This article analyzes the root causes of the galvanic effect that occurs during the chemical immersion silver (electroless silver) surface finish process on PCBs and presents practical production controls to mitigate it. The discussion explains why the galvanic effect is inherently possible in immersion silver, examines the key influencing factors via experiments, and summarizes effective control measures for manufacturing.
With RoHS restrictions limiting the use of lead, cadmium, mercury, and hexavalent chromium in PCB fabrication and assembly, manufacturers have shifted from hot air solder leveling (HASL) to lead-free finishes. Common lead-free finishes include HASL, OSP (organic solderability preservative), ENIG (electroless nickel/immersion gold), ENEPIG (electroless nickel/electroless palladium/immersion gold), immersion tin, immersion silver, and electroplated tin. Among these, immersion silver offers distinct advantages: it yields a very flat finish suitable for high-density interconnects, fine-pitch SMT, BGA, and die attach; it is relatively simple to operate, cost-effective, requires moderate maintenance, and supports high throughput with compact equipment. As a result, immersion silver is widely adopted as a PCB final finish.
Principles of Immersion Silver and the Galvanic Effect
Immersion silver is a displacement reaction that deposits metallic silver onto exposed copper surfaces. The reaction can be described as:
Ag+ + e? → Ag E = 0.799
Cu2+ + 2e? → Cu E = 0.340
Overall: 2 Ag+ + Cu → 2 Ag + Cu2+ E = 0.459
Under normal conditions, this reaction proceeds spontaneously, depositing a protective silver layer of sufficient thickness to protect the copper during storage and assembly.
However, PCB immersion silver is performed after solder mask imaging. Due to developing and exposure characteristics, the solder mask edge typically exhibits an undercut profile rather than an ideal vertical wall. This undercut creates a narrow crevice at the solder mask–copper interface. The crevice restricts solution exchange, limiting the local availability of Ag+ ions, while the displacement reaction continues, consuming copper electrons in the confined region and reducing silver on adjacent exposed pads.
Because Ag+ supply in the crevice is inadequate, the crevice acts as part of a galvanic cell. Copper under the solder mask becomes a sacrificial anode, supplying electrons to reduce Ag+ on exposed copper pads. The silver thickness along the crevice edge becomes locally higher than in fully accessible areas. This galvanic effect is fundamentally similar to crevice corrosion, driven by restricted mass transport and potential differences at the solder mask–copper interface.
Factors Driving the Galvanic Effect
The galvanic effect requires two conditions: a displacement reaction must be present (which is intrinsic to immersion silver), and there must be localized ion supply insufficiency (e.g., in crevices under the solder mask). Displacement reactions also occur in immersion gold processes, but the magnitude differs because typical ENIG immersion gold thickness is about 0.05–0.10 μm, whereas immersion silver thickness is typically 0.15–0.5 μm. The thicker silver layer and longer dwell times used in immersion silver increase susceptibility to galvanic effects under restrictive conditions.
In solder mask processes, two variables strongly influence crevice-driven behavior:
- Solder mask undercut magnitude, driven by imaging and developing conditions.
- Adhesion at the solder mask–copper interface, influenced by surface preparation and mask chemistry.
In short, the principal variables are immersion silver thickness, solder mask–copper adhesion, and solder mask undercut.
Experimental Design and DOE
A designed experiment (DOE) was conducted based on the above analysis. The selected factors included:
- Immersion silver line speed (controls silver thickness).
- Pre-solder mask surface preparation scratches or abrasion profile (affects adhesion).
- Solder mask ink type.
- Developing speed.
- Exposure level (together with developing and ink type, affects undercut).
Notes: (a) At 1.9 m/min line speed, the average silver thickness was approximately 0.30 μm; at 1.3 m/min, it was approximately 0.40 μm. (b) All test results met the specification of 20% of trace depth per the referenced standard.
Main Effects Analysis
Main effects analysis showed that immersion silver line speed (and therefore silver thickness) had the largest impact on the galvanic effect among the five tested factors. Exposure level also had a significant effect, consistent with its influence on solder mask undercut. Developing speed, solder mask ink selection, and pre-mask abrasion patterns did not exhibit statistically significant effects in this DOE.
Production Controls
Based on the analysis and experimental results, normal solder mask process parameters can meet the requirements for minimizing galvanic effects during immersion silver. In production, focus on the following controls:
- First-article silver thickness control: target 0.15–0.30 μm.
- First-article verification for the galvanic effect at solder mask–copper transitions.
- In-process sampling of silver thickness and SPC control of the immersion silver line.
- Prevent handling anomalies such as panel jams that can increase dwell time and local thickness.
Conclusions
Silver thickness is a critical parameter for immersion silver and must be tightly controlled. Since thickness in the 0.15–0.30 μm range provides adequate copper protection for storage and assembly, it is recommended to control the average immersion silver thickness to not exceed 0.30 μm whenever possible.
Different solder mask inks have different process windows. For immersion silver applications, compare undercut behavior across the relevant exposure and developing conditions for each ink and select the parameter set that yields the smallest undercut while maintaining adhesion and imaging quality.
Production Experience and Process Behavior
Over a nine-week bath life cycle for a 700 L immersion silver tank, a total board area of 12,200 m2 was processed, corresponding to a unit consumption of 17.5 m2/L. This included approximately 7,000 m2 of "thicker silver" boards (0.225–0.30 μm) and 5,200 m2 of "thinner silver" boards (0.15–0.20 μm). On 2 mm × 2 mm pads, the average silver thickness was 0.203 μm. Early- and late-life tests of thickness, appearance, adhesion, galvanic effect, post-IR reflow resistance (260 °C, two cycles), wet aging resistance (85 °C/85% RH, 12 hours), dry bake resistance (155 °C, 4 hours), and solderability met customer requirements. Routine quality tests throughout bath life, including solderability after dry bake, corrosion resistance, and ionic cleanliness, also indicated stable and reliable performance.
Immersion silver deposition is governed by the reduction rate of silver ions. Deposition rate increases with silver ion concentration, solution agitation, and temperature. Different target thicknesses can be achieved by adjusting line speed (dwell time in the immersion tank) and bath temperature. Because the process is a displacement reaction, Ag+ must be replenished and Cu2+ gradually accumulates in the bath over time; copper ion concentration is a key factor affecting bath life.
Silver addition correlates linearly with total panel area processed. On average, 1 m2 of PCB consumes about 0.53 g of silver (0.53 g/m2). Given an average silver thickness of 0.203 μm on standard pads, the effective surface area fraction receiving immersion silver is approximately 13%. Based on stoichiometry, the "theoretical" copper ion accumulation rate would be 0.156 g/m2. In practice, the measured Cu2+ accumulation deviates from the theoretical rate and stabilizes around 1.6 g/L due to drag-out of copper ions by processed boards. As copper concentration rises, the rate of Cu2+ removal via drag-out increases and eventually balances the copper generated by the displacement reaction. The final equilibrium copper concentration depends on line configuration and equipment. In general, a Cu2+ load of approximately 2 g/L is adequate for normal production.
A characteristic of immersion silver is that silver thickness tends to decrease as pad size increases. Because performance depends on thickness, acceptance criteria vary by process. IPC-4553 specifies a 1.5 mm × 1.5 mm pad as the standard test pad size for thickness measurement. When pad area increases from 1 mm2 (1 mm × 1 mm) to 25 mm2 (5 mm × 5 mm), the average silver thickness may decrease, for example from about 0.23 μm to about 0.16 μm. The difference between large and small pads on the line described here was approximately 30%. This difference is smaller than what is often observed in other immersion silver processes, implying that if small pads meet thickness targets, larger pads typically retain adequate thickness and performance.
Galvanic Effect: Mechanism and Mitigation
The galvanic effect observed on copper traces under solder mask is analogous to crevice corrosion. Under normal immersion conditions, copper acts as both anode and cathode in microscopic regions, and uniform silver deposition occurs across exposed surfaces. When a crevice exists at the solder mask–copper interface, Ag+ supply in the crevice is restricted, and copper under the mask becomes a sacrificial anode, supplying electrons to reduce Ag+ on adjacent exposed pads. The number of electrons required is proportional to the amount of reduced Ag+, so the galvanic driving force increases with exposed pad area and with target silver thickness.
Under typical production settings (for example, approximately 2.5 minutes dwell yielding about 0.25 μm silver), scanning electron microscopy reveals full silver coverage of the exposed copper pad and even coverage extending slightly under the solder mask due to undercut. After solder mask removal for inspection, many samples show no evident galvanic attack on copper traces; however, mild etching may occasionally appear behind the solder mask/copper boundary rather than directly under the undercut.
If dwell time is intentionally doubled to around 5 minutes to produce a thicker silver layer (e.g., about 0.48 μm), the galvanic effect intensifies. In such cases, grooves approximately 20 μm wide and 10 μm deep can form on copper traces under the solder mask, indicating localized galvanic attack driven by the extended displacement reaction in the presence of restricted Ag+ supply.
Practical methods to prevent or reduce the risk of galvanic effects include:
- Select an immersion silver chemistry with controlled aggressiveness (appropriate pH and corrosion profile) that does not require excessive silver thickness to meet corrosion resistance requirements.
- Control micro-etch within the specified range to balance adhesion and minimize over-roughening.
- In PCB design, avoid directly connecting large exposed copper areas to fine traces without relief features that reduce galvanic coupling during finishing.
- Improve solder mask–copper adhesion by optimizing pre-treatment, imaging, curing, and developing parameters, and by using solder mask formulations with good chemical resistance to the immersion silver line.
By treating immersion silver as a controlled displacement reaction and managing crevice formation and adhesion at the solder mask boundary, manufacturers can maintain solderability and surface protection while minimizing galvanic attack on copper traces.
