Surface Roughness of Occlusal Splints: Conventional Heat-Cure vs. 3D-Printed vs. CAD/CAM Milled

1. Conventional Heat-Cure (HC) PMMA Splints

• Heat-cured PMMA fabricated via compression molding consistently shows intermediate Ra values in every comparative study. In the 2025 in-vivo JCDR study (Nayak et al.), the conventional group started at Ra ~0.22 µm at baseline and progressed to ~1.07 µm at 6 months of clinical use.
• In the PMC 2025 study (Analyzing Color, Surface Roughness...), the HC group showed Ra of ~0.09 µm on unpolished surfaces, ranking in the middle among all six groups tested.
• The compression-molding process introduces residual porosity and processing shrinkage, which creates microscopic surface irregularities not seen with subtractive (milled) methods.
• Polishing significantly reduces Ra in HC splints, but polished surfaces showed decreased microhardness - a trade-off to consider clinically.
• In the 2026 study by Haugli et al. (PMID 42064378), the autopolymerization/conventional workflow (PalaXtreme) showed low roughness values overall and the least biofilm in the 72-hour experiment - showing that conventional materials can still be competitive when properly finished.

2. 3D-Printed (Additive Manufacturing) Splints

• 3D-printed splints consistently show the highest raw Ra values before polishing. In Nayak et al. 2025, the 3D-printed group had Ra 0.26 µm at baseline (already the highest) and showed the most dramatic roughness increase over time: reaching 1.92 µm at 6 months - nearly double the conventional group at the same timepoint (p<0.001).
• The layer-by-layer build orientation (typically 45° at 100 µm layer thickness for DLP) leaves stacking artifacts that form micro-ridges on the surface.
• However, post-processing dramatically changes this picture:
  • Rueda et al. 2025 (PMID 40181636) showed that polishing and resin-coating of 3D-printed materials produced the smoothest surfaces among all surface treatment conditions tested - even competitive with milled materials.
  • Curing in glycerin reduced staining; optical polish tanks improved surface smoothness of internal surfaces.
  • The unpolished "as-printed" condition is the worst performer; properly finished 3D prints can approach milled surfaces.
• Wuersching et al. 2023 (PMID 36576565) found that 4 out of 5 printable resins tested showed lower Ra and lower bacterial CFU than conventional powder/liquid PMMA - but this comparison was against non-optimally processed conventional material, and the result was material-specific (KeySplint Soft was an outlier with highest Ra among 3D-printed materials).
• The 2025 PMC study confirmed that 3D splints showed both the highest and some of the lowest Ra values depending on brand/resin chemistry - "CAD-CAM fabrication does not necessarily result in superior surface roughness, and the outcome largely depends on material properties."

3. CAD/CAM Milled (Subtractive) Splints

• Milled splints are fabricated from pre-polymerized, industrially pressed PMMA pucks (e.g., Therapon, ProArt CAD Splint, Astron CLEARsplint Disc). The industrial polymerization under high pressure and heat produces a homogeneous matrix with minimal porosity, which is the root cause of their superior surface finish.
• In Nayak et al. 2025: CAD/CAM milled showed Ra 0.17 µm at baseline - the lowest of all three groups - and only progressed to 0.23 µm at 6 months, far more stable than both other methods (p<0.001).
• In the meta-analysis by Valenti et al. 2024 (PMID 38916682 - the highest-level evidence, Systematic Review + Meta-Analysis of 13 studies): Subtractive (milled) specimens had significantly lower average surface roughness compared to conventional (Hedge's g = -1.25; 95% CI -1.84 to -0.66) - a large effect size.
• The PMC 2025 study ranked unpolished milled splints as having the two smoothest surfaces (ML-A and ML-B ranking lowest Ra), with ML-B showing "superior overall performance" combining low Ra, high microhardness, and best color.
• Haugli et al. 2026 confirmed Therapon (milled) had highest Vickers hardness and low surface roughness, though it paradoxically showed the most biofilm in the 72-hour experiment - suggesting hardness and roughness alone do not predict biofilm behavior, and that surface free energy and material chemistry also matter.
• Milled materials also showed less staining than 3D-printed materials (Rueda et al. 2025).

Direct Comparison Table

Which Is Better for Surface Roughness?

1. Post-processing equalizes the gap: A properly polished and resin-coated 3D-printed splint can achieve roughness comparable to milled materials (Rueda et al. 2025, PMID 40181636). The "as-printed" state should never be the final clinical state.
2. Material matters more than method for 3D printing: The 2023 Wuersching study (PMID 36576565) showed some 3D-printed resins outperforming conventional PMMA - it is not a class effect but a material-specific one.
3. Surface roughness threshold for biofilm: The clinically relevant threshold for bacterial adhesion is Ra ~0.2 µm. Milled splints reliably stay below this. Conventional and 3D-printed splints can exceed it over time, particularly 3D-printed ones without regular polishing.
4. Flexural strength caveat: The same Valenti 2024 meta-analysis found that conventional heat-cure PMMA had higher flexural strength than milled materials (Hedge's g = 2.32), meaning the milled splint's smooth surface advantage comes at the cost of slightly lower fracture resistance - a clinically relevant trade-off for bruxism patients.
5. Clinical use degradation: The in-vivo data (Nayak et al. 2025) is particularly telling - over 6 months of actual use, the gap between milled and 3D-printed widens dramatically (0.23 µm vs. 1.92 µm), making milled splints the most clinically durable option for surface quality maintenance.