Abstract

The sorption of Cu ions on pure fly ash-based geopolymer, fly ash-based geopolymer with Pb ions addition (modified) and zeolite NaX was studied. Taguchi's approach was applied to determine the impact of solute concentration (c), temperature (T), mixing time (t) and sorbent type (S) on the sorption of Cu ions onto different sorbents under batch conditions. Optimum experimental conditions and influence of controllable factors were determined using the larger-the-better approach. The influence in descending order is c > S > t > T, for both removal and loading. Also, the impact of sorbent type and solute concentration on the process equilibrium was examined. The equilibrium amount of Cu retained on the sorbents in equilibrium (qe) was as follows: pure geopolymer – 1.169 mmol g−1, modified geopolymer – 1.186 mmol g−1, and zeolite NaX – 1.695 mmol g−1. The experimental data were modelled using Jovanovic, Khan, Baudu, and Fritz-Schlünder isotherm models and their goodness of fit were compared. The Baudu isotherm model was the most useful in predicting the equilibrium of Cu sorption on pure and modified geopolymer. Goodness of fit of the selected isotherm models for the sorption of Cu ions on zeolite NaX was in the order: Fritz-Schlünder > Khan > Jovanovic > Baudu.

INTRODUCTION

The discharge of large amounts of wastewaters, contaminated with Cu, Ni, As, Pb, Co, Cd, Cr, and Zn, is a serious problem in the scope of environmental protection. Heavy metals possess high solubility in the aquatic environments and can be absorbed by living organisms, allowing them to enter the food chain and cause serious health disorders for humans (Siegel 2002; Malik 2004; Tchounwou et al. 2012; Abdel-Rahman et al. 2016; Cai et al. 2019; Rai et al. 2019). Hence, the removal of heavy metal from wastewaters prior their discharge into the natural environment is more than necessary. Sorption of heavy metals using geopolymers and zeolites as sorbents is an interesting and cost-effective method for heavy metal removal from wastewaters. Also, according to the literature zeolite could be modified with inorganic salts, resulting in an increase of its affinity towards metals and other specifically sorbing ions due to the larger surface area and increased number of active sites (Doula 2006; Kragović et al. 2012).

Zeolite NaX is a microporous, crystalline solid with a well-defined three-dimensional silica-alumina structure and extra-framework exchangeable sodium cations. Zeolites are appropriate materials for removing heavy metal ions from wastewaters and a very promising support for design and preparation of environmentally friendly catalysts (Svilović et al. 2017).

Geopolymers have been the focus of attention in the field of eco-friendly materials that can be produced from many materials, even the waste. Various aluminosilicate materials such as kaolinite, feldspar and industrial solid residues (fly ash, metallurgical slag, mining wastes, etc.) are typical raw materials used for the preparation of geopolymer composites (Singh et al. 2015). Fly ash is an industrial by-product generated during the combustion of coal in energy production. Since large amounts of fly ash are being released into the biosphere, the disposal of fly ash has become a serious environmental problem. On the other hand, fly ash is a valuable raw material for geopolymer production (Mužek et al. 2012; Zhuang et al. 2016). Due to their excellent properties, such as favourable mechanical and chemical stability, resistance to acid attack and freezing, low shrinkage after forming, and high temperature resistance (Temuujin et al. 2011; Živica et al. 2011; Azimi et al. 2015; Martin et al. 2015; Zhang et al. 2015; Yan et al. 2016), geopolymers are interesting sorbent materials (Li et al. 2006; Wang et al. 2007; Al-Zboon et al. 2011; Chen et al. 2013; López et al. 2014; Mužek et al. 2014; Alshaaer et al. 2015; Khan et al. 2015; Ge et al. 2015; Liu et al. 2016; Luukkonen et al. 2016a, 2016b, 2017; Mužek et al. 2016a, 2016b; Kara et al. 2018; Sudagar et al. 2018). Geopolymers consist of an amorphous three-dimensional structure constituted of [SiO4] and [AlO4] tetrahedra; hence it is expected that geopolymers have unique properties like zeolite sorbents and that their sorption capacity could be increased if they are modified with inorganic salts.

The Taguchi method applies an orthogonal array for experimental design and the signal to noise (S/N) ratio for quality assessment. There are three types of S/N ratio: smaller the best, nominal the best and larger the best (Yen & Li 2015). When compared with the other optimization methods, The Taguchi method demands a smaller number of experiments to find the optimum conditions (Siju et al. 2013).

The goals of this work were to (a) obtain optimum operating conditions for effective removal of Cu ions, (b) obtain optimum operating conditions for maximum loading of Cu ions, (c) make a comparison of different adsorption models that can describe the sorption of Cu ions on sorbents used, and (d) analyse remaining solutions in batch reactors.

MATERIALS AND METHODS

Preparation of sorbents and solutions

Pure fly ash-based geopolymer, modified fly ash-based geopolymer and zeolite NaX were used as a sorbent material in this study. Both pure and modified fly ash-based geopolymers were synthesized from fly ash class F. Detailed chemical analysis of the used fly ash was published in a previous work (Mužek et al. 2012). Sodium hydroxide (Kemika, p.a.) and technical-grade sodium silicate (a type S water glass) (Time d.o.o.) were used as alkaline activator. A certain amount of fly ash was mixed with 16 M NaOH solution and water glass solution in order to prepare paste specimens. Solution/ash ratio was 0.40. Extra water was added during the synthesis of pure fly ash-based geopolymer to provide good workability. A solution containing Pb ions was used instead of extra added water in the synthesis of modified geopolymer. The modified geopolymer has been synthesized to investigate the possibility of incorporation of Pb ions in the structure and to see whether Pb ions are going to be leached out of the geopolymer structure by using that geopolymer as sorbent. The solution of Pb ions was prepared by dissolving appropriate weight of Pb(NO3)2·3H2O (Kemika, p.a.) in distilled water. The content of Pb, accounting for the total mass of fly ash, was 0.1% (Mužek et al. 2013). The mixture obtained was mixed for 10–15 min and poured into polypropylene cylindrical containers (49.1 × 40 mm) which were hermetically sealed to prevent moisture evaporation. The pastes prepared were heat cured at 85 °C for 24 h in an oven and afterwards kept at room temperature for 3 days. After being taken out from the hermetically sealed containers geopolymers were washed at least three times with acetone to remove the excess alkaline activator solution and then crushed and sieved to obtain a particle size less than 0.09 mm. Zeolite NaX (Sigma-Aldrich) with rSi/Al = 1.23 was also crushed and sieved to obtain a particle size less than 0.09 mm.

Solutions containing Cu ions were prepared by dissolving the appropriate weight of Cu(NO3)2·3H2O (Kemika, p.a.) in distilled water. The volume of solutions used in all the experiments was 0.2 dm3. The initial concentrations of Cu solutions were checked by a Perkin Elmer Lambda 201 UV/VIS spectrophotometer.

Taguchi's methodology and statistical analysis

Experiments were planned according to Taguchi's L9 orthogonal array (Table 1), which has nine rows corresponding to the number of experiments and four columns corresponding to the controllable factors. The concentration, sorbent type, time and temperature were chosen as controllable factors and their impact on amount of Cu removal and loading was studied.

Table 1

Taguchi L9 design of experiments

TestFactor
cStT
TestFactor
cStT

As can be seen in Table 2, each factor used had three testing conditions – represented by levels 1, 2 and 3.

Table 2

Controllable factors and associated levels

FactorLevel 1Level 2Level 3
Concentration; c (mmol dm−313 
Sorbent; S Zeolite NaX Modified geopolymer Pure geopolymer 
Time; t (min) 10 30 50 
Temperature; T (°C) 27 31 35 
FactorLevel 1Level 2Level 3
Concentration; c (mmol dm−313 
Sorbent; S Zeolite NaX Modified geopolymer Pure geopolymer 
Time; t (min) 10 30 50 
Temperature; T (°C) 27 31 35 

Taguchi design batch sorption studies were performed in V = 0.35 dm3 batch reactors. Each batch reactor was filled with sorbent and Cu solution. Concentration of sorbent in suspension was 5 g dm−3. The mixture was agitated with a turbine impeller. The values of initial concentration, the sorbent used and experimental conditions (time and temperature) are given in Table 2 and combined according to Table 1.

The samples taken from suspensions at defined times were centrifuged and filtered, and the concentration of the Cu ions in the filtrates was determined by a UV/VIS spectrophotometer at wavelength of 810 nm.

In this study the larger-the-better quality characteristic was used (Yen & Li 2015): 
formula
(1)
where S/N represents signal-to-noise ratio; subscript LB represents larger-the-better; n is the number of repetitions (three) under the same experimental conditions and y is a measurement result, i.e. percentage of Cu removed by sorbents or amount of Cu retained on the sorbents (loading).
The percentage of Cu removed by sorbents is presented as R% and it was calculated by Equation (2): 
formula
(2)
where c0 is the initial concentration of Cu ions in solution (mmol dm−3), and ct is the final concentration of metal in solutions at time t (mmol dm−3).
The amount of Cu retained on the sorbents, qt (mmol g−1), was calculated by Equation (3) 
formula
(3)
where V is the volume of solution (dm3), and m is the mass of sorbent (g).
The next step in the Taguchi method is to calculate the average S/NLB ratio of each controllable factor at level i, denoted as S/NFL, in order to determine the optimal conditions: 
formula
(4)
where represents S/NLB ratio for factor F on the level i, the subscript j is the j-th appearance of the i-th level. For the nine tests, each level for every factor appears three times, nFi (Yen & Li 2015).

Verification of the most effective factor for both removal and loading was achieved by statistical analysis using sum of squares, mean square and percentage of contribution.

Sum of squares for each controllable factor was calculated as follows: 
formula
(5)
where SSF is sum of square of selected controllable factor (parameter), l is number of experiments excluding repetition, and z is number of levels.
Mean square was calculated as: 
formula
(6)
where MSF is a mean square for selected controllable factor, and DoFF is a degree of freedom for factor which was calculated as: 
formula
(7)
Percent of controllable factor was calculated as: 
formula
(8)
where SST is total sum of square: 
formula
(9)

Equilibrium studies

Equilibrium studies of the Cu ions sorption were carried out in batch reactors at 27 °C for all the sorbents used. Solutions containing Cu ions (3.458, 8.034, 13.627, 18.034, 23.881, and 29.983 mmol dm−3) were prepared by dissolving the appropriate weight of Cu(NO3)2·3H2O (Kemika, p.a.) in distilled water. Concentration of sorbent in suspension was the same as for Taguchi analysis, 5 g dm−3.

The mixtures were agitated until the equilibrium was reached. The equilibrium was determined by measuring the Cu ions concentration that remained in the filtered samples of solutions taken out from the system at the certain contact time, using a UV/VIS spectrophotometer. The amount of the Cu retained on the sorbent, qe, was calculated according to Equation (10): 
formula
(10)
where ce is concentration of Cu ions in solution at equilibrium (mmol dm−3).

After reaching the equilibrium, the solutions were additionally analysed for the concentrations of exchangeable cations (Na+, K+, Ca2+, Mg2+, and Pb2+), and pH values were monitored before and after equilibrium studies.

Adsorption isotherms

A sorption isotherm assumes a thermodynamic equilibrium relationship between the amount of material bound to the surface of adsorbent and the amount of the material present in the solution or in the gas phase. Sorption isotherms are often used to provide an insight into the sorption mechanism and the surface properties as well as the degree of sorbent affinity (Foo & Hameed 2010). In this paper various types of isotherm models, Jovanovic, Khan, Baudu and Fritz-Schlünder (Table 3), were used to fit the equilibrium data of Cu sorption from solutions on various sorbent materials using non-linear regression analysis.

Two-parameter isotherm model 
Jovanovic  (11) 
Three-parameter isotherm model 
Khan  (12) 
Four-parameter isotherm model 
Baudu  (13) 
Five-parameter isotherm model 
Fritz-Schlünder  (14) 
Two-parameter isotherm model 
Jovanovic  (11) 
Three-parameter isotherm model 
Khan  (12) 
Four-parameter isotherm model 
Baudu  (13) 
Five-parameter isotherm model 
Fritz-Schlünder  (14) 

qmax is maximum amount of Cu retained on the sorbents in equilibrium (mmol g−1); kJ is Jovanovic isotherm model constant (dm3 mmol−1); kK is Khan isotherm model constant; nK Khan model exponent; kB Baudu equilibrium constant, x and y are the Baudu parameters; k1, k2, m1, and m2 are the Fritz-Schlünder parameters.

Instrumentation

The concentrations of copper nitrate solutions were measured by a Perkin Elmer Lambda EZ 201 UV/VIS spectrophotometer. Concentrations of exchangeable cations (Na+, K+, Ca2+, and Mg2+) after equilibrium were measured by an ion chromatograph (Metrohm 761 Compact IC), and those of Pb2+ using the atomic absorption spectrometer AAS Z-2000. Also, the pH values of the solutions before and after equilibrium were measured using a HANNA instruments pH meter for all the experiments conducted.

RESULTS AND DISCUSSION

Process optimization

It is known that catalytic performance and activity of catalysts depend on metal loading, and according to some studies activity of catalysts increases with the increase of metal loading (Jabłońska & Palkovits 2016). Also, it is desirable to remove as much as possible heavy metals from wastewaters. So, the goal of this investigation was to find the optimum experimental conditions which will provide the highest Cu removal and the highest loading, and for that reason the larger-the-better quality characteristic was used. The results for S/NLB ratios calculated by Equation (1) along with the percentage of Cu removed by the sorbents, i.e. removal (%) Equation (2), the average removal (average of three repeated experiments denoted as R1, R2 and R3) and standard deviation are shown in Table 4.

Table 4

Results of experimental part of Taguchi design, S/NLB ratios and standard deviation (STD) – removal

TestRemoval (%)
Average removal (%)STDS/NLB ratio
R1R2R3
99.859 99.859 99.859 99.859 1.4 × 10−14 39.988 
99.233 97.548 97.548 98.233 0.968 39.844 
99.963 99.939 99.939 99.947 0.011 39.995 
90.736 90.626 90.626 90.662 0.052 39.149 
55.696 57.759 57.759 57.071 0.972 35.124 
45.382 46.413 45.382 45.726 0.486 33.202 
61.308 62.599 61.308 61.738 0.608 35.810 
38.079 38.156 38.110 38.115 0.031 31.622 
30.331 29.686 30.977 30.331 0.527 29.634 
TestRemoval (%)
Average removal (%)STDS/NLB ratio
R1R2R3
99.859 99.859 99.859 99.859 1.4 × 10−14 39.988 
99.233 97.548 97.548 98.233 0.968 39.844 
99.963 99.939 99.939 99.947 0.011 39.995 
90.736 90.626 90.626 90.662 0.052 39.149 
55.696 57.759 57.759 57.071 0.972 35.124 
45.382 46.413 45.382 45.726 0.486 33.202 
61.308 62.599 61.308 61.738 0.608 35.810 
38.079 38.156 38.110 38.115 0.031 31.622 
30.331 29.686 30.977 30.331 0.527 29.634 

The results show that average removal is in the range from 30.331 to 99.947% and S/NLB ratio in the range from 29.634 to 39.995, depending on controllable factors.

The results for S/NLB ratios calculated by Equation (1) along with the amount of Cu retained in the solid phase qt calculated by Equation (3), the average qt (average of three repeated experiments denoted as qt1, qt2, qt3) and standard deviation are shown in Table 5.

Table 5

Results of experimental part of Taguchi design, S/NLB ratios and standard deviation (STD) – loading

Testqt (mmol g−1)
Average qt (mmol g−1)STDS/NLB ratio
qt1qt2qt3
0.657 0.657 0.657 0.657 0.000 −3.649 
0.655 0.641 0.641 0.646 6.365 × 10−3 −3.796 
0.657 0.657 0.657 0.657 7.542 × 10−5 −3.649 
1.483 1.485 1.483 1.484 8.485 × 10−4 3.429 
0.911 0.945 0.945 0.934 0.016 −0.593 
0.743 0.759 0.743 0.748 7.956 × 10−3 −2.522 
1.603 1.637 1.603 1.615 0.016 4.163 
0.996 0.998 0.997 0.997 8.219 × 10−4 −0.202 
0.793 0.776 0.810 0.793 0.014 −2.015 
Testqt (mmol g−1)
Average qt (mmol g−1)STDS/NLB ratio
qt1qt2qt3
0.657 0.657 0.657 0.657 0.000 −3.649 
0.655 0.641 0.641 0.646 6.365 × 10−3 −3.796 
0.657 0.657 0.657 0.657 7.542 × 10−5 −3.649 
1.483 1.485 1.483 1.484 8.485 × 10−4 3.429 
0.911 0.945 0.945 0.934 0.016 −0.593 
0.743 0.759 0.743 0.748 7.956 × 10−3 −2.522 
1.603 1.637 1.603 1.615 0.016 4.163 
0.996 0.998 0.997 0.997 8.219 × 10−4 −0.202 
0.793 0.776 0.810 0.793 0.014 −2.015 

The results show that average qt is in the range from 0.646 to 1.615 mmol g−1 and S/NLB ratio in the range from −3.796 to 4.163, depending on controllable factors.

From these S/NLB ratios, related S/NFL ratios of each controllable factor were calculated and shown in Figures 1 and 2 for removal and loading, respectively.

Figure 1

Response distribution of S/NFL ratios – removal.

Figure 1

Response distribution of S/NFL ratios – removal.

Figure 2

Response distribution of S/NFL ratios – loading.

Figure 2

Response distribution of S/NFL ratios – loading.

Optimum operating conditions for removal and loading can be determined from Figures 1 and 2. The optimum working conditions for maximum removal are: c1; S1; t3; T3; which means concentration of 3 mmol dm−3, NaX as sorbent, 50 min for contact time and temperature of 35 °C, while the optimum conditions for maximum loading are: c3; S1; t3; T3; which means concentration of 13 mmol dm−3, NaX as sorbent, 50 min for contact time and temperature of 35 °C. In conclusion, the same temperature, contact time and sorbent provide the best results for removal and loading, while the optimum conditions differ only in concentration. The best removal is achieved for the lowest concentration and the best loading for the highest concentration used.

S/NFL ratios of each controllable factor along with the degree of freedom and rank as well as range, sum of squares, mean square, and percentage of contribution were calculated and presented in Tables 6 and 7 for removal and loading, respectively.

Table 6

Response table for SNFL ratios and contribution of each factor – removal

FactorcStT
Level 1 39.942 38.315 34.937 34.915 
Level 2 35.825 35.530 36.209 36.285 
Level 3 32.355 34.277 36.977 36.922 
Range 7.587 4.038 2.040 2.007 
DoFF 
SSF 86.594 20.618 6.374 6.293 
MSF 43.297 12.809 3.187 3.146 
pCF (%) 69.33 20.53 5.11 5.04 
Rank 
FactorcStT
Level 1 39.942 38.315 34.937 34.915 
Level 2 35.825 35.530 36.209 36.285 
Level 3 32.355 34.277 36.977 36.922 
Range 7.587 4.038 2.040 2.007 
DoFF 
SSF 86.594 20.618 6.374 6.293 
MSF 43.297 12.809 3.187 3.146 
pCF (%) 69.33 20.53 5.11 5.04 
Rank 
Table 7

Response table for SNFL ratios and contribution of each factor – loading

FactorcStT
Level 1 −3.698 1.314 −2.124 −2.085 
Level 2 0.105 −1.530 −0.794 −0.718 
Level 3 0.649 −2.728 −0.026 −0.141 
Range 4.347 4.042 2.098 1.944 
DoFF 
SSF 33.645 25.873 6.762 5.985 
MSF 16.822 12.936 3.381 2.99 
pCF (%) 46.56 35.80 9.36 8.28 
Rank 
FactorcStT
Level 1 −3.698 1.314 −2.124 −2.085 
Level 2 0.105 −1.530 −0.794 −0.718 
Level 3 0.649 −2.728 −0.026 −0.141 
Range 4.347 4.042 2.098 1.944 
DoFF 
SSF 33.645 25.873 6.762 5.985 
MSF 16.822 12.936 3.381 2.99 
pCF (%) 46.56 35.80 9.36 8.28 
Rank 

The range parameter was used to determine the impact of the controllable factors on processes examined. It was calculated as the difference between highest and lowest SNFL ratio of each controllable factor. Considering the obtained range, every controllable factor was associated with rank. The largest range implied the most effective factor and that controllable factor has a rank 1 so it should be utilized first. In this study concentration was found to be the most effective factor and it is followed by sorbent type, time and temperature.

The data for SSF, MSF and pCF in Tables 6 and 7 yield rankings that are totally consistent with those of S/NLF range from the same tables. Concentration is, again, the most influential process parameter and then sorbent type, time and temperature. But as can be seen from Tables 4 and 5, this influence is not equally effective. For example, for the first level concentration (3 mmol dm−3) average removals are 99.859, 98.233 and 99.947% so the average of them all is 99.246%; for third level concentration (13 mmol dm−3) these values are: 61.738, 38.115 and 30.331%; average 43.395%. It follows that removal efficiency is more than two times higher at the first level than at the third. For loading, effect of concentration is less pronounced. The average loading values for first level are: 0.657, 0.646 and 0.657 mmol g−1, leading to average of 0.653 mmol g−1. For the third level loadings are: 1.615, 0.997 and 0.793 mmol g−1, average 1.135 mmol g−1. This difference is lower than two times showing that concentration is more significant for removal than for loading although it is the most significant parameter for both. In the case of sorbent the effect is reversed, i.e. the sorbent type is more significant for loading than for removal. Time and temperature had approximately the same effect on both removal and loading.

Equilibrium studies

According to Table 8 and Figure 3, the results obtained show a good agreement of all isotherm models with the experimental data for the sorption of Cu ions on all sorbents used.

Table 8

Sorption isotherms and statistical comparison values

Isotherm modelParameterPure fly ash-based geopolymerModified fly ash-based geopolymerZeolite NaX
 qe (mmol g−11.169 1.186 1.695 
Jovanovic kJ 2.432 1.702 1.034 
qmax (mmol g−11.150 1.157 1.650 
r2 0.993 0.992 0.972 
SSE 0.002 0.480 0.483 
Khan kK 2.531 57.334 1.946 
qmax (mmol g−11.439 1.068 1.772 
nK 1.052 0.988 1.008 
r2 0.998 0.994 0.975 
SSE 4.248 × 10−4 0.480 0.480 
Baudu kB 2.301 3.597 2.611 
qmax (mmol g−11.382 1.405 1.806 
−4.183 × 10−4 −0.777 −1.000 
−0.055 −0.018 0.091 
r2 0.998 0.994 0.963 
SSE 4.332 × 10−4 8.771 × 10−4 0.490 
Fritz-Schlünder qmax (mmol g−11.147 1.692 1.752 
K1 0.819 2.025 1.691 
K2 0.084 2.123 1.534 
m1 0.110 0.171 0.819 
m2 0.000 0.204 0.849 
r2 0.728 0.994 0.975 
SSE 0.063 9.219 × 10−4 0.480 
Isotherm modelParameterPure fly ash-based geopolymerModified fly ash-based geopolymerZeolite NaX
 qe (mmol g−11.169 1.186 1.695 
Jovanovic kJ 2.432 1.702 1.034 
qmax (mmol g−11.150 1.157 1.650 
r2 0.993 0.992 0.972 
SSE 0.002 0.480 0.483 
Khan kK 2.531 57.334 1.946 
qmax (mmol g−11.439 1.068 1.772 
nK 1.052 0.988 1.008 
r2 0.998 0.994 0.975 
SSE 4.248 × 10−4 0.480 0.480 
Baudu kB 2.301 3.597 2.611 
qmax (mmol g−11.382 1.405 1.806 
−4.183 × 10−4 −0.777 −1.000 
−0.055 −0.018 0.091 
r2 0.998 0.994 0.963 
SSE 4.332 × 10−4 8.771 × 10−4 0.490 
Fritz-Schlünder qmax (mmol g−11.147 1.692 1.752 
K1 0.819 2.025 1.691 
K2 0.084 2.123 1.534 
m1 0.110 0.171 0.819 
m2 0.000 0.204 0.849 
r2 0.728 0.994 0.975 
SSE 0.063 9.219 × 10−4 0.480 
Figure 3

Comparison of experimentally obtained and calculated qe values of selected isotherm models for: (a) pure geopolymer, (b) modified geopolymer, and (c) zeolite NaX.

Figure 3

Comparison of experimentally obtained and calculated qe values of selected isotherm models for: (a) pure geopolymer, (b) modified geopolymer, and (c) zeolite NaX.

When the pure geopolymer was used as sorbent (Figure 3(a)), Jovanovic, Khan, and Baudu isotherm models were in a good agreement with experimental data, while the Fritz-Schlünder isotherm model slightly deviated from experimental data. Calculated statistical parameters r2 = 0.728 and SSE = 0.063 (Table 8) confirmed these findings. Taking into account all the values of statistical parameters, it can be concluded that the Baudu isotherm model (r2 = 0.998 and SSE = 4.332 × 10−4) is the most useful in predicting the equilibrium of Cu sorption on pure geopolymer for selected experimental conditions.

Figure 3(b) indicates that all of the isotherm models used are in a good agreement with experimental data obtained for the sorption of Cu ions on modified geopolymer. High values of r2 and low SSE values (Table 8) confirmed these findings.

The goodness of fit of the selected isotherm models to explain the equilibrium data for the sorption of Cu ions on zeolite NaX (Figure 3(c)) was in the order: Fritz-Schlünder > Khan > Jovanovic > Baudu. All models showed slightly lower r2 values and slightly higher SSE values (Table 8) unlike the sorption of the selected heavy metal on geopolymer sorbents.

The Jovanovic isotherm model is keeping the same assumptions as that of the Langmuir isotherm, also considering the surface binding vibrations of the adsorbed species. The Jovanovic isotherm represents another approximation for monolayer localized adsorption without lateral interactions and it reduces to the Langmuir isotherm at high concentrations but does not obey Henry's law (Sivarajasekar & Baskar 2014). The Khan isotherm model is a generalized model suggested for pure solutions, having a unique characteristic of covering both extremes of the Langmuir and Freundlich isotherms. When nK is equal to unity, the Khan isotherm model reduces to the Langmuir isotherm model (Rangabhashiyam et al. 2014). The Baudu isotherm model is a transformed Langmuir model which is only valid to the range of (1 + x + y) < 1 and (1 + x) < 1 (Sivarajasekar & Baskar 2014). The Fritz-Schlünder isotherm model is a five-parameter empirical expression which could be applied to a wide range of equilibrium data and is valid only in the range of m1 ≤ 1 and m2 ≤ 1. For m1 = m2 = 1, the Fritz-Schlünder isotherm model reduces to the Langmuir isotherm model (Rangabhashiyam et al. 2014; Sivarajasekar & Baskar 2014; Saadi et al. 2015).

According to the data presented in Table 8 a reversible monolayer adsorption on the homogeneous geopolymer surface occurred (Rangabhashiyam et al. 2014; Sivarajasekar & Baskar 2014; Saadi et al. 2015). For zeolite NaX, the Fritz-Schlünder isotherm model showed the best fit with values m1 ≠ m2 ≠ 1 due to the heterogeneity of the zeolite surface (Kyziol-Komosińska et al. 2015).

The analysis of the remaining Cu solution in a batch reactor shows that the removal of Cu ions from aqueous solutions is a very complex process consisting of ion exchange, adsorption and sorbent degradation (Mužek et al. 2016b). The total quantity of ingoing Cu ions is lower than the total quantity of outgoing cations (Na+, K+, Ca2+, Mg2+, and Pb2+), which is most prevalent for the pure geopolymer, and the lowest for modified geopolymer. These findings were confirmed by detecting and measuring the concentrations of Mg, Ca, K, and Pb ions (Table 9). The most interesting fact is that Pb ions from the modified geopolymer did not leach out from the structure in high concentration level. The small concentrations of Pb ions detected could be due to the degradation of the geopolymer structure (Mužek et al. 2016b) or due to the ion exchange with Cu ions.

Table 9

Concentration of ions in equilibrium

Concentration (mmol dm−3)
Initial Cu solutionCu retainedCu sorbedNaH0HeqKCaMgPb
Pure geopolymer 
3.458 0.323 3.135 9.602 5.248 × 10−6 3.388 × 10−9 0.144 <LOD <LOD 
8.034 2.439 5.595 10.787 1.047 × 10−5 2.089 × 10−6 0.422 0.622 0.087 
13.627 7.782 5.845 12.131 1.698 × 10−5 3.802 × 10−6 0.550 0.704 0.067 
18.034 12.189 5.845 12.206 2.239 × 10−5 4.571 × 10−6 0.594 1.111 0.056 
23.881 18.201 5.680 11.923 2.692 × 10−5 5.754 × 10−6 0.588 0.542 0.054 
29.983 24.218 5.783 12.772 3.467 × 10−5 8.318 × 10−6 0.696 0.735 <LOD 
Modified geopolymer 
3.458 0.000 3.458 7.634 5.248 × 10−6 1.698 × 10−8 0.057 <LOD 0.008 4.826 × 10−6 
8.034 2.356 5.678 10.702 1.047 × 10−5 3.311 × 10−6 0.356 0.390 0.061 5.019 × 10−6 
13.627 7.949 5.678 10.994 1.698 × 10−5 5.623 × 10−6 0.419 0.365 0.036 1.974 × 10−5 
18.034 12.271 5.763 11.048 2.239 × 10−5 7.079 × 10−6 0.439 0.346 <LOD 1.014 × 10−5 
23.881 18.119 5.762 10.952 2.692 × 10−5 9.333 × 10−6 0.457 0.225 0.012 1.308 × 10−5 
29.983 24.051 5.932 11.138 5.248 × 10−6 1.230 × 10−5 0.484 0.234 <LOD 2.722 × 10−5 
Zeolite NaX 
3.458 0.000 3.458 7.981 5.248 × 10−6 4.786 × 10−9 
8.034 1.508 6.526 14.574 1.047 × 10−5 1.514 × 10−6 
13.627 5.576 8.051 15.703 1.698 × 10−5 3.467 × 10−6 
18.034 9.898 8.136 17.858 2.239 × 10−5 4.467 × 10−6 
23.881 15.576 8.305 18.364 2.692 × 10−5 5.888 × 10−6 
29.983 21.508 8.475 18.596 3.467 × 10−5 7.413 × 10−6 
Concentration (mmol dm−3)
Initial Cu solutionCu retainedCu sorbedNaH0HeqKCaMgPb
Pure geopolymer 
3.458 0.323 3.135 9.602 5.248 × 10−6 3.388 × 10−9 0.144 <LOD <LOD 
8.034 2.439 5.595 10.787 1.047 × 10−5 2.089 × 10−6 0.422 0.622 0.087 
13.627 7.782 5.845 12.131 1.698 × 10−5 3.802 × 10−6 0.550 0.704 0.067 
18.034 12.189 5.845 12.206 2.239 × 10−5 4.571 × 10−6 0.594 1.111 0.056 
23.881 18.201 5.680 11.923 2.692 × 10−5 5.754 × 10−6 0.588 0.542 0.054 
29.983 24.218 5.783 12.772 3.467 × 10−5 8.318 × 10−6 0.696 0.735 <LOD 
Modified geopolymer 
3.458 0.000 3.458 7.634 5.248 × 10−6 1.698 × 10−8 0.057 <LOD 0.008 4.826 × 10−6 
8.034 2.356 5.678 10.702 1.047 × 10−5 3.311 × 10−6 0.356 0.390 0.061 5.019 × 10−6 
13.627 7.949 5.678 10.994 1.698 × 10−5 5.623 × 10−6 0.419 0.365 0.036 1.974 × 10−5 
18.034 12.271 5.763 11.048 2.239 × 10−5 7.079 × 10−6 0.439 0.346 <LOD 1.014 × 10−5 
23.881 18.119 5.762 10.952 2.692 × 10−5 9.333 × 10−6 0.457 0.225 0.012 1.308 × 10−5 
29.983 24.051 5.932 11.138 5.248 × 10−6 1.230 × 10−5 0.484 0.234 <LOD 2.722 × 10−5 
Zeolite NaX 
3.458 0.000 3.458 7.981 5.248 × 10−6 4.786 × 10−9 
8.034 1.508 6.526 14.574 1.047 × 10−5 1.514 × 10−6 
13.627 5.576 8.051 15.703 1.698 × 10−5 3.467 × 10−6 
18.034 9.898 8.136 17.858 2.239 × 10−5 4.467 × 10−6 
23.881 15.576 8.305 18.364 2.692 × 10−5 5.888 × 10−6 
29.983 21.508 8.475 18.596 3.467 × 10−5 7.413 × 10−6 

H0 – initial concentration of H ions; Heq – equilibrium concentration of H ions; LOD – limit of detection.

In comparison to the clinoptilolite (0.2 mmol g–1) (Perić et al. 2004) zeolite X has higher sorption capacity (1.695 mmol g–1) but lower than zeolite A (2.25 mmol g–1) (Biškup & Subotić 2004).

In comparison to other geopolymer-like sorbents, such as zeolite tuff-based geopolymer powder (0.52 mmol g–1) (El-Eswed et al. 2012); porous geopolymeric sphere (0.83 mmol g–1) (Ge et al. 2015); zeolite-based geopolymer powder (0.70 mmol g–1) (Andrejkovičová et al. 2016), metakaolin-based geopolymer powder (0.77 mmol g–1) (Cheng et al. 2012), and cork-based geopolymer powder (0.88 mmol g–1) (Sudagar et al. 2018), pure and modified fly ash-based geopolymer showed higher sorption capacity (1.169 mmol g–1 and 1.186 mmol g–1 respectively) but lower than fly ash-based geopolymer powder (1.52 mmol g–1) (Al-Harahsheh et al. 2015).

CONCLUSION

The obtained results provide information about not only optimal values of process parameters used, for both removal and loading, but also the significance of all parameters used on processes examined. It could be concluded that optimal parameter values for removal are: 3 mmol dm−3; NaX; 50 min; 35 °C; while optimum parameters values for maximum loading are: 13 mmol dm−3; NaX; 50 min; 35 °C. Also it is found that the most influential factor, in descending order, for both removal and loading is concentration, and then the sorbent, time and temperature.

According to the results of equilibrium sorption studies the Baudu isotherm model (r2 = 0.998 and SSE = 4.332 × 10−4) is the most useful in predicting the equilibrium of Cu sorption on pure geopolymer for selected experimental conditions. The results obtained indicate that all of the used isotherm models are in a good agreement with experimental data obtained for the sorption of Cu ions on modified geopolymer.

The goodness of fit of the selected isotherm models to explain the equilibrium data for the sorption of Cu ions on zeolite NaX was in the order: Fritz-Schlünder > Khan > Jovanovic > Baudu.

The analysis of the remaining solution in a batch reactor shows that the total quantity of ingoing Cu ions is lower than the total quantity of outgoing cations (Na+, K+, Ca2+, Mg2+, and Pb2+) for all the sorbents used. The analysis also shows that Pb ions from the modified geopolymer leach out from the structure in a very low concentration. These findings reveal the double application of the modified geopolymers – during the synthesis a heavy metal ion can be incorporated within the structure of geopolymer resulting in a low concentration of heavy metal that will leach out from the structure during the sorption of another heavy metal ion from aqueous solutions. Also, modified geopolymer shows greater sorption capacity as opposed to pure geopolymer.

ACKNOWLEDGEMENTS

The present study has been financially supported by Croatian Science Foundation under the Project HETMIX (Grant Number IP-11-2013-8959). The authors declare that they have no conflict of interest.

REFERENCES

REFERENCES
Abdel-Rahman
L. H.
,
Abu-Dief
A. M.
,
Abd-El Sayed
M. A.
&
Megalea Zikry
M.
2016
Nano sized Moringa oleifera an effective strategy for Pb(II) ions removal from aqueous solution
.
Chemistry and Materials Research
8
(
4
),
8
22
.
Al-Harahsheh
M. S.
,
Al Zboon
K.
,
Al-Makhadmeh
L.
,
Hararah
M.
&
Mahasneh
M.
2015
Fly ash based geopolymer for heavy metal removal: a case study on copper removal
.
Journal of Environmental Chemical Engineering
3
(
3
),
1669
1677
.
Alshaaer
M.
,
Zaharaki
D.
&
Komnitsas
K.
2015
Microstructural characteristics and adsorption potential of a zeolitic tuff–metakaolin geopolymer
.
Desalination and Water Treatment
56
(
2
),
338
345
.
Al-Zboon
K.
,
Al-Harahsheh
M. S.
&
Hani
F. B.
2011
Fly ash-based geopolymer for Pb removal from aqueous solution
.
Journal of Hazardous Materials
188
(
1–3
),
414
421
.
Andrejkovičová
S.
,
Sudagar
A.
,
Rocha
J.
,
Patinha
C.
,
Hajjaji
W.
,
Ferreira da Silva
E.
,
Velosa
A.
&
Rocha
F.
2016
The effect of natural zeolite on microstructure, mechanical and heavy metals adsorption properties of metakaolin based geopolymers
.
Applied Clay Science
126
,
141
152
.
Azimi
E. A.
,
Al Bakri Abdullah
M. M.
,
Liew Yun
M.
,
Heah Cheng
Y.
,
Hussin
K.
&
Aziz
I. H.
2015
Review of geopolymer materials for thermal insulating applications
.
Key Engineering Materials
660
,
17
22
.
Cai
L.-M.
,
Wang
Q.-S.
,
Luo
J.
,
Chen
L.-G.
,
Zhu
R.-L.
,
Wang
S.
&
Tang
C.-H.
2019
Heavy metal contamination and health risk assessment for children near a large Cu-smelter in central China
.
Science of the Total Environment
650
,
725
733
.
Chen
Y.-L.
,
Tong
Y.-Y.
,
Pan
R.-W.
&
Tang
J.
2013
The research on adsorption behaviors and mechanisms of geopolymers on Sr2+, Co2+ and Cs+
.
Advanced Materials Research
704
,
313
318
.
Cheng
T. W.
,
Lee
M. L.
,
Ko
M. S.
,
Ueng
T. H.
&
Yang
S. F.
2012
The heavy metal adsorption characteristics on metakaolin-based geopolymer
.
Applied Clay Science
56
,
90
96
.
El-Eswed
B.
,
Alshaaer
M.
,
Yousef
R. I.
,
Hamadneh
I.
&
Khalili
F.
2012
Adsorption of Cu(II), Ni(II), Zn(II), Cd(II) and Pb(II) onto kaolin/zeolite based-geopolymers
.
Advances in Materials Physics and Chemistry
2
(
4B
),
119
125
.
Foo
K. Y.
&
Hameed
B. H.
2010
Insights into the modeling of adsorption isotherm systems
.
Chemical Engineering Journal
156
(
1
),
2
10
.
Ge
Y.
,
Cui
X.
,
Kong
Y.
,
Li
Z.
,
He
Y.
&
Zhou
Q.
2015
Porous geopolymeric spheres for removal of Cu(II) from aqueous solution: synthesis and evaluation
.
Journal of Hazardous Materials
283
,
244
251
.
Kara
I.
,
Tunc
D.
,
Sayin
F.
&
Akar
S. T.
2018
Study on the performance of metakaolin based geopolymer for Mn(II) and Co(II) removal
.
Applied Clay Science
161
,
184
193
.
Kragović
M.
,
Daković
A.
,
Sekulić
Ž
,
Trgo
M.
,
Ugrina
M.
,
Perić
J.
&
Gatta
G. D.
2012
Removal of lead from aqueous solutions by using the natural and Fe(III)-modified zeolite
.
Applied Surface Science
258
,
3667
3673
.
Kyziol-Komosińska
J.
,
Rosik-Dulewska
C.
,
Franus
M.
,
Antoszczyszyn-Szpicka
P.
,
Czupiol
J.
&
Krzyżewska
I.
2015
Sorption capacities of natural and synthetic zeolites for Cu(II) ions
.
Polish Journal of Environmental Studies
24
(
3
),
1111
1123
.
Li
L.
,
Wang
S.
&
Zhu
Z.
2006
Geopolymeric adsorbents from fly ash for dye removal from aqueous solution
.
Journal of Colloid and Interface Science
300
(
1
),
52
59
.
López
F. J.
,
Sugita
S.
,
Tagaya
M.
&
Kobayashi
T.
2014
Metakaolin-based geopolymers for targeted adsorbents to heavy metal ion separation
.
Journal of Materials Science and Chemical Engineering
2
(
7
),
16
27
.
Luukkonen
T.
,
Runtti
H.
,
Niskanen
M.
,
Tolonen
E. T.
,
Sarkkinen
M.
,
Kemppainen
K.
,
Rämö
J.
&
Lassi
U.
2016a
 Simultaneous removal of Ni(II), As(III), and Sb(III) from spiked mine effluent with metakaolin and blast-furnace-slag geopolymers
.
Journal of Environmental Management
166
,
579
588
.
Luukkonen
T.
,
Sarkkinen
M.
,
Kemppainen
K.
,
Rämö
J.
&
Lassi
U.
2016b
 Metakaolin geopolymer characterization and application for ammonium removal from model solutions and landfill leachate
.
Applied Clay Science
119
,
266
276
.
Luukkonen
T.
,
Tolonen
E. T.
,
Runtti
H.
,
Kemppainen
K.
,
Perämäki
P.
,
Rämö
J.
&
Lassi
U.
2017
Optimization of the metakaolin geopolymer preparation for maximized ammonium adsorption capacity
.
Journal of Materials Science
52
(
16
),
9363
9376
.
Malik
A.
2004
Metal bioremediation through growing cells
.
Environment International
30
(
2
),
261
278
.
Martin
A.
,
Pastor
J. Y.
,
Palomo
A.
&
Fernandez Jimenez
A.
2015
Mechanical behaviour at high temperature of alkali-activated aluminosilicates (geopolymers)
.
Construction and Building Materials
93
,
1188
1196
.
Mužek
M. N.
,
Zelić
J.
&
Jozić
D.
2012
Microstructural characteristics of geopolymers based on alkali-activated fly ash
.
Chemical and Biochemical Engineering Quarterly
26
(
2
),
89
95
.
Mužek
M. N.
,
Svilović
S.
&
Zelić
J.
2013
Removal of copper ions in wastewater by geopolymers synthesized from fly ash
. In:
Proceedings of International Conference 14th Ružička Days ‘Today Science-Tomorrow Industry’
(
Jukić
A.
, ed.).
Faculty of Food Technology Osijek, Croatian Society of Chemical Engineers
,
Zagreb/Osijek, Croatia
, pp.
142
152
.
Mužek
M. N.
,
Svilović
S.
&
Zelić
J.
2014
Fly ash-based geopolymeric adsorbent for copper ion removal from wastewater
.
Desalination and Water Treatment
52
(
13–15
),
2519
2526
.
Mužek
M. N.
,
Svilović
S.
,
Ugrina
M.
&
Zelić
J.
2016a
 Removal of copper and cobalt ions by fly ash-based geopolymer from solutions-equilibrium study
.
Desalination and Water Treatment
57
(
23
),
10689
10699
.
Mužek
M. N.
,
Svilović
S.
&
Zelić
J.
2016b
 Kinetic studies of cobalt ion removal from aqueous solutions using fly ash-based geopolymer and zeolite NaX as sorbents
.
Separation Science and Technology
51
(
18
),
2868
2875
.
Perić
J.
,
Trgo
M.
&
Vukojević Medvidović
N.
2004
Removal of zinc, copper and lead by natural zeolite – a comparison of adsorption isotherms
.
Water Research
38
(
7
),
1893
1899
.
Rai
P. K.
,
Lee
S. S.
,
Zhang
M.
,
Tsang
Y. F.
&
Kim
K.-H.
2019
Heavy metals in food crops: health risks, fate, mechanisms, and management
.
Environment International
125
,
365
385
.
Rangabhashiyam
S.
,
Anu
N.
,
Giri Nandagopal
M. S.
&
Selvaraju
N.
2014
Relevance of isotherm models in biosorption of pollutants by agricultural byproducts
.
Journal of Environmental Chemical Engineering
2
(
1
),
398
414
.
Saadi
R.
,
Saadi
Z.
,
Fazaeli
R.
&
Fard
N. E.
2015
Monolayer and multilayer adsorption isotherm models for sorption from aqueous media
.
Korean Journal of Chemical Engineering
32
(
5
),
787
799
.
Siegel
F. R.
2002
Environmental Geochemistry of Potentially Toxic Metals
.
Springer
,
Berlin, Germany
.
Siju
K. H.
,
Rathod
P. P.
,
Goshwami
J. J.
,
Patel
T. M.
&
Rathod
G. R.
2013
Parametric optimization for performance of diesel engine using blends of karanja biodisel by Taguchi method
.
International Journal for Scientific Research & Development
1
(
3
),
750
754
.
Singh
B.
,
Ishwarya
G.
,
Gupta
M.
&
Bhattacharyya
S. K.
2015
Geopolymer concrete: a review of some recent developments
.
Construction and Building Materials
85
,
78
90
.
Sudagar
A.
,
Andrejkovičová
S.
,
Patinha
C.
,
Velosa
A.
,
McAdam
A.
,
Ferreira da Silva
E.
&
Rocha
F.
2018
A novel study on the influence of cork waste residue on metakaolin-zeolite based geopolymers
.
Applied Clay Science
152
,
196
210
.
Svilović
S.
,
Horvat
I.
,
Ćosić
M.
&
Stipišić
R.
2017
Investigation on the effects of process variables on copper exchange on NaX in a batch stirred reactor
.
Tehnički Glasnik
11
(
4
),
171
174
.
Tchounwou
P. B.
,
Yedjou
C. G.
,
Patlolla
A. K.
&
Sutton
D. J.
2012
Heavy metals toxicity and the environment
.
Experientia Supplementum
101
,
133
164
.
Temuujin
J.
,
Rickard
W.
,
Lee
M.
&
van Riessen
A.
2011
Preparation and thermal properties of fire resistant metakaolin-based geopolymer-type coatings
.
Journal of Non-Crystalline Solids
357
(
5
),
1399
1404
.
Wang
S.
,
Li
L.
&
Zhu
Z. H.
2007
Solid-state conversion of fly ash to effective adsorbents for Cu removal from wastewater
.
Journal of Hazardous Materials
139
(
2
),
254
259
.
Yan
D.
,
Chen
S.
,
Zeng
Q.
,
Xu
S.
&
Li
H.
2016
Correlating the elastic properties of metakaolin-based geopolymer with its composition
.
Materials & Design
95
,
306
318
.
Zhuang
X. Y.
,
Chen
L.
,
Komarneni
S.
,
Zhou
C. H.
,
Tong
D. S.
,
Yang
H. M.
,
Yu
W. H.
&
Wang
H.
2016
Fly ash-based geopolymer: clean production, properties and applications
.
Journal of Cleaner Production
125
,
253
267
.
Živica
V.
,
Balkovic
S.
&
Drabik
M.
2011
Properties of metakaolin geopolymer hardened paste prepared by high-pressure compaction
.
Construction and Building Materials
25
,
2206
2213
.