scispace - formally typeset

Journal ArticleDOI

Image formation mechanisms in scanning electron microscopy of carbon nanotubes, and retrieval of their intrinsic dimensions.

01 Jan 2013-Ultramicroscopy (Elsevier)-Vol. 124, pp 35-39

TL;DR: It is shown how SEM images can be modelled by accounting for surface enhancement effects together with the absorption coefficient for secondary electrons, and the electron-probe shape, enabling retrieval of the intrinsic nanotube dimensions.
Abstract: We present a detailed analysis of the image formation mechanisms that are involved in the imaging of carbon nanotubes with scanning electron microscopy (SEM). We show how SEM images can be modelled ...

Content maybe subject to copyright    Report

http://www.diva-portal.org
Preprint
This is the submitted version of a paper published in Ultramicroscopy.
Citation for the original published paper (version of record):
Jackman, H., Krakhmalev, P., Svensson, K. (2013)
Image formation mechanisms in scanning electron microscopy of carbon nanotubes,and retrieval
of their intrinsic dimensions..
Ultramicroscopy, 124: 35-39
Access to the published version may require subscription.
N.B. When citing this work, cite the original published paper.
Permanent link to this version:
http://urn.kb.se/resolve?urn=urn:nbn:se:kau:diva-16425

Image formation mechanisms in scanning electron microscopy of carbon nanotubes,
and retrieval of their intrinsic dimensions.
H. Jackman
a,
, P. Krakhmalev
b
, K. Svensson
a
a
Department of Physics and Electrical Engineering, Karlstad University, SE-651 88 Karlstad, Sweden
b
Department of Mechanical and Materials Engineering, Karlstad University, SE-651 88 Karlstad, Sweden
Abstract
We present a detailed analysis of the image formation mechanisms that are involved in the imaging of carbon nanotubes
with scanning electron microscopy (SEM). We show how SEM images can be modelled by accounting for surface enhance-
ment effects together with the absorption coefficient for secondary electrons, and the electron-probe shape. Images c an
then be deconvoluted, enabling retrieval of the intrinsic nanotube dimensions. Accurate estimates of their dimensions
can thereby be obtained even for structur es that are compa rable to the electr on-probe size (on the order of 2 nm). We
also present a simple and robust mo del for obtaining the outer diameter of nanotubes without any detailed knowledge
about the electron-probe shape .
1. Introduction
The development of scanning probe microscopy (SPM)
instruments combined with sca nning electron microscopy
(SEM) and transmission electron microscopy (TEM) en-
ables free standing nanos tructures (such as carbon nan-
otub es) to be studied in terms of their mechanical and
electrical properties. The microscope is mainly used to ob-
tain structural information while the SPM instruments are
used to perform the material analysis by applying forces or
electrical currents [1]. The space available inside SEM in-
struments is comparatively large and several probes can
be used [2]. In TEM the space is much more limited
and usually only one SPM probe is fitted [1], while two
probes have also recently been demonstrated [3]. One ma-
jor drawback with the SEM is that the resolution is limited
by the electron-probe shape and the imaging mechanisms.
Nanoscale structures are thereby substantially broadened
in SEM-images.
An accurate measure of the nanotube dimensions is
crucial for the analys is of mechanical, and electrical, prop-
erties. One route is then to combine the SEM in-situ anal-
ysis with regular TEM imaging [4, 5]. The material test-
ing would then first be performed inside an SEM, and the
sample of interest subsequently transferred into a TEM
to obtain the detailed structural informatio n. Apart from
providing only post-mortem information, the method is
also difficult, as the exact sample loc ation and orienta-
tion should prefera bly be analyzed ins ide both the SEM
and TEM. This r equires a transfer o f the whole sample,
or SPM probe, with remnant carbon nanotubes on it [4].
Corresponding author
Email address: henrik.jackman@kau.se (H. Jackman)
Alternatively one can also tr ansfer parts of the analyzed
carbon nanotube onto a regular TEM-grid for subsequent
TEM analysis [5]. It would instead be preferable to do all
of the analysis inside an SEM instrument [6].
The smallest electron-probe size in a modern, field emis-
sion gun (FEG) SEM can be of the order of 1 nm a t
optimum settings [7], and individual car bon nanotubes
can thereby fairly easy be imaged, but the SEM shows
a much broader structure tha n the true one (as obtained
from TEM imaging). The problem is illustrated in Fig. 1
where the same carbon nanotube has be e n imaged both
in an SEM (Fig. 1 (a)) and a TEM (Fig. 1 (b)). If the
broadening in the SEM imag e ca n be taken accounted for,
then the whole analysis could be done from o nly in-situ
SEM expe riments.
In this work we show how the SEM image formation
can be fully modelled for nanofibres , such as carbon nan-
otub es, by a convolution of the secondary electron yield
with the electron-probe shape. With a known electron-
probe shape the SEM image can then b e deconvoluted and
reveal more details. We also present an easy way to obtain
the outer diameter of carbon nanotubes ima ged in SEM,
without doing the full dec onvolution. The simple model
is fairly robust with respect to the detailed electron-probe
shape and enables an accurate estimation of the nanotube
outer diameter. A full investigation can thereby be per-
formed inside the SEM without the need for complemen-
tary TEM imaging.
2. Materials and Methods
2.1. Sample preparation
Two types of commercial CNTs have been studied:
NC2100 and NC2101, obtained from Nanocyl. Both types
Preprint submitted to Elsevier April 20, 2012

(a)
(b)
r
r
−6 −4 −2 0 2 4 6
−1
−0.5
0
0.5
1
Position, r [nm]
a.u
SEM profile
TEM profile
d
TEM
(c)
Figure 1: Images of the same CNT acquired by (a) SEM and (b)
TEM. (c) Shows the integrated intensity profiles from the two rect-
angles in (a) and (b), and the diameter estimated from the TEM
profile is indicated.
were produced by catalytic chemical vapor deposition (CCVD)
with the difference being that NC2101 was functionalized
with a carboxyl group (-COOH) to reduce bundling. The
CNTs are marketed as double-walled with an average di-
ameter of 3.5 nm and lengths varying between 1-10 µm.
As received CNTs were first dispersed in ethano l and then
sonicated for 15 minutes to reduce bundling. Sonication
for lo nger periods of time can introduce de fects so that
the concentric cylinder structure is broken [8]. No dam-
ages could be seen in TEM images of the CNTs due to the
sonication.
The solution containing the CNTs was dro p-casted onto
a holey carbon support film for TEM (R 2/1 produced by
Quantifoil) and then allowed to dry. CNTs were in this
way distributed uniformly over the film enabling imaging
of single CNTs without any underlying substrate.
2.2. Microscopy
The samples were first studied in a JEOL (JEM 210 0)
TEM equipped with a LaB
6
cathode and a digital camera
from Gatan (SC1000 Orius). Single CNTs, with well de-
fined concentric cylinder structure and with no amorphous
carbon deposits, were chosen and imaged. The acceler-
ation voltage used was 100 kV, since imag ing at higher
voltages can c ause dama ge to the CNTs even if they a re
imaged for short periods of time [9]. Even at 100 kV the
CNTs ca n b ecome damag ed if subjected to a high current
density beam for longer times. Therefore the current den-
sity of the beam and the time of exposure to the beam
were minimized so that the CNTs did not g et damaged.
The best performance of the TEM is obtained at the
Scherzer defocus [10]. One result of this defocus is that
bright Fres ne l fringes appear at the edges o f the CNT,
as shown in Fig. 1 (b) and (c). From intensity profiles,
obtained from TEM images, the diameter, d
T EM
, was esti-
mated by taking the distance between the points where the
bright Fresnel contrast returns to background intensity, as
shown in Fig. 1 (c). The width of the dark regions varies
with focus, but for small deviations from the Scher zer de-
focus they stay approximately constant [11]. Estimating
the diameter this way yields the width of the CNT that
interacts with the electron beam, both in SEM and TEM,
i.e the outer diameter. The error whe n estimating the di-
ameter o f a CNT from TEM images is less than 10% for
diameters larger than 1 nm [11]. Since all CNTs used in
this study had d > 1 nm, the diameters measured in the
TEM should deviate less tha n 10 % from the true diameter.
After the CNTs had been imaged in the TEM they
were imaged in a LEO 1530 FEG-SEM. The samples were
mounted in a custom ma de scanning transmission electron
microscopy holder enabling imaging of the CNTs with-
out any contribution from an underlying substrate. T he
electron-probe size and the secondary electron yield bo th
decrease with increasing acceleration voltage [7]. A higher
acceleration voltage thereby improves the resolution but
decreases the signal-to-noise ration in the SEM image. An
acceleration voltage of 12 kV was found to be a good trade-
off between resolutio n and signal-to-noise ratio and was
used when acquiring all images in this study.
Electron beam-induced deposition (EBID) can be a
problem during SEM analysis and it will increase the di-
ameter of the CNTs [12]. Such effects were minimized
here by leaving the samples in the SEM chamber for at
least 10 hours, r e sulting in a chamber pressure down to
5×10
7
mbar, befor e exposing them to the electron beam.
To check if EBID had occurred in the SEM, all CNTs were
imaged one more time in the TEM and the EBID was
found to be ne gligible.
2.3. Modelling
An SEM image can be thought of as the detected sec-
ondary electron (SE) intensity as a function of the lateral
coordinates, I(r). This can be described as a convolution
of the SE yield at each sample position, δ(r), with the
electron-probe shape, i(r):
I(r) = [δ i](r) =
Z
δ(r
0
)i(r r
0
)dr
0
(1)
At low magnifications, the influence of i(r) on the recorded
image can be neglected, while at higher magnifications i(r)
will smooth out any sharp sample details. The exact shape
of i(r) is g enerally unknown and depends on a number of
parameters in the micros cope. Estimation of CNT diame-
ters can be done on integrated intensity profiles, as shown
2

in Fig. 1 (c). In these intensity profiles, a radial s ymmet-
ric i(r) can be described as one dimensional and Gaussian
and Lorentzian functions can be used to describ e i(r). An
area normalised Gaussian function is given by:
i
G
(r) =
1
σ
2π
exp
r
2
2σ
2
(2)
where the full width at half maximum (Γ) is equal to Γ
G
=
2
2 ln 2σ , while a n area normalised Lorentzian function is
given by:
i
L
(r) =
1
π
1
2
Γ
L
r
2
+ (
1
2
Γ
L
)
2
(3)
where Γ
L
is the full width at half maximum.
The SE yield from a homogeneous material depends on
the generation and scattering of SE. Three types of SE are
generated inside an SEM: SE1 , SE2 and SE3 [7]. Electrons
labelled SE1 are excited directly by the primary beam,
SE2 are excited by backscattered electrons (BSE) near the
specimen surface, and SE3 are excited by backscattered
electrons outside the specimen. When the e xperimental
SEM images wer e obtained in this study an in-lens de-
tector was used which effectively excludes SE3 [13], and
the contribution from SE3 was neglected when modelling
I(r). In order for SE2 to be emitted, there must be a
considerable amount of backscattered electrons from the
incoming beam. Considering CNTs with a diameter up
to about 10 nm, an insignificant frac tion of the incident
beam will be backscattered, and the SE2 contribution was
also neg le cted here. The generation of SE1 is assumed to
be propo rtional to the energy loss of the incoming elec-
trons [14, 7]. Having a thin specimen with a low average
atomic number, such as a CNT, the incident electrons will
lose a negligible amount of energy travelling thro ugh the
specimen. This makes the probability o f generating SE
approximately constant throughout the thickness o f the
sample.
For SE to be detected they ne e d to escape the speci-
men and reach the detector. Because of their low energy,
the SE are more easily absorbed than the incoming high-
energy electrons. The probability of escaping a specimen
decreases exp onentially with the distance travelled in the
solid, z:
P
escape
e
αz
(4)
with α being the abso rption coefficient fo r SE of the speci-
men material. Combining the relations for ge neration and
escape of the SE an expression for the SE yield, as a func-
tion of the thickness t , was derived in [14]:
δ
t
1
2α
1 e
αt
(1 αt) (αt)
2
Z
αt
e
ξ
ξ
(5)
where the thickness can be a function of the lateral coor-
dinates, t = t(r).
For a flat specimen of constant tilt ang le , φ, Eq. 5
describe the SE yie ld. However for a CNT, φ w ill vary
across its surface and δ incre ases with increasing φ. The
relation between δ and φ is given approximately by [16]:
δ
φ
sec φ (6)
where the tilt angle can be a function of the lateral co-
ordinates, φ = φ(r). For a spe cimens with low average
atomic number, Z, the dependence on φ becomes more
rapid than sec φ [16]. On the other hand, the dependence
also becomes slower with lower V
SEM
, so although carbon
has a low Z, Eq. 6 is a reasonable approximation of the
depe ndence of δ on φ since we us e d V
SEM
= 12 kV in this
study.
By combining Eq. 5 and 6, the total SE yield is given
by:
δ
sim
(r) = δ
t
(r)δ
φ
(r) (7)
−6 −4 −2 0 2 4 6
−2
−1
0
1
2
3
Position, r [nm]
a.u
data3
CNT
δ
sim
, α=10
−1
nm
−1
δ
sim
, α=20
−1
nm
−1
(a)
−6 −4 −2 0 2 4 6
0
0.2
0.4
0.6
0.8
1
Position, r [nm]
a.u
δ
sim
*i
L
δ
sim
*i
G
(b)
Figure 2: (a) The total SE yi el d, δ
sim
, for two different α-values
(solid lines), calculated for the CNT dimensions indicated by the
dashed lines. (b) Two simulated intensity profiles, I
sim
, for the CNT
dimensions in (a) using i
sim
= i
G
and i
sim
= i
L
with Γ
G
= Γ
L
= 2.1
nm (for α = 1/20 nm
1
).
To obtain δ
sim
(r) for a carbon nanotube we need a
value for the absorption co efficient α. Experimental data
i scarce, but α may be as low as 1/20 nm
1
for carbon
and around 1/10 nm
1
for polymers [15]. In Fig. 2 (a) we
show the obtained δ
sim
for both values of α. T he influence
of α on δ
sim
is small and we have used the value fo r carbo n
(1/20 nm
1
) for the remaining analysis.
Simulated intensity profiles, I
sim
(r), were obtained through
a convolution of δ
sim
and i
sim
, where linear combinations
of i
G
and i
L
were used to describe the probe shape i
sim
.
The convolution was done by sweeping a function of the
probe shape, i
sim
(r r
0
), over the total SE yield, δ
sim
(r
0
),
3

and taking the integral of the product of these two func-
tions:
I
sim
(r) = [δ
sim
i
sim
](r) =
r+5Γ
Z
r
δ
sim
(r
0
)i
sim
(rr
0
)dr
0
(8)
Ultimately the integration above s hould be performed in
the interval [−∞, ], but since the equations describing
i(r) rapidly goes to zero we used the interval [r5Γ, r+5Γ]
to reduce the computational intensity. The integration in
Eq. 8 was performed numeric ally in MATLAB, were both
i
sim
(r r
0
) and δ
sim
(r
0
) were divided into small elements.
Pair of e le ments were multiplied and the products summed
together in each step. An example of two simulated inten-
sity profiles is shown in Fig. 2 (b), for purely Gauss ian
and Lorentzian probe shapes.
This method is not limited to CNTs, as it can be ap-
plied on any homogeneous nanofiber structure. In the case
of filled nanotubes it can easily be used by adding the δ
contribution of the filler to Eq. 7.
3. Results and Discussion
The inner and outer diameter of the CNTs obtained
from the TEM images was used as parameters when sim-
ulating intensity profiles, I
sim
. Good agreement between
I
sim
and experimental SEM intensity profiles (I
SEM
) was
obtained whe n using a c ombination of i
G
and i
L
to model
the probe shape (with Γ
G
= Γ
L
= Γ). On average a goo d
agreement was fo und when:
i
sim
=
1
2
i
L
+
1
2
i
G
(9)
A result of such a simulation is shown in Fig. 3 (c), where
the SEM profile of the CNT seen in Fig. 3 (b) was simu-
lated. The average Γ from all analysed profiles was found
to be 2.05 ± 0.05 nm.
The probe shape that worked best for reproducing I
SEM
should not be seen as an instrumental constant, since i(r)
depe nd on a number of SEM parameters. Therefore i(r)
likely varies between differ ent SEM instruments and even
for a certain SEM as the experimental conditions are al-
tered, by changes in foc us and a stigmatic corr e ctions. Know-
ing the exact probe shape, one can deconvolute a full SEM
image to obtain the SE yield, and in turn from that re-
trieve the intrinsic na notube dimensions. We have used
the two dimensional version of Eq. 9 as the input point
spread function in the MATLAB-function deconvblind to
deconvolute SEM images. The r e sult o f one such decon-
volution is shown in Fig. 4, where Fig. 1 (a) was used as
the raw SEM-image.
Even though the deconvoluted image in Fig. 4 (b)
contains more deta il, it is still quite far from the sharp
function describing δ in Eq. 7 (as can be seen in Fig. 4 (c)).
One reason that the deconvolution cannot fully retrieve
the underlying str uctures, is the image no ise in the shape
r
−6 −4 −2 0 2 4 6
−0.2
0
0.2
0.4
0.6
0.8
1
1.2
Position, r [nm]
a.u
i
L
0.5i
L
+0.5i
G
i
G
I
SEM
(c)
Figure 3: Images of the same CNT acquired by (a) TEM and (b)
SEM. In (c) the SEM profile, I
SEM
(·), is plotted along with three
simulated intensity profiles, obtained by using different probe shapes
(with Γ
G
= Γ
L
= 2.0 nm).
of streaks. This may be the res ult of vibrations or other
sources of instabilities, s uch as external electromagnetic
fields. Another reason is that Eq. 9 is an approximation
of the exact probe shape i (r). Other attempts to describe
the probe shape assumes a Gaussian probe shape and fails
to determine probe sizes below than 2 nm [17].
Instead of having to rely on the exact shape of i(r),
we have looked for an ea sier way to extract the outer di-
ameter of the nanotubes which is an important parameter
when studying the mechanical properties. We have pr evi-
ously shown, from empirical studies, that the width of the
SEM profile can be used to estimate the outer diameter
of the nanotubes [6]. Here we investigate this further by
using our knowledge of the image formation mechanisms.
The electron-yield from a nanotube has a sharp step at
the positions of the outer edges (see Fig. 2 (a)) which are
the positions that will give us the outer diameter. If we
look at the convolution of a step function, H(r r
0
), and
an arbitrary symmetric probe shape with a distinct maxi-
mum, I
H
(r) = [H i](r), the second derivative,
d
2
dr
2
I
H
(r),
becomes zero when the maximum of i(r) meets the edge
of the step function at r
0
. If the SE yield of a CNT would
be constant across the width, then the diameter would
4

Citations
More filters

DOI
01 Jan 2014-
TL;DR: Based on the current report, JRC will prepare a set of recommendations for the revision of the EC nanomaterial definition, as part of the review process foreseen in the 2011 EC Recommendation.
Abstract: This report provides the JRC assessment of feedback on the experiences of stakeholders with the EC nanomaterial definition, published in 2011 (EC Recommendation 2011/696/EU). The report is a follow-up report of the previous JRC report (EUR 26567 EN, 2014), which compiled feedback collected by JRC in 2013 and early 2014, partly through a dedicated survey. Based on the current report, JRC will prepare a set of recommendations for the revision of the EC nanomaterial definition, as part of the review process foreseen in the 2011 EC Recommendation. JRC Scientific and Policy Report Towards a review of the EC Recommendation for a definition of the term "nanomaterial" Part 2: Assessment of collected information concerning the experience with the definition

49 citations


Journal ArticleDOI
Abstract: We have studied the mechanical behavior of multi-walled carbon nanotubes for bending strains beyond the onset for rippling and buckling. We found a characteristic drop in the bending stiffness at the rippling and buckling onset and the relative retained stiffness was dependent on the nanotube dimensions and crystallinity. Thin tubes are more prone to buckle, where some lose all of their bending stiffness, while thicker tubes are more prone to ripple and on average retain about 20% of their bending stiffness. In defect rich tubes, the bending stiffness is very low prior to rippling, but these tubes retain up to 70% of their initial bending stiffness.

14 citations


Dissertation
01 Feb 2015-
Abstract: The main aim of this project was to develop, formulate, characterise and optimise novel pre-formed thin polymer film that will deliver therapeutically relevant drugs via the buccal mucosa route of paediatric patients, using OME as model drug. The development focused on obtaining formulations with optimized drug loading, drug release and permeation, stability and low toxicity. Five different film forming polymers hydroxypropyl methylcellulose (HPMC), methylcellulose (MC), sodium alginate (SA), carrageenan (CA) and metolose (MET) were used initially and subsequently with polyethylene glycol (PEG 400) as plasticiser and L-arg (to stabilise OME). Polymeric gels (1% w/w) were prepared using water and ethanol (10% v/v and 20% v/v) as the casting solvents with PEG 400 at different concentrations (0 and 0.5 % w/w) and the films were obtained by drying the gels in an oven (40 °C). SA and MET films were chosen for drug loading and further investigation (OME stabilisation). These films showed a good balance between flexibility and toughness required for ease of transportation and patient handling. Drug loaded gels showed that OME was unstable, with gels turning red after 20 minutes and therefore required addition of L-arg. From the results obtained, plasticised (0.5 % w/w PEG 400) MET films prepared from ethanolic (20% v/v) gels and containing OME : L-arg ratio of 1:2 showed the most ideal characteristics (transparency, ease of peeling and flexibility) and was the formulation of choice for further investigation. Results obtained for the hydration and in vitro mucoadhesion studies showed that plasticised films had higher swelling capacity and mucoadhesivity than unplasticised films. In addition, BLK films showed higher swelling index and adhesion than DL films, whilst gelatine equilibrated with PBS showed higher values compared with simulated saliva (SS). Dissolution data from optimised DL MET films showed OME release was sustained over 1 hour. Fitting the release data to kinetic models showed that the Korsmeyer-Peppas equations best fit the dissolution data for both PBS and SS media. The permeability profile of optimised DL film using pig buccal tissue, showed that the amount of OME permeating over 2 hours was 275ug/cm2 suggesting that pig buccal membrane is generally quite permeable and also that the OME is released from the films. Application of SCF caused significant changes to the functional and physical properties of the MET films and converted the original DL MET films from a sustained release formulation (1 hour) to a rapid release system, releasing > 90% of OME within 15 minutes and the release of OME from these films followed Higuchi kinetic model. Finally, incorporation of β cyclodextrin (βCD) into DL MET films containing OME:L-arg 1:1, improved the stability of the drug over 28 days under ambient conditions compared to 14 days for the corresponding DL MET films containing only L-arg at a higher loading (OME: L-arg 1:2). The optimised formulations have potential as paediatric buccal delivery system for OME.

2 citations


Journal ArticleDOI
A. A. Dmitrievskiy1Institutions (1)
Abstract: Regularities of changes in the mechanical properties (micro- or nanohardness, fracture toughness at indentation, and steady-state creep rate) and electrical characteristics (Hall constant, conductivity, and concentration of electrically active defects) of silicon single crystals under low-dose (F < 1012 cm–2) low-intensity (I ~ 106 cm–2∙s–1) beta-irradiation are described. The mechanism of nonmonotonic beta-induced softening of silicon is discussed.

1 citations


Book ChapterDOI
Weiguo Bian1, Mingyu Wang1, Zhan Yang1Institutions (1)
08 Aug 2019-
TL;DR: It is concluded from a large number of tests that when the secondary electron image gray histogram has obvious double peaks and is located in the trough, the threshold obtained is optimal and it is possible to better observe the pictures under the SEM.
Abstract: Observing the sample under a scanning electron microscope (SEM) requires adjustment of brightness and contrast to obtain a clear image. The traditional method is manually adjusted by the operator, which inevitably has errors. In this paper, an adaptive threshold processing method based on image-based normalized gray histogram is proposed. This method can acquire the threshold of the image according to the state of the currently obtained secondary electron images. When the brightness and contrast of the image change, the threshold can also be changed accordingly. It is concluded from a large number of tests that when the secondary electron image gray histogram has obvious double peaks and is located in the trough, the threshold obtained is optimal. Therefore, it is possible to better observe the pictures under the SEM.

References
More filters

Journal ArticleDOI
Min-Feng Yu1, Oleg R. Lourie1, Mark J. Dyer, Katerina Moloni2  +2 moreInstitutions (2)
28 Jan 2000-Science
TL;DR: The tensile strengths of individual multiwalled carbon nanotubes (MWCNTs) were measured with a "nanostressing stage" located within a scanning electron microscope and a variety of structures were revealed, such as a nanotube ribbon, a wave pattern, and partial radial collapse.
Abstract: The tensile strengths of individual multiwalled carbon nanotubes (MWCNTs) were measured with a “nanostressing stage” located within a scanning electron microscope. The tensile-loading experiment was prepared and observed entirely within the microscope and was recorded on video. The MWCNTs broke in the outermost layer (“sword-in-sheath” failure), and the tensile strength of this layer ranged from 11 to 63 gigapascals for the set of 19 MWCNTs that were loaded. Analysis of the stress-strain curves for individual MWCNTs indicated that the Young's modulus E of the outermost layer varied from 270 to 950 gigapascals. Transmission electron microscopic examination of the broken nanotube fragments revealed a variety of structures, such as a nanotube ribbon, a wave pattern, and partial radial collapse.

4,691 citations


BookDOI
David B. Williams1, C. Barry Carter1Institutions (1)
01 Jan 1996-
TL;DR: The client would like to get a larger, approximately 3 cm in diameter, well fixed tissue sample, together with a detailed report of the clinical presentation, gross, and microscopic lesions, along with the submission of samples prepared in a similar manner by the client for processing.
Abstract: We wrote it to be read by, and taught to, senior undergraduates and starting graduate students, rather than studied in a research laboratory. We wrote it using the same style and sentence construction that we have used in countless classroom lectures, rather than how we have written our countless (and much-less read) formal scientificpapers. In this respect particularly, wehave been deliberate in notreferencing the sources of every experimental fact or theoretical concept (although we do include some hints and clues in the chapters). However, at the end of each chapter we have included groups of references that should lead you to the best sources in the literature and help you go into more depth as you become more confident about what you are looking for. We are great believers in the value of history as the basis for under- standing the present and so the history of the techniques and key historical references are threaded throughout the book. Just because a reference is dated in the previous century (or even the antepenultimate century) doesn’t mean it isn’t useful! Likewise, with the numerous figures drawn from across the fields of materials science and engineering and nanotechnology, we do not reference the source in each caption. But at the very end of the book each of our many generous colleagues whose work we have used is clearly acknowledged.

4,259 citations


Book
01 Jan 1984-
Abstract: Electron Optics of a Scanning Electron Microscope.- Electron Scattering and Diffusion.- Emission of Backscattered and Secondary Electrons.- Electron Detectors and Spectrometers.- Image Contrast and Signal Processing.- Electron-Beam-Induced Current and Cathodoluminescence.- Special Techniques in SEM.- Crystal Structure Analysis by Diffraction.- Elemental Analysis and Imaging with X-Rays.

966 citations


Journal ArticleDOI
Florian Banhart1Institutions (1)
Abstract: The paper reviews the principles of interaction of energetic particles with solid carbon and carbon nanostructures. The reader is first introduced to the basic mechanisms of radiation effects in solids with particular emphasis on atom displacements by knock-on collisions. The influence of various parameters on the displacement cross sections of carbon atoms is discussed. The types of irradiation-induced defects and their migration are described as well as ordering phenomena which are observable under the non-equilibrium conditions of irradiation. The main part of this review deals with alterations of carbon nanostructures by the electron beam in an electron microscope. This type of experiment is of paramount importance because it allows in situ observation of dynamic processes on an atomic scale. In the second part, radiation effects in the modifications of elemental carbon, in particular in graphite which forms the crystallographic basis of most carbon nanostructures, are treated in detail. It follows a review of the available experimental results on radiation defects in carbon nanostructures such as fullerenes, nanotubes and carbon onions. Finally, the phenomena of structure formation under irradiation, in particular the self-assembling of spherical carbon onions and the irradiation-induced transformation of graphitic nanoparticles into diamond, are presented and discussed qualitatively in the context of non-equilibrium structure formation.

962 citations


Journal ArticleDOI
01 Jan 1996-Carbon

562 citations


Performance
Metrics
No. of citations received by the Paper in previous years
YearCitations
20191
20152
20141
20131