Perchlorate reduction by autotrophic bacteria in the presence of zero-valent iron

A series of batch experiments were performed to study the combination of zero-valent iron (ZVI) with perchlorate-reducing microorganisms (PRMs) to remove perchlorate from groundwater. In this method, H2 produced during the process of iron corrosion by water is used by PRMs as an electron donor to reduce perchlorate to chloride. Perchlorate degradation rates followed Monod kinetics, with a normalized maximum utilization rate (rmax) of 9200 microg g(-1) (dry wt) h(-1) and a half-velocity constant (Ks) of 8900 microg L(-1). The overall rate of perchlorate reduction was affected by the biomass density within the system. An increase in the OD600 from 0.025 to 0.08 led to a corresponding 4-fold increase of perchlorate reduction rate. PRM adaptation to the local environment and initiation of perchlorate reduction was rapid under neutral pH conditions. At the initial OD600 of 0.015, perchlorate reduction followed pseudo-first-order reaction rates with constants of 0.059 and 0.033 h(-1) at initial pH 7 and 8, respectively. Once perchlorate reduction was established, the bioreductive process was insensitive to the increases of pH from near neutral to 9.0. In the presence of nitrate, perchlorate reduction rate was reduced, but not inhibited completely.     

Role of humic acid and ouinone model compounds in bromate reduction by zerovalent iron

Experiments were conducted to examine the role of humic acid and quinone model compounds in bromate reduction by Fe(0). The reactivity of Fe(0) toward bromate declined by a factor of 1.3-2.0 in the presence of humic acid. Evidence was obtained that the quick complexation of humic acid with iron species and its adsorption passivated the iron surface and decreased the rate of bromate reduction by Fe(0). On the other hand, in the long run, the reduced functional groups present in humic acid were observed to regenerate Fe(II) and reduce bromate abiotically. Compared with the case of humic acid only, the simultaneous presence of Fe(II) and humic acid significantly increased the bromate removal rate. Fe(III)/Fe(II) acted as a catalyst in the oxidation of humic acid by bromate. Anthraquinone-2,6-disulfonate (AQDS) and lawsone did not cause any significant effect on the bromate reduction rate by Fe(0). However, the redox reactivity of lawsone in the presence of Fe(III) was evident, while AQDS did not show any under the tested conditions. The difference was attributable to the presence/ absence of reducing functional groups in the model compounds. The electron spin resonance further demonstrated that the redox functional groups in humic acid are most likely quinone-phenol moieties. Although the bromate reduction rate by regenerated Fe(II) is a few times slower than that by Fe(0), the reactive Fe(II) can be, alternatively, reductively formed to maintain iron surface activation and bromate reduction to prolong the lifetime of the zerovalent iron.     

Understanding soluble arsenate removal kinetics by zerovalent iron media

Zerovalent iron filings have been proposed as a filter medium for removing arsenic compounds from potable water supplies. This research investigated the kinetics of arsenate removal from aqueous solutions by zerovalent iron media. Batch experiments were performed to determine the effect of the iron corrosion rate on the rate of As(V) removal. Tafel analyses were used to determine the effect of the As(V) concentration on the rate of iron corrosion in anaerobic solutions. As(V) removal in column reactors packed with iron filings was measured over a 1-year period of continuous operation. Comparison of As(V) removal by freely corroding and cathodically protected iron showed that rates of arsenate removal were dependent on the continuous generation of iron oxide adsorption sites. In addition to adsorption site availability, rates of arsenate removal were also limited by mass transfer associated with As(V) diffusion through iron corrosion products. Steady-state removal rates in the column reactor were up to 10 times faster between the inlet-end and the first sampling port than between the first sampling port and the effluent-end of the column. Faster removal near the influent-end of the column was due to a faster rate of iron oxidation in that region. The presence of 100 microg/L As(V) decreased the iron corrosion rate by up to a factor of 5 compared to a blank electrolyte solution. However, increasing the As(V) concentration from 100 to 20,000 microg/L resulted in no further decrease in the iron corrosion rate. The kinetics of arsenate removal ranged between zeroth- and first-order with respect to the aqueous As(V) concentration. The apparent reaction order was dependent on the availability of adsorption sites and on the aqueous As(V) concentration. X-ray absorption spectroscopy analyses showed the presence of iron metal, magnetite (Fe3O4), an Fe(III) oxide phase, and possibly an Fe(II,III) hydroxide phase in the reacted iron filings. These mixed valent oxide phases are not passivating and permit sustained iron corrosion and continuous generation of new sites for As(V) adsorption.     

Reductive debromination of polybrominated diphenyl ethers by zerovalent iron

Polybrominated diphenyl ethers (PBDEs) are a new class of global, persistent, and toxic contaminants, which need proper remediation technologies. PBDE degradation in the environment is not well understood. In this study, degradation of PBDEs with zerovalent iron was investigated with six BDEs, substituted with one to 10 bromines. Within 40 days 92% of BDE congener 209 was transformed into lower bromo congeners. During the initial reaction period of BDE 209 (<5 days), hexa- to heptabromo BDEs were the most abundant products, but tetra- to pentabromo congeners were dominant after 2 weeks. The amount of mono- to tribromo BDEs was steadily increased during the experiments. BDEs 28, 47, 66, and 100 also showed a stepwise accumulation of lower bromo congeners. No oxidation products were detected in all experiments. The results showed that a stepwise debromination from n-bromoto (n-1)-bromodiphenyl ethers was the dominant reaction in all congeners. The reaction rate constants of lower bromo BDEs decreased as the number of bromines decreased. The initial reductive debromination rate constants were positively correlated with the heats of formation of BDEs. The preferential accumulation of specific congeners was observed in the experiment with BDEs 28, 47, 66, and 100, where the most abundant products were BDEs 15, 28, 37, and 47, respectively. Reactions proceeded to form more stable and less brominated products that have lower heats of formation. Almost all the possible isomers from a specific parent BDE were found in all the experiments, which was probably due to the small difference of heat of formation between the products (2-5 kcal/mol). Reactions of all congeners proceeded fast at the initial phase (<5 days) followed by a slow reaction. The rate of reductive debromination of BDE 209 was slower with environmentally relevant sulfide minerals (iron sulfide and sodium sulfide). However, the product congener pattern, produced by sulfide mineral catalysis, was nearly similar with that of zerovalent iron treatment. This may be a possible source of lower brominated BDEs in the environment. Debromination of PBDEs by zerovalent iron has high potential values for remediation of PBDEs in the environment.     

Chromium-removal processes during groundwater remediation by a zerovalent iron permeable reactive barrier

Solid-phase associations of chromium were examined in core materials collected from a full-scale, zerovalent iron permeable reactive barrier (PRB) at the U.S. Coast Guard Support Center located near Elizabeth City, NC. The PRB was installed in 1996 to treat groundwater contaminated with hexavalent chromium. After eight years of operation, the PRB remains effective at reducing concentrations of Cr from average values >1500 microg L(-1) in groundwater hydraulically upgradient of the PRB to values <1 microg L(-1) in groundwater within and hydraulically downgradient of the PRB. Chromium removal from groundwater occurs at the leading edge of the PRB and also within the aquifer immediately upgradient of the PRB. These regions also witness the greatest amount of secondary mineral formation due to steep geochemical gradients that result from the corrosion of zerovalent iron. X-ray absorption near-edge structure (XANES) spectroscopy indicated that chromium is predominantly in the trivalent oxidation state, confirming that reductive processes are responsible for Cr sequestration. XANES spectra and microscopy results suggest that Cr is, in part, associated with iron sulfide grains formed as a consequence of microbially mediated sulfate reduction in and around the PRB. Results of this study provide evidence that secondary iron-bearing mineral products may enhance the capacity of zerovalent iron systems to remediate Cr in groundwater, either through redox reactions at the mineral-water interface or by the release of Fe(II) to solution via mineral dissolution and/or metal corrosion.     

Uranium precipitation in a permeable reactive barrier by progressive irreversible dissolution of zerovalent iron

A permeable reactive barrier (PRB) containing zerovalent iron [Fe(O)] was installed at a former uranium milling site in Monticello, UT. A large-scale column experiment was conducted at the site to test the feasibility of Fe(O) to treat U prior to installing the PRB. Effluents from the field column experiment had pH values near 7.34, moderate decreases in C(IV) and Ca concentrations, and an elevated Fe concentration (27.1 mg/L). In contrast, groundwater exiting the PRB had a pH value of 9.82, decreases in C(IV) and Ca concentrations, and a low concentration of Fe (0.17 mg/L). A geochemical model was used to explain the chemical changes that occurred in both the field column experiment and the PRB. The model simulated the systems by the progressive irreversible dissolution of Fe(O). Modeling results indicated that a longer residence time in the PRB compared with the shorter residence time in the column contributed to the disparate effluent qualities. Prior to modeling, a controlled laboratory column experiment was conducted to help evaluate the dominant chemical mechanisms by which Fe(O) removes U from aqueous solutions. Results of the laboratory column experiment indicated that only a small amount of U could be adsorbed to ferric minerals, and, therefore, this mechanism was not considered in the model.     

EDTA degradation induced by oxygen activation in a zerovalent iron/air/water system

A method for the removal of ethylenediaminetetraacetic acid (EDTA) at room temperature and 1 atm is demonstrated. EDTA (1 mM, 50 mL) containing 2.5 g of granular zerovalent iron (ZVI) (20-40 mesh) was degraded in 2.5 h. Using a recently developed form of O2 activation, reactive oxygen species are generated in situ, resulting in the degradation of EDTA when complexed with FeII. ESI-MS measurements indicate that degradation of EDTA yields low-molecular carboxylic acids. The presence of oxygen is crucial: the observed pseudo-first-order rate constants for EDTA removal are kobs = 1.02 h(-1) (kSA = 1.85 +/- 0.046 L h(-1) m(-2)) and kobs = 0.04 h(-1) (kSA = 0.00724 +/- 0.002 L h(-1) m(-2)) under air and under N2 purge, respectively. kSA represents surface area normalized rate constants. Large excesses of EDTA in the reaction mixture slowthe rate of degradation. Increasing the concentration of EDTA from 1.0 to 10.0 mM while holding all other parameters constant gave observed rates of kobs = 1.02 +/- 0.26 h(-1) (kSA = 1.85 +/- 0.046 L h(-1) m(-2)) and kobs = 0.044 +/- 0.01 h(-1) (kSA = 0.00796 +/- 0.002 L h(-1) m(-2)), respectively. The rate-limiting step is determined to be homogeneous oxygen activation.     

Oxidation on zerovalent iron promoted by polyoxometalate as an electron shuttle

Most studies on zerovalent iron (ZVI) were mainly focused on the reductive transformation of halo- or nitrocompounds. Oxidation reactions occurring on ZVI have been recently recognized. In this study, we demonstrate that the oxidation pathways on ZVI can be accelerated by the presence of polyoxometalate (POM: nanosized metaloxygen cluster anion) serving as an electron shuttle. The ions, SiW12O40(4-) and PW12O40(3-), can mediate the electron transfer from the Fe0 surface to 02 while enhancing the production of H2O2, which subsequently initiates the OH radical-mediated oxidation through a Fenton-type reaction. The oxidation reaction was completely quenched by adding methanol as an OH radical-scavenger. On the other hand, PMo12O40(3-) completely inhibited the oxidative degradation by irreversibly scavenging an electron and holding it. We systematically investigated the effects of iron loading, the concentration of POM, and pH on the oxidative degradation kinetics of 4-chlorophenol in the POM-mediated ZVI system. The POM-mediated oxidations on ZVI were additionally tested for 12 organic contaminants and the rates were compared. Their oxidative degradation on ZVI was mostly enhanced in the presence of POM (SiW12O40(4-)). The present study provides a good model system upon which the ZVI-based oxidation technologies can be successfully enhanced and modified for further developments.     

Rapid dechlorination of polychlorinated dibenzo-p-dioxins by bimetallic and nanosized zerovalent iron

Polychlorinated dibenzo-p-dioxins and furans (PCDD/Fs), especially the 2,3,7,8-substituted congeners, are extremely toxic, persistent, and recalcitrant to remediation. Dechlorination of PCDD/Fs by zerovalent iron (ZVI) is thermodynamically feasible, but useful rates of reaction have not been previously reported. Here we show that ZVI (both micro- and nanosized ZVI, without palladization) dechlorinates PCDD congeners with four or more chlorines in aqueous systems, but the reaction is too slow to achieve complete dechlorination within a practical period of time. In contrast, palladized nanosized ZVI (Pd/nFe) rapidly dechlorinates PCDDs, including the mono- to tetra-chlorinated congeners. The rate of 1,2,3,4-tetrachloro dibenzo-p-dioxin (1,2,3,4-TeCDD) degradation using Pd/nFe was about 3 orders of magnitude faster than 1,23,4-TeCDD degradation using unpalladized ZVI. The distribution of products obtained from dechlorination of 1,2,3,4-TeCDD suggests that palladization shifts the pathways of contaminant degradation toward a greater role of H atom transfer rather than electron transfer.     

Transesterification by lipase entrapped in electrospun poly(vinyl alcohol) fibers and its application to a flow-through reactor

We entrapped lipase in electrospun poly(vinyl alcohol) fibers of approximately 1 mum in diameter and evaluated the transesterification activity by converting (s)-glycidol to glycidyl n-butyrate with vinyl n-butyrate. The initial transesterification rate of the entrapped lipase was 5.2-fold faster than that of non-treated lipase. The fibrous membrane could be used as a component of a flow-through reactor for continuous transesterification.     

Anionic homopolymers efficiently target zerovalent iron particles to hydrophobic contaminants in sand columns

The transport of microscale carbonyl iron powder suspensions modified with anionic homopolymers was studied in water-saturated sand columns containing well-dispersed hydrophobic sand grains. Sand grains functionalized with hexadecyltrimethoxysilane were coated with a eutectic mixture of dichlorobenzenes that was solid at -10 °C and was mixed by grinding with unmodified sand grains. The dichlorobenzene coating liquefied at the temperature of the transport experiments, and the coated grains were thus mimetic of uniform droplets of dense nonaqueous phase liquid (DNAPL) contaminants. By comparing iron particle transport in uncontaminated columns with those that contained a small fraction of DNAPL-coated sand grains, sticking coefficients for both types of grains could be estimated. The anionic polyelectrolytes tested (polyacrylate, carboxymethylcellulose, alginate, and metasilicate) all gave low particle sticking coefficients (0.004-0.05) to unmodified sand, as expected from earlier studies. However, iron particles modified with the two moderately hydrophobic polymers (carboxymethylcellulose and polyacrylate) had 30-fold higher sticking coefficients (0.40 and 0.13, respectively) to the model DNAPL surface than they did to the sand surface. In contrast, no significant difference between the two kinds of collector grains was found with the more polar polymers (metasilicate and alginate). The trend in sticking coefficients was correlated with the surface energy of the polymer-modified iron surface as measured by the static contact angle method. From these data one can conclude that the hydrophobicity of the polymer dispersant is a key factor in targeting zerovalent iron to DNAPL source zones in soil and groundwater.     

Congener-specific dechlorination of dissolved PCBs by microscale and nanoscale zerovalent iron in a water/methanol solution

Polychlorinated biphenyl (PCB)-contaminated sediments remain a significantthreatto humans and aquatic ecosystems. Dredging and disposal is costly, so viable in situ technologies to dechlorinate PCBs are needed. This study demonstrates that nanoscale zerovalent iron (ZVI) dechlorinates PCBs to lower-chlorinated products under ambient conditions, provides insight into structure-activity relationships between PCB isomers, and compares the reactivity of nanoscale ZVI to that of palladized microscale ZVI. Six PCB congeners were studied (22', 34', 234, 22'35', 22'45', and 33'44') to compare the initial rate of dechlorination of each and to monitor the order in which chlorines are removed. Using 200 g/L of nanoscale ZVI in a 30% MeOH/water mixture, observed surface-area-normalized pseudo-first-order PCB dechlorination rate constants ranged from 1 x 10(-6) to 5.5 x 10(-4) L yr(-1) m(-2) depending on the PCB congener tested. Using 200 g/L of palladized (0.05 wt %) microscale ZVI, surface-area-normalized pseudo-first-order PCB dechlorination rate constants were significantly faster and ranged from 3.8 x 10(-2) to 1.7 x 10(-1) L yr(-1) m(-2), but these rates were not sustainable. For nanoscale ZVI, nonorthosubstituted congeners had faster initial dechlorination rates than orthosubstituted congeners in the same homologue group. Chlorines in the para and meta position were predominantly removed over chlorines in the ortho position, which suggests that more-toxic coplanar PCB congeners are not likely to form from less-toxic noncoplanar, orthosubstituted congeners. Complete dechlorination was not observed over the course of the experiments. PCB dechlorination is rapid enough that nanoscale ZVI may offer novel in situ remedial alternatives for PCB-contaminated sediments.     

Chloride effect on TNT degradation by zerovalent iron or zinc during water treatment

Addition of corrosion promoters, such as sodium and potassium chloride, accelerated TNT degradation during water treatment using zerovalent zinc and iron. It was theorized that corrosion promoters could be used to accelerate electron generation from metallic species, create new reactive sites on the surface of metals during contaminated water treatment, and minimize passivating effects. The surface area normalized pseudo-first-order rate constant for the reaction of zerovalent zinc with TNT in the absence of KCl was 1.364 L x m(-2) x h(-1). In the presence of 0.3 mM and 3 mM KCI, the rate constant increased to 10.5 L x m(-2) x h(-1) and 51.0 L x m(-2) x h(-1), respectively. For the reaction with zerovalent iron and TNT, the rate constant increased from 6.5 (L/m2 x h) in the absence of KCl to 37 L x m(-2) x h(-1) using 3 mM KCl. The results demonstrate that chloride based corrosion promoters enhance the rate of TNT degradation. The in-situ breakage of the oxide layer using corrosion promoters was applied as a treatment to maintain the long-term activity of the metallic species. Zinc maintained a high reactivity toward TNT, and the reactivity of iron increased after 5 treatment cycles using 3 mM KCI. Zinc and iron scanning electron micrographs indicate that TNT degradation rate enhancement is caused by the pitting corrosion mechanism.     

Reduction of octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine by zerovalent iron: product distribution

RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine) and HMX (octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine) are cyclic nitramines ((CH2NNO2)n; n = 3 or 4, respectively) widely used as energetic chemicals. Their extensive use led to wide environmental contamination. In contrast to RDX, HMX tends to accumulate in soils due to its unique recalcitrance. In the present study, we investigated the potential of zerovalent iron (ZVI) to transform HMX under anoxic conditions. HMX underwent a rapid transformation when added in well-mixed anoxic ZVI-H2O batch systems to ultimately produce formaldehyde (HCHO), ammonium (NH4+), hydrazine (NH2NH2), and nitrous oxide (N2O). Time course experiments showed that the mechanism of HMX transformation occurred through at least two initial reactions. One reaction involved the sequential reduction of N-NO2 groups to the five nitroso products (1NO-HMX, cis-2NO-HMX, trans-2NO-HMX, 3NO-HMX, and 4NO-HMX). Another implied ring cleavage from either HMX or 1NO-HMX as demonstrated by the observation of methylenedinitramine (NH(NO2)CH2NH(NO2)) and another intermediate that was tentatively identified as (NH(NO2)CH2N(NO)CH2NH-(NO2)) or its isomer (NH(NO)CH2N(NO2)CH2NH(NO2)). This is the first study that demonstrates transformation of HMX by ZVI to significant amounts of NH2NH2 and HCHO. Both toxic products seemed to persist under reductive conditions, thereby suggesting that the ultimate fate of these chemicals, particularly hydrazine, should be understood prior to using zerovalent iron to remediate cyclic nitramines.     

Removal of arsenic from groundwater by zerovalent iron and the role of sulfide

Several recent investigations have shown encouraging potential for the removal of arsenic (As) from groundwater by granular zerovalent iron (Fe0). In contrast to previous studies conducted, we have investigated the applicability of this method and the nature of As bonding under conditions with dissolved sulfide. Three column tests were performed over the period of 1 year using solutions with either As(V) or As(II) (2-200 mg/L) in the input solution. Arsenic outflow concentrations decreased from initially 30-100 microg/L to concentrations of below 1 microg/L with time. XANES (X-ray absorptions near edge structure) and EXAFS (expanded X-ray absorption fine structure) spectra indicated that As in the solid phase is not only directly coordinated with oxygen, as is the case in adsorbed or coprecipitated arsenite and arsenate. Samples with high sulfur content showed additional bonding, for which Fourier transformations of EXAFS data exhibited a peak between 2.2 and 2.4 A. This bonding most likely originated from the direct coordination of sulfur or iron with As, which was incorporated in iron sulfides orfrom adsorbed thioarsenites. The formation of this sulfide bonding supports the removal of As by Fe0 because sulfide production by microbial sulfate reduction is ubiquitous in permeable reactive barriers composed of Fe0.     

An emission pattern of a thermophilic bacteria attached to or imbedded in porous supports

There are many problems with thermophilic bacteria contamination of milk in the dairy industry. This is, in part, a result of fouling by milk components on stainless steel surfaces, which provide good harboring facilities for these bacteria to attach, imbed and grow. The interactions between milk fouling and bacteria deposited in or on the fouling deposit therefore become important issues. There have been a number of previous studies on the biofilm development in dairy processing plants. Here, a different approach to investigate the bacteria emission from a porous layer has been taken. In this approach, various process fluids were flushed over the top of a model milk foulant layer that contains high percentages of milk proteins, fat and some bacteria cells, in order to investigate the behavior of the 'resident' microorganisms and how they are 'released' into the flushing liquids. Definitive results were obtained, which have created sufficient interest for a different approach taken later, where fabric layers were used as the support for the bacteria cells to explore the 'generic' behavior of the porous layer-bacteria system. This study has shown that Bacillus stearothermophilus could multiply on or within a porous layer and 'migrate' from the layer into the fluid during processing. This "migration" is somewhat peculiar in terms of its time-responses but these are reproducible in all the tests performed. The phenomena observed may have an impact on future microbial safety practice in food factories.     

Nitrate reduction by zerovalent iron: effects of formate, oxalate, citrate, chloride, sulfate, borate, and phosphate

Recent studies have shown that zerovalent iron (Fe0) may potentially be used as a chemical medium in permeable reactive barriers (PRBs) for groundwater nitrate remediation; however, the effects of commonly found organic and inorganic ligands in soil and sediments on nitrate reduction by Fe0 have not been well understood. A 25.0 mL nitrate solution of 20.0 mg of N L(-1) (1.43 mM nitrate) was reacted with 1.00 g of Peerless Fe0 at 200 rpm on a rotational shaker at 23 degrees C for up to 120 h in the presence of each of the organic acids (3.0 mM formic, 1.5 mM oxalic, and 1.0 mM citric acids) and inorganic acids (3.0 mM HCl, 1.5 mM H2SO4, 3.0 mM H3BO3, and 1.5 mM H3PO4). These acids provided an initial dissociable H+ concentration of 3.0 mM available for nitrate reduction reactions under conditions of final pH < 9.3. Nitrate reduction rates (pseudo-first-order) increased in the order: H3PO4 < citric acid < H3BO3 < oxalic acid < H2SO4 < formic acid < HCl, ranging from 0.00278 to 0.0913 h(-1), corresponding to surface area normalized rates ranging from 0.126 to 4.15 h(-1) m(-2) mL. Correlation analysis showed a negative linear relationship between the nitrate reduction rates for the ligands and the conditional stability constants for the soluble complexes of the ligands with Fe2+ (R2 = 0.701) or Fe3+ (R2 = 0.918) ions. This sequence of reactivity corresponds also to surface adsorption and complexation of the three organic ligands to iron oxides, which increase in the order formate < oxalate < citrate. The results are also consistent with the sequence of strength of surface complexation of the inorganic ligands to iron oxides, which increases in the order: chloride < sulfate < borate < phosphate. The blockage of reactive sites on the surface of Fe0 and its corrosion products by specific adsorption of the inner-sphere complex forming ligands (oxalate, citrate, sulfate, borate, and phosphate) may be responsible for the decreased nitrate reduction by Fe0 relative to the chloride system.     

TiO2-photocatalyzed As(III) oxidation in a fixed-bed, flow-through reactor

Compliance with the U.S. drinking water standard for arsenic (As) of 10 microg L(-1) is required in January 2006. This will necessitate implementation of treatment technologies for As removal by thousands of water suppliers. Although a variety of such technologies is available, most require preoxidation of As(III) to As(V) for efficient performance. Previous batch studies with illuminated TiO2 slurries have demonstrated that TiO2-photocatalyzed AS(III) oxidation occurs rapidly. This study examined reaction efficiency in a flow-through, fixed-bed reactor that provides a better model for treatment in practice. Glass beads were coated with mixed P25/sol gel TiO2 and employed in an upflow reactor irradiated from above. The reactor residence time, influent As(III) concentration, number of TiO2 coatings on the beads, solution matrix, and light source were varied to characterize this reaction and determine its feasibility for water treatment. Repeated usage of the same beads in multiple experiments or extended use was found to affect effluent As(V) concentrations but not the steady-state effluent As(III) concentration, which suggests that As(III) oxidation at the TiO2 surface undergoes dynamic sorption equilibration. Catalyst poisoning was not observed either from As(V) or from competitively adsorbing anions, although the higher steady-state effluent As(III) concentrations in synthetic groundwater compared to 5 mM NaNO3 indicated that competitive sorbates in the matrix partially hinder the reaction. A reactive transport model with rate constants proportional to incident light at each bead layer fit the experimental data well despite simplifying assumptions. TiO2-photocatalyzed oxidation of As(III) was also effective under natural sunlight. Limitations to the efficiency of As(III) oxidation in the fixed-bed reactor were attributable to constraints of the reactor geometry, which could be overcome by improved design. The fixed-bed TiO2 reactor offers an environmentally benign method for As(III) oxidation.