Table 1. The comparison among Creep Corrosion, Dendrite and CAF

In the electronic industry, there have been a number of test methods developed to evaluate the corrosion resistance from the view point of PCB materials, PCB assembly process and gas condition. The MFG test method was carried out in 1980’s. It’s the primary test method used currently in the electronic industry.
The MFG test is a laboratory test where the temperature (℃), relative humidity (%RH), concentration of gaseous pollutants (ppb level), and other critical variables (such as volume exchange rate and airflow rate) are carefully defined, monitored and controlled. [5] The purpose of MFG test is to use the combination of four most common corrosive gases in the environment, H2S, Cl2, NO2, SO2 to simulate and accelerate atmospheric corrosion due to exposure.


But there is still no accepted industry standard created for MFG test to correlate to real service life yet. From the literature published, the critical factor that causes creep corrosion is the concentration of H2S. IPC 3-11g Corrosion of Metal Finish Task Group also had a draft discussion for setting 1500ppb possibility as the minimum concentration of H2S in IPC/APEX 2011. Further progress will be updated in the next publication.

In Group B, two types NC(No Clean) and OA(Organic Acid) Flux are sprayed on PCB with the simulation of wave soldering condition by baking the flux residue on the PCB. PCBs of Group B are put in the oven at 125℃ for 5 minutes and then at 270℃ for 1 minute after the NC OA Flux are sprayed on it to simulate wave soldering condition and have flux residue remaining on the PCB.


Result and Discussion
a.niformity Test
Before and after the uniformity test, 5 weightings are conducted by the microbalance and the average is calculated. After the weight gain, corrosion product is verified by the methods of Weight Gain Analysis, Coulometeric Reduction(CR) and X-Section with SEM/EDS. The data from the three methods are consistent with each other. It can be defined that the gas inside of the chamber is uniform.

The effect of the mixed flowing gas with various H2S concentration (500 ppb, 1000 ppb, 1700 ppb) at 5 days duration showed that a much higher corrosion rate is observed on Cu coupons in both the Individual and Mixed Flowing Gas Test.

The corrosion rate of Cu coupons rapidly increases with H2S concentration after reaching at 1000 ppb. Ag coupons have more active corrosion in lower H2S concentration than higher H2S concentration. Flaking corrosion also happens on the Cu coupon with heavy corrosion product in the high H2S concentration test condition.


b.SMD vs NSMD effect on the creep corrosion
According to previous experiment, creep corrosion can be observed on both SMD and NSMD features. The degree of creep corrosion on SMD board features grows laterally across solder mask and is greater than that at the metal/laminate interface on NSMD board feature. The difference is explained as following discussion, which also refers to some published papers.
1)The migration on the laminate around NSMD areas has to overcome the land between Cu pad and solder mask.
2)SMD is a much smoother surface so that the corrosion product can travel much more readily across the planar surface.
3)A gap is made at the interface between soldermask and Cu pad shown in figure 17 and figure 18 due to the poor soldermask processing (exposed Cu at edge of soldermask), so that creep corrosion emanates from SMD features. Figure 19 shows the creep corrosion of SMD and NSMD board features.

Surface finish effect on the creep corrosion (Group A)
 ROHS legislation has been implemented and the use of lead in electronic products is prohibited. The PCB manufactures are driven to transition from lead coating final finishes to lead free alternatives. The most common lead free surface finishes applied today are ImAg (Immersion Silver) and OSP (Organic Solderability Preservatives) to date based on solder joint reliability validation. With creep corrosion concerns in high reliability product like telecom, network product and so on, the 2 available surface finish were found not to be good candidates to prevent its occurrence. Therefore, one improved ImAg surface finish with post-treatment is designed for comparison besides the above 2 candidates in the creep corrosion resistance study. Group A PCB is NSMD design with 3 different types of surface finish, and Group B PCB is Comb-Pattern design with ImAg surface finish.


Creep corrosion is observed on all the three types of surface finish based on the visual inspection and the analysis through Cross-Section,
ImAg and Post-Treatment ImAg
The comparison in SEM/EDS analysis between ImAg and Post-Treatment ImAg coating show that the corrosion rate of ImAg is higher than Post-Treatment ImAg. , the corrosion of ImAg creeps onto the trace from pad, but the corrosion of Post-Treatment ImAg only occur on the pad and doesn’t creep to the trace. The pad of ImAg has been corroded and flaked as shown in , therefore, it is concluded Post-Treatment ImAg has the better performance than normal ImAg. 


The silver layer is getting denser with better coverage of Cu pad preventing corrosion happening. The Post-Treatment ImAg with more organic preservative is designed to have better performance than ImAg under the corrosive gas exposure.
The comparison of the 2 HT OSPBoth HT OSP 1 and HT OSP 2 have worse performance than ImAg, but there is not enough evidences to judge which one is the worst. HT OSP 1 has creep and f laking corrosion. The corrosion product on the pad is thinner and smoother. HT OSP 2 has creep corrosion but the pad isn’t flaked. The corrosion product on the pad of HT OSP 2 is thicker and rougher as shown in  Figure 24. Figure 25 shows the side view of HT OSP 1 and HT OSP 2. Both of them have creep corrosion from pad to the trace and the pad crater. HT OSP 2 has thicker corrosion product on the pad.

ImAg has a higher porosity ratio than Post-Treatment ImAg and it is assumed that OSP might have higher porosity ratio than ImAg in Figure 26. Currently there are three testing methods for surface finish porosity in the industry. These are Gas Exposure Method, Electrolysis Imaging Method and Salt Spray Test Method. But all of them are not the ideal and reliable test methods to identify quantitatively the porosity.

e.Flux Residue effect on the creep corrosion (Group B)
Wave-soldering flux residues can promote creep corrosion and this has been published in Dr. C. Xu’s paper. [1][10] It is expected we will get a similar result in this study. The SEM/EDS of corroded pad with flux residue is shown in Figure 31. The creep corrosion is found in Figure 32 with SEM/EDS analysis. Flux residue will cause moisture absorption and the H+ ionic contamination will dissociate Cu oxide and accelerate creep corrosion.


Creep corrosion typically occurs when copper is exposed to an environment containing sulfur. Cu2S is the primary creep corrosion product. Cu2S is produced by the attack of the copper at the edge of the soldermask. Cu2S film can migrate across any surface that it contacts. Creep appears to begin by growth of dendrites. As the corrosion products increase in thickness, the resistance decreases until functional shorting occurs.

Conclusion and Further Research
Creep corrosion can be driven by multiple factors. In addition to environmental factors, such as pollution, temperature, humidity the complicated PCB manufacturing process is also another concern. There might be many potential influences on creep corrosion during the process and not only surface finish, flux, and board design. The corrosion occurrence on the PCB is very sensitive to surface chemical properties. Ionic cleanliness and the roughness and surface chemistry of the soldermask might be other factors that influence the rate of creep corrosion growth. That will be the topic for further study.