By Peter Meyers
As a general class, the arsenic adsorbents are all based on similar adsorbent chemistry. An insoluble metal oxide/hydroxide is the adsorbent, which is contained in some type of granular substrate that is porous enough to expose the adsorption sites to the water, yet robust enough to maintain its shape under the conditions of use. Arsenic in the form of arsenate, an anion, is adsorbed by co-precipitation onto the metal oxide, while arsenate (trivalent arsenic) is attracted by chelating forces. Precipitated arsenate, in most cases, is tightly bound to the oxide and does not come back off under any normal conditions found in potable water. All of the arsenic adsorbents promise extremely long throughputs by targeting the arsenic species, while ignoring common ions, such as sulfate, chloride, sodium, calcium, etc.
Several older media and a host of new media are commercially available for arsenic removal. The market is red hot due to the impending change in U.S. Environmental Protection Agency (EPA) regulations reducing allowable arsenic levels in drinking water from 50 ppb down to 10 ppb. All the current media promise many hundreds of thousand gallons per cubic foot (cu.ft.) of throughput on ideal water supplies. Real life throughputs, however, are often reduced by the presence of interfering substances and other water quality factors. Most of these interferences apply to all potential arsenic adsorption medias. Without prejudice toward any medium, here are the major factors that limit throughput.
By far the biggest influence on throughput is caused by pH. All arsenic adsorptive medias work better at lower pH and none work well above a pH of 9.0. In fact, of the media that claim to be regenerable, all employ some type of highly alkaline solution to strip arsenic off the exhausted medium. Although some media can also be regenerated with highly acidic solutions, it should be noted that many dissolve in acid and/or lose their active metal oxide, rendering them unfit for further use. Arsenite (As+3) is better adsorbed at a higher pH but the effect is limited within the potable water range.
Silica is either second or third by all media manufacturers in their list of interferences. Ionized silica competes with arsenic for adsorption sites and can dramatically reduce throughputs. Silica is essentially non-ionized below a pH of approximately 7.5, therefore having less effect. However, above a pH of 8, a substantial portion of the silica present is ionized and loads onto exchange sites, blocking them for future arsenic adsorption. Various media have varying degrees of sensitivity to silica, but all are profoundly affected at a pH above 8.5.
Phosphate weighs in close to silica on the list of interferences. Interestingly, the media that claim a reduced sensitivity to silica seem, by and large, to be more sensitive to phosphate. Ionized phosphate, like silica, directly competes with arsenic for adsorption sites. Although seldom mentioned, sulfide also competes for adsorption sites. The combined effect of these interfering ions can in some cases reduce the capacity for arsenic by more than 90 percent.
The effect of a high pH and the presence of silica and/or phosphate can easily reduce a media’s throughput from more than 500,000 gallons/cu.ft. to less than 100,000 gallons/cu.ft. Be sure to provide the system supplier with reliable information about the expected range of concentrations of these primary interferences so that the throughput can be accurately predicted in advance; otherwise the replacement cost of a medium may come as an unwelcome surprise.
Many additional factors can have moderate to minor effects on arsenic adsorption. Some are identified problem contaminants themselves; their removal, along with arsenic, may be viewed as beneficial even when the overall throughput capacity of the media is reduced. Other factors have a small influence on throughput (either negative or positive) and can generally be ignored when predicting throughput capacities.
As a group, the oxy anions (negatively charged species containing a transition metal cation and oxygen) are all adsorbed to some degree along with arsenic, thus competing for adsorption sites and reducing throughput. The most topical of these anions is perhaps vanadate (containing vanadium, a significant poison in its own right), chromate (poisonous in its +6 oxidation state, yet an essential nutrient when present in its cationic trivalent state) and selenate (containing selenium). These anions are removed to varying degrees by the arsenic adsorbents and removal is generally dependent on the pH and oxidation state of the metal. In most cases the presence of these contaminants does not result in huge reductions of throughput.
Inorganic ions such as sulfate and bicarbonate generally have very little effect on adsorption. However, very high concentrations (well above the potable water range) can have enough of an effect that they need to be considered. Some of the hybrid adsorbents may be partially (or fully) regenerated by highly saline solutions. Although cations do not generally play a role in the adsorption of anionic species, the presence of calcium seems to help minimize the effect of silica and phosphate (probably by combining with and therefore reducing the percentage of ionization). At higher levels, however, calcium can combine with sulfate (or carbonate) to form precipitants that can cause other problems.
As a general class, foulants are contaminants that either coat the media (blocking access to adsorption sites) or plug up the flow passages thru the media (thus causing unequal distribution known as channeling). Turbidity and/or suspended solids are removed by filtration within the media and thus can build up over time. The very long throughputs do not afford regular opportunities for purging suspended solids. Backwash is generally not 100 percent successful to remove suspended solids and in some cases can be inadequate to restore performance. This is because foulants may undergo physical/chemical changes, becoming larger and heavier over time and perhaps sticking to the media particles.
Iron and manganese are two well-known foulants. Many ground waters contain significant concentrations of iron and manganese in their soluble forms. However, potable waters are often chlorinated and/or aerated. These processes cause iron and manganese to precipitate as very fine colloidal materials. As such they are not well removed by pre-filters but they gradually coat the adsorption medias and block the adsorption sites. Since iron binds with arsenic, the presence of iron precipitants greatly complicates the removal. On one hand, additional arsenic may be removed by filtration but any iron that is not filtered out can drag arsenic thru the media causing premature or sporadic leakage. Various medias have widely differing tolerances to iron and manganese but all are affected to some degree.
Keep in mind that because the adsorptive medias generally have very long throughputs, they are vulnerable to foulants even at low concentrations. Even when a media is capable of filtration as well as adsorption, the act of filtering out contaminants is sure to reduce its adsorptive capabilities, at least to a minor extent.
All adsorptive medias are susceptible to fouling by organic matter, oils and greases and manmade pollutants that precipitate and/or coat the surface of the media. While some medias are probably more resistant than others, the presence of these physical contaminants should generally be addressed as separate pretreatment steps.
Despite interferences that may result in 90 percent reduction of throughput, the remaining 10 percent can still be hundreds of thousands of gallons per cubic foot. Resulting in exhaustion times measured in weeks or months under ideal conditions, these medias may last significantly longer, even up to several years.
Although by no means the only proven method to remove arsenic, the arsenic adsorbents require very little operator attention and are very simple devices lending themselves well to point-of-use (POU) and point-of-entry (POE) systems. In order to ensure a successful, long-lived installation, it is always advisable to consult knowledgeable sources and follow their direction in design and pre-treatment of these systems.
- Wang, L, Chen, A., Fields, K., “Arsenic Removal from Drinking Water by Ion Exchange and Activated Alumina Plants”, EPA, EPA/600R-00/088 October 2000 Report
- Kommineni, S., Chowdhury, Z, Amy, G. “Modeling of Water Quality Interferences on Arsenic Adsorbents”, 2004 AWWA WQTC Conference
- Driehaus, W., “Arsenic Removal with Iron Based Adsorbents: How to Deal with Silica”, 2004 AWWA WQTC Conference
- Chanda, M., O’Driscoll, K., Rempel, G., “Ligand Exchange Sorption of Arsenate and Arsenite Anions by Chelating Resins in Ferric Ion Form”, Reactive Polymers, 7 (1988) p. 251-261
- Meng, X., Bang, S., Korfiatis, G., “Effects of Silicate, Sulfate and Carbonate on Arsenic Removal by Ferric Chloride, Water Research (2000) Vol. 34, No. 4, p 1255-1261
- Meng, X., Korfiatis, G., Christodaoulator, C., Bang, S., “Treatment of Arsenic in Bangladesh Well Water Using a Household Co-precipitation and Filtration System”, Water Research (2001) Vol. 35, p 2805-2810
- Brandhuber, P., “The Impact of the Presence of Silica on the Treatment of Arsenic in Drinking Water”, 2004 AWWA WQTC Conference
- Meng, X., Korfiatis, G., Bang, S., Bang, K., “Combined Effects of Anions on Arsenic Removal by Iron Hydroxides”, Journal of Toxicology and Environmental Health, October, 2001
About the author
Peter Meyers is technical director for ResinTech Inc., of West Berlin, N.J. ResinTech is an ion exchange resin manufacturer and, through its two divisions Aries and ACM Company, also offers activated carbon, inorganic selective exchangers, cartridge filters and PEDI regeneration services. Meyers has more than 30 years of experience covering a wide range of ion exchange applications from demineralizers, polishers, and softeners to industrial process design and operation. He can be reached at (856) 768-9600 or email: [email protected]