When to apply ENIG.

ENIG (Electroless Nickel Immersion Gold) is applied to the bare copper of a printed circuit board during the fabrication process. When in the fabrication process should be a concern to the contract manufacturer or OEM responsible for assembling the components onto the bare board. If a fabrication drawing states that the printed circuit board shall have SMOBC (Solder Mask Over Bare Copper) then ENIG must be applied after the application of solder mask. There are some exceptions to the rule.

  1. The fabrication drawing or OEM fabrication specification clearly states that ENIG shall be applied to all outer layer copper surfaces prior to the application of solder mask. I have had one customer specify this as a requirement. The desire was to have an extra layer of plated nickel covering all copper surfaces under the solder mask. The assembled boards operated in a corrosive environment. Over time the environment attacked the copper through the solder mask and conformal coat. This  customer felt that it was worth the extra cost to plate everything for the extra layer of metal.
  2. The manufacturer of the bare board doesn’t know what they are doing. This is sometimes the case with printed circuit board manufacturers that do not process ENIG on site. In my previous post I list several manufacturing pit falls that impact the ENIG process. These companies may not appreciate the level of surface perfection required for ENIG since they don’t run the process on site. They may manufacturer a good board with HASL (Hot Air Solder Leveling) but have issues with ENIG. HASL is aggresive enough to burn away or remove many forms of contamination that may be on the copper to be plated. Hence their standard process may result in mystery non-conformances that come and go at the outside service or on the assembly line. In order to avoid problems with an outside service that applies the ENIG for them. Bare board manufacturers that process ENIG on site understand that the preceding process steps have a huge impact on the ENIG process. If they haven’t figure that out then you may want to change vendors.

The drawback to applying ENIG prior to the application of solder mask is a lower bond strength of the mask to the metal surface. The mask bond strength of a SMOBC board is greater than mask over gold. Prior to the application of solder mask the boards are cleaned and scrubbed. The cleaning component removes copper oxidation. This is typically done with a low concentration of sulfuric acid. At this point the copper is very smooth with next to no topograph. Solder mask shall not adhere very well unless the copper is roughened up. This is typically done with a pumice scrubber immediately after cleaning. This is where the problems occur with ENIG applied prior to solder mask. The immersion gold is only 2 to 5 micro-inches thick. The scrubbing operation needs to be minimized to prevent the gold from being scrubbed away. The minimized operation results in a lower bond strength between the mask and gold when compared to the mask and copper. Some assemblers may find that the solder mask flakes off of ENIG plated areas under the solder mask. When you apply solder mask over ENIG it happens. If you specify solder mask to be applied after ENIG then you have to live with it. If you specify SMOBC and you get ENIG under the mask then you need to talk to your vendor.

Tags: , ,

5 Responses to “When to apply ENIG.”

  1. BillMcGee Says:

    I have some pcb’s unpopulated that aren’t soldering with reflow. They are ENIG finish. They were sent with out dessicant and appear dull like oxides are on the surface. Supplier agreed problem and will take them back. Is there a way to clean them easily. I tried a light rubbing with scotchbrite. they look okay. I also ordered pcb cleaner microclean. Any/allsuggestions appreciated. Trying to keep a line running short term.

  2. David Duross Says:

    Hello Bill,

    I would avoid using Scotchbtite on an ENIG surface. The immersion gold is very thin at about 2 to 5 micro-inches thick. It is so thin that a light scrub with Scotchbrite shall expose the electroless nickel. The gold may appear brighter but the nickel undercoat has now been compromised. The mechanics of the surface finish are as follows…

    1) The Immersion Gold is applied to keep the Electroless Nickel active. Active nickel can be soldered to.

    2) The Immersion Gold is a porous surface. Most people don’t realize this. The way the immersion process works is that to plate one atom of gold you need to exchange it for one atom of nickel (more or less). As the gold thickness builds the nickel has to travel through the gold that has been applied. The immersion gold is a self limiting process. Once the thickness reaches a critical thickness the nickel cannot pass through the gold layer and the process stops. This critical point is somewhere between 10 to 15 micro-inches.

    3) The Immersion Gold has to be between 2-5 micro-inches in thickness. Too little and the nickel does not have enough protection. Too much and you can’t solder to the boards anyway. When the gold is plated above the 5 micro-inch limit the nickel to phosphorous ratio at the nickel to gold interface becomes unbalanced. This is another bit of information most people don’t know about. Electroless nickel has phosphorous in it. When the immersion gold is applied an atom of nickel is removed. The atoms of phosphorus stay behind. When the ratio of phosphorous exceeds a critical point the phosphorous starts to attack the nickel under the gold. This leads to hyper-corrosion or black-pad.

    4) When nickel is exposed to the environment it passivates very quickly. We see this in the manufacturing process if the panels are out of solution between the nickel and gold tanks for too long. The gold will plate but peel from the nickel when tape tested. It only takes a minute or two to passivate the nickel. Passivated nickel cannot be soldered to unless you re-activate it.

    That being said, the boards that have been Scotchbited shall have questionable soldering. Not all of the nickel shall solder well since it has become passive. This could lead to problems in the field. I’ve looked at your company web site. Some of your products go into medical devices. You may be able to get the boards to solder but may ultimately have joint fractures in the field after the devices have been in use for a few months. I wouldn’t risk it.

    The problem you are having depends on what is causing the gold discoloration.

    1) If the gold is discolored only around the edge of the solderable pads then this is a skipped plating non-conformance known as the edge effect. The electroless nickel plating process operates at a high enough temperature that any solvents in the laminate shall out-gas in the process bath. The out-gassing impedes the electroless nickel plating process at the edge of the pad only. The nickel is either thin or missing at the pad edge. Hence the dis-colored gold at the pad edges only.

    2) If the entire Immersion Gold surface is discolored then it could be surface contamination. This could be contamination from rinsing with contaminated water in the bare board fabrication process. Contamination of this type is typically an issue with the bare board’s fabrication process. Contamination from rinsing could be the result of rinsing the boards with city water. What comes in from the city or town well or reservoir is anyone’s guess on any given day.

    3) Solvent fume residue being redeposited onto the ENIG surface is also a possibility. Fume residue can also result in the entire gold surface to become discolored. This would depend upon when the board fabricator applied the ENIG in their process flow. It is also an indicator of poor ventilation in the baking operation. Solder mask and letter screen final baking operations performed post ENIG can result in this problem if the oven has poor ventilation. The same is true for dry baking boards in a poorly ventilated oven.

    4) Low nickel everywhere could result in discolored gold. If you have access to a XRF (X-Ray Florescence) scope you can measure the plating thickness. A board house or service bureau that provides ENIG should have one and may measure a board for you for short money. Electroless nickel should be between 120 to 240 micro-inches if it is below 120 micro-inches its a problem.

    One thing I need to state is this. What I have written above applies only if the product you are attempting to solder to is a fresh lot of boards from the manufacturer. If there is a problem right out of the box then what I have documented above may potentially be applied as a root cause. What I have documented above cannot be applied if you received a lot of boards, soldered some without issue, placed the balance into your spares inventory and are now trying to solder the boards from spares months later. If this is a case where these are spares that do not solder then all bets are off. This now potentially becomes a shelf-life and inventory management issue on the assemblers end.

    If you proceed with the assembly process you would be accepting all of the liability. You would need to find some way to qualify the product for long term reliability.

    That being said you may want to consider taking the non-conforming boards and having them sent out for Hot Air Solder Leveling, aka HASL. If this is a lead-free application you can have SN100 applied via HASL. If this is not a lead-free application you can have Sn63/Pb37 applied via HASL. This is not a perfect solution. However, if the gold is contaminated and the nickel is still active then the HASL operation may be robust enough to burn through the contamination and form a nickel-tin inter-metallic. In other words if the boards are leveled and the pads wet with solder then it is safe to assume that you have a potential method of rework. If you proceed with this option you need to consider a few things things.

    1) Dry bake the boards in a well ventilated oven.

    2) Are the vias plugged or open. If the vias are open and free of solder mask or completely plugged with mask then you should be OK. If the vias are partially plugged with ink then you may have solder balls form in the vias as a result of the HASL rework. The solder balls may result in assembly shorts. There are ways around this but it depends upon the size of the vias.

    3) Verify the nickel thickness with a XRF scope or by micro-section. If the nickel is below thickness to begin with its no good anyway.

    Feel free to ask any follow up questions.

  3. David Duross Says:

    Also, it is important to keep in mind that the HASL process is a destructive one. If the boards are re-worked through a HASL operation you should consider the following…

    1) Electrical Test the bare boards after the HASL re-work has been done. A few boards can be done on a flying probe tester. Many boards would require a dedicated electrical test fixture. It depends upon the cost vs time to test.

    2) Knowing that the HASL process is a destructive one you may want to sacrifice a few boards and have them micro-sectioned. If the boards pass electrical test then do a random sampling. If any board fail electrical test for opens that may be traced to a through hole then that is the hole that should be sectioned. If the sections cannot meet IPC-6012 after rework then you have a problem.

    3) Visually have the re-worked boards inspected to IPC-600 acceptance criteria.

    Re-working product needs to be qualified after the re-work has been done.

  4. David Duross Says:

    I do know of a source here on the East Coast of the US that could help you out for a fee in the event you need someone to attempt a re-work. Let me know and I’ll forward their contact info to you.

  5. BillMcGee Says:

    David thanks for the extremely thorough explaination and information for that matter. Parallel to the post I also did some additional research on the subject. Based on all my findings and your input this was a new supplier and we have decided to not use them. I appreciate your efforts in giving me options and methods to try. I am overseas and my resources are limited. We will remake the boards with our “A” hoigh quality supplier. Best regrds . Bill Mc

Leave a Reply