Distribution of arsenic(III), arsenic(V) and total inorganic arsenic in porewaters from a thick till and clay-rich aquitard sequence, Saskatchewan, Canada

Inorganic arsenic species were measured in the porewaters collected from eighteen piezometers installed between 3 and 91.4 m below ground in a thick till and clay-rich aquitard sequence located in southern Saskatchewan to investigate the distribution of, and controls on arsenic speciation in the sequence. Aqueous concentrations of As(V), As(III) and total As are in the range of 0.31-97, 0.71-21 and 3.2-98 ppb, respectively. Profiles of As(III) and As(V) concentration distribution with depth broadly track that of total As: erratic increases to 15.2 m, then more uniform concentrations to 88 m. Aqueous arsenic is accumulated at the upper redox transition zone (6-14 m). The alkaline porewater at 91.4 m contains the highest concentrations of As(V) and total As, which might result from the facilitated desorption of arsenate from the host solid due to decrease of positive surface charge of the oxides in alkaline solution. The ratio of As(V)/As(III) is greater than unity in the uppermost oxidized porewater (3 m), less than unity from 4.6 to 71.6 m, and greater than unity in the lowest four porewaters (76.2 to 91.4 m). In the 3 m porewater low As(III) but high As(V)/As(III) is due to the oxidized nature of the near surface weathered till. The high As(V)/As(III) in the deepest porewater at 91.4 m likely results from the enhanced and heterogeneous oxidation of As(III) to As(V) on clay mineral surfaces in the alkaline solution. Total As and arsenic speciation may not be controlled by As, Fe or Mn concentrations in the host till or clay. Dissolved As(V) and total As positively covary with aqueous chloride, whereas dissolved As(III) is independent of aqueous chloride. Aqueous As(III), and to a less extent As(V) and total As are positively correlated with dissolved Mn in the till. In the clay, aqueous As(V) and total As show strong negative covariation with Mn. However, aqueous As(III), As(V) and total As exhibit almost no correlation with total dissolved Fe in the till. The As(V)/As(III) ratio has strong negative correlation with dissolved Mn, but positive covaration with dissolved chloride. Generally good agreement between the redox potentials (Eh) calculated from aqueous As(V) and As(III) concentrations and those measured by a Pt electrode throughout most of the unoxidized till suggests the suitability of using As(V)-As(III) redox couple as a redox indicator for the studied aquitard system. However, large negative bias of the calculated Eh from the measured Eh in the oxidized till/upper unoxidized till and the clay is attributed to errors associated with the field measurements of Eh. Copyright (C) 2000 Elsevier Science Ltd.