Fig. 1: Illustration and structural characterizations of 1D semimetal contact.
From: One-dimensional semimetal contacts to two-dimensional semiconductors
a Chemical vapor deposition of the ultralong CNTs. b Fabrication processes of the 2D FET with 1D semimetal contact: (i) sulfur treatment of CNTs; (ii) optically identification of the CNT position on the entire Si substrate; (iii) transfer two segments of a same metallic SWCNT (red circles in panel ii) onto the prepared rectangular TMD channel; (iv) vacuum annealing and metal deposition. c Cross-sectional STEM image and EELS map of a representative sample showing the MoS2 channel and CNT electrode. Scale bar: 2 nm. d False-colored SEM image of the back-gate MoS2 FET showing the MoS2 channel (blue) and SWCNT electrodes (white). The dotted boxes present MoS2 FET with 1D semimetal contacts (orange), CNT FETs with different channel lengths for transfer length method analysis (blue), and MoS2 FETs for four-probe method measurements (green). Electrical properties of the devices defined by electrodes A, B, C, D, E, 1 and 2 are shown in Fig. 2. Scale bar: 8 µm. e Raman spectra of individual CNT electrodes and the MoS2 channel on Si/SiO2 substrate using a 532 nm laser. The vertical dashed lines are used to mark the Raman peaks. The radial breathing mode (RBM) peaks of the SWCNTs verify their single-walled structure and same chirality. f Plane-averaged charge density difference and electrostatic potential of the (5,5) CNT/MoS2 heterostructure along the Z direction (pointing from MoS2 to CNT). The inset is the charge density difference of heterostructure with an isosurface is 7 × 10−4 e/Å3, where yellow and cyan correspond to charge accumulation and depletion, respectively. EF is the Fermi level. g Projected density of states (PDOS) of MoS2, (5,5) CNT before contact formation, and the 1D-2D heterojunction. The Φn and Φp represent the Schottky barrier for electrons and holes respectively.
