Harnessing Robotics, Automated Processes, and AI in Solar Cell Research

Solar technology is one of the fastest-growing sectors in renewable energy. With solar installations on the rise, the demand for more efficient and scalable solutions has skyrocketed. While silicon-based solar panels currently lead the market, emerging technologies like organic photovoltaics (OPVs) and perovskite solar cells are opening up exciting new possibilities.

Image Credit: BELL KA PANG/Shutterstock.com

Although silicon solar panels are effective, they are not always suitable for every application. Newer technologies, such as perovskite and organic solar cells, use combinations of specifically designed organic and hybrid materials to optimize light absorption and energy conversion. These advancements have quickly brought their efficiencies on par with silicon in a relatively short time.

The NREL efficiency chart shows the efficiency increase of Perovskite Solar Cells and OPVs, which is catching up to silicon. Image Credit: Ossila Ltd

Beyond efficiency, OPVs and perovskite solar cells offer unique benefits over silicon, such as being flexible, lightweight, and semi-transparent. Perovskites can be layered on top of existing silicon cells to enhance performance. These innovations create new opportunities for more adaptable solar applications but also introduce challenges that require further research and refinement.

Finding the Right Molecules and Device Combinations

Despite their impressive potential, perovskite and organic photovoltaics need improvements in stability, scalability, and efficiency. These devices must withstand real-world conditions—like heat, humidity, and UV exposure—which currently pose significant barriers. Enhancing the materials and manufacturing processes is crucial for achieving consistent, long-term performance.

One advantage of perovskites and OPVs is that you can easily adjust the material properties by altering the composition of the active layer. These solar cells also require multiple layers, each with a mix of organic or inorganic molecules, offering further opportunities for optimization.

Image Credit: Ossila Ltd

This versatility creates a vast experimental space with thousands of potential material, additive, and processing combinations. These combinations need to be tested to find the “best” device, but on this scale, manual experimentation alone is almost impossible. This highlights the need for more efficient research methods to identify the most promising configurations.

Automation and Robotics: A Game-Changer for Solar Research

Given the complexity of modern solar cell development, automation is becoming essential. Automated platforms equipped with robotic systems can fabricate, test, and refine solar cells efficiently and precisely.

These systems excel at repetitive tasks that are challenging for humans to perform at scale, including:

• Spin coating and drop casting thin films
• UV-Vis characterization of thin films and solution
• Current density-voltage (J-V) characterization
• Solution creation, extraction, and deposition
• Thin-film characterization

Robotic systems perform these processes accurately, ensuring consistent results while reducing human error. They can operate in various lab environments and be integrated into specialized setups like vacuum chambers or inert conditions for handling air-sensitive materials like perovskites. Installing automated systems within controlled environments, such as a glove box, allows for even greater versatility and experimental flexibility, as researchers can precisely regulate the conditions under which experiments are conducted.

AI and Machine Learning: The Perfect Partner for Automation

Combining automated experimentation with AI-driven machine learning opens a new frontier in solar research. AI and machine learning can provide incredible analysis and learning possibilities but require huge data sets to train these models. Automation enables researchers to conduct large-scale experiments, quickly producing the extensive data needed for effective machine-learning applications.

With vast amounts of experimental data generated by robotics, AI can analyze trends, predict material stability, and optimize solar cell performance. For instance, machine learning has identified key properties, such as the effective energy gap, which influence the resilience of organic photovoltaics under environmental stress. This combination of automation and AI accelerates the discovery of stable, efficient solar cell designs ​(Du et al. 2024, Gu et al. 2020).

Results from Research

In a recent study, automated research platforms were used to evaluate over 40 combinations of materials for OPV devices, highlighting which materials have better stability under light and air exposure. By using automation within glove boxes, researchers preserved the integrity of sensitive materials, allowing for accurate, large-scale data collection and optimization ​(Du et al. 2024).

Another study employed high-throughput robotic systems to assess the stability of multi-cation perovskite materials, uncovering a temperature-dependent stability that challenged previous assumptions ​(Zhao et al. 2021). Using robotics allowed for rapid experimentation, testing 336 different perovskite-antisolvent combinations in just two days​ (Gu et al. 2020). This led to deeper insights into solvent interactions and crystallization processes, which are pivotal for perovskite solar cell stability and efficiency.

This automation ensures reproducibility by controlling variables such as deposition speed and solution volume, which is vital for scaling up to industrial production.

References

1. Du, X. et al. (2024) ‘Revealing processing stability landscape of organic solar cells with automated research platforms and machine learning’, InfoMat, 6(7).
2. Zhao, Y. et al. (2021) ‘Discovery of temperature-induced stability reversal in perovskites using high-throughput robotic learning’, Nature Communications, 12(1).
3. Gu, E. et al. (2020) ‘Robot-based high-throughput screening of antisolvents for lead halide perovskites’, Joule, 4(8), pp. 1806–1822.

This information has been sourced, reviewed and adapted from materials provided by Ossila Ltd.

For more information on this source, please visit Ossila Ltd.