Copper Foil Profile After Etching: Width, Angle and Undercut
The copper foil that remains after etching is not a rectangular bar with vertical walls. The etchant attacks the top surface and the sides at the same time, so the cross section has a trapezoidal shape with a sloped wall and a lip of resist overhanging it. That shape, the etch profile, decides the final line width, the impedance of a controlled line and the mechanical support the trace gives to the laminate beneath it.
Design rules are usually written in terms of a finished line width, and the process that produces that width has to work in the opposite direction. The artwork is drawn wider than the target, the etchant removes a known amount sideways and downwards, and the difference between what the designer drew and what the customer measures is the sum of the etch behaviour of the line and the compensation applied to it.
What the Profile Consists Of
Three numbers describe most profiles. The first is the top width, measured where the copper meets the resist edge. The second is the bottom width, measured where the copper meets the laminate. The third is the thickness of the copper at the end of the etch. A profile is called positive when the bottom is narrower than the top, which is the normal result of an isotropic etch, and the difference between the two widths is the undercut.
The wall angle is the angle between the copper wall and the surface of the laminate, and it is a convenient way to describe the same geometry. A vertical wall is 90 degrees, and a typical spray etched profile runs between 50 and 75 degrees depending on the copper thickness and the chemistry. The angle has to be measured on a microsection, because a top-down measurement cannot see it.
Vertical and Lateral Etch Rates
Etching is a diffusion limited process at the surface of the copper. Fresh etchant reaches the metal through a boundary layer, the reaction consumes it, and the products have to diffuse away. In a spray etcher the impingement of the droplets thins that boundary layer, and the rate is higher where the spray is more effective, which is one reason uniformity across a panel is never perfect.
The lateral rate under the resist edge is slower than the vertical rate on an open area, because the etchant has to travel sideways into the gap. That is why a wide opening etches down faster than a narrow one, and why two lines of different width on the same panel end up with different profiles. The effect grows as the line narrows, and it is the reason fine line etching needs a different chemistry and a different machine setting from a power plane.
Etch Factor and What It Predicts
The etch factor is the ratio of the etch depth to the lateral undercut, and it is the single number that compares processes. A bath and machine combination that etches 35 um of copper downwards while undercutting 12 um sideways has an etch factor of about 3, and a process with an etch factor of 2 will undercut the same thickness by 17 um. Immersion etching gives the lowest factors, spray etching with well maintained nozzles gives higher ones, and the value falls as the bath loads with copper.
The factor is used in two ways. During design it converts a target finished width into an artwork width for a given copper thickness. During production it is a control figure: a drop in the etch factor at a constant bath analysis usually points to a spray problem rather than to a chemistry problem, because the lateral attack is more sensitive to how the etchant reaches the surface than to its composition.
Undercut and Line Width Tolerance
The undercut matters most where a line has a width tolerance. A nominal 100 um line with 35 um copper and an etch factor of 3 loses about 12 um per side, so the finished width is close to 76 um unless compensation has been applied. The same nominal line on 70 um copper loses about 23 um per side, and the arithmetic is what makes heavy copper fine line work difficult rather than merely slow.
The tolerance question is really a question about the distribution. If the etch factor varies by 20 percent across a panel, the finished width varies by a few micrometres for a thin line and by more for a thick one. Where a design asks for a tight width tolerance on a thick copper layer, the process engineer has to decide between tighter bath control, a different etchant, or a conversation with the designer about the impedance budget.
The Effect of Copper Thickness
Etch time scales roughly with the thickness to be removed, and the undercut accumulates over the whole of that time. Doubling the copper thickness doubles the etch time and therefore roughly doubles the undercut, and the line becomes narrower and less rectangular at the same time. This is why a 1 oz layer can hold a 100 um line comfortably while a 2 oz layer of the same nominal artwork produces a line that is visibly thinner.
Where thick copper is unavoidable, the answers are to widen the artwork, to reduce the copper that has to be removed by using a tenting or a pattern plating approach, or to accept a wider line and the change in impedance it brings. The pattern plating alternative and its own tolerance stack are compared in the notes on copper thickness verification.
Spray Etching Mechanics and Uniformity
A spray etcher is judged on how evenly it delivers etchant. Nozzles are arranged in oscillating banks, the pattern overlaps, and the pressure is set so that the droplets wet the surface without bouncing. A blocked nozzle, a worn nozzle tip or a drop in pump pressure shows up as a light or a dark band on the panel, and the band is usually visible on the etch coupon before it becomes a yield loss.
Panel loading interacts with the spray pattern. Two panels travelling close together shield each other, and a panel with a large open copper area consumes etchant locally and starves the fine lines beside it. That is the practical reason for the etch compensation tables that a fabricator builds for each product: they encode the geometry of the panel and the machine behaviour rather than only the chemistry.
Bath Chemistry and Regeneration
Cupric chloride is the common etchant for this work, and it is controlled by specific gravity and by oxidation reduction potential rather than by a single titration. The specific gravity is held in a narrow band, typically around 1.30 to 1.36, the temperature is set between 45 and 52 °C, and the oxidation reduction potential is maintained in a window that keeps the etchant in its active cupric state. Free acid is added to control the side reactions that produce sludge.
Regeneration keeps the etch rate stable as the bath dissolves copper. Without it the rate falls, the operator extends the conveyor dwell, and the undercut grows on the lines that were already at the limit. The chemistry side of the loop is described in the notes on cupric chloride bath control, and the recovery side in the notes on copper recovery.
Measuring the Profile
The profile is measured on a microsection cut perpendicular to the trace, at a known location. The section is prepared so that the copper edge is not rounded by polishing, and the three widths are read at a magnification that puts the whole wall in the field of view. Several traces of different width should be measured on the same section, because a single measurement cannot separate a bath problem from a geometry effect.
A cross section coupon placed on every panel is the production version of the same measurement. Where a coupon is not available, the etch factor can be estimated from the resistance of a test trace and its known thickness, but that method reports an average and hides the worst case. The width control discipline that uses the resulting numbers is set out in the notes on etch factor and line width control.
Process Control and Records
The daily controls are the specific gravity, the temperature, the oxidation reduction potential and the free acid of the bath, the spray pressure of each bank, the conveyor speed and the etchant top-up volume. Each is recorded per shift, and each has a range that has been tied to a measured etch factor rather than to a supplier recommendation alone.
The records that make a difference over a year are the trend of the etch factor against the bath age and the nozzle service history. A profile problem that appears after a nozzle change is diagnosed in minutes with those two charts, and the same problem without them is usually answered by widening the artwork, which costs impedance margin on every board of the next revision. The fine line side of the same process is covered in the notes on fine line etching.
FAQ
Is a steeper wall always better? A steeper wall holds a tighter line width, which is better for impedance. It also concentrates stress at the copper and laminate interface, so the benefit is not unlimited.
Why does undercut grow during a shift? The bath loads with copper, the etch rate falls, and operators extend the dwell to compensate. The extra dwell attacks the side walls for longer, so the undercut grows even though the top surface looks correct.
Can the profile be measured without a microsection? An estimate can be made from a resistance measurement on a test trace. The wall angle and the worst case undercut require a section.
