Table of contents

An intercomparison of two ¹²⁷I₂ stabilized He-Ne lasers at λ ≈ 543 nm, recommended as length standards, has been made between the DFM and the PTB, and is the first international comparison at this wavelength reported so far. Both lasers use saturated absorption of iodine in an external cell, one (PTB) using frequency-modulation spectroscopy, the other (DFM) using the first or third harmonic detection techniques. The range of the frequency differences between the two lasers was from 11,8 kHz to 18,4 kHz, with a mean of Δf = 14,2 kHz and a standard uncertainty of 1,9 kHz. A minimum value for the two-probe relative standard uncertainty of 2,5 × 10^-13 (140 Hz) is reached after 5 000 s. The hyperfine line splittings were measured by the beat-frequency technique, and good agreement was found between the two completely different laser systems. Measurements of frequency dependence on modulation amplitude, iodine pressure, and pump power are also reported.

An expression for the attenuation of a plated coaxial line that is applicable over the entire usable frequency range of the line is derived. The result compares well with experiment and shows that accurate line attenuation estimates are possible for plated lines even in the presence of a complicated plating surface such as diffused copper and gold.

Some fifteen years ago, the Institut National de Métrologie (France) opted to produce its secondary mass standards in XSH alacrite and adopted a mechanical cleaning method using ethanol and isopropanol. This paper concerns the influence of this cleaning method and presents some of the results obtained on two alacrite kilograms. The influence of the cleaning process is not the same on the two standards which are nevertheless as similar as possible from the manufacturing point of view. Alcohol cleaning leads to mass variations of several tens of micrograms causing the standard to be unstable for several months. It finally reaches a mass value that is stable, but is not reproducible.

Accurate measurements have been made of the quantized Hall resistance (QHR) corresponding to the quantum number i = 2, R_H (2), of two GaAs-based heterostructures using alternating current in the frequency range from 1 Hz to 1,6 kHz. The QHR is compared with a conventional ac reference resistance standard (Vishay type) of the same nominal value. The measurements are made with an alternating current comparator bridge at 1 Hz and with a coaxial ac bridge from 400 Hz to 1,6 kHz. The quantum Hall effect (QHE) sample is connected to the coaxial bridge using a double series connection. An analysis of the circuit properties of the sample shows that this technique of connection is sufficient to define the QHR as a four terminal-pair impedance. It also avoids perturbing effects from small capacitances between the terminals of the sample and from inductive reactances in series with the terminals. At frequencies up to at least 1,6 kHz a central region on the Hall resistance plateau has been observed for which the Hall resistance R_H (2) is independent of the magnetic flux density to within 1 part in 10⁸. The ratio of R_H (2) to the reference resistance varies by less than 2 parts in 10⁷ from 1 Hz to 1,6 kHz. A significant part of this variation is probably due to the frequency dependence of the reference resistance itself.

In a previous study (Part I) we showed that the material used for national prototype and reference masses, Pt-10% Ir, is susceptible to mercury contamination even at the very low levels of mercury present in air in weighing laboratories. In this study we measure the mass uptake per unit area of Pt-10% Ir exposed to mercury vapour, as a function of time, using the mass response of a quartz crystal microbalance with electrodes of Pt-10% Ir. This system can detect with accuracy mass increases equivalent to less than 0,1 μg on a prototype kilogram. It is shown that there is a rapid adsorption of the first monolayer of mercury, but this mercury then diffuses into the solid allowing further mercury to adsorb. The quantity of mercury absorbed does not tend towards any immediate limit, but instead continues to grow in proportion to the square root of time. A consideration of the morphology of polished surfaces suggests this growth will continue for between 10 and 600 years. We conclude that our earlier recommendation to clean all environments where reference masses are housed, particularly with regard to mercury, is essential for maintaining the stability of Pt-10% Ir masses. Significant mass instability due to mercury sorption will occur for atmospheric mercury levels well below current health and safety limits.

Modulation broadening of the iodine absorption components of a ¹²⁷I₂ stabilized He-Ne laser at the transition 11-5, R(127) is studied in this report. The mathematical model used for computer simulation is described. Simulation results are presented for the frequently used d, e, f and g components and compared with the results of the measurements. The effect of ¹²⁷I¹²⁹I components on ¹²⁷I₂ absorption spectra is investigated. The level of the relative ¹²⁷I¹²⁹I line strength can be readily found by fitting the simulation results into the measured frequency shift data. The limitations of the conventional matrix analysis in the presence of ¹²⁹I are considered.

An international voltage standard comparison has been carried out by six metrological laboratories using a Zener diode as the travelling voltage standard. The results show agreement within the uncertainty of the measurements: this is limited only by the stability of the Zener, which shows a very linear drift. The data present a residual standard deviation of 0,047 μV from the linear regression fit for the 1,018 V output. This makes it possible to confirm the performance of the new 1V TPYCEA Josephson voltage standard.