review development of advanced electron holographic … · 2017-04-23 · review development of...

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

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

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

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

Review

Development of advanced electron holographic techniquesand application to industrial materials and devices

Kazuo Yamamoto1,*, Tsukasa Hirayama1 and Takayoshi Tanji21Nanostructures Research Laboratory, Japan Fine Ceramics Center, 2-4-1 Mutsuno, Atsuta-ku,Nagoya 456-8587, Japan and 2EcoTopia Science Institute, Nagoya University, Furo-cho, Chikusa-ku,Nagoya 464-8603, Japan*To whom correspondence should be addressed. E-mail: [email protected]

Abstract The development of a transmission electron microscope equipped with afield emission gun paved the way for electron holography to be put topractical use in various fields. In this paper, we review three advancedelectron holography techniques: on-line real-time electron holography,three-dimensional (3D) tomographic holography and phase-shifting elec-tron holography, which are becoming important techniques for materialsscience and device engineering. We also describe some applications ofelectron holography to the analysis of industrial materials and devices:GaAs compound semiconductors, solid oxide fuel cells and all-solid-statelithium ion batteries.

Keywords real-time electron holography, 3D tomographic holography, phase-shiftingelectron holography, dopant profiling, solid oxide fuel cell, lithium ionbattery

Received 30 November 2012, accepted 9 February 2013; online 26 March 2013

Introduction

The advent of electron holography microscopes andthe achievements of the two major research pro-jects, Tonomura Electron Wavefront Project andOak Ridge National Laboratory electron holographyproject, strongly stimulated research activities inthe field of electron holography in the 1990s [1–3].At present, off-axis electron holography with anelectron biprism is commonly used not only for cor-recting aberration of electron lenses but also forvisualizing electromagnetic fields. An ‘object wave’modulated by the fields and a ‘reference wave’passed through a vacuum interfere with each otherby the biprism and form interference fringe patternthat is called ‘hologram’. The phase modulation Df

of the object wave can be reconstructed from thehologram by an optical lens system or an image

processing with a computer. The phase Df due tothe electromagnetic fields is expressed as

Dfðx; yÞ ¼ CE Vðx; yÞ tðx; yÞ

� 2p eh

ððBnðx; yÞ dx dz; ð1Þ

where CE, V, t, e, h and Bn are the electron-energy-dependent constant [3], the electric potentialin a sample, the sample thickness, the electroncharge, the Planck’s constant and the component ofthe magnetic flux density normal to the planedefined by the electron beam paths, respectively.Thus, we can quantitatively display the potentialmap in the electric devices using the first term andthe magnetic flux distribution using the secondterm. However, it was known that electron holog-raphy provided only two-dimensional static images

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

Microscopy 62(Supplement 1): S29–S41 (2013)doi: 10.1093/jmicro/dft006

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

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

with low sensitivity for electron phase measure-ment. These drawbacks were overcome by strenu-ous efforts to develop new techniques in the 1990and 2000s. Nowadays, advanced electron holog-raphy techniques enable real-time dynamic observa-tion [4,5], 3D reconstruction of electromagneticfields [6,7] and highly sensitive phase measurement(e.g. detection at 1/300 of the electron wavelength)[8,9]. As a result, electron holography has becomean indispensable microscopy technique for bothscience and industry. This paper reviews the abovethree advanced electron holographic techniquesand some leading applications [10–12] over the past20 years.

Development of advanced electron

holographic techniques

On-line real-time observation of magnetic

domains

Dynamic observation often provides valuable infor-mation in materials science. However, such observa-tions were difficult with electron holography becauseholography needs a two-step imaging process. In1991, Matsuda et al. succeeded in observing fluxondynamics by digitally reconstructing frame by frameelectron holograms recorded on video tape [13]. Thiswas a breakthrough in the use of electron holographyfor dynamic observation. Nevertheless, frame-by-frame reconstruction is time consuming. Moreover,the transmission electron microscope (TEM) oper-ator cannot observe the reconstructed images whileoperating the microscope.In 1994, our group developed an on-line real-time

electron holography system and observed thedynamic behavior of magnetic domains [4,5]. In thissystem (Fig. 1), time-varying holograms are detectedby a TV camera attached to an electron holographyTEM (Hitachi, HF-2000). They are transferred as avideo signal to a liquid-crystal spatial light modulatorlocated in a Mach–Zehnder interferometer for imagereconstruction. The reconstructed images are sentback to a monitor placed beside the TEM. Using thissystem, the TEM operator is able to observe bothtime-varying holograms and their reconstructedimages at the same time. The time resolution isdetermined by the video rate (1/30 s) of the TV

camera shown in Fig. 1, and thus, we can observethe dynamics slower than this rate in real time.Examples of reconstructed interference micro-

graphs of a thin permalloy film, along with illustra-tive diagrams of magnetic flux lines and domains,are shown in Fig. 2. In this experiment, the object-ive lens was turned off, and then the specimen wasplaced into the TEM and tilted several degrees. Thefocus was adjusted by the first intermediate lens,and a magnetic field was applied to the specimenby increasing current to the objective lens. Thedynamic behavior of the magnetic flux and domainsduring magnetization and demagnetization is clearlyevident.

Three-dimensional reconstruction

of electromagnetic fields

Although optical holography enables 3D visualiza-tion of objects, it is rather difficult to obtain a 3Dmodel of a specimen by electron holography. This

Fig. 1 Schematic diagram of on-line real-time electron holographysystem. I: side view of electron holography TEM. II: plan view ofMach–Zehnder interferometer for image reconstruction. Hologramsformed by TEM are detected by TV camera and transferred toliquid-crystal spatial light modulator in the interferometer.Reconstructed images are sent to monitor next to TEM.

S30 M I C RO S CO P Y , Vol. 62, Supplement 1, 2013

is because electron holography provides the phaseinformation projected along the direction of theelectron beam. To overcome this problem, weapplied the theory and method of computedtomography.Two types of algorithms are basically used for re-

construction in computed tomography: noniterativeand iterative. The noniterative ones are based onthe fact that the Fourier transformation of a one-dimensional projection of a two-dimensionalsection is identical to the corresponding centralline of the two-dimensional Fourier transformationof the section. These algorithms can achieve high-speed and high-quality reconstruction when a largeamount of projection data is available without anylarge-angle missing cone. In contrast, the iterativealgorithms are a bit slower because they require anumber of iterative calculations; however, they canincorporate physical constraints.To reconstruct the 3D electric potential distribu-

tion, we used an iterative method because therange of the tilt angle in the TEM is limited (±60°).We used fine particles of latex on a thin carbon film

because such particles are amorphous (i.e. they donot produce complicated phase noise due to elec-tron diffraction) and have an electric inner potentialappropriate for the phase measurement. Anexample hologram and its reconstructed phaseimage are shown in Fig. 3a and b, respectively [6].The specimen was tilted in the TEM, and 24 holo-grams were taken every 5°. Bird’s eye views of the3D reconstruction from two different directions areshown in Fig. 4a and b. The reconstructed data canbe used to display contour and/or phase maps ofany section from any direction.To reconstruct the 3D magnetic vector fields, we

used a noniterative method. In a fundamental study,we found that the x and the y components of themagnetic flux density projected in the z directionare the derivatives of the two-dimensional phasedistribution with respect to (−y) and x, respectively.This indicates that the x and the y components canbe reconstructed by using a computed tomographyalgorithm for a scalar distribution. Furthermore, the

Fig. 3 Amorphous latex particles on thin carbon film. (a)Hologram. (b) Reconstructed phase image.

Fig. 4 Bird’s eye views of 3D reconstruction of particles from twodifferent directions.

Fig. 2 Dynamic observation of magnetic domains. (a)–(c)Reconstructed interference micrographs of thin permalloy film.(a0)–(c0) Illustrative diagrams of magnetic flux lines and domains.Dynamic behavior of film during magnetization and demagnetizationis clearly evident.

K. Yamamoto et al. Development of advanced electron holographic techniques S31

z component of the magnetic flux density can bedetermined by using equation divB = 0.An example hologram of a single-magnetic-domain

particle of barium ferrite and its reconstructedphase image are shown in Fig. 5a and b, respectively[7]. The 3D magnetic field in the vicinity of thesurface of the particle was reconstructed using �50holograms recorded by double-tilting the specimen.The 3D reconstructed magnetic field along theplanes A and B in Fig. 5a are shown in Fig. 6a and b,respectively. The magnetic fields are clearly seen.3D electric potential distributions across a p–n

junction in semiconductor devices were successful-ly obtained by Twitchett-Harrison et al. in 2007 [14].These 3D techniques should be useful for studyingindustrial materials and devices as well.

High-resolution and high-sensitivity phase

measurement by phase-shifting electron

holography

Observation of weak electromagnetic fields on ananometer scale requires simultaneous improve-ment of the spatial resolution and phase detectionsensitivity. Both of these characteristics, however,are essentially limited in conventional phase recon-struction method based on Fourier transformation.The spatial resolution is determined by the spacingof the interference fringes, so narrower fringes arenecessary to obtain higher resolution. However, thelack of electron coherency and brightness reducesthe fringe contrast, resulting in degraded sensitivity.The phase-shifting electron holography developed

by Ru et al. [15,16] can be used to reconstruct theobject wave without Fourier transformation. Theobject wave can be reconstructed from a series ofholograms in which the interference fringes areshifted one after another. The spatial resolution isdetermined by the pixel size of the hologram.Therefore, an object wave with higher resolutioncan be obtained from coarse interference fringeshaving a higher fringe contrast, meaning that thespatial resolution and sensitivity can be simultan-eously improved. However, Fresnel diffraction at anelectron biprism (Fig. 7a) creates two problemsduring phase-shifting reconstruction. One is thenon-uniform interference fringes resulting from theoverlapping of the Fresnel fringes, as shown inFig. 7b and c. In the phase-shifting reconstructionprocess, the phase is calculated by fitting the inten-sity change in each pixel of the hologram to acosine curve. Thus, when the non-uniform fringes(intensity profile shown in Fig. 7c) are shifted, theintensity at each pixel does not vary in accordancewith the ideal cosine curve, resulting in a phase

Fig. 5 Single-magnetic-domain particle of barium ferrite. (a)Hologram. (b) Reconstructed phase image.

Fig. 6 3D reconstructed magnetic fields along (a) plane A and (b)plane B in Fig. 5a. Magnetic fields are clearly evident.


calculation error. The other problem is that theFresnel diffraction directly distorts the electronwaves. As shown in the reconstructed phase image(Fig. 7d), the large phase distortion prevents obser-vation of the weak electromagnetic fields with aphase of <0.5 rad. The image processing method wedeveloped overcomes these problems by correctingfor the effect of the Fresnel diffraction [8,9].The intensity of the interference fringes is given

by the sum of the reference wave and the objectwave:

Iðx; yÞ ¼jCr þCoj2¼jAðx; yÞ exp½�ikax�þ Bðx; yÞ exp½ikaxþ Dfðx; yÞ�j2

¼jAðx; yÞj2 þ jBðx; yÞj2þ 2Aðx;yÞBðx; yÞ cos½2kaxþ Dfðx; yÞ�; ð2Þ

where A is the amplitude of the reference wavemodulated by the Fresnel diffraction, B is the ampli-tude of the object wave modulated by the Fresnel

diffraction and the object (specimen), Df is thephase of the object wave, α is the angle of electronwaves deflected by the electron biprism and k = 2π/λ, where λ is the electron wavelength. The top andbottom envelope curves of the non-uniform interfer-ence fringes, Etop and Ebottom (indicated in Fig. 7c),are derived from Eq. (2):

Etop ¼ jAðx; yÞj2 þ jBðx; yÞj2 þ 2Aðx; yÞBðx; yÞ;Ebottom ¼ jAðx; yÞj2 þ jBðx; yÞj2 � 2Aðx; yÞBðx; yÞ:

ð3ÞThe background jAðx; yÞj2 þ jBðx; yÞj2 and ampli-

tude 2Aðx; yÞBðx;yÞ terms of the non-uniform inter-ference fringes are independently obtained from(Etop +Ebottom)/2 and (Etop −Ebottom)/2, respectively,so we can normalize the intensity of the experimen-tal holograms. In this study, the Etop and Ebottom

curves of each hologram were calculated usingLagrange’s three-order interpolation [8].Figure 8a and b shows a TEM image of a latex

particle sitting on a carbon film edge and one of theholograms in the series, respectively. The intensity ofthe hologram in Fig. 8c was normalized using thesetwo envelope curves. This intensity profile expressesthe only cosine function of Eq. (2), so a phase calcu-lation error does not arise from the phase-shifting re-construction process.The problem of the Fresnel diffraction directly

distorting the electron waves is overcome by sub-tracting the phase distortion of the reference holo-grams obtained in a vacuum without a specimen fromthe reconstructed phase of the holograms. The non-uniform fringes of the reference holograms are alsonormalized using the envelope curves. A 24-times-phase-amplified image reconstructed without thesecorrections is shown in Fig. 8d. Quantitative in-formation about the weak electric field aroundthe charged latex particle obviously cannot beextracted.A phase image reconstructed using the correc-

tions is shown in Fig. 8e. The weak electric field isclearly observable. The phase in this reconstructedphase image can be amplified by 100 times, asshown in Fig. 8f. A phase detection sensitivity of2π/300 rad was achieved for 10 nm spatial reso-lution. These Fresnel corrections are essential forusing phase-shifting electron holography to observe

Fig. 7 Effect of Fresnel diffraction on amplitude and phase ofelectron waves. (a) Illustration of Fresnel diffraction at electronbiprism. (b) Interference fringes without object. (c) Intensity profileof non-uniform fringes of (b). (d) Reconstructed phase imageshowing phase distortion due to Fresnel diffraction.


weak electromagnetic fields. This technique hasbeen applied to nanometer-scale magnetic multi-layer materials showing giant magnetic resistance[17], compound semiconductors materials such asGaAs [10] and atomic structure observations withspherical aberration correction [18].

Application of electron holography to

industrial materials and devices

Dopant profiling of GaAs compound

semiconductors

Frabboni et al. were the first to experimentallyobserve a p–n junction in silicon (1985) [19,20], and

McCartney et al. measured the potential dropacross the junction [21]. In 1999, Rau et al. suc-ceeded in observing two-dimensional electric po-tential distribution in the cross section ofSi-transistors [22]. This potential distribution isformed by dopant distribution. Therefore, suchquantitative observation is called ‘dopant profiling’and caught the attention of a large number of semi-conductor engineers and electron microscopists.We then started making observations of com-

pound semiconductor materials; however, the p–njunctions or dopant distributions were still unob-servable. We suspected that the problem had some-thing to do with the damaged layers caused by thefocused ion beam (FIB) etching and decided toobserve the damaged layer itself. In our preliminaryobservations we found that the damaged layer con-tained small crystalline particles, which created acomplicated phase distribution due to electron dif-fraction. We, therefore, tried to remove thedamaged layer [10]. First, a thin aluminum foil wasattached to a copper plate using epoxy resin. Then,a small cross-sectional specimen was extractedfrom the test sample using FIB micro-sampling andfixed to a cross section of the thin aluminum foil bytungsten deposition. The specimen was thinned inan FIB system, and finally, both the top and bottomsurfaces of the specimen were milled using a low-energy Ar-ion beam for 5 min. Cross-sectional ob-servation before and after the Ar-ion millingshowed that the thickness of the damaged layerwas reduced from 20 to 2 nm.After the Ar-ion milling, we tried to observe a

model sample of p–n–p multi-layers deposited on aGaAs semi-insulating substrate by phase-shiftingelectron holography [10]. An example hologram afterFresnel fringe correction [8] is shown in Fig. 9a, aphase image reconstructed by phase-shifting elec-tron holography using 13 holograms is shown inFig. 9b, and a schematic is shown in Fig. 9c. The pand n regions are shown with clear contrast inFig. 9b. Furthermore, the low- and high-dopant-concentration n-regions are distinguished by thehigh contrast. This unexpected high contrast wasexplained by considering the thickness of the deple-tion layers and the electrically dead layers.These dopant profiling and sample preparation tech-

niques are thus useful for developing compound

Fig. 8 Charged latex particle on carbon film edge. (a) TEM image.(b) Hologram. (c) Hologram corrected using envelope curves ofnon-uniform fringes. (d) Reconstructed phase distribution withoutcorrections (24 times phase amplified), (e) with corrections (24times phase amplified), (f) with corrections (100 times phaseamplified).


semiconductor devices such as lasers and high-speed transistors.

Electron holography of a hetero-interface

in a solid oxide fuel cell (SOFC)

Solid oxide fuel cells (SOFCs), a large-scale energysource that can operate at high temperatures, areadvantageous compared with other types of fuelcells because they can use many kinds of fuel gas.Moreover, they do not require a Pt catalyst to gener-ate high levels of electricity. However, several pro-blems have to be overcome before they can be putinto general use. One problem is the ‘overpotentialeffect’, which is a decrease in the voltage from theexpected level in a stable state as an operatingcurrent. One possible cause of this problem is theaccumulation of O anions near the anode. Anotheris the lack of reserving O anions near the cathodedue to the poor charge transfer efficiency.In order to reveal such problems, it is valuable to

observe the ion motion at the interface. Numerousstudies have been conducted on various types ofSOFC systems with different electrode and electro-lyte materials in an effort to clarify and characterizethe reactions at the tri-phase boundary (TPB) andthe overall operation of the fuel cell [23–26].However, there have been only a few in situ TEMinvestigations in which reactions at the TPB wereobserved [27]. Electron holography has been usedto observe the electrostatic inner potentials in anO-ionic conductor when an external potential isapplied. Ion motions have been observed in thegadolinium-doped-ceria (GDC) with Pt electrodessystem [11].

To simulate fuel cell operation, we are experimen-tally applying an external potential to specimenSOFC cells with heating in a fuel gas environment.Initially, we observed the inner potentials by apply-ing an external potential to specimen cells in avacuum at room temperature. When an externalelectric potential is applied to solid electrolytes invacuum, O anions are drawn toward the anode.This shift in the O anions causes a shift in the elec-trostatic potential, so the ion flow could be blockedunless new O anions are reserved from the cathode.This is the same phenomenon as the overpotentialeffect.A noble specimen holder for the TEM used in this

experiment was developed to enable an electricfield to be applied to the specimen at high tempera-tures [28]. The holder (Fig. 10) has four electrodes:two for heating the specimen and two for applyinga potential of up to ±5 V. Its use enables the speci-men to be heated up to 1000°C. A cell sample wasdeposited by pulsed laser deposition (PLD) on a Siwafer (30 nm-thick Pt film, a 3 µm-thick GDC layerand a 60 nm-thick Pt electrode). A specimen forcross-sectional electron holographic observationwas prepared using an FIB micro-sampling tech-nique and mounted on the side of the Ta sheetheater (Fig. 10). The heater was 6 mm long, 0.5 mmwide and 0.03 mm thick. The Au lead wire was 3µm in diameter. The specimen cut from the waferwas 18 µm wide, 10 µm long and 3 µm thick.A TEM image of the Pt/GDC interface is shown in

Fig. 11a, a hologram of the area around the inter-face is shown in Fig. 11b and the simply recon-structed phase image is shown in Fig. 11c. Theimage shown in Fig. 11d was unwrapped from thesimply reconstructed phase image (Fig. 11c).

Fig. 10 Noble specimen holder developed to enable electric field tobe applied to specimen at high temperatures.

Fig. 9 Model sample of p–n–p multi-layers of GaAs deposited onGaAs semi-insulating substrate. (a) Example hologram after Fresnelfringe correction. (b) Phase image reconstructed by phase-shiftingelectron holography using 13 holograms. (c) Schematic. In (b), pand n regions are seen with clear contrast. Low- andhigh-dopant-concentration regions are distinguished by highcontrast.


The external potential applied and experimentalset-ups are shown in Fig. 12a–c. Figure 12d showsthe reconstructed phase obtained from the hologramwithout external potential (Fig. 11d), which corre-sponds to the mean inner potential distribution ofthe pure Pt and GDC. The delay in the electronphase corresponds to the positive potential. Since Pthas a larger mean inner potential than GDC, it had alarger phase delay, so it is shown darker. When weapplied external potential of −1.0 V to the Pt elec-trode, the reconstructed phase changed slightly asshown in Fig. 12e. Subtracting the phase distributionin Fig. 12d from that in Fig. 12e reveals the pureeffect of the additive potential on the electrostaticinner potential, as shown in Fig. 12f.Figure 13a and c shows the difference in the

phase profile when an external potential (−1.0 and1.0 V, respectively) was applied to the Pt electrode.

The difference between the phase applied −1.0 Vand that before applied any potential is shown inFig. 13a as a profile along the long side of the rect-angular inlet in Fig. 12f. The external voltageshould reduce the potential from the negative elec-trode to the positive one because the potential isdrawn so as to correspond to that for electrons.Therefore, the inner potential was expected tosimply decrease. The profile of the phase difference(Fig. 13a), however, shows that an electric doublelayer �4 nm thick (shown in red) formed at theinterface due to the dielectric polarization and thatthe O anions in the GDC were repulsed toward thepositive side and accumulated in an area �20 nmfrom the anode. This reduced the potential some-what, as illustrated in Fig. 13b.In contrast, a voltage of +1.0 V applied to the

same electrode as in Fig. 13a reversed the

Fig. 11 Electron holography of Pt/GDC interface. (a) Ordinarily TEM image. (b) Hologram of area around interface. (c) Simply reconstructedphase image. (d) Unwrapped phase image.


polarization of the electric double layer, resulting inthe accumulation of O-ion vacancies (positivecharges). This is evidenced by the small dip in thepotential at �6 nm from the interface. The thick-ness of the double layer was �3 nm in this case.The difference in the thickness of the electricdouble layers shown in Fig. 13a and c may havebeen due to the poor resolution of the hologram,�3 nm. However, the difference in relaxation lengthshown in these two figures is significant, and it mayhave been due to the difference in polarization of

the applied voltage. These results reflect the charac-teristics of the electrolyte GDC and electrode Pt.Although the thickness of the specimen and the

exact value of the potential were not determined,this was the first time that electric double layerswere observed at the interface. The results confirmthat the localization of O anions caused by an exter-nal electric field in solid electrolytes can beobserved with in situ electron holography. Theaccumulation of anions or vacancies was due toconducting the experiment in a vacuum at room

Fig. 13 Difference in phase profile when negative potential (–1.0 V) was applied to left-side electrode (a) and expected distributions of Oanions and vacancies (c); those when positive potential (+1.0 V) was applied (b) and (d). Electric double layers were observed also withopposite polarities.

Fig. 12 Experimental setups (a)–(c) and extraction of pure applied potential effect. Original reconstructed phase (unwrapped) obtainedwithout an external potential (d) was subtracted from the phase obtained with potential applied (e), resulting in the phase difference imageshowing pure external potential effect on electrostatic inner potential (f).


temperature. Experiments in an oxygen gas atmos-phere at high temperatures will make the effect ofvacuum atmosphere clearer, and close on theactual cell’s reaction.

In situ electron holography observation of

electric potential in an all-solid-state Li-ion

battery

All-solid-state Li-ion batteries are promising energystorage devices because of their safety, lifetime, costand energy density characteristics. The mainproblem that prevents practical use is their lowpower density, which is attributed to the large resist-ance of Li-ion transfer around the electrode/solid-electrolyte interfaces. In general, when the Li-ionconcentration changes in the battery materialsduring the charge–discharge reaction, the electrodepotential of the materials should change because ofelectrochemical oxidation or reduction. To observethe local electrode potential change around theinterfaces, we used in situ electron holography in aTEM to observe the battery reaction [12,29].The battery sample we prepared is schematically

illustrated in Fig. 14a. An 800-nm-thick LiCoO2

positive electrode was deposited by PLD on a90-µm-thick solid-electrolyte sheet of Li1+x+yAlyTi2−ySixP3−xO12 (LATSPO) [30]. Then, Au and Ptwere coated as current collectors on one side andthe other side, respectively. The negative electrode

was formed in situ around the LATSPO/Pt interfaceduring the first charging process in the TEM.Partial Li insertion reaction into the LATSPO nearthe interface (pink-colored region in Fig. 14a)irreversibly decomposed the LATSPO, resulting inthe ‘in situ formed negative electrode’ [31]. Thiscaused the electrode to bond to the electrolyteon an atomic scale, thereby reducing the inter-facial resistance, which was 100 Ω cm2 at thenegative-electrode/LATSPO interface and 4000 Ωcm2 at the LiCoO2/LATSPO interface [29]. Thepositive- and negative-side regions (indicated by redboxes in Fig. 14a) were thinned by FIB, and the TEMimages of these regions are shown in Fig. 14c and d,respectively. The LiCoO2 was grown as columnarstructures on the LATSPO sheet. The LATSPO con-sisted of a mixture of amorphous and crystallizedgrains. Figure 14b plots the initial cyclic voltammo-gram (CV) measured in the TEM. A pair of redoxpeaks is observed around 1.6 V, which means that thesample worked as a rechargeable battery in the TEM.Holograms around both interfaces were taken

during the CV measurement at a given voltage (indi-cated by the arrows in Fig. 14b). The potential pro-files along the positive side A–B and the negativeside C–D, surrounded by dotted lines in Fig. 14cand d, were reconstructed from the holograms.Figure 15a–f and g–l shows the profiles near thepositive- and negative-side interfaces, respectively.

Fig. 14 Preparation of all-solid-state Li-ion battery. (a) Schematic illustration of the sample. (b) Initial CV measured in TEM at potentialsweep rate of 40 mV min−1. (c) TEM image around LiCoO2/LATSPO interface. (d) TEM image around LATSPO/Pt interface.


The spatial scale on the horizontal axis is markedfrom the electrode/current–collector interfaces. Thepotential (shown on the vertical axis) is shown as arelative value against the flat potential level (whichwas set to 0 V) to enable us to see how much thepotential is distributed around the positive- andnegative-side interfaces.At 0.42 V (Fig. 15g), no potential was distributed

on the negative side, so the voltage was mainlyapplied to the positive side. Figure 15a shows thepositive-side potential distribution at 0.72 V. Ourprevious study [12] showed that the linear slope inthe LiCoO2 results from a shift in the electrical

band structure caused by extracting Li ions, whichindicates a subtle difference in the Li-ion concen-tration in the electrode. As shown in Fig. 15a,there was a steep potential drop at the positive-side interface and a gradual potential slope at theLATSPO sheet (indicated by the dashed redcurve). This gradual slope was possibly due to theLi-ion poor region resulting from the movement ofLi ions. At 1.05 V (Fig. 15b and h), a negative-sidepotential change appeared within 3 μm in distancefrom the interface, and the potential slope changedat a certain point (red arrow position), �700 nmfrom the interface. When the voltage was increasedto 1.53 V (Fig. 15c and i) and 1.95 V (Fig. 15d andj), the positive-side potential was mostly main-tained while the negative-side one was distributedmore deeply, especially to the right of the redarrow.When Li ions are charged in the negative elec-

trode, the electrode potential generally decreases.Therefore, the deeper potential was probably dueto Li-ion charging in the negative electrode. Whenthe voltage was reduced to 1.48 V (Fig. 15e and k)and 1.26 V (Fig. 15f and 1) in the discharged state, adifferent potential distribution formed on bothsides. On the positive side, a flat potential distribu-tion was newly observed inside the LATSPO aroundthe interface. It was 300 nm at 1.48 and 210 nm at1.26 V. The spatial resolution of the holographymeasurement was 100 nm. The potential resolutionwas estimated from the difference between the ex-perimental profile and the trend curve fitted with apolynomial, and it was �0.02 V in the LATSPOregion. Thus, the reduction in the flat potential dis-tribution was significant. On the negative side, thegentle slope on the left side of the red arrow almostflattened out compared with those in Fig. 15h and jin the charged state. A slightly lower potential(�0.08 V) appeared in a region �1600 nm wide,moving this region �300 nm to the left between1.48 and 1.26 V. On the right side of the red arrow,the deep potential slope remained.The 700-nm-wide negatively charged region at the

right side of the red arrow is the ‘in situ formednegative electrode’ as we described above. The po-tential slope inside the negative electrode probablyresulted from the difference in density of the negativeelectrode materials. On the left side of the red arrow

Fig. 15 In situ observation of electric potential distributions duringinitial CV measurement. (a)–(f) Distribution on positive side alongline A–B in Fig. 14c, (g)–(l) on negative side along line C–D inFig. 14d. Voltages applied between Au and Pt current correctorscorrespond to points (a)–(l) in Fig. 14b. Potential (V) is shown onthe vertical axis.


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

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

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

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

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

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

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

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

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

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

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

(the negative side), a gentle slope region formedduring the charge process. However, it mainly disap-peared during the discharge process, and a1600-nm-wide low-potential region formed insteadinside the LATSPO. Because this region has no signifi-cant TEM contrast, a small number of Li ions mighthave been trapped in certain places such as the grainboundaries, which would retard Li-ion movement inthe solid electrolyte. Moreover, a smooth potentialchange appeared at the negative-electrode/LATSPOinterface (red arrow position), while the potentialdropped sharply at the positive-side interface. Thispotential drop is probably related to the difference ininterfacial resistance between the interfaces.On the positive side, a gradual potential was ob-

served in the charged states. In the discharged states,the potential in the LATSPO around the interface wasslightly decreased, resulting in the flat distribution.The difference in both cases was the direction ofLi-ion movement. This interface had the high inter-facial resistance of Li-ion transfer (4000 Ωcm2), so the300-nm-wide flat region was probably due to the accu-mulation of Li ions in front of the interface. In thiscase, the decrease to a 210-nm-wide flat region at1.26 V was due to the decrease in the number of Li-ionmovements, as supported by the CV measurementresults.In short, in situ electrochemical electron holog-

raphy can be used to visualize how Li ions moveduring the charge–discharge reaction and to identifythe retardant region of the Li-ion transfer. Such infor-mation is essential to developing high-performancebattery devices.

Concluding remarks

We have reviewed three advanced electron holographytechniques: on-line real-time electron holography,3D reconstruction of electromagnetic fields and highsensitivity phase-shifting electron holography. Wealso presented several applications of electron hol-ography to the analysis of industrial materials anddevices: GaAs compound semiconductors, SOFCsand all-solid-state Li-ion batteries. Electron holog-raphy is becoming a more and more important mi-croscopy technique for materials science and deviceengineering, and its use should lead to the break-throughs essential for creating a sustainable society.

Funding

Experiments shown in Figs. 1–6 were carried outby the authors in “Tonomura Wavefront Project” ofJapanese Research & Development Corporation ofJapan. The study of SOFC was partially supportedby a Grant-in-Aid for Scientific Research on PriorityArea, Nanoionics’ (439), and grant No. 20246013 bythe Ministry of Education, Culture, Sports, andTechnology, Japan, and also by ‘MEXT Program forDevelopment of Environmental Technology usingNanotechnology’. The study of Li-ion batteries wasfinancially supported by Chubu Electric Power Co.,Inc.

Acknowledgements

The authors thank Dr G. Lai, Prof. J. Chen, Dr K. Ishizuka,Dr A. Fukuhara, Dr H. Sasaki, Dr S. Ootomo, Mr T. Matsuda, Dr F. Iwase,Mr R. Nakasaki, Dr H. Ishii, Dr A.H. Tavabi, Ms Z. Yasenjiang,Mr H. Moritomo, Mr K. Oura, Mr S. Enomoto, Prof. Y. Iriyama,Dr T. Asaka, Dr H. Fujita, Dr C.A.J. Fisher, Mr K. Nonaka, Mr K.Miyahara, Dr Y. Sugita, Prof. Z. Ogumi, Mr I. Kawajiri andProf. M. Hibino for their experimental support and discussion. Theyalso thank Dr Akira Tonomura for his supervision and encouragementthroughout this work.

References

1 Tonomura A, Allard L F, Pozzi G, Joy D C, and Ono Y A (eds) (1995)Electron Holography (Elsevier Science B. V., Amsterdam, Lausanne,New York, Oxford, Shannon, Tokyo).

2 Tonomura A (1999) Electron Holography, 2nd edn (Springer-Verlag,Berlin, Heidelberg, New York).

3 Völkl E, Allard L F, and Joy D C (eds) (1999) Introduction to ElectronHolography (Kluwer Academic/Plenum Publishers, New York).

4 Chen J, Hirayama T, Lai G, Tanji T, Ishizuka K, and Tonomura A(1993) Real-time electron-holographic interference microscopy witha liquid-crystal spatial light modulator. Opt. Lett. 18: 1887–1889.

5 Hirayama T, Chen J, Tanji T, and Tonomura A (1994) Dynamic ob-servation of magnetic domains by on-line real-time electron holog-raphy. Ultramicroscopy 54: 9–14.

6 Lai G, Hirayama T, Ishizuka K, Tanji T, and Tonomura T (1994)Three-dimensional reconstruction of electric-potential distribution inelectron-holographic interferometry. Appl. Opt. 33: 829–833.

7 Lai G, Hirayama T, Fukuhara A, Ishizuka K, Tanji T, and Tonomura A(1994) Three-dimensional reconstruction of magnetic vector fields usingelectron-holographic interferometry. J. Appl. Phys. 75: 4593–4598.

8 Yamamoto K, Kawajiri I, Tanji T, Hibino M, and Hirayama T (2000)High precision phase-shifting electron holography. J. ElectronMicrosc. 49: 31–39.

9 Yamamoto K, Hirayama T, Tanji T, and Hibino M (2003) Evaluationof phase-shifting electron holography by using hologram simulation.Surf. Interface Anal. 35: 60–65.

10 Sasaki H, Yamamoto K, Hirayama T, Ootomo S, Matsuda T, Iwase F,Nakasaki R, and Ishii H (2006) Mapping of dopant concentration ina GaAs semiconductor by off-axis phase-shifting electron holog-raphy. Appl. Phys. Lett. 89: 244101.

11 Tavabi A H, Yasenjiang Z, and Tanji T (2011) In situ off-axis electronholography of metal-oxide hetero-interfaces in oxygen atmosphere.J. Electron Microsc. 60: 307–314.


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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

12 Yamamoto K, Iriyama Y, Asaka T, Hirayama T, Fujita H, Fisher C AJ, Nonaka K, Sugita Y, and Ogumi Z (2010) Dynamic visualization ofthe electric potential in an all-solid-state rechargeable lithiumbattery. Angew. Chem. Int. Ed. Engl. 49: 4414–4417.

13 Matsuda T, Fukuhara A, Yoshida T, Hasegawa S, and Tonomura A(1991) Computer reconstruction from electron holograms and obser-vation of fluxon dynamics. Phys. Rev. Lett. 66: 457–460.

14 Twitchett-Harrison A C, Yates T J V, Newcomb S B, Dunin-Borkowski R E, and Midgley P A (2007) High-resolution three-dimensional mapping of semiconductor dopant potentials. NanoLett. 7: 2020–2023.

15 Ru Q, Endo J, Tanji T, and Tonomura A (1991) Phase-shifting elec-tron holography by beam tilting. Appl. Phys. Lett. 59: 2372–2374.

16 Ru Q, Lai G, Aoyama K, Endo J, and Tonomura A (1994) Principle andapplication of phase-shifting electron holography. Ultramicroscopy 55:209–220.

17 Tanji T, Hasebe S, Nakagami Y, Yamamoto K, and Ichihashi M(2004) Observation of magnetic multilayers by electron holography.Microsc. Microanal. 10: 146–152.

18 Yamamoto K, Sugawara Y, McCartney M R, and Smith D J (2010)Phase-shifting electron holography for atomic image reconstruction.J. Electron Microsc. 59: S81–S88.

19 Frabboni S, Matteucci G, and Pozzi G (1985) Electron holographicobservations of the electrostatic field associated with thin reverse-biased p-n junctions. Phys. Rev. Lett. 55: 2196–2199.

20 Frabboni S, Matteucci G, and Pozzi G (1987) Observation of electro-static fields by electron holography: the case of reverse-biased p-njunctions. Ultramicroscopy 23: 29–38.

21 McCartney M R, Smith D J, Hull R, Bean J C, Voelkl E, and Frost B(1994) Direct observation of potential distribution across Si/Si p-njunctions using off-axis electron holography. Appl. Phys. Lett. 65:2603–2605.

22 Rau W D, Schwander P, Baumann F H, Höppner W, and Ourmazd A(1999) Two-dimensional mapping of the electrostatic potential intransistors by electron holography. Phys. Rev. Lett. 82: 2614–2617.

23 Nielsen J and Jacobsen T (2007) Three-phase-boundary dynamics atPt/YSZ microelectrodes. Solid State Ionics 178: 1001–1009.

24 Backhaus-Ricoult M, Adib K, St. Clair T, Luerssen B, Gregoratti L,and Barinov A (2008) In situ study of operating SOFC LSM/YSZ cath-odes under polarization by photoelectron microscopy. Solid StateIonics 179: 891–895.

25 Kishimoto H, Sakai N, Yamaji K, Horita T, Brito M E, Yokokawa H,Amezawa K, and Uchimoto Y (2008) Visualization of oxygen trans-port behavior at metal electrode/oxide electrolyte interface usingsecondary ion mass spectrometry. Solid State Ionics 179: 347–354.

26 Cheng Z and Liu M (2007) Characterization of sulfur poisoning ofNi–YSZ anodes for solid oxide fuel cells using in situ Raman micros-copy. Solid State Ionics 178: 925–935.

27 Jeangros Q, Faes A, Wagner J B, Hansen T W, Aschauer U, VanHerle J, Hessler-Wyser A, and Dunin-Borkowski R E (2010) In situredox cycle of a nickel–YSZ fuel cell anode in an environmentaltransmission electron microscope. Acta Mater. 58: 4578–4589.

28 Moritomo H, Oura K, Tanji T, and Enomoto S (2006) New specimenholder with 4 electrodes. Proc. IMC16 2: 1154.

29 Yamamoto K, Iriyama Y, Asaka T, Hirayama T, Fujita H, Nonaka K,Miyahara K, Sugita Y, and Ogumi Z (2012) Direct observation oflithium-ion movement around an in-situ-formed-negative electrode/solid-state-electrolyte interface during initial charge-discharge reac-tion. Electrochem. Commun. 20: 113–116.

30 Fu J (1997) Fast Li+ ion conduction in Li2O-Al2O3-TiO2-SiO2-P2O5

glass-ceramics. J. Am. Ceram. Soc. 80: 1901–1903.

31 Iriyama Y, Yada C, Abe T, Ogumi Z, and Kikuchi K (2006) A newkind of all-solid-state thin-film-type lithium-ion battery developed byapplying a D.C. high voltage. Electrochem. Commun. 8: 1287–1291.


review development of advanced electron holographic … · 2017-04-23 · review development of...

Documents