Absorption losses at a nanorough silver back reflector of a solar cell were measured with high accuracy by photothermal deflection spectroscopy. Roughness was characterized by atomic force microscopy. The observed increase of absorption, compared to the smooth silver, was explained by the surface plasmon absorption. Two series of silver back reflectors ͑one covered with thin ZnO layer͒ were investigated and their absorption related to surface morphology.

It has been 20 years since Lurie et al. first published their model of electromigration of an analyte under simultaneous interaction with two cyclodextrins as chiral selectors. Since then, the theory of (enantio)separation in dual and complex mixtures of (chiral) selectors is well understood. In spite of this, a trial-and-error approach still prevails in analytical practice. Such a situation is likely caused by the fact that the entire theory is spread over numerous papers and the relations between various models are not always clear. The present review condenses the theory for the first time. Available mathematical models and feasible practical approaches are summarized and their advantages and limitations discussed.

No abstract

The optical absorption coefficient of amorphous and microcrystalline silicon was determined in a spectral range 400-3100 nm and a temperature range 77-350 K. Transmittance measurement and Fourier transform photocurrent spectroscopy were used. The measured data served as an input for our optical model of amorphous/microcrystalline tandem solar cell. Differences in the current generated in the amorphous and microcrystalline parts were computed, for an operating temperature between )20°C and +80°C. Optical spectra of microcrystalline silicon were compared to the spectrum of silicon on sapphire (without hydrogen and hydrogenated) and the observed difference was interpreted in terms of a different density of defects and higher disorder of microcrystalline Si.

A neutral marker of the EOF can gain a nonzero effective mobility because of its possible interaction with a charged complexing agent, such as a chiral selector in CE. We determined effective mobilities of four compounds often used as EOF markers (dimethyl sulfoxide, mesityl oxide, nitromethane, and thiourea) in the BGE-containing sulfated β-CD (60 g/L). All the compounds studied were measurably mobilized by their interaction with the selector. The highest effective mobility (-3.0·10(-9) m(2) s(-1) V(-1)) was observed for thiourea and the lowest (-1.5·10(-9) m(2) s(-1) V(-1)) for dimethyl sulfoxide and nitromethane. The mobilities were determined by a new two-detector pressure mobilization method (2d method), which we propose, and the results were confirmed by standard CE measurements. In the 2d method, one marker zone is situated in the BGE containing the charged selector, while the second marker zone is surrounded with a selector-free BGE, which prevents its complexation. The initial distance between the two marker zones equals the capillary length from the inlet to the first detector. After a brief voltage application, the final distance between the marker zones is determined based on known capillary length from the first to the second detector. The difference between these two distances determines the effective mobility.

No abstract

For Gaussian peaks, the migration time of the analyte results as the position of the top of the peak and the zone variance is proportional to the peak width. Similar relations have not yet been derived for the Haarhoff-van der Linde (HVL) function, which appears as a fundamental peak-shape function in electrophoresis. We derive the relations between the geometrical measures of the HVL-shaped peak, that is the position of its maximum, its width and a measure of its asymmetry, and the respective parameters a1, a2, and a3, of the corresponding HVL function. Under the condition of the HVL-shaped peak, the a1 parameter reflects the true migration time of the analyte, which may differ from the peak top position significantly. Our procedure allows us to express the parameters without the need of any external data processing (nonlinear regression). We demonstrate our approach on simulated peaks and on experimental data integrated by the ChemStation software (delivered with the CE instrumentation by Agilent Technologies). A significant improvement is achieved reading the migration time of the experimental and simulated peaks, which draws the error of the HVL-shaped peak migration time evaluation down to the resolution of the data sampling rate.

No abstract