Mastering Material Sets
10/01/2008
Simple Testing to Evaluate Ball Attach Fluxes.
BY JIM HISERT, Indium Corp. and SIGURD R. WATHNE PE, Sikama International, Inc.
When testing a flux on the production line is impractical due to the number of materials included in the evaluation processes, it makes sense to perform a simple comparison. This test procedure can be used to initially compare fluxes with minimal time, capital expense, and equipment.
The best way to test a flux is in the production line under actual working conditions. This can be impractical if too many materials are included in the evaluation process. There are, however, ways to understand the capabilities of range of flux materials without scrapping a large amount of production parts and time. One test procedure can be used to initially compare fluxes with minimal time, capital expense, and equipment. The key data is the quality of a flux to promote wetting of various alloys on a variety of surface finishes.1 This will be calculated as a change in solder diameter after reflow. Although solder spread is the numerical outcome of testing, cleanability of water-soluble fluxes and post reflow residue of no-clean fluxes may become apparent.
Performing this test requires only 5 pieces of equipment: a razor blade, tape, tweezers, calipers, and a hot plate. A series of multiple hot plates will speed testing because multiple temperatures can be used without waiting for temperature change and equilibrium. In this testing, a 5-zone conduction system* was used to reflow the samples. With a reflow system such as this, a precise nitrogen atmosphere can be maintained, if desired. The testing for this report was done in air to force a worst case scenario for each flux.
Solder spheres of different alloys are required for the test. The spheres were chosen to test different reflow temperature ranges, and consisted of the following alloys: 95.5Sn/3.8Ag/0.7Cu (solidus 217??C), 63Sn/37Pb (solidus 183??C), 88Pb/10Sn/2Ag (solidus 267??C), 99.99In (solidus 157??C), and 52In/48Sn (solidus 118??C). The size of the spheres tested measured .71mm, .76mm, and 1.02mm in diameter. Spheres of this size are commonly used in BGA manufacture, and are large enough to work with easily.
Various surface finishes were also used. Applicable surfaces should be chosen based on current designs and the future metallizations that might be used in later designs. Each flux/alloy combination was tested on the following: ENIG, immersion Ag, Sn, Cu, oxidized Cu, electrolytic Ni, OSP, and Alloy 42 (42Ni/Fe). Nickel and Alloy 42 are both difficult surfaces to solder to, so they are good choices to test the limits of flux activity.
1-5g of each flux in question is also required. This study included 12 ball-attach fluxes (7 water-soluble and 5 no-clean fluxes). This testing demonstrates the range of application of a flux, so it is valuable even if only testing a single flux.
Procedure
Secure the substrates to a flat surface with two parallel strips of tape covering approximately 1/8" of the top and bottom of the coupon. The tape will be used to set the depth of flux on the substrate and can be polyimide or cellophane as long as the thickness of the tape is consistent. An appropriate flux depth for testing .635-1.016mm diameter spheres is .0381mm to .0508mm.
Use a razor blade to spread the flux onto the coupon in the area not covered by tape. After a uniform layer of flux is prepared, place the solder spheres into the flux. It is good practice to use the same number of spheres on each coupon, because it will be easy to see if two or more solder spheres coalesce during reflow. It is best to keep the spheres spaced far enough apart so they do not join together during reflow. 7 spheres/1" coupon seem to work well.2 The tape used to control flux volume can be stripped from the coupon after the spheres are placed.
Different fluxes respond to reflow profiling in different ways, but to keep this testing simple and fair to all fluxes, profiling was not used. The coupons were placed on the pre-heated hot plate and removed after a set amount of time. The advantage of using conductive heating for this test is that you don’t need to use a one-size-fits-all profile. Fluxes that can accomplish proper wetting quickly are not subjected to longer profiles of a more conservative flux.
During experiment verification, it was found that differences in time above liquidus were the leading causes of variation in results (up to 74%). This parameter needs to be controlled during testing. The second cause of experimental variation originates from lot changes in substrates. By using a batch of coupons differently processed, solder spread was affected.
 Figure 1. The results of solder deposit measurement determine spread diameter. |
After the coupons were removed from the hot plate and allowed to cool, the water-soluble fluxes were washed in water. The solder deposits were later measured to determine spread diameter (Figure 1).
Data Collection and Processing
The resulting spread of each flux/alloy/surface combination can be determined by using one of two formulas:
 |
For experiments where it is difficult to determine solder joint height or if there is variable tolerance because of substrate dimensions, the spread diameter method is preferred. The spread diameter method was used for this testing.
Results and Further Testing
To provide an unbiased cross-section of ball-attach fluxes in the industry, each flux received a code that only reveals the particular cleaning method that should be used (no-clean or water wash). The percent increase in solder diameter for each flux is shown on Figure 2. Similar data was generated for each solder alloy that was tested. It is important to note that wetting tests with copper substrates are only valid when soldering to copper surfaces. The best flux for ENIG parts may not even be considered if the only flux test used is based on copper or oxidized copper.
 Figure 2. The percentage increase in solder diameter for each flux tested. |
After a database of wetting results for existing products is in place, it becomes easy to see the advantages of any new fluxes. This information helps to find new applications for existing fluxes and proper uses for prototype fluxes.
All the fluxes used are currently being tested for transfer efficiency over time. By combining the two tests along with reflow/voiding analysis and residue characterization, each flux will be understood for applications yet to be introduced.
Conclusion
The typical IPC wetting test is good for beginners, but a more intense test method is needed to really understand a flux.3 This test has been influential for material recommendations and will continue to be used to gain knowledge of new fluxes, as well as to compare existing fluxes. Performing this test first-hand offers a technological advantage in itself. A material master becomes that by mastering a set of materials. Combining this and other simple tests can save a lot of evaluation work when it comes time to select a flux or alloy for the next big project. The test method outlined here is both multifaceted and easy to perform. After becoming proficient, it is possible to test a flux with 40 different surface finish/alloy combinations in approximately one hour. In only an hour/flux, you can become a master of your material set.
*Falcon 5-Zone, Sikama International
References
- J Hisert, A Mackie, “The Evolution Revolution in Flux”, Advanced Packaging, July 2007
- JIS standard 3197
- IPC test method 2.4.45
JIM HISERT, application engineer, may be contacted at Indium Corp. 1676 Lincoln Ave. Utica, NY, 13502; 315/853-4900; [email protected]; SIGURD R. WATHNE, president, may be contacted at Sikama International, 118 East Gutierrez St., Santa Barbara, CA, 93101-2314; 805/962-1000; [email protected].