WO1999058239A1 - Contaminant adsorption and oxidation via the fenton reaction - Google Patents

Abstract

Contaminated water is treated by adsorbing contaminant onto a sorbent to concentrate the contaminant and then oxidizing the contaminant via the Fenton and related reactions. Iron is attached to the sorbent or can be added in solution with an oxidant. Both systems, iron attached to the sorbent or iron in solution, can be used to oxidize contaminants on or near the surface of the sorbent. The process can be used to treat contaminated water in above-ground and below-ground treatment systems.

Description

CONTAMINANT ADSORPTION AND OXIDATION VIA THE FENTON REACTION

Field of the Invention

The present invention is directed to a process for treating contaminated water using a combination of adsorption and oxidation.

Background of the Invention There are many limitations in ground water cleanup programs, which limiting factors are considered in the lifetime costs of implementing remedial actions (O'Brien et al, 1997) . For this reason, permeable reactive barriers and funnel and gate systems are currently the most cost effective methods for cleaning ground water.

There are currently several chemical oxidation systems in which soil or aquifer material has been remediated using hydrogen peroxide in the Fenton mechanism. These systems include injecting hydrogen peroxide in si tu, surface soil application and soil slurry reactors. Each of these systems has potential limitations which ultimately affect the feasibility of the system in treating groundwater.

Blowes et al, in U.S. Patents 5,362,394 and 5,514,279, disclose treating contaminated water by excavating a trench in the aquifer in the path of a contaminant plume, and placing a body of active material which causes the contaminant, by chemical reaction, to change its oxidation- reduction state and to precipitate harmlessly in the body of the material . This process merely involves flowing waste through the active material without concentrating the contaminant . Leachate generation is a potential limitation in surface soil application of hydrogen peroxide, resulting in the downward transport of contaminants. In soil slurry reactors, the treatment volume of contaminated soil is generally small, representing a limitation to the overall treatment process.

Competition kinetics can significantly reduce treatment efficiency and minimize effectiveness when scavengers react with hydroxyl radicals more rapidly than the target compound, as shown in Table 1. Scavenging can be minimized by using low ionic strength or low hardness make-up water for the hydrogen peroxide solution.

Table 1

Chemical Reactions Involving H₂0₂, Iron, 4-POBN,

2CP and Scavengers

H₂0₂ + Fe(II) → Fe(III) + OH^" + -OH (1)

H₂0₂ + Fe(III) → Fe (II) + ^• 0₂ ^" + 2H⁺ (2)

4-P0BN + -OH → -4-POBN (3)

2CP + ^• OH → reaction products (4)

^•OH + Σⁿ ₁_₁S₁ → products of scavenging reactions (5)

^•0₂- + Fe(III) → Fe(II) + 0₂ (6)

H₂0₂ + 2Mn(II) + 2H₂0 → 2Mn00H(s) + 4H⁺ (7)

H₂0₂ + 2Mn00H(s) + 4H⁺ → 2Mn(II) + 0₂ + 4 H₂0 (8) catalase

H₂0₂ → 0₂ + 4H₂0 (9) where • OH hydroxyl radical

• 0₂ ^" superoxide radical

4-POBN spin-trap compound

•4-POBN radical adduct

2CP 2-chlorophenol

S concentration of individual scavengers _x second-order rate constant (L/mol-s) for ^•OH with S_x

Reaction Reaction Rate Constant and General Comments

1 k, = 53.01 L/mol-s (Ingles, 1972), 76 L/mol-s

(Walling, 1975)

2 Rate constant not reported; reaction involves soluble and solid phase iron

3 k₃ = 3.8 x 10⁹ L/mol-s, pH 7 (Buxton et al, 1988)

4 k₄ = 1.2 x 10¹⁰ L/mol-s (Getoff and Solar, 1986)

5 ∑ ik,_. [S - pseudo-first-order rate constant

(T^"1) for -OH scavenging by all constituents of the solution except the probe

6 k₆ = 2.7 x 10⁸ L/mol-s

Limited reaction kinetics is the condition in which low concentrations of the target compound limits the second- order oxidation reaction. Correspondingly, the clean-up goal for the target compound in the ground water can be difficult to achieve. Exacerbating the issue are the numerous scavengers which effectively compete against low concentrations of target compound for hydroxyl radicals. While adsorption using activated carbon and oxidation using the Fenton mechanism has been widely used separately in ground water remediation and wastewater treatment, problems associated with oxidation in subsurface systems involve poor reaction kinetics, excessive scavenging and excessive non-productive hydrogen peroxide consuming reactions. Problems associated with adsorption in subsurface systems relate to exhausting the sorption capacity of carbon. To replace the carbon, it must be excavated and transported to a specialized facility for disposal. Long-term risks associated with this disposal is environmentally undesirable. If the carbon is reactivated rather than disposed of, additional costs are incurred in this reactivation.

Enzymatic and manganese reactions with hydrogen peroxide can consume hydrogen peroxide in reactions which do not yield hydroxyl radicals ( cf . Table 1 and Figure 2) . Selection criteria for granulated activated carbon should, therefore, include low manganese content. The iron content of the granulated activated carbon can be increased to enhance the Fenton mechanism. The effect of the enzymatic reactions are relatively short term because hydrogen peroxide inhibits catalase enzyme activity via the formation of an intermediate enzyme-substrate compound (Nicholls and Schonbaum, 1963; Aggarwal et al, 1991). Summary of the Invention

It is an object of the present invention to overcome the aforementioned deficiencies in the prior art.

It is another object of the present invention to provide a method for treating contaminated water. It is a further object of the present invention to provide a method for treating any waste stream which includes compounds which adsorb and can be oxidized by the hydroxyl radical . For the purposes of the present invention, the term

"contaminated water" includes any water, waste stream, or ground water which has compounds which adsorb and can be oxidized by the hydroxyl radical. According to the present invention, contaminated water is treated by first adsorbing contaminants onto a suitable non-treated sorbent and subsequently oxidizing the compounds near the surface of the sorbent . In one embodiment, contaminants are adsorbed onto activated carbon containing iron, or which has been amended with iron in solution, which concentrates contaminants from the contaminated water onto the reactive medium. Hydrogen peroxide or other oxidizing agent is then added, which reacts with iron to generate hydroxyl radicals which oxidize the adsorbed contaminants.

In the process of the present invention, contaminants are oxidized using a Fenton-driven mechanism and destroyed in si tu, and the sorbent is reactivated/regenerated in si tu . That is, the contaminants which can be treated by the process of the present invention are substances which can be oxidized by hydroxyl radicals through the Fenton mechanism. This process makes it possible to treat mixed wastes. For example, benzene, xylene, toluene, and halogenated compounds can be treated in the same waste stream, whereas conventional zero-valent iron treatment only involved dehalogenation. The adsorption/oxidation system of the present invention provides for adsorbing and oxidizing contaminants on the sorbent surface. This process is also much more efficient that conducting the Fenton reaction in bulk liquid.

Adsorption immobilizes and concentrates the contaminants onto the iron-treated sorbent. Treatment involves adding an oxidizing agent to the surface or solution of the iron-treated sorbent, which produces hydroxyl radical as a reaction intermediate. The hydroxyl radical oxidizes the contaminants sorbed to the iron-treated sorbent. This treatment process overcomes some of the limitations of other chemical oxidation processes involving oxidation of contaminants .

Brief Description of the Drawings

Figure 1 is a schematic of the adsorption/oxidation process of the present invention. Figure 1A shows the initial time and concentration and the flow (Q₀) of water containing contaminants at an initial concentration (C₀) into the sorbent/iron medium, which results in contaminant adsorption to the granulated activated carbon yielding an acceptable effluent concentration (C_a) . Figure IB shows that contaminants concentrate on the carbon/iron medium and reach an equilibrium concentration (X . Figure IC shows that hydrogen peroxide perfused into the carbon/iron medium initiates the Fenton mechanism. This results in the formation of hydroxyl radical, which oxidizes sorbed contaminants, thus decreasing the concentration of contaminants and regenerating the granulated activated carbon (X₂) . Figure ID shows cessation of hydrogen peroxide application and contaminated water continuing to flow through the reactive medium for another cycle.

Figure 2 is a cross-sectional diagram of a hydraulic barrier and adsorption/oxidation treatment unit.

Figure 3 is a plan-view diagram of a hydraulic barrier and adsorption/oxidation treatment unit. Figure 4 is a schematic of hydrogen peroxide reactions in soil slurry containing 2-chlorophenol . (a) is non-hydroxyl radical producing reactions; (b) is cycling between Fe(II) and Fe(III) oxidation states; (c) is production of hydroxyl radical via the Fenton reaction; (d) is competition between 2-chlorophenol and scavengers (S_±) for hydroxyl radical; and (e) is the reduction of Fe(III) via the superoxide radical .

Figure 5 shows time-dependent concentrations of hydrogen peroxide in batch reactors containing different types of granulated activated carbon with three successive applications of 100 mL 0.9% hydrogen peroxide into 1 g granulated activated carbon with 2CP = 35.4 g/Kg. Figure 6 shows time-dependent concentrations of hydrogen peroxide in batch reactors containing granulated activated carbon with different concentrations of iron (none, low, medium, high, containing 24.0, 5500.0, 9790.0, 12050.0 mg/Kg total iron, respectively) ; three successive applications of 100 mL 0.9% hydrogen peroxide into 1 g granulated activated carbon with 2CP = 35.4 g/kg.

Figure 7 shows time-dependent concentrations of hydrogen peroxide in batch reactors containing granulated activated carbon, Fe = 5500.0 mg/Kg, with different initial concentrations of hydrogen peroxide (9370, 21400, 71,400 mg/L, respectively. Three successive applications of 100 mL 0.9% hydrogen peroxide into 1 g granulated activated carbon with 2CP = 35.4 g/kg. Detailed Description of the Invention

Contaminated water is treated by a combination of adsorption of organic compounds in the water onto a sorbent to concentrate the contaminants and subsequent oxidation on the sorbent. This process is illustrated schematically in Figure 1. Contaminated ground water flows (Q₀) through a granulated activated carbon bed where contaminates adsorb onto the carbon. This results in acceptable concentrations of contaminants (C_A) in the effluent. Subsequently, the purified water continues through the treatment unit and back into the aquifer (Figure 1A) . Adsorption immobilizes and concentrates the contaminants onto the sorbent, which also contains iron which is capable of facilitating Fenton-driven oxidation reactions. Prior to breakthrough of contaminants from the reactive granulated activated carbon (Figure IB) , an oxidant, such as hydrogen peroxide, is injected, which reacts with iron, generating a strong oxidant, the hydroxyl radical. The hydroxyl radical oxidizes sorbed contaminants (Figure ID) . Ideally, the sorbent is treated in si tu , and the sorption capacity of the sorbent is regenerated (Figure ID) . The application of oxidant is performed at appropriate intervals to maintain an acceptable concentration in the contaminated water passing through the reactive unit. The process of the present invention treats contaminated water by destroying the contaminants in si tu, providing an efficient and economical treatment option. The process can be used above ground or below ground. For treating groundwater, the water treatment system can be constructed entirely below grade and can be entirely gravity driven, all of which reduces operation and maintenance costs. The system can be easily monitored. A broad range of water contaminants, including halogenated solvents, polycyclic aromatic hydrocarbons, petroleum constituents, etc., have a sufficiently high reaction rate constant with hydroxyl radical and, therefore, are viable target compounds to be oxidized. This indicates that the proposed treatment technology has wide application. Oxidation

In the Fenton mechanism, hydrogen peroxide reacts with Fe(II) to yield hydroxyl radical and Fe(III), as shown in Table 1, Reaction 1. The Fe(III) is reduced to Fe(II) via reaction with hydrogen peroxide, as shown in Table 1, Reaction 2. Reactions 1 and 2 cycle iron between the ferrous and the ferric oxidation states, producing hydroxyl radical continuously until the hydrogen peroxide is fully consumed. These reactions may involve either dissolved iron (homogeneous reactions) or solid phase iron oxides (heterogeneous reactions) . Since the hydroxyl radical is a powerful oxidant and reacts with compounds at near diffusion- controlled rates (Walling, 1975; Haag and Yao, 1992), hydrogen peroxide has been used to generate hydroxyl radical and oxidize undesirable contaminants in soils and aquifers (Watts et al, 1993; Ravikumar and Gurol, 1994; Yeh and Novak, 1995) .

Reactions 1 and 2 indicate that the overall Fenton mechanism is acid generating. pH affects hydrogen peroxide stability (Schumb et al, 1955) and iron solubility. Oxidation efficiency is optimum under acidic conditions (Watts et al, 1991) . In any oxidation system involving Fenton-derived hydroxyl radical, pH should be monitored, and steps taken to mitigate acidic conditions. A similar reaction involving the hydrogen peroxide oxidation/reduction cycling of Mn²⁺ and MnOOH(s) is thermodynamically favorable (Pardiek et al, 1992) and kinetically fast, but does not yield hydroxyl radical, as shown in Table 1, Reactions 7 and 8. Naturally-occurring soil microorganisms contain enzymatic catalysts, such as catalase and peroxidase, which also readily decompose hydrogen peroxide without producing the hydroxyl radical . The reactions between manganese or enzymatic catalysts and hydrogen peroxide reduce the amount of hydrogen peroxide available for Fenton reactions. Numerous non-target chemical species present in solid and aquifers, both naturally occurring (i.e., C0₃ ^2", HC0₃ ^", Cl^", etc) and anthropogenic (i.e., H₂0₂, mixed waste constituents, etc.), will also react with hydroxyl radical. The non-target chemical species scavenge hydroxyl radical which would otherwise oxidize the target contaminants. Hydrogen peroxide is generally present at high concentrations in Fenton systems and has a moderate reaction rate constant (2.7 x 10⁷ L/mol-s; Buxton et al, 1988) and, therefore, is responsible for scavenging a significant fraction of hydroxyl radical produced in Fenton systems. Oxygen is a significant byproduct of reactions involving hydrogen peroxide in soils or aquifers. Reaction 6 in Table 1 indicates that 0₂ ^~ reacts with Fe(III) to yield 0₂. In aqueous systems, the rapid rate of degradation of high concentrations of hydrogen peroxide, in conjunction with the relatively low solubility of dissolved oxygen, the formation of bubbles, i.e., oxygen gas, is certain. In the field applications of Fenton systems, the formation of gaseous oxygen in porous media may result in gas blockage of fluid flow.

A wide range of organic compounds of environmental significance and their reaction rate constants with hydroxyl radical have been reported (Haag and Yao, 1992; Buxton et al , 1988; Dorfman and Adams, 1973) . Organic compounds that are common ground water and soil contaminants at Superfund Sites have relatively high reaction rate constants (i.e., 10⁸-10¹⁰ L/mol-s) , indicating their potential for oxidation by hydroxyl radical . Adsorption

Any type of sorbent may be used in the process of the present invention, depending upon the contaminants to be removed from the water. The criteria for the sorbent are that it be capable of concentrating the contaminant sought to be treated/removed, and that it provide iron in some form for the Fenton mechanism during oxidation of the contaminant. The primary role of granulated activated carbon, as of any sorbent, in the process of the present invention is to immobilize and concentrate target compounds on the same surface on which the hydroxyl radical is produced. Subsequently, the target compounds on or near the surface of the sorbent are oxidized. In addition to activated carbon, which can be granulated, powdered, etc., sorbents for use in the process of the present invention include ion exchange resins, both anionic, cationic, or both, zeolites and other molecular sieves, alumina, silica, silicates, aluminum phosphates, and the like. One skilled in the art can readily determine which adsorbent is effective in adsorbing and concentrating a particular contaminant.

Granulated activated carbon is a preferred sorbent for removing organic compounds from waste streams . For example, the pore size distributions and surface chemistry for a given granulated activated carbon are directly related to the starting raw material and the activation conditions. pH and concentration of transition metals in the carbon vary and, therefore, affect the reactivity of oxidants in granulated activated carbons.

The oxidant can be any conventional oxidizing agent that works through the Fenton mechanism for oxidizing contaminants. While hydrogen peroxide has been illustrated in the specific examples, any other oxidizing agent that produces hydroxyl radicals in the presence of iron can be used, including ozone, permanganate salts, persulfate salts, and the like. Iron can be added to the sorbent to enhance the

Fenton mechanism and, therefore, enhance hydroxyl radical production. The amount of iron will affect the ability to carry out Fenton reactions and, therefore, the iron concentration of the sorbent can be optimized. For example, the concentration of iron may be adjusted so that the density of iron sites (i.e., the spatial distribution of hydroxyl radical production sites) is similar to the density of sorption sites on the sorbent to assure the spatial probability of hydroxyl radical and contaminant interaction. One method of iron attachment to a sorbent involves raising the pH using a sodium hydroxide solution to precipitate ferric iron in the pores of the sorbent. Additionally, the use of other forms of iron and the use of chelators and ligand agents can be used to attach iron to sorbent surfaces. Alternatively, a solution of iron and oxidant can be perfused through the sorbent to oxidize contaminants adsorbed thereto.

The adsorption/oxidation process described above is, thus, used in systems in which contaminated water can be diverted through a sorbent/iron treatment unit. Details of an adsorption/oxidation treatment process are provided below in the context of a hydraulic barrier in conjunction with a carbon/iron treatment unit, but this example is for illustrative purposes only and is not limiting of the invention. This treatment process can also be used in above- ground treatment systems and be constructed in existing or planned containment systems to serve as a pressure release mechanism to improve hydraulic control. For example, a containment system or hydraulic barrier can be designed to leak while meeting stringent ground water quality criteria.

Other applications include a below grade pump and treatment system, a passive landfill leachate treatment system, or any above ground treatment process.

Field Application in a Hydraulic Barrier/Treatment Unit Flow blockage through the granulated carbon/iron medium may be a limitation of the process. For example, the reaction product oxygen (Reactions 6, 8 and 9 in Table 1) will result in gas formation which may fill the void spaces and inhibit water flow. This problem can be avoided using an upflow regimen allowing gaseous oxygen bubbles to rise in the carbon/iron unit and escape into the headspace of the unit or distribution gallery and into the air. Manganese oxide may be used to ensure that all hydrogen peroxide is consumed after it leaves the reactive media. This step minimizes the introduction of hydrogen peroxide into the distribution gallery. Biofouling may occur because of the high surface area and substrate concentration associated with the granulated activated carbon. However, biofouling in injection wells for in si tu bioremediation have been remedied with hydrogen peroxide and, therefore, may not be a problem. Precipitation of solid phase material may result in fouling of the carbon/iron unit. The occurrence of this potential problem will be site specific. Precipitation may be eliminated by treatment with dilute acid. Alternatively, since the Fenton mechanism is acid generating, this may be sufficient to dissolve any precipitate which forms on the granulated activated carbon. The acid generating Fenton mechanism may release excess acid from the treatment unit.

Two approaches are described below which may be used to maintain a constant pH from the treatment unit . While potential limitations on the process exist, the potential effects on treatability, operation and maintenance and the associated costs must be evaluated on a site-specific basis.

Limited treatment volume, hydroxyl radical scavenging, low reaction kinetics, and non-productive hydrogen peroxide consumption may reduce the effectiveness of hydrogen peroxide application in soil slurry reactors. These limitations are minimized in the treatment system of the present invention. For example, contaminants partition from the ground water onto the carbon/iron medium, thus achieving stringent treatment criteria. Through this process, contaminants are concentrated on the carbon/iron reactive medium which enhances reaction kinetics. Further, hydrogen peroxide is applied in a scavenger-reduced solution to minimize the role of scavengers in the treatment unit . Flow blockage and pH reduction may also result, but design options can be implemented to minimize these potential limitations. Sorption/Oxidation System Design

Cross-section and plan-view diagrams of a hydraulic barrier and treatment unit according to the present invention illustrate the hydraulic and treatment components of the proposed system, as shown in Figures 2 and 3. Contaminated water flowed into a gravel-filled collection gallery 20 and was directed through the carbon/iron treatment medium 21. The ground water then passed through the hydraulic barrier 22 via a pipe 23 and back into the aquifer, through a gravel filled distribution gallery 24. The collection and distribution galleries facilitate water flow since head loss is minimized in gravel relative to the head loss through the porous medium. The combined head loss through the alternative flow regimen must be less than the head loss of the original flow regimen to ensure continuity and to minimize ground water flow stagnation. For example, the equivalent porous medium of the alternative flow regimen is comprised of the collection/distribution galleries, pipe flow, and treatment unit. The influence of the hydraulic barrier, collection/distribution galleries and treatment unit on the water gradient, flow pattern and capture zone must be evaluated on a site-specific basis. Using contaminated ground water as an example, ground water flow in the pipe through the hydraulic barrier provides minimal disturbance to the wall . Since the system can be constructed below ground and is gravity driven, it is not subject to freezing or power outages (O'Brien et al, 1997) . The granulated activated carbon/iron medium can be readily accessed for sampling or replenishment if necessary. Because of the oxidation of sorbed contaminants, ideally the granulated activated carbon is regenerated in si tu, and replenishment may be unnecessary. The granulated activated carbon/iron medium can be slurried, pumped and remixed in the treatment unit if recycling of the granulated activated carbon/iron medium is determined to be beneficial. Contaminants adsorbed to the carbon/iron medium were oxidized via hydrogen peroxide perfusion at selected intervals . Hydrogen peroxide was introduced by gravity into the system through a port at the surface which leads to a slotted distribution header 25 at the bottom of the carbon/iron unit. Gaseous oxygen formed in the carbon/iron unit will rise because of buoyancy and will escape into the headspace of the reactor unit, where it is vented into the atmosphere. The particle size of the activated carbon should be large enough to ensure mobility of gas bubbles in the carbon/iron medium and to minimize head loss. Assuming oxygen bubbles are diverted to the distribution gallery, an open chamber can be designed which will allow separation of bubbles and water. Passive gas capture and venting designs may also be used. It is undesirable to introduce hydrogen peroxide into the gravel-filled distribution gallery, since hydrogen peroxide decomposition and oxygen blockage may result. To ensure that hydrogen peroxide is degraded prior to leaving the reactor unit, a layer of manganese oxide ore or manganese-rich granulated activated carbon can be installed above the carbon/iron bed to rapidly decompose the remaining hydrogen peroxide and ensure that no hydrogen peroxide will be introduced into the distribution gallery. Laboratory results indicate that the Fenton mechanism is an acid-generating process. This is consistent with Reactions 1-2 shown in Table 1, which indicate a net production of hydrogen ion. Acid production may be problematic, and its control at field scale may be advantageous. Different approaches can be used to control the pH in the oxidation system. A layer of limestone placed on top of the sorbent/iron unit and/or in the distribution gallery will neutralize a low pH solution. Another pH control method uses an automated pH-stat. This system comprises continuous pH measurement and adjustment of pH using an acceptable source of base, such as sodium hydroxide. It is important to note that oxidant perfusion into the treatment unit occurs infrequently, and the volume of water relative to the volume of water between oxidation events is small. Therefore, pH control may be necessary only when the treatment unit is undergoing oxidation. One option, of course, is simply to remove the solution containing spent oxidant for disposal or treatment elsewhere. Monitoring treatment performance involves an upgradient well in the influent area and one downgradient well in the effluent area. A well 26 in the upper treatment unit could be useful for obtaining ground water quality data on the treatment unit. Specifically, monitoring for the contaminant provides information on breakthrough of the carbon/iron unit and indicates when oxidation is required. This information is helpful in establishing an oxidation schedule. Further, assuming halogenated contaminants were present, monitoring chlorides concentration in the treatment bed using monitoring well 27 during an oxidation treatment is useful in determining when oxidation is complete. For example, assuming the treatment bed is operated in batches or continuously, chloride concentration would eventually diminish as sorbed contaminants are oxidized. This simple monitoring system can be used to verify that the water quality leaving the treatment unit satisfies ground water quality cleanup goals.

Oxidation of sorbed contaminant occurs when an oxidant is perfused through the sorbent/iron unit. This is accomplished by, for example, introducing hydrogen peroxide into a port at the surface. The hydrogen peroxide then flows downward and out of a slotted distribution header and into the sorbent/iron media, as shown in Figures 2 and 3. Perfusing hydrogen peroxide can be effected either in continuous flow or in batch modes. The sorption/oxidation treatment system can be designed as two parallel units. For example, when hydrogen peroxide is perfused through one sorbent/iron unit, contaminants in the water can be treated via the second sorbent/iron unit. This provides the flexibility of operating one unit in a sorption mode and one unit in an oxidation/standby mode to ensure complete use of hydrogen peroxide. Other design configurations are also possible, such as series, batch, or continuous. The design options are also applicable to above-ground treatment systems .

The frequency at which oxidant is applied and the concentration depends on the mass loading rate, the mass of sorbent and treatment efficiency. The treatment efficiency depends on numerous parameters, including pH, hydrogen peroxide concentration, iron concentration, contact time, scavenging, non-productive oxidant degradation reactions, reaction rate constants, concentration of target compounds, etc. These parameters vary significantly from site to site, and the frequency necessarily reflects such variability.

There are several advantages to adsorbing contaminants in contaminated water onto iron amended sorbent and subsequent oxidation via the Fenton mechanism. In si tu treatment of contaminants minimizes water pumping and associated costs . The process involves contaminant destruction, not simply mass transfer onto the sorbent, which requires subsequent handling for transport and treatment or disposal. Ideally, the sorbent is regenerated each time oxidant is applied, although the sorption capacity will not be completely restored to that of virgin sorbent. Mixed wastes can be sorbed and oxidized, since a wide range of organic compounds sorb to activated carbon and react with hydroxyl radical. Since co-disposal of organic compounds is common, a ground water plume comprised of contaminants, such as halogenated volatiles, polycyclic aromatic hydrocarbons, and fuel compounds (BTEX) , can be treated together. Significant process control can be achieved in the system, including concentration and hydraulic retention time of oxidant and type of sorbent (particle size, oxidant reactivity, manganese content, contaminant sorption, iron concentration, etc.). Performance monitoring can be simplified since ground water wells can be placed directly in the collection and distribution galleries for pre-treatment and post-treatment evaluation, respectively. Significant treatment efficiency can be obtained, since the process involves a concentration step, minimizes the role of scavengers, and optimizes pH, since many oxidants can be used in dilute solution and since the Fenton mechanism is acid generating. The entire system can be gravity driven, including the delivery of oxidant . Thus, there need be no above-ground structures and iron, oxidants and sorbents are relatively inexpensive. This collectively minimizes costs. The loss of iron in the sorbent may potentially be a concern in low pH or redox conditions. Preliminary experiments indicate that iron mobility does not occur except at pH < 2.5. Assuming the pH in the treatment unit is maintained above 3, iron mobility can be limited. Iron contained in the mineral matrix of carbon is immobile relative to iron amended to the carbon.

Another potential problem may result in ground water contaminated with Cr³⁺ . Assuming Cr³⁺ is present in contaminated water, it may possibly accumulate in the treatment unit and be oxidized to Cr⁶⁺ upon application of hydrogen peroxide. This may increase the mobility and toxicity of chromium. It may be preferable to avoid treating water that is contaminated with metals by the process of the present invention.

Experimental Experiments were conducted to remove 2-chlorophenol from contaminated water by adsorption and oxidation via the oxidant-driven Fenton mechanism. The release of chloride ion indicated the transformation and dehalogenation of 2- chlorophenol . Adsorption and oxidation treatments in four commercially-available granulated activated carbons were compared, using for comparison the effects of 2-chlorophenol concentration, iron amendment of granulated activated carbon, and hydrogen peroxide concentration oxidation of 2- chlorophenol . Transformation products of 2-chlorophenol were identified using gas chromatography/mass spectrographic analysis .

As shown in Figure 1, contaminated ground water flows (Q₀) through a granulated activated carbon (GAC) bed and contaminants adsorb onto the carbon, with resulting acceptable concentrations thereof (C_A) in the effluent. Subsequently, the purified water continues through the treatment unit and back into the aquifer (Figure 1A) .

Adsorption immobilizes and concentrates the contaminant onto the GAC, which also contains iron capable of facilitating Fenton-driven oxidation reactions. Prior to breakthrough of contaminants from the reactive GAC (Figure IB) , hydrogen peroxide is injected, which reacts with the iron in the GAC, generating the hydroxyl radical, a strong oxidant, which oxidizes sorbed contaminants (Figure IC) . Ideally, the GAC is treated in si tu, and the sorption capacity of the GAC is regenerated (Figure ID) . Hydrogen peroxide is applied at appropriate intervals to maintain an acceptable concentration in the water passing through the reactive unit. Oxidation Because -OH is a powerful indiscriminate oxidant which reacts with compounds at near diffusion-controlled rates (Walling, 1975; Haag and Yao, 1992) , H₂0₂ has been used to generate -OH (rxns 1, 2) and oxidize undesirable contaminants in soils and aquifers (Watts et al, 1993; Ravikumar and Gurol , 1994; Yeh and Novak, 1995) . The Fenton mechanism involves either dissolved Fe in homogeneous reactions or solid phase Fe in heterogeneous reactions. In either case, H₂0₂ cycles Fe between oxidation states yielding ^•OH and other byproducts (rxns 1, 2) . The overall Fenton mechanism is acid generating. H₂0₂ stability increases with decreasing pH in Fenton systems and oxidation efficiency is optimum under acidic conditions (Watts et al, 1991) .

H₂0₂ + Fe(III) → Fe(II) + ^• 0₂ ^~ + 2H⁺ (rxn 1)

H₂0₂ + Fe(II) → Fe(III) + OH^" + -OH (rxn 2)

A wide range of organic compounds (TCE, BTEX, PCP, naphthalene, general pesticides) which commonly contaminate ground water and soil at Superfund Sites (USEPA, 1991) have moderate to high reaction rate constants (10⁸-10¹⁰ M^s^"1) (Haag and Yao, 1992; Buxton et al, 1988; Dorfman and Adams, 1973) . This range of rate constants indicates the vulnerability of these compounds to react with ^• OH in Fenton-based oxidation systems . The reaction of -OH with 2-chlorophenol (2CP) (rxn

3) predominantly involves -OH insertion yielding a phenoxyl radical, but direct H-abstraction via -OH is also possible

(Getoff and Solar, 1986) . The phenoxyl radical has several resonance structures which can lead to the formation of different final products. Reactions which follow addition and abstraction reactions includes disproportionation and dimerization. Therefore, several possible transformation pathways follow. k₃

2CP + -OH → phenoxyl radical (rxn 3) where k₃ = 1.2 x 10¹⁰ M^s^"1 (Getoff and Solar, 1986)

Performance evaluation of Fenton-based remediation is generally determined from the rate of disappearance of a target analyte or the appearance of a decomposition product

(Lipczynska-Kochany et al, 1995; Watts et al, 1990; Gates and Siegrist, 1995). Chloride ion (CI^") , for example, is a byproduct of the radical-mediated oxidation of chlorinated compounds, and its measurement is a reliable indicator of contaminant transformation. However, transformation of either the parent chlorinated compound or oxidation byproducts may proceed without simultaneous CI^" release (Getoff and Solar, 1986) . Also, complete CI^" release does not assure mineralization of the parent compound as indicated by intermediates measured in such systems.

Limitations to Fenton-based oxidation may include excessive H₂0₂ decomposition and the associated non-productive reactions (those that do not result in -OH production), excessive scavenging of -OH, insufficient Fe, and limited reaction rate kinetics (Huling et al, 1998a; 1998b) . For example, oxidation/reduction cycling of Mn(II) and MnOOH(s) by H₂0₂, and enzymatic reactions with H₂0₂ does not yield -OH. These non-productive H₂0₂ reactions reduce the amount of H₂0₂ available for Fenton reactions and can be a significant source of treatment inefficiency. Non-target chemical species (Si) present in soil and ground water, both naturally occurring (N0₃ ^", S0₄ ^2", Cl^", HP0₄ ^2", HC0₃ ^", C0₃ ^"2) (Buxton et al , 1988; Pignatello, 1992; Lipczynska-Kochany et al, 1995) (rxn

4) and anthropogenic (H₂0₂) (rxn 5), will react with -OH. These chemical species scavenge ^• OH which may otherwise react with the target contaminants. H₂0₂ is generally present at high concentrations in Fenton systems and has a moderate reaction rate constant and, therefore, is responsible for scavenging a significant fraction of ^• OH produced in Fenton systems (Huling et al, 1998a) .

Z =ι K ^• OH + ∑ⁿ ₁₌₁ S_x → scavenging byproducts (rxn 4)

^•OH + H₂0₂ → scavenging byproducts (rxn 5) where k₅ = 2.7xl0⁷ M^s^"1 (Buxton et al, 1988)

The reaction rate equations for -OH and 2CP (eqs 1 and 2, respectively) can be written as follows : d[2CP]/dt = -k₃ [2CP] [OH] (eq 1) d[-OH]/dt = k₂ [Fe(II)] [H₂0₂] - (eq 2)

(k₃ [2CP] [-OH] + ∑ⁿ ₁₌₁ k₁ [S [-OH] + k₅ [H₂0₂] [-OH])

Reaction rates between ^• OH and target compounds may be limited simply due to low concentrations. For example,

^• OH concentrations in environmental Fenton-based remediation systems are very low due to its extreme reactivity. Similarly, concentrations of target compounds may be low due to the low concentrations required for ground water cleanup. Adsorption on the GAC immobilizes and concentrates the target compound (s) which reduces concentration related reaction rate limitations .

A wide range of organic compounds and their reaction rate constants with ^• OH have been reported (Haag and Yao, 1992; Buxton et al, 1988; Dorfman and Adams, 1973) .

Organic compounds (TCE, BTEX, PCP, naphthalene, general pesticides) which commonly contaminate ground water and soil at Superfund Sites (USEPA, 1991) have moderate to high reaction rate constants (10⁸-10¹⁰ M^"1s^"1) . This range of reaction rates indicates a good potential for these compounds to react with hydroxyl radical in Fenton-based oxidation system. Hydroxyl radical adds to the aromatic or heterocyclic rings, as well as to the unsaturated bonds of alkenes or alkynes (Lipczynska-Kochany et al, 1995) . Substituents on these rings exert their influence on the reaction rates since the hydroxyl radical reaction rate depends on the availability of electrons on the ring (Shetiya et al, 1976) . For example, the electron availability over the ring is greater in phenol than benzene because of the interaction of the unshared electron pair of the hydroxyl with the delocalized 7r orbitals of the ring. Therefore, the reaction rate of phenol is greater than that of benzene.

2-chlorophenol was selected as the target contaminant because it is relatively non-volatile, soluble, conveniently analyzed and has a published reaction rate constant with hydroxyl radical. The reaction of hydroxyl radical with 2-chlorophenol (rxn 3; Figure 4) yields a phenyl radical which has several resonance structures and may react with 2-chlorophenol to release chloride ion (Getoff and Solar, 1986) . 2-chlorophenol + ^• OH → reaction products k₁₀=1.2 x lO^M^s^"1. Hydroxyl radical species react with phenols and related compounds preferentially on the ortho- and para- positions and less on the meta- and iso-sites. The phenoxyl radical has several resonance structures which can lead to the formation of different products. Direct hydrogen- abstraction via hydroxyl radical is also possible. The first stage of oxidation of chlorophenols leads to the formation of various hydroxy benzenes (phenol, catechol, resorcinol, hydroquinone and hydroxyhydroquinone) (Getoff and Solar, 1986) . Other reactions which follow the addition and abstraction reactions include disproportionation and dimerization .

Performance evaluation of Fenton-based remediation is generally based on the rate of disappearance of a target analyte or the appearance of a decomposition product (Lipczynska-Kochany et al, 1995; Watts et al, 1990; Gates and Siegrist, 1995) . Chloride ion (CI^") , for example, is a byproduct of the radical-mediated oxidation of chlorinated compounds, and its presence is an indicator that the compound was transformed. Transformation of either the parent chlorinated compound or oxidation byproducts may proceed without simultaneous release of chloride ion. Further, complete release of chloride ion does not ensure mineralization of the parent compound. Nevertheless, chloride ion production is a reliable indicator of contaminant transformation. In the examples given herein, production of chloride ion was used as a diagnostic to quantify hydroxyl radical mediated transformation reaction of 2-chlorophenol.

Adsorption, Oxidation and Granulated Activated Carbon

Granulated activated carbon (GAC) is widely used and is successful in facilitating the removal of organic compounds from waste streams. Granulated activated carbon has a high surface area, about 800-1000 m²/g, low bulk density, high porosity, high sorption capacity, is available in a large range of particle sizes, is produced from different sources of raw material, and its activation varies between manufacturers. Correspondingly, the performance of granulated activated carbon varies in chemical and physical characteristics and in sorption performance. More specifically, the following characteristics vary: (1) reactivity with hydrogen peroxide; (2) sorption isotherm;

(3) rate of scavenging hydroxyl radical; and

(4) iron concentration.

All of these characteristics affect adsorption and oxidation, and, therefore, some granulated activated carbons may be more effective than others for the treatment process of the present invention.

Reaction of an oxidant with granulated activated carbon can be attributed to different mechanisms, including:

(1) reaction with iron, manganese, and other transition metals;

(2) reduction of hydrogen peroxide by the reduced surfaces of activated carbon; (3) reaction between hydroxyl radical and hydrogen peroxides (scavenging) ; and

(4) sorption of hydrogen peroxide by granulated activated carbon. The surface of granulated activated carbon contains many functional groups and is generally reduced, both of which facilitate reactions with hydrogen peroxide. The concentration of iron and manganese in granulated activated carbon varies, which is attributed to the source of the activated carbon, as well as to the activation process. Methods, Materials, and Analytical Procedures Sorption

The reactors used were 125 mL Ehlermeyer flasks containing 1.0 gram granulated activated carbon and 40 mL 6 mM solution of 2-chlorophenol. The reactors were placed on an orbital shaker table for 24 hours, which allowed complete (> 99%) sorption of 2-chlorophenol. The granulated activated carbon slurry was decanted, and the solutions were analyzed for chloride ion, 2-chlorophenol, and total iron (Fe_τ) . Oxidation

The remaining granulated activated carbon was amended with 100 mL hydrogen peroxide, of 0.7%, 0.9%, 1.2%, or 7.2%, w/w concentration, or deionized water in three successive applications, unless otherwise noted. The reactors were wrapped in foil to prevent photodecay, covered with parafilm to minimize volatile losses and evaporation, and placed on an orbital shaker table at 100 rpm. The granulated activated carbon slurry was decanted, and the solution was analyzed for hydrogen peroxide, chloride ion, 2- chlorophenol , and Fe_τ. Control reactors containing granulated activated carbon and hydrogen peroxide, but not 2- chlorophenol, were used to measure background chloride ion. The pH of 2-chlorophenol control solutions was adjusted to between 2 and 7 and the solutions were analyzed for 2- chlorophenol. These data indicated no transformation. The granulated activated carbon slurry pH was measured by placing a pH probe (Orion Sure-Flow ROSS Combination pH) into the slurry for five minutes to instrument stabilization. Samples were collected by pipetting 1.5 mL from a completely mixed suspension and filtered using a Gelman 0.2 μm filter which stopped all reactions and removed colloidal particles interfering with subsequent analyses . Hydrogen peroxide was measured immediately, and 2-chlorophenol subsamples were stored at 4°C for analysis when the experiment was completed. The 2- chlorophenol was obtained from Aldrich Chemical . EXP3 granulated activated carbon is commercially available bituminous-based carbon obtained from Calgon Chemical Corp. (Pittsburgh, PA) . EXP4 granulated activated carbon was derived from the same stock of Bakers carbon but was activated differently to minimize degradation of hydrogen peroxide (Rich Hayden, personal communication, 1997) . Two additional granulated activated carbons were obtained from

Norit America, Inc., Pryor, OK (type A4) and Carbochem, Inc., of Haverfers, PA, (type LQ-900S) . The particle size distribution was 8 x 30 for all granulated activated carbons. The iron content of Calgon Chemical Corp. (special) granulated activated carbon was enhanced at low, medium, and high concentrations by mixing 15 grams granulated activated carbon into solutions comprising 38 mL and 0.744, 3.73, and 6.44 grams FeS0₄.7H₂0, respectively. Ferrous iron was precipitated by adjusting the acidic solution to pH 5.5 with sodium hydroxide. The slurry of granulated activated carbon and iron was filtered and rinsed with deionized water through a number 35 sieve which retained > 99.99% of the granulated activated carbon. The granulated activated carbon was air dried and placed into 40 mL glass vials until used. Representative samples of the granulated activated carbon/iron stock were analyzed by inductively coupled argon plasma after metals were extracted from the granulated activated carbon by digesting a 0.25 gram sample in 40 mL of 19% nitric acid for 40 minutes in a microwave oven at 150°C and 145 psia.

Analyses of 2-chlorophenol were performed by a Waters high performance liquid chromatography Alliance Separations Module (Model No. 2690) using a Waters 996 Photodiode Array detector and a Nova-Pak C18 stainless steel column. The mobile phase was 30% acetonitrile and 0.3% acetic acid in deionized water, the flow rate was 1.0 L/min, and the injection volume was 199 μL . The wavelength used was 200 nm and the average retention time was 5.5-5.8 minutes.

The 2-chlorophenol standard curve ranged from 0 to 400 μM, r² = 0.999, 0.14 μM detection limit. Check standards, blanks, duplicates, and spikes were run with each sample set, and the analytical quality was found to be in control. Hydrogen peroxide was analyzed using a modified peroxytitanic acid calorimetric procedure. Hydrogen peroxide standards ranged from 0-3 mM, r² = 0.999, detection limit = 0.1 mg/L; 0.9 mL of the standard hydrogen peroxide solution was added to 0.1 mL of titanium sulfate reagent and allowed to react for one hour. Appropriate dilutions were made of solution where the concentration exceeded the range of the reported calibration curve. Filtered granulated activated carbon slurry samples in triplicate were prepared in a similar manner. Absorbance of the hydrogen peroxide-titanium sulfate mixture was measured at 407 nm (A₄₀₇) using a Milton

Roy Spectronic 401 spectrophotometer . Regression analysis of the spectrophotometric response and the hydrogen peroxide concentration yielded a standard curve with an analytic equation of [H₂0₂] (mg/L) = 44.5 A₄₀₇, r²=0.999. Titanium sulfate reagent was obtained from Pfaltz and Bauer, Inc., and the 30 w/w% aqueous solution of hydrogen peroxide was from the Aldrich Chemical Company. Chloride ion was analyzed by Waters capillary electrophoresis method N-601 CI . Iron was analyzed by EPA Method No. 3500-Fe D, Phenanthroline method. Transformation products of 2-chlorophenol on the granulated activated carbon were identified using gas chromatography and mass spectroscopy . This analysis involved a derivatization technique using N-methyl-N- [ (tert-butyldimethyl) silyl] tri- fluoroacetaminde from Aldrich Chemical Co., Milwaukee, WI , which yielded tert-butyl-dimethylsilyl ethers and esters

(Heberer et al, 1997; Mawhinney 1983; Mawhinney et al, 1986). For each treatment, 100 μL of acetonitrile and 100 μL of N- methyl-N- [ (tert-butyldimethyl) silyl] trifluoroacetaminde were added to 10 mg of the granulated activated carbon sample.

This mixture was heated for one hour at 60°C. 2- chlorophenol, carbonic acid, sulfuric acid, and eleven dioic and hydroxy acids were derivatized to confirm the identification of the reaction products. Derivatized extracts were injected into a Finnigan 4600 gas chromatograph/mass spectrometer. A Hewlett Packard 7673 autoinjector delivered 1.0 μL of the extract under splitless conditions onto a J&W Scientific, DB5-MS capillary column (60 m; 0.25 mm id; 0.25 μm film thickness) . The column was temperature programmed from 100°C to 300°C at 6°C/minute. The mass spectrum was scanned from 42 to 650 m/z in 0. /5 seconds. The injection and transfer oven temperatures were 275°C. The treatment ratio, T.R., was calculated as the ratio of moles of contaminant oxidized to the number of moles H₂0₂ consumed over the same time frame (i.e., ΔC1^"/ΔH₂0₂) . The number of moles of 2-chlorophenol oxidized was assumed to be stoichiometrically 1:1 to chloride ion measured in solution and corrected for background.

Results The concentration of total iron and manganese in commercially available granulated activated carbon varied significantly, from 24.0 to 5520 mg/Kg and 0.55-94.3 mg/Kg, respectively, as shown in Table 2. These data also indicated that total iron can be significantly increased, i.e., by a factor of 230-500.

Metals analyses data indicate that total Fe was significantly increased using the Fe precipitation method (Table 2) . Total Mn concentration on the GAC was significantly less than total Fe . Analysis of the GAC, via ICAP, involved a digestion step, and, therefore, the Fe and Mn concentration data represented total Fe and Mn rather than available Fe and Mn for reaction. It is unclear what fraction of the total Fe or Mn was available to react with H₂0₂. Although limited Mn concentrations were measured for the GAC reported here, analysis of other GACs involved in similar studies in laboratory of the present inventors indicates that total Mn concentrations are much higher in other GAC and provide a plausible, non-productive sink for

H₂0₂ (data not shown) .

Table 2 Metal Analysis Results of GAC via ICAP

* low, med. and high iron concentration on the GAC resulted from iron amendment

Addition of Fe to EXP4 carbon was performed for the purposes of enhancing oxidation and improving treatment efficiency. Three Fe concentrations were evaluated (low, medium, high) (Table 3) . The baseline total Fe concentration in EXP4 GAC was 24.0 mg/Kg. H₂0₂ degradation conformed to pseudo first-order kinetics and was greatest in the unamended (Fe) GAC. Despite the low Fe concentration on the unamended GAC, greater initial slurry pH contributed to increased H₂0₂ degradation. In the Fe amended GAC, H₂0₂ half-lives decreased, and total Cl^" recovery and TR increased with increased Fe concentration.

Table 3

Effect of GAC Iron Concentration on Treatment Ratio in

Calgon Chemical Corp. EXP4 GAC

Three Applications of 100 mL H₂0₂

( 1 ) [2CP] _± = 35 . 4 g/Kg

( 2 ) avg . (n = 3 ) [H₂0₂] _initial = 9222 , 9370 , 9420 , 9340 mg/L

(3 ) ΣCl^" released corrected for background Cl^"

(4 ) ΣC17∑H₂0₂ (avg . n = 3 )

The low initial slurry pH in Fe amended reactors is attributed to the acidity associated with FeS0₄- 7H₂0 used to alter the Fe content of the GAC. The pH decline with time may be attributed to different mechanisms: acid production associated with the Fenton mechanism; hydrogen ion release from the oxidation of 2CP; and production of acidic compounds, such as carboxylic acids. Since the solubility of ferrous Fe is inversely proportional to pH, some Fe may become soluble (mobile) under acidic conditions. Fe_τ measurec in unfiltered slurry samples (i.e., soluble or solid phase) containing GAC with 24.0 or 5500 Fe_τ mg/Kg was < 1.0 %, and with 9790 and 12050 Fe_τ mg/Kg, was 3 and 3.5 %, respectively.

Three H₂0₂ concentrations (0.94, 2.1, 7.1% w/w) in conjunction with Calgon Chemical Corp. GAC EXP4 (low Fe) were evaluated with respect to 2CP sorption and oxidation. The degradation rate of H₂0₂ conformed to pseudo-first order degradation kinetics and half-lives increased with increasing H₂0₂ concentration (Table 4) . The overall H₂0₂ degradation rate decreased 60-78% with increasing H₂0₂ application to the GAC. The decrease in H₂0₂ degradation rate may be partially attributed to the decrease in pH; however, the precise mechanism is unknown. Table 4

Effect of Hydrogen Peroxide Concentration on the

Adsorption and Oxidation Treatment Efficiency in Calgon

Chemical Corp. EXP4 GAC Low Iron

[Fe_τ]=5500 mg/Kg); 35.4 g/Kg 2CP; Three Applications of 100 mL H₂0₂

(1) ΣC1^" released corrected for background Cl^", percent chloride recovery from 2CP in parentheses

(2) ΣC1^"/ΣH₂0₂ for each application; average in parentheses An increase in [H₂0₂] will increase [-OH] linearly as per the source term in eq 2 (i.e., k₂ [Fe(II)] [H₂0₂] ) , and correspond to an increase in the rate of 2CP oxidation (eq 1) . However, increased radical scavenging may reduce treatment efficiency. For example, increased concentrations of scavengers such as Cl^" and H₂0₂ would increase the rate of scavenging (i.e., rxns 4 and 5) resulting in a nonlinear response in treatment efficiency. In these data, the TR increased between reactors containing 0.28 and 0.63 M H₂0₂, but either remained the same or decreased between reactors containing 0.63 M and 2.1 H₂0₂ (Table 5). The increase in TR was partially attributed to pH since the final pH was lower. The decrease in TR was attributed to increased scavenging from H₂0₂ (i.e., greater Si) . Reaction rate kinetics may have been limited during the last application of H₂0₂ (2.1 M) . Overall, the extent of 2CP oxidation, as indicated by total Cl^" recovery, increased with increased [H₂0₂] , but it was less efficient at the higher H₂0₂ concentration (2.1 ) . Table 5

Effect of 2CP Concentration on the Adsorption and Oxidation

Treatment Efficiency in Calgon Chemical Corp. EXP4 GAC

(1) ΣCl^" released corrected for background Cl^"

(2) 100 mL [H₂0₂]₁ = 0.21 M; H₂0₂ t_M = 3.5 (hr^"1)

(3) Three applications of 100 mL [H₂0₂] _x = 0.21 (4) ΣC1^"/ΣH₂0₂ (95% confidence interval; lower interval value for TR = ΣCL^" _upper/∑H₂P_21ower, where Cl^" _lower-Cl^" _upper, and H₂0_21ower-H₂0_2upper are the respective lower-upper 95% confidence interval concentration values for Cl^" and H₂0₂, respectively The TR is influenced by the initial 2CP concentration on the GAC (Table 5) . Where the initial H₂0₂ and Fe concentrations and the soil slurry pH (5.5) were equal in five reactors, the data suggest an increase in oxidation efficiency was attributed to greater 2CP concentration on the GAC. A statistically significant difference in TR was not observed at low concentrations (2.4-11.8 g/Kg) . A significant increase in the TR was observed in the reactor where the initial 2CP (131 g/Kg) and Fe (9790 mg/Kg) concentrations were increased and the average pH lower (3.1) . In this reactor, it cannot be specifically determined what affect elevated Fe and 2CP concentrations had on treatment efficiency, since the pH was lower. However, increased Fe and 2CP concentrations on the surface of the GAC would result in a greater probability of reaction between 2CP and -OH. GC/MS analysis of GAC (EXP4 med., Table 3) was performed to identify decomposition products resulting from the oxidation of 2CP. Controls were used to differentiate compounds extracted from GAC not attributed to 2CP or its oxidation products. Carbonate and unknown nitrogen derivatives (CND) were extracted from the GAC indicating background compounds (Table 6) . 2CP and CND were found in the GAC where 2CP was applied. Extraction of the Fe amended GAC yielded a tBDMS sulfate derivative (SD) in addition to 2CP and CND. Under oxidizing conditions, several organic acid byproducts were measured. The most abundant were confirmed to be oxalic and maleic acids, while minor acids were identified as malonic and fumaric. Based on the mass spectrum, tentatively identified compounds include another abundant compound, 2-chloromaleic acid, and minor compounds included hydroxychlorobenzoic acid, two isomers of dihydroxychlorobenzene and a dimer of chlorophenol . Measurement of these chlorinated byproducts indicates that under oxidative conditions, transformation of 2CP involved ring cleavage without the release of Cl^" . These results are consistent with Getoff and Solar (1986) who reported 2CP oxidation via -OH yields byproducts which may include Cl^" release, hydroxy benzenes (phenol, catechol, resorcinol, hydroquinone, hydroxyhydroquinone) , dimers, and chlorinated isomers .

Table 6 GC/MS Analysis of GAC⁽¹⁾

(1) Calgon Chemical Corp. EXP4 ; 40 mL 2CP 6880 μM or 40 mL DI ; three applications of 100 ml H₂0₂ or 100 mL DI

Sequential adsorption/oxidation was evaluated by adsorbing 2CP to the GAC (Calgon Chemical Corp. EXP4 , 5500 mg/Kg Fe) in three successive events using similar procedures (volume, concentration, equilibrium time) , and oxidizing the GAC suspension between sorption events. H₂0₂ (100 mL, 0.59 M) was applied twice to the GAC (119 g/Kg 2CP) and 62% of the Cl^" from the sorbed 2CP was recovered. 2CP was re-adsorbed (90 g/Kg 2CP) , H₂0₂ applied (100 mL, 2.9 M) and 125% of the Cl^" as 2CP re-amended to the GAC was recovered. This was due to residual Cl^" retained on the carbon after oxidation with 0.59 M H₂0₂. The overall Cl^" recovery from both oxidations was 89%. 2CP (97 g/Kg) was re-amended to the GAC. These data indicate that aggressive oxidation did not alter the GAC surface to a degree that significantly interfered with the 2CP adsorption reaction.

Contaminants in water are adsorbed onto sorbents and oxidized in the presence of iron, which may be present on the sorbent or added with the oxidant, via Fenton-driven reactions. The selection of sorbent affects treatment effectiveness, since the concentration of oxidant reactants, such as iron and manganese, varies between manufacturers of granulated activated carbon. The iron content of the sorbent can be altered to enhance the Fenton-driven oxidation reactions. The rate and extent of oxidation depends on oxidant concentration, which affects hydroxyl radical concentration and scavenging. The efficiency of oxidation increases with increased contaminant concentration on the surface of the sorbent .

The foregoing description of the specific embodiments will so fully reveal the general nature of the invention that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without undue experimentation and without departing from the generic concept, and, therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. The means, materials, and steps for carrying out various disclosed functions may take a variety of alternative forms without departing from the invention.

Thus the expressions "means to..." and "means for...", or any method step language, as may be found in the specification above and/or in the claims below, followed by a functional statement, are intended to define and cover whatever structural, physical, chemical or electrical element or structure, or whatever method step, which may now or in the future exist which carries out the recited function, whether or not precisely equivalent to the embodiment or embodiments disclosed in the specification above, i.e., other means or steps for carrying out the same function can be used; and it is intended that such expressions be given their broadest interpretation. REFERENCES

Aggarwal et al, J. Hazardous Materials 3 (27) : 301-314 (1991)

Buxton et al, J. Phys . Chem. Reference Data 2(17) .513-886 (1988)

Dorfman and Adams, National Bureau of Standards, Report No. NSRDS-NBS-46 (1973)

Gates and Siegrist, J. Environmental Engineering pp. 639-644 (Sep 1995)

Gethoff and Solar, Radiat . Phys. Chem., Int. J. Radiat . Appl . Instrum. Part C 5-6 (28) :443-450 (1986)

Haag and Yao, Environ. Sci. Technol. 5 (26) : 1005-1013 (1992)

Heberer and Stan, Anal. Chim. Acta 341:21-34 (1997)

Huling et al , Environ. Sci. Technol. 21 (32) : 3436-3441 (1998a)

Huling et al , J. Environ. Eng. Accepted (October, 1998b)

Ingles (1972)

Lipczynska-Kochany et al, Chemosphere 30(l):9-20 (1995)

Mawhinney, J. Chromatogr. 257:37-44 (1983)

Mawhinney et al, J. Chromatogr. 361:117-130 (1986)

Nicholls and Schonbaum, "Catalases" in Enzymes, Boyer et al, Eds., John Wiley: New York, Vol. 8, pp. 147-225 (1963)

O'Brien et al, "Implementation of a Funnel and Gate System", presented at International Containment Technology Conference and Exhibition, St. Petersburg, FL, February 9-12, 1997

Pardieck et al, J. Contaminant Hydrology 9(3):221-242 (1992)

Pignatello, Environ. Sci. Technol. 26 (5) : 944-951 (1992)

Ravikumar and Gurol, Environ. Sci. Technol. 3(28) :394-400 (1994)

Schumb et al, Hydrogen Peroxide , American Chemical Society Monograph Series, Reinhold Publishing Corporation New York, NY, p. 759 (1955)

Shetiya et al, Indian J. Chem. 14A:575-578 (1976)

USEPA, 1991 Subsurface Contamination Reference Guide

Walling, Accounts of Chemical Research 8:125-131 (1975) Watts et al, Hazardous Waste and Hazardous Materials 4 (7) :335-345 (1990)

Watts et al, "Treatment of Contaminated Soils Using Catalyzed Hydrogen Peroxide" in Chemical Oxidations : Technology for the Nineties, W. W. Eckenfelder et al (Eds) (1991)

Watts et al, Water Environ. Research 7 (65) : 839-844 (1993)

Yeh and Novak, Water Environ. Research 67 (5) : 828-834 (1995)

Claims

WHAT IS CLAIMED IS:

1. A process for treating contaminants in water comprising : passing water containing contaminants through a sorbent amended with or containing iron to which the contaminants adsorb and are concentrated; adding an oxidant to the sorbent .

2. A process for treating contaminants in water comprising: passing water containing contaminants through a sorbent to which the contaminants adsorb and are concentrated; adding a composition comprising iron and an oxidant to the sorbent .

3. The process according to either of claims 1 or 2 , wherein the sorbent is amended with iron selected from the group consisting of ferrous iron and ferric ion.

4. The process according to either of claims 1 or 2 , wherein the sorbent is activated carbon.

5. The process according to either of claims 1 or 2 , wherein the oxidant is hydrogen peroxide.

6. The process according to either of claims 1 or 2 , wherein the iron is selected from the group consisting of ferrous iron and ferric ion.

7. The process according to either of claims 1 or 2 , wherein the water containing contaminants is treated above ground .

8. The process according to either of claims 1 or 2 , wherein the water containing contaminants is treated below ground .

9. A composition for treating contaminants in water by concentrating said contaminants and oxidizing said contaminants comprising a sorbent which is amended with iron.

10. The composition according to claim 9, wherein said iron is selected from the group consisting of ferric iron and ferrous iron.

11. The composition according to claim 9, wherein the sorbent is activated carbon.

12. The composition according to claim 9, wherein the iron is chelated.

13. An apparatus for treating contaminated water comprising a bed of sorbent amended with iron and means for supplying an oxidant to said bed.