Carbon nanotube-dependent synthesis of armchair graphene nanoribbons

March 01, 2022 - Sub-nanometer armchair graphene nanoribbons (GNRs) with moderate band gap have great potential towards novel nanodevices. Scientists from the Sun Yat-sen university the Shanghai Tech university (CN), the universities of Ulm (Germany) and Vienna (Austria), as well as the National Institute of Advanced Industrial Science and Technology in Tsukuba (JP) have explored how the diameter and metallicity of SWCNTs influence the synthesis of the GNRs which may enable the high-yield production of certain armchair graphene nanoribbons in large scale.

Graphene nanoribbons (GNRs) exhibit unique electronic properties, such as tunable bandgap, high mobility, and high current carrying capacity, which are superior to gapless graphene and enable the development of a new generation of electronic nanodevices1. Graphene nanoribbons are distinguished by their width and edges. Armchair GNRs (AGNRs) and zigzag GNRs are the two main groups of GNRs. Zigzag GNRs have a metallic character, while AGNRs have a width-dependent electronic property. AGNRs are classified according to the number of dimer lines (n) across the bandwidth: n = 3p, 3p + 1, or 3p + 2, where p is an integer. The AGNRs with n = 3p or 3p + 1 have a relatively large band gap. The AGNRs with n = 3p + 2, on the other hand, usually have a small band gap below 0.1 eV2.

In general, both top-down and bottom-up methods can be used to produce GNRs. The width of the obtained GNRs is usually larger than 10 nm and the edge cannot be well controlled. In contrast, in the bottom-up method, starting from designed precursor molecules, the edge and width of the synthesized GNRs can be precisely controlled3,4. Advanced polyaromatic hydrocarbon molecules (PAHs) are ideal raw materials as precursors. The GNRs can be formed by polymerization of the monomeric precursor molecules, e.g., on the substrate surface3,4.

In addition, single-walled CNTs (SWCNTs) are promising nanoreactors 5,6 that provide a tubular confinement space for nanometer-scale chemical reactions. Various types of metastable one-dimensional (1D) and qusi-1D nanostructures have been synthesized in CNTs, such as linear carbon chains7, sulfur chains8, lanthanum chains9, and ion chains10. GNRs have also been prepared by converting PAHs into SWCNTs11,12,13 but only recently have GNRs with specific width and armchair edge been successfully synthesized by using ferrocene molecules as precursors14,15. In this case, the mechanism for the formation of AGNRs is completely different from the previous one based on the polymerization of monomers, since ferrocene molecules are first decomposed and then catalytically converted into AGNRs. Therefore, not only the precursor molecules but also the SWCNTs play an important role in the synthesis. The limited space of SWCNTs enables effective tailoring of the internal nanostructures17.

The scientists employed a procedure shown in Fig. 1. The samples were characterized using the Thermo Fisher sub-Å low-voltage electron microscope (SALVE) equipped with a CEOS-developed chromatic and spherical aberration (Cc/Cs) corrector that fully corrects fifth-order axial geometric aberrations (including Cs and C5), third-order off-center geometric aberrations, and first-order chromatic aberrations, and also equipped with a Thermo Fisher Ceta 4K CMOS camera.

Raman spectroscopy of individual graphene nanoribbons

6 types of SWCNTs with different average diameters of 1.0, 1.1, 1.3, 1.43, 1.56, and 1.7 nm were designated as e1.0, H1.1, e1.3, e1.43, e1.56, and e1.7, respectively. The opened SWCNTs were measured with two excitations to evaluate their quality and diameter difference. As shown in Fig. 2, the radial breathing mode (RBM) of the SWCNTs from 100 to 300 cm-1 indicates the differential diameter distribution of the samples and confirms the average diameters obtained from the absorption. The split G-band with the G+ mode and the G- mode at 1.590 and 1.570 cm-1 originates from the in-plane vibrations along the tube axis and the circumferential direction, respectively17. The small D-bands of e1.3, e1.43, e1.56, and e1.7 indicate high quality SWCNTs. The D-bands of e1.0 and H1.1 are larger than the others, so the defects could affect the growth of GNRs.

The transformed samples show additional Raman modes compared to the opened SWCNTs (Fig. 3). Previously, the team had already found that these Raman modes belong to 6-AGNRs and 7-AGNRs, which are clearly observed upon excitation by lasers with wavelengths of 633 and 568 nm, respectively, due to the resonance effect15. The remaining peaks between 1,200 and 1,500 cm−1 are related to the oscillations near the edge of the AGNRs18. Similarly, the G-band of the AGNRs appears at about 1,600 cm−1, which strongly overlaps with the G-band of the SWCNTs, making it very difficult to resolve.

As can be seen in Fig. 3(c), the Raman modes of the 7-AGNRs are still clearly visible even with off-resonance excitations, but the signal of the 6-AGNRs is not found because the excitation energy is too far from their resonance window. With these excitations, the G-band at 1.606 cm-1 of the 7-AGNRs can finally be observed directly, since the intensity of the G-band of the SWCNTs decreases even more than that of the 7-AGNRs. The G-band intensity of the 7-AGNRs is very sensitive to environmental effects when excited in the resonant state. Therefore, excitation outside the resonance window was chosen to evaluate the synthesis process. As shown in Fig. 3(d), surprisingly, the synthesis saturates extremely fast, although it should be taken into account that the GNRs already start to grow when the furnace temperature is increased to the optimal temperature.

Optimizing the growth yield of AGNRs

In general, two factors can be expected to influence the growth yield of AGNRs in different SWCNTs. First, the degree of filling of the precursor molecules. It is influenced by the diameter of the SWCNTs. The spatial configuration of the molecules inside the SWCNTs mainly depends on the van der Waals interaction and the ideal distance between the inner molecular aggregation and the tube [20, 26]. Secondly, on the diameter of the carbon nanotubes. In the case of ferrocene, the distance between cyclopentadienyl groups is about 0.3 nm and the diagonal length of the molecule is 0.4 nm. Therefore, ferrocene molecules should be able to be encapsulated in SWCNTs with a diameter of 1.0 to 1.7 nm. Moreover, ferrocene filling is saturated in 2 days, since the yield of filled GNRs cannot be improved by increasing the filling time.

Consequently, the first factor can be neglected and the second factor would play a more important role. Comparing the intensity of the Raman modes belonging to the AGNRs, it is clear that the SWCNTs with different diameter distribution play a key role in the growth of both 6-AGNRs and 7-AGNRs. To compare the diameter-dependent synthesis, the relative intensities of the CHipb and G bands of the 6-AGNRs and 7-AGNRs are plotted as insets in Figs. 3(a) and 3(b), respectively. It can be clearly seen that the e1.3 sample has the best suitability as a template for the growth of the two AGNRs, which is understandable since the width difference between these two AGNRs is only about 0.1 nm. The best diameters for the 6-AGNRs and 7-AGNRs would be about 1.3 and 1.4 nm, respectively, considering the van der Waals interaction. This also explains why few 6-AGNRs and 7-AGNRs are synthesized when too narrow (e1.0, H1.1) and too large (e1.56, e1.7) samples are used.

AC-HRTEM to verify the AGNR synthesis

To verify the diameter-dependent synthesis of GNRs, aberration-corrected high-resolution transmission electron microscopy (AC-HRTEM) was used for direct observation in addition to Raman spectroscopy. As shown in Figures 4(a)-4(d), GNRs with different widths are clearly visible inside individual SWCNTs, confirming the formation of GNRs inside SWCNTs. The width of the GNRs and the diameter of the SWCNTs can both be accurately identified by their contrast profiles, as shown in the lower part of each image. Summarizing the obtained widths and diameters in Fig. 4(f), the width of the GNRs actually changes with the diameter of the SWCNTs, which is consistent with the Raman spectroscopy results. The distance between the edge of the GNR and the wall of the host SWCNT is about 0.35 nm, which is slightly higher than the theoretical value and corresponds to the diameter difference between the inner and outer tubes of the double-walled carbon nanotubes. Moreover, the overall yield of GNRs from many AC-HRTEM images is statistically above 90%, indicating high efficiency of the conversion of molecules into GNRs, as shown in Fig. 4(e). It is noticeable that most of the GNRs have a width between 0.6 and 0.8 nm, corresponding to 6-AGNRs and 7-AGNRs, again consistent with our Raman results. Although the GNRs can be quite flat, the twisted structure is also the most commonly observed state of the GNRs, similar to previous results12,15. The twisted structure is an important feature of the GNR in the AC-HRTEM, distinguishing it from the inner wall of double-walled carbon nanotubes. Time series of AC-HRTEM images of a narrow GNR with a width of about 0.5 nm (Figs. 4(g)-4(k)) show the motion of a twisted GNR under irradiation. Part of the GNR is observed in the form of a chain, and sometimes a flat shape is seen as it rotates. Such narrow GNRs remain intact under the beam for minutes during observation, indicating excellent stability that is well protected by the SWCNT.

It is well known that SWCNTs exhibit unique electronic properties that are closely related to their chirality and can be broadly divided into two categories: semiconducting and metallic SWCNTs. Therefore, it is of great interest to study the metallicity-dependent synthesis of GNRs.

Resource: Zhang, Y., Cao, K., Saito, T. et al. Carbon nanotube-dependent synthesis of armchair graphene nanoribbons. Nano Res. 15, 1709–1714 (2022). https://doi.org/10.1007/s12274-021-3819-8

References
  1. Dutta, S., & Pati, S. K. (2010). Novel properties of graphene nanoribbons: a review. Journal of Materials Chemistry, 20(38), 8207-8223.

  2. Siu, X., Li, G., Lipatov, A., Sun, T., Mehdi Pour, M., Aluru, N. R., ... & Sinitskii, A. (2020). Chevron-type graphene nanoribbons with a reduced energy band gap: Solution synthesis, scanning tunneling microscopy and electrical characterization. Nano Research, 13, 1713-1722.

  3. Liu, X. M.; et al. Nano Res. 2020, 13, 1713.

  4. Xu, X., Di Giovannantonio, M., Urgel, J. I., Pignedoli, C. A., Ruffieux, P., Müllen, K., ... & Narita, A. (2021). On-surface activation of benzylic CH bonds for the synthesis of pentagon-fused graphene nanoribbons. Nano research, 14(12), 4754-4759.

  5. Khlobystov, A. N. (2011). Carbon nanotubes: from nano test tube to nano-reactor. ACS nano, 5(12), 9306-9312.

  6. Botos, A., Biskupek, J., Chamberlain, T. W., Rance, G. A., Stoppiello, C. T., Sloan, J., ... & Khlobystov, A. N. (2016). Carbon nanotubes as electrically active nanoreactors for multi-step inorganic synthesis: sequential transformations of molecules to nanoclusters and nanoclusters to nanoribbons. Journal of the American Chemical Society, 138(26), 8175-8183.

  7. Shi, L., Sheng, L., Yu, L., An, K., Ando, Y., & Zhao, X. (2011). Ultra-thin double-walled carbon nanotubes: A novel nanocontainer for preparing atomic wires. Nano Research, 4, 759-766.

  8. Li, Y., Bai, H., Li, L., & Huang, Y. (2018). Stabilities and electronic properties of nanowires made of single atomic sulfur chains encapsulated in zigzag carbon nanotubes. Nanotechnology, 29(41), 415703.

  9. Guan, L., Suenaga, K., Okubo, S., Okazaki, T., & Iijima, S. (2008). Metallic wires of lanthanum atoms inside carbon nanotubes. Journal of the American Chemical Society, 130(7), 2162-2163.

  10. Meyer, R. R., Sloan, J., Dunin-Borkowski, R. E., Kirkland, A. I., Novotny, M. C., Bailey, S. R., ... & Green, M. L. (2000). Discrete atom imaging of one-dimensional crystals formed within single-walled carbon nanotubes. Science, 289(5483), 1324-1326.

  11. Anoshkin, I. V., Talyzin, A. V., Nasibulin, A. G., Krasheninnikov, A. V., Jiang, H., Nieminen, R. M., & Kauppinen, E. I. (2014). Coronene Encapsulation in Single‐Walled Carbon Nanotubes: Stacked Columns, Peapods, and Nanoribbons. ChemPhysChem, 15(8), 1660-1665.

  12. Chamberlain, T. W., Biskupek, J., Rance, G. A., Chuvilin, A., Alexander, T. J., Bichoutskaia, E., ... & Khlobystov, A. N. (2012). Size, structure, and helical twist of graphene nanoribbons controlled by confinement in carbon nanotubes. Acs Nano, 6(5), 3943-3953.

  13. Talyzin, A. V., Anoshkin, I. V., Krasheninnikov, A. V., Nieminen, R. M., Nasibulin, A. G., Jiang, H., & Kauppinen, E. I. (2011). Synthesis of graphene nanoribbons encapsulated in single-walled carbon nanotubes. Nano letters, 11(10), 4352-4356.

  14. Chernov, A. I., Fedotov, P. V., Talyzin, A. V., Suarez Lopez, I., Anoshkin, I. V., Nasibulin, A. G., ... & Obraztsova, E. D. (2013). Optical properties of graphene nanoribbons encapsulated in single-walled carbon nanotubes. ACS nano, 7(7), 6346-6353.

  15. Kuzmany, H., Shi, L., Martinati, M., Cambré, S., Wenseleers, W., Kürti, J., ... & Pichler, T. (2021). Well-defined sub-nanometer graphene ribbons synthesized inside carbon nanotubes. Carbon, 171, 221-229..