Anda di halaman 1dari 11

Microscopy 62(1): 193203 (2013)

doi: 10.1093/jmicro/dfs096
........................................................................................................................................................................................................................................................

Review

Atomic-resolution environmental TEM for quantitative


in-situ microscopy in materials science
Nanoscience and Nanotechnology Center, Institute of Scientific and Industrial Research (ISIR),
Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan
*To whom correspondence should be addressed. E-mail: takeda@sanken.osaka-u.ac.jp
..............................................................................................................................................................................................

Abstract

We have compiled our recent in-situ quantitative environmental transmission electron microscopy (ETEM) studies on typical gold nanoparticulate catalysts for the low-temperature oxidation of CO to describe the
issues surrounding the application of ETEM, with a special regard to
catalyst chemistry. Thanks to the recent development of high-resolution
environmental transmission electron microscopes that can work robustly
to accumulate observation data in controlled environments, we can deal
with the electron irradiation effects and heterogeneity of real catalysts.
We established a structural evolution diagram that summarizes the structure of catalysts under electron irradiation as a function of the electron
current density and the electron dose, D. By extrapolating to = 0,
D = 0, we could deduce the intrinsic catalysis structure (without electron
irradiation) in various environments, including reaction environments. By
numerically and statistically analyzing a substantial number of ETEM
images of gold nanoparticles, we established a morphology phase
diagram that summarizes how the majority of gold nanoparticles change
their morphology systematically as a function of the partial pressures of
CO and O2. Similar diagrams will be helpful in elucidating the phenomena
that directly correlate with the catalytic activity determined from ETEM
observations. Using these quantitative analyses, we could analyze Cs-corrected ETEM images of the catalysts. The surfaces of gold nanoparticles
were structurally reconstructed under reaction conditions, via interactions with CO molecules. CO molecules were observed on the surfaces of
catalysts under reaction conditions using high-resolution ETEM. Finally,
we discuss the potential of environmental transmission electron microscopy for quantitative in-situ microscopy at the atomic scale.

..............................................................................................................................................................................................

Keywords

environmental TEM, in-situ


Cs corrector, gold catalyst

TEM,

catalyst,

electron

irradiation,

..............................................................................................................................................................................................

Received

31 October 2012, accepted 1 December 2012; online 16 January 2013

..............................................................................................................................................................................................

Introduction
Since the invention of the transmission electron
microscope, several branches of transmission electron microscopy (TEM) have nucleated and grown.
Along with the essential branch that has pursued
higher spatial resolution, several pioneers have
created the invaluable branch of TEM that involves

the observation of both the interior and the surface


of materials in real space and time, at magnications much higher than those achievable using
optical microscopes. This in-situ TEM has advanced
greatly with the development of TEM technology.
Earlier in-situ studies were based on the
diffraction-contrast imaging method [14]. After

........................................................................................................................................................................................................................................................

The Author 2013. Published by Oxford University Press [on behalf of The Japanese Society of Microscopy]. All rights reserved.
For permissions, please e-mail: journals.permissions@oup.com

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

Seiji Takeda* and Hideto Yoshida

194

M I C R O S C O P Y , Vol. 62, No. 1, 2013

The development of ETEM occurred rather


slowly [11] up to the year 2000 (approximately),
when compared with the progress in electron
optics and peripheral instruments for chemical analyses in TEM. Nevertheless, several signicant
attempts were made to achieve higher resolution in
gases, using different designs [2335], and attempts
have been made recently for liquids. These studies
signaled the potential of ETEM to be useful in
chemical sciences and technology, to see what is
happening in reaction environments. By observing
the nucleation and growth of whiskers, or nanowires and nanotubes, in controlled environments
such as those present during chemical vapor deposition and vaporliquidsolid phase growth, several
researchers have derived clues from ETEM observations that help us to understand the synthetic processes [3640]. In catalyst science and technology,
pioneering works have already attempted to correlate structural data measured using TEM with the
catalytic activity determined by catalysis chemical
measurements [41]. Recently, the distribution of
metal atoms and their clusters has been correlated
with catalytic activity, using high resolution STEM
[42] and electron tomography [43]. ETEM has provided initial results showing that metal catalysis
particles change their morphology in specic gases
[44]. Careful ETEM analysis led to the proposal of
the mechanisms of catalysis [45]. Regarding the formation process of ETEM images in gases, the rst
theoretical study [46,47] showed that we could
observe atoms in a specimen, even in gases. Over
the past 8 years, we have been working on
atomic-resolution ETEM for gas environments for
materials science. In this paper, we compile our
recent applications of ETEM to metal nanoparticulate catalysts.

Toward quantitative in-situ microscopy


at the atomic scale in catalysis chemistry
There are several issues to address regarding the
application of ETEM to catalyst chemistry. Some of
the issues are shared with other applications of
ETEM, including the design of robust ETEMs for
high-resolution observations in well-controlled gas
environments. It is noteworthy that quantitative
in-situ TEM could determine basic quantities of

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

Iijima [5] and Cowley [6] established that crystalline


lattices could be observed directly using highresolution TEM, in-situ TEM moved forward to an
era of atomic-scale observations. The studies in this
era revealed the cooperative motions of atoms, for
instance, on surfaces [7,8] and in crystalline materials [9,10] in vacuum. The experimental window of
in-situ TEM through which we can observe
dynamic phenomena in materials was further
extended from vacuum to gas environments by environmental TEM (ETEM) [11]. The denitive, pioneering work in this eld was published by
Hashimoto et al. [12] in 1968; they observed the
growth of needles of zinc oxide at elevated temperatures in oxygen gases in situ, in collaboration
with JEOL. These pioneering TEM and ETEM
studies truly remind us that an act of observation
can be regarded as an experiment in which the
result of the observation constitutes the outcome of
the experiments. Therefore, we need make no distinction between experiments and observations [13].
With improvements in the spatial resolution of
TEM, the atomic positions in localized structures in
a crystalline specimen could be determined.
Attempts were made to determine static atomic
arrangements in grain boundaries [14,15], interfaces
[16,17], the core of dislocations [18], surfaces [19],
extended defects [20], and even in agglomerates of
hydrogen atoms [21] in metals, semiconductors and
ceramics. For instance, an unknown structure of
nanometer size could be determined using TEM,
without any previous knowledge from other experimental techniques [20], and the structural data from
TEM were useful in elucidating the mechanism of
the material processes [22]. In addition, the analysis
of TEM images was evaluated and rened using
computational techniques based on the theories of
solid state physics and chemistry [20]. The combination of TEM and ab initio computation has
become a standard methodology in analyzing the
structures of interfaces, surfaces and defects in
crystalline materials. With aberration correction
technology, we can now precisely determine the arrangement of individual atomic columns of different
chemical species in vacuum; this is possible even in
dynamic systems, thanks to the temporal resolution
available in both TEM and scanning transmission
electron microscopy (STEM).

S. Takeda and H. Yoshida Quantitative atomic-resolution in-situ ETEM

accumulate observation data is essential. To


achieve higher spatial resolution in ETEM, the path
length of electrons in gases must be shortened
[46,47]. The equiphase surface of electron waves is
disturbed with random noise at the entrance
surface of a specimen because of electron scattering by gas molecules above the specimen. The coherence of the incident electron wave is, therefore,
degraded. After passing through the specimen, the
electron wave is further scattered by gas molecules.
This smears the amplitude and phase of the electron wave, properties that correspond to the electrostatic potential of the atoms in the specimen. In
our specially designed E-cells, which were developed in collaboration with FEI Company, the thickness of the gas layer between the apertures can be
smaller than the distance between the pair of pole
pieces, with plenty of room provided around the
specimen to allow the installation of apparatus for
additional measurements and material syntheses.
Furthermore, conventional specimen holders for
TEM can be used in the ETEM mode. A cylinder
with a ap to manually change the observation
mode from ETEM to TEM (and vice versa) is indicated by an arrowhead in Fig. 1a. We can also
change the ratio of the path lengths above and
below a specimen, to distinguish between the two
effects on the electron wave mentioned above. An
ETEM can be used as a conventional highresolution and analytical TEM for high-angle
annular dark eld (HAADF)-STEM, electron energy

Design of robust ETEM for high-spatial


resolution in well-controlled
environments
High-spatial and time resolutions are necessary preconditions for the application of ETEM to catalyst
chemistry, but they are not sufcient to ensure
success. An ETEM that works robustly to

Fig. 1. Views of the outside of the specially designed specimen


chambers (environmental cells) in a (a) 200 kV ETEM (Installed in
December 2005, Osaka University) and a (b) 300 kV ETEM with a
spherical aberration corrector for the objective lens (Installed in
March 2010, Osaka University).

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

objects in vacuum that cannot currently be


observed directly. For instance, the activation energies for the thermal migration of intrinsic point
defects such as lattice vacancies in pure crystalline
metals were determined experimentally by accumulating a substantial number of dark-eld TEM
images under different experimental conditions, although the spatial resolution of the dark-eld
images was not the best at that time [48]. This type
of study showed that quantitative TEM could go
beyond seeing is believing type experiments.
ETEM is another microscopic technique that can be
used not only for recording beautiful pictures and
movies of materials in gases, but also for providing
quantitative in-situ microcopy information for
materials science. Given a well-designed ETEM,
one could further develop the methodology for the
quantitative in-situ microscopy of materials and
devices operating in gases, as is described in the
following subsections.
In this review, we describe studies examining
gold nanoparticulate catalysts. As it is well known,
Haruta [49] discovered over two decades ago that
gold the most stable metallic element shows remarkable catalytic activity for CO oxidation, even
below room temperature. Unlike platinum and palladium, gold must be supported in the form of
nanoparticles on crystalline metal oxides such as
TiO2 and CeO2. The mechanism of catalysis by gold
nanoparticles remains unclear. We studied the gold
catalysts Au/CeO2 and Au/TiO2; they were prepared
using the deposition precipitation method, and
exhibited high catalytic activity at room temperature. The conversion of CO to CO2 reached 100% at
room temperature. Details of the catalyst sample
and the ETEM analyses have been described (in
terms of the catalyst chemistry) in our previous
papers [50,51]. Here, we describe how we addressed
the issues of ETEM from a microscopy perspective.

195

196

M I C R O S C O P Y , Vol. 62, No. 1, 2013

Electron irradiation effects


A system under ETEM observation consists of a catalysis specimen and gas molecules. They are irradiated heavily by electrons of high energy (with
energies from keV to MeV). It is well known that

high-energy electrons transfer their energy and momentum and induce various elemental excitations
in the system, for example the excitation of plasmons and core electrons and the displacement of
atoms (knock-on damage). The cross sections of
electron scattering are different for each type of excitation. Although some excitations thermally relax
instantly, the system gradually undergoes irreversible processes with increases in the electron dose
and never returns to the initial state. In contrast, in
an ordinary catalytic reaction, the energy transfer
between a catalysis specimen and the gas molecules is far smaller than a few electron volt.
Therefore, electron irradiation effects cannot be
ignored in quantitative in-situ microscopy for catalyst chemistry.
The essential experimental quantities to describe
electron irradiation effects are the incident electron
energy, the electron current density and the electron dose D (D = ), where is the irradiation
time. Other important quantities are related to the
specimen and its environment; this includes the
temperature (as far as it is denable), the pressures
of the gases and the orientation of the crystalline
axes of a specimen in relation to an incident electron beam. The rst step toward quantifying the
electron irradiation effects is to systematically
measure a quantity of interest as a function of both
and D. By extrapolating the value at = 0 and
D = 0 using the systematic measurements, it is possible to deduce the intrinsic quantity (without electron irradiation). This extrapolation method can be
used to derive the structural changes produced by
the impact of even a single incident electron [58].
It is known that Au/TiO2 catalysts which are the
most commonly used gold nanoparticulate catalysts
are fragile under TEM observation. The structures
at the interface between the gold nanoparticles and
the TiO2 supports gradually become reorganized as
the observation time or electron dose is increased
[51]. Figure 2 summarizes the structures observed
in Au/TiO2 catalysts using ETEM, in different environments produced as a function of and D. In this
structural evolution diagram [51], gold nanoparticles
with decorations and pillars were observed in the
blue and red regions, respectively. Either gold nanoparticles with pillars or encapsulated gold nanoparticles were observed in the yellow regions. A blue,

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

loss spectroscopy (EELS) and/or energy dispersive


X-ray (EDX) chemical analysis as well as TEM.
Although some of the specication and design of
the special E-cell is presented here [33,34], the
details of the design will be described elsewhere.
The rst prototype ETEM of this design was
developed based on a 200-kV TEM equipped with a
eld emission gun (FEI Tecnai F20) and began
operations in Osaka University in December, 2005
(Fig. 1a). The ETEM was used for high-resolution
in-situ observations of the nucleation and growth
of carbon nanotubes (CNTs) in source gases at
high temperatures [39,5254]. The atomistic structures of catalytic nanoparticles were determined
in-situ during the growth of CNTs. Thanks to the
efforts of FEI Company, the gas pressure exceeded
2500 Pa (N2 gas) during stable ETEM observations,
and all the operations of the ETEM vacuum system
could be simply controlled using a single operation
box. To analyze the atomic arrangements on the
surfaces and interfaces of solids in gases, we have
recently (from March 2010, in Osaka University)
started using a newly developed ETEM (FEI Titan
ETEM G2) that has an objective lens with a corrector for spherical aberration (Cs corrector), as shown
in Fig. 1b. A Cs corrector is essential in an ETEM
[46,5557] for any determination of structure. All
the special features of the prototype 200 kV ETEM
(Fig. 1a) were transferred to the Cs-corrected
ETEM (Fig. 1b). The information limit even at
200 kV, in gases with pressures above 1000 Pa
reached 0.12 nm. Importantly, the base systems of
the ETEM were improved signicantly by FEI
Company, including the gas injection and evacuation systems. The software-controlled operation
system for the ETEM allows us to acquire substantial, systematic and reproducible high-resolution
ETEM data more easily. These features are useful
in various scientic, technological and industrial
elds, including catalysis chemistry. The rst issue
has, therefore, now been addressed.

S. Takeda and H. Yoshida Quantitative atomic-resolution in-situ ETEM

197

Fig. 2. Structure evolution diagrams for Au/TiO2 catalysts under


electron irradiation, as a function of dose, D and electron current
density , in (a) vacuum, (b) O2 at 100 Pa and (c) 1 vol% CO in air at
100 Pa, at room temperature [51]. (Reproduced with permission
from John Wiley and sons).

red or yellow triangle indicates the D at which decoration, pillars or encapsulation around a gold
nanoparticle were rst detected at a given current
density. HAADF-STEM, EELS and EDX analyses
conrmed that the decoration, pillar and encapsulation structures consisted of the TiO2 support material [51]. Under vacuum, the TiO2 supports were
heavily damaged and lost oxygen, as shown in the
black region in (a). The loss of oxygen from TiO2
under electron irradiation was measured quantitatively using EELS in combination with an analysis
using the rate theory of point defects [51,59]. In the
green regions of low and low D (Fig. 2), structural
reorganizations were not detected (on the length
scales observable using ETEM, which are limited
by the spatial resolution). By extrapolating the
structure at = 0 and D = 0 using the structure evolution diagram, the intrinsic structure in the
absence of electron irradiation can be deduced for
different environments, including reaction environments. Under reaction conditions, the major {111}
and {100} facets of gold nanoparticles were
exposed, and the particles displayed a polygonal
interface with the TiO2 support that was bounded
by sharp edges parallel to the <110> directions
(Fig. 3). Similar diagrams may be helpful in

deducing intrinsic structures and/or intrinsic phenomena in various chemical reactions that can be
produced in TEMs and ETEMs [51].
It has been suggested that gold nanoparticles
change their morphology in gases [29,31,60], although the results from these studies do not agree.
When and D were maintained in the range indicated in green in a structural evolution diagram
[51], we observed a reversible, environmentdependent change in the morphology of gold nanoparticles, even under electron irradiation. In an Au/
CeO2 catalyst, which is more robust than Au/TiO2
[50], the morphology of gold nanoparticles changed
from faceted to a roundish morphology after the
introduction of O2 gas in vacuum. After removing
the O2 gas, the morphology of the gold nanoparticles reverted to a faceted morphology.
Subsequently, 1 vol% CO/air was introduced under
vacuum. The major {111} and {100} facets appeared
clearly in 1 vol% CO/air. After removing the 1 vol%
CO/air, the morphology of the gold nanoparticles
appeared to be faceted, as in the proceeding observations in vacuum. The same morphology change
was observed in Au/TiO2 using aberration-corrected
ETEM, as shown in Fig. 3. As was described in
[50], and as will be described in the Section
Heterogeneity of real catalysts, a reversible morphology change in the majority of gold nanoparticles
indicates that the oxygen gas activation sites are
located at the interface between the gold nanoparticles and the metal oxide support.

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

Fig. 3. Systematic morphology change of gold nanoparticles in Au/


TiO2, depending on the environment [51]. The images were
recorded using a Cs-corrected ETEM. (Reproduced with permission
from John Wiley and sons).

198

M I C R O S C O P Y , Vol. 62, No. 1, 2013

gold nanoparticles, which is correlated with the


catalytic activity via the statistical analysis of substantial amounts of ETEM data. We summarize our
procedures below.
Gold nanoparticles frequently change their
morphology in the presence of gas mixtures. To describe the morphology quantitatively, we rst introduced a numerical index called the morphology
index, M [61]. M describes the shape of a supported
metal nanoparticle as projected in ETEM images.
Familiar circularity is insufcient to describe the
shape because the outline includes the straight line
described by the interface. To calculate the morphology index, the two-dimensional outline of a nanoparticle excluding the interface is tted to the arc
of an ellipse, using the least squares method; M is
then dened as the mean error between the tting
arc and the coordinates of the outline, and the M of
the nanoparticle under various conditions is normalized to that in vacuum. A smaller M means that
the nanoparticle is round.
We observed a number of gold nanoparticles and
estimated the morphology indexes in different
gases. We determined the morphology of several
gold nanoparticles in a given gas, as summarized in
Fig. 4. When more than 70% of the gold nanoparticles were classied with the same morphology, we
determined that the morphology of the gold nanoparticles in that gas environment had been statistically decided. When the morphology of gold
nanoparticles in a gas was disperse, or the

Heterogeneity of real catalysts


The microstructures of chemically synthesized catalysts are not structurally homogeneous, in contrast
with the model catalysts that are prepared in atomically controlled deposition under high vacuum.
Gold nanoparticles in real catalysts have different
atomic structures, although their size can be well
controlled. The question is, therefore, whether
a gold nanoparticle under high-resolution ETEM
observation is responsible for the catalytic activity
or not. Because there is no current microscopic
method that can directly estimate the catalytic activity of a small area of a real catalyst during ETEM
observations, we must determine the behavior of

Fig. 4. Morphology phase diagram of gold nanoparticles supported


on CeO2, as a function of the total pressure of the gases (CO and O2),
P(CO) + P(O2) and the ratio of partial pressures P(CO)/P(O2) [50].
The circles, triangles and squares represent rounded, statistically
mixed and faceted morphologies, respectively. (Reproduced with
permission from John Wiley and sons).

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

It is crucial to verify that catalytic reactions are


occurring during ETEM observations. Without any
consideration of this, a phenomenon observed
using ETEM might be regarded as only an electron
irradiation-induced phenomenon of catalyst samples
in gases. In gold nanoparticulate catalysts, the surfaces of the nanoparticles are thought to be
covered with CO molecules in reaction environments [50]. To conrm that CO gas molecules are
adsorbed on the surfaces in a steady state even
under electron irradiation, we estimated [50,57]
the rates of adsorption and desorption of CO on the
surfaces. For a typical reaction environment in
ETEM (1 vol%/air and 100 Pa), we estimated the
rates of thermal desorption [102 s1 (340 K)] and
electron
irradiation-induced
desorption
that
included electron-stimulated desorption (102
101 s1) and knock-on desorption (102 s1). The
rate of adsorption was estimated from the sticking
rate, which was 103 s1. CO was consumed
because of its oxidation by the catalyst. The oxidation rate of CO per surface gold atom was estimated
to be 2.4 101 s1 using catalytic chemical measurements [50]. The adsorption rate was, therefore,
larger than the sum of the desorption rates and
the oxidation rate. This analysis led us to the conclusion that CO dominated the surfaces of the nanoparticles in the reaction environment, even under
electron irradiation. The catalytic reaction of gold
nanoparticulate catalysts in ETEM is further discussed in the Section Pressure and temperature
gaps.

S. Takeda and H. Yoshida Quantitative atomic-resolution in-situ ETEM

Pressure and temperature gaps


The pressure gap is commonly observed between a
catalyst chemical reaction at atmospheric pressure
and an analysis that is carried out under high
vacuum. It was thought that the environments surrounding catalysts in ETEM were different from
those in typical catalytic reactions, owing to the
relatively low gas pressures and low temperatures.
Using an ETEM specimen holder with two parallel
thin windows [30], it was possible to achieve gas
pressures around the specimen that exceeded atmospheric pressure. For atomic-scale observations,
the gas pressure range in recently developed

Cs-corrected ETEMs now extends far above 2000


Pa. In practice, gold nanoparticulate catalysts are
used for the oxidation of CO in the partial pressure
range of 110 Pa. It is known that the catalytic activity of a gold nanoparticulate catalyst (Au/TiO2) is
linearly proportional to the partial pressure of CO,
even at pressures below 10 Pa. As mentioned
earlier, gold nanoparticulate catalysts are highly
active even below room temperature. Therefore,
neither a pressure gap nor a temperature gap exists
for gold nanoparticulate catalysts in ETEM observations [57]. ETEM can provide in-situ observations
at elevated temperatures [44,45,55,56,61]. The temperature dependence of the morphology of supported platinum nanoparticles above 150C
correlated with the catalytic activity [61]. The
ETEM observation of catalysis samples is technically possible at temperatures up to 1000C, if differently designed specimen holders are used.

Imaging of atoms and molecules during


catalytic reactions
We showed that the reaction environments for gold
nanoparticulate catalysts can be reproduced in a
high-resolution ETEM. Because high-resolution
ETEMs generally enlarge an extremely small area
of a specimen for observation, we should be careful
to determine whether or not the areas under observation reect the macroscopic nature of the specimen; in our case, the property in question is the
catalytic activity in gold nanoparticulate catalysts.
As mentioned above, numerical, quantitative and
statistical analyses showed that the majority (over
70%) of gold nanoparticles exhibited a faceted
morphology under reaction conditions, whereas in
O2-rich environments, they appeared to be round
because of the interaction of the gold atoms with
active oxygen species. This morphology change in
the majority of the gold nanoparticles was most
likely responsible for the catalytic activity. We,
therefore, observed the gold nanoparticles that
exhibited the same morphology change using
ETEM at higher magnication.
Figure 5 shows aberration-corrected ETEM
images obtained in a reaction environment, under
vacuum [57]. Each gold atomic column appears as a
dark dot. In vacuum (Fig. 5a), the distance between

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

percentage of the gold nanoparticles belonging to


the major morphology was smaller than 70%, we
concluded that the morphology was statistically
mixed. In this way, we established a morphology
phase diagram [50] for the Au/CeO2 catalyst (Fig. 4)
that represents the dependence of the morphology
of the gold nanoparticles on the partial pressures of
CO [P(CO)] and O2 [P(O2)]. The squares, triangles
and circles in Fig. 4 represent faceted, statistically
mixed and rounded morphologies, respectively.
This is the rst time that these systematic and quantitative microstructural data have been obtained for
an operating nanoparticle catalyst [50]. By establishing the morphology diagram (Fig. 4) and considering the electron irradiation effects mentioned in the
preceding section, we concluded that the reversible
morphology change correlated well with the catalytic activity of supported gold nanoparticles on CeO2
and TiO2. The morphology phase diagram in Fig. 4
showed that faceted gold nanoparticles were stable
at higher CO partial pressures; with decreases in
the partial pressure of CO, the faceted gold nanoparticles became rounded. When the partial pressure of O2 was sufciently low, the majority of the
gold nanoparticles returned to the faceted form,
even at low CO partial pressures. The active oxygen
species are most likely produced at the perimeter
interfaces, and with the aid of electron irradiation,
they make the surface of the gold nanoparticles
rough [50]. It is noteworthy that a small amount of
CO could stabilize the faceted morphology (Fig. 4).
CO molecules adsorb more strongly than oxygen
molecules on the surfaces of gold nanoparticles.

199

200

M I C R O S C O P Y , Vol. 62, No. 1, 2013

Fig. 5. Au{100}-hex reconstructed surface under catalytic conditions


[57]. A gold nanoparticle supported on CeO2 in (a) vacuum and (b) a
reaction environment (1 vol % CO in air at 45 Pa), at room
temperature. The two {100} facets in the rectangular regions
indicated by I and II are shown enlarged in vacuum and in CO in air.
The images were recorded using a Cs-corrected ETEM. (Reproduced
with permission from The American Association for the
Advancement of Science).

distorted normal square lattice (Fig. 5b). These


changes in the positions of the Au atomic columns
corresponded well with those of the Au{100}-hex
reconstructed surface structure. It was conrmed
that the reconstruction appeared in the pressure
range from 100 to 2000 Pa of 1 vol% CO/air, as
shown in Fig. 6. The corresponding partial pressure
of CO ranged from 1 to 20 Pa. Our observations
were, therefore, carried out under reaction conditions, to ensure no pressure gap. The reconstruction in the reaction environment was recovered to
give the unreconstructed surface by reducing the
partial pressure of CO to vacuum. As mentioned
earlier, CO molecules dominate the surface of gold
nanoparticles in reaction environments, even under
electron irradiation, given the appropriate electron
current density and electron energy [50,57].
Therefore, our observations indicated that the reconstruction of the {100} facets was induced by the
interaction of CO molecules and gold atoms on the
facets in the reaction environment.
The CO molecules interacting with the gold atoms
in the facets could be visualized using Cs-corrected
ETEM (Fig. 7). The reconstructed structure with
adsorbed CO molecules was conrmed using image
simulations and ab initio calculations [57]. In previous theoretical studies, it was suggested that CO
molecules adsorbed on low-coordination sites such
as steps, edges and corners more strongly than on
at facets, unlike ordinary Pd and Pt catalysts.
However, the present aberration-corrected ETEM
observations, in combination with ab initio calculations, showed that the surface structure of the {100}
facets could be modied to adsorb CO molecules

Fig. 6. No pressure gap [57]. The Au{100}-hex reconstructed surface remained stable under CO gas with lower (100 Pa) to higher (2000 Pa)
pressures of 1 vol% CO/air (Au/CeO2). The images of the rectangular regions are enlarged and shown at the bottom. The images were
recorded using a Cs-corrected ETEM. (Reproduced with permission from The American Association for the Advancement of Science).

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

the topmost and second topmost {100} surface


layers (0.20 nm) was the same as the interplanar
distance between the {200} planes in crystalline
bulk gold. The average distance between adjoining
Au atomic columns on the topmost surface layer
was 0.29 nm, the same as the corresponding distance in a {100} plane in crystalline bulk gold. We
found that during the catalytic reaction, the Au
atomic columns on the topmost and second
topmost {100} layers shifted to peculiar positions.
The Au atoms on the topmost surface layer formed
an undulating hexagonal lattice, whereas those on
the second topmost surface layer formed a slightly

S. Takeda and H. Yoshida Quantitative atomic-resolution in-situ ETEM

201

with a higher density. Hence, we have now obtained


a reliable atomistic view of the adsorption of CO
molecules on supported gold nanoparticulate catalysts. Gas molecules have now been imaged in highresolution ETEM observations, as demonstrated
[57]. The high-resolution ETEM techniques developed here based on well-designed instrumentation
will be useful to identify catalytically active sites
in reaction environments and will, thus, make a
further contribution to catalytic chemistry.

Concluding remarks
We have demonstrated the potential of recently
developed ETEMs for quantitative in-situ microscopy at the atomic scale, especially for the chemistry of catalysts in gases. The issues associated with
the application of ETEM to catalyst chemistry were
addressed using systematic, numerical and statistical ETEM analyses. Taking into account the electron
irradiation effects, the heterogeneity of real catalysts, the pressure gap and the temperature gap, we

were able to perform quantitative in-situ structural


studies on typical gold nanoparticulate catalysts
with high activity for the oxidation of CO at room
temperature. Using these quantitative analyses, the
positions of atoms on the surface of catalysts, including adsorbed gas molecules, could be determined in reaction environments using ETEM. To
further develop the ETEM instrument, faster and
more sensitive cameras are needed to observe
faster phenomena using lower-energy electron
doses. Importantly, in-situ TEM observations have
advanced in combination with the simultaneous
characterization of specimens using TEM; examples
of such observations include optical spectroscopy
[62], mechanical indentations [63], electronic measurements [64,65] and others [66]. These features
could be installed in ETEMs, so there are numerous
possible applications. An example of another interesting application is EELS in ETEM [67], which is
useful for the analysis of gases and samples, especially at higher energy resolutions. The effect of gas

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

Fig. 7. Glimpse of gas molecules on the surface of catalysts under reaction conditions [57]. (a) Observed and (b) simulated image of the Au
{100}-hex reconstructed structure with CO molecules. (c) Simulated image of Au{100}-hex reconstructed structure without CO molecules.
Lateral intensity scans of (a), (b) and (c) are shown in (d), (e) and (f ), respectively. Line scans normal to the surface of (a), (b) and (c) are
shown in (g), (h) and (i), respectively. The oscillation intensity in observation (d) can be accounted for by considering the adsorbed CO
molecules, as shown in (e). The Au{100}-hex reconstructed surface without CO molecules could not reproduce the oscillating intensity, as
shown in (f ). The images were recorded using a Cs-corrected ETEM. (Reproduced with permission from The American Association for the
Advancement of Science).

202

M I C R O S C O P Y , Vol. 62, No. 1, 2013


.......................................................................................................................

on the image quality of ETEM is under quantitative


investigation [68,69]. Hence, the eld of quantitative
in-situ ETEM is only in its beginning stages.

11 Flower H M (1972) High voltage electron microscopy of environmental reactions. J. Microsc. 97: 171190.
.......................................................................................................................

12 Hashimoto H, Naiki T, Eto T, and Fujiwara K (1968) High temperature gas reaction specimen chamber for an electron microscope.
Jpn. J. Appl. Phys. 7: 946952.
.......................................................................................................................

Funding

14 Ichinose H and Ishida Y (1981) Observation of [110] tilt boundary


structures in gold by high-resolution HVEM. Philos. Mag. A 43:
12531264.
.......................................................................................................................

15 Bourret A and Desseaux J (1979) The low-angle [011] tilt boundary


in germanium 1. High-resolution structure determination. Philos.
Mag. A 39: 405418.
.......................................................................................................................

16 Cherns D, Anstis G R, Hutchison J L, and Spence J C H (1982)


Atomic-structure of the NiSi2/(111)Si interface. Philos. Mag. A 46:
849862.
.......................................................................................................................

17 Tung R T, Gibson J M, and Poate J M (1983) Formation of ultrathin


single-crystal silicide lms on Si: surface and interfacial stabilization
of Si-NiSi2 epitaxial structures. Phys. Rev. Lett. 50: 429432.

.......................................................................................................................

18 Kolar H R, Spence J C H, and Alexander H (1996) Observation of


moving dislocation kinks and unpinning. Phys. Rev. Lett. 77:
40314034.
.......................................................................................................................

Acknowledgements
The authors are indebted to their collaborators in the projects, especially Yoshikazu Homma, Masatake Haruta and Masanori Kohyama. They
are thankful to Yasufumi Kuwauchi and Tetsuya Uchiyama for their
invaluable contributions to the ETEM studies. They also thank Tomoki
Akita, Keju Sun, Satoshi Ichikawa, Shingo Tanaka, Koji Tanaka, Hideo
Kohno, Yutaka Ohno, FEI Company, especially Joerg R. Jinschek,
Stephan Kujawa and Stan Konings and FEI Japan, especially Ryohei
Matsumoto for their invaluable contributions.

References
.......................................................................................................................

1 Hirsch P, Howie A, Nicholson R, Pashley D W, and Whelan M J


(1977) Electron Microscopy of Thin Crystals (Robert E. Krieger
Publishing Co., Malabar, Florida).
.......................................................................................................................

2 Thomas G and Goringe M J (1979) Transmission Electron


Microscopy of Materials. (John Wiley, New York).

19 Gibson J M, McDonald M L, and Unterwald F C (1985) Direct


imaging of a novel silicon surface reconstruction. Phys. Rev. Lett.
55: 17651767.
.......................................................................................................................

20 Takeda S, Kohyama M, and Ibe K (1994) Interstitial defects on (113)


in Si and Ge line defect conguration incorporated with a selfinterstitial atom chain. Philos. Mag. A 70: 287312.
.......................................................................................................................

21 Muto S, Takeda S, and Hirata M (1995) Hydrogen-induced platelets


in silicon studied by transmission electron microscopy. Philos. Mag.
A 72: 10571074.
.......................................................................................................................

22 Eaglesham D J, Stolk P A, Gossmann H J, and Poate J M (1994)


Implantation and transient B diffusion in Si: the source of the interstitials. Appl. Phys. Lett. 65: 23052307.
.......................................................................................................................

23 Rodriguez N M, Oh S G, Downs W B, Pattabiraman P, and Baker R T


K (1979) An atomic oxygen environmental cell for a transmission
electron microscope. Rev. Sci. Instrum. 61: 18631868.
.......................................................................................................................

24 Doole R C, Parkinson G M, and Stead J M (1991) A high resolution


gas reaction cell for the JEM 4000. Inst. Phys. Conf. Ser. 119:
157160.

.......................................................................................................................

.......................................................................................................................

3 Hirsch P B, Horne R W, and Whelan M J (1956) Direct observations


of the arrangement and motion of dislocations in aluminum. Philos.
Mag. 1: 677684.

25 Boyes E D and Gai P L (1997) Environmental high resolution electron microscopy and applications to chemical science.
Ultramicroscopy 67: 219232.

.......................................................................................................................

4 Ferreira Lima C A and Howie A (1976) Defects in electron-irradiated


germanium. Philos. Mag. 34: 10571071.
.......................................................................................................................

5 Iijima S (1971) High-resolution electron microscopy of crystal lattice


of titanium-niobium oxide. J. Appl. Phys. 42: 58915893.
.......................................................................................................................

6 Cowley J M and Iijma S (1972) Electron-microscope image contrast


for thin crystals. Zeitschrift Fur Naturforschung Part A-Astrophysik
Physik Und Physikalische Chemie A 27: 445.
.......................................................................................................................

7 Bovin J O, Wallenberg R, and Smith D J (1985) Imaging of atomic


clouds outside the surfaces of gold crystals by electron microscopy.
Nature 317: 4749.
.......................................................................................................................

8 Iijima S and Ichihashi T (1986) Structural instability of ultrane


particles of metals. Phys. Rev. Lett. 56: 616619.
.......................................................................................................................

9 Hashimoto H, Takai Y, Yokota Y, Endoh H, and Fukada E (1980)


Direct observations of the arrangement of atoms around stackingfaults and twins in gold crystals and the movement of atoms accompanying their formation and disappearance. Jpn. J. Appl. Phys. 19:
L1L4.
.......................................................................................................................

10 Sinclair R F, Ponce A, Yamashita T, Smith D J, Camps R A, Freeman


L A, Erasmus S J, Nixon W C, Smith K C A, and Catto C J D (1982)
Dynamic observation of defect annealing in CdTe at lattice
resolution. Nature 298: 127131.

.......................................................................................................................

26 Sharma R and Weiss K (1998) Development of a TEM to study in


situ structural and chemical changes at an atomic level during gassolid interactions at elevated temperatures. Microsc. Res. Tech. 42:
270280.
.......................................................................................................................

27 Daulton T L, Little B J, Kim J W, Newell S, Lowe K, Furukawa Y,


Jones-Meehan J, and Lavoie D L (2002) Quantitative environmental
cell transmission electron microscopy: studies of microbial Cr(VI)
and Fe(III) reduction. JEOL News 37E: 613.
.......................................................................................................................

28 Kamino T, Yaguchi T, Konno M, Watabe A, Marukawa T, Mima T,


Kuroda K, Saka H, Arai S, Makino H, Suzuki Y, and Kishita K (2005)
Development of a gas injection/specimen heating holder for use with
transmission electron microscope. J. Electron Microsc. 54: 497503.
.......................................................................................................................

29 Giorgio S, Joao S S, Nitsche S, Chaudanson D, Sitja G, and Henry C


R (2006) Environmental electron microscopy (ETEM) for catalysts
with a closed E-cell with carbon windows. Ultramicroscopy 106:
503507.
.......................................................................................................................

30 Creemer J F, Helveg S, Hoveling G H, Ullmann S, Molenbroek A M,


Sarro P M, and Zandbergen H W (2008) Atomic-scale electron microscopy at ambient pressure. Ultramicroscopy 108: 993998.
.......................................................................................................................

31 Kawasaki T, Ueda K, Ichihashi M, and Tanji T (2009) Improvement


of windowed type environmental-cell transmission electron

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

This research was partly supported by Grant-in-Aid


for Specially Promoted Research Grant No.
19001005 from the Ministry of Education, Culture,
Sports, Science and Technology, Japan. A part of
this study was supported by Low-Carbon Research
Network (Handai satellite) and the Management
Expenses Grants for National Universities
Corporations from the Ministry of Education,
Culture, Sports, Science and Technology of Japan
(MEXT).

13 Reif F (1965) Statistical Physics: Berkeley Physics Course, Vol. 5.


(McGraw-Hill, Inc., New York, USA).
.......................................................................................................................

S. Takeda and H. Yoshida Quantitative atomic-resolution in-situ ETEM

203

.......................................................................................................................

microscope for in situ observation of gas-solid interactions. Rev. Sci.


Instrum. 80: 113701.
.......................................................................................................................

32 De Jonge N, Bigelow W C, and Veith G M (2010) Atmospheric pressure scanning transmission electron microscopy. Nano Lett. 10:
10281031.
.......................................................................................................................

33 Takeda S, Yoshida H, and Kujawa S (2010) Recent advancement


of high resolution environmental TEM. IMC17 Proceedings Rio de
Janeiro, Brazil,, pp 290291.
.......................................................................................................................

.......................................................................................................................

35 Yaguchi T, Suzuki M, Watabe A, Nagakubo Y, Ueda K, and Kamino T


(2011) Development of a high temperature-atmospheric pressure
environmental cell for high-resolution TEM. J. Electron Microsc. 60:
217225.
.......................................................................................................................

36 Helveg S, Lpez-Cartes C, Sehested J, Hansen P L, Clausen B S,


Rostrup-Nielsen J R, Ablid-Pedersen F, and Nrskov J K (2004)
Atomic-scale imaging of carbon nanobre growth. Nature 427:
426429.
.......................................................................................................................

37 Sharma R and Iqbal Z (2004) In situ observations of carbon nanotube formation using environmental transmission electron microscopy. Appl. Phys. Lett. 84: 990992.
.......................................................................................................................

38 Sharma R, Rez P, Treacy M M J, and Stuart S (2004) In situ observation of the growth mechanisms of carbon nanotubes under diverse
reaction conditions. J. Electron Microsc. 54: 231237.
.......................................................................................................................

39 Yoshida H, Takeda S, Uchiyama T, Kohno H, and Homma Y (2008)


Atomic-scale in-situ observation of carbon nanotube growth from
solid state iron carbide nanoparticles. Nano Lett. 8: 20822086.
.......................................................................................................................

40 Kim B J, Tersoff J, Kodambaka S, Reuter M C, Stach E A, and Ross


F M (2008) Kinetics of individual nucleation events observed in
nanoscale vapor-liquid-solid growth. Science 322: 10701073.
.......................................................................................................................

41 Datye A K and Smith D J (1992) The study of heterogeneous catalysts by high-resolution transmission electron-microscopy. Catal.
Rev. 34: 129178.
.......................................................................................................................

42 Herzing A A, Kiely C J, Carley A F, Landon P, and Hutchings G J


(2008) Identication of active gold nanoclusters on iron oxide supports for CO oxidation. Science 321: 1331.
.......................................................................................................................

43 Prieto G, Zecevic J, Friedrich H, de Jong K P, and de Jongh P E


(2013) Towards stable catalysts by controlling collective properties
of supported metal nanoparticles. Nat. Mater 12: 3439.

.......................................................................................................................

44 Hansen P L, Wagner J B, Helveg S, Rostrup-Nielsen J R, Clausen


B S, and Topse H (2002) Atom-resolved imaging of dynamic
shape changes in supported copper nanocrystals. Science 295:
20532055.
.......................................................................................................................

45 Simonsen S B, Dahl S, Johnson E, and Helveg S (2008)


Ceria-catalyzed soot oxidation studied by environmental transmission electron microscopy. J. Catal. 255: 15.
.......................................................................................................................

46 Takeda S and Yoshida H (2004) HRTEM image simulation of carbon


nanotubes under actual growth environment. Microsc. Microanal.
10: 1819.
.......................................................................................................................

47 Yoshida H and Takeda S (2005) Image formation in a transmission


electron microscope equipped with an environmental cell: singlewalled carbon nanotubes in source gases. Phys. Rev. B 72: 195428.
.......................................................................................................................

48 Kiritani M and Takata H (1978) Dynamic studies of defect mobility


using high voltage electron microscopy. J. Nucl. Mater. 6970:
277309.
.......................................................................................................................

49 Haruta M, Yamada N, Kobayashi T, and Iijima S (1989) Gold catalysts prepared by coprecipitation for low-temperature oxidation of
hydrogen and of carbon-monoxide. J. Catal. 115: 301309.
.......................................................................................................................

50 Uchiyama T, Yoshida H, Kuwauchi Y, Ichikawa S, Shimada S, Haruta


M, and Takeda S (2011) Systematic morphology changes of gold
nanoparticles supported on CeO2 during CO oxidation. Angew.
Chem. Int. Ed. 50: 1015710160.

.......................................................................................................................

52 Yoshida H, Uchiyama T, and Takeda S (2007) Environmental transmission electron microscopy observations of swinging and rotational
growth of carbon nanotubes. Jpn. J. Appl. Phys. 46: L917L919.
.......................................................................................................................

53 Yoshida H, Shimizu T, Uchiyama T, Kohno H, Homma Y, and Takeda


S (2009) Atomic-scale analysis on the role of molybdenum in ironcatalyzed carbon nanotube growth. Nano Lett. 9: 38103815.
.......................................................................................................................

54 Yoshida H, Kohno H, and Takeda S (2012) In situ structural analysis


of crystalline Fe-Mo-C nanoparticle catalysts during the growth of
carbon nanotubes. Micron 43: 11761180.
.......................................................................................................................

55 Hansen T W, Wagner J B, and Dunin-Borkowski R E (2010)


Aberration corrected and monochromated environmental transmission electron microscopy: challenges and prospects for materials
science. Mater. Sci. Technol. 26: 13381344.
.......................................................................................................................

56 Ciston J, Si R, Rodriguez J A, Hanson J C, Martinez-Arias A,


Fernandez-Garcia M, and Zhu Y (2011) Morphological and structural
changes during the reduction and reoxidation of CuO/CeO2 and Ce1
xCuxO2 nanocatalysts: in situ studies with environmental TEM,
XRD, and XAS. J. Phys. Chem. C 115: 1385113859.
.......................................................................................................................

57 Yoshida H, Kuwauchi Y, Jinschek J R, Sun K, Tanaka S, Kohyama M,


Shimada S, Haruta M, and Takeda S (2012) Visualizing gas molecules
interacting with supported nanoparticulate catalysts at reaction conditions. Science 335: 317319.
.......................................................................................................................

58 Yamasaki J, Takeda S, and Tsuda K (2002) Elemental process of


amorphization induced by electron irradiation in Si. Phys. Rev. B 65:
115213.
.......................................................................................................................

59 Noda N and Takeda S (1996) Disordering of natural superlattice in


(Ga,In)P induced by electron irradiation. Phys. Rev. B 53: 71977204.
.......................................................................................................................

60 Tanaka T, Sano K, Ando M, Sumiya A, Sawada H, Hosokawa F,


Okunishi E, Kondo Y, and Takayanagi K (2010) Oxygen-rich Ti1xO2
pillar growth at a gold nanoparticleTiO2 contact by O2 exposure.
Surf. Sci. 604: L75L78.
.......................................................................................................................

61 Yoshida H, Matsuura K, Kuwauchi Y, Kohno H, Shimada S, Haruta


M, and Takeda S (2011) Temperature-dependent change in shape of
platinum nanoparticles supported on CeO2 during catalytic reactions. Appl. Phys. Express 4: 065001.
.......................................................................................................................

62 Ohno Y and Takeda S (2002) Analysis of polarization by means of


polarized cathodoluminescence spectroscopy in a TEM. J. Electron
Microsc. 51: 281290.
.......................................................................................................................

63 Minor A M, Syed Asif S A, Shan Z, Stach E A, Cyrankowski E, Wyrobek


T J, and Warren O L (2006) A new view of the onset of
plasticity during the nanoindentation of aluminium. Nat. Mater. 5:
697702.
.......................................................................................................................

64 Ohnishi H, Kondo Y, and Takayanagi K (1998) Quantized conductance through individual rows of suspended gold atoms. Nature 395:
780783.
.......................................................................................................................

65 Kohno H, Mori Y, Ichikawa S, Ohno Y, Yonenaga I, and Takeda S


(2009) Transformation of a SiC nanowire into a carbon nanotube.
Nanoscale 1: 344346.
.......................................................................................................................

66 Dehm G, Howe J M, and Zweck J (2012) In-Situ Electron


Microscopy, Applications in Physics, Chemistry and Materials
Science. (Wiley-VCH, Weinheim, Germany).
.......................................................................................................................

67 Crozier P A and Chenna S (2011) In situ analysis of gas composition


by electron energy-loss spectroscopy for environmental transmission
electron microscopy. Ultramicroscopy 111: 177185.
.......................................................................................................................

68 Jinschek J R and Helveg S (2012) Image resolution and sensitivity in


an environmental transmission electron microscope. Micron 43:
11561168.
.......................................................................................................................

69 Hansen T W and Wagner J B (2012) Environmental transmission


electron microscopy in an aberration-corrected environment.
Microsc. Microanal. 18: 684690.

Downloaded from http://jmicro.oxfordjournals.org/ at KOREA ADVANCED INST OF SCIENCE AND TECHNOLOGY on July 31, 2014

34 Kujawa S (2011) Can gas-solid or liquid-solid interfaces be visualized in


situ at atomic scale? Proceedings of the Sixty-Seventh Annual Meeting
of the Japanese Society of Microscopy Fukuoka, Japan,, p 115.

51 Kuwauchi Y, Yoshida H, Akita T, Haruta M, and Takeda S (2012)


Intrinsic catalytic structure of gold nanoparticles supported on TiO2.
Angew. Chem. Int. Ed. 51: 77297733.