Environ. Sci. Technol. 2002, 36, 3850-3854
Adsorbent-Adsorbate Interactions in the Adsorption of Cd(II) and Hg(II) on Ozonized Activated Carbons M . S AÄ N C H E Z - P O L O A N D J. RIVERA-UTRILLA* Departamento de Quı´mica Inorga´nica, Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spain
The present work investigated the effect of surface oxygenated groups on the adsorption of Cd(II) and Hg(II) by activated carbon. A study was undertaken to determine the adsorption isotherms and the influence of the pH on the adsorption of each metallic ion by a series of ozonized activated carbons. In the case of Cd(II), the adsorption capacity and the affinity of the adsorbent augmented with the increase in acid-oxygenated groups on the activated carbon surface. These results imply that electrostatic-type interactions predominate in this adsorption process. The adsorption observed at solution pH values below the pHPZC of the carbon indicates that other forces also participate in this process. Ionic exchange between -Cπ-H3O+ interaction protons and Cd(II) ions would account for these findings. In the case of Hg(II), the adsorption diminished with an increase in the degree of oxidation of the activated carbon. The presence of electron-withdrawing groups on oxidized carbons decreases the electronic density of their surface, producing a reduction in the adsorbentadsorbate dispersion interactions and in their reductive capacity, thus decreasing the adsorption of Hg(II) on the activated carbon. At pH values above 3, the pH had no influence on the adsorption of Hg(II) by the activated carbon, confirming that electrostatic interactions do not have a determinant influence on Hg(II) adsorption.
Introduction Cadmium and mercury are considered two of the metals that most damage the environment (1). Their high toxicity and the greater demands of new environmental legislation have prompted the development of different technologies to treat water bodies or industrial effluents contaminated with this type of heavy metal (2). Activated carbons are widely used to treat waters contaminated with heavy metals (3, 4) because of their great capacity to adsorb these species. This capacity is mainly derived from (i) their porous texture, which gives them a large surface area, and (ii) their chemical nature, which can be appropriately modified by physical or chemical treatments to increase their adsorption capacity. The surface chemistry of activated carbons has a major influence on their capacity to adsorb inorganic compounds, especially metallic cations (5). Research efforts to elucidate the role of activated carbon surface chemistry in the mechanism of metallic ion adsorp* Corresponding author phone: +34-958-248523; fax: +34-958248526; e-mail:
[email protected]. 3850
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 17, 2002
TABLE 1. Textural Characteristics of the Carbon Samples sample
SBET (m2/g)
Sext (m2/g)
Vmicro (cm3/g)
Vmeso (cm3/g)
F F10 F120
1000 1023 632
54 87 64
0.474 0.472 0.297
0.019 0.031 0.023
tion have also been stimulated by considerable industrial and environmental interest in the use of activated carbon as support for the preparation of metal catalysts (6). Indeed, the characteristics of these catalysts greatly depend on the above mechanism. Despite numerous studies on this issue (5), doubts still remain about the influence of the chemical nature of the activated carbon surface on the adsorption of metals in aqueous solution. The adsorption mechanism and the type of adsorbent-adsorbate interactions that govern these processes are poorly understood. The present work aimed to analyze the role of the surface chemistry in creating the activated carbon surface-metallic species interactions that govern the adsorption of Cd(II) and Hg(II). A study was undertaken to determine Cd(II) and Hg(II) adsorption isotherms on a series of ozonized activated carbons with increasing concentrations of surface oxygenated functional groups and to assess the influence of the solution pH on the adsorption process. This study is part of a wider project to develop a novel system to treat industrial effluents, based on the combined use of metallic catalysts supported on activated carbon and ozone.
Experimental Section Filtrasorb 400 (Calgon Carbon Corp.), an activated carbon of bituminous origin was used, with particle size of 500-800 µm. This carbon, denoted as “F”, was treated with ozone in a fixed-bed reactor loaded with 2 g of activated carbon under a constant ozone flow of 76 mg of O3/min, operating at 25 °C and 1 atm. Ozone was produced from oxygen in an Ozocav ozone generator and fed to the reactor for exposure times of 10 (F10 sample) or 120 min (F120 sample). Carbon samples were texturally characterized from the N2 adsorption isotherms obtained at 77 K. The surface area (SBET) was calculated by applying the BET equation to the corresponding isotherms. Micropore (Vmic) and mesopore (Vme) volumes and the external surface area (Sext) were calculated by applying the R method, based on the standard isotherm for carbonaceous materials proposed by Rodrı´guezReinoso et al. (7). Textural characteristics of the carbon samples are listed in Table 1. The procedure described by Boehm was used to determine the acid and basic groups (8). The pH of the point of zero charge (pHPZC) was established with the mass titration method (9). The oxygen content of the samples was measured from the amounts of CO and CO2 desorbed when the samples were heated to 1273 K in He atmosphere in temperatureprogrammed desorption (TPD) experiments. The chemical characteristics of the carbon samples are listed in Table 2. Details of all these techniques and a major discussion of the results shown in Tables 1 and 2 can be found in a previously published study (10). The adsorption isotherms, in aqueous solution, of Cd(II) and Hg(II) on the activated carbon samples were obtained at 298 K using Cd(NO3)2‚4H2O and Hg(NO3)2‚H2O (Merck), respectively, as precursor salts. In Erlenmeyer flasks, 50 mL of metal solution (5-80 mg/L) was put in contact with 0.1 g of activated carbon. The flasks were agitated until equi10.1021/es0255610 CCC: $22.00
2002 American Chemical Society Published on Web 07/31/2002
TABLE 2. Chemical Characteristics of the Carbon Samples sample
carboxyl (µequiv/g)
lactone (µequiv/g)
phenol (µequiv/g)
carbonyl (µequiv/g)
acid sites (µequiv/g)
basic sites (µequiv/g)
pHPZC
CO/CO2
oxygen (%)
F F10 F120
none 228 1330
38 76 494
57 228 1028
139 170 518
234 702 3370
570 437 none
8.82 5.96 2.59
4.64 2.62 1.03
1.82 4.30 24.9
TABLE 4. Calculated Distribution of Cd(II) Species Present at Different Solution pH Values pH (%)
Cd2+
Cd(OH)+ Cd(OH)2 Cd(OH)3CdCl+ CdCl2 CdCl3-
FIGURE 1. Adsorption isotherms of Cd(II) on activated carbon: F (]), F10 (0), F120 (4).
TABLE 3. Results Obtained by Applying Langmuir Equation to the Cd(II) Adsorption Isotherms sample
Xm (mg/g)
BXm (L/g)
F F10 F120
6.75 9.48 13.38
0.61 4.74 7.23
librium was reached (72 h). The time required to achieve equilibrium was determined from previous experiments by measuring the adsorbate concentration versus adsorption time. It was observed that the adsorbate concentration was constant after 72 h. The equilibrium concentration was determined by atomic absorption using Variant Spectra A-200 equipment. All adsorption experiments were performed in ultrapure water at pH 6; this pH was reached by adding the appropriate amounts of NaOH and HCl. To analyze the influence of the pH on the adsorption capacity of the activated carbon, 50-mL solutions of Cd(II) (25 mg/L) or Hg(II) (50 mg/L) at different pH (2, 4, 6, 8, 10, or 12) were put in contact, in different Erlenmeyer flasks, with 0.1 g of activated carbon. The flasks were agitated at 298 K until equilibrium was reached (72 h), when the equilibrium concentration was determined. The different pH values of the solutions were obtained by adding appropriate amounts of HCl and NaOH.
Results and Discussion Adsorption of Cadmium. Figure 1 shows the adsorption isotherms of Cd(II) on the three activated carbon samples (F, F10, and F120). The Langmuir equation was applied to calculate the equilibrium isotherm capacity (Xm) and the relative affinity of Cd(II) for the different carbon samples (BXm). These findings are presented in Table 3.
Ce Ce 1 ) + X BXm Xm The adsorption capacity of the oxidized carbon (F120) was greater than that of the original carbon (F) (Table 3) despite the reduction in surface area of the former produced
2
4
6
8
10
12
81.35 0.07 0.00 0.00 18.21 0.14 0.23
96.81 2.97 0.00 0.00 0.22 0.00 0.00
24.57 75.43 0.00 0.00 0.00 0.00 0.00
0.33 99.67 0.00 0.00 0.00 0.00 0.00
0.01 99.99 0.00 0.00 0.00 0.00 0.00
0.00 99.99 0.00 0.01 0.00 0.00 0.00
by the oxidation process (Table 1). This reduction in surface area might be due to surface functional groups blocking micropores. Ozonation produced a change in the chemical nature of the carbon from basicity (carbon F: pHPZC ) 8.82) to acidity (carbon F120: pHPZC ) 2.59), which enhances the adsorption process because of the greater attractive electrostatic interactions between the surface of the activated carbon and the Cd(II) species present at the pH of the experiments. The shapes of the isotherms (Figure 1) suggest that for the sample F120, and to a lesser extent for sample F10, there are high energy adsorption sites that lead to strong adsorption at low equilibrium concentrations. The presence, in the ozonized carbon samples, of a higher amount of acidic groups ionized at the working pH (pH ) 6), mainly carboxylic groups (pKa ≈ 3-6), would justify this behavior observed in Figure 1. In fact, some authors (11, 12) have proposed that cationic species are specifically adsorbed on carboxyl and lactone groups present on the carbon surface. Table 4 displays the calculated distribution of Cd(II) species present in aqueous solution at different pH values. These calculations were made using the stability constants proposed by Stumm and Morgan (13). Regardless of the pH used, all the species presented a positive sign. At pH 2 and pH 4, the predominant species was Cd2+(aq) (81.35% and 96.81%, respectively), whereas at pH values above 4, the predominant species was Cd(OH)+. Because Cd(II) species are cationic at the working pH, the adsorption process must have been of an electrostatic type. Depending on the pHPZC value of the activated carbon sample (Table 2), the surface charge density was positive (carbon F), almost zero (carbon F10), or negative (carbon F120) at pH 6. Thus, electrostatic attractions between Cd(II) species and carbon increased in the order F < F10 < F120, which was the order in which the values of Xm and BXm increased (Table 3). We highlight the adsorption of Cd(II) on carbon F. Carbon F is basic (pHPZC ) 8.82) and presented a positive sign at the working pH so that Cd(II) adsorption should be impeded by repulsive electrostatic interactions between the activated carbon surface and the Cd2+ and Cd(OH)+ species present at this pH. However, carbon F showed an adsorption capacity of 6.75 mg of Cd(II)/g of activated carbon. These results indicate that other types of interactions, besides electrostatic ones, must be established in the adsorption of Cd(II). It is widely accepted (5) that the adsorption of metallic ions on activated carbons is produced by a mechanism of VOL. 36, NO. 17, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3851
TABLE 5. Values of the Ratio of Metal Ion Adsorbed to H+ Released per Gram Carbon (R) as a Function of the Amount of Cd(II) Adsorbed (X) F
F10
F120
R R R X (equiv of Cd/ X (equiv of Cd/ X (equiv of Cd/ (mg/g) equiv of H+) (mg/g) equiv of H+) (mg/g) equiv of H+) 0.69 1.53 2.96 4.08 4.99 6.40
0.94 1.08 1.03 1.04 0.86 0.90
2.17 3.60 6.25 6.72 8.46 9.23
0.91 0.86 0.95 0.97 0.96 0.96
2.33 4.35 6.85 8.93 11.81 13.00
0.98 1.04 0.89 0.88 0.94 1.06
ionic exchange between ionizable protons of surface oxygen groups and metallic cations. The experimental data in Table 5, which describe the relationship between the Cd(II) ion equivalents adsorbed and the proton equivalents released by the activated carbon, indicate that the above mechanism determines the adsorption process of Cd(II) in the three activated carbon samples studied. These results were obtained by determining the pH of Cd(II) solutions before adding carbon and again when equilibrium was reached after adding carbon. In this way, the number of protons given up by 1 g of carbon was calculated, and the relationship (R) between the latter and the known number of metallic ion equivalents adsorbed per gram of carbon was determined. The value of R was obtained for different amounts of Cd(II) adsorbed (X). In all cases, R was close to unity (Table 5), which suggests that adsorption occurred by a mechanism of ionic exchange between carbon surface protons and Cd(II) ions in the three carbon samples at the range of concentrations studied. Ionic exchange can be expected in carbons of acidic nature, such as sample F120. This mechanism is harder to explain in the case of basic carbons, such as sample F, which do not have enough oxygenated groups with ionizable protons (Table 2). However, our group recently demonstrated that delocalized π electrons, responsible for the basicity of these carbons, play a determinant role in the adsorption mechanism (14). These delocalized π electrons (-Cπ) are known (15-17) to be largely responsible for the basicity of the carbon because the following equilibrium is established in aqueous solution:
-Cπ + 2H2O a -Cπ-H3O+ + OH-
(1)
Thus, we have proposed (14) that in basic carbons, like carbon F, -Cπ-H3O+ interaction protons must exchange with Cd(II) cations for Cd(II) adsorption to take place. This hypothesis was supported by the values of R obtained for carbon F, which were close to unity (Table 5). The solution pH is one of the parameters with greatest influence on the adsorption of metallic ions by activated carbons (5), because the surface charge density of the carbon and the charge of the metallic species present both depend on the pH. We therefore studied the influence of the solution pH on the adsorption of Cd(II) in the three activated carbon samples, observing that the amount of Cd(II) adsorbed progressively increased with higher pH values (Figure 2). Most Cd(II) species present positive charge, regardless of the solution pH (Table 4), so that this greater adsorption would be due to progressive ionization of oxygenated functional groups that intensify attractive electrostatic interactions between the carbon surface and the Cd(II) species present at each pH. At pH 2, the amount of Cd(II) adsorbed on the three samples was very low, possibly because of the high concentration of protons in solution that restricts the exchange 3852
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 17, 2002
FIGURE 2. Influence of solution pH on Cd(II) adsorption: F (]), F10 (0), F120 (4).
FIGURE 3. Adsorption isotherms of Hg(II) on activated carbon: F (]), F10 (0), F120 (4).
TABLE 6. Results Obtained by Applying Langmuir Equation to the Hg(II) Adsorption Isotherms sample
Xm (mg/g)
BXm (L/mg)
F F10 F120
62.11 48.78 38.61
10.55 27.80 2.31
between the Cd(II) species and the protons on the carbon surface, shifting the ionic exchange equilibrium to the left. In conclusion, electrostatic interactions play a major role in the mechanism of Cd(II) adsorption, especially on acidic carbons. The adsorption capacity of ozonated carbon for Cd(II) ions is enhanced because an increase in (i) attractive electrostatic adsorbent-adsorbate interactions and (ii) the number of protons available for ion exchange with Cd(II). The adsorption of Cd(II) on basic carbons is accounted, in part, for by exchange between metallic ions and -Cπ-H3O+ interaction protons of the activated carbon surface. Adsorption of Mercury. Figure 3 shows the adsorption isotherms of Hg(II) on the three activated carbon samples (F, F10, and F120). The Langmuir equation was applied to calculate Xm and BXm values. These findings are presented in Table 6. In contrast to the case of Cd(II), the Xm values reduced with increased oxidation of the activated carbon. The predominant Hg(II) species at the pH of the isotherm determinations (pH 6) was Hg(OH)2 (Table 7). This species presents no charge, so that the influence of electrostatic interactions does not seem to be determinant in Hg(II) adsorption, unlike in the case of Cd(II). Similar results to ours (Table 6) were reported by Macı´asGarcı´a et al. (18) in their study of the adsorption of HgCl2 on H2O2-oxidized carbons. They proposed that Hg(II) was adsorbed on the activated carbon via a mechanism of HgCl2 molecular adsorption with the subsequent reduction of the
TABLE 7. Calculated Distribution of Hg(II) Species Present at Different Solution pH Values pH (%)
Hg2+
Hg(OH)+ Hg(OH)2 Hg(OH)3HgCl+ HgCl2 HgCl3HgCl42-
2
4
6
8
10
12
0.00 0.00 0.00 0.00 0.00 90.49 8.44 1.07
0.00 0.00 0.04 0.00 0.31 99.56 0.09 0.00
0.00 0.03 99.97 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100 0.00 0.00 0.00 0.00 0.00
0.00 0.00 99.99 0.01 0.00 0.00 0.00 0.00
0.00 0.00 99.87 0.13 0.00 0.00 0.00 0.00
HgCl2, thanks to the presence of phenol and hydroquinone groups on the carbon surface:
FIGURE 4. Influence of solution pH on Hg(II) adsorption: F (]), F10 (0), F120 (4).
2(C-OH) + 2HgCl2 a 2(CdO) + Hg2Cl2 + 2HCl (2)
Hg(OH)2. Thus, at the pH range studied (pH 2-12), the predominant Hg(II) species was molecular (zero charge). Figure 4 depicts the influence of the solution pH on the adsorption of Hg(II) on the original and ozonized activated carbon samples. At pH values above 3, the adsorption process was not affected by the solution pH, confirming that electrostatic interactions do not play a determinant role in Hg(II) adsorption on activated carbon. This is because, at the pH range studied, the predominant species is not ionic (Table 7). Similar results to ours (Figure 4) were observed (23, 24) in other studies of the influence of the pH on Hg(II) adsorption by activated carbons. The authors explained the reduced adsorption at acid pH by the lesser adsorption of HgCl2 species (present at acid pH) than of Hg(OH)2 species (present at pH > 6). However, their hypothesis is not supported by our results, because the amount of Hg(II) adsorbed by the three samples at pH 4 is much greater than that adsorbed at pH 2, although the predominant species at both pH values is HgCl2. We believe, therefore, that the decreased adsorption at pH 2 may be due to the lesser extent of the reduction of HgCl2 to Hg2Cl2 at this pH as a result of the elevated HCl concentration present in the medium (see reaction 2). The adsorption of Hg(II), unlike Cd(II), decreases with greater oxidation of the activated carbon. Oxygenated (electron-withdrawing) functional groups reduce the basicity and reductive properties of the activated carbon, therefore impeding the adsorption-reduction mechanism that determines this process. Adsorbent-adsorbate dispersion interactions also play an important role in the Hg(II) adsorption process.
Sinha and Walker (19) postulated a similar mechanism to explain the increase in Hg(II) adsorption on sulfur-treated carbons. In an electronic microscopy study, Adams (20) recently confirmed the presence of Hg+ and Cl- on the surface of activated carbons in contact with HgCl2, supporting this reduction mechanism:
2(C-SH) + 2HgCl2 a 2(CdS) + Hg2Cl2 + 2HCl (3) This mechanism had already been presented in studies carried out by Lo´pez-Gonzalez et al. (21). Prior to the above studies, Rivin et al. (22) had discussed the possibility that π electrons of the basal plane may be involved in the reduction of Hg on the activated carbon surface. In the present study, Xm values decreased with the reduction in basicity of the activated carbon and thus with a reduction in its surface electronic density (Table 6). These results may indicate that dispersion interactions between the activated carbon surface and the Hg(II) species present in solution [Hg(OH)2] play a major role in the adsorption process. The basicity of an activated carbon should be due to the presence of basic oxygen-containing functional groups (e.g., pyrones or chromenes) and/or graphene layers that act as Lewis bases and form electron donor-acceptor (EDA) complexes with H2O molecules (reaction 1) (5). The latter basic sites are located at π electron-rich regions within the basal planes of carbon crystallites away from the crystallite edges (5). In our view, this delocalized π electron system (-Cπ) can act as a reduction center of different species (reaction 4):
2e- + 2Hg(OH)2 a Hg2(OH)2 + 2OH-
(4)
Thus, the decrease in Hg(II) adsorption with more oxygen (electron-withdrawing) functional groups on the carbon surface (Table 6) may be due to (i) a reduction in the basicity of the carbon because of the decrease in its surface electronic density, weakening dispersion interactions between activated carbon surface and Hg(OH)2; (ii) a decrease in the reductive properties of the activated carbon (because of a diminished π electron density), impeding reduction of the Hg(OH)2 adsorbed. Both processes explain the present results. Table 7 displays the calculated distribution of Hg(II) species present in aqueous solution at different solution pH values. These calculations were made using the stability constants proposed by Stumm and Morgan (13). At pH values below 6, the predominant species was HgCl2; whereas at solution pH values above 6, the predominant species was
Acknowledgments The authors are thankful for the financial support provided by MCT-DGI and FEDER (Project PPQ2001-3246-C02-01).
Literature Cited (1) International Programme on Chemical Safety. Environmental Health Criteria; World Health Organization: Geneva, 1988. (2) Ramahlo, R. S. Tratamiento de Aguas Residuales; Editorial Reverte´ S.A.: Barcelona, 1991. (3) Chermisinoff, P. N., Ellerbusch, C., Eds. Carbon Adsorption Handbook; Ann Arbor Science Publishers: Ann Arbor, MI, 1978. (4) Bautista-Toledo, I.; Rivera-Utrilla, J.; Ferro-Garcı´a, M. A.; Moreno-Castilla, C. Carbon 1994, 32, 93-100. (5) Radovic, L. R.; Moreno Castilla, C.; Rivera Utrilla, J. Chemistry and Physics of Carbon, Vol. 27; Marcel Dekker: New York, 2001; pp 227-403. (6) Radovic, L. R.; Rodrı´guez-Reinoso, F. Chemistry and Physics of Carbon, Vol. 25; Marcel Dekker: New York, 1997; pp 243358. (7) Rodrı´guez-Reinoso, F.; Martı´n-Martı´nez, J. M.; Prado-Burguete, C.; McEnaney, B. J. Phys. Chem. 1987, 91, 515-516. (8) Boehm, H. P. Adv. Catal. 1966, 16, 179-287. (9) Noh, J. S.; Schwartz, J. A. Carbon 1990, 28, 675-682. VOL. 36, NO. 17, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3853
(10) Valde´s, H.; Sa´nchez-Polo, M.; Rivera-Utrilla, J.; Zaror, C. A. Langmuir 2002, 18, 2111-2116. (11) Aggarwal, D.; Goyal, M.; Bansal, R. C. 1999, 37, 1989-1997. (12) Cigdem, A.; Basyilmaz, E.; Bektas, S.; Genc, O ¨ .; Yu ¨ rum, Y. Fuel Process. Technol. 2000, 68, 111-120. (13) Stumm, W.; Morgan, J. J. Aquatic Chemistry: Chemical Equilibria Rates in Natural Waters; Wiley: New York, 1996. (14) Rivera-Utrilla, J.; Sa´nchez-Polo, M. Water Res. (in press). (15) Leon y Leon, C. A.; Radovic, L. R. Chemistry and Physiscs of Carbon, Vol. 24; Marcel Dekker: New York, 1994; pp 213310. (16) Dougherty, D. A. Science 1996, 271, 163-168. (17) Montes-Mora´n, M. A.; Angel-Mene´ndez, J.; Fuente, E.; Sua´rez, D. J. Phys. Chem. B 1998, 102, 5595-5601. (18) Macı´as-Garcı´a, A.; Valenzuela-Calahorro, C.; Espinosa-Mansilla, A.; Go´mez-Serrano, V. Anal. Quim. 1995, 91, 547-552.
3854
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 17, 2002
(19) Sinha, R. K.; Walker, P. L., Jr. Carbon 1972, 10, 754-756. (20) Adams, M. D. Hydrometallurgy 1991, 26, 201-210. (21) Lo´pez-Gonzalez, J. D.; Moreno-Castilla, C.; Guerrero-Ruı´z, A.; Rodriguez-Reinoso, F. J. Chem. Technol. Biotechnol. 1982, 32, 575-579. (22) Rivin, D. Rubber Chem. Technol. 1971, 44, 307-343. (23) Namasivayam, C.; Kadirvelu, K. Carbon 1999, 37, 79-84. (24) Hall, K. R.; Eagleton, L. C.; Acrivos, A.; Ver Meulen, T. Ind. Eng. Chem. Fundam. 1966, 5, 212-218.
Received for review February 4, 2002. Revised manuscript received June 20, 2002. Accepted July 2, 2002. ES0255610
