Fine Line Etching: Undercut and Etch Factor Control

Introduction

As PCB designs push toward high-density interconnect architectures, fine-line etching with line and space widths below 3 mil has become a common requirement. In these geometries, the ratio of etched depth to lateral etch—commonly known as the etch factor—plays a decisive role in final line width accuracy. Any deviation caused by side etch not only changes the physical profile but also directly impacts electrical performance, particularly impedance and current carrying capacity.

For a process engineer, controlling side etch is a balancing act. Aggressive etching improves throughput but may produce trapezoidal cross-sections with overly narrow top widths. Conversely, conservative etching can leave residual copper or poor line definition. This study focuses on how side etch affects fine-line performance and provides a practical process window to achieve an etch factor of 3.0 or higher, based on real production observations.

Impact of Undercut on Impedance and Current Capacity

In controlled impedance circuits, the characteristic impedance is largely determined by the cross-sectional area and geometry of the trace. When side etch reduces the top width of a 3 mil line to 2.2 mil while the bottom remains wider, the effective signal conductor becomes non-uniform. This leads to impedance variations along the trace, causing reflections and signal integrity issues in high-speed applications.

Current carrying capacity is also compromised. A narrowed top section reduces the total copper cross-sectional area, increasing resistance and generating more heat for a given current flow. For example, a 3 mil line with 10% side etch loses roughly 10% of its cross-sectional area, but the impact is more severe when side etch reaches 20%, as the effective current path becomes concentrated at the bottom. In power distribution circuits, this can lead to hotspots and premature failure.

Process Parameters for Etch Factor > 3.0

To achieve an etch factor above 3.0, each process variable must be optimized together. Etchant concentration is the first lever. In cupric chloride systems, maintaining free acid concentration between 2.0 and 2.5 N and cupric ion content around 150–180 g/L provides a stable etch rate with reduced isotropic behavior. Lower acid levels tend to increase side etch, while too high a concentration can cause rough edges and undercut.

Spray pressure and conveyor speed directly affect the chemical and mechanical balance. A spray pressure of 2.0–2.5 bar with a conveyor speed adjusted to achieve a dwell time of 20–25 seconds in the etch chamber yields a uniform etch front. If the conveyor runs too fast, incomplete etching forces a second pass, which increases side etch. If it runs too slow, over-etching occurs. In our trials, a combination of 2.2 bar spray pressure and 3.5 m/min conveyor speed on a standard horizontal etcher produced an etch factor of 3.4.

Role of Stripping Process

Many engineers overlook the stripping step, but it can be a hidden source of side etch. If the dry-film resist is not fully cured or the stripping chemistry attacks the resist edges, the exposed copper boundary becomes uneven before etching even begins. After lamination, a final UV cure of 1.5–2.0 J/cm² is recommended to harden the resist sidewalls.

Stripping should use a low-alkaline solution at 45–50°C with a conveyor speed that avoids over-stripping. The key is to remove the resist cleanly without attacking the copper oxide or the under-resist layer. In several case studies, switching from a sodium hydroxide-based stripper to a formulated organic stripper reduced edge roughness by 25% and improved the effective etch factor by 0.4.

Real-World Validation: A High-Layer-Count Backplane

A practical example comes from a high-layer-count backplane with 2.8 mil line and space features. The original process produced an etch factor of 2.4, with top line width measuring only 2.1 mil. This caused impedance variations of ±8% across a 100-ohm differential pair. After adjusting the etchant concentration to 2.3 N free acid, increasing spray pressure to 2.3 bar, and reducing conveyor speed from 4.2 to 3.2 m/min, the etch factor rose to 3.2.

Additionally, the stripping process was modified from a high-speed sodium hydroxide bath to a controlled organic stripper at 48°C. This stabilized the line width distribution, reducing the standard deviation from 0.3 mil to 0.1 mil. The final impedance tolerance improved to ±3%, and the current carrying capacity increased by 11% due to the more rectangular cross-section. This confirms that careful parameter control delivers both electrical and thermal benefits.

Conclusion

Side etch remains the dominant challenge in fine-line PCB etching, but it is fully manageable through deliberate process design. Maintaining an etch factor above 3.0 requires tight control of etchant chemistry, spray pressure, conveyor speed, and stripping parameters. For line widths below 3 mil, each variable must be tuned in relation to the others rather than optimized in isolation.

In practice, a process window of 2.0–2.5 N free acid, 2.0–2.5 bar spray pressure, 3.0–3.6 m/min conveyor speed, and a low-alkaline stripping process at 45–50°C consistently yields etch factors above 3.0. Real-world validation on production backplanes has shown impedance variations below ±5% and improved current handling. For process engineers, adopting this parametric framework provides a clear route to reliable fine-line manufacturing.