Rigid-Flex PCB Air-Gap Process Optimization: Lamination, Film and Yield
Rigid-flex PCB air-gap process optimization targets one core problem: height differences on the substrate surface. After pure adhesive drilling, panel layout, lamination, mechanical drilling and plating, the air-gap surface forms raised and recessed areas. Dry film then fails to bond at the copper bottom corners, causing bubbles, peeling and open circuits. This study tests process methods that reduce those defects. The optimized flow is built around selective non-bonding, clean window edges, a controlled press curve, and inspection that proves the air gap did not fill with resin.
Why Rigid-Flex Air-Gap Design Creates Surface Height Differences
A rigid-flex air-gap design leaves certain layers unbonded in part of the flex area. As a result, the bending section can move more freely. However, the same structure leaves the surface uneven after panel layout and lamination.
A Rigid-Flex PCB with an air-gap structure runs through a fixed sequence: pure adhesive drilling, panel layout and lamination, mechanical drilling, and plating. Each via metallization step adds to the height difference. Therefore, the air-gap surface ends up with raised and recessed zones at the copper bottom corners.
Figure 1: Introduction to the Design of Layered (Air Gap) Stacked Structures
Figure 2： Schematic diagram illustrating the mechanism behind the elevation differences created by the layered design
Figure 3: Schematic diagram illustrating the mechanism behind the height difference formed after via metallization
Rigid-flex boards with an air-gap structure undergo a process involving pure adhesive drilling, panel layout and lamination, mechanical drilling, and electroplating; as a result, the substrate surface develops areas with height differences characterized by raised and recessed sections.
At the bottom corners of the copper on the air-gap surface, the dry film may not adhere properly, appearing as bubbles.
The dry film fails to bond tightly to the copper surface, resulting in bubbles after film application, peeling after development, and open circuits or notches after etching.
The initial SET yield for the target product was only 87.4%. To address the issues caused by air gaps, this paper explores the processes and methods for manufacturing rigid-flex circuit boards with air-gap designs through testing and research.
Figure 4: Schematic diagram showing the height difference between the shelves
Experimental Section
Laboratory Instruments and Equipment
| Experimental Equipment & Instruments | Model | Manufacturer |
|---|---|---|
| Horizontal electroplating line | HKD-ALH20P | Kede |
| Mechanical drilling machine | ND-6L180E | Hitachi |
| Dry film etching machine | 08SF20NFAAA01A1 | Dongguan Yuyu Circuit Board Equipment |
| Wet film laminating machine | YTL-M630 | Duantian Chemical |
| Vacuum laminating machine | MWZK-ONE610 | Zhongshan Mingwei Automation Equipment Co., Ltd. |
| Precision semi-automatic printer | ATMA 0E56 | Dongyuan |
| Scanner and inspection machine | Ultra fusion200 | KLA |
Table 1. Experimental Instruments and Equipment
Test Materials
Resist dry film (30/40/50 μm thickness), wet film (A-9000), and two-layer (1+1 stack-up) flexible printed circuit boards.
| Material No. | Stack-up Structure | Material Type | Material Supplier | Specification |
|---|---|---|---|---|
| 40630 | 1+1 | L1 FCCL | Thinflex | 18 μm Cu / 12.5 μm PI |
| AD Adhesive | Taiflex | 25 μm AD Adhesive | ||
| L2 FCCL | Thinflex | 18 μm Cu / 12.5 μm PI |
Table 2. Two-Layer Board Stack-up Information
Table 3: Experimental Board Design Specifications
Experimental Process Flow
Main experimental process flow: Layout of two-layer boards → Drilling of through-holes → Copper plating of holes → Pretreatment → Film application → Exposure → Etching → Optical inspection.
flow chart
The study primarily focuses on layout, pretreatment, and pattern formation (including film application, exposure, and development/etching).
A Study of Different Layout Methods
Common lamination methods for flexible printed circuits include high-speed press lamination and transfer press lamination.
Transfer press lamination includes methods such as aluminum-sheet lamination, 160-layer lamination, and three-in-one lamination.
| Lamination Method | Lay-up Method | Experiment No. |
|---|---|---|
| Fast Lamination | Aluminum sheet lay-up | 1 |
| Fast Lamination | Aluminum sheet lay-up | 2 |
| Conventional Lamination | 160 lay-up | 3 |
| Conventional Lamination | Three-in-one lay-up (30 μm matte release film) | 4 |
| Conventional Lamination | Three-in-one lay-up (50 μm transparent release film) | 5 |
Table 4: Experimental Plan for Lamination Methods
Figure 5: Method for stacking aluminum sheets under high pressure
Figure 6: Method for laminating 160 layers of plates using a press
Figure 7: Three-in-one laminate pressing method using a press
Table 5: Results of graphic production using different stacking methods
Note: Figures 4, 5, and 6 show the lamination methods described in this article.
Based on the experimental results, the bubble rate and yield rate are as follows: pressure transfer + 3-in-1 lamination (50 μm transparent film) > pressure transfer + aluminum sheet lamination > pressure transfer + 3-in-1 lamination (30 μm matte film) > pressure transfer + 160 lamination > quick press + aluminum sheet lamination.
However, reliability testing of the rapid lamination + aluminum sheet stacking method revealed issues such as insufficient compression and panel rupture.
Therefore, pressure lamination combined with 160-layer stacking proved to be the optimal layout method.
However, changing the layout alone does not improve the target product’s yield rate; other methods must provide further enhancements and improvements.
A Study of Common Processes in the Pattern Transfer Process
The current main process flow for pattern transfer is: pretreatment → film application → exposure → development → etching.
This paper fixes the production of the dry film for circuit lines as a wet film application process; a wet film applicator applies the film, as shown in Figure 8.
Figure 8: Wet Film Coating Machine
Table 5: SEM Results for Different Pretreatments
Figure 9: Schematic diagrams of horizontal and vertical film application
Based on the known factors affecting the existing pattern transfer process, a DOE (Design of Experiments) analysis was conducted, as shown in Table 6.
This analysis investigated the impact of different pretreatments, dry film thicknesses, and film application methods on the yield of the target circuit pattern fabrication, thereby identifying the most effective optimal parameters and conditions for pattern transfer.
| Pretreatment Method | Dry Film Thickness (μm) | Lamination Direction |
|---|---|---|
| Chemical Cleaning | 30 | Horizontal Lamination |
| Roughening | 40 | Vertical Lamination |
| Mechanical Brushing | 50 | N/A |
Table 6: Experimental Factors and Levels
| Pretreatment Method | Dry Film Thickness (μm) | Lamination Direction | SET Yield (%) |
|---|---|---|---|
| Roughening | 30 | Horizontal | 94% |
| Roughening | 30 | Vertical | 73% |
| Chemical Cleaning | 30 | Horizontal | 86% |
| Chemical Cleaning | 30 | Vertical | 66% |
| Mechanical Brushing | 30 | Horizontal | 69% |
| Mechanical Brushing | 30 | Vertical | 59% |
| Roughening | 40 | Horizontal | 81% |
| Roughening | 40 | Vertical | 72% |
| Chemical Cleaning | 40 | Horizontal | 86% |
| Chemical Cleaning | 40 | Vertical | 94% |
| Mechanical Brushing | 40 | Horizontal | 73% |
| Mechanical Brushing | 40 | Vertical | 83% |
| Roughening | 50 | Horizontal | 72% |
| Roughening | 50 | Vertical | 52% |
| Chemical Cleaning | 50 | Horizontal | 86% |
| Chemical Cleaning | 50 | Vertical | 86% |
| Mechanical Brushing | 50 | Horizontal | 66% |
| Mechanical Brushing | 50 | Vertical | 80% |
Table 7: DOE Experimental Results
A statistical analysis of the DOE test data shown in Figure 9 reveals the following yield rates: for different dry film thicknesses, the yield rate for 40 μm > 30 μm > 50 μm; for different pretreatments, the yield rate for super-coarsening > chemical cleaning > mechanical brushing; and for different film application methods, the yield rate for vertical application > horizontal application.
A comprehensive analysis indicates that super-roughening combined with a 40 μm thick dry film (AQ-4088) and vertical application achieves the best circuit yield.
However, defects such as poor exposure and open circuits still occur in the air-gap regions, requiring the incorporation of additional improvement methods into the standard film application process.
Figure 10: Analysis of DOE Experimental Results
Study using the dry film + back pressure method
Principle of dry film lamination: A laminator presses a photosensitive dry film onto the copper surface.
The dry film is the key material for circuit formation. It consists of a three-layer structure: a polyester film, a photosensitive resist layer, and a polyethylene protective film, as shown in Figure 12.
Figure 11: Schematic of a dry film structure
Figure 12: Dry film lamination process
The key component of the dry film is the photoresist layer, while the polyester film and polyethylene protective film both serve as protective layers.
The photoresist layer exhibits excellent filling properties, allowing a dry film laminator to laminate it onto the board surface. Figure 13 shows the dry film lamination process:
Figure 13: Air gap (layered) region air pocketscompaction issues
When an air-gap (layered) design is used, the board surface becomes uneven.
The photoresist layer of the dry film cannot fill the height differences created by the layered design, resulting in bubbles or areas where the film does not compact properly.
Attempted Process Improvement: Secondary Pressing
Based on the phenomenon shown in Figure 13, we will attempt to apply a second pass of pressure to the board with the dry film already applied to enhance the bond between the dry film and the board surface.
We will conduct tests to determine whether this improves issues such as bubbles and poor adhesion. Table 8 below shows the test plan.
| Experimental Scheme | Specific Method | Remarks |
|---|---|---|
| Re-pressing after lamination | Re-press before exposure | / |
| Re-press after exposure | / | |
| Re-press both before and after exposure | / |
Table 8: Dry Film Re-Pressing Plan
Table 9: Dry Film Back Pressure Test Results
The experimental results show that after applying the dry film to the board surface with an air-gap (layered) design, re-pressing the board with the hot roller of the laminator did not improve the presence of bubbles or poor adhesion on the board surface; instead, it resulted in a large number of exposure defects and over-etching.
Applying dry film followed by re-pressing cannot resolve pitting and surface unevenness caused by the air-gap (layered) design; therefore, other process solutions need testing.
Air-Gap Lamination in Rigid-Flex: Drilling, Layup and Pressing
Air-gap lamination rigid-flex control starts before the panel is built. First, pure adhesive drilling removes material at the designated positions. Next, panel layout and lamination bond the stack. Mechanical drilling then opens the vias, and plating completes the metallization.
Every step leaves a mark on the surface. For example, an uneven press result shows up again after drilling and plating. Therefore, the whole sequence must be judged as one chain, not as separate operations. This is the same discipline described in our Rigid-flex PCB manufacturing overview.
Which layers are intentionally left unbonded in the flex zone
In a 1+1 or 2+2 rigid-flex, the layers intentionally left unbonded in the flex window are the bonding sheet (no-flow prepreg) between the flex core and the rigid outer layer, the coverlay adhesive over the flex conductors, and any stiffener adhesive under a stiffener that crosses the bend. The PI film and copper of the flex core remain; the adhesive bondline is omitted. A glue-stop layer or selective prepreg cutout defines the non-bonded window, while the rigid-to-flex transition stays fully bonded.
Test stack-up used in this study
The trial used two-layer (1+1) flexible printed circuit boards. L1 and L2 are both FCCL with 18 μm Cu and 12.5 μm PI. A 25 μm adhesive layer sits between them.
| Material No. | Stack-up | Layer | Material type | Supplier | Specification |
|---|---|---|---|---|---|
| 40630 | 1+1 | L1 | FCCL | Thinflex | 18 μm Cu / 12.5 μm PI |
| 40630 | 1+1 | AD | Adhesive | Taiflex | 25 μm |
| 40630 | 1+1 | L2 | FCCL | Thinflex | 18 μm Cu / 12.5 μm PI |
When a project needs a thinner bend zone, Custom Flex PCB stacks usually follow the same material logic.
Selective Non-Bonding, Pre-Stack Registration and Press Curve
Step 1: Selective non-bonding and glue-stop layer
Use a glue-stop layer or selective prepreg cutout to define the unbonded window. The coverlay opening should be 0.2–0.5 mm larger than the air-gap window so adhesive flow does not enter the gap. The glue-stop width should be at least 0.3 mm around the window.
Step 2: Pre-stack registration and riveting
Register the stack with fiducials and tooling pins. Rivet outside the air-gap window. Use release film and conformal padding so local pressure peaks do not collapse the window edge.
Step 3: Press curve and resin flow control
Use a slow ramp, staged pressure, and vacuum lamination. The press curve should gel the adhesive before full pressure, then hold to cure. Monitor resin bleed at the window edge; if bleed reaches the gap, reduce pressure or widen the glue-stop.
Process Window and DOE Parameters
| Parameter | Starting window | DOE check | Risk if out |
|---|---|---|---|
| Lamination temperature | 170–190°C | Ramp to peak and hold | Under-cure or void |
| Pressure | 1.5–3.0 MPa | Step pressure | Resin starvation or excess flow |
| Ramp rate | 2–3°C/min | 1–5°C/min | Bubbles or resin squeeze-out |
| Resin bleed / glue overflow | 0.10–0.30 mm | Window edge measurement | Adhesive enters air gap |
| Window tolerance | ±0.10 mm | Laser vs route capability | Film bridging or step |
| Laser/routing opening precision | ±0.05–0.10 mm | Capability study | Offset opening |
Why Dry Film Will Not Sit Flat at the Air-Gap Edge
At the copper bottom corners of the air-gap surface, dry film may not adhere properly. It appears as bubbles right after film application. Later, the film peels after development. Finally, open circuits or notches appear after etching.
One surface defect therefore triggers three quality problems. The root cause is the same each time: the film cannot wrap the step created by the air gap. Our PCB Etching process guide explains how a broken resist layer turns into a trace defect. Wet film can conform better, but it needs its own process window; do not assume it solves a bad step.
Defect Map: Cause and Process Direction
| Symptom | When it appears | Direct cause | Process direction |
|---|---|---|---|
| Bubbles after film application | After lamination of dry film | Film not tight at the copper bottom corner | Match film type to surface step |
| Peeling | After development | Weak adhesion to copper | Control pretreatment and press quality |
| Open circuit or notch | After etching | Resist layer breached | Keep the mask complete over the step |
All three symptoms trace back to the height difference created by the air gap. Therefore, surface control comes first, and film selection comes second.
Yield Quantification for Air-Gap Defects
| Defect | Mechanism | Detection | Countermeasure | Target |
|---|---|---|---|---|
| Delamination at air-gap edge | Weak bond or moisture at window edge | C-SAM, cross-section | Plasma clean, glue-stop, press curve | 0 major |
| Bubbles after film lamination | Film cannot wrap copper step | Visual after lamination, AOI after develop | Thinner dry film, wet film, step coverage | <0.5% |
| Resin bleed into air gap | Over-press or high resin flow | Cross-section, window measurement | Glue-stop, prepreg cut, pressure control | 0 bleed >0.2 mm |
| Window offset | Registration drift or laser offset | AOI, dimensional check | Tooling pins, fiducials, laser offset | ±0.10 mm |
| Open circuit after etch | Resist breach at step | AOI, electrical test | Film match, edge coverage | <0.3% |
Inspection, Yield Verification and Process Window
The study used a horizontal electroplating line (HKD-ALH20P, Kede), a mechanical drilling machine (ND-6L180E, Hitachi), a dry film etching machine (08SF20NFAAA01A1A, Dongguan Yuyu), a wet film laminating machine (YTL-M630), a vacuum laminating machine (MWZK-ONE610, Zhongshan Mingwei), a precision semi-automatic printer (ATMA 0E56, Dongyuan) and a scanner and inspection machine (Ultra fusion200, KLA).
The initial SET yield of the target product was only 87.4%. In other words, the air-gap surface accounted for the main share of loss. Also, the tested resist films covered 30, 40 and 50 μm dry film plus A-9000 wet film, so both routes could be compared on the same board design. For the optimized flow, the target is ≥95% SET yield with air-gap defects below 0.5% and no resin bleed beyond 0.2 mm. The next stage of this work is covered in rigid-flex PCB air-gap yield.
Inspection, Acceptance and Reliability Testing
| Check | Method | Acceptance criterion | Sampling |
|---|---|---|---|
| Air-gap window and glue-stop | Cross-section | No resin bleed into gap; bondline matches drawing; window offset within ±0.10 mm | 1 per panel per shift |
| Delamination | C-SAM | No unbonded area >2 mm in bonded zone; no gap-edge lift | First article + 5% per lot |
| Via plating | Cross-section | Full barrel coverage; no voids or cracks | 1 per lot |
| Thermal cycling | -40 to +125°C, 100 cycles | No delamination, blister, or crack; resistance change <10% | 3 pcs per lot |
| Bend fatigue | Mandrel bend at qualified radius per IPC-2223 | No open circuit; resistance stable; no coverlay lift | 3 pcs per design |
Design and Process Rules That Travel Together
Because the rigid-flex air-gap design decides where the steps appear, a design choice changes the film result directly. Therefore, design review and process review should run in the same meeting. This link is also covered in our Rigid-Flex PCB Technology notes.
The same logic applies to neighbouring steps. Coverlay bonding strength and etch control both depend on a clean, flat copper interface. In addition, adding an air-gap step increases the number of operations and the yield ramp risk. Thus, when the bend requirement is modest, a simpler stack may cost less. For general stack and outer-layer questions, see our PCB Manufacturing FAQ.
Design-to-process linkage rules
| Design rule | Process requirement | Reason |
|---|---|---|
| Air-gap length and bend radius | Keep gap inside bend neutral zone; if R < 1.5–2× stack thickness, review air-gap need | Lower copper strain |
| Coverlay opening | Opening edge 0.2–0.5 mm outside air-gap edge | Prevent adhesive flow into gap |
| Glue-stop width | ≥0.3 mm around window | Control resin bleed |
| Via placement | No vias in unsupported air gap | Avoid barrel crack |
| Trace direction | Follow IPC-2223 bend-area guidance and keep grain orientation consistent | Reduce fatigue risk |
FAQ
Why does the air-gap area show height differences after lamination?
An air-gap board goes through pure adhesive drilling, panel layout and lamination, mechanical drilling, and plating. Each via metallization adds thickness at the via location. As a result, the substrate surface ends up with raised and recessed zones at the copper bottom corners.
Why does dry film bubble at the air-gap edge?
Dry film cannot adhere tightly at the copper bottom corner of the air-gap surface. It therefore forms bubbles after film application. The film later peels after development, and open circuits or notches appear after etching.
What was the initial SET yield before optimization?
The target product started at an initial SET yield of only 87.4%. The loss came from air-gap driven surface steps that damaged the resist layer before etching.
Which film materials were tested in this study?
The trial used dry film in 30, 40 and 50 μm thickness, together with A-9000 wet film. Both were evaluated on two-layer 1+1 flexible printed circuit boards.
Can vias be placed in the air-gap area?
Avoid vias in the unsupported air-gap span. A via in that zone has no bonded dielectric support, so resin starvation, barrel cracking, and flex fatigue risk rise. If a via is unavoidable, place it in the bonded flex or rigid zone, use teardrops and cap plating, and verify with cross-section and thermal cycling.
Does the air gap affect controlled impedance?
Yes. Air has a lower dielectric constant than adhesive or prepreg, so the air layer changes the effective Dk and raises impedance for a given trace geometry. Model the air gap in a 2D field solver, keep the reference plane continuous, and verify with an impedance coupon. Do not use a standard bonded-stack model without adjustment.
When should we choose air-gap over book-type construction?
Choose air-gap when the bend must be dynamic, low-stiffness, or tight-radius and the stack can tolerate the extra lamination steps. Choose book-type when the bend is static, impedance control is critical, or cost and yield ramp are the main constraints. Book-type is fully bonded and simpler; air-gap trades process complexity for mechanical freedom.
Summary / Key Takeaways
- Rigid-flex PCB air-gap process optimization is a surface-step control problem first, and a film selection problem second.
- Define the unbonded window with a glue-stop layer or selective prepreg cutout; keep the coverlay opening and adhesive flow path outside the gap.
- Control the press curve: slow ramp, staged pressure, vacuum lamination, and resin bleed at the window edge.
- Match dry film or wet film to the measured step; bubbles after lamination are the earliest warning.
- Use cross-section and C-SAM for process control; qualify with thermal cycling and bend fatigue per IPC-2223.
- Design rules matter: no vias in the unsupported gap, model impedance with the air layer, and choose book-type when dynamic bend is not required.
Related Posts
- Wet Film + Dry Film Process for Rigid-Flex PCB Air-Gap Yield — Related article
- Rigid-Flex PCB manufacturing Guide — Related article
- Coverlay Bonding Strength Optimization in Rigid-Flex PCB — Related article
- Flexible PCB & Rigid Flex PCB Etching Process — Related article
- Manufacturing Process of Semi-Rigid Flex PCB — Related article
Continue reading: Part 2 «
