ROLE OF THE ATMOSPHERIC ADSORBATES IN THE SURFACE CONDUCTIVITY OF HYDROGEN-TERMINATED DIAMOND

Abstract

The progresses achieved in the production of synthetic diamond have allowed the use of this material for high technological applications. Its combination of characteristic properties, such as high thermal conductivity, chemical inert ness, high support of electric eld before breakdown, corrosion resistance and bio-compatibility, make diamond a unique material. Among the various properties, one of the most curious and controversial is the p-type surface conductivity of diamond obtained when it is terminated in hydrogen and successively exposed to air: despite this characteristic has been already used for a wide variety of applications, such as, for example, for p-type semiconducting layers in the metal-semiconductor eld e ect transis tors (MESFETs), the interpretation of this surface conductivity has been only partially explained and based on two steps, that involve the negative electron a nity and the surface transfer doping; it is still unknown, however, which molecules, or group of molecules, are responsible of conduction: in this sense there are evidences that oxygen related species may have a signi cant role in conduction. One critical point in the study of surface conductivity of hydrogenated dia mond can be found in the poor e ort done to correlate the macroscopic e ects, such as the variation of conducibility of diamond upon air exposure or in dif ferent temperature/pressure conditions, and the microscopic e ects, such as the adsorption/desorption e ects or variation in morphology. Respect to previous investigations, our innovative approach is focused to com bine the two aspects, studying at the same time the macroscopic behaviour, investigate in particular in terms of resistance measurements, and the micro scopic behaviour, where electron spectroscopies such as the X-ray Photoemis sion spectroscopy (XPS) and the Electron Energy Loss Spectroscopy (EELS) are particularly suited to the task, due to their well known surface sensitivity. We focused our attention on three main aspect that we consider fundamental in the knowledge of hydrogen-terminated diamond: one aspect concerned the study of sample preparation, how it can a ect the surface morphology and the conductive properties; we compared di erent sample preparation in terms of electron di raction pattern and we found that heating the sample during hydrogenation improved sample order respect to a cold preparation. Despite this we found with XPS measurements that a signi cative portion of the sam ple, that we estimated to be 33±3 % is covered by oxygen (inactive in the conduction mechanism); moreover, with the aid of Montecarlo simulations we have found that hydrogen treatment may involve not only the surface layer but also the subsurface region. A second aspect regarded the study of surface conductivity of samples and its homogeneity: our measurements show that major issues can be found in terms of homogeneity of hydrogenation, investigated by a -innovative and unconvential- combination of electron re ectivity and resistance pattern: we as cribed these inhomogeneity e ects to the relative position caused by the manual placement of the samples in the hydrogenation chamber. Despite these issues in sample preparation, we were able to perform a conductive test, measuring the resistance as a function of temperature and pressure: the measurements showed as the resistance for sample in non-conductive state (i.e sample in which of airborne species adsorbed, responsible of conduction, are removed by an annealing) is eight order of magnitude higher than the value for sample in the conductive state, in agreement with similar works found in literature[1]: The last aspect concerned the explanation of the role of oxygen in the con duction mechanism, done through the elemental analysis of the surface for both the conducive and non-conductive state of hydrogen terminated diamond samples. Regarding this, spectra taken in the conductive case showed the in creasing presence of carbon-oxygen related species (such as C OH, C O, C=O) respect to the non-conductive case. Moreover, by the comparison of the oxy gen amounts in the conductive and in the non-conductive case we were able to estimate that a fraction of 23±3% of the conductive sample was covered by oxygen-related species active in the conduction mechanism. This value can be related to a hole sheet density in agreement with the ones reported in literature[2-4]. We also compared airborne exposure and oxygen exposure of sample: by studying the evolution of oxygen and C-O related species quanti ties upon di erent exposure and annealing treatment, we deduced a di erent absorption mechanism for air exposed and oxygen exposed sample, suggesting that oxygen is a necessary but not su cient condition to activate the conduc tion mechanism. The scenario presented shows how the study of microscopic properties of hydro gen terminated diamond requires a preparation of samples with more stronger constraints, in terms of homogeneity and presence of impurities, compared to the study of macroscopic properties. This is the starting point to get more and more completed information from elemental analysis of the surface: in this sense a study of alternative methods to hydrogenate the diamond surface is needed.