Estimated reading time: 4 minutes

The Chemical Connection: The Consequences of Additive Metallizing on Etching Steps

This month, I’m taking another look at additive manufacturing. What does mSAP, SAP, and additive metallizing (plating) look like in today’s advanced PCB fabrication? I must confess I don’t have a lot of insight, as my whole career has been devoted to removing copper from panel surfaces in even and controlled ways, and not to putting copper back on the panel. However, at some point in the additive process, especially in additive plating, copper must be removed from between the conducting surfaces to complete the circuits. Here, I can address some of the consequences of additive metallizing on subsequent etching steps.

Traditional additive metallization, or pattern plating, has been the primary way to get taller conductor heights for higher current capacities and less heat generation than is possible with just subtractive etching. The traditional process starts with a thin laminated foil (one-quarter or one-eighth ounce), then laminates and develops a plating resist with the desired circuit layout. The panel is put in an electroplating bath, and copper is plated until the desired circuit height is reached, followed by a solder or tin coating to serve as an etch resist. The base foil is etched with an alkaline etchant, leaving behind the desired circuit design.

A more recent technique is to sputter or flash plate an even thinner base copper, usually 2 to 4 microns, and electroplate the circuit lines to the desired thickness, but with no tin or solder. Then the panel goes through a flash etch in an acid etchant (usually cupric) at such a speed that etches away the base copper while having a minimal effect on the plated circuit lines.

Obviously, before the base copper can be etched away, the plating resist must be removed, and this is where problems can arise. Getting even plating over a large format panel, with several of these large format panels on a vertical flight bar, is difficult at best. It is not uncommon for areas of the panels to be over-plated, trapping resist under the over-plated areas that interfere with the etching of the base copper.

The first thought is to increase spray pressures in both the stripper module and rinses, but it turns out that this is not as simple as it seems. One customer was using a power wash at 400 psi after stripping and requested a 400-psi rinse after the strip chamber for their proposed new stripper. This would involve special high-pressure pumps with a very high flow rate, stainless steel plumbing, very good filtration (it would have to have a recirculating sump to keep water consumption at a reasonable level), etc. When all is said and done, it is very expensive. We decided to put together a 400-psi test stand to get an idea of what would be involved in actually putting one of these things together, which turned out to be a good idea.

We got some test panels from the customer, and to my surprise, the 400-psi rinse seemed to have no effect whatsoever. Somewhat apprehensively, I stuck my hand under the spray and felt hardly any impact even though the pressure gauges were reading 400 psi. To achieve 400 psi with the pump they had available, our engineers kept reducing the nozzle openings until they reached 400 psi. Unfortunately, the opening size was so small that spray droplets were broken up into such tiny droplets that they had hardly any impact force at all, even at 400 psi.

We had invented a “velvet touch” high pressure spray system. A bigger pump was obtained, adding nozzles with a more reasonable opening size, and found that the customers’ objectives could be reached with a 150-psi rinse—a much more reasonable proposition, but still requiring more effort than expected.

Another unanticipated consequence occurred when people started flash etching to remove very thin base copper from their boards. It should have been a simple proposition, a quick spray of etchant to remove two or three microns of base copper and on to the next process. However, we started getting reports of residues of base copper left on the panel in long strips, like tiger stripes. This did not occur on their normal copper weight panels (1-ounce, half-ounce, etc.), just their flash etch panels run at high conveyor speeds, so it was not due to clogged nozzles or misaligned conveyor wheels.

Measurements showed that the distance between the stripes was about the same as the distance between the spray tubes, so we speculated that the conveyor speed needed for the flash etch might result in a harmonic being reached with the spray tube oscillation rate. Most etchers and developers use some type of spray tube oscillation to sweep etch solution back and forth across the board surface as it travels through the etch chamber.

At the end of each oscillation stroke, even though it doesn’t look like it, the tube must decelerate to a stop, then accelerate in the opposite direction. There is a brief time when the tube is stopped and, at this moment, the area directly under the spray gets a little more etch time. At the oscillation rates and conveyor speeds used for normal copper weight boards, these areas tend to even out as the panel goes through the etcher, as a different part of the panel is below the spray tube at the end of each oscillation stroke.

It is conceivable, however, that at the high conveyor speed needed for flash etching, a harmonic could be reached where the same part of the panel could be under the spray tube at the end of each oscillation stroke and that little bit of extra etching could accumulate and cause striping. This proved to be the case when altering the oscillation speed by 25% in either direction caused the tiger stripes to go away.

I doubt there is anyone who could have anticipated a harmonic relationship between conveyor speed and etcher oscillation rate when flash etching was first thought of (anyone who could have thought of it would probably not be working in the PCB business anyway). So, when taking another look at additive manufacturing, keep in mind that there will be consequences further down the process line.Don Ball is a process engineer at Chemcut.

This column previously appeared in the May 2026 issue of I-Connect007 Magazine.