Research Highlights

This study introduces an innovative azeotropic solvent strategy for large-area printing of self-assembled monolayers (SAMs) as hole selection layers in organic solar cells (OSCs). Using an azeotropic mixture of IPA and toluene, researchers successfully prepared high-quality, uniform, and stable Cbz-2Ph SAMs, effectively improving the work function of ITO substrates.

Based on the azeotrope-treated SAM, OSCs with fully printed p-i-n stack structures achieved efficiencies of 18.89% for small-area devices (0.04 cm²) and 17.76% for large-area devices (1.008 cm²), setting a new record for 1 cm² fully printed OSCs.

The azeotrope-treated SAM devices also demonstrated excellent stability, maintaining T80 lifetime exceeding 2000 hours under prolonged illumination, significantly outperforming traditional PEDOT:PSS-based devices (T80 = 169 hours).

Research Team

This research was conducted by Gengxin Du and colleagues at City University of Hong Kong, with corresponding authors Professor Hin-Lap Yip, Professor Alex K.-Y. Jen, and Professor Francis R. Lin.

Research Background

Organic solar cells (OSCs) have garnered significant attention due to their lightweight properties, adjustable optical transparency, solution processability, and potential for sustainable production. Recent advances in material design, device physics, and interface engineering have pushed single-junction OSC power conversion efficiency (PCE) beyond 20%.

However, scaling up OSCs remains challenging due to inherent limitations in carrier transport dynamics in organic materials and increased film uniformity and defect density issues in large-area production processes.

These factors severely limit the efficiency of large-area OSCs, hindering commercialization. Consequently, efforts are intensifying to develop large-area processing techniques (≥1 cm²) that minimize scaling losses and are compatible with future roll-to-roll production.

In conventional p-i-n structure OSCs, the hole selection layer (HSL) plays a crucial role in modifying anode work function, facilitating hole extraction, and templating active layer morphology.

However, the commonly used HSL PEDOT:PSS has several drawbacks, including non-ideal energy levels, acidity, hygroscopicity, and significant parasitic absorption, all of which compromise device stability and efficiency.

Furthermore, phase separation in PEDOT:PSS composites generates pinholes during device processing and operation, leading to electrical shorts, which is particularly critical in large-area devices. To address these challenges, researchers have explored using self-assembled monolayers (SAMs) as alternatives to PEDOT:PSS.

SAMs form strong chemical bonds on ITO substrates, providing tunable energy levels, higher optical transmission, and enhanced device stability.

Nevertheless, SAM processing still faces multiple challenges. In particular, processing SAM molecules (which are typically amphiphilic and form micelles in parent solution) remains challenging, especially on rough ITO surfaces.

While SAM-based OSC modules have been reported, achieving low defect density on ITO typically shows poor reproducibility and requires dynamic processing techniques (such as spin-coating) that are difficult to apply on large substrates. These limitations make manufacturing large-area OSCs using high-quality SAM HSLs extremely challenging.

Recent developments in co-SAM and co-solvent strategies have improved SAM quality in perovskite solar cells. The co-SAM strategy introduces an additive SAM to fill vacancies in the host SAM, achieving denser monolayer formation.

On the other hand, the co-solvent strategy introduces a small amount of high-solubility N,N-dimethylformamide (DMF) into isopropanol (IPA), which helps break down SAM micelles into dispersed molecules, providing an optimized and more controllable SAM deposition process.

However, differences in surface tension and saturated vapor pressure (see Supporting Information (SI), Tables S1 and S2) between co-solvents lead to instabilities, which are particularly problematic in large-area printing where extended processing times result in solvent composition deviation over time due to different evaporation rates.

This instability exacerbates issues such as Marangoni flow and hinders uniform film formation.

Moreover, residual high-boiling-point additives are difficult to remove and may affect the morphology of the OSC active layer processed on top.

Solution

To mitigate the issues arising from complex evaporation dynamics of mixed solvents in depositing high-quality SAMs on ITO substrates, this study introduces an azeotropic strategy using slot-die coating for SAM deposition.

Azeotropes maintain fixed solvent composition equilibrium at the azeotropic point. This study employed an azeotrope composed of toluene (42 wt%) and IPA (58 wt%), achieving better dispersion of Cbz-2Ph SAM molecules and enhanced shelf stability of SAM inks.

The azeotrope-treated SAM showed higher coverage on ITO substrates and achieved optimized energy alignment with the donor material (PM6). Based on the high-quality azeotrope-SAM, researchers printed PM6:BTP-eC9 bulk heterojunction (BHJ) layer and PNDIT-F3N electron selection layer (ESL), achieving 18.89% PCE, superior to IPA-treated SAM and spin-coated PEDOT:PSS devices. 

Furthermore, azeotrope-SAM devices exhibited higher operational stability with T80 lifetime exceeding 2000 hours. More encouragingly, on large-area devices (1.008 cm²) with fully printed p-i-n stacks, researchers achieved 17.76% PCE, marking the highest efficiency for 1-cm² devices with all functional layers printed (excluding top/bottom electrodes).

This research not only demonstrates the potential of SAM HSL-based printed OSCs but also suggests that the azeotropic strategy could revolutionize large-scale OSC manufacturing.

Experimental Process and Steps

1. Materials Preparation: All reagents were used as received without further purification. PM6 and BTP-eC9 were purchased from Solarmer. Isopropanol (IPA), toluene, and chlorobenzene (CB) were purchased from Sigma-Aldrich. PNDIT-F3N was purchased from eFlexPV Limited.
2. Cbz-2Ph SAM Synthesis: Synthesis steps are detailed in the supporting information.
3. Device Fabrication: Prior to use, ITO substrates were sequentially ultrasonicated in diluted detergent solution, deionized water, acetone, and isopropanol for 20 minutes each, followed by drying in an oven at 80°C. The substrates were treated with UV-ozone for 20 minutes before use. PEDOT:PSS was spin-coated onto ITO substrates (15 mm × 15 mm for 0.04 cm² devices; 25 mm × 25 mm for 1 cm² devices) and annealed at 150°C for 10 minutes in air.

SAM solution was prepared by dissolving SAM powder in IPA, toluene, or IPA and toluene azeotrope (weight ratio 0.58:0.42) at a concentration of 1 mg/ml and stirred for 2 hours before use.

All layers were printed using a slot-die coating platform under ambient conditions with controlled relative humidity (approximately 15%). SAM printing parameters were: speed 25 mm/s, gap between coating head and substrate 100 μm, flow rate 5 μL/min, and substrate temperature 80°C.

After annealing at 100°C for 10 minutes, substrates were transferred to a glovebox for temporary storage. PM6:BTP-eC9 (1:1.2) or PM6:BTP-eC9:L8BO-2F (1:0.96:0.24) blends were dissolved in chlorobenzene at 10 mg/ml for slot-die coating and 22 mg/ml for spin-coating (without any additives).

Solutions were stirred at 50°C for at least 1 hour. For slot-die coating of the active layer, printing parameters were: speed 50 mm/s, printing gap 150 μm, flow rate 10 μL/min, and substrate temperature 70°C.

For spin-coated active layers, the rotation speed was 2500 rpm. After coating, the active layer was annealed at 100°C for 5 minutes.

PNDIT-F3N was dissolved in methanol (containing 0.5 vt% acetic acid) at 0.5 mg/ml for spin-coating and 3 mg/ml for slot-die coating.

Slot-die coating parameters for PNDIT-F3N were: printing speed 5 mm/s, printing gap 100 μm, flow rate 5 μL/min, and room temperature substrate.

For devices using D18/BTP-eC9 as the active layer, D18 film was spin-coated from chlorobenzene onto the SAM or PEDOT:PSS layer at 2600 rpm, followed by BTP-eC9 spin-coated from chloroform at 2000 rpm onto the D18 layer.

The concentrations of D18 and BTP-eC9 in solvents were 6 mg/ml and 10 mg/ml, respectively. D18 and BTP-eC9 solutions were heated at 80°C and 45°C, respectively.

For spin-coated PNDIT-F3N, the rotation speed was 2000 rpm. Samples were then transferred to an evaporation chamber for Ag (100 nm) deposition. The effective area of OSCs was 0.04 cm², defined by the overlap area of anode and cathode.

For J-V performance measurements, a test mask with a precise area of 0.0324 cm² was used.

For 1 cm² devices, the coating head width of the slot-die coating platform was adjusted to 30 mm, while it was 12 mm for 0.04 cm² devices.

For 1 cm² devices, a mask design with 9 mm width (this value is larger than the width of sub-cells in common OPV or perovskite modules, which is typically 5-7 mm due to the balance between geometric fill factor and carrier transport) and 11.2 mm length was used to explore the feasibility of further scaling up.

Research Characterization

To evaluate how the azeotropic strategy affects SAM formation, researchers printed Cbz-2Ph SAM onto ITO substrates using a slot-die coating platform (SI, Figure S5).

Contact angle measurements were performed by dropping water on Cbz-2Ph SAM films to evaluate film quality. Ideally, all phosphonate groups should be anchored to the ITO surface with carbazole heads exposed outward, forming a dense, hydrophobic surface. As shown in Figures 1d and 1e, azeotrope-SAM exhibited larger contact angles compared to IPA-SAM, indicating denser and more uniform SAM coverage.

Current-Voltage (J-V) Characteristics Data and Chart Interpretation

The literature provides multiple data sets and charts related to current-voltage (J-V) characteristics to demonstrate the impact of the azeotropic strategy on OSC performance:

Figure 3b: J-V Curves for Different Preparation Methods

Chart Description: This figure shows J-V curves for three different devices: based on azeotrope-treated SAM (azeotrope-SAM), IPA-treated SAM (IPA-SAM), and spin-coated PEDOT:PSS devices.

Interpretation: The graph clearly shows that azeotrope-SAM-based devices demonstrate the best photovoltaic performance, with steeper J-V curves, higher open-circuit voltage (Voc), larger short-circuit current density (Jsc), and higher fill factor (FF). In comparison, IPA-SAM-based devices show inferior performance, even lower than PEDOT:PSS-based devices. This indicates that the azeotropic strategy effectively improves overall OSC performance.

Figure 3d: Dark Current for Different Preparation Methods

Chart Description: This figure compares the leakage current under dark conditions for devices based on azeotrope-SAM and IPA-SAM.

Interpretation: Azeotrope-SAM-based devices show lower dark current, indicating higher film quality with reduced defects and pinholes, effectively suppressing leakage current. This further demonstrates the advantages of the azeotropic strategy in forming dense, uniform SAM films.

Figure 4a: Large-Area Device J-V Curves

Chart Description: This figure shows J-V curves for two types of large-area devices (1 cm²): one using only azeotrope-SAM as HSL and another with all functional layers printed by slot-die coating (i.e., fully printed p-i-n stack).

Interpretation: Impressively, the fully printed p-i-n stack devices (PCE = 17.76%) even outperform devices using only azeotrope-SAM as HSL (PCE = 17.52%). This indicates that slot-die coating technology has advantages in large-area device fabrication, capable of forming more uniform films and reducing thickness non-uniformity caused by spin-coating on large-area substrates.

Tables 1 and 2: Detailed Photovoltaic Parameters for Different Preparation Methods

Table Description: Table 1 lists detailed photovoltaic parameters for small-area devices (0.04 cm²), including Voc, Jsc, FF, and PCE, comparing different HSL preparation methods. Table 2 lists detailed photovoltaic parameters for large-area devices (1 cm²), comparing devices using only azeotrope-SAM as HSL versus fully printed p-i-n stack devices.

Interpretation: Table 1 shows that azeotrope-SAM-based devices outperform IPA-SAM and PEDOT:PSS-based devices in all photovoltaic parameters, consistent with the J-V curve analysis in Figure 3b. Table 2 further demonstrates the feasibility of fully printed p-i-n stack devices in large-area OSC fabrication, with performance even surpassing devices using only azeotrope-SAM as HSL.

Figure S17 and Table S6: Device Performance with D18:BTP-eC9 as Active Layer

Chart Description: Figure S17 shows J-V curves for OSCs using D18:BTP-eC9 as active layer under different HSL preparation methods. Table S6 lists the detailed photovoltaic parameters for these devices.

Interpretation: These data demonstrate that the azeotropic strategy is applicable not only to PM6:BTP-eC9 active layers but also to other high-performance active layer materials such as D18:BTP-eC9. This highlights the broad applicability of the azeotropic strategy.

It is recommended to use Enlitech AM1.5G standard spectrum solar simulator for IV measurements

External Quantum Efficiency (EQE) Data and Chart Interpretation

Figure 3c in the literature shows EQE curves and integrated current lines for OSCs based on azeotrope-treated SAM (azeotrope-SAM) and IPA-treated SAM (IPA-SAM).

Interpretation:
From the chart, it can be observed that azeotrope-SAM-based devices exhibit higher EQE values in the 500-800 nm wavelength range.

Since the absorption in the active layer is almost identical between the two devices (Figure S15), the EQE enhancement suggests that ITO substrates modified with azeotrope-SAM can more effectively extract charge carriers, thereby improving the device’s short-circuit current density (Jsc).

Transient photocurrent measurement results (Figure S16a) further confirm this, showing shorter charge extraction time for azeotrope-SAM-based devices, indicating faster charge carrier transport from active layer to electrode.

Moreover, transient photovoltage measurement results (Figure S16b)show longer carrier recombination lifetime for azeotrope-SAM-based devices, meaning fewer charge carriers are lost to recombination during transport, further enhancing device performance.

For EQE measurements, it is recommended to use Enlitech QE-R Solar Cell Quantum Efficiency Optical System

Other Characterizations

• Atomic Force Microscopy (AFM): AFM height images show that azeotrope-SAM samples have rougher surfaces with root mean square (RMS) roughness of 2.42 nm, with morphology and RMS roughness closely matching bare ITO samples (SI, Figure S9, RMS roughness = 2.69 nm), indicating the formation of a conformal monolayer coating. IPA-SAM samples show smoother surfaces with RMS roughness of 1.79 nm, possibly due to multilayer film formation from pre-aggregated Cbz-2Ph molecules during printing.
  
• Kelvin Probe Force Microscopy (KPFM): Surface potential distribution results captured by KPFM clearly show more uniform surface potential distribution observed in azeotrope-SAM samples, while IPA-SAM shows relatively non-uniform surface potential distribution, indicating more random arrangement of Cbz-2Ph molecules in the film. Relative to bare ITO (SI, Figure S10), azeotrope-SAM shows larger energy shift, indicating more effective ITO work function modification.
  
• Ultraviolet Photoelectron Spectroscopy (UPS): UPS measurement results (Figures 2d, 2e and Figure S11 (SI)) support this observation, where ITO/azeotrope-SAM shows deeper work function of -5.13 eV and highest occupied molecular orbital (HOMO, -5.48 eV) well-matched with PM6’s ionization potential, potentially improving OSC open-circuit voltage (Voc).
  
• Cyclic Voltammetry: Surface density of Cbz-2Ph on ITO was further quantified using cyclic voltammetry, showing azeotrope-SAM molecular density of 8.51×10^13 molecules/cm², denser than IPA-SAM (6.63×10^13 molecules/cm²).
• X-ray Photoelectron Spectroscopy (XPS): XPS results also show higher P/In elemental ratio for azeotrope-SAM (SI, Figure S13 and Table S4), indicating more effective anchoring of Cbz-2Ph molecules on ITO surface when printed using appropriate solvents. These results qualitatively and quantitatively demonstrate that the azeotropic strategy can promote formation of denser SAMs on ITO for further OSC manufacturing.
  

Device Performance Characterization

SAM-based OSCs were fabricated with the structure ITO/Cbz-2Ph/PM6:BTP-eC9/PNDIT-F3N/Ag, effective area of 0.04 cm², comparing performance between azeotrope-SAM and IPA-SAM. Azeotrope-SAM-based OSCs showed narrower PCE distribution with highest PCE reaching 18.89%, significantly outperforming IPA-SAM (17.12%) and PEDOT:PSS (17.86%). Its Voc (0.859 V), Jsc (28.62 mA cm⁻²), and FF (76.83%) all surpass IPA-SAM (Voc of 0.846 V, Jsc of 27.86 mA cm⁻², FF of 72.64%).

EQE testing shows azeotrope-SAM enhances carrier extraction efficiency in the 500-800 nm range, with transient photocurrent and photovoltage measurements confirming faster charge extraction and reduced recombination. Dark current measurements show azeotrope-SAM reduces leakage current, improving device stability and consistency.

In batch testing, azeotrope-SAM achieved average PCE of 18.51 ± 0.18%, with significantly better reproducibility than IPA-SAM (16.70 ± 0.25%). Additionally, D18:BTP-eC9 OSCs showed average PCE of 18.70% on azeotrope-SAM, further confirming its applicability.

In large-area manufacturing, azeotrope-SAM showed higher FF (71.18%) and stable PCE scaling performance. Devices reaching 1.008 cm² achieved PCE of 17.76%, superior to spin-coating techniques. Azeotrope-SAM-based devices demonstrated T80 lifetime exceeding 2000 hours under high temperature and prolonged illumination, far surpassing PEDOT:PSS (169 hours). Peel tests indicate enhanced HSL/BHJ interface adhesion, contributing to improved stability and lifetime.

In-Depth Analysis of Azeotropic Strategy and Stability

Azeotropic Strategy

Beyond detailed characterization results, the literature delves into the rationale and advantages of adopting the azeotropic strategy. Since SAM molecules (like Cbz-2Ph) are amphiphilic, meaning they have both hydrophilic and hydrophobic parts, dispersing these molecules in solvents is challenging.

Using IPA alone leads to Cbz-2Ph micelle formation, while using toluene alone results in severe aggregation.

Mixing IPA and toluene improves Cbz-2Ph dispersion, while the azeotropic ratio (42 wt% toluene and 58 wt% IPA) ensures constant solvent composition during solution processing, leading to uniform SAM film formation.

Another advantage of the azeotropic strategy is enhanced shelf life of SAM inks. After 45 days of storage, SAM in pure IPA forms larger aggregates, while SAM in IPA:toluene mixed solvent shows good storage stability. This is crucial for large-scale production as it can reduce material waste and improve manufacturing process reproducibility.

Stability

Beyond efficiency, azeotrope-SAM-based OSCs demonstrate excellent stability. In accelerated aging tests under simulated 1-sun illumination, azeotrope-SAM devices showed T80 lifetime exceeding 2000 hours, compared to just 169 hours for PEDOT:PSS-based devices. Peel tests reveal that the force or energy required to peel PM6 films from ITO/azeotrope-SAM substrates is an order of magnitude higher than from ITO/PEDOT:PSS substrates, indicating stronger adhesion between SAM and active layer, thereby improving device stability.

Conclusion

The azeotropic strategy offers a promising approach for large-scale production of efficient and stable OSCs. By addressing challenges associated with SAM molecule amphiphilicity, the strategy can form dense, uniform, and stable SAM films, thereby improving device performance and lifetime.

This research proposes an innovative azeotropic solvent strategy for large-area printing of self-assembled monolayers as hole selection layers in organic solar cells. Through the implementation of an IPA and toluene azeotropic mixture, the team successfully achieved high-quality, uniform, and stable Cbz-2Ph SAM films, effectively enhancing ITO substrate work function.

The devices fabricated using this approach demonstrated exceptional performance, with small-area (0.04 cm²) devices achieving 18.89% efficiency and large-area (1.008 cm²) devices reaching 17.76% efficiency. Moreover, the devices exhibited remarkable stability, maintaining performance for over 2000 hours under sustained illumination.

These achievements represent significant progress in the field of organic solar cells, particularly in addressing the challenges of large-area fabrication and long-term stability. The azeotropic strategy presented here could potentially revolutionize the mass production of organic solar cells, bringing us closer to their commercial realization.

Original Publication: Energy & Environmental Science
Title: Efficient organic solar cells with a printed p-i-n stack enabled by an azeotrope-processed self-assembled monolayer
DOI: https://doi.org/10.1039/D4EE04479G