Comparing dissolved reactive phosphorus measured by DGT with ferrihydrite and titanium dioxide adsorbents: anionic interferences, adsorbent capacity and deployment time.

Two adsorbents (Metsorb and ferrihydrite) used in binding layers with the diffusive gradients in a thin film technique were evaluated for the measurement of dissolved reactive phosphorous (DRP) in synthetic and natural waters. Possible interferences were investigated with Cl(-) (up to 1.35 mol L(-1)) and SO(4)(2-) (up to 0.056 mol L(-1)) having no affect on either DGT binding layer, and HCO(3)(-) (up to 5.7 mmol L(-1)) having no effect on Metsorb-DGT, over 4 days. However, HCO(3)(-) interfered with the ferrihydrite-DGT measurement at concentrations typical of many natural waters (≥0.7 mmol L(-1)) after a deployment period of 1-2 days. The capacity of the Metsorb binding phase for DGT response was ∼37,000 ng P, whereas the capacities of a low-mass (17.8 mg of adsorbent per DGT sampler) and high-mass (29.2mg of adsorbent per DGT sampler) ferrihydrite binding phase were substantially lower (∼15,000 ng P and ∼25,000 ng P, low-mass and high-mass, respectively). Increasing the capacity of the ferrihydrite adsorbent allowed the ferrihydrite-DGT to be utilized for up to 3 days before interference by HCO(3)(-) was observed. Seawater deployments demonstrated that even high-capacity ferrihydrite-DGT devices underestimated the DRP concentration by 37%, whereas Metsorb-DGT measurements were accurate. The Metsorb-DGT is superior to the ferrihydrite-DGT for determining DRP over deployment times greater than 1 day and in waters with ≥0.7 mmol L(-1) HCO(3)(-). Based on the experience obtained from this detailed validation process, the authors propose a number of key requirements that need to be considered when developing new DGT binding layers, with testing the performance over longer deployment times being critical.