A fully sealed luminescent tube of 40 cm length and 4 cm diameter based on carbon nanotube field emission is demonstrated. The device shows a homogeneous illumination over the whole length and circumference of the tube and reaches the luminance of conventional fluorescent tubes while being mercury-free, continuously dimmable and with a high illuminance capability. The realization has been made possible with the development of a chemical vapor deposition method to grow nanotubes homogeneously on long metallic wires, which provides an additional possibility to control the mean length and density of the emitters. This control has proven to be of utmost importance as it makes possible to adjust the emission voltage and emission site density needed to reach the target intensity and specifications of the device.

The photoabsorption spectra of silicon cluster cations Si-n(+), for n=2-13, are deduced from the analysis of the photodissociation spectra of xenon-silicon cluster cation complexes Sin+Xe. The spectra are measured for photon wavelengths between 290 and 700 mn (1.78 to 4.28 eV) for nless than or equal to6 and between 230 and 700 nm (1.78 to 5.40 eV) for n>6. The electronic structures and photoabsorption spectra of the Si-n(+) clusters are calculated within the framework of the density-functional theory, starting from optimized geometries published in the literature. Several explanations are given for the discrepancies between the experimental results and the simulations. In particular, we address the problem of the reliability of the photodissociation technique with an aim to deduce the photoabsorption spectra in the case of very small nonmetallic clusters.

The growth of multiwall carbon nanotubes is characterized in situ in a chemical vapor deposition reactor by measuring the current extracted by field emission from the growing nanostructures. The lengthening of the nanotubes provokes an increase of the emitted current at constant applied voltage, and the use of a phosphor screen allows to observe the individual emitters during the growth. A simple model permits furthermore to estimate the growth rate. The nanotubes grow with a closed cap under 10(-2)-10(-4) mbar of C2H2 at 700degreesC with a growth rate over 1 mum/s that increases with the C2H2 pressure. The growth of the nanotubes is neither simultaneous nor homogeneous over the cathode and involves a different activation time for every emitter.

It is common practice in the literature to reduce the field emission properties of film field emitters to a few parameters to make possible a comparison between different samples. In particular, the macroscopic field needed to extract a given current density and the field enhancement factor are often used. The purpose of this contribution is to attract attention to the fact that such comparisons have to be done with utmost care, as the values of these parameters depend on the experimental configuration, and especially on the inter-electrode distance, used for the measurements. Current and future solutions are considered and discussed.

Communication: A simple yet powerful method to produce carbon nanotubes is described. The growth is carried out directly on a heated support by the decomposition of gaseous hydrocarbons in a cold atmosphere without the requirement of an additional heating source. The technique allows the production of homogeneous films of well-graphitized nanotubes of shapes predetermined by the shape of the heated support. and also allows their lengths and density to be controlled. This is important for the maximization of the emission site density when films are used as cathodes in field emission devices. A 40 cm long mercury-free field emission luminescent tube could be realized.

Growing carbon nanotubes were observed during the chemical vapor deposition process by field emission microscopy. The experiments were performed in a configuration similar to the one used to realize a luminescent tube, with a wire as nanotube support and cathode placed in the center of symmetry of a glass tube with a phosphor layer that served as anode. To activate the growth, the wire was resistively heated to 700degreesC and a partial pressure of 10(-2...4) mbar of C2H2 was introduced in the chamber. 10 mum-long nanotubes grew after an activation time that was as short as 10 s with a growth rate of typically 1 mum (.) s(-1).

The growth of carbon nanotubes by chemical vapor deposition (CVD) is followed and characterized by performing field-emission microscopy directly in a modified CVD reactor, where the hydrocarbon gas is introduced at a partial pressure below 10−2 mbar and a high voltage is applied between the heated substrate and a phosphor screen. This allows us not only to detect a field-emission current that increases with the length of the nanotubes, but also to observe the growth of the individual emitters by following the evolution of their field-emission patterns on the phosphor screen. Nanotubes grow after an activation time of a few s and reach a length of 5–10 μm in typically 10 s. We deduce a growth rate in the range 1–10 μm/s that increases with the gas pressure.

We report on measurements of the plasmon losses of individual single-wall carbon nanotubes by electron energy-loss spectroscopy in a high-resolution transmission electron microscope. The experimental data are compared to simulated excitation probabilities calculated using the hydrodynamic theory of the interaction between a probe electron and a two-dimensional quasifree electron gas confined on a cylindrical shell. Depending on the nanotube geometry, the first- or the second-order oscillation mode dominates the loss spectrum. The resonance energy of the dominant resonance mode is found to depend on the radius of the nanotube.

Carbon-encapsulated cobalt particles were produced by a modified arc-discharge technique and subsequently purified with an acid treatment. The arc discharge was realized between a graphite cathode and a graphite crucible with a solid Co target, under partial He pressure with an additional quenching jet. The produced particles are spherical and covered by typically 3–5 graphitic carbon shells. The mean diameter could be varied between 5 and 45 nm by changing the deposition parameters (gas pressure and flow rate of the quenching jet). The magnetic properties were analyzed with a variable temperature SQUID magnetometer. The encapsulated particles showed ferromagnetic hysteresis loops with marked size-dependent properties.

We present simulations of transport and field emission properties of multi-wall carbon nanotubes [1], by using a transfer-matrix methodology [2] for taking account of three-dimensional aspects of the potential-energy distribution. Band-structure effects result from the periodic repetition of a basic unit of the nanotubes and by using pseudopotentials [3] for the representation of carbon atoms. The configuration considered for the transport simulations consists of (10,10) or (15,15) single-wall slots, which are used to introduce electrons in a single layer of a (5,5)@(10,10)@(15,15)@(20,20) multiwall nanotube (see Fig. 1 for a representation of the junction between the nanotubes). The study focuses on the distance this electronic flux can cross before spreading significantly to neighboring layers. The results indicate that the layers of multi-wall nanotubes may be used as independent current conduction channels over distances that will be discussed. Simulations of field emission from flator convexterminated (5,5)@(10,10)@(15,15) nanotubes are presented (see Fig. 2 for a representation of the potential-energy distribution around the convexterminated structure). The total-energy distributions exhibits features that are related to band-structure effects, to resonant states at the apex of the nanotubes and to standing waves in the structure. The calculations reveal that the emission properties of multi-wall nanotubes are better than those of their single-wall components, essentially because of a lower global polarizability of multi-wall structures. The consequences of closing the nanotubes, of saturating the dangling bonds by hydrogen and the efficiency of a photo-stimulation process to control the emission will be presented. The authors acknowledge the National Fund for Scientific Research (FNRS) of Belgium and NSF grant number DMI-0078637 administrated by UHV Technologies, Inc., Mt. Laurel, NJ for financial support. References

Electron energy loss spectroscopy is a well adapted tool for the investigation of the valence excitations of individual nanometer-size particles. The interpretation of the loss spectra of such small particles. however, relies in most cases on a quantitative comparison with simulated excitation probabilities. Here we present a formalism developed for the interpretation of the energy loss data of single-wall carbon nanotubes based on the hydrodynamic theory of plasmon excitations by high-energy electrons. The nanotubes are modeled as a two-dimensional electron gas confined on the circumference of a cylinder. The plasmon excitation probabilities. directly comparable to measurements, are discussed for various parameters.

Tuning the electron emission properties of films containing carbon nanotubes is demonstrated. Microcontact printing is used to pattern silicon substrates with catalyst, which facilitates the growth of the carbon nanotubes by the chemical vapor deposition of acetylene. The catalyst is applied in liquid form to an elastomeric stamp prior to transfer to the substrate, and varying the concentration of the catalyst on the solution (“the ink”) directly influences the density of nanotubes on the patterned film.

We show that the field of application of cold electron film emitters can be extended to nonplanar geometries by demonstrating a cylindrical field emission diode. The cathode is a metallic wire on which multiwall carbon nanotubes are grown by the catalytic decomposition of acetylene over a Fe catalyst. The emitter shows excellent performances and can be used to realize a luminescent, mercury-free, tube.

Photodissociation spectra of vanadium cluster ion-xenon atom complexes Vn+Xe (n=5–8) have been measured between 290 and 670 nm. Spectra have been obtained by recording the depletion signal induced on the mass-selected cluster current intensity by the absorption of a photon. Due to the weak interaction between the ionic cluster and the rare-gas atom, photodissociation spectra are regarded as the absorption spectra of the vanadium cluster cations themselves. The absorption bands are broad, but several peaks can be resolved for the smallest sizes. The influence of the rare-gas atom on the electronic structure of the vanadium cluster cation is probed by performing the measurements on krypton instead of on xenon complexes. The features of the spectra do not change, but a blueshift of 0.12 eV is observed from krypton to xenon.

Carbon nanotube films have been grown on cylindrical substrates using catalyst chemical vapor deposition of acetylene. It is shown here that Kanthal rods allow the controlled growth of carbon nanotube films, provided that a liquid catalyst solution containing metallic salts (Fe, Ni or Co nitrate) diluted in ethanol is deposited on the oxidized surface of the substrate. Nickel, Nickel-Chromium and Molybdenum rods were found to be less suitable materials for the controlled deposition of carbon nanotubes.

Specific heat measurements on free iron, cobalt and nickel clusters in different size ranges from 130 to 400 atoms are presented and the experimental method is discussed in detail. These measurements are achieved in a Stem-Gerlach experiment where the magnetization of transition metal clusters reflects their vibrational temperature. Hence, it can be used as a thermometer after a calibration procedure. The specific heat is measured by heating the clusters in flight with a laser. In the temperature range of the experiment (80-600 K), the main feature of the specific heat of Ni200-240 clusters is abroad peak centered at T=340 K which adds to an approximately constant baseline of 6 cal/(mol K). We attribute this peak to a ferro- to paramagnetic transition since its shape and area. are well described by the Weiss mean field model. The specific heat of Co200-240 clusters does not show any prominent feature within the temperature range 80-900 K except a steady increase from 5.5 cal/(mol K) at T=300 K to 15 cal/(mol K) at T=900 K. In iron clusters, the specific heat exhibits a peak which is poorly described by a Weiss mean field theory. Furthermore, the specific heat value:of Fe250-290 clusters at room temperature is up to 50% lower than the Dulong-Petit value. We discuss the possibility that iron clusters undergo a-magnetic transition between a high moment to a low moment state, which have different lattice parameters.

Electron energy loss spectroscopy ina high-resolution transmission electron microscope has recently been used with success to characterize the electronic properties of closed cage nanometer-size graphitic particles, in the plasmon region, the experimental data reveal interesting size-dependent variations, which are not yet fully understood. The difficulties encountered in the interpretation of the spectra are principally due to the lack of a complete theoretical treatment of the anisotropic dielectric response in nanometer-size particles. In order to obtain a better understanding of the experimental data we propose a model based on nonrelativistic local dielectric response theory for electrons penetrating through a nested concentric-shell fullerene or the so-called "carbon onion." The anisotropy of the electronic properties of the sphere is taken into account via the frequency-dependent dielectric tensor of graphite. The model can be applied to simulate electron energy loss spectra as well as line scans through energy filtered images and allows thus a direct comparison to experimental data.

The magnetic moments and heat capacities of small iron, cobalt and nickel clusters in a molecular beam have been measured as a function of their size and temperature using Stern-Gerlach deflections. Single-sided deflections are found indicating that spin relaxation occurs within the isolated clusters. Superparamagnetic behaviour is observed when all the degrees of freedom of the clusters are in equilibrium. Marked decreases in the magnetic moment are observed with increasing temperature and increasing size, slowly converging to the bulk magnetization curve as the cluster size increases. At a given temperature, oscillation of the moment is observed as a function of size, which suggests the filling of layers of atoms in the cluster. Heat capacity measurement of nickel and cobalt clusters in the beam reveals that the magnetic contribution to the heat capacity is well approximated using mean-field concepts, combined with measured magnetic properties. In contrast, iron clusters show anomalies both in the magnetic moments and in the heat capacities.

The magnetic properties of free Gd-N clusters (N = 13, 21, and 22) have been measured between 90 and 600 K in a Stern-Gerlach experiment. Gd-13 and Gd-21 are found to be superparamagnetic (SP) whereas Gd-22 is not. Within the SP model, the magnetic moment per atom at 100 K is 5.4 mu(B) for Gd-13 and 5 mu(B) for Gd-21, lower than the bulk value, 7.55 mu(B). We measure Curie temperatures of 420 +/- 20 K for Gd-21 and over 500 K for Gd-13, significantly higher than 293 K, the bulk value. A canted magnetic moment model is compatible with the experimental data.