The Photoacoustic Mercury Analyser and Antarctica

Updated 9 November 2023

Introduction

I joined the Department of Scientific and Industrial Research in 1974 where I initially worked on Atomic Spectrometry. This was mostly atomic absorption spectrometry and atomic emission spectrometry of solutions to determine their elemental composition. Later I became interested in other applications of atomic absorption spectrometry. This included the measurement of low concentrations of arsenic and mercury. Arsenic determination was by the release of the hydride into a heated quartz cell which produced arsenic vapour. Arsenic vapour concentration was measured by atomic absorption from an electrodeless discharge arsenic lamp. Mercury determination was by the release of cold mercury vapour into a long pyrex cell at room temperature. In both cases the atomic absorption was measured along the axis of the cells. My colleague Dr Byron Weissberg specialised in the latter method to determine mercury in rainfall.

I had an interest in the photoacoustic effect and I applied it to the visible and infrared absorption spectrometry of solids. I built a number of photoacoustic cells and the associated electronics for recording pulse shapes and amplitudes on a chart recorder. A Xenon strobe light source was used for illumination. A cell was also briefly mounted in a FTIR spectrometer. The signal from the microphone was able to be directly processed by the instrument electronics to obtain a spectrum. The setup could be used for FTIR spectrometry of opaque solids. Byron suggested I try mercury vapour in one of my cells with a mercury light source and it worked. Gradually I developed the technique and the geometry of the cells until it could respond to less than 1 picogram of mercury present in the sample cell. This proved to be sufficiently sensitive for routine trace mercury determinations. The prototype instrument is shown below.

The Photoacoustic Effect

When a sample is heated it absorbs energy and expands. When the heat source is removed the sample cools and contracts as energy is conducted and radiated away. By turning the source of energy on and off rapidly, say at 50 Hz, the expansion and contraction also occurs rapidly and is heard as sound. In the case of solids the rapid heating and cooling affects a boundary layer of air causing it to expand and contract making sound. In gaseous samples absorption of modulated light causes expansion and contraction of the gas at the modulation frequency. This signal is easily detected as sound by a standard electret microphone. A suitable electret microphone costs about $3.50.

In the following photo a microphone is located inside the brass chamber at left. A small hole in the fitting stops pressure transients damaging the microphone when the chamber is closed or opened by the window at right. It is sealed before use. The window is a mylar film to allow use in the infrared. The centre aluminium spacer minimises the dead volume and allows a gas pressure bypass to the microphone. The sample is stuck onto to the right-end using double sided sellotape.

This small chamber worked when directly connected to an auxiliary signal input of an FTIR spectrometer, without any special setup. An absorbed energy spectrum was easily displayed. A phonographic pre-amplifier might be useful in this application. This application was in the late 1970s so there probably not much here that is relevant today.

The Photoacoustic Effect in Mercury vapour

The photoacoustic mercury detector used a 15 watt germicidal mercury light source modulated at 50 Hz. The photoacoustic cell was illuminated transversely and was of a similar length to the light source. An aluminium foil reflector surrounded both. A microphone was positioned in a branch at one end of the cell. Cigarette filters partially blocked each end of the cell while allowing the passage of the carrier gas and mercury vapour. The presence of a diatomic carrier gas such as nitrogen or air quenched any fluorescence, producing heat, and therefore sound.

A second microphone and cell connected in opposition canceled any external sounds. 50 Hz narrow-band filtering also discriminated against any ambient noise. Amplification of the sound signal and rectification produced an analog signal with an amplitude directly proportional to the sample concentration. A sample and hold circuit gave a static output for measurement with a digital volt meter. This reading represented the maximum pulse amplitude and, in turn, the concentration of mercury.

Low levels of mercury were pre-concentrated on a gold collector which can be seen just in front of the transformer below at right. A timer circuit heated a nichrome coil around the gold collector. A pulse of mercury vapour was released into one cell, the second cell remaining blank.

In front, below is a mercury vapour reduction apparatus with an added bypass to reduce noise in the carrier gas flow, when a reading was made. The electronics comprised, at left, an AC amplifier with a 50 Hz Twin-T narrow band filter. The sample and hold circuit is on the middle board, being added later in the instrument development. At right is the gold collector heater-timer circuit and the 12 volt power supply.

In this final prototype the germicidal lamp used an 8 watt ballast in series with a 1N4007 diode. This provided half-wave rectified power at 50 Hz to the lamp. An isolating transformer should be used to keep the voltage imbalance away from the mains supply. This was caused by the half wave rectified DC,

The preferred cell was a 6 mm diameter silica tube or a clear FEP (Fluorinated ethylene-propylene) tube. The carrier gas flowed past a septum where mercury vapour could be admitted. The carrier gas can be nitrogen or air. With air a reducing vapour needs to be added to prevent oxidation of the mercury vapour in the presence of UV light. The reducing vapour can be a hydrocarbon or hydrogen gas. With hydrogen, a very low volume, is added to the air supplied from an aquarium pump. A reservoir and filter was used to remove any pump noise. Hydrogen was derived from a fuel cell running as an electroliser. Only tiny quantities of hydrogen were needed so the cell could be run at a small current.

The circuit is relatively simple as the amplifiers a, b, c and d are a single TL084 integrated circuit. The remaining two integrated circuits could be a single TL082. The microphone balancing is done by exposing both microphones to the same 50 Hz sound source. A trimming resistor was applied to the most sensitive microphone. A continuous reading could be taken from the output of the integrator. The active filter was adjusted, where the output of the integrator maximised when only one microphone is exposed to a 50 Hz sound source.

The sample and hold circuit and the built-in digital meter are optional, as there are many ways to process or log the signal from the integrator output. The gold collector can be heated as needed, with an alcohol flame.

The references below cover the development and use of this instrument.

Using Audacity to replace most of the electronics

There are software equivalents to this circuit that could be adapted, perhaps starting with audio software like Audacity. A phono preamplifier might be needed. The signal processing involves converting the stereo input into two separate mono tracks, inverting one track, mixing the two stereo tracks into one new mono track, rectifying and applying a 1 Hz low pass filter.

• Audio Track/Split Stereo Track to Mono. (Use the menu to the left of the audio tracks.)
• Highlight one track.
• Effect/Invert
• Highlight both mono tracks.
• Tracks/Mix/Mix and Render to New Track
• Rectify new track with Adam-V:Rectivert Effects plugin
• Effect/Low pass filter - 1 Hz full-wave

The output should correspond to the signal from the integrator The microphones should connect directly to the preamplifier and a suitable voltage source via a 10k resistor. This sequence could be written as a macro. This might be useful as a laboratory demonstration of the photoacoustic effect. A single channel input would simplify the commands, leaving just Rectivert and the low pass filter.

Antarctica

The Photoacoustic Mercury Analyser was used for several seasons in Antarctica. Air and snow were sampled to establish baseline levels of mercury. The instrument was able to operate at extremely low temperatures, as well as in a laboratory setting.

Publications

1. The Photoacoustic Effect in Mercury Vapour. J. E. Patterson, PhD Thesis, Victoria University, Wellington, 1988.
2. A sensitive photoacoustic mercury detector, J. E. Patterson, Anal. Chim. Acta, 136, 1982, 321.
3. A differential photoacoustic mercury detector, J. E. Patterson, Anal. Chim. Acta, 164, 1984, p119-126.
4. Relation between Photoacoustic Amplitude and Quenching of Mercury 3P1 --- 1S0 Fluorescence by hydrogen in Argon, J. E. Patterson, J. Chem. Soc., Faraday Trans. 2, 83, 1987, 255.
5. Mercury in Human Breath from Dental amalgams, J. E. Patterson, B. G. Weissberg and P. J. Dennison. Bull. of Environ. Contam. And Toxicol. 34, 1985, 459-468.
6. Atmospheric Mercury Concentrations inside Scott Base. S. J. De Mora, J. E. Patterson and D. M. Bibby, New Zealand Antarctic Record, 8, 1, 1987, 5.
7. Levels of Atmospheric Mercury at Two Sites in the Wellington Area, New Zealand. D. M. Bibby and J. E. Patterson, Environmental Technology Letters, 9, 1988, 71.
8. Gas Chromatographic Detector Based on Enhancement of Mercury 253.7 nm Photoacoustic Signals by Hydrocarbons in a Hydrogen Carrier. J. E. Patterson, Anal. Chim. Acta, 226, 1989, 99.
9. Mercury Content of Antarctic Surface Snow: Initial Results. Dick, A.L., Sheppard, D.S. and Patterson, J.E. Atmospheric Environment 24 A(4) 973-978 (1990).
10. Mercury content of Antarctic Ice and Snow: Further results. Sheppard, D.S., Patterson, J.E., Lyon, G.L. and McAdams, M.K. Atmospheric Environment, 25A (8) 1657-1660 (1991).
11. Baseline atmospheric mercury studies at Ross Island, Antarctica. de Mora, S.J., Patterson, J.E. and Bibby, D.M.,Antarctic Science 5 (3), 323-321 (1993)/
12. Construction of a Photoacoustic Mercury Detector J. E. Patterson CD 2339, DSIR Chem Div, June, 1984
13. Photoacoustic Detection of Mercury. J. E. Patterson NZIC Conference Poster, August, 1984.
14. Atmospheric Mercury, D M Bibby and J E Patterson, Atmospheric Chemistry Workshop Poster, Wellington, 23-24 May, 1985.
15. The Photoacoustic Detection of Mercury. J E Patterson, Conference poster presented at a meeting entitled "Flash photolysis and its applications" in honour of Sir George Porter P.R.S. at the Royal Institution of Great Britain, London, 1986.
16. Mercury in Antarctic Snow. D S Sheppard, J E Patterson and A L Dick. NZIC conference paper, Wellington, 20-23 August, 1990.