Yao N. Focused Ion Beam Systems: Basics and Applications
Подождите немного. Документ загружается.
E
s
is the sublimation energy of the atoms, and is the density of the target. It
is important to note that the sputtering yield is linearly dependent on target
density and inversely proportional to the sublimation energy.
Molecular sputtering theory for organic substances such as polymers and
biological specimens is defined by the ratio of removed molecules per incident
ion, and is given by a number weighted average of sputtering yields for
individual atoms:
Y
m
¼
X
%
A
Y
A
; ð13:6Þ
where Y
A
is given by (13.3). The molecular thinning rate (nm·cm
2
/mA·min) in
terms of the sputtering yield is given by
R
M
¼ 6:23 ·
S
M
M
· G
–R
ð13:7Þ
where G
–R
is
G
R
¼
X
A
–R
E
B
ð13:8Þ
A
–R
is the removal mass of a particular molecule and E
B
is the bonding
energy with which the molecule was bound to the bulk material. Values for
removal mass are affected by the packing density factor , which accounts for
the increased volume within the sample due to vacuoles and lipid vesicles. As
a result, the removal mass, A
–R
, is actually calculated as A
–R
for use in
(13.8). Once again, an inverse relationship between the mass removal rate and
the bonding energy is predicted. The effect of atomic packing is confirmed by
comparing the sputtering of enamel and dentine. Due to a decreased packing
factor, enamel with its increased hardness shows a lower removal rate despite
a similar composition to dentine as demonstrated by Kanaya et al.[19].
Enamel shows a predicted thinning rate of 95 nm·cm
2
/ mA·min with an
experimental value of 86.6 while dentine with a predicted value of
122 nm·cm
2
/ mA·min has a measured value of 123 nm·cm
2
/ mA·min.
13.3.2 Artefacts and considerations
Use of FIB to section and mill biological samples is preferable to ultra-
microtomy due to stress-free milling. Samples will not show the effects
of shear stress or scratch marks due to the cutting by diamond knives.
Despite these advantages, radiation damage as a result of milling or sec-
tioning of a biological specimen using a focused ion beam often results in
noticeable artefacts. Such damage includes ion implantation, sample heating,
Focused ion beam systems348
amorphization, destabilization of native structures, and milling artefacts.
Despite the unavoidable nature of these phenomena steps can be taken to
minimize their occurrence as well as their effects.
Ion implantation is commonplace in any sample that has been milled or
imaged with a focused ion beam. The effects of ion implantation range from
sample heating due to energy dissipation, defect formation, and amorphi-
zation of the surface layer. The penetration depth depends on the stopping
ability of the sample, which depends strongly on Lindhard’s power potential
theory reviewed earlier. Greater penetration depths result in thicker amor-
phous layers, and therefore increased artefacts when performing surface
imaging. Drobne et al.[8] noted the presence of such amorphous layers when
sectioning the digestive system of a crustacean, although no significant
reduction in the quality of the image was reported. Redeposition of sputtered
material may resemble amorphization of surface layers; however, these
artefacts are unrelated, with redeposition only becoming a significant issue
when dealing with deep cross sections or trenches. Ishitani et al.[20] suggest
that lowered incident ion beam energy will decrease implantation depth and
minimize thickness of the amorphous layer, which has been confirmed by
Monte Carlo simulations. However, increased damage density at the surface
is a likely consequence. Amorphous layers produced by milling with a high
energy ion beam can be removed and smoothed by subsequent scanning with
a lower energy beam to increase the integrity of imaging.
An important point to consider is the stability of complex cellular structures
under ion beam bombardment. High energy ions may result in the dena-
turation of proteins that form cellular support structures, or destroy cross
linkages produce during fixation, which may destabilize cellular features.
Stability is also decreased by sample heating indirectly caused by ion
implantation. Transfer of kinetic energy from ions to the atoms resulting in
the emission of secondary electrons often leads to thermal damage as these
electrons are reabsorbed by surrounding atoms. When the electron stopping
power of the sample is high, the increase in thermal energy becomes an
important factor, especially in frozen samples on a cold stage without che-
mical fixation. Although Averbeck et al.[21] has performed simulations that
show that sample heating in hard materials can be considered negligible, these
local effects may still have a large impact on biological samples.
On a larger scale, milling artefacts of biological specimen may result in
striations or ridge-like filaments in the direction of the ion beam [3,9]. The
factors leading to the formation of these structures are poorly understood,
however certain measures can be taken to reduce the presence of such
artefacts. After initial milling at 5 to 7 nA with a beam energy of 30 keV Ga
þ
,
Applications for biological materials 349
Drobne et al.[9] performed a final cleaning mill at an aperture corresponding
to 0.3 to 1.0 nA to remove surface layers containing milling artefacts. In a
similar method, Young et al.[3] utilized low energy ions to remove thin
surface layers, which revealed surface melting, amorphization, and redepo-
sition. These artefacts that arise as a result of ion–sample interactions are
unavoidable. Although their formation is not completely understood, their
presence can be minimized through optimization of the FIB for life science
applications.
13.4 Life science applications of FIB
As with the focused ion beam applications in the realm of traditional
materials such as semiconductors, metals, and ceramics, the most important
advantage of FIB is the ability to remove surface layers to reveal underlying
microstructures. This has presented scientists with an unprecedented ability
to analyze the structure of biological specimens quickly and efficiently
without the artefacts resulting from other sectioning methods such as
ultramicrotomy or freeze-fracturing. In-situ imaging provided by FIB sec-
ondary ion imaging through collection of secondary electrons in a manner
similar to that of SEM allows quasi-real time observation of the surface for
improved sectioning accuracy. In addition, due to physical effects in elec-
tron emission, FIB secondary ion imaging often avoids the charging noted
during SEM imaging, as discussed in the previous chapters. Indirect imaging
of biological samples is possible when imaging shadowgrams produced by
soft X-ray contact microscopy [30]. SIM imaging allows for direct imaging
while sectioning of the silicon nitride relief provides a direct height mea-
surement of the relief, which gives a sense of the size and thickness of the
organism being imaged. Despite these advantages, FIB utilization within the
field of biological studies remains a relatively new field. Thus far, life science
applications have been limited to adaptations of existing techniques to fulfill
the needs of those studying biological materials. The most popular uses to
date are simple sectioning of biological materials and secondary ion mass
spectroscopy (SIMS).
13.4.1 Sectioning of biological materials
Generation of cross sections and removal of surface layers are extremely
crucial to the study of structural elements in biological specimen ranging
from single cells to complete organs and organisms. In the past, sectioning of
soft materials required embedment of the sample within a hard plastic matrix
Focused ion beam systems350
to avoid scratching and extreme deformation due to the applied stresses of
diamond knives. Use of FIB provides stress-free sectioning, negating the need
for embedding of biological specimen with strong cell walls and exoskeletons
as described earlier in this chapter. Through the use of embedments, an
increasing range of samples suitable to FIB sectioning are available, such as
gland cells, digestive systems, and microvilli of the small intestine of small
terrestrial organisms [1,3,8,22]. Sample preparation was performed in
accordance with traditional preparation of biological specimens, while FIB
sectioning was performed in much the same way as traditional sectioning of
hard materials. Charging was noticed in dual-beam systems during imaging
with the electron beam; however, pre-bombardment with low current ion
beams was able to avoid charging (Figures 13.7 and 13.8)[22].
A natural extension of normal sectioning is three-dimensional recon-
struction using serial sectioning. This technique, which has been used
extensively to study semiconductors and metallic alloys, is slowly finding its
way into the field of life sciences. Three dimensional reconstruction of
single cells with the FIB is a key element in improving the knowledge of cell
structure and the relationships of different cellular components. Recon-
struction using serial sectioning has been slowly developed and refined since
the 1960s. Keddie et al.[23] and Davison et al.[24] utilized serial sectioning
Figure 13.7 SEM micrograph of digestive gland cell by Drobne et al.[22]
before ion beam bombardment shows charging. N stands for nucleus.
L stands for lipid vesicle.
Applications for biological materials 351
by ultramicrotomy to study cell sections by SEM. Models were then gen-
erated to study the lipid membrane network within yeast cells. These
methods relied on traditional sample preparation, and throughout the years
have been slowly refined to include cryotechniques such as freeze-sub-
stitution for increased accuracy in quantization of the volume fraction of
various cellular components in yeast cells [25]. However, these methods still
rely on sectioning by ultramicrotomy. The difficulties of serial sectioning in
this manner are presented in work by Fiala et al.[26]. Due to random
distortions in the sections, such as tilt, rotation, and skewing, digital images
must first be processed using complex mathematical transformations to
achieve correct alignments. Despite the ability to correct for the nonuni-
form alignment of serial sections, it is difficult to guarantee section thick-
ness, which increases the difficulty of using shape based interpolation when
visualizing three-dimensional volumes and also decreases the accuracy when
quantizing cellular components.
The use of FIB sectioning suggests a solution to these problems. Serial sec-
tioning in the FIB provides self-aligned images with stress-free cutting and
relatively low amounts of sample surface destruction at grazing angles. Inclu-
sion of pre-milled fiducial marks provides a relatively accurate method for
quantizing section depth, eliminating the error involved with inhomogeneous
Figure 13.8 SEM micrograph of digestive gland cell by Drobne et al.[22]
after ion beam bombardment shows no appreciable charging. N stands for
nucleus. L stands for lipid vesicle.
Focused ion beam systems352
section thickness. Despite the relative lack of published material concerning
this technique, the steps required to implement three-dimensional tomography
in the FIB are already in place. With the production of conduction polymer
embedments by inclusion of carbon nanotubes, sectioning can be performed by
FIB, and serial imaging by SEM to provide serial sections with high spatial
resolution. New developments in sectioning and etching could lead to nano-
biomachining in a similar manner to device preparation in the semiconductor
industry [4].
13.4.2 Secondary ion mass spectroscopy (SIMS)
Study of cell composition not only includes the structure analysis but chemical
analysis as well. Chemical analysis becomes extremely important when
studying the destination of proteins after production, and the uptake of
molecules from the extracellular environment. The primary advantage of
SIMS is the ability to detect elements with parts per billion sensitivity in a
small area. Combined with serial imaging, the ability to generate a three-
dimensional elemental map of biological structures becomes a possibility.
SIMS has found acceptance in the study of chemical uptake by cancer cells. By
analyzing the destination of certain chemicals and the specificity of cancer cells
for certain drugs, more effective cancer treatments can be developed. One such
method is the preferential uptake of
10
B by cancerous brain tissue [27–29].
10
B
undergoes fission to give
7
Li and
4
He when a thermal neutron is absorbed.
Release of radiation is enough to kill cells within 10 mm of the reaction site. As
a result, the location of
10
B within the cancer cell is important, with location
closest to the nucleus providing the most effective destruction of cancer cells
[28,29]. SIMS has shown concentration of
10
B at cellular and nuclear mem-
branes as well as within single cancer cells surrounded by normal brain tissue
[27]. These results suggest that the injection of compounds containing
10
B will
provide an effective cancer treatment, eradicating even single cancer cells.
SIMS has found applications in many similar situations, and its usage will
continue to grow as high spatial resolution of chemical mapping becomes
increasingly important in molecular microbiology.
References
[1] K. Yonehara, N. Baba and K. Kanaya. J. Electron Microsc. Tech., 12 (1989),
71–7.
[2] K. Kanaya, Y. Muranaka, K. Yonehara and K. Adachi. Micron Microscopica
Acta, 23 (1992), 45–64.
Applications for biological materials 353
[3] R. J. Young, T. Dingle, K. Robinson and J. A. Pugh. J. Microsc., 172 (1993),
81–8.
[4] M. Ballerini, M. Milani, M. Costato, I. C. Edmond Turcu and F. Squadrini.
Proc. SPIE, 3260 (1998), 221–30.
[5] M. Milani, M. Ballerini and F. Squadrini. Proc. SPIE, 3922 (2000), 212–21.
[6] M. Ballerini, M. Milani, D. Batani and F. Squadrini. Proc. SPIE. 4261 (2001),
92–104.
[7] M. Milani, D. Ballerini and D. Batani, et al. Eur. Phys. J. Appl. Sci., 26 (2004),
123–31.
[8] D. Drobne, M. Milani, M. Ballerini, et al. J. Biomed. Optics, 9 (2004), 1238–42.
[9] D. Drobne, M. Milani, A. Zrime c, et al. Scanning, 27 (2005), 30–4.
[10] M. Milani, D. Drobne, A. Zrimec. Scanning, 27 (2005), 60–1.
[11] D. N. Dunn, G. J. Shiflet and R. Hull. Rev. Sci. Instrum., 73 (2002), 330–4.
[12] C. J. Dawes. Introduction to Biological Electron Microscopy: Theory and
Techniques (Burlington, Vermont: Ladd Research Industries, 1988).
[13] M. Jones. Organic Chemistry (New York: Norton, 2005).
[14] L. C. Sawyer and D. Grubb. Polymer Microscopy, 2nd edn (Oxford: Chapman &
Hall 1996).
[15] Z. R. Li. Industrial Applicat ions of Electron Microscopy (New York: Mercel
Dekker Inc., 2003).
[16] M. Nastasi, J. W. Mayer and J. K. Hirvonen. Ion–Solid Interactions:
Fundamental and Applications (Cambridge: Cambridge University Press, 1996).
[17] J. Lindhard, V. Nielsen and P. V. Thomsen. Mat. Fys. Dan. Vid. Selek., 33 (1963),
1–42.
[18] K. Kanaya, Y. Muranaka, K. Yonehara and K. Adachi. Micron Microscopica
Acta, 23 (1992), 45–64.
[19] K. Kanaya, N. Baba, C. Shinohara and T. Ichijo. Micron Microscopica Acta, 15
(1984), 17–35.
[20] T. Ishitani and T. Yaguchi. Microsc. Res. Tech., 35 (1996), 320–33.
[21] R. S. Averbeck and M. Ghaly. J. Appl. Phys., 76 (1994), 3908.
[22] D. Drobne, M. Milani, A. Zrimec, V. Leser and M. Berden Zrimec. J. Microsc.,
219 (2005), 29–35.
[23] F. M. Keddie and L. Barajas. J. Ultrastructure Res., 29 (1969), 260–75.
[24] M. T. Davison and P. B. Garland. J. Gen. Microbiol., 98 (1977), 147–53.
[25] S. K. Biswas, M. Yamaguchi, N. Naoe, T. Takashima and K. Takeo. J. Electron
Microsc., 52 (2003), 135–43.
[26] J. C. Fiala and K. M. Harris. J. Am. Med. Informat. Assoc., 8 (2001), 1–16.
[27] A. C. Oyedepo, S. L. Brooke, P. J. Heard, et al. J. Microsc., 213 (2004), 39–45.
[28] D. N. Slatkin. Brain , 114 (1991), 1609–29.
[29] K. Haselsberger, H. Radner and G. Pendl. Cancer Lett., 131 (1998), 109–11.
[30] M. Milani, D. Drobne, F. Tatti, et al. Scanning, 27 (2005), 249–53.
[31] P. Loyd. 3D Cryo-Dual Beam (Hillsboro; OR: FEI co mpany, 2004).
[32] H. Mulder. GIT Imaging Microsc., 2 (2003) , 8–10.
Focused ion beam systems354
14
Focused ion beam systems as a multifunctional
tool for nanotechnology
toshiaki fujii, tatsuya asahata, and takashi kaito
SII NanoTechnology Inc.
14.1 Introduction
In 1979, Dr. Seliger proposed the concept of the focused ion beam (FIB)
using liquid-metal gallium as an ion source [1]. The FIB tool focuses ions
generated from an ion source using an electric field, irradiates the ion beam
on to specimen surfaces, and observes microscopic specimen surfaces by
scanning. The scan region of the ion beam can be selectively sputter etched
when ions heavier than electrons are used. A scanning electron microscope
(SEM) can be used for observation or analysis, but FIB can be used for both
observation and processing. Many research organizations and companies are
now involved in FIB development.
The Scientific Instruments Division of Seiko Instruments Inc. (currently
SII Nano Technology Inc.) started research and development at the begin-
ning of the 1980s and developed a technology called ion beam induced
chemical vapor deposition (CVD). This technology makes it possible to
accumulate thin films.
In 1984 SII introduced the world’s first FIB photo mask repair tool
called the SIR series [2]. White defects, the shaded part of the photo
mask used in making integrated circuits that falls off, are filled in and
repaired by an ion beam induced CVD of carbon fi lm. Later, there were
advances in technical development [3], such as the capability to repair
black defects left over from shading ma terial in parts that transmit l ight,
and precision proces sing that corresponds to a miniaturi zation of the
design rules.
In 1986, SII introduced the world’s first multi-purpose commercial FIB
tool called the SMI series. It was first put to practical use as a circuit edit
Focused Ion Beam Systems: Basics and Applications, ed. N. Yao.
Published by Cambridge University Press. ª Cambridge University Press 2007.
355
tool for semiconductor integrated circuits [4,5]. New wire formation could
be done by localized sputter etching with an FIB, by cutting wire or
exposing wiring for drilling beneath the passivation layer, and by depos-
iting tungsten film.
At the same time, US and Japanese semiconductor device makers were
competing to develop a 1 Mb DRAM. In particular, the 3D structuring of
elements and wires to improve memory integration progressed at the same
time as the miniaturization of photo-lithography. To verify the cross-
sectional structure of test devices, a method was used to prepare and observe
specimens with a microscope. But both specifying and observing a location
by this method has proved difficult.
On the other hand, a cross-sectioning and observation function using
FIB could observe the topography with a microscope function, determine
the cross section position, and realize a previously nonexistent function for
specifying the process location [6]. The cross section process observation
function is the third application of FIB equipment. This function was able
to largely contribute to the development of the semiconductor production
process.
Incidentally, the ion beam diameter of the FIB tool at that time was between
50 and 100 nm. Today, the tool is produced with a minimum beam size of
4 nm. This extremely narrow ion beam enables micro-processing at the order
of nanometers. This introduces ways in which the latest FIB system can
contribute to nanotechnology.
14.2 FIB tools
The FIB tool can observe microscopic areas and perform both etching and
deposition processes (Figure 14.1). This section introduces these functions as
related to nanotechnology.
14.2.1 Observation
Secondary electrons and secondary ions are produced from the specimen sur-
face by ion beam irradiation. A microscopic image can be obtained through
detecting these secondary electrons or secondary ions, converting them to
electric signals, and displaying two-dimensional distributions. The idea is the
same with SEM. High-resolution observation can be achieved by narrowing
the ion beam irradiation area.
Focused ion beam systems356
Figure 14.2 shows the differences between electron beam irradiation and
ion beam irradiation. Both SEM and FIB can obtain microscope images
using similar principles. However, they use signals of different characteristics
to form these images.
It is common to use a secondary electron for microscope observation. The
amount of secondary electrons generated by the electron beam depends upon
the topography. Imaging by electron beam irradiation gives a topographical
contrast image. Consequently, the amount of secondary electrons generated
by the ion beam relies not only on the topography but also on material or
structural differences in the sample, and thus the image obtained is a struc-
tural contrast image.
Figure 14.3 shows a cross section of a Cu wire. Grain observation is possible
through the channeling phenomenon of ion beam irradiation. Ion beam
irradiation is different to electron beam irradiation. As shown in Figure 14.2,
ion beam irradiation generates secondary ions from the specimen surface,
detects secondary ions, and enables observation of a secondary ion image.
Observation
Etching Deposition
Carbon wallTEM sample preparation
Figure 14.1 The basic functions of FIB.
Electron
Secondary electron Secondary electron
Sputtered particles
Secondary ion
Auger electron
lon
X-ray
Reflecting electron
Figure 14.2 The differences between electron beam irrad iation and ion beam
irradiation.
Focused ion beam systems as a multifunctional tool for nanotechnology 357