Ion balance in waters through inductively coupled plasma optical emission spectrometry

Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) has been employed to carry out the determination of both major anions and cations in water samples. The anion quantification has been performed by means of a new automatic accessory. In this device chloride has been determined by continuously adding a silver nitrate solution. As a result solid silver chloride particles are formed and retained on a nylon filter inserted in the line. The emission intensity is read at a silver characteristic wavelength. By plotting the drop in silver signal versus the chloride concentration, a straight line is obtained. As regards bicarbonate, this anion has been on-line transformed into carbon dioxide with the help of a 2.0 mol L−1 nitric acid stream. Carbon signal is linearly related with bicarbonate concentration. Finally, information about sulfate concentration has been achieved by means of the measurement of sulfur emission intensity. All the steps have been simultaneously and automatically performed. With this setup detection limits have been 1.0, 0.4 and 0.09 mg L−1 for chloride, bicarbonate and sulfate, respectively. Furthermore, it affords good precision with RSD below 6 %. Cation (Ca, Mg, Na and K) concentration, in turn, has been obtained by simultaneously reading the emission intensity at characteristic wavelengths. The obtained limits of detection have been 8 × 10−3, 2 × 10−3, 8 × 10−4 and 10−2 mg L−1 for sodium, potassium, magnesium and calcium, respectively. As regards sample throughput, about 30 samples h−1 can be analysed. Validation results have revealed that the obtained concentrations for these anions are not significantly different as compared to the data provided by conventional methods. Finally, by considering the data for anions and cations, precise ion balances have been obtained for well and mineral water samples.

Applying reactive models to column experiments to assess the hydrogeochemistry of seawater intrusion: optimising ACUAINTRUSION and selecting cation exchange coefficients with PHREEQC

Three sets of laboratory column experimental results concerning the hydrogeochemistry of seawater intrusion have been modelled using two codes: ACUAINTRUSION (Chemical Engineering Department, University of Alicante) and PHREEQC (U.S.G.S.). These reactive models utilise the hydrodynamic parameters determined using the ACUAINTRUSION TRANSPORT software and fit the chloride breakthrough curves perfectly. The ACUAINTRUSION code was improved, and the instabilities were studied relative to the discretisation. The relative square errors were obtained using different combinations of the spatial and temporal steps: the global error for the total experimental data and the partial error for each element. Good simulations for the three experiments were obtained using the ACUAINTRUSION software with slight variations in the selectivity coefficients for both sediments determined in batch experiments with fresh water. The cation exchange parameters included in ACUAINTRUSION are those reported by the Gapon convention with modified exponents for the Ca/Mg exchange. PHREEQC simulations performed using the Gains-Thomas convention were unsatisfactory, with the exchange coefficients from the database of PHREEQC (or range), but those determined with fresh water – natural sediment allowed only an approximation to be obtained. For the treated sediment, the adjusted exchange coefficients were determined to improve the simulation and are vastly different from those from the database of PHREEQC or batch experiment values; however, these values fall in an order similar to the others determined under dynamic conditions. Different cation concentrations were simulated using two different software packages; this disparity could be attributed to the defined selectivity coefficients that affect the gypsum equilibrium. Consequently, different calculated sulphate concentrations are obtained using each type of software; a smaller mismatch was predicted using ACUAINTRUSION. In general, the presented simulations by ACUAINTRUSION and PHREEQC produced similar results, making predictions consistent with the experimental data. However, the simulated results are not identical to the experimental data; sulphate (total S) is overpredicted by both models, most likely due to such factors as the kinetics of gypsum, the possible variations in the exchange coefficients due to salinity and the neglect of other processes.