- Email: [email protected]
A simple technology for arsenic removal from drinking water using hydrotalcite
Science of the Total Environment 366 (2006) 926 – 931 www.elsevier.com/locate/scitotenv
Technical note
A simple technology for arsenic removal from drinking water using hydrotalcite G.P. Gillman ⁎ CSIRO Land and Water, Private Mail Bag, Aitkenvale PO. Qld. 4814, Australia Received 1 August 2005; received in revised form 18 December 2005; accepted 16 January 2006 Available online 20 March 2006
Abstract The use of a synthetically prepared clay material, hydrotalcite (HT), for the removal of arsenite (As(III)) and arsenate (As(V)) from drinking water is described. Percolation through HT of water containing 500–1000μg/L As (levels often found in Ascontaminated well water) produced leachate with As levels well below 10 μg/L. The technology could be coupled to that used in less-developed regions for removing organisms from drinking water, viz. leaching through porous pots and filter candles. The ‘spent’ HT is easily converted into valuable phosphatic fertilizer that would have an insignificant effect on soil arsenic levels, thereby reducing the overall cost of manufacture and distribution. © 2006 Elsevier B.V. All rights reserved. Keywords: Arsenic removal; Drinking water; Hydrotalcite; Phosphate fertilizer
1. Introduction The presence of arsenic in drinking water, at levels shown to be a significant health hazard, is receiving increasing attention worldwide (Nordstrom, 2002). Levels of arsenic in excess of the World Health Organization recommended limit of 10μg/L can occur naturally in water bodies used as sources of drinking water, but the problem is exacerbated by the sinking of wells in areas of arsenic-rich geological strata. Values in the range 10–1000 μg/L have been reported and the UN Synthesis Report on Arsenic in Drinking Water (www.who.int/water_sanitation_health/dwq/arsenic3) provides an excellent overview of the extent of the
⁎ Tel.: +61 7 47538576; fax: +61 7 47538600. E-mail address: [email protected]. 0048-9697/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2006.01.036
problem and of technologies being evaluated for its solution. This paper canvasses the use of a synthetic clay mineral, hydrotalcite (HT), as a material for removing arsenic from water at household or communal level in less-developed regions. Hydrotalcite is essentially an aluminium-substituted brucite (Mg(OH)2) mineral and forms when a proportion of the magnesium ions in the structure is replaced by aluminium, leading to an excess of structural positive charge. The positively charged sheets align, with interlayer spacings being occupied by anions to maintain electrical neutrality. This process is the reverse of the adsorption reactions involving smectitic clays, where structural negative charge is balanced by interlayer cations. Hydrotalcite occurs naturally in specimen amounts under favourable conditions, but can be easily synthesised using inexpensive raw materials such as bauxite and magnesite (Gillman and Noble, 2005).
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
The term ‘hydrotalcite’ is generally reserved for the form wherein carbonate ions occupy the interlayer spacings. When other interlayer anions such as nitrate, chloride and phosphate are dominant, the materials are referred to as ‘hydrotalcite-like compounds’. In this paper, the general term ‘hydrotalcite’ (HT) will be applied irrespective of the charge-balancing anion and an anion prefix used to describe a particular form, e.g. NO3–HT and Cl–HT for nitrate and chloride hydrotalcite, respectively. Hydrotalcite specificity for anions follows the order CO3 > PO4 > SO4 > Cl > NO3 (Tamagawa, 2003) and, in the absence of carbonate, a chloride or nitrate HT will readily adsorb phosphate from solution as was demonstrated by Ookubu et al. (1993). The similarity between the structure of the arsenic anions arsenite and arsenate to that of phosphate anions indicates that HT could be used to effectively remove arsenic from water. This paper therefore presents results of laboratory experiments comparing the capacity of several HT materials to adsorb arsenic, followed by prototype technology studies that might be adaptable for use in regions referred to above. 2. Adsorption of As(III) by Cl–HT, NO3–HT and CO3–HT 2.1. Materials and methods Arsenic measurements were made with a Varian UltraMass 700 ICP-MS, externally calibrated using a commercially available standard solution, and also an In internal standard to correct for instrument drift and matrix effects. The limit of detection for As under the conditions used was 0.42μg/L. To obtain an approximate 500 μg/L As(III) stock solution, a saturated arsenic trioxide (As2O3) solution that was approximately 10,000 mg/L As was diluted by a factor of 20,000. The actual As(III) concentration was determined by ICP-MS to be 432 μg/L. The synthesis of HT is easily effected by coprecipitation of a Mg/Al double hydroxide. The ratio of Mg/Al should be held between 0.25 and 0.33, and pH must be higher than that at which either Mg or Al hydroxide would separately precipitate. In the three syntheses described below, pH was maintained at or above 9.5. XRD analysis was carried out on the three HT compounds to confirm that true HT materials had been formed. An NO3–HT slurry was prepared by adding 500mL of a mixed nitrate solution (1M Mg(NO3)2, 0.5M Al (NO3)3) via a peristaltic pump at 450 mL/min to 250mL
927
of concentrated ammonia (25%) with vigorous stirring. The final pH of the HT precipitate was 9.5. Following a number of centrifuge-washings to the point of dispersion to virtually remove all ammonium nitrate, the NO3–HT ‘cake’ (moisture content 87%) was stored in a sealed container. A CO3–HT slurry was prepared by bubbling CO2 through an NO3–HT slurry (see above) for 2 h. After centrifuge-washing to the point of dispersion to remove residual nitrate, the CO3–HT ‘cake” (moisture content 89%) was stored in a sealed container. A Cl–HT solid was prepared using the method of Ookubu et al. (1993). A 350mL solution of mixed chloride (0.86 M MgCl2, 0.4 M AlCl3) and, separately, a 2.5 M NaOH solution were pumped into a beaker under vigorously stirred conditions, with adjustment of pump rates to maintain pH of the resultant slurry at about 9.5. The amount of alkali consumed was 420 mL. The slurry was aged at 80 °C for 8 h, followed by centrifugewashing to the point of dispersion and drying of the ‘cake’ at 80°C. In separate experiments, 0.1, 0.5, 1.0 and 2.5 g (o. d. basis) of NO3–HT cake or Cl–HT powder were added to 500mL aliquots of the stock As(III) solution that initially contained 432 μg/LAs. In another experiment, 2.5g (o.d. basis) of CO3–HT cake was added to 500 mL of As stock solution. The suspensions were stirred in an open beaker for 1h and a small sample of supernatant solution then taken for As analysis. After a further 17 h stirring, another supernatant solution sample was taken. The samples were passed through a 0.45 μ membrane filter prior to As determination. 2.2. Results and discussion XRD patterns of the three synthesised hydrotalcite clays are presented in Fig. 1. Basal spacings calculated from the 003 reflections (near 10° 2ϴ) were 0.89nm, 0.77 nm and 0.78nm for NO3–HT, CO3–HT and Cl– HT, respectively, values that are in very good agreement with published basal spacings. The sharpness of the 003 and 006 peaks demonstrates the layered structure of these materials. The concentrations of As(III) remaining in solution following contact with NO3–HT and Cl–HT, over a range of HT addition rates, and for two contact times, are summarized in Table 1. (Addition of HT in carbonate form did not alter solution As concentration and CO3– HT will not be further discussed.) When added in sufficient quantity, both NO3–HT and Cl–HT reduced As(III) concentrations from those
928
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
Fig. 1. XRD patterns of dried powder samples of HT used in this study. (a) NO3–HT, (b) CO3–HT and (c) Cl–HT.
commonly found in water from tube-wells passing through arsenic-bearing rock strata, to below the WHO recommended upper limit of 10μg/L−1.
The adsorption appears to be diffusion controlled, with 18 h of contact resulting in removal of more arsenic than after 1h of contact. Even then, only a fraction of the
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931 Table 1 Residual As(III) concentration (μg/L) in water originally containing 432μg/L As(III) after contact with increasing amounts of NO3–HT and Cl–HT for 1h and 18h HT addn. rate (mg HT/500mL H2O)
NO3–HT 1h
18h
1h
18h
100 500 1000 2500
228 202 92 58
– 128 36 6
397 206 261 168
459 162 18 4
Cl–HT
known anion exchange capacity (AEC) of the HT had been occupied by arsenite anions. The exchange of NO3–HT by As(III) is described by the following equation: Mg2 AlðOHÞ6d NO3 þ 0:33AsO3− 3
→Mg2 AlðOHÞ6d 0:33AsO3 þ NO−3 :
ð1Þ
The anion exchange capacity (AEC) of the HTs used in these experiments is in the vicinity of 300cmol(+)/kg, which would theoretically allow adsorption of about 1 mol (75 g) As(III)/kg of HT. However, the observed reduction of As concentration from 400 μg/L to 5μg/L represents an adsorption of only about 0.1% of this adsorption capacity, i.e. 80 mg As(III)/kg HT. The chemical similarity between phosphate and arsenite ions would indicate that the latter would be specifically adsorbed into HT, but clearly the extremely low concentration of As in solution lessen its competitive advantage against incumbent nitrate anions. In fact, the author observes the same ‘concentration dependency’ with phosphate adsorption into HT. In any event, the very limited amounts of As that will be present in ‘spent’ HT raises the possibility of it being employed for another purpose, as will be discussed later.
929
water and then further diluting this solution by a factor of 200. On analysis, this stock solution contained 400μg/L As. A ‘vessel’ was fabricated from 10 cm diameter PVC pipe that was about 20cm in length, by gluing a PVC disk at one end. A hole drilled in this disk allowed the insertion of a commercially available filter candle that was 5 cm in diameter and 8 cm in length. The apparatus stood on a collecting vessel as shown in Fig. 2. An amount of Cl–HT powder sufficient to cover the candle (350 g) was placed in the vessel and 1L of water containing As(V) at approximately 400 μg/L added. When water began to leave the vessel via the filter candle at about 200 mL/h, the volume of As(V) solution was maintained at about 1L by adding more solution at the same rate from an overhead container. The filtered water was collected in 500 mL aliquots for As determination by ICP-MS. 3.1.2. Results and discussion The concentrations of As(V) in the aliquots of filtered water are summarized in Table 2. The HT reduced As concentrations to very acceptable levels during the passage of 8.5 L of arsenic-containing water, but at that point, the rate of filtration slowed considerably when HT began to block the pores of the filter. Attempts to clear the filter with a stiff brush were only partly successful and it became clear that, apart from using a more appropriate filtering material (if available), it would be necessary to use a granular form of HT that would be resistant to dispersion.
Arsenic Solution Filter Candle
3. A ‘porous pot’ technique for arsenic removal Commonly used methods of separating microorganisms from drinking water in less-developed regions include the filtering of water through porous pots, or filter candles inserted in the base of vessels. It might therefore be feasible to add HT clay to these vessels in arsenic-affected areas, thereby removing this toxic substance as part of everyday water treatment practice.
Cl-HT
Filtered Solution
3.1. Experiment 1 3.1.1. Materials and methods A stock As(V) solution was prepared by dissolving 208mg of Na2AsO4·7H2O in 250mL of distilled
Fig. 2. Filter candle apparatus containing Cl–HT powder. Untreated water in the upper container can only reach lower container via the filter candle and after extended exposure to the Hydrotalcite placed in the upper container.
930
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
3.2. Experiment 2 3.2.1. Materials and methods Stable aggregates of Cl–HT were prepared by kneading wetted Cl–HT with a small amount of polyacrylamide crystals, drying, gently crushing and separating the 0.85–2mm fraction. The granules did not disintegrate after standing in water for 14 days. A pouch with an open top was fabricated from commercially available filter cloth and 170 g of Cl–HT aggregates added. This configuration allowed the drop-wise entry of As-bearing water in the top and its entry at the base after percolating through the entrapped HT. An As(III)/As(V) solution was prepared by mixing As (III) and As(V) solutions, aiming at an As(III)/As(V) ratio of 2:3. The total As concentration was found by analysis to be 990μg/L. This solution was percolated through the Cl–HT granules at approximately 50mL/h until 11L had passed and samples of percolate were collected at intervals for total As determination by ICP-MS. 3.2.2. Results and discussion The concentrations of total As in the filtrate samples are summarized in Table 3. The stabilized Cl–HT granules were able to reduce the relatively high concentration of As to levels that would be acceptable in practice. Most probably, even lower concentrations in the filtrate could be achieved by reducing the percolation rate or increasing the amount of Cl–HT in the filter pouch. Better still, an apparatus with smaller diameter and greater path length would allow greater contact time between the water and HT. 4. General discussion It is well known that the adsorptive properties of HT are greatly enhanced if the carbonate form (CO3– Table 2 Concentrations of As in leachate after sequential volumes of 400μg/L As(V) solution was passed through a filter candle apparatus containing 350g of Cl–HT Volume filtered (mL)
Arsenic concentration (μg/L)
500 1000 1500 2000 5000 5500 6000 6500 7000 8500
<1 2.3 3.0 2.7 2.8 2.5 2.3 2.2 2.0 2.4
Table 3 Concentrations of As in leachate after sequential volumes of a mixed As(III)/As(V) solution containing 990 μg/L As by continuous percolation through Cl–HT granules stabilized with polyacrylamide Volume filtered (mL)
Arsenic concentration (μg/L)
500 1500 2600 3450 4150 5080 6780 8330 9430 10,980
79 21 17 10 10 18 18 13 22 25
HT) is calcined, with the expulsion of CO2 and water to form a mixed Mg/Al oxide. When introduced into a salt solution, this mixed oxide will re-form into a hydrotalcite-like compound with the anions from the salt solution comprising the interlayer anions (Carlino, 1997). Preliminary experiments with calcined CO3– HT and a 10,000 mg/L As solution showed that As was adsorbed by the HT at 120 mg As/g calcined HT. This is equivalent to about 95 mg As adsorbed per gram of re-constituted HT, in line with the theoretical maximum for HT anion (arsenite) content. Though cost of calcination would make it more expensive to produce, it appears that calcined HT could be used in far less amounts, though the capacity for ‘carbonate poisoning’ from dissolved CO2 in the water would be high in the context of the technology discussed here. It is also likely that, re-formed in submicron particle size, the HT would quickly block filtration apparatus. The possibility of being able to use calcined CO3–HT should not be overlooked if appropriate filtration or separation technology were available. As mentioned earlier, any considerations of the use of HT in removal of As from drinking water would have to include those covering the fate of used HT. The experiments show that Cl–HT would be ‘spent’, at these very low As concentrations, after the adsorption of only about 80mg As/kg HT. This used HT could be easily saturated with phosphate by contacting it with, e.g., 0.1M phosphate solution: Mg2 AlðOHÞ6d NO3 þ 0:5HPO2− 4 →Mg2 AlðOHÞ6d 0:5HPO4 þ NO−3
ð2Þ
where the minute amount of As in the ‘spent’ HT is neglected. Phosphate saturated HT (PO4–HT) is an effective phosphatic fertilizer for soils that ‘fix’ soluble
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
phosphate, or where phosphate leaching is a problem (Gillman and Noble, 2005). The final P content of the product could be controlled by how much residual P from the phosphate saturating solution is allowed to remain, but a phosphate content of 10% P would be a realistic level, where 60% of the phosphate would be present in interlayer form (protected from fixation and leaching) and the remainder in soluble form to promote early plant establishment. A P application rate of 30 kg/ha of this Ascontaminated PO4–HT to soil would correspond to an HT application rate of 300kg HT/ha, resulting in an arsenic application rate of 24 g/ha. Incorporated into the upper 10cm of soil, each application of the HT product would raise the arsenic soil content by an estimated 24g per million kilograms of soil. This is an insignificant increase, but, should there be any concerns about arsenic entry into the food chain from this source, the fertilizer could be restricted to non-edible crops such as fibre and forest plantations. The introduction of an As removal technology based on HT would be very dependent on local requirements, cost and its attractiveness vis-a-vis alternative technologies. If used at ‘household’ level, a system of delivery of HT in appropriate (sachet?) form, combined with collection for central processing of used HT would be necessary. At ‘village’ level, a system of treating well water with larger amounts of pouched HT (possibly directly into the well) could be considered.
931
If HT could be produced at US$700 per ton and resold as fertilizer at, say, $100 per ton, and if 200g of HT was sufficient to treat 20 L of water, the product cost (i.e. exclusive of distribution costs) could be in the order of 12cents/day for a family consuming 20 L of drinking water. Finally, there is the possibility of producing HT incountry, thereby introducing a new industry, along with transport and fertilizer marketing. The product can be readily manufactured from waste materials such as the magnesium retained in brine from evaporated seawater in table salt manufacture and the aluminium in recycled cans. However, as mentioned above, relatively low-cost magnesite and bauxite are very good sources of these two elements. References Carlino S. Chemistry between the sheets. Chem Br 1997;33:59–62. Gillman GP, Noble AN. Environmentally manageable fertilizers: a new approach. Environ Qual Manag 2005;15(2):59–70. Nordstrom DK. Worldwide occurrences of arsenic in groundwater. Science 2002;296:2143–5. Ookubu A, Ooi K, Hayashi H. Preparation and phosphate ionexchange properties of a hydrotalcite-like compound. Langmuir 1993;9:1418–22. Tamagawa, S. Determination of permanent positive charge and retention of anions in soil. PhD thesis. School of Tropical Biology, James Cook University. Townsville. Queensland. Australia; 2003.
Technical note
A simple technology for arsenic removal from drinking water using hydrotalcite G.P. Gillman ⁎ CSIRO Land and Water, Private Mail Bag, Aitkenvale PO. Qld. 4814, Australia Received 1 August 2005; received in revised form 18 December 2005; accepted 16 January 2006 Available online 20 March 2006
Abstract The use of a synthetically prepared clay material, hydrotalcite (HT), for the removal of arsenite (As(III)) and arsenate (As(V)) from drinking water is described. Percolation through HT of water containing 500–1000μg/L As (levels often found in Ascontaminated well water) produced leachate with As levels well below 10 μg/L. The technology could be coupled to that used in less-developed regions for removing organisms from drinking water, viz. leaching through porous pots and filter candles. The ‘spent’ HT is easily converted into valuable phosphatic fertilizer that would have an insignificant effect on soil arsenic levels, thereby reducing the overall cost of manufacture and distribution. © 2006 Elsevier B.V. All rights reserved. Keywords: Arsenic removal; Drinking water; Hydrotalcite; Phosphate fertilizer
1. Introduction The presence of arsenic in drinking water, at levels shown to be a significant health hazard, is receiving increasing attention worldwide (Nordstrom, 2002). Levels of arsenic in excess of the World Health Organization recommended limit of 10μg/L can occur naturally in water bodies used as sources of drinking water, but the problem is exacerbated by the sinking of wells in areas of arsenic-rich geological strata. Values in the range 10–1000 μg/L have been reported and the UN Synthesis Report on Arsenic in Drinking Water (www.who.int/water_sanitation_health/dwq/arsenic3) provides an excellent overview of the extent of the
⁎ Tel.: +61 7 47538576; fax: +61 7 47538600. E-mail address: [email protected]. 0048-9697/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2006.01.036
problem and of technologies being evaluated for its solution. This paper canvasses the use of a synthetic clay mineral, hydrotalcite (HT), as a material for removing arsenic from water at household or communal level in less-developed regions. Hydrotalcite is essentially an aluminium-substituted brucite (Mg(OH)2) mineral and forms when a proportion of the magnesium ions in the structure is replaced by aluminium, leading to an excess of structural positive charge. The positively charged sheets align, with interlayer spacings being occupied by anions to maintain electrical neutrality. This process is the reverse of the adsorption reactions involving smectitic clays, where structural negative charge is balanced by interlayer cations. Hydrotalcite occurs naturally in specimen amounts under favourable conditions, but can be easily synthesised using inexpensive raw materials such as bauxite and magnesite (Gillman and Noble, 2005).
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
The term ‘hydrotalcite’ is generally reserved for the form wherein carbonate ions occupy the interlayer spacings. When other interlayer anions such as nitrate, chloride and phosphate are dominant, the materials are referred to as ‘hydrotalcite-like compounds’. In this paper, the general term ‘hydrotalcite’ (HT) will be applied irrespective of the charge-balancing anion and an anion prefix used to describe a particular form, e.g. NO3–HT and Cl–HT for nitrate and chloride hydrotalcite, respectively. Hydrotalcite specificity for anions follows the order CO3 > PO4 > SO4 > Cl > NO3 (Tamagawa, 2003) and, in the absence of carbonate, a chloride or nitrate HT will readily adsorb phosphate from solution as was demonstrated by Ookubu et al. (1993). The similarity between the structure of the arsenic anions arsenite and arsenate to that of phosphate anions indicates that HT could be used to effectively remove arsenic from water. This paper therefore presents results of laboratory experiments comparing the capacity of several HT materials to adsorb arsenic, followed by prototype technology studies that might be adaptable for use in regions referred to above. 2. Adsorption of As(III) by Cl–HT, NO3–HT and CO3–HT 2.1. Materials and methods Arsenic measurements were made with a Varian UltraMass 700 ICP-MS, externally calibrated using a commercially available standard solution, and also an In internal standard to correct for instrument drift and matrix effects. The limit of detection for As under the conditions used was 0.42μg/L. To obtain an approximate 500 μg/L As(III) stock solution, a saturated arsenic trioxide (As2O3) solution that was approximately 10,000 mg/L As was diluted by a factor of 20,000. The actual As(III) concentration was determined by ICP-MS to be 432 μg/L. The synthesis of HT is easily effected by coprecipitation of a Mg/Al double hydroxide. The ratio of Mg/Al should be held between 0.25 and 0.33, and pH must be higher than that at which either Mg or Al hydroxide would separately precipitate. In the three syntheses described below, pH was maintained at or above 9.5. XRD analysis was carried out on the three HT compounds to confirm that true HT materials had been formed. An NO3–HT slurry was prepared by adding 500mL of a mixed nitrate solution (1M Mg(NO3)2, 0.5M Al (NO3)3) via a peristaltic pump at 450 mL/min to 250mL
927
of concentrated ammonia (25%) with vigorous stirring. The final pH of the HT precipitate was 9.5. Following a number of centrifuge-washings to the point of dispersion to virtually remove all ammonium nitrate, the NO3–HT ‘cake’ (moisture content 87%) was stored in a sealed container. A CO3–HT slurry was prepared by bubbling CO2 through an NO3–HT slurry (see above) for 2 h. After centrifuge-washing to the point of dispersion to remove residual nitrate, the CO3–HT ‘cake” (moisture content 89%) was stored in a sealed container. A Cl–HT solid was prepared using the method of Ookubu et al. (1993). A 350mL solution of mixed chloride (0.86 M MgCl2, 0.4 M AlCl3) and, separately, a 2.5 M NaOH solution were pumped into a beaker under vigorously stirred conditions, with adjustment of pump rates to maintain pH of the resultant slurry at about 9.5. The amount of alkali consumed was 420 mL. The slurry was aged at 80 °C for 8 h, followed by centrifugewashing to the point of dispersion and drying of the ‘cake’ at 80°C. In separate experiments, 0.1, 0.5, 1.0 and 2.5 g (o. d. basis) of NO3–HT cake or Cl–HT powder were added to 500mL aliquots of the stock As(III) solution that initially contained 432 μg/LAs. In another experiment, 2.5g (o.d. basis) of CO3–HT cake was added to 500 mL of As stock solution. The suspensions were stirred in an open beaker for 1h and a small sample of supernatant solution then taken for As analysis. After a further 17 h stirring, another supernatant solution sample was taken. The samples were passed through a 0.45 μ membrane filter prior to As determination. 2.2. Results and discussion XRD patterns of the three synthesised hydrotalcite clays are presented in Fig. 1. Basal spacings calculated from the 003 reflections (near 10° 2ϴ) were 0.89nm, 0.77 nm and 0.78nm for NO3–HT, CO3–HT and Cl– HT, respectively, values that are in very good agreement with published basal spacings. The sharpness of the 003 and 006 peaks demonstrates the layered structure of these materials. The concentrations of As(III) remaining in solution following contact with NO3–HT and Cl–HT, over a range of HT addition rates, and for two contact times, are summarized in Table 1. (Addition of HT in carbonate form did not alter solution As concentration and CO3– HT will not be further discussed.) When added in sufficient quantity, both NO3–HT and Cl–HT reduced As(III) concentrations from those
928
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
Fig. 1. XRD patterns of dried powder samples of HT used in this study. (a) NO3–HT, (b) CO3–HT and (c) Cl–HT.
commonly found in water from tube-wells passing through arsenic-bearing rock strata, to below the WHO recommended upper limit of 10μg/L−1.
The adsorption appears to be diffusion controlled, with 18 h of contact resulting in removal of more arsenic than after 1h of contact. Even then, only a fraction of the
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931 Table 1 Residual As(III) concentration (μg/L) in water originally containing 432μg/L As(III) after contact with increasing amounts of NO3–HT and Cl–HT for 1h and 18h HT addn. rate (mg HT/500mL H2O)
NO3–HT 1h
18h
1h
18h
100 500 1000 2500
228 202 92 58
– 128 36 6
397 206 261 168
459 162 18 4
Cl–HT
known anion exchange capacity (AEC) of the HT had been occupied by arsenite anions. The exchange of NO3–HT by As(III) is described by the following equation: Mg2 AlðOHÞ6d NO3 þ 0:33AsO3− 3
→Mg2 AlðOHÞ6d 0:33AsO3 þ NO−3 :
ð1Þ
The anion exchange capacity (AEC) of the HTs used in these experiments is in the vicinity of 300cmol(+)/kg, which would theoretically allow adsorption of about 1 mol (75 g) As(III)/kg of HT. However, the observed reduction of As concentration from 400 μg/L to 5μg/L represents an adsorption of only about 0.1% of this adsorption capacity, i.e. 80 mg As(III)/kg HT. The chemical similarity between phosphate and arsenite ions would indicate that the latter would be specifically adsorbed into HT, but clearly the extremely low concentration of As in solution lessen its competitive advantage against incumbent nitrate anions. In fact, the author observes the same ‘concentration dependency’ with phosphate adsorption into HT. In any event, the very limited amounts of As that will be present in ‘spent’ HT raises the possibility of it being employed for another purpose, as will be discussed later.
929
water and then further diluting this solution by a factor of 200. On analysis, this stock solution contained 400μg/L As. A ‘vessel’ was fabricated from 10 cm diameter PVC pipe that was about 20cm in length, by gluing a PVC disk at one end. A hole drilled in this disk allowed the insertion of a commercially available filter candle that was 5 cm in diameter and 8 cm in length. The apparatus stood on a collecting vessel as shown in Fig. 2. An amount of Cl–HT powder sufficient to cover the candle (350 g) was placed in the vessel and 1L of water containing As(V) at approximately 400 μg/L added. When water began to leave the vessel via the filter candle at about 200 mL/h, the volume of As(V) solution was maintained at about 1L by adding more solution at the same rate from an overhead container. The filtered water was collected in 500 mL aliquots for As determination by ICP-MS. 3.1.2. Results and discussion The concentrations of As(V) in the aliquots of filtered water are summarized in Table 2. The HT reduced As concentrations to very acceptable levels during the passage of 8.5 L of arsenic-containing water, but at that point, the rate of filtration slowed considerably when HT began to block the pores of the filter. Attempts to clear the filter with a stiff brush were only partly successful and it became clear that, apart from using a more appropriate filtering material (if available), it would be necessary to use a granular form of HT that would be resistant to dispersion.
Arsenic Solution Filter Candle
3. A ‘porous pot’ technique for arsenic removal Commonly used methods of separating microorganisms from drinking water in less-developed regions include the filtering of water through porous pots, or filter candles inserted in the base of vessels. It might therefore be feasible to add HT clay to these vessels in arsenic-affected areas, thereby removing this toxic substance as part of everyday water treatment practice.
Cl-HT
Filtered Solution
3.1. Experiment 1 3.1.1. Materials and methods A stock As(V) solution was prepared by dissolving 208mg of Na2AsO4·7H2O in 250mL of distilled
Fig. 2. Filter candle apparatus containing Cl–HT powder. Untreated water in the upper container can only reach lower container via the filter candle and after extended exposure to the Hydrotalcite placed in the upper container.
930
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
3.2. Experiment 2 3.2.1. Materials and methods Stable aggregates of Cl–HT were prepared by kneading wetted Cl–HT with a small amount of polyacrylamide crystals, drying, gently crushing and separating the 0.85–2mm fraction. The granules did not disintegrate after standing in water for 14 days. A pouch with an open top was fabricated from commercially available filter cloth and 170 g of Cl–HT aggregates added. This configuration allowed the drop-wise entry of As-bearing water in the top and its entry at the base after percolating through the entrapped HT. An As(III)/As(V) solution was prepared by mixing As (III) and As(V) solutions, aiming at an As(III)/As(V) ratio of 2:3. The total As concentration was found by analysis to be 990μg/L. This solution was percolated through the Cl–HT granules at approximately 50mL/h until 11L had passed and samples of percolate were collected at intervals for total As determination by ICP-MS. 3.2.2. Results and discussion The concentrations of total As in the filtrate samples are summarized in Table 3. The stabilized Cl–HT granules were able to reduce the relatively high concentration of As to levels that would be acceptable in practice. Most probably, even lower concentrations in the filtrate could be achieved by reducing the percolation rate or increasing the amount of Cl–HT in the filter pouch. Better still, an apparatus with smaller diameter and greater path length would allow greater contact time between the water and HT. 4. General discussion It is well known that the adsorptive properties of HT are greatly enhanced if the carbonate form (CO3– Table 2 Concentrations of As in leachate after sequential volumes of 400μg/L As(V) solution was passed through a filter candle apparatus containing 350g of Cl–HT Volume filtered (mL)
Arsenic concentration (μg/L)
500 1000 1500 2000 5000 5500 6000 6500 7000 8500
<1 2.3 3.0 2.7 2.8 2.5 2.3 2.2 2.0 2.4
Table 3 Concentrations of As in leachate after sequential volumes of a mixed As(III)/As(V) solution containing 990 μg/L As by continuous percolation through Cl–HT granules stabilized with polyacrylamide Volume filtered (mL)
Arsenic concentration (μg/L)
500 1500 2600 3450 4150 5080 6780 8330 9430 10,980
79 21 17 10 10 18 18 13 22 25
HT) is calcined, with the expulsion of CO2 and water to form a mixed Mg/Al oxide. When introduced into a salt solution, this mixed oxide will re-form into a hydrotalcite-like compound with the anions from the salt solution comprising the interlayer anions (Carlino, 1997). Preliminary experiments with calcined CO3– HT and a 10,000 mg/L As solution showed that As was adsorbed by the HT at 120 mg As/g calcined HT. This is equivalent to about 95 mg As adsorbed per gram of re-constituted HT, in line with the theoretical maximum for HT anion (arsenite) content. Though cost of calcination would make it more expensive to produce, it appears that calcined HT could be used in far less amounts, though the capacity for ‘carbonate poisoning’ from dissolved CO2 in the water would be high in the context of the technology discussed here. It is also likely that, re-formed in submicron particle size, the HT would quickly block filtration apparatus. The possibility of being able to use calcined CO3–HT should not be overlooked if appropriate filtration or separation technology were available. As mentioned earlier, any considerations of the use of HT in removal of As from drinking water would have to include those covering the fate of used HT. The experiments show that Cl–HT would be ‘spent’, at these very low As concentrations, after the adsorption of only about 80mg As/kg HT. This used HT could be easily saturated with phosphate by contacting it with, e.g., 0.1M phosphate solution: Mg2 AlðOHÞ6d NO3 þ 0:5HPO2− 4 →Mg2 AlðOHÞ6d 0:5HPO4 þ NO−3
ð2Þ
where the minute amount of As in the ‘spent’ HT is neglected. Phosphate saturated HT (PO4–HT) is an effective phosphatic fertilizer for soils that ‘fix’ soluble
G.P. Gillman / Science of the Total Environment 366 (2006) 926–931
phosphate, or where phosphate leaching is a problem (Gillman and Noble, 2005). The final P content of the product could be controlled by how much residual P from the phosphate saturating solution is allowed to remain, but a phosphate content of 10% P would be a realistic level, where 60% of the phosphate would be present in interlayer form (protected from fixation and leaching) and the remainder in soluble form to promote early plant establishment. A P application rate of 30 kg/ha of this Ascontaminated PO4–HT to soil would correspond to an HT application rate of 300kg HT/ha, resulting in an arsenic application rate of 24 g/ha. Incorporated into the upper 10cm of soil, each application of the HT product would raise the arsenic soil content by an estimated 24g per million kilograms of soil. This is an insignificant increase, but, should there be any concerns about arsenic entry into the food chain from this source, the fertilizer could be restricted to non-edible crops such as fibre and forest plantations. The introduction of an As removal technology based on HT would be very dependent on local requirements, cost and its attractiveness vis-a-vis alternative technologies. If used at ‘household’ level, a system of delivery of HT in appropriate (sachet?) form, combined with collection for central processing of used HT would be necessary. At ‘village’ level, a system of treating well water with larger amounts of pouched HT (possibly directly into the well) could be considered.
931
If HT could be produced at US$700 per ton and resold as fertilizer at, say, $100 per ton, and if 200g of HT was sufficient to treat 20 L of water, the product cost (i.e. exclusive of distribution costs) could be in the order of 12cents/day for a family consuming 20 L of drinking water. Finally, there is the possibility of producing HT incountry, thereby introducing a new industry, along with transport and fertilizer marketing. The product can be readily manufactured from waste materials such as the magnesium retained in brine from evaporated seawater in table salt manufacture and the aluminium in recycled cans. However, as mentioned above, relatively low-cost magnesite and bauxite are very good sources of these two elements. References Carlino S. Chemistry between the sheets. Chem Br 1997;33:59–62. Gillman GP, Noble AN. Environmentally manageable fertilizers: a new approach. Environ Qual Manag 2005;15(2):59–70. Nordstrom DK. Worldwide occurrences of arsenic in groundwater. Science 2002;296:2143–5. Ookubu A, Ooi K, Hayashi H. Preparation and phosphate ionexchange properties of a hydrotalcite-like compound. Langmuir 1993;9:1418–22. Tamagawa, S. Determination of permanent positive charge and retention of anions in soil. PhD thesis. School of Tropical Biology, James Cook University. Townsville. Queensland. Australia; 2003.