This article is an openaccess article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).
The adsorption isotherm, the adsorption kinetics, and the thermodynamic parameters of ammonium removal from aqueous solution by using clinoptilolite in aqueous solution was investigated in this study. Experimental data obtained from batch equilibrium tests have been analyzed by four twoparameter (Freundlich, Langmuir, Tempkin and DubininRadushkevich (DR)) and four threeparameter (RedlichPeterson (RP), Sips, Toth and Khan) isotherm models. DR and RP isotherms were the models that best fitted to experimental data over the other two and threeparameter models applied. The adsorption energy (E) from the DR isotherm was found to be approximately 7 kJ/mol for the ammoniumclinoptilolite system, thereby indicating that ammonium is adsorbed on clinoptilolite by physisorption. Kinetic parameters were determined by analyzing the
Discharging of wastewater streams containing high ammonium concentrations into the receiving body causes serious problems in the natural nutrient cycle between the living world and the soil, water, and atmosphere [
In the last three decades, an increasing number of investigations have been conducted on ammonium removal from wastewater by ion exchange because of the ammonium selectivity of clinoptilolite [
Clinoptilolite is an abundant natural zeolite found in igneous, sedimentary and metamorphic deposits in the form of aluminosilicate minerals with high cationexchange capacity. The adsorption capacity of clinoptilolite is significantly affected by physical and chemical pretreatment and loading or regeneration of clinoptilolite. The pretreatment of natural zeolites by acids, bases, surfactants,
Isotherms and kinetic results are valuable information to determine the suitability and effectiveness of the adsorption process [
Although many studies have been conducted on adsorption isotherm and kinetics, very few works [
The main objective of the present study is to examine ammonium removal by clinoptilolite, which was initially pretreated with aqueous sodium chloride solution, and to analyze the equilibrium modeling by two and threeparameter adsorption isotherms, the kinetic modeling by
Clinoptilolite used in the experiments was obtained from Balıkesir in the Northwestern part of Turkey. The chemical properties of the clinoptilolite can be found in our previous study [
Synthetic samples were prepared to give NH_{3}N concentrations of 30, 60, 100, 160 and 250 mg/L by adding required NH_{4}Cl salt to distilled water for both isotherm and kinetic studies. For kinetic studies, samples of 5 g were equilibrated with 500 mL ammonium nitrogen solution at 10, 25, and 40 °C for 100 min. Samples were periodically taken for measurement of aqueous phase of ammonia concentrations. Batch mode adsorption isotherm studies were carried out in conical flasks containing 200 mL of the solutions and 2 g clinoptilolite at temperatures of 10, 25, and 40 °C. The flasks were placed in a magnetic stirrer and agitated for 4 h at a fixed agitation speed of 200 rpm. All experiments were carried out at an initial pH of 4.5 where clinoptilolite ion exchange occurs conveniently when the pH is between 4 and 8 in which range ammonium is at ionized form [
The amount of ammonium nitrogen in the solid phase, Q (mg/g), was calculated by using the following equation:
where C_{0} and C_{t} are the initial and retained ammonium concentration (mg/L) in solution at time t, respectively;
The adsorption isotherm and adsorption kinetic parameters were determined by using nonlinear regression analysis. The nonlinear method is a better way to obtain the kinetic parameters than the linear method, and thus it should be primarily adopted to determine the kinetic parameters [
where the subscripts “exp” and “cal” denote the experimental and calculated values of
In order to quantitatively compare the applicability of isotherm and kinetic models for fitting the experimental data, nonlinear coefficient of determination (R^{2}) and average relative error (
Expression of twoparameter and threeparameter adsorption isotherm models.
Isotherm models  Expression *  Adjustable model parameters *  Constraints * 



Freundlich [ 


Langmuir [ 

  
DR [ 


Tempkin [ 

  


RedlichPeterson [ 

0 < 

Sips [ 

0 < 

Toth [ 

0 < 

Khan [ 

* where; Q_{e} is equilibrium solid phase concentration (mg/g) and C_{e} is equilibrium liquid phase concentration (mg/L) in all isotherm models. In all models,
Adsorption isotherms describe the relationship between the amount of adsorbed ion on adsorbent and the final ion concentration in the solution. Experimental results have been analyzed using four twoparameter isotherm models including Freundlich, Langmuir, Tempkin and DR; and four threeparameter adsorption isotherm models including RP, Sips, Toth and Khan isotherm models (
Comparison of adsorption isotherm models at 10 °C; (
The determined isotherm parameters obtained for all temperatures are shown in
Comparison of twoparameter and threeparameter isotherms for different temperatures.
Twoparameter isotherms  10 °C  25 °C  40 °C  Threeparameter isotherms  10 °C  25 °C  40 °C  





2.055  1.961  1.660 

2.387  2.217  0.863  

0.424  0.434  0.464 

0.610  0.585  0.084  
R² (nonlinear)  0.986  0.987  0.962 

0.709  0.701  0.905  
HYBRID  7.211  6.575  17.166  R² (nonlinear)  0.993  0.994  0.999  
ΔQ, (%)  8.093  7.692  12.487  HYBRID  27.832  25.424  14.187  

ΔQ, (%)  4.799  4.458  4.048  

0.056  0.052  0.045 



15.902  16.163  16.305 

33.075  33.905  17.338  
R² (nonlinear)  0.971  0.973  0.995 

0.058  0.054  0.051  
HYBRID  92.123  85.887  14.316 

0.551  0.560  0.913  
ΔQ, (%)  9.868  9.620  3.438  R² (nonlinear)  0.991  0.992  0.995  

HYBRID  39.054  35.520  18.599  

19.108  19.330  19.406  ΔQ, (%)  6.348  5.943  4.366  

7.024  7.268  7.379 


R² (nonlinear)  0.988  0.989  0.993 

64.023  67.410  18.309  
HYBRID  33.750  31.798  18.461 

1.540  1.631  11.042  
ΔQ, (%)  3.980  3.504  5.888 

0.256  0.258  0.798  

R² (nonlinear)  0.992  0.993  0.996  

0.736  0.673  0.469  HYBRID  34.297  31.120  16.590  

748.321  773.183  748.566  ΔQ, (%)  5.690  5.288  4.256  
R² (nonlinear)  0.982  0.981  0.999 


HYBRID  51.874  54.940  3.193 

0.180  0.161  0.145  
ΔQ, (%)  6.777  7.499  1.976 

545.242  548.264  545.194  

0.599  0.589  0.583  
R² (nonlinear)  0.988  0.989  0.969  
HYBRID  52.619  48.473  131.342  
ΔQ, (%)  8.141  7.727  13.157 
The magnitude of energy of adsorption (E) in the DR isotherm is around 1–16 kJ/mol and useful for the assessment of the adsorption mechanism. If this value is below 8 kJ/mol, physisorption is considered to occur. In the present case, the values of E slightly increased as the temperature increased from 10 to 40 °C and were found to be between 7.024 and 7.379 kJ/mol, which indicated physical adsorption.
Adsorption kinetics are required for selecting optimum operational conditions of water and wastewater treatment facilities for fullscale processes. The results obtained from the experiments at 10 °C were examined to describe the reaction kinetics according to the
Comparison of adsorption kinetic models at 10 °C.
Comparison of adsorption kinetic models at 10 °C.
C_{i} (initial conc.), mg/L  30  60  100  160  250 

Q_{e}, _{exp.} (mg/g)  2.716  5.248  8.054  10.734  14.504 


2.713  5.383  8.485  13.775  20.832  

1.162  1.446  1.961  3.419  4.162 

0.996  1.014  0.857  0.848  1.085 
0.173  0.123  0.231  0.220  0.174  
HIBRID  0.654  0.532  0.205  1.474  2.918 
Average Δq (%)  3.979  2.959  1.166  2.321  4.043 
R^{2}  0.989  0.996  0.997  0.989  0.987 


2.961  5.819  8.518  11.499  15.621  
0.228  0.138  0.236  0.175  0.141  

0.901  1.002  0.839  1.004  1.113 
HIBRID  0.892  0.607  0.176  1.624  3.154 
Average Δq (%)  4.469  3.329  1.144  3.086  5.023 
R^{2}  0.980  0.993  0.997  0.988  0.986 


2.771  5.236  8.095  10.992  14.700  

26.427  33.567  45.292  50.886  53.777 

1.280  18.790  35.663  59.037  93.223 

0.159  0.182  0.267  0.357  0.611 

0.009  0.050  0.059  0.048  0.051 
RF, %  95.38  64.11  55.95  46.29  36.58 
SF, %  4.62  35.89  44.05  53.71  63.42 
HIBRID  0.780  1.208  0.216  1.142  2.554 
Average Δq (%)  4.682  3.903  1.233  2.516  2.841 
R^{2}  0.990  0.992  0.998  0.994  0.991 
It is thought that instead of assuming order of the reaction as 1 or 2, the direct calculation of rate constant and order of the adsorption reaction is a more appropriate method [
where
The determined model parameters are listed in
Modified secondorder equation can be obtained from the
The modified secondorder reaction constants
The model describes the adsorption process with respect to both chemical and mathematical points of view [
where D_{1} and D_{2} are the amount of rapidly and slowly adsorbed fraction of ammonium (mg/l), respectively, and K_{D1} and K_{D2} are rapid and slow rate constants (min^{–1}). It should be noted that the sum of D_{1}/m_{ads} and D_{2}/m_{ads} has the same physical meaning as the calculated value of Q_{e}, and K_{D1} is greater than K_{D2}. Rapidly and slowly adsorbed fractions (%), RF and SF, can be calculated as:
and
respectively.
Model parameters for the slow and rapid steps and statistical comparison parameters obtained from experimental data are presented in
For all kinetic models, calculated
Thermodynamic parameters were evaluated by considering the thermodynamic equilibrium constants. The standard free energy were calculated using the following equation:
where
For the Tempkin and Langmuir isotherms, the values of Tempkin constant (
where
The other thermodynamic parameters,
The values of change in enthalpy (∆H°) and entropy (∆S°) were calculated from the slope and intercept of the plot of ln
Plots of ln (Q
Plot of ln
The results of change in standard free energy, enthalpy and entropy are given in
The positive values of ∆G° imply that the adsorption of ammonium on clinoptilolite is not spontaneous. ∆G° value increased with the temperature, indicating that the spontaneous nature of adsorption is inversely proportional to the temperature. Since the value of standard enthalpy change, ∆H°, is negative, the process is exothermic; therefore, an increase in the temperature leads to a lower adsorption of ammonium at equilibrium, and physical by nature and involves weak forces of attraction. The negative values of ∆S° suggest that the system exhibits random behavior.
Comparison of thermodynamic parameters for adsorption of ammonia on Clinoptilolite.
Temperature, °C  Slope × 10^{−3}  Intercept  R^{2} (linear)  



10 °C  1.596  −1.10  
25 °C  1.454  1.34  −4.20  0.877  −0.93  −11.11  −34.95 
40 °C  1.011  −0.03  


10 °C  0.736  0.72  
25 °C  0.673  1.31  −4.90  0.873  0.98  −10.92  −40.73 
40 °C  0.469  1.97  


10 °C  0.056  6.77  
25 °C  0.052  0.65  −5.18  0.977  7.35  −5.43  −43.03 
40 °C  0.045  8.07 
* Thermodynamic equilibrium constant.
The adsorption of ammonium on clinoptilolite was evaluated as a function of two and threeparameter isotherms, adsorption kinetics and thermodynamic aspect. Generally, equilibrium data fitted better in threeparameter isotherm models than twoparameter isotherms models. DR and RP isotherms effectively described the experimental data for twoparameter and threeparameter isotherm models, respectively. Adsorption energy for ammonium–clinoptilolite system was found to be approximately 7 kJ/mol, which lies within the range of 1–8 kJ/mol for the physisorption processes, indicating that ammonium is adsorbed on clinoptilolite predominantly by physisorption process.
The kinetics of adsorption were found to conform to all kinetics studied with a good correlation. The best model that described the kinetic data was the
The author declares no conflict of interest.