Reading time ( words)
The potential risk for whisker growth is one of the concerns when using a pure-tin lead finish. In recent years, a large amount of testing and analytical work has been performed; and the driving forces for whisker growth at various environmental conditions are better understood. This article discusses growth mechanisms for whisker growth during three accelerated tests recommended by JEDEC standards.
By Sheila Chopin and Peng Su, Ph.D.
The phenomenon of whisker growth on electroplated pure-tin finishes has been investigated extensively. Data from these studies suggest that the major driving force for whisker formation is the increased stress level in the finish, which can be affected by a range of factors. The plating process, for example, can affect the stress state by plating tin finishes with different grain sizes, thicknesses, and contamination levels. Application conditions, such as temperature and humidity levels, can also induce certain microstructure changes that affect the rate of whisker growth. This article concentrates on three JEDEC-recommended test conditions. To a certain degree, these tests represent some common field-application conditions. Mitigation techniques developed based on test results can be effective in real-life environments.
Test conditions recommended by JEDEC are summarized in Table 1. For the air-to-air temperature cycling (AATC) test, a temperature range of -40º to 85ºC is allowed; however, the range of -55º to 85ºC is used for all studies within this article.
The driving force for whisker growth during the thermal-cycling test is the most straightforward of the three tests. The variation of temperature generates thermal stresses in the tin finish because of a mismatch of coefficient of thermal expansion (CTE) between tin and the leadframe material. Due to wide temperature ranges used, a high thermal stress is introduced into the finish in a short period of time, making it possible to ignore stresses generated by slower mechanisms. The crystallographic orientation of tin grains is another important factor when determining the level of thermal stress. Tin lattice is anisotropic, meaning that on different crystal planes, or along different crystal directions, mechanical properties such as Young’s modulus and CTE, could vary. For the case of tin finish, because it usually consists of one layer of grains, we need only be concerned with the horizontal components of the CTE and E. A simplified one-dimensional view of this model is shown in Figure 1.
Figure 1. One-dimensional view of the effect of grain orientation.
When grains 1 and 2 have different E and CTE values, stress levels within the two grains could be different, even with the same thermal strain ɛ. This difference generates stress-concentration points at the grain boundary, where whiskers would have a higher likelihood to nucleate. The choice of chemistry and plating-process parameters is important to determine the microstructure of tin finishes. Even with the same plating chemistry, controlling certain process parameters can result in tin finishes of different physical appearances and grain-orientation mixes. Figures 2a/b show the surface of tin finishes with two different sets of plating parameters, both after 1,000 cycles of AATC.
Figure 2a. Whisker growth on tin finish plated with a non-optimized process.
One tool that can provide grain-orientation information is X-ray diffraction (XRD). Process variations can manifest on XRD spectrums as differences in orientation mixes or peak-intensities. Initial success has been achieved in correlation-orientation data to whisker growth during the AATC test, suggesting that XRD can be used as an effective development tool to locate the optimal process window and investigate the effects of process variations.
Figure 2b. Whisker growth on tin finish with optimized plating process.
Low Temperature/Humidity Storage
For the low-humidity, isothermal storage test, the driving force for whisker growth is more complex. Because of the long test durations, many mechanisms can contribute to the nucleation and growth process of whiskers. Post-plating recrystallization, surface oxidation, grain orientation, and intermetallic-compound (IMC) growth could all affect whisker growth rates. Among these factors, however, IMC growth at the Sn/Cu interface appears to be the most significant source of stress for the Sn/Cu material set.
Sn/Cu IMC has a high growth rate at room temperature, and the growth at grain boundaries is even faster. The extrusion of IMC grains into the tin finish increases the level of compressive stress, assisting the nucleation and growth of whiskers at grain boundaries. Currently, the most common mitigation technique for irregular IMC growth is a bake process (150ºC for 1 hour), performed immediately after plating. At this elevated temperature, the diffusivity of Cu is increased, resulting in a more uniform layer of IMC during the bake process. The benefit of the improved IMC morphology is that the extrusion of IMC grains into the finish is reduced, which decreases the level of compressive stress, and therefore, whisker growth.
High Temperature/Humidity Storage
Tin has a low melting temperature of 232ºC. At the testing temperature of 55ºC, if the humidity level is low, whisker growth is usually not observed because stresses can be relaxed by the relatively fast self-diffusion process at this temperature. However, when the humidity level is high, whisker growth is greatly increased. Careful examination of the lead surface during this test shows that corrosion of the tin finish usually occurs first; and then whiskers nucleate and grow within corroded areas. Additionally, corrosion of the tin finish usually initiates at locations where the copper substrate is exposed, such as lead tips and dam-bar cut regions (Figure 3). This suggests that the corrosion is likely induced by the difference in the galvanic potential between Sn and Cu. Tin is more anodic than copper. When moisture is present, the corrosion rate of Sn is significantly accelerated because of the formation of the Sn/Cu cell.
Figure 3. A schematic view of one formed lead, with the location of corrosion and whisker growth during the high temperature/humidity test. Red regions indicate the location of corrosion.
The effect of the corrosion is also very localized. In one study, more 10,000 leads were inspected over a 4,000-hour test period. All observed whiskers were located within the corroded area, while the rest of the finish (without corrosion) remained whisker-free. Why corrosion would accelerate the nucleation and growth process is still not well understood. It has been shown, however, that the board-mounting process can mitigate the risk of corrosion. Similar components as those mentioned above were board-mounted with tin/lead or Sn/Ag/Cu solder pastes, and tested at the same condition. After 4,000 hours of testing, few leads showed signs of corrosion and whisker growth.
To understand the difference of whisker growth between board-mounted and loose components, galvanic potential of reflowed tin/lead and Sn/Ag/Cu pastes were measured and compared to that of tin. Results show that both solder pastes are more anodic than tin, indicating that when components are mounted with either of these pastes, they can help reduce the corrosion rate of tin. Considering the effects of corrosion, for application conditions where corrosion is a concern, system manufacturers must take additional measures. For example, anti-corrosive coatings can also be applied when situations allow to control or eliminate surface corrosion and corrosion-induced whisker growth.
Testing and analytical results from several published evaluations suggest that compressive stress is the most probable driving force for whisker growth on electroplated tin finishes. Due to the complex process of whisker nucleation and growth, a systematic approach must be taken to reduce or eliminate whisker growth. Based on the current understanding of growth mechanisms, an effective mitigation strategy will include a well-controlled plating process, a post-plating bake step, and joint efforts by component and system manufacturers to reduce the rate of corrosion of the tin finish.
For a list of references, contact the authors.
Sheila Chopin, Materials Technology Integration manager, Freescale Semiconductor, Technology Solutions Organization, may be contacted via e-mail: firstname.lastname@example.org. Peng Su, Ph.D., senior packaging technologist, Freescale Semiconductor, Technology Solutions Organization, may be contacted via e-mail: email@example.com.