Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM

Abstract

Probing mechanical properties in the micrometer regime is of current interest in materials science. A focused ion beam microscope was employed to fabricate miniaturized specimens, while an indenter installed in a scanning electron microscope was utilized to actuate the samples and record the load and displacement data during the deformation. Examples for miniaturized compression, tension, bending, as well as newly developed bending fatigue and bending fracture experiments are presented, demonstrating the unique flexibility of in-situ mechanical testing in the scanning electron microscope at small length scales.

Full text

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
Daniel Kiener
a,b,d
, Christian Motz
b
, Gerhard Dehm
b,c
, Reinhard Pippan
b
a
Materials Center Leoben, Forschungs GmbH, Leoben, Austria
b
Erich Schmid Institute of Material Science, Austrian Academy of Sciences, Leoben, Austria
c
Department of Materials Physics, Montanuniversität Leoben, Leoben, Austria
d
now at: National Center for Electron Microscopy, Lawrence Berkeley National Laboratory, Berkeley, U.S.A.
Overview on established and novel FIB based
miniaturized mechanical testing using in-situ SEM
Dedicated to Professor Dr. Franz Jeglitsch on the occasion of his 75
th
birthday
Probing mechanical properties in the micrometer regime is
of current interest in materials science. A focused ion beam
microscope was employed to fabricate miniaturized speci-
mens, while an indenter installed in a scanning electron mi-
croscope was utilized to actuate the samples and record the
load and displacement data during the deformation. Exam-
ples for miniaturized compression, tension, bending, as
well as newly developed bending fatigue and bending frac-
ture experiments are presented, demonstrating the unique
flexibility of in-situ mechanical testing in the scanning
electron microscope at small length scales.
Keywords: Size-effect; Micro-mechanical testing; Fo-
cussed ion beam (FIB); in-situ SEM
Introduction
It is common knowledge that the strength of a polycrystal-
line bulk sample is governed by its smallest microstructural
dimension [1 – 3]. With the still ongoing trend of miniaturi-
zation, new methods for testing materials with smaller and
smaller dimensions down to the micrometer regime have
been developed. Classical work on bending of thin polycrys-
talline Ni foils [4], torsion of polycrystalline Cu wires [5],
and indentation testing [6, 7] shows that there are distinct
size effects on the mechanical properties in this size regime.
Some of these approaches were even further developed to
probe features in the nanometer size regime, for example
dedicated micro-electro-mechanical system (MEMS) de-
vices to test free-standing nanostructured thin films in ten-
sion [8], or instrumented in-situ indentation in a transmis-
sion electron microscope (TEM) to investigate incipient
plasticity [9]. While bending [4], torsion [5], and the MEMS
[8] techniques were successfully applied to polycrystalline
samples, the fabrication of suitable samples imposed limita-
tions to these approaches when aiming to determine the me-
chanical properties of free-standing micrometer sized single
crystals. As a matter of fact, melt cast single crystals cannot
be directly fabricated in the shape of a sheet or wire with mi-
crometer dimensions. Nanoindentation techniques provide
the resolution of probing quasi single crystal properties in
the nanometer regime in polycrystalline material [10], but
the complex stress – strain state beneath the indenter compli-
cates the data evaluation. Single crystal thin films can be
produced to thicknesses of several micrometers and the ef-
fect of the confining substrate can be minimized by newly
developed transfer techniques [11]. Still, only one dimen-
sion of these films is in the micrometer regime, while the
other two dimensions are in the millimeter range.
These fabrication limitations were overcome by the broad-
er availability of focused ion beam (FIB) microscopes in ma-
terials science [12]. Using a FIB for local material removal
and a flat ended diamond tip for compression testing in a
nanoindenter, Uchic et al. [13, 14] were the first to test FIB
micro-structured single crystal specimens with all dimensions
in the micrometer regime. While their results showed an inter-
esting size effect of increasing strength with reduced sample
dimension, which is still a hot topic in the materials science
community debate, they also opened up the field to FIB-based
micromechanical testing [15]. A number of groups more or
less closely mimicked their micro-compression approach and
applied it to various materials [16 – 26], but other loading
techniques were also developed, inspired by their work. For
example, there were methods developed for micro-bending
[27] and micro-tensile testing [28] of miniaturized single
crystals. The aim of this overview is to focus on newly devel-
oped micro-mechanical test methods rather than discussing
size-effects in materials. However, before reporting on novel
miniaturized testing concepts a short review of the current un-
derstanding of micro-compression testing of single-crystal-
line face centered cubic (fcc) metals is provided, and chal-
lenges and limitations of micro-mechanical testing with a
focus on micro-compression testing are discussed.
2. Common observations and challenges in micro-
compression testing
2.1. Common observations for fcc single crystals
One of the striking observations of the micro-compression
experiments carried out by Uchic et al. [13, 14] was the inter-
mittency of the plastic flow. In load controlled experiments,
discrete strain bursts connected by regimes of nearly elastic
loading were observed. Exemplary data provided by Frick
et al. [29] for Ni(111) showing this behavior is presented in
Fig. 1a. The strain bursts are usually interpreted as disloca-
tion avalanches [30, 31], and statistical analysis of the size
and frequency of these bursts shows a power law scaling
[30, 31]. Apparently, once a dislocation source is activated
at a certain stress level, significant plastic deformation can
be achieved by easy slip. There seems to be literally no hard-
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1074 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
ening involved, until this source is shut down. At this point
the system is loaded in a nearly elastic manner until another
source is activated at higher stress levels.
In displacement controlled testing, the occurrence of load
drops is observed. Once a source is activated, it will emit
dislocations and thereby generate strain. These dislocations
move faster than the usual strain rate of such a quasi-static
experiment. This results in a reduction in the contact force
seen as a vertical drop in the load – displacement data.
When the displacement rate controlled flat punch catches
up with the sample, it will be reloaded in a close to elastic
manner. An example of this behavior is shown in Fig. 1b
[20, 32] for Cu(111) micro-compression samples. There-
fore, strain bursts and load drops are generally regarded as
equivalent. However, a frequently overlooked but signifi-
cant difference between the two loading modes can be ra-
tionalized when looking at the stress – strain curve of the
smallest sample (diameter 1.29 lm) shown in Fig. 1b.
There are several minor load drops that occur at lower stres-
ses than the previous major load drop. This cannot be ob-
served in load controlled mode, where the load would stay
at the level of the previous major burst [33].
A second important observation by Uchic et al. [13, 14]
was the inverse scaling of strength with sample size. This
has been widely confirmed in various micro-compression
studies and is also obvious from the data shown in Fig. 1a
and b. A more comprehensive dataset for Cu [20, 32, 34,
35], Ni [14, 25, 29], Au [16, 18, 36], and Al [24] is shown
in Fig. 1c. To compare the different materials in an intuitive
dimensionless plot, as very recently suggested by Dou and
Derby [37], the shear strength swas normalized by the ani-
sotropic shear modulus Gand the sample dimension dby
the Burgers vector b. Note that the studies differ in the re-
presentative strain used to determine the flow stress level,
ranging from the plastic limit [24] to 10 % strain [16, 32].
Nevertheless, a size effect is observed in all cases. This is
generally rationalized by a scaling of the dislocation source
size with the sample size, causing higher stresses for smal-
ler samples [28, 38 – 42]. Values of a power law fitting to
the data range between d
–0.4
and d
–1
. The power law expo-
nent itself is influenced by several experimental factors, for
example the initial dislocation density [43] and strain rate
[44], as well as details of the data evaluation, for example
the representative strain value used to determine flow stres-
ses [45] and the fitting method [46].
Similar observations in terms of strength scaling, inter-
mittency and stochastic flow, but with differences in the
hardening behavior and the scaling exponents, are also ob-
served in body centered cubic (bcc) materials. A detailed
treatment is beyond the scope of this article, the interested
reader is referred for example to a recent review [47].
2.2. Need for in-situ experiments
An important and frequently discussed issue in any small
scale testing is the alignment between sample and testing
equipment [48, 49]. There are ways to improve alignment
issues ex-situ [15], but clearly the benefits from performing
measurements in-situ is that, beside minimizing these align-
ment issues, direct visual control of the experiment signifi-
cantly increases the reliability of information gained during
a mechanical test. Ex-situ testing relies on linking of the
specimen characteristics prior to testing and after ex-situ
characterization by the measured mechanical data. This
can be done for continuous and homogenous sample defor-
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1075
B
Basic
(a)
(b)
(c)
Fig. 1. Common features observed in micro-compression testing of
face centered cubic (fcc) materials produced by FIB machining. (a)
Ni(111) micro-compression samples of various size tested under load
control show discrete strain bursts (redrawn from Frick et al. [29]). (b)
Cu(111) samples loaded in displacement control mode depict load
drops (redrawn from [20, 32]). (c) Normalized plot of strength versus
dimension for micro-compression results obtained from various fcc
metals [14, 16 –18, 20, 24, 25, 29, 32, 34– 36]. In general higher
strength is observed for smaller samples.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
mation, comparable to what is known from bulk testing.
With increasing complexity of the deformation behavior,
even the simple question for the true stress can be a matter
for discussion, since for example the actual minimal sample
cross-section or the contact area are unknown. This is easily
rationalized for micrometer sized single crystal samples
that deform by dislocation glide, given the stochastic [24]
and discrete nature [31] of deformation on this length scale.
For amorphous materials these geometrical problems are
also of major concern. A recent analysis showed that the
postulated size effect in the deformation of micrometer
sized amorphous compression samples could be to a large
extent caused by an improper estimation of the contact size
[26]. The above clearly shows that the issues mentioned are
in general geometrical problems, independent of the inves-
tigated material. As a consequence, all the samples pre-
sented here were tested in-situ in a scanning electron micro-
scope (SEM) to have permanent documentation of at least
two of the three sample dimensions.
3. Available in-situ testing setups
There exist several methods for direct in-situ observation in
the micrometer regime, using either imaging or X-ray
methods. Here we focus on the most general and flexible
technique, namely direct visual control in an SEM. Com-
pared to optical microscopy, the SEM offers higher resolu-
tion and reduced environmental influences. Meanwhile
there is a growing number of groups that have installed an
instrumented indenter system in an SEM [20 – 22, 34, 50 –
53]. Similar approaches have even been realized in a TEM
[8, 9, 25, 54 – 56], offering higher resolution and the possi-
bility to observe the sample deformation direct or in diffrac-
tion mode [42, 57]. A drawback of the TEM approach is the
limited thickness to maintain electron transparency of the
investigated material. Concerning the X-ray methods, mi-
cro-Laue techniques were applied to characterize the sam-
ple microstructure in the initial state for Si, Au, Ni, Cu and
NiTi samples in [19, 58] and in-situ during the experiment
for Au and Cu samples in [59, 60], while micro-Raman
spectroscopy [61] was used for local in-situ stress measure-
ment of Si samples.
3.1. SEM and attached micro-indenter
The requirements for the SEM to perform micro-mechani-
cal testing are rather moderate. Besides a large vacuum
chamber to contain sample and loading equipment, just sev-
eral flanges with signal feedthroughs for piezo voltage sup-
ply as well as load and displacement measurement are re-
quired. In the present work, a tungsten filament SEM (Carl
Zeiss SMT AG, Oberkochen, Germany; LEO 440 Stereo-
scan) typically operated at 10 – 20 keV was used together
with a micro-indenter system (ASMEC, Radeberg, Ger-
many; UNAT) for sample loading. To save operational
space, the indenter positioning system is partially contained
in an attached cap. For this reason, these stages can stay at-
tached to the SEM once the micro-indenter is removed for
ex-situ use. Additionally, to improve the signal-to-noise ra-
tio, the whole SEM was put on an active vibration isolation
system (Halcyonics, Goettingen, Germany). Figure 2
shows the open SEM chamber with the micro-indenter in
an inclined position suited for indentation experiments un-
der a predefined (viewing) angle. Note that the sample sur-
face has to be tilted accordingly.
In the following, the purpose of the indicated axis will be
briefly explained. On the indenter side, four axes (x
2
,y
2
,m
2
,
t
2
) allow to position the indenter tip in the electron beam
(x
2
,y
2
) at the desired working distance (m
2
) and inclination
angle (t
2
). Due to constructional limitations given by the
SEM column, the minimum working distance is about
10 mm. This mechanical alignment of adjusting the inden-
ter tip in the electron beam is done at moderate magnifica-
tions of the order of 1000 ·. Changing the field of view at
higher magnifications during testing is accomplished using
the electron beam shift of the SEM.
The sample is moved into the coincidence point using the
SEM stages. The linear axes x
1
,y
1
,z
1
,(m
1
) are in principle
sufficient for positioning. The rotational axis r
1
ensures
alignment between indenter loading axis (x
3
) and specimen
axis (x
1
) in the x-direction. The tilting axis t
1
is required to
ensure orthogonality between sample surface and indenter
loading axis. The actual test is performed with the instru-
mented x
3
axis of the indenter. A schematic arrangement
of the different axis in the aligned system is given in Fig. 3f,
while the movement ranges are provided in Table 1.
For all tests presented in this work except of the indenta-
tion shown in Fig. 3a and b, samples were situated at edges
or freestanding on the tip of needles. In these cases, the in-
denter can be used in the horizontal position (t
2
= 0). This
simplifies alignment and reduces uncertainties from an in-
clined view.
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1076 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic
Fig. 2. View into the open SEM chamber with the micro-indenter in-
stalled, including indication of the available axis and the coincidence
point.
Table 1. Overview of the available axis and their moving range.
SEM Indenter
Axis Range Axis Range
x
1
±30 mm x
2
±5 mm
y
1
±30 mm y
2
±3 mm
z
1
±4 mm x
3
±30 lm
m
1
±30 mm m
2
±25 mm
t
1
–108– +808t
2
08–258
r
1
3608

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
3.2. Indenter tips
Besides commercially available sharp tips (cube corner,
Berkovich) used for in-situ indentation (Fig. 3a and b),
beam bending, and fracture experiments, there is a variety
of custom-made tips. The easiest and most frequently used
approach is to flatten a sharp indenter tip by FIB milling to
obtain a flat punch of the desired dimensions (Fig. 3c) [13,
15, 20, 62]. Flat punches are also commercial available
(e. g. Synton MDP, Nidau, Suisse) [18]. In our lab a large
variety of custom and commercial flat punches made of dia-
mond, tungsten or steel with diameters of 2 lm, 8 lm,
10 lm, 20 lm, 50 lm, and 200 lm are available. Choosing
the flat punch large enough to cover the whole specimen is
important. On the other hand, the smaller the tip, the larger
is the view of sight and the less surrounding material needs
to be removed by FIB milling to prevent contact of the flat
punch and nearby bulk material. Metallic punches reduce
charging problems caused by the electron beam, which can
become an issue especially when both sample and tip are
non-conductive. Changing the contact partners by using flat
punches of different materials may even open new possibi-
lities for studying adhesion issues.
In addition to these rather common tips, there are several
custom designed tips that also require FIB milling for fabri-
cation. These are for example differently sized micro-grip-
pers dedicated to micro-tensile testing (Fig. 3d) and side-
slits with defined wedges for reverse bending and bending
fatigue experiments (Fig. 3e).
4. Sample fabrication
Sample fabrication using FIB milling offers a unique tech-
nique to shape various materials with a precision of better
than tens of nanometers and surface roughness of some
nanometers [63]. The general approaches and important
parameters as well as the possible material modifications
from this milling process will be briefly described. How-
ever, the details of FIB fabrication of the individual speci-
mens will not be discussed, as this is beyond the scope of
this work.
A FIB (Carl Zeiss SMT AG, Oberkochen, Germany;
Zeiss 1540 XB) operated with Ga
+
ions at 30 keV is
applied. For coarse milling typical ion currents are of the
order of several nA. These are stepwise reduced to final
polishing currents of about 0.1 nA. When using perpendi-
cular ion impact for material removal, a higher ion da-
mage can be expected from simulations and experimental
studies [63]. Furthermore, perpendicular ion impact in
conjunction with the crystal orientation/impinging angle
dependent sputter yield [64] leads to a broadening of the
milled face with increasing depth [65]. This finally re-
sults in unwanted specimen taper as seen in many mi-
cro-compression experiments. To circumvent this, lathe
milling techniques [15] or other techniques that facilitate
grazing ion impact by back-tilting the specimen [20] can
be applied. These are, however, more complicated to
realize, especially for smaller sample dimensions. For
example, Uchic et al. so far used their lathe milling for
compression samples with diameters of 1 lm or larger
[13, 15].
It was shown recently that the strength of defect free
whiskers that sustained stresses close to the theoretical
strength in their pristine state [66], was reduced signifi-
cantly [67] by the presence of artefacts created by FIB
milling [63, 68]. In fact, the material behaved as being
pre-strained [62], thus containing a higher amount of mo-
bile dislocations than expected for a perfect whisker or a
well annealed melt grown single crystal. This is a critical
observation when trying to understand the size dependent
strength of single crystals in the micrometer regime, since
the fabrication technique might artificially alter the dislo-
cation density and as a consequence the mechanical re-
sponse. Therefore, we will focus on materials that per se
contain a higher density of defects, implying that the influ-
ence of FIB damage does not play an important role in
these cases.
Other limitations of FIB processing are possible forma-
tion of new phases [64, 69] or the well documented liquid
metal embrittlement. It was for example shown that sub-
monolayer wetting of Al grain boundaries by Ga causes em-
brittlement of these boundaries [70, 71].
Newly developed He
+
ion microscopes are promising for
future development of micro sample fabrication. While the
much lower milling rates when using He
+
ions will not al-
low actual sample fabrication, removal of the damaged sur-
face layer caused by Ga
+
ion milling is a very desirable op-
tion achievable when using He
+
ion microscopes.
At the same time, alternative fabrication methods are
being developed. The main goal here is to reduce the re-
quired amount of ion milling or to replace it completely.
Moreover, strong parallelization of the sample fabrication
process is desirable. Among these emerging methods are
micro-electrodischarge machining [15], electroplating
[16], photolithography etching techniques [22], directional
solidification with subsequent etching [62], and embossing
[46].
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1077
B
Basic
Fig. 3. (a, b) Sharp diamond cube corner tip for indentation, beam
bending and bending fracture experiments (not shown here). (c) Flat
ended conical 20 lm diamond tip for micro-compression experiments.
(d) Custom made tungsten sample gripper for micro-tensile loading.
(e) Side slit with wedges for reverse bending and bending fatigue load-
ing. (f) Schematic depicting the available axis of the applied in-situ set-
up.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
5. Examples of in-situ deformation techniques
in an SEM
In the following exemplary case studies of different in-situ
deformation techniques will be provided. No examples for
in-situ indentation in an SEM will be shown, since there is
extensive literature available [51, 52, 72].
5.1. Compression testing
Compression testing is known to be strongly influenced by
boundary conditions, which are difficult to control in mi-
cro-compression testing [32]. Nevertheless, there are situa-
tions where these limitations are of minor concern and the
technique can be applied successfully. This holds true espe-
cially if the microstructural dimensions are significantly
smaller than a micron-sized compression sample. The other
general concern is related to the defects generated during
FIB milling. It is reasonable to assume that materials that
per se contain a high defect density should not be suscepti-
ble to FIB damage [73]. Material systems that fulfil these
requirements are, for example, sputter deposited hard coat-
ings with grain sizes of some tens of nanometers [74], or
nuclear reactor steels exposed to neutron radiation damage.
The lifetime reduction of a structural steel due to radia-
tion damage is probed by mechanical testing. While real re-
actor tests are time consuming and expensive, similar mate-
rial damage can be created by proton irradiation at lower
costs and without strongly activating the material. A limita-
tion is given by the low penetration depth of the protons
compared to classical neutron irradiation. Thus, small scale
testing techniques are required. Indentation testing of the ir-
radiated material reveals an increase in hardness, which can
be correlated to an increased yield strength [75, 76]. It is,
however, difficult to access important quantities such as
the hardening capability, the fracture toughness, or the
strain to failure of the irradiated material by indentation
techniques. Thus, micro-compression tests of samples situ-
ated in individual grains in the irradiated and unirradiated
material were performed on FIB fabricated micro-compres-
sion samples [77]. Since this complex alloy consists of a
martensitic microstructure, the critical dimension is given
by the lamellar width, and not the sample dimension itself.
The grains were screened with respect to the lamella spa-
cing and grains with lamellae oriented edge-on were se-
lected, but no detailed orientation information was col-
lected. Subsequently, specimens with dimensions of
8lm·8lm·15 lm were fabricated to sample a volume
as large as possible within a single grain. Since the material
was on purpose irradiated with a 2 MeV proton beam at
doses corresponding to 7 displacements per atom, no influ-
ence from FIB damage is expected for the samples tested
in the irradiated condition.
The stress versus strain results of four samples FIB fabri-
cated under identical conditions are shown in Fig. 4a. The
proton irradiated samples exhibit higher stresses than the
unirradiated specimens. While the data of the unirradiated
samples are in close agreement, there are differences be-
tween the two curves of the irradiated samples. This is ex-
pected to result from the grain orientation, which is not
known. Taking the derivative of the stress versus strain
data, the strain hardening rate was obtained and is depicted
in Fig. 4b. There is some scatter in the data which results
from noise in the measurement as well as discrete plasticity
events [31] leading to “spikes”, since the derivative was
taken directly from the measured data without any curve fit-
ting. However, a general trend that the unirradiated material
shows lower initial hardening and higher hardening cap-
ability at strains above about 3 % is observed.
5.2. Tensile testing
While tensile loading can be difficult to perform on brittle
or very hard materials, it is the preferable method to test
ductile materials. Here we show another application of a
previously developed tensile setup [28]. By FIB fabricating
a micro-tensile sample with a specially shaped head and a
corresponding individual gripper (Fig. 3d, Fig. 5a), tensile
loading can be performed. The result shown here depicts
the mechanical behaviour of a nanocrystalline Cu micro-
tensile specimen with dimensions of 3 lm·3lm·15 lm
and an average grain size of 0.2 lm ± 0.1 lm as produced
by severe plastic deformation [78]. Such material is known
to possess very high dislocation densities, thus the damage
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1078 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic
(a)
(b)
Fig. 4. (a) Stress versus strain curves of 8 lm·8lm·15 lm micro-
compression samples situated in single grains in the irradiated (red
lines) and unirradiated (green lines) material. Higher yield stresses are
observed for the irradiated material [77]. (b) Derivative of the data
from (a) showing that the irradiated material has higher initial harden-
ing and less work hardening capability at higher strains compared to
the unirradiated material.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
introduced by FIB milling can be neglected. The sample has
no perfectly regular dimensions, because orientation depen-
dent sputter rates [64] lead to the preferential removal of
certain grain orientations, resulting in irregular edges.
These stress concentrations also pre-define the necking re-
gion, as can be seen in the in-situ SEM image at 6.1 % strain
(Fig. 5a) as well as recorded with a higher resolution ex-situ
in the unloaded condition after having been strained to
6.7 % (Fig. 5b). Nevertheless, comparison between the
stress versus strain curve of this micro-tensile test and a
macroscopic tensile test with similar microstructure and
sample dimensions of 1.2 mm ·1mm·8 mm loaded in a
commercial tensile machine also suited for in-situ tensile
testing in an SEM (Kammrath & Weiss GmbH, Dortmund,
Germany) shows close agreement. This is because even for
the micro-tensile specimen there are several grains across
the sample cross-section and it is the grain size that controls
the dislocation generation and motion. For comparison,
samples with grain sizes of 100 lm, 10 lm, and 1 lm tested
using the tensile stage [79] are also included in Fig. 5c.
Compared to the commercial straining stage, where the
displacement is measured at the cross-head, it is worth
pointing out the very reliable strain determination in the
case of the micro-tensile test, where the elongation is mea-
sured at the indenter tip. This is shown by the high elastic
slope of the micro-tensile test compared to the macroscopic
samples. Further improvement is possible by image correla-
tion, as will be discussed in Section 5.1. Another interesting
feature is visible in the inset in Fig. 5c. While the measured
curve of the macroscopic sample is smooth, the curve of the
microscopic sample shows stress drops indicated by arrows.
This is typical for dislocation avalanches in displacement
controlled tests and can usually only be observed if the
sampled volume is small enough [31]. Similar observations
with more pronounced load drops were reported for strain-
ing of Cu wires with comparable diameter but larger grain
size [80, 81].
An alternative method to accomplish tensile loading with
a simple indentation setup is depicted in Fig. 6a for a single
crystal Cu(100) sample. A *3lm thick lamella is struc-
tured to a U-shaped configuration. The bar on the left hand
side is further thinned to become the actual tensile speci-
men with dimensions of 1.5 lm·1lm·5.5 lm, while
the right hand side acts as a hinge. Actuating the indenter
tip along x
3
bends the structure around the hinge and causes
the left hand side to elongate under a tensile stress.
Loading the hinge with a sharp tip causes at the same
time indentation of the lamella. This yields additional mea-
sured displacement, which has to be corrected for. Fig-
ure 6b shows the measured load versus displacement data
at the indentation tip (dashed line). Applying the equili-
brium of momentum at the hinge, the lower load at the ten-
sile sample was calculated and is given by the dotted line.
The actual strain of the tensile specimen was measured in-
situ during loading, resulting in the final load versus displa-
cement curve depicted by the solid line in Fig. 6b. Another
possibility of correcting for the additional indentation is to
subtract the displacement from reference indents made in
the same lamella [27]. This approach was performed for
the bending experiments explained in the next chapter.
The stress versus strain curve presented in Fig. 6c shows a
low loading slope compared to the micro-tensile tests
shown previously (Fig. 5c). This is presumably because
the image resolution is not high enough to clearly discrimi-
nate between the indentation displacement and the elonga-
tion of the tensile specimen in this initial loading regime.
After straining to *5 %, a decrease in the technical stress
is observed. This is caused by the formation of a single
large slip step indicated by an arrow in Fig. 6d. This is not
common for a macroscopic multiple slip oriented sample,
but has been observed several times for miniaturized speci-
mens and explained by the stochastic nature of dislocation
sources in this size regime [82].
The flow stress before necking of *350 MPa corre-
sponds to a resolved shear stress of *140 MPa. This value
is significantly higher than that of *50 MPa observed for
freestanding Cu micro-tensile specimens of similar dimen-
sions [28]. This indicates that the possible bending compo-
nent and the limited lateral movement of the bending lamel-
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1079
B
Basic
(a)
(b)
(c)
Fig. 5. (a) In-situ SEM image of a Cu micro-tensile specimen with an
average grain size of 0.2 lm at a strain of 6.1 %. (b) Ex-situ SEM im-
age of the unloaded sample showing the necking zone at the maximum
strain of 6.7 %. (c) Stress versus strain curves of macroscopic Cu sam-
ples with grain sizes of 100 lm, 10 lm, 1 lm, and 0.2 lm deformed
using a regular tensile machine (black lines) [79] and a Cu micro-ten-
sile specimen with a grain size of 0.2 lm (red line). Close agreement
between the macroscopic and the microscopic sample with a grain size
of 0.2 lm is observed. The inset shows a detail of the stress versus
strain curve, individual load drops indicated by arrows are caused by
discrete plastic events.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
la in the x
1
and z
1
directions by the stiff indenter tip exerts a
significant constraint on the deformation of the single crys-
tal specimen, resulting in higher flow stresses as studied in
detail for micro-compression and micro-tensile experi-
ments in [35]. However, the limited lateral movement of
the miniaturized tensile sample should be of minor influ-
ence when testing fine grained materials, where the micro-
structural dimensions are far below the geometrical speci-
men dimensions.
The sample design of Fig. 6 requires a noticeable amount
of coarse FIB milling to provide enough free space for later
fine milling in order to avoid material redeposition and as-
sure precise geometries without any contact between mov-
ing parts and redeposited material. Generally, the geometry
is more time consuming in fabrication and larger in dimen-
sions compared to the miniaturized tensile sample shown
in Fig. 5. An additional drawback is the superimposed in-
dentation and bending process, which makes the data eval-
uation more tedious.
The advantages are a higher load resolution compared to
uniaxial loading, since the contact point of the indenter
can be chosen along the lamella. This is a consequence
from the equilibrium of momentum around the hinge.
Moreover, a symmetric version of this tensile approach
can be designed with two tensile specimens instead of one
specimen and a hinge. This minimizes possible bending of
the tensile samples, but only allows measuring an average
strength for both samples. Last but not least, since only a
sharp tip needs to be positioned on the lamella, this test
can be reasonably performed ex-situ with an indenter sys-
tem offering imaging possibility by scanning with the tip.
5.3. Beam bending
Bending beams as shown in Fig. 7a have already been used
to investigate the influence of strain gradients on the me-
chanical properties of single crystal Cu in the micrometer
regime [20, 27]. Here the results from micro bending tests
performed on single crystal W[110](111) samples with
thicknesses of 5 lm and 1 lm (Fig. 7a and b) and nanocrys-
talline Cu specimens with thicknesses of 1.9 lm and 3.9 lm
(Fig. 7c and d) are presented. Interestingly, the single crys-
tal W beams show remarkable plasticity, with the disloca-
tion activity being confined to the highly stressed zone
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1080 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic
(a) (b)
(c) (d)
Fig. 6. (a) In-situ tensile test of a 1.46 lm diameter single crystal Cu(100) specimen using a modified bending configuration actuated with a cube
corner tip. (b) Measured load versus displacement data (dashed), corrected load at the tensile specimen (dotted), and indentation depth corrected
load versus displacement data (solid). (c) Stress versus strain data for the micro-tensile sample calculated from the corrected load versus displace-
ment data (solid line in (b)). See text for details concerning the low slope during loading. (d) Inclined ex-situ SEM image of the sample, showing the
formation of one pronounced slip step.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
close to the sample base (Fig. 7b). In the case of the nano-
crystalline bending beams shown in Fig. 7c and d, no dis-
tinct deformation features or differences between the two
beams with different thickness can be deduced, although
there is a difference in the applied strain gradient. The load
versus displacement data of the four tests is presented in
Fig. 7e. It was corrected for the additional displacement
caused by indentation of the beam. This displacement was
taken from reference imprints placed in the base region be-
tween the two samples [27]. Note that for the single crystal
W[110](111) samples intentional unloading steps were per-
formed.
For a better comparison between the different samples,
the data was replotted as normalized bending moment
F·l
B
/(w·t
2
) versus normalized displacement u/l
B
in
Fig. 7f [27, 83]. Here, Fand uare the load and displace-
ment, respectively, while the bending beam is described by
the bending length l
B
, the width w, and the thickness t.
These dimensions are indicated in Fig. 7a and given in the
diagrams Fig. 7e and Fig. 7f for the individual specimens.
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1081
B
Basic
(a) (b)
(c) (d)
(e) (f)
Fig. 7. (a) Inclined view of a FIB fabricated W[011](111) double bending beam with thicknesses of 5.0 lm and 1 lm, respectively. The width was
5lm in both cases. (b) Inclined ex-situ SEM image showing glide steps on a deformed 1.0 lm thick W[011](111) bending beam. (c) Ex-situ back-
scatter electron image of a deformed 3.9 lm thick nanocrystalline Cu beam. (d) Ex-situ backscatter electron image of a 1.9 lm thick nanocrystal-
line Cu beam after bending. (e) Load versus displacement curves for four different bending beams. (f) Normalized bending moment F·l
B
/(w·t
2
)
versus normalized displacement u/l
B
for the four presented bending beams. See text for details.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
For the single crystal W[110](111) samples a higher nor-
malized flow stress (normalized bending moment) for the
thinner sample is observed, as reported for single crystal
Cu [27]. Considering the nanocrystalline Cu, there is no sig-
nificant difference between the two specimens, although a
higher strain gradient is present in the thinner sample.
Again, with respect to the high defect density of this nano-
crystalline material, no influence from FIB milling was ex-
pected and is not observed.
5.4. Bending fracture
Instead of pure bending of a beam, a defined notch can also
be introduced by FIB milling to perform a bending fracture
test as schematically shown in Fig. 8a. Figure 8b –d present
images collected during the fracture test of a W single crys-
tal. With increasing bending of the beam, the pre-crack
grows. Also, the cube corner diamond tip slightly, but no-
ticeably indents the W beam. During the test several inten-
tional holding steps were conducted to record high quality
SEM images required to precisely detect the actual crack
length and deformation phenomena. These holding periods
in the displacement controlled experiment manifest as sig-
nificant load drops in the load versus displacement data
(Fig. 8e). The positions corresponding to Fig. 8b – d are
marked.
The most striking observation from the load versus dis-
placement data and the in-situ images is the significant
plastic deformation of the single crystal W fracture speci-
men, in contrast to bulk W which is known to fail by brittle
fracture at room temperature [84, 85]. This remarkable
amount of plasticity is clearly revealed by ex-situ backscat-
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1082 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic
(a) (b)
(c) (d)
(e) (f)
Fig. 8. (a) Schematic drawing of the fracture specimen FIB milled into the bulk W single crystal. (b –d) SEM images taken during loading to pre-
cisely evaluate the crack extension. (e) Load versus displacement curve of the fracture test. A remarkable amount of plastic deformation is ob-
served. The positions along the loading path corresponding to Fig. 8b – dare indicated. (f) Ex-situ backscatter electron image showing the formation
of slip lines along the crack path.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
ter electron investigation (Fig. 8f) showing the formation of
slip steps adjacent to the crack path.
Since proper determination of any fracture mechanics
characteristics requires knowledge of the actual crack
length, in-situ observation is essential for this kind of test-
ing. Finally, such types of experiments in conjunction with
a variation in the size of the sample as well as the tempera-
ture will permit completely new insights to the fracture pro-
cesses and the brittle to ductile transition regime.
5.5. Bending fatigue
In the previous examples, only unidirectional loading was
applied. Fully reversed loading cannot be done by simple
indentation, compression, or tension setups. But taking an
indenter tip equipped with a side slit and well defined con-
tact points in combination with a bending beam as depicted
in Fig. 9a allows to bend the beam in ±x
1
by actuating the
indenter in ± x
3
. In the example provided in Fig. 9 the in-
denter was actuated displacement controlled with a sinusoi-
dal displacement versus time curve at a frequency of
0.033 s
–1
. 100 loading cycles were performed with ampli-
tudes of ± 1 lm, ± 2 lm, ± 4 lm, and ± 8 lm each.
The subsequent in-situ SEM images in Fig. 9b – f show a
single crystal single slip oriented Cu(– 234) beam with di-
mensions of 3 lm·3lm·50 lm in the aligned position
before first loading with an amplitude of ± 1 lm (Fig. 9b).
In Fig. 9c the sample is bent by +1 lm in the maximum up-
ward position, while it is relaxed again in Fig. 9d. Figure 9e
shows the sample deflected by – 1 lm in the chosen maxi-
mum downward position, while it is released after complet-
ing the first loading cycle in Fig. 9f. To bend the sample by
a defined amount, the indenter has to be additionally actu-
ated to compensate for the gap length l
gap
= 6.8 lm between
sample and wedges, as indicated in Fig. 9e. This gap is re-
quired to align sample and tip without previous deforma-
tion. Moreover, the use of a larger gap also allows testing
of beams with larger dimensions using the same tip. The
gap length is also indicated in the load versus displacement
curve in Fig. 9g. After moving the tip without measuring
any load, the wedge contacts the sample and bends it nearly
elastically, as can be seen by the narrow hysteresis between
loading and unloading portion. The position of the in-situ
images is indicated along the loading path.
After cycling the sample a hundred times in the nearly
elastic regime without indication of emerging slip steps
in the in-situ SEM images, the bending amplitude was
increased stepwise to ± 2 lm, ± 4 lm, and ± 8 lm for
100 cycles each. With increasing bending angle, increas-
ing plasticity was observed. This is exemplarily shown in
Fig. 10 for the second loading cycle at an amplitude of
±2 lm. Figure 10a – d again shows the specimen in the
unloaded, loaded, reverse loaded, and finally unloaded po-
sitions. Clear indication for one set of slip lines according
to the single slip orientation is observed in the highly
stressed region. The recorded load versus displacement
curve in Fig. 10e shows, in contrast to the one given in
Fig. 9, a broad hysteresis between loading and unloading
part of the curve indicating a high amount of plastic defor-
mation. The gap length is again indicated. Ex-situ SEM
investigation of the fatigue specimen after cycling with
amplitudes of ± 8 lm shows a high amount of fine slip
traces and necking of the sample in the highly stressed re-
gion close to the base. A detailed analysis of these fatigue
experiments will be described in a forthcoming publica-
tion [86].
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1083
B
Basic
Fig. 9. (a) Low magnifica-
tion overview of the bending
fatigue setup with a single
crystal Cu(– 234) sample with
dimensions of 3 lm·3lm·
50 lm and a tungsten holder
with FIB cut edges during
alignment. (b – f) Subsequent
in-situ SEM images showing
the sample in the first loading
cycle with an amplitude of
±1 lm: Before loading (b), at
maximum upward bending
displacement of +1 lm (c),
after unloading in neutral po-
sition (d), at maximum down-
ward bending displacement of
–1 lm (e), and after the load-
ing cycle in neutral position
(f). (g) Recorded load versus
displacement curve of the
bending cycle depicted above
showing nearly elastic beha-
vior. The positions of images
(b – f) along the loading path
are indicated. The gap length
of l
gap
= 6.8 lm is shown
in (e) and indicated in (g),
the bending length of l
B
=
40.14 lm is indicated in (c).

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
6. Further complementary methods and outlook
Simple monitoring during an in-situ SEM test provides reli-
able alignment [48, 49] in conjunction with valuable in-
sights with respect to, for example, the deformation mode
[52], the acting slip system [34], and the failure mode [22].
However, there are several other analysis techniques avail-
able when using in-situ SEM tests.
6.1. Image correlation
Obviously, the images recorded during the in-situ test can
be used for subsequent analysis. For example, by compar-
ing subsequent SEM images taken during in-situ straining,
a precise measure of the global sample strain as well as the
strain localization on the sample surface can be determined
[87, 88]. The group of Michler [22] facilitated real-time im-
age correlation techniques to improve the accuracy of their
strain measurement. Moreover, the sample sink-in [16, 18,
29, 48], a common problem in micro-compression testing
for samples that are not situated on a rigid substrate [32,
74, 89], can be analyzed and compared to analytical solu-
tions [90].
To further illustrate this issue, a grid with *1lm spa-
cing was written on a single crystal Cu(100) sample with di-
mensions of 8 lm·8lm·12 lm using the FIB with a
Ga
+
ion current of 20 pA (Fig. 11a). This mesh was discre-
tized and is plotted in the uncompressed state (red) and after
compression of Dx
3
=1lm (blue) in Fig. 11b. Besides
movement of the flat punch and displacement of the mesh
positions in the upper half of the pillar, noticeable deforma-
tion is seen below the sample base. Subsequent snapshots
of the deformed grid after compression by Dx
3
=2lm and
Dx
3
=3lm in comparison with the previous state after
compression by Dx
3
=1lm and Dx
3
=2lm are shown in
Fig. 11c and Fig. 11d, respectively. As can be seen by the
detail taken from Fig. 11c shown in Fig. 11e, the deforma-
tion of the sample base is not homogeneous, but resembles
similarities to the slip line field (also known as Prandtl
field).
Evaluating the average sample displacement at the base
position indicated in Fig. 11e leads to sample sink-in displa-
cements of 234 ± 15 nm for displacements from 0 lmto
1lm in Fig. 11b, 188 ± 56 nm for displacements from
1lmto2lm in Fig. 11c, and 217 ± 50 nm for displace-
ments from 2 lmto3lm in Fig. 11d. In contrast, the analyt-
ical Sneddon solution [90] predicts 113 nm, 23 nm, and
27 nm, respectively. These significant differences can be
rationalized by considering some facts. First Sneddon de-
rived this solution for a semi-infinite halfspace penetrated
by a cylindrical punch. Common sample geometries deviate
from this idealized geometry and will behave more softly
[48]. This is especially the case for the sample shown, since
for viewing purposes the base had to be partially removed,
thus resembling more closely a “half-half-space”. This is
only the case for this compression test, but was not the usual
sample configuration. Second, there is a not negligible
amount of plastic deformation taking place beneath the
sample base, leading to hardening of the base material, as
reported previously [29]. Thus, the sample behaves as hav-
ing a larger effective diameter.
Besides correcting for experimental uncertainties, such
image correlation techniques can also be applied to deter-
mine the local strain distribution on the sample surface
[87, 91]. Figure 11f shows an overlay between one of the
cross-correlated in-situ SEM images and the corresponding
local deformation map showing the e
xx
component of the
strain field obtained using MeX 4.0 software (Alicona Ima-
ging GmbH, Krambach, Austria). While the global com-
pressive strain determined from the load versus displace-
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1084 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic
Fig. 10. (a – d) In-situ SEM
images showing the same
specimen as in Fig. 9 during
the second loading cycle with
an amplitude of ± 4 lm after
having sustained 100 loading
cycles with amplitudes of
±1 lm and ± 2 lm: Before
loading (a), at maximum up-
ward bending displacement
of +4 lm (b), at maximum
downward bending displace-
ment of – 4 lm (c), and after
the loading cycle in neutral
position (d). Note the clearly
visible slip steps in the highly
stressed region. (e) Recorded
load versus displacement
curve of the bending cycle de-
picted above showing clearly
plastic behavior. The posi-
tions of images (a – d) along
the loading path are indicated.
(f) Inclined ex-situ SEM im-
age showing the fatigue sam-
ple after it was additionally
cycled 100 times with even
higher amplitudes of ± 8 lm.
A large number of fine slip
steps as well as indication of
necking are observed.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
ment curve was 0.06 between the two images, only at white
spots in Fig. 11f a compressive strain higher than 0.05 was
reached. Compressive strains of *0.05 are observed at an
inclination of *458corresponding to the expected slip
plane. This is of special interest, since this allows determin-
ing strain localization and active slip systems even before
slip steps are observed in the in-situ images. Moreover,
again localized deformation at the sample top is shown
[34, 60]. In accordance with the previous discussion, the
image cross-correlation also shows a significant amount of
deformation beneath the sample base, as indicated by an ar-
row in Fig. 11f.
6.2. Electron backscatter diffraction
In-situ SEM images show the actual deformation of the
samples with the active slip systems, which can be extended
as shown above by post image correlation techniques. A
striking feature of in-situ micro-Laue techniques [59, 60]
is the capability to locally monitor changes in the crystallo-
graphic structure of the sample during testing. While micro-
Laue requires use of a synchrotron beamline, similar infor-
mation can be gained using electron backscatter diffraction
(EBSD) techniques in an SEM [92]. It has already been
shown that the data gained by in-situ micro-Laue X-ray dif-
fraction and ex-situ EBSD investigation of the identical mi-
crometer sized single crystal specimen [60, 93] are in excel-
lent agreement and complement each other with respect to
lateral and angular resolution. While ex-situ EBSD has
been demonstrated for miniaturized compression [34, 60],
bending [27], and tensile [28, 93] samples, in-situ EBSD re-
quires tilting of the whole testing setup by typically 658–
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1085
B
Basic
Fig. 11. (a) 8 lm·8lm·
12 lm single crystal Cu(100)
compression sample with a
FIB made grid having
*1lm spacing. (b) Overlay
of the discretized meshes of
the unloaded sample (red)
and after Dx
3
=1lm com-
pression (blue). (c) Compari-
son of the mesh after Dx
3
=
1lm (red) and Dx
3
=2lm
(blue) deformation. (d) The
deformed grid at Dx
3
=2lm
(red) and Dx
3
=3lm (blue)
compression. (e) Detail of the
deformed sample base from
(c). (f) Overlay between in-
situ SEM image and the cor-
responding local strain map-
ping showing the e
xx
strain
component. Strain along the
expected 458inclined slip
plane, at the sample top, and
sink-in at the sample base (ar-
row) were observed.
Fig. 12. Application window for the existing in-situ SEM indenter
used in this study as a function of sample dimension and flow stress
of the material. Possible extension towards nanomechanics by using a
more sensitive load cell is indicated by arrows.

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
708around t
1
(see Fig. 2). Therefore, to load the sample
without blocking the electron beam, the indenter needs to
be mounted on the back of the SEM chamber to align the
x
3
direction with the y
2
direction (Fig. 2). This is not possi-
ble for the present SEM chamber, since no flange exists at
this position, but can be realized when designing a new
setup.
7. Towards in-situ nanomechanics in an SEM
Through the foregoing examples it was shown that, with the
exception of torsional loading, all common mechanical test-
ing techniques have been successfully downscaled to the
micrometer regime by the combination of a FIB worksta-
tion for fabrication issues and the use of a single indenter
system equipped with various custom tips and installed in
an SEM for in-situ loading of the samples. This unique
combination offers large flexibility in combination with a
minimum amount of instruments, thus being very efficient.
However, as for any setup, there is a typical application
window, as shown in Fig. 12, given by the load and displa-
cement range of the setup in conjunction with the flow
stress and dimensions of the sample to be tested. In our pres-
ent case, the maximum load is 300 mN and the maximum
displacement 30 lm. The lower limits are given by the ex-
perimental noise, which is below 10 lN (see for example
Fig. 10e) and better than ± 5 nm for the indenter mounted
in the SEM without sample contact. This is also about the
resolution provided by the SEM used. This allows covering
a large area of what is commonly attributed to as micro-me-
chanics.
Moreover, there is no principal restriction to extending
these techniques into the nanometer regime. FIB fabrication
of samples with dimensions of several tens of nanometers
can be achieved. However, the increasing fraction of FIB
modified material surface (ion damage) with respect to the
total tested volume might set a limit to this fabrication
method [63]. Nevertheless, there is a growing field of nano-
materials including nanowires and whiskers where mechan-
ical properties need to be determined. Load cells with a
compact design and higher resolution to extend testing to-
wards lower loads have become commercially available
[53, 94]. The possible extension of the operation window
using such load cells is indicated by arrows in Fig. 12. The
compact design reduces the masses in the SEM chamber
that need to be actively damped and allows operating at a
lower working distance in the SEM, thus gaining better re-
solution. Finally, using a field emission SEM or a dual
beam FIB workstation as an all-in-one tool increases the
imaging resolution for reliable strain determination of nan-
ometer sized samples.
The authors thank Dr. Thomas Chudoba and his co-workers for the ex-
cellent support during hardware and software adaptation of the in-situ
micro-indenter in order to meet the special requirements for the per-
formed work. Partial financial support by the Austrian Fonds zur För-
derung der wissenschaftlichen Forschung, Project P17375-N07, and
within the research activities of the K2 Competence Centre on “Inte-
grated Research in Materials, Processing and Product Engineering”,
operated by the Materials Center Leoben Forschung GmbH under the
frame of the Austrian COMET Competence Centre Program, is ac-
knowledged. D.K. gratefully acknowledges financial support by the
Austrian Science Fund (FWF) through the Erwin Schrödinger scholar-
ship J2834-N20.
References
[1] E.O. Hall: Proc. Phys. Soc. Lond. B 64 (1951) 747.
[2] N.J. Petch: J. Iron Steel 174 (1953) 25.
[3] E. Arzt: Acta Mater. 46 (1998) 5611.
DOI:10.1016/S1359-6454(98)00231-6
[4] J.S. Stölken, A.G. Evans: Acta Mater. 46 (1998) 5109.
DOI:10.1016/S1359-6454(98)00153-0
[5] N.A. Fleck, G.M. Muller, M.F. Ashby, J.W. Hutchinson: Acta Met.
Mater. 42 (1994) 475. DOI:10.1016/0956-7151(94)90502-9
[6] N.A. Stelmashenko, M.G. Walls, L.M. Brown, Y.V. Milman: Acta
Met. Mater. 41 (1993) 2855.
DOI:10.1016/0956-7151(93)90100-7
[7] W.D. Nix, H. Gao: J. Mech. Phys. Solids 46 (1998) 411.
DOI:10.1016/S0022-5096(97)00086-0
[8] M.A. Haque, M.T.A. Saif: Experim. Mech. 43 (2003) 248.
DOI:10.1007/BF02410523
[9] A.M. Minor, S.A.S. Asif, Z.W. Shan, E.A. Stach, E. Cyrankowski,
T.J. Wyrobek, O.L. Warren: Nat. Mater. 5 (2006) 697.
DOI:10.1038/nmat1714
10] M.J. Cordill, M.D. Chambers, M.S. Lund, D.M. Hallman, C.R. Per-
rey, C.B. Carter, A. Bapat, U. Kortshagen, W.W. Gerberich: Acta
Mater. 54 (2006) 4515.
DOI:10.1016/j.actamat.2006.05.037
[11] P.A. Gruber, C. Solenthaler, E. Arzt, R. Spolenak: Acta Mater. 56
(2008) 1876. DOI:10.1016/j.actamat.2007.12.043
[12] M.W. Phaneuf: Micron 30 (1999) 277.
DOI:10.1016/S0968-4328(99)00012-8
[13] M.D. Uchic, D.M. Dimiduk, J.N. Florando, W.D. Nix: Science 305
(2004) 986. DOI:10.1126/science.1098993
[14] D.M. Dimiduk, M.D. Uchic, T.A. Parthasarathy: Acta Mater. 53
(2005) 4065. DOI:10.1016/j.actamat.2005.05.023
[15] M.D. Uchic, D.M. Dimiduk: Mater. Sci. Eng. A 400 – 401 (2005)
268. DOI:10.1016/j.msea.2005.03.082
[16] J.R. Greer, W.C. Oliver, W.D. Nix: Acta Mater. 53 (2005) 1821.
DOI:10.1016/j.actamat.2004.12.031
[17] J.R. Greer, W.C. Oliver, W.D. Nix: Acta Mater. 54 (2006) 1705
[Corrigendum]. DOI:10.1016/j.actamat.2005.12.004
[18] C.A. Volkert, E.T. Lilleodden: Phil. Mag. 86 (2006) 5567.
DOI:10.1080/14786430600567739
[19] R. Maaß, D. Grolimund, S. Van Petegem, M. Willimann, M. Jensen,
H. Van Swygenhoven, T. Lehnert, M.A.M. Gijs, C.A. Volkert, E.T.
Lilleodden, R. Schwaiger: App. Phys. Lett. 89 (2006) 151905.
DOI:10.1063/1.2358204
[20] D. Kiener, C. Motz, T. Schöberl, M. Jenko, G. Dehm: Adv. Eng. Ma-
ter. 8 (2006) 1119. DOI:10.1002/adem.200600129
[21] J. Michler, K. Wasmer, S. Meier, F. Ostlund, K. Leifer: App. Phys.
Lett. 90 (2007) 043123. DOI:10.1063/1.2432277
[22] B. Moser, K. Wasmer, L. Barbieri, J. Michler: J. Mater. Res. 22
(2007) 1004. DOI:10.1557/JMR.2007.0140
[23] C.P. Frick, S. Orso, E. Arzt: Acta Mater. 55 (2007) 3845.
DOI:10.1016/j.actamat.2007.02.034
[24] K.S. Ng, A.H.W. Ngan: Acta Mater. 56 (2008) 1412.
DOI:10.1016/j.actamat.2007.12.016
[25] Z.W. Shan, R.K. Mishra, S.A.S. Asif, O.L. Warren, A.M. Minor:
Nat. Mater. 73 (2008) 115. DOI:10.1038/nmat2085
[26] B.E. Schuster, Q. Wei, T.C. Hufnagel, K.T. Ramesh: Acta Mater. 56
(2008) 5091. DOI:10.1016/j.actamat.2008.06.028
[27] C. Motz, T. Schöberl, R. Pippan: Acta Mater. 53 (2005) 4269.
DOI:10.1016/j.actamat.2005.05.036
[28] D. Kiener, W. Grosinger, G. Dehm, R. Pippan: Acta Mater. 56
(2008) 580. DOI:10.1016/j.actamat.2007.10.015
[29] C.P. Frick, B.G. Clark, S. Orso, A.S. Schneider, E. Arzt: Mater. Sci.
Eng. A 489 (2008) 319. DOI:10.1016/j.msea.2007.12.038
[30] D.M. Dimiduk, C. Woodward, R. LeSar, M.D. Uchic: Science 312
(2006) 1188. DOI:10.1126/science.1123889
[31] F.F. Csikor, C. Motz, D. Weygand, M. Zaiser, S. Zapperi: Science
318 (2007) 251. DOI:10.1126/science.1143719
[32] D. Kiener, C. Motz, G. Dehm: Mater. Sci. Eng. A 505 (2009) 79.
DOI:10.1016/j.msea.2009.01.005
[33] O.L. Warren, S.A. Downs, T.J. Wyrobek: Z. Metallkd. 95 (2004)
287. http://www.ijmr.de/directlink.asp?MK017951
[34] D. Kiener, C. Motz, G. Dehm: J. Mater. Sci. 43 (2008) 2503.
DOI:10.1007/s10853-008-2531-3
[35] D. Kiener, W. Grosinger, G. Dehm: Scripta Mater. 60 (2009) 148.
DOI:10.1016/j.scriptamat.2008.09.024
[36] J.R. Greer, C.R. Weinberger, W. Cai: Mater. Sci. Eng. A 493 (2008)
21. DOI:10.1016/j.msea.2007.08.093
[37] R. Dou, B. Derby: Scripta Mater. In Press, Accepted Manuscript).
DOI:10.1016/j.scriptamat.2009.05.012
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
1086 Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8
B
Basic

W2009CarlHanserVerlag, Munich, Germanywww.ijmr.de Notforuseininternetorintranetsites. Notforelectronic distribution.
[38] T.A. Parthasarathy, S.I. Rao, D.M. Dimiduk, M.D. Uchic, D.R. Trin-
kle: Scripta Mater. 56 (2007) 313.
DOI:10.1016/j.scriptamat.2006.09.016
[39]
S.I. Rao, D.M. Dimiduk, M. Tang, T.A. Parthasarathy, M.D. Uchic,
C. Woodward: Phil. Mag. 87 (2007) 4777.
DOI:10.1080/14786430701591513
[40] B. von Blanckenhagen, E. Arzt, P. Gumbsch: Acta Mater. 52 (2004)
773. DOI:10.1016/j.actamat.2003.10.022
[41] J.A. El-Awady, M. Wen, N.M. Ghoniem: J. Mech. Phys. Solids 57
(2009) 32. DOI:10.1016/j.jmps.2008.10.004
[42] S.H. Oh, M. Legros, D. Kiener, G. Dehm: Nat. Mater. 8 (2009) 95.
DOI:10.1038/nmat2370
[43] S.I. Rao, D.M. Dimiduk, T.A. Parthasarathy, M.D. Uchic, M. Tang,
C. Woodward: Acta Mater. 56 (2008) 3245.
DOI:10.1016/j.actamat.2008.03.011
[44] A.S. Schneider, B.G. Clark, C.P. Frick, P.A. Gruber, E. Arzt: Mater.
Sci. Eng. A 508 (2009) 241. DOI:10.1016/j.msea.2009.01.011
[45] P.J. Guruprasad, A.A. Benzerga: J. Mech. Phys. Solids 56 (2008)
132. DOI:10.1016/j.jmps.2007.03.009
[46] S. Buzzi, M. Dietiker, K. Kunze, R. Spolenak, J.F. Loffler: Phil.
Mag. 89 (2009) 869. DOI:10.1080/14786430902791748
[47] M.D. Uchic, P.A. Shade, D. Dimiduk: Ann. Rev. Mater. Res. 39
(2009). DOI:10.1146/annurev-matsci-082908-145422
[48] H. Zhang, B.E. Schuster, Q. Wei, K.T. Ramesh: Scripta Mater. 54
(2006) 181. DOI:10.1016/j.scriptamat.2005.06.043
[49] Y.S. Choi, M.D. Uchic, T.A. Parthasarathy, D.M. Dimiduk: Scripta
Mater. 57 (2007) 849. DOI:10.1016/j.scriptamat.2007.06.057
[50] M.D. Uchic, D.M. Dimiduk, R. Wheeler, P.A. Shade, H.L. Fraser:
Scripta Mater. 54 (2006) 759.
DOI:10.1016/j.scriptamat.2005.11.016
[51] R. Rabe, J.M. Breguet, P. Schwaller, S. Stauss, F.J. Haug,
J. Patscheider, J. Michler: Thin Solid Films 469-470 (2004).
DOI:10.1016/j.tsf.2004.08.096
[52] B. Moser, J.F. Loffler, J. Michler: Phil. Mag. A 86 (2006) 5715.
DOI:10.1080/14786430600627301
[53] K.A. Rzepiejewska-Malyska, G. Buerki, J. Michler, R.C. Major, E.
Cyrankowski, S.A.S. Asif, O.L. Warren: J. Mater. Res. 23 (2008)
1973. DOI:10.1557/jmr.2008.0240
[54] K. Svensson, Y. Jompol, H. Olin, E. Olsson: Rev. Sci. Instrum. 74
(2003) 4945. DOI:10.1063/1.1614872
[55] M.S. Bobji, C.S. Ramanujan, J.B. Pethica, B.J. Inkson: Meas. Sci.
Technol. 17 (2006) 1324. DOI:10.1088/0957-0233/17/6/006
[56] I.M. Robertson, P.J. Ferreira, G. Dehm, R. Hull, E.A. Stach: MRS
Bull. 33 (2008) 122.
[57] J. Ye, R.K. Mishra, A.M. Minor: Scripta Mater. 59 (2008) 951.
DOI:10.1016/j.scriptamat.2008.06.052
[58] R. Maaß, S.V. Petegem, J. Zimmermann, C.N. Borca, H.V. Swygen-
hoven: Scripta Mater. 59 (2008) 471.
DOI:10.1016/j.scriptamat.2008.04.034
[59] R. Maaß, S. Van Petegem, H. Van Swygenhoven, P.M. Derlet, C.A.
Volkert, D. Grolimund: Phys. Rev. Lett. 99 (2007) 145505.
DOI:10.1103/PhysRevLett.99.145505
[60] R. Maaß, S. Van Petegem, D. Grolimund, H. Van Swygenhoven, D.
Kiener, G. Dehm: App. Phys. Lett. 92 (2008) 071905.
DOI:10.1063/1.2884688
[61] K. Wasmer, T. Wermelinger, A. Bidiville, R. Spolenak, J. Michler:
J. Mater. Res. 23 (2008) 3040. DOI:10.1557/jmr.2008.0363
[62] H. Bei, S. Shim, G.M. Pharr, E.P. George: Acta Mater. 56 (2008)
4762. DOI:10.1016/j.actamat.2008.05.030
[63] D. Kiener, C. Motz, M. Rester, G. Dehm: Mater. Sci. Eng. A 459
(2007) 262. DOI:10.1016/j.msea.2007.01.046
[64] L.A. Giannuzzi, F.A. Stevie: Introduction to focused ion beams,
Springer, New York (2005).
[65] C.A. Volkert, A.M. Minor: MRS Bull. 32 (2007) 389.
[66] H. Bei, S. Shim, E.P. George, M.K. Miller, E.G. Herbert,
G.M. Pharr: Scripta Mater. 57 (2007) 397.
DOI:10.1016/j.scriptamat.2007.05.010
[67] S. Shim, H. Bei, M.K. Miller, G.M. Pharr, E.P. George: Acta Mater.
57 (2009) 503. DOI:10.1016/j.actamat.2008.09.033
[68] H. Bei, S. Shim, M.K. Miller, G.M. Pharr, E.P. George: App. Phys.
Lett. 91 (2007) 111915. DOI:10.1063/1.2784948
[69] J. Marien, J.M. Plitzko, R. Spolenak, R.-M. Keller, J. Mayer: J. Mi-
croscopy 194 (1999) 71.
[70] S. Schmidt, W. Sigle, W. Gust, M. Rühle: Z. Metallkd. 93 (2002)
428. http://www.ijmr.de/directlink.asp?MK020428
[71] W. Sigle, G. Richter, M. Ruhle, S. Schmidt: App. Phys. Lett. 89
(2006) 121911. DOI:10.1063/1.2356322
[72] T. Chudoba, P. Schwaller, R. Rabe, J.M. Breguet, J. Michler: Phil.
Mag. A 86 (2006) 5265. DOI:10.1080/14786430600746424
[73] S. Wurster, R. Pippan: Scripta Mater. 60 (2009) 1083.
DOI:10.1016/j.scriptamat.2009.01.011
[74] H. Wörgötter, D. Kiener, J.M. Purswani, D. Gall, G. Dehm: Berg-
und Hüttenmännische Monatshefte 153 (2008) 257.
DOI:10.1007/s00501-008-0386-1
[75] D. Tabor: Sheet Met. Ind. 31 (1954) 749.
[76] D.M. Marsh: Proc. R. Soc. Lond. A 279 (1964) 420.
DOI:10.1098/rspa.1964.0114
[77] P. Hosemann, J.G. Swadener, D. Kiener, G.S. Was, S.A. Maloy,
N. Li: J. Nuc. Mater. 375 (2008) 135.
DOI:10.1016/j.jnucmat.2007.11.004
[78] R. Pippan, F. Wetscher, M. Hafok, A. Vorhauer, I. Sabirov: Adv.
Eng. Mater. 8 (2006) 1046. DOI:10.1002/adem.200600133
[79] D. Kiener, M. Rester, S. Scheriau, B. Yang, R. Pippan, G. Dehm: Int.
J. Mat. Res. 98 (2007) 1047. DOI:10.3139/146.101578
[80] G. Khatibi, A. Betzwar-Kotas, V. Gröger, B. Weiss: Fatigue Fract.
Eng. Mater. Struct. 28 (2005) 723.
DOI:10.1111/j.1460-2695.2005.00898.x
[81] B. Yang, C. Motz, W. Grosinger, W. Kammrath, G. Dehm: Int. J.
Mat. Res. 99 (2008) 716. DOI:10.3139/146.101690
[82] K.S. Ng, A.H.W. Ngan: Scripta Mater. 59 (2008) 796.
DOI:10.1016/j.scriptamat.2008.06.019
[83] C. Motz, D. Weygand, J. Senger, P. Gumbsch: Acta Mater. 56
(2008) 1942. DOI:10.1016/j.actamat.2007.12.053
[84] P. Gumbsch: J. Nuc. Mater. 323 (2003) 304.
DOI:10.1016/j.jnucmat.2003.08.009
[85] M. Faleschini, H. Kreuzer, D. Kiener, R. Pippan: J. Nuc. Mater.
367 – 370 (2007) 800. DOI:10.1016/j.jnucmat.2007.03.079
[86] D. Kiener, C. Motz, D. Weygand, R. Pippan: under prep. (2009).
[87] A. Tatschl, O. Kolednik: Mater. Sci. Eng. A 339 (2003) 265.
DOI:10.1016/S0921-5093(02)00111-9
[88] A. Tatschl, C. Motz, R. Pippan, O. Kolednik, in: T.W. Clyne, F. Si-
mancik (Eds.), The Characterization of Crystal Plasticity by Orienta-
tion Image Microscopy and Local Deformation Measurements. Met-
al Matrix Composites and Metallic Foams. 2005. p. 151.
[89] A.S. Budiman, S.M. Han, J.R. Greer, N. Tamura, J.R. Patel,
W.D. Nix: Acta Mater. 56 (2008) 602.
DOI:10.1016/j.actamat.2007.10.031
[90] I.N. Sneddon: Int. J. Eng. Sci. 3 (1965) 47.
DOI:10.1016/0020-7225(65)90019-4
[91] O. Kolednik: Adv. Eng. Mater. 8 (2006) 1079.
DOI:10.1002/adem.200600150
[92] V. Randle, O. Engler: Introduction to Texture Analysis: macrotex-
ture, microtexture and orientation mapping, Gordon & Breach, Am-
sterdam (2000).
[93] C. Kirchlechner, D. Kiener, N. Vaxelaire, J.-S. Micha, O. Ulrich,
O. Thomas, G. Dehm, J. Keckes: under prep. (2009).
[94] D.S. Gianola, C. Eberl: JOM 61 (2009) 24.
DOI:10.1007/s11837-009-0037-3
(Received January 19, 2009; accepted June 3, 2009)
Bibliography
DOI 10.3139/146.110149
Int. J. Mat. Res. (formerly Z. Metallkd.)
100 (2009) 8; page 1074 – 1087
#Carl Hanser Verlag GmbH & Co. KG
ISSN 1862-5282
Correspondence address
Dr. Daniel Kiener
National Center for Electron Microscopy
Lawrence Berkeley National Laboratory
One Cyclotron Road, MS 72-150, Berkeley, CA 94720, U.S.A.
Tel.: +1 510 486 4614
Fax: +1 510 486 5888
E-mail: dkiener@lbl.gov
You will find the article and additional material by enter-
ing the document number MK110149 on our website at
www.ijmr.de
D. Kiener et al.: Overview on established and novel FIB based miniaturized mechanical testing using in-situ SEM
Int. J. Mat. Res. (formerly Z. Metallkd.) 100 (2009) 8 1087
B
Basic