Radiant energy – Inspection of solids or liquids by charged particles – Methods
Reexamination Certificate
1999-06-22
2002-03-26
Anderson, Bruce (Department: 2881)
Radiant energy
Inspection of solids or liquids by charged particles
Methods
C250S309000, C250S310000, C250S492210, C250S492200, C438S690000, C438S691000
Reexamination Certificate
active
06362475
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention is directed, in general, to analysis of monolithic structures and, more specifically, to a scanning electron microscope/energy dispersive spectroscopy (SEM/EDS) sample preparation method and a monolithic sample produced by way of the method.
BACKGROUND OF THE INVENTION
In semiconductor processing today it is often necessary to spectroscopically examine portions of a semiconductor die to determine the results of new or conventional processes. The examination may be to confirm the results of an experimental process, or even to determine the nature of a particular failure or defect in a semiconductor device. Of course, because of the nature of integrated circuits, the examination must often be performed on samples cut from the die in question. Scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS) is frequently used in the determination of the composition of target material in a feature of a semiconductor die. In SEM/EDS, X-rays generated by the primary electron beam's interaction with the target assist in determining the composition of the target.
Referring initially to
FIG. 1
, illustrated is a cross-sectional view of a conventional SEM sample symbolically showing effects of the SEM process. A primary electron beam
100
with width
180
is directed to a sample
110
having a thickness
111
. The sample
110
may be cut from any orientation within a material of interest. In this embodiment, the beam
100
has a width
180
and a tear shaped interaction volume
120
, with depth
160
. For example, at a given beam energy and in a given material, a beam width
180
of 10 nm can produce an interaction volume with a maximum width
170
of 1 &mgr;m. The beam
100
interacts with essentially all of the material in the interaction volume
120
, producing secondary electrons
130
, backscattered electrons
140
, and X-rays
150
. The specific X-rays present in the spectra are traceable by EDS to particular elements present at the site of the SEM examination.
As device sizes are made ever smaller, a defect that was insignificant in the past becomes quite significant. However, as the semiconductor features and defects become smaller, it has become increasingly difficult to obtain the lateral resolution needed from EDS for compositional identification of small defects and thin layers. With a decrease in feature size, the feature becomes a smaller percentage of the target, thereby reducing the X-ray emanations from the feature. It is well known that the interaction volume
120
depth
160
and width
170
are dependent upon the accelerating energy of the primary electron beam
100
. However, because the beam
100
interacts with the entire interaction volume
120
, precise data on the nature of a feature of interest
160
at the examination site may be unclear. In order to reduce the interaction volume, the primary beam accelerating voltage has been reduced to account for the reduced feature size. However, several elements commonly used in semiconductor fabrication have overlapping peaks at lower beam energies, and the ability to distinguish between the elements is therefore lost at these low beam energies. At higher beam voltages more peaks in the EDS spectra allow for better discrimination between the elements present. Therefore, reducing beam energy does not effectively improve element discrimination. Of course, tool manufacturers have worked to improve the sensitivity of their tools. Unfortunately, these efforts have met with limited success, and the available sensitivity is still not sufficient for current and projected needs.
Referring now to
FIG. 2A
, illustrated are EDS spectra of a conventional semiconductor SEM sample taken with an incident electron beam of 5 keV. As can be seen, the results make it difficult to resolve between oxygen
210
and titanium
220
. There also appears to be copper
230
present, as well as aluminum
240
, silicon
250
and phosphorous
260
. Referring now to
FIG. 2B
, illustrated are EDS spectra of the same area as shown in
FIG. 2A
with an incident electron beam of 10 keV. Comparing
FIG. 2B
to
FIG. 2A
, it is clear that oxygen
211
, titanium
221
, aluminum
241
, silicon
251
and phosphorus
261
are all present. It is also now evident that copper is not present. It is readily apparent that the definition of the peaks
211
,
221
,
241
,
251
,
261
is much better at the higher (10 keV) beam energy.
Referring now to
FIG. 3A
, illustrated are EDS spectra taken at 5 keV of a different semiconductor SEM sample. At this voltage, titanium
310
and possibly silicon
320
appear to be present. However, one who is skilled in the art will readily observe because of the relative magnitudes of the peaks at X-ray energy levels of 0.6 keV and above that it is extremely difficult to discern the presence of any particular element. Referring now to
FIG. 3B
, illustrated are EDS spectra taken at 10 keV of the same area as shown in FIG.
3
A. One who is skilled in the art will readily appreciate that the spectra of
FIG. 3B
clearly allows one to differentiate peaks associated with titanium
311
, silicon
321
, and tungsten
331
. Comparing
FIG. 3B
to
FIG. 3A
, one can readily understand that the presence of tungsten is simply not recognizable from the spectra of FIG.
3
A. Thus higher beam energies make identification of elements much easier.
Accordingly, what is needed in the art is a sample preparation method that minimizes the interaction volume while retaining the feature of interest.
SUMMARY OF THE INVENTION
To address the above-discussed deficiencies of the prior art, the present invention provides a method of preparing a monolithic structure for scanning electron microscope/energy dispersive spectroscopy (SEM/EDS) and a sample produced by way of the method. In one embodiment, the method includes: (1) aiming a focused ion beam at any location beneath an area of interest of any size or depth in the monolithic structure and (2) employing the focused ion beam to remove at least a portion of an interaction volume of material beneath the area of interest. The area of interest preferably remains substantially intact for the subsequent spectroscopy.
The present invention therefore introduces the broad concept of employing a focused ion beam to remove at least some of the interaction volume beneath an area of interest that would otherwise generate unwanted background noise during subsequent SEM/EDS analysis. For purposes of the present invention, a “monolithic structure” is defined as a structure formed in or on a body of material. A monolithic structure may be a semiconductor wafer (such as a silicon or gallium arsenide wafer), but can be composed of any material. For purposes of the present invention, an “area of interest” is defined as any portion of the monolithic structure that is to be the subject of analysis.
In one embodiment of the present invention, the area of interest is a portion of a surface of the monolithic structure. Alternatively, the area of interest could be located above or below the surface of the monolithic structure.
A preferred embodiment of the present invention comprises employing the focused ion beam to remove an entirety of the interaction volume. Of course, removal of only a portion of the interaction volume may be advantageous in some applications.
In one embodiment of the present invention, the employing comprises employing the focused ion beam to remove at least the portion of the interaction volume such that the area of interest becomes membranous. The membranous area of interest may, but need not, have a thickness comparable to samples employed in conventional transmission electron microscope (TEM)/EDS analysis.
In one embodiment of the present invention, the monolithic structure comprises a silicon wafer. Alternatively, the monolithic structure could be a gallium arsenide wafer or a wafer or body of another suitable composition.
In one embodiment of the present invention, the monolithic structure comprises at least one layer located on a sili
Bindell Jeffrey B.
Stevie Frederick A.
Vartuli Catherine
Agere Systems Guardian Corp.
Anderson Bruce
Wells Nikita
LandOfFree
Scanning electron microscope/energy dispersive spectroscopy... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Scanning electron microscope/energy dispersive spectroscopy..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Scanning electron microscope/energy dispersive spectroscopy... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2854848