- Email: [email protected]
XEDS STEM tomography for 3D chemical characterization of nanoscale particles
Ultramicroscopy 131 (2013) 24–32
Contents lists available at SciVerse ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
XEDS STEM tomography for 3D chemical characterization of nanoscale particles Arda Genc a,n, Libor Kovarik b, Meng Gu b, Huikai Cheng a, Paul Plachinda a, Lee Pullan a, Bert Freitag c, Chongmin Wang b a
FEI Company, 5350 NE Dawson Creek Drive, Hillsboro, OR 97124, USA Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352, USA c FEI Company, Achtseweg Noord 5, P.O. Box 80066, 5600 KA Eindhoven, The Netherlands b
art ic l e i nf o
a b s t r a c t
Article history: Received 29 December 2012 Received in revised form 25 March 2013 Accepted 29 April 2013 Available online 11 April 2013
We present a tomography technique which couples scanning transmission electron microscopy (STEM) and X-ray energy dispersive spectrometry (XEDS) to resolve 3D distribution of elements in nanoscale materials. STEM imaging when combined with XEDS mapping using a symmetrically arranged XEDS detector design around the specimen overcomes many of the obstacles in 3D chemical imaging of nanoscale materials and successfully elucidates the 3D chemical information in a large field of view of the transmission electron microscopy (TEM) sample. We employed this technique to investigate 3D distribution of Nickel (Ni), Manganese (Mn) and Oxygen (O) in a Li1.2Ni0.2Mn0.6O2 (LNMO) nanoparticle used as a cathode material in Lithium (Li) ion batteries. For this purpose, 2D elemental maps were acquired for a range of tilt angles and reconstructed to obtain 3D elemental distribution in an isolated LNMO nanoparticle. The results highlight the strength of this technique in 3D chemical analysis of nanoscale materials by successfully resolving Ni, Mn and O elemental distributions in 3D and discovering the new phenomenon of Ni surface segregation in this material. Furthermore, the comparison of simultaneously acquired high angle annular dark field (HAADF) STEM and XEDS STEM tomography results shows that XEDS STEM tomography provides additional 3D chemical information of the material especially when there is low atomic number (Z) contrast in the material of interest. & 2013 Elsevier B.V. All rights reserved.
Keywords: Tomography XEDS STEM Silicon drift detector Li ion battery 3D chemical mapping Li1.2Ni0.2Mn0.6O2
1. Introduction TEM characterization of thin foils provides valuable information on the microstructure and chemistry of the materials at the nanometer scale. While typical thin foils used in TEM scale in the 50–100 nm thickness range, important 3D chemical and structural information of the material is often missing and convoluted within the analyzed thickness of the TEM foil. This lack of depth sensitivity in TEM limits the full characterization of nanoscale materials with complex morphology and chemistry and hinders the efforts in developing new materials with novel properties. Therefore 3D TEM tomography techniques have been of considerable interest and increasingly employed by microscopists deeply involved in characterization of nanoscale materials [1,2]. Despite the extensive use of TEM tomography, it still remains a challenge to directly observe the composition of materials in 3D using TEM. In materials science, computed tilt tomography has been utilized based on bright field (BF) TEM or HAADF STEM techniques [3].
n
Corresponding author. Tel.: +1 503 726 7500; fax: +1 503 726 7509. E-mail addresses: [email protected], [email protected] (A. Genc).
0304-3991/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ultramic.2013.03.023
For crystalline materials, BF TEM has its limitations and often does not fulfill the projection requirement due to the strong contributions from diffraction contrast with sample tilt [5]. The signal collected in HAADF STEM has a monotonic increase with sample thickness and it fulfills the projection requirement for tomographic reconstruction [4]. However, HAADF STEM tomography of low-contrast materials might be difficult such as in multiphase materials consisting of elements with similar atomic numbers and post acquisition image filtering techniques have been investigated to improve the quality of the tomograms [6]. Neither HAADF STEM nor BF TEM technique provides direct 3D chemical information of the material due to the origin of the signal collected in these techniques. 3D tomography using spectrum imaging or mapping by electron energy-loss spectrometry (EELS) and XEDS have also been pursued [7– 10]. Spectrum imaging can be achieved by using energy filters in TEM enabling fast acquisition of many images in tomography tilt series. STEM coupled with XEDS and/or EELS mapping requires long dwelling of each point and consequently much longer acquisition times in tomography tilt series. Tomography based on EELS or energy filtered TEM (EFTEM) suffers from multiple inelastic scattering artifacts and strong contributions from background signal due to variations in sample thickness [11,12]. Therefore EELS based tomography is
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
cumbersome especially when extracting signals from elements with the edges of poor jump ratios caused by the shape of the edge or sample thickness. As compared to EELS, XEDS shows a constant variation in peak to background ratio and monotonic increase of signal with thickness for high energy X-ray ionization edges [13,14]. Yet the detector geometry is a limitation in XEDS tomography due to shadowing of the holders while employing a single detector configuration. Many reports in the literature have employed specially designed holders and prepared rod shaped TEM samples to avoid detector shadowing and to improve tilt response in single detector geometry [8,9]. However, preparation of a rod shaped specimen can be very time consuming or even impossible for nanoparticle materials which require a supporting material and a grid. Researchers in materials science have also been using complementary technique of atom probe tomography (APT) to obtain chemical information of materials in 3D, but there are serious limitations on the type of samples can be utilized by this technique. APT samples need to be durable under high electrical field and always prepared in needle shape [15–17]. Therefore, sample preparation significantly constrains the wide application of APT in characterization of various materials, especially when the sample of interest is not necessarily in bulk form such as in nanoparticle materials. Here we present a XEDS STEM tomography technique to investigate the 3D distribution of elements in a nanoparticle LNMO battery cathode material. Li ion batteries are essential due to their high energy density and one of the new approaches to improve battery performance is to incorporate multivalance transition-metal ions into metal oxide cathodes by forming nanoparticle complex oxides [18,19]. The overall battery performance is strongly correlated to the role of each alloying element and their 3D distribution in the nanoparticle. In this regard, STEM tomography combined with XEDS offers a powerful capability to characterize complex morphology and composition in 3D, otherwise difficult to achieve by employing conventional 2D TEM techniques. This newly applied technique of XEDS STEM tomography overcomes many of the problems faced in 3D characterization of nanoscale materials by utilizing a symmetrically arranged XEDS detector design around the sample combined with a high brightness electron gun [20]. This design is optimized for high X-ray collection efficiency, fast XEDS mapping and improved tilt response during XEDS STEM tomography experiments [21,22].
25
ion batteries. A detailed description of the synthesis of the material was reported previously [18]. Relative atomic numbers (Z) of the consitutent atoms in Li1.2Ni0.2Mn0.6O2 are ZLi ¼3, ZNi ¼28, ZMn ¼25 and ZO ¼8. TEM sample was prepared by dropping a solution containing well-dispersed nanoparticles on a lacey Carbon film. The BF TEM image in Fig. 1a shows an overview of the as synthesized LNMO nanoparticles featuring plate-like shape and surface facets. Fig. 1b is a HAADF STEM image of an isolated nanoparticle with dimension of ∼250 nm in size which is selected for the tomography experiment.
2. Materials and methods We investigated the capability of the XEDS STEM tomography technique by examining the 3D distribution of Ni, Mn and O elements in a LNMO nanoparticle used as a cathode material in Li
Fig. 2. A schematic of the symmetrically arranged four XEDS detectors around the specimen and HAADF detector in STEM.
Fig. 1. (a) BF TEM image showing an overview of the as synthesized LNMO nanoparticles standing on a lacey Carbon film and (b) HAADF STEM image of the isolated single LNMO nanoparticle used in the tomography experiment.
26
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
A 200 kV FEI Tecnai Osiris™ TEM microscope was used for fast XEDS spectrum imaging (SI) and simultaneous HAADF STEM imaging. This system offers a high brightness electron gun, four symmetrically arranged windowless silicon drift detectors (SDD) with large solid angle (Ω¼ 0.9 sr) and improved tilt response. Fig. 2 shows a schematic indicating the geometry of HAADF and four XEDS detectors in the STEM imaging mode. Signals from all four detectors are combined into one spectrum to improve the collection efficiency. The SIs are obtained by progressive scanning of individual frames and summing multiple frames where each frame is obtained with a dwell time of 25 μs. Progressive scanning helps to reduce specimen damage as the beam is not parked on any specimen region for a long period of time. The fast mapping capability makes it possible to obtain composition maps in much shorter time periods from large specimen areas and reduces the total time needed for a tomography tilt series. A Fischione single tilt tomography holder was used in the acquisition of the HAADF and XEDS STEM tomography tilt series. Data acquisition consisted of a STEM tomography procedure where each 728 728 pixel SI was acquired for 5 min using Bruker Esprit™ software. The total acquisition time was ∼3 h which is the total duration for simultaneous HAADF and XEDS STEM tilt series acquisition including the
operations of tracking/focusing. Each tomography tilt series consisted of 29 HAADF images and XEDS elemental maps acquired at the tilt range of 7701 and tilt increment of 51. HAADF STEM images were acquired at the detector inner and outer collection angles of 40 mrad and 280 mrad. Elemental maps were acquired at beam current of 1.5 nA with 1 nm probe size and accelerating voltage of 200 kV. The semi-convergence angle (α) of the probe was 10 mrad. Processing of the tilt series for 3D analysis was carried out using a cross correlation method for image shift and tilt alignments and simultaneous iterative reconstruction technique (SIRT) for 3D reconstructions with 40 iterations in FEI Inspect3D™ software. Amira™ software was employed to generate the 3D volume rendering of the reconstructions and analysis of the volumes.
3. Results and discussion 3.1. XEDS STEM tomography of LNMO nanoparticle HAADF images and XEDS elemental maps were simultaneously acquired from an isolated single LNMO nanoparticle standing on a lacey Carbon film. Fig. 3 shows four X-ray spectra extracted from a
Fig. 3. A set of four X-ray spectra at the energy range from 0 keV to 10 keV extracted from a 01 tilt SI in the XEDS tilt series. The spectra were generated by summing of 2 2, 4 4, 6 6 and 8 8 pixel size regions in the SI revealing the change in the maximum intensities of O, Mn and Ni Kα X-ray ionization edges at 0.523 keV, 5.89 keV and 7.47 keV, respectively. The Cu Kα intensity at 8.04 keV is spurious X-rays from the grid bar.
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
27
Fig. 4. A sub-set of the Ni Kα elemental map tilt series acquired from a LNMO nanoparticle at the +701 and −701 tilt range. The symmetric design of the four XEDS detectors used in the tomography experiment allows for optimum collection angle at 01 tilt and provides sufficient XEDS signal under all tilt conditions.
01 tilt SI at the energy range between 0 keV and 10 keV in the XEDS tilt series. These spectra were obtained by summing of 2 2, 4 4, 6 6 and 8 8 pixel size regions in the SI and extracted from the same location where the line profile analysis was conducted in this work. The Cu Kα intensity at 8.04 keV is spurious X-rays from the grid bar. O, Mn and Ni Kα X-ray ionization intensities are noted at 0.523 keV, 5.89 keV and 7.47 keV energies, respectively. As seen in the series of spectra in Fig. 3, the signal collected from mapping at dwell times of few tenths of microseconds per pixel was sufficient enough to process the X-ray data for tomographic reconstructions. However, in order to minimize the effects of noise and improve the appearance of maps, O, Mn
and Ni Kα elemental map tilt series were post-processed by using a moving average filter at the 3 3 pixel range in Bruker Esprit™ software. This step also helped to achieve a successful alignment of the elemental map tilt series for 3D XEDS reconstructions. No other post processing algorithm such as any principle component analysis (PCA) was required to enhance the signal to noise ratio of the elemental maps. Fig. 4 shows a sub-set of the Ni (Ni-Kα) elemental maps acquired from a LNMO nanoparticle at the tilt range of 7701. The use of symmetrically arranged four SDDs allows tilting of the TEM sample without any shadowing effect at the positive and negative tilt directions avoiding the detection problems related to
28
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
Fig. 5. Volume rendered 3D visualizations of XEDS reconstructions. (a) 3D elemental distribution of Mn (assigned red), (b) 3D elemental distributions of Ni (assigned green), (c) composite of 3D elemental distributions of Mn and Ni, (d) 3D elemental distribution of O (assigned blue), and (e) composite of 3D elemental distributions of O and Mn.
specimen tilting. In this configuration, X-ray detection has a minimum dependence on the specimen tilt. XEDS STEM tomographic reconstructions from a LNMO nanoparticle are shown in a series of volume rendered projections in
Fig. 5a–e. 3D reconstructions were generated by processing of Mn, Ni and O elemental map tilt series of Kα X-ray ionization edges, and illustrate the 3D elemental distributions of Mn (red), Ni (green) and O (blue) in the entire nanoparticle. In general, Fig. 5
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
indicates that SIRT reconstructions yield clearly interpretable results for the 3D XEDS analysis of LNMO nanoparticle. For instance, 3D visualization of Mn in Fig. 5a shows relatively homogeneous distribution of Mn in the nanoparticle. In contrast to the Mn distribution, Ni is selectively enriched in grain boundaries and segregated at certain surface locations on the outer rim of the nanoparticle as shown in 3D elemental distribution of Ni in Fig. 5b and composite 3D elemental distributions of Ni and Mn in Fig. 5c. Preferential segregation of Ni occurs at surface orientations which favor termination of transition-metal ions and this observation is supported by the high resolution STEM analysis of the surface facets [18]. A careful examination of the volume rendered 3D visualization of Mn in Fig. 5a also reveals that Mn is locally deficient in the same locations where Ni is enriched. This competing behavior between Ni and Mn was expected since both elements occupy similar lattice sites in the trigonal α-NaFeO2 structure of the material [18,19]. As shown in Fig. 5d and e, 3D elemental distribution of O is comparably homogenous indicating no signs of any compositional enrichment in the 3D structure of the material. XEDS analysis in 3D clearly illustrates the elemental distribution of Mn, Ni and O and resolves the Ni surface segregation in the LNMO nanoparticle. What is most striking here is the selective surface segregation of Ni in the nanoparticle, which is delineated by the application of XEDS STEM tomography technique. Such surface segregation is expected to have significant implications on the Li ion transport behavior by forming a Li ion diffusion barrier and eventually affect the charge/discharge rate of the battery material [18,19].
29
3D XEDS chemical mapping brings the flexibility of extracting element specific 3D information from the material of interest and further provides a direct method for tracing and visualizing the compositional change in 3D. To get more insight of the elemental distributions in 2D and 3D, a line profile analysis of Ni and Mn intensities were carried out using 3D XEDS reconstructions and compared with those obtained from the analysis of 2D elemental maps. A schematic of the line profile analysis of Ni variation in 3D is presented in Fig. 6a. Ni and Mn line profiles were generated by integrating the intensities of individual voxels across a portion of the Ni segregated surface region in the nanoparticle as illustrated by the yellow line drawn between yellow and red spheres in Fig. 6a. The graph in Fig. 6b shows the variation in both Ni and Mn intensities in 3D. Not surprisingly, line profile analysis in Fig. 6b reveals a clear increase in Ni intensity due to the surface segregation in a distance of ∼12 nm from the nanoparticle surface, whereas Mn intensity decays rapidly in the same region. Similar analysis was conducted in 2D using elemental maps of Ni and Mn and at approximately the same location of the Ni inhomogenity in the nanoparticle and along similar direction of the 3D profile as illustrated in Fig. 6c. Interestingly, line profile analysis of 2D elemental maps of Ni and Mn (Fig. 5d) indicates a relatively much broader Ni segregated region in a distance of ∼30 nm from the nanoparticle surface, almost 3 orders of magnitude broader than 3D. This result suggests that integrated Ni and Mn XEDS intensity variations in 2D were affected by the 3D geometry of the nanoparticle in this region and convoluted by the 3D shape of the material.
Fig. 6. Line profile analysis of Ni and Mn in 2D and 3D. (a) A schematic showing the location (yellow line drawn between two spheres) of the line profile analysis carried out in 3D XEDS reconstruction of Ni, (b) the graph of Ni and Mn intensity line profiles integrated from the 3D elemental distributions, (c) a schematic showing the location of the line profile analysis (yellow line) conducted in 2D elemental map of Ni and (d) the graph of Ni and Mn intensity line profiles generated from 2D XEDS elemental maps.
30
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
Spatial resolution of these projections parallel to the tilt axis is defined by STEM resolution and proportional to the size of the electron beam, which is approximately 1 nm. Volumetric resolution of the projections perpendicular to tilt axis (parallel to electron beam) is defined by the number of tilt increments and estimated to be ∼10 nm due to the limited angular sampling frequency used in this experiment. Besides the line profile analysis using voxel intensities in 3D, it is also possible to do a quantitative 3D measurement of the composition within the nanoparticle, by analyzing the intensity of individual voxels. A quantitative 3D
chemical analysis is feasible but it must be done with great caution that absolute intensities are successfully preserved during the SIRT reconstruction process and precisely measured in 3D. At this point, a quantitative 3D XEDS analysis is beyond the scope of this paper. Another important point to consider in XEDS STEM tomography is the possibility of electron beam induced damage in the TEM samples due to the long exposure to the electron beam. As described previously, XEDS STEM maps were obtained by using very short dwell times (typically on the order of tens of microseconds) per pixel, whereas the improved XEDS detector collection
Fig. 7. Comparison of HAADF and XEDS STEM observations in 2D and 3D. (a) HAADF STEM image, (b) 2D elemental map of Ni, (c) composite of Ni and Mn 2D elemental maps, (d) composite of Ni and Mn 3D elemental distributions, (e) composite of Ni and Mn 3D elemental distributions viewed at different projections, (f) volume rendered visualization of HAADF STEM 3D reconstruction and (g) 3D HAADF STEM reconstruction viewed at different projections.
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
efficiency makes it possible to obtain composition maps in much shorter acquisition times reducing the total time needed for a tomography tilt series. The combination of very short dwelling of the electron beam and fast mapping capability significantly reduces the specimen damage by minimizing the total electron dose that sample receives during the tomography tilt series. The comparison of the images of the LNMO nanoparticle before and after the tomography tilt series have shown no apparent change in the shape and size of the nanoparticle indicating a low electron beam induced damage in the TEM sample (Supplementary Fig. 1). 3.2. Comparison of HAADF and XEDS STEM tomography of LNMO nanoparticle In order to compare the contributions from HAADF and XEDS STEM signals to the 3D projections, conventional HAADF STEM image and XEDS STEM 2D elemental maps of Ni and Mn were compared with the 3D tomograms as shown in Fig. 7a–g. Z-contrast image taken at 01 tilt in Fig. 7a shows mainly the intensity variations in the image caused by the difference in the local composition of the material. The signal collected in HAADF STEM shows dependence on atomic number through the Rutherford elastic scattering and scattered intensity varies with ∼Zn, where Z is atomic number and n is exponent (n ¼1.6–1.9, depending on the inner and outer detector angles) [23]. However, for elements with close atomic numbers, HAADF STEM images suffer from the effect of low Z-contrast. It is worth noting here that by considering a simple proportionality of the intensities in HAADF, the atomic number dependent intensity ratio between Ni and Mn is calculated to be ∼1.21 (∼ZNin/ZMnn), therefore a poor Z-contrast is expected in the LNMO system containing Ni and Mn. As seen in Fig. 7b and c, simultaneously acquired 2D Ni elemental map and a composite of Ni and Mn elemental maps of LNMO nanoparticle both feature the inhomogeneity in Ni distribution. On the other hand, this inhomogeneity is not as evident in 2D HAADF STEM image due to the poor Z contrast. 3D information from the nanoparticle is given in a series of projections in Fig. 7d–g. As shown in Fig. 7d and e, XEDS tomography provides a complete picture of the Ni distribution in the nanoparticle depicting that Ni is selectively segregated at the faceted surface locations. Fig. 7f and g show the volume rendered projections of 3D HAADF STEM reconstructions which were generated by processing of simultaneously acquired STEM image tilt series. Compared to XEDS, HAADF STEM 3D projections indicate a relatively weaker contrast from Ni segregated regions in the nanoparticle due to the poor Z-contrast in LNMO. The results from the analysis of 3D projections are consistent with the observations in 2D, suggesting that XEDS STEM tomography provides additional 3D chemical information of the material especially when there is low atomic number (Z) contrast in the material of interest.
4. Conclusions XEDS STEM tomography technique is successfully utilized to resolve 3D compositional variations of Ni, Mn and O elements in a LNMO nanoparticle used as a cathode material in Li ion batteries. In particular, the improved XEDS detector design used in this work enables rapid acquisition of XEDS tomograms while avoiding some of the limitations associated with using a single detector design. This new technique permits a larger field of view of the TEM samples and significantly reduces total acquisition time of a complete XEDS mapping tilt series. The analysis of the results from LNMO nanoparticle clearly shows that 3D XEDS STEM tomography gives new information for battery research which is not accessible by only 2D chemical
31
imaging. The XEDS signal provides enhanced chemical contrast compared to the HAADF STEM signal in tomography applications, adding a new powerful technique for 3D nanoscale chemical studies in materials science.
Acknowledgments The authors would like to thank Dr. Dapeng Wang, Ilias Belharouak and Khalil Amine at Argonne National Laboratory for the sample synthesis. Research described in this paper is part of the Chemical Imaging Initiative at Pacific Northwest National Laboratory. It was conducted under the Laboratory Directed Research and Development Program at PNNL, a multiprogram national laboratory operated by Battelle for the U.S. Department of Energy. Part of the work was conducted in the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by DOE's Office of Biological and Environmental Research and located at PNNL. PNNL is operated by Battelle for the DOE under Contract DE-AC05-76RLO1830.
Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.ultramic.2013.03.023.
References [1] I. Arslan, T.J.V. Yates, N.D. Browning, P.A. Midgley, Embedded nanostructures revealed in three dimensions, Science 309 (2005) 2195–2198. [2] P.A. Midgley, M. Weyland, 3D electron microscopy in the physical sciences: the development of Z-contrast and EFTEM tomography, Ultramicroscopy 96 (2003) 413–431. [3] M. Weyland, P.A. Midgley, T.J.V. Yates, I. Arslan, R.E. Dunin-Borkowski, J.M. Thomas, Nanoscale scanning transmission electron tomography, Journal of Microscopy 223 (2006) 185–190. [4] M. Weyland, P.A. Midgley, J.M. Thomas, Z-contrast tomography: a technique in three dimensional nanostructural analysis based on Rutherford scattering, Chemical Communications (2001) 907–908. [5] M. Weyland, P.A. Midgley, J.M. Thomas, Electron tomography of nanoparticle catalysts on porous supports: a new technique based on Rutherford scattering, Journal of Physical Chemistry B 105 (2001) 7882–7886. [6] V. Ortalan, M. Herrera, D.G. Morgan, N.D. Browning, Application of image processing to STEM tomography of low-contrast materials, Ultramicroscopy 110 (2009) 67–81. [7] M.A. Aronova, Y.C. Kim, R. Harmon, A.A. Sousa, G. Zhang, R.D. Leapman, Threedimensional elemental mapping of phosphorus by quantitative electron spectroscopic tomography (QuEST), Journal of Structural Biology 161 (2008) 322–353. [8] K. Jarausch, P. Thomas, D.N. Leonard, R. Twesten, C.R. Booth, Four-dimensional STEM-EELS: enabling nano-scale chemical tomography, Ultramicroscopy 109 (2009) 326–337. [9] T. Yaguchi, M. Konno, T. Kamino, M. Watanabe, Observation of threedimensional elemental distributions of a Si device using a 3601-tilt FIB and the cold field-emission STEM system, Ultramicroscopy 108 (2008) 1603–1615. [10] G. Mobus, R.C. Doole, B.J. Inkson, Spectroscopic electron tomography, Ultramicroscopy 96 (2003) 433–451. [11] L. Yedra, A. Eljarrat, R. Arenal, E. Pellicer, M. Cabo, A. Lopez-Ortega, M. Estrader, J. Sort, M. Dolors BaroS. EstradeF. Peiro, EEL spectroscopic tomography: towards a new dimension in nanomaterials analysis, Ultramicroscopy 122 (2012) 12–18. [12] N.Y. Jin-Phillipp, C.T. Koch, P.A. van Aken, Towards quantitative core-loss EFTEM tomography, Ultramicroscopy 111 (2011) 1255–1261. [13] A. Genc, H. Cheng, J. Winterstein, L. Pullan, B. Freitag, 3D chemical mapping using tomography with an enhanced XEDS system, Microscopy and Analysis 116 (2012) 23–25. [14] D. Huber, H.L. Fraser, D.O. Klenov, H.S. von Harrach, N.J. Zaluzec, Relative sensitivity of XEDS vs EELS in the AEM, Late Breaking Poster Microscopy and Microanalysis Suppl. 2 (2010) 16. [15] T.F. Kelly, D.J. Larson, Atom probe tomography 2012, Annual Review Materials Research 42 (2012) 1–31. [16] M.K. Miller, K.F. Russell, Atom probe specimen preparation with a dual beam SEM/FIB miller, Ultramicroscopy 107 (2007) 761–766. [17] M.K. Miller, K.F. Russell, G.B. Thompson, Strategies for fabricating atom probe specimens with a dual beam FIB, Ultramicroscopy 102 (2005) 287–298.
32
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
[18] M. Gu, I. Belharouak, A. Genc, Z. Wang, D. Wang, K. Amine, F. Gao, G. Zhou, S. Thevuthasan, D.R. Baer, J.G. Zhang, N.D. Browning, J. Liu, C. Wang, Conflicting roles of nickel in controlling cathode performance in lithium ion batteries, Nano Letters 12 (2012) 5186–5191. [19] M. Gu, I. Belharouak, J. Zheng, H. Wu, J. Xiao, A Genc, K. Amine, S. Thevuthasan, D.R. Baer, J.G. Zhang, N.D. Browning, J. Liu, C. Wang, Formation of the spinel phase in the layered composite cathode used in Li-ion batteries, Nano (2012), http://dx.doi.org/10.1021/nn305065u. [20] P. Schlossmacher, D.O. Klenov, B. Freitag, H.S. von Harrach, Enhanced detection sensitivity with a new windowless XEDS system for AEM based on silicon drift detector technology, Microscopy Today 18 (2010) 14–20.
[21] P. Schlossmacher, D.O. Klenov, B. Freitag, H.S. von Harrach, A. Steinbach, Nanoscale chemical compositional analysis with an innovative S/TEM-EDX system, Microscopy and Analysis (Nanotechnology Suppl.) 24 (2010) S5–S8. [22] L.J. Allen, A.J. D’Alfonso, B. Freitag, D.O. Klenov, Chemical mapping at atomic resolution using energy-dispersive x-ray spectroscopy, MRS Bulletin 37 (2012) 47–52. [23] P. Hartel, H. Rose, C. Dinges, Conditions and reasons for incoherent imaging in STEM, Ultramicroscopy 63 (1996) 93–114.
Contents lists available at SciVerse ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
XEDS STEM tomography for 3D chemical characterization of nanoscale particles Arda Genc a,n, Libor Kovarik b, Meng Gu b, Huikai Cheng a, Paul Plachinda a, Lee Pullan a, Bert Freitag c, Chongmin Wang b a
FEI Company, 5350 NE Dawson Creek Drive, Hillsboro, OR 97124, USA Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352, USA c FEI Company, Achtseweg Noord 5, P.O. Box 80066, 5600 KA Eindhoven, The Netherlands b
art ic l e i nf o
a b s t r a c t
Article history: Received 29 December 2012 Received in revised form 25 March 2013 Accepted 29 April 2013 Available online 11 April 2013
We present a tomography technique which couples scanning transmission electron microscopy (STEM) and X-ray energy dispersive spectrometry (XEDS) to resolve 3D distribution of elements in nanoscale materials. STEM imaging when combined with XEDS mapping using a symmetrically arranged XEDS detector design around the specimen overcomes many of the obstacles in 3D chemical imaging of nanoscale materials and successfully elucidates the 3D chemical information in a large field of view of the transmission electron microscopy (TEM) sample. We employed this technique to investigate 3D distribution of Nickel (Ni), Manganese (Mn) and Oxygen (O) in a Li1.2Ni0.2Mn0.6O2 (LNMO) nanoparticle used as a cathode material in Lithium (Li) ion batteries. For this purpose, 2D elemental maps were acquired for a range of tilt angles and reconstructed to obtain 3D elemental distribution in an isolated LNMO nanoparticle. The results highlight the strength of this technique in 3D chemical analysis of nanoscale materials by successfully resolving Ni, Mn and O elemental distributions in 3D and discovering the new phenomenon of Ni surface segregation in this material. Furthermore, the comparison of simultaneously acquired high angle annular dark field (HAADF) STEM and XEDS STEM tomography results shows that XEDS STEM tomography provides additional 3D chemical information of the material especially when there is low atomic number (Z) contrast in the material of interest. & 2013 Elsevier B.V. All rights reserved.
Keywords: Tomography XEDS STEM Silicon drift detector Li ion battery 3D chemical mapping Li1.2Ni0.2Mn0.6O2
1. Introduction TEM characterization of thin foils provides valuable information on the microstructure and chemistry of the materials at the nanometer scale. While typical thin foils used in TEM scale in the 50–100 nm thickness range, important 3D chemical and structural information of the material is often missing and convoluted within the analyzed thickness of the TEM foil. This lack of depth sensitivity in TEM limits the full characterization of nanoscale materials with complex morphology and chemistry and hinders the efforts in developing new materials with novel properties. Therefore 3D TEM tomography techniques have been of considerable interest and increasingly employed by microscopists deeply involved in characterization of nanoscale materials [1,2]. Despite the extensive use of TEM tomography, it still remains a challenge to directly observe the composition of materials in 3D using TEM. In materials science, computed tilt tomography has been utilized based on bright field (BF) TEM or HAADF STEM techniques [3].
n
Corresponding author. Tel.: +1 503 726 7500; fax: +1 503 726 7509. E-mail addresses: [email protected], [email protected] (A. Genc).
0304-3991/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ultramic.2013.03.023
For crystalline materials, BF TEM has its limitations and often does not fulfill the projection requirement due to the strong contributions from diffraction contrast with sample tilt [5]. The signal collected in HAADF STEM has a monotonic increase with sample thickness and it fulfills the projection requirement for tomographic reconstruction [4]. However, HAADF STEM tomography of low-contrast materials might be difficult such as in multiphase materials consisting of elements with similar atomic numbers and post acquisition image filtering techniques have been investigated to improve the quality of the tomograms [6]. Neither HAADF STEM nor BF TEM technique provides direct 3D chemical information of the material due to the origin of the signal collected in these techniques. 3D tomography using spectrum imaging or mapping by electron energy-loss spectrometry (EELS) and XEDS have also been pursued [7– 10]. Spectrum imaging can be achieved by using energy filters in TEM enabling fast acquisition of many images in tomography tilt series. STEM coupled with XEDS and/or EELS mapping requires long dwelling of each point and consequently much longer acquisition times in tomography tilt series. Tomography based on EELS or energy filtered TEM (EFTEM) suffers from multiple inelastic scattering artifacts and strong contributions from background signal due to variations in sample thickness [11,12]. Therefore EELS based tomography is
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
cumbersome especially when extracting signals from elements with the edges of poor jump ratios caused by the shape of the edge or sample thickness. As compared to EELS, XEDS shows a constant variation in peak to background ratio and monotonic increase of signal with thickness for high energy X-ray ionization edges [13,14]. Yet the detector geometry is a limitation in XEDS tomography due to shadowing of the holders while employing a single detector configuration. Many reports in the literature have employed specially designed holders and prepared rod shaped TEM samples to avoid detector shadowing and to improve tilt response in single detector geometry [8,9]. However, preparation of a rod shaped specimen can be very time consuming or even impossible for nanoparticle materials which require a supporting material and a grid. Researchers in materials science have also been using complementary technique of atom probe tomography (APT) to obtain chemical information of materials in 3D, but there are serious limitations on the type of samples can be utilized by this technique. APT samples need to be durable under high electrical field and always prepared in needle shape [15–17]. Therefore, sample preparation significantly constrains the wide application of APT in characterization of various materials, especially when the sample of interest is not necessarily in bulk form such as in nanoparticle materials. Here we present a XEDS STEM tomography technique to investigate the 3D distribution of elements in a nanoparticle LNMO battery cathode material. Li ion batteries are essential due to their high energy density and one of the new approaches to improve battery performance is to incorporate multivalance transition-metal ions into metal oxide cathodes by forming nanoparticle complex oxides [18,19]. The overall battery performance is strongly correlated to the role of each alloying element and their 3D distribution in the nanoparticle. In this regard, STEM tomography combined with XEDS offers a powerful capability to characterize complex morphology and composition in 3D, otherwise difficult to achieve by employing conventional 2D TEM techniques. This newly applied technique of XEDS STEM tomography overcomes many of the problems faced in 3D characterization of nanoscale materials by utilizing a symmetrically arranged XEDS detector design around the sample combined with a high brightness electron gun [20]. This design is optimized for high X-ray collection efficiency, fast XEDS mapping and improved tilt response during XEDS STEM tomography experiments [21,22].
25
ion batteries. A detailed description of the synthesis of the material was reported previously [18]. Relative atomic numbers (Z) of the consitutent atoms in Li1.2Ni0.2Mn0.6O2 are ZLi ¼3, ZNi ¼28, ZMn ¼25 and ZO ¼8. TEM sample was prepared by dropping a solution containing well-dispersed nanoparticles on a lacey Carbon film. The BF TEM image in Fig. 1a shows an overview of the as synthesized LNMO nanoparticles featuring plate-like shape and surface facets. Fig. 1b is a HAADF STEM image of an isolated nanoparticle with dimension of ∼250 nm in size which is selected for the tomography experiment.
2. Materials and methods We investigated the capability of the XEDS STEM tomography technique by examining the 3D distribution of Ni, Mn and O elements in a LNMO nanoparticle used as a cathode material in Li
Fig. 2. A schematic of the symmetrically arranged four XEDS detectors around the specimen and HAADF detector in STEM.
Fig. 1. (a) BF TEM image showing an overview of the as synthesized LNMO nanoparticles standing on a lacey Carbon film and (b) HAADF STEM image of the isolated single LNMO nanoparticle used in the tomography experiment.
26
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
A 200 kV FEI Tecnai Osiris™ TEM microscope was used for fast XEDS spectrum imaging (SI) and simultaneous HAADF STEM imaging. This system offers a high brightness electron gun, four symmetrically arranged windowless silicon drift detectors (SDD) with large solid angle (Ω¼ 0.9 sr) and improved tilt response. Fig. 2 shows a schematic indicating the geometry of HAADF and four XEDS detectors in the STEM imaging mode. Signals from all four detectors are combined into one spectrum to improve the collection efficiency. The SIs are obtained by progressive scanning of individual frames and summing multiple frames where each frame is obtained with a dwell time of 25 μs. Progressive scanning helps to reduce specimen damage as the beam is not parked on any specimen region for a long period of time. The fast mapping capability makes it possible to obtain composition maps in much shorter time periods from large specimen areas and reduces the total time needed for a tomography tilt series. A Fischione single tilt tomography holder was used in the acquisition of the HAADF and XEDS STEM tomography tilt series. Data acquisition consisted of a STEM tomography procedure where each 728 728 pixel SI was acquired for 5 min using Bruker Esprit™ software. The total acquisition time was ∼3 h which is the total duration for simultaneous HAADF and XEDS STEM tilt series acquisition including the
operations of tracking/focusing. Each tomography tilt series consisted of 29 HAADF images and XEDS elemental maps acquired at the tilt range of 7701 and tilt increment of 51. HAADF STEM images were acquired at the detector inner and outer collection angles of 40 mrad and 280 mrad. Elemental maps were acquired at beam current of 1.5 nA with 1 nm probe size and accelerating voltage of 200 kV. The semi-convergence angle (α) of the probe was 10 mrad. Processing of the tilt series for 3D analysis was carried out using a cross correlation method for image shift and tilt alignments and simultaneous iterative reconstruction technique (SIRT) for 3D reconstructions with 40 iterations in FEI Inspect3D™ software. Amira™ software was employed to generate the 3D volume rendering of the reconstructions and analysis of the volumes.
3. Results and discussion 3.1. XEDS STEM tomography of LNMO nanoparticle HAADF images and XEDS elemental maps were simultaneously acquired from an isolated single LNMO nanoparticle standing on a lacey Carbon film. Fig. 3 shows four X-ray spectra extracted from a
Fig. 3. A set of four X-ray spectra at the energy range from 0 keV to 10 keV extracted from a 01 tilt SI in the XEDS tilt series. The spectra were generated by summing of 2 2, 4 4, 6 6 and 8 8 pixel size regions in the SI revealing the change in the maximum intensities of O, Mn and Ni Kα X-ray ionization edges at 0.523 keV, 5.89 keV and 7.47 keV, respectively. The Cu Kα intensity at 8.04 keV is spurious X-rays from the grid bar.
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
27
Fig. 4. A sub-set of the Ni Kα elemental map tilt series acquired from a LNMO nanoparticle at the +701 and −701 tilt range. The symmetric design of the four XEDS detectors used in the tomography experiment allows for optimum collection angle at 01 tilt and provides sufficient XEDS signal under all tilt conditions.
01 tilt SI at the energy range between 0 keV and 10 keV in the XEDS tilt series. These spectra were obtained by summing of 2 2, 4 4, 6 6 and 8 8 pixel size regions in the SI and extracted from the same location where the line profile analysis was conducted in this work. The Cu Kα intensity at 8.04 keV is spurious X-rays from the grid bar. O, Mn and Ni Kα X-ray ionization intensities are noted at 0.523 keV, 5.89 keV and 7.47 keV energies, respectively. As seen in the series of spectra in Fig. 3, the signal collected from mapping at dwell times of few tenths of microseconds per pixel was sufficient enough to process the X-ray data for tomographic reconstructions. However, in order to minimize the effects of noise and improve the appearance of maps, O, Mn
and Ni Kα elemental map tilt series were post-processed by using a moving average filter at the 3 3 pixel range in Bruker Esprit™ software. This step also helped to achieve a successful alignment of the elemental map tilt series for 3D XEDS reconstructions. No other post processing algorithm such as any principle component analysis (PCA) was required to enhance the signal to noise ratio of the elemental maps. Fig. 4 shows a sub-set of the Ni (Ni-Kα) elemental maps acquired from a LNMO nanoparticle at the tilt range of 7701. The use of symmetrically arranged four SDDs allows tilting of the TEM sample without any shadowing effect at the positive and negative tilt directions avoiding the detection problems related to
28
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
Fig. 5. Volume rendered 3D visualizations of XEDS reconstructions. (a) 3D elemental distribution of Mn (assigned red), (b) 3D elemental distributions of Ni (assigned green), (c) composite of 3D elemental distributions of Mn and Ni, (d) 3D elemental distribution of O (assigned blue), and (e) composite of 3D elemental distributions of O and Mn.
specimen tilting. In this configuration, X-ray detection has a minimum dependence on the specimen tilt. XEDS STEM tomographic reconstructions from a LNMO nanoparticle are shown in a series of volume rendered projections in
Fig. 5a–e. 3D reconstructions were generated by processing of Mn, Ni and O elemental map tilt series of Kα X-ray ionization edges, and illustrate the 3D elemental distributions of Mn (red), Ni (green) and O (blue) in the entire nanoparticle. In general, Fig. 5
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
indicates that SIRT reconstructions yield clearly interpretable results for the 3D XEDS analysis of LNMO nanoparticle. For instance, 3D visualization of Mn in Fig. 5a shows relatively homogeneous distribution of Mn in the nanoparticle. In contrast to the Mn distribution, Ni is selectively enriched in grain boundaries and segregated at certain surface locations on the outer rim of the nanoparticle as shown in 3D elemental distribution of Ni in Fig. 5b and composite 3D elemental distributions of Ni and Mn in Fig. 5c. Preferential segregation of Ni occurs at surface orientations which favor termination of transition-metal ions and this observation is supported by the high resolution STEM analysis of the surface facets [18]. A careful examination of the volume rendered 3D visualization of Mn in Fig. 5a also reveals that Mn is locally deficient in the same locations where Ni is enriched. This competing behavior between Ni and Mn was expected since both elements occupy similar lattice sites in the trigonal α-NaFeO2 structure of the material [18,19]. As shown in Fig. 5d and e, 3D elemental distribution of O is comparably homogenous indicating no signs of any compositional enrichment in the 3D structure of the material. XEDS analysis in 3D clearly illustrates the elemental distribution of Mn, Ni and O and resolves the Ni surface segregation in the LNMO nanoparticle. What is most striking here is the selective surface segregation of Ni in the nanoparticle, which is delineated by the application of XEDS STEM tomography technique. Such surface segregation is expected to have significant implications on the Li ion transport behavior by forming a Li ion diffusion barrier and eventually affect the charge/discharge rate of the battery material [18,19].
29
3D XEDS chemical mapping brings the flexibility of extracting element specific 3D information from the material of interest and further provides a direct method for tracing and visualizing the compositional change in 3D. To get more insight of the elemental distributions in 2D and 3D, a line profile analysis of Ni and Mn intensities were carried out using 3D XEDS reconstructions and compared with those obtained from the analysis of 2D elemental maps. A schematic of the line profile analysis of Ni variation in 3D is presented in Fig. 6a. Ni and Mn line profiles were generated by integrating the intensities of individual voxels across a portion of the Ni segregated surface region in the nanoparticle as illustrated by the yellow line drawn between yellow and red spheres in Fig. 6a. The graph in Fig. 6b shows the variation in both Ni and Mn intensities in 3D. Not surprisingly, line profile analysis in Fig. 6b reveals a clear increase in Ni intensity due to the surface segregation in a distance of ∼12 nm from the nanoparticle surface, whereas Mn intensity decays rapidly in the same region. Similar analysis was conducted in 2D using elemental maps of Ni and Mn and at approximately the same location of the Ni inhomogenity in the nanoparticle and along similar direction of the 3D profile as illustrated in Fig. 6c. Interestingly, line profile analysis of 2D elemental maps of Ni and Mn (Fig. 5d) indicates a relatively much broader Ni segregated region in a distance of ∼30 nm from the nanoparticle surface, almost 3 orders of magnitude broader than 3D. This result suggests that integrated Ni and Mn XEDS intensity variations in 2D were affected by the 3D geometry of the nanoparticle in this region and convoluted by the 3D shape of the material.
Fig. 6. Line profile analysis of Ni and Mn in 2D and 3D. (a) A schematic showing the location (yellow line drawn between two spheres) of the line profile analysis carried out in 3D XEDS reconstruction of Ni, (b) the graph of Ni and Mn intensity line profiles integrated from the 3D elemental distributions, (c) a schematic showing the location of the line profile analysis (yellow line) conducted in 2D elemental map of Ni and (d) the graph of Ni and Mn intensity line profiles generated from 2D XEDS elemental maps.
30
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
Spatial resolution of these projections parallel to the tilt axis is defined by STEM resolution and proportional to the size of the electron beam, which is approximately 1 nm. Volumetric resolution of the projections perpendicular to tilt axis (parallel to electron beam) is defined by the number of tilt increments and estimated to be ∼10 nm due to the limited angular sampling frequency used in this experiment. Besides the line profile analysis using voxel intensities in 3D, it is also possible to do a quantitative 3D measurement of the composition within the nanoparticle, by analyzing the intensity of individual voxels. A quantitative 3D
chemical analysis is feasible but it must be done with great caution that absolute intensities are successfully preserved during the SIRT reconstruction process and precisely measured in 3D. At this point, a quantitative 3D XEDS analysis is beyond the scope of this paper. Another important point to consider in XEDS STEM tomography is the possibility of electron beam induced damage in the TEM samples due to the long exposure to the electron beam. As described previously, XEDS STEM maps were obtained by using very short dwell times (typically on the order of tens of microseconds) per pixel, whereas the improved XEDS detector collection
Fig. 7. Comparison of HAADF and XEDS STEM observations in 2D and 3D. (a) HAADF STEM image, (b) 2D elemental map of Ni, (c) composite of Ni and Mn 2D elemental maps, (d) composite of Ni and Mn 3D elemental distributions, (e) composite of Ni and Mn 3D elemental distributions viewed at different projections, (f) volume rendered visualization of HAADF STEM 3D reconstruction and (g) 3D HAADF STEM reconstruction viewed at different projections.
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
efficiency makes it possible to obtain composition maps in much shorter acquisition times reducing the total time needed for a tomography tilt series. The combination of very short dwelling of the electron beam and fast mapping capability significantly reduces the specimen damage by minimizing the total electron dose that sample receives during the tomography tilt series. The comparison of the images of the LNMO nanoparticle before and after the tomography tilt series have shown no apparent change in the shape and size of the nanoparticle indicating a low electron beam induced damage in the TEM sample (Supplementary Fig. 1). 3.2. Comparison of HAADF and XEDS STEM tomography of LNMO nanoparticle In order to compare the contributions from HAADF and XEDS STEM signals to the 3D projections, conventional HAADF STEM image and XEDS STEM 2D elemental maps of Ni and Mn were compared with the 3D tomograms as shown in Fig. 7a–g. Z-contrast image taken at 01 tilt in Fig. 7a shows mainly the intensity variations in the image caused by the difference in the local composition of the material. The signal collected in HAADF STEM shows dependence on atomic number through the Rutherford elastic scattering and scattered intensity varies with ∼Zn, where Z is atomic number and n is exponent (n ¼1.6–1.9, depending on the inner and outer detector angles) [23]. However, for elements with close atomic numbers, HAADF STEM images suffer from the effect of low Z-contrast. It is worth noting here that by considering a simple proportionality of the intensities in HAADF, the atomic number dependent intensity ratio between Ni and Mn is calculated to be ∼1.21 (∼ZNin/ZMnn), therefore a poor Z-contrast is expected in the LNMO system containing Ni and Mn. As seen in Fig. 7b and c, simultaneously acquired 2D Ni elemental map and a composite of Ni and Mn elemental maps of LNMO nanoparticle both feature the inhomogeneity in Ni distribution. On the other hand, this inhomogeneity is not as evident in 2D HAADF STEM image due to the poor Z contrast. 3D information from the nanoparticle is given in a series of projections in Fig. 7d–g. As shown in Fig. 7d and e, XEDS tomography provides a complete picture of the Ni distribution in the nanoparticle depicting that Ni is selectively segregated at the faceted surface locations. Fig. 7f and g show the volume rendered projections of 3D HAADF STEM reconstructions which were generated by processing of simultaneously acquired STEM image tilt series. Compared to XEDS, HAADF STEM 3D projections indicate a relatively weaker contrast from Ni segregated regions in the nanoparticle due to the poor Z-contrast in LNMO. The results from the analysis of 3D projections are consistent with the observations in 2D, suggesting that XEDS STEM tomography provides additional 3D chemical information of the material especially when there is low atomic number (Z) contrast in the material of interest.
4. Conclusions XEDS STEM tomography technique is successfully utilized to resolve 3D compositional variations of Ni, Mn and O elements in a LNMO nanoparticle used as a cathode material in Li ion batteries. In particular, the improved XEDS detector design used in this work enables rapid acquisition of XEDS tomograms while avoiding some of the limitations associated with using a single detector design. This new technique permits a larger field of view of the TEM samples and significantly reduces total acquisition time of a complete XEDS mapping tilt series. The analysis of the results from LNMO nanoparticle clearly shows that 3D XEDS STEM tomography gives new information for battery research which is not accessible by only 2D chemical
31
imaging. The XEDS signal provides enhanced chemical contrast compared to the HAADF STEM signal in tomography applications, adding a new powerful technique for 3D nanoscale chemical studies in materials science.
Acknowledgments The authors would like to thank Dr. Dapeng Wang, Ilias Belharouak and Khalil Amine at Argonne National Laboratory for the sample synthesis. Research described in this paper is part of the Chemical Imaging Initiative at Pacific Northwest National Laboratory. It was conducted under the Laboratory Directed Research and Development Program at PNNL, a multiprogram national laboratory operated by Battelle for the U.S. Department of Energy. Part of the work was conducted in the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by DOE's Office of Biological and Environmental Research and located at PNNL. PNNL is operated by Battelle for the DOE under Contract DE-AC05-76RLO1830.
Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.ultramic.2013.03.023.
References [1] I. Arslan, T.J.V. Yates, N.D. Browning, P.A. Midgley, Embedded nanostructures revealed in three dimensions, Science 309 (2005) 2195–2198. [2] P.A. Midgley, M. Weyland, 3D electron microscopy in the physical sciences: the development of Z-contrast and EFTEM tomography, Ultramicroscopy 96 (2003) 413–431. [3] M. Weyland, P.A. Midgley, T.J.V. Yates, I. Arslan, R.E. Dunin-Borkowski, J.M. Thomas, Nanoscale scanning transmission electron tomography, Journal of Microscopy 223 (2006) 185–190. [4] M. Weyland, P.A. Midgley, J.M. Thomas, Z-contrast tomography: a technique in three dimensional nanostructural analysis based on Rutherford scattering, Chemical Communications (2001) 907–908. [5] M. Weyland, P.A. Midgley, J.M. Thomas, Electron tomography of nanoparticle catalysts on porous supports: a new technique based on Rutherford scattering, Journal of Physical Chemistry B 105 (2001) 7882–7886. [6] V. Ortalan, M. Herrera, D.G. Morgan, N.D. Browning, Application of image processing to STEM tomography of low-contrast materials, Ultramicroscopy 110 (2009) 67–81. [7] M.A. Aronova, Y.C. Kim, R. Harmon, A.A. Sousa, G. Zhang, R.D. Leapman, Threedimensional elemental mapping of phosphorus by quantitative electron spectroscopic tomography (QuEST), Journal of Structural Biology 161 (2008) 322–353. [8] K. Jarausch, P. Thomas, D.N. Leonard, R. Twesten, C.R. Booth, Four-dimensional STEM-EELS: enabling nano-scale chemical tomography, Ultramicroscopy 109 (2009) 326–337. [9] T. Yaguchi, M. Konno, T. Kamino, M. Watanabe, Observation of threedimensional elemental distributions of a Si device using a 3601-tilt FIB and the cold field-emission STEM system, Ultramicroscopy 108 (2008) 1603–1615. [10] G. Mobus, R.C. Doole, B.J. Inkson, Spectroscopic electron tomography, Ultramicroscopy 96 (2003) 433–451. [11] L. Yedra, A. Eljarrat, R. Arenal, E. Pellicer, M. Cabo, A. Lopez-Ortega, M. Estrader, J. Sort, M. Dolors BaroS. EstradeF. Peiro, EEL spectroscopic tomography: towards a new dimension in nanomaterials analysis, Ultramicroscopy 122 (2012) 12–18. [12] N.Y. Jin-Phillipp, C.T. Koch, P.A. van Aken, Towards quantitative core-loss EFTEM tomography, Ultramicroscopy 111 (2011) 1255–1261. [13] A. Genc, H. Cheng, J. Winterstein, L. Pullan, B. Freitag, 3D chemical mapping using tomography with an enhanced XEDS system, Microscopy and Analysis 116 (2012) 23–25. [14] D. Huber, H.L. Fraser, D.O. Klenov, H.S. von Harrach, N.J. Zaluzec, Relative sensitivity of XEDS vs EELS in the AEM, Late Breaking Poster Microscopy and Microanalysis Suppl. 2 (2010) 16. [15] T.F. Kelly, D.J. Larson, Atom probe tomography 2012, Annual Review Materials Research 42 (2012) 1–31. [16] M.K. Miller, K.F. Russell, Atom probe specimen preparation with a dual beam SEM/FIB miller, Ultramicroscopy 107 (2007) 761–766. [17] M.K. Miller, K.F. Russell, G.B. Thompson, Strategies for fabricating atom probe specimens with a dual beam FIB, Ultramicroscopy 102 (2005) 287–298.
32
A. Genc et al. / Ultramicroscopy 131 (2013) 24–32
[18] M. Gu, I. Belharouak, A. Genc, Z. Wang, D. Wang, K. Amine, F. Gao, G. Zhou, S. Thevuthasan, D.R. Baer, J.G. Zhang, N.D. Browning, J. Liu, C. Wang, Conflicting roles of nickel in controlling cathode performance in lithium ion batteries, Nano Letters 12 (2012) 5186–5191. [19] M. Gu, I. Belharouak, J. Zheng, H. Wu, J. Xiao, A Genc, K. Amine, S. Thevuthasan, D.R. Baer, J.G. Zhang, N.D. Browning, J. Liu, C. Wang, Formation of the spinel phase in the layered composite cathode used in Li-ion batteries, Nano (2012), http://dx.doi.org/10.1021/nn305065u. [20] P. Schlossmacher, D.O. Klenov, B. Freitag, H.S. von Harrach, Enhanced detection sensitivity with a new windowless XEDS system for AEM based on silicon drift detector technology, Microscopy Today 18 (2010) 14–20.
[21] P. Schlossmacher, D.O. Klenov, B. Freitag, H.S. von Harrach, A. Steinbach, Nanoscale chemical compositional analysis with an innovative S/TEM-EDX system, Microscopy and Analysis (Nanotechnology Suppl.) 24 (2010) S5–S8. [22] L.J. Allen, A.J. D’Alfonso, B. Freitag, D.O. Klenov, Chemical mapping at atomic resolution using energy-dispersive x-ray spectroscopy, MRS Bulletin 37 (2012) 47–52. [23] P. Hartel, H. Rose, C. Dinges, Conditions and reasons for incoherent imaging in STEM, Ultramicroscopy 63 (1996) 93–114.