Before each new batch of SAC Solder Paste is introduced, we must use a laser particle size analyzer to measure the actual D10, D50, and D90 data, because the D50 of solder powder with the same nominal mesh size often fluctuates by ±2.3μm between different suppliers. If the width-to-thickness ratio of the stencil opening is less than 1.5, solder powder with a D50 exceeding 26.5μm will undergo particle stacking under squeegee pressure, resulting in increased edge burrs of the solder paste after printing and a decrease in height consistency by more than 12%.

When the sphericity of the solder powder is lower than 92%, the rolling resistance of the solder paste on the stencil aperture wall increases significantly, especially in the fine-pitch area of 0.3mm pitch QFN. After printing, the standard deviation of solder volume per pad expands from ±4.2mg to ±7.8mg. Meanwhile, the Flux is easily splashed in the early stage of reflow after being squeezed by irregular particles, causing the tombstoning rate of nearby 0201 components to increase by 0.18%.

On-site SEM observation reveals that the surface oxide film thickness of solder powder with an oxygen content exceeding 0.55wt% reaches 8 to 12nm. This film not only weakens the wetting tension between the solder paste and the copper pad, but also induces local agglomeration during the squeegee shearing process, resulting in missing prints with an area of more than 0.15mm on 2 out of every 10 boards when printing with a 300-mesh stencil. Under the same process conditions, batches with an oxygen content ≤0.42wt% have no missing prints for 500 consecutive boards.

For solder paste with a stirring time of less than 3 minutes, the secondary agglomeration ratio of the internal solder powder rises to 17%. At this time, even if the D50 is qualified, a large number of non-uniform particle clusters with a diameter >8μm can still be seen under a microscope after printing. These clusters are broken at the moment of demolding to form tails. When the tail length exceeds 30% of the pad width, the risk of subsequent placement offset increases by 2.4 times.

We have compared three types of solder powder solder paste with a nominal Type 4. The measured D50 of two of them are 24.1μm and 25.9μm respectively, while the third one, although marked as 25μm, has a measured D90 of 42.7μm. As a result, the printing yield of the first two on a 0.4mm opening stencil reaches 99.2%, while the latter causes bridging in the BGA area of 17 consecutive boards due to coarse particle jamming, and the frequency of stencil cleaning after shutdown increases from once every 4 hours to once every 1.2 hours.

If the storage temperature of the solder paste is higher than 25℃ for more than 48 hours, the softening of the internal rosin resin will accelerate the sedimentation of the solder powder. At this time, although the sampled D50 remains unchanged, the enrichment of particles >35μm in the bottom slurry is 3.7 times that in the top. Directly putting it on the machine for printing will inevitably cause the solder volume of the first 10 boards to increase by more than 15%. It must be re-stirred and left to return to a temperature of 22±2℃ before use.