Scalable perovskite module manufacturing requires interconnection processes that are not only precise, but also compatible with cost-effective and accessible production platforms. Interconnection can be achieved through careful registration control during printing or through scribing to create interconnects using mechanical or laser patterning. Ultrafast femto- and pico- second lasers are widely used in laboratory conditions for high-quality module patterning as they minimise thermal damage whilst enabling narrow scribe lines. However, their high capital cost is a challenge for low-cost perovskite module manufacturing. Nanosecond laser systems provide a more affordable and industrially established alternative; however, their longer pulse duration increases thermal interaction with the device stack, leading to melting, heat-affected regions, edge damage, and potential shunting pathways. The key challenge is therefore not simply replacing ultrafast lasers with lower-cost systems but defining whether a controllable nanosecond laser process window can be achieved while balancing processing cost, interconnection quality, and long-term module stability. This current work explores these challenges in the context of plastic flexible perovskite module fabrication.

Flexible substrates are attractive for scalable and cost-effective perovskite module manufacture due to their lightweight nature and compatibility with roll-to-roll processing. Flexible substrates are a particular challenge for laser scribing, where thermal and mechanical sensitivity can narrow the available processing window. The mechanical instability of the Indium Tin Oxide (ITO) layer when laser processing introduces difficulties when considering roll-to-roll manufacture. This work has found that laser processing of the ITO-PET introduces heat affected regions, damaging the surrounding material, introducing defects at the initial processing stage. Cracking in the surrounding material could propagate during roll-to-roll manufacture, decreasing longevity of the final modules. Additionally, scribe edge heights of up to 1.5 µm have been measured. As the deposited stack has a thickness of only 0.6 µm, this allows for the electrodes to be in direct contact introducing shorting and shunting pathways. Additional difficulties were observed when producing reliable and repeatable scribes, with slight variations in layer thickness and laser focus shown to change scribe morphology when using the same laser powers. This would reduce the repeatability of module manufacture. Through a systematic study of laser pulse power, frequency and speed we have been able to determine the ideal process windows for the nanosecond laser. These optimised settings have been used to explore the module fabrication process. To show the particular issues for PET, this work will highlight the differences in laser scribing glass and PET substrates and will also showcase results obtained for post scribe cleaning processes which can improve subsequent device performance.