In-situ warpage measurement of 300mm wafer with

speckle-free digital image correlation method

Yuling Niu1, Seungbae Park1 , Hohyung Lee1
1Department of Mechanical Engineering, State University of New York at Binghamton

Background Core Steps Discussion

Wafer Warpage: During the manufacturing process, Experiment Set-up Measurement Verification 1. Conventional speckle method for wafer
wafer undergoes many microfabrication process steps, In project plan, two sets of sensors are applied for different Before the wafer measurement starts, two verification tests Overall warpage contours were the same, but for
which would make wafer experience thermal cycles. field of view. were done to verify the speckle-free method. microelectronic devices built on the wafer tiny warpage
difference may lead to big problems.
Field of View Measurement Warpage Test
Sensor Usage
(mm) Sensitivity (m)
A plastic package (fig. 9) with convex surface is selected
Global Camera 350 x 280 11.67 Whole wafer for profiling. The surface is measured by both speckle-free
Microcircuits
DIC and Wyko (nanometers of measurement sensitivity).
Local Camera 10 x 8 0.33
on the wafer

Figure 1. lll-V-OI on Si wafer fabrication process flow chart Global camera

Digital Image Correlation: Local camera

Figure 14. C4 wafer absolute warpage room temperature for two methods

400 C4 Wafer Warpage C4 Wafer Warpage

150
Figure 9. Plastic package Figure 10. Warpage contour from DIC & Wyko 350
Speckle-free Speckle-free
300 100
Speckle Speckle

Absolute warpage, um
250

Absolute warpage, um
200 50

150

100 0

50

Figure 11. Comparison

Two lays of anti-
0 -50

of surface profile
reflection glass
-50

between speckle-free -100 -100

0 50 100 150 200 250 300 0 50 100 150 200 250 300
DIC and Wyko Profiler
Wafer Sample Length, mm Sample Length, mm
Figure 2. Schematic of 3D DIC
Figure 15. C4 wafer absolute warpage along the cross in fig. 14
Figure 6. Schematic of speckle-free DIC
The thermal chamber provides the temperature range from -
73 to 315. High contrast ratio and resolution projector Surface Treatment Effect 2. Speckle-free DIC Method Limitation
generates patterns on the wafer (fig. 4). For wafer level specimen, surface is flat and the surface For in plane displacement measurement, the correlation
treatment might influence the actual result. So an optical algorithm is based on the tracking of the grey value
mirror was used to check the surface treatment effect. pattern G(x, y) in small local neighborhood facets.
Figure 3. 3 DIC surface treatment methods: (a) etching; (b) spraying; (c) painting
Surface Treatment Effect
0.60

Objectives 0.45
Before painting by Wyko
After painting by Wyko
After painting by DIC
Before painting by DIC

0.30

Absolute Warpage, um
0.15

0.00

-0.15

Figure 7. Control system for the experiment: (a) oven control program; (b) linear -0.30

stage control program -0.45

-0.60

Pattern Size and Density 0 10 20 30

Sample Length, mm
40 50
Figure 16. Schematic of speckle-free DIC method limitation
Figure 12. Comparison of surface treatment effect with optical mirror specimen
Speckle-free method can adjust pattern parameters easily
and achieve optimized speckle patterns. Wafer Warpage Measurement Conclusion
Figure 4. 300mm warpage contour & patterns projected on the wafer
Once the speckle-free DIC method solidified, the actual
300mm wafer was measured.

Verified speckle- 300mm Wafer

Built speckle-free free DIC method Warpage
DIC measurement & optimized Measurement
system conventional DIC subjected to
method thermal loading

Figure 13. In-situ warpage measurement of 300mm wafer with speckle-free DIC method
Figure 8. Pattern size and density effects: pattern generation program
Figure 5. Digital Image Correlation System