Microsc. Microanal. 21, 617–625, 2015 doi:10.1017/S1431927615000574
© MICROSCOPY SOCIETY OF AMERICA 2015
Statistical Study of Beam-Induced Motion of Gold Adatoms by a Scanning TEM Wei Zhou, Xin Li, and Guo-zhen Zhu* State Key Laboratory of Metal Matrix Composites, School of Materials Science and Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, P. R. China
Abstract: In order to achieve reliable structural characterization by transmission electron microscopy, beaminduced structural changes should be clarified for any target material system. As an example, the movement of heavy adatoms on a thin carbon support has been repeatedly reported under the electron beam while the underlying reason for such motion is still in debate. By applying statistical analysis to the group behavior of gold adatoms, we investigated their motion under different beam conditions and detected features corresponding to beam-induced motion, under typical scanning transmission electron microscopy observation conditions. Our results are consistent with the theoretical prediction proposed by Egerton (2013). Key words: beam-induced motion, adatoms, HAADF, statistical study
I NTRODUCTION Scanning transmission electron microscopy (STEM) is one, if not the only, comprehensive tool for seeing atomic structure and investigating the electronic structure of materials on the atomic scale (Pennycook & Nellist, 2011). This technique has been demonstrated with excellent performance for studying bulk materials, surfaces and interfaces, nanoparticles, and individual dopants. It can also probe the atomic structure and bonding nature of reactive nanoparticles (Liu, 2005), surface monolayers (Shibata et al., 2008; Jeon et al., 2011; Zhu et al., 2012; van der Zande et al., 2013), and adatoms (Voyles et al., 2002; Batson, 2008; Zan et al., 2011; Warner et al., 2013), which can be easily modified by the high-energy incoming electrons under most observation conditions. These structures with reduced dimensions, including active sites of nanocatalysts (Yan et al., 2006; Khanal et al., 2012) and surface monolayers on substrates (Hansen et al., 2011; van der Zande et al., 2013), are incredibly important to the future of material science. In order to achieve a reliable structural observation, beam responses, depending on the target material system, must be elucidated on a case-by-case basis. Although the power of STEM was first demonstrated from seeing heavy adatoms on a 2-nm-thick supporting carbon film (Crewe et al., 1970), a thorough description of the beam response from heavy adatoms remains largely unclear. Initially, experimental characterization of adatoms was limited by the spatial resolution of an STEM. Therefore, it was technically impossible to precisely locate individual adatoms. With the launch of field-emission guns (FEG) (Isaacson et al., 1977) and aberration correctors (Haider et al., 1998; Krivanek et al., 1999), an STEM is now capable of resolving single atoms. As a Received August 12, 2014; accepted June 4, 2015 *Corresponding author. [email protected]
consequence, the movement of adatoms and the change of adatom clusters under the electron beam have been reported (Isaacson et al., 1977; Krivanek et al., 2010). Furthermore, the tracking of individual adatoms has been carried out by recording experimental images with a time resolution of ~200 ms (Batson, 2008). While much effort has been expended, the underlying reason for this motion, whether it is thermal-induced, beaminduced, or both, is still in debate. The facts that cooling the specimen down reduces such motion (Wall, 1979) and such motion was detected at a low acceleration voltage of 30 kV (Isaacson et al., 1977) suggest that thermal vibration strongly affects the motion of adatoms. Recently, a theoretical proposal stated that the beam-induced motion has a threshold of incident energy generally 300 keV. The assumption above is additionally supported by the fact that the total number of adatoms within the entire frame was found without much change over time, as shown in Figure 4a. Another fact we noticed is that gold adatoms can attach and detach to the surface of gold nanoparticles, which may modify the total number of adatoms within a target area. However, such effect has been partially eliminated by expanding the calculated boundaries of nanoparticles during our selection of intensity windows. Thus, the nearly constant number of gold adatoms within an area including gold nanoparticles implies that the absorption and desorption of adatoms into gold nanoparticles nearly reached equilibrium under current experimental conditions. In order to minimize the effect caused by the attachment and detachment of adatoms at surfaces of nanoparticles, areas of interest were selected within the empty space between nanoparticles (see Fig. 4b) in order to investigate the surface-diffusion behavior of adatoms. Within each rectangular area, the number of adatoms kept constant as shown in Figure 4a, implying the directionless surface diffusion of adatoms. This fact is in full agreement with the predicted behavior resulting from both thermal-induced and beaminduced motion. The thermal-induced motion of adatoms can be described by the random walk model, which describes the directionless movement of adatoms and results in a constant number of adatoms within any selected area. On the other hand, the beam-induced motion of adatoms leads to a directionless diffusion of adatoms when the incident beam is perpendicular to the specimen according to Egerton’s model (Egerton, 2013). It should be noted that the surface roughness of a carbon support, including the existence of the so-called low-energy sites, may also affect the movement of individual adatoms. However, the average effect contributed from surface roughness should be homogenous. As a consequence, no preferential direction of the adatoms’ movement should be introduced. Therefore, a directionless motion of adatoms is expected by combining all factors without the knowledge of the dominating reason. As a result, the total number of adatoms within any selected area should be kept as a constant without considering the statistical fluctuation and the attachment and detachment of gold adatoms into gold nanoparticles. These factors may slightly modify the total number of adatoms within a selected area, which may be responsible for the experimental fluctuation of the counting number of adatoms in Figure 4a.
Statistical Study of Beam-Induced Motion of Gold Adatoms
Figure 3. A sequence of images when the incident beam is perpendicular to the sample. Eight STEM-HAADF images were selected after adjusting the contrast and correcting the drift. Numbers indicate the acquisition time of the frame. STEM-HAADF, scanning transmission electron microscopy high-angle annular dark-field.
621
622
Wei Zhou et al.
Figure 4. Statistical results from the image sequence in Figure 3. a: The counting result of adatoms from the selected areas labeled in (b); (b) the target areas for adatom counting. The incident beam was perpendicular to the specimen.
Beam NOT Perpendicular to the Supporting Film With a 20° tilt of the specimen, a different interaction pattern between the incident beam and the carbon support was expected according to Egerton’s model (Egerton, 2013). Under a probe current of ~35 pA, a selected region of 14.4 × 15.3 nm was tracked for 646 s, as shown in eight selected STEM-HAADF images in Figure 5. In contrast, a decrease in the total number of gold adatoms was detected with the increase in the beam-interaction time, which implies a strong effect introduced from the incident beam (see the curve of the whole area in Fig. 6a). The reasons for the decreasing number of adatoms can be: (a) the attachment of adatoms onto gold nanoparticles; (b) the beam-induced diffusion; and (c) the detachment of adatoms by beaminduced surface sputtering. The contribution from beam-induced sputtering is trivial because the required activation energy is much higher, equal to an incident beam energy of ~645 keV (Gan et al., 2008). In order to avoid any possible effect from gold nanoparticles, the investigation was carried out within a few areas occupying the empty space between nanoparticles. Those areas must have a reasonable number of adatoms so that a reliable statistical analysis can be achieved. The same tendency of decreasing number of adatoms was observed in each selected area (A or B), as shown in Figure 6a. The decreasing tendency is in agreement with the theoretical prediction (Egerton, 2013) that beaminduced diffusion has a preferential direction under the current condition. With the above preferential diffusion direction, adatoms refreshed from the neighboring regions can be stopped due to either the lack of the incident beam or blocking of adatoms by gold nanoparticles. On the other hand, a small area inside the entirely scanned frame can have enough adatoms refreshed from the neighboring area under the condition of directionless diffusion. Although it is cursory to conclude that the electron beam dominates the motion of adatoms, the above result
confirms the existence of beam-induced effects, which is consistent with the predicted result by Egerton (2013). Accordingly, the beam-induced motion dominates the behavior of adatoms at or below room temperature when the beam-induced surface-diffusion energy Esd > 0.5 eV. Under the current experimental condition, the surface-diffusion energy is estimated to be 2.3 eV as for gold adatoms on graphene (Gan et al., 2008). Thus, the beam-induced motion of gold adatoms can be detected at room temperature and may be additionally enhanced by tilting the sample (Egerton, 2013). In conclusion, the behavior of adatoms resulting from beam–adatom interaction can thus be observed. It should be noted that the motion of adatoms is correlated when the spacing between them is less than a few Ångstroms. As shown in Figure 7a, a small adatom cluster consisting of ~30 adatoms was recorded under the electron beam continuously for 1,170 s including 60 frames. The cohesiveness between adatoms is retained under the condition that the electron beam is 20° to the carbon support and the formation of a single particle was not detected. This above behavior implies that gold adatoms preferentially locate at some low-energy sites. In order to prove that, the area containing this adatom cluster was divided into small zones of 3 × 3 pixels for counting the appearance of adatoms and the pixel intensity of adatoms over time. Small zones of 3 × 3 pixels were used to improve reliability since the image offset has the uncertainty of 1 pixel. It is no surprise to discover that some of these small zones have much higher frequency of detecting gold adatoms (e.g., 20 frames) than others (e.g., 0 frames) as shown in Figure 7b. We believe that these preferential sites are due to the carbon support. It is reasonable to assume that the surface of our carbon support was not perfectly flat and there were locations nearly perpendicular to the incident beam. At those locations, the beam-induced diffusion of the adatoms was minimized and directionless. As a result, adatoms preferentially stayed at
Statistical Study of Beam-Induced Motion of Gold Adatoms
Figure 5. A sequence of images under the condition that the incident beam was not perpendicular to the carbon support. These HAADF images were processed to visualize adatoms. Numbers in the images indicate the acquisition time. HAADF, high-angle annular dark-field.
623
624
Wei Zhou et al.
Figure 6. Statistical results of the image sequence in Figure 5. a: Represents the counting result from the selected areas labeled in (b). The incident beam was not perpendicular to the specimen.
Figure 7. Cohesiveness between adatoms retained under the condition that the electron beam was tilted 20° with respect to the carbon support. a: STEM-HAADF image sequence of an adatom cluster. Numbers indicate the acquisition time of the frame. b: The enlarged image in the rectangle region with the size of 90 × 90 pixels; c: the superimposed image of all 60 frames of the rectangle region in b; d: the counting result of the appearance of adatoms over time, which indicated the existence of preferential sites. STEM-HAADF, scanning transmission electron microscopy high-angle annular dark-field.
Statistical Study of Beam-Induced Motion of Gold Adatoms
these low-energy sites, which prevented those adatoms from coalescing into one particle.
CONCLUSIONS By statistically analyzing the STEM-HAADF image sequences of gold adatoms on a carbon support under different beam conditions, we studied the beam-induced behavior of adatoms and provided experimental evidence of the dependence of beam response on the geometric configuration of the beam and the specimen. The evidence of beam-induced motion is clearly detected when the incident beam is not perpendicular to a carbon support. Our results support the previously reported theoretical model, and provide insight into the understanding of beam responses of adatoms.
ACKNOWLEDGMENTS The authors acknowledge funding from the National Natural Science Foundation of China (No. 51401124). The microscopy was carried out at the Frontier Research Center for Materials Structure in Shanghai Jiao Tong University.
REFERENCE BATSON, P.E. (2008). Motion of gold atoms on carbon in the aberration-corrected STEM. Microsc Microanal 14(01), 89–97. CREWE, A.V., WALL, J. & LANGMORE, J. (1970). Visibility of single atoms. Science 168(3937), 1338–1340. EGERTON, R.F. (2013). Beam-induced motion of adatoms in the transmission electron microscope. Microsc Microanal 19(2), 479–486. GAN, Y., SUN, L. & BANHART, F. (2008). One-and two-dimensional diffusion of metal atoms in graphene. Small 4(5), 587–591. HAIDER, M., ROSE, H., UHLEMANN, S., KABIUS, B. & URBAN, K. (1998). Towards 0.1 nm resolution with the first spherically corrected transmission electron microscope. J Electron Microsc 47(5), 395–405. HANSEN, L.P., RAMASSE, Q.M., KISIELOWSKI, C., BRORSON, M., JOHNSON, E., TOPSOE, H. & HELVEG, S. (2011). Atomic-scale edge structures on industrial-style MoS2 nanocatalysts. Angew Chem Int Ed Engl 50(43), 10153–10156. ISAACSON, M., KOPF, D., UTLAUT, M., PARKER, N.M. & CREWE, A.V. (1977). Direct observations of atomic diffusion by scanning transmission electron microscopy. Proc Natl Acad Sci 74(5), 1802–1806.
625
JEON, I., YANG, H., LEE, S.H., HEO, J., SEO, D.H., SHIN, J. & SEO, S. (2011). Passivation of metal surface states: Microscopic origin for uniform monolayer graphene by low temperature chemical vapor deposition. ACS Nano 5(3), 1915–1920. KHANAL, S., CASILLAS, G., VELAZQUEZ-SALAZAR, J.J., PONCE, A. & JOSE-YACAMAN, M. (2012). Atomic resolution imaging of polyhedral Pt Pd core–shell nanoparticles by Cs-corrected STEM. J Phys Chem C 116(44), 23596–23602. KRIVANEK, O., DELLBY, N. & LUPINI, A. (1999). Towards sub-Å electron beams. Ultramicroscopy 78(1), 1–11. KRIVANEK, O.L., CHISHOLM, M.F., NICOLOSI, V., PENNYCOOK, T.J., CORBIN, G.J., DELLBY, N., MURFITT, M.F., OWN, C.S., SZILAGYI, Z.S., OXLEY, M.P., PANTELIDES, S.T. & PENNYCOOK, S.J. (2010). Atomby-atom structural and chemical analysis by annular dark-field electron microscopy. Nature 464(7288), 571–574. LIU, J. (2005). Scanning transmission electron microscopy and its application to the study of nanoparticles and nanoparticle systems. J Electron Microsc 54(3), 251–278. PENNYCOOK, S.J. & NELLIST, P.D., eds. (2011). Scanning Transmission Electron Microscopy: Imaging and Analysis. Springer-Verlag New York: Springer. SHIBATA, N., GOTO, A., CHOI, S.Y., MIZOGUCHI, T., FINDLAY, S.D., YAMAMOTO, T. & IKUHARA., Y. (2008). Direct imaging of reconstructed atoms on TiO2 (110) surfaces. Science 322(5901), 570–573. VAN DER ZANDE, A.M., HUANG, P.Y., CHENET, D.A., BERKELBACH, T.C., YOU, Y., LEE, G.H., HEINZ, T.F., REICHMAN, D.R., MULLER, D.A. & HONE, J.C. (2013). Grains and grain boundaries in highly crystalline monolayer molybdenum disulphide. Nat Mater 12(6), 554–561. VOYLES, P.M., MULLER, D.A., GRAZUL, J.L., CITRIN, P.H. & GOSSMANN, H.J.L. (2002). Atomic-scale imaging of individual dopant atoms and clusters in highly n-type bulk Si. Nature 416(6883), 826–829. WALL, J. (1979). Biological Scanning Transmission Electron Microscopy in Introduction to Analytical Electron Microscopy. New York: Plenum. WARNER, J.H., LIU, Z., HE, K., ROBERTSON, A.W. & SUENAGA, K. (2013). Sensitivity of graphene edge states to surface adatom interactions. Nano Lett 13(10), 4820–4826. YAN, W., BROWN, S., PAN, Z., MAHURIN, S.M., OVERBURY, S.H. & DAI, S. (2006). Ultrastable gold nanocatalyst supported by nanosized nonoxide substrate. Angew Chem Int Ed Engl 118(22), 3696–3700. ZAN, R., BANGERT, U., RAMASSE, Q. & NOVOSELOV, K.S. (2011). Metalgraphene interaction studied via atomic resolution scanning transmission electron microscopy. Nano Lett 11(3), 1087–1092. ZHU, G.Z., RADTKE, G. & BOTTON, G.A. (2012). Bonding and structure of a reconstructed (001) surface of SrTiO3 from TEM. Nature 490(7420), 384–387.
© MICROSCOPY SOCIETY OF AMERICA 2015
Statistical Study of Beam-Induced Motion of Gold Adatoms by a Scanning TEM Wei Zhou, Xin Li, and Guo-zhen Zhu* State Key Laboratory of Metal Matrix Composites, School of Materials Science and Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, P. R. China
Abstract: In order to achieve reliable structural characterization by transmission electron microscopy, beaminduced structural changes should be clarified for any target material system. As an example, the movement of heavy adatoms on a thin carbon support has been repeatedly reported under the electron beam while the underlying reason for such motion is still in debate. By applying statistical analysis to the group behavior of gold adatoms, we investigated their motion under different beam conditions and detected features corresponding to beam-induced motion, under typical scanning transmission electron microscopy observation conditions. Our results are consistent with the theoretical prediction proposed by Egerton (2013). Key words: beam-induced motion, adatoms, HAADF, statistical study
I NTRODUCTION Scanning transmission electron microscopy (STEM) is one, if not the only, comprehensive tool for seeing atomic structure and investigating the electronic structure of materials on the atomic scale (Pennycook & Nellist, 2011). This technique has been demonstrated with excellent performance for studying bulk materials, surfaces and interfaces, nanoparticles, and individual dopants. It can also probe the atomic structure and bonding nature of reactive nanoparticles (Liu, 2005), surface monolayers (Shibata et al., 2008; Jeon et al., 2011; Zhu et al., 2012; van der Zande et al., 2013), and adatoms (Voyles et al., 2002; Batson, 2008; Zan et al., 2011; Warner et al., 2013), which can be easily modified by the high-energy incoming electrons under most observation conditions. These structures with reduced dimensions, including active sites of nanocatalysts (Yan et al., 2006; Khanal et al., 2012) and surface monolayers on substrates (Hansen et al., 2011; van der Zande et al., 2013), are incredibly important to the future of material science. In order to achieve a reliable structural observation, beam responses, depending on the target material system, must be elucidated on a case-by-case basis. Although the power of STEM was first demonstrated from seeing heavy adatoms on a 2-nm-thick supporting carbon film (Crewe et al., 1970), a thorough description of the beam response from heavy adatoms remains largely unclear. Initially, experimental characterization of adatoms was limited by the spatial resolution of an STEM. Therefore, it was technically impossible to precisely locate individual adatoms. With the launch of field-emission guns (FEG) (Isaacson et al., 1977) and aberration correctors (Haider et al., 1998; Krivanek et al., 1999), an STEM is now capable of resolving single atoms. As a Received August 12, 2014; accepted June 4, 2015 *Corresponding author. [email protected]
consequence, the movement of adatoms and the change of adatom clusters under the electron beam have been reported (Isaacson et al., 1977; Krivanek et al., 2010). Furthermore, the tracking of individual adatoms has been carried out by recording experimental images with a time resolution of ~200 ms (Batson, 2008). While much effort has been expended, the underlying reason for this motion, whether it is thermal-induced, beaminduced, or both, is still in debate. The facts that cooling the specimen down reduces such motion (Wall, 1979) and such motion was detected at a low acceleration voltage of 30 kV (Isaacson et al., 1977) suggest that thermal vibration strongly affects the motion of adatoms. Recently, a theoretical proposal stated that the beam-induced motion has a threshold of incident energy generally 300 keV. The assumption above is additionally supported by the fact that the total number of adatoms within the entire frame was found without much change over time, as shown in Figure 4a. Another fact we noticed is that gold adatoms can attach and detach to the surface of gold nanoparticles, which may modify the total number of adatoms within a target area. However, such effect has been partially eliminated by expanding the calculated boundaries of nanoparticles during our selection of intensity windows. Thus, the nearly constant number of gold adatoms within an area including gold nanoparticles implies that the absorption and desorption of adatoms into gold nanoparticles nearly reached equilibrium under current experimental conditions. In order to minimize the effect caused by the attachment and detachment of adatoms at surfaces of nanoparticles, areas of interest were selected within the empty space between nanoparticles (see Fig. 4b) in order to investigate the surface-diffusion behavior of adatoms. Within each rectangular area, the number of adatoms kept constant as shown in Figure 4a, implying the directionless surface diffusion of adatoms. This fact is in full agreement with the predicted behavior resulting from both thermal-induced and beaminduced motion. The thermal-induced motion of adatoms can be described by the random walk model, which describes the directionless movement of adatoms and results in a constant number of adatoms within any selected area. On the other hand, the beam-induced motion of adatoms leads to a directionless diffusion of adatoms when the incident beam is perpendicular to the specimen according to Egerton’s model (Egerton, 2013). It should be noted that the surface roughness of a carbon support, including the existence of the so-called low-energy sites, may also affect the movement of individual adatoms. However, the average effect contributed from surface roughness should be homogenous. As a consequence, no preferential direction of the adatoms’ movement should be introduced. Therefore, a directionless motion of adatoms is expected by combining all factors without the knowledge of the dominating reason. As a result, the total number of adatoms within any selected area should be kept as a constant without considering the statistical fluctuation and the attachment and detachment of gold adatoms into gold nanoparticles. These factors may slightly modify the total number of adatoms within a selected area, which may be responsible for the experimental fluctuation of the counting number of adatoms in Figure 4a.
Statistical Study of Beam-Induced Motion of Gold Adatoms
Figure 3. A sequence of images when the incident beam is perpendicular to the sample. Eight STEM-HAADF images were selected after adjusting the contrast and correcting the drift. Numbers indicate the acquisition time of the frame. STEM-HAADF, scanning transmission electron microscopy high-angle annular dark-field.
621
622
Wei Zhou et al.
Figure 4. Statistical results from the image sequence in Figure 3. a: The counting result of adatoms from the selected areas labeled in (b); (b) the target areas for adatom counting. The incident beam was perpendicular to the specimen.
Beam NOT Perpendicular to the Supporting Film With a 20° tilt of the specimen, a different interaction pattern between the incident beam and the carbon support was expected according to Egerton’s model (Egerton, 2013). Under a probe current of ~35 pA, a selected region of 14.4 × 15.3 nm was tracked for 646 s, as shown in eight selected STEM-HAADF images in Figure 5. In contrast, a decrease in the total number of gold adatoms was detected with the increase in the beam-interaction time, which implies a strong effect introduced from the incident beam (see the curve of the whole area in Fig. 6a). The reasons for the decreasing number of adatoms can be: (a) the attachment of adatoms onto gold nanoparticles; (b) the beam-induced diffusion; and (c) the detachment of adatoms by beaminduced surface sputtering. The contribution from beam-induced sputtering is trivial because the required activation energy is much higher, equal to an incident beam energy of ~645 keV (Gan et al., 2008). In order to avoid any possible effect from gold nanoparticles, the investigation was carried out within a few areas occupying the empty space between nanoparticles. Those areas must have a reasonable number of adatoms so that a reliable statistical analysis can be achieved. The same tendency of decreasing number of adatoms was observed in each selected area (A or B), as shown in Figure 6a. The decreasing tendency is in agreement with the theoretical prediction (Egerton, 2013) that beaminduced diffusion has a preferential direction under the current condition. With the above preferential diffusion direction, adatoms refreshed from the neighboring regions can be stopped due to either the lack of the incident beam or blocking of adatoms by gold nanoparticles. On the other hand, a small area inside the entirely scanned frame can have enough adatoms refreshed from the neighboring area under the condition of directionless diffusion. Although it is cursory to conclude that the electron beam dominates the motion of adatoms, the above result
confirms the existence of beam-induced effects, which is consistent with the predicted result by Egerton (2013). Accordingly, the beam-induced motion dominates the behavior of adatoms at or below room temperature when the beam-induced surface-diffusion energy Esd > 0.5 eV. Under the current experimental condition, the surface-diffusion energy is estimated to be 2.3 eV as for gold adatoms on graphene (Gan et al., 2008). Thus, the beam-induced motion of gold adatoms can be detected at room temperature and may be additionally enhanced by tilting the sample (Egerton, 2013). In conclusion, the behavior of adatoms resulting from beam–adatom interaction can thus be observed. It should be noted that the motion of adatoms is correlated when the spacing between them is less than a few Ångstroms. As shown in Figure 7a, a small adatom cluster consisting of ~30 adatoms was recorded under the electron beam continuously for 1,170 s including 60 frames. The cohesiveness between adatoms is retained under the condition that the electron beam is 20° to the carbon support and the formation of a single particle was not detected. This above behavior implies that gold adatoms preferentially locate at some low-energy sites. In order to prove that, the area containing this adatom cluster was divided into small zones of 3 × 3 pixels for counting the appearance of adatoms and the pixel intensity of adatoms over time. Small zones of 3 × 3 pixels were used to improve reliability since the image offset has the uncertainty of 1 pixel. It is no surprise to discover that some of these small zones have much higher frequency of detecting gold adatoms (e.g., 20 frames) than others (e.g., 0 frames) as shown in Figure 7b. We believe that these preferential sites are due to the carbon support. It is reasonable to assume that the surface of our carbon support was not perfectly flat and there were locations nearly perpendicular to the incident beam. At those locations, the beam-induced diffusion of the adatoms was minimized and directionless. As a result, adatoms preferentially stayed at
Statistical Study of Beam-Induced Motion of Gold Adatoms
Figure 5. A sequence of images under the condition that the incident beam was not perpendicular to the carbon support. These HAADF images were processed to visualize adatoms. Numbers in the images indicate the acquisition time. HAADF, high-angle annular dark-field.
623
624
Wei Zhou et al.
Figure 6. Statistical results of the image sequence in Figure 5. a: Represents the counting result from the selected areas labeled in (b). The incident beam was not perpendicular to the specimen.
Figure 7. Cohesiveness between adatoms retained under the condition that the electron beam was tilted 20° with respect to the carbon support. a: STEM-HAADF image sequence of an adatom cluster. Numbers indicate the acquisition time of the frame. b: The enlarged image in the rectangle region with the size of 90 × 90 pixels; c: the superimposed image of all 60 frames of the rectangle region in b; d: the counting result of the appearance of adatoms over time, which indicated the existence of preferential sites. STEM-HAADF, scanning transmission electron microscopy high-angle annular dark-field.
Statistical Study of Beam-Induced Motion of Gold Adatoms
these low-energy sites, which prevented those adatoms from coalescing into one particle.
CONCLUSIONS By statistically analyzing the STEM-HAADF image sequences of gold adatoms on a carbon support under different beam conditions, we studied the beam-induced behavior of adatoms and provided experimental evidence of the dependence of beam response on the geometric configuration of the beam and the specimen. The evidence of beam-induced motion is clearly detected when the incident beam is not perpendicular to a carbon support. Our results support the previously reported theoretical model, and provide insight into the understanding of beam responses of adatoms.
ACKNOWLEDGMENTS The authors acknowledge funding from the National Natural Science Foundation of China (No. 51401124). The microscopy was carried out at the Frontier Research Center for Materials Structure in Shanghai Jiao Tong University.
REFERENCE BATSON, P.E. (2008). Motion of gold atoms on carbon in the aberration-corrected STEM. Microsc Microanal 14(01), 89–97. CREWE, A.V., WALL, J. & LANGMORE, J. (1970). Visibility of single atoms. Science 168(3937), 1338–1340. EGERTON, R.F. (2013). Beam-induced motion of adatoms in the transmission electron microscope. Microsc Microanal 19(2), 479–486. GAN, Y., SUN, L. & BANHART, F. (2008). One-and two-dimensional diffusion of metal atoms in graphene. Small 4(5), 587–591. HAIDER, M., ROSE, H., UHLEMANN, S., KABIUS, B. & URBAN, K. (1998). Towards 0.1 nm resolution with the first spherically corrected transmission electron microscope. J Electron Microsc 47(5), 395–405. HANSEN, L.P., RAMASSE, Q.M., KISIELOWSKI, C., BRORSON, M., JOHNSON, E., TOPSOE, H. & HELVEG, S. (2011). Atomic-scale edge structures on industrial-style MoS2 nanocatalysts. Angew Chem Int Ed Engl 50(43), 10153–10156. ISAACSON, M., KOPF, D., UTLAUT, M., PARKER, N.M. & CREWE, A.V. (1977). Direct observations of atomic diffusion by scanning transmission electron microscopy. Proc Natl Acad Sci 74(5), 1802–1806.
625
JEON, I., YANG, H., LEE, S.H., HEO, J., SEO, D.H., SHIN, J. & SEO, S. (2011). Passivation of metal surface states: Microscopic origin for uniform monolayer graphene by low temperature chemical vapor deposition. ACS Nano 5(3), 1915–1920. KHANAL, S., CASILLAS, G., VELAZQUEZ-SALAZAR, J.J., PONCE, A. & JOSE-YACAMAN, M. (2012). Atomic resolution imaging of polyhedral Pt Pd core–shell nanoparticles by Cs-corrected STEM. J Phys Chem C 116(44), 23596–23602. KRIVANEK, O., DELLBY, N. & LUPINI, A. (1999). Towards sub-Å electron beams. Ultramicroscopy 78(1), 1–11. KRIVANEK, O.L., CHISHOLM, M.F., NICOLOSI, V., PENNYCOOK, T.J., CORBIN, G.J., DELLBY, N., MURFITT, M.F., OWN, C.S., SZILAGYI, Z.S., OXLEY, M.P., PANTELIDES, S.T. & PENNYCOOK, S.J. (2010). Atomby-atom structural and chemical analysis by annular dark-field electron microscopy. Nature 464(7288), 571–574. LIU, J. (2005). Scanning transmission electron microscopy and its application to the study of nanoparticles and nanoparticle systems. J Electron Microsc 54(3), 251–278. PENNYCOOK, S.J. & NELLIST, P.D., eds. (2011). Scanning Transmission Electron Microscopy: Imaging and Analysis. Springer-Verlag New York: Springer. SHIBATA, N., GOTO, A., CHOI, S.Y., MIZOGUCHI, T., FINDLAY, S.D., YAMAMOTO, T. & IKUHARA., Y. (2008). Direct imaging of reconstructed atoms on TiO2 (110) surfaces. Science 322(5901), 570–573. VAN DER ZANDE, A.M., HUANG, P.Y., CHENET, D.A., BERKELBACH, T.C., YOU, Y., LEE, G.H., HEINZ, T.F., REICHMAN, D.R., MULLER, D.A. & HONE, J.C. (2013). Grains and grain boundaries in highly crystalline monolayer molybdenum disulphide. Nat Mater 12(6), 554–561. VOYLES, P.M., MULLER, D.A., GRAZUL, J.L., CITRIN, P.H. & GOSSMANN, H.J.L. (2002). Atomic-scale imaging of individual dopant atoms and clusters in highly n-type bulk Si. Nature 416(6883), 826–829. WALL, J. (1979). Biological Scanning Transmission Electron Microscopy in Introduction to Analytical Electron Microscopy. New York: Plenum. WARNER, J.H., LIU, Z., HE, K., ROBERTSON, A.W. & SUENAGA, K. (2013). Sensitivity of graphene edge states to surface adatom interactions. Nano Lett 13(10), 4820–4826. YAN, W., BROWN, S., PAN, Z., MAHURIN, S.M., OVERBURY, S.H. & DAI, S. (2006). Ultrastable gold nanocatalyst supported by nanosized nonoxide substrate. Angew Chem Int Ed Engl 118(22), 3696–3700. ZAN, R., BANGERT, U., RAMASSE, Q. & NOVOSELOV, K.S. (2011). Metalgraphene interaction studied via atomic resolution scanning transmission electron microscopy. Nano Lett 11(3), 1087–1092. ZHU, G.Z., RADTKE, G. & BOTTON, G.A. (2012). Bonding and structure of a reconstructed (001) surface of SrTiO3 from TEM. Nature 490(7420), 384–387.
