Before the list: what a supplier’s equipment table is actually telling you
Ask ten thin-film engineers what equipment a perovskite solar cell production line uses and you will get ten different lists. The differences are rarely about the physics — everyone agrees on the process steps. The differences come from what a given supplier has actually built, what it lists on its website, and what it is willing to integrate between machines. A useful answer to “what equipment is used?” therefore has two layers: the process chain every line has to cover, and the questions that turn a supplier’s catalogue into a line you can plan around.
One more thing before the walk-through. A published specification table is not marketing; it is an interface definition. The substrate size on a coating machine tells you what your upstream station has to deliver. The frame size on a handling machine tells you what has to be supplied and returned. A throughput figure tells you what the next station must absorb. Read every table this way and half of the “integration risk” in a line project disappears before any purchase order is signed.
Substrates and front contacts: where the line starts
Perovskite cells are built on a conductive starting surface — typically a coated glass or, for flexible work, a coated polymer web — and most production lines receive that surface ready-made from a glass or web supplier rather than making it in-house. The equipment at this end of the line is therefore about handling: unpacking, inspection, cleaning if your process calls for it, and presentation to the first coating station at a defined position, height, and rate.
The questions to ask at this stage are unglamorous and decisive. What substrate format does the line accept, and what thickness range? How are substrates loaded — by hand, by magazine, by conveyor? What is the reject path for a substrate that fails inspection before it ever sees a coating? A line that cannot feed itself reliably will not be limited by its most expensive machine.
The coating station: the heart of the line
Depositing the perovskite absorber is the step most line projects are really about, and the equipment differs by what you are making. Small-area development work is commonly done by spin coating, which is inexpensive to set up and quick to iterate on, but it does not scale to the formats a production line uses. Its real job is to develop the ink, the formulation, and the film target that a production coater will later have to reproduce.
For scale-up, the route most production lines evaluate is a slot-die style process. PowerDe’s slit coating equipment listing describes exactly this principle: coating liquid pressed along the gaps of a mould under pressure and transferred to a moving substrate. The page states that the process is used widely in solar cells, lithium-ion batteries and perovskite work for “fast coating speed, good coating uniformity, and wide coating window”. The published speed range of 1–10 m/min and thickness range of 5–200 µm are best read as the window you are allowed to explore, not as a promise about your ink. The questions that matter are how uniformity is measured across the width, how the oven profile is set and recorded, and how long a change between two formulations takes. Two rows on that same table — coating width and equipment pressure — are published in units that do not parse as written, which is exactly the kind of item to close in writing before you plan a layout around it.
Development and production need different machines for the same reason. A glove-box spin coating system lets you control the atmosphere at the moment of coating while you fix a formulation; a slit coater is what you buy once the formulation is fixed and the question becomes repeatability at speed. Buying them in the wrong order is an expensive way to discover that your production coater cannot reproduce your lab results.
For continuous processing on flexible substrates, roll-to-roll magnetron sputtering is listed in the same thin-film solar equipment series. In a perovskite line it is usually considered for contact layers rather than the absorber itself, and the planning questions shift to web width, line speed, tension handling, and how warm the substrate is allowed to get — polymer webs have temperature limits that glass does not.
Drying and crystallisation: the step that turns a wet film into a cell
A coated film is not a perovskite film until it has been dried and crystallised under a controlled thermal profile. The equipment here is an oven or drying section whose temperature zones, atmosphere, and transport speed determine the film structure — and therefore the device you are actually making. The slit-coating listing above includes a one-metre electric heating oven, but the entry carries an unexplained note, which is a good example of a specification line that has to be closed in writing before layout: what type of oven, how many zones, what temperature range, what control accuracy, and how the profile is recorded for each run.
If your process needs a controlled atmosphere during drying, say so at the enquiry stage. An oven that reaches the temperature is not the same machine as an oven that reaches the temperature in the atmosphere your film requires.
Laser patterning: how small cells become a usable module
A single small cell delivers far less than a module needs, in both voltage and current. Modules reach usable voltages by connecting many cells in series inside one continuous stack, and that is done with a laser. Three scribing steps are standard practice in monolithic interconnection: one isolates the front contact, one clears the absorber to expose the contact beneath it, and one opens the rear contact to complete each cell’s path. The same machine therefore performs three different removal tasks on three different materials, and settings that work on one stack can damage another.
PowerDe lists a laser cutting machine within the thin-film equipment series, and the listing does not publish detailed process parameters. That is not a problem as long as you treat it the way you should treat any laser: ask the supplier to cut samples of your own stack, define the acceptance criteria in writing before the trial — scribe width, residue, damage to the layers beside the cut — and judge the machine on your material rather than on a demonstration video.
Rear contacts and lamination: closing the stack
The rear contact can be deposited by evaporation or sputtering, laminated as a foil, or printed, depending on the stack and the production philosophy. Whatever the route, the line then needs equipment that closes the stack: bonding or laminating machines that join layers at a defined temperature, pressure, and alignment, and — for modules that must survive years outdoors — encapsulation that excludes water and oxygen.
The membrane-electrode hot-press bonding machine in PowerDe’s thin-film series publishes a working temperature up to 150 °C, a 2 t press force, a fitting error of 0.2 mm or less, and a cycle of at least 1.5 pieces per minute. Those numbers describe the machine’s envelope — the temperature ceiling, the force, the alignment it holds, and the pace it is built to run at — not the recipe your stack requires, which is always narrower than the machine’s limits. The listings describe membrane-electrode assembly rather than perovskite module lamination specifically, so the question to ask is direct: has this platform been used on your type of stack, and will the supplier trial it with your materials?
Testing and metrology: the equipment that decides what “production” means
Every line needs measurement at two levels: inline checks that catch a problem while it is still one bad substrate rather than a bad shift, and offline characterisation that tells you whether the process itself is drifting. Visual inspection, electrical measurement, and stability testing are usually supplied by specialist metrology vendors rather than by the companies that build the coaters and lasers. The integration work is yours: deciding where each measurement sits, what happens to a unit that fails, and who has the authority to stop the line.
A useful discipline is to define the measurement before the machine list is final. A line that produces output with no measurement is an expensive way to manufacture scrap you discover later.
Between the machines: handling, controls, and utilities
Ask what is between the machines, because that is where line projects are won and lost. Substrates have to move between stations without accumulating queues, and queues have to exist somewhere — buffer stations absorb the difference between a coating station that runs continuously and a laser or test station that indexes. Controls have to speak to each other: which station is allowed to signal start, stop, and fault to which other station, and what data is recorded per unit.
PowerDe’s automated production line solutions category covers line-level integration and custom automation, which is the honest way to describe this part of the work: it is engineered per project, not ordered from a catalogue. Alongside it, collect the utilities sheet for every machine before you freeze the layout — power, exhaust, compressed air, cleanliness class, floor space — because a machine that fits the process but not the building is still a machine you cannot install.
How to read a specification table as a line planner
The perovskite production line equipment listing published by PowerDe is a compact demonstration of why tables should be read as interfaces. It lists a substrate size of 108.5 × 108.5 mm at 450–850 µm thickness, a composite film of 121.5 × 121.5 mm at 20–127 µm, and a 125 × 125 mm frame at 0.8 mm. Those six numbers tell you what the upstream station must deliver and what the downstream station must receive. The capacity figure of 800 pieces per hour and a cycle of 4.5 seconds or less tell you what the next station must absorb; the fragmentation rate and mean time between failures are planning data for spares, maintenance windows, and how much manual intervention the line assumes.
The same listing describes a series used for research and development across perovskite solar, OLED, lithium battery, quantum-dot LED, and OPV work. A description that broad is a reason for a question, not an assumption: which process steps does the delivered configuration actually perform, and which does it not? Every table also has gaps — footprint, weight, utilities, cycle breakdown, measurement methods — and the gaps are what you ask for in writing.
What to request in writing, and how to validate it
For every machine on the shortlist, request three things before any commercial discussion: the published parameters with their measurement methods and instruments, the unpublished items listed above, and a trial on your own substrates and films against acceptance criteria you wrote before the trial started. A supplier who runs a trial on your material is giving you evidence; a supplier who offers a video is giving you a brochure.
Validate in the same order you would build the line: first the coating and drying result on your own stack, then the laser steps on the coated result, then the closing and measurement steps. Record everything — settings, times, defects, and photographs at fixed reference points — and send the record, not a summary, to the supplier. Ambiguities about atmosphere, measurement points, or cycle time are cheap to close on paper and expensive to close after installation.
What this list cannot tell you
An equipment list says which machines exist. It does not say whether they work together, whether they hold their settings through a full shift, or whether the supplier will still answer the phone when a station stops at two in the morning. The order that keeps projects honest is process first — your stack, your format, your target output written down; interfaces second — what each station delivers to the next; machines third. If you are mapping a line plan against the equipment that is actually built and available, send your process description and format to the PowerDe team and ask for a written answer on each stage.
