Every selective soldering pot slowly eats copper — and the copper it eats changes everything about how the process behaves. Each board that crosses the nozzle wave leaves a little of its pad copper, lead copper, and plated through-hole copper behind in the melt. Over weeks and months, that dissolved copper shifts the alloy away from the specification it shipped with, and the effects show up where you least want them: slower wetting, stiffer fillets, more bridging, and a bath that no longer behaves like the datasheet. Understanding copper dissolution is therefore not metallurgy for its own sake — it is the key to keeping a selective soldering line stable. For the equipment and process context this article builds on, start with the I.C.T selective soldering machine overview, then come back to the details below.

This article explains what copper dissolution actually is, where the copper comes from, how it changes both the bath and the joints, how to detect the drift early, how to slow it down, and what to do when the level is already too high. The theme runs through the whole bath-management series: measure rather than guess, and act on trends rather than symptoms.

What Copper Dissolution Actually Is

Dissolution is not the same as alloying

It helps to separate two things that sound similar. Alloying is deliberate: the metallurgist who designed your solder chose its copper content on purpose, because a controlled amount of copper shapes the alloy's melting behavior, joint strength, and copper-scavenging behavior during soldering. Dissolution is uncontrolled: it is copper that enters the melt from the boards themselves, on top of what the formulation calls for. A SAC-type alloy arrives with a defined copper percentage; after months of selective soldering, the melt usually contains more copper than that, plus whatever trace elements arrived with the boards. The bath is still "an alloy" — just not the one you bought.

Why selective soldering accelerates dissolution

Three features of selective soldering make it an efficient copper-dissolution machine. First, the nozzle wave is small and energetic: a focused, pumped stream of molten alloy at high local velocity, which strips copper from wetted surfaces faster than a broad, gently rolling wave. Second, contact is localized but repeated: dense joint groups see long effective dwell times, and the same copper area is exposed to fresh, hot alloy throughout the joint cycle. Third, the pot is small: every milligram of dissolved copper is diluted by a few kilograms of melt rather than tens of kilograms. Put together, a selective pot running copper-heavy boards can move measurably out of specification in a fraction of the time a large wave pot would need.

The mechanism in one paragraph

Molten tin-rich solder is a solvent for copper. Wherever molten alloy touches solid copper, copper atoms dissolve into the melt, and the rate depends on temperature, contact time, and how quickly fresh unsaturated alloy is swept across the surface — which is exactly what a pump does. The dissolved copper does not stay neatly dissolved forever: some of it forms intermetallic compounds, and some of it circulates back out into joints. The net direction, for a working pot, is steadily upward copper content, until something — an analysis, a defect wave, or a maintenance decision — intervenes.

Where the Copper Comes From

Pads, leads, and through-hole barrels

The copper that ends up in your pot comes from three main places on the board. Exposed copper pad surfaces — on OSP-finished boards in particular — touch the wave directly around every selective joint. Component leads and terminations carry copper or copper alloys into the joint as well. And plated through-hole barrels contribute copper along their full length wherever the wave wets the barrel during hole fill. Which of the three dominates depends on the board family: a board with large thermal pads and thick barrels delivers more copper per joint than a light single-row connector board.

Board mix matters more than board count

Two lines running the same number of boards can drift at completely different rates, because copper delivery scales with copper area exposed and with thermal mass — not with board count. A line that takes on thick power boards with heavy copper planes will see its bath chemistry move much faster than a line running small control boards. This is why a change in the production mix belongs on the list of events that should trigger an off-cycle alloy analysis, as covered in the alloy analysis article in this series. The board mix is the input; the bath chemistry is the state; and the two are connected through exactly this dissolution mechanism.

Rework and repair as a hidden source

Boards are not the only copper that enters the melt. Rework and repair soldering — hand touch-up with the same alloy, local re-soldering of corrected joints, and any machine-assisted correction work that draws from the same material stream — contributes copper too, because the copper it touches comes from the same pads and leads. Factories that feed touch-up irons and pot-fed tools from the same alloy inventory sometimes see faster drift than their board mix alone predicts. The practical response is not to invent new chemistry for rework; it is to make rework consumption part of the bath's material history, so that when the analysis comes back higher than expected, the explanation does not have to be guessed. A short note in the machine file each time rework volume is unusual is enough to keep the picture honest.

How Copper Changes the Bath

Melting and flow behavior shift first

Copper beyond the alloy's design range changes how the melt behaves at temperature. Excess copper raises the fraction of intermetallic compounds circulating in the bath and stiffens the alloy's behavior in the joint area: fillets form pastier, spread slows, and the wave looks and flows differently at the nozzle. None of this announces itself with an alarm; it appears gradually, which is why lines that do not test their alloy often blame the flux, the nozzle, or the recipe for symptoms that are actually chemistry.

Dross behavior changes too

A copper-loaded bath tends to manage its surface worse: operators commonly report more dross, more surface skinning, and more effort to keep the wave clean and bright. Dross also carries tin away, so a bath fighting both copper pickup and tin depletion is fighting on two fronts at once. Good nitrogen coverage and stable wave conditions slow both problems, but they do not reverse the copper content — only dilution or a changeover does that.

Intermetallics accumulate in the melt

Some of the dissolved copper combines with tin into intermetallic compounds inside the pot. Over a long bath life, these compounds add up: they circulate in the melt, they can end up in joints, and they are one of the reasons an aged pot behaves differently from a fresh one even when the major-element numbers look acceptable. Intermetallic accumulation is also the reason pot age belongs in your analysis records: a ten-month bath and a two-month bath with the same copper reading are not equally trustworthy.

What the wave looks like at the nozzle

There is a visible layer to all of this. A healthy melt at the nozzle looks bright and mirror-like, and the wave rises cleanly and steadily. A bath carrying excess copper and intermetallic load often looks duller, and the wave can appear less crisp at the edges — subtle cues that experienced operators register before any instrument does. Treat these observations as prompts for measurement rather than as evidence by themselves: nozzle wear, pump performance, and nitrogen coverage produce similar visual changes, and only an analysis can separate chemistry from hardware. The nozzle hardware that faces the melt all day — racks, ceramic nozzles, and their wetted surfaces — is also worth inspecting during the same pot work, because a nozzle that has been running in a drifted bath behaves differently from one running in fresh alloy.

How Copper Changes the Joints

Wetting and spread degrade gradually

The first joint-level symptom of copper drift is usually wetting: solder that used to spread cleanly now climbs more slowly, wets more reluctantly, and leaves fillets that look stiff and rounded rather than smooth and feathered. On the same recipe, the same flux, and the same boards, that change is a strong hint that the melt has moved. It is not proof — flux aging and tip oxidation produce lookalike symptoms — but it is one more reason the troubleshooting sequence should include an alloy analysis rather than ending at parameter changes.

Bridging and icicle tendency rise

Excess copper is widely associated with more bridging and icicle formation in selective soldering, because the altered flow behavior holds the wave differently at the nozzle and at fine-pitch joint groups. A process that used to clear a dense connector field cleanly starts leaving bridges between the last pins, and the operator compensates with pump speed or dwell — which treats the symptom and may accelerate the underlying drift. When bridging appears on a board family that used to run clean, and nozzle condition and nitrogen check out, chemistry is the next suspect.

Microstructure and long-term reliability

Joints made from a drifted alloy can look duller and grainier than that line's historical baseline, and their internal microstructure — the distribution of intermetallics in the joint — changes with it. Whether that matters for reliability depends on the application, the standard the product is built to, and the customer's acceptance criteria; the defensible position is simply that joints should be made from the specified alloy, and that any deviation is a process change that needs evidence, not an assumption that it is harmless. That is the reasoning behind conservative action limits from suppliers, and behind the habit of confirming limits against the current supplier datasheet and relevant industry specifications such as IPC J-STD-006 published by IPC.

Detecting Copper Drift Early

What operators notice first

Operators are a sensitive early-warning instrument, but only if their observations are recorded and compared against the line's own history. The typical progression: the wave looks slightly less bright; dross seems heavier; joints on the usual boards need a touch more dwell; a dense connector field that always ran clean starts bridging intermittently. None of these observations alone means anything — but a cluster of them, on an unchanged recipe, is the classic signature of a bath that has drifted. The correct response is an analysis, not a parameter hunt.

Coupons and first-article trends

The cheapest structured detector is a repeatable coupon or first-article routine: a representative board or test panel, wetted and inspected the same way every time, with results recorded. When coupon wetting, hole fill, and bridging results slide while everything else is constant, the bath is the prime suspect. This instrument also protects you in the other direction: if the coupons stay stable, you can rule chemistry out and look at flux, nozzles, or preheat with confidence.

Scheduled analysis confirms what the symptoms suggest

The alloy analysis report closes the loop. It shows the copper content against the alloy's specification range, alongside the other elements and contaminants, and it turns a suspicion into a number with a trend. For a selective soldering line, the combination of scheduled analysis plus a coupon routine plus operator observations recorded per shift is the practical detection system — each covers a gap the others leave.

Separating chemistry symptoms from hardware symptoms

The hardest part of early detection is that bath drift and hardware wear produce overlapping symptoms. A worn nozzle, a tired pump, a failing nitrogen heater, and a copper-loaded melt can all present as "the wave does not behave like it used to." The discipline that separates them is subtraction: inspect and clean the nozzle, verify pump performance against its baseline, confirm nitrogen coverage and purity as the machine specifies — and then, if the hardware checks out and the symptoms persist, the chemistry is the remaining suspect and the analysis confirms or clears it. Running that sequence in the same order every time turns a vague complaint into a short diagnostic. It also prevents the expensive mistake of correcting chemistry that was never the problem, or replacing hardware that was never worn.

Reading the trend, not the single point

One analysis report is a photograph; the trend is a film. A copper number inside specification but climbing steadily for three consecutive reports tells a different story than the same number stable for a year, and the response should differ accordingly: the first calls for a shortened interval and a look at what changed in the production mix, the second for simply keeping the schedule. This is why analysis records belong in one continuous file per pot, with results plotted in production order. When the trend is visible, decisions about correction stop being debates about single data points and become straightforward readings of direction and rate — which is exactly what a quality system needs when it asks why the bath was managed the way it was.

Managing the Rate of Dissolution

Temperature discipline is the biggest lever

Dissolution rate rises steeply with temperature, so pot setpoint discipline is the single most effective way to slow copper pickup. That means running the temperature the process actually needs — verified against the supplier's guidance and the board family's thermal demand — rather than the temperature that quietly crept upward while someone chased hole fill on one difficult board. Every degree above what the joint requires is paid for in faster copper dissolution and faster dross growth. Temperature changes belong in the change-control process, with a coupon check and a record, not in the operator's day-to-day toolkit.

Contact time and dwell matter more than they look

The wave dissolves copper for as long as it touches it, so dwell time per joint, pump speed, and nozzle path planning all feed the rate. A process that lingers on dense joint groups to guarantee hole fill pays for that margin in copper; a process with an over-specified dwell pays the same price everywhere. Reviewing joint-cycle times when a new board family is introduced — and after any bridging-driven parameter inflation — keeps this lever honest. Machines with programmable pot and nozzle control, such as the I.C.T SS550 inline selective wave soldering machine, make it practical to hold dwell where the process needs it instead of where habit left it.

Alloy choice sets the working envelope

Some solder alloys are formulated specifically to manage copper pickup — tin-copper-nickel systems, for example, are designed around a saturated copper balance that slows further copper absorption from the boards. Choosing such an alloy does not repeal the physics, but it can widen the working envelope between analysis intervals. The choice of alloy is a process decision with bath-consequence consequences either way; match it to the board mix, confirm the specification against the supplier's current datasheet, and record it as part of the machine file. Our solder bar selection article walks through that choice in detail, and the incoming inspection routine in our solder bar inspection article makes sure the alloy that arrives matches the one that was ordered.

Flux and cleanliness play a supporting role

Flux activity and board cleanliness shape how efficiently heat and solder reach the copper, which makes them quiet contributors to the dissolution rate. An under-fluxed joint resists wetting, invites dwell inflation, and — indirectly — feeds copper into the pot for longer. A clean, properly fluxed process transfers heat efficiently and gets out of the melt quickly. Flux quantity changes therefore belong in the same recorded, verified loop as temperature changes.

Nitrogen coverage and surface condition

Nitrogen does not change how fast copper dissolves into the melt directly, but it changes two things that matter to the same discipline: dross formation and wave stability. A pot running with proper coverage generates less oxide skin, loses less tin to the surface, and keeps the wave brighter and more consistent — all of which reduce the pressure to compensate with temperature and dwell, the very parameters that drive copper pickup. Checking nitrogen coverage and purity as part of the same maintenance window as sampling keeps the whole bath-management loop in one place, and it means that when the analysis does come back high, the team already knows the hardware side was verified. Air leaks around the pot, damaged coverage ducts, and empty supply bottles are mundane causes of symptoms that look exactly like a chemistry problem, and they are cheap to rule out first.

A word on productivity trade-offs

Some of the measures that slow dissolution look, on a production floor, like they cost time: lower setpoints, shorter dwell, disciplined parameter change control. The trade-off resolves when you count the other side of the ledger. A pot that drifts slowly needs fewer corrections, fewer emergency changeovers, and fewer defect-driven troubleshooting sessions; a pot pushed hot and long to gain seconds per board pays those seconds back as analysis surprises and maintenance downtime. The right comparison is not "minutes saved today against nothing" but "minutes saved today against the scheduled and unscheduled work the drift will demand later." Framed that way, temperature and dwell discipline usually pays for itself, and the lines that hold it tend to be the lines that rarely fight their bath.

Correcting a Copper-Loaded Pot

Topping up cannot remove copper

The most common mistake in copper management is reaching for a fresh solder bar when the analysis comes back high in copper. Adding alloy dilutes the copper somewhat, but the added tin mostly replaces tin that was lost to dross — the copper stays, and with a small selective pot the dilution effect of a practical top-up is modest. Top-up is the right correction for tin depletion when copper is in range; it is not a correction for copper above the limit. Knowing which problem you have is exactly why the analysis comes first.

Partial replacement as the working correction

When copper is past the supplier's action level but the bath is otherwise healthy, partial replacement is usually the proportionate response: drain a defined portion of the melt, refill with fresh alloy of the same specification, verify with a new analysis, and record the whole sequence. The size of the drain, and whether one round is enough, should come from the supplier's guidance for their alloy — they can estimate the dilution arithmetic for their own product. Partial replacement fits into a planned maintenance window, which keeps the line's availability intact.

Full changeover and requalification for the deep end

A pot that is far past its limits, has accumulated years of unmanaged drift, or has an unknown history is a candidate for full changeover: drain, clean, refill, and requalify. The requalification is the part that saves time in the long run — temperature verification, flux and wetting confirmation on coupons, and a first-article pass on a known board before production release. Our lead-free alloy changeover article covers that sequence step by step.

Building the Prevention Into the Line

Set the baseline when the pot is young

The cheapest moment to manage copper is the first month of a pot's life. Test early after commissioning, again after the first significant production block, and use those two points to estimate the drift rate for your actual board mix. From there, set the analysis interval with supplier guidance and your own data — not with a number borrowed from another line. Machines designed for straightforward pot access, such as the I.C.T SS430 offline selective soldering machine and the economical selective wave soldering machine range, make early-life sampling a small task instead of a production interruption.

Define the trigger events in writing

Write down the events that trigger an off-cycle analysis, so the decision is never improvised: alloy or supplier change, board-mix change, abnormal dross, unexplained wetting or bridging changes, temperature excursions, and any corrective refill. A written trigger list converts bath management from tribal memory into a process, and it is the first thing an auditor or customer will ask to see.

Keep records and one owner

Copper management is a records discipline. Every analysis, every temperature or dwell change, every top-up and refill goes into the machine file with dates and results, and one named process owner holds the file and the decision authority. The trend chart that comes out of those records is worth more than any single report: it shows the drift rate, it shows whether interventions worked, and it tells you — with evidence — when the next correction is due.

Frequently Asked Questions