110-1X2 Surface) (100)Facet (111)Facet [110] [220]

ISSN 1349-6832

July2011 Vol.46 No.1

(110-1x2 Surface) (100)Facet (111)Facet _ [110] _ [220]

[001] Volume 46 Number 1 Study of Nanoparticles at UTSA: July, 2011 One Year of Using the First JEM-ARM200F Installed in USA

Alvaro Mayoral†,†††, Rodrigo Esparza†, Francis Leonard Deepak†,††, Gilberto Casillas†, Sergio Mejía-Rosales††††, Arturo Ponce† and Miguel José-Yacamán†

† Department of Physics and Astronomy, University of Texas at San Antonio †† International Iberian Nanotechnology Laboratory, Avda Mestre Jose Veiga ††† Laboratorio de Microscopías Avanzadas (LMA), Instituto de Nanociencia de Aragón (INA), Universidad de Zaragoza, c/Mariano Esquillor, Edificio I+D †††† Center for Innovation and Research in Engineering and Technology, and CICFIM-Facultad de Ciencias Físico-Matemáticas, Universidad Autónoma de Nuevo León Contents

Study of Nanoparticles at UTSA: One Year of Using the First JEM-ARM200F The first results from our group that have been obtained using the Installed in USA ...... 1 newly installed aberration-corrected STEM microscope (JEOL JEM- Exploring Biological Samples in 3D ARM200F) are reported. Studies were carried out on noble metal Beyond Classic Electron Tomography . . 6 nanoparticles and their corresponding alloys and / or core-shell structures. In this paper we focus on some of the exciting areas of research Application of Scanning Electron Microscope that have been under investigation in our group. These include studies to Dislocation Imaging in Steel ...... 11 on Au nanoparticles, bimetallic Au-Co nanoparticles and core-shell Au-Pd Atmospheric Scanning Electron nanoparticles. In addition, studies that were carried out on very small clusters namely Pd-Ir and other similar systems have also been high- Microscopy (ASEM) Realizes Direct lighted in this report. EM-OM Linkage in Solution: Aqueous Immuno-Cytochemistry ...... 17 Information Derived from PGSE-NMR ~Ion Diffusion Behavior, Molecular Introduction CESCOR software, to finally obtain a Association, Molecular Weight/ twelve-fold Ronchigram with a flat area of Composition Correlation of Synthetic The University of Texas at San Antonio 50 mrad. The pixel spacing was calibrated Polymers~ ...... 23 organized recently an Electron Microscopy using Si [110] lattice images in the HAADF Development of JEM-2800 High core facility named the "Kleberg Advanced mode, and confirmed by using gold standard Throughput Electron Microscope. . . . . 31 Microscopy Facility". The most relevant particles. Images were commonly recorded for 10 to 16 s. The probe current used for Introduction of New Products ...... 37 instrument is a JEOL JEM-ARM200F (with FEG) with probe aberration correction; this acquiring the HAADF as well as the BF- instrument was the first one installed outside STEM images was 9C (23.2 pA) and the CL Japan and is operational since February 2010. aperture size was 40 m. High angle annular In this paper we report some of the results dark-field (HAADF) STEM images were after one year of its operation. acquired with a camera length of 8 cm/6 cm and the collection angle of 50-180 mrad/67- 250 mrad was used. The BF-STEM images Experimental were obtained using a 3 mm /1 mm aperture and a collection angle of 11 mrad / 3.8 mrad Cover micrograph For the electron microscopy analysis, the was used (camera length in this case was 8 nanoparticles samples were dispersed in cm). The HAADF as well as the BF images HAADF-STEM image of a gold decahedral ethanol and a drop of this suspension was were acquired using a digiscan camera. In particle oriented near a five-fold axis. Around deposited onto a holey carbon grid. The sam- order to reduce the noise of the images and the particle a cobalt oxide layer is growing. ples were characterized using the aberration to obtain clearer images the raw data was fil- One of the (110) facets of the nanoparticles (C ) corrected JEOL JEM-ARM200F tered using the 2D Wiener filter and the shows the 1 2 surface reconstruction corre- s (ARM) 200 kV FEG-STEM/TEM, equipped Richardson-Lucy / Maximum Entropy algo- sponding to a missing atom row. Cobalt atoms with a CEOS C corrector on the illumina- rithm implemented by Ishizuka. The EDS occupy the empty gold atom positions. This is s tion system. The probe correction was per- analysis was performed using EDAX instru- probably one the best examples of epitaxial formed through a dodecapole corrector mentation attached to the JEOL-ARM growth. (See page 5 ) (CEOS GmbH) aligned through the microscope. Spectra, line scans as well as chemical maps for the various elements were obtained using the EDAX Genesis software. One UTSA Circle, San Antonio, Texas 78249, USA For the EDS analysis the probe size used was 6C (145 pA) and the CL aperture size [email protected] was 40 m. One- and two-dimensional elec- Serving Advanced Technology

( 1) JEOL News Vol. 46 No.1 1(2011) (a) (b)

7.35°

Fig. 1 (a) Au decahedra oriented along the 5-fold axis of the particle, (b) geometry angular deficien- cy in the decahedra. For a decahedron, the angle between adjacent (111) plane is 7.35° on projection of (110) orientation.

tron energy loss spectroscopy (EELS) analysis was performed with a GIF-Tridiem spectrometer. The probe size used was 6C (145 pA) and the CL aperture size was 40 m (camera length – 6 cm).

Results and Discussion Monometallic nanoparticles

Nanoparticles are a fundamental part of nanotechnology. In fact, metal nanoparticles were used since the 1940's to improve the quality of gasoline. The surprisingly high catalytic activity of particles in the nano size range was recognized early and hence it was used extensively [1]. However, for many 2 nm years it was assumed that the only role of the particles was to provide a higher surface area. The recent explosion of interest in nanotechnology has made clear the fundamental Fig. 2 The smallest icosahedra observed so far in which the five-fold sym- physical principles behind the particle metry is apparent. behavior. The nanoparticles are intermediate between clusters (a few atoms) and bulk systems. In fact clusters are different from nanoparticles in the fact that most of the an example of a gold decahedral particle in Indeed many of the most useful nanoparticles atoms are on the surface. In a particle which the atomic columns are revealed by are bimetallic, one important problem is to although an important fraction of the atoms the HAADF-STEM image. A periodic strain determine its detailed structure and chemical are on the surface there is still a core of non- can be observed along the one of the twin composition. In our group since the last few surface atoms. boundaries and this is related to the fact that years we have been carrying out a very exten- The use of HAADF-STEM has opened up if we pack five fcc regular tetrahedra in a sive research program on bimetallic nanoparti- new and exciting possibilities to understand cyclic twinning (see Fig. 1b) a gap of 7° 35' cles. We have studied a number of systems nanoparticles. The atomically resolved images will be produced. The strain on the particle including Au/Pd, Au/Co, Au/Pt, Pd/Pt and in combination with EDS and EELS data pro- will be necessary to accommodate the gap. It Au/Ag among many others. Normally it is vide a very powerful set of tools capable of is possible from images of this kind, to assumed [2] that nanoparticles with two met- understanding the atomic behavior. measure the strain more acurrately. In Fig. 2, als will show a structure which is either core- One of the most common findings in we show one of the smallest icosahedra so shell or an alloy. In the first case, one metal nanoparticles is that the crystal structure is far observed in which the fivefold symmetry will be the core and the second metal will be different from that of the bulk. The most is apparent. the shell surrounding the core. However commonly observed case is the formation of experimental results from our group clearly the five fold symmetry structures such as the Bimetallic nanoparticles indicate [6, 7] that the structure is much more icosahedron and decahedron [2-4]. A very complex, than a simple alloy or core shell fundamental problem is to understand how The properties of metallic nanoparticles can structure. We will describe these details in the these structures become stable. Fig. 1 shows be improved by adding a second metal [5]. following sections.

JEOL News Vol. 46 No.1 2(2011) ( 2) Z contrast will be proportional to the atomic number Experimental examples of bimetallic and the contrast due to the atomic composi- particles When studying nanomaterials with two tion will be clearly observed. In order to different elements, the most interesting case illustrate this effect we have made calcula- When applied to real nanoparticles that have being those of transition and noble metals, tion of the intensity of columns of atoms of been obtained in our laboratory, the atomic con- the TEM techniques are somewhat limited different elements with the same number of trast is clearly observed. Fig. 4 (a) and (b) since the contrast is more related to coherent atoms. We have carried out multislice calcu- shows an Au/Pd nanoparticle grown by chemi- scattering. Thus one of the major limitations lations using the software developed by cal synthesis (6). The variation in the composi- is the diffraction contrast which will be easi- HREM Inc.[8] The results are shown in tion is clearly observed (Fig.4(b)), the central ly observed (if present). Bright field images Figs. 3(a) and (b). We have plotted a relative core is an alloy rich in Pd, then a second layer might give some information if there is strain contrast of a number of elements where the of alloy rich in gold and finally a exterior shell or change of orientation between the core Z contrast is apparent. On the other hand the which is an alloy rich on Pd. It is possible to and shell structure. However, STEM- intensity goes as I ~ Z 1.45. With the probe plot the chemical composition along the particle HAADF is the most important technique in corrected aberration a resolution of 0.8 Å can using EDS and EELS, the result of which is the study of bimetallic nanoparticles. The achieved therefore variations of the composi- shown in Fig. 5 and Fig.6. EDS line scan in the intensity of the signal which is generated tion at atomic scale can be observed. case of the three-layer Pd-Au-Pd nanoparticles

(a) (b)

Pd Cu S B

Ag Zn Fe C

Data: Data5_B Model: Allometric2 Equation: y=a+b* x ^c Weighting: y No weighting

Chi^2/DoF = 1.03877 R^2 = 0.99885

a 8.03367 ±0.67518 y = a + b* x ^c Pt Ge Co O b 0.15572 ±0.02232 a=8.03367 c 1.46189 ±0.03208 b=0.15572 c=1.46189

Au Mo Ni Si

Fig. 3 (a) Calculations of the relative HAADF-STEM contrast of different elements (columns with the same number of atoms). The atomic number contrast is very clear, (b) Variation of the intensity of the contrast with the atomic number.

(a) (b)

Spat ial Drift

A Z u r = ic 7 h 9 P la d y A Z r e = ic r Spectrum Image h 4 6 la y e r B

C 2 nm

Fig. 4 A truncated cuboctahedra three-layered Fig. 5 (a) and (b) Characterization of the Au and Pd elemental distribution across the nanoparticle by (Pd-Au-Pd) nanoparticle. STEM-EDS line-scanning technique across the individual three-layer nanoparticle. The Pd-L and the Au-L,M line scan signals can be clearly seen varying in intensity along the different regions of the nanoparticles. (a) Shows the analyzed area and the direction of analysis.

( 3) JEOL News Vol. 46 No.1 3(2011) reveal the following interesting features as can Shown in Fig. 7, is the case of the four layered to the two particles. be seen in Fig. 5. The Pd-L and the Au-L,M nanoparticles wherein it can be seen that near signals can be clearly traced across the region the surface there is an extra layer of atoms of Surface reconstruction in of the individual Pd and Au layers with the gold and this suggests a four layered structure. nanoparticles maximum intensity of the signals varying This is a different case from the previously between the respective layers (Fig. 5(b)). The mentioned three-layered nanoparticles and has One of the most interesting discoveries of corresponding EELS map of Pd reveals clearly not been reported till date. the surface science research back in the seven- the presence of Pd in the core and in the outer- A very interesting result is shown in Fig. 8, ties was that, the surfaces of crystals can have most shell of the tri-layered nanoparticles as is wherein two Au / Pd particles that have coa- a different atomic arrangement than those of shown in Fig. 6. These results clearly confirm lesced are forming an elongated structure. the bulk [9]. This phenomenon is known as the complex chemical nature of the bimetallic Most remarkably the 3 layer structure is pre- surface reconstruction and in the mid 80's particles. In addition it is possible even to dis- served by the coalescence. This means that Marks et al. [10] showed that indeed surface tinguish layer by layer, when it is Pd rich or Au during the initial coalescence process the Pd reconstruction is also observed in gold rich as shown in Fig. 7. Thus it is worthy to atoms diffuse along the boundary. However nanoparticles. In our group we have been mention that in case of these nanoparticles many bright spots, corresponding to gold investigating the conditions in which surface occasionally four-layered structure are seen. atoms can be observed on the region common reconstruction appears. We have observed inter-

Spatial Drift Spectrum image

Fig. 6 EELS map of Pd reveals clearly the presence of Pd in the core and in the outermost shell of the tri-layered nanoparticles.

D C

B 2 nm 2 nm A

Fig. 7 Aberration-corrected STEM images of the four layered Au/Pd Fig. 8 Coalescence of two / three-layered Pd-Au-Pd nanoparticles. nanoparticles. The contrast of the four distinct regions Notice also the Au intercalation into the Pd layer. (marked A, B, C and D) can be clearly seen.

JEOL News Vol. 46 No.1 4(2011) ( 4) faces of Au with Co. Shown in Fig. 9 (a) is from our group over the last year with the new 2G12RR013646-11 from the National Center the <110> surface of a Au / Co nanoparticle. ARM microscope. It is clear from the discussion for Research Resources. The content is solely The bright atoms correspond to gold and the provided, on the results that have been obtained the responsibility of the authors and does not light atoms correspond to Co. It can be concerning metal nanoparticles/ bimetallic and necessarily represent the official views of the observed that gold shows a (1 2) reconstruc- core-shell nanoparticles and other related sys- National Center for Research Resources of the tion. A model of the reconstructed surface in the tems, that it has indeed been an exciting one year. National Institutes of Health. Finally, the missing row is shown in Fig. 9(b) (compare the With unprecedented advantages in terms of reso- authors would like to thank Dr. Eduardo image Fig. 9a and the model Fig. 9b). The inter- lution and using the HAADF-STEM imaging in Pérez-Tijerina from UANL for providing of esting result is that 10 atoms accommodate on combination with additional spectroscopic tools some of the samples included in this paper. the places of the missing gold atoms. This kind associated with it, namely EDS and EELS it has of analysis suggests ways in which surfaces can been possible to unravel some of the fundamental References be tailored for improved catalytic activity. questions in the case of nanoparticles. Indeed with all these new developments, we can look [1] G. Somorjai, Chem. Rev., 96, 1223 (1996). Observation of individual forward to many such more exciting areas of [2] A. L. Mackay, Acta Crystallogr., 15, 916 atoms/clusters research in the coming years. (1962). [3] J. L. Rodriguez-Lopez, J. M. Montejano- It is indeed interesting to note that the ultimate Acknowledgements Carrizales, M. Jose-Yacaman, Modern resolution of observing individual atoms / small Physics Letters, 20, 725 (2006). clusters have been achieved using the ARM. The authors would like to acknowledge [4] L. D. Marks, Reports on Progress in Shown in Fig. 10, are the very small clusters of THE WELCH FOUNDATION AGENCY Physics, 57, 603 (1994). Pd-Ir (inside the circle). The size of the cluster is PROJECT # AX-1615. "Controlling the [5] R. Ferrando, J. Jellinek, R. L. Johnston, just about ~ 2 nm. It is interesting to note that Shape and Particles Using Wet Chemistry Chem. Rev., 108, 845 (2008). based on the Z contrast, the two different atoms Methods and Its Application to Synthesis of [6] D. Ferrer, D. A. Blom, L. F. Allard, J. Mat. can be distinguished very clearly (Pd, Z = 46, Ir Hollow Bimetallic Nanostructures". The Chem., 18, 2442 (2008). = 77). In addition in the vicinity of these small authors would also like to acknowledge the [7] D. Ferrer, A. Torres-Castro, X. Gao et al. clusters it is also possible to locate individual NSF PREM Grant # DMR 0934218, Title: Nano Lett., 7, 1701, 2007. atoms of Ir / Pd (outside the circle). Oxide and Metal Nanoparticles - The [8] K. Ishizuka, N. Uyeda, Acta Cryst. A, 33, Interface between life sciences and physical 740 (1977). Conclusions sciences. The authors would also like to [9] H. Wagner. Springer Tracts in Modern acknowledge RCMI Center for Interdisciplinary Physics (Springer, Berlin), 85, (1979). This paper summarizes the most important Health Research CIHR. The project described [10] L. D. Marks, Phys. Rev. Lett., 51, 1000 results on noble metal nanoparticles obtained was supported by Award Number (1983).

(a) (b)

_ [110] _ [220]

[001]

(110-1x2 Surface) (100)Facet (111)Facet

Fig. 9 (a) Image of surface reconstruction observed in Au-Co nanoparticles and (b) the corresponding model showing the 12 reconstruction. 2 nm

Pd Pd Ir

Fig. 10 Clusters of Pd-Ir are shown (inside the circle). It is possible to distinguish each element based on the Z contrast Ir (Pd = 46, Ir = 77). Notice also the presence of individual atoms (outside and close to the circle).

( 5) JEOL News Vol. 46 No.1 3(2011) Exploring Biological Samples in 3D Beyond Classic Electron Tomography

Marcel Cunha, Cédric Messaoudi and Sergio Marco

Institut Curie, Centre de Recherche and INSERM

Developments of electron tomography methods are of great importance to unravel the tridimensional organization and the interior detail of objects of nanometric scale. Here we describe two recently developed approaches to surpass some of the milestones of contemporary electron tomography: The low signal to noise ratio of tomograms and the detection of chemical elements in 3D. The techniques applied rely on STEM tomography and EFTEM tomography, capabilities of the 200kV electron microscope JEOL JEM-2200FS equipped with an in-column energy filter and STEM detectors. In the resin embedded samples tested, STEM-HAADF tomography was applied with success to improve the signal to noise ratio and determine cytoskeletal structures on mammalian kidney cells (LLC-PK1). For elemental detection, EFTEM tomography revealed the precise location and 3D arrangement of iron particles bound to cell wall of the pathogenic fungus Fonsecaea pedrosoi. Both techniques resulted in datasets clear enough for direct interpretation and threshold- based segmentation.

Introduction and organelles leading to a better under- ing wave generated by a tomograph which standing of life processes [5]. Thus, meth- also records the projected images [6]. To Imaging techniques of tridimensional ods to obtain 3D information of objects are study samples at nanometric scale, the reconstructions provide details of the interi- nowadays indispensable for large aspects of transmission electron microscope (TEM) is or of reconstructed objects otherwise inac- modern science. presently the most used tomograph cessible. Such methods allow the quantifi- 3D reconstruction methods are usually although other alternatives are slowly being cation of volumes and might give access to based on the stacking of images obtained applied to life sciences, such as those based the information of the different elements from sections of the studied object or on the on synchrotron radiation [7]. Volume constituting the object. They are used in combination of data from its projections. reconstructions methods based on the com- many fields such as archeology, geo- The electron microscopy techniques avail- bination of TEM images acquired from a physics, material sciences, quality control able nowadays which use the first approach single specimen tilted at different angles are and biology. In medicine, these methods are destructive to the specimen, since they called transmission electron tomography represented by X-ray tomography [1], sin- are based on the progressive slicing by (TET) [8]. The automation of image acqui- gle photon emission computed tomography knives or by sample erosion by an ion sition using TEMs started a quarter of cen- [2], magnetic resonance imaging [3] or beam. The second approach, used for elec- tury ago and was perfected more recently positron emission tomography [4] are regu- tron tomography, is based on the analysis of with the use of fast computers and sensitive larly used for exploration and diagnosis. In the projections of a thick specimen (from and larger digital cameras sensors. This cell biology, volume reconstruction meth- 100 to 1000 nanometers), which is pre- enabled TET as a routine in a reasonable ods provide a more accurate vision of the served and could be used for successive and number of well-equipped electron relationships between cellular structures complementary analyses. This is one reason microscopy laboratories. In biology, TET that makes approaches based on reconstruc- has recently been combined with cryo-fixa- U759 Orsay, F-91405 France tion techniques from projections largely tion methods [9,10] consisting in the freez- favored. These methods require image for- ing of samples at speed, pressure and tem- [email protected] mation by the use of any kind of penetrat- perature conditions to reach a quality of ice

JEOL News Vol. 46 No.1 6(2011) ( 6) known as the vitreous state of the water. and the relation of its tridimensional struc- cal samples (Figure 1A,B). However, in Since cells are mainly composed by water, ture with nearby organelles or associated most of 2D studies performed in stained its non-crystaline vitreous state formation proteins. biological samples, the SNR improvement by cryo-fixation allows a preservation of Cellular structures are not only related to does not justify the use of STEM. For cellular structures in a close-to-native state. the cellular constitution and maintenance instance, in our data set, different compo- Cryo-fixation can then be followed by a but also to interaction with other cells, from nents of the cytoskeleton such as micro- replacement of water for resins (cryo-sub- the same or other organism. The interaction tubules or actin could be identified in both stitution) or by direct observation of the between different organisms may result in cases. In addition, for an equivalent surface, sample into the electron microscope at liq- growth of one in detriment of the other the acquisition time required to record uid nitrogen or lower temperatures (cryo- designing infection. The virulence of patho- STEM images is larger than that required in electron microscopy and cryo-electron genic fungi depends on several factors, and standard TEM mode. Nevertheless, the tomography when combined with electron an important one is the presence of a pig- SNR enhancement that can be achieved by tomography). Cryo-electron tomography ment named melanin, mostly located on the STEM-HAADF is useful to improve 3D (CET) [11] is one of the most informative fungal cell wall, providing strength and reconstructions computed by electron 3D approaches in structural cell biology, shape to the cell and related to resistance of tomography because it improves image despite its limitations on sample size and antifungal treatments [15,16]. Since alignment (without fiducial markers) and resolution, which is approximately 5 nm for melanin interacts with iron forms, its spatial leads to less noisy reconstructions allowing most used TET conditions, is currently localization can be tackled by studying the a reliable and reproductive volume interpre- becoming more and more popular among 3D distribution of iron exogenously added tation. biologists, but it is still far from being wide- via cationized ferritin. This requires the The setup of STEM-HAADF tomography ly used in biology labs. computation of 3D maps of iron by EFTEM is becoming more user friendly with the In parallel to the development of CET, tomography (EFTET). new suites of software for the handling of other electron tomographic approaches, the microscope as well as for the steps of valid for both resin embedded biological calibration of acquisition parameters and specimens and material sciences samples, Experimental, Results and focusing. STEM images suppose that all the are in progress. For this, the two main chal- Discussion, etc. points of the sample are placed at the same lenges are: (I) to increase the quality (reso- focal plane during the scanning. This condi- lution and signal-to-noise ratio) of the 3D tion does not occur during tomographic reconstructions; and (II) to get 3D chemical Data acquisition and image acquisition due to sample tilting, which distribution of elements. The first is consid- analysis implies a significant focus difference erably important to define small cellular between the upper and the lower region of structures like cytoskeletal filaments and Images were acquired on a JEOL JEM- the specimen in the perpendicular direction other protein-made supramolecular struc- 2200FS equipped with an in-column energy to the tilt axis. These focus variations tures and its subunits without the need to filter and a ssCCD 2k 2k Gatan camera. should be dynamically corrected, line by isolate such from the cellular environment. For standard TET, Z-loss tilted series were line, by the acquisition software [21]. The The latter is becoming increasingly interest- acquired using an energy window of 10 eV. dynamic correction of focus, area by area, ing for biologist to help to unravel the role In the case of STEM-HAADF tomography results in STEM-HAADF tomographic tilt of elements (e.g. Iron, phosphorous) related images were recorded via JEOL's series in which all the points in all the to its location in the cell and its association TEMography software suite for STEM. For images have been acquired in focus. This, with other cellular structures or organelles. EFTEM tomography tilt series at 0, 20, together with SNR improvement, leads to The improvement of the TET approaches 560, 590, 620, 650, 680 and 710 eV were final reconstructions in which the stained in resin embedded samples goes presently acquired also with JEOL's acquisition soft- biological objects are more contrasted in through its combination with Scanning ware suite. Image alignment was performed STEM-HAADF tomography than in stan- Transmission Electron Microscopy in its using Etomo [17,18] and TomoJ [19,20] dard TET, as they appears as white objects different modes (STEM-BF, ADF or depending on the addition or not of fiducial against an uniform dark background (Figure HAADF) [12] allowing the enhancement of markers (5nm gold beads). 3D reconstruc- 1C,D). Therefore, the reconstructed vol- the signal-to-noise ratio of the projections; tions were performed in Etomo or TomoJ umes are suitable for objective segmenta- and also through its combination with using WBP or ART algorithms respectively tion methods such as thresholding [22] Energy filtered TEM (EFTEM) to go fur- both using a GPU accelerated implementa- which, instead of relying on the subjective ther than the structural information by tion. For EFTET, background subtraction manual tracing of contour lines, are based obtaining 3D chemical maps. STEM and and image reconstruction were performed in the assignation of colors or grey levels as EFTEM are frequently used in material sci- after multivariate analysis using an Image J function of the voxel values (Figure 2). ences [13] but they are less common in the plug-in developed by our group (unpub- study of biological specimens. In this man- lished software). uscript, we illustrate the potential that com- EFTEM tomography bining these two imaging modes with elec- Comparison between stan- tron tomography has in biology by two Fonsecaea pedrosoi cells from 5-day old examples consisting in the visualization of dard TET and STEM-HAADF cultures were filtered in a 40–60G porous cytoskeleton elements in mammalian cells tomography plate filter followed by centrifugation and the iron localization around a fungal (13,600g, 30 min, 4°C) to isolate the coni- cell wall. Brush border microvilli cells derived dial forms. Fungal conidia were fixed with The cytoskeleton [14] is not only the cel- from proximal tubule of porcine kidney 2.5% glutaraldehyde in PBS, exhaustively lular scaffold, but also plays essential roles epithelium (LLC-PK1) were used for visu- rinsed with PBS, and incubated in the pres- in a series of cellular phenomena like cellu- alization of cytoskeleton elements by ence of 10 g/mL cationized ferritin, at pH lar division, cell growth, movement and STEM and STET. A standard procedure 7.2, for 1 h at room temperature. Cells were intracellular transport of organelles. In consisting to fix samples with phalloidin postfixed for 30 min in 1% osmium tetrox- eukaryotic cells, the cytoskeleton contains and aldehydes, post-fix with tannic acid and ide, dehydrated in acetone, and embedded three main kinds of filaments: (I) actin fila- osmium (TOTO), dehydrate and embedding in Epon [23]. 150 nm thick sections were ments, constituted by double helix of 7 nm in epon resin was applied before obtaining used for observations under the electron of diameter; (II) intermediate filaments, 100 nm thick sections. microscope by TET and in EFTET. having about 10 nm of diameter; and (III) Since contrast in STEM-HAADF is The main advantage of EFTET is that it microtubules, which are cylinders of about directly related to the atomic number of the complements the structural information 25 nm diameter. A high number of contem- elements present in the sample (Z-contrast obtained by standard TET by providing the porary questions regarding the above cited image), this imaging mode is able to pro- 3D localization of studied chemical ele- phenomena are related to the precise deter- duce data presenting higher signal-to-noise ments. Therefore the process today is mination of cytoskeleton structures in cell ratio than standard TEM in stained biologi- almost fully automatized, one drawback in

( 7) JEOL News Vol. 46 No.1 7(2011) Fig. 1 Comparison between TET and STEM- HAADF tomography. A) 0° tilt image TEM; B) 0° tilt image STEM in a similar region of the same sample from the cell shown in (A). C) Virtual section extracted from TET tomogram computed from tilt series corresponding to (A). D) Virtual section extracted from STEM-HAADF tomogram computed from the tilt series corresponding to (B). E) Virtual cropped sections of areas of interest from (A). F) Virtual cropped sections of areas of interest from (B). N=nucleus, tw= terminal web, mv=microvilli. White arrow=actin filament; white arrowhead=microtubules; black arrows=intermediate filaments. Scale bars: A-D 500 nm, E-F 200 nm.

Fig. 2 Anaglyph from threshold segmentation TET and STET. White arrow=actin filament; white arrowhead=microtubules; black arrows=intermediate filaments. Scale bars 200 nm. Require red-cyan anaglyph glasses.

JEOL News Vol. 46 No.1 8(2011) ( 8) EFTET is that acquisition is time consum- ferritin aggregates which are associated Arpin from the Institut Curie in Paris ing due to the need of a high number of with melanin pigments. However, the com- (France) for providing LLC-PK1 cells. Dr. images to be acquired at different energy putation of the 3D electron map after the Sonia Rozental from the Universidade loss values for each tilt angle (Figure 3). In use of MVA (Figure 4) demonstrates that Federal do Rio de Janeiro (Brazil) for the addition, this may result in a high electron not every electron dense area correspond to kind donation of fungal samples. The dose incident on the sample. Objects too iron. This map was validated by the quan- “Fondation pour la Recherche Médicale” sensitive to radiation might not be ideal for tification of the signal measured at different for the fellowship of MC, and “Region Ile EFTET studies, although working at liquid regions of the sample. de France” for its contribution on the pur- nitrogen or lower temperature may reduce chase of the JEOL JEM-2200FS micro- the negative effects of the radiation damage scope used for every experiment reported in [24]. In the case of EFTET, one important Conclusions this manuscript. procedure is computation of the reconstruction, since image alignment must be very This study has shown that STEM- precise, what can be increasingly difficult HAADF tomography is suitable and should due to the low signal-to-noise ratio fre- be encourage for the study of small, low quently occurring in data from biological contrasted, resin embedded biological sam- samples. An additional improvement to ples since provide an improvement of the achieve a good EFTEM tomogram is asso- SNR on reconstructed volumes. This results ciated to the prerequisite of unspecific in the observation of delimited borders References background subtraction which should be against a near-homogenous background on based on robust algorithms because it can the 3D reconstructions, facilitating the [1] Seeram E., Computed Tomography: lead to the overestimation of characteristic interpretation of the images or allowing Physical Principles, Clinical signal of the studied element. The use of them to be used for automatic or semi-auto- Applications, and Quality Control different law models for background sub- matic segmentation. Regarding EFTEM (CONTEMPORARY IMAGING traction and of multivariate statistical analy- tomography, it adds the important informa- TECHNIQUES), 3rd edition, sis (MVA) can prevent for this overestimation of the 3D localization of defined chem- Sauders Elsevier, (2008). tion allowing the computation of accurate ical elements which complements the struc- [2] Wernick M.N. and Aarsvold J.N., maps [25]. Despite these obstacles for the tural data obtained by standard TET. Both, Emission Tomography: The Fundamentals inexperienced user, EFTET should be STEM-HAADF and EFTEM tomographies of PET and SPECT, Elsevier Academic encouraged since it allows the study of the are techniques that will be more applied in Press, (2004). distribution of specific chemical elements future researches in cell biology and micro- [3] Liang Z. & Laut erbur P.C., Lauterbur., in the sample volume, and its use become biology and they are also an important basis Principles of Magnetic Resonance justified to prevent incorrect interpretations for approaches that are currently in devel- Imaging: A Signal Processing only based on contrast. Once 3D chemical opment which allow the specific labeling of Perspective, IEEE Press Sereies in bio- maps are correctly computed, the returned proteins, by specific heavy metal associated medical Engineering, (2000). information gives access to precise elemen- clonable tags [27]. [4] Christian P. E. Christian & Kristen M. tal positioning in a sample [26]. Waterstram-Rich, Nuclear Medicine and In our samples, as the TEM image shown PET/CT Technology and Techniques, in Figure 3, different dark particles can be Acknowledgements Springer, (2007). observed associated to the fungal cell wall. [5] Bárcena M. Koster A.J. Electron These particles are usually interpreted as Authors want to thanks Dr. Monique tomography in life science. Semin Cell

Fig. 3 Untilted images extracted from F. pedrosoi tomographic series. A) 0° tilt-images recorded at different energies values between 0 and 710 eV. C= cytoplasm, CW= cell wall. Scale bars 400 nm.

( 9) JEOL News Vol. 46 No.1 9(2011) Dev Biol. 20(8), 920-930 (2009). [14] Frixione E. Recurring views on the 110(9), 1231-1237 (2010). [6] Herman, G.T., Fundamentals of com- structure and function of the cytoskele- [22] Sezgin M. and Sankur B. Survey over puterized tomography: Image recon- ton: a 300-year epic. Cell Motil image thresholding techniques and struction from projection, 2nd edition, Cytoskeleton. 46(2), 73-94 (2000). quantitative performance evaluation, Springer, (2009). [15] Ikeda R., Sugita T., Jacobson E.S. Journal of Electronic Imaging. 13(1), [7] Larabell C.A. Nugent K.A. Imaging Shinoda T. Effects of melanin upon 146–165 (2004). cellular architecture with X-rays. Curr susceptibility of Cryptococcus to anti- [23] Cunha M.M., Franzen A.J., Alviano Opin Struct Biol. 20(5), 623-31 (2010). fungals. Microbiol Immunol. 47(4), D.S., Zanardi E., Alviano C.S., De [8] Frank J., Electron Tomography: 271-277 (2003). Souza W. and Rozental S. Inhibition of Methods for Three-Dimensional [16] Franzen A.J., Cunha M.M., Miranda melanin synthesis pathway by tricycla- Visualization of Structures in the Cell, K., Hentschel J., Plattner H., da Silva zole increases susceptibility of Springer, (2006). M.B., Salgado C.G., de Souza W. Fonsecaea pedrosoi against mouse [9] Vanhecke D., Graber W. Studer D. Rozental S. Ultrastructural characteri- macrophages. Microsc Res Tech. 15, Close-to-native ultrastructural preser- zation of melanosomes of the human 68(6), 377-84 (2005). vation by high pressure freezing. pathogenic fungus Fonsecaea [24] Aronova M.A., Sousa A.A., Zhang G. Methods Cell Biol. 88, 151-64 (2008). pedrosoi. J Struct Biol. 162(1), 75-84 and Leapman R.D.. Limitations of [10] Dubochet J., Adrian M., Chang J. J., (2008). beam damage in electron spectroscopic Homo J. C., Lepault J. McDowall A. W., [17] Kremer J.R., Mastronarde D.N., tomography of embedded cells. J Schultz P. Cryo-electron microscopy of McIntosh J.R. Computer visualization Microsc. 239(3), 223-232 (2010). vitrified specimens. Q Rev Biophys. of three-dimensional image data using [25] Quintana C., Marco S., Bonnet N., 21(2), 129-228 (1988). IMOD. J Struct Biol. 116(1), 71-6 Risco C., Gutiérrez M.L., Guerrero A. [11] Mader A., Elad N. and Medalia O. (1996). and Carrascosa J.L.. Optimization of Cryoelectron tomography of eukaryot- [18] http://bio3d.colorado.edu/imod/ phosphorus localization by EFTEM of ic cells. Methods Enzymol. 483, 245-65 [19] Messaoudi C., Boudier T., Sanchez nucleic acid containing structures. (2010). Sorzano C.O. and Marco S. TomoJ: Micron. 29(4), 297-307 (1998). [12] Friedrich H., McCartney M.R. Buseck tomography software for three-dimension- [26] Weyland M., Yatesa T.J.V., Dunin- P.R. Comparison of intensity distribu- al reconstruction in transmission electron Borkowskia R.E., Laffonta L. and tions in tomograms from BF TEM, microscopy. BMC Bioinformatics. 8, 288 Midgley P.A., Nanoscale analysis of ADF STEM, HAADF STEM, and cal- (2007). three-dimensional structures by elec- culated tilt series. Ultramicroscopy [20] http://u759.curie.u-psud.fr/software- tron tomography. Scripta Materialia, 106(1), 18-27 (2005). su759.html 55(1), 29-33 (2006). [13] Midgley P.A. Weyland M. 3D electron [21] Biskupek J., Leschner J., Walther P. and [27] Mercogliano C.P. DeRosier D.J.. microscopy in the physical sciences: Kaiser U. Optimization of STEM Concatenated metallothionein as a the development of Z-contrast and tomography acquisition - a comparison clonable gold label for electron EFTEM tomography. Ultramicroscopy of convergent beam and parallel beam microscopy. J Struct Biol. 160(1), 70- 96(3-4), 413-31 (2003). STEM tomography. Ultramicroscopy 82 (2007).

Fig. 4 Iron localization in F. pedrosoi by EFTEM tomography after addition of ferritin for melanin detection. A) Sections of the reconstructed iron map after MVA filtering extracted each 16 nm. Scale bar 500 nm. B) Anaglyph of threshold based segmented iron map. Require red-cyan anaglyph glasses.

JEOL News Vol. 46 No.1 10(2011) (10) Application of Scanning Electron Microscope to Dislocation Imaging in Steel

Masaaki Sugiyama† and Masateru Shibata††

† Advanced Technology Research Laboratories, Nippon Steel Corporation †† SM Business Unit, JEOL Ltd.

Dislocation imaging using the scanning electron microscope with super hybrid lens will be demonstrated to investigate the dislocation cell walls and single dislocations inside cells introduced by shear deformation in conventional steel. The resolution of the dislocation by electron channeling contrast imaging method is similar to that obtained by conventional TEM observation, and a new approach for the study of dislocations which is possible to detect the heterogeneity of the deformation microstructure will be expected on the view point of the advantage of SEM-BSE techniques. There are two different imaging configurations for doing ECCI, one is a fore-scatter geometry, and the other is a back-scatter one. In the present study, the latter case is utilized, which has advantages for little restriction to the sample size and shape and several applications to the stage design in SEM. Since there are some discussions in the contrast mechanism, the improvement of the backscattering electron detector will bring us several ideas for the application of microscopy to the dislocation study with a combination of the conventional TEM technique.

Introduction their mechanical properties must be consid- attempt to obtain dislocation imaging using ered using a more macroscopic view, taking SEM with and without a field emission gun, With the increase of the use of high- in the picture of the whole material, includ- along with the imaging and contrast simula- strength steel, the formability control of ing the effect of grain boundary, the crystal- tion of the clusters for misfit dislocations at steel materials with complex phases lography of grains, and other hetero- the interface, were demonstrated, for exam- becomes important, resulting in the require- geneities of microstructure. The TEM thin- ple, in strained Si-Ga layers on silicon [1]. ment for the understanding of the local foil analysis of dislocations has some limi- Taking into account the effects of surface defect structure, such as dislocations and tations regarding these multi-scale forma- stress relaxation, back-scattered electron several boundaries based on both experi- tion and deformation aspects, due to the intensity within electron channeling con- mental and computational aspects. The thin-foil effect and some difficulty in the trast images of dislocations has been also characterization of dislocations and other preparation of TEM thin foils themselves, calculated using a two-beam dynamical dif- defects has been mainly studied using trans- including the heterogeneous large area. fraction model and the simple treatment of mission electron microscopy (TEM) utiliz- Recently, in order to examine the disloca- multiple scattering [2]. The imaging of ing diffraction imaging contrast and this has tion structure over a larger area of the mate- screw dislocations near the surface of Si, brought us a great success in understanding rials in question, the electron channeling Ni, and Ga thin films using the ECCI local dislocation structure and further sever- contrast imaging (ECCI) technique using method have been also studied in the al interaction models among dislocations scanning electron microscopy (SEM) has 1990’s, in which the g • b=0 criterion for and other defects such as solute atoms, been discussed in various papers. The theo- the dislocation image has been discussed impurities, clusters, and precipitation. On ry underlying channeling contrast, of [3]. Based on these results, it was estab- the other hand, most structure materials, course, goes back to the beginning of TEM lished that the intensity of back-scattered such as steels, consist of polycrystals, and contrast interpretation in the 1960’s, and it electrons is strongly dependent on the ori- is considered that the defect regions in dis- entation of the incident beam with respect 20-1 Shintomi, Futtsu, Chiba 293-8511, JAPAN locations should give rise to a modulation to the crystal lattice. Thus, it was pointed of the back-scattered electron (BSE) inten- out that the information regarding the crys- [email protected] sity. In the 1990’s, several examples in tal orientation and its diffraction conditions

(11) JEOL News Vol. 46 No.1 11(2011) observed was necessary to understand the field for the deformed microstructure of fraction vector, Burgers vector, and deviation dislocation structure using the ECCI structure materials. There are two ways parameter on channeling contrast are found method, and the electron channeling pat- progress on the ECCI techniques with fore- to be markedly similar to the corresponding terns (ECP’s) and electron backscatter dif- scatter and back-scatter geometry measure- parameters controlling TEM diffraction con- fraction (EBSD) method were extensively ments. In comparison with data obtained by trast [10]. investigated at the same time [4]. TEM and SEM, the contrast change of edge With the progress on the basic approach for In the 1990’s, on the other hand, the dislo- dislocations in FeAl alloy has been investi- the mechanism of dislocation imaging, the cation images obtained using the ECCI gated under conditions of g • b=0 and g • b applying of SEM instead of TEM will be method did not provide sufficient contrast u=0 criteria in 2001 [8]. However, the expected to observe dislocations, because of regarding the resolution, which was neces- image formation process in ECCI was not a lot of advantages on the flexibility for the sary in order to investigate the dislocation so simple, and the overall image quality for observation condition. In order to overcome structure for the deformed steel materials. As ECCI was not as high as that for TEM. the problems of the actual application of the a result, the dislocation cell structure com- The strong and clear contrast for single dis- ECCI method in the field of conventional posing of a set of dislocations introduced locations by the ECCI method must be materials, it is valuable to improve the detec- under fatigued experiments has been mainly obtained. In 2006, dislocation images tor of the back-scattered electron signals and investigated regarding the application of the observed for a crack tips and edges under other instrumental conditions. In the present ECCI method [5]. The study of dislocation low tilting conditions have been reported in paper, it is demonstrated that single disloca- structure using the ECCI method has not deformed NiAl single crystal using a tions in deformed and non-deformed steels been achieved for use as a conventional CamScan 44 FE SEM [9]. In contrast to have been observed with strong contrast and method when compared to TEM for the rou- TEM, electron channeling imaging of dislo- in a wide range of areas using a new detector tine imaging and analysis of dislocations. cations is optimized with s=0, and the image for back-scattered electrons made by JEOL However, with the progress of instrumenta- contrast falls off rapidly with both positive Ltd. tion in the SEM fields in around the 2000’s, and negative deviations from the perfect the imaging of individual dislocations has Bragg’s condition. The dislocation contrast been gradually achieved over a wide range of using ECCI is very sensitive for the Bragg’s ECCI Experimental conditions using the ECCI method utilizing a condition, resulting in the misunderstanding Configuration standard commercially available SEM, using for distribution of dislocations, in compari- a straightforward experimental configuration, son with the thin foil observation under The observation of dislocations has been in which the sample is tilted with high angles TEM. In order to consider the simple chan- carried out using a JEOL JSM-7001F with such as 60 or 70 degrees [6,7]. neling condition, the fore-scatter geometry super hybrid lens and a Schottky thermal In high tilting sample geometry using for the ECCI method is now superior that the field emission gun operated at 0.1-25keV. fore-scatter electrons, it is easy to compare back-scatter one. The careful study using sin- A new detector such as an Si detector for crystal orientation data, such as EBSD, gle crystals of 4H-SiC [10] and GaN [11] back-scattered electrons has been mounted because of the same sample stage geometry, have been carried out using the fore-scatter on the pole piece of the objective lens, as however, it is not so convenient for in situ geometry detector, and the experimental dif- illustrated in Figure 1. Since the intensity observation, such as in a deformation fraction parameters and contrast features are of the back-scattered electrons from the sur- experiment, as well as for large bulk sam- determined for the threading screw disloca- face in this configuration is smaller than ples after being deformed, in which experi- tions. It is concluded from the study of 4H- that in the high tilting condition (60-70°), ments have been expected in the research SiC that the combined influences of the dif- the large collection angle is preferable to

JEOL News Vol. 46 No.1 12(2011) (12) obtain a good contrast of ECCI. The new of sample. TEM analysis. On the other hand, the per- detector is possible to set a condition of pendicular cell walls appear close to the short working distance (WD) by 2mm, the Dislocation Cell Walls and triple junction, as seen in the center part of observation conditions are estimated under the photograph. It is considered that the other several conditions, as indicated on the table Single Dislocations Using slip system on {110} planes was activated in Figure 1. The maximum detection solid the ECCI Method under the deformation. This is an example angle () is obtained by 2.7str. under the showing the heterogeneity of the deforma- working distance of 3mm. The present tion, and in comparison with the grain orien- experiment for the ECCI technique was car- In metals, dislocations have been intro- tation relationship and each dislocation cell ried out using short working distances from duced during deformation as a result of acti- walls in the large area using ECCI method, it 3 to 4mm. vated process of slip system in relation to the will be possible to get the information about When increasing at a large tilting angle for crystal structure. The most of steels used for the heterogeneity of the microstructure the sample, it is possible to expand the work- the structure materials such as automobiles, developed under deformation. ing distance than 5mm, similar to the con- ships, and buildings etc. are composed of the Since the electron diffraction pattern is ventional SEM. The EBSD for the orienta- ferrite phase with a body centered cubic crys- not obtained in the conventional SEM, the tion measurement and the EDS for the com- tal structure, in which screw dislocations are crystal orientation information using the position analysis were carried out using both mainly introduced during the deformation at EBSD measurement was investigated. detectors equipped on the SEM. The other room temperature. The burgers vector of a Figure 3 is a series of SEM-BSE micro- advantage for the design of super hybrid lens screw dislocation is <111> direction and the graphs (a,b) and the corresponding EBSD is to achieve the minimum external magnetic slip planes are parallel to {110} plane in orientation maps (c,d). The black right and electrostatic field on the sample position BCC structure. upper part of Fig.3 (a) is a square vacant under the observation, resulting in the easy The most excellent elongation behavior in region used as a marking, which was alignment of the stigmatism and focusing the thin plates is of -fiber plates with the formed by focused ion beam fabrication. even for strong magnetic materials such as {111} texture in the BCC ferrite phase, The normal direction of the center grain steel. which are mainly applied as the panel plates marked by yellow square is confirmed to be The steel used in this study consists of of the automobile body. When the simple <111> direction, judging from the ND conventional low carbon steel with and shear deformation test has been carried out direction mapping in Fig.3(c). For above without a crystal orientation texture. In for the model bulk plate with {111} texture, the photograph (a) and (b), the direction comparison with the dislocation microstruc- it is known that a characteristic directional was determined to be parallel to <10 -1> ture, both observation using TEM and SEM dislocation cell walls parallel to {110} direction by the EBSD mapping in Fig.3(d), were carried out for the electro polished planes have been formed with an increasing Based on the analysis, the crystal orienta- samples with 3mm in diameter, using the of the shear deformation [12]. Figure 2 is a tion information in the center grain is deter- SEM stage for a thin foil TEM sample. The SEM-BSE micrograph showing dislocation mined as indicated in Fig.3 (b). As clearly surface of a bulk steel plate was also electro cell walls introduced after 60% simple seen in Fig.3 (b), the formation of two polished to observe the dislocation shear deformation to the ferrite steel with directional traces of dislocation cell walls microstructure by the ECCI method, after {111} texture orientation. The magnifica- with (01-1) and (-110) plane are confirmed, mechanical polishing. It is not necessary to tion observed was 2,000 times, and the which are reasonable planes predicted from take account too much for surface condi- accelerated voltage was 25kV and WD was the activated principle slip systems in BCC tions of steel, that is, a slight surface oxide equal to 4mm. The dislocation cell walls ferrite steel. film is not affected to the contrast of ECCI, are clearly seen on the whole grains, and Using the present SEM-BSE system, mor- which is a large advantage against the low the morphology of them is one directional phology of single dislocations between dislo- voltage observation less than 1keV of the lamellar structure, which is correspondence cation cell walls is clearly distinguished thin film surface microstructure on the top to the observation result in the thin foil using the ECCI method with more high mag-

Incident electron

Objective Lens Back Scattered 2 Electron Detector WD 1

Specimen

Acc WD 1 2 12 (kV) (mm) (°)(°)(°) (str) 10 ~ 25 2 80 64161.7 10 ~ 25 2.5 72 45 27 2.5 10 ~ 25 3 64 30342.7 10 ~ 25 4 51 20312.0

Fig. 1 Schematic diagram showing geometry of specimen and back scat- Fig. 2 SEM-BSE micrograph showing dislocation cell walls introduced to tered electron detector in FE-SEM with super hybrid lens. In the each crystal grain after 60% simple shear deformation in steel. A cen- table, accelerating voltages (Acc), working distance (WD), take- ter part of the micrograph is a triple junction of grain boundaries. off angles, and detection solid angles are also listed.

(13) JEOL News Vol. 46 No.1 13(2011) nification in the same sample, as seen in deeper region in the crystal and the characteristic trast. It was confirmed by TEM observation Fig.4. The magnification of the photographs; contrast of a single dislocation will be white. using the same sample that the morphology (a) is 30,000 times and (b) is 50,000 times With enlarging the figure, the single disloca- and density of dislocations and dislocation under conditions that WD is 4mm and the tion and its tangled morphology are remark- cell walls are almost similar between the data accelating voltage is 25keV. In Fig.4 (a), the ably observed in Fig.4 (b). The top of a of SEM and TEM. center area between the dislocation cell walls screw dislocation on the sample surface is The ECCI contrast of the single dislocation is satisfied with a strong channeling condi- represented by a pair contrast of white and is very sensitive to the orientation change of tion, resulting in a black contrast of the black, and the dislocation line expanding to the sample. A series of SEM-BSE photo- matrix. The incident electrons proceed to the the bulk is a white line with a gradient con- graphs with stage tilting conditions of 0,1,2,3

(a)

001 101

ND Direction (b) - <101> (d) RD Direction (01 - 1) - (110)

RD Direction ND Direction

Fig. 3 Orientation analyses for dislocation cell walls using EBSD measurement for same sample investigated by SEM-ECCI method. Micrographs (a)(b) are BSE images, and (c)(d) are orientation mappings images taken from ND direction and RD direction, respectively. The small yellow squares are markings indicating the same grain.

(a) (b)

Fig. 4 SEM-BSE micrographs showing single dislocations distributed among dislocation cell walls, in which micrograph (b) is enlarged image of (a).

JEOL News Vol. 46 No.1 14(2011) (14) degrees, respectively in Fig.5(a)(b)(c) and resulting in the different internal defect tion of the metal is not avoided. In the purpose (d). The tilting axis is parallel to the horizon microstructure of the martensite, which is trans- to observe the dislocation imaging, only surface line on the photograph. Since the incident formed from the austenite phase during the cool- residual stress after mechanical polishing is beam is parallel to <111> direction on the ing. If the carbon concentration is low in the pre- taken care for the sample preparation. After the grain observed, all dislocation lines must be vious austenite phase, the internal defects of bulk sample is mechanically polished and imaged. The dislocations are screw one with martensite becomes to be random dislocations, ground by the alumina polishing solution, the the burgers vector of b=[111], that is not satis- and on the other hand, it is known that the inter- surface of the sample is performed by only the fied with the condition of g • b=0 criteria. It is nal defect of the martensite phase changes from electrolytic polishing. found that the dislocation contrast changes the dislocations to the internal twined microstruc- As a similar observation example, dislocation within one degree, where the Bragg’s angle is ture with increasing of carbon content in the cell structures and mechanical twins of 30nm about one degree under the accelerating volt- austenite phase. The SEM observation is an easy thickness have been observed by the ECCI age of 25keV. The detail of the contrast mech- way to examine the internal microstructure of the method using a SEM-EBSD based setup sys- anism will be discussed in other paper. martensite phase, which has been mainly carried tem in twinning induced plasticity (TWIP) out by the TEM observation. The present demon- steels [13]. The advantage to use the EBSD sys- Application of the ECCI stration shows that such an examination will be tem instead of ECP’s to orientate the crystal possible using the recent ECCI method instead of into optimal diffraction conditions is proposed Method for Other Defects the TEM one. When the sample is slightly tilted, in the article. The recent advance of the BSE in Steels the other dislocation image in the neighboring detector and corresponding improvement for ferrite grain marked B appear with a slight the SEM system provide us the sufficient quali- The ECCI method is useful to observe other change of the crystal orientation condition. With ty of dislocation images and other defects. defects, which are mainly investigated by con- careful tilting experiment, it may estimate the As another example showing a new application ventional TEM. Figure 6 is a SEM-BSE micro- difference of dislocation density depending on of the ECCI method instead of the TEM one, the graph showing complex steel microstructure the amount of local deformation condition around grain boundary morphology and naturally intro- including ferrite and martensite. The area A is a the martensite grain. In addition, the advantage of duced dislocations are indicated in Fig.7. The sam- martensite grain transformed from austenite, the ECCI method is easy to investigate a large ple is a conventional ferrite steel with a bulk size, enlarged by the inserted micrograph (a). It is area with a heterogeneous microstructure such whose grain size is several tens of microns with found by the SEM-BSE observation that the as the complex phases of ferrite and martensite random orientation, as seen in SEM-BSE micro- inner defect of the martensite composes of with different internal defects. graph of Fig.7(a). Under the observation with a twined structure, which is identified to be a high It must be noticed here the sample preparation high magnification as 50,000 times, there are carbon martensite. During the steel making technique. The SEM-BSE image is not sensi- some dislocations in a grain and the grain bound- process, the carbon enriches from a ferrite to an tive to the surface oxide film, which is another ary in Fig.7(b). The sufficient contrast and reso- austenite phase in the complex phase region, advantage of ECCI method, because the oxida- lution for the dislocations enable us to investigate

(a) (b)

100nm

Fig. 5 Change of dislocation image contrast taken by ECCI method with different tilting conditions, (a) 0°, (b) 1°, (c) 2°, (d) 3°, respectively, based on the flat position of the SEM sample stage. 100μm

(15) JEOL News Vol. 46 No.1 15(2011) 100μm the generation and annihilation of a dislocation tion image using the conventional SEM, the contrast imaging technique to the study of under the deformation process. improvement of the sample stage tilting system dislocations associated with cracks in bulk in SEM will be expected. specimens”, Ultramicroscopy, 75, 137-145 Summary (1998). Acknowledgments [7] Simkin, B.A., Crimp, M.A. and Bieler, A single dislocation image using ECCI T.R.,”A factor to predict microcrack nucle- method has been clearly obtained using the The authors acknowledge the technical sup- ation at -grain boundaries.”, Scripta conventional FE-SEM equipped with the super port by SM Business Unit in JEOL to demon- Metall Mater.,49,149-154 (2003). hybrid lens, which will be also useful for the strate the maximum ability of the new BSE [8] Crimp, M.A., Simkin, B.A. and Ng, B.C, magnetic material such as steel. Since the reso- detector in the JSM-7001F. The author also “Demonstration of the g • b u edge dislo- lution of dislocation images is similar to that expresses the thanks to Dr. Uenishi and Dr. cation invisibility criterion for electron obtained by TEM, the advantage that wide area Ikematsu in Nippon Steel Corporation with the channeling contrast imaging”, Phil. Mag. can be observed will be a powerful tool to deformation experiments and useful discussions. Lett, 81, 833-837 (2001). understand the heterogeneity of microstructure [9] Crimp, M.A.,”Scanning Electron under deformation in steel. In the present References Microscopy Imaging of Dislocations in paper, it is demonstrated that the dislocation Bulk Materials, Using Electron Channeling cell walls and single dislocation images are [1] Wilkinson,A. J., Anstis,G. R., Czernuszka, Contrast”, Microscopy research and observed in the back-scatter geometry of ECCI J. T., Long, N. J. and Hirsh, P. B. “Electron Tech.,69, 374-381 (2996). method, where the sample is mounted on the channeling contrast imaging of interfacial [10] Picard, Y.N. and Twigg, M.E., “Diffraction flat position of the stage in SEM and incident defects in strained silicon-germanium layers contrast and Bragg reflection determination electrons are irradiated perpendicular to the on silicon”, Phil. Mag. A, 68, 59-80 (1993). in forescattered electron channeling contrast specimen surface. In this geometry, the sample [2] Wilkinson, A.J. and Hirsh, P.B. “The images of threading screw dislocations in shape and size are not restricted in comparison effects of surface stress relaxation on elec- 4H-SiC”, J. of Applied Phys., 104, 124906 with the fore-scatter geometry, and several in tron channeling contrast images of disloca- (2008). situ observations will be utilized. However, the tions”, Phil. Mag. A, 72, 81-103 (1995). [11] Picard, Y.N., Twigg, M.E., Caldwell, J.D., interpretation of the contrast mechanism for the [3] Czemuszka, J.T., Long, N,J., Boyes, E.D. Eddy, Jr. C.R., Mastro, M.A. and Holm, ECCI method will be more difficult, because and Hirsh, P.B.,”Imaging of dislocations R.T. “Resolving the Burgers vector for indi- several kinds of back-scattering electrons in a using backscattered electrons in a scanning vidual GaN dislocations by electron chan- material must be considered to understand the electron-microscope”, Phil. Mag. Lett., 62, neling contrast imaging”, Scripta case of back-scatter geometry. The detail for 227-232 (1990). Materialia, 61, 773-776 (2009). the contrast origin and several kinds of imaging [4] Dingley, D. “Progressive steps in the devel- [12] Uenishi, A, Teodosiu, C. and Nesterova, data for dislocations will be published in others. opment of electron backscatter diffraction E.V.,”Microstructural evolution at high At the present, the limitation of the observa- and orientation imaging microscopy”, J. of strain rates in solution-hardened interstitial tion for dislocations still remains because the Microscopy, 213, 214-224 (2003). free steels”, Materials Science and contrast is strongly affected by the sample tilt- [5] Ahmed, J, Wilkinson, A.J. and Roberts, Engineering A.,400-401,499-503 (2005). ing condition. The two ways of ECCI methods, S.G, “Characterizing dislocation structures [13] Gutierrez-Urrutia, I., Zaefferer, S. and which are fore-scatter and back-scatter geome- in bulk fatigued copper single crystals using Raabe, D. ”Electron channeling contrast try, will be investigated to make clear the origin electron channeling contrast imaging imaging of twins and dislocations in twin- of the dislocation contrast in detail and also the (ECCI)”, Phil. Mag. Lett.,76,237-245 ning-induced plasticity steels under con- diffraction controlling technique will be investi- (1997). trolled diffraction conditions in a scanning gated for the application. Since it is demonstrat- [6] Ng, B.C., Simkin, B.A. and Crimp, M.A, electron microscopy.”,Scripta Materialia, ed to obtain the sufficient contrast of disloca- “Application of the electron channeling 61, 737-740 (2009).

(a) (a)

100nm

100μm

(b) (b) 100μm B 100nm

100μm

1 μm 100μm Fig. 6 SEM-BSE micrographs showing morphology of complex microstructure with martensite and ferrite phases in steel. Mark A indicates the martensite grain with internal defects of Fig. 7 SEM-BSE micrograph showing convention- twined microstructure as identified by the enlarged micrograph (a), and mark B indicates al polycrystalline steel with random orienta- the locally deformed region in the ferrite grain next to the martensite, the tangled dislocation grains (a) and enlarged BSE micrograph tions are seen in the enlarged micrograph (b). taken by magnification of 50,000 times showing naturally introduced dislocations in a grain near to triple grain boundary (b).

JEOL News Vol. 46 No.1 16(2011) (16) Atmospheric Scanning Electron Microscopy (ASEM) Realizes Direct EM-OM Linkage in Solution: Aqueous Immuno-Cytochemistry

Chikara Sato

Biomedical Research Institute, National Institute of Advanced Industrial Science and Technology (AIST)

The JEOL ClairScopeTM Atmospheric SEM (ASEM) is the next generation in environmental EM, inverting the traditional SEM column to view the sample from beneath. The open ASEM dish on the top of the column allows directly linked correlative microscopy in open solution: optical microscope (OM) images from above, immediately followed by EM of the same sample area at high resolution, from below. The 35 mm ASEM dish allows neuronal primary culture and efficient antibody labeling-washing cycles. Because the epitopes of cells are preserved in aqueous solution, we tested antibodies, including mouse monoclonal antibodies, used in immuno-fluorescence microscopy. All were successful, enabling observation at SEM resolutions. The resolution of the ASEM is 8 nm, which is advantageous for the observation of fine intracellular structures and bacteria. Observable depth is 2-3 µm at 30 kV, and tissue sections placed on the ASEM dish can be observed immediately. Stained with platinum blue, cell nuclei were especially prominent. This is applicable not only to basic biology, but also to intra-operative cancer diagnosis and detection of pathogens. The ASEM is also highly applicable in non-bioscience fields including materials sciences. Introduction organelles. imaging in solution has been one of the most Localization microscopy at high resolution important goals for high-resolution imaging of Proteins in cells dynamically form molecu- has been accomplished with various immuno- cells and tissues. With the advent of new lar complexes with other proteins, which can electron microscopy (immuno-EM) tech- semiconductor technologies, tough, practically then form higher-order supermolecular com- niques, such as immuno-transmission electron electron-transparent films have greatly plexes and/or be integrated into cellular microscopy (immuno-TEM), that have higher improved environmental cells [6, 7]. organelles. They change location by associ- resolution than SEM [3]. However, since Membrane windows are much stronger, ating or dissociating with other proteins or these methods require that the sample be enabling the development of various types of molecules, including cytoskeletal proteins. placed in vacuum or extremely low air pres- environmental cells with microelectrodes or Observation of subcellular supercomplexes sure, pretreatment of samples for protection microfluidic systems [7, 8]. and organelles in cells is important for usually requires significant amounts of time Environmental cells have achieved signifi- understanding protein behavior and function. and effort. Various sample preparation meth- cant breakthroughs in ligand- and affinity- The cytoplasm is generally crowded with pro- ods, such as post-embedding, pre-embedding, labeled EM, visualizing the localizations of teins, especially around the vesicular struc- and the replica method [4], have been devel- cellular proteins in solution. With Scanning tures; solid substances, comprised mainly of oped for immuno-TEM, but the pretreatment TEM (STEM), the distribution of the epider- high-molecular-weight complexes and of samples in post-embedding methods fre- mal growth factor (EGF) receptor in COS7 attached water molecules, usually represent at quently includes dehydration and resin cells was determined by means of 10 nm- least 30% of total intracellular weight. embedding, which can sometimes decrease gold-EGF labeling, at 4 nm resolution, using Therefore, localization microscopy using spe- antigenicity. Thin-sectioning of samples an environmental capsule holding aqueous cific labeling is also necessary for the determi- directly embedded in hydrophilic resin is pos- samples between two electron-transparent nation of protein distribution. For this, labeled sible without dehydration, though it requires films [6]. Mitochondria were labeled with 0.8 probes such as antibodies, complementary considerable technical skill. Pre-embedding nm gold-streptavidin conjugate, and silver oligonucleotides, lectins, and specific ligands preserves antigenicity by labeling fully enhanced, enabling determination of their can be directly conjugated with fluorescence, hydrophilic samples; however, the samples localization in HeLa cells using a SEM-based or labeled again with a secondary fluorescent must then be dehydrated and embedded in environmental capsule with windows [9]. tag for visualization by optical microscopy resin, involving the possible shrinkage of deli- Since the hydrophilic environment preserves (OM). However, diffraction-limited OM has a cate subcellular structures. Also, the thin-sec- sensitive epitopes, it should allow labeling resolution limit of approximately 200 nm. tioning required by both pre- and post-embed- with various antibodies. However, the types of Recent super-resolution OM techniques go ding methods can restrict information to with- cells applicable to these systems are somewhat beyond this [1, 2], but they still do not have in a thin slice, if continuous-sectioning is not limited, since only a very small volume of high enough resolution to locate protein distri- performed. Culture-substrate-parallel thin sec- medium can be placed in the closed capsules. butions in fine subcellular structures or tioning of attached cells is not easy in prac- Another trend has been the development of tice, though it can contain a great deal of correlative microscopy, which combines vari- Higashi 1-1-1, Tsukuba, Ibaraki 305-8566, Japan information. ous EM systems with OM [10, 11]. Among Since the first environmental cells were them, Sartori et al. focused on cryo-correlative [email protected] reported in 1944 [5], electron microscope microscopy [10]. For correlative microscopy

(17) JEOL News Vol. 46 No.1 17(2011) ABOptical microscope OM Air ASEM dish Air SiN Film ASEM Detector Dish Vacuum Magnetic lens

Electron Vacuum Electron beam Detector

Gun SEM Inverted SEM

Fig. 1 Schematic diagrams of the ASEM system and the ASEM dish. (A) An Optical Microscope and a Scanning Electron Microscope are arranged in a column. The SEM is inverted, scanning the sample from below. The OM views the open culture dish from above. (B) The ASEM dish separates atmosphere from vacuum. Electrons penetrate the SiN film fabricated into its bottom plate; backscattered electrons are captured by the backscattered electron imaging (BEI) detector. The OM observes samples quasi-simultaneously through a water-immersion objective lens.

A B ASEM Dish CO22

CDE

Fig. 2 Cell culture, staining and washing using the ASEM dish. (A) The 35 mm ASEM dish is a disposable plastic Petri dish with a SiN film window fabricated into a silicon chip set in the bottom plate. Below: bottom view of the silicon chip with SiN film window. (B) The ASEM dish, with various coatings, can be used for cell culture in a CO2 incubator. (C - E) Cell fixation, washing, labeling, and staining are easy to perform with standard laboratory equipment.

in solution, Dukes et al. determined the distri- dish, as described by Nishiyama et al [18]. performed as described [18]. bution of EGF receptors tagged with EGF- For our experiments, cells were cultured on an quantum dots on COS7 cells, at a resolution of ASEM dish with or without polyethyl- Results 3 nm, using STEM [12]. However, a simulta- eneimine coating, and fixed with 4% neous direct link between OM and EM is gen- paraformaldehyde or 1% glutaraldehyde in Configuration of the ASEM correla- erally not easy. Immuno-correlation PBS (pH 7.4). Except for non-staining or tion microscope microscopy frequently requires tagging an WGA-labeling experiments, cells were perfo- epitope with both fluorescence and gold, rated with 0.1% or 0.5% Triton X-100. They The JEOL ClairScopeTM Atmospheric which has been made possible by the develop- were stained with heavy metal solution and SEM is the next generation in environmental ment of many kinds of labeling systems, e.g., washed three times. For immuno-EM, COS7 EM; it inverts the traditional SEM column to fluoro-nanogold, quantum dots, and sophisti- cells were labeled with anti--tubulin primary view the sample from beneath. This main- cated photo conversion [13-15] [16, 17]. antibody. After fluoronanogold-secondary tains the vacuum necessary for the electron antibody attachment was confirmed using flu- pathway under the thin membrane of the Materials and Methods orescence OM, the gold was enlarged using ASEM dish window (Fig. 1). The 35 mm gold sedimentation [19] and observed using ASEM sample dish is completely open to the The ASEM, with its inverted SEM, was the inverted SEM. Wheat germ agglutinin atmosphere; it is a traditional plastic Petri developed simultaneously with the ASEM (WGA)-15 nm colloidal gold labeling was dish except for the SiN film window in its

JEOL News Vol. 46 No.1 18(2011) (18) base (Fig. 2). The open configuration allows bridge the resolution gap between OM and hyde fixation, specifically stains nuclei. Osmium OM imaging from above, quickly followed nanoscale structural determination methods. tetroxide stains membranes, and emphasizes by EM of the same sample area at high reso- lipid droplets as spherical white dots in the cyto- lution, from below (correlative microscopy) Preferential staining in solution plasm. The endoplasmic reticulum, storing Ca2+ (Fig. 1B). The optical axes of both micro- ions, is an important organelle for cell excitation scopes are aligned and fixed to allow quasi- The 35 mm ASEM dish allows efficient stain- and neuroplasticity. It was stained using plat- concurrent observation of the same area, ing and washing. We employed various tradi- inum blue, and compared with ER-specific fluo- while the sample stage can shift two-dimen- tional heavy-metal EM staining protocols for rescence-labeling using anti-protein disulfide sionally. ASEM observation in solution, and compared isomerase antibody (Fig. 4). This correlative OM/EM microscope has a them with non-stain controls (Fig. 3). In COS7 practical target-size from 200 m to 8 nm, cells, uranyl acetate stains protein and nucleic Neuronal primary culture mainly corresponding to the mesoscopic scale. acids. Platinum-blue stains nucleic acids, and, It provides a fast, relatively easy-to-use way to especially in combination with 1% paraformalde- The detachable ASEM dish has a 3 mL

Paraformaldehyde 1% Paraformaldehyde 4% Glutaraldehyde 1% No stain Pllatinum blue Uranyl Acetate 1% 2% 0.6% 1% 2% Osmium tetroxide

10 m

Fig. 3 Preferential staining using heavy metals. Optimum staining conditions for nuclei were determined using COS7. (A) Cells were fixed with different aldehyde buffers, and stained. Among the various staining reagents, platinum-blue especially emphasizes the nuclei. Within the platinum-blue stained groups, nuclei fixed with 1% paraformaldehyde appear at the highest contrast from the background. For comparison, cells were stained with uranyl acetate or osmium tetroxide. Osmium tetroxide staining emphasizes lipid droplets as spherical white dots in the cytoplasm.

(19) JEOL News Vol. 46 No.1 19(2011) capacity and can be used in a CO2 incubator observation of molecular complexes or fine platinum-blue and placed section-down on the (Fig. 2). We successfully coated the dish subcellular structures, including growth SiN film. The nuclei in the surface were clear, using protocols for glass dishes, and accom- cones and spines. Growth cones were clearly suggesting the applicability of the ASEM to plished a wide range of high-level primary observed using the ASEM at 30 kV accelera- intra-operative cancer diagnosis without cryo- culture on the SiN film (Fig. 5). The success tion voltage (Fig. 5), with an observable sectioning. with coatings is presumably because the sur- depth of 2-3 m from the SiN film [20]. face of the SiN film is oxidized, creating a layer of SiO during the fabrication process. Subcellular structure of bacteria Primary culture of neurons is generally dif- Observation of tissue surface ficult, making neurons inappropriate for Filamentous bacteria used to ferment cellu- environmental-cell systems. Cultured neu- Using the ASEM, brain tissue was also visu- lose for alcohol were observed using the rons on standard dishes can be observed by alized to a depth of 2-3 m in solution (Fig. ASEM after staining with phosphotungstic OM, but resolution is not sufficient for the 6). A slab of goldfish brain was stained with acid. In solution, they manifest as connected

Optical Microscope A B

Fig. 4 Endoplasmic reticulum (ER) visualized using C D E the ASEM. COS7 cells were fixed, perforat- ed, and labeled with anti-protein disulfide isomerase IgG2b and with Alexa Fluor® 488-tagged secondary antibody, and imaged with OM. The cells were then stained with platinum-blue and imaged with the ASEM. (A) Fluorescent image. (B) enlargement. (C) ------EM image of the same cells. (D) 5,000 ASEM 10 m 5 m 1 m magnification. (E) 20,000.

A C

Fig. 5 Cerebral cortex dissociation culture. Neurites of primary-culture pyramidal neurons on day 1, isolated from E16 rat cerebral cortex. (A) Fluorescence microscope image of betaIII tubulin and (B) filamentous actin (F-actin) in neurites. (C) SEM of B the neurites stained with platinum-blue. At 10,000 magnification, filopodia and lamellipodia are moderate-density structures, while the backbone region of the neurite is a high-density area. White squares in A and B correspond to the SEM image in C. Neuron cultures were kindly prepared by 1 m ---- Dr. Takashi Shiga (Tsukuba University).

ABCB

10 m ------10 m ------5 m ------

Fig. 6 Tissue sample stained to observe nuclei. A half-brain surgically resected from a goldfish, fixed with 1% paraformaldehyde and stained with platinum-blue. Tissue was laid section-down on the SiN film of the ASEM dish and observed using the ASEM. Surface nuclei are brightly distinguished from the background. (A) 1,000 magnification. (B) 2,000. (C) 5,000. Outlines of nuclei are clear, without cryo-sectioning. Neurite- like fibers are also visible.

JEOL News Vol. 46 No.1 20(2011) (20) A B C B C

50 m ------10 m ------1 m ------

Fig. 7 Intracellular structure of bacteria. Filamentous bacteria used in the fermentation of cellulose for alcohol are stained with phosphotungstic acid, and observed as connected cells forming filaments in solution. Each cell includes fine, granular structures. Filamentous bacteria were kindly provided by Dr. Shigeki Sawayama (Kyoto University).

cells forming filaments. Each cell includes structures using the immuno-ASEM correlative means that tumor excision is delayed. The fine, granular structures, which could be microscope. Sparsely distributed epitopes can ASEM has the potential to make diagnosis involved in fermentation activity (Fig. 7). be observed with the fluorescence microscope, much quicker, since it does not require cryo- and a specific area targeted in the cell. Then, thin-sectioning, the most difficult and time-con- WGA-labeled EM high resolution images can be captured using suming part of the process. the inverted SEM. Since the ca. 40 nm gold Fast sample examination without time-con- Glycosylation is known to play an important particles, even after enhancement, are generally suming pretreatment is realized with the role in embryonic development and cancer too small to observe at the low magnifications ASEM, and operation of the machine is rela- metastasis. We labeled surface glycans con- used for targeting, this is an efficient means to tively easy to learn, due to the fixed focal point nected to membrane proteins, using lectin. determine the distribution of proteins. Densely (the height of the ASEM dish’s SiN film is COS7 cells labeled with 15 nm colloidal gold- distributed epitopes can be efficiently observed by always the same). The ASEM is widely appli- conjugated WGA were observed using the exploiting the ASEM’s broad, 100 – 100,000 cable not only to basic biology but also to drug ASEM (Fig. 8). White spots indicating gold magnification range. Quasi-simultaneous cor- discovery, intra-operative cancer diagnosis, signals were ubiquitously distributed over the relative microscopy brings another benefit in pathogen diagnosis, food science, cosmetics, whole cell, with higher concentrations on cell that, using variously colored fluorescent trans- polymer chemistry, nanoscience, and materials- fringes. Gold signals were especially dense in fection markers, cells expressing recombinant sciences. the pseudopods, including filopodia and proteins can be targeted with the OM for SEM lamellipodia. These results correspond well imaging. Since only a small percentage of Acknowledgements with the expected distribution of the glycans. recombinant plasmids are usually transfected From the observable gap between gold parti- into cells, this can be very useful. I thank Dr. Toshihiko Ogura at the National cles, the resolution is estimated to be 8 nm. Two other advantages are brought by the Institute of Advanced Industrial Science and open ASEM dish: a large variety of cell types Technology (AIST), Mr. Hidetoshi Nishiyama Immuno-EM are easily culturable, and labeling-washing- and Mr. Mitsuo Suga (JEOL Ltd., Japan) for staining cycles are quick and efficient. Culture valuable discussions in the development of the The open ASEM dish allows the direct conditions are stable, due to the large volume ClairScopeTM. I express cordial thanks to Dr. application of various immuno-cytochemical of the open dish and to the large surface area Takashi Shiga (Tsukuba University) for neu- techniques and conditions developed for OM, of the culture medium, which facilitates gas ral primary cultures, and to Dr. Shigeki including labeling and washing. Since epi- exchange in a CO2 incubator. Throughput of Sawayama (Kyoto University) for filamentous topes are as well preserved in the fully immuno-EM can be very high, allowing the bacteria. I thank Dr. Yuusuke Maruyama for hydrophilic environment (water solution) as quick observation of variously conditioned immuno-electron microscopy (AIST), Dr. Y. with immuno-OM, samples labeled with a cells, and facilitating drug screening. In this Watanabe (Yamagata Research Institute of wide variety of antibodies are observed using study, all pretreatments were carried out in Technology, Japan) and Y. Konyuba (JEOL the ASEM [21]. aqueous buffer, facilitating antigen preserva- Ltd., Japan) for the manufacture of SiN mem- Cytoskeletal proteins are involved in cell tion. More than 50 different antibodies used branes; Mr. M. Koizumi, and Mr. K. Ogawa morphology, protein trafficking, cell division for fluorescence optical microscopy were test- for the development and maintenance of the [22, 23], and neuronal networking [24]. ed. Our success rate with these was 100%, ClairScopeTM, and A. Kohtz, M. Sato, and S. Tubulin, a typical cytoskeletal protein which enabling observation at SEM resolutions. It Manaka (AIST) for technical assistance. makes up microtubules, was immunolabeled should be noted that half were mouse mono- with fluoro-nanogold [13], followed by gold colonal antibodies. enhancement [19] (Fig. 9). Since the open Without any need for sectioning, immuno- References structure of the ASEM dish allows efficient ASEM images through a depth of a few m labeling and washing (medium change), the from the SiN membrane. Decreased accelera- [1] Huang, B., M. Bates, and X. Zhuang, whole procedure can be finished within a few tion voltage reduces the imageable thickness "Super-resolution fluorescence microscopy" hours. Using the ASEM, microtubules were [20], so the depth of the target can be deter- Annu Rev Biochem, 78, 993-1016 (2009) visualized as white lines extending from the mined by duplicated scans at lower accelera- [2] Hell, S.W., "Far-field optical nanoscopy" center to the periphery of cells at low magnifi- tion voltages. This results in imaging compa- Science, 316(5828), 1153-8 (2007) cation (Figure 9 B). At higher magnifications, rable to confocal OM, though the ASEM has a [3] Griffiths, G., "Fine structure immunocyto- the lines were revealed to be uniform, distin- fixed focal point. chemistry, Springer-Verlag" Fine structure guishable gold particles about 40 nm in diam- Without thin-sectioning, the ASEM success- immunocytochemistry, Springer-Verlag, eter (Figure 9 C). It should be noted that back- fully depicted the cell nuclei of tissue samples, 1993. ground signals were extremely low in this using platinum-blue staining. The size of nuclei [4] Fujimoto, K., "Freeze-fracture replica elec- immuno-EM method. is the most important indicator for intra-opera- tron microscopy combined with SDS tive cancer diagnosis, and is traditionally exam- digestion for cytochemical labeling of inte- Discussion ined using OM, with cryo-sectioning and stain- gral membrane proteins. Application to the ing with eosin and hematoxylin. The 15-30 immunogold labeling of intercellular junc- Here, we describe dual-labeling of cellular minutes required for this method, however, tional complexes" J Cell Sci, 108 ( Pt 11),

(21) JEOL News Vol. 46 No.1 21(2011) A B

------2 m 1 m

C D Fig. 8 WGA-colloidal-gold labeling of COS cells; 8 nm resolution of the ASEM. Cells were fixed with glutaraldehyde, and surface glycans were labeled with WGA-colloidal gold con- jugates (15 nm in diameter). Using the →｜｜← ASEM, the gold was imaged through SiN film 100 nm in thickness at a magnification of 5,500 (A), 20,000 (B), and 100,000 (C), where (D) cropped from (C) exhibits a ------resolution of 8 nm. 0.1 m

ABC

Fig. 9 Immuno-staining of microtubules. Immuno-EM of the cytoskeleton using the correlative ASEM: SEM and OM. C2C12 muscle progenitor cells fixed with aldehyde were labeled with anti-microtubule antibody, and further tagged with fluorescence and gold. Fluorescence (A) is directly compared with the SEM image in buffer at 800 (B) and at 10,000 (C).

3443-9 (1995) [12] Dukes, M.J., D.B. Peckys, and N. de Jonge, and tissues in open medium through sili- [5] Abrams, I.M. and J.W. McBain, "A Closed "Correlative fluorescence microscopy and con nitride film" J Struct Biol, 169(3), 438- Cell for Electron Microscopy" Science, scanning transmission electron microscopy 49 (2010) 100(2595), 273-4 (1944) of quantum-dot-labeled proteins in whole [19] Hainfeld, J.F.P., R. D.; Stein, J. K.; [6] de Jonge, N., et al., "Electron microscopy cells in liquid" ACS Nano, 4(7), 4110-6 Hacker, G. W.; Hauser-Kronberger, C.; of whole cells in liquid with nanometer (2010) Cheung, A. L. M., and Schofer, C., "Gold- resolution" Proc Natl Acad Sci U S A, [13] Powell, R.D., et al., "A covalent fluores- based autometallography" Proc. 57th Ann. 106(7), 2159-64 (2009) cent-gold immunoprobe: simultaneous Mtg., Micros. Soc. Amer., Springer- [7] Williamson, M.J., et al., "Dynamic detection of a pre-mRNA splicing factor Verlag, New York, NY, 1999: p. 486-487. microscopy of nanoscale cluster growth at by light and electron microscopy" J [20] Suga, M., et al., "Atmospheric Electron the solid-liquid interface" Nat Mater, 2(8), Histochem Cytochem, 45(7), 947-56 Microscope: Limits of Observable Depth" 532-6 (2003) (1997) Microsc Microanal, 15(Suppl 2), 924 [8] Ring, E.A. and N. de Jonge, "Microfluidic [14] Giepmans, B.N., et al., "The fluorescent (2009) system for transmission electron toolbox for assessing protein location and [21] Murai, T.,et al., "Low cholesterol triggers microscopy" Microsc Microanal. 16(5), function" Science, 312(5771), 217-24 membrane microdomain-dependent CD44 622-9. (2006) shedding and suppresses tumor cell migra- [9] Thiberge, S., et al., "Scanning electron [15] Giepmans, B.N., et al., "Correlated light tion" J Biol Chem. 286(3) (1999-2007) microscopy of cells and tissues under fully and electron microscopic imaging of multi- [22] Fischer, M.,et al., "Rapid actin-based plas- hydrated conditions" Proc Natl Acad Sci U ple endogenous proteins using Quantum ticity in dendritic spines" Neuron, 20(5), S A, 101(10), 3346-51 (2004) dots" Nat Methods, 2(10), 743-9 (2005) 847-54 (1998) [10] Sartori, A., et al., "Correlative microscopy: [16] Gaietta, G., et al., "Multicolor and electron [23] Honkura, N., et al., "The subspine organi- bridging the gap between fluorescence microscopic imaging of connexin traffick- zation of actin fibers regulates the structure light microscopy and cryo-electron tomog- ing" Science, 296(5567), 503-7 (2002) and plasticity of dendritic spines" Neuron, raphy" J Struct Biol, 160(2), 135-45 (2007) [17] Sosinsky, G.E., et al., "Markers for corre- 57(5), 719-29 (2008) [11] Agronskaia, A.V., et al., "Integrated fluo- lated light and electron microscopy" [24] Hirokawa, N., Y. Shiomura, and S. Okabe, rescence and transmission electron Methods Cell Biol, 79, 575-91 (2007) "Tau proteins: the molecular structure and microscopy" J Struct Biol, 164(2), 183-9 [18] Nishiyama, H., et al., "Atmospheric scan- mode of binding on microtubules" J Cell (2008) ning electron microscope observes cells Biol, 107(4), 1449-59 (1998)

JEOL News Vol. 46 No.1 22(2011) (22) Information Derived from PGSE-NMR ~Ion Diffusion Behavior, Molecular Association, Molecular Weight / Composition Correlation of Synthetic Polymers~

Yasuhiro Hashimoto†,†††, Mie Nayuki†, Noriko Horiike†, and Akira Yoshino††

† Asahi Kasei Corporation, Analysis & Simulation Center †† Asahi Kasei Corporation, Yoshino Laboratory ††† Asahi Kasei Corporation, Strategic Planning & Development

Introduction cations and anions as a whole. Thus the components, separation of the spectra by information had been obtained using diffusion coefficients using DOSY is possi- Pulse-Field Gradient Spin Echo NMR PGSE-NMR methods. ble. Thus, there is no longer any need to (PGSE-NMR) can be used to calculate self- isolate the components in advance using diffusion coefficients of molecules. Since it is (2) Evaluation of molecular association [9] HPLC or any other separation method. possible to determine the self-diffusion coeffi- Upon association of molecules, the appar- Combining DOSY with other two-dimen- cient from each individual peak that can be ent molecular sizes will change, which sional NMR methods like COSY, enables a observed in one-dimensional NMR spectra, we should be observed as changes in the self- detailed structure analysis even for com- can evaluate the self-diffusion coefficients diffusion coefficients. Thus by applying plex mixtures. even for a mixture of multiple molecular DOSY method, molecular association can species. We can also separate each signal by be detected. This is not a complete list, and there are innu- their diffusion coefficients using the DOSY merable applications of combined analysis of method [1-4]. Further, it is an important point (3) Evaluation of molecular weight / composi- self-diffusion coefficients and chemical struc- that the self-diffusion coefficients can be con- tion correlation [10-12] ture information. The ability of PGSE-NMR to verted into information about the size and The distribution of self-diffusion coeffi- relate the structure (composition), physical molecular weight of the molecules. This cients can be converted into the distribution properties, and performance of a wide variety of method also allows us to link the chemical of molecular masses. Furthermore, compo- materials holds great potential. In 2006, we had composition information obtained from the sition information obtained from the one- acquired a JEOL JNM-ECA400 series NMR one-dimensional NMR to the self-diffusion dimensional NMR spectra can be correlat- spectrometer and a probe capable of generating coefficients (molecular weight and molecule ed with the diffusion coefficients, i.e. the gradient fields up to 13 T/m, and have been size) provided by PGSE NMR. Thus, these molecular weight of the molecules. Thus, it evaluating a wide variety of samples, from liq- methods are expected to be very useful in a is possible, for example, to use PGSE- uids to solids. Here we report several examples wide range of material analyses, as will be NMR in applications where GPC-NMR is of our recent applications of PGSE-NMR for described below. conventionally used, such as evaluating the material analysis, including ion diffusion in correlation between the composition of a electrolytes, molecular association, and assess- (1) Evaluation of ion diffusion in an electrolyte copolymer and its molecular weight ment of the molecular weight / composition cor- [5-8] relation of synthetic polymers. A principal function of an electrolyte is the (4) Reaction tracking [13] conveyance of the ions. This conduction In situations where the size of the molecules phenomenon can be evaluated using a changes, such as in synthesis reactions, reac- Evaluation of Ion Diffusion direct physical parameter; the self-diffusion tion tracking can be performed by measuring in Electrolytes coefficient of the ion species. Even for a the self-diffusion coefficients of the mole- mixture of multiple ion species, the diffu- cules using PGSE-NMR techniques. In con- The ion conductivity of an electrolyte is sion coefficients can be determined for trast to GPC analysis, where the reaction determined by the number and mobility of the each individual ion specie. For example, in must be stopped for analysis, this ability of dissociated ions. As indicated by the Nernst- the case of rechargeable lithium ion batter- PGSE-NMR to link the information of the Einstein equation, ion conductivity is a func- ies, the movement of the lithium ions is self-diffusion coefficients with the information of the self-diffusion coefficients and the particularly important, but this information tion of chemical structures, obtained from sum of the number of cations and anions pres- cannot be obtained by assessing the ion the one-dimensional NMR spectra, elimi- ent, as well as the degree of dissociation. conductivity measured for the dissociated nates the need to halt reactions midway to Using PGSE-NMR, the self-diffusion coeffi- perform GPC analyses. cients can be determined separately for the 2-1, Samejima, Fuji-city, Shizuoka, 416－8501 Japan anions and cations. This makes it possible to (5) Spectrum separation of mixtures evaluate the individual contribution of the E-mail : [email protected] Even if the sample is a mixture of multiple cations and of the anions to the ion conductivi-

(23) JEOL News Vol. 46 No.1 23(2011)

Na+ About 4mm Water

S membrane C membrane

Anode

Cathode chain

Fig. 1 Chlor-alkali electrolysis process. Fig. 2 Cluster structure and chemical structure of the perfluoro ion exchange membrane.

A B 0 2.4 EW912－① C membrane (EW=1150)

) 2.2 EW972 -1 s -1 2 m 2.0 EW1022-① ) -10 0

(10 1.8 Na ln(E/E D -2 1.6 EW912-② 1.4 S membrane (EW=950) EW1022-② -3 1.2 0 1x109 2x109 3x109 4x109 5x109 3.6 3.8 4.0 4.2 4.4 4.6

γ2δ2g2（Δ-δ/3） Ionic cluster diameter (nm)

23 -9 2 Fig. 3 (A) Na - PGSE-NMR diffusion plot for the S membrane (EW = 950) and the C membrane (EW = 1150). S membrane: DNa = 1.0 10 m /s, -10 2 C membrane: DNa = 4.7 10 m /s. Measurements were carried out with the membrane immersed in water at 90 °C. (B) Relationship between the cluster diameter, determined using SAXS, and the Na+ self-diffusion coefficient, determined using PGSE-NMR, for membranes with different EW. Both SAXS and PGSE-NMR were performed with the membrane immersed in water at room temperature. and are polymers with the same EW, but having different membrane formation conditions to control the cluster diameter. In both cases, a PGSE Spin Echo (PGSE-SE) sequence was used for the measurements.

ty, which cannot be obtained by measuring the may have an influence on the cation selectivi- weight (EW), the relationship between the ion conductivity alone. In this way, it is possi- ty. We comprehensively evaluated the cluster cluster diameter (determined using SAXS) and ble to extract information about the degree of structure using small-angle x-ray scattering the Na+ diffusion coefficient (determined using dissociation. (SAXS), and the self-diffusion coefficients of PGSE-NMR) was clarified (Fig. 3B). It is clear Here, we present diffusion data of Na+, Cl-- in the cations and anions using PGSE-NMR. that as the diameter of the clusters that form + chlor-alkali electrolysis, water in fuel cells, Since both T1 and T2 are short (10 ms or the path for Na becomes larger, the speed of and Li+ in a lithium ion battery. less) for 23Na, a long diffusion time () could the Na+ diffusion increases. The non-linearity not be used in the diffusion measurement. in this relationship might be due to the fact that Chlor-Alkali Electrolysis [14] Furthermore, due to the small gyromagnetic the increase in the cluster diameter does not ratio, it was necessary to apply a large magnet- only make the ion path larger, but also has an Chlor-alkali electrolysis is the process of ic field gradient to attenuate the signals. By effect on the formation of networks and cou- producing chlorine and sodium hydroxide using a 13 T/m gradient magnetic field probe, pling among individual clusters. Furthermore, from NaCl aqueous solution by a membrane we were able to obtain a good 23Na PGSE- direct proportional relationship between these electrolysis (Fig. 1). As shown in Fig. 2, a flu- NMR diffusion plot (Fig 3A). Na+ diffusion coefficients and the ion conduc- orine-based electrolyte membrane, with a car- Distinct difference were observed between the tivity were observed. This means that the Na+ boxylic acid group (C membrane) and a sul- S membrane and the C membrane with regard to ion conductivity is in accordance with the pre- + -9 fonic acid group (S membrane), is commonly Na diffusion (S membrane: DNa = 1.0 10 viously mentioned Nernst-Einstein equation. 2 -10 2 used. Due to the amphipathic nature, the m /s, C membrane: DNa = 4.7 10 m /s). This is in contrast to the fact that the observed exchange groups within the membranes form This small Na+ self-diffusion coefficient in the relationship between the ion conductivity and water-containing micelles (clusters), inside C membrane was a result of the low water the diffusion coefficient of water for the fuel which the ions move. An important character- content and the small cluster diameter in the cell membranes is not linear. This will be dis- istic required for the membranes is the selec- membrane. Using membranes for which the cussed later. tive permeability of Na+ cations and the block- cluster diameter was controlled by changing In order to evaluate the cation selectivity, we ing of anions. The structure of these clusters the film-forming conditions and equivalent evaluated the diffusion of Na+ (23Na-NMR),

JEOL News Vol. 46 No.1 24(2011) (24) -9 2 H2O(D = 1.18×10 m /s) inside the membrane H2O outside the membrane

1H-NMR PPM 3 2 1 0 -1 -2 Cl- outside the membrane

Cl-(D = 0.82×10-9m2/s) inside the membrane 35Cl-NMR PPM 7.5 5.0 2.5 0.0 -2.5 -5.0 -7.5 Na+ outside the membrane Electrolyte membrane O Na+(D = 0.54×10-9m2/s) inside the membrane Anode Cathode 23Na-NMR PPM + - ・Anode H2 → 2H + 2e 10.0 7.5 5.0 2.5 0.0 -2.5 -5.0 -7.5 -10.0 + ・Cathode 2H + 1/2O2 → H2O

Fig. 4 Self-diffusion coefficient evaluated using PGSE-23Na-, 35Cl-, 1H-NMR. Fig. 5 Solid polymer fuel cell. Measurements were performed with membranes immersed in 3.5N NaCl aqueous solution at 80 °C.

4.0 A Water content 3.5 Ion conductivity 3.0

2.5 ) -1 s 2 2.0 m

-10 1.5 (10 D 1.0 Conductivity (S/cm) 0.5

0.0 0 20 40 60 80 100 Relative humidity (%RH) 0.10 B 100%RH 96%RH 0.08

86%RH 0.06

0.04 66%RH

Conductivity (S/cm) 47%RH 0.02

11%RH 32%RH 0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

D (10-10m2s-1)

Fig. 6 (A) Dependency of the water content and ion Fig. 7 (A) Dependency of the self-diffusion coefficient conductivity of the electrolyte membrane on rela- of water on relative humidity. tive humidity. (B) Relationship between the self-diffusion coef- (B) Relationship between water content and ion ficient of water and ion conductivity. Membrane: conductivity. Membrane: S membrane (proton S membrane (proton type), EW = 950, measured type), EW = 950. at room temperature.

-- 35 1 Cl ( Cl-NMR), and H2O ( H-NMR) using an Polymer Electrolyte Fuel Cells [15] inflection point in the graph of this relation- S membrane soaked in NaCl (80 °C). 35Cl is a ship, indicating the presence of two types of low-frequency nucleus, and is outside the fre- An outline of a Polymer Electrolyte Fuel Cell water; type a, which makes a small contribu- quency range of a conventional probe (JEOL (PEFC) is shown in Fig. 5. The polymers tion to the ion conductivity, and type b with a Ltd. TH5 probe), but we managed to perform commonly used as the exchange membrane large contribution. Thus, it was crucial to ana- the measurements by adjusting the tuning are sulfonate membranes. In PEFC, the H+ in lyze the states of the water to understand the beyond the normal limit. The chemical shifts the membrane move from the anode to the proton conduction phenomena. of the ions inside and outside the membrane cathode, accompanying water molecules. It is Then we measured the diffusion coefficient of were different in the spectra, and we were able also known that the ion conductivity of the water under various humidity conditions in to selectively evaluate the diffusion of the ions original fluorine type membranes depends on order to clarify the dependency of proton con- in the membrane (Fig. 4). the relative humidity (Fig. 6A), with a rapid ductivity on humidity, and the results were ana- These results can be directly related to the drop of the ion conductivity observed under lyzed along with information of the states of the electrolyte performance; such as the mem- low humidity. Therefore, it has been necessary water and the membrane cluster structure brane resistance (Na+ diffusion coefficient), to operate PEFC with the membranes in a obtained from IR and SAXS measurements. All cation selectivity (Na/Cl diffusion coefficient high-humidity environment. the PGSE-NMR, SAXS and IR measurements ratio), and the amount of water permeation The relationship between ion conductivity were performed in-situ while controlling the (H2O diffusion coefficient). The ion conduc- and water content is shown in Fig 6B. As the humidity. To achieve this, a saturated brine tivity measurement can not give each parame- water content becomes higher, the ion conduc- solution was sealed into the sample measure- ter individually. This demonstrates the power tivity increases, revealing the importance of ment cells. Since the saturation vapor pressure of the PGSE-NMR method. the presence of water. Furthermore, there is an varies according to the type of salt, it is possible

(25) JEOL News Vol. 46 No.1 25(2011) bound water (IR peak areas) (Fig 9B), reveals that the increase in bound water at low humidi- Free water ty (see a in Fig 9B) only makes a slight contribution to the ion conductivity, while the Bound water growth of free water makes a large contribu- (Bound water) tion to the increase in ion conductivity (see b in Fig 9B). This suggests that the water types a and b in Fig 6 discussed earlier originate from the bound and free water, respectively. The cluster structure was analyzed using (Free water) SAXS. The obtained peaks (indicated by the arrow in Fig 10) were assigned to the scattering interference peaks for each cluster. The cluster distance (d) was determined from the scattering angle using the Bragg's formula. Furthermore, the cluster diameter was deter- (Free water) mined by assuming the clusters to be rigid- (Bound water) body spheres and performing fitting. The dependency of the SAXS profile on the rela- Fig. 8 IR spectra measured on varied relative tive humidity is shown in Fig. 10. It is clear humidity. that for this electrolyte membrane the cluster Membrane: S membrane (proton type), structure exhibits an extremely dynamic alter- EW=950, measured at room temperature. nation dependent on the external environment. The increase in the peak intensity at high humidity may be due to the increase of the water content. Also, a shift of the peak to the Fig. 9 (A) IR peak area for free and bound water low-angle side was observed, suggesting the at varied relative humidity. growth of d. Fig 11A shows the cluster diame- (B) Relationship between IR peak area and ter, determined from the fitting with the SAXS the ion conductivity. Membrane: S membrane (proton type), EW = 950. profile dependent on relative humidity. We can observe a rapid increase in the cluster diameter at relative humidity of over 70% RH. Further, the value obtained by subtracting the cluster A 2.5 diameter from d (= separation between adja- 82%RH cent clusters) drops drastically as the relative 80%RH humidity increases, revealing that the individ- 61%RH 2.0 ual clusters are in close proximity (Fig 11B). 49%RH These rapid changes at around 70% relative 37%RH 19%RH humidity correspond to the region of rapid 1.5 increase in free water discussed earlier. Cluster diameter (nm) These results observed for the states of the

1.8 water support the following conclusions with Fig. 10 SAXS profile measured at varied rela- B regard to the non-linear relationship between the tive humidity. 1.6 ion conductivity and the diffusion coefficient Membrane: Sulfonate membrane (pro- shown in Fig 9B, and the dependence of the H+ ton type), EW = 950, measured at 1.4 conduction mode on the relative humidity. room temperature. cluster diameter 1.2 ● In the range of 0 -30% RH, the water is cluster distancedistance directly bonded to the sulfonic acid group's Fig. 11 The dependency of the cluster diameter 1.0

determined using SAXS on relative Cluster distance (nm) proton by hydrogen bonding. This bound humidity (A), and of the inter-cluster dis- 0.8 water has a low mobility and does not make tance (B). 020406080100 a large contribution to the ion conductivity. ● Membrane: S membrane (proton type), Relative humidity (%RH) In the range of 30 -70% RH, there is no EW = 950. increase in the bound water, and only a slight increase in free water. The free water makes a large contribution to the ion con- to control the humidity inside the sample cell. diffusion coefficient. ductivity. However, since the amount of The salts used include LiCl-H2O (11% RH), To further investigate the source of this increase is small, there is no large increase CaCl2-6H2O (32% RH), KNCS (47% RH), change, the humidity dependency of the results in ion conductivity. NaNO2 (66% RH), NaCl (76% RH), KCl (86% of IR and SAXS were investigated. In the IR ● In the range of 70 -100% RH, there is a sud- RH), and K2SO4 (96% RH). spectrum, two types of peaks were observed; den jump upward in the ion conductivity Fig. 7A shows the dependency of the self- in positions 1640 cm-1 and 1710 cm-1 (Fig. 8). accompanying a rapid increase in the free diffusion coefficient of water in the membrane The 1710 cm-1 peak corresponds to the water water. We can see that the ion conductivity on humidity. As the humidity lowers, there is directly bonded to the sulfonate proton by increases more than the corresponding rise a rapid decrease in the diffusion coefficient of hydrogen bonding (bound water). The 1640 of the diffusion coefficient in this range. water, revealing that this is the source of the cm-1 peak is assigned to the free water demon- At humidity levels below 60% RH the ion drop in the ion conductivity. Fig. 7B shows the strating the same behavior as bulk water [16]. conduction phenomenon appears to be the correlation between the self-diffusion coeffi- Under conditions of low humidity, the amount H+ movement accompanying water (Vehicle cient of water and the ion conductivity. In con- of bound water increases rapidly as the humid- mechanism). In contrast, at humidity levels trast to the previously mentioned direct linear ity rises (up to about 30% RH). As the humidi- of over 70% RH, there is an increase in the relationship exhibited between the ion conduc- ty rises above this level, there is no increase in free water and the diameter of the clusters tivity and the Na+ diffusion coefficient for the bound water (Fig. 9). In comparison, the makes the path for the ions larger. In addi- chlor-alkali electrolysis, the relationship in this free water steadily increases above 30% RH, tion, the individual clusters form a network, case is represented by a curve. As the humidity with a rapid increase appearing at levels over so it is believed that the ion conductivity is rises, it appears that the increase in ion con- 70% RH. The relationship between the ion manifested as the H+ hopping between water ductivity becomes greater than the rise of the conductivity and the amounts of free water and molecules (Grotthus mechanism). This may

JEOL News Vol. 46 No.1 26(2011) (26) be the cause of the observed increase in ion ride composed of 1-4 linked glucose mole- fusion coefficient (Fig. 14 A, B, C). This sug- conductivity exceeding the contribution of cules. A six-membered sugar ring is called gests that in the time scale of DOSY measure- the diffusion coefficient under conditions of CD, a 7 sugar ring molecule, CD, and the ments, a rapid exchange occurs between the high humidity. 8-sugar ring molecule, CD. The outside free and CD associated PNP. Thus the average region of the CD ring is hydrophilic, while the self-diffusion coefficients of the free and includ- As demonstrated above, PGSE-NMR does interior is hydrophobic, and this feature gives ed PNP was observed. As the proportions of not simply offer information on the self-diffu- CD the property of trapping and retaining a PNP and CD were varied, the proportion of sion coefficient; the results can be combined variety of compounds (guest molecules) inside PNP captured in the CD also changed, causing with those from IR and SAXS to perform a its ring [19, 20] (Fig. 13). These inclusion com- the alternation of the diffusion coefficient. At comprehensive analysis, even enabling inter- plexes are widely used in the pharmaceutical, the lower PNP content, the percentage that was pretation of the proton conduction mechanism, food product and chemical saccharide fields, captured by the CD was larger, and the making PGSE-NMR an extremely useful tool. taking advantage of the solubility, extended- observed value approached the CD diffusion release and masking effect properties. coefficient. In this way, by analyzing the varia- Conventionally, the detection of the association in the diffusion coefficient relative to the Rechargeable Lithium Ion Batteries tion of a CD host and a guest molecule was PNP/CD ratio it is also possible to obtain infor- [17] preformed by methods like tracking changes in mation about the association constant. We also UV or 1-dimensional NMR spectra. The diffi- note that the self-diffusion coefficient of CD The movement of the Li+ ions is important for culty, however, is that these changes in the was independent of changes in the mixture ratio. rechargeable batteries, but it is only possible to spectra do not necessarily indicate the inclu- This is due to the fact that there is no significant evaluate the total mobility of all ions (cations sion directly. In comparison, a DOSY method change in the Stokes radius of CD, even when and anions) with ion conductivity measure- allows us to obtain parameters that are related associated with PNP. ments. PGSE-NMR can be used to evaluate the to the apparent molecular size, i.e. the diffu- Next, an example of the inclusion of 4,4'- Li+ and the anions separately, and calculations sion coefficients, which should enable a more biphenyl dicarboxylic acid sodium salt (BDC) can be performed to determine the Li+ trans- direct evaluation of the inclusion phenomenon. in CD will be shown. As can be seen from port number and degree of dissociation. Upon association, the diffusion coefficient of Fig. 15A, a diffusion coefficient value Furthermore, as presented in the following the host and guest molecules should become observed for BDC agrees with that of CD. In example, it is possible to analyze not just the ion identical. addition, there are also peaks obtained that self-diffusion coefficients, but also the diffusion First, we present an example of an analysis match the diffusion coefficients for the BDC behavior. Fig. 12 shows the diffusion plots for of the inclusion of para-nitrophenol (PNP) in and the CD alone. This should correspond to the self-diffusion coefficient of Li+ within two CD. DOSY measurements were performed the free molecules that are not involved in an types of polyketone electrolyte membranes with against a sample mixture of these molecules in inclusion complex. This indicates that unlike the varying diffusion time (). The self-diffu- a 1:1 (mole ratio) in water (Fig. 14B). The PNP, BDC is stably contained in the inclusion sion coefficient of Li+ within the Type 1 elec- results show that although the diffusion coeffi- complex (does not show fast exchange). 1H- trolyte membrane does not exhibit any depend- cient of PNP is smaller than that of PNP alone NMR peaks corresponding to the CD-BDC ency on the diffusion time. In contrast, the Li+ in an aqueous solution, it is not equal to that of complex was assigned by their diffusion coef- diffusion in the Type 2 membrane shows a clear CD. Then, the mixture ratio of these two ficient. Then the molar proportion of BDC and dependency on the diffusion time. The value of compounds was varied and DOSY measure- CD was clearly identified to be 1:2. The dif- + the Li self-diffusion coefficient was DLi = 5.6 ments were performed. The results show that fusion coefficient for the inclusion complex is -11 2 -11 10 m /s for Type 1, and DLi = 0.4 10 the diffusion coefficient of PNP was depend- smaller than that of the free CD, which also m2/s ( = 500 ms) for the Type 2 membrane; so, ent on the mixture ratio; as the PNP fraction suggests that the complex is not simply a 1:1 the value of the Type 1 electrolyte membrane decreases, the value approaches the CD dif- pairing of BDC and CD; rather, a dimer was more than 10 times greater than that of Type 2. Type 1 was an amorphous membrane, while Type 2 was a crystalline membrane. And because of this difference the two should show different diffusion behavior in which the amor- 0.0 0.0 A ■ =50ms B ■ =50ms ▲ =100ms □ =100ms phous material (Type 1) should show free diffu- -0.5 -0.5 + ● =200ms ● =200ms sion of Li and Type 2 material should pass ○ =500ms ○ =300ms -1.0 -1.0 ▲ =400ms through barriers, such as crystal sites. This may △ =500ms ) ) 0 0 explain the non-uniform diffusion indicated by -1.5 -1.5 Type 1 In(E/E

the self-diffusion coefficient becoming smaller In(E/E Type 2 as the diffusion time becomes longer. In this -2.0 -2.0 way, the ability of PGSE-NMR to obtain infor- -2.5 -2.5 mation on the ion diffusion behavior as well as -3.0 -3.0 the self-diffusion coefficients makes it possible 11 11 11 11 11 to establish additional design parameters for 0 1x10 2x10 3x10 4x10 5x10 0 1x1011 2x1011 3x1011 4x1011 5x1011 membranes. Fig. 12 Diffusion plot of 7Li --PGSE-NMR measured at varied diffusion times for polyketone-type electrolyte membrane. Evaluation of Molecular -11 2 (A) Type 1 electrolyte membrane: DLi = 5.6 10 m /s -11 2 Association (B) Type 2 electrolyte membrane: DLi = 0.4 10 m /s ( = 500 ms)

The association of molecules should alter the apparent molecular size. And the self-diffusion coefficient is inversely proportional to the Stokes radius of the molecule. Therefore, it should be possible to detect the association by monitoring the self-diffusion coefficients. Here, we present an example of an analysis of a cyclodextrin inclusion complex. Cyclodextrin Guest molecule Inclusion complex Evaluation of Cyclodextrin Association [18] Fig. 13 Inclusion of a guest molecule in CD.

Cyclodextrin (CD) is a cyclic oligosaccha-

(27) JEOL News Vol. 46 No.1 27(2011) ABC

Diffusion coefficient for PNP alone PNP

PNP PNP

CD CD CD

PNP: CD＝5 : 1 PNP: CD＝1 : 1 PNP: CD＝0.1 : 1

Fig. 14 1H-DOSY spectrum for a PNP-CD inclusion complex. (A) PNP: CD = 5:1 (mole ratio), (B) PNP: CD = 1:1 (mole ratio), (C) PNP: CD = 0.1:1 (mole ratio). DOSY measurements were performed at room temperature.

BDC: CD＝1 : 2 BDC: CD＝1 : 2

Free

BDC Dimer BDC CD inclusion CD complex

Fig. 15 1H-DOSY spectra. (A) BDC: CD = 1:2 (mole ratio), (B) BDC: CD = 1:2 (mole ratio).

NaO O NaO O

ONa O O ONa

Fig. 16 Formation of the BDC and CD dimer inclusion complex.

JEOL News Vol. 46 No.1 28(2011) (28) ー(P1)xー(P2)yー

P1 P2 ) -1 s 2 (m

D Ratio of P1/P2 P1/P2 distribution P1 or P2 content (%)

Molecular weight Molecular weight Molecular weight

Fig. 17 Schematic drawing of the distribution of copolymer composition dependency on M (g mol-1) molecular weight. Fig. 18 Relationship between the molecular weight of polystyrene (M) and the self-diffusion coefficient (D). D = (2.1 10-8) M-0.6

St Bd

Slice data

St Bd

1 Fig. 19 H-DOSY spectrum of styrene-butadiene copolymer. The sample was dissolved in CDCl3, and measurements were performed at room temperature. inclusion complex, as illustrated in Fig. 16. (exchange), and the inclusion structure. In parameter for the polymer properties. Thus, This dimer structure is thought to be the reason addition, because the time scale for the DOSY there is a strong demand for detailed analysis for the stability of this inclusion complex. The measurements is dependent on the diffusion of the composition and structure. For example, trapping of the BDC molecules by 2 molecules time (), as discussed for the evaluation of for the molecular mass dependence of the of CD in the solution also agrees with the lithium diffusion above, by freely varying the composition, the conventional technique had results from X-ray structure analysis of the time scale of the measurements, and tracking been used to perform NMR analysis on each of crystal [21]. the -dependence of the diffusion coefficient, the GPC fractions, or to apply a GPC-NMR In contrast, in the case of the BDC-CD sys- it is possible to obtain information on the method in which GPC and NMR are linked tem, the diffusion coefficients do not match exchange rate of the complex. on-line. However, not only are these methods between BDC and CD (Fig. 15B), unlike the complicated and cumbersome, these methods BDC-CD system. In the same way as also have the disadvantages that once an frac- observed for PNP-CD, there appears to be Evaluation of Correlation tionated portion has been dried, it can no rapid exchange in the time scale of the DOSY between Molecular Weight longer be dissolved in the solvent; or, as for measurements. Since the inner diameter of GPC-NMR, it is necessary to use a costly CD is large for the BDC, so the molecules do and Composition of Synthetic NMR solvent as the eluent. In comparison, not fit precisely, and a stable complex is not Polymers with DOSY the self-diffusion coefficient and formed. its distribution can be determined for each We have presented an example of an analy- In the case of a synthetic copolymer com- individual peak from each component. sis of CD association using DOSY, and have posed of several monomer constituents, the Therefore, evaluation of the correlation shown that the intuitive and straightforward composition ratio of the monomers is often between the molecular weight (diffusion coef- assessment of molecular association is capa- used as a parameter to specify the structure of ficient) and the composition (NMR spectrum) ble. Furthermore, this method also offers the the polymer. However, the molecular mass is possible. Here we present an example of ability to acquire comprehensive information dependency (Fig. 17A, B) or distribution (Fig. analysis of a styrene-butadiene copolymer such as the association constants, stability 17C) of this composition is also an important using DOSY.

(29) JEOL News Vol. 46 No.1 29(2011) select the best method according to the information that is required.

Acknowledgements

We would like to express our sincere appreciation to Dr. Kikuko Hayamizu of the National Institute of Advanced Industrial Science and Technology (AIST) for guid- ance and advice on a wide range of areas related to the evaluation of ion self-diffusion coefficients, from measurement to data Molecular weight interpretation. We are also grateful to Mr. Satoshi Sakurai of JEOL Ltd. for invaluable assistance with the PGSE-NMR measurements, particularly the measurements for 23Na and 35Cl.

References

[1] C. S. Johnson Jr., Progress in Nuclear Magnetic Resonance Spectroscopy 34, Molecular weight 203 (1999). [2] K. Hayamizu, Nihondenshi News 38, 2 (2006) (in Japanese) Fig. 20 Dependency on molecular weight of the styrene and the butadiene unit content. [3] S. Sakurai, JEOL Application Note (A) As determined from the slices of the DOSY spectra of the styrene peak (7.1 ppm) NM146 (2002) (in Japanese) and the butadiene peak (5.4 ppm). CONTIN protocol was used for the inverse LaPlace [4] S. Sakurai, JEOL Application Note transform. NM134 (2000) (in Japanese) (B) As determined from the styrene / butadiene ratio from the 1H-NMR of the GPC frac- [5] K. Hayamizu, Y. Aihara, S. Arai and C. tion (conventional method) G. Martinez, J. Phys. Chem. B 103, 519 (1999). [6] M. Saito, N. Arimura, K. Hayamizu and T. Okada, J. Phys. Chem. B 108, 16064 Correlation between Composition The comparison of the results from the con- (2004). and Molecular Weight for Styrene- ventional method (GPC fraction NMR) [7] M. Saito, K. Hayamizu and T. Okada, J. with the DOSY results is shown in Fig. 20. Phys. Chem. B 109, 3112 (2005). Butadiene Copolymer The results from these two methods are in [8] K. Hayamizu, Kobunshi Ronbunshu 63, close agreement, confirming the reliability of 19 (2006). The relationship between the molecular the DOSY method. It is clear that the high [9] S. Simova and S. Berger, J. Incl. Phen. weight and the diffusion coefficient for stan- molecular weight components are styrene unit 53, 163 (2005). dard polystyrene of differing molecular rich, with the components having a molecular [10] K. Ute, Kagaku 54, 69 (1999) (in weights (usually used for GPC molecular weight of about 1,000,000 being nearly 100% Japanese) weight calibration) is shown in Fig. 18. On the styrene. In contrast, the low molecular weight [11] K. Ute, Kobunshi 51, 824 (2002) (in log-log plot, a nearly straight line is obtained. components in the range of 100,000 to Japanese) Using this relationship as the standard curve, 300,000 have styrene and butadiene units in [12] K. Ute, Seisan to Gijutsu 59, 26 (2007) calculation of the molecular weights from the nearly equal proportions. (in Japanese) obtained diffusion coefficients is possible. Applying the DOSY method in this way [13] M. Nakakoshi, Nihondenshi News 38, 10 However, in the same way as the molecular enables us to make detailed characterizations (2006) (in Japanese) weight calibration curves for GPC, when it is of a synthetic polymer. It is not only signifi- [14] Y. Hashimoto, N. Horiike and S. not possible to generate the calibration curve cantly simpler than the conventional methods, Yamazaki, Lecture Summary of 46th using the identical polymer as in the test sam- it is especially effective for analyzing com- Annual Meeting of the NMR Society of ple, it is difficult to determine the true molecu- pounds that become denatured when the frac- Japan, 92 (2007) (in Japanese) lar weight. Therefore, it is necessary to keep in tions are concentrated and dried. For the [15] Y. Hashimoto, N. Sakamoto and H. mind that the results will be a "polystyrene- styrene-butadiene copolymer used for this Iijima, Kobunshi Ronbunshu 63 166 equivalent weight", in the same way as for study, once the GPC fraction solution is dried, (2006). GPC. it cannot be re-dissolved, so care is required in [16] R. Buzzoni, S. Bordiga, G. Ricchiardi, G. We used DOSY on a styrene-butadiene the concentration process. DOSY does not Spoto and A. Zecchina, J. Phys. Chem. copolymer to determine the dependency of the have this disadvantage. 99, 11937 (1995). styrene / butadiene unit composition on the DOSY provides information about the [17] Y. Hashimoto, N. Horiike, S. Yamazaki, molecular weight (Fig 19). The 7.1 ppm aro- dependency of the composition on the molec- H. Shobukawa and A. Yoshino, The 48th matic ring peak for the styrene unit, and the ular weight, as shown in Fig. 17 A and B. Battery symposium in Japan, 780 (2007). 5.4 ppm double-bond peak of the butadiene The analysis of the composition distribution, [18] N. Horiike, Y. Hashimoto and S. unit were used to obtain diffusion coefficients however in Fig. 17 C is not possible. For Yamazaki, Preprints for the Society of and the distribution. Using the calibration such a case, a two-dimensional GPC-LC Polymer Science, Japan, 55, 3002 (2006). curve mentioned above, the molecular weight method, composed of a combination of GPC [19] A. Harada, J. Li and M. Kamachi, Nature and its distribution were obtained for styrene (molecular weight isolation) and LC (compo- 356, 325 (1992). and butadiene units individually. The results of sition isolation), is required. The drawbacks [20] A. Harada, J. Li and M. Kamachi, Nature the DOSY experiment were compared to the are that it is not easy to define the LC condi- 364, 516 (1993). results of the conventional method in which tions, and it is necessary to search for the [21] S. Kamitori, S. Muraoka, S. Kondo and NMR composition analysis were performed optimal conditions for each sample. K. Okayama, Carbohydrate Research for each GPC fraction. Therefore, it is necessary to consider and 312, 177 (1998).

JEOL News Vol. 46 No.1 30(2011) (30) Development of JEM-2800 High Throughput Electron Microscope

Mitsuhide Matsushita†, Shuji Kawai†, Takeshi Iwama†, Katsuhiro Tanaka†, Toshiko Kuba†† and Noriaki Endo†

† EM Business Unit, JEOL Ltd. †† Electron Optics Sales Division, JEOL Ltd.

Introduction cost of the future (S)TEM is reduced, as well sion that the operation of the (S)TEM is diffi- as in the future, (S)TEM can maintain a stable cult. In the JEM-2800, we have newly devel- In recent years, new material developments operating rate after installation. The users also oped a graphical user interface (GUI) and utilizing nanometer-order structure control hope to effectively apply the future (S)TEM to operation panels so that the user can intuitively technologies are actively being carried out, as various data acquisitions such as 3D observa- perform the necessary operations without cum- well as product developments using these tion by Tomography. Based on these require- bersome feeling. Figure 2 shows the operation newly developed materials. In the process of ments, the JEM-2800 has been developed so panels of this instrument. The left side of Fig. developing these materials and products, it that this new instrument can be accepted by 2 shows the trackball for moving the speci- becomes indispensable to use a scanning-trans- many companies and research laboratories men, the center of Fig. 2 shows the main oper- mission and / or a transmission electron micro- with keywords of “high throughput” and “high ation panel, and the right side of Fig. 2 shows scope ((S)TEM) for the methods of morpho- usability.” the panel for adjusting the specimen height (Z) logical observation and analysis of a local area. and tilting the specimen. Although the number Furthermore, in high-tech industries such as of knobs and switches is smaller compared the semiconductor industry, the (S)TEM is Features with those of the operation panel of the con- installed in the site near the production line as ventional (S)TEM, the user can intuitively an analysis tool for improving the yield rate of External view operate the instrument because appropriate production and analyzing defect causes, and functions are automatically assigned to these the user is making full use of the (S)TEM at Figure 1 shows an external view of the knobs and switches according to the state of any time of the day and night. However, instal- JEM-2800. The installation environment of a the instrument. lation of the (S)TEM in such a site often caus- (S)TEM differs largely depending on environ- Figure 3 shows the GUI of the instrument. es problems from the viewpoints of cost and ments of user sites. In addition, there are many The JEM-2800 GUI can display all the obser- usability. In order to solve these problems, the cases where an installation room does not suf- vation images such as (S)TEM and SEM JEM-2800 High Throughput Electron ficiently meet the installation requirements of a images, set each observation condition, and Microscope has been newly developed. This (S)TEM. To overcome this situation, the display the operational status of each part of report describes the product concept and fea- microscope column of the JEM-2800 is cov- the JEM-2800 on a real-time basis. In addition, tures of this instrument. ered with an enclosure so that the (S)TEM can the user can operate the instrument under the be stably used. The enclosure is designed by environment of a bright room. Since the JEM- Product Concept taking account of smooth maintenance so that 2800 enables you to simultaneously observe the down time can be minimized at the mainte- and record scanning images, such as STEM- Table 1 shows the needs of (S)TEM for nance. The instrument height is about 2.6 m, BF image, STEM-DF image and SEM image, users in the semiconductor industry. The users making it lower compared with the general enhancement of throughput can be expected. in the semiconductor industry feel that the pre- 200 kV TEM. The user sits at the front of the Depending on the procedure of observation or vious or present (S)TEM requires a profession- operation console shown to the right side of analysis, it is sometimes required to frequently al and proper operator. The users also feel that Fig. 1, and can perform all the operations of switch between TEM and STEM. In the JEM- the implementation cost of (S)TEM is higher the instrument. 2800, simply clicking the observation-mode compared with that of other instruments such switching buttons on the top left of the GUI as SEM. User interface enables you to easily switch the observation On the other hand, the users hope that a modes between TEM, STEM-BF, STEM-DF, future (S)TEM has a user interface allowing A user who used a SEM or a user who uses a SEM and Diffraction. In the case of switching anyone to operate, and that the implementation (S)TEM for the first time often has an impres- image observation such as from TEM to

(31) JEOL News Vol. 46 No.1 31(2011) Table 1 Needs of STEM and TEM for users in the semiconductor industry.

2010Y 2020Y 27 22 18 15.3 12.8 10.7 8.9 7.4

ITRS 2009: MPU physical gate length (nm)

Previous or present STEM / TEM Future STEM / TEM Professional and proper operators are required Easy operation Complicated operation procedure SEM like operation Specially techniques for analysis High cost Reduction of analysis cost Cost is higher than the SEM Shortening in observation or analyzing time Long time is spent for analysis No proper operator needed Prolonged operator training Reduce operator training time Present performance (0.1 ~ 0.2 nm) is adequate in most cases Performance is more better than the present TEM and STEM Limited data acquisition techniques Various data acquisition techniques 2D analysis / imaging 3D imaging by tomography Manual operation Critical dimension by automation Strain analysis, etc.

● Everyone needs these functions

Fig. 1 External view of JEM-2800. Left-hand side of the figure is the microscope column in enclosure. Enclosure height is about 2.6 m. Right-hand side of the figure is an operation console. All operations are performed from the operation console.

Fig. 2 Operation panels of JEM-2800. Trackball panel, main operation panel and specimen height and tilt control panel are shown from the left-hand side. Main operation panel upper-side is used for alignment, and bottom-side is used for acquisition setting of images.

Fig. 3 Graphical user interface (GUI) of JEM-2800. Simultaneous acquisition of STEM-BF, STEM-DF and SEM images can be performed. Thumbnail view- er is prepared for the lower part of GUI.

JEOL News Vol. 46 No.1 32(2011) (32) STEM, the observation mode changes in about astigmatism correction. Alignment of the speci- enables you to operate the instrument from a 10 seconds while keeping the same magnifica- men orientation is to align the (crystal) zone remote site. This optional function offers the tion and the same field of view, thus you can axis of the specimen with respect to the present- same GUI and operation panel even in the continue the observation without requiring the ly observed field of view. Furthermore, auto- remote site. If you arrange the respective oper- time for re-searching the observation field of matic focusing and automatic astigmatism cor- ation panels and PCs in the (S)TEM main unit view due to the switching of the observation rection can be executed irrespective of the and the remote site, you can also perform the mode. observation mode of TEM or STEM. These image observation and analysis in real time automatic functions help to greatly reduce cum- while performing discussion. Automatic functions bersome operations of the user and variations of the observation conditions. Navigation system JEM-Navi When performing the observation of multiple fields of view, it is necessary to repeat the same Remote operation A user who is unaccustomed to the (S)TEM operation procedure. Figure 4 shows the gener- operation often becomes confused on the oper- al operation flow of a (S)TEM. The JEM-2800 There is also a need to observe an image or ation procedure. In order to guide the user is provided with automatic adjustment functions perform discussion while observing the image without miss-operations, the navigation system for frequently used operations, including bright- from a place different from the installation site is built in the JEM-2800. Figure 5 shows the ness & contrast adjustment, alignment of the of the (S)TEM. The JEM-2800 is provided example screen of the JEM-Navi. When the specimen orientation (zone axis), focusing and with an optional remote operation function that user reads the operation procedure and clicks a

Procedure of observation Insert a specimen holder in a column Search target view

Contrast & brightness Auto contrast &brightness

Z- Control Auto Z-control

Orientation Auto orientation

Focus Auto focus

Stigmator Auto stigmator

Take a photograph, analyze by using EDS and / or EELS

On GUI

Fig. 4 General operation flow of TEM and STEM. Almost operation proce- Fig. 5 Example screen of the operation navigator (JEM-Navi). dures can be performed automatically. Corresponding switch of operation panel and / or the portion of GUI blinks by clicking the link button of the operation navigator.

0.1nm ||

2nm Fig. 6 High resolution TEM image of gold single crystal. Accelerating voltage is 200 kV.

(33) JEOL News Vol. 46 No.1 33(2011) link button in the document on the monitor lution STEM image of a silicon (111) single STEM detector. Figure 9 shows STEM screen, the software leads the user by blinking crystal that exhibits lattice fringes with spacing images of a semiconductor device with differ- the lamp of switches or a portion of the screen of 0.19 nm. In recent years, there has been a ent collection angles. They are a STEM-BF that the user should operate. In the future, the high demand for high-resolution imaging even image, a STEM-LAADF image, and a STEM- main terms in the navigation document will be for soft materials such as porous material and HAADF image from the left side of the figure. linked to the glossary; that is, clicking a term macro-molecular material. The JEM-2800 Many elements are used in the FET portion of allows for displaying the explanation of the meets this demand, enabling you to observe recent semiconductor devices, and artificially relevant term on the screen. such a high-resolution image at an accelerating controlled strain is introduced into the silicon voltage of 100 kV even in the standard config- substrate. The STEM-LAADF image can dis- High-resolution image uration. Figure 8 shows a high-resolution tinguish the elements different from the low- TEM image of a gold single crystal acquired at dielectric constant (Low-k) material portion The JEM-2800 enables you not only to pro- an accelerating voltage of 100 kV. Lattice since image contrast is different for each dif- vide high usability, but also to obtain high fringes with spacing of 0.14 nm can be visual- ferent element depending on the atomic num- quality data. Figure 6 shows a high-resolution ized in this image. ber. Furthermore, in the silicon substrate por- TEM image of a gold single crystal acquired at On the other hand, there is also a high tion, it is possible to clearly observe the por- an accelerating voltage of 200 kV. Lattice demand for observing an image with suffi- tion into which the distortion is introduced fringes with spacing of 0.1 nm are clearly visi- ciently good contrast. The JEM-2800 allows with a contrast different from the surrounding ble in this image. Figure 7 shows a high-reso- large change of the collection angles of the portion free of distortion.

0.14nm || || 0.192nm

1nm

Fig. 7 High resolution STEM-DF image of silicon single crystal. Fig. 8 High resolution TEM image of gold single crystal. Accelerating volt- Accelerating voltage is 200 kV. age is 100 kV.

(a) (b) (c)

Low-k

Lattice defects

50nm

Fig. 9 STEM images of semiconductor device acquired by different collection angles. (a) < 11 mrad, (b) 14 to 63 mrad and (c) 46 to 208 mrad. Lattice defects and low-k layers are clearly observed in (a) STEM-BF image and (b) STEM-LAADF image.

JEOL News Vol. 46 No.1 34(2011) (34) High speed EDS analysis detected. It is also found that spectral peaks of sil- ImageCenter and managed in a lump. These icon and tungsten are clearly separated irrespec- data are read from each client to perform data One of the advantages to perform EDS analysis tive of their close characteristic X-ray energies to processing. Figure 12 shows the GUI dis- using a (S)TEM is that you can perform a high- each other. played on a client site. Images and data can be resolution EDS analysis; in recent years, howev- Furthermore, the JEM-2800 enables you to searched using keywords such as photography er, analysis speed-up is also requested. The JEM- perform various analyses such as EELS, date/time and photography conditions. After 2800 can install an SDD (silicon drift detector) of Tomography and local-distortion analysis in that, you can perform analysis or other tasks 100 mm2 having a solid angle of 0.95 sr. With accordance with the needs of the user. using the necessary software. this installation, the instrument can provide detection sensitivity about as five or more times as that Data management system Summary compared with a 50 mm2 Si(Li) detector. In addition, as you can also change the illumination con- Some companies also request to efficiently In the future, opportunities to use a (S)TEM dition from a large current probe to a high-resolu- manage the observation images and analysis will further increase in various fields. In such a tion probe, you can set the optimum analysis con- results. A data management system circumstance, the JEM-2800 High Throughput dition according to the size of the analysis area (ImageCenter) has been developed for the Electron Microscope will contribute to more and the resolution required. Figure 10 shows ele- JEM-2800. Figure 11 shows a schematic diversified fields since this instrument can pro- mental mapping images of a semiconductor drawing of this data management system. The vide all users with high performance and high device acquired with EDS. It is seen that trace results of data acquired or analyzed by a throughput owing to its new various functions doped elements of hafnium and tantalum can be (S)TEM are automatically unified in the focused on high usability.

(a)

(b) (c) (d) (e)

(f) (g) (h) (i)

(j) (k) (l) (m)

Fig. 10 EDS peak separated maps of 32 nm PMOS. (a) STEM-DF, (b) C-K, (c) N-K, (d) O-K, (e) Al-K, (f) Si-K, (g) Ti-K, (h) Ni-K, (i) Cu-K, (j) Ge-K, (k) Hf-L, (l) Ta-L and (m) W-L.

(35) JEOL News Vol. 46 No.1 35(2011) LAN TEM Image Center

⅐ SEI, STEM and TEM images ⅐ EDS, EELS ⅐ Tomography (TEMography) All data are automatically stored into a server (Image Center).

Client 1 Client 2 ⅐ CD (Image Excite) Client 3 ⅐ Particle analysis ⅐ Creation of reports

Fig. 11 Schematic drawing of TEM data management system. All the images and analysis data are stored in a file server (Image Center). Users can access Image Center from each client PC.

CD software Image Excite

Particle analysis software Region Gauge

Data management by data

Picture information is displayed

Fig. 12 Client GUI of Image Center. Stored images are shown. Data search can be performed by keywords such as date, specimen name, etc.

JEOL News Vol. 46 No.1 36(2011) (36) Introduction of New Products

High Throughput Electron Microscope JEM-2800

The JEM-2800 is a new transmission electron microscope that achieves nano-area analysis through Automation and Convenience, while being Easy-to-use so that expert results can be achieved by operators of any skill level. The advanced electron optical system of the JEM-2800 makes it possible to perform high-resolution TEM and STEM imaging, EDS, EELS, tomography and in-situ observation on the same sample without sacrificing any of these capabilities. This next-generation TEM provides innovative solutions through its high performance and user-friendly operation.

Automation Performance The JEM-2800 can be operated using fully automatic func- The JEM-2800, with its new electron optic system, achieved tions for contrast, brightness, specimen Z position (height), both high resolution imaging and high speed analysis. A vari- crystal zone axis alignment, focus and astigmatism correction. ety of preset beam data are available to insure the optimum Each function can be simply executed by pushing the corre- settings for the sample or analyzing technique, enhancing the sponding button. speed and accuracy of analysis. The microscope is image rota- tion free and positional shift of the area of view in the entire magnification range from low to high resolution imaging, and when the mode is switched between TEM and STEM. This Auto Function includes rapid, easy imaging and analysis in all imaging modes. ACB AZ AO AF ASTG

The Auto Function buttons on the GUI

All -- in -- one The JEM-2800 has the capability to observe TEM, STEM and SE (secondary electron) images along with electron diffraction 5nm patterns. In scanning mode BF-STEM, DF-STEM and SE images can be observed simultaneously. With the addition of EDS and EELS systems, microscope operation from initial High resolution TEM images of 32nm NMOS 5nm search to final analysis can be performed seamlessly . Quick Turn Around Time JEM-Navi These features combined enable for quick specimen observa- The JEM-2800 incorporates a new operation navigation tion and data collection, realizing quick turn around time system, "JEM-Navi". This system makes the JEM-2800 between specimens. easy to be used for operators with any level of skill.

(37) JEOL News Vol. 46 No.1 37 (2011) Introduction of New Products

Thermal Field Emission Scanning Electron Microscope JSM-7800F

The new JSM-7800F has been developed to be an ultimate research tool suitable for institutions requiring a large variety of material research. The newly developed super hybrid objective lens (SHL) provides the resolution of 0.8nm at 15kV and 1.2nm at 1kV. The SHL objective lens allows researchers to study magnetic samples at high magnification. The in-lens thermal FEG produces an extremely stable probe current. Addition to the observation of fine surface structures, the JSM-7800F enables researchers to analyze sub-micron structures with EDS, WDS, and EBSD.

Extreme resolution The super hybrid objective lens provides extreme resolution of 0.8nm at 15kV and 1.2nm at 1kV. With a very low incident electron energy, extremely fine surface structures are revealed. The distribution of materials can be observed even at 0.5kV. Fast and high precision analysis The aperture angle optimizing lens keeps the electron probe small even at large probe currents. A large probe current allows you to analyze samples quickly without sacrificing the precision and quality of the analyses. A variety of analytical systems including EDS, WDS, and EBSD are available. Distortion free EBSD patterns are obtained for high precision crystal structure analysis. High stability, consistency in acquired data The in-lens thermal FEG produces an extremely stable probe current. The Highest Performance is always available when you need it. Results obtained by multiple users or on different days are con- sistent and easy to compare. No limitation in specimens The super hybrid lens is a field free lens at the analytical working distance. Fe3O4 nano-particle cluster Magnetic samples can be observed and analyzed at high magni- The SEM image courtesy of Dr. Takanari Togashi, Prof. Tadafumi Adschiri, fication. Advanced Institute for Material Research, Tohoku University Non conductive samples are easily observed. Come closer to the nano-world The super hybrid objective lens brings you closer to the nano- world. The objective lens delivers superb high resolution even at extremely low electron energies. This is necessary to observe and analyze nano size structures. The super hybrid objective lens can be used to observe and analyze magnetic materials at high magnifications. Detectors A variety of detectors can filter electrons from the specimen by energy and emission angle.

JEOL News Vol. 46 No.1 38 (2011) (38) Certain products in this brochure are controlled under the “Foreign Exchange and Foreign Trade Law” of Japan in compliance with international security export control. JEOL Ltd. must provide the MGMT.SYS. MGMT.SYS. Japanese Government with “End-user’s Statement of Assurance” RvA C024 R vA C425 and “End-use Certificate” in order to obtain the export license ISO 9001 & 14001 REGISTERED FIRM DNV Certification B.V., THE NETHERLANDS needed for export from Japan. If the product to be exported is in this category, the end user will be asked to fill in these certificate forms. ISO 9001 & ISO 14001 Certificated

http://www.jeol.com/ 1-2 Musashino 3-chome Akishima Tokyo 196-8558 Japan Sales Division Telephone:+81-42-528-3381 Facsimile:+81-42-528-3386

ARGENTINA GREECE SCANDINAVIA COASIN S.A.C.I.yF. N. ASTERIADIS S.A. SWEDEN Virrey del Pino 4071, 56-58,S. Trikoupi Str. P.O. Box 26140 JEOL (Skandinaviska)AB 1430 Buenos Aires GR-10022, Athens, Greece Hammarbacken 6A, Box 716, 191 27 Sollentuna Argentina Telephone: 30-1-823-5383 Sweden Telephone: 54-11-4552-3185 Facsimile: 30-1-823-9567 Telephone: 46-8-28-2800 Facsimile: 54-11-4555-3321 Facsimile: 46-8-29-1647 HONG KONG AUSTRALIA & NEW ZEALAND FARMING LTD. SERVICE & INFORMATION OFFICE JEOL(AUSTRALASIA) Pty.Ltd. Unit 1009, 10/F., MLC Millennia Plaza NORWAY Suite 1, L2 18 Aquatic Drive 663 King's Road, North Point, Hong Kong JEOL (Skandinaviska)AB - Frenchs Forest NSW 2086 Telephone: 852-2815-7299 Lorenvangen 23, 0580 Oslo, NORWAY Australia Facsimile: 852-2581-4635 Telephone: 47-2-2-64-7930 Telephone: 61-2-9451-3855 Facsimile: 47-2-2-65-0619 Facsimile: 61-2-9451-3822 INDIA BLUE STAR LTD. (HQ: Mumbai) FINLAND AUSTRIA Analytical Instrments Department, JEOL (Skandinaviska)AB LABCO GmbH Sahas' 414/2 Veer Savarkar Marg Ylakaupinkuja 2, FIN-02360 Espoo, Finland Dr.-Tritremmel-Gasse 8, Prabhadery Mumbai 400 025, India Telephone: 358-9-8129-0350 A-3013 Pressbaum, Austria Telephone: 91-22-6666-4000 Facsimile: 358-9-8129-0351 Telephone: 43-2233-53838 Facsimile: 91-22-6666-4001 Facsimile: 43-2233-53176 SINGAPORE BLUE STAR LTD. (Delhi) JEOL ASIA PTE. LTD. BANGLADESH Analytical Instruments Department, 2 Corporation Road #01-12 Corporation Place A.Q. CHOWDHURY SCIENCE & SYNERGY PVT. LTD. E-44/12 Okhla Industrial Area, Singapore 618494 House No. 12, Road No. 5A Phase-II, New Delhi 110 020, India Telephone: 65-6565-9989 Sector No. 11, Uttara Dhaka - 1230 Telephone: 91-11-4149-4000 Facsimile: 65-6565-7552 Bangladesh Facsimile: 91-11-4149-4005 Telephone: 880-2-9980790, 8953450, 8953501 SOUTH AFRICA Facsimile: 880-2-8854428 BLUE STAR LTD. (Calcutta) ADI Scientific (Pty) Ltd. Analytical Instruments Department, 370 Angus Crescent, BELGIUM 7, Hare Street Calcutta 700 001, India Northlands Business Park, 29 Newmarket Road JEOL (EUROPE) B.V. Telephone: 91-33-2213-4133 Northriding, Ranburg, Republic of South Africa Planet II, Gebouw B Facsimile: 91-33-2213-4102 Telephone: 27-11-462-1363 Leuvensesteenweg 542, Facsimile: 27-11-462-1466 B-1930 Zaventem BLUE STAR LTD. (Chennai) Belgium Analytical Instruments Department, SPAIN Telephone:32-2-720-0560 No. 46, Garuda Building, IZASA. S.A. Facsimile:32-2-720-6134 Cathedral Road, Chennai 600 086, India Argoneses, 13, 28100 Alcobendas, Telephone: 91-44-4244-4000 (Poligono Industrial), Madrid, Spain BRAZIL Facsimile: 91-44-4244-4190 Telephone: 34-91-663-0500 JEOL Brasil Instrumentos Cientificos Ltda. Facsimile: 34-91-663-0545 Av. Itaberaba, 3563 INDONESIA 02739-000 Sao Paulo, SP Brazil PT. TEKNOLABindo Penta Perkasa SWITZERLAND Telephone: 55-11-3983 8144 Komplek Gading Bukit Indah Blok I/11 JEOL (GERMANY) GmbH Facsimile: 55-11-3983 8140 JI. Bukit Gading Raya Kelapa Gading Permai, Oskar-Von-Miller Strasse 1, Jakarta 14240, Indonesia 85386 Eching, Germany CANADA Telephone: 62-21-45847057/58/59 Telephone: 49-8165-77346 JEOL CANADA, INC. Facsimile: 62-21-45842729 Facsimile: 49-8165-77512 3275 Premier Rue #8, St-Hubert, Quebec J3Y 8Y6, Canada ITALY TAIWAN Telephone: 1-450-676-8776 JEOL (ITALIA) S.p.A. JIE DONG CO., LTD. Facsimile: 1-450-676-6694 Centro Direzionale Green Office 7F, 112, Chung Hsiao East Road, Via dei Tulipani, 1 Section 1, Taipei, Taiwan 10023 CHILE 20090 Pieve Emanuele (MI) Italy Republic of China TECSIS LTDA. Telephone: 39-02-9041431 Telephone: 886-2-2395-2978 Avenida Kennedy 5454 - Piso 5 Facsimile: 39-02-90414343 Facsimile: 886-2-2322-4655 Vitacura, Santiago, Chile Telephone: 56-2-401-8520 KOREA For Semiconductor Products: Facsimile: 56-2-410-8541 JEOL KOREA LTD. JEOL TAIWAN SEMICONDUCTORS LTD. Dongwoo Bldg. 7F, 458-5, Gil-Dong, 11F-1, No. 346, Pei-Da Road, Hsin-Chu City 300, CHINA Gangdong-Gu, Seoul, 134-010, Korea Taiwan, Republic of China JEOL(BEIJING) CO., LTD. Telephone: 82-2-511-5501 Telephone: 886-3-523-8490 Room B1010/1110, Wantong New World Plaza No. 2 Facsimile: 82-2-511-2635 Facsimile: 886-3-523-8503 Fuchengmenwai Street, Xicheng District, Beijing 100037, P.R.China KUWAIT THAILAND Telephone: 86-10-6804-6321 Ashraf & CO. Ltd. W.L.L. BECTHAI BANGKOK EQUIPMENT & CHEMICAL CO., Ltd. Facsimile: 86-10-6804-6324 P.O.Box 3555 Safat 13036, Kuwait 300 Phaholyothin Rd. Phayathai, Bangkok 10400, Telephone: 965-1805151 Thailand JEOL LTD., SHANGHAI OFFICE Facsimile: 965-24335373 Telephone: 66-2-615-2929 Shanghai Equatorial Hotel Office Building 803, Facsimile: 66-2-615-2350/2351 65 Yanan Road West, Shanghai 200040, P.R. China MALAYSIA Telephone: 86-21-6248-4868/4487/4537/4404 JEOL(MALAYSIA) SDN.BHD.(359011-M) THE NETHERLANDS Facsimile: 86-21-6248-4075 205, Block A, Mezzanine Floor, JEOL (EUROPE) B.V. Kelana Business Center, Lireweg 4, NL-2153 PH Nieuw-Vennep, JEOL LTD., GUANG ZHOU OFFICE 97, Jalan SS 7/2, Kelana Jaya, The Netherlands N3104, World Trade Center Building 47301 Petaling Jaya, Selangor, Malaysia Telephone: 31-252-623500 371-375, Huan Shi East-Road, Guang Zhou, Telephone: 60-3-7492-7722 Facsimile: 31-252-623501 510095, P.R.China Facsimile: 60-3-7492-7723 Telephone: 86-20-8778-7848 TURKEY Facsimile: 86-20-8778-4268 MEXICO TEKSER A.S. JEOL DE MEXICO S.A. DE C.V. Kayisdagi, Inonu Mah.Kartal Cad. No:55/3 JEOL LTD., WUHAN OFFICE Av. Amsterdam #46 DEPS. 402 34755 Atasehir / Istanbul, Turkey Room 3216, World Trading Bldg. Col Hipodromo, 06100, Mexico D.F. Telephone: 90-216-5736470 686 Jiefang Street, Hankou, Wuhan, Hubei 430032 Mexico Facsimile: 90-216-5736475 P.R.China Telephone: 52-5-55-211-4511 Telephone: 86-27-8544-8953 Facsimile: 52-5-55-211-0720 UAE Facsimile: 86-27-8544-8695 BUSINESS COMMUNICATIONS LLC. (Abu Dhabi) PAKISTAN (Karachi) P.O. Box 2534, Abu Dhabi UAE JEOL LTD., CHENGDU OFFICE ANALYTICAL MEASURING SYSTEM (PVT) LTD.(AMS LTD.) Telephone: 971-2-6348495 1807A Zongfu Building, 14-C Main Sehar Commercial Avenue Lane 4, Facsimile: 971-2-6316465 NO. 45 Zhongfu Road, Chengdu, Sichuan, 610016 Khayaban-e-Sehar, P.R. China D.H.A-VII, Karachi-75500, Pakinstan BUSINESS COMMUNICATIONS LLC. (Dubai) Telephone: 86-28-86622554 Telephone: 92-21-35345581/35340747/35346057-8 P.O. Box 233, Dubai, UAE Facsimile: 86-28-86622564 Facsimile: 92-21-35345582 Telephone: 971-4-2220186 Facsimile: 971-4-2236193 CYPRUS PANAMA JEOL (EUROPE) SAS PROMED S.A. USA Espace Claude Monet, 1 Allee de Giverny Parque Industrial Costa del Este JEOL USA, INC. 78290, Croissy-sur-Seine, France Urbanizacion Costa del Este 11 Dearborn Road, Peabody, MA 01960, U.S.A. Telephone: 33-13015-3737 Apartado 0816-01755, Panama, Panama Telephone: 1-978-535-5900 Facsimile: 33-13015-3747 Telephone: 507-303-3100 Facsimile: 1-978-536-2205/2206 Facsimile: 507-303-3115 EGYPT JEOL USA, INC. WEST OFFICE JEOL SERIVCE BUREAU PHILIPPINES PHILAB INDUSTRIES INC. 5653 Stoneridge Drive Suite #110 3rd Fl. Nile Center Bldg., Nawal Street, Pleasanton, CA 94588, U.S.A. Dokki, (Cairo), Egypt 7487 Bagtikan Street, SAV Makati, Telephone: 1-925-737-1740 Telephone: 20-2-3335-7220 1203 Metro, Manila Phillippines Facsimile: 1-925-737-1749 Facsimile: 20-2-3338-4186 Telephone: 63-2-896-7218 Facsimile: 63-2-897-7732 VENEZUELA FRANCE GOMSA Service and Supply C.A. JEOL (EUROPE) SAS PORTUGAL Izasa Portugal Lda. Urbanizacion Montalban III Espace Claude Monet, 1 Allee de Giverny - Residencias Don Andres - Piso 7 - Apartomento 74 78290, Croissy-sur-Seine, France R. do Proletariado, 1 Avenida 3, entre calles 7 y 6 Telephone: 33-13015-3737 2790-138 CARNAXIDE, Portugal Montalban, Caracas, Venezuela Facsimile: 33-13015-3747 Telephone: 351-21-424-73-00 Telephone: 58-212-443-4342 Facsimile: 351-21-418-60-20 Facsimile: 58-212-443-4342 GERMANY JEOL (GERMANY) GmbH RUSSIA VIETNAM Oskar-Von-Miller-Strasse 1a, 85386 JEOL LTD. Moscow Office TECHNICAL MATERIALS AND RESOURCES IMPORT-EXPORT JOINT STOCK Eching, Germany Krasnoproletarskaya Street, 16, COMPANY(REXCO) Telephone: 49-8165-77346 Bld. 2, Entrance 5, 127473, Moscow, Hanoi Branch, Facsimile: 49-8165-77512 Russian Federation No. 13-Lot 12 Trung Yen, Trung Hoa Street, Cau Giay Dist, Hanoi, Vietnam Telephone: 7-495-748-7791 Telephone: 84-4-562-0516,17/562-0535 GREAT BRITAIN & IRELAND Facsimile: 7-495-748-7793 Facsimile: 84-4-853-2511 JEOL (U.K.) LTD. JEOL House, Silver Court, Watchmead, SAUDI ARABIA Welwyn Garden City, Herts AL7 1LT, U.K. ABDULREHMAN ALGOSAIBI G.T.C. (Riyadh) Telephone: 44-1707-377117 King Abdulaziz Avenue, Facsimile: 44-1707-373254 P.O. Box 215, Riyadh 11411, Saudi Arabia Telephone: 966-1-479-3000 Facsimile: 966-1-477-1374

(Ser. No. 129) No. 9701G143C Printed in Japan, Kp