 to 7.5×10⁻⁴ Ω·cm and about 3.8×10⁻⁴ Ω·cm) and the thickness ratio of about 0.50. The indium mass fraction of about 0.827 in In2O3, about 0.53 g of indium in a 100 nm, 1 m² ITO film, about 2.1 t per GW and about 0.19% of world refinery production. 1 mA/cm² as about 1.7 pointsOur calculation. The ITO density of 7.1 g/cm³, composition of 90 wt% In2O3, module efficiency of 25% and operating voltage of 1.7 V are assumptions set by this article, not values from the references. Only the amount remaining in the film is counted, not material lost in sputtering. The resistivity comparison uses two examples made under different conditionsOur calculation
Adoption of indium-free transparent conductive films in mass production, how thin the recombination-layer TCO can be made while remaining stable in production, and long-term outdoor durabilityNot stated because they could not be confirmed in the primary sources within the scope of this article (commentary by this article)Not yet confirmed
Dividing transparent conductive films into three positions. The tug of war between transparency and conductivity, and the general design relationship that more carriers mean more near-infrared absorption. The reading that half the thickness means roughly half the indium. The framing of the constraint that you want to crystallise but the substrate cannot take the heat. Reading nanoparticle buffer-layer development as ink and pigment technology. The framing that quantity and procurement should be considered separately. The two-way split into strengths and challengesThis article's own framing and commentary based on published content. The relationship between carrier concentration and near-infrared absorption is this article's general explanation, not a direct citation from the referencesCommentary
That Figs. 1, 2, 4 and 5 are explanatory drawings, and that the hero image and Fig. 3 are AI-generated imagesA note by this article (commentary)Commentary

Last updated 25 September 2026. Sources are limited to primary material (the USGS Mineral Commodity Summaries and peer-reviewed papers). No market estimates from research firms are used. The utilisation efficiency of sputtering targets, adoption of indium-free transparent conductive films in mass production, and long-term outdoor durability are not stated, because they could not be confirmed in published primary sources. The calculation in Section 8 rests on this article's own assumptions for ITO density, composition and module efficiency, and shows only the amount remaining in the film. All figures are for explanation. Figs. 1, 2, 4 and 5 are vector drawings, and the hero image and Fig. 3 are AI-generated images; none of them shows a real cross-section, micrograph or physical product.

🌐 Japanese