Yao N. Focused Ion Beam Systems: Basics and Applications
Подождите немного. Документ загружается.
planar/cross-sectional view samples, (iv) multiple sampling from an adjacent
area is possible, (v) the method is applicable for magnetic materials, (vi) high
reliability of the sampling. The advantage (v) results from a significant
decrease in the TEM specimen volume. A magnetic power attracts the
magnetic specimen in the strong magnetic TEM-lens field and may pull it
away from the TEM grid. The smaller the specimen, the less the attracting
power. In addition, disturbance of the magnetic lens field is also smaller. FIB-
AD welding the m-sample to the TEM grid greatly serves to secure a dis-
sipation path for beam heating during the TEM/STEM inspection. The
several mm-sizing is also of great use for radioactive samples because their
radiation activities are reduced in proportion with the volume to the level of
easy handling [75].
Recently, a pillar m-sampling has been reported for three-dimensional (3D)
STEM observation of the same point of interest [74,75]. The FIB/STEM side-
entry holder has been specially designed to give the pillar sample a tilt (fi)of
±20
and a rotation (fl)of360
(see Figure 4.17)[75]. The conventional-size
m-sample is FIB milled, transferred to the needle stage edge using the
manipulated needle, and fixed to it using FIB-AD. The m-sample is FIB milled
to be a 2–5 mm-square pillar that contains the inspection point of interest.
Here, the pillar sample is STEM observable at any rotation angle. We can
Pillar sample
Top of the
needle stage
b : rotation
a : tilt
5
HD-2000 200kv ×1.5k SE
20.0m
Figure 4.17 FIB/STEM side-entry holder for 3D analysis (Hitachi High-
Technologies Corporation) [75]. The m-sample is mounted on the needle
stage end using the m-sampling technique and is FIB fabricated as a pillar of
a few mm squares. The holder allows to the pillar sample to tilt (fi) by ±20
and rotate (fl) by 360
.
Focused ion beam systems118
reconstruct computerized tomography (CT) images using the STEM images at
all rotation angles. Since nothing surrounds the inspection region except along
the direction of the pillar base, extreme reduction has been achieved in system
peak intensities, i.e., X-rays excited from the surroundings by high-energy stray
electron bombardment, in energy dispersive X-ray (EDX) analysis.
Another similar technique to the m-sampling is an ‘‘ex-situ lift-out,’’ where
the FIB milled TEM lamella is lift out by the glass needle using an optical
microscope under the atmosphere pressure [3,76]. For TEM analysis, the ex-
situ lift-out specimen of about 0.1 mm in thickness is usually positioned onto
a carbon coated TEM grid. Then, the ‘‘ex-situ lift-out’’ is in general faster
than the ‘‘m-sampling,’’ but does not allow to FIB rethin the sample.
The FIB works as micro ‘‘cutter’’ and ‘‘attacher’’ in micro-fabrication.
Figure 4.18 shows a typical example, i.e., a micro universal joint. Each of the
micro parts has been FIB milled from 25 mm-diameter aluminum wire and
assembled with them using the manipulated FIB micro ‘‘cutter’’ and ‘‘atta-
cher.’’ This is the very FIB manufacturing system. Both by extracting
m-devices having electrical/mechanical functions and by transplanting them
at any desired positions, we will manufacture new hybrid systems, which are
difficult to fabricate in the same process. Device transplantation is a task with
more challenge in the future line [70].
Another interesting question is ‘‘how smooth and flat a sample can
FIB fabricate?’’ TEM 0.3–0.4 mm-thick and very flat specimens of less than
several nm in roughness over 10 mm square in size have been successfully FIB
milled to observe structures of vortices in high-Tc superconductor material
(YBaBiCaCuO) by Lorentz microscopy [77].
10 µm
Figure 4.18 Micro universal-jo int. Each of the parts is FIB fabricated from a
25 mm diameter aluminum wire and is sequentially assembled with each
other using both the manipulated needle and the FIB ‘‘cutter’’ and
‘‘attacher’’ (Hitachi FIB FB-2100 with the micro-sampling system).
(Courtesy of T. Tanaka, Osaka Sangyou University.)
Imaging using electrons and ion beams 119
4.4.2 SEM/STEM/TEM microscopy of FIB cross-sectioned specimens
A focused ion beam (FIB) has been actively applied for preparation of about
0.1 mm-thick specimens for transmission electron microscopes (TEMs). For
device failure analyses, however, it is difficult to prepare the exact-point TEM
specimens. The reason is that the failures are usually beneath the surface and
their exact locations are unknown. Then, even step-by-step FIB cross-
sectioning may sputter away the failures in the TEM specimen preparation.
One of the solutions for exact-point TEM microscopy is high-voltage (HV)
SEM imaging in TEM [78].
A TEM has an HV-SEM imaging mode in addition to the conventional
transmission imaging mode (e.g., Hitachi 200 kV TEM system, HF-2000). In
the HV-SEM imaging mode, SEs emitted from the specimen travel upward
through the objective lens bore under the influence of a combination of the
forces from the magnetic field of the lens and the electrostatic field of the SE
detector ( þ10 kV) and are collected to form the SEM images. Figure 4.19
shows an electron–sample interaction when a 200 kV electron beam illumi-
nates the sample. Devices such as 4M and 16M dynamic random access
memory (DRAM) used in the present study have transistors and capacitors
that are composed of Al and poly-Si surrounded by silicon oxide (SiO
2
),
respectively. When the 200 kV primary electron beam strikes the region of Al
or Si, BSEs are more generated in their regions than their surroundings. The
reason is that both Al and poly-Si are larger in Z
2
than SiO
2
. The SEs are
excited not only by the primary electrons, but also by BSEs. The SEs gen-
erated deep in the sample, however, are absorbed within the sample because
their energies are so small. On the other hand, the BSEs have high energies so
that they can travel through 1 mm thick silicon oxide and emerge from the
surface. The SEs excited in the surface area of the sample by the BSE and
emitted from it reveal the planar position of each BSE’s birthplace in the SIM
Specimen
SE detector
Backscattered
electrons
HV primary
electrons
SiO
2
Al, Si
SE
SE
Figure 4.19 Generation of secondary electrons (SE) and backscattered
electrons using an HV electron beam [78].
Focused ion beam systems120
image. Thus, the HV-SEM imaging provides information on sample surfaces
as well as inner structures to a depth of about 1 mm. Figure 4.20 shows a
typical comparison of 30 kV-Ga SIM and 200 kV HV-SEM images. The
information depth for SIM imaging is only a few nm, as mentioned before.
30 kV- SIM image
200 kV- SEM image
1m
1m
Figure 4.20 Typical comparison of 30 kV Ga-SIM and 200 kV HV-SEM
images of an FIB milled cross section.
1
m
m
1
m
m
1
m
m
1
m
m
(a)
(b)
(c)
(d)
Figure 4.21 Series of HV SEM images taken every 0.1–0.2 mm in FIB milling
of a 16M-DRAM specimen (Hitachi 20 0 kV TEM; HF-2000) [78].
Imaging using electrons and ion beams 121
Figure 4.21 demonstrates a series of HV-SEM images of the 16M-DRAM
device, which have been taken at every 0.1–0.2 m m step in FIB milling. The
figure shows that the device structures around the contact change in three
dimensions (3D) as milling proceeds. The depth of field is about 1 m m. The
image resolution gradually worsens with the depth and reaches about 0.1 m m
at a depth of about 1 mm. Fine-step FIB cross-sectioning is repeatedly carried
out until the specified region of interest surely appears in both front- and
back-side HV-SEM images of the TEM specimen (of about 0.1 mm or less in
thickness). Electron energy loss spectroscopy (EELS) of transmitted electrons
in the STEM system allows providing of the elemental and chemical infor-
mation in the site-specific region and has tentatively been applied to the FIB
prepared specimen [78].
Recently, the FE-STEM of low voltage (5–30 kV) also has been applied to
characterize the device features with ease [79,80]. A standard in-lens FE-SEM
(e.g., Hitachi S-5000) was modified by the addition of a STEM detector and a
bright field (BF) aperture. Using a ‘‘m-sampling’’ technique in an FIB system
(see Section 4.3.1), a thin specimen including the specific area of interest was
(a) 110 nm
(b) 110 nm
1 µm
Figure 4.22 (a) STEM view of an FIB milled SRAM and (b) the highly
magnified view of metal line showing grain structure. The arrow indicates a
barrier metal [79].
Focused ion beam systems122
dug out, lifted, transferred from the wafer, and fixed onto a TEM grid
mounted on a side-entry specimen stage. The m -sampling technique using the
compatible FIB-SEM specimen holder permits SEM/STEM inspection and
FIB milling repeatedly, so that the precise area of interest is easily prepared
for STEM observation.
The electron straightforward penetration at accelerating voltage V ¼ 5–30 kV
is rather short in contrast to that at V ¼ 200–300 kV. So, an interesting point
is ‘‘How fine are the device features that are observable in the low voltage
FE-STEM?’’ Figures 4.22(a) and (b) show the STEM views of the FIB milled
static random access memory (SRAM) [79] and its highly magnified view of
the metal lines, respectively. It has been found that the low voltage FE-
STEM is applicable to obtain rather fine structure STEM images for
0.1 mm-thick specimens.
References
[1] K. Nikawa. J. Vac. Sci. Technol. B, 9 (1991), 2566–77.
[2] T. Ishitani and T. Yaguchi. Microsc. Res. Tech., 35 (1996), 1320–33.
[3] L. A. Giannuzzi and F. A. Stevie. Micron., 30 (1999), 197–204.
[4] T. Ishitani and K. Ohya. Scanning, 25 (2003), 201–9.
[5] T. Ishitani, H. Hirose and H. Tsuboi. J. Electron Microsc., 44 (1995), 110–14.
[6] M. Utlaut. Handbook of Charged Particle Optics, ed. J. Orloff (New York: CRC
Press, 1997), pp. 429–88.
[7] Y. Sakai, T. Yamada, T. Suzuki et al. Appl. Phys. Lett., 73 (1998), 611–13.
[8] Y. Sakai, T. Yamada, T. Suzuki and T. Ichinokawa. Appl. Surf. Sci., 73 (1999),
96–100.
[9] T. Ishitani and H. Tsuboi. Scanning , 19 (1997), 489–97.
[10] J. Orloff, M. Utlaut and L. Swanson. High Resolution Focused Ion beams: FIB
and its Applications (New York: Kluwer Academic/Plenum Publishers, 2003).
[11] J. I. Goldstein, D. E. Newbury, P. Echlin, et al. Scanning Electron Microscopy
and X-ray Microanalysis (New York and London: Plenum Press, 1992).
[12] L. Reimer. Scanning Electron Microscopy, 2nd edn (Berlin: Springer, 1998).
[13] D. C. Joy. Inst. Phys. Conf. Ser. No. 93, 1 (1988), 23–32.
[14] T. Ishitani. Jpn. J. Appl. Phys., 34 (1995), 3303–6.
[15] L. Reimer. Image Formation in Low-Voltage Scanning Electron Microscopy
(Washington, DC: SPIE Optical Engineering Press, 1993).
[16] G. Dearnaley, J. H. Freeman, R. S. Nelson and J. Stephen. Ion Implantation
(Amsterdam: North-Holland Publishers, 1973).
[17] H. Ryssel and I. Ruge. Ion Implantation (New York: John Wiley & Sons, 1986).
[18] Z. J. Ding and R. Shimizu. Scanning, 18 (1996), 92–113.
[19] W. O. Hofer. Scanning Microsc. Suppl., 4 (1987), 265–310.
[20] D. Hasselkamp, H. Rothard, K.-O. Groeneveld et al. Particle Induced Electron
Emission II (Berlin: Springer-Verlag, 1992).
[21] E. S. Parilis, L. M. Kishinevsky, N. Yu. Turaev et al. Atomic Collisions on Solid
Surfaces (Amsterdam: North-Holland, 1993), pp. 391–473.
[22] R. A. Baragiola, ed. Ionizat ion of Solids by Heavy Particles (New York: Plenum
Press, 1993).
Imaging using electrons and ion beams 123
[23] R. A. Baragiola. Low Energy Ion-Surface Interactions, ed. J. W. Rabalais
(Chichester: John Wiley and Sons, 1994), pp. 187–262.
[24] J. Schou. Physical Processes of the Interaction of Fusion Plasmas with Solids,ed.
W. O. Hofer and J. Roth (San Diego: Academic Press, 1996), pp. 177–216.
[25] H. Kudo. Ion-Induc ed Electron Emission from Crystalline Solids (Berlin: Springer,
2002).
[26] H. D. Hagstrum. Phys. Rev., 96 (1954), 325–35.
[27] H. D. Hagstrum. Phys. Rev., 96 (1954), 336–65.
[28] R. A. Baragiola, E. V. Alonso, J. Ferron and A. Oliva-Florio. Surf. Sci. 90, 240–
55.
[29] E. J. Sternglass. Phys. Rev., 108 (1957), 1–12.
[30] J. Schou. Phys. Rev. B, 22 (1980), 2114–74.
[31] M. Ro
¨
sler. Ionization of Solids by Heavy Particles, ed. R. A. Baragiola (New
York: Plenum Press, 1993), pp. 27–58.
[32] P. M. Eche nique, F. Flores and R. H. Ritchie. Solid State Physics: Advances in
Research and Ap plications, Vol. 43, ed. H. Ehrenrech and D. Turnbull (New
York: Academic Press, 1990), pp. 229–308.
[33] I. Nagy, A. Arnau and P. M. Echenique. Phys. Rev. B, 38 (1988), 9191–3.
[34] K. Ohya and T. Ishitani. J. Electron Microsc., 52 (2003), 291–8.
[35] D. R. Penn. Phys. Rev. B, 35 (1987), 482–6.
[36] J. C. Ashley. J. Electron Spectrosc. Relat. Phenom., 50 (1990), 323–34.
[37] K. Ohya and J. Kawata. Jpn. J. Appl. Phys., 32 (1993) , 1244–7.
[38] H.-J. Fitting and J. Reinhardt. Phys. Status Solidi A, 88 (1985), 245–59.
[39] R. A. Baragiola. Interaction of Charged Particles with Solids and Surfaces,ed.
A. Gras-Marti, H. M. Urbassek, N. R. Arista and F. Flores (New York: Plenum
Press, 1990), pp. 443–58.
[40] G. Lakits, F. Aumayr, M. Heim and H. Winter. Phys. Rev. A, 42 (1990), 5780–3.
[41] J. Lo
¨
rincik, Z. Sroubek, H. Eder, F. Aumayr and H. Winter. Phys. Rev. B, 62
(2000), 16116–25.
[42] K. Ohya and T. Ishitani. Surf. Coat. Technol., 158–159 (2002), 8–13.
[43] W. Eckstein. Computer Simulation of Ion–Solid Interactions (Berlin: Springer-
Verlag, 1991), pp. 4–32.
[44] J. F. Ziegler, J. P. Biersack and U. Littmark. The Stopping and Ranges of Ions in
Solids (New York: Pergamon Press, 1985), pp. 109–40.
[45] T. Koshikawa and R. Shimizu. J. Phys. D: Appl. Phys., 7 (1974), 1303–15.
[46] J. P. Ganachaud and M. Cailler. Surf. Sci., 83 (1979), 498–518.
[47] Z. J. Ding, and R. Shimizu. Surf. Sci., 197 (1988), 539–54.
[48] M. Kotera, R. Ijichi, T. Fujiwara, H. Suga and D. B. Wittry. Jpn. J. Appl. Phys.,
29 (1990), 2277–82.
[49] A. Dubus, J.-C. Dehaes, J. P. Ganachaud, A. Hafni and M. Caill er. Phys. Rev.
B, 47 (1993), 11056–73 .
[50] J.-Ch. Kuhr and H.-J. Fitting. Phys. Stat. Sol. A, 172 (1999), 433–48.
[51] K. Nishimura, J. Kawata and K. Ohya. Nucl. Instrum. and Meth. Phys. Res. B,
164–165 (2000), 903–9.
[52] T. K. Olson, R. G. Lee and J. C. Morgan. Proc. 18th Int. Symp. Testing and
Failure Analysis (1992), 373–380.
[53] M. W. Phaneuf. Micron, 30 (1999), 277–88.
[54] T. Ishitani, Y. Madokoro, M. Nakagawa and K. Ohya. J. Electron Microsc., 51
(2002), 207–13.
[55] K. Ohya and T. Ishitani. Nucl. Instrum. Meth. Phys. Res. B, 202 (2003), 305–11.
Focused ion beam systems124
[56] K. Ohya, F. Aumayr and H. Winter. Phys. Rev. B, 4 (1992), 3101–4.
[57] K. Ohya and J. Kawata. Scan. Microsc., 9 (1995), 331–53.
[58] M. Kudo, Y. Sakai and T. Ichinokawa. Appl. Phys. Lett., 76 (2000), 3475–7.
[59] K. Ohya. Proc. 3rd Int. Conf. Atomic Level Characterizations for New Materials
and Devices: Nara (2001), 53–6.
[60] K Ohya and T. Ishitani. J. Electron Microsc., 53 (2004), 229–35.
[61] K. Ohya. J. Phys. Soc. Jpn., 61 (1992), 3013–14.
[62] J. Kawata and K. Ohya. Radiat. Eff. Def. Sol., 130–131 (1994), 131–6.
[63] K. Ohya and J. Kawata. Nucl. Instrum. Meth. Phys. Res. B, 90 (1994), 552–5.
[64] P. D. Prewett and G. L. R. Mair. Focused Ion Beams From Liquid Metal Ion
Sources (Taunton, UK: Research Studies Press Ltd, 1991).
[65] B. A. Brusilovsky. Vacuum, 35 (1985), 595–615.
[66] Y. Yahiro, K. Kaneko, T. Fujita, W.-J. Moon and Z. Horita. J. Electron
Microsc., 53 (2004), 571–6.
[67] E. S. Mashkova, V. A. Molchanov and D. D. Odintsov. Doklad. Akad. Nauk
SSSR, 151 (1963), 1074–5; Engl. transl.: Sov. Phys. Doklad., 8 (1964), 806–7.
[68] J. Lindhard. Mat. Fys. Medd. Dan Vid. Selsk., 34 (1965).
[69] T. Ishitani and T. Ohnishi. J. Vac. Sci. Technol. A, 9 (1991), 3084–9.
[70] T. Ishitani, T. Ohnishi, Y. Madokoro and Y. Kawanami. J. Vac. Sci. Technol. B,
9 (1991), 2633–7.
[71] T. Ohnishi, H. Koike, T. Ishit ani et al. Proc. 25th Int. Symp. Testing and Failure
Analysis (1999), 449–53.
[72] T. Yaguchi, R. Urao, T. Kamino et al . Mater. Res. Soc. Symp. Proc., 636 (2001),
D9.35.1–D9.35.6.
[73] Y. Nemoto, Y. Miwa, M. Kikuchi et al. J. Nucl. Sci. Technol., 39 (2002),
996–1001.
[74] T. Yaguchi, M. Konno, T. Kamino et al . Microsc. Microanal., 9 (2003), 118–19.
[75] T. Yaguchi, Y. Kuroda, M. Konno et al. Hitachi Scientific Instrument News, 46
(2003), 18–20.
[76] L. A. Giannuzzi, B. W. Kempshall, S. D. Anderseon, B. I. Prenitzer and T. M.
Moore. Microelectronic Failure Analysis Desk Reference Supplement (2002)
29–35.
[77] A. Tonomura, H. Kasai, O. Kamimura et al. Phys. Rev. Lett., 88 (2002), 237001.
[78] T. Ishitani, Y. Taniguchi, S. Isakizawa et al. J. Vac. Sci. Technol. B, 16 (1998),
2532–7.
[79] M. Nakagawa, R. Dunne, K. Koike et al. J. Electron Microsc., 51 (2002), 53–7.
[80] M. V. Moore. Scanning, 25 (2003), 159–60.
Imaging using electrons and ion beams 125
5
Characterization methods using FIB/SEM
DualBeam instrumentation
steve reyntjens
FEI Company, The Netherlands
lucille a. giannuzzi
FEI Company, Hillsboro, OR
5.1 Introduction
In this chapter, we consider some of the characterization possibilities that are
typical and in some cases even unique to focused ion beam (FIB) and
DualBeam (FIB/scanning electron microscope (SEM)) instruments. It is
well known that FIB based instruments may be used to section and provide
analysis of the third dimension for subsequent SEM or FIB imaging. Fol-
lowing the preparation of any cross section, the steps to obtain serial sections
are relatively straightforward. Why not, after opening up the inside of the
bulk substrate by milling a trench and imaging a polished sidewall of that
trench, polish off a thin layer of material, and open up a second surface for
imaging? One can continue in the same fashion, and open up a third, a
fourth, a fifth, etc., surface, and in so doing create a series of serial sections
through a volume. In this manner, one may analyze a selected volume of the
sample, instead of ‘‘just’’ a single cross-sectional surface. This type of analysis
is ‘‘3D characterization,’’ and it constitutes the better part of this section. 3D
characterization is more than a series of SEM images; it can equally be
applied to FIB images [1], and even complete elemental maps (via energy
dispersive spectrometry (EDS)) [2] or 3D crystallographic data (via electron
backscatter diffraction (EBSD)) [3,4].
As an option to alternating between polishing and imaging as described
above (i.e., automated ‘‘slice and view’’), collecting images while implementing
simultaneous patterning and imaging (SPI mode) can be considered as a
Focused Ion Beam Systems: Basics and Applications, ed. N. Yao.
Published by Cambridge University Press. ª Cambridge University Press 2007.
126
shortcut to 3D SEM imaging. At present, this is limited to SEM imaging
while FIB polishing within a DualBeam instrument. As an aside, SPI mode
imaging is as much a ‘‘process monitoring’’ or ‘‘end point detection’’ tool
(helping the operator decide on the progress of the FIB machining) as a
characterization method.
Characterization by scanning transmission electron microscopy (STEM)
using the SEM is also possible. In many cases, the preparation method of
choice for STEM imaging is FIB or DualBeam. The preparation process and
characterization step may be combined very elegantly and conveniently in a
DualBeam microscope.
Although less common, FIB based analytical techniques exhibit some very
specific possibilities. Near the end of this section, we will briefly elaborate on
the merits and difficulties of techniques such as particle induced X-ray
emission (PIXE) and secondary ion mass spectroscopy (SIMS).
We conclude the section with some options and extensions to an FIB
or DualBeam microscope that can further enhance their characterization
capabilities. Electrical probing is possible when using in-chamber probes
connected to measurement instruments outside the chamber via electrical
feed-throughs. Similarly, micromanipulators in or on the chamber provide
means for the investigation of mechanical properties of microstructures.
5.2 3D characterization
With overall increasingly complex man-made or natural devices and objects,
the nature of many microscopic investigations is changing from surface
information only, to tomographic three-dimensional analysis. In this intro-
duction, 3D characterization using a sequential sectioning technique in
the FIB is situated among other existing techniques for 3D imaging. Without
attempting a complete overview, we will show that FIB tomography
fills an obvious gap in the length scales that are embraced by the existing
techniques.
Figure 5.1 shows a graph representing some different techniques for 3D
imaging and their associated typical resolution performance and sample size.
Three-dimensional imaging techniques have been important in medical
technology for quite some time. Magnetic resonance imaging (MRI) is a well-
known example used for the 3D imaging of internal organs. The resolution of
the MRI technique is of the order of a few hundred micrometers. Confocal
microscopy is a technique that uses an optical through-focus series of images
of the same sample, and subsequent offline reconstruction into a 3D model.
Characterization methods using FIB/SEM DualBeam instrumentation 127