A continuous adsorption study in a fixed-bed column was carried out using granulated slag (GS) as an adsorbent for the removal of methylene blue (MB) from aqueous solution. The effects of various parameters, such as initial dye concentration, flow rate, bed depth, and pH were investigated. Obtained results confirmed that the breakthrough time and exhaustion time were dependent on these factors. The adsorption capacity of GS was calculated at the 50% breakthrough point for different conditions. The highest breakthrough capacity (q,exp = 0.296 mg.g–1) was obtained with a 15 cm bed height and a 2 mL.min–1 rate by using a 10 mg.L–1 initial MB concentration at pH 7.5. Bohart–Adams, Bed Depth Service Time (BDST), and Thomas models were applied to experimental data to determine the characteristic parameters of the column. The Thomas model was found suitable for the description of the whole breakthrough curve, while the Bohart–Adams model was only used to predict the initial part of the dynamic process. The data were in good agreement with the BDST model. Thus, the granulated slag can be used as an adsorbent in the treatment of wastewater. Desorption was carried out with a deionized water as the desorbing agent, and reuse study was investigated.

NOMENCLATURE

  • C0

    Initial MB concentration (mg.L–1)

  • Ct

    Effluent MB concentration (mg.L–1)

  • Cb

    Breakthrough concentration (mg.L–1)

  • F

    Linear velocity (cm.min–1)

  • Ka

    Rate constant in BDST model (L.mg–1.min–1)

  • KBA

    Kinetic constant of Bohart–Adams model (L.mg–1.min–1)

  • KTH

    Kinetic constant of Thomas model (L.mg–1.min–1)

  • N0

    Saturation concentration of Bohart–Adams model (mg.L–1)

  • N0'

    Adsorption capacity in BDST model (mg.L–1)

  • q50%

    Breakthrough capacity (mg.g–1)

  • q

    Adsorption capacity of Thomas model (mg.g–1)

  • Q

    Flow rate (mL·min–1)

  • t

    Effluent time (min)

  • tb

    Breakthrough time (min)

  • te

    Exhaustion or saturation time (min)

  • W

    Adsorbent mass (g)

  • Z

    Bed height of column (cm)

INTRODUCTION

Wastewaters discharged from many industries related to textiles, paper, plastics, food, and cosmetics cause water pollution and serious problems to the environment (Pearce et al. 2003; Aksu 2005; Zulfikar & Setiyanto 2013; Milenova et al. 2014). Due to their ease of production and cost-effectiveness, synthetic dyes are preferred in the textile industry compared to natural dyes (Abdul Halim & Kar Mee 2011; Kumar & Tamilarasan 2013). They have complex molecular structures, which make them more stable and difficult to biodegrade (Seshadri et al. 1994; Fewson 1998; Li et al. 2009). Therefore, wastewaters containing dyes are treated before being discharged into water bodies (Elkassimi et al. 1998) by using different conventional methods (Blackburn 2004; Sampedro et al. 2004; Crini 2006; Mustafa et al. 2008). The adsorption technique is widely used and recommended in the treatment of effluents due to its inexpensive nature and ease of use (Aksu 2005). Batch adsorption experiments are generally used for the treatment of small volumes of effluent. Instead, adsorption in a fixed bed column is more applicable due to its low operating cost and the ability of columns to adapt to versatile processes (Cheknane et al. 2012).

Recently, many works have been studied for the development of low-cost adsorbents for water treatment including natural materials, biosorbents, and waste materials from industry and agriculture (Crini 2006; Han et al. 2009; Atshan 2014). Granulated slag (GS) is an industrial by-product in the production of cast iron, which causes a disposal problem. It was converted into an effective adsorbent and used for the removal of organic/inorganic pollutants (Ramakrishna & Viraraghvan 1997; Dimitrova & Mehandgiev 1998; Kostura et al. 2005; Das et al. 2007; Zhou & Haynes 2010). The presence of CaO, SiO2, and Al2O3 in granulated slag contributes to its good adsorbent properties.

The focus of this work is to investigate the possibility of using granulated slag as an adsorbent for the removal of methylene blue (MB) from aqueous solution in a fixed bed column. The effects of initial dye concentration, flow rate, and bed depth on the breakthrough characteristics of the adsorption system were determined. The obtained results were presented in term of breakthrough curves. Three models, namely Bohart–Adams, Bed Depth Service Time (BDST), and Thomas were employed to predict the breakthrough curves and to determine the characteristic parameters useful for column design.

MATERIALS AND METHODS

The GS is a byproduct from the manufacture of Arcelor-Mittal (Algeria) with a rate of 97% vitrification, but low water content (Ader 1981). The material was dried for 24 h in an oven at 105–110 °C. The chemical composition of GS was determined in the Arcelor-Mittal laboratory using Fluorescence X (TW16006 PHILIPS). X-ray spectra showed that the used slag has an amorphous structure (Arabi et al. 2012). Size distribution was measured using a laser granulometer (Malvern Master Sizer), and the diameter was between 50 and 80 μm. The specific surface area was determined by the Brunauer–Emmett–Teller (BET) method (Brunauer et al. 1938) using a Micromeritics ASAP 2010 apparatus, and was found to be equal to 2.42 m2.g–1.

The basic dye, methylene blue, is the most commonly used material for dying cotton, wood, and silk. It was chosen in this study because of its strong adsorption onto solids. The chemical structures with general data of the dye are represented in Table 1. Serial dilutions of the stock solution were made to obtain specific concentrations required for the adsorption study.

Table 1

General data on methylene blue

Dye Formula Molecular weight (g.mol−1Chemical structure Wave length (nm) 
Methylene blue C16H18 ClN3319.5  663 
Dye Formula Molecular weight (g.mol−1Chemical structure Wave length (nm) 
Methylene blue C16H18 ClN3319.5  663 

Kinetic models

In order to predict the breakthrough curves of the adsorption process in the fixed bed, and estimate the parameters necessary for the design of a large-scale fixed-bed adsorber, the Bohart–Adams, BDST, and Thomas models have been used.

The Bohart–Adams model (Bohart & Adams 1920)

This is based on the surface reaction theory, to describe the relationship between Ct/C0 and t in a continuous system, and is applied to describe the initial part of the breakthrough curve. The expression is given below (Ahmad & Hameed 2010): 
formula
1

The BDST model (Hutchins 1973)

The BDST model is used to predict the bed capacity by utilizing the different breakthrough values. The modified version of the equation used in this evaluation is given in Equation (2) (Janet et al. 2015): 
formula
2

Thomas model (Thomas 1944)

This model is one of the most general and widely used models in the column performance theory. The linearized form of the model is given as (Han et al. 2009): 
formula
3

Error analysis

A linear regressive method is used to compare the models using experimental data. To evaluate the validity of the models, there are different types of errors that indicate the appropriateness between the experimental and calculated adsorption capacity values. Therefore, the conformity of a model is better when the correlation coefficient is higher (Mathialagan & Viraraghavan 2002) and the error is lower (Tan et al. 2007). For this purpose, the experimental data were used to determine the error represented by the relative mathematical formula SS (Equation (4)) (Han et al. 2009): 
formula
4
where is the calculated ratio of MB concentrations according to dynamic models, and is the experimental ratio of MB concentrations. N is the number of the experimental point.

Experimental setup

Fixed bed column studies were performed at room temperature using a glass column of 1.1 cm internal diameter and 25 cm in length. A known quantity of adsorbent was then placed in the column to submit the desired bed height (5, 10 and 15 cm), keeping the same particle size (50–80 μm) obtained by sieving in order to maintain the same porosity for the adsorbent. A dye solution of known concentrations of 5, 10, and 15 mg.L–1 was pumped upward through the bed at different flow rates of 1, 2, and 3 mL.min–1 with a peristaltic pump (ISMATEC). Samples were collected at regular time intervals from the exit of the column for all the experiments. The concentration of the solution in the effluent was analyzed using a UV/VIS spectrophotometer (TECHCOMP) at 663 nm. Operation of the column was stopped when saturation was achieved. The experiments were performed in triplicate, and mean values were taken into account.

Analysis of column data

The operation of an adsorption column in dynamic mode can be determined by using the time to achieve the breakthrough and the shape of the breakthrough curve (Kanadasan et al. 2010). A typical breakthrough curve (Figure 1) is generally expressed by plotting Ct (effluent concentration) and Ct/C0 (the normalized concentration, defined as the ratio of Ct to C0 as a function of time (min) or treated volume V (mL).

Figure 1

Typical breakthrough curve (Al-Degs et al. 2009).

Figure 1

Typical breakthrough curve (Al-Degs et al. 2009).

The concentration at breakthrough point Cb, is chosen arbitrarily at a low value (in this study, when Ct reaches 10% of its initial value). When Ct approaches 90% of its initial concentration, the adsorbent is considered to be essentially exhausted (Faust & Aly 1987; Suksabye et al. 2008).

Breakthrough capacity q50% (at 50% or Ct/C0 = 0.5) expressed in mg of dye adsorbed per gram of adsorbent was calculated by the following equation (Sadaf & Bhatti 2014): 
formula
5

RESULTS AND DISCUSSION

Characterization of the adsorbent

The results of the study of the chemical composition are shown in Table 2.

Table 2

Chemical composition of GS of Arcelor-Mittal (Algeria)

Constituents SiO2 CaO Al2O3 MgO Fe2O3 TiO2 MnO K2Na2BaO 
(massic %) 35.68 42.17 7.92 7.49 0.5 0.29 3.04 1.518 0.51 0.29 1.74 
Constituents SiO2 CaO Al2O3 MgO Fe2O3 TiO2 MnO K2Na2BaO 
(massic %) 35.68 42.17 7.92 7.49 0.5 0.29 3.04 1.518 0.51 0.29 1.74 

From the results, the GS seems to be constituted essentially of four elements: CaO, SiO2, Al2O3, and MgO.

The effect of various parameters influencing adsorption was determined and the obtained results are presented in Table 3.

Table 3

Column data and parameters with different conditions

Initial concentration C0 (mg.L–1Bed height Z (cm) Flow rate Q (mL.min–1Breakthrough time tb (min) Exhaustion time te (min) Breakthrough capacity q,exp (50%) (mg.g–1
10 134 282.97 0.19 
10 10 91.17 260.21 0.23 
15 10 39.15 165.64 0.16 
10 31.17 196.67 0.18 
10 15 158.55 292.42 0.29 
10 10 40.33 236.57 0.19 
10 10 145.91 309.93 0.22 
Initial concentration C0 (mg.L–1Bed height Z (cm) Flow rate Q (mL.min–1Breakthrough time tb (min) Exhaustion time te (min) Breakthrough capacity q,exp (50%) (mg.g–1
10 134 282.97 0.19 
10 10 91.17 260.21 0.23 
15 10 39.15 165.64 0.16 
10 31.17 196.67 0.18 
10 15 158.55 292.42 0.29 
10 10 40.33 236.57 0.19 
10 10 145.91 309.93 0.22 

From these results, we can deduce the effects as follows.

Effect of bed height (amount of adsorbent) on the breakthrough curve

The dynamic adsorption of MB is largely dependent on the bed height, which is directly proportional to the quantity of GS in the column (Kanadasan et al. 2010). To produce 5, 10 and 15 cm of bed height, 5, 10 and 15 g of GS were used, respectively. The breakthrough capacity (q50%) increased from 0.181 to 0.296 mg.g–1 when the bed height (adsorbent mass) increased from 5 to 15 cm, this is due to an increase in the surface area of the adsorbent, which provided more activated sites for MB binding (Zulfadhly et al. 2001). At the same time, the higher bed column resulted in a decrease in the solute concentration in the aqueous solution (Han et al. 2009). Breakthrough time, exhaustion time and the volume of treated solution also increase with the bed height. These results were in agreement with those reported in previous studies (Al-Degs et al. 2009).

Effect of initial concentration on the breakthrough curve

The effect of different initial MB concentrations (5, 10 and 15 mg.L–1) on the adsorption process was investigated. It can be deduced that a slower breakthrough curve will be obtained at a lower initial concentration. This may be explained by the fact that a lower concentration gradient caused slower transport due to a decreased diffusion or mass transfer coefficient (Uddin et al. 2009; Lezehari et al. 2012; Sadaf & Bhatti 2014). Higher initial concentrations led to a higher driving force for mass transfer, hence the adsorbent saturation was achieved more quickly, which resulted in a decrease in the breakthrough and exhaustion time and adsorption zone length (Malkoc et al. 2006; Baral et al. 2009).

Effect of flow rate on the breakthrough curve

In order to perform this investigation, the effect of different flow rates (1, 2 and 3 mL·min–1) was studied. From Table 3 it can be seen that the breakthrough time decreases as the flow rate increases. The faster breakthrough point at higher flow rates is due to reduced contact time between dye molecules and adsorbent (Sadaf & Bhatti 2014). Therefore, the MB solution will leave the bed before the equilibrium can be reached, which will result in a decreasing amount of MB being adsorbed. (Kanadasan et al. 2010). This result was in agreement with the findings of other researchers in the literature (Taty-Costodes et al. 2005; Li et al. 2011).

The adsorption capacity of the GS was found to be lower than that of many other adsorbents (Han et al. 2007; Zhang et al. 2011; Tarawou et al. 2014); the main reason could be the poor porosity and low surface area of the adsorbent (Grassi et al. 2012).

The pH values of influent and effluent during adsorption

The breakthrough curve with various influent pH (3, 5, 7.5, 9 and 12) is shown in Figure 2. The appropriate pH was adjusted by adding 0.1 M NaOH or HCl.

Figure 2

Breakthrough curves for adsorption of MB on GS at different influent pH (Z = 10 cm, C0 = 10 mg.L–1, Q = 2 mL.min–1).

Figure 2

Breakthrough curves for adsorption of MB on GS at different influent pH (Z = 10 cm, C0 = 10 mg.L–1, Q = 2 mL.min–1).

As shown in Figure 2, it may be noted that pH 3 and 12 did not promote the adsorption of MB onto GS. Our adsorbent consists essentially of silicates and aluminosilicates of calcium. As in Runzhang et al. (1988), the highly acidic and alkalic influents provoke the solubilization of these constituents and therefore the destruction of the structure of our GS, which results in a clogging of the column. On the other hand, it is obvious that pH 7.5 is the most appropriate for obtaining a higher breakthrough time and thus a higher adsorption capacity of MB.

Figure 3 shows the effluent pH values after adsorption for different influents.

Figure 3

The effluent pH values after adsorption of MB by GS.

Figure 3

The effluent pH values after adsorption of MB by GS.

It can be observed that there is always an increase in effluent pH values during adsorption.

Breakthrough curve assessment

For the Bohart–Adams model, the characteristic parameters such as No and KBA can be calculated from the linear plot of ln (Ct/C0) against t (Equation (1)) at different bed heights, concentrations and flow rates. The obtained results are represented in Table 4 and Figure 4, where there is an increase in the initial concentration from 5 to 15 mg.L–1, the saturation concentration of MB increases from 259.5 to 294.4 mg.L–1. On the other hand, with increasing flow rate the saturation concentration of MB decreased. The R2 values vary between 0.64 and 0.98. Nevertheless, most of the R2 values are superior to 0.9 and SS values range between 0.001 and 0.135), which indicates that the data fits into the model perfectly (Kanadasan et al. 2010).

Table 4

Bohart–Adams model parameters at various conditions

C0 (mg.L–1Z (cm) Q (mL.min–1kBA (L.mg–1.min–1) ×103 No (mg.L–1R2 SS 
10 2.94 259.5 0.89 0.008 
10 10 5.91 274.68 0.98 0.001 
15 10 1.76 294.4 0.64 0.034 
10 131.2 212.8 0.93 0.038 
10 15 0.85 430.5 0.99 0.018 
10 10 1.77 558.3 0.93 0.002 
10 10 12.68 215.9 0.76 0.135 
C0 (mg.L–1Z (cm) Q (mL.min–1kBA (L.mg–1.min–1) ×103 No (mg.L–1R2 SS 
10 2.94 259.5 0.89 0.008 
10 10 5.91 274.68 0.98 0.001 
15 10 1.76 294.4 0.64 0.034 
10 131.2 212.8 0.93 0.038 
10 15 0.85 430.5 0.99 0.018 
10 10 1.77 558.3 0.93 0.002 
10 10 12.68 215.9 0.76 0.135 
Figure 4

Linear regression of Bohart–Adams model with experimental data at different conditions: (a) Z, (b) C0 and (c) Q.

Figure 4

Linear regression of Bohart–Adams model with experimental data at different conditions: (a) Z, (b) C0 and (c) Q.

For the BDST model, a plot of t versus Z (Equation (2)) is expected to yield a linear curve, in which No' and Ka could be evaluated from the slope and y-axis intersection point, respectively. The results are listed in Table 5 and shown in Figure 5.

Table 5

Bed depth service time of column at different bed height (C = 10 mg.L–1 and Q = 0.9 mL.min–1)

Ct/C0 Ka (L.mg–1.min–1N0′ (mg.L–1R2 
10% −0.0065 269.24 0.99 
90% −0.0014 209.88 0.96 
Ct/C0 Ka (L.mg–1.min–1N0′ (mg.L–1R2 
10% −0.0065 269.24 0.99 
90% −0.0014 209.88 0.96 
Figure 5

Linear regression of the BDST model at different breakthrough points with C0=10 mg.L–1 and Q = 2 mL.min–1.

Figure 5

Linear regression of the BDST model at different breakthrough points with C0=10 mg.L–1 and Q = 2 mL.min–1.

From these results it can be deduced that at different Ct/C0 ratios, the values of the correlation coefficient are high, which show good agreement of the experimental data with the BDST model.

For the Thomas model, maximum adsorption capacity of the adsorbent (q,cal) and kinetic constant (kTH) are determined by fitting the experimental data into Equation (3). Linear regression results, with correlation coefficients and SS values, are listed in Table 6 and presented in Figure 6. It can be observed that for a given experimental condition, the experimental and calculated values are similar. Also, most of the R2 values were greater than 0.89 with smaller SS (less than 0.016), which validates the use of the Thomas model to predict the maximum adsorption capacity of the bed (Kanadasan et al. 2010).

Table 6

Thomas model parameters at various conditions

C0 (mg.L–1Z (cm) Q (mL.min–1KTH (L.mg–1.min–1) ×10 2 q,cal (mg.g–1R2 q,exp (mg.g–1SS 
10 0.60 0.199 0.90 0.195 0.004 
10 10 0.20 0.280 0.89 0.236 0.012 
15 10 0.17 0.181 0.74 0.160 0.014 
10 0.20 0.248 0.71 0.181 0.016 
10 15 0.28 0.285 0.90 0.296 0.003 
10 10 0.27 0.223 0.98 0.191 0.001 
10 10 0.16 0.274 0.83 0.224 0.008 
C0 (mg.L–1Z (cm) Q (mL.min–1KTH (L.mg–1.min–1) ×10 2 q,cal (mg.g–1R2 q,exp (mg.g–1SS 
10 0.60 0.199 0.90 0.195 0.004 
10 10 0.20 0.280 0.89 0.236 0.012 
15 10 0.17 0.181 0.74 0.160 0.014 
10 0.20 0.248 0.71 0.181 0.016 
10 15 0.28 0.285 0.90 0.296 0.003 
10 10 0.27 0.223 0.98 0.191 0.001 
10 10 0.16 0.274 0.83 0.224 0.008 
Figure 6

Linear regression of the Thomas model with experimental data at different conditions: (a) Z, (b) C0 and (c) Q.

Figure 6

Linear regression of the Thomas model with experimental data at different conditions: (a) Z, (b) C0 and (c) Q.

Comparison of Bohart–Adams and Thomas models

According to the values of R2 and SS listed in Tables 4 and 6 at the same experimental conditions, the values of R2 from Bohart–Adams (0.64–0.98) were larger than those from the Thomas model (0.71–0.98). Comparing the SS values, we notice that the SS values of the Thomas model are higher than those of the Bohart–Adams only when C0 = 10 mg.L–1, Z = 10 cm, Q = 2 mL.min–1. These results showed that both models can be selected to predict the column process of MB onto GS.

Column regeneration

The column with a bed depth of 10 cm and flow rate of 2 mL.min–1 saturated with 10 mg.L–1 of MB was selected for a desorption study. Desorption was carried out, and deionized water was used as the desorbing agent. The concentration of MB was measured at different time intervals as shown in Figure 7.

Figure 7

First cycle desorption profile of MB.

Figure 7

First cycle desorption profile of MB.

It was observed that the desorption cycle took 50 min, after which further desorption was negligible.

Reuse study of regenerated column

To check the adsorption efficiency of the regenerated GS column, it was reloaded with a dye solution of 10 mg.L–1 at a rate of 2 mL.min–1.

The breakthrough curve obtained was compared with that of GS (10 cm depth; 10 mg.L–1 MB concentration, flow rate of 2 mL.min–1), and is shown in Figure 8.

Figure 8

Comparative breakthrough curve for original and regenerated GS.

Figure 8

Comparative breakthrough curve for original and regenerated GS.

For GS, the breakthrough time was 91.17 min and exhaustion time was 260.21 min. However, it was decreased for regenerated GS (breakthrough time was 14.30 min and exhaustion time was 249.46 min).

CONCLUSIONS

In the present study, the GS packed bed was used to analyze the column dynamics in the adsorption process. The effect of bed height, initial MB concentration and flow rate on breakthrough curves has been investigated. A higher uptake of MB was observed at higher bed depth (Z = 15 cm). Both breakthrough time and exhaustion time increased with increasing bed height, but decreased with increasing MB initial concentration. Besides, a faster breakthrough curve has been found at higher flow rates. According to the obtained results, the granulated slag can be proposed as an adsorbent in the treatment of wastewater for the removal of dye from an aqueous solution. The dynamic behavior of the column was predicted by the Bohart–Adams, BDST, and Thomas models. All models were found suitable for describing the whole or a definite part of the dynamic behavior of the column.

REFERENCES

REFERENCES
Abdul Halim
H. N.
Kar Mee
K. L.
2011
Adsorption of basic Red 46 by granular activated carbon in a fixed bed column
.
Int. Conf. Environ. Indust. Innov.
12
,
263
267
.
Ader
R.
1981
Psychoneuroimmunology
.
Academic Press
,
New York
,
321
pp.
Al-Degs
Y. S.
Khraisheh
M. A. M.
Allen
S. J.
Ahmad
M. N.
2009
Adsorption characteristics of reactive dyes in columns of activated carbon
.
J. Hazard. Mater.
165
,
944
949
.
Atshan
A. A.
2014
Adsorption of methyl green dye onto bamboo in batch and continuous system
.
Iraq J. Chem. Pet. Eng.
15
(
1
),
65
72
.
Baral
S. S.
Das
N.
Ramulu
T. S.
Sahoo
S. K.
Das
S. N.
Chaudhury
G. R.
2009
Removal of Cr(VI) by thermally activated weed Salvinia cucullata in a fixed-bed column
.
J. Hazard. Mater.
161
,
1427
1435
.
Bohart
G. S.
Adams
E. Q.
1920
Some aspects of the behaviour of the charcoal with respect chlorine
.
J. Am. Chem. Soc.
42
,
523
544
.
Brunauer
S.
Emmett
P. H.
Teller
E.
1938
Adsorption of gases in multimolecular layers
.
J. Am. Chem. Soc.
60
,
309
319
.
Das
B.
Prakash
S.
Reddy
P. S. R.
Misra
V. N.
2007
An overview of utilization of slag and sludge from steel industries
.
Resour. Conserv. Recycl.
50
,
40
57
.
Dimitrova
S. V.
Mehandgiev
D. R.
1998
Lead removal from aqueous solution by granulated blast-furnace slag
.
Water Res.
32
,
3289
3292
.
Elkassimi
M.
Meziane
D.
Abouarnadasse
S.
Azizi
H.
1998
Elimination des colorants de l'industrie de textile par le charbon de bois
.
2ème Conférence Maghrébine de Génie des Procédés
,
Gabès
(
Tunisie
),
ENIG
.
Faust
S. D.
Aly
O. M.
1987
Adsorption Processes for Water Treatment
.
Butterworth Publishers
,
Boston
.
Grassi
M.
Kaykioglu
G.
Belgiorno
V.
Lofrano
G.
2012
Emerging compounds removal from wastewater
.
Springer Briefs Green Chem. Sustain.
10
(
1007
),
978
994
.
Hutchins
R. A.
1973
New method simplifies design of activated carbon systems
.
Chem. Eng.
80
,
133
138
.
Janet
A.
Kumaresan
R.
Uma Maheshwari
S.
2015
Removal of dyes in adsorption column
.
J. Chem. Pharm. Res.
7
(
3
),
1718
1723
.
Kanadasan
G.
Mashitah
M. D.
Vadivelu
V. M.
2010
Fixed bed adsorption of methylene blue by using palm oil mill effluent waste activated sludge
. In:
3rd ISWA Asia Pacific Young Water Professional Conference 2010 Achieving Sustainable Development in the New Era
,
Singapore
.
Kostura
B.
Kulveitova
H.
Lesko
J.
2005
Blast furnace slags as sorbents of phosphate from water solutions
.
Water Res.
39
,
1795
1802
.
Li
L.
Dai
W.
Yu
P.
Zhao
J.
Qu
Y.
2009
Decolorisation of synthetic dyes by crude laccase from Rigidoporus lignosus W1
.
J. Chem. Tech. Biotechnol.
84
(
3
),
399
404
.
Mathialagan
T.
Viraraghavan
T.
2002
Adsorption of cadmium from aqueous solution by perlite
.
J. Hazard. Mater.
B94
,
291
303
.
Milenova
K. I.
Nikolov
P. M.
Georgieva
A. L.
Batakliev
T. T.
Georgiev
V. F.
Rakovsky
S. K.
2014
Discoloration of reactive dyes in wastewaters by ozonation
.
J. Int. Sci. Publ. Ecol. Saf.
8
,
1314
7234
.
Mustafa
A. I.
Alam
S.
Amin
N.
Bahadur
N. M.
Habib
A.
2008
Phenol removal from aqueous system by jute stick
.
Pak. J. Anal. Environ. Chem.
9
,
92
95
.
Ramakrishna
K. R.
Viraraghvan
T.
1997
Use of slag for dye removal
.
Waste Manage.
17
(
8
),
483
488
.
Runzhang
Y.
Qiongying
G.
Shixi
O.
1988
Study on structure and latent hydraulic activity of slag and its activation mechanism
.
Silicates Indust.
53
(
3–4
),
55
59
.
Sampedro
I.
Romero
C.
Ocampo
J. A.
Brenes
M.
Habib
A.
2004
Removal of monomeric phenols in dry mill olive residue by saprobic fungi
.
J. Agr. Food. Chem.
52
,
4487
4492
.
Seshadri
S.
Bishop
P. L.
Agha
A. M.
1994
Anaerobic/aerobic treatment of selected azo dyes in wastewater
.
Waste Manage.
15
,
127
137
.
Tarawou
T.
Young
E.
Ere
D.
2014
Adsorption of methylene blue dye from aqueous solution using activated carbon produced from water hyacinth in a fixed–bed column system
.
Sch. Acad. J. Biosci.
2
(
9
),
607
612
.
Thomas
H. C.
1944
Heterogeneous ion exchange in flowing system
.
J. Am. Chem. Soc.
66
,
1664
1666
.
Zhang
W.
Dong
L.
Yan
H.
Li
H.
Jiang
Z.
Kan
X.
Yang
H.
Li
A.
Cheng
R.
2011
Removal of methylene blue from aqueous solutions by straw based adsorbent in a fixed-bed column
.
Chem. Eng. J.
173
,
429
436
.
Zulfikar
M. A.
Setiyanto
H.
2013
Adsorption of congo red from aqueous solution using powdered eggshell
.
Int. J. Chem. Tech. Res.
5
(
4
),
1532
1540
.