No description available.

The behaviour of thin film semiconducting and magnetic devices depends upon the chemical and physical status of the as-grown structure. Since the dimensions of many devices can be in the Angstrom and nanometre region, characterisation techniques capable of measuring chemical and physical parameters in this regime are necessary if an understanding of the effect of specimen structure on observed properties is to be achieved. This thesis uses high resolution x-ray scattering techniques to characterise sub-micron layered structures of semiconducting and magnetic materials. Double crystal diffraction is routinely employed in the semiconductor industry for the on line inspection of sample quality. While material parameters such as sample perfection and layer composition may be rapidly deduced, the non-destructive measurement of layer thickness is more difficult (particularly for multilayered samples) and lengthy simulation procedures are often necessary to extract the thickness information from a double crystal diffraction profile. However, for semiconductor structures which act as Bragg case interferometers, oscillations (known as thickness fringes) appear in the diffracted profile. The period of these fringes can be directly related to layer thickness. Attempts to Fourier transform diffraction data, in order to automatically extract the frequency" of thickness fringes, have previously been only partially successful. It is shown that the relatively weak intensity of the thickness fringes and the presence of the substrate peak in the analysed diffraction data, drastically reduce the quality of the subsequent Fourier transform. A procedure for the manipulation of diffraction data is suggested, where an "average” envelope is fitted to the thickness fringes and used to normalise the data. The application of an auto-correlation is shown to further increase the quality of the Fourier transform of the normalised data. The application of Fourier transform techniques to the routine analysis of double crystal diffraction data is discussedA novel technique for the measurement of absolute lattice parameters of single crystals is presented, which is capable of determining lattice constants with an absolute accuracy of around 2 parts in 10(^5). The technique requires only the use of a conventional triple crystal diffractometer with motorised 20 circle movement and the provision for a fine, precise rocking motion of the analyser. To demonstrate the technique, exemplary measurements on GaAs and InAs crystals are presented. Triple crystal diffi-action analysis has been performed on three material systems of current technological interest; the Hg(_1-x)Mn(_x)Te on GaAs, the Cd(_1-x)Hg(_x)Te on CdTe/Cd(_1-x)Zn(_x)Te and the low temperature grown GaAs systems. Studies on the Hg(_1-x)Mn(_x)Te on GaAs system reveal that the principal contribution to the rocking curve widths of layers grown using the direct alloy growth (DAG) method, arise from the tilt (i.e., mosaicity) of layer sub-grains. This finding is confirmed by double crystal topography which shows that the layers are highly mosaic with a typical grain size of (130±5)µm. Topographic studies of Hg(_1-x)Mn(_x)Te on GaAs, grown using the interdiffused multilayer process (IMP), show that sample quality is significantly improved with single crystal material being produced using this growth method. Triple crystal diffraction studies of the Cd(_1-x)Hg(_x)Te on CdTe/Cd(_0.96)Zn(_0.04)Te systems reveal several findings. These are that the main contribution to rocking curve widths is from lattice tilts and that the tilt distribution increases as the layer thickness decreases. Further, the quality of the Cd(_0.96)Zn(_0.04)Te substrate analysed is superior to that of the CdTe and that Cd(_1-x)Hg(_x)Te layers grown on Cd(_0.96)Zn(_0.04)Te substrates are generally of a higher quality than those grown on CdTe. Triple crystal analysis of MBE and ALE grown GaAs films, deposited at low growth temperatures, show that, at equivalent temperatures, superior quality films are grown by the ALE technique. Narrow lattice dilation and tilt distributions are reported for GaAs films grown at temperatures as low as 300ºC by the ALE method. While diffraction techniques are highly suitable for the study of relatively perfect crystalline material, they are not appropriate to the analysis of heavily dislocated or even amorphous specimens. This is not the case for the Grazing Incidence X-Ray Reflectivity (GIXR) technique, whose sensitivity is not dependent upon sample structure. The GIXR technique is currently attracting increasing interest following the development of commercial instruments. In this thesis, GIXR has been used to probe the layer thickness and interfacial roughness of a series of magnetic multilayer samples and Si/Si(_x)Ge(_1-x) superlattices. The technique is shown to be capable of measuring layer thickness to an accuracy of one monolayer. Modelling of specular GIXR data for the Si/Si(_x)Ge(_1-x) superlattices has shown that the magnitude of interfacial roughness is different for the two types of interface within the high Ge content superlattice samples, the Si(_x)Ge(_1-x)→Si interface possessing a long range sinusoidal roughness of (0.9±0.3)nm, in addition to die short range roughness of (0.5±0.2)nm present at all interfaces. By collecting the diffuse scatter from a GIXR experiment, conformal, or correlated, roughness has been observed in both the multilayer and superlattice samples.

Les recherches sur les matériaux en couches minces dédiées à l'industrie solaire restent un sujet d'intérêt avec le nombre croissant de types de matériaux incorporés en tant qu'absorbeur dans un dispositif solaire. Le besoin de techniques de caractérisation est donc aigu pour l'optimisation des matériaux et leur incorporation dans des cellules photovoltaïques. Dans cette thèse, une méthode de photo-courant basée sur la spectroscopie de photo-courant à transformée de Fourier (FTPS) est utilisée pour effectuer des mesures sur des matériaux en couches minces et des cellules solaires. Notre FTPS a été développée pour réaliser 3 types de mesures: 1.) mesure de réflexion et de transmission (R/T), 2.) spectroscopie du coefficient d'absorption, et 3.) mesure de réponse spectrale, efficacité quantique externe et densité de photo-courant court-circuit. Cette dernière est spécifiquement utilisée pour les cellules solaires. Nous avons utilisé les résultats de R/T pour effectuer une simulation numérique donnant l'épaisseur, l'indice de réfraction, la rugosité du film et le coefficient d'absorption optique. Une modélisation de la densité d'états (DOS) en utilisant le logiciel DeOSt automatisé avec l'algorithme TLBO (Teacher Learner Based Optimization) a été développée pour trouver les valeurs des paramètres de DOS les mieux adaptées afin de reproduire le ∝ expérimental. Une analyse de sensibilité a été faite pour trouver les paramètres DOS les plus importants parmi 15-17 paramètres. Nous avons mesuré plusieurs échantillons de a-Si: H déposés sous différentes conditions de dépôt, et utilisé nos résultats pour étudier leur DOS. Une comparaison des mesures de α sur a-Si: H déposé sur un substrat de verre et incorporé dans une cellule solaire a également été réalisée. Cette étude a conclu qu'une correction du spectre de coefficient d'absorption doit être effectuée pour les mesures sur les cellules solaires. / Investigations on thin film materials dedicated to the solar industry are still a matter of interest with the growing numbers of material types incorporated as absorbers in a solar cell device. The need of characterization techniques is therefore acute for the optimization of materials and their incorporation in solar devices. In this thesis, a photocurrent method based on Fourier Transform Photocurrent Spectroscopy (FTPS) is used to perform the measurements of thin film materials and solar cells. Our FTPS was further developed to perform 3 types of measurements: 1.) reflection and transmission (R/T) measurement, 2.) absorption coefficient spectroscopy and 3.) spectral response, external quantum efficiency, and short circuit photocurrent density measurements. This latter is specifically used for solar cells. We used the R/T results to perform numerical simulations giving the thickness, refractive index, film roughness, and optical absorption coefficient. A modeling of the density of states (DOS) using the software DeOSt automated with the Teacher Learner Based Optimization (TLBO) algorithm was achieved to find the best suited DOS parameter values to reproduce the experimental spectrum of alpha. A sensitivity analysis was performed to find the most important DOS parameters among 15-17 parameters. For the experimental studies, we have measured several a-Si:H thin film samples prepared under different deposition conditions, and used their absorption coefficient; spectra to study their DOS. A comparison of absorption coefficient; measurements on a-Si:H thin films deposited on a glass substrate and incorporated in a solar cell device stack was also conducted. This study concluded that a correction of the absorption coefficient spectrum measured on solar cells had to be done.