Monthly Archives: October 2017

One limitation of this study is the

One limitation of this study is the observation of bubble translation and streaming flow in only a two-dimensional plane. The lenses of the microscope were focused on the midaxial plane of the test tube to measure bubble cluster formation and translation. Therefore, the images of the bubble clusters outside of the plane were blurred, making estimation of bubble cluster size and translation velocity difficult. Streaming velocities also were measured in a two-dimensional plane using micro-PIV. Because the flow velocities were not measured in three dimensions, the out-of-plane velocity, which could have an effect on the accuracy of the measured streaming velocities, was not measured. However, the out-of-plane velocity is small because the acoustic beam generates axisymmetric flow and the beam axis coincides with the velocity measurement plane. Another limitation is the spatial resolution of the micro-PIV system. It might have had an effect on the measurement of wall shear stress because the wall shear rate was estimated from the near-wall velocity, which was measured at a half interrogation window size (88μm). Increasing the spatial resolution of the velocity measurement system would increase the accuracy of the measured velocity gradient at the wall. A third limitation of this study is related to the ultrasound parameters used. Streaming flow was studied for the focused ultrasound beam with two frequencies (2.25 and 1.0MHz) in this study. Further studies on other ultrasound parameters using a wider range of pressures and frequencies should be performed. Finally, streaming flow in the distilled water, which was used in this study, would be different from that in blood, because of the difference of viscosity. Medium viscosity would affect streaming velocity by changing attenuation of bcr-abl tyrosine kinase inhibitors pressure [13]. Also, streaming from microbubbles should be affected by medium viscosity via changing the pulsation amplitude of microbubbles by acoustic waves [31]. Using a water-glycerin mixture as blood-analogous test fluid would simulate acoustic streaming in a blood vessel more accurately.

Conclusion
We studied the streaming flow from the microbubbles of a UCA activated by acoustic waves in a vessel model. Sonication of the fluid without a UCA present did not generate noticeable flow in the vessel model. When the UCA was added, the microbubbles formed bubble clusters and translated in the acoustic beam’s axial direction upon ultrasound sonication. Cluster formation and translation were faster with sonication using 2.25MHz ultrasound, a frequency close to the resonance frequency of a UCA bubble. Translation of the bubble clusters induced strong flow in the beam’s direction, and the flow jet impinged on the vessel wall and formed recirculating vortices. The maximum streaming flow was about 60mm/s for acoustic waves with the pressure amplitude of 600kPa. The velocity of bubble cluster translation was similar to the streaming velocity, but streaming flow lagged behind cluster translation. Bubble cluster formation and translation were noticeably slower for the 1.0 MHz acoustic wave with the same pressure amplitude as the 2.25 MHz acoustic wave because of the reduced secondary radiant force from the acoustic wave of nonresonant frequency. A reduced pressure amplitude also reduced the streaming flow. The effects of the acoustic wave’s frequency and amplitude on wall shear stress were more distinct.

Acknowledgment

Introduction
Foodborne disease remains a public problem worldwide. Gram-positive and-negative bacteria are the commonest pathogens causing foodborne diseases [1,2]. The current strategies are to reduce bacterial load through washing by water, sanitization and disinfection. However, water washing and sanitization reduce pathogenic bacteria with a limited success [3]. The disinfectants might cause genotoxicity and carcinogenicity [4]. Thus, there is an urgent need to develop alternatives to eradicate foodborne bacteria.

gli1 br Acknowledgment The authors are

Acknowledgment
The authors are grateful for the funding provided to their laboratory by the Tunisian Ministry of Higher Education and Scientific Research.

Introduction
Pancreatic changes, including changes in the parenchyma, ductal structures, and microscopic structures, are observed in both diseased and non-diseased conditions as a part of the degenerative process [1,2]. The influence of age on the pancreatic gli1 has long been known [3,4]. The pancreatic volume begins to decrease after the sixth decade, and the percentage of pancreatic fat increases [5,6]. Microscopically, lipomatosis, fibrosis, and focal necrosis are also observed with increasing frequency with aging [3].
Transabdominal ultrasonography is a non-invasive imaging modality that has been used for pancreatobiliary diseases for many decades. When combined with elastography, transabdominal ultrasonography can provide more information about the mechanical quality of the tissues. Because tissues with different pathologic processes have different mechanical properties, the sonographer can accurately characterize lesions based on differences in their elasticities [7–11]. There are two main types of ultrasound elastography that depend on the measured physical quantities: strain elastography and shear wave elastography [12]. With strain elastography, quantitative data can be acquired using the strain ratio, strain histogram, or neural network analysis. Strain histograms are used to analyze the stiffness by converting color images that represent different levels of stiffness into gray-scale and calculating parameters that include the Mean, Standard deviation, Kurtosis and Skewness.
A correlation of the Mean value of the gray-scale histogram and the degree of pancreatic fibrosis has been reported [13,14]. Although the majority of strain elastography reports concern endoscopic ultrasonography [13,15], transabdominal ultrasonography may be another imaging method of choice for screening the pancreases of normal subjects that is associated with fewer complications and reduced invasiveness. The feasibility of the use of tissue elastography with transabdominal ultrasound in the diagnosis of pancreatic diseases has also been reported [16,17] and an increase in diagnostic accuracy can be achieved when the method is combined with B-mode imaging.

Methods
The subjects were consecutive adults over the age of 18 who underwent transabdominal ultrasonography. Those without known pancreatic disease, without histories of alcohol consumption >80g per day [18,19], no alcohol-related diseases, and no newly diagnosed pancreatic lesions on the ultrasound for the current study were included. Subjects whose pancreas was not well visualized and those in whom we were unable to perform strain elastography due to technical difficulties were excluded. Information regarding alcohol intake, tobacco use, the presence of diabetes mellitus (fasting plasma glucose >126mg/dl) and indications for transabdominal ultrasonography, was obtained from the medical record. Alcohol consumption was divided into categories of none, occasional, and regular, and those who consumed >80g per day were excluded from the study. Tobacco consumption was divided into none, previous smoker, and current smoker. We categorized subjects by age as follows: <40years, 41–60years, 61–80years and >80years. We recruited subjects from each age group to achieve an even distribution of ages. The research was conducted in accordance with the World Medical Association’s Declaration of Helsinki regarding experiments involving human subjects, was approved by the ethics committee of Nagoya University Hospital, and is registered in UMIN-CTR (UMIN000016975). All subjects provided informed consent to participate in the study.

Results
Between October 2014 and March 2015, 106 subjects were included in this study. Four of these subjects were excluded; two were unable to co-operate in terms of respiration, which resulted in failure to perform transabdominal elastography, one was excluded due to poor pancreatic visualization, and one was excluded due to history of acute pancreatitis. Thus, 102 subjects (65 women and 37 men) aged between 20 and 85years (median age 62) were included in the analysis. A flow diagram of the study is presented in Fig. 3. There were 20 patients aged <40years, 26 aged between 41 and 60years, 45 aged between 61 and 80years, and 11 aged >80years. In the 41 subjects with liver disease, those with gallbladder disease (30 subjects) and malignancies such as colonic, gastric or ovarian carcinoma (7 subjects), the transabdominal ultrasonographies were performed as a regular surveillance procedure. In 24 subjects, the transabdominal ultrasonographies were performed to check for non-pancreatic diseases.

dopamine-2 receptor antagonist Park et al demonstrated both

Park et al. [18] demonstrated both theoretically and experimentally in a composite plate that it is possible to minimize the amplitude dispersion effect and successfully reconstruct the original input waveform by using a narrow-band excitation signal. The frequency was selected so as to excite only the fundamental antisymmetric mode. Xu and Giurgiutiu [19,20] and Santoni et al. [21] also studied single-mode tuning effects on Lamb wave time-reversal in thin metallic plates with surface-bonded PZT patch transducers, and concluded that single-mode Lamb waves are rigorously time reversible under narrow-band excitation, and hence are recommended for time-reversal based damage detection. The time-reversal of a mixed mode () Lamb wave creates two extra wave packets symmetrically placed around the main wave packet. Their experimental results conducted on an aluminum plate using surface mounted PZT transducers, however, showed that a two-mode () Lamb wave yielded a better reconstruction by the TRP than either of the single modes ( or ).
The use of the TRP of Lamb waves for damage detection was illustrated by Sohn et al. [22], examining the deviation of the reconstructed signal from the known input signal in a composite plate containing delamination damage. Poddar et al. [23] used the time-reversal method for experimentally detecting notch, block mass and surface corrosion type defects in metallic plates. At the same time, it has also been reported by several researchers that the similarity index of the reconstructed signal with the original input signal does not show appreciable difference between the undamaged and damaged states for notch and surface damages in metallic plates [24,25]. This aspect was recently examined by the authors [26] through finite dopamine-2 receptor antagonist (FE) simulations verified by some experimental data. Using frequency sweeping with narrow-band modulated tone burst excitations, they showed that (i) the single-mode tuning does not lead to the best reconstruction of the original input signal in the undamaged plate, and (ii) the damage index computed using the conventional main wave packet of the reconstructed signal does not indeed show any appreciable change in presence of notch-type damage. To overcome this drawback, the authors introduced the concept of the best reconstruction frequency at which the similarity index in the undamaged plate is maximum over a given range of excitation frequency. It can be determined for a given structure-transducer-adhesive system in undamaged state by frequency sweeping over a desired frequency range depending on the size of defects to be detected. At this frequency, the wave contains both and modes in general. A refined method of computing the damage index was proposed using an extended wave packet ranging between the two side bands accompanying the main wave packet, which showed excellent sensitivity to damage when used at the best reconstruction frequency. The FE model employed in the aforementioned study, however, assumed a perfect bonding between the transducers and the plate. Apart from the excitation frequency, several parameters such as the bonding between piezoelectric wafer transducers and the host plate, the transducer size, the plate thickness as well as the tone burst count in the excitation signal affect the amplitude dispersion and consequently the time reversibility. These effects must be understood thoroughly for successfully applying the TRP of Lamb waves for damage detection.
The effect of adhesive layer on the performance of piezoelectric transducers for impedance based SHM of metallic plate structures was experimentally studied by Qing et al. [27]. It was shown that an increase in the adhesive thickness alters the electromechanical impedance and the resonant frequency of the piezoelectric elements. Ha and Chang [28] studied the effect of adhesive layers on the sensor signal for Lamb wave propagation through numerical simulations as well as experiments. It was shown that the amplitude of the sensor signal may increase with the decrease in the shear modulus of the adhesive layer or with the increase in its thickness, when the excitation frequency is near the resonant frequency of the surface-mounted sensor, which is opposite to what is expected due the shear lag effect. This happens due to what was termed as the resonant effect. They also studied these effects at elevated temperatures [29]. Lanzara et al. [30] studied the effect of bond degradation on the performance of piezoelectric actuators and sensors. To the best of authors’ knowledge, no study on the effect of adhesive layer on the time reversibility of Lamb waves has been reported so far. The effect of number of cycles in the modulated tone burst excitation on the time reversibility was examined by Poddar et al. [23] and Agrahari and Kapuria [26]. But, these studies were conducted at specific frequencies, and the obtained trend may not be valid at all frequencies. Also, there has been no experimental study reported so far on the effect of the transducer thickness and the plate thickness on the time reversibility, which may have significant effect on the quality of the reconstructed signal after the TRP.

Evaluation of measured ultrasonic signals showed that ongoing deterioration

Evaluation of measured ultrasonic signals showed that ongoing deterioration of microstructure of mortar bars can be successively detected by both time and frequency domain analysis of recorded signals. The parameters recorded and computed (P-wave velocity, attenuation of ultrasonic signal, energy, etc.) during the experiment show different scales of changes that can reflect both normal development of concrete microstructure (hardening of concrete mixture) and/or degradation of microstructure due to developing ASR. Measured expansion of the test mortar bars of 0.5% in 40-days lasting experiment corresponded to a drop in P-wave velocity of about 400m/s (approx. 9% of maximum value) in the case of the specimens subjected to the effect of 1N NaOH. The amplitudes of the first onset and maximum signal amplitude exhibited similar trend but the first onset with 80% change of the value seems to be much more sensitive to the development of material’s microstructure change than the maximum signal amplitude with only 10% change. This can be explained by the fact that the amplitude of the first onset, as well as arrival time, corresponds directly to P-wave propagation, which is affected only by the actual composition and microstructure of the material tested. The amplitude of the first onset is also 10-times more sensitive than the velocity of P-wave propagation, which is given by the integral character NF-κB Signaling Library of the sensor (sensor area). Lower changes of ultrasonic sounding parameters were detected in mortar bars with the innocuous aggregate in comparison with mortar bars containing aggregate with higher ASR potential.

Acknowledgements
This study was financially supported through research project GAP104/12/0915 with funding provided by the Czech Science Foundation and by the Czech Academy of Sciences, project RVO 67985831. The technical help of Mr. Filler, Mrs. Erdingerová, and Mr. Nemejovský was very much appreciated.

Introduction
The purpose of this work was to improve the existing models that allow spatial averaging effects of piezoelectric hydrophones to be accounted for. Miniature hydrophones are widely used to detect the spatial and temporal characteristics of ultrasound fields [1]. The hydrophone is used to measure the averaged acoustic pressure over the active area of the element. This technique is recommended for the measurement of many field parameters that are considered important in the International Electrotechnical Commission (IEC) standards [2]. The effective radius of the hydrophone should ideally be comparable to or smaller than one quarter of the acoustic wavelength, as indicated in paragraph 5.1.6.1 of IEC 62127-1 [2]. This is to ensure that the phase and amplitude variations over the active NF-κB Signaling Library do not significantly contribute to the measurement uncertainties. In the far field of the ultrasonic transmitter, the criteria can be relaxed based on the dimensions of the transducers, the acoustic wavelength and the distance between the hydrophone and the transmitter surface [2,3]. The error caused by the finite aperture size is termed the spatial averaging error [4]. Umchid and Gopinath also noted that the effective diameter of the hydrophone should be on the order of the half-wavelength at the highest frequency to be measured to eliminate the effect of spatial averaging [5]. In water, this requires the use of an active aperture on the order of 50μm at 15MHz. However, although a special 50-μm-diameter hydrophone has been reported [6], the majority of commercially available piezoelectric hydrophones has nominal active apertures (or diameters) within 0.2–1mm, which is too large for measurement in an ultrasonic fields above a few MHz. Hydrophones that do not meet the requirements for a point detector produce spatially averaged values of the acoustic pressure.
For the applications which require sensors with high temporal and spatial resolution, the interferometric [7–9] and fiber-optic methods [10–13] are noteworthy. A fiber-optic hydrophone with a tip diameter of approximately 7μm has been reported by Lewin et al. [12]. The hydrophone enables extension of the working frequency to 100MHz without inducing spatial averaging effects. Although the fiber-optic hydrophones are promising in biomedical ultrasound applications, they are primarily found in well-equipped university- or testing houses laboratories and are relatively expensive. The spatial averaging effects due to the finite aperture size of piezoelectric hydrophones have attracted much attention from investigators. Boutkedjirt and Reibold [14,15] addressed this issue as an inverse problem and used numerical methods to deconvolve the averaging effects. Other methods for correcting the spatial averaging effects have been proposed although their discussions were restricted to idealized models. For the measurement arrangement in which the hydrophone is located along the acoustic axis of the transmitting transducer, Daly and Rao [16] derived a set of closed-form expressions based on the Lommel diffraction formulation for both planar and spherically focused transmitters. In addition, Preston et al. [17] and Smith [18] used a beam-plot method based on the theoretical model of the pressure distribution in the focal plane to estimate the spatial averaging effects. Zeqiri and Bond [19] and Radulescu et al. [20] numerically calculated the theoretical pressure distribution over the hydrophone in the focal plane, with the purpose of estimating the same effects. Harris [21,22] used the generalized spatial impulse response function of the velocity potential to evaluate the averaging effects in the transient field of planar transducers. Markiewicz and Chivers [23]used the spatial impulse response method [21,22] to investigate the typical errors involved in far field measurements. Furthermore, Beissner [24] derived an exact integral expression for the spatial averaging effects in the steady state of a planar transducer field when the acoustic axes of the transmitter and the hydrophone are coaxial. Goldstein et al. [25] estimated the spatial averaging effects of a hydrophone along the acoustic axis by determining the effective radius of a planar transducer used as a transmitter. Later Goldstein [26] determined the magnitudes of the axial and lateral pressures for the field characterization of a planar transducer in a steady-state field that contained no hydrophone.

br Kinematics principles of a single diamond abrasive Schematic

Kinematics principles of a single diamond abrasive
Schematic diagram of RUD process and the abrasive trajectories generated with and without ultrasonic are illustrated in Fig. 1. The hollow nickel-bonded tool, which impregnated with diamond abrasives, was rotated at a high constant speed, meanwhile vibrated at an ultrasonic frequency in its axial direction. Additionally, the longitudinal feeding motion of the diamond tool at an invariant speed resulted in the abrasive penetrations into the material, thus promoting the nucleation and propagations of the crack systems, which would benefit the extending downwards of the drilled hole. An internal coolant system pumped water through the central hole of the tool, washing away the machining debris and cooling the tool simultaneously (Fig. 1(a)). The trajectory of the abrasive located on the tool end-face could be deduced as:where r is the tool radius, ω is the angular velocity, A is the vibration amplitude, is the feed speed, f is the ultrasonic frequency, and t is the time. Based on the above equation, the periodic fluctuation features of the abrasive trajectories are presented in Fig. 1(b). Obviously, the abrasives traveled along their specific sinusoidal paths in the microscopic view, while generally feeding downwards coupled with the tool revolution. In this context, the performance of each individual abrasive during the hole entrance chipping formation process could be simplified as one abrasive parallel scratching the polished surface, since the small feed speed of the diamond tool would have puny effects on the instantaneous cutting depth of each abrasive in each ultrasonic cycle.
Marinescu demonstrated that, under small cutting depth, the abrasive-material extrusion provided insufficient ginsenoside rh2 to arouse the pre-existing defect nucleation in the interior material, and serious plastic deformation could be evoked, which would expand with the cutting depth increasing [6]. Based on this argument, the illustrations of the ductile deformation regions generated with and without ultrasonic just underneath the contact site are represented in Fig. 2. Apparently, the constant scratching depth of the abrasive in CD process would presumably result in the invariable plastic deformation depth in the interior material. While the fluctuated abrasive trajectory in RUD process brought about the effective rake angle changed periodically, and its relationship with the normal value could be described as . Obviously, at stage #I, when the abrasive just engaged into the specimen surface, the maximum was achieve, which was larger than its normal value , and generally shrank with the abrasive moving downwards. Therefore, with decreasing and the cutting depth increasing, the plastic deformed region would broaden, due to the enlarged extrusion loading between the abrasive and the material. Noting suprachiasmic nucleus (SCN) the maximum extrusion load was achieved at the trajectory bottom (namely stage #0), resulting in the plastic deformed region reached its maximum, which was equivalent to that aroused in the CD process, if the maximum cutting depths were the same (see Fig. 2). Reversely, at stage #II, although the effective rake angle was still decreased, the reduced depth of cut would relieve the abrasive-material extrusion, which benefited the decreasing of the ductile deformation of the material. It should be emphasized that, in formal drilling processes, because of the various penetrations, the abrasives situated on the tool end face engaged the specimen with their respective trajectories and cutting depths, which would contribute to the entrance chipping formation simultaneously.
Gu et al. demonstrated that lateral cracking nucleated at the bottom of the plastic deformation region, and propagated parallel to the specimen surface [7]. Accordingly, these specific deformation characteristics of the material induced by the ultrasonic superposition would significantly affect the nucleation and propagations of the crack systems, as well as the entrance chipping formation mechanisms in RUD process.

br Multi actuated atomic force microscopy

Multi-actuated atomic force microscopy

Experimental methods
An atomic force microscope is designed and implemented based on the proposed multi-actuation methodology. The schematic view and the setup of this atomic force microscope are shown in Fig. 3. In this section key aspects of the design of the instrument are described. These include the optical beam deflection setup, the multi-actuated scanner, the data acquisition and controls. These subsystems and components are explained in what follows.

Results and discussion

Conclusion
A multi-actuation methodology for large-range and high-speed atomic force microscopy is presented. In the proposed approach, various nanopositioners with different range and bandwidth specifications are combined in a cascaded series form. The dynamic couplings between the different actuators are treated through control. An intuitive data-based control design methodology is presented where the controllers are derived directly from the measured positioner response and without intermediate modeling. A multi-actuated atomic force microscope is designed and implemented based on the proposed methodology. Two flexure-based out-of-plane and two lateral positioners work together to simultaneously achieve a large-range and high-speed performance. Etching of calcite in diluted sulfuric LY3009104 is visualized using this AFM. The fast retreat (∼254nm/s) of calcite terraces and formation, deepening and merger of neighboring pits are visualized. The morphology of the dissolving calcite surfaces are observed to be affected by the crystalline structure of calcite even at low pH levels. Furthermore, through the analysis of the time-lapse images the dissolution rate of calcite is estimated. The AFM presented here combines high-speed imaging capability with lateral/out-of-plane scan ranges larger than any previously reported in an instrument of this type. Various practical features, such as simultaneous optical view of the sample and probe, a conveniently large sample stage, and compatibility with small cantilevers, further enhance its utility as a research tool. This design enables studies of various dynamic nanoscale processes in air and in aqueous environments.

Acknowledgments
The authors would like to thank the Center for Clean Water and Clean Energy at MIT and KFUPM for funding this research under Project no. R16-DMN-12. The authors also thank National Instruments for their hardware and software support.

Introduction
Cathodoluminescence (CL) excited by electron beam irradiation is used for many applications such as analysis of materials [1], field emission displays [2], laser sources [3] and light source of near field optical microscope and super resolution microscope [4–7]. Cathodoluminescence analysis is a crucial technique for understanding the optical properties of nanowires [8–10], nanodiamonds [11–14], nanoparticles [15,16], and plasmonic nanoantennas [17,18] because it has nanometric spatial resolution. The spatial resolution of CL analysis is much higher than that of photoluminescence techniques because the electron beam can be focused to regions a few nanometer in size. The spatial resolution of CL analysis is determined by the spatial extent of electron beam scattering and light propagation excited with the electron beam. In order to evaluate and improve the spatial resolution of CL analysis, it is required to analyze the scattering of electron beam and the light propagation of CL.
We have developed a new analysis method for the light intensity distribution of CL using a combination of Monte Carlo simulation and the finite-difference time-domain (FDTD) method [19]. The analysis method has three steps. The first step is a scattering analysis of the incident electron beam by Monte Carlo simulation [20–22]. The trajectory and the energy loss of each electron are simulated with the analysis method. The second step is the calculation of the arrangements of dipole excited with the electron beam. The position of the dipole is determined by the trajectory of incident electron analyzed by Monte Carlo simulation. CL emission is expressed as dipole radiation in the analysis method. Finally, light propagation from each dipole is analyzed by the FDTD method. The intensity distribution of light emitted from each dipole is summed up to obtain the intensity distribution in and near the thin film.

The random height distribution of the surface topography makes it

The random height distribution of the surface topography makes it difficult to extract the tip radius quantitatively and accurately directly from the AFM images. Liu et al. [24] found that the power spectral density curve of the AFM image obtained with a blunt tip was lower than that with a sharper tip, indicating a decline in AZD2281 with a worn tip. By using this method they qualitatively evaluated the tip wear through estimation of the tip radius. It seemed that the frequency domain analysis method had good potential as a way to estimate the tip radius. For defined structure surfaces, the tip radius can be directly calculated from the sample image artifacts caused by the tip–sample interaction, but artifacts introduced by instrumental noise and poor response to the AFM control system may degrade the accuracy of this method. Fast Fourier transform (FFT) analysis can separate the noise signal (usually high frequency) from the useful component. As such, spectrum analysis for the tip assessment may decrease the error.
Our objective here is to develop a simple and effective FFT analysis method for AFM images of the defined structure to quantitatively evaluate the AFM tip radius. In the present study, a nano-stepped structure that has accurate and simple dimensions was chosen as the tip characterizer. The geometrical model based on the AFM tip scanning the step structure was established. The shape of the AFM tip in the model was assumed to be a hemispherical cone, which is the frequently presented shape in AFM experiments. In addition, this hemispherical tip is often used as an indenter for nanoindentation [4,5]. Tips with different radii were simulated to scan the step structure and the FFT results of the corresponding scanned contours were calculated. The relationships between tip radius and variations in the harmonics of the step structure were also investigated.

Modeling and estimation methods based on FFT analysis

Experimental details
Experiments were carried out using a Dimension Icon AFM (Bruker Company) in the tapping mode under ambient conditions. Two used and worn single crystal silicon tips (RTESP; Bruker Company), named tip-1 and tip-2, respectively, were used to scan the step for evaluation. A nano grating sample with step structures was employed: a platinum-coated silicon substrate with a square pit array on its surface (VGRP-15M; Bruker Company). Dimensions of each pit were 10μm in length and 180nm in depth. For each scan, a 50×50μm2 (256×256pixels) scan size on the pit array sample was first captured and then a selected local area of 0.5×2μm2 (128×512pixels) in the imaged area was employed for characterization, as shown in Fig. 6(a) and (b), respectively. SEM imaging and blind reconstruction were also used to estimate the size of the tips used to verify catabolic reactions method. The sample used for the blind reconstruction approach was a rough titanium sample, which produced many sharp features (RS-15M; Bruker Company).
Because in the FFT method the data used for evaluating the tip radius is the AFM image scanned by the corresponding tip, the scanning quality would affect the accuracy of the evaluation result. If the feedback loop of the AFM system is slow, the tip can be off track of the height variation of step structure, and hence the artifact (low frequency) would appear in the AFM image, which in turn misleads the subsequent tip evaluation. To minimize this error, the response of the measuring system to the height variation of the step structure should be evaluated. In our AFM (Dimension Icon), the differences between the round-trip scanning lines for each measuring profile were used to examine the tip tracing state [25]. The scan parameters were adjusted to obtain the minimum difference between the round-trip scanning lines, which renders the measuring results more accurate and lowers the influence of feedback system property. In this case, the scanning results were used for calculating the tip radii.

In the simulations a model of the Cu

In the simulations, a model of the Cu/LiPON/Cu sample with the same shape and size as shown in Fig. 4 was constructed. An example of the calculated potential profile inside the Cu/LiPON/Cu model in the direction parallel to the edge of the specimen is displayed in Fig. 5(a). To simplify the simulation, we assumed that the potentials in the Cu electrodes and the LiPON electrolyte were flat and that the electric double layer (EDL) had a linear rather than curved potential slope. The electric potential within the specimen was also assumed to be uniform in the direction of the SBHA supplier beam.
In the first simulation, the potential in the LiPON layer, VLiPON, was set at half the applied voltage VCu-Cu, and the width t of the EDL was set at 25nm, as shown in Fig. 5(a). These values were then varied until the simulated phase distributions agreed with those obtained from experiment. The 3D potential distribution outside the specimen was calculated by the 3D boundary-charge method (3D-BCM), dividing all the surface plates of the simulation model into 26,050 surface elements, and the same potentials applied across the Cu electrodes, GC, and W components. Finally, 2D phase images were calculated by integrating the 3D potential distribution along the electron trajectory. Integration was carried out from +100µm to −100µm as in the case of the simulated profiles shown in Fig. 3. The effect of electric fields in the reference-wave region was also taken into account.
Thin black and red lines in Fig. 5(b) are the reconstructed phase profiles obtained from the dotted box region in Fig. 4. Thick green lines are the simulated phase profiles at different voltages when the potential in Cu and LiPON regions is flat, VLiPON is half of the applied voltage VCu-Cu, and the EDL width is 25nm. In both the experimental and simulation profiles the phase distributions within Cu regions are sloped, even though flat potentials were assumed in the simulation. The gradient in the experimental profile is thus likely due to electric leakage fields above and beneath the specimen. Sloping of phase distribution curves is also seen in the LiPON region, presumably due to the same effect. It can also be noticed that the simulated phase lines are slightly above the measured lines, implying that potentials assumed in the simulations (VLiPON) were higher than the actual potentials. Simulations were repeated with lower values of VLiPON until a better match with experiment was obtained.
Fig. 5(c) shows a model with a lower VLiPON, and Fig. 5(d) compares the simulated phase profiles obtained using this model with the measured profiles. They are in much better agreement than those in Fig. 5(b). This suggests that the potential is less than half the applied voltage and that a larger potential change occurs at the interface on the high-voltage side, and a smaller change on the low-voltage side compared to the potential in the short-circuited specimen [19].
If the detected potential change around the interface is assumed to originate solely from the distribution of Li ions, Li-ion and Li-vacancy distributions in LiPON can be inferred. The relationship between changes in electric potential and the behavior of Li-ion and ion vacancies, as well as the width of the electric double layers, are discussed in a recent paper [19]. The good agreement between experimental and simulated phase distributions suggests that this method is a useful way of examining charge distributions in solid-state ion conductors.

Damage layers formed during FIB milling and narrow Ar-ion beam milling
FIB milling is a powerful and widely used method for preparing TEM specimens. However, FIB causes damage to the surface of the specimen because of the high energy of the Ga ion beam used. Earlier work has shown that phase images of semiconductors are improved by in situ annealing [25]. In the case of ion conductors, these damage layers have high electronic conductivity, which interferes with the formation of the “true” potential distribution inside the specimen under an applied voltage.

To illustrate this possibility Fig a shows a quickly scanned

To illustrate this possibility, Fig. 12a shows a quickly-scanned HAADF image of a calcium hydroxide specimen, recorded using the Nion HERMES STEM at Arizona State University. This instrument incorporates a high-resolution monochromator and energy-loss spectrometer, giving an purchase MK1775 resolution (< 15meV at E0 = 60keV) sufficient to detect vibrational-mode energy-loss peaks that are characteristic of hydrogen or specific chemical bonds [1]. The region within the green square was previously scanned using a coarse digital raster (s = 10nm, d = 1nm) and with a probe current high enough to cause substantial radiation damage, visible as dark spots that indicate reduced specimen thickness (mass loss). Despite this localized damage, the 0.45eV vibrational peak (due to hydrogen) remained almost constant. Subsequent “leapfrog” scanning of guanine [50] showed a similar array of dark spots (indicating substantial damage within the 1nm probe) and little decrease in the H-bond (0.3–0.4eV) signal, but only if the step size was increased from s = 10nm to s = 30nm. However, increasing s reduces the spatial resolution of the image; for leapfrog scanning to be beneficial, it must provide a larger signal for the same spatial resolution. Fig. 13a shows the predicted decay of the H-bond signal (integrated over a diameter s) as a function of irradiation dose D per pixel, based on the theory presented above. For a regular scan (d=s) the energy-loss signal falls exponentially with increasing radiation dose. For a leapfrog scan (d << s), it falls more rapidly but then more slowly as signal is collected (via scattering delocalization) from largely undamaged material between each beam position. For D > 2, the leapfrog signal exceeds that from the regular scan and this advantage becomes more dramatic as the irradiation continues. However, the time-integrated signal from the leapfrog scan remains less than that from the regular scan, so the leapfrog image will be more noisy, degrading the effective spatial resolution (approximately s for both types of scan).
The situation for a 6eV signal (e.g. π* peak, characteristic of double bonds) is illustrated in Fig. 13b. Once again, the leapfrog signal is higher for a large dose but the corresponding time-integrated signal is only about half that for a regular scan.

Discussion
We have confined our attention to energy losses due to valence-electron or phonon excitation, where the delocalization distance is much larger than atomic dimensions and is determined by bmax and the 1/r2 dependence of the point spread function. These properties derive from the kinematics of electron interaction rather than internal dynamics of an atom, as expressed in terms of the generalized oscillator strength for example [51]. Such dynamical effects will influence the PSF for r < bmin but this contribution to the inelastic signal is small for low values of energy loss. The situation is different for inner-shell losses, where the region 0 < r < bmin represents a larger part of the PSF and greatly affects the probability of core-electron excitation. Here the delocalization lengths of inelastic and elastic scattering are comparable to each other, producing a coupling that complicates the interpretation of atomic-resolution inelastic images recorded from crystalline specimens. Such complications are unlikely to be important for the low energy losses considered here. Vibrational (phonon) energy losses have been studied for many years using medium-energy electrons reflected from the surfaces of crystalline specimens [52]. The EELS signal is found to contain two components: a dipole mode concentrated around Bragg-reflected beams, and an “impact” component that can be measured between the Bragg beams [53,54]. The latter has a broad angular distribution, indicating a high degree of localization around atomic cores. In fact, calculations of the transmission-mode phonon signal have predicted atomic resolution [55–58] and recent STEM measurements have confirmed that a resolution better than 2nm is possible [59]. These measurements were made (using a crystalline BN specimen) in a dark-field mode, with the incident beam tilted so that the central and diffracted beams were intercepted by the spectrometer entrance aperture, giving a high-resolution signal about 1% of the bright-field phonon signal. Under such conditions, the equations derived above will not apply; we have assumed bright-field spectroscopy in which the dipole signal is predominant.

V5 peptide CLs and light guided modes are only

CLs and light guided modes are only present at scattering angles below a few tens of µrad [8–10,15]. In normal STEM, the semi-convergence angles of the beam are normally so large [∼several mrad] that the convergent beam V5 peptide diffraction (CBED) discs have a significant overlap. This also means that the extremely forward scattered CLs, surface losses and guided light modes are still present in the entire diffraction plane. Hence, normal STEM cannot be used to measure the bandgap of high refractive index materials in the standard acceleration voltage range of 60–300kV for most modern TEMs. However, in this work we use a monochromated and probe corrected beam in low-magnification (Low-Mag) STEM mode. In low magnification STEM, semi-convergence and –collection angles in the µrad range are used. Off-axis conditions, i.e. dark field EELS, is further used to suppress the ZLP and to be outside the angular range where CLs and light guided modes hit the spectrometer. Furthermore, the low semi-convergence and -collection angles decrease the detection of phonons (they have a large angular distribution) and interband transitions with a significant momentum transfer. Dark field EELS means that we detect electrons with a momentum transfer. However, because the semi-angles are far down into the µrad range, we can still detect a part of reciprocal space that is close to the centre of the first Brillouin zone (BZ), the Gamma point (Γ), and far from any of the BZ boundaries. Since the valence and conduction bands both are very flat at the BZ centre, the measured bandgap would still be very close to the material\’s direct bandgap. One of the benefits of the presented methodology is that bandgap of high refractive index materials can be determined at any acceleration voltage and over the entire range of conventional TEM specimen thicknesses. Compared to EELS performed by conventional TEM in diffraction mode, low-mag STEM has important advantages: Most importantly, in Low-Mag STEM a much higher spatial resolution can be achieved than in conventional TEM mode. The resolution is typically a few nm and dependent on the convergence angle of the beam. Secondly, the camera lengths in conventional TEM mode are not large enough to provide the same small semi-collection angles that can be used in Low-Mag STEM.

Materials and method
The GaAs-based materials characterized in this work were grown by molecular beam epitaxy (MBE). A high angle annular dark field (HAADF) STEM image of the sample structure is shown in Fig. 1a). A 400nm thick Al0.25Ga0.75As layer is grown on top of a (001) GaAs substrate followed by a stack of 20 quantum dot (QD) layers. Each of the QD layers were grown as 2 monolayers thick InAs and they are separated by 20nm thick Al0.25Ga0.75NxAs1-x (x<0.01) spacer layers. On top of the QD stack there is a 330nm thick Al0.25Ga0.75As layer followed by a 50nm thick GaAs top layer. Cross-section TEM samples were prepared both by focused ion beam (FIB) and by Ar ion-milling. The FIB samples were prepared by a FEI Helios Nanolab dual-beam FIB equipped with an Oxford Omniprobe. The coarse Ga+ ion-beam thinning was done at 30kV acceleration voltage followed by final thinning at 5kV and 2kV. A high resolution HAADF STEM image from one of the InAs QD layers is shown in Fig. 1b) and demonstrates that the samples made purely by FIB have very minor beam damage from the sample preparation. A PIPS II was used to thin the Ar+ ion-milled samples. Coarse thinning was performed at 3kV acceleration voltage before progressively reducing the acceleration voltage, finishing at 100eV. The samples were cooled by liquid N2 during milling.
The HAADF STEM images in Fig. 1 were taken with a double Cs corrected, coldFEG JEOL ARM 200CF, operated at 200kV. EELS was performed with a double Cs corrected, monochromated Titan cube, operated at 80 or 120kV and equipped with a Gatan Imaging Filter (GIF) Quantum ERS spectrometer. An energy dispersion of 0.01eV/pixel, and a 2.5mm GIF entrance aperture were used at all times. The GaAs-based material was oriented a few degrees away from the [110] zone axis to avoid strong channelling effects, and to make sure that we are off most Kikuchi bands in the off-axis set-up.