Comment Comment on: ‘‘The dissolution rates of gibbsite in the presence of chloride, nitrate, silica, sulfate and citrate in open and closed systems at 20 °C’’ by M. Dietzel and G. Bo ¨hme David J. Wesolowski a, * , Donald A. Palmer a , Pascale Be ´ne ´zeth b a Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6110, USA b Laboratoire des Me ´canismes de Transfert en Ge ´ologie, CNRS-UMR5563, 14 Avenue Edouard Belin, 31400 Toulouse, France Received 18 April 2005; accepted in revised form 4 October 2005 The recent paper by Dietzel and Bo ¨ hme (2005) reports the rate of dissolution of gibbsite under far-from equilibri- um conditions and low pH at 20 °C, indicating that the rates are about the same in chloride and nitrate solutions, faster in silica-bearing solutions, and fastest in sulfate- and citrate-bearing solutions. This information is highly relevant to environmental geochemistry and soil science. However, the authors failed to mention our earlier study of the dissolution of gibbsite in NaCl–H 2 SO 4 and NaCl– HCl solutions (0.1 molal ionic strength, 0.01 molal H + ) at 5 °C(Ridley et al., 1997), in which we demonstrated that the release rate of Al 3+ from gibbsite under the specified conditions, via the reaction AlðOHÞ 3;gibbsite þ 3H þ $ Al 3þ þ 3H 2 O ð1Þ was approximately 10 times faster in the sulfate medium than in the chloride medium, and appeared to be controlled by a zero-order rate-determining reaction in batch experi- ments at far-from-equilibrium conditions. Dietzel and Bo ¨hme also failed to cite our recent study (Xiao et al., 2002) of the formation constants of aluminum sulfate com- plexes via the reactions: Al 3þ þ SO 4 2 $ AlSO 4 þ ð2Þ Al 3þ þ 2SO 4 2 $ AlðSO 4 Þ 2  ð3Þ in which we reported values of log K r(2) = 3.84 ± 0.12 and log K r(3) = 5.58 ± 0.12 at 25 °C, significantly higher than the values of 3.35 and 4.90, respectively, used by the authors in their speciation calculations. The authors used values from much earlier studies that were thoroughly dis- cussed in our paper. There is no question that the literature on the behavior of aluminum in aqueous systems is vast and constantly growing, but in calculating solution species distributions in order to interpret their dissolution rate results, the authors have chosen equilibrium constants for the relevant reactions that ignore much of the recent literature on this maturing subject. They tabulate an equilibrium constant for reaction (1) at 25 °C, log K r(1) = 8.04 (Singh, 1974), that differs significantly from the value of 7.74 ± 0.06 obtained in our more recent study (Palmer and Wesolowski, 1992). In this study, we used a commercial gibbsite that was exten- sively pretreated by rinsing/settling, followed by reaction with 1 molal HCl to remove high-energy particles and sur- face defects (Bloom and Weaver, 1982; Wesolowski, 1992, 2002), and final rinses in DI water. Dietzel and Bo ¨hme show that a similar pretreatment of their commercial start- ing material greatly affects the dissolution rates of gibbsite, though they fail to reference these previous papers in dis- cussing the benefits of pretreatment. In Palmer and Weso- lowski (1992), we argued that the earlier results of Singh (1974) and several previous studies were most likely super- saturated with respect to well-crystallized gibbsite, due to failure to pretreat the solid. However, we showed excellent agreement with studies that followed a similar pretreatment step (Bloom and Weaver, 1982; Peryea and Kittrick, 1988; Nagy and Lasaga, 1993). Dietzel and Bo ¨ hme report a value for log K r(1) from their own near-equilibrium experiments at 20 °C of 7.82 ± 0.04, significantly lower than the value calculated from the tem- perature function (8.05 ± 0.07) for this reaction given by Palmer and Wesolowski (1992). This function was based www.elsevier.com/locate/gca Geochimica et Cosmochimica Acta 70 (2006) 2140–2142 0016-7037/$ - see front matter Ó 2006 Published by Elsevier Inc. doi:10.1016/j.gca.2005.10.038 * Corresponding author. Fax: +1 865 574 4961. E-mail address: wesolowskid@ornl.gov (D.J. Wesolowski).