The bonding of polyoxymethylene (POM) components is constrained by a semicrystalline surface that has low polar free energy and high chemical resistance. On untreated injection-moulded POM, water contact angle measured under ISO 19403-2:2017 is typically 70–80°, total surface free energy is 36–40 mN/m, and the polar component is below 10 mN/m. Injection moulding also deposits mold-release agents, waxes, and low-molecular-weight processing aids in a thin boundary layer that interferes with adhesive wetting. Because POM homopolymer and copolymer both resist common solvent etching and do not develop sufficient mechanical interlock from abrasion alone, adhesive bond strengths remain low unless oxidation or topography is created at the surface. Accurate process selection therefore requires matching surface treatment to POM bonding: plasma, flame and etching.

The untreated surface fails most structural adhesive qualification tests because adhesives such as two-part acrylics, toughened epoxies, and reactive polyurethanes require a total surface free energy above 50 mN/m and a polar component above 15 mN/m to achieve intimate wetting. POM also exhibits crystallinity between 60% and 80% depending on grade and thermal history, which reduces amorphous free volume and prevents adhesive diffusion. Solvent wiping with 99.9% isopropanol can remove mold release but does not oxidize the surface or create micro-roughness. In lap-shear testing under ASTM D3163 or ISO 4587:2003, untreated POM bonded with structural acrylic or epoxy frequently fails adhesively below 1 MPa, with failure occurring at the adhesive-POM interface rather than cohesively within the adhesive.

Why Does Untreated POM Fail Adhesive Qualification Tests?

The adhesion failure mode is explained by the low polar contribution to POM surface energy. Contact-angle measurements under ASTM D5946 or ISO 19403-2:2017 show that the dispersive component dominates the total surface free energy, and the polar component is insufficient for hydrogen bonding with adhesive chemistries. Without polar interaction, the adhesive film retracts or forms weak boundary regions at the interface. Mechanical interlocking is also limited because the moulded POM surface is smooth and hard, and solvent-based bonding systems do not swell the amorphous phase enough to create diffusion bonding. Even when aggressive solvent wiping is applied, molded-in stresses and surface contamination can remain, producing high batch-to-batch variance in adhesion tests.

In production lines operating with 13.56 MHz RF low-pressure plasma chambers, POM surfaces are exposed to oxygen or oxygen–argon mixtures at 0.2–1.0 mbar and 100–300 W RF power for 60–300 s. Plasma-generated atomic oxygen abstracts hydrogen from the polymer backbone, forming carbon radicals that react with oxygen to produce hydroxyl, carbonyl, and carboxyl groups. This oxidation raises total surface free energy to 50–65 mN/m and increases the polar component to 15–30 mN/m. XPS analysis of treated POM shows an increase in carbon–oxygen bonding states in the 287–289 eV region, corresponding to carbonyl and carboxyl functionality. Plasma treatment also removes mold-release contaminants by sputtering and chemical volatilization, which reduces the weak boundary layer. Batch vacuum plasma chambers require careful rack configuration because stacked or shielded parts can remain untreated; witness coupons placed at chamber edges and lower rack positions are used to verify uniformity before production release.

Atmospheric plasma provides an in-line alternative for POM components that are flat or have limited three-dimensional geometry. Dielectric barrier discharge systems operating at 20–50 kHz with a nozzle gap of 2–5 mm and travel speed of 50–300 mm/s typically generate post-treatment surface free energies of 48–56 mN/m. Atmospheric plasma is line-of-sight and sensitive to ambient humidity; high relative humidity above 60% can reduce radical density and produce less reproducible oxidation. Hydrophobic recovery begins after treatment as polar groups reorient into the bulk or become contaminated by airborne organic species. For both vacuum and atmospheric plasma, the practical adhesion window is usually 4 h after treatment when parts are held at 23 ± 2°C and 50 ± 5% RH. Longer storage requires dry, low-lint packaging, but bond strength may still decline within 24–72 h. Pre-cleaning in an aqueous alkaline bath at 50–60°C is therefore specified before plasma treatment to eliminate mold release and prevent variable surface oxidation.

Flame Oxidation of Injection-Moulded POM Housings and Structural Ribs

Flame oxidation relies on short exposure of the POM surface to a premixed propane–air flame. Production parameters are typically an air-to-propane ratio of 9:1–11:1, gas flow of 20–60 L/min, burner gap of 5–15 mm, and dwell time of 0.05–0.2 s. The flame temperature in the oxidizing zone approaches 1,700–1,900°C, but the POM surface must remain well below its melt temperature because POM depolymerizes by unzipping and releases formaldehyde when overheated. A drilled-port or ribbon burner mounted on a servo-controlled slide maintains gap distance, because a variation of ±1 mm can change measured surface energy by more than 5 mN/m. Post-treatment surface free energy for optimized flame conditions is typically 45–58 mN/m, with polar component from 10–20 mN/m.

Flame treatment should not be confused with flame polishing. Flame polishing melts the POM surface into a glossy skin and reduces the polar functionality needed for bonding. For flame oxidation, the part surface is passed through the oxidizing region just beyond the inner cone; contact with the luminous outer cone is avoided because soot deposition can recreate a weak boundary layer. Thin-walled sections below 2.0 mm require a surface temperature limit of 120°C, while thicker sections are normally controlled below 140°C using an infrared pyrometer. Multiple short passes with inter-stage cooling are preferred over a single long exposure. Flame treatment on unfilled POM can cause local annealing of molded-in stress and warpage on flat housings; for parts with ISO 2768-mK tolerances, dimensional validation after first-off treatment is required. Recessed areas, snap-fit undercuts, and internal ribs are not treated because flame treatment is line-of-sight.

Chemical etching of POM follows a different mechanism than oxidative gas-phase treatment. Strong oxidizing acids, historically chromic acid solutions, preferentially attack amorphous POM domains and create micro-roughness that supports mechanical interlocking. A typical chromic acid immersion process operates at 60–70°C for 5–15 min, depending on bath age and part geometry. Etch depth is controlled between 1–5 µm per surface to avoid dimensional changes on precision parts. Weight loss can vary from 0.2–1.0 wt%, and bath analysis by redox titration is performed at intervals not exceeding 8 h. After etching, parts are rinsed in 18 MΩ·cm deionized water at 50–60°C, neutralized if required, and force-dried below 60°C. Final rinse conductivity is typically held below 10 µS/cm to prevent acid residue in blind holes or moulded-in features.

Hexavalent chromium etchants are subject to authorization under REACH Annex XIV, so closed-loop etch lines with fume suppression and ion-exchange recovery are used where chromic acid remains permitted. Chromium-free acid blends based on sulfuric and phosphoric acid at 70–85°C can etch POM but often require longer immersion times and produce less aggressive micro-roughness; published data for this specific configuration is limited. Chemical etching treats all wetted surfaces, including internal channels and complex recesses that are shadowed from plasma or flame. However, trapped air in blind holes below 2 mm diameter can prevent acid ingress unless parts are agitated at 10–30 rpm and oriented to release air. Ultrasonic agitation at 25–40 kHz improves wetting but can damage sharp POM edges by cavitation. Etched POM surfaces typically show surface free energies of 50–60 mN/m and can retain wettability longer than plasma-treated surfaces if protected from contamination.

If Complex Internal Channels Require Treatment, Etching May Outperform Line-of-Sight Plasma

Line-of-sight processes leave shadowed recesses untreated. Recesses with an aspect ratio above 1:1 or holes smaller than 2 mm diameter generally show water contact angles above 60° after plasma or flame treatment. Etching is a wet immersion process and reaches surfaces that cannot be oxidized by a flame or plasma nozzle, but it introduces dimensional change and rinse complexity. For precision gears and thin-walled parts with ISO 2768-mK tolerances, plasma treatment is usually selected because dimensional change is not measurable when substrate temperature is kept below 120°C. For adhesive-bonded manifolds, internal flow channels, or parts with undercuts, chemical etching may be the only process that creates uniform topography. The selection decision is therefore governed not only by surface energy but by geometry, tolerance stack, production rate, and chemical handling restrictions.

Comparative Processing Data for Plasma, Flame, and Etching

Following any of the three surface treatments, production release testing should include contact-angle measurement on flat witness coupons under ISO 19403-2:2017 or ASTM D5946, and lap-shear testing under ASTM D3163 or ISO 4587:2003. Adhesive application is allowed only when total surface free energy is at least 50 mN/m and water contact angle is not more than 30°. SPC control charts for water contact angle are established with limits of target ±5° from first-off, in-process, and last-off samples. For bonded assemblies, environmental durability is evaluated under ISO 4892-2 or ASTM D4587, with typical exposure at 85°C/85% RH for 500 h. Retention of lap-shear strength is adhesive-dependent and must be validated for each POM grade and surface treatment combination; published data for this specific configuration is limited.

POM linear expansion coefficients of 90–110 × 10⁻⁶ K⁻¹ require bond-line control from 0.2–0.5 mm or a toughened adhesive to accommodate thermal strain. Plasma- and flame-treated parts are scheduled for bonding within 4 h at 23 ± 2°C and 50 ± 5% RH. Etched parts can tolerate up to 24 h if stored in low-lint, sealed packaging. Acid etching should not be applied to impact-modified POM grades without first screening a molded plaque, because modifier phases may dissolve or swell in the etch bath and create voiding at the surface. If contact angle or lap-shear results fall below control limits on production samples, the treatment system must be checked for gas flow, power, burner gap, bath temperature, or rack loading before further parts are processed.