Guest Editorial -For Plateworld.com
Don Baudrand, Don Baudrand Consulting, e-mail:firstname.lastname@example.org
Nickel Plating Failure
A CONSULTING REPORT
Objective: Investigate and identify possible/probable causes of failure of nickel-plating on mild steel items.
I was given a comprehensive tour led by the people responsible for manufacturing and engineering who introduced me to the nature of the problems.
The nature of the failure: Plating failure occurred after extended use. Failure was separation of the electroless nickel plating from the mild steel items.
Observations: The failure was characteristic of aggressing chemical attack of the plated deposit with extensive non-oxygen corrosion (corrosion beneath the plated film) and pit corrosion on the surface of the nickel plating.
Factors influencing failure: Sharp edges, burrs, porosity in the solder joints aggressive cleaning materials and possible loss of adhesion are inherent in the plating process if procedures are not adequate or are not followed.
The aggressive cleaning procedure is a major cause of failures of plated items. The additional possibility of stray electrical currents, thermal cycling and flexing adds to the corrosion potential. There is a great deal of variation of circumstances including varying water hardness, chloride in the water, the many different cleaning solutions available, and the variance in recommended amount of cleaner to use contribute to the results.
To protect the items the plating must be nearly perfect in procedure, control and maintenance for all aspects of the preparation and plating procedure. Contributing to the plating difficulties are sharp edges burrs, type of solder-flux, the type of solder, the procedures for soldering, and cleaning solder porosity. There are also possible factors not seen by me such as the rinsing pattern inside the washer, the heat distribution in the soldering furnace, etc.
There may be no is single cause, but many contributing causes. I concluded that the plating procedure was adequate to produce good quality results, were it not for the aggressiveness of the cleaning chemicals and cleaning procedures. and the other factors that will be discussed in detail. I recommended some changes.
I am aware of failures that have occurred where no cleaning took place and no aggressive materials were used. In the following report I will deal with the nature of corrosion and the many factors involved.
This report includes details of the plating process as used by your supplier. The report includes observations and suggestions for some remedial changes in procedures.
Corrosion takes place due to electrochemical (battery) action. The source of the electrical current can be external by so-called stray currents or direct electrical contact or, likely in this case, dissimilarities such as two dissimilar metals in contact with each other forming a battery, often called corrosion cells. Corrosion cells are due to trace materials in or on the surface, differences in structure of the electroless nickel deposit, grain size differences and trace materials in the deposit. Some materials are necessary in the deposits that are present as a part of the plating process such as phosphorus as an alloying metal and trace materials that are used as stabilizers and codeposited with the nickel Phosphorus alloy. For example stabilizers such as lead, molybdenum, or numerous other similar materials, metallic or non-metallic. Pit corrosion can start from these because of the electrochemical cell between nickel and the material. High phosphorus electroless nickel processes are the most chemical resistant of any plated deposit. This fact highlights the aggressive nature of the corrosive materials used for cleaning, and the need to look for and eliminate stray currents from the system.
For corrosion to take place, water (or moisture) must be present in some form. There are two basic types of corrosion. One requires oxygen; the other does not. Oxygen is supplied from air or dissolved in water or from corrosion cells on the surface. Non-oxygen corrosion sometimes called crack corrosion. This is similar to filiform corrosion under painted surfaces. Non-oxygen corrosion is the type that occurs underneath the coating. It starts as a cell at a discontinuous spot such as a hole in the plating or where a burr is broken off, or any porosity such as in the solder and proceeds to corrode under the plating resulting in flaking or complete removal of the plated deposit.
Regarding items that show nickel flaking, it is probable that some stray currents of electrical contact may have occurred. Once the corrosion starts, the non-oxygen corrosion starts in the pores burrs, etc. Oxygen corrosion will take place simultaneously accelerated by any electrical currents including proximity to magnetic fields. I do not have a sure correction for this situation except to say that the corrosion is aided by porosity, sharp edges, etc.
To continue with corrosion: My observations are that from the photomicrographs and personal observation that indeed pitting corrosion occurs. This is greatly accelerated by the aggressive cleaning processes. The good chemical resistance of high phosphorus electroless nickel deposits emphasizes the aggressiveness of the corrosive cleaning. These products are designed to de-scale hard water products.
Electroless nickel deposits are hard materials. Typically the hardness of high phosphorus deposits has a hardness of about 500 to 550 Knoop hardness number (200 gram load). For the engineers, this is about equivalent to 45-50 Rockwell-C. There is little ductility. Thermal cycling may crack the deposit allowing corrosion to start. Flexing the unit may cause cracking. This must be kept in mind when handling and installing the items . Copper work hardens. Repeated flexing may lead to cracking of the copper resulting on crack propagation in the nickel.
The plating process
Plating was observed in detail. First, addressing the possible causes of adhesion failure due to the plating process:
The last rinsing stage in your shop should be such that the gaps are rinsed well prior to drying and packaging for shipment. The cleaning and rinsing appeared me to be adequate for the plating process provided there is no entrapped of charred materials in the intersections of the grid from the IMI last cleaning and rinsing process.
Comments: When I surveyed the plating line and procedures my attention was drawn to all of the above items and many more items. I am familiar with all the requirements and operating parameters for this type of plating necessary to produce adherent quality plating deposits.
Survey results: The soak cleaner was at the proper concentration and correct temperature and time for good cleaning. The rack was held over the cleaning tank after the cleaning was complete, spray rinsed by hand including the top of the rack, and both sides of the grid, then transferred to a tank equipped with spray rinse nozzles.
Then an immersion water rinse moving the rack up and down three times, and agitated by hand by the operator. Low-pressure blowers (oil free) are used for agitation to increase rinsing efficiency. I consider this adequate rinsing.
The next step is electroclean using anodic current. Followed by rinsing in the same way as after the soak clean using separate rinses stations. Again this is adequate for the purposes.
The acid/de-oxidizer solution was correct time and temperature. The rinses that followed were used the same as for the soak clean.
The rack was transferred to a separate line without due loss of time. This line duplicates the cleaning/ acid from the previous line of process tanks and used in an identical manner.
The next step is the cyanide copper strike. (This is a true low efficiency copper strike) the time and temperature were appropriate for this step as were the rinses following, the mild acid dip followed, then rinses and into the electroless nickel tank. The temperature was 190 F, correct for this operation. Mild agitation is used from sparger directed away from the grids and pass through a filter using 1-micron filter cartridge. All this is correct for the application. I could see no differential agitation patterns. The rinsing after electroless nickel was adequate. The final rinse uses DI water.
The testing and solution maintenance was correct for the processes involved. The records were maintained well. The rack design was correct for the application. It showed some creative ingenuity to design such a rack that works so well
Conclusion: I found nothing in the plating process that was improper or that would result in failure due to poor adhesion, or porosity due to the plating process. The adhesion appeared to be good. No undercutting corrosion was visible. There is likely some minor porosity inherent in the plating. This is due to imperfections in the basis metal, or co-deposition of insoluble particles in the plating solution. The continuous filtration through a 1-micron filter is designed to eliminate particulate matter. The deposit from the rack I followed through the process did not show roughness visibly or to the feel. However there could be micro roughness not visible. I will recommend a change that will help deal with that potential problem. Roughness does not always result in pitting. The plated steel panel could also be used to verify thickness, although there may be a slight variation from plated copper. A wider specification may be required. I am not convinced that 1 mil is the correct thickness for the best performance. This is something that should be investigated sometime in the future, because of the brittle nature of electroless nickel a thick deposit, i.e. 1-mil will crack more easily than a slightly thinner deposit. I understand that the thickness matches a competitors practice. My recommendation will deal with this matter as it is tied to the copper plating thickness.
Testing and Qualification
A porosity test would be useful. The problem is that I know of no porosity test for nickel onto copper. I will research this and report later. The test for porosity called the ferroxyl test is useful for nickel onto iron alloys. A 1-mil thick electroless nickel plate on a steel panel should show no porosity. The panel must be a clean polished panel. There are such panels available commercially called Hull-Cell test panels. They are polished steel zinc plated to protect them from corrosion until use. The zinc is stripped, electroless nickel deposited to the desired thickness. This becomes the test panel. A color change takes place for each pore, if any. This is a standard, widely used test. This only tests the intrinsic porosity and does not necessarily represent porosity from plating copper grids because I feel that most of the porosity starts with the grids. The thickness may differ slightly from that which uses copper. The thickness specification may have to be widened slightly.
An accelerated test was developed by Don Wiley and or associates that is similar to the one used by another plater. With some modifications, the tests can be made to be the same. This test can approximate the useful life to the grids. Both are acquainted with the way to modify your current test method to produce more predictable results. This type of test would show porosity as well as a life test. The coating would fail rapidly where porosity is present.
Changes and recommendations:
I suggest that a thicker deposit of copper may solve many of the problems. I recommend that an evaluation be made using approximately 0.0003" (three tenths of a mil) copper plating instead of a copper strike. The copper strike is deliberately an inefficient process designed to improve adhesion on steel, solder, and other metals other than copper as a basis metal. It works well for the intended use. Copper plating solutions do well over solder, as well as the copper strike.
For plating over solder that may be somewhat porous, a thicker deposit would provide better protection, and help fill pores in the solder or copper. The copper solution has a lower viscosity than nickel plating solutions making it easier to enter pores and deposit copper. The thicker copper at the interface of the nickel also acts as a cushion under the brittle electroless nickel. Copper from a copper plating solution is very soft compared with sheet copper. When heat/cold cycles occur or if there is deformation, the electroless nickel will be more protected from cracking by the ductile copper.
Another consideration is that when metals are heated there is inter-diffusion of the metals. In the case of copper over tin and silver there is rapid diffusion of the three metals. The result is that there is a very weak eutectic alloy of copper and tin that will break apart easily resulting in blisters. A thicker copper plate will provide more copper to form a stronger alloy less likely to break apart.
I agree with the radius of the tip edges of the grid components. It should be done such that there are no sharp edges left, and no deep scratches. Sharp edges deep scratches and burrs are likely to cause failure of the nickel deposits.
Improved solder and soldering conditions: The less tin exposed the better the adhesion will be on the solder.
Better flux; preferably a non-chloride flux. Chlorides attach to various transition metals such as iron alloys and copper leaving a corrosive material on the surface. Presumably a non-chloride etch will remove most of the chlorides. Mid-America uses a non-chloride etch in the plating process. I believe you are looking into this
Improve the washing system used prior to the soldering operation. The cleaner was allowed to reach pH 7 during usage. There is little or no cleaning at that pH. I suggest a pH range of 11-13, no less than 11. (Slightly over 13 will not be problem) This means additions to the cleaner and possible more frequent changes. We have contacted the supplier to re-calibrate the drop test kit and to recommend procedures to keep the cleaner up to over pH 11. There was a hand wiping operation after the rinse station. I was told that the rinsing is in the process of improvement.
Faster cooling after soldering would limit the "run-out" of tin from the solder joint. I believe that this is being corrected. (Tin-copper week eutectic alloy prevention)
Another approach to reducing the sharp edge and remove burrs is chemical deburring. It may be useful to investigate chemical suppliers that sell deburring processes. If you cannot find one I will give it a try. It may or may not be better than the mechanical process you have selected.
Bead blasting to clean the burnt solder flux may be applicable, but it would take evaluation to determine the right bead size, pressure, time, positions, etc. Also there is a possibility of driving the charred flux into the surface.
Material handling in the manufacturing area may contribute to cleaning problems. Fingerprints are notoriously difficult to clean short of etching. I know there is light etching twice in the plating process, but is it enough? Fingerprints are acidic and tend to etch as well as leaving oils. Gloves are suggested for operators or anyone that may handle the copper components or finished product. This may not be popular with the employees. Some may like gloves.
I recommend that your supply the cleaning solution to its customers for the end use cleaning. This cleaning solution should be formulated to be mild yet effective in removing scale and microbial contamination. Two ways to do this: contact a reliable cleaning solution manufacturer giving the criteria you want a cleaner to accomplish and that it be less aggressive than the present cleaners on the market. Or have such a cleaner formulated and made in your facility, or have a formulation that can be made under contract to a "toll chemical manufacturer" other than a cleaner maker that may be a competitor.
Pursuant to finding or formulating a suitable cleaner, the cleaners now used could be analyzed for total acidity, reported as phosphoric acid. Then compared by panel tests on electroless nickel plated specimens for degree of aggressiveness and cleaning performance by coating a portion with hard water minerals and any other contaminants characteristic of life use. Select the least aggressive or nearly less aggressive cleaner for a model and test the effectiveness.
I have reviewed the changes and the proposed changes in the information given to me and found them to be well thought out and useful.
You may download this article FREE in .pdf form, save it or share it with a colleague. Click here.