Graphene: Chains and Edges

August 19, 2009 – Ulm and AIST in Tsukuba show, independently, how the electron beam thins a graphene bridge down to a single chain of carbon atoms. AC-HRTEM at 80 kV on the image-side corrected FEI Titan in Ulm follows bridges, adsorbates and edges atom by atom; the JEOL 2010F at AIST and DFT confirm why chains are the end product. First come the edges: a zigzag edge switches to the predicted pentagon–heptagon reconstruction and back, which Berkeley compares with its own edge study.

Two research groups have shown, in parallel and independently, a simple and reliable route to an exotic form of carbon: the single-atom linear carbon chain[1],[2]. Both start from a narrow bridge, or constriction, in graphene and let the electron beam of the microscope thin it down atom by atom, until only one or two carbon chains are left (Fig. 1). The same experiments also show how the edges of graphene behave under the beam. The news thus has two parts and two results, each checked against an independent study: first the chains, compared with Jin et al. at AIST in Tsukuba, and then the edges they grow from, compared with Girit et al. at Berkeley.

Linear carbon chains exist in two electronically different forms, polyyne (–C≡C–)ₙ with alternating single and triple bonds and cumulene (=C=C=)ₙ with double bonds only. Their predicted properties have attracted wide interest in nanotechnology, for example as the thinnest possible conducting wire[3]–[5]. Spectroscopic evidence for such chains had been found on surfaces and inside carbon nanotubes[6],[7], but a single chain had hardly ever been handled and observed as an object of its own in an experimental set-up.

In Ulm, Andrey Chuvilin, Jannik Meyer, Gerardo Algara-Siller and Ute Kaiser used the image-side Cs-corrected FEI Titan 80-300 at 80 kV. At AIST in Tsukuba, Chuanhong Jin, Kazu Suenaga and Sumio Iijima, together with Haiping Lan and Lianmao Peng of Peking University, used a JEOL 2010F with a CEOS image corrector at 80 or 120 kV. In both microscopes, the beam knocks carbon atoms out of the edges of holes in graphene (knock-on damage)[8], while the closed graphene lattice stays intact. Two holes that grow towards each other leave a bridge between them, which becomes narrower and narrower. The process resembles the fabrication of metallic quantum wires by thinning a metal bridge[9],[10].

Part 1: Carbon chains under the electron beam
In Ulm, the team looked for two nearby holes in a clean region of a single graphene layer and recorded a continuous image series, with 1 s exposures every 4 s. As the holes grow, the graphene bridge between them narrows down. Below a width of about 1 nm, the whole bridge, not only its edges, begins to rebuild itself. In one example, the bridge first takes up pentagons and heptagons, then turns into three parallel carbon chains, rebuilds into a network of polygons with five to eight sides, and finally becomes a double and then a single chain before the two holes merge (Video 1). The chains appear darker than the graphene lattice, exactly as the image simulation predicts for a single-atom chain. In more than 50 % of all observed bridges, a single chain of typically 10–15 carbon atoms was the final product, and it stayed stable for up to 2 min under the intense beam. Videos 2 and 3 show two more bridges.

Video 1: Transformation of a graphene ribbon into a single carbon chain, FEI Titan at 80 kV (SI video S1).

Video 2: Time evolution of a second graphene constriction, similar to Video 1 (SI video S2).

Video 3: Time evolution of a third graphene constriction that turns into a chain (SI video S3).

One of the intermediate structures was stable for 2 min, longer than any other polygon network in the data, at a dose of about 4 × 109 e−/nm2. An averaged image of ten frames shows that it is locally the structure of pentaheptite, a planar carbon allotrope of pentagons and heptagons that had been predicted in 1996[11] but never observed (Fig. 2). Bridges are a good place to find such structures: they are free to rebuild in one direction, and the beam keeps randomizing some atoms, so that configurations in a local energy minimum survive longer than others.

That carbon chains appear so often cannot be a coincidence of atoms removed at random. Several bridges even turn into two parallel chains, which shows that the chain is an energetically favoured configuration. The groups at AIST and Peking University reached the same conclusion with a different microscope, a different sample and density functional theory (DFT). The first validation box compares both studies and holds the AIST figures (Fig. 3, Fig. 4, Fig. 5 and Fig. 6).

▸ Validation: FEI Titan at 80 kV (Ulm) vs. JEOL 2010F at 120 kV (AIST in Tsukuba)

Chains do not only form from graphene. They also appear where thin amorphous carbon adsorbates on the graphene shrink under the beam (Fig. 7, Videos 4 and 5). Here, the low-contrast graphene underneath serves as a support that makes the process visible. Some of the adsorbates are single-layer networks of carbon pentagons, hexagons and heptagons. The chains show a slight preference for the zigzag direction of the graphene below, but again no zigzag shape of the chain itself. The precursor of the chains thus ranges from crystalline graphene to amorphous carbon, which points to a rather general self-organization of carbon at low atom density under continuous energy input.

Video 4: Chain formation from a constriction in amorphous adsorbates on graphene, first example (SI video S4).

Video 5: Chain formation from a constriction in amorphous adsorbates on graphene, second example (SI video S5).

▸ Recipe: samples and microscopes

Challenges named by the authors
Neither study can tell polyyne from cumulene: at the present resolution, the signal-to-noise ratio is not sufficient to distinguish alternating from equal bond lengths. Angled chains, such as alkane or polyacetylene, were ruled out only statistically, assuming random orientations of the chains. Ulm finds the chains surprisingly robust: one minute in the beam corresponds to about 2 × 109 e−/nm2, five orders of magnitude above the usual dose limit of about 104 e−/nm2 for organic molecules[18]. Why, is not understood; experiments at other electron energies should help. The non-local switching of a whole edge also remains to be explained. Little is known so far about the electrical properties of the sp-hybridized chains.

Part 2: Graphene at the edge
Wherever the beam opens a hole in graphene, it creates an edge, and the chains grow out of these edges. The edges keep changing under the beam. One edge in Ulm switched several times between a clean zigzag edge and a reconstructed edge of alternating pentagons and heptagons with twice the period of the zigzag edge (Fig. 8f,g and Video 6), one of the reconstructions predicted by Koskinen and co-workers[16]. It switched back and forth between two exposures, along about 5 nm of edge in one step. Random knock-on events of single atoms cannot produce this; more likely, both are stable configurations, and the beam supplies the energy to switch between them, possibly through a lattice vibration. At the same time, Girit et al. followed graphene edges at room temperature at 80 kV on the TEAM 0.5 microscope at Berkeley[19]; their armchair and zigzag edges stand next to the Ulm edges in Fig. 8a,b. The second validation box compares the two studies.

Video 6: A graphene edge switching several times between the zigzag edge and the reconstructed edge of pentagons and heptagons (SI video S6).

▸ Validation: FEI Titan at 80 kV (Ulm) vs. TEAM 0.5 at 80 kV (Berkeley)

The edges are where the chains are born. Chain loops can be found reliably by watching the edges of holes for some time (Fig. 8c–e). A chain separates from the edge, may bridge a gap and often merges back into the edge later. The double loop of Fig. 8e also shows a few smaller loops, signs of a further edge reconstruction. How an edge recedes under the beam was filmed at the same time in Oxford: Warner and co-workers watched zigzag edges in few-layer graphene on an FEI Titan3 at 80 kV and saw the carbon atoms leave row by row, starting at the step between two zigzag rows and running along the edge like a zipper (Fig. 9)[20]. The same row-by-row loss thins a narrow ribbon: when two holes grow towards each other, the beam strips the bridge between them row by row from both edges until only a chain is left, and finally nothing, when the chain breaks and the holes merge.

The thinnest wire
Many electronic applications of graphene need ribbons and constrictions only nanometres wide. Narrow ribbons rebuild themselves, which has to be taken into account, while carbon chains are the ultimate constriction and might serve as conducting channels in future devices. Because they form by self-organization at a predictable place, electron-beam fabrication could become a route to such devices. Within the SALVE project, a microscope with higher resolution at low voltages is being developed, which could make it possible to tell the two kinds of linear chains apart. The work in Ulm was supported by the German Research Foundation (DFG) under SFB 569 and by the SALVE project.

Resource:
Chuvilin, A., Meyer, J. C., Algara-Siller, G., Kaiser, U. (2009).
From graphene constrictions to single carbon chains.
New J. Phys. 11(8), 083019.
https://doi.org/10.1088/1367-2630/11/8/083019

Download this news article as PDF – for reading on paper, taking notes or filing.

References
  1. Chuvilin, A., Meyer, J. C., Algara-Siller, G., & Kaiser, U. (2009). From Graphene Constrictions to Single Carbon Chains. New Journal of Physics, 11(8), 083019. https://doi.org/10.1088/1367-2630/11/8/083019

  2. Jin, C., Lan, H., Peng, L., Suenaga, K., & Iijima, S. (2009). Deriving Carbon Atomic Chains From Graphene. Physical Review Letters, 102(20), 205501. https://doi.org/10.1103/PhysRevLett.102.205501

  3. Lang, N. D., & Avouris, P. (2000). Carbon-Atom Wires: Charge-Transfer Doping, Voltage Drop, and the Effect of Distortions. Physical Review Letters, 84(2), 358. https://doi.org/10.1103/PhysRevLett.84.358

  4. Tongay, S., Senger, R. T., Dag, S., & Ciraci, S. (2004). Ab-Initio Electron Transport Calculations of Carbon Based String Structures. Physical Review Letters, 93(13), 136404. https://doi.org/10.1103/PhysRevLett.93.136404

  5. Okano, S., & Tománek, D. (2007). Effect of Electron and Hole Doping on the Structure of C, Si, and S Nanowires. Physical Review B, 75(19), 195409. https://doi.org/10.1103/PhysRevB.75.195409

  6. Derycke, V., Soukiassian, P., Mayne, A., Dujardin, G., & Gautier, J. (1998). Carbon Atomic Chain Formation on the β-SiC(100) Surface by Controlled sp → sp3 Transformation. Physical Review Letters, 81(26), 5868. https://doi.org/10.1103/PhysRevLett.81.5868

  7. Zhao, X., Ando, Y., Liu, Y., Jinno, M., & Suzuki, T. (2003). Carbon Nanowire Made of a Long Linear Carbon Chain Inserted Inside a Multiwalled Carbon Nanotube. Physical Review Letters, 90(18), 187401. https://doi.org/10.1103/PhysRevLett.90.187401

  8. Banhart, F. (1999). Irradiation Effects in Carbon Nanostructures. Reports on Progress in Physics, 62(8), 1181. https://doi.org/10.1088/0034-4885/62/8/201

  9. Kondo, Y., & Takayanagi, K. (1997). Gold Nanobridge Stabilized by Surface Structure. Physical Review Letters, 79(18), 3455. https://doi.org/10.1103/PhysRevLett.79.3455

  10. Bettini, J., Sato, F., Coura, P. Z., Dantas, S. O., Galvão, D. S., & Ugarte, D. (2006). Experimental Realization of Suspended Atomic Chains Composed of Different Atomic Species. Nature Nanotechnology, 1(3), 182–185. https://doi.org/10.1038/nnano.2006.132

  11. Crespi, V. H., Benedict, L. X., Cohen, M. L., & Louie, S. G. (1996). Prediction of a Pure-Carbon Planar Covalent Metal. Physical Review B, 53(20), R13303. https://doi.org/10.1103/PhysRevB.53.R13303

  12. Perdew, J. P., Chevary, J. A., Vosko, S. H., Jackson, K. A., Pederson, M. R., Singh, D. J., & Fiolhais, C. (1992). Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient Approximation for Exchange and Correlation. Physical Review B, 46(11), 6671. https://doi.org/10.1103/PhysRevB.46.6671

  13. Blöchl, P. E. (1994). Projector Augmented-Wave Method. Physical Review B, 50(24), 17953. https://doi.org/10.1103/PhysRevB.50.17953

  14. Kresse, G., & Furthmüller, J. (1996). Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Physical Review B, 54(16), 11169. https://doi.org/10.1103/PhysRevB.54.11169

  15. Henkelman, G., Uberuaga, B. P., & Jónsson, H. (2000). A Climbing Image Nudged Elastic Band Method for Finding Saddle Points and Minimum Energy Paths. The Journal of Chemical Physics, 113(22), 9901–9904. https://doi.org/10.1063/1.1329672

  16. Koskinen, P., Malola, S., & Häkkinen, H. (2008). Self-Passivating Edge Reconstructions of Graphene. Physical Review Letters, 101(11), 115502. https://doi.org/10.1103/PhysRevLett.101.115502

  17. Okada, S. (2008). Energetics of Nanoscale Graphene Ribbons: Edge Geometries and Electronic Structures. Physical Review B, 77(4), 041408. https://doi.org/10.1103/PhysRevB.77.041408

  18. Malac, M., Beleggia, M., Egerton, R., & Zhu, Y. (2008). Imaging of Radiation-Sensitive Samples in Transmission Electron Microscopes Equipped With Zernike Phase Plates. Ultramicroscopy, 108(2), 126–140. https://doi.org/10.1016/j.ultramic.2007.03.008

  19. Girit, Ç. Ö., Meyer, J. C., Erni, R., Rossell, M. D., Kisielowski, C., Yang, L., Park, C.-H., Crommie, M. F., Cohen, M. L., Louie, S. G., & Zettl, A. (2009). Graphene at the Edge: Stability and Dynamics. Science, 323(5922), 1705–1708. https://doi.org/10.1126/science.1166999

  20. Warner, J. H., Rümmeli, M. H., Ge, L., Gemming, T., Montanari, B., Harrison, N. M., Büchner, B., & Briggs, G. A. D. (2009). Structural Transformations in Graphene Studied With High Spatial and Temporal Resolution. Nature Nanotechnology, 4(8), 500–504. https://doi.org/10.1038/nnano.2009.194