Vacuum Coating Machine Process: Steps, Vacuum Levels, and Cycle Control

What the Process Actually Controls

Strip away the hardware and a vacuum coating machine process does one job: it removes everything from the chamber that you did not ask to be there, then lets a chosen material arrive at the part surface under controlled energy. The film you get is the result of three variables — background pressure, ion energy at the substrate, and the rate and time of deposition. Every other setting, from argon flow to jig rotation, is only a way of tuning those three.

That is why two shops can run "the same" PVD recipe and see different results. A chamber held at 3 x 10⁻³ Pa without substrate heating still carries a water-vapour load that competes with the coating for bonding sites. The same chamber at 4 x 10⁻⁴ Pa after a proper bake-out behaves like a different machine. Pressure alone is not the number that matters; the ratio between residual gas load and deposition rate is.

Vacuum is not the goal of the process. It is the filter that decides what else is allowed to touch the part before the film does.

The Coating Cycle, Step by Step

Most batch cycles follow the same sequence, whether the machine uses cathodic arc, magnetron sputtering, or thermal evaporation. The differences sit in the middle of the cycle, not at the ends.

1. Pre-cleaning and loading. Ultrasonic degreasing, rinsing, and drying remove oils and polishing compound. Parts are then fixed on jigs that also define rotation — a fixture that hides a surface from the plasma will hide it from the coating too.
2. Roughing and high-vacuum pump-down. A rotary vane or dry pump takes the chamber from atmosphere to roughly 1–10 Pa, then a roots, diffusion, or turbomolecular pump carries it to base pressure, commonly between 5 x 10⁻³ and 5 x 10⁻⁴ Pa.
3. Heating and outgassing. Chamber walls and substrates are heated, often to 150–400 °C depending on the film, to drive off adsorbed water and volatiles. Skipping this stage is one of the most common reasons a coating blisters weeks later.
4. Ion bombardment (glow discharge cleaning). Argon is admitted and a negative bias of several hundred volts is applied to the parts for a few minutes. This sputters away oxide and the last molecular layer of contamination and leaves a chemically active surface.
5. Deposition. The source is energised — arc spots on a cathode, a magnetron target, or an evaporation boat — and material travels to the substrate. Reactive gases such as nitrogen, oxygen, or acetylene are added here when a nitride, oxide, or carbon-based film is required.
6. Thickness control. Film thickness is set by rate multiplied by time, confirmed by quartz crystal monitoring on optical coatings and by calibration runs on decorative or wear films.
7. Cooling under vacuum. Parts are held in vacuum until they fall below roughly 80–100 °C. Venting early oxidises the fresh film, and in aluminium metallising it is the classic cause of yellowing.
8. Venting, unloading, and inspection. Dry nitrogen or filtered air breaks the vacuum, parts are unloaded, and the batch is checked for colour, thickness, adhesion, and surface defects.

How the Deposition Method Reshapes the Process Window

The middle of the cycle is where methods diverge, and that choice sets the practical limits on rate, uniformity, and defect risk.

A decorative line chasing colour consistency usually settles on multi-arc deposition, while an optical filter line cannot use arc at all, because a single droplet destroys the layer stack. Process selection is therefore a trade between speed, surface quality, and the geometry of the parts in the fixture.

The Numbers Worth Recording on Every Batch

Process drift is rarely dramatic. It creeps. A machine that logs the same eight values on every run can usually trace a defect back to its origin within one or two batches instead of one or two weeks.

• Base pressure before argon admission, plus the time taken to reach it
• Working pressure during deposition, with the tolerance band you accept
• Substrate temperature at the start and at the end of deposition
• Bias voltage and duty cycle, which set ion energy and film density
• Source current or power for each cathode or evaporation boat
• Reactive gas flow and the resulting film stoichiometry
• Jig rotation speed and the number of passes through the plasma
• Deposition time or accumulated thickness reported by the monitor

Where Common Faults Actually Begin

Most coating defects are process defects rather than equipment defects, and they appear in predictable places.

• Poor adhesion — incomplete pre-cleaning, an under-heated load, or an ion etch that is too short or too weak.
• Pinholes and particles — flaking material from chamber shields, dirty fixtures, or a target that was not conditioned after a change.
• Colour deviation between batches — drifting reactive gas flow, target erosion, or thickness variation caused by changing fixture density.
• Non-uniform thickness — uneven target wear, wrong source-to-substrate distance, or jigs that shadow the plasma.
• Yellowed or darkened aluminium — residual oxygen and water vapour in an evaporation cycle, often from venting before the load has cooled.
• Target poisoning in reactive sputtering — too much reactive gas, which builds a compound layer on the target and collapses the deposition rate.

Matching the Process to the Part

The part decides more of the recipe than the machine does. Three groups cover most production requests, and each one pushes the process in a different direction.

Tools, moulds, and wear parts

Cutting tools, forming dies, and transmission components need 1–5 µm of a hard nitride or carbon-based film and a substrate temperature high enough to build a metallurgical bond. Cycles run long, adhesion testing is non-negotiable, and a dedicated hard-coating platform is usually more economical than adapting a decorative machine.

Automotive lighting and trim

Reflectors and lamp housings combine a base coat, an aluminium reflective layer, and a protective top coat. The base coat matters as much as the metal, because it hides moulding marks that would otherwise show through a mirror finish. Wheels, trim strips, and badges follow the same logic, with colour consistency as the main acceptance criterion.

Decorative metal, sanitary ware, and hardware

Fittings, tableware, watch bands, and door hardware live or die on colour repeatability, uniformity across polished surfaces, and increasingly on anti-fingerprint performance. Sputtering gives tighter control over colour and thickness than arc deposition, at the cost of a slower cycle and more careful target maintenance.

Cycle Time, Throughput, and the Cost Behind the Process

Pump-down and outgassing can consume 30–50 percent of a batch cycle, and deposition often takes less than half the time on the clock. That ratio explains why jig design, loading density, and pumping speed move cost per part more than a marginal gain in deposition rate. When cycle time becomes the constraint, the honest options are a larger chamber, a second batch position, or a continuous inline system for parts that can be handled in one orientation.

Operating cost also hides in the work nobody schedules: pump oil changes, target replacement and utilisation, shield cleaning downtime, and the power drawn by diffusion pumps that stay hot between shifts. A machine that is technically capable but demands weekly mechanical intervention rarely delivers the lowest cost per coated part. That is the practical argument for measuring equipment against your own part mix and batch size, as covered in this guide on how to choose a PVD vacuum coating machine, before comparing headline deposition rates.

What a Stable Process Looks Like

A mature coating process is boring in the best sense. Base pressure returns to the same figure in the same time, colour readings stay inside a narrow band across shifts, and maintenance intervals are predictable rather than reactive.

For manufacturers specifying new capacity, that means buying a documented process window rather than a chamber with a target inside. PVD vacuum coating systems should be judged on how repeatably they hold pressure, bias, and temperature through a full production day, because those three numbers are what the film actually remembers.