Langmuir 2004, 20, 9217-9222
9217
Ozonation of Naphthalenesulphonic Acid in the Aqueous Phase in the Presence of Basic Activated Carbons J. Rivera-Utrilla*,† and M. Sa´nchez-Polo‡ Departamento de Quı´mica Inorga´ nica, F. Ciencias, Universidad de Granada, 18071, Granada, Spain, and Swiss Federal Institute for Environmental Science and Technology (EAWAG), Ueberlandstrasse, 133, CH-8600, Du¨ bendorf, Switzerland Received May 24, 2004. In Final Form: July 30, 2004 The present study aimed to explore the possibility of increasing the purification efficacy of ozone in the removal of high-toxicity contaminants by using carbons of basic character and to analyze the mechanism involved in this process. These carbons were prepared by treating a commercial activated carbon (Witco, W) with ammonia (W-A), ammonium carbonate (W-C), or urea (W-U), under high pressure and temperature. The ammonia and carbonate treatments slightly increased the mesoporosity and, to a greater degree, the macroporosity of carbon W, whereas the urea treatment produced an increase in the porosity across the whole range of pore sizes. In addition, treatment of the activated carbon with these nitrogenating agents produced a marked change in the chemical nature of its surface. Thus, according to the pH of the point of zero charge (pHPZC) values obtained for each sample, carbon W was neutral (pHPZC ) 7.12), but the treated carbons were basic, especially carbon W-U (pHPZC ) 8.85). This basicity results from an increased concentration of basic oxygenated and nitrogenated surface functional groups, as confirmed by the results of elemental and XPS analyses. An increase in the degradation of 1,3,6-naphthalenetrisulfonic acid was observed when the activated carbon samples were added to the system. This degradation was especially enhanced in the presence of carbon W-U. The increased NTS degradation rate in the presence of the activated carbon is due to an increased concentration of highly reactive radicals in the system. When the catalytic activity of the activated carbon samples was related to their chemical and textural characteristics, it was found that: (i) The catalytic activity increased with an increase in the surface basicity. Interestingly, in the sample with greatest catalytic activity in NTS ozonation, carbon W-U, most of the nitrogenated surface groups introduced were pyrrol groups. These groups increase the electronic density of the basal plane of the activated carbon, thereby enhancing the reduction of ozone on the surface and the generation of highly reactive radicals in the system. (ii) The greater catalytic activity of carbon W-U may also be partly related to its greater surface area and higher volume of mesopores and macropores; these large pores facilitate access of the ozone to the surface active centers of the carbon, increasing its catalytic activity. The presence of the activated carbon samples during NTS ozonation also favored the removal of total organic carbon present in the solution, due to (a) transformation of organic matter into CO2 through the generation of highly reactive species catalyzed by the presence of the activated carbons (catalytic contribution) and (b) adsorption of NTS oxidation byproducts on the activated carbon (adsorptive contribution). The results obtained show that activated carbons treated with nitrogenating agents are very promising catalysts for application in the ozonation of aromatic compounds.
1. Introduction The past few decades have witnessed an increase in the contamination levels of water reserves, mainly due to the industrial discharge of liquid organic residues, many of which are characterized by high concentrations of contaminants of low biodegradability and high toxicity. Unfortunately, conventional biological treatment systems are inadequate to effectively remove these types of compound1-3 to the degree required by current legislation. Therefore, more sophisticated systems are required to reduce the environmental impact of these effluents and, thereby, comply with the regulations. The latest advances in water purification are mainly founded on catalytic and photochemical procedures, known as advanced oxidation processes (AOPs).4-6 AOPs are * Author to whom correspondence should be addressed. Tel: +34958-248523. Fax: +34-958-248526. E-mail address:
[email protected]. † Universidad de Granada. ‡ Swiss Federal Institute for Environmental Science and Technology. (1) Tchobanoglous, G.; Burton, F. L. Wastewater engineering: Treatment, disposal and reuse, 3rd ed.; McGraw-Hill, Inc.: New York, 1991. (2) Paula, M.; Shie, V.; Young Y. Biorrem. J. 2000, 4, 1. (3) Correa, J.; Domı´nguez, V. M.; Martı´nez, M.; Vidal, G. Environ. Int. 2003, 29, 459.
largely based on the generation in the medium of the hydroxyl radical (‚OH) and have been increasingly successful due to the highly oxidant nature of this species (E° ) 2.8 V), capable of transforming the dissolved organic matter into CO2. Recent studies by our research group showed that the generation of ‚OH radicals during the ozonation of aromatic contaminants is enhanced in the presence of activated carbon.7,8 It has been proposed that the metallic centers of mineral matter, the electrons of the basal plane, and the oxygenated basic groups present on the surface of activated carbon (chromene and pyrone) are mainly responsible for ozone decomposition in aqueous media.7,8 The ozone reduction on the activated carbon surface generates OH- and H2O2, which, according to the mechanism of ozone decomposition in aqueous phase,9,10 can initiate the process of its decomposition into ‚OH radicals. In accordance with results obtained in previous studies,7,8 the present investigation was designed to explore (4) Von Gunten, U. Water Res. 2003, 37, 1443. (5) Peyton, G. R.; Huang, F. Y.; Burleson, J. L.; Glaze, W. H. Environ. Sci. Technol. 1982, 16, 448. (6) Espuglas, S.; Yue, P. L.; Pervez, M. I. Water Res. 1994, 28, 1323. (7) Rivera-Utrilla, J.; Sa´nchez-Polo, M. Appl. Catal., B 2002, 39, 319. (8) Rivera-Utrilla, J.; Sa´nchez-Polo, M.; Mondaca, M. A.; Zaror, C. A. J. Chem. Technol. Biotech. 2002, 77, 883.
10.1021/la048723+ CCC: $27.50 © 2004 American Chemical Society Published on Web 09/04/2004
9218
Langmuir, Vol. 20, No. 21, 2004
the possibility of increasing the purification efficacy of ozone in the removal of high-toxicity contaminants by using highly basic carbons and to analyze the mechanism involved in this process. These carbons were prepared by treating a commercial activated carbon with different nitrogenating agents under conditions of high pressure and temperature. The model contaminant selected for the study was 1,3,6-naphthalenetrisulfonic acid (NTS), present in textile industry effluents and characterized by a low reactivity to ozone (kD ) 6.72 M-1 s-1) and a high affinity for free radicals (kOH ) 3.68 × 109 M-1 s-1).11,12 Because of these properties, NTS is an ideal compound for detecting the presence of highly reactive radicals in the medium. 2. Experimental Section 2.1. Materials and Methods. The ozone was produced from oxygen using an OZOKAV ozone generator with a maximum capacity of 76 mg min-1. The reactor used was covered for temperature control and was equipped with gas inlet and outlet, reactive alimentation, and sampling accessories. A more detailed description of the experimental system used was previously described.12 The original activated carbon used was a Witco commercial one, designated W, with a particle size of 0.5-0.8 mm. It was treated with ammonia (W-A), ammonium carbonate (W-C), or urea (W-U) in an autoclave at 135 °C and 2 atm of pressure. For this purpose, 25 g of activated carbon was placed in an Erlenmeyer flask with 20 g of nitrogenating agent dissolved in 100 mL of distilled water. The contact was maintained during 1 h. Then, the carbon residue was filtered and washed with distilled water to reach a constant pH. Finally, the samples were dried in an oven at 110 °C and stored in a desiccator until their use. Details of this procedure are given elsewhere.13 The method used to perform the 1,3,6-naphthalenetrisulfonic acid (NTS) ozonation in the presence of activated carbon has been described in detail elsewhere.7 The NTS ozonation experiments were conducted at a solution pH of 2. This solution pH was selected to avoid the contribution of OH- ions from the solution to ozone decomposition process into ‚OH radicals.9,10 2.2. Analytical Methods. The ozone concentration in the gas phase was determined by spectrophotometry using Spectronic Genesis 5 equipment. The dissolved ozone concentration in aqueous solutions was determined colorimetrically by the Karman-Indigo method.14 The NTS (Fluka) concentration was followed by HPLC using a Merk-Hitachi apparatus with UV detector. These methods have been described in detail elsewhere.7,8 The four activated carbon samples used as catalysts of NTS oxidation were texturally and chemically characterized using N2 and CO2 adsorption at 77 and 273 K, respectively, mercury porosimetry, elemental analysis, determination of the pH of the point of zero charge (pHPZC), selective titrations, X-ray photoelectron spectroscopy (XPS), and diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). These techniques have been described in detail elsewhere.7,15
3. Results and Discussion 3.1. Chemical and Textural Characterization of Activated Carbon Samples. Table 1 exhibits the results of the textural characterization of activated carbon W and the nitrogenating agent-treated samples. The surface area of carbon W (SN2) was increased by treatment with (9) Staehelin, J.; Hoigne´, J. Environ. Sci. Tech. 1982, 16, 676. (10) Sotelo, J.; Beltra´n, F.; Benı´tez, F. Ind. Eng. Chem. Res. 1987, 26, 39. (11) Sa´nchez-Polo, M.; Rivera-Utrilla, J.; Zaror, C. A. J. Chem . Technol. Biotech. 2002, 77, 148. (12) Rivera-Utrilla, J.; Sa´nchez-Polo, M.; Zaror, C. A. Phys. Chem. Chem.. Phys. 2002, 4, 1129. (13) Bimer, J.; Saltbut, D. P.; Berlozocki, S.; Bondon, J. P.; Broniek, E.; Siemienieska, T. Fuel 1998, 77, 519. (14) Bader, H.; Hoigne´, J. Water Res. 1981, 15, 449. (15) Valde´s, H., Sa´nchez-Polo, M.; Rivera-Utrilla, J.; Zaror, C. A. Langmuir 2002, 18, 2111.
Rivera-Utrilla and Sa´ nchez-Polo Table 1. Textural Characterization of Activated Carbon Samples sample
S N 2a (m2 g-1)
SCO2b (m2 g-1)
Vmicroc (cm3 g-1)
V2d (cm3 g-1)
V3e (cm3 g-1)
W W-A W-C W-U
812 904 825 1057
669 578 626 814
0.238 0.206 0.222 0.289
0.040 0.047 0.042 0.064
0.050 0.091 0.102 0.122
a S ) Apparent surface area determined applying BET equation N2 to N2 adsorption isotherm. b SCO2 ) Apparent surface area determined applying Dubinin-Raduskevich equation to CO2 adsorption isotherm. c Vmicro ) Micropore volume determined applying Dubinin-Raduskevich equation to CO2 adsorption isotherm. d V2 ) Volume of pores with diameter of 50-6.6 nm. e V3 ) Volume of pores with diameter above 50 nm.
nitrogenating agents, especially urea (W-U). In addition, the micropore volume of carbon W, determined from CO2 adsorption isotherms (Vmicro), was slightly decreased in carbons W-A and W-C, which could be due to a blocking effect of groups formed at the entrance to the pores, preventing access of the CO2 into a fraction of the micropores. However, carbon W-U presented an increased microporosity because, as mentioned above, the urea attacks the carbon and gasifies it, with the resulting development of its porosity. Comparison between the surface-area values determined from N2 (SN2) and CO2 (SCO2) adsorption isotherms showed that all four samples had higher values of SN2 than of SCO2. Given that the total surface area is determined with N2 and only the micropore surface area with CO2,16 these results indicate that a large fraction of the surface area of these samples corresponded to mesopores and macropores. Thus, this fraction was 17% in carbon W and 36% in carbon W-A, possibly because the micropores of the latter suffered a greater blocking effect. The pore volume values determined by mercury porosimetry (V2 and V3) are included in Table 1. It can also be deduced from these results that treatment with ammonia or ammonium carbonate develops, to a small degree, the mesoporosity and macroporosity of the carbon, whereas treatment with urea developed, to a large degree, the porosity of the carbon across the whole range, i.e., the micropores, mesopores, and macropores. The greater development of the porosity of carbon W after urea treatment may be due to the decomposition of the urea during the treatment into highly reactive nitrogenating agents, such as ammonia, ammonium cyanate, cyanic acid, polyimides, aminodihydroxitriazine, and trihydroxitriazine.13,17 Some of these compounds, obtained from the decomposition of the urea, can react with the activated carbon surface and produce its gasification, forming CH4, HCN, and CH3CN, with the resulting development of the carbon porosity.13,17 Tables 2, 3, and 4 exhibit the results of the exhaustive characterization of the surface chemistry of the carbons. The aim was to determine the modifications in the carbon surface chemistry that result from the nitrogenating agent treatments and to identify the groups that may act as catalytic centers in the ozonation process. The results of the elemental analysis of the samples (Table 2) indicate that treatment of the carbon with the nitrogenating agents reduced their carbon content because they introduced groups composed of N, H, or O. Thus, the (16) Rodrı´guez-Reinoso, F.; Linares-Solano, A. Microporous structure of activated carbons as revealed by adsorption methods. In Chemistry and Physics of Carbon; Walker, P. L., Ed.; Marcel Dekker: New York, 1989; Vol. 21, 146. (17) Jansen, J. J. R.; van Bekmannh, H. Carbon 1995, 33, 1021.
Ozonation of Naphthalenesulphonic Acid
Langmuir, Vol. 20, No. 21, 2004 9219
Table 2. Elemental Analysis of Activated Carbon Samples (% dry weight) sample
C
N
H
O (by difference)
W W-A W-C W-U
92.55 91.30 88.22 76.88
0.00 0.50 0.51 1.64
0.17 0.41 0.31 0.37
4.24 4.78 8.13 19.30
Table 3. Acidic Oxygen Surface Groups of Activated Carbon Samples Determined by Selective Titrations (in µeq g-1) sample
pHPZC
carboxyl
lactone
phenol
W W-A W-C W-U
7.12 8.50 8.38 8.85
42 17 11 18
82 55 58 63
59 52 55 42
Table 4. XPS Analysis of N1s Region of Activated Carbon Samples (%) sample
pyridine (398.5 ( 0.2 eV)
pyridone (399.5 ( 0.2 eV)
pyrrol (400.5 ( 0.2 ev)
W-A W-C W-U
11 56 8
79
10 44 50
42
percentage of these heteroatoms in the samples increased with the treatments, especially when urea was the treatment agent. To determine the character of the groups introduced by the treatments, the pHPZC of the samples was measured (Table 3). The results indicate that the treatment of carbon W with nitrogenating agents produced a marked change in the chemical character of its surface. Thus, whereas carbon W had a neutral character (pHPZC ) 7.12), the treated carbons were basic, especially carbon W-U (pHPZC ) 8.85). The functional groups generated on the carbon surface by the treatments were identified by means of selective titrations,18 XPS,19-21 and DRIFTS.21-23 The results obtained using selective titrations (Table 3) indicate that carbon W presents a low concentration of carboxylic, lactonic, and phenolic groups. Moreover, when this carbon was treated with nitrogenating agents, there was a slight reduction in the concentration of these three groups, especially the carboxylic group, which was due to their transformation into ammonium carboxylate. The deconvolution of the XPS spectra corresponding to the N1s region of the activated carbon samples (Table 4) was carried out following the procedure proposed by several authors.19-21 The results indicated that the concentration of nitrogenated functional groups introduced into the carbon depends on the treatment used. Thus, carbon W-A was characterized by a high concentration of pyridone groups, carbon W-C by a high concentration of pyridine and pyrrol-type groups, whereas the ureatreated sample was characterized by high concentrations of pyrrol and pyridone groups. (18) Boehm, H. P. Adv. Catal. 1966, 16, 179. (19) Raymundo-Pin˜ero, E.; Cazorla-Amoro´s, D.; Linares-Solano, A. Carbon 2003, 41, 1925. (20) Burg, P.; Frydich, P.; Cagniant, D.; Nanse, G.; Bimer, J.; Jankwska, A. Carbon 2002, 40, 1521. (21) Biniak, S.; Szymansky, G.; Siedlewski, J.; Swiatkowski, A. Carbon 1997, 35, 1799. (22) Go´mez-Serrano, V.; Pirit-Almeida, F.; Dura´n-Valle, C. J.; PastorVillegas J. Carbon 1999, 37, 1517. (23) Moreno-Castilla, C.; Ferro-Garcı´a, M. A.; Joly, J. P.; BautistaToledo, I.; Carrasco-Marı´n, F.; Rivera-Utrilla, J. Langmuir 1995, 11, 4386.
Figure 1. NTS ozonation in the presence of activated carbon. pH 2, T ) 298 K. (]), Without activated carbon; (O), W; (4), W-A; (+), W-C; (0), W-U. The solid symbols correspond to the adsorption kinetics of NTS on activated carbon W (b) and W-U (9).
According to the elemental analysis results (Table 2), the oxygen content of the activated carbon increased with the nitrogenating treatments, especially when ammonium carbonate or urea was used. These results indicate that oxygenated groups were formed on the carbon surface during the treatments. The data displayed in Table 3 show that these groups were not of acid character; therefore, the oxygenated groups of the treated samples must present a neutral or basic character. DRIFTS experiments were conducted to analyze these groups, and the spectra obtained reveal no appreciable differences between that of the original carbon and those of the treated samples (figure not shown). This technique is very widely used to characterize carbons, but there are difficulties involved, and the physical-chemical characteristics of activated carbons are known to complicate the results interpretation. 3.2. NTS Ozonation in the Presence of Activated Carbon Samples. After the activated carbon samples were characterized, NTS ozonation processes were studied to determine their possible catalytic activity. Figure 1 depicts the results of NTS ozonation in the presence of the activated carbon samples under study. The NTS degradation rate increased when activated carbon was added to the system and was especially enhanced in the presence of carbon W-U. As mentioned above, in previous studies,11,12 it was found that NTS shows a low reactivity against ozone (kD ) 6.72 M-1 s-1) and a high affinity for free radicals (kOH ) 3.68 × 109 M-1 s-1). The electron-withdrawing character of the sulfonic groups deactivates the aromatic ring against the reaction with ozone (Reaction 1).12
The adsorption kinetics of NTS on the carbon samples was studied to determine the influence on its elimination
9220
Langmuir, Vol. 20, No. 21, 2004
Rivera-Utrilla and Sa´ nchez-Polo
Table 5. Values of the Heterogeneous Reaction Constants of the Activated Carbon Samples sample
khetero (M-1 s-1)
W W-A W-C W-U
81.7 ( 0.2 85.6 ( 0.2 88.5 ( 0.2 120.2 ( 0.2
of the NTS adsorption on the activated carbon. The results obtained indicate that there was practically no adsorption on the activated carbon after 60 min of contact (Figure 1). Therefore, it can be affirmed that the increase in the NTS elimination rate in the presence of carbon is due exclusively to an increased concentration in the system of ‚OH radicals, which are highly reactive against NTS. The catalytic activity of each carbon during ozonation was studied by quantifying the increase in the NTS removal rate due to the presence of the activated carbon in the system. Thus, the NTS degradation rate in the presence of activated carbon, designated the total reaction rate (-rtotal), can be defined as the sum of the homogeneous reaction rate (-rhomo), calculated in absence of activated carbon,12 and the heterogeneous reaction rate (-rhetero), due exclusively to the presence of activated carbon in the system (Equation 1).
(
(-rtotal) ) (-rhomo) + (-rhetero) ) -
)
dCM dt
(
-
O3 + H2O + 2e- f O2 + 2 OH- (2)
+
homo
)
dCM dt
hetero
(1)
Furthermore, the heterogeneous reaction rate can be mathematically represented by means of eq 2:24
(
)
dCM (-rhetero) ) dt
hetero
) kheteroCMCO3
(2)
where khetero represents the heterogeneous reaction constant when activated carbon catalyzes the reaction; CO3 and CM represent the ozone and NTS concentrations, respectively, during the NTS ozonation process. Because the ozone concentration remains constant (1.04 × 10-4 M) during the NTS ozonation process, due to the ozone gas stream being continuously fed to the reactor, eq 2 can be transformed into eq 4:
kobs ) kheteroCO3
(
(-rhetero) ) -
)
dCM dt
hetero
(3) ) kobsCM
macropore volume favored this process. These large pores would facilitate access of the ozone to active centers of the carbon surface, with a resulting enhancement of their catalytic activity. When the catalytic activity was considered in relation to the chemical properties of the samples, it was observed to increase in the basic activated carbon samples. The catalytic role in NTS ozonation of chromene and pyronetype oxygenated basic groups on the activated carbon surface was analyzed in a previous study.7 Interestingly, in the sample with the greatest catalytic activity in the NTS ozonation process (W-U), a high proportion of the surface nitrogenated groups were pyrrol groups, with a low proportion of pyridine groups (Table 4). These results appear to indicate that pyrrol-type groups increase the catalytic activity of the carbon. This hypothesis could be explained as follows: in the pyrrol group, the pair of nitrogen electrons forms part of the electronic cloud of the ring and is, therefore, delocalized among the five atoms that form the molecule. As a result, pyrrol has six π electrons on five centers, so that these are π-excessive aromatic rings.25 Therefore, the presence of pyrrol groups on the activated carbon surface increases the electronic density of its basal plane and this carbon will consequently have a greater capacity to produce a reduction of the ozone dissolved on its surface (Reaction 2).
(4)
In this way, when (-rhetero) is plotted against CM, the data fit a straight line, with a regression coefficient of approximately 0.99. The value of kobs is obtained from this representation, and by substituting the value of the ozone concentration (1.04 × 10-4 M) in eq 3, the value of khetero is deduced for each carbon sample under study (Table 5). These values indicate that the catalytic activity of the samples increased in the order W < W-A < W-C < W-U. When the catalytic activity (Table 5) of the samples was considered in relation to their textural characteristics (Table 1), it was observed to be directly related to their porosity. In a previous paper,7 it was found that the catalytic activity of activated carbons in the NTS ozonation are not related to their surface area; however, a high (24) Sa´nchez-Polo, M.; Leyva-Ramos, R.; Rivera-Utrilla, J. Carbon, submitted for publication.
Moreover, the increase in the π-electron system of the carbon due to the presence of pyrrol groups produces a greater degree of interaction with the water molecules (Reaction 3).26
Cπ + 2 H2O f Cπ-H3O+ + OH- (3) In both processes, OH- ions are generated in the medium, which act as initiators of the ozone decomposition process9,10 into ‚OH radicals, which are highly reactive against NTS, increasing the degradation rate (Reactions 4-6).
O3 + OH- f ‚O2- + HO2‚ (4) ‚O2- + O3 f O2 + ‚O3- (5) ‚O3- + H+ f ‚OH + O2 (6) This reasoning may explain, in part, the high catalytic activity of carbon W-U due to its elevated concentration of pyrrolic groups. In addition, the greater catalytic activity of carbon W-C versus carbon W-A, despite the larger surface area of the latter, may be due to the higher concentration of pyrrol groups in carbon W-C. Thus, the lower nitrogen percentage of carbon W-C compared with carbon W-U and, therefore, its much lower pyrrol concentration (Tables 2 and 4) explains that carbon W-U presents a higher catalytic activity than carbon W-C. In contrast, the presence of pyridine and pyridone groups on the activated carbon surface does not favor the above reactions because, in pyridine, the nitrogen does not donate an excess of electronic density to the aromatic ring. In (25) Vollhardt, K. P. C.; Schore, N. E. Quı´mica Inorga´ nica; Ediciones Omega S. A.: Barcelona, 1996. (26) Radovic, L. R.; Moreno-Castilla, C.; Rivera-Utrilla, J. Carbon materials as adsorbents in aqueous solutions. In Chemistry and Physics of carbon; Radovic, L. R., Ed.; Marcel Decker: New York, 2000; Vol. 27, p 227.
Ozonation of Naphthalenesulphonic Acid
Langmuir, Vol. 20, No. 21, 2004 9221
Table 6. Ozonation Effect on Oxygenated Surface Groups Concentration (in µeq g-1) sample
pHPZC
carboxyl
lactone
phenol
W (Ozonated) W-U (Ozonated)
1.7 2.3
600 520
1520 960
1130 966
this heterocycle, the single pair of electrons of N are localized in an sp2-hybridized orbital and they do not form part of the electronic cloud of the aromatic ring.25 In contrast, because the nitrogen is more electronegative than the carbon, it attracts electronic density from the ring by both induction and resonance. Therefore, in the pyridine and pyridone groups, the π cloud is partially anchored on the N, being π-deficient heterocycles.25 Hence, the presence of these groups in the carbon surface reduces the electronic density of its graphene layers, which does not potentiate the decomposition of the ozone into radicals, in accordance with the above reactions. For this reason, carbons W-A and W-C, with their high concentration of pyridone and pyridine rings, respectively, presented a low reactivity in NTS ozonation in comparison with carbon W-U. To determine the participation of the different functional groups in the catalytic activity of the activated carbon in NTS ozonation, we analyzed the transformations of the carbon surface chemical groups during ozonation, which result from interaction of the ozone with the carbon surface. Table 6 displays some of the chemical characteristics of carbon W and carbon W-U after their use as catalysts in the ozonation process. The acidity of these carbons considerably increased with their ozonation to pHPZC values of 1.7 and 2.3, much lower than the values of pHPZC presented by these carbons before the ozonation. This is because the ozonation of the carbon considerably increases the concentration of carboxylic, lactonic, and phenolic groups (Table 6). The ozonation of carbon affected not only the oxygenated groups but also the nitrogenated groups. Thus, Table 7 shows the changes observed in sample W-U as an example. It can be observed that the pyridone groups were not affected by the ozonation, whereas a large number of the pyrrol-type groups were oxidized and transformed into N-oxide-type groups. Therefore, these results appear to indicate that the ozone may attack the pyrrolic groups of the activated carbon graphene planes during ozonation, yielding N-oxide-type groups and the hydroperoxide radical (Reaction 7).
Figure 2. TOC evolution during NTS ozonation in the presence of activated carbons. pH 2, T ) 298 K. (]), Without activated carbon; (O), W; (4), W-A; (+), W-C; (0), W-U.
3.3. Evolution of the Concentration of Dissolved Total Organic Carbon (TOC) during NTS Ozonation in the Presence of Carbon Samples. The capacity to remove dissolved organic matter (TOC) is a widely used parameter to evaluate the purification efficacy of a water treatment system. Figure 2 depicts the evolution of TOC concentration with NTS ozonation time in absence and presence of the carbon samples. In absence of the carbon, the TOC remained constant during the ozonation process, indicating that the ozone had inadequate oxidant power to transform the dissolved organic matter into CO2. However, in the presence of the carbon samples, the ozonation produced a major reduction in TOC. Thus, after 60 min of treatment, the percentage of TOC removed in the presence of the carbon was more than 50% of that initially present, reaching 83% in the case of carbon W-U. The reduction in the TOC concentration during NTS ozonation in the presence of activated carbons is due to two simultaneously occurring processes: (i) adsorption of NTS oxidation byproducts on the carbons, designated the adsorptive contribution, and (ii) mineralization of the organic matter, due to the generation of highly reactive radicals catalyzed by the presence of carbon in the system, designated the catalytic contribution. To determine both contributions to the overall removal of dissolved organic matter according to the NTS ozonation time, experiments were conducted using the reactor in discontinuous mode in absence of catalyst. For this purpose, an NTS solution ([TOC] ) 8 mg L-1) was treated with ozone for 30 or 60 min. Then, 1 mL of NaNO2 was added to remove the dissolved ozone. Once the dissolved ozone was removed, 0.5 g of activated carbon was added to the system, which was continuously agitated for 30 or 60 min, respectively. The reduction in TOC observed in these experiments can be considered to be exclusively due to the adsorption of NTS-oxidation byproducts on the activated carbons. Knowledge of the contribution of the adsorption to the overall removal of organic matter allows the corresponding contribution of the catalytic process to be determined. For this purpose, the contribution of the adsorption process must be subtracted from the difference between the initial and final TOC after 30 or 60 min of NTS ozonation in the presence of the carbon under study. Table 8 exhibits the values of the adsorptive and catalytic contributions to the global removal of organic matter for the different carbon samples studied at 30 and
This reaction, which explains the transformations in the pyrrol group observed during ozonation, may always contribute favorably to the NTS ozonation process, given that the hydroperoxide radical attacks the ozone, enhancing its decomposition into radicals that are highly effective in NTS degradation. This might contribute to the greater catalytic activity of pyrrol groups present in the carbon surface (Table 5).
Table 7. Ozonation Effect on Nitrogen Surface Groups Concentration (%) sample
pyridine (398.5 ( 0.2 eV)
pyridone (399.5 ( 0.2 eV)
pyrrol (400.5 ( 0.2 eV)
N-oxide (402.5 ( 0.2 eV)
W-U W-U (ozonated)
8 7
42 47
50 16
30
9222
Langmuir, Vol. 20, No. 21, 2004
Rivera-Utrilla and Sa´ nchez-Polo
Table 8. Adsorptive and Catalytic Contributions to the Overall TOC Process Removal W t (min)
adsorb.
30 60
2.44 3.44
(mg L-1)
W-A catal
adsorb.
0.76 0.77
2.20 3.24
(mg L-1)
60 min of ozonation. Interestingly, the adsorptive contribution was greater than the catalytic one in all samples. Moreover, as expected, the adsorptive contribution increased with longer ozonation time. This increase is due to (i) a reduction in the size of the NTS degradation byproducts, facilitating their access to the carbon micropores and thereby enhancing their adsorption and (ii) a longer contact time, facilitating the diffusion of the compounds to the interior of the pores. The adsorptive contribution of carbon W-U was greater than that of carbons W, W-A, and W-C, regardless of the treatment time considered (Table 8). This greater adsorption capacity of carbon W-U is largely due to the greater development of the porosity and surface area of the urea-treated carbon and to the increase in π-π dispersive interactions between the π-electron system of the carbon graphene layers and the π-electrons of the ring of the aromatic degradation byproducts from NTS.26 The urea treatment increases the electronic density of the basal planes of the carbon due to the formation of pyrrolic groups, thereby potentiating the adsorption process. Interestingly, the catalytic contribution to the overall removal of dissolved organic matter remained practically constant during the ozonation time. This fact reveals that the oxidation of the carbon samples during the NTS ozonation and the adsorption of NTS-oxidation byproducts on the carbon surface reduce its capacity to transform the dissolved organic matter in the water into CO2 during its ozonation. Thus, the mineralization of the organic matter is produced, practically, in the first 30 min of treatment. This rapid catalytic deactivation of the carbon samples during NTS ozonation may limit their use in water treatment plants. Therefore, new activated carbon treatments must be studied and developed to avoid their deactivation.
W-C catal
adsorb.
1.39 1.41
2.00 3.92
W-U
(mg L-1)
catal
adsorb.
1.33 1.36
3.47 6.06
(mg L-1)
catal 2.58 2.63
The catalytic contribution to the overall removal of TOC was greater in carbon W-U than in the other samples studied (Table 8). As commented above, this is due to the greater capacity of this sample to reduce the ozone and generate ‚OH radicals, which are highly reactive against NTS-degradation compounds, transforming them into CO2 with the consequent purification of the water 4. Conclusions The porosity of activated carbon is affected to a small degree by treatments with nitrogenating agents. The greatest textural modifications are produced in activated carbon treated with urea. The treatments increase the basicity of the carbon by generating basic, oxygenated surface groups and pyrrolic, pyridinic, and pyridonic nitrogenated groups. The concentration of these groups on the carbon surface depends on the nitrogenating agent used. The presence of activated carbon increases the NTS degradation rate and reduces the concentration of dissolved organic matter in the water. The catalytic activity of activated carbons in NTS ozonation is increased after their treatment with nitrogenating agents mainly due to the creation of pyrrol groups on the carbon surface. Acknowledgment. The authors are grateful for the financial support provided by MCT-DGI and FEDER (Project No. PPQ2001-3246-C02-01). M.S.-P. expresses his gratitude to the Junta de Andalucı´a for providing a research fellowship. LA048723+
