Article pubs.acs.org/est
Adsorption of Arsenic on Polyaluminum Granulate Jasmin Mertens,*,†,‡ Jérôme Rose,§,¶ Ralf Kag̈ i,‡ Perrine Chaurand,∥,¶ Michael Plötze,⊥ Bernhard Wehrli,†,‡ and Gerhard Furrer† †
Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zürich, Universitätsstrasse 16, 8092 Zürich, Switzerland Eawag, Swiss Federal Institute of Aquatic Science and Technology, Ü berlandstrasse 133, 8600 Dübendorf, Switzerland § CNRS, CEREGE, UMR 7330 (FR ECCOREV), 13545 Aix-en-Provence, Cedex 4, France ∥ Aix-Marseille University, CEREGE, UMR 7330, 13545 Aix-en-Provence, Cedex 4, France ⊥ Institute for Geotechnical Engineering (IGT), ETH Zürich, 8093 Zürich, Switzerland ¶ International Consortium for the Environmental Implications of Nanotechnology iCEINT, Europôle de l’Arbois, 13545 Aix-en-Provence, Cedex 4, France ‡
S Supporting Information *
ABSTRACT: The kinetics and efficiencies of arsenite and arsenate removal from water were evaluated using polyaluminum granulates (PAG) with high content of aluminum nanoclusters. PAG was characterized to be meso- and macroporous, with a specific surface area of 35 ± 1 m2 g−1. Adsorption experiments were conducted at pH 7.5 in deionized water and synthetic water with composition of As-contaminated groundwater in the Pannonian Basin. As(III) and As(V) sorption was best described by the Freundlich and Langmuir isotherm, respectively, with a maximum As(V) uptake capacity of ∼200 μmol g−1 in synthetic water. While As(III) removal reached equilibrium within 40 h, As(V) was removed almost entirely within 20 h. Micro X-ray fluorescence and electron microscopy revealed that As(III) was distributed uniformly within the grain, whereas As(V) diffused up to 81 μm into PAG. The results imply that As(V) is adsorbed 3 times faster while being transported 105 times slower than As(III) in Al hydroxide materials.
■
diffusion, and surface-complex formation.15 Adsorption processes of arsenic are known to be relatively fast.5,12 Slower processes are related to surface precipitation or to diffusion into the pore space of particles.16 Optimizations of aluminum materials, regarding pore size and distribution have improved arsenic uptake and kinetics.9 The coagulant polyaluminum chloride (PACl) with high amount of Al nanoclusters is a cost-effective and efficient coagulant for As removal.17,18 PACl consists of Al monomers, oligomers, and the nanoclusters Al13 (AlO4Al12(OH)24H2O127+), and Al30 (Al2O8Al28(OH)56(H2O)2618+),19 that exhibit high reactivity due to a great number of hydroxyl and water groups.20 The goal of this study was to investigate whether these Al nanoclusters exhibit favorable behavior to remove As from solution as main constituents of the adsorbent polyaluminum granulate. Since it is not clear whether the Al nanoclusters maintain their chemical structure entirely we use the term “Al oxide surfaces”. A mechanistic understanding of arsenic behavior is needed to improve removal performance of sorbents. Previous studies interpreted arsenic transport processes in iron and aluminum
INTRODUCTION Arsenic (As) is classified as carcinogenic, and a major exposure source for people is drinking water. To reduce the risk of Asrelated diseases, the EU drinking-water directive allows a maximum contaminant level of 10 μg/L.1 This threshold is exceeded in many water resources around the world, particularly in SE Asia.2,3 The biggest affected area in Europe is the Pannonian Basin (Central Europe) with arsenic concentrations up to 210 μg/L.4 Coagulation-coprecipitation and adsorption processes using iron (Fe)- or aluminum (Al)-based materials are used most frequently to remove As. Adsorbents can be easily applied in flow-through water-treatment processes, due to fast separation of As from solution. Fe- and Al-based adsorbents include ferrihydrite,5 hydrous ferric oxide,6 granular ferric hydroxide,7 ferric oxide (Fe2O3-physical vapor synthesis) and aluminum oxide (Al2O3-ALO101),8 mesoporous alumina,9,10 activated alumina,11 and amorphous aluminum hydroxide.12 Adsorption of arsenic onto activated alumina has been rated as best available technology, due to low operating costs and high As removal efficiency,13 and has been used effectively in point-ofuse water-treatment devices.14 The reaction of an adsorbate with the solid-solution interface is crucial for the pollutant’s mobility control. It includes the transport of the adsorbate to the surface by convection and molecular diffusion, interaction with the surface, intraparticle © 2012 American Chemical Society
Received: Revised: Accepted: Published: 7310
December 15, 2011 May 24, 2012 June 7, 2012 June 7, 2012 dx.doi.org/10.1021/es204508t | Environ. Sci. Technol. 2012, 46, 7310−7317
Environmental Science & Technology
Article
sorbents based on macroscopic data.5,9,11 We assessed the transport mechanisms of As(V) and As(III) transport mechanisms in porous aluminum hydroxides based on the analyses of the aqueous As fraction in combination with a thorough characterization of the sorbent using micro-X-ray fluorescence (μXRF) spectroscopy and scanning electron microscopy (SEM).
■
EXPERIMENTAL SECTION Chemicals. All chemicals were reagent grade from Sigma Aldrich or Fluka. To prepare arsenite (As(III)) or arsenate (As(V)) stock solutions (13.3 mM) the required amounts of NaAsO2 or NaHAsO4·7H2O were added to high-purity 18 MΩ water (Millipore, US). As(III) stock solution was kept in the dark to avoid As(III) oxidation. Synthetic groundwater was prepared after Roberts et al.21 by adding solid-state 1 mM CaCO3, 8 mM NaHCO3 and 0.6 mM MgCl2 to 18 MΩ water imitating groundwater with geogenic As contamination in the Pannonian Basin (Supporting Information Table S1). The pH was adjusted by air bubbling22 to 7.5 ± 0.1, a common pH in East-Pannonian groundwater. Al Granulate Preparation. Polyaluminum granulate (PAG) was generated by coagulation from PACl solutions after four repetitive cycles of aggregation, centrifugation, washing with ultrapure water to remove chloride ions, and consecutive airdrying. Altot concentrations of 1.2 M (= 40 mM Al30) were obtained by adding 234.3 g Locron-S powder (Clariant, Muttenz, Switzerland) in 1 L. Dried material was ground and pressed into pellets using a pill press with 103 kg cm−2 pressure. PAG Characterization. The specific surface area was determined with the BET method from 11-point N2 adsorption isotherm data at 77.3 K in the p/p0-range 0.05−0.3 (Autosorb 1MP, Quantachrome, Odelzhausen, Germany). Before analysis, the samples (0.28 g) were outgassed for 15 h at 150 °C. Size distribution and volume of macro- and mesopores was determined using low (375 kPa) and high (400 MPa) pressure mercury porosimeters (Pascal 140 and 440, Porotec, Hofheim, Germany). Skeleton density was measured in triplicates with a pycnometer using cyclohexane. Randomly oriented powder samples were analyzed to evaluate whether crystalline precipitates formed inside the granulate material. For that purpose, bulk mineralogy was determined using a θ−θ-X-ray diffractometer (D8 Advance, Bruker AXS, Karlsruhe, Germany) with Co−Kα radiation (40 kV, 30 mA) operating in step scan mode over an angular range of 4−70° with 0.02° (2θ) steps and 4 s count time. Prior to analysis, aliquot (∼1 g) samples from bulk dried granulate were ground to obtain
