Makers of OilSorb™ and Other State-of-the-Art Filtration Media

Request Biomin's Cost Savings Model and Guides

Using organoclays to enhance carbon filtration

by George Alther

Abstract

Organoclays have found increased acceptance as pre-treatment for activated carbon adsorption systems in both groundwater and wastewater cleanup. The reason is that activated carbon tends to become quickly blinded by large organic molecules of low solubility, particularly oils. However, it is also well established that activated carbon is more efficient at low concentrations of organic contaminants than at higher ones, i.e. at less than 1 ppm. With organoclays it is exactly the opposite, they are better at removing organics at higher concentrations, above 3 ppm. Therefore it is cost effective in these applications to use two or more vessels in series, the first one filled with organoclay, the remainder with activated carbon. The economics make sense, even though the organoclay is not regenerated, because of the reduction in down time every time a carbon vessel has to be changed out. Use of organoclays increases the volume treated by carbon in many applications seven to nine fold. In the case of other organic contaminants, as the aqueous solubility increases, the efficiency decreases, except in the case of methylene chloride, which it removes at far higher efficiency then carbon. This article presents the results of a series of tests, including K_d determinations, jar tests, and mini-column tests. These tests determined the adsorption capacity and efficiency of organoclay and activated carbon for the removal of benzene, toluene, xylene and naphtalene from water. These tests were followed by adding the four compounds into one container to see if the combination of organoclay, followed by carbon, would be more efficient then each sorbent alone. The tests also compared the efficiency of organoclay versus carbon for the removal of various oils from water. (2002 Elsevier Science Ltd. All rights reserved).

1. Introduction

Organoclays consist of bentonite that has been mod­ified with quaternary amine cations. The positively charged nitrogen cation exchanges with sodium and calcium ions on the clay surface [1]. This exchange provides an organophilic material, that will not swell in water, but will swell in hydrocarbon fluids such as kerosene, diesel fuel, and so on. The quaternary amine chains extend away from the clay surface into the water [2]. The amine chain is now neutral because the cation is attached to the clay surface, rendering the entire system into a non-ionic surfactant with a solid base, i.e. it becomes a non-ionic organoclay. This chemical modification allows it to function as a non-ionic surfactant and remove oil and other organic compounds of low polarity by partition [3]. As oil droplets and other organic compounds, pass an ozganoclay;pa-l^-xicle whose chains extend into the water, these chains will extend like the tentacles of a squid into the water, partitioning into the oil droplet and fixating it (the droplet) onto themselves. Since this activity takes place outside of the clay, no blinding occurs. During the ion exchange of the quaternary amine with the sodium and calcium on the clay surface, only a fraction of the total number of ions is exchanged [4]. The remainder are still present, and give the clay the ability to also remove inorganic metal cations such as lead, zinc, copper and others. While the removal capacity is hard to predict, it nevertheless is a side benefit for the end user to have the benefit of an advanced media that removes organics and inorganics simultaneously. Usually not all the quaternary amine chains are attached to the clay platelets by ion exchange. Some of them tie to the attached chains by a form of tail-tail interaction (Boyd, personal communication): The result of this tail-tail interaction is. that a positive charge extends into the water, causing negatively charged metal anions such as hexavalent chrome, selenite and arsenate to be removed by ionic attraction. Many humic acids are also removed by such cationic organoclays because they contain multiple negative charges.

If the removal of oil is the main purpose of using the organoclay, and if it is placed into a carbon vessel, the clay is granulated and then blended with anthracite to keep the pore spaces open. The ratio between organoclay and anthrazite is 30% versus 70% [5]. Anthrazite is used because it has the same bulk density as that of organoclay.

It also adds BTU (British Thermal Units) if the spent media is incinerated.

1.1. Description of organoclays

In order to understand the proposed research that forms the basis of this article, we need to understand organoclays in some detail. Organically modified clays, also called organoclays, consist of bentonite which is modified with quaternary amines [1]. The major constituent of the bentonite, a chemically altered volcanic ash, is the clay mineral montmorillonite. It has a cation exchange capacity of 70-95 meq/100 g. The nitrogen end of the quaternary amine, which has a positive charge, is exchanged onto the clay surface for sodium and calcium [1]. The replaced sodium and calcium ions are ionically bonded with the free chloride ions from the amines, forming salts which are washed out during the initial backwash [2]. The clay becomes a solid based non-ionic surfactant. The amine chains extend into the water, removing the oils and other non-polar or slightly polar organics by a partition phenomena [4,6-9]. The objectives of this research are to compare the removal capacity of organoclays and bituminous activated carbon in order to guide the engineers that design groundwater remediation systems in how to apply these two sorbents for maximum efficiency at minimum cost.

1.2. Laboratory methods

1.2.1. Preliminary tests, sorption experiments

For a preliminary comparison the iodine numbers of a non-ionic organoclay, a cationic organoclay, and activated carbon were determined. They are:

Non-Ionic Organoclay: 275 mg/g (1.1 meq/g)

Cationic Organoclay: 190 mg/g (0.76 meq/g)

Bituminous Activated Carbon: 600-800 mg/g

Virgin Bituminous Activated Carbon: 900-1100 mg/g.

The method used for iodine number determination is: ASTM D 4607-94.

Next, the sorption distribution pp efficient, K_d, was determined as a quick comparison between the sorbents. A conventional batch equilibration procedure was used. Each batch reactor (15 ml nominal volume glass centrifuge tube with Teflon-lined cap) contained 2 g of sorbent, 14 ml of de-ionized, organic free water and 8.75 μl of (C 14) BTX. The batch reactors were equilibrated for 24 h and then centrifuged for 1 h at 2000 g to separate the solid and aqueous phases. Two ml of the supernatants and 6 ml of the scintillation cocktail were then placed into scintillation vials. These samples were then analyzed using a scintillation counter. The instrument was first calibrated with solutions of known BTX concentrations prior to analyzing the supernatants. The counts per minute (CPM) readings were used to find the aqueous phase concentration of each sorbent sample. In each sample trial, "blank" batch reactors that contained only BTX and water, but no sorbent, were also prepared and analyzed. Losses from the blanks were accounted for in each sample trial. All experiments were performed in triplicate.

The equilibrium sorbed concentration of BTX on the sorbent is represented by Cs. The equilibrium aqueous concentration is represented by Caq. The sorption coefficient, Kd, can then be defined as:

K_d = Cs/Caq

The higher the value of K_d, the greater the amount of benzene, toluene or xylene sorbed onto the sorbent. The percent removal efficiency, R, for each sorbent pair is defined by:

    R=(Mi-Mf/Mi) x 100

where Mi =initial mass of solute in aqueous phase prior to any sorption; and Mf, final (equilibrium) mass of solute in aqueous phase after sorption.

1.2.2. The mini-column technique:

This technique consists of pumping a solution through a small sorbent sample, for example 1 g, in a mini-column into which the sorbent is packed, until the influent concentration of the contaminant equals that in the effluent [10]. The solid sorbent is a fine powder, in this case organoclay or powdered activated carbon (PAC). In these experiments, synthetic water solutions were prepared by spiking deionized water with organics, including benzene, toluene, xylene, and naphthalene. These solutions were pumped through individual mini columns to determine the effectiveness of each sorbent, i.e. organoclay and PAC. The advantage of the mini column technique is that the equilibrium concentration is the same as the influent concentration, and therefore under ready control. This technique is more, real world than isotherms, even though the two sorbents, when used in the field in flow-through adsorption vessels, are of granular nature, which experience longer contact time to reach equilibrium than powders.

1.2.3. The jar test

The third test used is the "Jar Test", which is a single point isotherm [10]. It is performed by contacting a known weight of sorbent with the contaminated water. A water sample containing no sorbent is also tested and used as control, so that the adsorbent performance can be determined. For example, 1 g sorbent is slowly added to a 100-ml jar filled with the contaminated water. The sorbent is then dispersed in the water by a shaking mechanism, a magnetic stirrer, or a paddle. After a predetermined time the stirring activity is stopped, the solids allowed to settle, followed by centrifugation to separate the solids from the liquid. The amount of the remaining contamination is then determined by chemical analysis. This test provides a quick performance evaluation, without having to perform a 10 point ASTM Isotherm Test [10]. Due to financial constraints, replication was not conducted.

2. Results

The iodine numbers for activated carbon are much higher then for organoclays. This is because carbon granules contain pores that have a much higher sorption capacity than the amine chains that stick off the organoclay, even though 1 g of bentonite has a surface area of 700 m²/g [1]. However, when larger organic compounds are to be removed, this difference becomes irrelevant because these compounds will fill, and blind, the pores of the carbon.

The table below shows the K_d and R results from these tests.

ND = not detected.

Note that R is a function of sorbent to water ratio. The amount of sorbent and water used in each experiment were the same. Both R and K_d can be used as a direct comparison between the sorbents. Bituminous activated carbon is the strongest sorbent, according to this type of test, as would be expected from the iodine numbers. However, the organoclays are not far behind. These results are the results of, and based on, triplicate analysis of measurements of a single-phase aqueous phase concentration. These results may change for different equilibrium aqueous concentrations owing to possible non-linear solute-sorbent sorption isotherms. Particularly noteworthy is that pH determines, to a substantial degree, the effectiveness of activated carbon, i.e. a pH between 4 and 7 gives higher results then a pH above 7. Non-ionic organoclay, on the other hand, which removes these compounds by partition, is not affected by pH. The values for capacity as mass percent are based on one data point. If these numbers were determined off a 10 point isotherm, they would probably be much higher.

Fig. 1 displays a graphic illustration of some of the mini-column test results. When testing the removal capacity of the sorbents for benzene, toluene, o-xylene, and naphtalene, the organoclay performs nearly as well as carbon, improving in performance as the solubility of the compounds decreases. The organoclay outperforms carbon with the less soluble PCB 1260 and motor oil (not shown), which is nearly insoluble. Surprisingly, organoclay removes methylene chloride much more effectively than does carbon. The same had been observed in an earlier study with vinyl chloride [11]. It is theorized that compounds such as methylene chloride have a high electro-negativity due to the presence of large amounts of chlorides. The organoclay possesses some positive charges on the clay surface, which would then chemically bond through electrostatic attraction with these negative charges and thus remove the organo-halogen compounds from the water. Therefore, two removal mechanisms account for the organoclays effectiveness, partition and ionic bonding [12].

These mini-column tests using single contaminants were followed by a set of tests using a ternary compound contaminant in the influent. The influent was spiked with benzene, toluene and naphthalene. This was done to observe the competition amongst the contaminants for sorption sites, and to observe if the least soluble ones keep the more soluble ones such as benzene, off the sorption sites. The synthetic contaminants were added at 900 mg/l of each compound into de-ioidized water. It was possible to add that much naphthalene because the benzene and toluene helped dissolve it. Usually its solubility is 10 mg/l.  This concentrate was pumped separately through an organoclay and a carbon mini-column. The effluent was tested for each solvent to determine the breakthrough, i.e. which compound broke through earlier and at what levels. Another mini-column was filled first with 0.5 g of organoclay, followed by 0.5 g of activated carbon on the effluent side, to determine if the organoclay/carbon succession was indeed more effective then each of the media alone.

Fig. 1 shows that benzene breaks through first in the ternary experiment, followed by toluene and naphthalene last. This observation is expected based on their solubility. This scenario is similar to three people wanting to enter a train at the same time. The skinniest one slips in first, the heaviest one last. This competition is not 100% proof, the total adsorbed amounts are higher than the individually adsorbed amounts of the three solvents at breakthrough, probably because the geometrically arranged packing of solvent molecules of different sizes in the carbon pores or around the amine chains favors a higher packing density.

The most important conclusion is that the Organoclay/Carbon mini-column system is much more effective then either sorbent alone, even though the amount of sorbent in each case is double than that in the combined column (Fig. 2). These results were obtained from a real world scenario, a ground water clean up project. The organoclay removes the oil completely, and also a significant amount of other solvents, increasing the activated carbon's effectiveness for the removal of the VOCs. These tests were conducted in three columns, 3 inches wide and 30 inches long. Glass wool keeps the media on the bottom from escaping. Water is then pumped through at less than 5 gpm/ft².

Figs. 3 and 4 contain summary data that show the effectiveness of organoclay in removing many different oils from water. It was observed that refined oils are removed much more efficiently then crude oils. The reason is that crude oils contain many polar substances which the non-polar organoclay removes with lower effectiveness. Fig. 3 shows its removal capacity for mineral oils. As long as these oils are non-polar, the non-ionic organoclay is an excellent adsorber. The effect of polarity is even more pronounced in the removal of plant oils from water (Fig. 4).

As long as the oil is refined the results are good. Fig. 4 shows this even more clearly when plant oil is removed from water. Fig. 5 shows a comparison of cationic organoclay with non-ionic organoclay for the removal of non-refined or crude oils, and some synthetic oils. The cationic organoclay is much better at removing polar crudes then non-ionic organoclay. The crudes tend to have charges of both types, as a result of the presence of lipids.

3. Case history

Three million gallons of wastewater were stored in a gills holding tank at a former manufactured gas plant.

The tank is made out of steel, placed onto a concrete slab. The bottom of the tank was covered with coal tar. Contaminated water sat on top of the tar. The water was part of the original seal that contained the gas in the telescoping tank. The purpose of removing the water was to facilitate the removal of the tar. An oily sheen was found floating on top of the water and adhering to the tank walls. The water was contaminated with polycyclic aromatic hydrocarbons, oil, benzene, toluene, ethyl benzene, xylenes and heavy metals. Only benzene and xylenes needed to be removed, according to local discharge standards, to a level of 134 and 74 ppb, respectively.

The treatment system that was initially installed included an oil/water separator, bag filters to remove suspended solids, and two granular activated carbon (GAC) vessels that could each hold 6000 lb of GAC. The activated carbon was spent within 3 days due to the oil in the water, its pores were blinded.

Then the carbon in the first vessel was replaced by 9000 lb of organoclay. The

organoclay lasted about 6 weeks, was changed out once, while the carbon was replaced twice. Discharge limits were met at all times.

The concentration of contaminants in the influent and effluent were:

3.1. Conclusions

The results from these tests and real world situations discussed in trade journal articles show that the costs of water treatment, be that industrial wastewater, or pump and treat groundwater cleanup systems, can be significantly lowered by using the organoclay/carbon approach. Pre-polishing with organoclay, first of all, lowers activated carbon consumption by 700%. Secondly, the organoclay removes a substantial amount of less soluble solvents, improving the effectiveness of activated carbon for removal of more soluble solvents even more. These data and the case history show that organoclays can be used to improve the efficiency of carbon, and thus lower the operations::doqs of. the project, even if there is no oil, or only a very small amount (1 ppm), present in the water. The data also shows that organoclays can be used to remove any kind of oil, which would be far too expensive if carbon alone were used. Therefore, the recommendation to the design engineer is that he consider using a clay/carbon system for improved efficiency of water cleanup. The initial test results, including iodine number and K_d values, are of little value for these field applications, but are useful for a preliminary comparison of the efficiency of different organoclays and activated carbon as sorbents.

Acknowledgements

The K_d testing was performed under the guidance of Professor James Smith, University of Virginia, in his laboratory.  The mini-column and jar tests were performed by Dr. Henry Novicki in the laboratory of PAC Inc., Coraopolis, PA.

[ Home ] [ About Us ] [ Products ] [ Services ] [ Technical Bulletins ] [ Case Studies ] [ Testimonials ] [ News ] [ Links ] [ Contact Us ] [ Request More Information ]

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.