By using the equilibrium computer program MINTEQA2, a quality control standard suitable for assessing measurement errors of pH, acid neutralizing capacity (ANC, formerly titration alkalinity) and conductivity of dilute, neutral pH solutions [ionic strength ≈ 4.6 × 10–4 mol kg–1 of solvent (water)] has been developed and tested. A dilute phosphate standard is prepared as a 200-fold dilution (by mass) of 0.025 mol kg–1(of solvent) KH2PO4 and Na2HPO4 standard buffer solution. At 25 °C and in equilibrium with 340 vppm (parts per million by volume) CO2, the standard has a theoretical pH of 6.89, a calculated ANC of 125 × 10–6 equiv kg–1(124.6 × 10–6 equiv dm–3) and a calculated conductivity of 37.6 × 10–4 S m–1. Theoretical pH and calculated ANC and conductivity values were also determined at equilibrium with laboratory CO2 concentrations ranging between 0 and 680 vppm. The dilute phosphate standard was measured repeatedly over a 6 month period at three laboratories. Measurements agreed well with the theoretical values (mean pH = 6.87, 2s=±0.100, n= 362; mean ANC = 124.5 × 10–6 equiv dm–3, 2s=±8.78, n= 294; mean conductivity = 37.7 × 10–4 S m–1 at 25 °C, 2s=±5.22, n= 308). The long-term systematic and random errors associated with the dilute phosphate standard are reduced for pH (because of reduced sensitivity to changes in CO2), and comparable for ANC and conductivity as compared to a dilute hydrogen carbonate standard (7.0 × 10–5 mol kg–1 NaHCO3 and 1.75 × 10–4 mol kg–1 KCl). For the dilute phosphate standard, among-day variation within a laboratory is the greatest source of error for ANC and conductivity measurements. For the dilute hydrogen carbonate standard, among-laboratory variation is the greatest source of error in pH measurements. Within-day variation is the greatest source of error in ANC measurements for the dilute hydrogen carbonate standard, while among-day variation within a laboratory is the greatest source of error for conductivity measurements of both dilute standards. Additional application of total speciation programs to the solving of design problems for quality control and assurance samples appears promising and justified.
Read full abstract