Journal of the Chilean Chemical Society
On-line version ISSN 0717-9707
J. Chil. Chem. Soc. vol.48 no.3 Concepción Sept. 2003
J. Chil. Chem. Soc., 48, N 3 (2003) ISSN 0717-9324
A PEACH STONE ACTIVATED CARBON CHEMICALLY MODIFIED TO
ADSORB AQUEOUS AMMONIA.
G. Soto-Garrido*, Croswel Aguilar, Rafael García, Renán Arriagada
Faculty of Chemical Sciences, Concepción University, Concepción, Chile
(Received: October 22, 2002 Accepted My 26, 2003)
The oxidation of an activated carbon prepared from peach stones as well as the adsorption of aqueous ammonia using untreated and modified carbons is described here in. Boehm method in conjunction with temperature-programmed desorption (TPD) were mainly used to assess the chemical changes. These methods were complemented by chemical analysis, pH measurements and FT-IR spectra. Textural characterization was obtained from CO2 and N2 adsorption. Chemical modifications performed under the same experimental conditions with aqueous solutions of HNO3 and H2O2 indicated evidences that carboxylic acids like group formation is enhanced by short periods of time of HNO3 treatment at temperatures about 353 K followed by a mild oven-dry process in atmospheric air. The activated carbon modified under such conditions resulted to be the best ammonia adsorbent.
Key Words: activated carbon, oxidation, adsorption, Boehm method, ammonia.
Several activated carbons of industrial manufacture are obtained from raw materials of carbonaceous nature like different kinds of woods, shells, and fruit pits [1,2]. It is well known and recognized now a day that chemical and textural properties of these carbons depend on their previous history . Thus, their physical and chemical behavior depends not only on the activation process itself but also on how the carbons are handled afterwards. Surface oxygen complexes [4 - 9] on active carbons have an important role in the adsorption properties of a particular carbon, they contribute also to improve its wetability. The wet oxidation of activated carbons has been studied by several authors [i.e.: 4, 7, 10, 11], different oxidizing aqueous solutions (HNO3, H2O2, ZnCl2, (NH4)2S2O8 ) at various concentrations and temperatures have been used. It is observed from literature reports that depending upon experimental conditions these oxidation will favor certain oxygenated surface groups [4,10,12,13].
The objectives of this study were to learn about the chemical characteristics of an industrial manufactured carbon obtained from peach stones and its overall changes after oxidation treatments with HNO3 and H2O2 under the same experimental conditions. Our interest was to chemically modify the as received carbon without altering its textural properties, and at the same time to favor the formation of carboxylic like groups to further study its subsequent effects on NH3 adsorption from aqueous solutions. Chemical changes were followed by the Boehm method  and temperature-programmed desorption (TPD) complemented with elemental analysis (C, H and N), pH measurements and Fourier transformed infrared spectroscopy (FT-IR). The changes of basic sites were observed from HCl neutralization .
2.1 Activated Carbon Oxidation
Peach stones-granular activated carbons from Petrochil S.A., Chile (CUDU1000) was used. To introduce chemical changes on this carbon surface it was reacted with 6.0-M solutions of HNO3 and H2O2 for 1 or 3 h at 353 K, and for 1 h at 298 K. The dry-modified carbons were kept in a desiccator, in the dark, for further use. Table 1 shows the identification and preparation conditions of studied carbons.
2.2 Textural characterization:
Surface SBET area, Dubinin-Raduskevich micropore volumes Vo (N2), Vo (CO2), and mesopore volumes (Vm) were obtained from the adsorption isotherms of nitrogen at 77 K and carbon dioxide at 273 K. Mesopore volumes were obtained by a difference of volumes at 0.95 relative pressure and the micropores volumes from nitrogen isotherms. The isotherms were measured in a Micromeritics Gemini 2375 Surface Area Analyzer.
2.3 Chemical characterization:
Elemental analyses (C, H, N) was carried out in a Perkin Elmer, PE-2400 analyzer, and oxygen was obtained by difference to 100%. Surface-acidic groups were determined by the Boehm method, and basic groups from the amount of HCl required to neutralize basic sites. The pH of the aqueous slurries were determined by the standard procedure ASTM D-24 . FT-IR Spectra were recorded on a Nicolet, Magna 550 spectrometer, the samples were prepared by mixing the carbon with KBr in a 1: 600 w/w ratio. Discs-samples were oven-dried during an hour at 323 K and kept in a desiccator for measurements . Temperature programmed desorption (TPD) were carried out from 298 K up to 1373 K using a thermal conductivity detector at 165-mA and helium flow of 50 cc/min.
2.4 Ammonia adsorption isotherms
Ammonia adsorption isotherms at 303 K were obtained using 0.200 g of activated carbon in contact with 50-mL aqueous ammonia solutions of different composition for 24 h in a thermal bath (GFL 1083) provided with constant agitation. Initial and equilibrium concentrations of ammonia were colorimetrically determined by the Berthelot reaction (Spectroquant® kit) in a UV/VIS Perkin Elmer Spectrometer (Lambda 40).
3. RESULTS AND DISCUSSION
3.1 Textural Properties
The experimental conditions used for the chemical modifications can be considered mild since they introduced no significant changes into micro and mesoporosity of treated activated carbons (AC). Table 2 shows the textural parameters of the studied carbons. It is observed that under our experimental conditions only minor changes occurred after reactions of CUDU1000 with 6 M HNO3 or H2O2 but the later showed a small decrease on surface area. Microporosities obtained from CO2 isotherms showed slightly smaller values than those of N2 adsorption consistent with using lower relative pressures (up to P / Po a). Literature reports show chemical modifications under various experimental conditions with varying results like those of Rodriguez-Reinoso et coll.  and of Pradhan et coll. and many other researchers, thus from our results and those found in the literature a chemical modification of activated carbons under mild conditions like oxidizing agent concentration, reaction temperature and reaction period will not significantly alter the textural properties of parent carbons. However, harsh conditions like higher temperatures (³ 373 K) and concentrations (³ 10 M) plus extended reaction periods will cause significant changes on the parent material 's porosity.
3.2 CHEMICAL CHARACTERISTICS
3.2.1 Untreated Carbon
Elemental analysis of as received carbon (Table 3, CUDU1000) yielded high hydrogen and oxygen content, also its TGA analysis from 298 to 823 K showed a mass loss of 25 % below 373 K thus the carbon had an important amount of water retained after its production. Its FT-IR spectrum had a broad absorption band between 3650-3000 cm-1 consistent with adsorbed water , this band disappeared upon drying the carbon. Thus, CUDU1000 was oven dry in air (4 h at 378 K) to produce CUDU_O and CUDU_O was vacuum dry (1h, 323 K) to produce CUDU_OV. FT-IR spectra of dried carbons showed only two important bands also present in CUDU1000 spectrum, at 1542 and 1190 cm-1. These bands are normally assigned to the overlap of aromatic ring modes of vibration associated with phenolic -OH, and to different skeletal structures; like C-O stretching, O-H bending of phenoxy and common features to oxidized AC . To achieve a better understanding on the behavior of the AC TPD, Boehm titration's and slurry-pH determination's were carried out. Important changes in chemical composition as well as in chemical behavior of dried carbons was found, Fig. 1, Tables 3 and 4. Their TPD profile's showed peaks between 315 - 375 K, 380 600 K, and 600-1200 K normally attributed to water-loss, low and high temperature evolved groups ; a major difference was observed in the middle peak areas. Furthermore, slurry-pH 's and Boehm ´s titration showed that CUDU1000, CUDU_O and CUDU_OV were different; so, both the oven dry and oven plus vacuum dry processes had produced changes upon the AC, the first a mild oxidation of CUDU1000, and the second a removal of labile acid groups on the carbon surface. From these observations one can deduce that the middle TPD peak of CUDU_O had at least two contributions; one from the groups that were readily removed with vacuum, and the other from group(s) that were more stable and remained at the AC surface. The most relevant observation upon using Boehm´s method after drying the original activated carbon in oven and vacuum is the variation on the amount of carboxylic like and basic groups . Upon treating the original material in oven, at 378 K, in the presence of air, basic groups were oxidized to carboxylic like groups; thus the decreased in basicity of CUDU_O with respect to CUDU1000. Furthermore, the most labile groups obtained by air oxidation of CUDU1000 could be formed by chemisorption of oxygen onto active carbon atoms producing a complex which later ionized in water in an organic acid like structure shown by scheme 1 . However, under vacuum some of the carboxylic like groups are decomposed by decarboxylation with the regeneration and formation of basic groups, as shown by carbon CUDU_OV on Table 4.
SHEME 1. Ionización of oxygen complex in water.
Fig. 1. TPD profiles, effect of drying process.
3.2.2 H2O2 Treatment
Table 4 and Fig. 2 show the chemical behavior of activated carbons produced after hydrogen peroxide treatment. The oxidation process as seeing from Table 4 increases the total acidity of CUDU1000 and is in agreement with the increase of carboxylic and phenolic like groups. Also, TPD profiles show an increment of the evolved gases, with peaks centered at the same temperatures as in the case of CUDU1000. An overall increase of areas of low as well as high temperature decomposition groups is attributed to carboxylic-lactonic like structures the first and phenolic like groups the later. FT-IR spectra of H2O2 oxidized carbons show bands at 1542 and 1190 cm-1 like those of the original carbon but the later band has broaden showing the increase in C - O stretching modes. In summary, these studies showed only small variations on the properties of H2O2 treated carbon, that is, no significant changes were observed upon oxidation which is similar to the behavior reported by other researchers [6, 9]. From Table 4, it is interesting to compared carbons CUDU_O and C298_1PO where the effect of oxidation by the hydrogen peroxide solution is markedly seeing as an important decrease of carboxylic like groups accompanied by a large increase on basic groups.
|Fig. 2. TPD profiles of CUDU1000 untreated and H2O2 treated.|
3.2.3 HNO3 Treatment
Boehm analysis, Table 4, shows that reaction at 298 K produced only minor changes, so at low temperature the HNO3 solution reacted mainly as an acid, causing only minor modifications on the AC. However, the CUDU1000 reaction at 353 K showed major changes related to the oxidizing effect of HNO3, like an overall increase of total surface groups and the largest formation of carboxylic like groups in AC C353_1NO. Also, the decrease in basicity from sample C353_1NV to C353_3NV can be explained as a result of the extended oxidation of basic groups after three hours treatment with 6 M nitric acid solution. Fig. 3 shows TPD profiles of these AC, the presence of new peaks is observed giving an account of the change in the chemical composition of the carbons, namely the increment of oxygenated groups. Elemental analysis of nitric acid modified carbons showed, as observed by others [10, 11], an increase in nitrogen content that can be attributed to nitro groups (NO2) formation which can be confirmed by the 1384 cm-1 absoption band  on the FT-IR spectra, Fig. 4. This band is not present in the spectrum of the parent carbon but it is well observed on the C353_3NV spectrum.
|Fig. 3. TPD profiles of CUDU1000 untreated and HNO3 treated.||Fig. 4. FT-IR of CUDU1000 and HNO3 treated carbon.|
4. Ammonia Adsorption
Ammonia adsorption isotherms shown in Fig. 5, represent different subgroups of L-class in Giles et al. classification . The general trend is that adsorption in carbons C353_1NO and C353_1NV is higher than in CUDU1000 while that on carbon C353_1PV is initially lower but when the ammonia concentration becomes higher than 2500 ppm its adsorption overcomes that of CUDU1000. Thus the chemical properties of carbon 's surface importantly influences the adsorption of ammonia; at low ammonia concentrations the adsorption is greatly favored by the presence of carboxylic like groups. From the shape of the isotherms one can deduce that the adsorption process must in general be different for each carbon and dependent on the different kind of sites available. Thus, carboxylic like groups largely enhance ammonia adsorption at low ammonia concentrations but at high ammonia concentrations all ACs increment their adsorption; the adsorption capacities of activated carbons at two different equilibrium concentrations show that at 500 ppm of ammonia the adsorption increase is C353_3NV < C3531PV < CUDU1000 < C3531NV < C3531NO whereas at 2400 ppm is C3531PV CUDU1000 < C353_3NV < C353_1NO < C353_1NV. Thus, the adsorption capacity of the activated carbons depends not only on their chemical characteristics but also on the adsorbate concentration. At low ammonia concentration one can observe that as the total amount of carboxylic and lactonic like groups increases the quantity of adsorbed ammonia increases with the exception of C353_3NV. The abnormal trend on this carbon can be attributed to the presence of nitro-groups which did not enhance ammonia adsorption at low concentrations but provided the acidic character of the carbon. All the adsorption isotherms show that as the ammonia concentration increases its adsorption by AC also increases; this could be interpreted as ammonia adsorption by hydrogen bond formation  which is enhanced at high ammonia concentrations or at least when it is high enough to compete with water molecules of the media. This observations lead us to propose that carboxylic and lactonic like groups play an important role in the adsorption process at low ammonia concentrations were they will retain the ammonia by an acid-base interaction but at higher concentrations (³1000 ppm) ammonia competes with water to form hydrogen bonds with other oxygenated groups or electron donor sites also present on the AC surface.
|Fig. 5. Ammonia adsorption isotherms at 303 K.|
The activated carbon CUDU1000 can be chemically modified using either mild air oxidation, H2O2 or HNO3 solutions. Modified activated carbons show an acid behavior in aqueous solution; the activated carbon treated with HNO3 resulted in greater acidity and long reaction periods increase the formation of nitro-groups. The drying process affects the oxidized activated carbons, thus oven drying favors stabilization of carboxylic like groups while vacuum drying causes partial elimination of these groups and a small increment of lactonic and phenolic like groups.
Under the same experimental conditions, H2O2 was a stronger oxidant than HNO3 since it was less selective, producing minor quantities of carboxylic and lactonic like groups and favoring the stabilization of basic groups; the small decrease observed on SBET can be attributed to porosity lost associated with oxidation that caused some porous destruction due to CO2 elimination. In general, all ACs adsorb ammonia, the higher the content of carboxylic and lactonic like groups the better was the adsorption capacity of the AC. The presence of basic groups did not enhance adsorption at low concentrations.
The authors wish to thank MECESUP for Ph.D. thesis fellowship of C. Aguilar, and grants FONDECYT 102-0468 and DIUC 98.022.014-1.0 for financial support. C. Aguilar also acknowledges a leave of absence from Trujillo University (Trujillo, Perú).
1. R.C. Bansal, J.B. Donnet, F. Stoeckli, Active Carbon. Marcel Dekker, New York (1988) [ Links ]
2. F. Rodríguez-Reinoso, In Introduction to Carbon Technologies, Chapter 2, p. 35 (Edited by H. Marsh, E.A. Heintz, F. Rodriguez- Reinoso) University of Alicante (1997) [ Links ]
3. A.J. Groszwek, Carbon 1987, 25:717. [ Links ]
4. J.L. Figuiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M.Órfao, Car bon 1999, 37:1379. [ Links ]
5. A.Dandekar, R.T.K. Baker, M.A. Vannice, Carbon 1998, 36: 1821. [ Links ]
6. V. Gómez-Serrano, M. Acedo-Ramos, A.J. López-Peinado, C. Valenzuela-Calahorro, Fuel 1994, 73 (3): 387. [ Links ]
7. M. Molina-Sabio, M.A. Muñecas-Vidal, F. Rodriguez-Reinoso, Characterization of Porous Solids II, p.329. Elsevier Science, Amsterdam (1991). [ Links ]
8. J.Rivera-Utrilla, M.A. Ferro-García, Adsorption Science and Tech nology 1986; 3:293. [ Links ]
9. Y. Matsumura, S. Hagiwara, H. Takahashi, Carbon 1976, 14:163. [ Links ]
10. B.K. Pradhan, N.K. Sandle, Carbon 1999, 37:1323 [ Links ]
11. Y. Otake, R.G. Jenkins, Carbon 1993, 31:109 [ Links ]
12. C. Moreno-Castilla, F. Carrasco-Marín, A. Mueden, Carbon 1997, 35: 1619. [ Links ]
13. H. Benaddi, T.J. Bandosz, J. Jagiello, J.A. Schwarz, J.N. Rouzaud, D. Legras, F. Béguin, Carbon 2000, 38: 669. [ Links ]
14. H.P. Boehm, Carbon 1994, 32: 759. [ Links ]
15. S. Weller, T.F. Young, J. Am. Chem. Soc. 1948, 70: 4155 [ Links ]
16. American Society for Testing Materials, ASTM: D 1512 95. [ Links ]
17. P.R. Solomon, R.M. Carangelo, Fuel 1982, 61:663. [ Links ]
18. R.C. Bansal, J.B. Donnet, F. Stoeckli, Active Carbon, Chap.2, p.41. Marcel Dekker, New York (1988). [ Links ]
19. J. Zawadski, Carbon 1980, 18:281. [ Links ]
20. C.H. Giles, T.H. MacEwan, S.N. Nakhwa, D. Smith, J. Chem. Soc. 1960; 3973 [ Links ]
21. W.H. Lee, P.J. Reucroft, Carbon 1999, 37:2 [ Links ]