Solution-processed OLEDs and organic solar cells are stacks of layers a few to a few hundred nanometres thick, and both the electrical and the optical behaviour of the device depend on those thicknesses. This guide explains what each layer does, which thicknesses the cited studies report, how to coat one layer on another without dissolving it, and a laboratory process flow from substrate cleaning to cathode.
In one paragraph: OLED and OPV stacks combine layers from a few nanometres (injection layers) to a few hundred nanometres (OPV active layers) [7][12]. In OPV, a thicker active layer absorbs more but collects charges less well; one P3HT:PCBM series peaked at 120 nm [12]. Optical interference means that even transparent layers change device output [14][15]. Multilayers need orthogonal solvents, cross-linking or oxide layers [8], and spin curves set the thickness of each layer [13][23].
1. Why Solution-Process OLED and OPV Layers
Organic light-emitting diodes (OLEDs) and organic photovoltaics (OPV) share a common architecture: a stack of very thin layers between two electrodes, at least one of them transparent. The first efficient OLED, reported by Tang and VanSlyke in 1987, was a double-layer structure of organic thin films prepared by vapor deposition [1]. Three years later, Burroughes and co-workers reported light-emitting diodes based on conjugated polymers [2], the class of materials that opened the route to solution-processed devices. On the photovoltaic side, Tang described a two-layer organic photovoltaic cell in 1986 and found that the interface between the two organic materials was crucial for the cell’s behaviour [3], and Yu and co-workers showed in 1995 that blending a semiconducting polymer with a fullerene acceptor forms a network of internal donor–acceptor heterojunctions that greatly improves carrier collection [4].
Solution processing is attractive because the layers can, in principle, all be coated from inks: reviews of OLED materials describe the potential of fully solution-processed devices for large-area and flexible products and for printable, roll-to-roll production [5], and printing and coating routes for polymer solar cells have been reviewed in detail [6]. In the laboratory, the same idea usually starts with spin coating on small substrates, followed by a vacuum-evaporated metal cathode.
This guide explains what each layer in a solution-processed stack does, which thickness ranges the cited literature reports, how labs build multilayers without dissolving the layer underneath, and a step-by-step laboratory process flow. Thickness and process numbers are quoted from the cited papers as examples, not as a recipe; materials, solvents and target values must be chosen for each device. Measuring the thickness of the layers is the subject of the companion article How to Measure Thin Film Thickness: Profilometry, Ellipsometry and Interference.
2. Device Stacks and What Each Layer Does
Both device types are built from the same functional layers, although the names differ:
| Layer | Role in an OLED | Role in an OPV cell | Typical deposition in a lab stack |
|---|---|---|---|
| Transparent anode (often ITO on glass) | Injects holes | Collects holes (conventional structure) | Bought as coated substrate; cleaned and surface-treated |
| Hole injection / transport layer (e.g. PEDOT:PSS) | Eases hole injection, smooths the anode | Hole-selective contact | Spin coated from an aqueous dispersion, then baked |
| Emissive layer (EML) / active layer | Where electrons and holes recombine and emit light | Absorbs light; donor–acceptor blend separates charges [4] | Spin coated from an organic solvent |
| Electron transport layer (ETL) | Transports electrons, can block holes | Electron-selective layer or optical spacer | Solution or vacuum, depending on materials |
| Electron injection layer (EIL) and cathode | Thin interlayer plus metal (e.g. LiF/Al) [10] | Thin interlayer plus metal [11] | Usually vacuum evaporation |
Simplified conventional (non-inverted) structures. Inverted structures reverse the electrode roles [27].
The papers listed here give concrete examples. Pu and co-workers fabricated an OLED with the structure ITO (130 nm)/PEDOT:PSS (40 nm)/cross-linked emissive layer (40 nm)/TPBi (40 nm)/NaF (4 nm)/Al (100 nm), in which all the organic layers were solution-processed and the NaF and Al were vacuum-deposited to form the cathode; a bilayer version used an 80 nm emissive layer directly under the cathode [7]. Shaban and co-workers built OPV cells with the structure ITO/PEDOT:PSS/P3HT:PC60BM/Mg-Al, with spin-coated organic layers and evaporated metal cathodes [13].
High-efficiency OLEDs generally need several layers to confine excitons and balance charge transport, which is exactly what makes solution processing hard: each new layer must be coated without disturbing the one below [8]. Geffroy and co-workers review the materials, device structures and display technologies involved [9].
3. Thickness Targets: What the Literature Reports
Device layers are measured in nanometres, and small changes matter. The table collects thickness values reported in the cited studies. They show the orders of magnitude involved; they are not universal targets.
| Layer | Reported thickness | Context | Source |
|---|---|---|---|
| ITO anode | 130 nm | Transparent electrode in the cited multilayer OLED stack; not a solution-deposited organic layer | [7] |
| PEDOT:PSS | 40 nm | Same device | [7] |
| Cross-linked emissive layer | 40 nm (multilayer) or 80 nm (bilayer) | Same study | [7] |
| Solution-processed ETL (TPBi) | 40 nm | Same device | [7] |
| NaF electron injection layer | 4 nm (vacuum-deposited) | Same device | [7] |
| Cs2CO3 electron injection layer | No thicker than 2 nm | Must fully react with evaporated Al | [7] |
| ZnO:Liq electron injection layer | About 10 nm still worked efficiently | Solution-processed, polymer LEDs | [7] |
| P3HT:PCBM active layer | 80–340 nm studied; best efficiency at 120 nm | Thickness series | [12] |
| P3HT:PC60BM active layer | About 290 nm at 900 rpm to about 160 nm at 3000 rpm | Spin-speed series | [13] |
| Al cathode | 100 nm | Vacuum-deposited | [7] |
Values as reported in the cited papers; the right thickness depends on materials, architecture and the optical design of the device.
Very thin interlayers
Injection layers are the extreme case. Hung and co-workers used a bilayer electrode of an ultrathin LiF layer next to the electron-transporting layer and an aluminium outer layer, and observed enhanced electron injection [10]. Brabec and co-workers reported that thin LiF interlayers under the metal electrode of bulk heterojunction solar cells significantly enhanced the fill factor [11]. Solution-processed injection layers can be even more demanding: Pu and co-workers note that a Cs2CO3 layer must be no thicker than 2 nm, which makes a uniform thickness and reproducible performance difficult, whereas a ZnO:Liq film of about 10 nm still worked efficiently [7]. A material whose performance is less sensitive to thickness is often the more practical choice for a solution process.
The active layer: absorption against collection
In OPV, a thicker active layer absorbs more light, but charges have farther to travel before they are collected. Lee and co-workers characterised P3HT:PCBM cells with active layers from 80 to 340 nm: the short-circuit current increased slightly and then saturated with thickness, while the fill factor dropped sharply above 120 nm, so that the best efficiency in their series was obtained at 120 nm [12]. In OLEDs, Kumar and Pereira compared solution-processed two-organic-layer devices with different emissive and electron transport layer thicknesses, processed from chlorobenzene, dichlorobenzene and chloroform, and discuss solvent, organic layer thickness and host together as factors in device characteristics [17].
Thickness is also optics
Optical interference within a multilayer device depends on the optical path lengths, refractive indices, interfaces and electrode structure. A layer that absorbs little light can still alter the field distribution and device response. Pettersson and co-workers modelled photocurrent spectra of organic photovoltaic devices by calculating the optical field inside the stack from complex refractive indices and layer thicknesses determined by spectroscopic ellipsometry [14]. Kim and co-workers inserted a solution-processed titanium oxide layer as an “optical spacer” between the active layer and the aluminium electrode and reported about a 50% increase in short-circuit current, attributed to the redistributed light intensity inside the device [15]. Gilot and co-workers found excellent agreement between optical modelling and experiment for ZnO optical spacers in P3HT:PCBM devices [16]. The practical message: the thickness of a layer that does not itself absorb can still change device output, so every layer should be measured, not only the active one.
4. Coating One Layer on Another: Orthogonal Solvents and Cross-Linking
In a vacuum-deposited stack, each new layer simply condenses on the previous one. In a solution process, the solvent of the next ink can dissolve, swell or mix with the layer underneath. The literature describes three main strategies [8]:
- Orthogonal solvents. Choose the next ink’s solvent so that it does not dissolve the layer below. Water/alcohol-soluble conjugated polymers can be processed from water or other polar solvents, which offers a way to avoid interfacial mixing when multilayer devices are made by solution processing [18].
- Cross-linking. Coat a soluble material and then convert it into an insoluble network. Pu and co-workers used a thermally cross-linkable host and showed that a solution-processed electron-transporting layer could then be coated on top of the cross-linked emissive layer [7]. Müller and co-workers demonstrated multi-colour OLED displays made by solution processing [19].
- Metal oxides. Nanocrystalline oxide layers are a third route [8]. Pu and co-workers note that some oxide films need high annealing temperatures, which limits where in the stack they can be placed [7].
Each strategy brings its own process step (a solvent choice, a cross-linking bake or a high-temperature anneal) that has to fit the thermal and chemical tolerance of the layers already on the substrate. A recent critical review of solution-processed OLEDs points out that selecting compatible materials for each layer is complicated by inconsistent documentation, solvent interactions and limited reproducibility across the literature, and argues for a structured, reproducibility-oriented approach [20]. In practice, that means documenting solvents, concentrations, spin programs and bake conditions for every layer.
5. A Laboratory Process Flow
The sequence below summarises a typical laboratory flow for a conventional small-area device. It is a framework for planning and documentation; the materials, values and atmosphere at each step come from the chosen materials system and the literature for it.
- Substrate cleaning. Patterned ITO glass is cleaned, typically in a sequence of detergent and solvents. Shaban and co-workers, for example, cleaned ITO ultrasonically for 15 min each in soapy water, toluene, acetone and isopropanol [13]. A full protocol is described in Substrate Cleaning Protocol for Thin-Film Devices: Solvents, UV-Ozone, Plasma and Queue Time.
- Surface treatment. The condition of the ITO surface matters: Kim and co-workers related different ITO treatments to device performance in polymer LEDs [21], and Sugiyama and co-workers showed that the ITO work function depends on the surface cleaning method [22]. UV-ozone and oxygen plasma are the usual options; see UV-Ozone vs Oxygen Plasma for ITO and Glass: Which Clean Before Coating?. Coat soon after treatment and keep the queue time constant.
- Hole injection layer. The aqueous PEDOT:PSS dispersion is filtered, spin coated and baked. In the OPV study of Shaban and co-workers, PEDOT:PSS was spin coated at 3000 rpm and annealed at 178 °C for 15 min in air before the substrates were moved into a glovebox [13]. Wetting problems at this step are discussed in Why Your Ink Dewets: Surface Energy, Contact Angle and Wetting Before You Coat.
- Emissive or active layer. Spin coated from an organic solvent, usually in an inert atmosphere. Thickness is set mainly by solution concentration and spin speed (next section).
- Post-deposition bake or anneal. For OPV blends, annealing changes the morphology of the bulk heterojunction. Ma and co-workers reported thermally stable cells after post-production annealing at 150 °C and attributed the improvement to annealing-induced changes in the bulk heterojunction, including improved nanoscale morphology and increased polymer crystallinity [24]; Li and co-workers reported high-efficiency cells obtained through self-organization of polymer blends [25]; Shaban and co-workers found 160 °C for 5 min optimal in their annealing series from 130 to 190 °C [13]. Hot plate steps in general are covered in Soft Bake, Post-Bake and Annealing on a Hot Plate: What Each Step Does, and annealing under nitrogen in Annealing Under Nitrogen: Inert Hot Plates for Air-Sensitive Films.
- Further solution layers. ETL or EIL from an orthogonal solvent or on a cross-linked layer (previous section) [7][18].
- Cathode. Thin interlayer and metal electrode, usually by thermal evaporation through a shadow mask [7][13].
- Protection and measurement. Organic devices are sensitive to their environment; the stability and degradation of polymer solar cells are reviewed by Jørgensen and co-workers [26]. Devices are therefore usually tested in an inert atmosphere or protected before testing; joining and sealing variables are discussed in Heat Sealing and Vacuum-Assisted Joining: Process Variables for Thin-Film Samples.
Two habits make this flow reproducible: measure the thickness of each new layer on witness samples (see How to Measure Thin Film Thickness: Profilometry, Ellipsometry and Interference), and keep the time between steps, especially between surface treatment and coating and between coating and baking, as constant as the process parameters themselves.
6. Hitting a Thickness with Spin Coating
Spin coating is widely used to prepare small-area organic layers. For a given solution, final thickness often decreases with increasing spin speed, although the measured relationship depends on the formulation and process conditions [23]. Shaban and co-workers reported about 290 nm at 900 rpm and about 160 nm at 3000 rpm for their P3HT:PC60BM series, describing the trend as a good approximation to the reciprocal-square-root relationship [13]. Use the concentration and preparation details specified in the paper’s methods when comparing or reproducing these results.
Because the trend is approximately a power law, a short speed series on your own solution is a practical way to hit a target thickness:
- coat the solution at three to five speeds across the usable range;
- measure each film (see How to Measure Thin Film Thickness: Profilometry, Ellipsometry and Interference);
- plot thickness against speed on log–log axes and read the speed for the target;
- if the target lies outside the range, change the concentration rather than pushing the speed to an extreme.
How to build and use such a curve is explained in Spin Speed vs Film Thickness: How to Build and Use a Spin Curve, and the stages and physics behind it in Spin Coating Explained: Stages, Physics and Film Thickness. Because interlayers of a few nanometres are near the limit of what a spin curve can control, they are often made from very dilute solutions and checked by device behaviour as well as by direct measurement [7].
7. Beyond Spin Coating and Further Reading
Spin coating wastes most of the ink and does not scale beyond small substrates. Printing and coating techniques for polymer solar cells are reviewed by Krebs [6], and fully printable OLEDs are a stated goal of solution-processing research [5]. Their thickness physics differs from spin coating: in slot-die coating, for example, the wet thickness is pre-metered by flow rate and coating speed, and in dip coating it depends on withdrawal speed. See Solution-Processed Thin Films: Choosing Spin, Dip, Slot-Die or Blade Coating for a comparison, Slot-Die Coating Explained: Pre-Metered Thickness and the Coating Window and Dip Coating Explained: Withdrawal Speed, Coating Regimes and Film Thickness for the individual methods, and From Spin Coating to Slot-Die and Blade: Scaling a Thin-Film Ink for transferring an ink from spin coating to a larger-area method.
Device architecture changes the flow as well. He and co-workers reported enhanced power-conversion efficiency in polymer solar cells using an inverted device structure [27], in which the electrode roles, and therefore the order and chemistry of the interlayers, are reversed.
Related case studies on this blog
8. Troubleshooting Solution-Processed Stacks
| Symptom | Possible causes | What to check |
|---|---|---|
| Lower layer dissolves or mixes when the next layer is coated | Solvent not orthogonal; incomplete cross-linking [7][8] | Solvent choice; cross-linking bake; rinse test on a witness sample |
| Device behaviour very sensitive to an interlayer | Interlayer must be extremely thin, e.g. Cs2CO3 no thicker than 2 nm [7] | Measure thickness; consider a less thickness-sensitive material [7] |
| OPV fill factor falls as the active layer gets thicker | Collection losses in thick films [12] | Thickness series; spin speed and concentration |
| Current changes when only a transparent layer changes | Optical interference in the stack [14][15] | Measure all layer thicknesses; optical model |
| Run-to-run drift | Queue times, atmosphere, undocumented solvent or bake changes [20][26] | Fixed step timing; record atmosphere; full process log |
| Dewetting, pinholes, comets | Surface energy, particles, dispensing | Surface treatment; filtering; see the defects guide |
Starting points summarised from the cited literature and general coating practice.
Coating defects in more depth are in Spin Coating Defects: Comets, Striations, Pinholes and Edge Beads Diagnosed; surface energy and wetting in Why Your Ink Dewets: Surface Energy, Contact Angle and Wetting Before You Coat.
9. Levon™ Coating, Cleaning and Hot Plate Equipment for Organic Device Layers
The tables quote the Levon™ product pages; “Not stated” means the page does not give the value. The EZ4-S, AC100-S and EZ6-S pages list thin-film research for perovskite solar cells and organic optoelectronic semiconductor devices among their application areas.
| Specification | EZ4-S | AC100-S | EZ6-S |
|---|---|---|---|
| Speed range | 100–9,999 rpm | Up to 12,000 rpm (no load); 1 rpm resolution | 100–12,000 rpm |
| Acceleration | 100–9,999 rpm/s | Up to 30,000 rpm/s | 100–30,000 rpm/s |
| Step time | Up to 3,000 s per step, 1 s resolution | Up to 3,000 s, 0.1 s resolution | Up to 3,000 s, 0.1 s resolution |
| Program memory | 5 programs × 5 steps | 100 recipes × up to 100 steps | 10 programs × 10 steps |
| Substrates and chucks | Maximum recommended substrate 4-inch round; vacuum chucks 10 mm, 25 mm and 2-inch | Chucks 10 mm, 25 mm and 2-inch | Chucks 10 mm, 25 mm and 2-inch |
| Glovebox | Glovebox-compatible design | Not stated | Not stated |
| Other published features | Aluminium housing (EZ4-S) or PP housing (EZ4-S-PP) | Waste bottle; integrated exhaust port for connection to external ventilation | 7-inch touchscreen; vacuum and lid safety interlock; upgradeable single-channel metering dispensing system |
| Dimensions (W × D × H) | 210 × 285 × 220 mm (EZ4-S-PP: 215 × 285 × 222 mm) | 340 × 410 × 385 mm; 15 kg | 295 × 386 × 250 mm; 12 kg |
Specifications as published on the Levon™ product pages on 24 October 2026. The supplied configuration is confirmed in the quotation. “No load” is stated by the AC100-S page.
| Specification | UC100-SE | HP6 |
|---|---|---|
| Function | UV-ozone cleaning | Hot plate |
| Key published values | 300 W U-shaped lamp, 185 nm + 254 nm; lamp lifetime ≥8,000 h | Ø170 mm plate for wafers up to 6 inches; RT to 250 °C; 0.1 °C resolution |
| Working area / control | Cleaning area ≤8 in; 35 mm illumination height; timer up to 99.99 h | Precision ±0.2 °C after the temperature is stable; uniformity ≤1% |
| Other published features | Pull-out, damped self-closing cleaning chamber; 5-level fan speed | Manual vacuum chuck (0.04–0.09 MPa); thermal-radiation-sealed lid; recessed design for glovebox integration |
| Dimensions (W × D × H) | 330 × 423 × 312 mm; 19 kg | 370 × 270 × 126 mm |
Specifications as published on the Levon™ product pages on 24 October 2026. The supplied configuration is confirmed in the quotation.
| Specification | SD200 | DP100-BE |
|---|---|---|
| Coating method | Slot-die | Dip (withdrawal) |
| Substrate | Up to 200 × 200 mm; thickness 0.05–10 mm | Up to 100 × 100 × 5 mm |
| Motion | Coating (Y) axis 0.002–50 mm/s over 0–355 mm; Z axis 0.001–20 mm/s over 0–50 mm, ±0.002 mm repeatability | 0.1–600 mm/min; stroke 0–100 mm; ±0.01 mm |
| Liquid supply / timing | Syringe pump 0–10 mL, 0–1,000 µL/s | Immersion and interval time 1–9,999.9 s; 1–999 cycles |
| Other published features | Granite platform 260 × 260 mm with vacuum adsorption; contact detection resolution ±1 µm | 100 recipes; 4.3-inch touchscreen; can be placed in a glovebox |
| Dimensions (W × D × H) | 820 × 600 × 680 mm; 100 kg | 227 × 350 × 416 mm |
Specifications as published on the Levon™ product pages on 24 October 2026. The supplied configuration is confirmed in the quotation.
How this connects to the guide: the EZ4-S, AC100-S and EZ6-S provide different speed, program and configuration options for laboratory coating. The UC100-SE is a UV-ozone cleaner, and the HP6 is a hot plate for heating and annealing. For the slot-die and dip coating routes in Section 7, the SD200 slot-die coater and DP100-BE dip coater provide alternatives to spin coating. Published specifications describe the instruments; they do not establish compatibility with a particular organic ink, multilayer stack or process atmosphere. Confirm the intended materials, utilities and configuration through the contact form.
10. Keep Exploring the Organic Electronics and Thin-Film Knowledge Hub
Start here
- Solution-Processed Thin Films: Choosing Spin, Dip, Slot-Die or Blade Coating — comparing spin, dip, slot-die and blade coating
- Spin Coating Explained: Stages, Physics and Film Thickness — stages and physics of spin coating
- Spin Speed vs Film Thickness: How to Build and Use a Spin Curve — building and using a spin curve
- Substrate Cleaning Protocol for Thin-Film Devices: Solvents, UV-Ozone, Plasma and Queue Time — substrate cleaning protocol
Process steps
- UV-Ozone vs Oxygen Plasma for ITO and Glass: Which Clean Before Coating? — UV-ozone or plasma before coating
- Soft Bake, Post-Bake and Annealing on a Hot Plate: What Each Step Does — soft bake, post-bake and annealing
- Annealing Under Nitrogen: Inert Hot Plates for Air-Sensitive Films — annealing under nitrogen
- How to Measure Thin Film Thickness: Profilometry, Ellipsometry and Interference — measuring the thickness of each layer
- Heat Sealing and Vacuum-Assisted Joining: Process Variables for Thin-Film Samples — heat sealing and vacuum-assisted joining
Links will be added as each article is published.
Equipment
- EZ4-S Spin Coater — compact, glovebox-compatible, up to 4-inch round
- AC100-S Programmable Spin Coater — 100 recipes, exhaust port
- EZ6-S Benchtop Spin Coater — 100–12,000 rpm, touchscreen
- UC100-SE UV Ozone Cleaner — 185 nm + 254 nm lamp
- HP6 Hot Plate — RT to 250 °C, glovebox integration
- SD200 Slot-Die Coater — substrates up to 200 × 200 mm
- DP100-BE Dip Coater — 0.1–600 mm/min
- Levon™ Spin Coaters — all spin coaters
Questions about a specific instrument can be sent through the contact form.
11. FAQ
What does solution processed OLED mean?
Some or all organic layers are coated from solution, typically by spin coating, instead of being evaporated in vacuum; the cathode is often still evaporated [5][7].
How thick are the layers in a solution-processed OLED?
In one cited device, PEDOT:PSS, emissive and electron transport layers were each 40 nm and the NaF injection layer 4 nm [7].
What is a good active layer thickness for a P3HT:PCBM solar cell?
It depends on the device; in one 80–340 nm series the best efficiency was at 120 nm, with the fill factor falling sharply above that [12].
How do you coat a second layer without dissolving the first?
Use an orthogonal solvent, cross-link the first layer, or use an oxide layer [7][8][18].
Why does the thickness of a transparent layer change the device?
Interference inside the stack depends on every layer’s thickness and refractive index, so a non-absorbing layer can still shift the optical field; spacer layers redistribute the light in the active layer [14][15][16].
How do I set the thickness of a spin-coated layer?
Build a spin curve: thickness falls roughly with the reciprocal square root of spin speed and rises with concentration [13][23].