The objective of this work was to verify the sorption properties of granular filter materials (GEH, READ-As) during the process of removing antimony from water. The pilot tests showed that the use of sorption materials could possibly decrease the antimony content in water to the values limited for drinking water (5 μg/L Sb). A more suitable adsorbent for removing antimony was READ-As. At a concentration of antimony in raw water ranging from 21.5 to 31.1 μg/L, a filtration rate of 5.58 m/h, the value of the bed volume of 3,967, and the adsorption capacity of 128.4 μg/g, which was achieved at a breakthrough concentration of 5 μg Sb/L, were determined. The surface characteristics of the sorption materials used through the physical adsorption of nitrogen, mercury porosimetry, X-ray microanalysis, and scanning electron microscopy (SEM) were studied.

INTRODUCTION

The Dúbrava group water supply system in Slovakia was built in connection with the construction of the Liptovská Mara Water Reservoir. The group water supply system was supplied with water from the Dúbrava water resource (capacity of about 40 L/s). The water resource included three springs (Brdáre, Močidlo, and Škripeň). Škripeň is currently the only spring used for the drinking water supply, since it contains no antimony. The two other springs are contaminated with antimony.

A deposit of antimony ore is considered to be the main cause of the increased antimony concentrations in the springs of Močidlo and Brdáre. Furthermore, a high concentration of antimony in water from mining activities as well as from rain falling on antimony-rich waste rock piles and a sludge-settling pond has led to higher concentrations of antimony in the relevant water resources.

Antimony is present in water as Sb(III) and Sb(V), depending on the water pH, the oxidation-reduction potential, the Sb(III)/Sb(V) ratio and the concentration of oxygen. It is found mainly in the form of antimonate as oxyanions and or it can be present in the form of antimonite (WHO 2003).

Antimony is a toxic heavy metal with effects similar to those of arsenic and lead. Intoxication by antimony is not as severe as that from arsenic, since the antimony compounds are absorbed more slowly. Findings on the health aspects of certain heavy metals in drinking water are included in several publications (AWWA 1990; US EPA 1984). The World Health Organization (WHO) and institutions dealing with the monitoring of carcinogenity have not yet classified antimony as a carcinogen.

In Slovakia, an acceptable concentration of heavy metals in drinking water is defined under Government Regulation No. 496/2010 on drinking water. The limit concentration for antimony is 5 μg/L. This limit value is in accordance with the WHO recommendations (WHO 2011) and EU Directive (98/83/EC 1998).

There are several technological methods for removing heavy metal in the water treatment process: precipitation, ion exchange, membrane processes, adsorption, electrochemical processes, and biological methods.

Adsorption using an appropriate sorption material is preferred for water treatment when considering small water resources. Sorption is a simple (regarding its operation) and effective method of heavy metal removal, mainly because of the wide range of sorption materials that can be used in this process. The most frequently tested sorbents are as follows: iron oxides and oxyhydroxides, activated alumina, sand covered by iron hydroxide, and media containing TiO2 or MnO2 layers on their surface, etc. (Bailey et al. 1999).

There are very few references related to the removal of antimony from water by sorption materials. The literature mostly describes the use of iron oxides and oxyhydroxides for the removal of arsenic from water. A number of experiments and model studies on the adsorption of arsenic and other heavy metals are described in various publications (Jekel & Seith 2000; US EPA 2003; Rubel 2003; Sperlich et al. 2005; Mohan & Pittman 2007; Westerhoff et al. 2008; Guan et al. 2008; Zeng et al. 2008; Nguyen et al. 2011). These studies describe sorption processes at different pH values, the initial heavy metal ion concentrations in water, the solid/liquid ratio, the particle size of a sorption material, and the temperature and composition of the water to be treated (concentration of iron, manganese, phosphorus, silicon, fluorides, sulfates, organic matter, etc.).

The objective of this work was to verify the sorption properties of granular filter materials (GEH, READ-As) in the Dúbrava water resource during the process of removing antimony from the water.

MATERIALS AND METHODS

Properties of sorption materials

The GEH was obtained from a supplier (GEH Wasserchemie, Osnabrück, Germany). GEH is a sorption material developed by the Chair of Water Quality Control at the Technical University of Berlin for the purpose of removing arsenic from water. GEH consists of ferric hydroxide and oxyhydroxide with a dry solid content of 57% (±10%) by mass and 43–48% by mass moisture content, GEH is not regenerated (Driehaus et al. 1998; GEH Wasserchemie, 2014). GEH is highly selective toward arsenate; therefore, it requires an initial oxidation step in the presence of arsenite. Increasing phosphate and sulfate concentrations in influent water greatly reduces arsenic removal (Driehaus et al. 1998).

READ-As was obtained from S-Metalltech. It was developed by Nihon Kaisui Co., Ltd in Japan. READ-As is a granular hydrous cerium oxide adsorbent (88% CeO2 and 12% ethylene-vinyl alcohol copolymer), which displays a high degree of selectivity for arsenic ions under a broad range of conditions and effectively adsorbs both arsenite and arsenate. The oxidation of arsenite to arsenate is not needed for arsenic removal nor is an adjustment of the pH required before or after treatment. This material can be regenerated by the addition of 25% sodium hydroxide, then 12% sodium hypochlorite solution and finallly, washing with water. Regenerated material needs neutralization with 35% HCl and has to be washed with water before using.

Depending on the amount of the adsorbent and the water chemistry, the regeneration of the adsorbent saturated with arsenic should be accomplished in 4–12 months. The sorbent can be used for eight to 14 regeneration cycles. This waste material is not hazardous (Shin Nihon Salt Co., Ltd 2000; Jain & Singh 2012; S-Metalltech 2014).

Table 1 includes the basic physical–chemical properties of the materials used in the test.

Table 1

Physical–chemical properties of the selected sorption materials

ParameterREAD-AsGEH
Matrix/active agent   + 52–57%  
Physical form Moist granular Moist granular 
Color Yellow Dark brown 
Bulk density (g/cm30.76 1.22–1.29 
Grain size (mm) 0.3–1.0 0.32–2.0 
ParameterREAD-AsGEH
Matrix/active agent   + 52–57%  
Physical form Moist granular Moist granular 
Color Yellow Dark brown 
Bulk density (g/cm30.76 1.22–1.29 
Grain size (mm) 0.3–1.0 0.32–2.0 

Characterization techniques

The textural properties of the sorption materials were studied through the physical adsorption of nitrogen at the temperature of liquid nitrogen (−197 °C) by the volumetric method using ASAP 2400 (Micromeritics). The samples were activated before measurement for 15 h at a temperature of 350 oC and in a vacuum of less than 2 Pa. The adsorption data were processed using a standard Brunauer–Emmett–Teller (BET)-isotherm with the linearization in the range of 0.05 ≤ p/p0 ≤ 0.3 (specific surface area SBET). The desorption branches of the isotherms were used for calculating the pore size distribution by the standard Barrett-Joyner-Halenda (BJH) method (maximum on the pore size distribution curve Rp). The total pore volume (Va) was estimated from the nitrogen-adsorbed quantity at a relative pressure of p/p0 = 0.99. The meso-macroporous texture was verified by mercury porosimetry using AutoPore IV 9500 (Micromeritics).

The sorption materials were analyzed with a Jeol JXA-840A X-ray microanalyzer using the energy dispersive spectroscopy (EDS) system. The samples were evaporated with a thin carbon layer to ensure their electrical conductivity. The quantitative analysis of every sample was determined in three different areas, and the mean value was then calculated. The microstructure was observed by scanning electron microscopy (SEM, TESLA BS 300).

Model adsorption system

The effectiveness of the antimony removal was verified using adsorption columns containing the selected sorption material. The adsorption column was made of glass with an internal diameter of 5.0 cm. The height of the media ranged from 48 to 49 cm, and the total height of the glass column was 80 cm.

Raw water (the Brdáre spring) passed through the adsorption system, and the concentration of antimony in the raw and treated water was monitored at the outlets of the adsorption columns. Simultaneously, the flow rates were measured at the outlet of each column. A system of several valves was used for feeding the water for the adsorption system and for the filter backwash (from bottom to top) as well as for regulating the filtration rates.

Quality of raw water (Brdáre spring)

The results of analysis of raw water (RW) from the Brdáre spring are listed in Table 2. The antimony concentration during the pilot tests ranged from 21.5 to 31.1 μg/L (an average of 27.7 μg/L ± 3.41 RSD). No other heavy metals were present in the Dúbrava water resource.

Table 2

Results of RW analysis (analysis of some selected parameters)

ParameterUnitRWParameterUnitRW
pH  8.2 Fe total mg/L 0.06 
Conductivity mS/m 22.4 Mn mg/L 0.001 
Alkalinity (ANC) mmol/L 2.871 Ca2+ mg/L 28.86 
Acidity (BNC) mmol/L 0.139 Mg2+ mg/L 16.14 
TOC mg/L 0.64  mg/L 15.38 
TDS mg/L 190 F mg/L 0.05 
Sb μg/L 31.0  mg/L 0.02 
ParameterUnitRWParameterUnitRW
pH  8.2 Fe total mg/L 0.06 
Conductivity mS/m 22.4 Mn mg/L 0.001 
Alkalinity (ANC) mmol/L 2.871 Ca2+ mg/L 28.86 
Acidity (BNC) mmol/L 0.139 Mg2+ mg/L 16.14 
TOC mg/L 0.64  mg/L 15.38 
TDS mg/L 190 F mg/L 0.05 
Sb μg/L 31.0  mg/L 0.02 

TOC: total organic carbon; TDS: total dissolved solids.

Analytical method

Antimony samples were collected into plastic bottles and immediately acidified with highly pure nitric acid (Merck). All bottles were submerged in 10% nitric acid solution over 3 days and triple rinsed with deionized water. Agilent 7500CE inductively coupled plasma mass spectrometry (ICP-MS) (ORS technology) was used to determine antimony concentration in solution. The detection limit for Sb by ICP-MS was 1 μg/L.

RESULTS AND DISCUSSION

Within the frame of the experiments the chemical composition and surface properties of the sorption materials used were studied. The chemical composition was determined using the X-ray microanalysis method; the values are listed in Table 3.

Table 3

Chemical composition of selected sorption materials

Compound in mass (%)
MaterialAl2 O3SiO2P2 O3SOxCl2 OK2 OCaOCeO2Fe2 O3
READ-As – 1.30 – – 0.39 – – 98.30 – 
GEH 1.74 3.05 0.21 0.54 2.27 0.08 0.18 – 91.92 
Compound in mass (%)
MaterialAl2 O3SiO2P2 O3SOxCl2 OK2 OCaOCeO2Fe2 O3
READ-As – 1.30 – – 0.39 – – 98.30 – 
GEH 1.74 3.05 0.21 0.54 2.27 0.08 0.18 – 91.92 

The shape and external surface of the READ-As and GEH sorption materials were scanned by an electron microscope. Figures 1 and 2 illustrate the differences in the character of the surfaces.

Figure 1

The microstructure of the READ-As (40 ×, 250 ×, 1,000 ×, and 5,000× magnification).

Figure 1

The microstructure of the READ-As (40 ×, 250 ×, 1,000 ×, and 5,000× magnification).

Figure 2

The microstructure of the GEH (40 ×, 500 ×, 1,000 ×, and 5,000× magnification).

Figure 2

The microstructure of the GEH (40 ×, 500 ×, 1,000 ×, and 5,000× magnification).

Table 4 includes the surface properties of sorption materials determined by physical adsorption of nitrogen and mercury porosimetry.

Table 4

Texture properties of sorbent materials

Physical adsorption of nitrogenMercury porosimetry
Material SBET (m2/g) Va (cm3/g) Rp (nm) Rp (nm) 
READ-As 42.8 0.154 20 20, 80–100 
GEH 190 0.295 2.35 < 2.5 
Physical adsorption of nitrogenMercury porosimetry
Material SBET (m2/g) Va (cm3/g) Rp (nm) Rp (nm) 
READ-As 42.8 0.154 20 20, 80–100 
GEH 190 0.295 2.35 < 2.5 

The adsorption isotherms of the READ-As and GEH samples (Figure 3) are significantly different.

Figure 3

Adsorption isotherms for sorption material GEH and READ-As used in the test.

Figure 3

Adsorption isotherms for sorption material GEH and READ-As used in the test.

The GEH sample, which has a specific surface area of 190 m2/g and a total pore volume of almost 0.3 cm3/g, is a typically mesoporous material with a hysteresis loop on a nitrogen adsorption–desorption isotherm at a relative pressure of p/p0 = 0.6–0.8, resulting in a relatively sharp pore distribution calculated from the desorption branch of the nitrogen isotherm (Figure 4, left) with a maximum at a pore radius of about Rp = 2.35 nm (diameter 4.7 nm). This texture is also supported by the results from the mercury porosimetry (Figure 5, left) which shows a pore structure with a pore radius smaller than about 2.5 nm. Such a texture also corresponds to the relatively smooth surface visible on the SEM pictures, even at the highest magnification (Figure 2).

Figure 4

Pore size distributions for sorption materials GEH (left) and READ-As (right) from nitrogen desorption isotherms.

Figure 4

Pore size distributions for sorption materials GEH (left) and READ-As (right) from nitrogen desorption isotherms.

Figure 5

Pore size distribution for sorption materials GEH (left) and READ-As (right) from mercury porosimetry.

Figure 5

Pore size distribution for sorption materials GEH (left) and READ-As (right) from mercury porosimetry.

On the other hand, the texture of the READ-As with a specific surface area of only 42.8 m2/g is completely different. The nitrogen adsorption isotherm (Figure 3) shows a narrow hysteresis loop at a relative pressure of over 0.7, indicating the presence of mesopores of a much wider diameter. The adsorption isotherm shows a rapid increase in the relative pressure over 0.9, which corresponds to the presence of macropores. The pore size distribution based on nitrogen desorption (Figure 4, right) is relatively wide with a maximum at Rp = 20 nm (diameter of 40 nm), which is close to the upper limit of the definition of a mesopore. A small maximum at about Rp = 1.9 nm is an artefact that does not correspond to actual pores (Groen et al. 2003).

What is interesting are the results of the mercury porosimetry of the READ-As sample (Figure 5, right), which shows a double porous structure. The lower pore structure at Rp = 20 nm precisely supports the results of nitrogen desorption. Moreover, the pore-size distribution from the mercury porosimetry shows a second pore structure in the range of the macropores with a maximum at Rp = 80–100 nm (a diameter of about 200 nm). The presence of macropores is also clearly seen on the SEM picture of the READ-As sample at the highest magnification. The total pore volume of 0.154 cm3/g is thus concentrated in the meso- and macroporous internal pore structures.

The pilot test was aimed at monitoring the effectiveness of the GEH and READ-As sorption materials for the removal of antimony from water. The antimony concentrations in the RW were in a range from 21.5 to 31.1 μg/L (an average of 27.7 μg/L with a standard deviation ± 3.41). The filtration conditions are shown in Table 5.

Table 5

Adsorption conditions of the model tests

ParameterREAD-AsGEH
Grain size (mm) 0.3–1.0 0.32–2.0 
Medium height (cm) 48 49 
Volume of adsorption column (cm3942.5 962.1 
Mass of adsorption column (g) 717.5 1204.8 
Average flow through column (mL/min) 182.6 181.8 
Average filtration rate (m/h) 5.58 5.55 
Empty bed contact time (EBCT) (min) 5.16 5.29 
ParameterREAD-AsGEH
Grain size (mm) 0.3–1.0 0.32–2.0 
Medium height (cm) 48 49 
Volume of adsorption column (cm3942.5 962.1 
Mass of adsorption column (g) 717.5 1204.8 
Average flow through column (mL/min) 182.6 181.8 
Average filtration rate (m/h) 5.58 5.55 
Empty bed contact time (EBCT) (min) 5.16 5.29 

Figure 6 shows the breakthrough curves of the antimony as a function of the water volumes treated in terms of bed volumes and the values of the bed volume or the adsorption capacity for each sorption material when reaching the limit concentration of antimony (5 μg/L).

Figure 6

Breakthrough curves of antimony (left) and the bed volumes and adsorption capacity achieved (right) at a breakthrough concentration of antimony of 5 μg/L.

Figure 6

Breakthrough curves of antimony (left) and the bed volumes and adsorption capacity achieved (right) at a breakthrough concentration of antimony of 5 μg/L.

As for the results presented in Figure 6, it can be concluded that READ-As is a more suitable material for antimony removal compared to the GEH sorbent used in the test. The following bed volumes were measured for the antimony concentration at the outlet of the adsorbent media: bed volume 3,967 for the READ-As (the time of the wave breakthrough reached 336 h) and 3,236 for the GEH (the time of the wave breakthrough reached 311 h). The adsorption capacities were as follows: READ-As = 128.4 μg/g and GEH = 63.0 μg/g.

According to the material balance of the antimony in these experiments, the amount of adsorbed antimony in the adsorption media when reaching the limit concentration of 5 μg/L Sb at the outlet of the media was 92,166 μg for the READ-As and 75,945 μg for the GEH.

The antimony sorption capacity depends also on the quantity and character of adsorption centers on adsorbents where sorption of antimony species proceeds. Because, the chemical character of both adsorbents is very different, it is very hard (or impossible) to compare their adsorption capacities on the basis of measured characteristics.

CONCLUSIONS

The technological tests performed on the groundwater from the spring in Dúbrava showed that the use of sorption materials can possibly decrease the content of antimony in water to the values limited by Government Regulation No. 496/2010 on drinking water. Based on the pilot tests, a more suitable sorption material for removing antimony from water is the READ-As.

At a concentration of antimony in RW ranging from 21.5 to 31.1 μg/L, the filtration rate of 5.58 m/h and empty bed contact time (EBCT) of 5.16 min, the value of the bed volume 3,967, and the adsorption capacity of 128.4 μg/g achieved at a breakthrough concentration of 5 μg Sb/L, were determined.

The results proved that the GEH materials can also be used to decrease the concentration of Sb in drinking water below the limit value of 5 μg/L. The adsorption capacities and bed volumes are lower for these sorption materials in comparison to the READ-As.

The filtration rate, height of the filter media (i.e., EBCT), and antimony concentrations in RW have an influence on the value of the adsorption capacity and bed volume. Therefore, pilot tests to verify these values directly on a water source are necessary.

The efficiency of antimony removal from water under dynamic conditions affects not only the interaction of the antimony with the surface of the sorption material, but also mainly the pore size (mesopores and macropores), which may affect the rate of internal transport. The internal transport in pores also depends on the ratio between the size of adsorbate species and pore size of adsorbent, so the increase of pore size positively improves the internal transport of antimony species. Therefore, the specific surface of each material and pore size distribution was determined by the physical adsorption of nitrogen. The mercury porosimetry completed the results of the physical adsorption in terms of the presence of macropores. On the basis of these measurements, it can be concluded that the READ-As has a smaller specific surface than the GEH, but the size of the mesopores, moreover, the presence of the macropores, increase the rate of internal transport and, by that, the ability to capture antimony on the surface too.

ACKNOWLEDGEMENTS

Experimental measurements were carried out with the financial support of the APVV-0379-07 project from the Slovak Research and Development Agency, VEGA 1/1243/12 and VEGA 1/0228/12 research projects from the Scientific Grant Agency of the Slovak Ministry of Education. Thank belongs to O. Solcova from Institute of Chemical Process Fundamentals of the AS-CR, Prague, for mercury-porosimetry characterization of sorption media.

REFERENCES

REFERENCES
AWWA
1999
Water Quality and Treatment. A Handbook of Community Water Supplies
, 5th edn.
McGraw-Hill Inc., USA
.
Bailey
S. E.
Olin
T. J.
Bricka
R. M.
Adrian
D. D.
1999
A review of potentially low – cost sorbents for heavy metals
.
Water Research
33
,
2469
2479
.
Driehaus
W.
Jekel
M.
Hildebrandt
U.
1998
Granular Ferric Hydroxide – a new adsorbent for the removal of arsenic from natural water
.
Journal of Water Supply: Research and Technology-Aqua
47
,
30
35
.
EU Directive 98/83/EC
1998
Quality of water intended for human consumption
.
Jekel
M.
Seith
R.
2000
Comparison of conventional and new techniques for the removal of arsenic in a full scale water treatment plant
.
Water Supply
18
,
628
631
.
Mohan
D.
Pittman
Ch. U.
Jr
2007
Arsenic removal from water/wastewater using adsorbents – a critical review
.
Journal of Hazardous Materials
142
,
1
53
.
Rubel
F.
Jr
2003
Design manual: removal of arsenic from drinking water by adsorptive media (EPA/600/R-03/019)
.
Shin Nihon Salt Co., Ltd
2000
Report on Performance of Read-F Arsenic Removal Unit (ARU)
.
S-Metalltech
2014
.
US EPA
1984
Antimony: An Environmental and Health Effects Assessment
.
US Environmental Protection Agency, Office of Drinking Water
,
Washington, DC, USA
US EPA
2003
Arsenic Treatment Technology Evaluation Handbook for Small Systems
,
Office of Water (4606M), EPA 816-R-03–014
.
Westerhoff
P.
Benn
T.
Chen
A. S. C.
Wang
L.
Cumming
L. J.
2008
Assessing arsenic removal by metal (hydr)oxide adsorptive media using rapid small Scale Column Test. EPA/600/R-08/051
.
WHO
2003
Antimony in Drinking-Water. Background Document for Preparation of WHO Guidelines for Drinking-Water Quality
.
WHO
,
Geneva
.
WHO
2011
Guidelines for Drinking-Water Quality
, 4th edn.
WHO
,
Geneva
.