Plasma Cleaning in Battery Cell Preparation

.
13 min read

On this page

It is also the station that determines whether the thousands of welds downstream are consistent. Weld quality problems that nobody can trace usually originate here, because what a laser or a resistance welder actually joins is not the terminal, it is whatever is on the terminal.

This article covers what contaminates a cell terminal, what plasma treatment does about it, what it cannot do, and why its role is growing rather than shrinking.

What Is Actually On a Cell Terminal

  • Oxide. Every exposed metal surface oxidizes. Aluminum does so within milliseconds of exposure to air and continues to build a film indefinitely.
  • Processing residues. Release agents, forming lubricants and cutting fluids from cell manufacture, present in quantities far too small to see and entirely sufficient to affect a weld.
  • Protective coatings. Some cells ship with a thin anti-corrosion layer intended to survive storage, which then has to be removed before joining.
  • Handling contamination. Fingerprints deposit oils and salts, and both persist through automated handling once present.
  • Packaging residue and dust. Fibers and particles from trays and wrapping, plus whatever settled during storage.

None of this is visible at the scale that matters. A contaminant film measured in nanometers is invisible, and it is thick enough to change how a weld forms.

The Aluminum Oxide Problem

Oxide deserves separate treatment because of one property that makes it genuinely difficult rather than merely inconvenient.

Aluminum melts at around 660 degrees Celsius. Aluminum oxide melts at around 2,072 degrees. The film sitting on the terminal therefore has a melting point roughly three times that of the metal beneath it, which means a welding process delivering enough energy to melt aluminum is trying to work through a ceramic layer that will not melt at that temperature.

What happens in practice is that the oxide breaks up unevenly. Energy couples into the surface inconsistently, penetration varies from joint to joint, oxide fragments become inclusions in the weld, and the process produces spatter. The result is not a failed weld that gets rejected, it is a population of welds with a wide distribution of strength, most of them acceptable and some of them marginal in a way no inspection reliably catches.

Oxide also raises electrical contact resistance, which matters for the joint in service and for any measurement taken through that surface.

What Plasma Treatment Does

Atmospheric plasma is generated by applying a high-voltage discharge to a gas stream, usually compressed air, which ionizes it. The resulting mixture of ions, electrons and reactive species is directed onto the surface through a nozzle.

Two mechanisms then operate.

Removal of organic contamination

Reactive oxygen species in the plasma attack hydrocarbon molecules on the surface, breaking them into carbon dioxide and water vapor which simply evaporate. Oils, fingerprints, release agents and light residues are removed at a molecular level rather than being wiped from one place to another, and nothing is left behind because the products are gases.

Surface activation

The treatment also raises the surface energy of the material by creating reactive sites. A high-energy surface wets readily, which means adhesives, coatings and molten metal spread across it rather than beading up. For welding this improves consistency; for adhesive bonding it substantially improves the strength of the joint.

What It Does Not Do

This is worth being clear about, because plasma treatment is sometimes specified as though it cleans everything.

It does not remove thick oxide layers. Plasma is a surface chemistry process, not an abrasive one, and while it can reduce very thin oxide films it will not strip a substantial one. Where heavy oxide is the problem, laser cleaning or mechanical preparation is required.

It does not remove particulates. A dust particle or a fiber sitting on a terminal is still there afterward, because nothing has physically displaced it.

And it is a line-of-sight process. The plasma reaches what the nozzle can see, so complex geometry, recesses and undercuts are treated unevenly, which affects nozzle positioning and the number of treatment passes required.

Why It Matters Before Welding

In laser welding, surface condition governs how energy couples into the material. Contamination absorbs differently from clean metal, so the effective energy delivered varies from joint to joint even with identical laser parameters. Organic residue also vaporizes explosively under the beam, producing porosity and spatter in the weld.

In resistance welding, contamination sits directly in the electrical path. Contact resistance varies with what is on the surface, so the heat generated at the joint varies with it, and the process is less repeatable than the equipment is capable of.

The consistent theme is variability rather than outright failure. A dirty terminal usually still welds. It welds differently from the one before it, and in a pack containing thousands of joints, a wide strength distribution is a reliability problem that appears years later under vibration and thermal cycling.

Why It Matters More Before Bonding

Adhesive performance depends directly on surface energy. A bond formed on a low-energy contaminated surface fails at the interface rather than within the adhesive, and it does so at a fraction of the strength the adhesive is capable of. Plasma activation raises surface energy immediately before bonding, which is the difference between a joint that carries structural load for a decade and one that does not.

The same applies to thermal interface materials, where wetting determines contact area and contact area determines heat transfer. Poor wetting produces voids, and voids are thermal insulators sitting exactly where heat needs to leave the cell.

Alternatives Compared

MethodWhat it removesInline suitabilityDrawbacks
Atmospheric plasmaOrganic contamination, and it activates the surfaceExcellent; milliseconds to seconds, no chamber or consumablesDoes not remove thick oxide or particulates; effect decays with time
Solvent wipingOrganic contaminationPoor; manual, inconsistent and needs drying timeSolvent handling, residue, and results that vary with the operator
Laser cleaningOrganic contamination and oxide layersGood; fast and preciseHigher capital, and enough energy to alter the surface if misapplied
Mechanical abrasionOxide and adherent depositsModerateGenerates particulate, which is the last thing wanted near an open cell
Dry ice blastingParticulates and some organicsModerateBulky equipment, and thermal shock to a cell terminal is undesirable
Chemical etchingOxide and organics thoroughlyPoor; a wet process with rinse and dry stagesEffluent, handling and incompatibility with a dry assembly line

The reason plasma dominates in-line cell preparation is the combination in the third column. It treats in a fraction of a second, needs no chamber, consumes only compressed air and electricity, produces no waste and leaves nothing to dry. Methods that clean more thoroughly are all slower, wetter or more aggressive, and on a line running thousands of cells an hour those characteristics decide the matter.

“Clean terminals are the foundation of reliable cell joining – plasma treatment controls organic contamination and surface energy before welding or bonding begins”

See it in action

Process Control and the Decay Window

Plasma treatment is a process with parameters, and treating it as an on-or-off device is how its benefit is lost.

  • Standoff distance. The gap between nozzle and surface determines how much of the reactive species reaches the target. Too far and treatment is weak; too close and the surface heats.
  • Traverse speed. Dwell time per unit area follows from speed, and it sets how complete the treatment is.
  • Power and gas flow. These determine the density of reactive species in the stream.
  • Coverage. Line-of-sight means the whole joining area must be reached, which on a terminal with any relief requires deliberate nozzle positioning.

The decay window

The most operationally important characteristic, and the one most often overlooked. Surface activation is not permanent. A treated surface begins to lose its raised energy immediately as it re-adsorbs contamination from the surrounding air, and over hours to days it returns toward its untreated state.

The practical consequence is that plasma treatment must be immediately followed by the operation it enables. Treating cells and then buffering them for an hour before welding or bonding gives away much of the benefit, and treating a batch at the start of a shift for use through the day gives away nearly all of it.

This is why the station belongs inline directly ahead of joining rather than at goods inward, and why a line layout that separates them by a long conveyor or an accumulation buffer is working against the process it installed.

Where It Sits in the Line

Cybernetik cell preparation capabilitySpecification
Surface preparationPlasma cleaning of cell terminals within the cell-to-module zone
Position in sequenceAfter cell feeding and barcode reading, before OCV, IR and ACIR testing and welding
Incoming verificationOCV, IR and ACIR testing on every cell with robotic sorting onto parallel conveyors
JoiningLaser and resistance welding with inline weld integrity testing on every joint
Weld verificationIntegrity testing on both top and bottom module faces
Placement verificationVision-confirmed insulation paper application and cell polarity checking
Line rateUp to 6,000 cells per hour
Cell formatsCylindrical 18650, 21700 and 32140, plus prismatic, pouch and blade
TraceabilityBarcode and RFID capture with MES connectivity and process logging

That position does two jobs. It puts surface preparation close enough to welding that the decay window is not consumed, and it puts it ahead of the electrical measurements, so contact resistance during OCV, IR and ACIR testing is not affected by whatever the terminal arrived with. A test probe contacting a contaminated terminal reads the contamination as well as the cell, which is one route by which sorting bands quietly widen.

Why manufacturers choose Cybernetik

  • Surface preparation as a station, not an assumption. Plasma cleaning built into the line rather than left to incoming cell condition.
  • Positioned against the decay window. Treatment immediately ahead of testing and joining, so the activated surface is used rather than allowed to recover.
  • Cleaned before measuring. Terminals prepared ahead of OCV, IR and ACIR testing, so contact resistance reflects the cell rather than its surface.
  • Verified downstream. Inline weld integrity testing on every joint and on both module faces, which is how the benefit of clean terminals is confirmed rather than assumed.
  • Format coverage. Cylindrical from 32 to 46 mm plus prismatic, pouch and blade, at up to 6,000 cells per hour.
  • Complete line responsibility. Cell handling through module build to end of line from one engineering team, with factory acceptance testing before dispatch.

Frequently asked questions

Because a welder joins whatever is on the terminal rather than the terminal itself. Cells arrive carrying oxide, release agents and lubricants from manufacture, protective coatings, handling oils and packaging residue. None of it is visible, and all of it changes how energy couples into the surface, producing welds with a wide strength distribution rather than outright failures.

Organic contamination, through reactive oxygen species that break hydrocarbons into carbon dioxide and water vapor which evaporate. It also raises surface energy, which improves how molten metal and adhesives wet the surface. It does not remove thick oxide layers or particulates, since it is a surface chemistry process rather than an abrasive one.

Because it melts at around 2,072 degrees Celsius while the aluminum beneath melts at around 660. A process delivering enough energy to melt the metal is working through a layer roughly three times more refractory, so the oxide breaks up unevenly, energy couples inconsistently, oxide fragments become weld inclusions, and the process produces spatter.

Not long. Surface activation begins decaying immediately as the treated surface re-adsorbs contamination from the air, returning toward its untreated state over hours to days. The station must therefore sit immediately ahead of the operation it enables, and treating cells then buffering them for an hour before welding gives away much of the benefit.

Yes, because cells are bonded into the pack structure with adhesive rather than held by module frames, and adhesive strength depends directly on surface energy. A bond formed on a contaminated low-energy surface fails at the interface at a fraction of the adhesive capability, and the same applies to thermal interface materials where poor wetting leaves voids that insulate exactly where heat needs to escape.

Share

Related Blogs