Geosynthetic Clay Liners Containing Bentonite Polymer Nanocomposite

Geo-Frontiers 2011 © ASCE 2011

Geosynthetic Clay Liners Containing Bentonite Polymer Nanocomposite J. Scalia1, S.M. ASCE, C. H. Benson2, Ph.D., P.E., F. ASCE, D.GE, T. B. Edil3, Ph.D., P.E., F. ASCE, D.GE, G.L. Bohnhoff4, S.M. ASCE, and C.D. Shackelford5, Ph.D., P.E., M.ASCE 1

Graduate Research Assistant, Geological Engineering, University of Wisconsin at Madison, Madison, WI; PH (608) 265-4491; email: [email protected] 2 Wisconsin Distinguished Professor and Chair, Geological Engineering, University of Wisconsin at Madison, Madison, WI; PH (608) 262-7242; email: [email protected] 3 Professor, Geological Engineering, University of Wisconsin at Madison, Madison, WI; PH (608) 262-3225; email: [email protected] 4 Graduate Research Assistant, Civil and Environmental Engineering, Colorado State University, Fort Collins, CO; PH (970) 491-3895; email: [email protected] 5 Professor, Civil and Environmental Engineering, Colorado State University, Fort Collins, CO; PH (970) 491-5051; email: [email protected] ABSTRACT Bentonite was modified at the nanoscale so that low hydraulic conductivity would be maintained under adverse conditions. Nanoscale modification consisted of polymerizing acrylic acid within a bentonite slurry to form a stable interconnected structure. This modified material is referred to as a bentonite-polymer nanocomposite (BPN). The BPN was then air dried and ground to simulate the granule-size distribution of granular bentonite in geosynthetic clay liners (GCLs). Index property tests indicate that BPN has different behavior than natural sodium bentonite (Na-bentonite). For example, free swell tests with natural Na-bentonite swelled to approximately 30 mL in deionized water, whereas BPN swelled to more than 70 mL in the same solution. GCLs were assembled and directly permeated with a range of calcium chloride (CaCl2) solutions known to cause large increases in the hydraulic conductivity (k) of natural Na-bentonite. Low hydraulic conductivities (k < 3x10-10 cm/s) were maintained by BPN for all solutions tested. In contrast, natural Na-bentonite tested under the same conditions had k more than 4 orders of magnitude higher (k > 2x10-5 cm/s). Additional study of the properties and the underlying mechanisms of BPN are ongoing. INTRODUCTION Sodium bentonites (Na-bentonites) are used for barriers for waste containment because they have low hydraulic conductivity (k) to water (typically k < 10-8 cm/s) and exhibit membrane behavior, the latter giving rise to hyperfiltration, chemicoosmotic flow, and reduced diffusion (Malusis et al. 2003). However, the effectiveness of Na-bentonite as a barrier for waste containment depends on the

2001


valence of the cations dominating the exchange complex (i.e., the collection of cations adsorbed on the mineral surface). When the exchange complex is dominated by monovalent cations, such as Na+, the montmorillonite (MMT) dominating the mineralogy of bentonite swells appreciably due to osmotic adsorption of water molecules between the MMT platelets (McBride 1994). Osmotic swelling reduces the size of paths that conduct flow and transport, resulting in low k. In contrast, if multivalent cations (i.e., valence ≥ 2+) dominate the exchange complex, osmotic swelling of MMT does not occur, resulting in much higher k (Shackelford et al. 2000, Jo et al. 2001, 2005, Kolstad et al. 2004). Unfortunately, Na-MMT is thermodynamically unstable in environments where multivalent cations (e.g., Ca2+) are present (Sposito 1989), including most naturally occurring pore waters. When present, multivalent cations replace monovalent cations originally in the exchange complex, thereby reducing or eliminating osmotic swell and the ability of the bentonite to function effectively (Kolstad et al. 2004). Monovalent solutions of high concentration can also limit osmotic swell (McBride 1994). Numerous laboratory studies have shown that Ca2+ for Na+ exchange results in reduced swelling capacity of the original Na-bentonite in geosynthetic clay liners (GCLs) upon hydration, and ultimately to decreased hydraulic performance (Lin and Benson 2000, Jo et al. 2001, 2005, Egglofstein 2001, Shackelford and Lee 2003, Kolstad et al. 2004, Lee and Shackelford 2005). Cation exchange often occurs slowly because the rate of exchange is controlled by the rate at which multivalent cations diffuse from the bulk pore water into the interlayer space (Jo et al. 2005, 2006). Field studies have confirmed that GCLs will undergo Ca2+ for Na+ exchange (ATU 1992, James et al. 1997, Benson et al. 2004, 2007, Meer and Benson 2007, Scalia and Benson 2010). The Ca2+ is derived from surrounding soils, and migrates into the GCL in response to hydraulic (suction) and/or chemical (diffusive) gradients. Even GCLs installed in the field with an overlying geomembrane (GM) have been shown to undergo Ca2+ for Na+ exchange (Meer and Benson 2007, Scalia and Benson 2010). Thus, in the long term, exchange of Ca2+ for Na+ should be anticipated. Trauger and Darlington (2000) developed a bentonite-polymer nanocomposite (BPN) by polymerizing an organic monomer in a bentonite slurry. This BPN slurry was deposited into a needle-punched non-woven geotextile (GT) and permeated with seawater. The hydraulic conductivity of the BPN-GT was 5x10-10 cm/s compared to 2x10-6 cm/s for a traditional GCL. Analysis of dried BPN showed an increase in the inter-particle spacing from 0.35 nm to from 1.0 to 1.5 nm, indicating that polymer was intercalated in the bentonite interlayer. The anionic polymer was hypothesized to bond with the Na+ ions in the bentonite interlayer. In this study, BPNs were manufactured with poly(acrylic acid) (PAA). PAA was chosen because PAA should interact with bentonite via hydrogen bonding, is relatively inexpensive, is similar to the super-absorbent and super-swelling polymers used in baby diapers (Buchholz 1998), and is relatively safe. The BPN and Nabentonite were tested for k and swell index in aqueous solutions having a range of Ca2+ concentrations. For k testing, GCLs composed of a thin layer of Na-bentonite or BPN sandwiched between two non-woven geotextiles were permeated under low stress condition.

2002


2003

MATERIALS AND METHODS Bentonite Natural Na-bentonite from Wyoming, USA was used for comparative testing with the BPN. The bentonite was GCL grade and was provided by a GCL manufacturer. The granule-size distribution of the Na-bentonite used in this study, and a granular bentonite from a new GCL studied by Shackelford et al. (2000), are shown in Figure 1. The Na-bentonite in this study has a similar range of granule sizes as the GCL bentonite studied by Shackelford et al. (2000), but has a smaller fraction of fine sand-size particles. The BPN used in this study was ground and screened to match a GCL as defined in Shackelford et al. (2000). 100 BPN Na-bentonite GCL bentonite (Shackelford et al. 2000)

Percent finer (%)

80

60

40

20

0 10

1

0.1

0.01

Granule size (mm)

Figure 1. Granule-size distributions of granular BPN, granular Na-bentonite, and granular bentonite removed from a GCL by Shackelford et al. (2000). Error bars are maximums and minimums from duplicate testing. Bentonite Polymer Nanocomposite BPNs containing PAA were created using a method comparable to that employed for conventional polymer nanocomposites (e.g., Muzny et al. 1996). Powdered Na-bentonite was mixed into a monomer solution prepared by dissolving acrylic acid in water, followed by slow neutralization with sodium hydroxide to allow dissipation of the heat of neutralization. Sodium persulfate was added as an initiator. Bentonite was added to the neutralized solution in concentrations between 30-50% by weight to form a bentonite-monomer slurry. During bentonite addition, the solution was vigorously agitated to increase the surface area available for PAA adsorption. Once the bentonite-monomer slurry was formed, polymerization was initiated by raising the temperature of the slurry above the decomposition temperature of the


initiator. This caused the initiator to decompose and form one or more free radicals. During polymerization, free radicals (R•) attack the double bond of the acrylic monomer to form new free radicals (RM•), which then react with additional monomers to propagate the polymer chain (RMMM•). After polymerization, the solution was oven dried. The resulting BPN was milled and screened. Liquids Both bentonite and BPN were evaluated with deionized (DI) water, 5, 50 and 500 mM calcium chloride (CaCl2) solutions. Lee et al. (2005) and Jo et al. (2001) showed that permeating GCLs with DI water or 5 mM CaCl2 results in low k of ~2x10-8 cm/s. In contrast, 50 mM and 500 mM CaCl2 solutions were shown by Lee et al. (2005) to result in k on the order of 10-6 or 10-5 cm/s. Thus DI water and 500 mM CaCl2 were selected to provide a base-line condition, and 50 and 500 mM CaCl2 were chosen to stress the BPN barrier layer. CaCl2 solutions were prepared by dissolving dihydrate calcium chloride salt (CaCl2·2H2O) in DI water. Concentrations of Ca2+ were verified by inductively coupled plasma-optical emission spectroscopy (ICP-OES) following USEPA Method 6010 B. The concentration of Ca2+ in DI water was below the ICP-OES method detection limit (MDL) of 0.05 mM Ca2+). Swell Index Swell index (SI) tests were performed on bentonite and BPN in general accordance with ASTM D 5890. Materials were ground to 100% passing a 200 mesh U.S. Standard Sieve with a mortar and pestle DI water as well as 5, 50, and 500 mM CaCl2 were used as hydration solutions. Swell indices are reported in Table 1.

2004


Table 1. Properties of bentonite and bentonite polymer nanocomposite (BPN) at varying calcium chloride (CaCl2) concentrations. Data are the average of 2 or more duplicate tests. Bentonite BPN CaCl2 Swell Hydraulic Swell Hydraulic concentraiton index conductivity index conductivity (mM) (mL/2 g) (cm/s) (mL/2 g) (cm/s) -9 < 0.05 30.5 1.7x10 73.0 1.9x10-10 -9 5 28.5 1.8x10 46.5 1.8x10-10 50 10.0 2.4x10-5 19.0 2.8x10-10 -5 500 8.5 4.5x10 7.0 1.3x10-10 Hydraulic Conductivity Hydraulic conductivity tests were conducted with bentonite or BPN using flexible-wall permeameters in general accordance with ASTM D 5084. The falling headwater-constant tailwater method was employed. Glass tubing with 5.2-mm inside diameter was used as the falling headwater reservoir. The tubing was fixed at 5° from horizontal to minimize the change in hydraulic gradient during testing. Gravity heads were used to apply cell and influent pressure. Backpressure was not applied to allow convenient collection of effluent for chemical analysis. Specimens were tested under an average effective stress of approximately 20 kPa and an average hydraulic gradient of 200. This hydraulic gradient is higher than in the field, but is typical for GCL testing (Shackelford et al. 2000). Test specimens of the GCL were constructed with granular bentonite or BPN in flexible-wall permeameters. Geotextiles (GTs) were used in lieu of porous stones and filter paper. A latex membrane was attached to a 154-mm-diameter base pedestal with O-rings. A 0.75-kg/m2 non-woven needle-punched GT was then placed on the base pedestal, and topped with a heat-bonded non-woven GT. The bentonite or BPN was then placed with a dry mass per unit area of 4.8 kg/m2 and manually leveled. Two GTs and a top pedestal were then placed atop the bentonite or BPN layer mirroring the lower assemblage. The top pedestal was then secured to the latex membrane with O-rings. In-cell hydration was conducted for 48 h prior to flow. After the permeameter was assembled and connected to the falling headwater apparatus, cell pressure was applied and all tubing was saturated with the permeant liquid. The inflow line of the permeameter was then opened to allow the specimen to hydrate while the effluent line remained closed. After 48 h, the lines were flushed to remove any air bubbles and flow was initiated by opening the effluent line. During testing, effluent was collected in fluorinated ethylene propylene (FEP) bags and sampled for effluent chemistry.

2005


2006

RESULTS AND DISCUSSION Hydraulic Conductivity Hydraulic conductivities of GCLs constructed from Na-bentonite or BPN are summarized in Table 1 and shown versus permeant CaCl2 concentration in Figure 2. Equilibrium k from long-term testing of GCLs by Lee et al. (2005) are also shown in Figure 2 for comparison. The k of the BPN is low (< 2.8x10-10 cm/s) and independent of CaCl2 concentration within the range of concentrations tested. In contrast, GCLs constructed with conventional bentonite had k similar to GCLs tested by Lee et al. (2005). 10-3 Na-Bentonite BPN Na-Bentonite (Lee et al. 05)

10-5 10-6 10-7 10-8

MDL Ca2+

Hydraulic Conductivity (cm/s)

10-4

10-9 10-10 10-11 10-3

10-2

10-1

100

101

102

103

Solution CaCl2 Concentration (mM)

Figure 2. Hydraulic conductivity of Na-bentonite, BPN, and Na-bentonite GCLs tested by Lee et al. (2005). Results based on permeation with deionized water are plotted at the ICP-OES method detection limit (MDL). Data are the average of 2 or more duplicate tests. Swell Index Swell index is shown as a function of the CaCl2 concentration of the hydrating solution in Figure 3a. In DI water, BPN exhibited more than twice the swell index compared to that of Na-bentonite (i.e., 73.0 vs. 30.5 mL/2 g). Similar to Nabentonite, however, the BPN swell index decreased with increasing CaCl2 concentration. At a 500 mM CaCl2 concentration, the swell index of BPN was similar to that of Na-bentonite (7.0 vs. 8.5 mL/2 g), and in the range typical of that for calcium bentonite ( 2x10-5 m/s) when permeated with 50 and 500 mM CaCl2. Additional study of the properties and the underlying mechanisms of BPN is ongoing. ACKNOWLEDGEMENTS Financial support for this study was provided by National Science Foundation (Grant No. CMMI-0758334) and CETCO Inc. through NSF's Grant Opportunities for Academic Liaison with Industry (GOALI) program. Funding for Benson was provided by his Wisconsin Distinguished Professorship. The findings and recommendations in this paper are solely those of the authors, and do not necessarily represent the policies or opinions of the sponsors. Endorsement by the sponsors is not implied and should not be assumed. REFERENCES ATU (1992). “ClayMax sodium bentonite liner found degraded at New Jersey site.” Aboveground Tank Update, 3(2), 1, 17-18. Benson, C., Jo, H., and Abichou, T. (2004). “Forensic analysis of excessive leakage from lagoons lined with a composite GCL.” Geosynthetics International, 11(3), 242-252. Benson, C., Thorstad, P., Jo, H., and Edil, T. (2007). “Case history: hydraulic performance of geosynthetic clay liners in a landfill final cover.” J. of Geotech. and Geoenviron. Engr., 133(7), 814-827. Buchholz, F. (1998). Absorbency and Superabsorbency, in Modern Superabsorbent Polymer Technology, F. Buchholz and A. Graham, eds., John Wiley, New York, 1-18. Egloffstein, T. (2001). “Natural bentonites-influence of the ion exchange and partial desiccation on permeability and self-healing capacity of bentonites used in GCLs.” Geotextiles and Geomembranes, 19, 427-444. James, A., Fullerton, D., and Drake, R. (1997). “Field performance of geosynthetic clay liner under ion exchange conditions.” J. of Geotech. and Geoenviron. Engr., 123(10), 897-901. Jo, H., Benson, C., and Edil, T. (2006). “Rate-limited cation exchange in thin bentonitic barrier layers.” Canadian Geotech. J., 43, 370-391

2008


Jo, H., Benson, C., Shackelford, C., Lee, J., and Edil, T. (2005). “Long-term hydraulic conductivity of a non-prehydrated geosynthetic clay liner permeated with inorganic salt solutions.” J. of Geotech. and Geoenviron. Eng., 131(4), 405-417. Jo, H., Katsumi, T., Benson, C., and Edil, T. (2001). “Hydraulic conductivity and swelling of non-prehydrated GCLs permeated with single species salt solutions.” J. of Geotech. and Geoenviron. Engr., 127(7), 557-567. Kolstad, D., Benson, C., and Edil, T. (2004). “Hydraulic conductivity and swell of nonprehydrated GCLs permeated with multi-species inorganic solutions.” J. of Geotech. and Geoenviron. Engr., 130(12), 1236-1249. Lee, J. and Shackelford, C. (2005). “Impact of bentonite quality on hydraulic conductivity of geosynthetic clay liners.” J. of Geotech. and Geoenviron. Engr., 131(1), 64-77. Lee, J., Shackelford, C., Benson, C., Jo, H.. and Edil, T. (2005). “Correlating index properties and hydraulic conductivity of geosynthetic clay liners.” J. Geotech. and Geoenviron. Engr., 131(11), 1319-1329. Lin, L. and Benson, C. (2000). “Effect of wet-dry cycling on swelling and hydraulic conductivity of geosynthetic clay liners.” J. of Geotech. and Geoenviron. Engr., 126(1), 40-49. Malusis, M., Shackelford, C., and Olsen, H. (2003). “Flow and transport through clay membrane barriers.” Engineering Geology, 70(2-3), 235-248. McBride M. (1994). Environ. Chemistry of Soils. Oxford University Press, New York. Meer, S. and Benson, C. (2007). “Hydraulic conductivity of geosynthetic clay liners exhumed from landfill final covers.” J. of Geotech. and Geoenviron. Engr., 133(5), 550-563. Muzny, C., Butler, B., Hanley, H., Tsvetkov, F., and Pfeiffer, D. (1996). “Clay platelet dispersion in a polymer matrix.” Materials Letters, 28(4-6), 379-384. Scalia, J. and Benson, C. (2010). “Hydraulic conductivity of geosynthetic clay liners exhumed from landfill final covers with composite barriers.” J. of Geotech. and Geoenviron. Engr., in press. Shackelford, C., Benson, C., Katsumi, T., Edil, T., and Lin, L. (2000). “Evaluating the hydraulic conductivity of GCLs permeated with non-standard liquids.” Geotextiles and Geomembranes, 18(2-3), 133-161. Shackelford, C. and Lee, J. (2003). “The destructive role of diffusion on clay membrane behavior.” Clays and Clay Minerals, 51(2), 187-197. Sposito, G. (1989). The Chemistry of Soils, Oxford University Press, New York. Trauger R. and Darlington J. (2000). “Next generation geosynthetic clay liners for improved durability and performance." 14th GRI Conference, Geosynthetic Institute, Folsom, PA, USA, 255-267.

2009