Reading time ( words)
The main task of the final finish is to protect the copper pad from tarnishing or oxidation while simultaneously keeping the surface active for the assembly. Electroless nickel/immersion gold (ENIG) is a widely accepted finish in the market that provides a good solderability and capability for Al-wire bonding. A main function of the gold layer is to prevent the oxidation of the nickel layer.
Typically, the deposit thickness follows the recommendations as specified in the IPC 4552, which suggests an acceptable minimum gold thickness as xmean - 3σ(sigma) ≥ 0.04 µm1. As a potential answer to the increasing demand for cost reduction in the printed circuit board (PCB) industry, lowering the gold target thickness may be an option to help reduce the precious metal cost of the finish.
The gold plating electrolytes applied for ENIG finishes can be categorized into three different types: fully immersion type gold electrolytes and mixed reaction gold electrolytes. For the immersion gold electrolytes, the deposition is fully driven by the immersion reaction dissolving nickel to provide the electrons for the gold plating. To reduce the immersive attack to the nickel, electrolytes were developed in the recent years that exhibit more autocatalytic properties. As the reaction mechanism is typically based on a combination of immersion reaction and autocatalytic reaction, these electrolytes can be described as mixed reaction gold type. Depending on the type of additive or reducing agent, the ratio of immersion/autocatalytic reaction varies. More autocatalytic properties provide the benefits of a linear thickness increase over time and lower thickness distribution compared to immersion type baths.
In particular, the thickness distribution of the gold can vary with the type of electrolyte, and for gold electrolytes with high immersion, it can strongly depend on the plating temperature. A decrease of the temperature by just 15°C can lead to a reduction of the deposit thickness of 20% and more. By that a poor temperature distribution in the plating tank can easily contribute to a poor gold thickness distribution on the panel. At the same time, a low gold thickness bears a high risk for porous layer formation.
Electrochemical measurements where the ENIG finish was exposed to corrosive electrolyte show that with low gold thickness, the measured current varies over a wide range, while from 0.05 nm and higher, the gold layer appears to be dense enough to ensure a good protection against the corrosive attack of the acidic electrolyte. The graph in Figure 1 illustrates the risk of higher gold layer porosity at low gold layer target thickness.
To simulate the effect of a porous gold layer on the ENIG-properties after aging, XPS surface analysis was performed on heat treated samples. The samples, with varying gold thickness layers of 40, 70, and 90 nm, were tempered at 150°C for 30 minutes and 120 minutes and then compared to a reference. The XPS results show, that with increasing curing times, a higher content of nickel and oxygen can be detected. This effect strongly depends on the gold layer thickness; with a low gold thickness, already even without the curing, nickel oxides can be detected on the ENIG surface. With additional thermal exposure, these values increase significantly. The difference is most obvious between 40 and 70 nm. At 90 nm, the migration of nickel is inhibited, so that even after 120 minutes of curing time, the nickel oxides on the surface can be kept at a low level.
These results indicate that with decreasing gold thickness, the gold layer is more porous and thereby becomes permeable for the nickel migration.
As lower gold thickness and higher gold layer porosity lead to an increased risk of nickel oxide formation on the ENIG surface, this is also likely to affect the solder wetting of the final finish. Solder wetting tests, such as solder spread test, confirm this observation. In this test, a solder depot is printed on the ENIG pad and reflowed. The panels were pre-aged by humid aging and 2x reflow aging before the soldering process to simulate the conditions in the assembly process.
Figure 3: Solder spreading for gold layer thickness of 40 (sample 1), 70 (sample 2), and 90 (sample 3) nm.
Even though the ENIG layers with a gold thickness of 40 nm can fulfill the acceptance criteria for the solder spread test of wetting angle below 25°C, an improved wettability can be observed clearly with higher gold thickness. The migration of nickel to the ENIG surface, which is connected to the lower gold thickness and higher porosity, inhibits the formation of the Cu/Sn intermetallic and thereby bears the risk for soldering defects. This is reflected in the reduced wetting performance at the lower end of the low gold thickness values.
Because of this risk of soldering defects for ENIG layers, the gold layer thickness needs to be kept in the specified ranges. To prevent strong thickness variations on the panel, the gold plating step needs to be well controlled, and temperature distribution in the tank well maintained.
As an additional measure to prevent soldering defects by low gold thickness, mixed reaction gold electrolytes with high autocatalytic properties like “Aurotech G-Bond”-series can also be considered. The autocatalyic properties that they exhibit can help to reduce the thickness distribution, and by that, the risk of accidental low thickness plating.
- “IPC4552B: Specification for Electroless Nickel/Immersion gold (ENIG) plating for printed boards,” by IPC, April 2021.
Britta Schafsteller is Global Product Manager of Selective Finishing at Atotech Deutschland GmbH& Co KG. Mario Rosin is Global Application Manager at Atotech Deutschland GmbH& Co KG. Gustavo Ramos is Global Product Director of Selective Finishing at Atotech Deutschland GmbH& Co KG. Joe McGurran is Product Manager at Atotech USA, LLC.