Ind. Eng. Chem. Res. 2008, 47, 5615–5622
5615
Isotherms, Isosteres, and Enthalpy of Adsorption of 1,2-Dichloroethane from Aqueous Solution onto Activated Carbons Robert Pełech† Institute of Chemical Organic Technology, Szczecin UniVersity of Technology, Pulaskiego 10, 70-322 Szczecin, Poland
In this work the adsorption of 1,2-dichloroethane from aqueous solutions onto DTO, AG5, and WD-ekstra activated carbons has been studied. The adsorption isotherms at the range of temperatures 298-343 K have been described by the Freundlich and Dubinin-Astakhov equations. The characteristic curves of the adsorption process have been determined. It was found out that the temperature dependence of the adsorption equilibrium was described well by the Dubinin-Astakhov model. The isosteric enthalpy of the adsorption was assigned. To describe the dependence of the adsorption isosteric enthalpy on the filling degree, the Do model and the combinations of the van’t Hoff and Dubinin-Astakhov equations in a form of the logarithmic equation were used. Introduction Wastewaters containing 1,2-dichloroethane (DCE) are discharged in a large amount from the production plant of vinyl chloride.1 Because of their significant resistance to biodegradation and their carcinogenic and mutagenic properties, their removal is purposeful.1 So far there have been practically no works connected with DCE removal from aqueous solution. A high degree of purification can be achieved by using the adsorption method on the activated carbons.2–4 The adsorption processes are usually carried out on an industrial scale in the columns packed with adsorbents or bath reactors. The fundamental information concerning the process is achieved experimentally by the investigation of the systems with a constant volume of a treated liquid.5 Application of adsorption from solutions on solid adsorbents have been considerably developed recently and become very important in many fields such as in purification processes and water treatment processes. Unfortunately, the theory of adsorption from solution has not been well developed yet. The equilibrium adsorption quantities from solution is generally represented by the Freundlich equation6,7 a ) kCen
(1)
where a is the equilibrium adsorption quantities (mmol/g), Ce is the equilibrium concentration (mmol/dm3), and k, n are the constants of the Freundlich equation. The adsorption from the flux with the constant temperature is not carried out in practice. Very often the temperature is changed along with the season, the plant production capacity, or as a result of out-of-control events. Considering the temperature effect on the adsorption process, it is possible to eliminate unexpected reduction of the adsorption amount of the sorbents and to limit uncontrolled elutes of contaminates, for example, by quickly changing the adsorption bed. That is why the knowledge of the adsorption equilibrium at different temperatures is an important element of designing the adsorption systems. The Freundlich equation is sometimes used in a modified form, on the strength of the potential theory of Polanyi’s adsorption.8,9 †
E-mail:
[email protected]. Fax: +48 91 449 43 65.
a)k′
() Ce Cs
n
(2)
where k′ is the constant of modified Freundlich equation and Cs is the saturation concentration (mmol/dm3). The equation above is a peculiar case of Dubinin-Astakhov equation (D-A):10,11
( ( EA ) )
a ) a0 exp -
nDA
(3)
where A is the adsorption potential (Gibbs free energy): A ) -∆G ) RT ln
Cs Ce
(4)
However, E is the characteristic adsorption energy (J/mol). Parameter E is recognized in Dubinin’s theory as independent of temperature, although in practice many experimental systems have shown a weak dependence of E on temperature. For nDA ) 1 eq 3 converts to form eq 2. Then a0 ) k′ is the limiting of equilibrium adsorption quantities (mmol/g). The D-A model is convenient to set the adsorption equilibrium at different temperatures. The relation between adsorption potential and the volume of adsorbed adsorbate may be presented by the characteristic curve: W ) f(A) W)
(5)
a F
(6)
where W is the volume of adsorbed adsorbate (cm3/g) and F is the density of the adsorbate (mol/cm3). On the basis of the adsorption characteristic curve for the specific relation adsorbate-adsorbent, the adsorption isotherm for any temperature may be determined. Table 1. Parameters of Using Activated Carbons adsorbent
DTO
AG-5
WD-ekstra
bulk density, Fb (g/dm ) apparent density, Fa (g/dm3) specific surface area (m2/g) porosity pore volume (cm3/g) micropore volume 99%) used in these studies was taken from Fluka, A.G. The DCE solubility was calculated from the relation Cs ) 0.02053T2 - 11.979T + 1831.6 (mmol/ dm3), determined on the basis of the data from ref 12; the molar density was calculated from the relation F ) (-0.0142T + 16.835) × 10-3 (mol/cm3), determined on the basis of the data from refs 13 and 14.
Ind. Eng. Chem. Res., Vol. 47, No. 15, 2008 5617
Figure 3. Isotherms adsorption of DCE onto AG5 activated carbon. Table 2. Parameters of Freundlich and D-A Equations and the Surface Covering for the Limiting Equilibrium Adsorption Quantities T (K)
k
n
298 313 323 343
0.515 0.435 0.379 0.290
0.565 0.588 0.630 0.673
298 313 323 343
0.634 0.523 0.436 0.327
0.507 0.541 0.564 0.598
298 313 323 343
0.672 0.580 0.481 0.356
0.545 0.589 0.590 0.630
E (J/mol) DTO 4384 4424 4305 4237 AG5 4882 4809 4763 4830 WD-ekstra 4543 4421 4549 4527
a0 (cm3/g)
S0 (m2/g)
0.50 0.52 0.56 0.67
992 1010 1097 1296
0.48 0.49 0.49 0.50
946 964 955 971
0.60 0.68 0.61 0.67
1185 1325 1191 1286
Figure 4. Isotherms adsorption of DCE onto WD-ekstra activated carbon.
Analytical Method. The concentration of DCE was determined chromatographically on a Thermoquest GC 8000Top instrument with a flame ionization detector (FID) by direct injections of solution. The volume of analytical sample was 2.0 µL. The determination were conducted with an external standard method and following analytical conditions: capillary column (J&W DB1) 27 m × 0.53 mm × 1.5 µm (100% dimethylpolysiloxane); carrier gas He 6 cm3/min; detector temperature 300 °C; injector temperature 150 °C (split 1:6); oven temperature program 40 °C (7 min) to 180 °C at 15 °C/min. The detection limit was >3 µmol/dm3. The average error of this is equal to 3.6%. Experimental Method. The adsorption isotherms of DCE were studied by the batch adsorption method. The sample of adsorbent (0.050-2.500 g) and 245.0 g of distilled water were put in eight glass flasks of volume 250 cm3. At the end, DCE in a form of methanol solution with the concentration of 48
5618 Ind. Eng. Chem. Res., Vol. 47, No. 15, 2008
Figure 5. Characteristic curve of DCE adsorption onto DTO activated carbon.
g/dm3 was added to the adsorbent suspension, according to the procedure presented in the works.6,15 Flasks with open-top screw caps and PTFE-silicone septa were used. Next, the flasks were placed in thermostatic water batch equipped with eight magnetic stirrers. They were intensively mixed for 24 h at a given temperature. After that, the mixers were turned off and the systems were left for 2 h for fining. After the fining, the analysis checking the contents of DCE in the solution was carried out. The analyte sample was drawn by PTFE-silicone septa. When the analyses were finished, a new value of temperature on a thermostat was introduced and the mixers were activated; the systems were left again for 24 h in order to establish equilibrium. The activity was repeated for the next temperatures. The measurements were started from the lowest temperature of 25 °C and finished at the highest of 70 °C. At the same time the
Figure 6. Characteristic curve of DCE adsorption onto AG5 activated carbon.
DCE concentration in a flask without any adsorbent was measuredscheck test. Results and Discussion Isotherms Analysis of DCE Adsorption. Figures 2–4 present the obtained adsorption isotherms. All the courses were described by the Freundlich equation and, what comes from eqs 1–3, by the D-A equation with the coefficient nDA ) 1. In Table 2 the constants of the Freundlich equation were compiled. Figures 5–7 show the characteristic curves of DCE adsorption on the tested adsorbents. It was found that the characteristic curves for particular temperatures overlap. That indicates that the DCE adsorption from aqueous solution onto the studied activated carbons proceeds according to the D-A model. The Freundlich equation coefficients, the adsorption characteristic energy, and the
Ind. Eng. Chem. Res., Vol. 47, No. 15, 2008 5619
Figure 7. Characteristic curve of DCE adsorption onto WD-ekstra activated carbon.
Figure 8. Dependence n constant Freundlich equation on temperature.
adsorption limiting volume for D-A model are presented in Table 2. Figure 8 presents the dependence of n constant of the Freundlich equation on the temperature. It was found that it is linearly depended on the temperature, according to eq 8. There was not found any significant temperature effect on the adsorption characteristic energy and the limiting volume. Dubinin and Stoeckli showed that the adsorption characteristic energy of the activated carbons is defined by the average micropore size.16,17 Values obtained of the adsorption characteristic energy indicate that the tested activated carbons possess the pores with the average size of 2 nm.18 The limiting surface coverage S0 for the limiting equilibrium adsorption quantities was determined, too. S0 ) a0NSp
(9)
Sp ) (FMN)-2/3
(10)
where FM is molar density (mol/m3), N is Avogadro’s number (6.022 × 1023 mol-1), and Sp is the cross-sectional area of DCE particle (m2). Values of the degree of coverage determined for the maximum saturation of the adsorbent were correlated in Table 2. For DTO and AG5 adsorbents, those values are approximate to the size of the adsorbent specific surface. That indicates that on the surface of those adsorbents the DCE monolayer is formed. In case of WD-ekstra carbon, that value is about 30% bigger than its specific surface. That indicates that on this adsorbent the multilayer DCE adsorption takes place. The volume of micropores of WD-ekstra activated carbon equals 0.11 cm3/g, whereas the volume of pores is 0.66 cm3/g (Table 1). WDekstra activated carbon possesses big mezzo volume and macropores. The multilayer adsorption of the DCE molecules is possible owing to that structure.
5620 Ind. Eng. Chem. Res., Vol. 47, No. 15, 2008
Figure 9. Isosteres of DCE adsorption onto DTO activated carbon. Table 3. Parameters of Do’s Equation and Eq 13 adsorbent
∆H0 (J/mol)
ε
γ
µ
∆Hmin (J/mol)
DTO AG5 WD-ekstra
31 133 37 621 32 596
0.740 0.620 0.566
23.9 20.6 27.2
0 0 0
1624 8385 8454
Isosteres and Isosteric Enthalpy of DCE Adsorption. From the following van’t Hoff equilibrium equation ∆H + ln K (11) RT the isosteric enthalpy adsorption (∆H (J/mol)) can be conveniently figured out based on the isostere,19 and K is the integrate constant. The plot of ln Ce is directly related to the same equilibrium adsorption quantity at different temperatures vs 1/T. When I plot ln Ce vs 1/T, I get a straight line, and the isosteric enthalpy could be calculated from the slope of the line. The ln Ce )
Figure 10. Isosteres of DCE adsorption onto AG5 activated carbon.
isosteres corresponding to different equilibrium adsorption quantities for DCE adsorption on the tested adsorbent are shown in Figures 9–11. Figure 12 shows the dependence of isosteric enthalpy on fraction equilibrium adsorption quantity Θ. The description of the isosteric enthalpy dependence on fractional equilibrium adsorption quantity was presented by the Do model20 (broken line). Do’s model rests on the following basic assumption: the concerned system exhibits type I isotherms, the Henry’s region exists at zero loading, and the patterns as well as the extent of surface heterogeneity of adsorbent represented by the isosteric enthalpy of adsorption are independent of the adsorbate. The energy of adsorbate-adsorbate interaction is a linear function of adsorbing loading. The model takes the enthalpy as a function of the fractional loading of adsorbate
Ind. Eng. Chem. Res., Vol. 47, No. 15, 2008 5621
εγΘ + µΘ (12) 1 + (γ - 1)Θ where ∆H0 is the isosteric enthalpy at zero loading, ε represents the ratio of the variation of the isosteric enthalpy with the loading fraction from zero to monolayer coverage, γ is called the pattern parameter as it characterizes the pattern surface heterogeneity, and µ is the adsorbate-adsorbate interaction energy between adsorbed molecules. The parameters ∆H0, ε, γ, and µ can be determined by fitting the experimental isosteric enthalpy to eq 12 with a nonlinear estimation method. To their estimation the software STATISTICA 6.0 was used. It was found that the model proposed by Do holds a good matching to obtained experimental data. However, the Do model takes it that the isotherm acts according to the Henry’s law at the range of low concentrations. But the tested systems at this range of concentrations have nonlinear equilibrium. Using the Do model in such a range may misrepresent the dependence of
(
∆H ) ∆H0 1 -
)
isosteric heat of adsorption on the degree of adsorbent filling. That dependence is also described by a combination of the eqs 3 and 11 in the form of ∆H ) ∆Hmin - E ln Θ
(13)
where ∆Hmin is the isosteric enthalpy at the limiting of equilibrium adsorption quantities. In Table 3 the values of the parameters of the Do equation and ∆Hmin in eq 13 were correlated. The zero of the µ value means no adsorbate-adsorbate interaction. The larger the γ value is, the quicker the isosteric enthalpy decreases at initial adsorption stage and the slower the isosteric enthalpy decreases near the monolayer coverage loading. This suggests that the fraction of the high-energy adsorption sites on the adsorbent surface is small when the γ value of the adsorbent is great. The highest values of γ for WD-ekstra carbon indicate that it holds
Figure 11. Isosteres of DCE adsorption onto WD-ekstra activated carbon.
Figure 12. Isosteric enthalpy dependence on fractional equilibrium adsorption quantities.
5622 Ind. Eng. Chem. Res., Vol. 47, No. 15, 2008
the highest surface heterogeneity. Such case can also result from the wide pores distribution of WD-ekstra carbon. Conclusion The isotherms of DCE adsorption on all activated carbons tested are of type 1, according to the Brunauer classification. The activated carbons tested show good effectiveness of absorbing DCE from aqueous solution. The equilibrium dependences of DCE adsorption onto the activated carbons in aqueous solution are described well by the D-A model. The dependence of isosteric adsorption on the degree of surface filling at the range of 0.01-0.16 is characterized by the Do model. The whole range can be determined by the logarithmic equation, and on its basis the isosteric enthalpy at the limiting of equilibrium adsorption quantities may be assigned. Literature Cited (1) Milchert, E. Technologie Produkcji Chloropochodnych Organicznych. Utylizacja Odpado´w; Szczecin University of Technology: Szczecin, 1997. (2) Milchert, E.; Goc, W.; Pełech, R. Adsorption of CCl4 from Aqueous Solution on Activated Carbons. Adsorpt. Sci. Technol. 2000, 18, 823. (3) Pełech, R.; Bembnowska, A.; Milchert, E. Adsorption of Hydrocarbon Chloroderivatives on DTO Commercial Activated Carbon from Multicomponent Aqueous Solution. Adsorpt. Sci. Technol. 2003, 21, 707. (4) Pełech, R.; Milchert, E.; Wro´bel, R. Adsorption Dynamics of Chlorinated Hydrocarbons from Multi-component Aqueous Solution onto Activated Carbon. J. Hazard. Mater. 2006, 137, 1479. (5) Pełech, R.; Bembnowska, A.; Milchert, E. Kinetics of Adsorption of Hydrocarbon Chloro-derivatives from Seven Component Aqueous Solution onto a Thin Layer of DTO Activated Carbon. J. Colloid Interface Sci. 2005, 290, 83. (6) Speth, T.; Miltner, R. Technical Note: Adsorption Capacity of GAC for Synthetic Organics. J. Am. Water Works Assoc. 1990, 82, 72. (7) Chern, J.; Chien, Y. Adsorption Isotherms of Benzoic Acid onto Activated Carbon and Breakthrough Curves in Fixed-bed Columns. Ind. Eng. Chem. Res. 2001, 40, 3775.
(8) Urano, K.; Koichi, Y.; Nakazawa, Y. Equilibria for Adsorption of Organic Compounds on Activated Carbons in Aqueous Solutions. J. Colloid Interface Sci. 1981, 81, 477. (9) Urano, K.; Yamamoto, E.; Tonegawa, M.; Fujite, K. Adsorption of Chlorinated Organic Compounds on Activated Carbon from Water. Water Res. 1991, 25, 1459. (10) Dubinin, M. Fundamentals of the Theory of Adsorption in Micropores of Carbon Adsorbents: Characteristics of Their Adsorption Properties and Microporous Structures. Carbon 1989, 27, 457. (11) Stoeckli, F.; Lopez-Ramon, V.; Moreno Catilla, C. Adsorption of Phenolic Compounds from Aqueous Solution, by Activated Carbons, Described by the Dubinin-Astakhov Equation. Langmuir 2001, 17, 3301. (12) Horvath, A. Halogenated Hydrocarbons SolubilitysMiscibility with Water; Marcel Dekker: New York, 1982. (13) Comelli, F. Densities and Excess Molar Volumes of Binary Mixtures Containing Propylene Carbonate + 10 Chlorohydrocarbons at 298.15K and Atmospheric Pressure. J. Chem. Eng. Data 1995, 40, 1184. (14) Slvaramprasad, G.; Venkateshwara Rao, M. Density and Viscosity of Ethanol + 1,2-DichIoroethane, Ethanol + 1,1,1-Trichloroethane, and Ethanol + 1,1,2,2-Tetrachloroethane Binary Mixtures. J. Chem. Eng. Data 1990, 35, 122. (15) Dobs, R.; Cohen, M. Carbon Adsorption Isotherms for Toxic Organics. EPA 600/8-80-023. Munic. Envir. Res. Lab., USEPA, 1980. (16) Dubinin, M.; Stoeckli, H. Homogeneous and Heterogeneous Micropore Structures in Carbonaceous Adsorbents. J. Colloid Interface Sci. 1980, 75, 34. (17) Dobruskin, V. Potential Theory of Adsorption of Nonelectrolytes from Dilute Aqueous Solutions. Benzene Adsorption. Langmuir 1996, 12, 5606. (18) Terzyk, A.; Gauden, P. New Relationships Between the Characteristic Energy of Adsorption and Average Effective Diameter of Carbon Slit-like Microporessthe Dependence on the Type of an Adsorbate. Colloids Surf., A 2001, 177, 57. (19) Li, H.; Xu, M.; Shi, Z.; He, B. Isotherm Analysis of Phenol Adsorption on Polymeric Adsorbents from Nonaqueous Solution. J. Colloid Interface Sci. 2004, 271, 47. (20) Do, D.; Do, H. A New Adsorption Isotherm for Heterogeneous Adsorbent Based on Based on the Isosteric Heat as a Function Loading. Chem. Eng. Sci. 1997, 52, 297.
ReceiVed for reView November 22, 2007 ReVised manuscript receiVed May 20, 2008 Accepted May 21, 2008 IE0715862
