Requirements in perovskite solar have been shifting quickly - larger substrates, thinner glass, and a material stack that degrades if it gets warm - and those shifts change what glass suppliers are being asked to deliver. A process that was acceptable on a laboratory coupon is not acceptable on a production panel, because the laser that cuts the glass also decides how much of the perovskite survives. These are the design trends and sourcing notes that matter when specifying laser-cut glass for perovskite and tandem modules.
Laser Cutting of Glass in Perovskite Solar
In perovskite solar, that translates into a specific set of requirements rather than a general preference for quality. A perovskite module is a thin-film device: the absorber sits between electrode layers on a transparent substrate, and the module is divided into series-connected cells by a sequence of scribe lines that run through the stack. The scribing pattern is what turns a large panel into a working module, and it is formed by laser processing at three separate stages. At the same time, the absorber is thermally fragile, so the laser has to remove material without depositing enough heat to degrade the cells either side of the cut. The glass specification and the laser process specification are therefore written together, because the substrate's thermal and optical properties set the process window.
What drives cost and lead time
Cost in this application is driven by the width of the area that the scribes take out of the active region. Each scribe line is a strip of panel that generates no power, so the accumulated dead zone across a module is a direct efficiency loss, and narrowing the lines is worth more than almost any other process improvement. The second driver is the laser source: ultra-fast lasers that remove material by ablation with minimal heat are more expensive per watt and more expensive per hour than the nanosecond sources used on established thin-film lines, but they are the only practical way to scribe a perovskite stack without damaging it. Lead time follows the process qualification rather than the glass itself, and a new substrate format usually requires a full re-qualification of the scribe recipe before production starts.
How laser cutting of glass works in practice
Glass is not cut by melting it, because a melt front leaves a micro-cracked edge and a heat-affected zone that will fail later. Production processes either introduce a controlled defect and then separate the part, or ablate the material directly with a pulsed beam whose pulse duration is short enough that heat has no time to diffuse into the surrounding material. The distinction matters for perovskite because the same short pulses that keep the glass cold also keep the absorber cold, and that is what makes the scribe lines narrow and clean. Where the cut has to reach through a coated substrate, the process is normally run from the film side or the glass side depending on which layer is more sensitive, and the choice is a process decision that has to be made with the module design in front of it rather than afterwards.
Design rules that reduce cost
Keep the scribe count as low as the electrical design allows, because every line is a permanent loss. Specify the substrate thickness from the mechanical requirement rather than from habit: thinner glass lowers material cost and improves the thermal behaviour of the scribe, but it also changes how the panel is handled through every subsequent step. Design the edge deletion and the busbar layout together, since both remove active area and the two are usually optimised by different people. Where the module is a tandem rather than a single junction, plan the scribe sequence before the layer stack is fixed, because the intermediate connection is what makes a tandem work and it is created by the same laser process. Finally, specify the edge quality in terms of what the encapsulation can tolerate, which is usually a chip size and a micro-crack depth rather than a cosmetic grade.
Standards, documentation and traceability
Module qualification follows the established photovoltaic standards, and the glass supplier's contribution is usually the substrate certificate and the edge inspection data. What a perovskite programme needs beyond that is process traceability: which laser parameters were used on which substrate lot, because the interaction between the coating and the beam varies with the deposition run. Reliability testing for the module covers damp heat, thermal cycling and mechanical load, and the edge of the glass is where several of those tests fail if the cutting process was marginal. A supplier who can provide a chip-size distribution and a micro-crack assessment per lot is worth more to the programme than one who can only certify the substrate.
Requirements specific to Perovskite Solar
Perovskite modules add constraints that established thin-film lines do not carry. The absorber degrades under heat and moisture, so every process step after deposition has to respect a thermal and humidity budget, and the glass cutting step is no exception. Where the module is a tandem with silicon, the substrate sees additional process steps that a single-junction line does not have. The panel format is moving towards larger substrates, which raises the demands on laser positioning accuracy and on how flat the substrate stays during processing. Encapsulation requirements are strict because the absorber is sensitive to both moisture and oxygen, which means the edge quality achieved by the cut is directly coupled to the reliability of the module.
- Scribe width and accumulated dead zone drive module efficiency
- Ultra-fast pulses keep both the glass and the absorber cold
- Chip size and micro-crack depth specified against encapsulation limits
- Process traceability linking laser parameters to substrate lot
IEC 61215 and IEC 61730 module qualification, SEMI conventions for substrate documentation, and lot level inspection data covering chip size distribution and micro-crack assessment.
Selection data at a glance
| Parameter | Typical capability | Notes |
|---|---|---|
| Substrate | TCO glass, ultra-clear float, borosilicate | Coated or uncoated |
| Thickness | 0.7-3.2 mm, 1.1-2.0 mm typical | Mechanical driven |
| Cut edge | no chips over 100 µm, micro-crack under 50 µm | Encapsulation limited |
| Heat affected zone | under 20 µm with ultra-fast pulses | Absorber sensitive |
| Scribe width | 20-50 µm typical | Dead zone accumulates |
| Positioning accuracy | ± 20 µm standard, ± 10 µm on request | Tandem alignment |
| Panel size | up to 1.2 x 0.6 m and larger formats | Line dependent |
| Thermal budget | process kept below the absorber limit | Typically under 150 C |
| Inspection | chip size distribution and micro-crack depth per lot | Edge focused |
Frequently asked questions
Why not use the nanosecond lasers already installed on thin-film lines?
Because the perovskite absorber is thermally fragile. A longer pulse deposits heat into the material around the cut, and the cells closest to the scribe line lose performance even though the cut itself looks clean. Ultra-fast pulses remove material before heat can diffuse, which is what keeps the dead zone narrow and the adjacent cells intact.
Does substrate thickness change the scribe process?
It changes the process window rather than the principle. A thinner substrate conducts heat away differently and is more sensitive to how it is supported during cutting, so the laser parameters normally have to be re-optimised when the thickness changes. It is one of the reasons a new format requires a fresh qualification.
What edge quality should a perovskite program specify?
One expressed in terms of what the encapsulation can survive: a maximum chip size at the edge and a maximum micro-crack depth into the face. Cosmetic grades are not useful here, because a chip that looks minor can breach the encapsulation seal and let moisture reach the absorber.
Need a quotation for laser cutting of glass?
Send your drawing, target specification or coating requirement and an optical engineer will reply within 24 hours with a price and a lead time.
Request a Quote Email an engineer