Carbon nanotubes (CNTs) are an innovative material for future electronic and energy devices because of their outstanding properties, such as high carrier mobility, sharp features in the electron density of states, and large specific surface area. Such devices often require both p- and n-type materials with controlled carrier density depending on the intended application. While doping technology has been well established in the current silicon-based-semiconductor industry, chemical doping of nanoscale carbon materials is currently under development. Further, ensuring the stability of doped states in CNTs is a technologically important issue because the electronic and energy devices (e.g., transistors, photovoltaics, and thermoelectric generators) generate condensed heat or should function while placed on heat sources. However, the doped states of CNTs, especially the n-doped states, are usually not stable at elevated temperatures, which limits the applications of n-type CNTs for such devices.
To overcome this limitation, the K. Ishida laboratory of Kobe University (Japan) and the Nanofilm Device group at National Institute of Advanced Industrial Science and Technology (Japan) collaboratively developed reducing agents for developing n-type CNTs with outstanding thermal stability in air. These groups have investigated doping technology for nanoscale carbon materials and conducting polymers, developed characterization technology for quantifying the doping levels, and applied the developed materials to molecular thermoelectric generators1–5.
The introduction of carriers in CNTs relies on charge transfer between the CNT π-electron systems and dopants on the external or inner surfaces of the nanotubes. Several interactions such as π–π, n–π, σ–π, and ion–π can be considered in the toolbox of charge-transfer options. In this study, we focused on 4 molecules (Fig. 1): 1,8-diazabicyclo[5.4.0]-7-undecene (DBU); 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD); 7-methyl-1,5,7-triazabicyclo[4.4.0]dec-5-ene (Me-TBD); and 1,1,3,3-tetramethylguanidine (TMG). These molecules are often termed as organic superbases because of their extremely high pKa. Using these bases for n-doping of CNTs was inspired by the reported charge-transfer interaction between DBU and fullerene, by which a negative charge was introduced in the fullerene through lone-pair electron transfer from DBU (namely, n–π∗ interaction).
Several processes are available for applying the doping regents to the CNT surfaces, including immersing CNTs into a dopant solution, casting dopant solutions on the CNT surfaces, and vapor deposition of the dopants onto CNTs. Among them, the former two wet processes are attractive because of their simplicity, low temperature, and low energy use. In this study, samples were prepared by immersing the CNT films in the dopant solutions (Fig. 2a) or casting dopant solutions on the film.
Further, we used a thermoelectric (TE) charge-carrier determination technique to evaluate the material polarity (p- or n-type). The sign of the Seebeck coefficient reflects the major carrier type (holes or electrons); positive and negative signs correspond to the p- and n-type polarities, respectively. It is well known that as-prepared CNTs without intentional doping show p-type polarity because of autoxidation in air, which is apparent from the positive sign of the slope of the TE data (positive Seebeck coefficient) shown in Fig. 2b. In contrast, the voltage polarity became negative after DBU doping; therefore, it was be demonstrated that the superbase can inject electrons into CNTs to create n-type materials.
While the n-type doping of CNTs could be very easily achieved by immersing the film into the DBU solution, we found that the stability of the DBU-doped CNTs under elevate temperatures is not sufficient; the negative Seebeck coefficient became positive after heating at 100 °C in air, as shown in Fig. 3. To investigate the improvement of the stability, we tested the other 3 bases (TBD, Me-TBD, and TMG) shown in Fig. 1. All these dopants successfully converted the CNTs into n-type materials. It is noteworthy that (Me-)TBD-doped CNT films retained their n-type polarity for more than 6 months (Fig. 3), which highlights the good thermal stability of the resultant materials.
Further, it is very interesting that the slight changes in the primary molecular structures are the source of the apparent difference in stability of the n-type CNTs; for example, TMG shares the guanidine moiety with TBD and Me-TBD, but the stability of TMG-doped CNT film is very poor compared to the other dopants (Fig. 3). DBU has a bicyclic-ring structure but has only 2 nitrogen atoms while TBD and Me-TBD have 3 nitrogen atoms. Our comparisons of the base molecular structures indicate that bicyclic-ring bases with guanidine moiety are useful reagents for creating n-type CNTs that are thermally stable in air. Therefore, molecular structural design is considered as a valuable tool for improving the doped states of CNTs, which is an exciting research target in physical chemistry. Unravelling the molecular origin of the differences in stability is also important to improve our understanding of the underlying mechanisms.
The research project was managed by Dr. S. Horike at Kobe University. These exciting findings have culminated after almost two years of efforts with the co-authors. The success of these bicyclic-ring guanidine bases is expected to pave the way for producing n-type CNTs for future molecular electronics applications.
For more details on this work, please read our paper in Nature Communications at
https://doi.org/10.1038/s41467-022-31179-6
Written By Shohei Horike at Kobe University
References
(1) Horike S. et al., Highly stable n-type thermoelectric materials fabricated via electron doping into inkjet-printed carbon nanotubes using oxygen-abundant simple polymers. Mol. Syst. Des. Eng. 2, 616–623, https://doi.org/10.1039/C7ME00063D (2017).
(2) Koshiba Y. et al., Preparation of poly(3,4-ethylenedioxythiophene) by vapor-phase polymerization at the interface between 3,4-ethylenedioxythiophene vapor and oxidant melt. Mol. Cryst. Liq. Cryst. 688, 53–59, https://doi.org/10.1080/15421406.2019.1651068 (2019).
(3) Mukaida M. et al., Enhanced Power Output in Polymer Thermoelectric Devices through Thermal and Electrical Impedance Matching. ACS Appl. Energy Mater. 2, 6973–6978, https://doi.org/10.1021/acsaem.9b01342 (2019).
(4) Horike S. et al., Large thermoelectric power factor in wafer-scale free-standing single-walled carbon nanotube films. Appl. Phys. Lett. 118, 173902, https://doi.org/10.1063/5.0047089 (2021).
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in