Effect of Plating Variables on Whisker Formation
in Pure Tin Films
Stephanie Miller
Advisors: Uttara Sahaym, M. Grant Norton
Materials Science & Engineering REU Program, Department of Mechanical & Materials Engineering, Washington State University
Abstract
Microelectronic failures due to spontaneous whisker formation on pure tin
plating have prompted investigation of mitigation processes. This experiment
explored the affect of plating variables on whisker growth. Increasing the
plating thickness above 5μm was found to decrease whisker formation , and
no whiskers were observed below at a thickness of 1μm. Varying the plating
bath temperature was found to have little effect on whisker growth, though
whisker morphology was affected.
a.
20μm
20μm
b.
20μm
20μm
c.
20μm
20μm
a.
5μm
5μm
b.
5μm
5μm
c.
5μm
5μm
d.
5μm
5μm
e.
5μm
5μm
f.
Figure 5. Surface features of varied thickness samples, a. 5μm/45°C sample, whiskers (10μm max)
after aging for 7 days, b. 5μm/45°C sample, long whiskers (40μm max) after aging for 28 days, c.
1μm/45°C sample, no whisker growth after aging for 28 days
Figure 1. Hook terminal of EM relay2
Figure 2. Ceramic chip capacitor2
Introduction
Tin whiskers are single-crystal filaments approximately one micron in thickness
and up to several millimeters in length. Their spontaneous formation on
microelectronic components is known to cause short-circuiting failures, and
recently there has been renewed interest in developing low-cost methods of
whisker prevention (Figures 1 & 2).1 This research examined the effect of
plating variables on whisker growth in tin-plated copper samples, with the goal
of finding a combination of plating parameters that would mitigate whisker
formation. Surface morphology was also compared among all the samples.
Procedure
Tin was electrodeposited onto four copper substrates to thicknesses of 1, 5,
10, and 15µm in an alkaline bath at a temperature of 45˚C. Four other samples
were electroplated to a thickness of 5µm at bath temperatures of 35, 55, 70,
and 85˚C. After electroplating, all of the samples were examined periodically
via scanning electron microscopy (SEM) to compare their surface
morphologies and determine whisker growth onset time.
Results & Discussion
• Below 85°C, increasing the plating thickness and increasing the bath
temperature increased surface roughness (Figures 3 & 4). The 85˚C sample had
a smooth surface of faceted grains.
• At thicknesses below 5µm, whisker growth was directly proportional to
surface roughness. Above 5µm, whisker growth decreased as surface
roughness increased (Figure 5).
• Pyramidal surface grains were observed on samples in the plating thickness
range of 5-15µm and within the temperature range of 45-70˚C (Figure 6).
• Dendritic structures were noted on the 10µm and 15µm samples, particularly
at their edges (Figure 6).
• Temperatures below 85°C had little effect on whisker formation onset time,
but did affect the morphology of the whiskers (Figure 7).
a.
2μm
b.
2μm
2μm
c.
Figure 6. Surface features of temperature varied
samples, a. Smooth surface of 85°C sample, b.
Pyramid structures noted on 5μm/55°C sample,
c. Pyramids on 5μm/70°C sample, d. Dendritic
structures noted on edges of 5μm/55°C sample
d.
5μm
5μm
Table I
Summary of the Effect of Plating Variables on Whisker formation
Plating
Thickness Temperature Aging
(μm)
(°C)
Time
1
45
28
5
45
28
10
45
28
15
45
28
5
35
14
5
55
14
5
70
14
5
85
14
Onset
Time
(days)
28+
7
14
21
7
7
7
14
Whisker
Length,
Avg
N/A
20
5
5
2
5
5
10
Whisker
Length,
Max
N/A
40
10
15
9
25
17
15
Notes
No whisker growth observed
Long, thin whiskers
Mostly nodules, some whiskers
Mostly nodules, a few whiskers
Short, thin whiskers
Long, thin whiskers
Thick whiskers, some nodules
Thick whiskers/nodules
2μm
2μm
5μm
5μm
Figure 7. Whisker morphologies, a. short thin whisker 5μm/35°C sample, b. long
thin whisker 5μm/45°C sample, c. nodule 5μm/55°C sample, d. thick whisker
5μm/55°C sample, e. short thick whisker 5μm/70°C sample, f. short thick whisker
5μm/85°C sample
Results & Discussion, continued
• The quickest onset time and maximum whisker growth were observed in 5μm
plating, with both of those factors decreasing in thicker plating layers (Table I).
The 1μm sample showed the least susceptibility to whisker formation.
Conclusions
• Varying the plating bath temperature did not have a substantial effect on
whisker growth onset time, but it did influence whisker morphology.
• Whisker formation was dependent on the plating thickness, with the quickest
onset time and maximum whisker length observed in the 5µm tin-plated
sample.
•The 1μm/45°C sample had the best result, with no observed whisker formation
after aging for 28 days.
• Below 85°C, increasing the plating thickness and increasing the bath
temperature increased surface roughness. At 85°C, the surface became smooth
and uniform.
• At thicknesses below 5µm, whisker growth increased as surface roughness
increased. Above 5µm, whisker growth decreased as surface roughness
increased.
Suggested Further Work
• Cross section analysis of plated samples
a.
10μm
10μm
b.
10μm
10μm
c.
10μm
10μm
d.
Citations
1 Osenbach, J. et al, “Sn Whiskers: Material, Design, Processing, and Post-Plate
Reflow Effects and Development of an Overall Phenomenological Theory,”
Transactions, 28 1 (2005).
2 Photos courtesy of NASA GSFC.,
<http://nepp.nasa.gov/WHISKER/PHOTOS/INDEX.HTML>
10μm
10μm
Figure 3. Surface features of varied thickness samples plated at 45°C. a. 1μm, b. 5μm, c. 10μm, d. 15μm
Acknowledgements
Special thanks to Hi-Rel Labratories, Inc. for their
support of this project.
a.
10μm
b.
10μm
10μm
c.
10μm
10μm
d.
Figure 4. Surface features of temperature varied samples plated to a thickness of 5μm, a. 35°C, b. 45°C, c. 55°C, d. 70°C, e. 85°C
10μm
10μm
e.
10μm
10μm
This work was also funded by National Science
Foundation's Research Experience for
Undergraduates program under grant number
DMR-0755055.