Environ. Sci. Technol. 2008, 42, 9296–9301
Compatibility of Polymers and Chemical Oxidants for Enhanced Groundwater Remediation M E G A N M . S M I T H , * ,† J E F F A . K . S I L V A , ‡ J U N K O M U N A K A T A - M A R R , †,‡ A N D J O H N E . M C C R A Y †,‡ Hydrologic Science & Engineering Program and Environmental Science & Engineering Division, Colorado School of Mines, Golden, Colorado 80401
Received March 16, 2008. Revised manuscript received September 30, 2008. Accepted October 3, 2008.
Polymer floods provide a promising method to more effectively deliver conventional groundwater treatment agents to organic contaminants distributed within heterogeneous aquifer systems. Combinations of nontoxic polymers (xanthan and hydrolyzed polyacrylamide) and common chemical oxidants (potassium permanganate and sodium persulfate) were investigated to determine the suitability of these mixtures for polymer-enhanced in situ chemical oxidation applications. Oxidant demand and solution viscosity were utilized as initial measures of chemical compatibility. After 72 h of reaction with both test oxidants, solution viscosities in mixtures containing hydrolyzed polyacrylamide were decreased by more than 90% (final viscosities ∼2 cP), similar to the 95% viscosity loss (final viscosities ∼1 cP, near that of water) observed in xanthan/ persulfate experiments. In contrast, xanthan solutions exposed to potassium permanganate preserved 60-95% of initial viscosity after 72 h. Permanganate depletion in xanthancontaining experiments ranged from 2% to 24% over the same test period. Although oxidant consumption in xanthan/ permanganate solutions appeared to be correlated with increasing xanthan concentrations, solutions of up to 2000 mg/L xanthan did not inhibit permanganate from oxidizing a dissolved-phase test contaminant (tetrachloroethene, PCE) in xanthan solution. These advantageous characteristics (high viscosity retention, moderate oxidant demand, and lack of competitive effects on PCE oxidation rate) render xanthan/ permanganate the most compatible polymer/oxidant combination of those tested for remediation by polymer-enhanced chemical oxidation.
Introduction The addition of a polymer flood to in situ chemical oxidant (ISCO) treatments at contaminated groundwater sites may increase the effectiveness of delivery of the oxidant to a larger region of the aquifer, especially in zones of lower permeability, if the increased viscosity of the polymer solution can be maintained. Aquifer heterogeneities (composed of regions with physically different medium properties) tend to focus flow through higher-permeability zones, causing bypassing * Corresponding author:
[email protected]; 303.384.2095; 303.273.3413 (fax). † Hydrologic Science & Engineering Program. ‡ Environmental Science & Engineering Division. 9296
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 24, 2008
of lower-permeability areas that may retain dissolved-phase contaminants and contribute to contaminant “rebound” after remediation. Shear-thinning polymer solutions can more thoroughly distribute injectate among layers of differing permeability (1, 2), creating a more uniform advancing injection front and potentially increasing contact between coinjected remediation agents and dispersed contaminants. Polymer enhancement of traditional groundwater remediation schemes could thus significantly improve treatment efficiency as a result of the increased “sweep area,” or amount of the aquifer contacted by the polymer-amended injectate. Outside of the oil industry’s varied success in utilizing polymer floods during enhanced oil recovery (2-4), a survey of the literature shows relatively few applications of polymer floods in environmentally relevant (and shallow) aquifer systems. These studies have been limited to the coupling of polymer and surfactant-based remediation methods (5-7). For example, sandbox experiments demonstrated the ability of shear-thinning xanthan polymer floods to more evenly distribute a previously injected surfactant solution throughout a dual-permeability field (5). This same technique of in-series surfactant and polymer flooding was utilized at a DNAPL-contaminated aquifer site with free product present (6). Robert et al. (7) measured increased contaminant recovery and provided further visual evidence of increased contact between a contaminant (trichloroethene, in pools and residual saturation) and a coinjected xanthan/surfactant mixture in a heterogeneous system with 1 order of magnitude variation in permeability. However, the coupling of a polymer flood with a destructive remediation technique (such as ISCO) has yet to be investigated, presumably due to the assumption that the remedial agent would degrade the polymer. Although chemical oxidants are effective at destroying numerous aqueous organic contaminants (8), the long-term effectiveness of oxidants in heterogeneous media has been questioned because postremediation rebound of contaminant concentrations has been observed at many field sites, presumably due to contaminant diffusion from untreated regions (9). Thus the increased sweep efficiency that polymer floods provide can aid the delivery of chemical oxidants to lower-permeability layers and increase the effectiveness of remediation efforts in heterogeneous porous media. A demonstration of basic polymer/oxidant stability is necessary to first show that the proposed polymer-enhanced treatments are viable. Our polymer/oxidant compatibility experiments utilized xanthan and hydrolyzed polyacrylamide (HPAM) solutions coupled with the chemical oxidants potassium permanganate (KMnO4) and sodium persulfate (Na2S2O8). These particular polymers were chosen for their nontoxic character, high solubility in water, and previous history of application in either petroleum or agricultural settings (4, 10). Both are anionic polymers, and they each exhibit the shear-thinning rheology (wherein viscosity is reduced under high-shear conditions and increased in lowshear conditions) necessary for improved injectability and sweep-efficiency enhancement. Additional characteristics of these polymers, including molecular structures, can be found in Figure S1a,b and Table S1 in Supporting Information. The test oxidant potassium permanganate was selected due to its documented history of use (11). The extensive research concerning both the details of permanganate reactions with various contaminants as well as possible disadvantages of its use (8, 12, 13) makes it a natural candidate for experimentation. The second test oxidant, sodium persulfate, was chosen not simply because of its growing popularity (due to lack of solid oxidation byproducts) but also to contrast a 10.1021/es800757g CCC: $40.75
2008 American Chemical Society
Published on Web 11/08/2008
TABLE 1. Experimental Conditions for Batch Compatibility Tests sample descriptions polymer control polymer/salt control polymer/oxidant test oxidant control
[polymer] (mg/L)
[oxidant] [added (mg/L) cations]a (mM)
160, 800, 1600 0 160, 800, 1600 0 160, 800, 1600 2000b 0 2000b
0 12.6, 8.4 0 0
a First value refers to added K+ ions; second refers to added Na+ ions. b Experiments contained either 12.6 mM KMnO4or 8.4 mM Na2S2O8.
different type of oxidation mechanism: while persulfate oxidation proceeds through free-radical production, permanganate generally reacts through electron transfer processes (14). Two primary screening criteria were considered as measures of compatibility and suitability for potential application: the stability of polymer/oxidant solution viscosity through time, and the level of increased oxidant demand placed on the oxidant by the polymers. A significant loss of viscosity would prevent the polymer solution from delivering oxidant to contaminated low-permeability zones, while an increased oxidant demand would reduce the amount of oxidant available for contaminant degradation. Any polymer/ oxidant combinations judged stable by these standards were then subjected to a secondary criterion: determining the competitive effect, if any, that polymer poses to the degradation rate of a test contaminant (here, tetrachloroethene).
Experimental Methods Materials. Hydrolyzed polyacrylamide (Superfloc) powders of varying molecular weight and charge density were obtained from Cytec Corp., and xanthan powders with and without a dispersant additive (Xanvis and Keltrol T, respectively) were obtained from CP Kelco Oil Field Group, Inc. and CP Kelco. Potassium permanganate (KMnO4) was obtained at 99+% purity from Cairox Corp., and sodium persulfate (Na2S2O8, also 99+% purity) was obtained from Sigma-Aldrich. Potassium chloride, sodium chloride, and tetrachloroethene (PCE, g99.9%) were obtained from Sigma-Aldrich. Methods. Polymer stock solutions of 2 g/L were prepared by slowly adding solid powders to water at room temperature (22 ( 2 °C). According to manufacturers’ instructions, solutions were continuously stirred for 2-3 h with an overhead mixer, allowing at least 2 h of rest or “aging”, before dilution to desired concentrations. Fresh solutions were prepared prior to each experiment and used quickly to avoid microbial growth. Deionized water was used as the solvent (except where noted) in order to prevent additional oxidant demand in the solutions. Stock solutions of both oxidants were created at 10 g/L concentration in deionized water, protected from light and heat, and used within 2 weeks. Batch tests containing stock solutions of each polymer formulation with each oxidant were created by use of 4:1 mixing ratios by volume (see details in Table 1). Tests were conducted in duplicate except where noted, in 40-mL amber glass volatile organic analysis (VOA) vials with poly(tetrafluoroethylene) (PTFE) -lined caps, which were continuously agitated and stored in darkness in a 22 ( 2 °C environment. Initial oxidant concentration was fixed at 2 g/L, while polymer concentrations were varied over a range of 160, 800, and 1600 mg/L (providing 3 orders of magnitude initial viscosity values, from 5 to 1250 centipoise (cP); see Figure S2 in Supporting Information for full data set). This range actually encompasses viscosities that are considered higher than “reasonable” (roughly estimated in the 10-50 cP range) for subsurface injection (6), but these higher values
were used in expectation of viscosity reductions after oxidant exposure. The oxidant concentrations used in these experiments are within range of those used in previous laboratory investigations (e.g., refs 15-17), although field applications of KMnO4 can vary from as low as 250 mg/L to saturation (∼60 g/L) (18). However, recent research suggests that lower oxidant concentrations, such as those used in this work, may be more efficient for contaminant mass destruction if delivered more effectively (19). The concentrations used here are high enough to permit molar oxidant concentrations orders of magnitude greater than molar polymer concentrations (as a result of the high formula weight of the polymers). Therefore, potentially deleterious effects of the oxidants on polymer properties should be observed at this mixing ratio. Oxidant concentrations were monitored at 4, 24, 48, and 72 h intervals, except where noted; a Hach DR-4000 spectrophotometer was used at 525 nm for permanganate and 450 nm for persulfate (20) quantification, and daily calibrations were run for each oxidant. Two to three individual samples were taken from each duplicate experiment at every measurement interval, and all measurements agreed within 10% variation. Although visible manganese dioxide solid generation was not observed during experiments, solutions were passed through 0.2 µm filters prior to analysis to avoid interference. Pure polymer solution absorbance was limited to lower wavelengths (∼190-250 nm) than those required for oxidant analysis and so did not interfere with oxidant quantification. The measured absorbance of pure polymer solutions at 525 and 450 nm generated negligible background values (
