Underfill Flow as Viscous Flow Between Parallel Plates Driven - Profile

136 IEEE TRANSACTIONS ON COMPONENTS, PACKAGING, AND MANUFACTURING TECHNOLOGY-PART C, VOL. 19, NO. 2, APRIL 1996
Consider the case where the flip-chip is oriented vertically
such that the underfill flows from top to bottom aided by
the hydrostatic pressure induced by gravity acting on the
underfill. Assuming underfill with a ycose value of 0.015
N/m was used on a chip with a 75 pm standoff (separation
distance), the capillary pressure given by (7) would be 395
Pa. The median hydrostatic pressure generated during the
flow would be pgH/2 where p is the mass density of the
underfill, g = 9.81 m/s2, and H is the chip size. A typical
underfill mass density of 1600 kg/m3 on a 10 mm square
flip-chip would generate a median hydrostatic pressure of 78
Pa. Since the hydrostatic pressure is only about 1/5 of the
capillary pressure, it is unlikely to significantly enhance the
flow rate, especially when the gap (separation distance) is
small. Vacuum enhancements can generate driving pressures
of up to one atmosphere (lo5 Pa). Since this is more than
250 times the capillary pressure, the vacuum enhancement will
have a significant effect on the flow rate. An example of the
use of vacuum to enhance underfill flow has been demonstrated
by Banerji et al. [21].
V. CONCLUSION
An exact model was developed for understanding the flow
time (t) for quasisteady, laminar flow between parallel
plates driven by capillary action, as a function of flow
distance (L), separation distance (h), wetting angle (e),
surface tension (y), and the absolute viscosity (p). The
flow time is given by
The flow time for viscous flow between parallel plates
driven by capillary action is inversely proportional to the
surface tension, separation distance and cosine of the wetting
angle, and directly proportional to the viscosity and
the square of the flow distance. This model agreed well
with experimental results of a typical underfill material
flowing between plates of glass and ceramic.
The flow rate of an underfill is strongly related to two
underfill material properties: surface tension (y) and
viscosity (p). The quantity
was defined as the coescient of planar penetrance (Q)
where 8 is the wetting angle of the underfill to the planes.
The ycos8 term can be determined by measuring the
capillary rise of a progressing meniscus.
This parameter measures the penetrating power of a
liquid into a gap between parallel plates, and has units
of velocity. This parameter was strongly related to the
temperature of the underfill material.
Gravity will not significantly enhance the flow rate of
underfill materials flowing under flip-chips. Vacuum enhancement,
however, can generate pressures of more than
250 times the capillary pressure, and will effectively
enhance underfill flow rates.
VI. RECOMMENDATIONS
In order to characterize the flow behavior of an underfill
material under a flip-chip, the coeficient of planar
penetrance should be specified at the recommended flow
temperature.
Gravity (inclined or vertical flow) should not be pursued
as a method to enhance underfill flow rates. Vacuum
enhancements, however, should be pursued.
This work does not consider the effects of surface roughness,
solder bumps, flux residues, or other obstructions
on the underfill flow. This model also does not consider
non-Newtonian flow behavior. A more complicated model
could be developed to include these factors, however its
usefulness would be limited due to additional complexity.
An empirical approach, based on the simple model shown
here, is recommended for optimizing processes including
these other factors.
ACKNOWLEDGMENT
The authors would like to thank L. Powers-Maloney and T.
Ogasawara for their characterization of the underfill properties,
to C. Avila, M. Evans, J. Glazer, H. Holder, and G. Margaritis
for their editing, encouragement and support, and to J. A.
Emerson for helpful comments on underfill surface tension
values.
REFERENCES
[I] R. R. Tummala and E. J. Rymaszewski, Microelectronics Packaging
Handbook. New York: Van Nostrand Reinhold, 1989, pp. 366-391.
[2] K. DeHaven and J. Dietz, “Controlled collapse chip connection
(C4)-An enabling technology,” in Proc. 44th ECTC, 1994, pp. 1-6.
[3] M. Kelly and J. Lau, “Low cost solder bumped flip-chip MCM-L
demonstration,” Circuit World, vol. 21, no. 3, pp. 25-28, Mar. 1995.
[4] B. Wun and J. Lau, “Characterization and evaluation of the underfill
encapsulants for flip chip assembly,” Circuit World, vol. 21, no. 4, pp.
14-17, July 1995.
[5] Y. Tsukada et al., “Reliability and stress analysis of encapsulated flipchip
joint on epoxy base printed circuit board, advances in electronic
packaging,” in Proc. 1992 Joint ASME/JSME Con$ Eleciron. Packaging,
1992, vol. 2, pp. 827-835.
[6] A. Rai et al., “COB (chip on board) technology: Flip chip bonding onto
ceramic substrates and PWB (printed wiring boards),” in Proc. I990 iat.
Symp. Microelectron. (ISHM), 1990, pp. 474481.
[7] A. Rai et al., “Flip-chip COB technology on PWB,” in Proc. 1MC 1992,
Yokohama, Japan, June 3, 1992, pp. 144-149.
[8] Y. Tsukada et al.> “Surface laminar circuit packaging,” in Proc. 42nd
ECTC, 1992, pp. 22-27.
[9] J. Lau, “Experimental and analytical studies of encapsulated flip-chip
solder bumps on surface laminar circuit boards,” Circuit World, vol. 19,
no. 3, pp. 18-24, 1993.
[lo] J. Kloeser el al., “Low-cost flip-chip bonding on FR-4 boards” Circuit
World, vol. 22, no. 1, pp. 18-21, Oct. 1995.
[ll] F. Nakano et al., “Resin-insertion effect on thermal cycle resistivity of
flip-chip mounted LSI devices,” in Proc. 1987 Int. Symp. Microelectron.
(ISHM), 1987, pp. 536-541.
[I21 D. Suryanarayana et al., “Enhancement of flip-chip fatigue life by
encapsulation,” IEEE Trans. Comp., Packag., Munufact. Technol., vol.
14, no. I, pp. 218-223, Mar. 1991.
[13] D. Suryanarayana et al., “Encapsulants used in flip-chip packages,”
IEEE Trans. Comp., Packag., Manufact. Technol., vol. 16, no. 8, pp.
858-862, Dec. 1993.
[14] D. W. Wang and K. I. Papathomas, “Encapsulant for fatigue life
enhancement of controlled collapse chip connection (C4),” IEEE Trans.
Comp., Packag., Manufact. Technol., vol. 16, no. 8, pp. 863-867, Dec.
1993.
[lS] J. Clementi et al., “Flip-Chip encapsulation on ceramic substrates,” in
Proc. 43rd ECTC, 1993, pp. 175-181.