Guides / Sourcing

Reading PEEK Outgassing Data Before You Trust It

September 26, 2026

Sources: manufacturer datasheets and published standards. See editorial standards.

A vacuum chamber or a fab tool does not care what a plastic’s datasheet says in general. It cares what the exact part sitting inside it releases once the chamber is sealed and pumped down. PEEK earns a place in vacuum hardware because it outgasses less than most engineering plastics, but “less than most” and “low enough for your tool” are two different claims, and the gap between them is where outgassing data gets misread.

What TML and CVCM actually measure

The standard screening test is ASTM E595, run by baking a sample at 125°C under vacuum for 24 hours and weighing it before and after. Two numbers come out of that bake:

• TML (total mass loss): everything the sample gave off, as a percentage of its starting weight
• CVCM (collected volatile condensable material): the fraction of that mass loss that condenses on a cold collector nearby, which is the part that actually matters for contaminating a wafer, a mirror or a sensor downstream

A common screening threshold from aerospace and vacuum work is under 1.0 percent TML and under 0.10 percent CVCM. Unfilled PEEK typically screens comfortably under both figures, which is why it shows up on vacuum-compatible material lists in the first place. The test says nothing, though, about your bake schedule, your chamber’s base pressure or how long your tool runs before the part is replaced. It is a standardized screening number, not a simulation of your process.

A generic figure is a starting point, not proof

The number on a resin producer’s datasheet came from one sample, tested once, usually a molded plaque of the base unfilled grade. A few things separate that plaque from the part on your drawing:

Processing history. Machined stock carries residual moisture and surface condition different from a freshly molded plaque, and a part cut from extruded rod has a different thermal history than one from resin pressed into a test coupon.

Surface area to volume ratio. A small machined fixture, comb or insulator with slots, holes and thin fingers has far more exposed surface per gram than a flat plaque. Outgassing scales with surface area more than with mass, so a geometrically complex part can outgas more per gram than the datasheet sample even in the same grade.

Lot to lot variation. Moisture content at the time of bagging, storage conditions before machining and how long a part sat exposed to air before installation all shift the result. This is the same lot traceability question covered in verifying the PEEK you are buying, just applied to a different property.

None of this means the generic figure is wrong. It means it is a starting point for a material family, not a certificate for the part in your hand.

Why filled grades do not behave like the base resin

Most fab hardware machined from PEEK is CF30, chosen for static dissipation and dimensional stability rather than for vacuum performance, and CF30 does not automatically inherit unfilled PEEK’s outgassing number.

Carbon fiber ends get exposed and cut open at every machined surface, and the interface between fiber and resin matrix is where adsorbed moisture tends to sit. That moisture bakes off under vacuum along with whatever the resin itself releases, and a heavily machined part with many fiber ends at the surface has more of that interface exposed than a molded coupon ever would.

Bearing grade PEEK adds PTFE and graphite on top of that, each with its own outgassing behavior, and glass filled grades carry a similar interface question with glass rather than carbon. The practical result is the same across all of them: a filled grade’s outgassing is a property of that specific compound, not an extension of the unfilled resin’s number, so it needs its own data rather than an assumption.

What to request before you trust a number

• Outgassing data for the specific grade you are buying, not the generic “PEEK” figure from a resin family datasheet
• Whether the data was measured on a machined part or a molded plaque, since geometry changes the result
• Lot-specific test data where the application is contamination sensitive, rather than a figure from an unrelated batch
• An independent ASTM E595 run on your actual machined geometry when the tool is high vacuum or particularly contamination sensitive, since a few hundred dollars of testing is cheap against a contaminated chamber
• Whether the tool’s install procedure calls for a bake-out or degas step before the part goes in, and to what schedule

Where this actually matters

Vacuum-adjacent PEEK parts, chuck insulators, feedthrough hardware and hot-zone fixtures, are covered at a component level in PEEK parts for semiconductor equipment. Outgassing sits alongside particle generation and metallic contamination as one more line item that belongs on the drawing rather than assumed from a material name.

Send the drawing with the vacuum or process conditions the part will see, and we match you with shops that can get outgassing data for the grade you specify.

Frequently asked questions

Keep reading

• Clearance, Grade and Fit for PEEK Pump Wear Rings and Bushings
• What a Shop Needs to Machine Spinal Cage Prototypes in PEEK