Bifunctional anchors connecting carbon nanotubes to metal electrodes for improved nanoelectronics.

Since their discovery in 1991, carbon nanotubes have attracted much attention due to their unique electric, mechanical, and chemical properties.1-5 Numerous breakthroughs have led to practical fabrication of carbon nanotube electronics devices, such as transistors,6 interconnects,7 spintronics,8 and sensors. In addition, exfoliated or single graphene sheets show promise as an alternative material for novel electronics devices.9,10 Furthermore, a strategy to fabricate a large-scaled carbon nanotube circuit device from single-walled CNTs has been proposed6,11,12 in which carbon nanotubes are controllably assembled in a specific pattern on the surface and deposited on the metal contacts. Similarly, single graphene sheets are promising for such applications. Two outstanding obstacles impede this strategy: (1) The carbon nanotubes or graphene sheet may move13,14 on the contact surface, leading to device unreliability. (2) The contact resistance between the metal contact and a carbon nanotube or graphene sheet can be too high for optimum performance (e.g., 10 MΩ without posttreatment).15 To alleviate such problems, we propose connecting the metal contact to the carbon nanotubes via bifunctional molecular anchors. We determine here molecular anchors that would solve these two major problems in CNT and graphene sheet architectures: providing greatly reduced contact resistance (60-fold decrease) and greatly increased mechanical stability. In order to determine acceptable anchors, we considered the following functional groups, -SH, -OH, -NH2, -COOH, -CONH2, and -SO3H. We assume that the CNT or graphene sheet would be functionalized with a modest coverage (∼1%) of one of these groups. For computational convenience, our model system (Figure 1) has a coverage of ∼8 atom % anchors per surface carbon atoms, a range accessible to experiment.17 We expect these functional groups to lose hydrogen atoms as they attach to the metal surface, making a strong covalent bond that provides good electrical contact (small contact resistance) between the CNT/graphene and metal while also providing a good mechanical connection (preventing thermal movement of the carbon nanotubes or graphene). To test this concept, we used quantum mechanics (QM using the PBE flavor of density functional theory) to test our designs for molecular anchors by predicting: (1) the structure of the CNT/grapheneanchor-metal interface, (2) interaction energies between anchor and CNT/graphene and between anchor and metal, and (3) the contact resistance between the CNT/graphene and metal for various anchors. To determine the interface structures and anchor energies, we used a 2 × 2 cell of a three-layer Pt(111) slab to describe the Pt surface plus a single graphene sheet to represent the carbon nanotube surface. We then optimized the structures for the Pt slab-anchorgraphene system with one anchor per cell. (The two bottom layers of Pt atoms were fixed; see Supporting Information.) We find (Table 1 and Figure 1) the (-N-) anchor and the conjugated anchors [(-COO-) and (-CON-)] have the best linkage strengths. To determine the contact resistance for anchored assemblies, we optimized the sandwich slab structure in Figure 2a. Then we calculated the current/voltage performance (electrical resistance) by combining Green’s function theory17-22 with the DFT Hamiltonian. This leads to