Evaluation of the radiometric performance of UV photodetectors.
Project Description
In order to improve the accuracy of detector-based radiometry in the UV several types of photodiode shall be probed for their suitability as radiometric standards Initially the focus will be on a newly developed type of Schottky-barrier silicon photcdiode (PtSi-n-Si), which, according to first investigations by PTB is virtually stable in the entire UV spectral range. Exploiting the different research capabilities at the participating laboratories, the aim of the collaboration is the complete characterization of the electrical and optical properties of the new PtSi-n-Si photodiodes in comparison to other types of photodiode under different operation condition (e g. using different radiation sources - lasers, incancescent lamps, synchrotron radiation). As PTB shall provide the PtSi-n-Si diodes and calibrate them initially, an indirect comparison of UV spectral responsivity scale will be a by-product of the collaboration.
Collaboration includes non-Euromet partners NIST (USA).
Final Report 2007-02-28
Objectives
To probe the suitability of novel UV photodiodes for use as transfer detector standards. To perform an informal comparison of the spectral responsivity scales of the participating laboratories by exchanging calibrated specimen of the photodiode types under investigation.
To organise a series of meetings to monitor the progress of the project.
Activities
A large area version of the PtSi-n-Si Schottky type photodiodes developed at ETH Zurich in cooperation with PTB was produced, and a first fabrication batch of these photodiodes was obtained by PTB for use in the project.
Four selected photodiodes from this lot were circulated among the actively participating laboratories for calibration in the 200 nm to 400 nm wavelength range.
Three international workshops on detector-based UV radiometry were organized in 1999 (Madrid), 2000 Boras, and 2002 (Gaithersburg).
Results and achievements
Based on the calibration results for the spectral responsivity of the circulated photodiodes, an informal comparison of the spectral responsivity scales of the partners was derived.
The outcome of this comparison was that in the wavelength interval between 280 nm and 400 nm, the results of the different participants seem to be in reasonable agreement. In this wavelength range, the relative standard deviation of the values of the different partners amounts to 0.47% on average, and an average χ² of 1.03±0.40 is obtained.
At wavelengths below 280 nm significant discrepancies persist, leading to an average relative standard deviation of 0.95% and a χ² of 1.85±0.58, which indicates that probably some of the uncertainties in this wavelength range were estimated too optimistic.
The established discrepancy between the participants is worst in the wavelength range between 250 nm and 275 nm. For wavelengths short of 250 nm, the results of NIST depart from the other laboratories by up to 3.5% at 200 nm wavelength, whereas the results of the latter show an agreement as good as for wavelengths in excess of 280 nm.