Anions in non saline waters - Determination of Anions in Natural and Treated Waters - page 197

Table 2.21 Precision results for nitrate determination using spongy cadmium

Standard deviation a mg L −1

Solution Mean absorbance

S w b S b c S c d

0.000mg L −1

nitrate-nitrogen 0.005 0.0024 (6) - -

0.140mg L −1

nitrate-nitrogen 0.047 0.0042 (6) 0.0072 (5) 0.01 10

2.100mg L −1

nitrate-nitrogen 0.726 0.0279 (6) NSe (5) 0.0416

a The degrees of freedom for the standard deviations are given in parentheses; b Within-batch;
c Between-batch; d Total standard deviation; a Not statistically significant
Source: Reproduced with permission from the Royal Society of Chemistry [391]

The analytical implications of this work in the determination of nitrite are as follows.
Although cadmium filings are more easily and reproducibly prepared than spongy
cadmium, and have been shown to give more precise results [390] for the batch
procedure, reduction by spongy cadmium should, perhaps, be preferred because it is
much less prone to interferences. Table 2.21 gives basic performance characteristics
determined for this method by using 0.5-0.6g of spongy cadmium and a shaking time of
20min. Values for S t , showed that the precision of the method ranged from a relative
standard deviation S r =7.8% at 0.14mg L −1 to S r =2% at 2.1mg L -1 of nitrate-nitrogen. Sw
can be used to calculate the criterion of detection using ts w √ 2 where t is Student's t. A
value of 0.006g L −1 is obtained for the 95% confidence level.
Reduction conditions must always be optimised for natural samples rather than
standards, and, ideally, for each individual water body that is measured. The results
suggest a realistic quality control procedure, which can be used to assess the accuracy of
measurements on a variety of water samples; two measurements could be made, each at
different effective reduction capacities. This measurement technique could be very easily
arranged in an automated system by having two channels with two different reduction
column lengths. Disparity between the duplicate results would indicate the possible
presence of an interfering species.
Olsen [397] demonstrated that 2.5-25µM phosphate may decrease reduction of nitrate
by 10-40% in cadmium-copper columns and stated that the effect of phosphate is gradual
and varies with the age of the column. These and other interferences are much less likely
in the method described by Jones [389] than in a column reduction method because
shaking has a continual scouring effect on the cadmium during the reduction process and
the cadmium is cleaned before each reduction. According to Nydahl [398], chloride ions
can have a strongly retarding effect on the rate of nitrate reduction. However, Jones [389]
found that, when the reduction solution was buffered above pH 8 and optimum cadmium
contact was used, reduction of 20µmol L −1 nitrate was more than 99% complete, even in
the presence of sodium chloride concentration corresponding to that of seawater.
Certain metals and phosphate may interfere with nitrate reduction. Precipitation of
metal salts onto the cadmium surface may be prevented by the addition of EDTA to the