Analytical and in situ Applications Using Aberration Corrected Scanning Transmission Electron Microscope (cid:3)

Sensitivity of energy dispersive X-ray spectroscopy (EDS) in scanning transmission electron microscopy (STEM) has been signi(cid:12)cantly improved with a recent detection system with large-sized silicon drift detectors (SDDs). The detection solid angle for the latest system, composed of two windowless SDDs, is better than 2 steradians. In combination with a spherical aberration corrector for STEM, the system allows us to observe an atomic resolution elemental map. On the other hand, in situ experiments under various conditions such as heating, cooling, and gas environments have been conducted by using chip-based specimen holders, which are developed owing to fabrication technology of micro electro mechanical systems (MEMS). The aberration corrected microscopy with such a special holder allows us to observe morphological and chemical changes of samples under various conditions at atomic resolution. In this paper, the observation and analysis results of in situ applications is reported by using the aberration corrected microscope with the EDS system and the in situ sample holders. [DOI: 10.1380/ejssnt.2018.286]


I. INTRODUCTION
Analytical capability with an energy-dispersive X-ray spectroscopy (EDS) in a transmission electron microscope (TEM) has been dramatically improved by using a recent silicon drift detector (SDD). Highly flexible size and shape of the sensor allow us to place the multiple large-sized detectors around an objective lens pole-piece at optimized positions for increasing X-ray detection efficiency [1].
Recently, we have developed a new highly efficient Xray detecting system for a spherical aberration corrected 300 kV TEM (JEM-ARM300F, JEOL Ltd.) with a cold field emission electron gun [2][3][4][5]. In the combination of a spherical aberration corrected scanning transmission electron microscopy (STEM), we can make atomic resolution elemental maps with high definition. Moreover, since the system allows chemical analysis in a short acquisition time, the system enables us to analyze electron beam sensitive materials.
Meanwhile, in situ experiments by using TEM under various conditions such as heating, cooling, and gas environments have became popular and important for the fields of material science, since they enable us to observe the on-going chemical reaction or phase transition of nanometer-sized samples under the conditions close to actual environments of samples. To realize in situ experiments in TEM, two ways have been proposed and conducted so far. One is to use a dedicated microscope, other is to use a special sample holder with reaction cell. In the latter, recently the chip-based sample holders have been developed. The chips were designed and developed utilizing micro electro mechanical systems (MEMS) technology. They provide highly stable and reproducible exper- * This paper was presented at the 11th International Symposium iments at atomic resolution with an aberration corrected S/TEM. In this paper, we report the performance of analysis by the JEM-ARM300F with the highly efficient X-ray detection system. And we show in situ observation and its analytical results by the analytical microscope with the MEMS-based gas-cell type sample holder.

II. ENERGY DISPERSIVE X-RAY SPECTROSCOPY
The X-ray detection system for JEM-ARM300F with a newly designed wide gap objective lens pole-piece (WGP) consists of two windowless SDDs (SDD1 and SDD2). The SDD1 is located in the right side of the specimen holder rod, and the SDD2 is on the sample holder rod axis. The sensor sizes of these detectors are 158 mm 2 in area. These large detectors are set closer to the analytical sample holder with thin width between the wide gap of the WGP for increasing X-ray detection solid angle. The solid angles of SDD1 and SDD2 are 1.106 sr and 1.108 sr, respectively. The total solid angle in the system reaches 2.214 sr. The take-off angles of SDD1 and SDD2 are as high as 30.45 • and 30.65 • , which minimize shadowing effect of sample holder and bumps of sample surface, resulting in higher signal. Figure 1 shows a comparison of analytical performance for the new X-ray detection system and previous one, which consists of JEM-ARM300F with WGP and two SDDs with 100 mm 2 -sized sensors. The gross intensity of Ni Kα line for NiO x thin film sample (Ted Pella Inc.) was measured by using the both system under the same analytical conditions: acceleration voltage = 300 kV, probe current = 500 pA, and X-ray acquisition time = 100 s. In the case of SDD2, the intensity of Ni Kα line showed similar value in 100 mm 2 -and 158 mm 2 -SDDs. On the other hand, in the case of SDD1, the intensity of Ni Kα line for 158 mm 2 -SDD showed about 2 times higher value than that for 100 mm 2 -SDD. Eventually, total intensity of SDD1+SDD2 for the new X-ray detection system with 158 mm 2 -SDDs showed about 1.3 times higher than that for the previous system with 100 mm 2 -SDDs. Figure 2 shows atomic resolution X-ray maps for SrTiO 3 [100], obtained at an acceleration voltage of 80 kV. The mapping size was 1024 × 1024 pixels. The probe current was approximately 32 pA. The acquisition time of X-ray was approximately 10 min. Since the numbers of pixels for the X-ray maps were very large, we can recognize each atomic column even in enlarged maps. The new X-ray detection system enables us to obtain such high definition maps with realistic measurement time. Figure 3 shows elemental maps of Pt/Pd catalysis particles (∼15 nm in diameter), observed at 200 kV. Probe current and acquisition time were approximately 12 pA and 3 min. The particle shows core/shell structure, that is, the core is made of Pd, and the shell is Pt at the beginning of the experiment. However, the particles become homogeneous alloy structure, as they suffer damage by electron beam irradiation. The elemental maps by the X-ray detection system reduces damage and enables us to see the core/shell structure showing thin shell layer of just 3 atomic layer thickness in the acquired elemental maps.

III. IN SIT U EXPERIMENTS
We have conducted in situ experiments under gas environment by using the MEMS-based gas-cell type sample holder (Atmosphere 200, Protochips Inc.). The combination of JEM-ARM300F and Atmosphere 200 provides highly stable results of in situ experiments even under high gas pressure conditions up to 1 atmosphere (10 5 Pa) at high temperature up to 1000 • C. Figure 4 shows a high angular annular dark field (HAADF) STEM image of Ti oxide (anatase) [111] obtained at 300 • C under low pressure N 2 gas (10 3 Pa) environment at 300 kV. The titanium column dumbbell separation of 134 pm is clearly resolved as shown in Fig. 4(a) spots of (88 pm) −1 as well as (104 pm) −1 as shown in Fig. 4(b). Figure 4(c) shows a HAADF-STEM image of the same sample at 300 • C under high pressure N 2 gas (10 5 Pa ≃ atmospheric pressure) condition. The image is relatively blurred compared with one under lower gas pressure condition. However, we can recognize the spots of (116 pm) −1 as well as (134 pm) −1 in the Fourier transform of the image [ Fig. 4(d)]. These images prove that atomic resolution imaging is possible even under high gas pressure condition up to 1 atmosphere.
In situ chemical analyses by using an EDS and an electron energy loss spectroscopy (EELS) are also available even at high gas pressure condition in an experiment using an in situ sample holder. Figure 5 shows the results of a redox reaction experiment for copper powder sample. The original sample was partially oxidized metallic copper. Elemental analysis by EDS and EELS under N 2 gas (10 3 Pa) at 300 • C shows the presence of both oxidized and metallic copper as shown in the lower panel in Fig. 5. After we changed the N 2 gas to H 2 gas (10 4 Pa), the morphology of the sample significantly changed as shown in middle panel of Fig. 5. The elemental analysis suggests that the most copper become metallic under the reduced environment. After that, we changed the H 2 gas to O 2 gas (10 4 Pa). In the O 2 gas, the morphology of the sample significantly changed again and the copper fully altered to be oxide. Furthermore, we found that the peak positions between initial and final copper oxides are different in EELS spectrum. This result suggests that the valence state of copper oxide changes after the redox process.

IV. CONCLUSION
We have developed a new X-ray detection system composed of two windowless 158 mm 2 -sized SDDs for an aberration corrected 300 kV microscope with a wide gap objective lens pole piece. The system significantly improved X-ray detection collection angle up to more than 2 steradians. It allows us to make atomic resolution X-ray maps with high definition and to analyze beam sensitive materials. The recent MEMS-based sample holder provides highly stable and reproducible in situ experiments. Experiment by 300 kV aberration corrected microscope with such an in situ sample holder, realized not only in situ observation at atomic resolution but also in situ chemical analyses by using EDS and EELS even under high gas pressure condition (1 atmosphere).