Anions in seawater - Determination of Anions in Natural and Treated Waters - page 385

the system. Comparable Teflon tubing and connectors are used downstream from the
impinger.
The helium carrier gas and nitric oxide from the sample pass sequentially through a
salt/ice cold trap to condense water vapour, an anhydrous sodium carbonate drying tube
to remove remaining traces of water vapour and acid gases and a Teflon submicron filter
to exclude particles. The gas stream then enters a Bendix Model 8101-C Nitrogen Oxides
Analyser operated in the 'NO only' mode, where nitric oxide is determined
chemiluminescently.
The instantaneous concentration in the gas flow is recorded on a strip chart recorder
operated at an input sensitivity of 10mV full scale, and a chart speed of 0.2in min−1
( 5 mm min−1 ). The integrated area is recorded by using an Analog Devices AD537J V to f
converter connected to the 1V output of the nitrogen oxides analyser and the output of the
V to f converter is totalled using a Heath Model IM 4100 frequency counter operated in
the totalise mode. The V to f converter is mounted internally in the frequency counter and
power is supplied from its logic circuits.
Linear calibration plots are obtained by this procedure covering the range up to 1nmol
L−1 for nitrate and for nitrite.
3.20.4 Continuous flow analysis
Hydes and Hill [239] used the copper-cadmium reduction method to determine nitrate in
seawater. The construction of a copper-cadmium (50-50 w:w) reductor column for use
during continuous flow analysis at sea is described. A100% yield could be obtained using
a 20cm×3mm column fitted with grains of copper-cadmium alloy between 500 and
350µm in size. The column maintained its reactivity during 3 months storage prior to an
oceanographic cruise, and during the 4-week cruise period. Its performance was similar
to that of the Stainton-type cadmium wire column but it had the advantages of easier
preparation and easier control of reductor volume.
3.20.5 Flow injection analysis
Flow injection analysis is another technique that has been applied to the determination of
nitrate and nitrite in seawater. Anderson [130] used flow injection analysis to automate
the determination of nitrate and nitrite in seawater. The detection limit of his method was
0.1µmol L−1 . However, the sampling rate was only 30 per h which is low for flow
injection analysis. Reactions seldom go to completion in a determination by flow
injection analysis [131,132] because of the short residence time of the sample in the
reaction manifold. Anderson selected a relatively long residence time so that the extent of
formation of the azo dye was adequate to give a detection limit of 0.1µmol L −1 . This
reduced the sampling rate because only one sample is present at a time in the post-
injector column in flow injection analysis. Any increase in reaction time causes a
corresponding increase in the time needed to analyse one sample.
Johnson and Petty [133] reduced nitrate to nitrite with copperised cadmium which was
then determined as an azo dye, but the method is automated by means of flow injections
analysis technique [134]. More than 75 determinations can be made per hour. The