Makers of OilSorb™ and Other State-of-the-Art Filtration Media
|
Makers of OilSorb™ and Other State-of-the-Art Filtration Media
|
|
|
Evaluation of Two Organoclays, Clinoptilolite and Hydroxy-Apatite as Sorbents for Heavy Metal Removal from Water
George R. Alther* Biomin, Inc.
Fred D. Tillman, Jr. and James A. Smith Program of Interdisciplinary Research in Contaminant Hydrogeology Department of Civil Engineering, University of Virginia *Corresponding author Running title: Sorbents for heavy metal removal from water Abstract A series of column tests where performed with four sorbent media to determine their effectiveness for the removal of heavy metals from water. The sorbents are an organoclay, an organoclay/anthracite blend, clinoptilolite (a zeolite), and hydroxy-apatite. Heavy metals studied were Ni, Cr, Cu, Cd, and Zn. Four column tests were conducted using an influent solution containing all four metals (one test for each sorbent). Four column tests were conducted using an influent solution containing only Zn (one for each sorbent), and two column tests were conducted using an influent solution containing only Ni (with clinoptilolite and hydroxy-apatite only). For the Zn-hydroxy-apatite experiment, the inflow solution also contained 200 mg/L Ca++ to simulate hard water. Aside from ion-selective resins, hydroxy-apatite is the only sorbent that is believed to retain its sorption capacity for metals in hard water. Metal effluent breakthrough data indicated that the organoclay and the organoclay/anthracite blend removed all heavy metals, but only to a limited extent, with effluent metal concentrations equal to 50% of the influent concentration appearing after 2 to 3 pore volumes of flow. These sorbents have the unique capability of removing both heavy metals and nonpolar organic pollutants from water simultaneously. Hydroxy-apatite demonstrated the greatest removal of metals from solution. Pore volumes of flow required to produce effluent concentrations equal to 50% of the influent concentration ranged between 50 (Ni) and 160 (Cu). Although hydroxy-apatite performed best for metal removal, it should be noted that clinoptilolite is superior to most media for ammonia removal.
Key terms: heavy metals, organoclay, anthracite, hydroxy-apatite, clinoptilolite, zeolite
Introduction
Sorption of heavy metals by sorbents such as bentonite, organically modified bentonite (organoclay), clinoptilolite (zeolite), and hydroxy-apatite, has been studied previously using batch experiments (Alther, 1998; Alther, 2000; Alther, 2001; Alther and Ellis, 2001a,b). Cation exchange capacities of standard sorbent media typically range between 70 meq/100 g for bentonite to 600 meq/100 g for synthetic zeolites. Organoclays typically have much lower cation exchange capacities than bentonite owing to the exchange of relatively large, organic quaternary ammonium cations onto the clay that are not easily displaced by ammonium cations. Mumpton (1981) presents the results of several studies on the sorption capacity of zeolites.
Organoclays are particularly effective at removing relatively non-polar organic pollutants such as oil, polychlorinated biphenyls, polycyclic aromatic hydrocarbons, chlorinated solvents, and gasoline hydrocarbons from water (Alther, 2002). When used as a pretreatment for activated carbon systems, organoclays can increase the life of the activated carbon by 7 to 10 times. However, little is known about the ability of organoclays to remove heavy metals from solution. In addition, there are relatively little data that quantify the sorption of metals to organoclays, zeolites, and hydroxy-apatite in flow-through, column reactors (which mimic field treatment systems).
Given the strict regulatory requirements on heavy-metal discharges to the environment, it is important to develop cost-effective sorbents that can treat metal-contaminated waters (or in some cases, water contaminated with both metals and organic pollutants). In this study, we evaluate the suitability of four sorbents (an organoclay, an organoclay/anthracite blend, a zeolite (clinoptilolite), and hydroxy-apatite for the removal of heavy metals from water. Experiments are performed with combinations of metals and with single metals in flow-through columns to better assess real world performance and possible kinetic sorption limitations.
Experimental Materials and Methods
Initially, all five metals were combined in one stock solution, with later experiments involving two of the metals individually. For all experiments, a column constructed of 30-inch long (76.2 cm) by 3-inch diameter (7.62 cm) poly-vinyl-chloride (PVC) and fitted with flexible couplings, reducing bushings and NPT thread-to-5/8” hose fittings was used. Two layers of circular type-304 stainless steel woven cloth (20x20 mesh) were placed inside each end-fitting to suspend the sorbent inside the column and prevent sorbent material from entering the tubing. A 55-gallon (208.2 L) self-supporting polyethylene tank was used to prepare and hold the inflow metal solution. A Master-Flex peristaltic pump was used to force the aqueous-metal solution up through the column to displace void-space air and ensure maximum contact with the sorbent material. Figure 1 shows a picture of the experimental column system.
Each packed column was back-flushed with clean water several times prior to each experiment in order to remove fines associated with the sorbent. After the fines where removed, the physical properties of each packed column were measured. For the organoclay/anthracite column, the porosity, pore volume, sorbent mass, experimental flow rate, and water residence time in the column were 0.377, 1.3 L, 2.9 kg, 160 mL/min, and 8.3 min respectively. For the organoclay column, the porosity, pore volume, sorbent mass, experimental flow rate, and water residence time in the column were 0.359, 1.2 L, 2.9 kg, 156 mL/min, and 8.0 min respectively. For the clinoptilolite column, the porosity, pore volume, sorbent mass, experimental flow rate, and water residence time in the column were 0.329, 1.4 L, 2.6 kg, 72 mL/min, and 20 min respectively. For the hydroxy-apatite column, the porosity, pore volume, sorbent mass, experimental flow rate, and water residence time in the column were 0.42, 1.3 L, 2.9 kg, 160 mL/min, and 8.3 min respectively.
The aqueous-metal inflow solution was prepared using dry-reagent compounds. In the first set of four experiments, a reservoir containing 60 mg/L Cd, 150 mg/L Cr, 120 mg/L Cu, 60 mg/L Ni and 60 mg/L Zn was pumped through columns containing the sorbents to be studied. In later experiments, a 785 mg/L Zn solution was pumped through columns containing each of the four sorbents. Lastly, a 437 mg/L Ni solution was tested on columns of clinoptilolite and hydroxy-apatite. For the Zn/hydroxy-apatite experiment, 200 mg/L Ca++ was also added to the influent water to simulate hard water. In each experiment, samples were collected at times ranging from 15 minutes to several hours at the outflow of the column in 50-mL graduated polypropylene tubes and capped. Effluent not sampled was collected in 5-gallon (18.9 L) plastic carboys and sent to hazardous waste treatment facility. Samples were analyzed for aqueous-metal species using Acetylene-Air Flame/Atomic Adsorption (Perkin-Elmer Model 5100PC), using single-element hollow cathode lamps. Two-point calibration curves were prepared from stock solution for each element. Samples collected at column outflow were diluted with water to bring concentration within linear range of instrument. Check samples were used periodically throughout sample analysis to ensure quality control.
Results
Figure 2 presents the breakthrough curves for each metal for the organoclay/anthracite column. The vertical axis is relative concentration, defined as the ratio of the inflow concentration to the outflow concentration. The horizontal axis is pore volumes of flow. Figures 3-5 present similar curves for each of the other three sorbents. Figure 6 presents breakthrough curve data for Zn (as a single solute) on all four sorbents. Figure 7 presents breakthrough curve data for Ni on hydroxy-apatite with 200mg/L Ca++ and clinoptilolite. In all cases, the breakthrough curves are approximately S-shaped (or are approaching an S-shape), which is characteristic of advective/dispersive transport through a water-saturated porous medium. The time for complete breakthrough varies with both metal type and sorbent, indicating that some metals are sorbing to a greater extent than others. In the absence of sorption, the relative solute concentration (defined as the influent concentration divided by the effluent concentration) is expected to equal 1 at 1 pore volume of flow. For all experiments, the relative concentrations of 0.5 correspond to pore volumes of 2 or greater, indicating that there is measurable sorption of every metal to every sorbent.
Table 1 presents summary information about the column experiments with all five metals present in the inflow solution. Table 2 presents summary information for Zn (as a single solute) transport through all four sorbent media. Table 3 presents summary information for Ni transport through clinoptilolite and hydroxy-apatite (with 200 mg/L Ca++ as the co-solute). These data were calculated from the breakthrough curves shown in Figures 2-5. We define breakthrough as the point where the relative concentration equals 0.95. For the transport of some metals through the clinoptilolite and hydroxy-apatite columns, relative concentrations of 0.95 were not reached during the experimental monitoring period, so the breakthrough curves were extrapolated to calculate the values in Table 1.
Discussion
Considering first the multiple-metal-sorption experiments (Figure 2 and Table 1), several important trends can be identified. For the organoclay/anthracite blend, it appears that sorption of all five metals is similar and relatively weak. Sorption increases for the pure organoclay (Figure 3, Table 1), with percent sorbed descending in the following order: Cd > Cr > Cu > Zn > Ni. The greater sorption of the metals to the organoclay compared to the organoclay/anthracite blend suggests that the organoclay is a better metal sorbent than the anthracite. Replacing the anthracite in the blend with organoclay therefore results in a better sorbent for the metals. Sorption of the metals is further increased when the sorbent is changed to clinoptilolite, with percent sorbed descending in the following order: Cu > Cr > Zn > Cd > Ni. Unlike organoclays or anthracite, zeolites like clinoptilolite have polar (charged) surfaces that can effectively sorb metals or charged metal-hydroxides, and this feature of the sorbent likely explains its improved performance relative to the organoclay or organoclay/anthracite blend. It may be possible to regenerate the clinoptilolite column with a 3% NaCO3 solution. Other zeolites, such as chabazite, may have a higher removal capacity than clinoptilolite, but these zeolites are more expensive and therefore may not be of practical use. The best metal removal from solution was observed for the hydroxy-apatite column (Figure 5, Table 1), with percent sorbed descending in the following order: Cu > Cr > Cd > Zn > Ni. The strong performance of this sorbent may be attributable to its relatively large internal porosity and correspondingly polar surface area.
Although the organoclay and organoclay/anthracite blend did not perform as well for metal removal as the clinoptilolite or the hydroxy-apatite, it should be noted that these sorbents may still be well suited for waste streams that contain significant amounts of oil and grease (or other dissolved, nonpolar organic pollutants) and relatively small amounts of metals. The organoclay or organoclay/anthracite blend will effectively remove the organic contamination and still be able to remove small amounts of heavy metals. The authors do not know of other comparable dual-purpose sorbents that can effectively treat such a mixed-contaminant wastewater stream.
In considering the transport of Zn as a single solute through the four porous media (Table 2 and Figure 6), the four sorbents show similar relative sorption behavior, e.g. sorption decreases in the following order: Hydroxy-apatite > clinoptilolite > organoclay > organoclay/anthracite. However, the magnitude of sorption, expressed on a percentage weight basis (Table 2), increases for every sorbent compared to the multiple-metal experiments (Table 1). This indicates that sorption of the metals is competitive. Each sorptive media will likely have better metal removal efficiency when only one or two metals are present in the waste stream compared to multiple metals (assuming everything else is equal). Similar conclusions can be drawn for Ni sorption (as a single solute) to clinoptilolite and hydroxy-apatite (Table 3 and Figure 7). Exposing the hydroxy-apatite to 200 mg/l of Ca++ seems to have had no deleterious effect on the media’s performance, although we have not performed a single-solute Ni/hydroxy-apatite experiment in the absence of Ca++. Such a test would re required to positively confirm this observation. It is a well known that zeolites, clays, and synthetic ion-exchange resins loose substantial capacity for removal of metals in hard water. The hydroxy-apatite may not have this limitation.
Conclusions
Four sorbents were studied for the removal of heavy metals from water. All four sorbents showed measurable metal sorption when studied in column experiments with inflow solutions containing all five metals and in a limited number of single-solute column experiments. Sorption of Zn and Ni to clinoptilolite and hydroxy-apatite exhibited competitive sorption, and it is likely that competitive sorption effects exist with the other sorbent-metal combinations. Hydroxy-apatite performed best in the experiments, and appears to be an effective sorbent for heavy metals. Its performance did not appear to be hindered by the presence of hardness (200 mg/L Ca++). In multi-metal column sorption experiments, each sorbent preferentially sorbed different metals, although in general, Cu and Cr uptake was strongest and Ni uptake was weakest. Although the organoclay and organoclay/anthracite blend generally did not perform as well as the clinoptilolite and the hydroxy-apatite, these sorbents may be well suited for removal of oil and/or other nonpolar organic pollutants in combination with small quantities of heavy metals.
Figure captions
Figure 1. Photograph of experimental setup showing PVC column, peristaltic pump, inflow reservoir and waste collection carboy. Aqueous-metal solution is pumped up through the column in order to displace pore-space air and prevent short-circuit flow. Figure 2. Breakthrough curves of individual metals from an inflow solution containing 60 mg/L Cd, 150 mg/L Cr, 120 mg/L Cu, 60 mg/L Ni and 60 mg/L Zn through a column containing organoclay/anthracite. Figure 3. Breakthrough curves of individual metals from an inflow solution containing 60 mg/L Cd, 150 mg/L Cr, 120 mg/L Cu, 60 mg/L Ni and 60 mg/L Zn through a column containing organoclay. Figure 4. Breakthrough curves of individual metals from an inflow solution containing 60 mg/L Cd, 150 mg/L Cr, 120 mg/L Cu, 60 mg/L Ni and 60 mg/L Zn through a column containing clinoptilolite. Figure 5. Breakthrough curves of individual metals from an inflow solution containing 60 mg/L Cd, 150 mg/L Cr, 120 mg/L Cu, 60 mg/L Ni and 60 mg/L Zn through a column containing hydroxy-apatite. Figure 6. Breakthrough curves from an inflow solution containing 785 mg/L of Zn through columns containing the listed sorbent materials. Figure 7. Breakthrough curves from an inflow solution containing 437 mg/L of Ni through columns containing clinoptilolite and hydroxy-apatite.
|
|
Biomin, Inc. P.O. Box 20028 Ferndale, MI 48220 Phone: 248-544-2552 Fax: 248-544-3733 E-Mail: biomin@aol.com See Master
Mind Marketing Inc with questions or comments regarding this web site.
|
