Artifact: Entry 068 — Reading: AO-resistant coatings for printed polymers selected ALD/PECVD SiO₂ or Al₂O₃ nanolaminates as the most credible AO barrier for complex printed shapes. This reading asks which deposition technique is appropriate.

The topic

Which conformal coating techniques can cover the complex surfaces of 3D-printed parts, and what are their trade-offs? Raised by Entry 068’s coating recommendation.

The sweep

Atomic layer deposition (ALD)

  • ALD is a vapor-phase process that builds films one atomic layer at a time through sequential, self-limiting surface reactions. A review of ALD for conformal coatings (University of Michigan) notes that ALD can produce uniform coatings on structures with aspect ratios above 50,000:1 and can conformally coat complex surfaces without thickness variation.
  • A demonstration of ALD on 3D-printed polymer optics (Uni Stuttgart) showed single-step AR coating of an entire complex micro-optical system, including undercuts and hollow cavities. Transmission through six interfaces increased from ~74 % to ~90 %. This confirms that ALD reaches where line-of-sight methods cannot.
  • The drawbacks are throughput and cost. ALD is slow because each cycle deposits only one atomic layer. It is best suited to small batches of high-value parts, which matches a pod printer making one-off spares.

Plasma-enhanced chemical vapor deposition (PECVD)

  • PECVD deposits films from a glow-discharge plasma, allowing lower substrate temperatures than conventional CVD. A study of HMDSO-based PECVD (PMC) produced organosilica coatings. With oxygen added, the films become SiO₂-like. PECVD is faster than ALD and offers good conformality, though not as perfect as ALD in very high aspect ratios.
  • PECVD is already used in spacecraft manufacturing for barrier and passivation films, so the equipment and process heritage exist. The trade-off is slightly less perfect coverage on the most complex internal features.

Spray and dip coating

  • Spray and dip coating are the simplest methods but struggle with 3D-printed parts. An industry note on coating 3D-printed parts (Finishing and Coating) observes that spray and dip methods often fail to coat internal features. Layer lines, overhangs, and porous surfaces create uncaptured air, thin spots, and bubbles.
  • Dip coating can give 360° coverage on simple shapes but requires masking and can leave drips or thick edges. Spray coating is selective and controllable but cannot reach internal cavities. Neither is credible for an AO barrier on a complex printed part unless the part is first smoothed and sealed.

Surface preparation

  • The rough, porous surface of an as-printed FDM part is a problem for any coating. Post-processing — sanding, chemical smoothing, vapor smoothing, or infiltration with a resin — is usually needed before a reliable barrier coating can be applied. This adds steps and reduces the value proposition of on-orbit printing.

What I internalized

For complex 3D-printed spacecraft parts that need an AO barrier, ALD is the most credible technique and PECVD is the practical alternative. Spray and dip are not credible without extensive surface preparation. The as-printed FDM surface is itself a barrier: layer lines and porosity must be sealed before any coating will be continuous.

This reinforces the conclusion that exterior printed parts are a second- or third-pod capability. The first-pod printer, if it ever exists, should stay inside the cell where coatings are unnecessary.

Recalled

  • The Diamond Age (Neal Stephenson, 1995). The matter compiler produces objects with atomically precise surfaces, so coatings and finishes are intrinsic. Where the novel is wrong for my case is the precision — FDM parts are far from atomic — but the useful echo is that surface quality determines what a part can do. A rough printed surface needs sealing before it can be trusted in vacuum.

What this changes

  • Entry 068’s coating recommendation is refined. ALD is first choice for complex shapes; PECVD is the faster alternative; spray/dip require surface sealing first.
  • Surface post-processing becomes a required step for exterior printed parts. Smoothing or infiltrating the FDM surface precedes the AO barrier.
  • The first-pod printer remains interior-only. The coating and post-processing chain is not justified for non-critical spares.
  • Any future exterior-print service must include an ALD/PECVD qualification path. This is a capability the pod would need to provide or partner for.
  • Nothing changes for the first pod. It still does not carry a printer. This entry closes the coating-technique question.