This document reflects the status of the research on root causes of tin whiskering. It explains the hypothesis and evidence of whisker growth and why whiskers can be prevented by short time storage at higher temperatures.
This document reflects the status of the research on root causes of tin whiskering. It explains the hypothesis and evidence of whisker growth and why whiskers can be prevented by short time storage at higher temperatures.
Whisker formation with lead free plating is a known phenomenon. Whiskers are single crystals, protruding from electroplated layers, originating from diffusion under the influence of compressive stress. Figure 1 gives an example of a whisker.
Long whiskers can have influence on quality of electronic devices as those can cause short circuits. It is therefore important to understand the root cause of whisker growth.
Knowledge of the growth mechanism of whiskers can be used to develop a whisker test. Until now, no standardized test is available for industry [18].
Figure 1(left): Example of a column type whisker showing striations along the length of the whisker [18]
Figure 2(right): Individual grains of Sn in a cross-section of a C19400 substrate plated with ~12 µm Sn with irregular intermetallic at the grain boundaries, (the marker depicts 15 µm) [18]
It is generally accepted that whiskers form because of compressive stress in the plating layer. Collected data on Cu based lead-frames support a hypothesis that the formation of irregular Cu6Sn5 intermetallics generate these compressive stresses. It can be calculated that the irregular growth of this intermetallic can introduce 58% additional volume in the plating layer, showing that the existence of compressive stress is plausible.
This Cu6Sn5 is irregular after storage at lower temperatures, because at these temperatures grain boundary diffusion predominates over bulk diffusion. At higher temperatures, bulk diffusion causes a homogeneous Cu6Sn5 layer, resulting in less stress. This also explains the observation that whiskers grow fastest at ambient temperature [28], [30], [31], [32].
Figure 2 presents a cross-sectional view, showing irregular growth of the Cu6Sn5 intermetallic in the grain boundaries of the columnar matte Sn deposit due to grain boundary diffusion at lower temperatures.
By selectively etching away the Sn, the morphology of the intermetallic can be shown, see Figure 3 and Figure 4.
Figure 3 shows the intermetallic structure after the sample had been stored for 6 months at room temperature and subsequently the Sn had been etched away. Here it can be seen that the intermetallic grows predominantly at the grain boundaries, while almost no intermetallic grows in the centre.
If one compares the morphology of the intermetallic phase after 1 h at 150 °C and subsequent etching as in Figure 3 with the one shown in Figure 4 , it is obvious that the intermetallic is much more regular at higher temperatures than at ambient. This is because at higher temperatures (> ~75oC), the predominating bulk diffusion forms a homogeneous Cu3Sn/Cu6Sn5 layer. This regular layer forms a diffusion barrier against further irregular growth of the Cu6Sn5 at lower temperatures.
Figure 3(left): Surface view of the intermetallic after storage at room temperature for 6 months and subsequent selective etching of Sn [18]
Figure 4(right): Surface view of the intermetallic after 1 hour at 150 °C and subsequent selective etching of Sn [18]
A further beneficial effect of a high temperature treatment is recrystallization of the Sn and annealing of already present stress. Therefore, a post bake treatment of 1h at 150 °C is introduced in several factories of component manufacturers to reduce whisker growth.
The above mentioned theory holds for Cu based leadframe materials. For other materials such as FeNi42 whisker growth is caused by the mismatch of the coefficient of thermal expansion (CTE) of the leadframe material and the Sn plating. Therefore, whisker formation can be observed in these samples predominantly after temperature cycling.
A quantitative explanation of the results will require a whisker model that recognizes more than the dominant effects. Finite element stress modeling, starting with the dedicated material property set, will support refinement and verification of the hypothesis. The goal is that this, ultimately, will result in an industry-wide accepted acceler-ated test for whisker growth.
[18] Pascal Oberndorff, Philips CFT, Eindhoven, The Netherlands, EFSOT Europe; pascal.oberndorff@philips.com
[28] Dittes, M; Oberndorff, P; Petit, L. ?Tin Whisker Formation ? Results, Test Methodsand Countermeasures?, Proc. 53. ECTC 2003, New Orleans, pp 822-826
[30] Oberndorff, P; Dittes, M; Petit, L; Chen, C.C.; Klerk, J; Kluizenaar, E.E de ?Tin Whiskers on Lead-Free Platings? Proc. SEMI Technology Symposium, Advanced Packaging Technologies II, August 2003, Singapore, pp 51-55
[31] Oberndorff, P; Dittes, M; Petit, L. ?Intermetallic Formation in Relation to Tin Whiskers? Proc. International Conference on Lead-Free Electronics 2003, Brussels
[32] Dittes, M.; Oberndorff, P.; Crema, P.; Schroeder, V. ?The Effect of Temperature Cycling on Whisker Formation?, IPC/JEDEC 4th Internation Conference on Lead-free Electronic Components and Assembly, October 2003, Frankfurt Germany
Dr. Pascal Oberndorff, Philips CFT, The Netherlands, EFSOT-Project, 2003
Whisker formation with lead free plating is a known phenomenon. Whiskers are single crystals, protruding from electroplated layers, originating from diffusion under the influence of compressive stress. Figure 1 gives an example of a whisker.
Long whiskers can have influence on quality of electronic devices as those can cause short circuits. It is therefore important to understand the root cause of whisker growth.
Knowledge of the growth mechanism of whiskers can be used to develop a whisker test. Until now, no standardized test is available for industry [18].
Figure 1(left): Example of a column type whisker showing striations along the length of the whisker [18]
Figure 2(right): Individual grains of Sn in a cross-section of a C19400 substrate plated with ~12 µm Sn with irregular intermetallic at the grain boundaries, (the marker depicts 15 µm) [18]
It is generally accepted that whiskers form because of compressive stress in the plating layer. Collected data on Cu based lead-frames support a hypothesis that the formation of irregular Cu6Sn5 intermetallics generate these compressive stresses. It can be calculated that the irregular growth of this intermetallic can introduce 58% additional volume in the plating layer, showing that the existence of compressive stress is plausible.
This Cu6Sn5 is irregular after storage at lower temperatures, because at these temperatures grain boundary diffusion predominates over bulk diffusion. At higher temperatures, bulk diffusion causes a homogeneous Cu6Sn5 layer, resulting in less stress. This also explains the observation that whiskers grow fastest at ambient temperature [28], [30], [31], [32].
Figure 2 presents a cross-sectional view, showing irregular growth of the Cu6Sn5 intermetallic in the grain boundaries of the columnar matte Sn deposit due to grain boundary diffusion at lower temperatures.
By selectively etching away the Sn, the morphology of the intermetallic can be shown, see Figure 3 and Figure 4.
Figure 3 shows the intermetallic structure after the sample had been stored for 6 months at room temperature and subsequently the Sn had been etched away. Here it can be seen that the intermetallic grows predominantly at the grain boundaries, while almost no intermetallic grows in the centre.
If one compares the morphology of the intermetallic phase after 1 h at 150 °C and subsequent etching as in Figure 3 with the one shown in Figure 4 , it is obvious that the intermetallic is much more regular at higher temperatures than at ambient. This is because at higher temperatures (> ~75oC), the predominating bulk diffusion forms a homogeneous Cu3Sn/Cu6Sn5 layer. This regular layer forms a diffusion barrier against further irregular growth of the Cu6Sn5 at lower temperatures.
Figure 3(left): Surface view of the intermetallic after storage at room temperature for 6 months and subsequent selective etching of Sn [18]
Figure 4(right): Surface view of the intermetallic after 1 hour at 150 °C and subsequent selective etching of Sn [18]
A further beneficial effect of a high temperature treatment is recrystallization of the Sn and annealing of already present stress. Therefore, a post bake treatment of 1h at 150 °C is introduced in several factories of component manufacturers to reduce whisker growth.
The above mentioned theory holds for Cu based leadframe materials. For other materials such as FeNi42 whisker growth is caused by the mismatch of the coefficient of thermal expansion (CTE) of the leadframe material and the Sn plating. Therefore, whisker formation can be observed in these samples predominantly after temperature cycling.
A quantitative explanation of the results will require a whisker model that recognizes more than the dominant effects. Finite element stress modeling, starting with the dedicated material property set, will support refinement and verification of the hypothesis. The goal is that this, ultimately, will result in an industry-wide accepted acceler-ated test for whisker growth.
[18] Pascal Oberndorff, Philips CFT, Eindhoven, The Netherlands, EFSOT Europe; pascal.oberndorff@philips.com
[28] Dittes, M; Oberndorff, P; Petit, L. ?Tin Whisker Formation ? Results, Test Methodsand Countermeasures?, Proc. 53. ECTC 2003, New Orleans, pp 822-826
[30] Oberndorff, P; Dittes, M; Petit, L; Chen, C.C.; Klerk, J; Kluizenaar, E.E de ?Tin Whiskers on Lead-Free Platings? Proc. SEMI Technology Symposium, Advanced Packaging Technologies II, August 2003, Singapore, pp 51-55
[31] Oberndorff, P; Dittes, M; Petit, L. ?Intermetallic Formation in Relation to Tin Whiskers? Proc. International Conference on Lead-Free Electronics 2003, Brussels
[32] Dittes, M.; Oberndorff, P.; Crema, P.; Schroeder, V. ?The Effect of Temperature Cycling on Whisker Formation?, IPC/JEDEC 4th Internation Conference on Lead-free Electronic Components and Assembly, October 2003, Frankfurt Germany
Dr. Pascal Oberndorff, Philips CFT, The Netherlands, EFSOT-Project, 2003
Top
Version for print