19 / 62
A D V A N C E D
M A T E R I A L S
&
P R O C E S S E S | F E B R U A R Y
2 0 1 6
1 9
DISCUSSION AND SUMMARY
In describing this work, Andreas
Graff of Fraunhofer CAM explains, “To
our knowledge, no one has previously
established a routine to prepare OLED
samples for S/TEM by FIB, and no one
has seen the composition in the crit-
ical layers in a real device before and
confirmed their dimensions by S/TEM.
Descriptions in the literature led us to
expect layer dimensions greater than
30 nm, yet we measured thicknesses in
the 20 nm range. We were able not only
to prepare the samples and measure
the layer thicknesses, but also to mea-
sure elemental distributions within the
layers. The fluorine distribution is nota-
bly different than we expected, perhaps
due to an aging effect that occurred
during the observations. Our client was
quite pleased with the results and fully
expects to incorporate this type of anal-
ysis in future development work.”
In a recent interview, Frank Alt-
man, Fraunhofer CAM’s group leader
for semiconductor diagnostics, empha-
sized the challenges posed by the rap-
idly increasing complexity of microelec-
tronic devices. “Our main challenge is
complexity, especially with the advent
of 3D integration and advanced pack-
aging technologies. Typically, we look
at electrically defective samples and
are tasked with discovering the physi-
cal root cause. Often we have only one
chance to analyze a defect, so it is very
important to make the right choices in
preparing the sample and executing the
analysis,” says Altman.
“First, we have to locate the de-
fect and expose it, then perform the
structural and material analysis. It can
be very challenging just to get access
to structures buried deeply in a chip,
or in multichip packages with through-
silicon vias, or TSVs. We have to remove
lots of material to gain access, then
find a specific TSV, microbump, mem-
ory cell, or transistor. The challenge is
further increased by the range of mate-
rials and scales we have to cover, from
centimeters to nanometers. We have
access to numerous preparation tools
for grinding, broad beam ion milling,
FIB, and delayering, but we first have to
decide which tools to use. For site spe-
cific preparation, FIB is always our first
choice. The availability of gas chemis-
tries for selective milling is very pow-
erful in deprocessing a circuit. From
the electron microscopy point of view,
polymers, organics, and other soft ma-
terials are a big challenge, and they are
becoming more common in microelec-
tronic devices. They are easily damaged
and they do not generate much con-
trast. The OLED display analysis is an
excellent example. We had to analyze
a stack of different amorphous layers
that are very difficult to differentiate,
with minimal contrast, and very easy to
damage or destroy. In this case, the Ti-
tan TEM’s low voltage performance and
the SuperX EDS detector were absolute-
ly essential to our success,” he adds.
~AM&P
For more information:
Frank Altmann
is head of diagnostic of semiconductor
technologies group at Center for Ap-
plied Microstructure Diagnostics (CAM)
of the Fraunhofer Institute for Micro-
structure of Materials and Systems
IMWS, Walter-Huelse-Strasse 1, 06120
Halle, Germany,
frank.altmann@imws. fraunhofer.de,
www.cam.fraunhofer.de.
LECO_AMH55_LE-150.ps
T:\MISC\ADS\LE-150\LE-150i.cdr
Friday, January 15, 2016 1:22:52 PM
Color profile: Disabled
Composite Default screen