We investigate the coupling between individual tubes in a rope of single-wall carbon nanotubes using four probe resistance measurements. By introducing defects through the controlled sputtering of the rope we generate a strong non-monotonic temperature dependence of the four terminal resistance. This behavior reflects the interplay between localization in the intentionally damaged tubes and coupling to undamaged tubes in the same rope. Using a simple model we obtain the coherence length and the coupling resistance. The coupling mechanism is argued to involve direct tunneling between tubes.The unique structural and electronic properties of carbon nanotubes make them interesting objects for basic science study as well as applications. The relation between their geometry and and electronic structure is of particular interest. Semiconducting or metallic behavior is possible depending on tube diameter and chirality [1]. Based on their unique properties, several applications in electronics have been proposed and some, such as field effect transistors [2,3] and diodes [4] have already been demonstrated. While the electronic structure of individual tubes has been characterized using scanning tunneling spectroscopy and found to be in agreement with the theoretical predictions [5], the interaction between tubes in ropes has received much less attention. Some studies have concluded that the coupling between tubes must be weak [6], but few attempted to directly measure this interaction [4,7]. Thus, most of the applications rely on single tubes bridging metal contacts [2,8]. However, the extensive use of nanotubes in future nano-electronics would also require a knowledge of the tube-tube electronic coupling.Here, we present a novel approach that allows us to determine the electrical coupling between tubes in a rope using four terminal transport measurements. The ropes are self-assembled bundles of carbon nanotubes, in which the tubes line up parallel to each other. The tubes in our ropes have diameters close to 1.4 nm and form a regular triangular lattice with a lattice constanct of d 0 = 1.7 nm [9]. Both, semiconducting and metallic tubes are present in a rope in a random distribution. In our experiment, the ropes are dispersed on an oxidized Si substrate and gold electrodes were subsequently fabricated on top of the ropes (inset Fig. 1). The key feature in our investigation involves a sputtering of the rope before deposition of the electrodes by an Ar + ion beam at an energy of 500 eV. The purpose of the sputtering is to introduce defects into the top nanometers of the rope. As will be shown later, this will enable us to vary the path taken by the electric current in a well defined manner. In order to estimate the extent of the sputter damage, a Monte Carlo simulation was performed [10]. From our sputtering conditions, we estimate that the damage reaches about 6 (±2) nm deep into the rope and the damage density is about one defect per 1000 atoms, which gives a distance of 5-10 nm between defects along the tubes. This defec...

The study of the intrinsic transport properties of carbon nanotubes suffers from the difficulties in fabricating noninvasive contacts. Here, we present a scheme for the investigation of transport phenomena in metallic single-wall carbon nanotubes by means of a special four-terminal measurement configuration. To suppress the impact of the contacts on the measured conductance in a tube, we found a combination of top and bottom contacts to the rope of single-wall nanotubes to be most appropriate. Our experimental findings demonstrate that a linear decrease of the sample resistance can be observed under these circumstances without the common increase of resistance for decreasing temperatures.

No abstract

In this letter, we describe a technique for determining the Andreev reflection probability of electrons impinging on a semiconductor–superconductor interface. A two-dimensional electron gas (2DEG) in an InGaAs/InP heterostructure is linked to a niobium superconductor. A point contact in the 2DEG emits ballistic electrons and detects the reflected carriers. The vast majority of the detected carriers are retroreflected holes because of our specific sample setup. We have found an Andreev reflection probability of up to 20%. The experimental results are compared with the predictions of two theoretical models.