As the final absorbance signal is controlled by the amount of nitrate in the reagent sample mixture, the injection volume of the sample was varied between 75 and 800 mL to determine the optimized value.
Signal in creases were observed for injection volumes up to 500 mL and remained stable a fterwards as shown in Fig. 2(e). Thus, the injection volume was maintained at 500 mL for all subsequent measurements to reduce the sample consumption
(e.g. for sediment pore water sample analysis).