What
is Acoustic Microscopy?
Acoustic Microscopy is a nondestructive method
used to detect and localize surface and subsurface flaws as well as cavities. Acoustic
Microscopy has been used for several decades to inspect large metal structures such as
long steel shafts and rotor forgings. During the last decade acoustic imaging has been
applied successfully to IC analysis, locating areas of poor die attachment and subsurface
anomalies.
History
The scanning acoustic microscope (SAM) was
first developed by Lemmons and Quate in the early 1970's. It was originally developed for
biomedical imaging purposes, specifically improved imaging contrast in tissue samples.
Contrast in the SAM is determined by the large variation in elastic properties in these
samples compared to a relatively small variation in dielectric properties, which determine
optical contrast. Tissue samples normally require complicated chemical staining for
optical microscopy. In years following its discovery, the frequency of the SAM was
continuously increased until the lateral resolution was comparable to that of optical
microscopes.
Acoustic Microscopy works best as an FA tool
for localizing delaminations and/or cracks in plastic packaged ICs. Delaminations and
cracks usually result from mismatched coefficients of thermal expansion in the packaging
materials. Figure 1 shows the thermal coefficients of expansion for various materials in
an IC package and where delaminations and cracks may occur as a result.
Theory of Acoustic Microscopy
Acoustic imaging uses the interactions between
high frequency beams of transverse sound waves (1-50 MHz) and the test sample to localize
areas of different acoustic impedance. Sound waves are reflected at interfaces of
different acoustic impedance and attenuated as they travel through the sample material.
Almost complete reflection occurs at interfaces where the difference in the acoustic
impedance is large such as the interface between metal and air. No reflection occurs at
interfaces where the acoustic impedance does not change. By analyzing the transmitted or
reflected sound beam, subsurface flaws may be localized.
There are 3 modes used to display pulse echo
information. In the A mode the transducer is not x-y scanned and the acoustic information
is generated from a spot. In the B mode the transducer is scanned along a line. A
2-dimensional image is generated in which one axis is the x-position of the transducer and
the other axis is the time of flight for reflected pulses. The amplitude of the reflected
pulse is mapped as intensity on the time of flight axis. Images created in the B-mode
resemble cross-sections of samples. In the C-mode the transducer is moved in a raster
fashion. A 2-dimensional image is generated in which the x-y position in the image
corresponds to an x-y position of the transducer. The intensity of a given x-y position on
the image is proportional to the amplitude of the reflected signal at a given time of
flight. Using this time window a C-mode image can be generated to the extent of the flaw
at a particular depth in the sample.
Acoustic Imaging, as used for microelectronic
failure analysis, comprises three techniques that utilize ultrasonic imaging technology.
The three acoustic microscopy technologies are the Scanning Laser Acoustic Microscope
(SLAM), the Scanning Acoustic Microscope (SAM), and the C-Mode Scanning Acoustic
Microscope (C-SAM). All three techniques are non destructive.
The Scanning Laser Acoustic Microscope (SLAM)
is a transmission microscope system that uses a transducer to generate an ultrasonic wave
and a laser beam to detect the transmitted waves. Because ultrasonic waves attenuate
rapidly in air, a liquid medium such as water or Flourinert™ is used to provide better
acoustic transmission characteristics. As the acoustic waves propagate through the sample,
voids and dissimilar interfaces will reduce the transmission of the acoustic waves. The
laser detector on the back side of the sample then detects the waves as they hit the
interface between the coverslip and the fluid couplant. Voids, dissimilar interfaces and
other defects that alter the transmission properties of the acoustic waves can then be
detected as regions of poor acoustic transmission. This signal can be turned into a map by
scanning the laser detector.
The Scanning Acoustic Microscope (SAM) is a
surface measurement system that looks at the reflected image from the surface of the
sample. At very high acoustic frequencies, one can obtain an image of the surface to be
examined similar to that of an optical image. This technique can also yield information
about the elastic properties of the material. A complete image can be produced in
approximately 10 seconds. This technique is not widely used for IC failure analysis.
The C-Mode Scanning Acoustic Microscope
(C-SAM) is an instrument that examines reflected waves from interfaces external and
internal to the sample. The transducer acts as both the sender and receiver (see Figure
2). A very short acoustic pulse enters the sample and return echoes are produced at
specific interfaces within the part (see Figure 3 .) The return times are a function of
the distance from the interface from the interface to the transducer. This can be seen in
Figure4, where an echo from the initial interface shows up as a signal after about 0.5
microseconds, and the echo from an internal delamination shows up as a signal about 1
microsecond later
|
|
 |
|
|
 |
|
|
 |
An electronic gate can then be set to examine
a specific period of time corresponding to echoes from a particular interface. The
transducer can then be scanned to produce a map of the interface. This is the C-Mode of
operation. This data contains information as to the quality of the interface within the
package. This mode is most useful for examining delaminations, voids, and cracks, as well
as die attach quality, in packaged ICs.
The lateral resolution (d) of a spherically
focused probe for sub-surface inspection can be estimated by
Here, l is the acoustic wavelength, F is the
focal length, and D is the diameter of the lens. Typically the effective F/D ranges from 1
to 1.6 for plastic package inspection. At a center frequency of 20 MHz the wavelength is
approximately 170 mm and the calculated resolution is 270 mm. This is an ideal resolution
that does not account for attenuation. The observed resolution is roughly 400mm at a depth
of 1.6 mm (or 3.2 mm round trip) in a standard PLCC package. Mold compound attenuation
varies considerably from one formulation to the next and this dramatically affects the
lateral resolution. Both the penetration and resolution are noticeably degraded by
temperature shock damage to the molding compound.
The acoustic attenuation in mold compound has
been reported to be 40 dB/cm at 15 MHz and to increase rapidly with increasing frequency.
The package preferentially attenuates the higher frequencies. As a result, acoustic
inspections of standard PLCC packages are limited to roughly 20 MHz. However, thin plastic
packages (mm) are often inspected at frequencies up to 90 MHz.
Attenuation also affects the depth resolution,
especially in plastic package inspection. The decay period for a broad-band echo pulse is
roughly 1.5l which creates a "dead zone" after each reflection in which later
reflections cannot be temporarily resolved without the use of frequency domain analysis
methods. This dead zone becomes thinner as the center frequency is increased.
In ceramic packages, the attenuation is much
less (<.05 dB/cm), but the speed of sound is almost three times greater than that of
mold compound. Hence, one wavelength at 50 MHz is Al2O3 is 200 mm and roughly equivalent
to one wavelength in mold compound at 20 MHz. A lateral resolution approaching one
wavelength is observed at a center frequency of 50 MHz and a depth of 1.6 mm in Al2O3.
The sensitivity of reflection acoustic imaging
is superior to the lateral resolution of the technique. For example, 25 mm (1 mil) bond
wires are often seen in 20 MHz C-SAM at 1.6 mm in plastic package inspections in which the
acoustic wavelength is 170 mm. A weakly reflecting object with lateral dimensions much
larger than a wavelength that is easily detected in reflection may produce only a
negligibly small contrast effect in transmission. However, the apparent size of a small,
non-specular reflector in the image will be determined by the lateral resolution
available.
The sensitivity of reflection mode acoustic
imaging to thin cracks with lateral dimensions much larger than a wavelength is an
important aspect of C-SAM inspection. Sound is a matter wave and depends on molecular
vibration for propagation. Theory predicts that crack openings greater than the particle
displacement amplitude produced by the interrogating sound wave should be detectable.
Experiments with steels, for example, indicate that air-filled cracks on the order of 10
nm are detectable.
Other than acoustic impedance, the major
variable of acoustic analysis is frequency. At higher frequencies sensitivity in detecting
small cavities is increased. Unfortunately, the maximum depth of useful information is
reduced at higher frequencies. Normally a compromise is needed to optimize sensitivity and
sound beam penetration. For silicon IC examinations frequencies of 30 to 50 MHz are used.
This yields a maximum theoretical sensitivity of 20 microns, therefore, the depth of
cavities and flaws cannot be located more accurately than 20 microns. The x-y resolution
of acoustic microscopy depends upon a number of variable including frequency, sample
absorption and scattering, and the x-y scanning step size of the transducer.
Figure 5 shows a table of acoustic parameters
for differing types of packaging materials.
|
|
Material |
velocity
(m/sec) |
density
(g/cc) |
Impedance
(105g/cm2sec) |
Al2O3 |
10400 |
3.8 |
40 |
Cu |
4400 |
8.9 |
39 |
Si |
8430 |
2.4 |
20 |
Mold Compound |
~3500 |
1.8 |
6.3 |
Water |
1480 |
1.0 |
1.5 |
Air |
343 |
1.2x10-3 |
4.1x10-4 |
Figure 6 shows the acoustic reflectivity
versus impedance for both ceramic and acoustic packages
Why Perform
Acoustic Microscopy?
Acoustic Microscopy is a non-destructive test
for locating delaminations, voids and cracks in packaged ICs, as well as examining die
attach. Acoustic Microscopy can yield this information without the need for delidding the
IC. Acoustic Imaging can also image voids and delaminations much easier than x-ray
imaging.
How is
Acoustic Microscopy Performed?
Acoustic Microscopy is performed using an
ultrasonic piezoelectric transducer. The transducer sends out signals and the reflected
signal is sensed with the piezoelectric material. Several manufacturers make commercial
instruments for acoustic imaging of integrated circuits. The two biggest manufacturers of
acoustic imaging equipment for integrated circuits are Sonoscan and Sonix. Both companies
offer turn-key systems with easy to use software interfaces.
When is
Acoustic Microscopy Performed?
Acoustic Microscopy is normally performed
before delidding the IC, however, if you need better spatial resolution while examining
for voids or delaminations in the die attach, you can remove the lid and then examine the
IC..
Photographs
Photo
1 - Graph showing a C-mode time window with a good IC and a delaminated IC (the
delaminated IC signal is dashed) after T. Moore.
Photo
2 - Photograph showing delaminations at each of the corners on a PLCC package. after
T. Moore.
Photo
3 - Photograph showing extensive delamination in the lead frame on a PLCC package.
after T. Moore.
Photo
4 - Photograph showing a difference in delaminations between two PLCC packages. after
T. Moore.
Photo
5 - Photograph showing good adhesion (on the left) and bad adhesion (on the right) for
a TO-220 package before and after thermal cycling. after T. Moore.
Photo
6 - Screen shot of C-mode scanning acoustic microscope. (photo courtesy Sonix Corp.)
Photo
7 - Screen shot of C-mode scanning acoustic microscope. (photo courtesy Sonix Corp.)
Photo
8 - Screen shot of C-mode scanning acoustic microscope. (photo courtesy Sonix Corp.)
Photo
9 - Screen shot of C-mode scanning acoustic microscope. (photo courtesy Sonix Corp.)
Photo
10 - Screen shot of C-mode scanning acoustic microscope. (photo courtesy Sonix Corp.)
References on Acoustic Microscopy
[1] R. A. Lemmons and C. F. Quate, Applied
Physics Letters., Vol. 25, 1974, pp. 251-253.
[2] R. A. Lemmons and C. F. Quate, Proc.
1973 IEEE Ultrasonic Symposium, 1973, pp. 18-21.
[3] L. W. Kessler, Proc. IEEE, Vol.
67, 1979, pp. 526-536.
[4] L. A. Kinsler, et al., Fundamentals of
Acoustics, John Wiley and Sons, 1982, p. 125
[5] Ultrasonic Testing, J. Szilard
ed., John Wiley and Sons, 1982, pp. 1-23.
[6] A. J. Miller, Journal of Acoustic
Imaging, Vol. 12., 1982, pp. 67-78
[7] J. A. Michael and W. W. Fultz,
"Non-Destructive Evaluation by Acoustic Microscopy as Applied to the Failure Analysis
of Microelectronics," Proc. Int. Symp. Testing and Failure Analysis, 1985,
pp. 181-186.
[8] N. J. Burton and D. M. Thacker,
"Practical Applications of Acoustic Microscopy in Failure Analysis," Proc.
Int. Symp. Testing and Failure Analysis, 1985, pp. 187-192.
[9] S. Okikawa et al., "Stress Analysis
for Passivation and Interlevel Insulation Film Cracks in Multilayer Aluminum Structures
for Plastic-Packaged LSI," Proc. Int. Symp. Testing and Failure Analysis,
1987, pp. 75-81
[10] K. Sharai, K. Kobayashi, T. Noguchi and
T. Goka, "Non-Destructive Inspection of Voids in Silver Hard Solder Layer of Power
Transistor," Proc. Int. Symp. Testing and Failure Analysis, 1988, pp. 47-52.
[11] M. J. Mirasole, "Characterization of
Die Attach Integrity Using Destructive and Nondestructive Techniques," Proc. Int.
Symp. Testing and Failure Analysis, 1988, pp. 77-88.
[12] J. E. Semmens and L. W. Kessler,
"Nondestructive Evaluation of Thermally Shocked Plastic Integrated Circuit Packages
Using Acoustic Microscopy," Proc. Int. Symp. Testing and Failure Analysis,
1988, pp. 211-215.
[13] T. M. Moore, "Identification of
Package Defects in Plastic-Packaged Surface-Mount ICs by Scanning Acoustic
Microscopy," Proc. Int. Symp. Testing and Failure Analysis, 1989, pp. 61-67.
[14] A. van der Wijk and K. van Doorselaer,
"Non-Destructive Failure Analysis of ICs Using Scanning Acoustic Tomography (SCAT)
and High-Resolution X-ray Microscopy (HRXM)," Proc. Int Symp. Testing and Failure
Analysis, 1989, pp. 69-74.
[15] T. M. Moore, R. McKenna, S. J. Kelsall,
"The Application of Scanning Acoustic Microscopy to Control Moisture/Thermal-Induced
Package Defects," Proc. Int. Symp. Testing and Failure Analysis, 1990, pp.
251-258.
[16] L. W. Kessler and S. R. Martell,
"Acoustic Microscopy Technology (AMT) Analysis of Advanced Materials for Internal
Defects and Discontinuities," Proc. Int. Symp. Testing and Failure Analysis,
1990, pp. 491-504.
[17] T. M. Moore, R. McKenna, S. J. Kelsall,
"Correlation of Surface Mount Plastic Package Reliability Testing to Nondestructive
Inspection by Scanning Acoustic Microscopy," Proc. Int. Reliability Physics Symp.,
1991, pp. 160-166.
[18] T. M. Moore, "Reliable Delamination
Detection by Polarity Analysis of Reflected Acoustic Pulses, Proc. Int. Symp. Testing
and Failure Analysis, 1991, pp. 49-54.
[19] T. M. Moore, "Inspecting IC Packages
with C-Mode Acoustic Microscopy," Microelectronic Failure Analysis Desk
Reference, 3rd Edition, T. W. Lee. and S. V. Pabbisetty eds., 1993, pp.
41-50.
