Article pubs.acs.org/ac
Direct Observation of Sublimation Behaviors in One-Dimensional In2Se3/In2O3 Nanoheterostructures Cheng-Lun Hsin,*,†,‡,§ Chun-Wei Huang,† Jui-Yuan Chen,† Kuo-Cheng Liao,∥ Po-Liang Liu,∥,⊥ Wen-Wei Wu,*,† and Lih-Juann Chen*,‡ †
Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu City, 300 Taiwan Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu City, 30013 Taiwan § Department of Electrical Engineering, National Central University, Taoyuan City, 32001 Taiwan ∥ Graduate Institute of Precision Engineering, National Chung Hsing University, Taichung City, 402 Taiwan ⊥ Department of Physics, Norwegian University of Science and Technology, 7491 Trondheim, Norway ‡
S Supporting Information *
ABSTRACT: Recently, in situ transmission electron microscopy (TEM) has provided a route to analyze structural characterization and chemical evolution with its powerful and unique applications. In this paper, we disclose the detailed phenomenon of sublimation on the atomic scale. In2Se3/In2O3 nanowires were synthesized via the vapor−liquid−solid mechanism and studied in an ultra-high-vacuum (UHV) TEM at high temperature in real time. During in situ observation of the sublimation process of the nanowires, the evolution and reconstruction of the exposed In2Se3 surface progressed in different manners with time. The surface structure was decomposed by mass-desorption and stepwise-migration processes, which are also energetically favored processes in the ab initio calculation. This study developed a new concept and will be essential in the development of atomic kinetics.
T
in nanomaterials from the information obtained in a study of transient structural transformation during sublimation. The surface steps were observed to be unstable and reconstructed through mass-desorption and stepwise-migration processes rather than the conventional macroscopic concept, which favors the individually unstable atom model. The atomic sublimation process has been unraveled and is expected to substantially advance contemporary scientific understanding. The model system selected to study the transient structural change was a chalcogenide compound, In2Se3. Detailed information about the nanostructures is provided in Supporting Information and elsewhere.4 The process of transient structural transformation of In2Se3 in the In2Se3/In2O3 core/shell nanowire during sublimation was investigated at the atomic scale in situ by use of a transmission electron microscope. Sublimation of In2Se3 occurred at 450 °C in UHV. For the mesoscopic-scale investigation, to determine the residual length of In2Se3, TEM images at the nanoscale were taken during the sublimation process, as presented in Figure 1af. From the recorded images, the disappearance of In2Se3 appears to be inclined to one side. A plot of the sublimation length as a function of time was extrapolated from these
he study of new phenomena at the nanoscale has been a widely developing research field in recent years.1−3 To disclose detailed properties of nanomaterials, microscopy techniques that can image in real time (with atomic resolution) were employed to study solid-state reactions,4−7 electromigration,8 nanomechanics,9−11 and structural12−17 and phase transformation.18−20 With the tremendous progress in nanotechnology, fundamental physics of nanowires has advanced remarkably, unraveling detailed information that can help us understand and develop new knowledge. With the employment of in situ TEM, we can directly observe the chemical reaction at atomic scale at the interface between different phases, providing unique, important, and interesting observations and analysis.21−25 The obtained data not only show time-dependent information but also allow us to look into the structural change at the surface/interface, providing a route to look into the details of chemical reaction. Conventionally, sublimation is an intense phase change between solid and vapor along with chemical reactions. The solid-to-vapor phenomenon is assumed to occur as long as the vapor pressure of the solid reaches a critical point, similar to the reverse process of epitaxy, a general process for developing novel semiconductor electronic devices. The question of whether the surface atoms are individually unstable or not has yet to be clarified. Here, we demonstrate that the sublimation process of one-dimensional nanostructures is vastly more complex than previously envisioned. We propose a new concept of the microscopic physics of dynamic behavior © 2015 American Chemical Society
Received: January 20, 2015 Accepted: May 5, 2015 Published: May 5, 2015 5584
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry
Figure 1. Series of TEM images illustrating the progression of sublimation of In2Se3. The images were obtained at elapsed times (a) 0, (b) 6.5, (c) 8, (d) 10, (e) 12, and (f) 14 min. The scale bar at the bottom corner corresponds to 200 nm. (g) Plot of sublimation length as a function of time.
Figure 2. Series of images clipped from the video illustrating the progression of transient structural transformation of In2Se3 at 450 °C. Pathways of the transient structural transformation are highlighted by red arrows, and the blue dashed lines are auxiliary markers that depict the same position in different frames. Black dotted lines illustrate the corresponding lattice planes. The first two numbers of the time scale are in units of minutes and seconds, respectively, and the following numbers with the prime are in units of 1/60 s: (a) 3:25:02′, (b) 3:35:50′, (c) 3:57:28′, (d) 4:51:16′, (e) 4:59:22′, (f) 5:08:12′, (g) 5:17:02′, (h) 5:19:12′. The scale bar at the bottom corner corresponds to 2 nm. (i) Schematic illustration of structural evolution of surface steps over the entire nanowire.
In2Se3 nanowires. The core/shell structure was obtained by oxidation after synthesis at ambient conditions. Nanostructure Characterization and in Situ Observation. The nanowires were drop-cast on a Mo grid with a thin C film to support the nanostructures for the high-temperature study. The observation was conducted in a JEOL 2000 V UHVTEM with a base pressure of 3 × 10−10 Torr and 200 keV acceleration voltage. Before the in situ TEM observation, the sample was placed into a heating holder for degassing in the pretreatment chamber at 200 °C for 5 min. During real-time observation, the temperature was controlled and the pressure would be raised to ∼1 × 10−7 Torr. Video recorder had a time resolution of 1/30 s. First-Principles Density Functional Theory Calculations. First-principles density functional theory (DFT) calculations were conducted by use of the Vienna ab-initio simulation package.27−29 Projected augmented wave pseudopotentials were used to describe the valence electrons of In (5s2
recorded images. The linear tendency indicated that the average reaction rate and mechanism should be attributed to a Langmuir-type vaporization process at the In2Se3/vacuum interface, where direct sublimation occurred with no flux recondensing from the vapor.26
■
EXPERIMENTAL SECTION Synthesis. Single-crystal, 1−30 Ω·cm, (001)-oriented silicon wafers were used as substrates. A 2-nm-thick Au thin film was deposited on each substrate as a catalyst. Synthesis of the In2Se3 nanowires was performed in a horizontal tube furnace. During the growth process, 0.82 g of Se powder was placed upstream and a 1.2 g In slug was placed downstream. The samples were positioned 2 cm apart from the In source at the rear. The Se and In sources were heated to 240 and 600 °C, respectively, and transported by a carrier gas composed of 80% Ar and 20% H2 at 200 sccm at 1 Torr. When the set temperature was reached, the process was held for another 2 h and then cooled down to room temperature to obtain the 5585
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry 5p1) and Se (4s2 4p4) within the PW91 generalized gradient approximation for exchange and correlation energy.30,31
■
RESULTS AND DISCUSSION A series of high-resolution images from the recorded video are presented in Figure 2 to illustrate the transformation at the atomic scale. The pathways leading to transient structural transformation in the next frame are highlighted by red arrows. The single/double blue dashed lines are auxiliary marks that depict the same position in different frames. The black dotted lines depict the corresponding lattice planes. The first two numbers of the time scale in the figures are in units of minutes and seconds, respectively, and the following numbers with the prime are in units of 1/60 s. At the beginning of the recording (Figure 2a), the surface of In2Se3 was composed of facets with low indices. The convex facets were observed to depart from the matrix, termed mass desorption, and transformed into concave facets within 1 s. Then the field of view was kept the same for ∼10 s (Figure 2b). At the top of the frame in Figure 2b, mass desorption occurred again for the lattice planes to come off, becoming the facet at the top of Figure 2c. It is clear that mass desorption is an important process in reconstruction of the steps at the surface as well as at the bottom of Figure 2c. The (1̅100) plane in Figure 2c sublimated toward the downside, making the (011̅0) plane the most exposed surface. With the progression of mass desorption, one of the low-index planes became dominant at the surface, as illustrated in Figure 2d. In Figure 2d,e, the double blue dashed line represents the identical field of interest. A surface marked with a single-dashed line is observed at the bottom of Figure 2e. Both of the steps were moving downward, resulting in a new (0110̅ ) plane becoming the surface plane, termed stepwise migration. The single blue dashed line in Figure 2e,f represents the same position. At the bottom of the field of view in Figure 2f, a small step departed from the matrix, followed by downward movement of the other surface step, resulting in the new surface structure was shown in Figure 2g. After stepwise migration of the surface step in Figure 2g, the surface of In2Se3 became flat with only the (0110̅ ) surface plane. The projected cross section of structural evolution of the In2Se3 surface steps is schematically illustrated in Figure 2i. Two sublimation processes (mass desorption and stepwise migration) are essential for deconstruction of the surface. For the process of mass desorption to introduce a new surface, the surface steps were composed of the (011̅0) and (1̅100) planes in Figure 3a, which are highlighted with the double-dashed lines. Within a very short time, atoms in the exposed rows of the (011̅0) plane came off, introducing a new (1̅100) plane, as observed at the bottom of Figure 3b. Subsequently, sublimation of the existing and exposed (1̅100) plane occurred. A transient step was induced with a new (0110̅ ) plane at the top of Figure 3c, followed by completion of the mass-desorption process. After the process, the surface in Figure 3d was composed of identical low-index planes with a diminished matrix volume. The removal stages of each atomic layer are well resolved in the recorded video (Supporting Information, Movie S2). Another important process in sublimation is stepwise migration. A pertinent example is illustrated in Figure 4. With the presence of the (11̅ 00) plane (black dotted line) in Figure 4a, the surface step became unstable. To eliminate the excess energy associated with the plane, the (011̅0) facet layers (yellow dashed line) decomposed along with the atomic planes
Figure 3. Series of images clipped from the video illustrating progression of the mass-desorption process: (a) 3:25:10′, (b) 3:25:20′, (c) 3:25:28′, (d) 3:26:00′. The double-dashed lines highlight the primal surface step. The scale bar at the bottom corner corresponds to 2 nm.
Figure 4. (a−d) Series of images clipped from the video illustrating progression of the stepwise-migration process: (a) 5:28:12′, (b) 5:28:18′, (c) 5:28:20′, (d) 5:28:22′. The single-dashed line highlights the surface. (e, f) Series of images clipped from the video illustrating sublimation progression of the (011̅0) plane: (e) 5:37:36′, (f) 5:37:52′. The scale bar at the bottom corner corresponds to 2 nm. Drawings are included to guide the eye in locating the evolution of the surface steps.
(Figure 4a−d). The migration of surface atoms resulted in removal of the (1̅100) plane. Ultimately, the surface was predominantly composed of the low-index, energetically favorable (011̅0) plane, causing the disappearance of the (11̅ 00) plane. In the recorded video, the reaction stages of the atomic planes are well resolved (Supporting Information, Movie S3). After formation of the energetically favorable plane, the matrix continued to sublime at the high temperature, while the exposed atomic layers vaporized and continued to sublimate layer-by-layer, termed layer sublimation, resulting in the formation of a nanotube. In the frames in Figure 4e,f, the 5586
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry
Gibbs free energy of the In2Se3 matrix/vacuum interface at sublimation, G, should be a positive value, causing sublimation of the In2Se3 matrix. These data indicated that this transient structural transformation also manifested the role of energy minimization in the dynamic behavior of materials, which will be helpful in understanding the reverse process, that is, epitaxy.
reference markers are indicated by arrows. In Figure 4e,f, the distance between the reference point and the surface plane (yellow dashed line) increased, substantiating that the atomic layers sublimated with the unit in atomic layers. The reaction stages of each atomic layer are also well recorded in the video (Supporting Information, Movie S4). The high-resolution TEM images in Figures 3 and 4 clearly illustrate the fluctuating surface step edges of the In2Se3 matrix. The appearance of two distinguishable In2Se3 (1̅100) and (011̅0) surfaces during sublimation was not expected. A complete survey of the structural trends of alignment primarily consisting of low-index (1̅100) and (011̅0) facets with edges along the [1̅1̅20] and [2̅110] directions are depicted in Supporting Information. A plausible mechanism for the zigzag behavior during sublimation is supported by theoretical simulations of the surface formation energy by use of the Vienna ab-initio simulation package.27−29 The In2Se3 (1̅100) and In2Se3(0110̅ ) surfaces are energetically equivalent and stable within the thermodynamically allowed ranges as μbulk In + bulk f bulk Hf [In2Se3] ≤ μIn ≤ μbulk In or μSe + H [In2Se3] ≤ μSe ≤ μSe . As expected, the region observed in each frame from the highresolution image is localized, and dominant facets all over the cross-sectional surface could be bimodal from the lowmagnification images. However, the chemical potentials of In and Se in sublimation under vacuum extend beyond the range of allowed chemical potentials, leading to the gradual decomposition of the surface steps (Supporting Information). There are several points that need to be addressed with direct observation of the gradual transient surface structural transformation from faceted steps to one dominant facet. First, during TEM imaging, the interaction between the substance and the incident high-energy electrons generates heat and increases the temperature. Thus, the equilibrium temperature could be higher than that measured (450 °C). Additionally, the electron beam could also decompose the structure of In2Se3.32 We also have performed experiments at lower and higher temperatures. At lower temperature, the sublimation could be too slow to proceed, or In2Se3 will dissociate first and then a small amount of nanoparticles will form on the surface. After several tens of seconds, these nanoparticles will shrink gradually. At higher temperature, the sublimation behaviors are the same as what we have observed but with faster progress speed. This observation implies that the most important factor, which will result in the sublimation process, would be the temperature raised by the holder rather than the electron beam. Second, the In2O3 oxide shell could also play an important role in the sublimation process. The high melting temperature oxide shell could withstand the thermal energy but would introduce a local heating effect on the matrix with electron irradiation.33 Third, the speed of the video camera was 30 frames·s−1 such that the time resolution was 1/30 s. With better camera speed limitation, decomposition involving transformation of the faceted surface steps into the favored low-index plane and continuous sublimation in the unit of atomic layers (in Figure 4) were observed. The time resolution can be improved with high-standard computer-controlled display and digital camera. Furthermore, it is clear that decomposition was accompanied by intermittent introduction of a high-index surface plane. A higher-speed recorder is required to obtain further information in a future study. Moreover, the nonequilibrium interfacial energies of the In2Se3 matrix/In2O3 shell (γso), In2Se3 matrix/ vacuum (γsv), and In2O3 shell/vacuum (γov) should follow the relationship γsv ≫ γso > γov > 0, where γsv = (∂G/∂A)T,P,n. The
■
CONCLUSIONS In conclusion, we elucidated the detailed phenomenon of sublimation in one-dimensional In2Se3/In2O3 nanoheterostructures by in situ high-resolution TEM at high temperature in real time. During in situ observation of the sublimation process of the nanowires, evolution of the exposed In2Se3 surface progressed with time via two important mechanisms: mass desorption and stepwise migration, which are also the energetically favored processes in the ab initio calculation. The present observation not only provides valuable insight into atomic movement during the solid−vapor phase transition in this material but also could help to clarify atomic dynamics in the epitaxy process of many other material systems.
■
ASSOCIATED CONTENT
S Supporting Information *
Three figures showing characterization of nanoheterostructures, first-principles DFT calculations, and (0001) stereographic projection of In2Se3, and one table listing chemical potentials and formation energies (PDF). Three high-resolution TEM videos of mass-migration, stepwise-migration, and layeredsublimation processes (AVI). The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00255.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail [email protected]. *E-mail [email protected]. *E-mail [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS W.-W.W., C.-L.H., and L.-J.C. acknowledge support from Ministry of Science and Technology Grants 103-2221-E-009056 -MY2, 103-2221-E-009-222-MY3, 101-2218-E-008-014MY2, 103-2633-E-008-001-, 103-2221-E-006-077-MY3, 1032221-E-007-003 and 102-2633-M-007-002.
■
REFERENCES
(1) Hashimoto, A.; Suenaga, K.; Gloter, A.; Urita, K.; Iijima, S. Nature 2004, 430, 870−873. (2) van Huis, M. A.; Kunneman, L. T.; Overgaag, K.; Xu, Q.; Pandraud, G.; Zandbergen, H. W.; Vanmaekelbergh, D. l. Nano Lett. 2008, 8, 3959−3963. (3) Yoshida, H.; Takeda, S.; Uchiyama, T.; Kohno, H.; Homma, Y. Nano Lett. 2008, 8, 2082−2086. (4) Hsin, C.-L.; Lee, W.-F.; Huang, C.-T.; Huang, C.-W.; Wu, W.-W.; Chen, L.-J. Nano Lett. 2011, 11, 4348−4351. (5) Hsin, C.-L.; Wu, W.-W.; Chu, L.-W.; Hsu, H.-C.; Chen, L.-J. CrystEngComm 2011, 13, 3967−3970. (6) Wen, C.-Y.; Reuter, M. C.; Bruley, J.; Tersoff, J.; Kodambaka, S.; Stach, E. A.; Ross, F. M. Science 2009, 326, 1247−1250. (7) Huang, C.-W.; Chen, J.-Y.; Chiu, C.-H.; Wu, W.-W. Nano Lett. 2014, 14, 2759−2763. 5587
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry (8) Chen, K.-C.; Wu, W.-W.; Liao, C.-N.; Chen, L.-J.; Tu, K. N. Science 2008, 321, 1066−1069. (9) Han, X. D.; Zhang, Y. F.; Zheng, K.; Zhang, X. N.; Zhang, Z.; Hao, Y. J.; Guo, X. Y.; Yuan, J.; Wang, Z. L. Nano Lett. 2006, 7, 452− 457. (10) Huang, J. Y. Nano Lett. 2007, 7, 2335−2340. (11) Huang, J. Y.; Zheng, H.; Mao, S. X.; Li, Q.; Wang, G. T. Nano Lett. 2011, 11, 1618−1622. (12) Jin fan, H.; Knez, M.; Scholz, R.; Nielsch, K.; Pippel, E.; Hesse, D.; Zacharias, M.; Gosele, U. Nat. Mater. 2006, 5, 627−631. (13) Helveg, S.; Lopez-Cartes, C.; Sehested, J.; Hansen, P. L.; Clausen, B. S.; Rostrup-Nielsen, J. R.; Abild-Pedersen, F.; Norskov, J. K. Nature 2004, 427, 426−429. (14) Kwon, D.-H.; Kim, K. M.; Jang, J. H.; Jeon, J. M.; Lee, M. H.; Kim, G. H.; Li, X.-S.; Park, G.-S.; Lee, B.; Han, S.; Kim, M.; Hwang, C. S. Nat. Nano 2010, 5, 148−153. (15) Suenaga, K.; Wakabayashi, H.; Koshino, M.; Sato, Y.; Urita, K.; Iijima, S. Nat. Nano 2007, 2, 358−360. (16) Warner, J. H.; Rummeli, M. H.; Ge, L.; Gemming, T.; Montanari, B.; Harrison, N. M.; Buchner, B.; Briggs, G. A. D. Nat. Nano 2009, 4, 500−504. (17) Chen, J.-Y.; Hsin, C.-L.; Huang, C.-W.; Chiu, C.-H.; Huang, Y.T.; Lin, S.-J.; Wu, W.-W.; Chen, L.-J. Nano Lett. 2013, 13, 3671−3677. (18) In, J.; Yoo, Y.; Kim, J.-G.; Seo, K.; Kim, H.; Ihee, H.; Oh, S. H.; Kim, B. Nano Lett. 2010, 10, 4501−4504. (19) Zheng, H.; Rivest, J. B.; Miller, T. A.; Sadtler, B.; Lindenberg, A.; Toney, M. F.; Wang, L.-W.; Kisielowski, C.; Alivisatos, A. P. Science 2011, 333, 206−209. (20) Huang, Y.-T.; Huang, C.-W.; Chen, J.-Y.; Ting, Y.-H.; Lu, K.-C.; Chueh, Y.-L.; Wu, W.-W. ACS Nano 2014, 8, 9457−9462. (21) Lai, H.-Y.; Huang, C.-W.; Chiu, C.-H.; Wang, C.-W.; Chen, J.Y.; Huang, Y.-T.; Lu, K.-C.; Wu, W.-W. Anal. Chem. 2014, 86, 4348− 4353. (22) Huang, Y.-T.; Yu, S.-Y.; Hsin, C.-L.; Huang, C.-W.; Kang, C.-F.; Chu, F.-H.; Chen, J.-Y.; Hu, J.-C.; Chen, L.-T.; He, J.-H.; Wu, W.-W. Anal. Chem. 2013, 85, 3955−3960. (23) Yim, J. W. L.; Xiang, B.; Wu, J. Q. J. Am. Chem. Soc. 2009, 131, 14526−14530. (24) Simonsen, S. B.; Chorkendorff, I.; Dahl, S.; Skoglundh, M.; Sehested, J.; Helveg, S. J. Am. Chem. Soc. 2010, 132, 7968−7975. (25) Kim, K. J.; Jang, J. S.; Lee, J.-H.; Jee, Y.-J.; Jun, C.-S. Anal. Chem. 2009, 81, 8519−8522. (26) L’vov, B. V. Thermochim. Acta 2001, 373, 97−124. (27) Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169−11186. (28) Kresse, G.; Furthmüller, J. Comput. Mater. Sci. 1996, 6, 15−50. (29) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758−1775. (30) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B 1992, 46, 6671− 6687. (31) Perdew, J. P.; Wang, Y. Phys. Rev. B 1992, 45, 13244−13249. (32) Huang, C.-W.; Hsin, C.-L.; Wang, C.-W.; Chu, F.-H.; Kao, C.Y.; Chen, J.-Y.; Huang, Y.-T.; Lu, K.-C.; Wu, W.-W.; Chen, L.-J. Nanoscale 2012, 4, 4702−4706. (33) Yang, Y.; Scholz, R.; Berger, A.; Kim, D. S.; Knez, M.; Hesse, D.; Gösele, U.; Zacharias, M. Small 2008, 4, 2112−2117.
5588
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Direct Observation of Sublimation Behaviors in One-Dimensional In2Se3/In2O3 Nanoheterostructures Cheng-Lun Hsin,*,†,‡,§ Chun-Wei Huang,† Jui-Yuan Chen,† Kuo-Cheng Liao,∥ Po-Liang Liu,∥,⊥ Wen-Wei Wu,*,† and Lih-Juann Chen*,‡ †
Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu City, 300 Taiwan Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu City, 30013 Taiwan § Department of Electrical Engineering, National Central University, Taoyuan City, 32001 Taiwan ∥ Graduate Institute of Precision Engineering, National Chung Hsing University, Taichung City, 402 Taiwan ⊥ Department of Physics, Norwegian University of Science and Technology, 7491 Trondheim, Norway ‡
S Supporting Information *
ABSTRACT: Recently, in situ transmission electron microscopy (TEM) has provided a route to analyze structural characterization and chemical evolution with its powerful and unique applications. In this paper, we disclose the detailed phenomenon of sublimation on the atomic scale. In2Se3/In2O3 nanowires were synthesized via the vapor−liquid−solid mechanism and studied in an ultra-high-vacuum (UHV) TEM at high temperature in real time. During in situ observation of the sublimation process of the nanowires, the evolution and reconstruction of the exposed In2Se3 surface progressed in different manners with time. The surface structure was decomposed by mass-desorption and stepwise-migration processes, which are also energetically favored processes in the ab initio calculation. This study developed a new concept and will be essential in the development of atomic kinetics.
T
in nanomaterials from the information obtained in a study of transient structural transformation during sublimation. The surface steps were observed to be unstable and reconstructed through mass-desorption and stepwise-migration processes rather than the conventional macroscopic concept, which favors the individually unstable atom model. The atomic sublimation process has been unraveled and is expected to substantially advance contemporary scientific understanding. The model system selected to study the transient structural change was a chalcogenide compound, In2Se3. Detailed information about the nanostructures is provided in Supporting Information and elsewhere.4 The process of transient structural transformation of In2Se3 in the In2Se3/In2O3 core/shell nanowire during sublimation was investigated at the atomic scale in situ by use of a transmission electron microscope. Sublimation of In2Se3 occurred at 450 °C in UHV. For the mesoscopic-scale investigation, to determine the residual length of In2Se3, TEM images at the nanoscale were taken during the sublimation process, as presented in Figure 1af. From the recorded images, the disappearance of In2Se3 appears to be inclined to one side. A plot of the sublimation length as a function of time was extrapolated from these
he study of new phenomena at the nanoscale has been a widely developing research field in recent years.1−3 To disclose detailed properties of nanomaterials, microscopy techniques that can image in real time (with atomic resolution) were employed to study solid-state reactions,4−7 electromigration,8 nanomechanics,9−11 and structural12−17 and phase transformation.18−20 With the tremendous progress in nanotechnology, fundamental physics of nanowires has advanced remarkably, unraveling detailed information that can help us understand and develop new knowledge. With the employment of in situ TEM, we can directly observe the chemical reaction at atomic scale at the interface between different phases, providing unique, important, and interesting observations and analysis.21−25 The obtained data not only show time-dependent information but also allow us to look into the structural change at the surface/interface, providing a route to look into the details of chemical reaction. Conventionally, sublimation is an intense phase change between solid and vapor along with chemical reactions. The solid-to-vapor phenomenon is assumed to occur as long as the vapor pressure of the solid reaches a critical point, similar to the reverse process of epitaxy, a general process for developing novel semiconductor electronic devices. The question of whether the surface atoms are individually unstable or not has yet to be clarified. Here, we demonstrate that the sublimation process of one-dimensional nanostructures is vastly more complex than previously envisioned. We propose a new concept of the microscopic physics of dynamic behavior © 2015 American Chemical Society
Received: January 20, 2015 Accepted: May 5, 2015 Published: May 5, 2015 5584
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry
Figure 1. Series of TEM images illustrating the progression of sublimation of In2Se3. The images were obtained at elapsed times (a) 0, (b) 6.5, (c) 8, (d) 10, (e) 12, and (f) 14 min. The scale bar at the bottom corner corresponds to 200 nm. (g) Plot of sublimation length as a function of time.
Figure 2. Series of images clipped from the video illustrating the progression of transient structural transformation of In2Se3 at 450 °C. Pathways of the transient structural transformation are highlighted by red arrows, and the blue dashed lines are auxiliary markers that depict the same position in different frames. Black dotted lines illustrate the corresponding lattice planes. The first two numbers of the time scale are in units of minutes and seconds, respectively, and the following numbers with the prime are in units of 1/60 s: (a) 3:25:02′, (b) 3:35:50′, (c) 3:57:28′, (d) 4:51:16′, (e) 4:59:22′, (f) 5:08:12′, (g) 5:17:02′, (h) 5:19:12′. The scale bar at the bottom corner corresponds to 2 nm. (i) Schematic illustration of structural evolution of surface steps over the entire nanowire.
In2Se3 nanowires. The core/shell structure was obtained by oxidation after synthesis at ambient conditions. Nanostructure Characterization and in Situ Observation. The nanowires were drop-cast on a Mo grid with a thin C film to support the nanostructures for the high-temperature study. The observation was conducted in a JEOL 2000 V UHVTEM with a base pressure of 3 × 10−10 Torr and 200 keV acceleration voltage. Before the in situ TEM observation, the sample was placed into a heating holder for degassing in the pretreatment chamber at 200 °C for 5 min. During real-time observation, the temperature was controlled and the pressure would be raised to ∼1 × 10−7 Torr. Video recorder had a time resolution of 1/30 s. First-Principles Density Functional Theory Calculations. First-principles density functional theory (DFT) calculations were conducted by use of the Vienna ab-initio simulation package.27−29 Projected augmented wave pseudopotentials were used to describe the valence electrons of In (5s2
recorded images. The linear tendency indicated that the average reaction rate and mechanism should be attributed to a Langmuir-type vaporization process at the In2Se3/vacuum interface, where direct sublimation occurred with no flux recondensing from the vapor.26
■
EXPERIMENTAL SECTION Synthesis. Single-crystal, 1−30 Ω·cm, (001)-oriented silicon wafers were used as substrates. A 2-nm-thick Au thin film was deposited on each substrate as a catalyst. Synthesis of the In2Se3 nanowires was performed in a horizontal tube furnace. During the growth process, 0.82 g of Se powder was placed upstream and a 1.2 g In slug was placed downstream. The samples were positioned 2 cm apart from the In source at the rear. The Se and In sources were heated to 240 and 600 °C, respectively, and transported by a carrier gas composed of 80% Ar and 20% H2 at 200 sccm at 1 Torr. When the set temperature was reached, the process was held for another 2 h and then cooled down to room temperature to obtain the 5585
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry 5p1) and Se (4s2 4p4) within the PW91 generalized gradient approximation for exchange and correlation energy.30,31
■
RESULTS AND DISCUSSION A series of high-resolution images from the recorded video are presented in Figure 2 to illustrate the transformation at the atomic scale. The pathways leading to transient structural transformation in the next frame are highlighted by red arrows. The single/double blue dashed lines are auxiliary marks that depict the same position in different frames. The black dotted lines depict the corresponding lattice planes. The first two numbers of the time scale in the figures are in units of minutes and seconds, respectively, and the following numbers with the prime are in units of 1/60 s. At the beginning of the recording (Figure 2a), the surface of In2Se3 was composed of facets with low indices. The convex facets were observed to depart from the matrix, termed mass desorption, and transformed into concave facets within 1 s. Then the field of view was kept the same for ∼10 s (Figure 2b). At the top of the frame in Figure 2b, mass desorption occurred again for the lattice planes to come off, becoming the facet at the top of Figure 2c. It is clear that mass desorption is an important process in reconstruction of the steps at the surface as well as at the bottom of Figure 2c. The (1̅100) plane in Figure 2c sublimated toward the downside, making the (011̅0) plane the most exposed surface. With the progression of mass desorption, one of the low-index planes became dominant at the surface, as illustrated in Figure 2d. In Figure 2d,e, the double blue dashed line represents the identical field of interest. A surface marked with a single-dashed line is observed at the bottom of Figure 2e. Both of the steps were moving downward, resulting in a new (0110̅ ) plane becoming the surface plane, termed stepwise migration. The single blue dashed line in Figure 2e,f represents the same position. At the bottom of the field of view in Figure 2f, a small step departed from the matrix, followed by downward movement of the other surface step, resulting in the new surface structure was shown in Figure 2g. After stepwise migration of the surface step in Figure 2g, the surface of In2Se3 became flat with only the (0110̅ ) surface plane. The projected cross section of structural evolution of the In2Se3 surface steps is schematically illustrated in Figure 2i. Two sublimation processes (mass desorption and stepwise migration) are essential for deconstruction of the surface. For the process of mass desorption to introduce a new surface, the surface steps were composed of the (011̅0) and (1̅100) planes in Figure 3a, which are highlighted with the double-dashed lines. Within a very short time, atoms in the exposed rows of the (011̅0) plane came off, introducing a new (1̅100) plane, as observed at the bottom of Figure 3b. Subsequently, sublimation of the existing and exposed (1̅100) plane occurred. A transient step was induced with a new (0110̅ ) plane at the top of Figure 3c, followed by completion of the mass-desorption process. After the process, the surface in Figure 3d was composed of identical low-index planes with a diminished matrix volume. The removal stages of each atomic layer are well resolved in the recorded video (Supporting Information, Movie S2). Another important process in sublimation is stepwise migration. A pertinent example is illustrated in Figure 4. With the presence of the (11̅ 00) plane (black dotted line) in Figure 4a, the surface step became unstable. To eliminate the excess energy associated with the plane, the (011̅0) facet layers (yellow dashed line) decomposed along with the atomic planes
Figure 3. Series of images clipped from the video illustrating progression of the mass-desorption process: (a) 3:25:10′, (b) 3:25:20′, (c) 3:25:28′, (d) 3:26:00′. The double-dashed lines highlight the primal surface step. The scale bar at the bottom corner corresponds to 2 nm.
Figure 4. (a−d) Series of images clipped from the video illustrating progression of the stepwise-migration process: (a) 5:28:12′, (b) 5:28:18′, (c) 5:28:20′, (d) 5:28:22′. The single-dashed line highlights the surface. (e, f) Series of images clipped from the video illustrating sublimation progression of the (011̅0) plane: (e) 5:37:36′, (f) 5:37:52′. The scale bar at the bottom corner corresponds to 2 nm. Drawings are included to guide the eye in locating the evolution of the surface steps.
(Figure 4a−d). The migration of surface atoms resulted in removal of the (1̅100) plane. Ultimately, the surface was predominantly composed of the low-index, energetically favorable (011̅0) plane, causing the disappearance of the (11̅ 00) plane. In the recorded video, the reaction stages of the atomic planes are well resolved (Supporting Information, Movie S3). After formation of the energetically favorable plane, the matrix continued to sublime at the high temperature, while the exposed atomic layers vaporized and continued to sublimate layer-by-layer, termed layer sublimation, resulting in the formation of a nanotube. In the frames in Figure 4e,f, the 5586
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry
Gibbs free energy of the In2Se3 matrix/vacuum interface at sublimation, G, should be a positive value, causing sublimation of the In2Se3 matrix. These data indicated that this transient structural transformation also manifested the role of energy minimization in the dynamic behavior of materials, which will be helpful in understanding the reverse process, that is, epitaxy.
reference markers are indicated by arrows. In Figure 4e,f, the distance between the reference point and the surface plane (yellow dashed line) increased, substantiating that the atomic layers sublimated with the unit in atomic layers. The reaction stages of each atomic layer are also well recorded in the video (Supporting Information, Movie S4). The high-resolution TEM images in Figures 3 and 4 clearly illustrate the fluctuating surface step edges of the In2Se3 matrix. The appearance of two distinguishable In2Se3 (1̅100) and (011̅0) surfaces during sublimation was not expected. A complete survey of the structural trends of alignment primarily consisting of low-index (1̅100) and (011̅0) facets with edges along the [1̅1̅20] and [2̅110] directions are depicted in Supporting Information. A plausible mechanism for the zigzag behavior during sublimation is supported by theoretical simulations of the surface formation energy by use of the Vienna ab-initio simulation package.27−29 The In2Se3 (1̅100) and In2Se3(0110̅ ) surfaces are energetically equivalent and stable within the thermodynamically allowed ranges as μbulk In + bulk f bulk Hf [In2Se3] ≤ μIn ≤ μbulk In or μSe + H [In2Se3] ≤ μSe ≤ μSe . As expected, the region observed in each frame from the highresolution image is localized, and dominant facets all over the cross-sectional surface could be bimodal from the lowmagnification images. However, the chemical potentials of In and Se in sublimation under vacuum extend beyond the range of allowed chemical potentials, leading to the gradual decomposition of the surface steps (Supporting Information). There are several points that need to be addressed with direct observation of the gradual transient surface structural transformation from faceted steps to one dominant facet. First, during TEM imaging, the interaction between the substance and the incident high-energy electrons generates heat and increases the temperature. Thus, the equilibrium temperature could be higher than that measured (450 °C). Additionally, the electron beam could also decompose the structure of In2Se3.32 We also have performed experiments at lower and higher temperatures. At lower temperature, the sublimation could be too slow to proceed, or In2Se3 will dissociate first and then a small amount of nanoparticles will form on the surface. After several tens of seconds, these nanoparticles will shrink gradually. At higher temperature, the sublimation behaviors are the same as what we have observed but with faster progress speed. This observation implies that the most important factor, which will result in the sublimation process, would be the temperature raised by the holder rather than the electron beam. Second, the In2O3 oxide shell could also play an important role in the sublimation process. The high melting temperature oxide shell could withstand the thermal energy but would introduce a local heating effect on the matrix with electron irradiation.33 Third, the speed of the video camera was 30 frames·s−1 such that the time resolution was 1/30 s. With better camera speed limitation, decomposition involving transformation of the faceted surface steps into the favored low-index plane and continuous sublimation in the unit of atomic layers (in Figure 4) were observed. The time resolution can be improved with high-standard computer-controlled display and digital camera. Furthermore, it is clear that decomposition was accompanied by intermittent introduction of a high-index surface plane. A higher-speed recorder is required to obtain further information in a future study. Moreover, the nonequilibrium interfacial energies of the In2Se3 matrix/In2O3 shell (γso), In2Se3 matrix/ vacuum (γsv), and In2O3 shell/vacuum (γov) should follow the relationship γsv ≫ γso > γov > 0, where γsv = (∂G/∂A)T,P,n. The
■
CONCLUSIONS In conclusion, we elucidated the detailed phenomenon of sublimation in one-dimensional In2Se3/In2O3 nanoheterostructures by in situ high-resolution TEM at high temperature in real time. During in situ observation of the sublimation process of the nanowires, evolution of the exposed In2Se3 surface progressed with time via two important mechanisms: mass desorption and stepwise migration, which are also the energetically favored processes in the ab initio calculation. The present observation not only provides valuable insight into atomic movement during the solid−vapor phase transition in this material but also could help to clarify atomic dynamics in the epitaxy process of many other material systems.
■
ASSOCIATED CONTENT
S Supporting Information *
Three figures showing characterization of nanoheterostructures, first-principles DFT calculations, and (0001) stereographic projection of In2Se3, and one table listing chemical potentials and formation energies (PDF). Three high-resolution TEM videos of mass-migration, stepwise-migration, and layeredsublimation processes (AVI). The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00255.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail [email protected]. *E-mail [email protected]. *E-mail [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS W.-W.W., C.-L.H., and L.-J.C. acknowledge support from Ministry of Science and Technology Grants 103-2221-E-009056 -MY2, 103-2221-E-009-222-MY3, 101-2218-E-008-014MY2, 103-2633-E-008-001-, 103-2221-E-006-077-MY3, 1032221-E-007-003 and 102-2633-M-007-002.
■
REFERENCES
(1) Hashimoto, A.; Suenaga, K.; Gloter, A.; Urita, K.; Iijima, S. Nature 2004, 430, 870−873. (2) van Huis, M. A.; Kunneman, L. T.; Overgaag, K.; Xu, Q.; Pandraud, G.; Zandbergen, H. W.; Vanmaekelbergh, D. l. Nano Lett. 2008, 8, 3959−3963. (3) Yoshida, H.; Takeda, S.; Uchiyama, T.; Kohno, H.; Homma, Y. Nano Lett. 2008, 8, 2082−2086. (4) Hsin, C.-L.; Lee, W.-F.; Huang, C.-T.; Huang, C.-W.; Wu, W.-W.; Chen, L.-J. Nano Lett. 2011, 11, 4348−4351. (5) Hsin, C.-L.; Wu, W.-W.; Chu, L.-W.; Hsu, H.-C.; Chen, L.-J. CrystEngComm 2011, 13, 3967−3970. (6) Wen, C.-Y.; Reuter, M. C.; Bruley, J.; Tersoff, J.; Kodambaka, S.; Stach, E. A.; Ross, F. M. Science 2009, 326, 1247−1250. (7) Huang, C.-W.; Chen, J.-Y.; Chiu, C.-H.; Wu, W.-W. Nano Lett. 2014, 14, 2759−2763. 5587
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
Article
Analytical Chemistry (8) Chen, K.-C.; Wu, W.-W.; Liao, C.-N.; Chen, L.-J.; Tu, K. N. Science 2008, 321, 1066−1069. (9) Han, X. D.; Zhang, Y. F.; Zheng, K.; Zhang, X. N.; Zhang, Z.; Hao, Y. J.; Guo, X. Y.; Yuan, J.; Wang, Z. L. Nano Lett. 2006, 7, 452− 457. (10) Huang, J. Y. Nano Lett. 2007, 7, 2335−2340. (11) Huang, J. Y.; Zheng, H.; Mao, S. X.; Li, Q.; Wang, G. T. Nano Lett. 2011, 11, 1618−1622. (12) Jin fan, H.; Knez, M.; Scholz, R.; Nielsch, K.; Pippel, E.; Hesse, D.; Zacharias, M.; Gosele, U. Nat. Mater. 2006, 5, 627−631. (13) Helveg, S.; Lopez-Cartes, C.; Sehested, J.; Hansen, P. L.; Clausen, B. S.; Rostrup-Nielsen, J. R.; Abild-Pedersen, F.; Norskov, J. K. Nature 2004, 427, 426−429. (14) Kwon, D.-H.; Kim, K. M.; Jang, J. H.; Jeon, J. M.; Lee, M. H.; Kim, G. H.; Li, X.-S.; Park, G.-S.; Lee, B.; Han, S.; Kim, M.; Hwang, C. S. Nat. Nano 2010, 5, 148−153. (15) Suenaga, K.; Wakabayashi, H.; Koshino, M.; Sato, Y.; Urita, K.; Iijima, S. Nat. Nano 2007, 2, 358−360. (16) Warner, J. H.; Rummeli, M. H.; Ge, L.; Gemming, T.; Montanari, B.; Harrison, N. M.; Buchner, B.; Briggs, G. A. D. Nat. Nano 2009, 4, 500−504. (17) Chen, J.-Y.; Hsin, C.-L.; Huang, C.-W.; Chiu, C.-H.; Huang, Y.T.; Lin, S.-J.; Wu, W.-W.; Chen, L.-J. Nano Lett. 2013, 13, 3671−3677. (18) In, J.; Yoo, Y.; Kim, J.-G.; Seo, K.; Kim, H.; Ihee, H.; Oh, S. H.; Kim, B. Nano Lett. 2010, 10, 4501−4504. (19) Zheng, H.; Rivest, J. B.; Miller, T. A.; Sadtler, B.; Lindenberg, A.; Toney, M. F.; Wang, L.-W.; Kisielowski, C.; Alivisatos, A. P. Science 2011, 333, 206−209. (20) Huang, Y.-T.; Huang, C.-W.; Chen, J.-Y.; Ting, Y.-H.; Lu, K.-C.; Chueh, Y.-L.; Wu, W.-W. ACS Nano 2014, 8, 9457−9462. (21) Lai, H.-Y.; Huang, C.-W.; Chiu, C.-H.; Wang, C.-W.; Chen, J.Y.; Huang, Y.-T.; Lu, K.-C.; Wu, W.-W. Anal. Chem. 2014, 86, 4348− 4353. (22) Huang, Y.-T.; Yu, S.-Y.; Hsin, C.-L.; Huang, C.-W.; Kang, C.-F.; Chu, F.-H.; Chen, J.-Y.; Hu, J.-C.; Chen, L.-T.; He, J.-H.; Wu, W.-W. Anal. Chem. 2013, 85, 3955−3960. (23) Yim, J. W. L.; Xiang, B.; Wu, J. Q. J. Am. Chem. Soc. 2009, 131, 14526−14530. (24) Simonsen, S. B.; Chorkendorff, I.; Dahl, S.; Skoglundh, M.; Sehested, J.; Helveg, S. J. Am. Chem. Soc. 2010, 132, 7968−7975. (25) Kim, K. J.; Jang, J. S.; Lee, J.-H.; Jee, Y.-J.; Jun, C.-S. Anal. Chem. 2009, 81, 8519−8522. (26) L’vov, B. V. Thermochim. Acta 2001, 373, 97−124. (27) Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169−11186. (28) Kresse, G.; Furthmüller, J. Comput. Mater. Sci. 1996, 6, 15−50. (29) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758−1775. (30) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B 1992, 46, 6671− 6687. (31) Perdew, J. P.; Wang, Y. Phys. Rev. B 1992, 45, 13244−13249. (32) Huang, C.-W.; Hsin, C.-L.; Wang, C.-W.; Chu, F.-H.; Kao, C.Y.; Chen, J.-Y.; Huang, Y.-T.; Lu, K.-C.; Wu, W.-W.; Chen, L.-J. Nanoscale 2012, 4, 4702−4706. (33) Yang, Y.; Scholz, R.; Berger, A.; Kim, D. S.; Knez, M.; Hesse, D.; Gösele, U.; Zacharias, M. Small 2008, 4, 2112−2117.
5588
DOI: 10.1021/acs.analchem.5b00255 Anal. Chem. 2015, 87, 5584−5588
