Ultrananocrystalline diamond (UNCD) films consist of randomly oriented 3-5nm crystallites surrounded by atomically abrupt 0.2nm wide, largely sp2 bonded grain boundaries. Such films, typically synthesized from hydrogen poor, 1% CH₄/99% Ar microwave plasmas, are highly electrically insulating but are rendered highly conducting (up to several hundred S/cm) by substituting up to 20% of N2 for Ar in the synthesis gas. Nitrogen incorporated into the film grain boundaries, although reaching total concentrations only up to about 0.2%, appears dramatically to increase conduction presumably by modifying the electronic states introduced into the band-gap of diamond. The small amount of lattice substitutional nitrogen is electrically inactive because it is situated 1.7 ev below the conduction band. The nature of the sudden transition from insulating to conducting films occurring over the narrow range of 5-8% added N2 has been the subject of considerable puzzlement for some time. Hall effect measurements have shown the films to be n-type with small activation energies and carrier concentrations up to 10+21/cm³. For a time it was thought that the carriers were associated with free spins but this idea was abandoned when it was found recently that the spin density is invariant with film nitrogen content.

The potential utility of these films which provide the only currently available source of n-type diamond material that is electrically conducting at ambient temperatures makes it important to gain a better understanding of the mechanism underlying the insulator-metal transition. This paper deals with a detailed study of the microstructure of the films as a function of plasma nitrogen content using high resolution TEM, small angle neutron scattering (SANS) as well as SEM and Raman techniques.

The salient result of the study is the finding that the metal-insulator transition is strongly correlated with the unexpected formation of diamond nanowires. The nanowires which appear suddenly and reproducibly when the N2 content reaches about 8% by volume, are 5nm in width and 100-150nm in length. They are encased in largely sp² bonded shells 1-2nm in thickness as revealed by EELS measurements.

The SANS results on films with lower nitrogen content show rod-like structures displaying a high neutron scattering intensity consistent with 3-5nm crystallites. These develop into sheet-like structures at the highest nitrogen concentrations. Thus the SANS data suggest that the insulator-metal transition is associated with a crossover from a randomly oriented microstructure to a two-dimensional structure that is partly ordered on the nanometer scale and grows parallel to the substrate surface.

The mechanism causing the transition from a randomly oriented to a more ordered microstructure resulting from nitrogen additions cannot be understood in detail as yet. However, it is likely that electron transport is very strongly facilitated by a larger fraction of highly "connected" sp² bonded carbon that accompanies the structural transformation. The results presented here are consistent with tight- binding density functional calculations that predict an increase in sp² bonded carbon due to nitrogen incorporation into UNCD grain boundaries (1). The transition to metallic conductivity can best be rationalized by theoretical developments due to Godet which postulate an exponential behavior of the density of states near the Fermi level ⁽²⁾. It is of interest to point out that classical Mott theory postulates that at the crossover region from an insulator to a metallic regime the screening length becomes comparable to the localization length. Following this approach, the critical concentration of carriers (10+19-10+20/cm3) to reach the metallic regime requires a 1-2nm wide sp² "bonded" grain boundary, in accord with experimental observations ⁽³⁾.

The implications of the electron transport results for the potential use of UNCD/Nanocarbon composites as high efficiency thermoelectric materials will be discussed.

⁽¹⁾P. Zapol, M. Sternberg, L. A. Curtiss, T. Frauenheim, and D. M. Gruen, Phys. Rev. B 65, 045403, (2002).

⁽²⁾P. Achatz, O. A. Williams, P. Bruno, D. M. Gruen, A. Bergmaier, J. A. Garrido, and M. Stutzmann, Phys. Rev. B, in press.

⁽³⁾I. S. Beloborodov, P. Zapol, D. M. Gruen, and L. A. Curtiss, Phys. Rev. B, Submitted

* Work performed under the auspices of the Division of Materials Science, Office of Basic Energy Sciences, U. S. Department of Energy, under Contract No. W-31-109-ENG-38.

Rigid conducting biocomposites are interesting transducing materials for the construction of electrochemical immunosensors, genosensors and enzymosensors, particularly if the transducer is bulk-modified with universal affinity biomolecules. Two "ouniversal" affinity biocomposites for electrochemical biosensing are presented, in which the common base material is a graphite-epoxy composite. The first approach relies on strept(avidin)-graphite-epoxy biocomposite transducer, as an universal immobilisation platform whereon biotinylated DNAs, enzymes or antibodies can be captured by means of strept(avidin)-biotin reaction. The second approach is based on Protein A-graphite-epoxy biocomposite. Protein A is able to bind the Fc region of antibodies serving as generic affinity matrix for immuno-immobilization onto the transducer.

Biocomposite electrodes offer many potential advantages compared to more traditional electrodes based on a surface-modified conducting phase. The capability of integrating various materials into a single one is their main advantages. From an electrochemical point of view, rigid conducting biocomposites show higher signal-to-noise ratio compared to the corresponding pure conductors, thus improving sensitivity. These materials can be easily prepared through "dry chemistry" using procedures that can be transferred to mass fabrication. As biological bulk-modified materials, the conducting biocomposites act not only as transducers, but also as reservoir for the biomaterial. After its use, the electrode surface can be renewed by a simple polishing procedure, stating a clear advantage of these approaches respect to classical biosensors. The different strategies for electrochemical genosensing, immunosensing and enzymosensing are discussed. Response parameters as well as ease of preparation, robustness, sensitivity, surface regeneration, costs, and transfer to mass production of these different approaches are also considered.