- Email: [email protected]
Adsorptive removal of phosphorus from aqueous solution using sponge iron and zeolite
Journal of Colloid and Interface Science 402 (2013) 246–252
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Adsorptive removal of phosphorus from aqueous solution using sponge iron and zeolite Cheng Jiang a, Liyue Jia b, Yiliang He a,⇑, Bo Zhang a, George Kirumba a, Jie Xie a a b
School of Environmental Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China School of Urban Construction, Hebei University of Engineering, Handan 056038, China
a r t i c l e
i n f o
Article history: Received 26 February 2013 Accepted 25 March 2013 Available online 11 April 2013 Keywords: Sponge iron Zeolite Phosphorus Adsorptive removal
a b s t r a c t Phosphorus adsorptive removal is an important and efficient treatment process in constructed subsurface flow wetlands. Many materials have been proposed for removal of excess phosphorus from wastewater. Selecting a substrate with a high phosphorus adsorption capacity is therefore important in obtaining significant phosphorus removal. In this study, the phosphorus removal capacities of sponge iron and zeolite were evaluated and related to their physico-chemical characteristics. The potential mechanisms affecting the adsorptive removal of phosphorus from aqueous solutions onto sponge iron and zeolite were investigated in batch experiments. The pseudo-second-order kinetics were useful since the adsorption rate data fitted well. The Freundlich and Langmuir models well described the adsorption isotherm data. The results of static experiments and dynamic experiments (column experiments) indicated that the adsorption of phosphorus onto sponge iron was more apt to chemical combination, but zeolite was more apt to electrostatic attraction or ion-exchange. For sponge iron, some iron (iii) (Fe3+) or iron (ii) (Fe2+) and phosphate ions (P) form Fe–P, the solid phases compound was fixed. For zeolite, aluminum oxide and silicon oxide formed complexes in aqueous solution. It was observed that positive or negative charge surface sites favored the adsorption of phosphate due to the electrostatic attraction or ion-exchange. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction Constructed wetlands have been described as ‘‘environmentally sensitive and cost-effective treatment systems for wastewater renovation’’ [1]. They can be thought of as ‘‘ecological engineers’’ [2] because the concept involves supplying the necessary components and capitalizing on the naturally occurring wetland processes to reduce the targeted pollutants. Constructed wetlands are widely used to treat a wide variety of wastewater throughout the world because of their low cost, low energy consumption, and good effect they promise in wastewater treatment [3]. Using constructed wetlands for the treatment of polluted water is increasing in popularity as an ecological engineering alternative to conventional and chemical based methods. The substrate, vegetation, and microbial communities all make up a wetland ecosystem. Substrates capture high levels of total phosphorus from the effluent. The concentration of total phosphorus is thus reduced after adsorption process. The choice of substrates is vital in wastewater treatment by means of constructed wetlands [4]. In phosphorus removal, the appropriate choice of filling substrate is a key factor [5]. It is important to select substrates with a high phosphorus immobility capacity and economical ⇑ Corresponding author. Fax: +86 21 54740825. E-mail addresses: [email protected], [email protected] (Y. He). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.03.057
friendliness in order to obtain a sustainable phosphorus removal in constructed wetlands. Recently, the removal of phosphorus from aqueous solutions and wastewater via substrates has been widely studied [6–13]. It is generally found that several of these substrates have the potential to enhance the phosphorus removal in constructed wetland systems. Our studies aimed at selecting substrates with a high phosphorus sorption capacity. Based on the static and dynamic experiments, this paper analyzes the adsorption capacity of two kinds of substrates: sponge iron and zeolite. Zeolite is an excellent and traditional absorbent, but its application in phosphorus removal in wastewater treatment is barely reported. Sponge iron is also a traditional environmental material, but its application as an adsorbent in phosphorus removal is a new research field. During the study, the adsorption capacity and practical applicability of phosphorus adsorption process were explored based on the experimental data. The Langmuir and Freundlich isotherm models were used to fit the experimental data and tested for their applicability. The pseudo-second-order kinetic models were employed to interpret the kinetics results under static conditions. The practical applicability of the sponge iron and zeolite in wastewater treatment was analyzed experimentally under dynamic conditions. This paper explains the science behind sponge iron and zeolite, and their immediate and long-term advantages over conventional substrates.
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
2. Materials and methods 2.1. Materials Sponge iron and zeolite were obtained from Henan Gongyi Hualong Filter Factory, China. Sponge iron was composed of 95% Fe2O3 and other oxides; the specific surface area was P80 m2/g. While zeolite was composed of 5% Al2O3 and 95% SiO2, the specific surface area was 500–800 m2/g. The two adsorption substrates were washed several times with distilled water to remove any surfaceadhered particles and to be sure that there were no soluble salts which could be dissolved during the batch studies. The washed substrates were then dried to a constant weight in an oven at 105 °C for 48 h. They were later converted into fine powder by grinding with a mechanical grinder. They were sieved with a 20mesh (0.85 mm) particle size sieve. Finally, they were stored in a desiccator for further use [14]. All reagents were of analytical grade, obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). In addition, all solutions were prepared using deionized water. Stock solutions of phosphorus (1000 mg L1) were prepared using anhydrous potassium dihydrogen phosphate (KH2PO4). The pH was pre-adjusted by HCl and NaOH solutions and measured by a pH meter with a combined pH electrode. The stock solution was diluted appropriately as necessary. The concentrations of the phosphorus were determined using the ascorbic acid method with a Unico Spectrophotometer (US-2102 PCS) [15]. To analyze phosphorus removal capacity and practical applicability of sponge iron and zeolite in wastewater treatment, samples were collected from Guangdong Heyuan Sewage Treatment Plant. In the wastewater, the initial concentrations of TP, TDP, PP, SRP, and DOP were 2.0–2.5 mg/L, 1.5–2.0 mg/L, 0.2– 0.5 mg/L, 1.5–2.0 mg/L, and 0.1–0.5 mg/L respectively. 2.2. Phosphorus analysis Total phosphorus (TP) in water can be separated into particulate phosphorus (PP) and total dissolved phosphorus (TDP). TDP can further be separated into inorganic and organic phosphorus. Whereas soluble reactive phosphorus (SRP) is an inorganic type, dissolved organic phosphorus (DOP) is organic in nature [16]. These different forms of phosphorus were determined and analyzed. TP was determined after 30 min of autoclave-mediated digestion (120 °C, 100 kPa, with K2S2O8 and H2SO4) of an unfiltered sample. The molybdenum-blue method was employed in photometric TP analysis. A separate analysis of TDP (a sample of 0.45lm filtrate digested and determined as in the TP analysis) that would also determine PP by subtracting the TDP from TP was employed [17]. SRP was analyzed using the molybdenum-blue method [18]. The sample was mixed with a reagent containing ammonium molybdate, potassium tartrate, and sulfuric acid. Ascorbic acid was then added, and a blue complex formed was measured using a flow through detector. For the determination of TDP, organic phosphorus was first oxidized to SRP with the addition of an alkaline potassium persulfate reagent. This was coupled with heating and UV oxidation. Polyphosphates were oxidized to SRP by sulfuric acid digestion. The sample was then analyzed in a similar manner to SRP. DOP was calculated by subtracting SRP from TDP [19]. All the above procedures were repeated, and the average values were used for analysis. 2.3. Static experiments The phosphorus static experiments, including the kinetic studies, adsorption isotherms and desorption experiment, were carried out by batch experiments in 100 mL conical flasks. The flasks were
247
capped and shaken horizontally on a shaker equipped with thermostat at 120 rpm at 25 ± 1 °C. This was followed by the filtration of the aqueous supernatants through a 0.45 lm membrane filter. The concentrations of phosphorus in the filtrate samples were spectrophotometrically determined. All the above procedures were repeated for several batches, and the average values used for analysis. 2.3.1. Adsorption isotherms Isotherm data were analyzed using Langmuir and Freundlich adsorption equations. The Langmuir and Freundlich parameters were determined and correlation coefficients calculated. The linear form of the Langmuir isotherm is represented below [20]:
Ce 1 Ce ¼ þ Q e Q m k1 Q m
ð1Þ
where Qe is the equilibrium phosphorus concentration on adsorbent (mg/kg), Ce is the equilibrium phosphate concentration in solution (mg/L), Qm is the monolayer phosphorus adsorption capacity of the adsorbent (mg/kg), and k1 is binding constant (L/mg) [21]. The linear form of Freundlich isotherm model can be written as [20]:
lg Q e ¼ lg K F þ
1 lg C e n
ð2Þ
where KF and 1/n are the Freundlich constants which are related to the adsorption capacity ((mg/kg)/(mg/L)1/n) and adsorption intensity respectively. Qe is still the equilibrium phosphorus concentration on adsorbent (mg/kg). Adsorption isotherm study was carried out by shaking 5 g of substrates (20 mesh) in 50 mL KH2PO4 solutions with different initial concentrations (3, 6, 10, 12, 15, 18, 24, and 40 mg/L) for 48 h. 2.3.2. Adsorption kinetics The pseudo-second-order kinetic expression for adsorption systems has been applied in a number of studies including some involving adsorption of anions [22]. Kinetics of the phosphorus ions adsorption was described using a pseudo-second-order mechanism. The linear form of the model is given as follows [23]:
t 1 t ¼ þ Q t k2 Q 2e Q e
ð3Þ
where k2 is the rate constant of pseudo-second-order kinetic model (kg/mg h). The values of equilibrium adsorption capacity (Qe) and rate constant (k2) are calculated from the intercept and the slope of the linear plot of t/Qt versus t, along with the value of determination coefficient, R2. Adsorption experiments for the kinetic study was carried out by shaking 5 g of substrates (20 mesh) in 50 mL KH2PO4 solutions with an initial concentration of 12 mg/L. The set experimental contact times were 2, 4, 8, 12, 18, 24, 30, and 48 h. 2.3.3. Phosphorus desorption Desorption experiments conducted to explore the reversibility of the sorption reactions were carried out with pre-sorbed P saturated substrates. Pre-sorbed P samples were prepared by the phosphorus adsorption experiments. To maximize the amount of phosphorus that the sponge iron and zeolite would remove from aqueous solution under our study conditions, the phosphorus desorption kinetics in the presence of the substrates was determined. Based on phosphorus desorption experiment, the properties of adsorption such as physical, chemical or both, would be investigated. The sponge iron and zeolite desorption process was also studied. The rate of phosphorus desorption from the substrates was
248
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
measured using the Batch technique. Three 1 g phosphorus saturated substrates were weighed into conical flasks. This was followed by the addition of 50 mL of 0.02 mol/L KCl solutions into each conical flask. Three sets of 50 mL–0.02 mol/L KCl solutions without substrates were included. The conical flasks were shaken at time intervals of 0.2, 0.5, 1, 1.5, 5, 10, and 20 h. 2.4. Dynamic experiments The column (a plexiglas column) experiments were carried out in vertical upflow columns of 595 mm length and 280 mm internal diameter. Fig. 1 shows the general layout of a filled vertical upflow column. All column experiments were run in a constant ‘‘temperature room’’ (25 ± 1 °C). The columns were filled in the following order: A wire screen was placed on the bottom of the column, followed by 200 mm of substrate. The vertical upflow system consisted of three waste tanks, metering pump, flow meter, and tubing. The bottom port of a column was the waste substrate recovery outlet. The substrate in the fixed bed column was pre-washed by deionized water for hours. In addition, the wastewater inlet flow rate was controlled by the metering pump and flow meter. The experimental wastewater with initial phosphorus concentration of 1.8–2.2 mg/L had a constant flow rate from the elevated water tank to the bottom of column. The effluent was then collected into water tank. The substrates’ size (20 mesh) was consistent with adequate hydraulic conductivity, minimizing the risk of clogging [24]. This is the determinant for long-term application in constructed wetlands systems. The hydraulic retention time (HRT) was controlled by wastewater inflow velocity via the metering pump and the flowmeter in column experiments. In the first part of the experiments, a design HRT of 1–8 h was adopted. Based on the experimental analysis, an optimal HRT of 4 h was obtained. In the second part of the experiments, the HRT was 4 h. The experiments would involve the analysis of phosphorous removal efficiency from wastewater onto the substrates. 3. Results and discussion 3.1. Adsorption isotherm The adsorption isotherms were obtained in order to understand the phosphorus adsorption mechanism of sponge iron and zeolite. The isotherms reflected the equilibria [25]. The adsorption equilibrium defines the distribution of the solute between the liquid and the solid phases after the adsorption reaction reaches equilibrium
at constant temperature. Two-parameter isotherm models (Langmuir and Freundlich) were used to fit the experimental data. The Langmuir (Eq. (1)) and the Freundlich (Eq. (2)) equations are used to describe the adsorption equilibrium for water and wastewater treatment applications [26]. The adsorption isotherms are shown in Fig. 2. The correlation coefficients (R2 > 0.99) given in Table 1 showed that the Langmuir equation gives a better fit for the adsorption isotherms of sponge iron and zeolite. Based on this fact, the Langmuir isotherm model was the best fit for this analysis. In this regard, the shape of the isotherms for the substrates/phosphorus systems indicated that there was selectivity for the substrates toward the incoming ion compared to the ions already bound in its lattice. In other words, the substrates favored the replacement of its ions by the solute. The Langmuir model assumed that the energy of adsorption was the same for all surface sites and did not depend on the degree of coverage. This was indicative of monolayer adsorption of the substrates [27]. More sites in the substrates were filled up. It thus became increasingly difficult for a bombarding solute molecule to find an available vacant site. This implies that there was no strong competition from the aqueous solution. The correlation coefficients (R2 > 0.99) in Table 1 show that the Freundlich equation yielded a fit to the adsorption isotherms of the substrates. The Freundlich model assumes that the frequency of sites associated with free energy of adsorption decreases exponentially with increasing free energy. In addition, the energy of adsorption may vary because real surfaces are heterogeneous surfaces (multilayer adsorption) with a non-uniform distribution [27,28]. The values of maximum phosphorus adsorption capacity and Langmuir constants were calculated by the slope and intercept of Langmuir equations (Table 1). The maximum phosphorus adsorption capacity for sponge iron and zeolite was found to be 1111.11 mg/kg and 303.03 mg/kg, respectively. The result indicated that the sponge iron favored more replacement of its ions by the phosphorus ion than zeolite. The 1/n constant for the Freundlich isotherm is a measure of exchange intensity or surface heterogeneity and ranges between 0 and 1. In this study, the values of 1/n were less than 1 (Table 1), which shows that the adsorption conditions were favorable. The value of n for the sponge iron was smaller than that of zeolite. High value of 1/n indicated a strong bond between the substrate and the phosphorus. The data showed that the KF constant for the sponge iron was higher than that of zeolite. The values of Freundlich constant KF in the result indicated the extent of phosphorus removal, hence a measure of adsorption capacity as Langmuir Qm. Whereas sponge iron had high values of Qm (1111.11 mg/kg) and KF (155.02), the corresponding values for zeolite were relatively lower (303.03 mg/kg and 87.49, respectively). Therefore, Langmuir Qm and Freundlich KF indicated and compared the adsorption performance, but from the Langmuir Qm, the specific surface area could be estimated. These implied a more favorable adsorption of phosphorus ion by the sponge iron.
3.2. Adsorption kinetic modeling
Fig. 1. Schematic diagram of experimental set-up (1. Water tank; 2. Metering pump; 3. Elevated water tank; 4. Flowmeter; 5. Waste recovery outlet; 6. Column filled substrate; 7. Water tank).
Available studies have shown that the pseudo-second-order kinetic equation (Eq. (3)) is a reasonably good fit for data over the entire fractional approach to equilibrium. Therefore, it has been employed extensively in the study of adsorption kinetics [29–32]. Table 2 shows that the model was the best fit to the experimental data with correlation coefficients (R2 > 0.999). The adsorption rate denoted by dQt/dt shows how much phosphorus was adsorbed from the aqueous solution onto the substrates within a unit time. Fig. 3 reveals that the adsorption rate (dQt/dt) decreases with time
249
400
400
300
300
Qe (mg P/Kg)
Qe (mg P/Kg)
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
200
100
0
Experimental Langmuir Freundlich 0
0.5
1
1.5
2
2.5
200
100
0
Experimental Langmuir Freundlich 0
1
2
3
4
Ce (mg/L)
Ce (mg/L)
(a)
(b)
5
6
Fig. 2. Substrates’ phosphorus adsorption isotherm at 25 ± 1 °C, (a) sponge iron phosphorus adsorption isotherm and (b) zeolite phosphorus adsorption isotherm.
Table 1 Langmuir and Freundlich constants for adsorption of phosphorus by sponge iron and zeolite.
Sponge iron Zeolite
Langmuir
Freundlich
k1 (L/ mg)
Qm (mg/ kg)
R2
KF (mg/kg)/ (mg/L)1/n
n
R2
0.164
1111.11
0.9984
155.02
1.08
0.9941
0.458
303.03
0.9913
87.49
1.31
0.9922
Removal efficiency (%)
Substrates
100
80 60
40 Sponge iron
20
Zeolite
until it gradually approaches the equilibrium state. This is due to the continuous decrease in the driving force (QeQt). The adsorption of phosphorus onto sponge iron and zeolite as a function of contact time is shown in Fig. 3. The pseudo-second-order predicted adsorption capacity at equilibrium, Qe. This was consistent with the experimental ones (Table 2). At equilibrium, the adsorption process is supposed to be in a dynamic state. In this regard, the rates of forward process and backward process are equal [33]. The time at which the adsorption equilibrium is attained is known as equilibrium time. According to the results, the adsorption process was rapid at the beginning. The equilibrium for both substrates was reached after 8 h. Many preliminary investigations also indicate that most adsorption of solutes occur within a short contact time. The rapid rate of adsorption thereafter slows down and becomes significantly lower at equilibrium. For longer periods, the adsorption capacity tends to remain constant. The results also suggest that the pseudo-second-order model was the most suitable model for the studied experimental conditions. The application of the model relied on the assumption that adsorption may be the rate-limiting step for chemical adsorption involving valency forces through sharing or exchange of electrons between the substrates and phosphorus. It was selected as a measure of the substrates’ selectivity and was consistent with the experimental data. The respective values for all the systems used are given in Table 2. The k2 values of the substrates implied a higher selectivity or adsorption for sponge
Table 2 Kinetics parameters for adsorption of phosphorus by sponge iron and zeolite. Substrates
0
0
k2 (kg/mg h) Qe (mg/kg) R2 0.00795 0.00445
116.279 109.89
24
32
40
48
Fig. 3. The TP removal efficiency of sponge iron and zeolite under different hydraulic retention time.
iron. It was observed that the k2 values of the sponge iron was a higher than the zeolite. 3.3. Phosphorus desorption study After saturation of the substrates, the investigation of phosphorus desorption for phosphorus retained in the substrates is also very crucial. The phosphorus desorption data of sponge iron and zeolite are presented in Table 3. Whereas a phosphorus release value of 5.7122 mg/kg was noted for sponge iron, a corresponding value of 12.0762 mg/kg was noted for zeolite at the equilibrium time. The desorption results suggested that phosphorus was tightly bound to the adsorbents. In desorption experiments, the substrates rapidly increased in volume once placed in water. In the process, the adsorbed phosphorus in the substrates slowly diffused into the water solution. With an increase in time, the equilibrium was reached gradually. Table 3 shows that the phosphorus desorption rate of sponge iron was lower than that of zeolite. It is evident from the low desorption values that the adsorption of phosphorus onto the
Table 3 The equilibrium desorption content and desorption rate of saturated substrates. Substrates
Saturated adsorption capacity (mg/kg)
Equilibrium desorption content (mg/kg)
Desorption rate (%)
Sponge iron Zeolite
1111.11 303.03
5.7122 12.0762
0.5141 3.9851
(QeQe,exp)/Qe
0.9998 0.028053 0.9991 0.03765
16
Time (h)
Qe,exp (mg/kg) Pseudo-second-order model
Sponge iron 113.017 Zeolite 105.753
8
250
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
substrates was chemical in nature, coupled with chemical bonding and electrostatic attraction. Chemisorption exhibits poor desorption, it may be due to fact that the adsorbate species are firmly held to the adsorbent by comparatively stronger bonds. For the sponge iron, the reaction products (Fe(OH)2 and Fe(OH)3) in water formed a high number of flocs. These could easily fix the phosphorus and unease the released adsorbed phosphorus. The formation of Fe–P compounds such as Fe1.6H2PO4(OH)3.8 compound. The desorption of Fe–P was difficult. However, the desorption value of zeolite was relatively higher. Aluminum oxide and silicon oxide form complexes in solution. The solid–solution interface of complexes leads to development of positive or negative charge on the surface. The charged surface sites on the adsorbent favored the adsorption of phosphate due to the electrostatic attraction or ion-exchange. For the substrates, sponge iron is more apt to chemical combination than zeolite. However, zeolite is more apt to electrostatic attraction or ion-exchange than sponge iron. 3.4. Adsorptive removal of phosphorus from wastewater Static experiments results showed that the adsorption of phosphorus onto sponge iron was more apt to chemical combination while zeolite was more apt to electrostatic attraction or ion-exchange. Dynamic analysis involved the study of adsorption mechanism of sponge iron and zeolite upon different phosphorus species in wastewater. The column batch experiments of phosphorus removal would provide more realistic laboratory results. This is because they had a greater resemblance to the flux conditions in full scale constructed wetlands than in short-term stirred batch experiments. This could result in overestimation of adsorption capacities [4,33]. The maximum adsorption capacity depends on the concentration of the tested solution [4]. In column experiments, many wastewater factors such as COD, water turbidity and hydraulic retention time, affect substrates’ adsorption of phosphorus from wastewater. In this study, the wastewater’s initial CODcr, BOD5, and water turbidity were 48.36 mg/L, 10.50 mg/L and 36.10 NTU, respectively. Static experiments showed much better results than dynamic experiments. For example, considering the adsorption capacity of TP, the equilibrium adsorption capacity using sponge iron decreased from 113.02 mg/kg under static experiments to 87.89 mg/kg under dynamic experiments. Similarly, using zeolite, the capacity decreased from 105.75 mg/kg under static experiments to 55.57 mg/kg under dynamic experiments. This is because other substances in wastewater affected phosphorus adsorption. The variation of hydraulic retention time in phosphorus removal by the two substrates is shown in Fig. 4. The adsorptive removal of TP, TDP, PP, SRP, and DOP was analyzed. Fig. 4 shows that
the adsorptive removal of various forms phosphorus stabilized after 4 h. In addition, with the exception of PP, the removal efficiency of phosphorus from the wastewater by sponge iron was higher than that of zeolite. This is shown in Fig. 4. The adsorptive removal of phosphorus stabilized after an adsorption time of 4 h. The phosphorus removal efficiency of each substrate increased with increasing adsorption time. When the adsorption time was 4 h, the TP removal efficiencies for sponge iron and zeolite were 94.93% and 73.62%, respectively. When the adsorption time was 8 h, the corresponding TP removal efficiencies increased to 96.56% and 80.52%. Similarly, when the adsorption time increased from 4 h to 8 h, TDP removal efficiencies for sponge iron and zeolite increased from 99.10% to 99.92% and 69.09% to 75.30%, respectively. In addition, the corresponding removal efficiencies of PP increased from 79.26% to 83.90% and 84.89% to 93.51% respectively. Similarly, whereas the SRP removal efficiencies for sponge iron and zeolite increased from 99.97% to 99.99% and 67.58% to 72.65% respectively, the corresponding DOP removal efficiencies increased from 78.03% to 98.04% and 75.60% to 88.35% for the substrates. After 8 h of adsorption, the removal efficiency of TP, TDP, SRP, and DOP using sponge iron was approximately 16.61%, 24.64%, 27.34%, and 9.88% higher that of zeolite, respectively. However, the removal efficiency of PP using sponge iron was 10.28% lower that of zeolite. In batch column experiments, when hydraulic retention time was 4 h, as shown in Fig. 5b, the TP removal efficiencies by sponge iron and zeolite were 97.79% and 68.67%, respectively. TDP, PP, SRP, and DOP were also analyzed. Whereas the removal efficiencies of these phosphorus species by sponge iron were 99.13%, 75.04%, 99.55% and 96.77%, the corresponding efficiencies for zeolite were 65.87%, 80.06%, 65.22%, and 71.32%. TP accounts for both particulate-bound phosphorus (PP adsorbed to sediment particles) and all forms of TDP. Whereas the adsorptive removal of PP was mainly due to electrostatic attraction or ion-exchange, the adsorptive re moval of TDP (HPO2 4 and H2 PO4 ) was dependent on the chemical compound. As shown in Fig. 5a, after a 4 h adsorption, the adsorption capacities of TP, SRP, and DOP by sponge iron were 87.89, 62.15, and 16.78 mg/kg, respectively. The corresponding capacities for zeolite were 55.57, 38.86, and 5.07 mg/kg, respectively. The adsorption capacities of TDP using sponge iron and zeolite were 78.96 mg/kg and 43.90 mg/kg, respectively. And the adsorption capacity of PP using sponge iron and zeolite was 8.94 mg/kg and 11.68 mg/kg, respectively. Whereas the removal capacity of TDP using sponge iron was 79.86% higher that of zeolite, the removal capacity of PP using sponge iron was 23.46% lower that of zeolite. Fig. 5b describes the adsorption removal efficiency of all forms of phosphorus.
100
80
60
TP TDP PP SRP DOP
40 20
1
2
3
4
5
6
7
8
Removal efficiency (%)
Removal efficiency (%)
100
80 60
TP TDP PP SRP DOP
40 20
1
2
3
4
5
Time (h)
Time (h)
(a)
(b)
6
7
8
Fig. 4. Adsorptive removal efficiency of phosphorus from wastewater onto the substrates under different HRT, (a) adsorptive removal efficiency of phosphorus onto sponge iron and (b) adsorptive removal efficiency of phosphorus onto zeolite.
251
100
Sponge iron Zeolite
120
Removal efficiency (%)
Adsorption capacity (mg/Kg)
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
80 60 40 20
100 80 60 40 20 0
0 TP
TDP
PP
(a)
SRP
DOP
Sponge iron Zeolite
TP
TDP
PP
SRP
DOP
(b)
Fig. 5. Adsorptive removal capacity of phosphorus from wastewater onto sponge iron and zeolite, (a) adsorptive removal capacity of phosphorus onto the substrates and (b) adsorptive removal efficiency of phosphorus onto the substrates.
This phenomenon could further confirm the static experiments’ results that the adsorption of phosphorus onto sponge iron was more apt to chemical combination while zeolite was more apt to electrostatic attraction or ion-exchange. Considering the structure of the substrates, sponge iron possessed some favorable characteristics, such as higher surface area and higher surface energy. To obtain higher contaminants removal efficiency, adequate surface area and intension should be taken into account during the production process of sponge iron [34]. When electrochemistry reaction took place, some Fe2O3 was transformed to Fe3+ or Fe2+. TDP can be further separated into inorganic (soluble reactive phosphorus or SRP) and organic (dissolved organic phosphorus or DOP) components [21]; HPO2 and H2 PO 4 4 are commonly used to quantify TDP. In wastewater, a large number of products including complexes, polymers, and precipitates are formed [35]. The equations below describe the equilibrium distributions of the soluble orthophosphate species as well as soluble iron species [35].
Fe3þ þ H2 O ¼ FeOH2þ þ Hþ
ð4Þ
Fe3þ þ 2H2 O ¼ FeðOHÞþ2 þ 2Hþ
ð5Þ
phase; the content of Fe–P was chemically adsorbed. It is well known that phosphorus ion has a relatively strong affinity for the substrate surfaces [36]. The chief constituents of zeolite are mainly metal oxides like Al2O3 and SiO2. These metal oxides form metal-hydroxide complexes in solution, and the subsequent acidic or basic dissociation of these complexes at the solid-solution interface leads to development of a positive or negative charge on the surface. Positive charge surface sites on the adsorbent favored the adsorption of phosphate due to the electrostatic attraction. The particulatebound phosphorus (PP, adsorbed to sediment particles) was strongly adsorbed via electrostatic attraction. The substrates/phosphorus systems had highly polar solute and substrate as well as a non-polar solvent. In the systems, the monofunctional ions substances with very strong intermolecular attraction were adsorbed from water through ion-ion attraction. It is possible that the adsorbed ions may have become associated into very large clusters just before adsorption took place [37]. This relatively high adsorption capacity shows the strong electrostatic force of attraction between the phosphorus molecules and the substrates bending sites [38]. In the wastewater, soluble reactive phosphorus or SRP was also adsorbed by ion–ion attraction or ion-exchange using zeolite.
þ 2Fe3þ þ 2H2 O ¼ FeðOHÞ4þ 2 þ 2H
ð6Þ
4. Conclusions
Fe3þ þ 3H2 O ¼ FeðOHÞ3ðaqÞ þ 3Hþ
ð7Þ
Fe3þ þ 4H2 O ¼ FeðOHÞ4 þ 4Hþ
ð8Þ
þ 3Fe3þ þ 4H2 O ¼ Fe3 ðOHÞ5þ 4 þ 4H
ð9Þ
The present study focused on the mechanism and adsorption capability of wetland substrates (sponge iron and zeolite) during phosphorus removal processes. Static experiments and dynamic experiments results showed that whereas the adsorption of phosphorus onto sponge iron was more inclined to chemical combination, that of zeolite was more inclined to electrostatic attraction or ion-exchange. In treating wastewater using sponge iron, electrochemistry reactions occurred. Consequently, a large number of products were formed, contributing to significant flocculation. The products so formed could easily fix phosphorus because of the strong affinity for the phosphorus ion by the substrate surfaces. In treating wastewater using zeolite led to the development of a positive or negative charge at the solid-solution interface. The strong electrostatic force of attraction between the phosphorus and the zeolite was due to ion–ion attraction or ion-exchange. Unlike conventional substrates, sponge iron and zeolite captured phosphorus through adsorption, allowing high levels of phosphorus removal to be achieved. In addition, unlike other substrates, their desorption rates are very low. The different adsorption mechanism and significantly higher adsorption capacities for sponge iron and zeolite could be applied to separate different forms phosphorus from wastewater. In addition, the choice on which substrate to use for different kinds of wastewater can be
H3 PO4 ¼ Hþ þ H2 PO4
ð10Þ
H2 PO4 ¼ Hþ þ HPO2 4
ð11Þ
3 þ HPO2 4 ¼ H þ PO4
ð12Þ
þ Fe3þ þ HPO2 4 ¼ FeHPO4
ð13Þ
Fe3þ þ H2 PO4 ¼ FeH2 PO2þ 4
ð14Þ
1:6Fe3þ þ H2 PO4 þ 3:8OH ¼ Fe1:6 H2 PO4 ðOHÞ3:8ðsÞ
ð15Þ
OH could be produced when dissolved oxygen acted as an electron receptor, and the reaction products (Fe(OH)2 and Fe(OH)3) in wastewater contributed to significant flocculation, which could easily fix phosphorus. Fe1.6H2PO4(OH)3.8(s) compound was solid
252
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
made. Further practical application of sponge iron and zeolite in the constructed wetland treatment systems should be enhanced for future sustainability in industrial/municipal wastewater treatment. Acknowledgments This work was supported by Important National Science and Technology Specific Projects (2013ZX07206001), the National Research Foundation Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE), Science and Technology Commission of Shanghai Municipality Projects (09DZ1200109), and Technological Research and Development Programs of the Ministry of Railways (J2011Z003). The authors thank fellow colleagues for their help and valuable advice. The authors also express their sincere gratitude to all the editors and reviewers of this manuscript. References [1] K.R. Hench, G.K. Bissonnette, A.J. Sexstone, J.G. Coleman, K. Garbutt, J.G. Skousen, Water Res. 37 (2003) 921–927. [2] C.C. Tanner, Ecol. Eng. 7 (1996) 59–83. [3] D.A. Hammer, Ecol. Eng. 1 (1992) 49–82. [4] A. Drizo, Y. Comeau, C. Forget, R.P. Chapuis, Environ. Sci. Technol. 36 (2002) 4642–4648. [5] D.M.R. Mateus, M.M.N. Vaz, H.J.O. Pinho, Ecol. Eng. 41 (2012) 65–69. [6] D. Wu, B. Zhang, C. Li, Z. Zhang, H. Kong, J. Colloid Interface Sci. 304 (2006) 300–306. [7] H. Ye, F. Chen, Y. Sheng, G. Sheng, J. Fu, Sep. Purif. Technol. 50 (2006) 283–290.
[8] J. Chen, H. Kong, D. Wu, X. Chen, D. Zhang, Z. Sun, J. Hazard. Mater. 139 (2) (2007) 293–300. [9] E. Oguz, J. Hazard. Mater. 114 (2004) 131–137. [10] A.K. Golder, A.N. Samanta, S. Ray, Sep. Purif. Technol. 52 (2006) 102–109. [11] L. Zeng, X. Li, J. Liu, Water Res. 38 (2004) 1318–1326. [12] A. Genz, A. Kornmüller, M. Jekel, Water Res. 38 (2004) 3523–3530. [13] M. Özacar, Chemosphere 51 (2003) 321–327. [14] M. Dog˘an, M. Alkan, Ü. Çakir, J. Colloid Interface Sci. 192 (1997) 114–118. [15] W.G. Walter, 22nd ed., A PHA, AWWA, WEF, Portland, 2012. [16] M.E. Ellison, M.T. Brett, Water Res. 40 (2006) 1258–1268. [17] O. Broberg, G. Persson, Hydrobiologia 170 (1988) 61–90. [18] J. Murphy, J.P. Riley, Anal. Chim. Acta 27 (1962) 31–36. [19] K.R. Rinker, R.T. Powell, Mar. Chem. 102 (2006) 170–179. [20] M.A. Wahab, S. Jellali, N. Jedidi, Bioresour. Technol. 101 (2010) 5070–5075. [21] L. Svarovski, Edited by, 3rd ed., Woburn, 1990. [22] N.I. Chubar, V.A. Kanibolotskyy, V.V. Strelko, G.G. Gallios, V.F. Samanidou, T.O. Shaposhnikova, V.G. Milgrandt, I.Z. Zhuravlev, Eng. Aspects 255 (2005) 55–63. [23] J. Wang, C. Zheng, S. Ding, H. Ma, Y. Ji, Desalination 273 (2011) 285–291. [24] J. Vymazal, Ecol. Eng. 18 (2002) 633–646. [25] Z. Talip, M. Eral, Ü. Hiçsönmez, J. Environ. Radioact. 100 (2009) 139–143. [26] N. Ünlü, M. Ersoz, J. Hazard. Mater. 136 (2006) 272–280. [27] M. Metaxas, V. Kasselouri-Rigopoulou, P. Galiatsatou, C. Konstantopoulou, D. Oikonomou, J. Hazard. Mater. B97 (2003) 71–82. [28] Z. Ma, Q. Li, Q. Yue, B. Gao, W. Li, X. Xu, Q. Zhong, Chem. Eng. J. 171 (2011) 1209–1217. [29] F.C. Wu, R.L. Tseng, R.S. Juang, Water Res. 35 (2001) 613–618. [30] C.Y. Chang, W.T. Tsai, C.H. Ing, C.F. Chang, J. Colloid Interface Sci. 260 (2003) 273–279. [31] A.S. Özcan, B. Erdem, A. Özcan, J. Colloid Interface Sci. 280 (2004) 44–54. [32] M.Y. Chang, R.S. Juang, J. Colloid Interface Sci. 278 (2004) 18–25. [33] D.M.R. Mateus, H.J.O. Pinho, Water Environ. Res. 82 (2010) 128–137. [34] J. Li, J. Li, Y. Li, J. Environ. Suppl. (2009) S60–S64. [35] K. Fytianos, E. Voudrias, N. Raikos, Environ. Pollut. 101 (1998) 123–130. [36] W. Stumm, Wiley, New York, 1992. [37] C.H. Giles, T.H. MacEwan, S.N. Nakhwa, D. Smith, J. Chem. Soc. (1960) 3973– 3993. [38] P. Kaewsarn, Q. Yu, Environ. Pollut. 112 (2001) 209–213.
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Adsorptive removal of phosphorus from aqueous solution using sponge iron and zeolite Cheng Jiang a, Liyue Jia b, Yiliang He a,⇑, Bo Zhang a, George Kirumba a, Jie Xie a a b
School of Environmental Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China School of Urban Construction, Hebei University of Engineering, Handan 056038, China
a r t i c l e
i n f o
Article history: Received 26 February 2013 Accepted 25 March 2013 Available online 11 April 2013 Keywords: Sponge iron Zeolite Phosphorus Adsorptive removal
a b s t r a c t Phosphorus adsorptive removal is an important and efficient treatment process in constructed subsurface flow wetlands. Many materials have been proposed for removal of excess phosphorus from wastewater. Selecting a substrate with a high phosphorus adsorption capacity is therefore important in obtaining significant phosphorus removal. In this study, the phosphorus removal capacities of sponge iron and zeolite were evaluated and related to their physico-chemical characteristics. The potential mechanisms affecting the adsorptive removal of phosphorus from aqueous solutions onto sponge iron and zeolite were investigated in batch experiments. The pseudo-second-order kinetics were useful since the adsorption rate data fitted well. The Freundlich and Langmuir models well described the adsorption isotherm data. The results of static experiments and dynamic experiments (column experiments) indicated that the adsorption of phosphorus onto sponge iron was more apt to chemical combination, but zeolite was more apt to electrostatic attraction or ion-exchange. For sponge iron, some iron (iii) (Fe3+) or iron (ii) (Fe2+) and phosphate ions (P) form Fe–P, the solid phases compound was fixed. For zeolite, aluminum oxide and silicon oxide formed complexes in aqueous solution. It was observed that positive or negative charge surface sites favored the adsorption of phosphate due to the electrostatic attraction or ion-exchange. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction Constructed wetlands have been described as ‘‘environmentally sensitive and cost-effective treatment systems for wastewater renovation’’ [1]. They can be thought of as ‘‘ecological engineers’’ [2] because the concept involves supplying the necessary components and capitalizing on the naturally occurring wetland processes to reduce the targeted pollutants. Constructed wetlands are widely used to treat a wide variety of wastewater throughout the world because of their low cost, low energy consumption, and good effect they promise in wastewater treatment [3]. Using constructed wetlands for the treatment of polluted water is increasing in popularity as an ecological engineering alternative to conventional and chemical based methods. The substrate, vegetation, and microbial communities all make up a wetland ecosystem. Substrates capture high levels of total phosphorus from the effluent. The concentration of total phosphorus is thus reduced after adsorption process. The choice of substrates is vital in wastewater treatment by means of constructed wetlands [4]. In phosphorus removal, the appropriate choice of filling substrate is a key factor [5]. It is important to select substrates with a high phosphorus immobility capacity and economical ⇑ Corresponding author. Fax: +86 21 54740825. E-mail addresses: [email protected], [email protected] (Y. He). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.03.057
friendliness in order to obtain a sustainable phosphorus removal in constructed wetlands. Recently, the removal of phosphorus from aqueous solutions and wastewater via substrates has been widely studied [6–13]. It is generally found that several of these substrates have the potential to enhance the phosphorus removal in constructed wetland systems. Our studies aimed at selecting substrates with a high phosphorus sorption capacity. Based on the static and dynamic experiments, this paper analyzes the adsorption capacity of two kinds of substrates: sponge iron and zeolite. Zeolite is an excellent and traditional absorbent, but its application in phosphorus removal in wastewater treatment is barely reported. Sponge iron is also a traditional environmental material, but its application as an adsorbent in phosphorus removal is a new research field. During the study, the adsorption capacity and practical applicability of phosphorus adsorption process were explored based on the experimental data. The Langmuir and Freundlich isotherm models were used to fit the experimental data and tested for their applicability. The pseudo-second-order kinetic models were employed to interpret the kinetics results under static conditions. The practical applicability of the sponge iron and zeolite in wastewater treatment was analyzed experimentally under dynamic conditions. This paper explains the science behind sponge iron and zeolite, and their immediate and long-term advantages over conventional substrates.
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
2. Materials and methods 2.1. Materials Sponge iron and zeolite were obtained from Henan Gongyi Hualong Filter Factory, China. Sponge iron was composed of 95% Fe2O3 and other oxides; the specific surface area was P80 m2/g. While zeolite was composed of 5% Al2O3 and 95% SiO2, the specific surface area was 500–800 m2/g. The two adsorption substrates were washed several times with distilled water to remove any surfaceadhered particles and to be sure that there were no soluble salts which could be dissolved during the batch studies. The washed substrates were then dried to a constant weight in an oven at 105 °C for 48 h. They were later converted into fine powder by grinding with a mechanical grinder. They were sieved with a 20mesh (0.85 mm) particle size sieve. Finally, they were stored in a desiccator for further use [14]. All reagents were of analytical grade, obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). In addition, all solutions were prepared using deionized water. Stock solutions of phosphorus (1000 mg L1) were prepared using anhydrous potassium dihydrogen phosphate (KH2PO4). The pH was pre-adjusted by HCl and NaOH solutions and measured by a pH meter with a combined pH electrode. The stock solution was diluted appropriately as necessary. The concentrations of the phosphorus were determined using the ascorbic acid method with a Unico Spectrophotometer (US-2102 PCS) [15]. To analyze phosphorus removal capacity and practical applicability of sponge iron and zeolite in wastewater treatment, samples were collected from Guangdong Heyuan Sewage Treatment Plant. In the wastewater, the initial concentrations of TP, TDP, PP, SRP, and DOP were 2.0–2.5 mg/L, 1.5–2.0 mg/L, 0.2– 0.5 mg/L, 1.5–2.0 mg/L, and 0.1–0.5 mg/L respectively. 2.2. Phosphorus analysis Total phosphorus (TP) in water can be separated into particulate phosphorus (PP) and total dissolved phosphorus (TDP). TDP can further be separated into inorganic and organic phosphorus. Whereas soluble reactive phosphorus (SRP) is an inorganic type, dissolved organic phosphorus (DOP) is organic in nature [16]. These different forms of phosphorus were determined and analyzed. TP was determined after 30 min of autoclave-mediated digestion (120 °C, 100 kPa, with K2S2O8 and H2SO4) of an unfiltered sample. The molybdenum-blue method was employed in photometric TP analysis. A separate analysis of TDP (a sample of 0.45lm filtrate digested and determined as in the TP analysis) that would also determine PP by subtracting the TDP from TP was employed [17]. SRP was analyzed using the molybdenum-blue method [18]. The sample was mixed with a reagent containing ammonium molybdate, potassium tartrate, and sulfuric acid. Ascorbic acid was then added, and a blue complex formed was measured using a flow through detector. For the determination of TDP, organic phosphorus was first oxidized to SRP with the addition of an alkaline potassium persulfate reagent. This was coupled with heating and UV oxidation. Polyphosphates were oxidized to SRP by sulfuric acid digestion. The sample was then analyzed in a similar manner to SRP. DOP was calculated by subtracting SRP from TDP [19]. All the above procedures were repeated, and the average values were used for analysis. 2.3. Static experiments The phosphorus static experiments, including the kinetic studies, adsorption isotherms and desorption experiment, were carried out by batch experiments in 100 mL conical flasks. The flasks were
247
capped and shaken horizontally on a shaker equipped with thermostat at 120 rpm at 25 ± 1 °C. This was followed by the filtration of the aqueous supernatants through a 0.45 lm membrane filter. The concentrations of phosphorus in the filtrate samples were spectrophotometrically determined. All the above procedures were repeated for several batches, and the average values used for analysis. 2.3.1. Adsorption isotherms Isotherm data were analyzed using Langmuir and Freundlich adsorption equations. The Langmuir and Freundlich parameters were determined and correlation coefficients calculated. The linear form of the Langmuir isotherm is represented below [20]:
Ce 1 Ce ¼ þ Q e Q m k1 Q m
ð1Þ
where Qe is the equilibrium phosphorus concentration on adsorbent (mg/kg), Ce is the equilibrium phosphate concentration in solution (mg/L), Qm is the monolayer phosphorus adsorption capacity of the adsorbent (mg/kg), and k1 is binding constant (L/mg) [21]. The linear form of Freundlich isotherm model can be written as [20]:
lg Q e ¼ lg K F þ
1 lg C e n
ð2Þ
where KF and 1/n are the Freundlich constants which are related to the adsorption capacity ((mg/kg)/(mg/L)1/n) and adsorption intensity respectively. Qe is still the equilibrium phosphorus concentration on adsorbent (mg/kg). Adsorption isotherm study was carried out by shaking 5 g of substrates (20 mesh) in 50 mL KH2PO4 solutions with different initial concentrations (3, 6, 10, 12, 15, 18, 24, and 40 mg/L) for 48 h. 2.3.2. Adsorption kinetics The pseudo-second-order kinetic expression for adsorption systems has been applied in a number of studies including some involving adsorption of anions [22]. Kinetics of the phosphorus ions adsorption was described using a pseudo-second-order mechanism. The linear form of the model is given as follows [23]:
t 1 t ¼ þ Q t k2 Q 2e Q e
ð3Þ
where k2 is the rate constant of pseudo-second-order kinetic model (kg/mg h). The values of equilibrium adsorption capacity (Qe) and rate constant (k2) are calculated from the intercept and the slope of the linear plot of t/Qt versus t, along with the value of determination coefficient, R2. Adsorption experiments for the kinetic study was carried out by shaking 5 g of substrates (20 mesh) in 50 mL KH2PO4 solutions with an initial concentration of 12 mg/L. The set experimental contact times were 2, 4, 8, 12, 18, 24, 30, and 48 h. 2.3.3. Phosphorus desorption Desorption experiments conducted to explore the reversibility of the sorption reactions were carried out with pre-sorbed P saturated substrates. Pre-sorbed P samples were prepared by the phosphorus adsorption experiments. To maximize the amount of phosphorus that the sponge iron and zeolite would remove from aqueous solution under our study conditions, the phosphorus desorption kinetics in the presence of the substrates was determined. Based on phosphorus desorption experiment, the properties of adsorption such as physical, chemical or both, would be investigated. The sponge iron and zeolite desorption process was also studied. The rate of phosphorus desorption from the substrates was
248
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
measured using the Batch technique. Three 1 g phosphorus saturated substrates were weighed into conical flasks. This was followed by the addition of 50 mL of 0.02 mol/L KCl solutions into each conical flask. Three sets of 50 mL–0.02 mol/L KCl solutions without substrates were included. The conical flasks were shaken at time intervals of 0.2, 0.5, 1, 1.5, 5, 10, and 20 h. 2.4. Dynamic experiments The column (a plexiglas column) experiments were carried out in vertical upflow columns of 595 mm length and 280 mm internal diameter. Fig. 1 shows the general layout of a filled vertical upflow column. All column experiments were run in a constant ‘‘temperature room’’ (25 ± 1 °C). The columns were filled in the following order: A wire screen was placed on the bottom of the column, followed by 200 mm of substrate. The vertical upflow system consisted of three waste tanks, metering pump, flow meter, and tubing. The bottom port of a column was the waste substrate recovery outlet. The substrate in the fixed bed column was pre-washed by deionized water for hours. In addition, the wastewater inlet flow rate was controlled by the metering pump and flow meter. The experimental wastewater with initial phosphorus concentration of 1.8–2.2 mg/L had a constant flow rate from the elevated water tank to the bottom of column. The effluent was then collected into water tank. The substrates’ size (20 mesh) was consistent with adequate hydraulic conductivity, minimizing the risk of clogging [24]. This is the determinant for long-term application in constructed wetlands systems. The hydraulic retention time (HRT) was controlled by wastewater inflow velocity via the metering pump and the flowmeter in column experiments. In the first part of the experiments, a design HRT of 1–8 h was adopted. Based on the experimental analysis, an optimal HRT of 4 h was obtained. In the second part of the experiments, the HRT was 4 h. The experiments would involve the analysis of phosphorous removal efficiency from wastewater onto the substrates. 3. Results and discussion 3.1. Adsorption isotherm The adsorption isotherms were obtained in order to understand the phosphorus adsorption mechanism of sponge iron and zeolite. The isotherms reflected the equilibria [25]. The adsorption equilibrium defines the distribution of the solute between the liquid and the solid phases after the adsorption reaction reaches equilibrium
at constant temperature. Two-parameter isotherm models (Langmuir and Freundlich) were used to fit the experimental data. The Langmuir (Eq. (1)) and the Freundlich (Eq. (2)) equations are used to describe the adsorption equilibrium for water and wastewater treatment applications [26]. The adsorption isotherms are shown in Fig. 2. The correlation coefficients (R2 > 0.99) given in Table 1 showed that the Langmuir equation gives a better fit for the adsorption isotherms of sponge iron and zeolite. Based on this fact, the Langmuir isotherm model was the best fit for this analysis. In this regard, the shape of the isotherms for the substrates/phosphorus systems indicated that there was selectivity for the substrates toward the incoming ion compared to the ions already bound in its lattice. In other words, the substrates favored the replacement of its ions by the solute. The Langmuir model assumed that the energy of adsorption was the same for all surface sites and did not depend on the degree of coverage. This was indicative of monolayer adsorption of the substrates [27]. More sites in the substrates were filled up. It thus became increasingly difficult for a bombarding solute molecule to find an available vacant site. This implies that there was no strong competition from the aqueous solution. The correlation coefficients (R2 > 0.99) in Table 1 show that the Freundlich equation yielded a fit to the adsorption isotherms of the substrates. The Freundlich model assumes that the frequency of sites associated with free energy of adsorption decreases exponentially with increasing free energy. In addition, the energy of adsorption may vary because real surfaces are heterogeneous surfaces (multilayer adsorption) with a non-uniform distribution [27,28]. The values of maximum phosphorus adsorption capacity and Langmuir constants were calculated by the slope and intercept of Langmuir equations (Table 1). The maximum phosphorus adsorption capacity for sponge iron and zeolite was found to be 1111.11 mg/kg and 303.03 mg/kg, respectively. The result indicated that the sponge iron favored more replacement of its ions by the phosphorus ion than zeolite. The 1/n constant for the Freundlich isotherm is a measure of exchange intensity or surface heterogeneity and ranges between 0 and 1. In this study, the values of 1/n were less than 1 (Table 1), which shows that the adsorption conditions were favorable. The value of n for the sponge iron was smaller than that of zeolite. High value of 1/n indicated a strong bond between the substrate and the phosphorus. The data showed that the KF constant for the sponge iron was higher than that of zeolite. The values of Freundlich constant KF in the result indicated the extent of phosphorus removal, hence a measure of adsorption capacity as Langmuir Qm. Whereas sponge iron had high values of Qm (1111.11 mg/kg) and KF (155.02), the corresponding values for zeolite were relatively lower (303.03 mg/kg and 87.49, respectively). Therefore, Langmuir Qm and Freundlich KF indicated and compared the adsorption performance, but from the Langmuir Qm, the specific surface area could be estimated. These implied a more favorable adsorption of phosphorus ion by the sponge iron.
3.2. Adsorption kinetic modeling
Fig. 1. Schematic diagram of experimental set-up (1. Water tank; 2. Metering pump; 3. Elevated water tank; 4. Flowmeter; 5. Waste recovery outlet; 6. Column filled substrate; 7. Water tank).
Available studies have shown that the pseudo-second-order kinetic equation (Eq. (3)) is a reasonably good fit for data over the entire fractional approach to equilibrium. Therefore, it has been employed extensively in the study of adsorption kinetics [29–32]. Table 2 shows that the model was the best fit to the experimental data with correlation coefficients (R2 > 0.999). The adsorption rate denoted by dQt/dt shows how much phosphorus was adsorbed from the aqueous solution onto the substrates within a unit time. Fig. 3 reveals that the adsorption rate (dQt/dt) decreases with time
249
400
400
300
300
Qe (mg P/Kg)
Qe (mg P/Kg)
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
200
100
0
Experimental Langmuir Freundlich 0
0.5
1
1.5
2
2.5
200
100
0
Experimental Langmuir Freundlich 0
1
2
3
4
Ce (mg/L)
Ce (mg/L)
(a)
(b)
5
6
Fig. 2. Substrates’ phosphorus adsorption isotherm at 25 ± 1 °C, (a) sponge iron phosphorus adsorption isotherm and (b) zeolite phosphorus adsorption isotherm.
Table 1 Langmuir and Freundlich constants for adsorption of phosphorus by sponge iron and zeolite.
Sponge iron Zeolite
Langmuir
Freundlich
k1 (L/ mg)
Qm (mg/ kg)
R2
KF (mg/kg)/ (mg/L)1/n
n
R2
0.164
1111.11
0.9984
155.02
1.08
0.9941
0.458
303.03
0.9913
87.49
1.31
0.9922
Removal efficiency (%)
Substrates
100
80 60
40 Sponge iron
20
Zeolite
until it gradually approaches the equilibrium state. This is due to the continuous decrease in the driving force (QeQt). The adsorption of phosphorus onto sponge iron and zeolite as a function of contact time is shown in Fig. 3. The pseudo-second-order predicted adsorption capacity at equilibrium, Qe. This was consistent with the experimental ones (Table 2). At equilibrium, the adsorption process is supposed to be in a dynamic state. In this regard, the rates of forward process and backward process are equal [33]. The time at which the adsorption equilibrium is attained is known as equilibrium time. According to the results, the adsorption process was rapid at the beginning. The equilibrium for both substrates was reached after 8 h. Many preliminary investigations also indicate that most adsorption of solutes occur within a short contact time. The rapid rate of adsorption thereafter slows down and becomes significantly lower at equilibrium. For longer periods, the adsorption capacity tends to remain constant. The results also suggest that the pseudo-second-order model was the most suitable model for the studied experimental conditions. The application of the model relied on the assumption that adsorption may be the rate-limiting step for chemical adsorption involving valency forces through sharing or exchange of electrons between the substrates and phosphorus. It was selected as a measure of the substrates’ selectivity and was consistent with the experimental data. The respective values for all the systems used are given in Table 2. The k2 values of the substrates implied a higher selectivity or adsorption for sponge
Table 2 Kinetics parameters for adsorption of phosphorus by sponge iron and zeolite. Substrates
0
0
k2 (kg/mg h) Qe (mg/kg) R2 0.00795 0.00445
116.279 109.89
24
32
40
48
Fig. 3. The TP removal efficiency of sponge iron and zeolite under different hydraulic retention time.
iron. It was observed that the k2 values of the sponge iron was a higher than the zeolite. 3.3. Phosphorus desorption study After saturation of the substrates, the investigation of phosphorus desorption for phosphorus retained in the substrates is also very crucial. The phosphorus desorption data of sponge iron and zeolite are presented in Table 3. Whereas a phosphorus release value of 5.7122 mg/kg was noted for sponge iron, a corresponding value of 12.0762 mg/kg was noted for zeolite at the equilibrium time. The desorption results suggested that phosphorus was tightly bound to the adsorbents. In desorption experiments, the substrates rapidly increased in volume once placed in water. In the process, the adsorbed phosphorus in the substrates slowly diffused into the water solution. With an increase in time, the equilibrium was reached gradually. Table 3 shows that the phosphorus desorption rate of sponge iron was lower than that of zeolite. It is evident from the low desorption values that the adsorption of phosphorus onto the
Table 3 The equilibrium desorption content and desorption rate of saturated substrates. Substrates
Saturated adsorption capacity (mg/kg)
Equilibrium desorption content (mg/kg)
Desorption rate (%)
Sponge iron Zeolite
1111.11 303.03
5.7122 12.0762
0.5141 3.9851
(QeQe,exp)/Qe
0.9998 0.028053 0.9991 0.03765
16
Time (h)
Qe,exp (mg/kg) Pseudo-second-order model
Sponge iron 113.017 Zeolite 105.753
8
250
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
substrates was chemical in nature, coupled with chemical bonding and electrostatic attraction. Chemisorption exhibits poor desorption, it may be due to fact that the adsorbate species are firmly held to the adsorbent by comparatively stronger bonds. For the sponge iron, the reaction products (Fe(OH)2 and Fe(OH)3) in water formed a high number of flocs. These could easily fix the phosphorus and unease the released adsorbed phosphorus. The formation of Fe–P compounds such as Fe1.6H2PO4(OH)3.8 compound. The desorption of Fe–P was difficult. However, the desorption value of zeolite was relatively higher. Aluminum oxide and silicon oxide form complexes in solution. The solid–solution interface of complexes leads to development of positive or negative charge on the surface. The charged surface sites on the adsorbent favored the adsorption of phosphate due to the electrostatic attraction or ion-exchange. For the substrates, sponge iron is more apt to chemical combination than zeolite. However, zeolite is more apt to electrostatic attraction or ion-exchange than sponge iron. 3.4. Adsorptive removal of phosphorus from wastewater Static experiments results showed that the adsorption of phosphorus onto sponge iron was more apt to chemical combination while zeolite was more apt to electrostatic attraction or ion-exchange. Dynamic analysis involved the study of adsorption mechanism of sponge iron and zeolite upon different phosphorus species in wastewater. The column batch experiments of phosphorus removal would provide more realistic laboratory results. This is because they had a greater resemblance to the flux conditions in full scale constructed wetlands than in short-term stirred batch experiments. This could result in overestimation of adsorption capacities [4,33]. The maximum adsorption capacity depends on the concentration of the tested solution [4]. In column experiments, many wastewater factors such as COD, water turbidity and hydraulic retention time, affect substrates’ adsorption of phosphorus from wastewater. In this study, the wastewater’s initial CODcr, BOD5, and water turbidity were 48.36 mg/L, 10.50 mg/L and 36.10 NTU, respectively. Static experiments showed much better results than dynamic experiments. For example, considering the adsorption capacity of TP, the equilibrium adsorption capacity using sponge iron decreased from 113.02 mg/kg under static experiments to 87.89 mg/kg under dynamic experiments. Similarly, using zeolite, the capacity decreased from 105.75 mg/kg under static experiments to 55.57 mg/kg under dynamic experiments. This is because other substances in wastewater affected phosphorus adsorption. The variation of hydraulic retention time in phosphorus removal by the two substrates is shown in Fig. 4. The adsorptive removal of TP, TDP, PP, SRP, and DOP was analyzed. Fig. 4 shows that
the adsorptive removal of various forms phosphorus stabilized after 4 h. In addition, with the exception of PP, the removal efficiency of phosphorus from the wastewater by sponge iron was higher than that of zeolite. This is shown in Fig. 4. The adsorptive removal of phosphorus stabilized after an adsorption time of 4 h. The phosphorus removal efficiency of each substrate increased with increasing adsorption time. When the adsorption time was 4 h, the TP removal efficiencies for sponge iron and zeolite were 94.93% and 73.62%, respectively. When the adsorption time was 8 h, the corresponding TP removal efficiencies increased to 96.56% and 80.52%. Similarly, when the adsorption time increased from 4 h to 8 h, TDP removal efficiencies for sponge iron and zeolite increased from 99.10% to 99.92% and 69.09% to 75.30%, respectively. In addition, the corresponding removal efficiencies of PP increased from 79.26% to 83.90% and 84.89% to 93.51% respectively. Similarly, whereas the SRP removal efficiencies for sponge iron and zeolite increased from 99.97% to 99.99% and 67.58% to 72.65% respectively, the corresponding DOP removal efficiencies increased from 78.03% to 98.04% and 75.60% to 88.35% for the substrates. After 8 h of adsorption, the removal efficiency of TP, TDP, SRP, and DOP using sponge iron was approximately 16.61%, 24.64%, 27.34%, and 9.88% higher that of zeolite, respectively. However, the removal efficiency of PP using sponge iron was 10.28% lower that of zeolite. In batch column experiments, when hydraulic retention time was 4 h, as shown in Fig. 5b, the TP removal efficiencies by sponge iron and zeolite were 97.79% and 68.67%, respectively. TDP, PP, SRP, and DOP were also analyzed. Whereas the removal efficiencies of these phosphorus species by sponge iron were 99.13%, 75.04%, 99.55% and 96.77%, the corresponding efficiencies for zeolite were 65.87%, 80.06%, 65.22%, and 71.32%. TP accounts for both particulate-bound phosphorus (PP adsorbed to sediment particles) and all forms of TDP. Whereas the adsorptive removal of PP was mainly due to electrostatic attraction or ion-exchange, the adsorptive re moval of TDP (HPO2 4 and H2 PO4 ) was dependent on the chemical compound. As shown in Fig. 5a, after a 4 h adsorption, the adsorption capacities of TP, SRP, and DOP by sponge iron were 87.89, 62.15, and 16.78 mg/kg, respectively. The corresponding capacities for zeolite were 55.57, 38.86, and 5.07 mg/kg, respectively. The adsorption capacities of TDP using sponge iron and zeolite were 78.96 mg/kg and 43.90 mg/kg, respectively. And the adsorption capacity of PP using sponge iron and zeolite was 8.94 mg/kg and 11.68 mg/kg, respectively. Whereas the removal capacity of TDP using sponge iron was 79.86% higher that of zeolite, the removal capacity of PP using sponge iron was 23.46% lower that of zeolite. Fig. 5b describes the adsorption removal efficiency of all forms of phosphorus.
100
80
60
TP TDP PP SRP DOP
40 20
1
2
3
4
5
6
7
8
Removal efficiency (%)
Removal efficiency (%)
100
80 60
TP TDP PP SRP DOP
40 20
1
2
3
4
5
Time (h)
Time (h)
(a)
(b)
6
7
8
Fig. 4. Adsorptive removal efficiency of phosphorus from wastewater onto the substrates under different HRT, (a) adsorptive removal efficiency of phosphorus onto sponge iron and (b) adsorptive removal efficiency of phosphorus onto zeolite.
251
100
Sponge iron Zeolite
120
Removal efficiency (%)
Adsorption capacity (mg/Kg)
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
80 60 40 20
100 80 60 40 20 0
0 TP
TDP
PP
(a)
SRP
DOP
Sponge iron Zeolite
TP
TDP
PP
SRP
DOP
(b)
Fig. 5. Adsorptive removal capacity of phosphorus from wastewater onto sponge iron and zeolite, (a) adsorptive removal capacity of phosphorus onto the substrates and (b) adsorptive removal efficiency of phosphorus onto the substrates.
This phenomenon could further confirm the static experiments’ results that the adsorption of phosphorus onto sponge iron was more apt to chemical combination while zeolite was more apt to electrostatic attraction or ion-exchange. Considering the structure of the substrates, sponge iron possessed some favorable characteristics, such as higher surface area and higher surface energy. To obtain higher contaminants removal efficiency, adequate surface area and intension should be taken into account during the production process of sponge iron [34]. When electrochemistry reaction took place, some Fe2O3 was transformed to Fe3+ or Fe2+. TDP can be further separated into inorganic (soluble reactive phosphorus or SRP) and organic (dissolved organic phosphorus or DOP) components [21]; HPO2 and H2 PO 4 4 are commonly used to quantify TDP. In wastewater, a large number of products including complexes, polymers, and precipitates are formed [35]. The equations below describe the equilibrium distributions of the soluble orthophosphate species as well as soluble iron species [35].
Fe3þ þ H2 O ¼ FeOH2þ þ Hþ
ð4Þ
Fe3þ þ 2H2 O ¼ FeðOHÞþ2 þ 2Hþ
ð5Þ
phase; the content of Fe–P was chemically adsorbed. It is well known that phosphorus ion has a relatively strong affinity for the substrate surfaces [36]. The chief constituents of zeolite are mainly metal oxides like Al2O3 and SiO2. These metal oxides form metal-hydroxide complexes in solution, and the subsequent acidic or basic dissociation of these complexes at the solid-solution interface leads to development of a positive or negative charge on the surface. Positive charge surface sites on the adsorbent favored the adsorption of phosphate due to the electrostatic attraction. The particulatebound phosphorus (PP, adsorbed to sediment particles) was strongly adsorbed via electrostatic attraction. The substrates/phosphorus systems had highly polar solute and substrate as well as a non-polar solvent. In the systems, the monofunctional ions substances with very strong intermolecular attraction were adsorbed from water through ion-ion attraction. It is possible that the adsorbed ions may have become associated into very large clusters just before adsorption took place [37]. This relatively high adsorption capacity shows the strong electrostatic force of attraction between the phosphorus molecules and the substrates bending sites [38]. In the wastewater, soluble reactive phosphorus or SRP was also adsorbed by ion–ion attraction or ion-exchange using zeolite.
þ 2Fe3þ þ 2H2 O ¼ FeðOHÞ4þ 2 þ 2H
ð6Þ
4. Conclusions
Fe3þ þ 3H2 O ¼ FeðOHÞ3ðaqÞ þ 3Hþ
ð7Þ
Fe3þ þ 4H2 O ¼ FeðOHÞ4 þ 4Hþ
ð8Þ
þ 3Fe3þ þ 4H2 O ¼ Fe3 ðOHÞ5þ 4 þ 4H
ð9Þ
The present study focused on the mechanism and adsorption capability of wetland substrates (sponge iron and zeolite) during phosphorus removal processes. Static experiments and dynamic experiments results showed that whereas the adsorption of phosphorus onto sponge iron was more inclined to chemical combination, that of zeolite was more inclined to electrostatic attraction or ion-exchange. In treating wastewater using sponge iron, electrochemistry reactions occurred. Consequently, a large number of products were formed, contributing to significant flocculation. The products so formed could easily fix phosphorus because of the strong affinity for the phosphorus ion by the substrate surfaces. In treating wastewater using zeolite led to the development of a positive or negative charge at the solid-solution interface. The strong electrostatic force of attraction between the phosphorus and the zeolite was due to ion–ion attraction or ion-exchange. Unlike conventional substrates, sponge iron and zeolite captured phosphorus through adsorption, allowing high levels of phosphorus removal to be achieved. In addition, unlike other substrates, their desorption rates are very low. The different adsorption mechanism and significantly higher adsorption capacities for sponge iron and zeolite could be applied to separate different forms phosphorus from wastewater. In addition, the choice on which substrate to use for different kinds of wastewater can be
H3 PO4 ¼ Hþ þ H2 PO4
ð10Þ
H2 PO4 ¼ Hþ þ HPO2 4
ð11Þ
3 þ HPO2 4 ¼ H þ PO4
ð12Þ
þ Fe3þ þ HPO2 4 ¼ FeHPO4
ð13Þ
Fe3þ þ H2 PO4 ¼ FeH2 PO2þ 4
ð14Þ
1:6Fe3þ þ H2 PO4 þ 3:8OH ¼ Fe1:6 H2 PO4 ðOHÞ3:8ðsÞ
ð15Þ
OH could be produced when dissolved oxygen acted as an electron receptor, and the reaction products (Fe(OH)2 and Fe(OH)3) in wastewater contributed to significant flocculation, which could easily fix phosphorus. Fe1.6H2PO4(OH)3.8(s) compound was solid
252
C. Jiang et al. / Journal of Colloid and Interface Science 402 (2013) 246–252
made. Further practical application of sponge iron and zeolite in the constructed wetland treatment systems should be enhanced for future sustainability in industrial/municipal wastewater treatment. Acknowledgments This work was supported by Important National Science and Technology Specific Projects (2013ZX07206001), the National Research Foundation Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE), Science and Technology Commission of Shanghai Municipality Projects (09DZ1200109), and Technological Research and Development Programs of the Ministry of Railways (J2011Z003). The authors thank fellow colleagues for their help and valuable advice. The authors also express their sincere gratitude to all the editors and reviewers of this manuscript. References [1] K.R. Hench, G.K. Bissonnette, A.J. Sexstone, J.G. Coleman, K. Garbutt, J.G. Skousen, Water Res. 37 (2003) 921–927. [2] C.C. Tanner, Ecol. Eng. 7 (1996) 59–83. [3] D.A. Hammer, Ecol. Eng. 1 (1992) 49–82. [4] A. Drizo, Y. Comeau, C. Forget, R.P. Chapuis, Environ. Sci. Technol. 36 (2002) 4642–4648. [5] D.M.R. Mateus, M.M.N. Vaz, H.J.O. Pinho, Ecol. Eng. 41 (2012) 65–69. [6] D. Wu, B. Zhang, C. Li, Z. Zhang, H. Kong, J. Colloid Interface Sci. 304 (2006) 300–306. [7] H. Ye, F. Chen, Y. Sheng, G. Sheng, J. Fu, Sep. Purif. Technol. 50 (2006) 283–290.
[8] J. Chen, H. Kong, D. Wu, X. Chen, D. Zhang, Z. Sun, J. Hazard. Mater. 139 (2) (2007) 293–300. [9] E. Oguz, J. Hazard. Mater. 114 (2004) 131–137. [10] A.K. Golder, A.N. Samanta, S. Ray, Sep. Purif. Technol. 52 (2006) 102–109. [11] L. Zeng, X. Li, J. Liu, Water Res. 38 (2004) 1318–1326. [12] A. Genz, A. Kornmüller, M. Jekel, Water Res. 38 (2004) 3523–3530. [13] M. Özacar, Chemosphere 51 (2003) 321–327. [14] M. Dog˘an, M. Alkan, Ü. Çakir, J. Colloid Interface Sci. 192 (1997) 114–118. [15] W.G. Walter, 22nd ed., A PHA, AWWA, WEF, Portland, 2012. [16] M.E. Ellison, M.T. Brett, Water Res. 40 (2006) 1258–1268. [17] O. Broberg, G. Persson, Hydrobiologia 170 (1988) 61–90. [18] J. Murphy, J.P. Riley, Anal. Chim. Acta 27 (1962) 31–36. [19] K.R. Rinker, R.T. Powell, Mar. Chem. 102 (2006) 170–179. [20] M.A. Wahab, S. Jellali, N. Jedidi, Bioresour. Technol. 101 (2010) 5070–5075. [21] L. Svarovski, Edited by, 3rd ed., Woburn, 1990. [22] N.I. Chubar, V.A. Kanibolotskyy, V.V. Strelko, G.G. Gallios, V.F. Samanidou, T.O. Shaposhnikova, V.G. Milgrandt, I.Z. Zhuravlev, Eng. Aspects 255 (2005) 55–63. [23] J. Wang, C. Zheng, S. Ding, H. Ma, Y. Ji, Desalination 273 (2011) 285–291. [24] J. Vymazal, Ecol. Eng. 18 (2002) 633–646. [25] Z. Talip, M. Eral, Ü. Hiçsönmez, J. Environ. Radioact. 100 (2009) 139–143. [26] N. Ünlü, M. Ersoz, J. Hazard. Mater. 136 (2006) 272–280. [27] M. Metaxas, V. Kasselouri-Rigopoulou, P. Galiatsatou, C. Konstantopoulou, D. Oikonomou, J. Hazard. Mater. B97 (2003) 71–82. [28] Z. Ma, Q. Li, Q. Yue, B. Gao, W. Li, X. Xu, Q. Zhong, Chem. Eng. J. 171 (2011) 1209–1217. [29] F.C. Wu, R.L. Tseng, R.S. Juang, Water Res. 35 (2001) 613–618. [30] C.Y. Chang, W.T. Tsai, C.H. Ing, C.F. Chang, J. Colloid Interface Sci. 260 (2003) 273–279. [31] A.S. Özcan, B. Erdem, A. Özcan, J. Colloid Interface Sci. 280 (2004) 44–54. [32] M.Y. Chang, R.S. Juang, J. Colloid Interface Sci. 278 (2004) 18–25. [33] D.M.R. Mateus, H.J.O. Pinho, Water Environ. Res. 82 (2010) 128–137. [34] J. Li, J. Li, Y. Li, J. Environ. Suppl. (2009) S60–S64. [35] K. Fytianos, E. Voudrias, N. Raikos, Environ. Pollut. 101 (1998) 123–130. [36] W. Stumm, Wiley, New York, 1992. [37] C.H. Giles, T.H. MacEwan, S.N. Nakhwa, D. Smith, J. Chem. Soc. (1960) 3973– 3993. [38] P. Kaewsarn, Q. Yu, Environ. Pollut. 112 (2001) 209–213.