Measurement
Negative-pressure formation: what the fixture does to the cell
Why formation fixtures pull a vacuum, how stack pressure and gas removal affect SEI and lithium plating, and how to prove the pressure is the same on every channel.
Negative-pressure formation is usually described as a fixture setting: close the chamber, pull a vacuum, run the first charge. That description hides the two things the fixture is actually doing. It is applying a mechanical stack pressure to the cell, and it is removing gas from the space around the cell so that the gas cannot sit between the electrodes. Both actions change the electrochemical result. Neither is a measurement, and neither is captured by the channel's voltage and current accuracy alone.
This note explains what the pressure does, why uniformity matters more than the setpoint, and how to verify a fixture without inventing a pressure window the cell supplier has not given you.
What the first charge is doing
Formation is the first controlled charge after filling. The electrolyte contacts the anode and cathode, the solid electrolyte interphase (SEI) begins to form, and some of the electrolyte decomposes. Gas is produced as part of that process, and gas that remains trapped inside a pouch cell creates regions where the electrode stack is no longer wetted or compressed. The local current density then differs from the rest of the cell, which is exactly the condition that produces uneven SEI and, at the negative electrode, local lithium plating.
A negative-pressure fixture addresses both problems. Reducing the pressure around the pouch lets the atmosphere press the cell flat against the fixture plate, and a vacuum path from the gas bag removes gas that would otherwise remain in the cell. The stack is held while the SEI forms, so particle contact and porosity stay closer to the values the cell design assumes.
Datasheet fact: the formation and grading cabinets are published as 512 channels per cabinet, with measurement accuracy of ±(0.2% RD + 0.2% Range), typical and model-dependent. The same page lists negative-pressure formation and self-discharge screening, with results passed to BTS 9.1 and a LIMS. That accuracy figure describes the electrical channel. It says nothing about the pressure setpoint, the pressure uniformity, or the leak rate of the fixture. Those belong to the process specification.
Force is pressure times area
The reason a small pressure change matters is that the fixture applies the pressure over the whole cell face. Force is pressure multiplied by area. A pressure differential of 20 kPa across a 200 cm² pouch face is 400 N, roughly the weight of a 40 kg mass. At 50 kPa the same face sees 1,000 N. The cell does not experience “50 kPa” as an abstract number; it experiences a distributed load that can flatten the stack, squeeze electrolyte out of the separator, or deform the tabs if the load is not reacted properly.
| Pressure differential | 100 cm² face | 200 cm² face | 300 cm² face |
|---|---|---|---|
| 10 kPa | 100 N | 200 N | 300 N |
| 30 kPa | 300 N | 600 N | 900 N |
| 50 kPa | 500 N | 1,000 N | 1,500 N |
Illustrative arithmetic: force = pressure × area. It is not a recommended process window. The correct window comes from the cell design and the supplier's formation specification.
The same arithmetic explains why “negative pressure” needs a definition. A gauge reading of −80 kPa means 80 kPa below ambient, which at sea level is about 20 kPa absolute. A gauge reading of −20 kPa is about 80 kPa absolute. The two are not interchangeable, and a fixture that holds “−80” in one lab may be describing a different absolute pressure from a fixture that holds “−80” in another. Record the reference: gauge or absolute, and the ambient pressure at the time.
What pressure does to SEI and plating
SEI formation is sensitive to how well the electrolyte wets the electrode and how uniformly the current is distributed. A stack that is held flat has a more uniform distance between the electrodes and a more uniform electrolyte film. A stack with a gas pocket has a region where the ionic path is longer, the local current density is lower during a fixed-current charge, and the rest of the electrode must carry more current. The average voltage looks normal; the local overpotential does not.
Too little pressure leaves gaps and allows gas to remain. Too much pressure can close porosity, reduce the electrolyte available at the reaction front, and damage particles or the separator. The useful range is a window, not a maximum. It is also format-specific: a coin cell, a stacked pouch cell and a prismatic cell do not share a pressure specification, and a number copied from one format is not evidence for another.
Engineering judgement: the fixture should be qualified on the cell format and recipe you actually run. Start from the cell supplier's formation specification, then verify that the fixture can hold that pressure across the full tray and for the full formation time. Do not use the vacuum gauge as the only evidence. The gauge tells you what the pump circuit reached; it does not tell you what the cell face experienced.
Why uniformity is harder than setpoint
A single gauge on a manifold can read the target while individual fixtures sit at different pressures, because the pressure drop is in the tubing, the valve and the fixture seal, not at the gauge. The first tray position and the last tray position are the two ends of that path. If the formation cabinet has 512 channels, the pressure circuit may serve many fixtures from one pump; the channel-to-channel spread is a process variable that the electrical record will not reveal.
Temperature adds a second source of spread. Gas pressure changes with temperature, and a fixture that is cold at the start of formation will reach a different pressure after the cabinet warms. If the pressure is set before the cabinet reaches thermal equilibrium, the cells see a pressure ramp that is not in the recipe. Log the fixture pressure and the cabinet temperature together, and give the pressure loop enough time to settle before the charge starts.
How to verify pressure consistency
- Calibrate the pressure sensor against a reference. Record the calibration date and the uncertainty. A gauge with 1% full-scale uncertainty is not adequate for a 10 kPa process window if the range is 100 kPa.
- Run a leak-down test on every fixture. Isolate the fixture, record the pressure decay over a defined interval, and set a pass limit. A fixture that passes the electrical check can still leak.
- Map the cell face, not just the port. Use a pressure-sensitive film or a dummy cell with several sensors to see whether the load is uniform across the face and across tray positions.
- Compare channels under the same recipe. Run the same dummy load on a sample of channels and record the pressure trace. A channel that reaches the setpoint later, or drifts more, is a maintenance candidate.
- Record the pressure with the electrical data. The per-cell record should carry fixture ID, pressure setpoint, measured pressure, vacuum duration and leak-test result, alongside voltage, current and temperature.
These checks are engineering practice, not a published specification. Their purpose is to make the pressure trace auditable, so that a group of cells with unusual capacity or plating can be separated into a fixture problem and a cell-chemistry problem.
What to write into the process record
The electrical channel record already has a time base, a channel ID and a step file. Add the fixture identity and the pressure trace to the same record, not to a separate spreadsheet. The formation page states that results are passed to BTS 9.1 and a LIMS-friendly export, with an audit trail. Use that path: the pressure setpoint and the measured pressure should be reviewable next to the voltage curve they influenced.
For each cell, record at least: cell serial, fixture ID, pressure reference, setpoint, measured value at start and end, leak-test result, cabinet temperature and recipe version. If a PLC outside BTS 9.1 controls the pressure, export its trend against the tester clock. A pressure record without a shared time base is difficult to defend.
Safety and failure modes
A vacuum fixture stores energy. A door that opens under vacuum, a failed seal, or a fixture released before the pressure has equalised can move hardware and cells. Interlock the fixture so the pump stops and the chamber vents before the door opens. The chamber datasheets list over-temperature protection, smoke detection and optional pressure relief; those protect the chamber, not the formation fixture, which needs its own risk assessment.
Loss of vacuum is also a process failure, not only a safety event. If the pump stops mid-formation, the stack pressure falls and the gas path changes. The recipe should define whether the run is aborted, held, or continued, and the record should show the vacuum loss. Do not let the software silently continue a formation step whose mechanical boundary condition has changed.
Takeaways
- Negative pressure does two jobs: it applies a distributed stack load and it removes gas. Treat the pressure setpoint and the pressure uniformity as process variables, not as a fixture accessory.
- The force on a pouch face is pressure × area; a 20 kPa differential on 200 cm² is 400 N, so a small gauge error or a blocked port can change the mechanical load substantially.
- Verify the fixture with a calibrated sensor, a leak-down test, a face-pressure map and a channel-to-channel comparison, and put the pressure trace in the same record as the voltage and current.
Request a configuration review
Tell us the channel count, current range and protocol. We will check the measurement class against the note and return a configuration, not a generic datasheet.
Request a configuration reviewReferences
- Peer-reviewed use of this instrument class: Surface engineering enables robust SEI growth towards a stable and efficient lithium-ion battery SiO x anode — Wu (2026), RSC Advances. DOI: 10.1039/d5ra08407e
- Formation & grading systems
- Datasheets and glossary