What's best for your board: immersion silver, immersion tin or maybe even ENEPIG? Learn the pros and cons of the newest alternative finishes.
Download Table 1 [PDF format].
One word describes the current state of the printed circuit industry – change. The North American and European shops that survived the three-year downturn and rush of business to Asia are now faced with a constant stream of new laws. Of course this includes RoHS and WEEE, and the upcoming EU regulations Registration, Evaluation and Authorization of Chemicals (REACH) and Energy-using Products (EuP). And don't forget all the technical changes also en route, such as via fill coppers and DC acid coppers that can give nearly 1:1 hole-to-surface coverage. Dramatic changes have taken place in drilling (ever smaller holes); etching (much more sophisticated texturing of copper surfaces for enhanced solder mask adhesion); and thickness of the multilayers required.
The choice of a "final finish" used to be simple too. You could choose either hot air solder leveling (HASL) or electrolytic nickel/gold. But with the rapid increase in technological improvements, it became necessary to develop more sophisticated surface finishes. The main driver was the need for pads with better planarity, as well as finished pads for wire bonding and press fit connectors. The surface finish is also called upon to serve as a contacting surface. Now legal influences (specifically RoHS) are driving shops to replace HASL with an alternative as quickly as possible.
As the boards continue to shrink in size, surface geography becomes more valuable, making bussing for electrolytic processes virtually impossible. This drove the industry to utilize the newer electroless processes being developed. There is also the perceived environmental benefit to eliminate the lead in the HASL process. Ridding the planet's electronics of leaded solders has become the Holy Grail, despite our own Environmental Protection Agency's studies showing that lead doesn't readily leach when in landfills.
"Final finish" implies there is one choice available, but this has not been the case. Even recently the HASL process was still used on over 70% of boards in North America, but it isn't being allowed on products going to Europe and other markets. The global industry has tried various organic solderability preservatives (OSPs); immersion tin; immersion silver; electroless nickel/immersion gold (ENIG); ENIG with an optional layer of thicker electroless gold (ENEG); electroless nickel/electroless palladium/immersion gold (ENEPIG); and a new gold direct on copper. Each has benefits and potential weaknesses.
OSPs come in thin and thicker versions, and now in high-temperature versions for lead-free solders. These are relatively inexpensive and are easily applied but limited in the number of heat cycles that can be obtained in subsequent assembly. Some OSPs require a nitrogen atmosphere at assembly. They are also not suitable for wire bonding or as a contacting surface. Concerns about solder voids with high-temperature OSPs are being evaluated at this time.
Immersion tins have had some early success as a solderable surface but may have a limited future. Their main weaknesses are their use of the carcinogenic ingredient thiourea, and the evidence of occasional whiskering as well as intermetallic formation. Whiskers are especially a concern with fine lines and spaces where they could be knocked out of a hole during part insertion, thus increasing the possibility of a subsequent electrical short. A copper/tin intermetallic can form during deposition and continue to grow, limiting the useful shelf life of the stored parts. In this age of cost awareness, another factor that has limited this product's growth is cost; it is almost as expensive as electroless nickel/immersion gold but without the additional benefits.
Immersion silver seems to have a bright future. It is easy to apply to boards, relatively inexpensive and usually performs well. Like the OSPs, thin (2-5 µin) and thicker (8-12 µin) deposit versions have been sold but the preference seems to be towards the thicker products. To prevent tarnishing, the processes have included anti-tarnish either as an ingredient within the silver bath or applied in a subsequent step. Current testing is looking for methods to provide complete coverage on the walls of through-holes and into blind vias. Other process concerns are the possible inclusion of voids in the solder joint, and the desire for better thickness uniformity per part. Some manufacturers have complained about issues with corrosion of the copper surface near holes. If severe enough, this could lead to shorts – failure of the board. Another complaint was recently named the "red plague." This refers to a discolored surface from copper contamination on the silver surface.
ENIG has been growing steadily in use. It is the most expensive of the final finishes but offers the most benefits. This process also requires the most steps. Parts must be clean and have a smooth copper surface on which to build. The electroless nickel is an autocatalytic process that deposits nickel on the palladium-catalysed copper surface. The process requires continuous replenishment of the nickel ion and the reducing agent. Good process control (constituent concentration, temperature and pH) is the key to a consistent reproducible deposit. It is very important that the nickel be able to plate a surface with consistent phosphorus levels. Most prefer a middle range of 6-8% P – too low would easily corrode, but too high makes subsequent soldering of parts more difficult.
Immersion golds are replacement chemistries. This means that they attach themselves to the nickel by replacing atoms of nickel with atoms of gold. The purpose of the immersion gold layer is to protect the nickel surface until it is soldered. The recommended gold thickness is 2-4 µin. As the purpose of the gold layer is to maintain the solderability of the nickel surface, it is necessary that it be thin (2-4 µin are preferred) and pore-free.
Immersion gold has had to deal with various issues as it has grown in use. Early problems included:
- Background plating (speckled areas on the solder mask not intended to be plated) and bridging or "Ni foot" (extraneous plating between lines thus causing shorts).
- Incompatibility with solder masks.
- Skip plating (where some pads aren't plated with ENIG, though sometimes gold does adhere to the copper) often due to surface contamination on the copper or a static charge.
- Black nickel, such that the nickel surface has become passivated or corroded, even if mostly still covered by the immersion gold layer. Black nickel is a serious issue because it is hard to detect. It results from the combination of a compromised nickel deposit coupled with prolonged dwell in an aggressive immersion gold bath. The compromised nickel deposit may be due to tin or solder mask residues on the incoming copper surface, a poorly maintained or a poorly designed Ni bath. A well maintained Ni bath holds a consistent temperature and pH as well as maintaining the different chemical components in their proper concentration ranges. A Ni bath should never be run beyond its useful recommended life in metal turnovers (MTOs). A phosphorous content below 4% by weight in the nickel will corrode during gold deposition and cause black nickel; over 11-12% makes subsequent soldering difficult or impossible. Even a good nickel can be ruined by excessive gold thickness, thus the limited thickness recommended by the IPC specs. When we have heard reports of black nickel, it has always turned out to be the scapegoat for poor soldering practice, or poor plating practice (including a recent case where a large OEM complained about some black Ni on pads that had a black material that prevented the ENIG from ever being able to plate on the copper surface. In short, black may not always be bad nickel.)
The most successful ENIG chemistries running today have had several design strengths: low palladium/no chloride activators for solder mask compatibility and avoiding background and skip plating; a mid-phos nickel that runs at lower temperatures, requires infrequent dummy plating, holds its ability to perform even after repeated heat-ups, and gives level plating around the shoulders of pads and lines; and an immersion gold that has gentle chemistry and is self-limiting in thickness.
To enhance the performance of the nickels, sophisticated controllers are being used to make additions and provide history to the quality department. To maintain a steady state in the nickel chemistry, the better controllers bleed in the ingredients in measured amounts vs. major ads made periodically.
With the newer combination immersion golds, the first microinch is the traditional replacement/immersion type. The next microinch or two are attached through an autocatalytic procedure that allows gold to attach itself to gold in a non-replacing build-up. This gives rise to a non-corrosive bath that produces a pore-free gold surface.
Direct gold over copper (also called DIG – direct immersion gold) was developed specifically for parts where nickel could create RF interference. The planarity is exceptional, again, dependent on the quality of the copper surface. To avoid the inherent problems of copper migration through the thin gold surface, it is necessary for these parts to go to final assembly within four months. Even that period is only possible with a combined replacement/autocatalytic process. Normal immersion gold will have trouble properly attaching to the copper surface and providing the necessary pore-free layer.
One last option has been developed for more sophisticated boards intended for gold wire bonding. A neutral pH autocatalytic electroless gold is in use that allows thicker gold deposits on top of ENIG without harming the solder mask (ENEG). The more traditional, high-pH electroless gold is great for non-solder masked parts.
Electroless palladiums have also been created but have not grown in use due to the higher, more volatile cost of palladium metal. ENEPIG is a very versatile surface finish with gold wire bonding capabilities. Recent studies also show that ENEPIG may prove to be an ideal soldering surface for SAC type alloys.
All fields in the PCB industry will continue to face constant change. The successful companies will be open to change while making wise choices along the way. Not all chemistries are created equal, and sophistication demands wiser choices in buying the process designed for the specific product's end use. PCD&M
Donald Walsh is national sales manager at Uyemura International (UIC); This email address is being protected from spambots. You need JavaScript enabled to view it.. George Milad is national accounts manager for technology; This email address is being protected from spambots. You need JavaScript enabled to view it.. Donald Gudeczauskas is technical director at UIC; This email address is being protected from spambots. You need JavaScript enabled to view it..
