No abstract

No abstract

We introduce a phase diagram for boron nitride film growth. It is based on studies of the influence of the ion energy and substrate temperature on the phase formation using mass-selected ion-beam deposition of B ϩ and N ϩ ions. For the formation of the cubic phase we find threshold values of 125 eV for the ion energy and 150°C for the substrate temperature. Furthermore, we find a characteristic ion energy and substrate temperature dependence of the compressive stress, yielding low stress values for high energies and/or temperatures. c-BN nucleation and growth is attributed to a subsurface process qualitatively described by the subplantation model. ͓S0163-1829͑97͒05519-7͔Regarding mechanical and electronic applications, cubic boron nitride ͑c-BN͒ may leave diamond far behind, because it is chemically inert against iron and oxygen at high temperatures, 1 and because it can be p and n type doped. 2 So far, low pressure synthesis of c-BN thin films is only possible with a variety of ion-assisted physical and chemical vapor deposition techniques. 3 These films are usually nanocrystalline, with a textured h-BN interface layer. Attempts to grow c-BN by chemical processes alone have failed so far. 4 It is therefore generally accepted that ion bombardment is necessary for c-BN nucleation and growth; however, a satisfactory understanding of the underlying mechanisms is still lacking. An unwanted side effect is the ion-induced high compressive film stress, which limits the achievable film thickness to only a few hundred nm, insufficient for most tribological applications. The growth of low-stress c-BN films is therefore highly desirable.The c-BN growth conditions for ion-beam-assisted deposition ͑IBAD͒ were investigated in systematic studies. 3,5 With IBAD, films are usually grown using evaporated boron atoms and additional ion bombardment, typically with a mixture of N 2 ϩ , N ϩ , and Ar ϩ ions. In addition, the ion angle of incidence may vary for different deposition systems. IBAD growth is therefore rather complex, and a variety of different processes such as condensation and thermal desorption, implantation of ions, recoil implantation of atoms deposited on the surface, and sputtering have to be considered. The c-BN growth regime is thus a complex function of three parameters: substrate temperature, ion energy, and the flux ratio of ions to neutral ͑boron͒ atoms. 3,5 Furthermore, c-BN growth takes place at conditions close to the resputter limit due to the intense heavy ion bombardment. Kester et al. introduced the average momentum transferred to the boron atoms deposited on the surface as a more universal parameter to characterize c-BN growth. 5 Sputter yield and nuclear stopping are quantities which depend on the square root of the ion energy ͑or the ion momentum͒ to a first approximation. On the other hand, nuclear stopping itself determines the range of low-energy ions in matter, the energy transfer to recoil atoms, and the deposited energy density. As a consequence, several models for c-BN nucleation and growth wer...

We have studied the growth and the properties of (t)a-C:F films prepared by the deposition of mass separated 12C+ and 19F+ ions as a function of the F concentration. The films are always strongly F deficient due to the formation of volatile F2 and CFx molecules during the deposition process. A maximum F content of about 25 at. % is obtained for an ion charge ratio of C+:F+=1:1. The observed mechanical, optical, electrical, and structural properties as well as the thermal stability of the films are strongly influenced by the F content. A three step progression of the film structure is evident for increasing F concentration: the amorphous three-dimensional network of tetrahedrally bonded carbon atoms of pure carbon films (ta-C) with diamondlike properties is doped for very low F concentrations (ta-C:F). A further increase of the F content results first in transformation to a graphitelike amorphous structure (a-C:F) before the deposited films become porous and to a polymerlike one for the highest F content.

No abstract

In this study we investigate the possibility of nucleating nanocrystalline cubic boron nitride (c-BN) thin films directly onto suitable substrates without the soft turbostratic BN (t-BN) interlayer that is usually observed. This would open a path to the epitaxial growth of c-BN films which is essential particularly for practicable applications in electronic devices. Appropriate substrates are required to exhibit a lattice that matches the c-BN crystallite structure, survives the ion bombardment imperative for c-BN film formation, and is not disturbed by the development of a heterogeneous interface layer. In accordance with these criteria, monocrystalline AlN is selected and employed as a potential substrate for direct c-BN film growth using mass selected ion beam deposition. A detailed examination of the BN/AlN interface microstructure by cross-sectional high-resolution transmission electron microscopy reveals that the AlN crystallinity is indeed retained, with no amorphous layer next to the BN film as commonly observed on Si substrates. Nanocrystalline BN grains with the cubic, and, more frequently, with the wurtzitic structure are found in direct contact with certain regions of the rugged AlN substrate, covering about one-third of its entire surface with no mediating t-BN or other interface layer. The c-BN and w-BN growth areas are textured and exhibit definite preferential orientation relationships with the faceted AlN substrate surface. The consequences of these findings for the understanding of the role of the t-BN interlayer in c-BN film nucleation are discussed.

The present study focuses on the interaction of C60 with the surfaces of highly oriented pyrolitic graphite (HOPG) and sp2-bonded boron nitride (BN). The nanocrystalline BN film was deposited by mass selected ion beams and features an sp2-bonded surface layer, which covers a cubic phase BN film. The first part of the experiment is the sequential deposition of C60, which is monitored by photoelectron spectroscopy in the x-ray (XPS) and ultraviolet (UPS) regime. The growth of the C60 layer on HOPG is close to a layer-by-layer growth mode, but on the BN surface island growth is favored. No charge transfer or chemical reaction (e.g., carbide formation) between the fullerene layer, and the underlying substrate is observed in either case. In the second part of the experiment the samples are heated at a rate of 10 K/min while simultaneously recording the UPS VB spectra. The complete desorption of C60 from the HOPG surface occurs in a small temperature interval between 510–530 K. For the sp2 BN surface the majority of C60 desorbs around 493 K, about half a monolayer (ML) remains, and the C60 concentration decreases gradually with increasing temperature; less than a tenth of a ML can be detected even at 1000 K. The first desorption event at 493 K is attributable to the multilayer desorption from islands. The remaining C60 directly in contact with the BN surface is then removed in a large temperature interval between 500 and 1000 K which indicates the presence of a multitude of adsorption sites. The presence of C60 on the BN film surface also induces a band bending and related B 1s and N 1s core level shifts. An upward band bending is present in the C60 overlayer, which indicates that defects are responsible for the pinning of the Fermi level at the interface.

The ripple topography of ion-beam-eroded surfaces offers a novel method to determine the shape of collision cascades and the distribution of deposited energy. From the energy dependence of the ripple spacing of Ar + -and Xe + -irradiated graphite surfaces at ion energies between 2 and 50 keV, the relations between mean depth, longitudinal and lateral straggling of the damage cascade were obtained. Their evolution with the ion energy was found to follow power laws for both ion masses and implies an energy-independent lateral spread of the damage cascade, while depth and longitudinal spread scale with the ion energy. This can be explained by the nuclear stopping power being nearly independent of energy in the observed region. High-resolution micrographs of single-ion impacts support this interpretation, as the hillock-shaped surface defects found in the experiments show a lateral extension being independent of the ion energy.