The corrosion test, also called the “copper mirror test,” differs from the copper coupon test because it qualitatively determines the corrosivity of the flux before reflow rather than post-reflow. This test is very simple and, in some ways, somewhat archaic. The test vehicle is a microscope slide that has been coated on one side with a very thin layer of copper. To perform the test you drop raw flux on the copper side of the slide. For comparison purposes, a control flux consisting of just rosin and alcohol solvent is also added to the slide. The control is necessary in the event that the mirror is defective. It will also tell you if there are environmental factors that might contribute to the corrosion. Rosin flux with no activators should never corrode a copper mirror.
Once prepared, the sample slides are aged for 24 hours under set temperature and humidity conditions. Then, they are examined to see how much, and to what degree, the flux has started to eat into the copper. The corrosiveness is identified in the third character of the J-STD designation code:
L = low activity/corrosivity
M = moderate activity/corrosivity
H = high activity/corrosivity.
WF-9940 
WF-9945 
WF-9958 
WF-9955 
Three of the Project 99 fluxes are low activity (L) because, other than staining the copper, they didn't react with it. However, Shadow Warrior (or WF-9958) (lower right) has a halo where it started to attack the copper a little bit; therefore, we have classified it as an M-type flux. The M designation alone means nothing regarding the post-soldering reliability of the flux; it is only an indicator of flux activity. For example: when comparing the The Alchemist (WF-9955) to Shadow Warrior, Shadow Warrior is the stronger flux.
Although the copper mirror test is widely used for flux classification, it is somewhat outdated and has no relationship to the production environment. Typically, wave fluxes are sprayed onto the circuit boards and then immediately go through the pre-heating cycle and then the wave itself. A wave flux would be never be applied to the circuit board and then aged 24 hours before going through the wave soldering or selective soldering machine, which essentially renders this test useless.
The benefit of an M flux is that it reduces the number of soldering defects when compared to similar L fluxes. For good reliability, it is best to look at the post-soldering corrosion test as well as the post-reflow SIR and electro-migration tests together to determine if a flux is safe to use for your assemblies.
If you would like to learn more about our Project 99 wave fluxes, or the procedures under which they were tested, contact me at ebastow@indium.com.
