Ultramicroscopy 159 (2015) 73–80
Contents lists available at ScienceDirect
Ultramicroscopy
journal homepage: www.elsevier.com/locate/ultramic
Critical conditions for atomic resolution imaging of molecular crystals
by aberration-corrected HRTEM
Kaname Yoshida a,n, Johannes Biskupek b, Hiroki Kurata c, Ute Kaiser b
a
Japan Fine Ceramics Center, 2-4-1 Mutsuno, Atsuta-ku, Nagoya 456-8587, Japan
Electron Microscopy Group of Materials Science, University of Ulm, Albert-Einstein-Allee 11, D-89081 Ulm, Germany
c
Institute for Chemical Research, Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan
b
art ic l e i nf o
a b s t r a c t
Article history:
Received 9 April 2015
Received in revised form
22 July 2015
Accepted 23 August 2015
Available online 24 August 2015
Atomically resolved imaging of organic molecules consisting of thin crystals by aberration-corrected (AC)
HRTEM was studied by experimental observations and image simulations. An atomically resolved image
of the hexadecachlorophthalocyanatocopper (CuPcCl16) molecule was obtained under low-dose conditions. The conditions for imaging organic frameworks were found to be more restricted than those for
heavier elements such as copper and chlorine. For the characterization of the benzene rings within
CuPcCl16 molecules, the specimen thickness had to be less than 5 nm. The effects of the defocus
conditions were examined by changing the image according to the location of the inclined specimen. The
optimal defocus range for atomic resolution imaging of organic molecules was limited to a narrow region
around the Scherzer defocus. Compared with scanning transmission microscopy, AC-HRTEM is more
suitable for low-dose imaging, but the optimum conditions were severely restricted.
& 2015 Elsevier B.V. All rights reserved.
Keywords:
Molecular crystal
Aberration-corrected HRTEM
Light element imaging
1. Introduction
The development of molecular manipulation techniques for
device production, such as organic molecular beam epitaxy
(OMBE), allows the formation of fine organic structures [1,2] that
are closing in on inorganic semiconductors. To improve the performance of organic devices, it is necessary to characterize their
nanostructure, which depends on both their molecular and crystal
structure [3–5]. This has stimulated the development of the
methodology of organic structural analysis at the nanoscale [6–
10]. Although high-resolution transmission electron microscopy
(HRTEM) has been demonstrated to be a powerful technique, its
application to organic materials is severely restricted owing to
radiation damage to the specimen. Electron irradiation damage
has been a serious problem ever since the development of the
electron microscope [11–13]. Uyeda et al. were the first to report
atomic resolution imaging of an organic crystal in 1979 [14]. By
minimizing the electron dose during observation and using an
ultrahigh-voltage TEM (UHVTEM) with a high resolving power,
they achieved high-resolution imaging of epitaxially grown hexadecachlorophthalocyanatocopper (CuPcCl16) thin films. However,
organic frameworks consisting of carbon and nitrogen atoms had
not been observed, even though the spatial resolution had been
n
Corresponding author. Fax: þ 81 52 871 3599.
E-mail address: kaname_yoshida@jfcc.or.jp (K. Yoshida).
http://dx.doi.org/10.1016/j.ultramic.2015.08.006
0304-3991/& 2015 Elsevier B.V. All rights reserved.
close to the carbon–carbon bond length ( 0.14 nm). UHVTEM was
unsuited for imaging light elements because of the lower image
contrast and intense electron irradiation, implying knock-on damage. In addition to the irradiation damage, the insufficient contrast associated with light elements is a serious problem in HRTEM
imaging. Although image reconstruction of CuPcCl16 had also been
reported by Kirkland et al. [15,16], 5-membered and 6-membered
rings had not been resolved clearly even in such high-resolution
images.
There are two ways in which light element imaging by HRTEM
can be achieved: (i) utilizing a high-sensitivity detector, and (ii)
enhancement of image contrast. Both are essential for atomic
imaging under very low dose conditions, which result in a noisy,
low-signal image. Although highly efficient cameras have been
studied extensively [17–20], their applicability to low-damage
imaging should be considered from the viewpoint of operability.
Our study focuses on the experimental inspection about direct
imaging of organic molecular frameworks with HRTEM. Since
aberration correction (AC) is now available in state-of-the-art instruments, even light and heavy atomic columns can be separated
from each other precisely in the AC-HRTEM image [21–24], and
their contrast can be enhanced by fine-tuning the image corrector
[25,26]. By choosing the appropriate aberration parameters, such
as defocus and spherical aberration (CS) of objective lens, the
atomic contrast of certain spatial frequencies in an HRTEM image
can be improved; however, the problem of radiation damage largely persists. Direct observation of the six-membered carbon rings
74
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
in a graphene sheet has already been achieved by AC-HRTEM at
80 kV e.g. [27–32] and by AC-STEM at 60 kV [33]. These studies
employed microscopes with low accelerating voltage to reduce the
knock-on sputtering caused by fast electron bombardment. However, this is not necessarily effective for any material [34]. We also
need to consider the radiation chemistry of the specimen for a
complete understanding of electron irradiation damage. In the
case of organic molecular crystals, irreversible intermolecular or
intramolecular reactions induced by radiolysis cause their crystal
structures to collapse [11,35,36]. Organic molecular crystals consist
of strong covalent bonds within a molecule and very weak intermolecular interactions so that the mechanism of their radiation
damage is very complex. The radiolytic process of sp2-carbon
materials, such as graphene, fullerene, and carbon nanotubes, is
less prone to cause structural collapse because of the stabilization
of their ionized state by the entire conjugated system. In addition,
the accelerating voltage affects the critical thickness of the weak
phase object approximation (WPOA). Since graphitic carbon materials are quite thin, a lower accelerating voltage is more effective
in atomic resolution imaging. On the other hand, because molecular crystals cannot be made sufficiently thin, a higher voltage
may be needed to obtain atomic structure images. The optimal
accelerating voltage for HRTEM observation varies with the type of
specimen. In order to determine the critical conditions of specimen thickness and defocus value for structural imaging of an organic molecular crystal by AC-HRTEM, we observed a CuPcCl16
thin crystal (Fig. 1(a)), known to be more radiation–resistant than
most other molecular crystals [13], by AC-HRTEM and compared
with simulated images.
Meanwhile, the new imaging mode of annular bright-field
scanning transmission electron microscopy (ABF-STEM) has enabled imaging of extremely light elements, such as hydrogen and
lithium atomic columns [37–41]. Furthermore, direct imaging of
the organic frameworks in the CuPcCl16 crystal with ABF- and
LAADF-STEM had been reported in 2012 [42]. However, illuminating conditions cannot be assigned independently to processes
for focusing and acquiring image. The STEM mode is particularly
disadvantage to low-dose observation owing to the fixed illuminating condition during focusing, unless minimum-dose-system
(MDS) [43] is optimized for STEM imaging. Thus, it is still necessary to improve and explore imaging possibilities for organic
molecules by HRTEM. It is desirable for there to be many choices of
imaging modes available for soft materials like organic crystals.
The combination of Zernike phase plate and AC-HRTEM can be
also a powerful technique for direct imaging of soft materials
[44,45]. However, ideal phase plates have not been developed so
far, so it is impossible to get high-resolution image without any
Fig. 2. AFM image of CuPcCl16 and its cross-sectional profile. The arrows in the
image indicate the crystallographic directions.
artifacts by the current stage.
2. Experimental
2.1. Thin film preparation
CuPcCl16 epitaxial thin films were fabricated by using a previously reported procedure [46] with slight modification: the
molecules were deposited onto a freshly cleaved KCl (001) surface
kept at 590 K by using a sublimation method under a vacuum of
o10 5 Pa. The deposition source of CuPcCl16 was purchased from
Tokyo Chemical Industry, with further purification of train sublimation. The deposition rate was kept constant at 0.2 nm/min
with the aid of a quartz microbalance, and the deposition was
stopped once the films reached an average thickness of 5 nm. The
morphology of the grown thin crystals was observed by an atomic
force microscope (AFM, Digital Instruments NanoScope IIIa). After
AFM observation, the CuPcCl16 thin film reinforced with very thin
amorphous carbon was transferred onto a microgrid support film
for subsequent high-resolution imaging.
2.2. AC-HRTEM observations
AC-HRTEM observations were performed using a JEOL
JEM2200FS (accelerating voltage: 200 kV) equipped with a CS
corrector (CEOS CETCOR). The free parameter CS was set to
þ15 mm. The second-order aberrations (coma and threefold astigmatism) and third-order aberrations (fourfold astigmatism and
star-aberration) were smaller than the error in the aberration
measurement (carried out by iterative tuning of the corrector). All
HRTEM images were taken by a slow-scan CCD camera
(14 14 mm, 2048 2048 pixels, Gatan UltraScan 1000). The
images were magnified by a factor of 450,000 on the camera.
Fig. 1. (a) Molecular structure and (b) crystal structure of CuPcCl16.
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
75
Fig. 3. Relationship between morphology and crystallographic orientation. (a) A low-magnification TEM image of a CuPcCl16 single crystal. (b) An electron diffraction pattern
obtained with the incident beam perpendicular to the film surface and a schematic of the setup. (c) An electron diffraction pattern obtained with the incident beam parallel
to the c axis and a schematic of the setup.
The pixel size on the digital image captured with this setup corresponded to 0.03 nm/pixel, a sufficient sampling frequency (Nyquist frequency: 17.7 nm 1) compared with the information limit
( 10 nm 1) of the microscope. To get a molecular image, the
specimen was tilted by 26.5° around the b axis so that the c axis
would be parallel to the incident beam direction. The phthalocyanine molecules overlap finely along this direction. The monoclinic crystal structure of CuPcCl16 [47] with unit-cell dimensions
of a ¼1.962 nm, b¼2.604 nm, c ¼0.376 nm, and β ¼ 116.51° is illustrated in Fig. 1(b).
To remove the random noise in raw HRTEM images, we applied
a Wiener filter [48] based on the two-dimensional fast Fourier
transform (2D FFT) method to all HRTEM images. The Wiener filtering combined with a Butterworth filter was executed using the
software Digital Micrograph (Gatan Company, Pleasanton, USA).
2.3. HRTEM image simulation
Fig. 4. Amplitudes and phases of exit wavefunctions simulated by the multi-slice
method.
Table 1
Optical parameters for each CS settings estimated under an information limit of
10 nm 1 and accelerating voltage of 200 kV.
CS (μm) Defocus Δf (nm) Radius of confusion disk, R
(nm)
Zero CS mode
0
Extended Scherzer 7 34
Lentzen's CS
7 15
7 4.0
∓11
∓7.1
0.1
0.27
0.059
HRTEM image simulation was executed exploiting the multislice method [49] of the software MacTempas X (Total Resolution
LLC, Berkeley, USA). By comparing an experimental HRTEM image
with a series of simulated through-focus images for various crystal
thickness, we determined the optimally focused area in each
image.
3. Results and discussion
3.1. Characterization of CuPcCl16 thin film
A typical AFM image and height profile of a CuPcCl16 single
crystal is shown in Fig. 2. The height profile indicates that the
CuPcCl16 crystals exhibit a wedge shape with a maximum thickness of ca. 30 nm. These results also indicate that the wedge angle
76
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
of the CuPcCl16 crystals is very narrow (o2°). The epitaxial relationship between monoclinic CuPcCl16 crystals and KCl substrates has already been reported [47,50]. The {001} planes of
epitaxially grown CuPcCl16 are known to be parallel to the substrate surface. Fig. 3(a) and (b) shows electron diffraction patterns
with the beam direction perpendicular to the thin film surface and
parallel to the c axis, respectively. To observe each independent
atomic column, the sample film was inclined at 26.5° around the b
axis in the TEM so as to make the c axis parallel to the optical axis.
3.2. AC-HRTEM observations
Fig. 5. Phase contrast transfer functions for (a) zero CS, (b) extended Scherzer and
(c) Lentzen's CS modes and (d) atomic form factors of constituent elements. PCTFs
with dumping factor (solid line), without dumping factor (dotted) and envelope
(dashed line).
Prior to AC-HRTEM observations, the exit wavefunction at the
lower surface of the specimen, which varies with crystal thickness
and accelerating voltage, was simulated by the multi-slice method.
Fig. 4 shows the amplitudes of simulated exit wavefunctions. With
increasing crystal thickness, the electrons converge into individual
atomic column positions and lower electron density areas appear
in the surrounding regions. The amplitude of the exit wavefunction increases monotonically up to a certain thickness, above
which it diverges from the atomic column positions. In particular,
the configuration of the five-membered rings in the phthalocyanine framework can only be maintained for a very thin crystal. In
order to observe small membered organic rings, the thickness
must be less than a critical value at which the wavefunction is
reflected in the molecular configurations. The critical specimen
thickness decreases with decreasing accelerating voltage. The
quantitative analysis of the image contrast deteriorates as the
variation in amplitude of the exit wavefunction increases, because
linearity of image contrasts is preserved only under the WPOA
conditions. The image simulations indicated that the thickness
should be less than 3 nm for an electron beam of 80 kV. However,
we could not find such thin areas suitable for observation at 80 kV
even at the edge of the wedge-shaped crystal. Thus, we employed
an accelerating voltage of 200 kV in our study.
To enhance the image contrast of light elements, we employed
an imaging mode proposed by Lentzen et al. [25]. For this imaging
mode, the CS value was adjusted so that Scherzer′s defocus [51]
and Lichte's defocus [52] are coinciding. The optimum CS is given
by CS ¼ (64/27)λ 3gmax 4, where gmax is the information limit and
λ is the wavelength of the electron beam. At an accelerating voltage of 200 kV, the optimal CS value for our microscope with its
information limit of 10 nm 1 was þ 15 mm. This CS value, combined with the optimal focus condition, not only enhanced the
phase contrast but also minimized image delocalization. The delocalization of the scattered beam for the optimum defocus image
in our study was estimated to be less than 0.059 nm from the CS
and gmax values. Radii of confusion disk and optical parameters for
each CS settings are listed in Table 1. Although confusion disk
under zero CS setting is also small, we employed Letzen's CS setting
as an optimum CS setting for direct imaging of CuPcCl16 molecules
tentatively. When the envelope functions are taken into consideration, phase contrast transfer functions (PCTFs) for each CS
settings vary remarkably in low frequency region as shown in
Fig. 5. Rather than small CS values, defocus value can be more
sensitive to AC-HRTEM imaging of molecular crystals, which have
large crystal units.
To minimize the radiation damage, all images were captured
with the slow-scan CCD camera at a freshly illuminated area of the
crystal just after aligning the microscope under very low dose
conditions ( o10 2 C/cm2 on the specimen) by using the pixelbinning system of the camera. The electron dose on the specimen
for capturing one image with a magnification of 450,000 was
0.4 C/cm2. In order to seek for optimal condition for observations of soft materials, an idea of dose limited resolution (DLR)
reported by Egerton [53] seems to be useful. But it is difficult to
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
77
Fig. 6. (a) A raw AC-HRTEM image of CuPcCl16 (inset is enlargement of boxed area). (b) A diffractogram of image (a). The two arrows indicate the crystallographic directions
(inset is enlargement of boxed reflection). Please note the change of contrast as a function of the defocus condition.
Fig. 7. (a) A diffractogram processed with Wiener and Butterworth filters. (b) A filtered AC-HRTEM image formed by an inverse Fourier transform from (a). (Inset is
enlargement of boxed area). Please note the change of contrast as a function of the defocus condition.
Fig. 8. Schematic of the correlation between specimen tilt and focus condition of
HRTEM image.
find the optimum conditions for actual high-resolution imaging of
organic molecules from the DLR referring characteristic dose (D1/e).
The D1/e values are estimated from decay of arbitrary diffraction
spots which are indexed comparatively low. Molecular structures
can be collapsed severely by electron dose of D1/e, so electron
dosage for high resolution imaging of organic molecules must be
quite smaller than D1/e. It is important to detect higher frequency
components for high-resolution imaging of crystalline samples.
Because the actual conditions are too complicated to optimize the
each parameter, we had selected the condition in which higher
order spot appeared in a diffractogram of acquired image
consequently.
A raw AC-HRTEM image taken at a very thin area and its diffractogram are shown in Fig. 6. The raw image was comparatively
noisy, because of the low electron dose used in image acquisition.
But the highest-order spot in the diffractogram, which corresponded to a 0 20 0 reflection, indicated that the detection limit of
this image was as high as 7.7 nm 1. This value arose from the
high degree of maintenance of the crystal structure during acquisition and/or the high signal-to-noise ratio of the image.
78
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
Fig. 9. (a–d) Images extracted from various areas in a filtered AC-HRTEM image.
The defocus value varied by 7.2 nm in going from (a) to (d). (e–h) Simulated TEM
images with a specimen thickness of 3.4 nm and a CS of þ 15 mm. The number below
each image indicates the defocus value.
Fig. 10. (a–d) Images extracted from various sections of a filtered AC-HRTEM image. The defocus value varied by 7.2 nm in going from (a) to (d). (e–h) Simulated
TEM images with a specimen thickness of 7.9 nm and a CS of þ 15 mm. The number
below each image indicates the defocus value.
However, some image processing had to be applied to the experimental image because the image was too noisy to resolve each
molecular configuration. As shown in Fig. 6(a), the variation in the
image contrast with the focal conditions was clearly visible, even
in the raw image. While the change in thickness affects the image
contrast, we assumed that the thickness of a given sample captured on any one image was uniform, as described later.
from this filtered diffractogram. Compared with the raw image,
piled noise artifacts were removed in the processed image. We
extracted the optimum focused image from the filtered AC-HRTEM
image by exploring along the a axis direction. Owing to the specimen tilt, the focus condition of the image was not uniform, but
rather, changed depending on the position along the a axis direction, as shown in Fig. 8. The optimum focusing area can be
determined extracted by comparing a through-focus image series
in one experimental image and the image simulations. Although
the Bragg filter is very common in HRTEM image processing, we
did not employ it here. A typical Bragg filter, which applies a
periodic two-dimensional net mask to a diffractogram, extracts the
spots in a diffractogram regardless of their frequency. The intensity
of the random noise caused by the amorphous layer depends on
the frequencies of the spots. To prevent the formation of artifacts
3.3. Image processing
In addition to the desired signal, the raw images also included
noise due to blurring by the amorphous carbon layer and quantum
fluctuations in the electron-detecting system (readout noise).
Fig. 7 shows a diffractogram processed by a Wiener and Butterworth filter (cut-off frequency at 4 8 nm 1) and an image derived
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
due to the periodically masked noise in a diffractogram, we opted
to use Wiener filtering in the present study.
3.4. Critical limit of defocus
Typical extracted AC-HRTEM images and simulated TEM images are shown in Fig. 9. The defocus interval, which is equivalent to
the distance between the lower surface of the specimen and the
object plane of the objective lens, of each extracted image was
estimated from the specimen tilt angle and the distance between
the areas along the a axis. We ignored variations in thickness over
the area covered in one TEM image (75 75 nm), because the
wedge angle of CuPcCl16 crystals is very small, as shown in Fig. 2.
The thickness variation in one TEM image could be less than 1 nm,
which is negligible in an HRTEM image of CuPcCl16. Comparison
with simulated through-focus images indicates that the thickness
of the observed area was less than 4 nm. The phthalocyaninato
frameworks consisting of carbon and nitrogen can be seen as a
moderate dark contrast in Fig. 9(b). The five-membered rings in
the organic frameworks were resolved under a narrow defocus
range around the optimum (underfocus: 7.1 nm). Structural images of complex crystals such as molecular crystals are formed by
many Fourier components (Fig. 6a). Furthermore, moderate contrasts of carbon and nitrogen columns are strongly affected even
by small changes in phase contrasts of heavier atomic columns,
such as chlorine and copper. Thus, a structural image of CuPcCl16
could be formed in a small range of defocus around the optimum.
Meanwhile, the images taken around zero focus did not agree
with the simulated images. Since the phase contrast decreases to a
very large extent around zero focus, zero-focus images become
very sensitive to small misalignments. In order to improve estimates of image contrast, a detailed understanding and correct
adjustment of certain optical parameters are important. Furthermore, inelastic scatterings such as thermal diffuse scattering,
which influence the picture of the exit wave, were not taken into
consideration in our simulation. This influence is known to be
more prominent for high-resolution images [54]. Both phasecontrast and amplitude-contrast effects may contribute competitively to the image taken under a defocus between zero and
Scherzer's defocus. To estimate an AC-HRTEM image taken under
very low phase-contrast conditions, we need to know the lens
parameters and exit wavefunctions more accurately.
3.5. Critical limit of thickness
Fig. 10 shows extracted images taken from a comparatively
thick area of a wedge-shaped crystal and the corresponding simulated images. The thickness was found to be 5–9 nm through a
comparison of experimental and simulated images. Relative to
images from a very thin area, the image contrast of organic frameworks in the thicker crystal was enhanced and could be visualized over a wide range of defocus. The image contrast of light
elements became insensitive to the deviation from the optimal
defocus, owing to the increase in phase contrast. On the other
hand, heavy atomic columns such as copper and chlorine in the
thicker region were not observed as simple dark dots even under
optimum defocus conditions. This was a consequence of the variation in the amplitude of the exit wavefunction. As shown in
Fig. 4, with increasing thickness, the electrons converged into each
atomic column position by dynamical scattering. Converged exit
waves are reflected on the HRTEM image as a brighter contrast at
the center of each atomic column, while phase shifts bring a
somewhat diffuse dark contrast depending on the projected potentials. Therefore, heavy atomic columns in thick crystals were
observed not as simple dark dots but as dark tori, although bright
spots in the images of Fig. 10(f) and (g) slant from the center of
79
atomic column locations due to slight misalignment. Especially for
observations under low dose conditions, this complication in the
image contrast is unfavorable because it makes focusing more
difficult. The thickness of the CuPcCl16 crystal should be less than
4 nm for AC-HRTEM imaging with a 200-kV electron beam.
4. Conclusion
Our results demonstrated that the direct imaging of light elements with AC-HRTEM is very sensitive to the defocus value and
specimen thickness. The thickness range permitted for AC-HRTEM
imaging was restricted considerably, as compared with ABF-STEM
imaging [42]. The imaging mode should be set to either ACHRTEM or ABF-STEM according to the type of specimen. ACHRTEM imaging can be applied to a wider variety of organic materials if a novel sample preparation technique can be developed
that reduces the damage during the thinning process.
Acknowledgment
The authors would like to thank Prof. Andrey Chuvilin of Ikerbasque for his valuable contribution in the early stages of this
work.
References
[1] R.R. Lunt, K. Sun, M. Kröger, J.B. Benziger, S.R. Forrest, Ordered organic–organic
multilayer growth, Phys. Rev. B 83 (2011) 064114, http://dx.doi.org/10.1103/
PhysRevB.83.064114.
[2] D.P. Puzzo, M.G. Helander, P.G. O’Brien, Z. Wang, N. Soheilnia, N. Kherani, Z. Lu,
G.A. Ozin, Organic light-emitting diode microcavities from transparent conducting metal oxide photonic crystals, Nano Lett. 11 (2011) 1457–1462, http:
//dx.doi.org/10.1021/nl104036c.
[3] M. Rajeswaran, T.N. Blanton, C.W. Tang, W.C. Lenhart, S.C. Switalski, D.J. Giesen,
B.J. Antalek, T.D. Pawlik, D.Y. Kondakov, N. Zumbulyadis, R.H. Young, Structural,
thermal, and spectral characterization of the different crystalline forms of
Alq3, tris(quinolin-8-olato)aluminum(III), an electroluminescent material in
OLED technology, Polyhedron 28 (2009) 835–843, http://dx.doi.org/10.1016/j.
poly.2008.12.022.
[4] M. Goswami, P. Nayak, N. Periasamy, P.K. Madhu, Characterisation of different
polymorphs of tris(8-hydroxyquinolinato)aluminium(III) using solid-state
NMR and DFT calculations, Chem. Cent. J. 3 (2009) 15, http://dx.doi.org/
10.1186/1752-153X-3-15.
[5] A. Girlando, S. Ianelli, I. Bilotti, A. Brillante, R.G. Della Valle, E. Venuti,
M. Campione, S. Mora, L. Silvestri, P. Spearman, S. Tavazzi, Spectroscopic and
structural characterization of two polymorphs of 1,1,4,4-Tetraphenyl-1,3-butadiene, Cryst. Growth Des. 10 (2010) 2752–2758, http://dx.doi.org/10.1021/
cg100253k.
[6] T. Komeda, H. Isshiki, J. Liu, Y.H. Zhang, N. Lorente, K. Kato, B.K. Breedlove,
M. Yamashita, Observation and electric current control of a local spin in a
single-molecule magnet, Nat. Commun. 2 (2011) 217, http://dx.doi.org/
10.1038/ncomms1210.
[7] T. Ichii, Y. Hosokawa, K. Kobayashi, K. Matsushige, H. Yamada, Molecular-resolution imaging of lead phthalocyanine molecules by small amplitude frequency modulation atomic force microscopy using second flexural mode,
Appl. Phys. Lett. 94 (2009) 133110, http://dx.doi.org/10.1063/1.3114380.
[8] Y.G. Kuznetsov, A.J. Malkin, T.A. Land, J.J. DeYoreo, A.P. Barba, J. Konnert,
A. McPherson, Molecular resolution imaging of macromolecular crystals by
atomic force microscopy, Biophys. J. 72 (1997) 2357–2364, http://dx.doi.org/
10.1016/S0006-3495(97)78880-5.
[9] I. Swart, L. Gross, P. Liljeroth, Single-molecule chemistry and physics explored
by low-temperature scanning probe microscopy, Chem. Commun. 47 (2011)
9011–9023, http://dx.doi.org/10.1039/C1CC11404B.
[10] O. Guillermet, S. Gauthier, C. Joachim, P. de Mendoza, T. Lauterbach,
A. Echavarren, STM and AFM high resolution intramolecular imaging of a
single decastarphene molecule, Chem. Phys. Lett. 511 (2011) 482–485, http:
//dx.doi.org/10.1016/j.cplett.2011.06.079.
[11] K. Stenn, G.F. Bahr, Specimen damage caused by the beam of the transmission
electron microscope, a correlative reconsideration, J. Ultrastruct. Res. 31 (1970)
526–550, http://dx.doi.org/10.1016/S0022-5320(70)90167-X.
[12] W.R.K. Clark, J.N. Chapman, R.P. Ferrier, The structure of α’-copper phthalocyanine and its susceptibility to radiation damage, Nature 277 (1979)
368–370, http://dx.doi.org/10.1038/277368a0.
[13] W.R.K. Clark, J.N. Chapman, A.M. MacLeod, R.P. Ferrier, Radiation damage
80
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
K. Yoshida et al. / Ultramicroscopy 159 (2015) 73–80
mechanisms in copper phthalocyanine and its chlorinated derivatives, Ultramicroscopy 5 (1980) 195–208, http://dx.doi.org/10.1038/277368a0.
N. Uyeda, T. Kobayashi, K. Ishizuka, Y. Fujiyoshi, High voltage electron microscopy for image discrimination of constituent atoms in crystals and molecules, Chem. Scr. 14 (1978&;1979) 47–61.
E.J. Kirkland, Improved high resolution image processing of bright field electron micrographs: I. Theory, Ultramicroscopy 15 (1984) 151–172, http://dx.doi.
org/10.1016/0304-3991(84)90037-8.
E.J. Kirkland, B.M. Siegel, N. Uyeda, Y. Fujiyoshi, Improved high resolution
image processing of bright field electron micrographs: II. Experiment, Ultramicroscopy 17 (1985) 87–103, http://dx.doi.org/10.1016/0304-3991(85)
90002-6.
R.S. Ruskin, Z. Yu, N. Grigorieff, Quantitative characterization of electron detectors for transmission electron microscopy, J. Struc. Biol. 184 (2013)
385–393, http://dx.doi.org/10.1016/j.jsb.2013.10.016.
M. Allegretti, D.J. Mills, G. McMullan, W. Kühlbrandt, J. Vonck, Atomic model of
the F420-reducing [NiFe] hydrogenase by electron cryo-microscopy using a
direct detector, eLife 3 (2014) e01963.
X. Li, P. Mooney, S. Zheng, C.R. Booth, M.B. Braunfeld, S. Gubbens, D.A. Agard,
Y. Cheng, Electron counting and beam-induced motion correction enable nearatomic-resolution single-particle cryo-EM, Nat. Methods 10 (2013) 584–590,
http://dx.doi.org/10.1038/nmeth.2472.
B.E. Bammes, R. Rochat, J. Jakana, D.H. Chen, W. Chiu, Direct electron detection
yields cryo-TEM reconstructions at resolution beyond 3/4 Nyquist frequency, J.
Struct. Biol. 177 (2012) 589–601, http://dx.doi.org/10.1016/j.jsb.2012.01.008.
C.L. Jia, M. Lentzen, K. Urban, Atomic-resolution imaging of oxygen in perovskite ceramics, Science 299 (2003) 870–873, http://dx.doi.org/10.1126/
science.1079121.
Z. Zhang, W. Sigle, F. Phillipp, M. Rühle, Direct atom-resolved imaging of
oxides and their grain boundaries, Science 302 (2003) 846–849, http://dx.doi.
org/10.1126/science.1089785.
C.L. Jia, K. Urban, Atomic-resolution measurement of oxygen concentration in
oxide materials, Science 303 (2004) 2001–2004.
C.L. Jia, S.B. Mi, K. Urban, I. Vrejoiu, M. Alexe, D. Hesse, Atomic-scale study of
electric dipoles near charged and uncharged domain walls in ferroelectric
films, Nat. Mater. 7 (2008) 57–61, http://dx.doi.org/10.1038/nmat2080.
M. Lentzen, B. Jahnen, C.L. Jia, A. Thust, K. Tillmann, K. Urban, High-resolution
imaging with an aberration-corrected transmission electron microscope, Ultramicroscopy 92 (2002) 233–242, http://dx.doi.org/10.1016/S0304-3991(02)
00139-0.
M. Lentzen, Progress in aberration-corrected high-resolution transmission
electron microscopy using hardware aberration correction, Microsc. Microanal. 12 (2006) 191–205, http://dx.doi.org/10.1017/S1431927606060326.
Z. Lee, J.C. Meyer, H. Rose, U. Kaiser, Optimum HRTEM image contrast at 20 kV
and 80 kV – Exemplified by graphene, Ultramicroscopy 112 (2012) 39–46.
S. Kurasch, J. Kotakoski, O. Lehtinen, V. Skákalová, J. Smet, C.E. Krill, A.
V. Krasheninnkikov, U. Kaiser, Atom-by-Atom observation of grain boundary
migration in graphene, Nano Lett. 12 (2012) 3168–3173, http://dx.doi.org/
10.1021/nl301141g.
B. Westenfelder, J.C. Meyer, J. Biskupek, S. Kurasch, F. Scholz, C.E. Krill,
U. Kaiser, Transformations of carbon adsorbates on graphene substrates under
extreme heat, Nano Lett. 11 (2011) 5123–5127, http://dx.doi.org/10.1021/
nl203224z.
J. Kotakoski, J.C. Meyer, S. Kurasch, D. Santos-Cottin, U. Kaiser, A.
V. Krasheninnikov, Stone-Wales-type transformations in carbon nanostructures driven by electron irradiation, Phys. Rev. 83 (2011) 245420, http:
//dx.doi.org/10.1103/PhysRevB.83.245420.
J. Kotakoski, A.V. Krasheninnikov, U. Kaiser, J.C. Meyer, From point defects in
graphene to two-dimensional amorphous carbon, Phys. Rev. Lett. 106 (2011)
105505, http://dx.doi.org/10.1103/PhysRevLett.106.105505.
C.O. Girit, J.C. Meyer, R. Erni, M.D. Rossell, C. Kisielowski, L. Yang, C.H. Park, M.
F. Crommie, M.L. Cohen, S.G. Louie, A. Zetti, Graphene at the edge: stability and
dynamics, Science 323 (2009) 1705–1708, http://dx.doi.org/10.1126/
science.1166999.
O.L. Krivanek, N. Dellby, M.F. Murfitt, M.F. Chisholm, T.J. Pennycook,
K. Suenaga, V. Nicolosi, Gentle STEM: ADF imaging and EELS at low primary
energies, Ultramicroscopy 110 (2010) 935–945, http://dx.doi.org/10.1016/j.
ultramic.2010.02.007.
[34] K. Yoshida, Y. Sasaki, Optimal accelerating voltage for HRTEM imaging of
zeolite, Microscopy 62 (2013) 369–375, http://dx.doi.org/10.1093/jmicro/
dfs087.
[35] L. Reimer, Transmission Electron Microscopy, fourth edition, Springer Verlag,
Berlin, Heidelberg, 1997.
[36] M.C.R. Symons, Chemical aspects of electron-beam interactions in the solid
state, Ultramicroscopy 10 (1982) 15–23, http://dx.doi.org/10.1016/0304-3991
(82)90183-8.
[37] S.D. Findlay, T. Saito, N. Shibata, Y. Sato, J. Matsuda, K. Asano, E. Akiba,
T. Hirayama, Y. Ikuhara, Direct imaging of hydrogen within a crystalline environment, Appl. Phys. Express 3 (2010) 116603, http://dx.doi.org/10.1143/
APEX.3.116603.
[38] S.D. Findlay, N. Shibata, H. Sawada, E. Okunishi, Y. Kondo, Y. Ikuhara, , Dynamics of annular bright field imaging in scanning transmission electron
microscopy, Ultramicroscopy 110 (2010) 903–923, http://dx.doi.org/10.1016/j.
ultramic.2010.04.004.
[39] R. Ishikawa, E. Okunishi, H. Sawada, Y. Kondo, F. Hosokawa, E. Abe, Direct
imaging of hydrogen-atom columns in a crystal by annular bright-field electron microscopy, Nat. Mater. 10 (2011) 278–281, http://dx.doi.org/10.1038/
nmat2957.
[40] E. Okunishi, I. Ishikawa, H. Sawada, F. Hosokawa, M. Mori, Y. Kondo, Visualization of light elements at ultrahigh resolution by STEM annular bright field
microscopy, Microsc. Microanal. 15 (2009) 164–165, http://dx.doi.org/10.1017/
S1431927609093891.
[41] R. Huang, T. Hitosugi, S.D. Findlay, C.A.J. Fisher, Y.H. Ikuhara, H. Moriwake,
H. Oki, Y. Ikuhara, Real-time direct observation of Li in LiCoO2 cathode material, Appl. Phys. Lett. 98 (2011) 051913, http://dx.doi.org/10.1063/1.3551538.
[42] M. Haruta, H. Kurata, Direct observation of crystal defects in organic molecular
crystals of copper hexadecachlorophthalocyanine by STEM-EELS, Sci. Rep. 2
(2012) 1–4, http://dx.doi.org/10.1038/srep00252.
[43] Y. Fujiyoshi, T. Kobayashi, K. Ishizuka, N. Uyeda, Y. Ishida, Y. Harada, A new
method for optimal-resolution electron microscopy of radiation–sensitive
specimens, Ultramicroscopy 5 (1980) 459–468, http://dx.doi.org/10.1016/
0304-3991(80)90046-7.
[44] M. Malac, M. Beleggia, R. Egerton, Y. Zhu, Imaging of radiation-sensitive
samples in transmission electron microscopes equipped with Zernike phase
plates, Ultramicroscopy 108 (2008) 126–140, http://dx.doi.org/10.1016/j.
ultramic.2007.03.008.
[45] R. Danev, K. Nagayama, Optimizing the phase shift and the cut-on periodicity
of phase plates for TEM, Ultramicroscopy 111 (2011) 1305–1315, http://dx.doi.
org/10.1016/j.ultramic.2011.04/004.
[46] Y. Saito, Epitaxial growth mechanism of chlorinated copper phthalocyanine on
KCl surfaces, Appl. Surf. Sci. 22/23 (1985) 574–581, http://dx.doi.org/10.1016/
0378-5963(85)90188-6.
[47] N. Uyeda, T. Kobayashi, E. Suito, Y. Harada, M. Watanabe, Molecular image
resolution in electron microscopy, J. Appl. Phys. 43 (1972) 5181–5189, http:
//dx.doi.org/10.1063/1.1661094.
[48] R. Kilaas, Optimal and near-optimal filters in high-resolution electron microscopy, J. Microsc. 190 (1998) 45–51, http://dx.doi.org/10.1046/
j.1365-2818.1998.3070861.x.
[49] J.M. Cowley, A.F. Moodie, The scattering of electrons by atoms and crystals. I. A
new theoretical approach, Acta Cryst. A 28 (1957) 609–619, http://dx.doi.org/
10.1107/S0365110 57002194.
[50] Y. Saito, M. Shiojiri, Molecular energetic of the epitaxial growth of chlorinated
copper phthalocyanine on KCl surfaces, J. Cryst. Growth 67 (1984) 91–96, http:
//dx.doi.org/10.1016/0022-0248(84)90135-0.
[51] O. Scherzer, The theoretical resolution limit of the electron microscope, J. Appl.
Phys. 20 (1949) 20–29, http://dx.doi.org/10.1063/1.1698233.
[52] H. Lichte, Optimum focus for taking electron holograms, Ultramicroscopy 38
(1991) 13–22, http://dx.doi.org/10.1016/0304-3991(91)90105-F.
[53] R.F. Egerton, Choice of operating voltage for a transmission electron microscope, Ultramicroscopy 145 (2014) 85–93, http://dx.doi.org/10.1016/j.
ultramic.2013.10.019.
[54] Z.L. Wang, Thermal diffuse scattering in high-resolution electron holography,
Ultramicroscopy 52 (1993) 504–511, http://dx.doi.org/10.1016/0304-3991(93)
90067-8.