Published June 25, 2014

Journal of Environmental Quality

TECHNICAL REPORTS Heavy Metals in the Environment

Formation of Manganese Oxide Coatings onto Sand for Adsorption of Trace Metals from Groundwater A. S. Tilak,* S. Ojewole, C. W. Williford,* G. A. Fox, T. M. Sobecki, and S. L. Larson

T

race metals occur naturally in soil at low levels in

Manganese oxide (MnOx(s)) occurs naturally in soil and has a high affinity for trace metals adsorption. In this work, we quantified the factors (pH; flow rate; use of oxidants such as bleach, H2O2, and O3; initial Mn(II) concentrations; and two types of geologic media) affecting MnOx(s) coatings onto Ottawa and aquifer sand using batch and column experiments. The batch experiments consisted of manual and automated titration, and the column experiments mimicked natural MnOx(s) adsorption and oxidation cycles as a strategy for in situ adsorption. A Pb solution of 50 mg L-1 was passed through MnOx(s)–coated sand at a flow rate of 4 mL min-1 to determine its adsorption capacity. Batch experimental results showed that MnOx(s) coatings increased from pH 6 to 8, with maximum MnOx(s) coating occurring at pH 8. Regarding MnOx(s) coatings, bleach and O3 were highly effective compared with H2O2. The Ottawa sand had approximately twice the MnOx(s) coating of aquifer sand. The sequential increase in initial Mn(II) concentrations on both sands resulted in incremental buildup of MnOx(s). The automated procedure enhanced MnOx(s) coatings by 3.5 times compared with manual batch experiments. Column results showed that MnOx(s) coatings were highly dependent on initial Mn(II) and oxidant concentrations, pH, flow rate, number of cycles (h), and the type of geologic media used. Manganese oxide coating exceeded 1700 mg kg-1 for Ottawa sand and 130 mg kg-1 for aquifer sand. The Pb adsorption exceeded 2200 mg kg-1 for the Ottawa sand and 300 mg kg-1 for the aquifer sand.

Earth’s crust and include soil, underlying geologic materials, groundwater, and surface water (Chang and Page, 2000). The concentration of these trace metals can often reach 1000 mg kg-1 in crustal rocks or contaminated soils (Davis et al., 1993). Trace metals such as cadmium (Cd), mercury (Hg), and lead (Pb) are typically found in low concentrations in soil and water; however, industrial, military, and construction practices have contaminated soil and water with excessive trace metals at many sites (Bohn et al., 1979). Examples include Pb-based paints in residential areas, Pb projectiles on military training sites, and byproducts from electroplating facilities and the manufacture of nuclear materials (Bricka et al., 1993). Trace metals of major concern in groundwater include Cd, copper (Cu), Pb, zinc (Zn), and Hg (Bedient et al., 1994). The most common and traditional method to remediate groundwater is the “pump and treat” system. In this method, contaminated water is pumped from the ground and transported to a water treatment plant where traditional water treatment methods are used to clean the contaminated water (NRC, 1997; Wantanaphong et al., 2005). The major disadvantages of this method are the long operational time required for long-term extraction of groundwater and the input of continuous energy (Bayer and Finkel, 2006). Permeable reactive barriers (PRBs) intercept contaminated groundwater while destroying organics, immobilizing metals, and limiting the spread of the contaminants (Khan et al., 2004). By this mechanism, the barriers prevent groundwater contaminants from migrating into uncontaminated aquifers (Roberts et al., 1996; McMahon et al., 1999). Typical PRBs consist of an excavated trench filled with natural and reactive materials. Permeable reactive barriers have several advantages over the pump and treat systems, such as operational, maintenance, and energy costs. No active energy costs or maintenance costs are needed for PRB due to is passive nature of treatment (Yin and Allen, 1999). A report published by the USEPA in 2001 computed the capital and operational costs for 32 pump and treat sites to be $4,900,000, whereas the operational and capital cost of PRB installed at 16 sites was computed to be $730,000 (USEPA, 2001).

Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

A.S. Tilak, Dep. of Geography and Earth Sciences, Univ. of Birmingham, Birmingham, UK; S. Ojewole, Invensys Operations Management, Houston, TX 77041; C.W. Williford, Univ. of Mississippi, Oxford, MS 38677; G.A. Fox, Dep. of Biosystems and Agricultural Engineering, Oklahoma State Univ., 120 Agricultural Hall, Stillwater, OK 74078; T.M. Sobecki, USACE Cold Regions Research and Engineering Lab., 72 Lyme Rd., Hanover, NH 03755-1290; S.L. Larson, U.S. Army Corp of Engineers, Engineering Research and Development Centre, Vicksburg, MS 39180. Assigned to Associate Editor Brett Robinson.

J. Environ. Qual. 42:1743–1751 (2013) doi:10.2134/jeq2013.04.0142 Supplemental data file is available online for this article. Received 20 Apr. 2013. *Corresponding author ([email protected]; [email protected]).

Abbreviations: DI, deionized; PRB, permeable reactive barrier.

1743

Reeter et al. (1999) reported that remediation costs for cleaning up of chlorinated hydrocarbons in groundwater using PRBs was 50 to 70% cheaper when compared with pump and treat systems. Manganese oxides (and/or oxyhydroxides), which include all oxides (MOx), hydroxides [M(OH)x], and oxyhydroxides (MOxOHy), are promising media for PRB as a reactant and adsorbent for heavy metals (Taffarel and Rubio, 2010). Manganese mineralogy is complicated due to its large number of oxides and hydroxides, in which Mn(II) and Mn(III) are extensively substituted for Mn(IV) (McKenzie, 1972). Generally, Mn(II) is found only in igneous rocks, whereas Mn(III) and Mn(IV) are commonly found in soils and sediments (McKenzie, 1972). Metal oxyhydroxides, particularly Fe and Mn, can be viewed as polymeric or polynuclear oxoacids or bases (Stumm and Morgan, 1996), the surface pH of which is a function of the surrounding solution composition and the presence of so-called “potential-determining ions.” Depending on pH, oxidation reduction potential, and ionic concentrations, oxidation of Mn2+ leads to precipitation as oxyhyroxides. Manganese has multiple oxidation states (represented as manganese oxide [MnOx(s)]), and the various oxides exhibit a diversity of atomic architecture that can accommodate cations. These cations participate in oxidation–reduction and cation exchange reactions (Post, 1999). The surface chemistry of MnOx(s) is of environmental importance because of its high adsorption capacity for trace metals (Drever, 1982; Nicholson and Eley, 1997). Manganese oxide is an important scavenger of trace metals in soils, sediments, and rocks despite being less abundant than Fe (O’Reilly and Hochella, 2003). Research conducted by McKenzie (1980) and McBride (1994) showed that the reactivity and high surface areas of Fe and Mn oxides make them very effective for adsorbing trace metals like Cu, Pb, and Zn. Other adsorbents include activated alumina, activated carbon, oxyhydroxides, zerovalent iron, ion exchange resins, peat, coal, clays, and zeolites (Morrison and Spangler, 1993; Yin and Allen, 1999). There are extensive studies in the literature on the removal of soluble Mn by MnOx(s)–coated media from water treatment plants (Maliyekkal et al., 2006; Kim and Jung, 2008; Knocke et al., 2010; Islam et al., 2010). Maliyekkal et al. (2006), through batch and column experiments, investigated the adsorption potential of MnOx(s)–coated alumina or defluoridation of drinking water. Results showed that optimum removal of fluoride ions occurred in the pH range 4 to 7. They also showed that the fluoride adsorption rate of MnOx(s)–coated alumina was greater than activated alumina. Kim and Jung (2008) used column experiments to demonstrate that soluble manganese was effectively removed by MnOx(s)–coated sand as compared with sand + MnOx(s)–coated sand (1:1) and granular activated carbon. Knocke et al. (2010) and Islam et al. (2010) performed batch experiments for soluble manganese removal from water treatment plants by MnOx(s)–coated media. Their results showed that initial manganese concentrations and pH were important experimental parameters for soluble manganese removal by MnOx(s)–coated media. Thus, the above studies demonstrate that pH is an important parameter for the removal of soluble Mn(II) by MnOx(s)–coated media. The study by Hargette and Knocke (2001) demonstrated the regeneration of MnOx(s) coatings on a given media on its depletion. However, there is a greater need to demonstrate the regeneration of MnOx(s) coatings onto a natural 1744

soil media using batch experiments. This process could be useful for regenerating the MnOx(s) coating of a PRB installed in the subsurface for attenuation of heavy metals. Also, none of the above-mentioned studies demonstrated the effect of different oxidants such as bleach, O3, and H2O2 on the formation of MnOx(s) coating onto natural soil media. The oxidant having maximum MnOx(s) coating onto natural soil media can be used for larger commercial and industrial purposes. This experimental study reports the results that were achieved through batch and column experimental procedures that differed from the above studies. The batch experiments (manual and automated) were performed to quantify the parameters affecting MnOx(s) coatings onto natural soil media (Ottawa and aquifer sand) and to determine (i) the effect of pH (6–8) on MnOx(s) coating and the pH having maximum MnOx(s) coating, (ii) the effect of geological media (Ottawa and aquifer sand), (iii) the effect of the use of oxidants (bleach, O3, and H2O2), (iv) the incremental increase in aqueous Mn(II) concentrations onto the same sand sample (regeneration of MnOx(s) coatings), and (v) the most effective coating strategy. Manganese oxide is commonly found in nature at the interface between the lithosphere and the hydrosphere. A relevant example is the formation of MnOx(s) coatings at the interface between soil particles and soil solution in alternately flooded and dried soils. The flooded/reducing conditions allow Mn transport and adsorption onto the soils, whereas the dry/oxidizing conditions allow oxide formation. The experimental approach described in this paper attempted to mimic nature by having a small-scale sand-bed vertical column (i.e., a continuous reactor) to quantify key parameters affecting MnOx(s) coating onto natural soil media: (i) initial Mn(II) and oxidant bleach concentrations, (ii) solution cycle duration (hours), (iii) MnOx(s)–coated material, (iv) the initial amounts of Ottawa sand, and (v) changes in hydraulic conductivity. Quantification of these key parameters from batch and column experiments will be useful for the preparation of MnOx(s)–coated media on a larger commercial scale.

Materials and Methods For the batch and column experiments, the MnOx(s) coatings onto Ottawa and aquifer sand were produced using three oxidants: O3 produced by an O3 generator manufactured by Pacific Ozone Technology, bleach (5.25% commercial grade sodium hypochlorite), and H2O2 (50% w/w in water) from Sigma Aldrich chemicals. A sodium hydroxide (NaOH) solution was prepared by diluting a 10 mol L-1 solution purchased from Fisher chemicals. For monitoring the pH, an Orion pH meter was used. We worked with two geological media. The first was Ottawa sand obtained from the U.S. Silica Company. This quartz sand, widely used in experiments, typically exhibits rounded grains, is white, and has little natural deposition of metal oxides, important for these studies. This particular sand was identified by the company as 140 grain fineness, referring to the fraction retained on a 140-mesh screen. Grain size distribution for our sand sample was determined at the U.S. Silica Company lab in Berkeley, West Virginia. The second media was excavated from a near-surface deposit at Goodwin Creek near Batesville, Mississippi, and characterized in the Civil Engineering soils lab at University of Mississippi. It is a sandy, permeable media, representative of a shallow groundwater aquifer (Table 1). Journal of Environmental Quality

Batch Experimental Procedure

Table 1. Properties for Ottawa and Aquifer sands.

Alternatives for MnOx(s), coating as reviewed by Feng et al. (2000), included redox precipitation (for birnessite), sol-gel processes, and hydrothermal processes. For this experimental work, we used ambient temperature methods more in line with those of Merkle et al. (1997). The interface of aqueous Mn(II) with solid phases of MnOx(s) reveals thermodynamically favorable pathways for a redox precipitation. The interface between aqueous Mn(II) and g-MnOOH(s) lies approximately between pE 5 to 12 and pH 6 to 9, bracketing effective experimental conditions. For the batch experimental work, the Ottawa and aquifer sands were rinsed with deionized (DI) water to remove organic debris and oxygen-containing air from the void space. Deionized water was added to bring the total volume to 600 mL. Two burettes were used: one was filled with 0.2 mol L-1 NaOH, and the other was filled with 5.25% bleach or 10% H2O2. A mixer agitated the slurry at 300 to 400 rpm. Nitrogen gas was purged into the slurry through a pipette tip to further remove dissolved oxygen from the mixture. Manganese (II), previously dissolved in approximately 100 mL deionized water, was added at 1.625 or 16.25 g (500 and 5000 mg Mn, respectively) to give ~1000 or ~10,000 mg L-1 Mn(II) at an initial solution volume of ~500 mL. Sodium hydroxide was added from its burette to the reactor to adjust the pH to the specified value. The oxidant bleach or H2O2 flow was manually controlled. An Orion pH meter was used to monitor pH. Oxidant addition and subsequent reaction liberated protons, suppressing pH, but the addition of NaOH maintained the specified pH. After approximately 5 min, the media began to darken for Ottawa sand from white to light brown. The flow from both the burettes required constant attention and adjustment to continue Mn(II) oxidation and maintain pH. The average running time for batch experiments was approximately 60 min. The amounts of NaOH and oxidant solution added were approximately 75 and 20 mL, respectively. The average flow rates for NaOH and the oxidant were 0.75 and 0.35 mL min-1, respectively. The cessation of pH change with oxidant addition indicated that the Mn(II) had been oxidized. The reaction was terminated, and final pH was recorded. The MnOx(s)–coated sand was washed three times with DI water, and the MnOx colloidal material was decanted from the sand. Repetitive coating experiments were performed in three stages on a given Ottawa sand sample using all three oxidants and at pH 8. The procedure of repetitive coating was the same as before, except that all experiments used a lower initial Mn(II) concentration of ~1000 mg L-1. Once a coating stage was completed, the sand was washed to remove colloidal MnOx, and a small sample was withdrawn. The remaining sand was returned to the reactor, and the batch experimental procedure was repeated. After every experiment, a sand sample was withdrawn to determine MnOx(s) coating.

Property/media

Automated Titrator Procedure For these experiments, a 250-mL plastic cup reactor containing 50 g of Ottawa sand was used. Two dispensing 10-mL burettes were used to pump Mn(II) and NaOH solutions from their reservoirs (amber reagent bottles) to their respective burettes and then to the plastic reactor cup. A pH electrode from Orion monitored the entire process. Only two oxidants were www.agronomy.org • www.crops.org • www.soils.org

UCS† classification Percent sand

Ottawa sand

Aquifer sand

poorly graded sand 97.6

sand 99.4

0.08 7, rather than slightly acidic pH SOH), which bind Mn(II) and promote rapid oxidation (Davies, 1986; Wehrli et al., 1989; Wehrli, 1990; Wehrli et al., 1995). Soluble Mn(II) removal occurs in presence or absence of an oxidant. In the absence of an oxidant, sorption capacity increases with increasing alkaline pH or surface MnOx concentration or both; in the presence of an oxidant, Mn(II) removal is due to sorption alone (Knocke et al., 1991). The addition of free chlorine in combination with oxide-coated filter media results in the efficient removal of soluble Mn(II) above pH 6 (Knocke et al., 1988; Knocke et al., 1991). The MnOx oxidation with free chlorine was modeled first by Nakanishi (1967) and shown by Coffey et al. (1993) in which the hypochlorite exists in associated or unassociated form, depending on pH. Mn2+ + MnO(OH)2(s) « MnO2MnO(s) + 2H+ [2] MnO2MnO(s) + HOCl/OCl- ® 2MnO2(s) + H+ + Cl– [3]

Using this relationship, they modeled the removal of manganese from solution and of MnOx(s) coating onto filter media. Merkle et al. (1997) expressed the overall reaction as: Mn2+ + MnO(OH)2(s) + HOCl + H2O ® [4] MnO2 + MnO(OH)2(s) + 3H+ + Cl– [5]

Reactions [4] and [5] imply that one site is regenerated as MnO(OH)2(s) with one MnO2 incorporated into the oxide phase. The sequence of reactions requires bidentate complexation with two hydroxyl groups per adsorption site (Merkle et al., 1997). The repetitive coating experiments were performed in three stages with Ottawa and aquifer sand and initial Mn(II) concentration of ~1000 mg L-1 (Table 3). These experiments were conducted at pH 8 based on the above experimental results. For the Ottawa sand and the three oxidants, and aquifer sand and oxidant bleach the MnOx(s) coatings increased with increasing stage (from stage 1 to stage 3). The highest MnOx(s) coatings (3956 mg kg-1) occurred with Ottawa sand and oxidant bleach (Table 3). This result was comparable to that obtained by Merkle et al. (1997), who repeatedly mixed gravel with alternate solutions of Mn(II) and bleach and achieved a coating of 4000 mg Mn kg-1. Overall, the oxidant bleach had the highest MnOx(s) coating onto Ottawa sand at each stage compared with O3 and H2O2. The repetitive experimental results are consistent with a model of adsorption and oxidation. A surface with greater initial MnOx(s) www.agronomy.org • www.crops.org • www.soils.org

Table 3. Manganese oxide coating using repetitive method and Ottawa sand with three oxidants at pH 8 and aquifer sand using the oxidant bleach at pH 8 at an initial Mn(II) concentration of ~1000 mg L-1 [500 mg Mn(II)] for each stage of coating. Type of geologic media Ottawa sand

Type of oxidant bleach

H2O2

O3

Aquifer sand

bleach

Experimental stages

MnOx(s) coating

1 2 3 1 2 3 1 2 3 1 2 3

mg kg-1 842 2677 3956 309 434 701 310 1564 2099 215 339 459

had greater capacity to adsorb and oxidize Mn(II) in subsequent stages. The repetitive coatings achieved more efficient use of Mn. Each of the three stages began with ~1000 mg L-1 (500 mg Mn), versus the higher initial Mn(II) concentration of ~10,000 mg L-1 (5000 mg Mn). Thus, the three-stage repetitive coatings only used 30% of the Mn, compared with single-stage coatings at an initial Mn(II) concentration of ~10,000 mg L-1. For the threestage process, a Mn conversion of 67% (to coating) was achieved, which was higher than the single-stage batch experiments. The automated titrator results shown in Table 4 for each pH (6–8) are the average of three samples (Supplemental Table S1). The highest MnOx(s) coating (5512 mg kg-1) was achieved using Ottawa sand and oxidant bleach at pH 8, whereas the highest MnOx(s) coating for aquifer sand was also achieved at pH 8 using oxidant bleach. It was clear from these experiments that pH 8 had the highest MnOx(s) coating and that oxidant bleach outperformed H2O2 at all pH values from 6 to 8 (Table 4). Lee et al. (2009) studied the removal of aqueous Mn(II) by two different manganesecoated samples over a wide pH range (5–10.5) and in the absence and presence of sodium hypochlorite (i.e., bleach). They observed that Mn(II) removal from aqueous solution was pH dependent and increased with increasing pH from 5 to 10.5 in the presence of sodium hypochlorite but not in its absence. These results showed that oxidant bleach can be a very effective oxidant for commercial and industrial use in terms of coating MnOx(s) onto a media and for removal of Mn(II) from aqueous solution. Also, the experiments with bleach produced MnOx(s) coating onto Ottawa sand at a rate 3.5 times greater than with manual batch experiments. Titrator experiments with H2O2 produced consistent MnOx(s) coatings that changed less with pH using Ottawa and aquifer sand compared with the manual batch experiments. For the column experiments, the MnOx(s) coating for Ottawa sand (24- to 96-h cycles) and aquifer sand (24- and 48-h cycles) is shown in Table 5. Visual images of the column experiments for Ottawa sand showed increased MnOx(s) coatings and darkening after 48 h compared with 24 and 0 h (Fig. 2). At a small scale, the photo and photomicrographs in Fig. 3a, 3b, and 3c show the appearance with varying levels of MnOx(s) coating onto the surface of the glass bead packing and sand grains. The monitored pH for these experimental cycles ranged from 6 to 9. For Ottawa sand, 1747

the surface lacks adsorption capacity. As the MnOx(s) coating gradually builds, the adsorption capacity and the rate of MnOx(s) coating per cycle increases. Finally, as the adsorption capacity stabilizes, the rate of the MnOx(s) coating per cycle moderates, producing an “S”-shaped curve, as seen in Fig. 4. The “S”-shaped curve obtained in this study was similar to other studies of MnOx(s) coatings onto media (Stumm and Morgan, 1996; Christophi and Axe, 2000). The “S”-shaped curve was due the net surface charge becoming more negative as the pH became alkaline. This resulted in increased adsorption capacity at higher alkaline pH (Stumm and Morgan, 1996; Christophi and Axe, 2000). The experimental results shown in Table 6 involved changing initial Mn(II), bleach concentrations, and flow rates to determine MnOx(s) coating onto Ottawa sand for a 24-h cycle. The results showed lower MnOx(s) coating irrespective of flow rate (4 and 40 mL min-1) due to the lower availability of oxidant bleach (Table 6). However, there was greater uniform distribution of MnOx(s) coating throughout the Ottawa sand column. The MnOx(s) coatings increased when initial bleach concentrations increased from 20 to 100 mg L-1 because greater Mn(II) was available for conversion to oxide coating (Table 6). Another set of experiments involved changing the initial amounts of Ottawa sand to quantify MnOx(s) coating (Table 7). The results showed that MnOx(s) coating for 40 g of Ottawa sand averaged 489 mg kg-1, whereas MnOx coating for 75 g of Ottawa sand averaged 350 mg kg-1. From these results, it can be concluded that there was little difference between MnOx(s) coating for 40 and 75 g of Ottawa sand. Also, there was uniform MnOx(s) distribution throughout the length of the column. For example, the differences in MnOx(s) coating between the inlet (bottom, section I) and the outlet (top of the column, section IV) for 40 and 75 g of coated Ottawa sand were 26 and 25%, respectively (Table 7). These similar reductions implied that flow rate, initial Mn(II) and bleach concentrations, and type of geological media

Table 4. Manganese oxide coating onto Ottawa and aquifer sand vs. pH (6–8) for the oxidants bleach and hydrogen peroxide using an automated titrator. Type of geologic media Ottawa sand

Type of oxidant bleach

H2O2

Aquifer sand

bleach

H2O2

pH

MnOx(s) coating

6 7 8 6 7 8 6 7 8 6 7 8

mg kg-1 2799 4016 5512 793 1115 1322 1783 2321 3540 1233 1239 1303

the MnOx(s) coatings increased with increasing cycles from 24 to 96 h, with maximum MnOx(s) coating occurring at the 96-h cycle (Table 5). For the aquifer sand, the MnOx(s) coating was greater for the 48-h cycle than for the 24-h cycle (Table 5). We observed a darkening area at the inlet (bottom of the column) compared with the outlet (toward the top of the column). To quantify this, the sand columns were sectioned into halves for the Ottawa sand (Table 5) and into fourths for the aquifer sand (Table 5), and the MnOx(s) content was determined. Our results showed the highest MnOx(s) coatings at the inlet of the column compared with the outlet for both sands. Manganese oxide coating at the inlet of the column (bottom, section I) was 1.5 to 2.6 times greater than at the outlet of column (top of the column, section IV). Although increasing MnOx(s) coating with the cycles was clear, a more subtle trend appears where the initial coating per cycle is slow because

Table 5. Average manganese oxide coating onto Ottawa and aquifer sand (column experiments).† Type of geologic media Ottawa sand

No of experimental cycles per h

MnOx(s)–coated sand sections

24

I II I II I II I II III IV I II III IV I II III IV

48 72 96

Aquifer sand

24

48

MnOx(s) coating for each soil section

Average MnOx(s) coating

—————————— mg kg-1 —————————— 170 142 113 468 359 250 1558 1074 590 1708 1279 850 471 312 98 60 63 49 28 133 82 93 62 41

† For the Ottawa sand cycles, soil was divided into two sections (section I: inlet of the column; section II: above the column). For the aquifer sand cycles, soil was divided into four sections (section I: inlet of the column; section II: above section I; section III: above section II; and section IV: outlet of the column). 1748

Journal of Environmental Quality

Fig. 2. Appearance of Ottawa sand after 0-, 24-, and 48-h cycles of four fluids: deionized water, manganese nitrate [100 mg L−1 Mn(II)], deionized water, and bleach (135 mg L−1 hypochlorite).

Fig. 4. Manganese oxide coatings onto Ottawa sand versus number of coating cycles (h).

Likewise, the rapid rise in voltage by the Pb electrode signaled the breakthrough of Pb ions, which occurred as the MnOx(s)–coated sand column became saturated with Pb. Tables 8 and 9 summarize the breakthrough test conditions, times in hours, and average Pb adsorption onto Ottawa and aquifer sand. Lead adsorption onto MnOx(s)–coated Ottawa sand for the 96-h cycle was higher (2226 mg kg-1) than for the other cycles because there was higher MnOx(s) coating for the 96-h cycle. Also, the Pb adsorption for aquifer sand was higher at the 48-h cycle (320 mg kg-1) because the greatest MnOx(s) coating occurred for the 48-h cycle. Lead adsorption for all cycles of Ottawa sand was greater than for aquifer sand. The hydraulic conductivity declined from uncoated sand (i.e., 0-h cycle to 24-h cycle and then through 48, 72, and 96 h) but at a rapid rate for the first two cycles (Fig. 5). For the Ottawa sand, the hydraulic conductivity decreased by 67% (i.e., the difference between the 0-h cycle and the 96-h cycle). For the aquifer sand, the hydraulic conductivity decreased by 32% (i.e., the difference between the 0-h cycle and the 48-h cycle).

Conclusions

Fig. 3. (a) Photo of glass beads with varying amounts of MnOx coating. (b) Photomicrograph of uncoated Ottawa sand. (c) Photomicrograph of coated Ottawa sand.

had a more pronounced effect on MnOx(s) coating than the initial amount of Ottawa sand in the column.

Lead Adsorption Capacity and Hydraulic Conductivity By tracking the response of the NO3- ions, the residence time of the system was determined to be approximately 20 min. www.agronomy.org • www.crops.org • www.soils.org

Batch experiments showed that pH was an important factor and that MnOx(s) coating increased with increasing pH from 6 to 8, with the greatest MnOx(s) coating at pH 8. Regarding the geological media, the Ottawa sand had greater MnOx(s) coating compared with aquifer sand. Oxidant bleach was highly effective compared with oxidant H2O2 in terms of MnOx(s) coating at all experimental conditions. Bleach provided greater buffering capacity, lower colloidal phase, and greater oxide coating compared with H2O2. This shows that bleach can be commercially used as an oxidant for MnOx(s) coating onto media. Incremental increases in initial Mn(II) concentrations onto the same soil sample thrice (repetitive method) achieved the highest MnOx(s) coatings compared with single-batch experiments and demonstrated the regenerative process of oxidizing and increasing MnOx(s) coatings at every incremental process, which could be potentially used in regenerating MnOx(s) coating of a PRB. This process also resulted in lower colloidal production and greater MnOx(s) coating, which, for environmental reasons, is particularly desirable for field applications. The automated titrator improved the productivity and reproducibility of MnOx(s) coatings, which is important for scientific investigation and industrial 1749

Table 6. Effect of changing flow rates and initial Mn(II) and bleach concentrations on manganese oxide coatings onto Ottawa sand for a 24-h cycle (column experiments). Type of geologic media Ottawa sand

Flow rate

Initial Mn(II) and oxidant bleach concentrations

mL min-1 4

mg L-1 20 and 26

40

20 and 100

40

100 and 132

MnOx(s)–coated sand ut sections

MnOx(s) coating on each soil section

Average MnOx(s) coating on soil sections

——————— mg kg-1 ——————— 26 24 26 23 22 256 221 218 211 199 922 605 610 483 404

I II III IV I II III IV I II III IV

Table 7. Effect of changing initial amounts of Ottawa sand on manganese oxide coatings for the 24-h cycle (column experiments). Type of geologic media Ottawa sand

Flow rate

Initial Mn(II) and oxidant bleach concentrations

Amount of Ottawa sand

mL min-1 40

mg L-1 100 and 132

g 40

40

100 and 132

75

production. It achieved better MnOx(s) coating (~3.5 times that of manual), reinforcing the importance of stable oxidant/alkaline reagent addition, and better pH control, which is important to maintaining the adsorption/oxidation reaction regime. The technique of alternately flowing Mn(II) and oxidant bleach solutions in column experiments showed that oxide coating slowly develops once some initial MnOx(s) is present. The surface then acts as a regenerative surface, with more adsorption of Mn(II) followed by its oxidation with bleach and consequently more MnOx(s) coating. The rate increases until the coating supports a steady state or limiting rate of adsorption. Thus, these column and batch experiments showed that parameters such a

MnOx(s) coated sand sections

MnOx(s) coatin on each soil section

Average MnOx(s) coating on soil sections

—————— mg kg-1 —————— 574 489 514 445 424 417 350 345 326 313

I II III IV I II III IV

pH, type and concentration of oxidant, type of geologic media, initial Mn(II) concentration, repetitive experiments, and column in situ strategy could be used for producing MnOx(s) coating onto media at a commercial and industrial scale.

Acknowledgments This work was supported by the U.S. Army Corp of Engineers, Engineering Research and Development Center (ERDC) at the Water Ways Experiment Station (WES) Vicksburg, MS. The authors thank Dr. June Mirecki and Dr. Tony Bednar at WES for access to the automated titrator and analytical support; Mr. David Burkhart of the MettlerToledo, maker of the automated titration system, for valuable assistance through instruction, equipment set up, and customization of the LabX

Table 8. Conditions and results for lead absorption onto Ottawa sand (column experiments). Cycles

Flow rate

pH

Breakthrough time

Initial Pb concentration

Average Pb adsorbed on MnOx(s) coated sand

24 h 48 h 72 h 96 h

mL min-1 4 4 4 4

6–8.9 5–7 5–8.2 4.5–8

h 4.5 4.7 7.5 18.5

mg L-1 50 50 50 50

mg kg-1 500 560 900 2226

Table 9. Conditions and results for lead absorption onto aquifer sand (column experiments). Cycles

Flow rate

24 h 48 h

mL min-1 4 4

1750

pH

Breakthrough time

Initial Pb concentration

Average Pb adsorbed onto MnOx(s) coated sand

5–7 4.5–7

h 1.5 2.7

mg L-1 50 50

mg kg-1 180 320 Journal of Environmental Quality

Fig. 5. Hydraulic conductivity measurements (cm h−1) versus MnOx(s) coatings onto Ottawa and aquifer sand cycles per hour.

software algorithm; and Dr. Ashraf Soltan of the Military Technical College of Cairo, Egypt for reviewing the manuscript.

References Bayer, P., and M. Finkel. 2006. Life cycle assessment of active and passive groundwater remediation technologies. J. Contam. Hydrol. 83(3–4):171–199. doi:10.1016/j.jconhyd.2005.11.005 Bedient, P.B., H.S. Rifai, and C.J. Newell. 1994. Groundwater contamination. Prentice Hall, Englewood Cliffs, NJ. Bohn, H.L., B.L. McNeal, and G.A. O’Connor. 1979. Soil chemistry. John Wiley & Sons, New York. Bricka, R.M., C.W. Williford, and L. Jones. 1993. Technology assessment of currently available and development techniques for heavy metals contaminated soil treatment, Technical Report IRRP-93–4, U.S. Army Corp of Engineers, Waterways Experiment Station, Vicksburg, MS. http://el.erdc.usace.army.mil/ elpubs/pdf/trirrp93-4.pdf (accessed 28 Nov. 2010). Chang, A.C., and A.L. Page. 2000. Trace elements slowly accumulating, depleting in soils. Calif. Agric. 54(2):49–55. doi:10.3733/ca.v054n02p49 Coffey, B.M., D.L. Gallager, and W.R. Knocke. 1993. Modeling soluble manganese removal by oxide coated filter media. J. Environ. Eng. 119(4):679–694. doi:10.1061/(ASCE)0733-9372(1993)119:4(679) Christophi, A.C., and L. Axe. 2000. Competition of Cd, Cu and Pb adsorption on goethite. J. Environ. Eng. 126(1):66–74. doi:10.1061/ (ASCE)0733-9372(2000)126:1(66) Davies, S.H.R. 1986. Mn (II) oxidation in the presence of lepidocrcite: The influence of other ions. In: J.A. Davies and K.F. Hayes, editors, Geochemical processes at mineral surfaces. Am. Chem. Soc., Washington, DC. p. 487–502. Davis, J.A., C.C. Fuller, J.A. Coston, K.M. Hess, and E. Dixon. 1993. Spatial heterogeneity of geochemical and hydrologic parameters affecting metal transport in groundwater, USEPA Environmental Research Brief EPA/600/S-93/006. http://nepis.epa.gov/Adobe/PDF/2000CAY2.PDF (accessed 21 July 2011). Drever, J.I. 1982. The geochemistry of natural water. Prentice Hall, Englewood Cliffs, NJ. Feng, Q., L. Liu, and K. Yanagisawa. 2000. Effects of synthesis parameters on the formation of birnessite-type manganese oxides. J. Mater. Sci. Lett. 19(17):1567– 1570. doi:10.1023/A:1006733308073 Hargette, A.C., and W.R. Knocke. 2001. Assessment of fate of manganese in oxide coated filtration systems. J. Environ. Eng. 127(12):1132–1138. doi:10.1061/ (ASCE)0733-9372(2001)127:12(1132) Henry, M., J.P. Jolivet, and J. Livage. 1992. Aqueous chemistry of metal cations: Hydrolysis, condensation and complexation. Struct. Bonding 77:153–206. doi:10.1007/BFb0036968 Islam, A.A., J.E. Goodwill, R. Bouchard, J.E. Tobaison, and W.R. Knocke. 2010. Characterization of filter media MnOx(S) surfaces and Mn removal capability. J. Am. Water Resour. Assoc. 102(9):71–83. Kim, J., and S. Jung. 2008. Soluble manganese removal by porous media filtration. Environ. Technol. 29(12):1265–1273. doi:10.1080/09593330802306139 Khan, F.I., T. Husain, and R. Hejazi. 2004. An overview and analysis of site remediation technologies. J. Environ. Manage. 71:95–122. doi:10.1016/j. jenvman.2004.02.003 Knocke, W.R., J.R. Hamon, and C.P. Thompson. 1988. Soluble manganese removal on oxide coated filter media. J. Am. Water Works Assoc. 80(12):65–70. Knocke, W.R., S. Occiano, and R. Hungate. 1990. Removal of soluble manganese from water by oxide coated filter media. AWWA Research Foundation and American Water Works Association, Denver, CO.

www.agronomy.org • www.crops.org • www.soils.org

Knocke, W.R., S.C. Occiano, and R. Hungate. 1991. Removal of soluble manganese by oxide coated filter media: Sorption rate and removal mechanisms issues. J. Am. Water Works Assoc. 83(8):64–69. Knocke, W.R., L. Zuravnsky, J.C. Little, and J.E. Tobaison. 2010. Adsorptive contractors for removal of soluble manganese during drinking water treatment. J. Am. Water Resour. Assoc. 102(8):64–75. Lee, S.M., D. Tiwari, K.-M. Choi, J.-K. Yang, Y.-Y. Chang, and H.-D. Lee. 2009. Removal of Mn(II) from aqueous solutions using manganese-coated sand samples. J. Chem. Eng. Data 54(6):1823–1828. doi:10.1021/je800854s Liu, D., J.J. Sansalone, and F.K. Cartledge. 2004. Adsorption characteristics of oxide coated buoyant media (rs < 1.0) for storm water treatment. I: Equilibria and kinetic models. J. Environ. Eng. 130(4):374–382. doi:10.1061/ (ASCE)0733-9372(2004)130:4(383) Maliyekkal, S.M., A.K. Sharma, and L. Philip. 2006. Manganse-oxide-coated alumina: A promising solvent for defluoridation of water. Water Res. 40:3497–3506. doi:10.1016/j.watres.2006.08.007 Martin, S.T. 2003. Precipitation and dissolution of iron and manganese oxides. In: V.H. Grassian, editor, Environmental catalysis. Imperial College Press, London. p. 61–81. McBride, M.B. 1994. Environmental chemistry of soils. Oxford Univ. Press, New York.. McKenzie, R.M. 1972. The manganese oxides in soils. Z. Pflanzenernaehr. Bodenkd. 131:221–242. doi:10.1002/jpln.19721310302 McKenzie, R.M. 1980. The adsorption of lead and other heavy metals on oxides of manganese and iron. Aust. J. Soil Res. 18(1):61–73. doi:10.1071/SR9800061 McMahon, P.B., K.F. Dennehy, and M.W. Sandstrom. 1999. Hydraulic and geochemical performance of permeable reactive barrier containing zero-valent iron. Ground Water 37(3):396–404. doi:10.1111/j.1745-6584.1999.tb01117.x Merkle, P.B., W.R. Knocke, and D.L. Gallager. 1997. Method for coating filter media with synthetic manganese oxide. J. Environ. Eng. 123(7):642–649. doi:10.1061/(ASCE)0733-9372(1997)123:7(642) Morrison, S.J., and R.R. Spangler. 1993. Chemical barriers for controlling groundwater contamination. Environ. Progress 12:175–181. doi:10.1002/ep.670120305 Nakanishi, H. 1967. Kinetics of continuous removal of manganese in a MnO2– coated sand bed. Kogyo Kagaku Zasshi 70(4):407. doi:10.1246/ nikkashi1898.70.4_407 Nicholson, K., and M. Eley. 1997. Manganese oxide geochemistry: Metal adsorption in freshwater and marine environments. In: K. Nicholson, J.R. Hein, B. Buehn, and S. Dasgupta, editors, Manganese mineralization: Geochemistry and mineralogy of terrestrial and marine deposits. Geological Soc., London. p. 309–326. National Research Council. 1997. Innovations in groundwater and soil clean up. National Academy Press, Washington, DC. O’Reilly, S.E., and M.F. Hochella, Jr. 2003. Lead adsorption efficiencies of natural and synthetic Fe and Mn oxides. Geochim. Cosmochim. Acta 67(23):4471–4487 . doi:10.1016/S0016-7037(03)00413-7 Post, J.E. 1999. Manganese oxide minerals: Crystal structures and economic and environmental significance. Proc. Natl. Acad. Sci. USA 96:3447–3454. doi:10.1073/pnas.96.7.3447 Reeter, C., S. Chao, and A. Gavaskar. 1999. Permeable reactive wall remediation

of chlorinated hydrocarbons in groundwater. U.S. Department of Defense, Environmental Security Technology Certification Program, Cost and Performance Rep. U.S. Department of Defense, Washington, DC.

Roberts, A.L., L.A. Totten, W.A. Arnold, D.R. Burris, and T.J. Campbell. 1996. Reductive elimination of chlorinated ethylene by zero valent metal ions. Environ. Sci. Technol. 30(8):2654–2659. doi:10.1021/es9509644 Stumm, W., and J.J. Morgan. 1996. Aquatic chemistry. Chemical equilibria and rates in natural waters. 3rd ed. John Wiley & Sons, New York. Taffarel, S.R., and J. Rubio. 2010. Removal of Mn2+ from aqueous solution by manganese oxide coated zeolite. Minerals Engineering. 23:1131–1138. doi:10.1016/j.mineng.2010.07.007 USEPA. 2001. Cost analysis for selected groundwater cleanup projects: Pump and treat systems and permeable reactive barriers. EPA-R-00-013. http://epa.gov/ tio/download/remed/542r00013.pdf (accessed 30 Aug. 2011). Wantanaphong, J., S.J., Mooney, and E.H. Bailey. 2005. Natural and waste materials as metal sorbents in permeable reactive barriers (PRBs). Environ. Chem. Lett. 3:19–23. doi:10.1007/s10311-005-0106-y Wehrli, B., B. Sulzberger, and W. Stumm. 1989. Redox processes catalyzed by hydrous oxide surface. Chem. Geol. 78:167–179. doi:10.1016/0009-2541(89)90056-9 Wehrli, B. 1990. Redox reactions of metal ions at mineral surfaces. In: W. Stumm, editor, Aquatic chemical kinetics: Reaction rate processes in natural waters. John Wiley & Sons, New York. p. 311–336. Wehrli, B., G. Friedl, and M. Alain. 1995. Reaction rates and products of manganese oxidation at the sediment-water interface. In: C.P. Huang, C.R. O’Melia, and J.J. Morgna, editors, Aquatic chemistry: Interfacial and interspecies processes. Am. Chem. Soc., Washington, DC. p. 111–134. Yin, Y., and H.E. Allen. 1999. In situ chemical treatment. Groundwater Remediation Technologies Analysis Center technology evaluation report, TE-99–01. Pittsburgh, PA. http://clu-in.org/download/toolkit/inchem.pdf (accessed 24 Aug. 2011).

1751