Rapid formation and transfer of carbon nanotube probes: The lowest resistance from nano-bridges

Aggregated carbon nanotubes (CNTs) are difficult to be manipulated. Various approaches have been carried out for making nanotube probes, including selection of longest protruding tube from a nanotube aggregate adhered onto substrate [1], direct synthesis of oriented CNTs on a porous surface via metal catalytic pyrolysis [2] and mechanical attachment of CNTs onto an AFM tip under dark-field optical microscope [3]. In this work, automatic attachment of CNTs onto metal tips at solvent–air interface is demonstrated. Adhered CNTs can be bridged between electrodes. Multi-walled boron-doped carbon nanotubes (BCNTs) produced by arcing BN-containing graphite rods were used as probe formation materials [4]. Distribution of tube length is centered at 20–30 lm and their aspect ratio is greater than 1000, meanwhile, BCNTs possess lower intrinsic resistance (10 –10 6 X m) than pure carbon tubes [4]. Fig. 1a–e depicts CNT-mounting procedure. BCNTs (50 mg) are dispersed in acetone solution (30 ml), followed by an ultrasonic stirring for 10 min (Fig. 1a–b). An electrochemically sharpened tungsten tip (ca. 1 lm) electrically connected to ground level is dipped into solution for 10–20 s (Fig. 1c–d), tip retracts (Fig. 1e). Optical microscope reveals that BCNT bundles always adhere onto tip. Our repeat procedures show successful alignment of BCNTs with tip by ca. 60–70%; a value was obtained based on 100 runs. Organic solvents such as toluene, chloroform, ethanol and benzene produce a similar result. Fig. 1f–g shows oriented nanotubes adhered on metal tips from different runs. The probe-tip configuration relies only on van der Waals binding (Fig. 1f–g), therefore a drastic vibration on tube-tip assembly often leads to tube deviation. Nevertheless, adhered BCNTs do not fall. Adhered bundle is a tapered-structure, i.e. BCNT fringes are clearly seen at tube-tip junction; becomes obscured at its free end (Fig. 1f–g). On the basis of previous data [4], the tiny bundles (Fig. 1f–g) correspond to ca. 10–25 BCNTs at tube-tip junction and ca. 3–8 BCNTs at free end respectively. Oriented tube-tip formation at sharper tips is realized and tubes mostly align with tip at solvent–air interface. Fig. 1h–j illustrates the probe transfer to another tip coated with Ag-paste. Fig. 1k displays a successful transfer. Fig. 2a–e highlights the transfer of BCNT-probe from tip to tip and corresponding sequence (Fig. 2f–j). Fig. 2f shows a tiny bundle attached by two carbonaceous aggregates (arrows A and B, Fig. 2f). Insertion of nanotube bundle into Ag-coating is facile before paste solidification (Fig. 2g). Small vibration from micro-manipulator is capable of pulling-out embedded nanotubes from Ag-paste, accompanied by a blister-like feature (arrow, Fig. 2i). The blister indicates that Ag is still in a paste form and nanotube bundle is truly embedded in Ag-coating. One of the carbonaceous aggregates has vanished after tube submergence (Fig. 2j). The length of free-standing bundle is ca. 20 lm and carbonaceous aggregates A and B locate approximately at distance of 3 and 12 lm from tube-tip junction respectively (Fig. 2f), namely, the depth of bundle insertion into paste exceeds 8 lm and remaining aggregate (arrow, Fig. 2j) is likely to be aggregate B; A has submerged into paste. Fig. 3 displays a tiny bundle bridged between Agelectrodes and its electrical resistance measurements were carried out at zero-bias to exclude the modification of band structure by electric field. The initial bridging resistance is ca. 10–11 kX between t = 0 and 9 min, due to wet paste (Fig. 3a–b). When paste is dried, the resistance rapidly drops to 80 X (Fig. 3c). This value maintains for 15 min (t = 8–23 min, Fig. 3c), then increases