7474
Ind. Eng. Chem. Res. 2009, 48, 7474–7481
Porosity Development in Activated Carbons Prepared from Walnut Shells by Carbon Dioxide or Steam Activation Juan F. Gonza´lez,*,† Silvia Roma´n,† Carmen M. Gonza´lez-Garcı´a,† J. M. Valente Nabais,‡ and A. Luis Ortiz§ Departamento de Fı´sica Aplicada and Departamento de Ingenierı´a Meca´nica, Energe´tica y de los Materiales, UniVersidad de Extremadura, AVda ElVas s/n, 06071 Badajoz, Spain, and Centro de Quı´mica de E´Vora e Departamento de Quı´mica, UniVersidade de E´Vora, Rua Roma˜o Ramalho no. 59, 7000-671 E´Vora, Portugal
The influence of carbon dioxide and steam as activating agents on the porosity development of activated carbons produced from walnut shells was investigated. The study was made covering a wide range of burnoff (12-76%) and employing different temperatures and times: in carbon dioxide activation, 850 °C varying the activation time in the range 60-480 min, and in steam activation, 700, 850, and 900 °C (for 30-120 min). It was found that the gasifying agent has a profound influence on the activated carbon porosity development. First, steam is more reactive and produces, in general, activated carbons with greater N2 adsorption capacity. Second, the increase in the fraction of mesopores with activation time is more pronounced for steam. While steam generates micro-, meso-, and macropores from the early stages of the process, carbon dioxide produces highly microporous carbons, with broadening of the microporosity only for long activation times. 1. Introduction Activated carbons (ACs) can be prepared from a large number of organic precursors. Among them, agricultural subproducts or wastes are currently being extensively studied because of their cheapness, availability in large quantities, and potential to create significant economic added value. Published studies have considered the use of such waste products as walnut shells,1 peach stones,2,3 olive stones,4-6 almond shells,7,8 and coffee endocarp,9 to cite just a few. It is widely accepted that the porosity of ACs depends not only on the raw material used as precursor, but also on the manufacturing process. With respect to physical activation, it has been shown that the molecular size and reactivity of the activating agents play important roles in the porosity development of the resulting ACs. In this line, many researchers have established the differences between the way in which carbon dioxide and steam develop porosity in chars.5,10,11 Reactions of carbon dioxide or water steam with carbon are endothermic and require temperatures in the range 700-950 °C to eliminate carbon atoms. The reactions involved (see below) are (1) in the case of CO2 activation and (1)-(3) for steam activation. The equilibrium (1) is present in both activations due to the large quantities of CO2 generated during the process and the high temperatures involved: C + CO2 f 2CO
(1)
C + H2O f CO + H2
(2)
CO + H2O f CO2 + H2
(3)
These equilibria are favored by temperature to different extents. Most authors agree that, at a given temperature, the reactivity of steam is greater than that of carbon dioxide.5,6,11-13 Also, * To whom correspondence should be addressed. E-mail: jfelixgg@ unex.es. Tel.: +34924289619. Fax: +34924289601. † Departamento de Fı´sica Aplicada, Universidad de Extremadura. ‡ Universidade de E´vora. § Departamento de Ingenierı´a Meca´nica, Energe´tica y de los Materiales, Universidad de Extremadura.
these agents give rise to ACs with very different pore size distributions, and the impact of varying the activation conditions on the pore structure of the carbons also depends on the activating agent. The present study was conducted with the objective of developing adsorbents from walnut shells that have similar features to the current commercially available ACs in terms of high surface areas and tailored pore size distributions to be used in a wide variety of adsorption applications. Walnut shells were selected as AC precursors because of their availability and suitability for AC production, as reported by previous studies. Earlier studies have investigated walnut shells as a raw material for AC production. However, most have centered on chemical activation with ZnCl2,14,15 KOH,16 or K2CO3.17 The few studies that have considered the physical activation of walnut shells were performed in the framework of a broader study of various lignocellulose materials18,19 and investigated the effect of a single activating agent.20 2. Materials and Experimental 2.1. Materials. ACs were prepared from walnut shells, which were previously crushed, ground, and sieved to a particle size of 1-2 mm. The ultimate analysis was performed using a LECO CHNS 1000 analyzer. The proximate analysis was carried out following the procedure described elsewhere.21-23 The results are given in Table 1. With respect to the proximate analysis, it can be seen that the walnut shells have a high content of volatile matter and low content of ash, which is clearly important for gasification and pyrolysis processes. On the other hand, from the ultimate analysis, we can highlight that this precursor also presents no sulfur and a low content of nitrogen. Table 1. Proximate and Ultimate Analyses of Walnut Shell proximate analysis (%) volatile fixed carbon moisture matter 15.86 a
11.02
71.81
Balanced.
10.1021/ie801848x CCC: $40.75 2009 American Chemical Society Published on Web 07/28/2009
ultimate analysis (%) ash
C
H
N
S
Oa
1.31 45.10 6.00 0.30 0.00 48.60
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009
Figure 1. Degree of burnoff as a function of activation time.
Figure 2. N2 adsorption isotherms (a) and RS-plots (b) of the carbons activated with CO2. Table 2. Lignocelullosic Composition of Walnut Shell
walnut shell
hemicelullose (%)
celullose (%)
lignin (%)
20.7
40.1
18.2
The lignocellulosic composition of walnut shell, determined according to the Van Soest method,24 is given in Table 2. From the values of hemicellulose, cellulose, and lignin, it can be seen that this precursor has a low content of lignin. According to Daud and Ali25 and Gergova et al.,26 materials with a low content of this fraction are prone to yield microporous activated carbons. 2.2. Experimental Section. The carbonization and activation processes were performed in a laboratory prototype device described elsewhere,12 which consists basically of a sealed vertical furnace with controlled atmosphere and temperature, and a system to store tars and condensable liquids. The carbonization was done at 600 °C under a constant N2 flow of 120 mL min-1 for 60 min. We chose these operating variables according to previous works which showed that these were the optimal conditions to yield the chars with higher porosity development.27
7475
Activations were performed by heating 5 g of the char under N2 flow at a rate of 20 °C min-1 up to the target temperature, then switching to the desired gas and maintaining for the appropriate time, and then switching back again to the N2 flow during the cooling down to room temperature. In particular, CO2 activations were carried out at 850 °C during times in the 60-480 min interval, with a flow rate of 40 cm3 min-1. Water steam activations were carried out either at 700 °C for 60-120 min, or 850 and 900 °C for 30-60 min, with a water flow rate of 0.2 g min-1 diluted in the N2 flow to achieve a steam partial pressure of 0.6. The resulting ACs were designated X-T-t, where X is the activating agent (C and S for CO2 and water steam, respectively) and T and t are the activation temperature and time, respectively. 2.3. Textural Characterization of the Char and Activated Carbons. The char and ACs were characterized by N2 adsorption at -196 °C using a semiautomatic adsorption unit (Autosorb-1, Quantachrome, USA) and by CO2 adsorption at 0 °C with a manual volumetric system. Adsorption measurements were made on 1.5 g of sample that was previously outgassed at 300 °C for 12 h to a residual vacuum of 10-5 torr. The N2 adsorption isotherms were used to (1) calculate the specific BET surface (Brunauer, Emmett and Teller, SBET),28 (2) evaluate the external surface (SEXT) using the Rs-method29 with the reference material proposed by Rodrı´guez-Reinoso et al.,30 (3) determine the volume of micropores (Vmi) through the Dubinin-Radushkevich equation,31 and (4) compute the mesopore volume (Vme) by subtracting Vmi from the nitrogen adsorbed at P/P0 ) 0.95.32 The value of the percentage of internal surface (% SINT) was calculated using the expression % SINT ) 100(SBET - SEXT)/SBET. The CO2 adsorption isotherms were used to (1) calculate the micropore volume (Wo) according to the Dubinin-Radushkevich equation and (2) determine the micropore size distribution, characteristic energy (Eo) and the mean equivalent micropore radius (req) by the Medek model.33 The mercury intrusion curves were collected at room temperature with a conventional porosimeter (Autopore 4900 IV, Micromeritics, USA) and were used to determine the meso(VmeP) and macropore (VmaP) volumes as well as the mercury density (FHg). Helium densities (FHe) were measured with a stereopyknometer (Quantachrome, USA) following the standard procedure.34 The total pore volume (Vt) was calculated from FHg and FHe by the usual method.35 2.4. Morphological Characterization. The char and ACs were observed under scanning electron microscopy (SEM; Hitachi S-3600N, Japan) to examine in detail the particle morphology. The SEM samples were prepared by depositing about 50 mg of the char and ACs individually on aluminum studs covered with conductive adhesive carbon tapes and then coating with Rd-Pd for 1 min to prevent charging during observations. Imaging was done in the high vacuum mode at an accelerating voltage of 20 kV, using secondary electrons. 3. Results and Discussion 3.1. Char. The carbonization of the precursor led to a char (solid yield of 25.1%) with the textural characteristics inferred from the N2 and CO2 adsorption isotherms and the mercury intrusion listed in Table 3. Since the N2 adsorption isotherm of the char was of type I according to the BDDT (Brunauer, Deming, Deming, and Teller) classification,36 it could be concluded that the char was essentially microporous in nature. Also, the char contained a significant volume of macropores, as revealed by the mercury
7476
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009
Table 3. Textural Characteristics of Walnut Shell Char SBET (m2 g-1)
Vmi (cm3 g-1)
Vme (cm3 g-1)
SEXT (m2 g-1)
VmaP (cm3 g-1)
VmeP (cm3 g-1)
FHe; FHg (g cm-3)
Vt (cm3 g-1)
W0 (cm3 g-1)
209
0.12
0.05
21
0.45
0.19
1.55; 0.67
0.85
0.46
Table 4. Textural Characteristics of the Activated Carbons Prepared with Steam and CO2 as Determined by N2 Adsorption at -196 °C
C-850-60 C-850-120 C-850-180 C-850-240 C-850-480 S-700-60 S-700-120 S-850-30 S-850-45 S-850-60 S-900-30 S-900-60 a
burnoff (%)
burnoffa (%)
SBET (m2 g-1)
Vmi (cm3 g-1)
Vme (cm3 g-1)
SEXT (m2 g-1)
SINTb (%)
20.6 35.1 39.8 43.7 59.6 27.2 42.3 37.8 50.2 56.2 39.7 74.3
5.2 8.8 9.9 10.9 15.0 6.8 10.6 9.5 12.6 14.1 9.9 18.6
542 743 841 1220 1304 542 611 699 966 1361 790 1339
0.30 0.39 0.49 0.68 0.72 0.30 0.31 0.37 0.47 0.74 0.41 0.72
0.05 0.09 0.18 0.18 0.21 0.04 0.14 0.08 0.11 0.20 0.04 0.09
49 94 145 223 269 30 131 99 163 203 46 142
91 87 83 82 79 95 79 86 83 85 94 89
Determined in relation to raw material. b % SINT calculated according to % SINT ) 100(SBET - SEXT)/SBET.
Figure 4. Micropore size distribution of char and selected samples.
Figure 3. N2 adsorption isotherms (a) and RS-plots (b) of the carbons activated with steam.
intrusion analysis. A comparison between the micropore volumes determined by CO2 (W0) and N2 (Vmi) adsorptions reveals the presence of certain discrepancies. Note that W0 is almost four times greater than Vmi, indicating clearly that the adsorption process depends markedly on the adsorbate and temperature. This discrepancy between W0 and Vmi has been observed in other chars37-39 and has been justified by alluding to certain constrictions that make it difficult for N2 to diffuse at -196 °C into the narrower micropores as its adsorption kinetics are extremely slow at that temperature and the measurement time is likely not long enough for equilibrium to be reached. 3.2. Activated Carbons. Figure 1 shows the degree of burnoff as a function of activating time for CO2 and steam activations. It can be seen that the degree of burnoff increased with increasing time and that this trend was more pronounced in the case of steam. This in turn indicates that the reactivity of
Figure 5. Mercury intrusion curves of char and selected samples.
the char toward steam is higher than toward CO2, as has been observed in other carbonaceous materials.4-6 In addition, in the case of steam activation, the activation rate evaluated by the slope of the curve is greater at higher temperatures. In principle, as Wigmans has demonstrated,13 steam has more advantageous kinetics than CO2 because its lower molecular size allows faster diffusion through the pore network and consequently easier access to the micropores. However, whether steam or CO2 should be used depends on the desired pore structure. 3.2.1. N2 Adsorption. Figure 2a shows the N2 adsorption isotherms of the ACs prepared by CO2 activation (i.e., samples C-850-60, C-850-120, C-850-180, C-850-240, and C-850-480). The corresponding Rs-plots are shown in Figure 2b. The degrees of burnoff and the textural characteristics determined from N2
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009
7477
Figure 6. Micrograph of walnut shell char. Magnification: (a) 900 and (b) 8000.
Figure 7. Micrograph of activated carbon C-850-120. Magnification: (a) 1000 and (b) 10 000.
adsorption isotherms are given in Table 4. From Figure 2a, it can be seen that the volume of adsorbed N2 increased with increasing activation time. The shape of the isotherms does not change significantly between C-850-60 and C-850-120 (the isotherms are type I, according to BDDT classification,36 typical of microporous materials), indicating that the pore size distribution of these carbons is very similar. However, the slope of the RS-plots in Figure 2b increased slightly, indicating a slight increase of the contribution of their external area to the total surface area. Dissimilar behavior is seen for C-850-180, C-850240, and C-850-480, which are very similar to each other, but different from those of C-850-60 and C-850-120. As it can be inferred from Figure 2a, the isotherms of C-850-180, C-850240, and C-850-480 approach the plateau more slowly, and the values of adsorbed N2 increase steadily over the entire P/P0 range, indicating a broader pore size distribution. In this case the isotherms would then be composite isotherms of types I and II.36 This shows that there is a clear widening of the micropores at higher degrees of burnoff, which is consistent with most of the literature results.8,40,41 The most striking observation is the drastic change in pore development for degrees of burnoff below and above 35% (see Table 4). For low burnoff values, the ACs are essentially microporous. Further activation yields carbons presenting wider microporosity with some mesoporosity. These samples also have a large value of SEXT, as deduced from the increasing slope in the RS-plots (Figure 2b).
These facts can be explained by considering that CO2 primarily reacts with the active sites located at the center of the pores (thus creating microporosity) and only attacks the pore walls when the activation time is large (broadening the microporosity and yielding more mesoporous carbons).4-6 Steam, like CO2, can produce ACs with tailored porosity by controlling the experimental activation conditions. Figure 3a and b show the N2 adsorption isotherms and the corresponding RSplots, respectively, for the ACs produced by steam activation. It can be seen that the adsorption capacity of the carbons increased with increasing activation time at the three temperatures studied (700, 850, and 900 °C). From Figure 3a, it can be seen that the N2 adsorption isotherms of S-700-60, S-85030, and S-900-30 are all type I.36 The small increase found in the N2 adsorption volume at medium-to-high values of relative pressure indicates that these three samples present low values of mesopores and wide micropores, and the low slopes of their RS-plots are indicative of the low contribution of their SEXT. On the other hand, although the N2 adsorption isotherms of samples S-700-120, S-850-45, S-850-60, and S-900-60 are of type I, they show some features typical of type II. For an activation time of 30 min, the increase in temperature (compare S-850-30 and S-900-30) does not markedly change the textural characteristics of the ACs (producing only a small increase in the N2 adsorption capacity). For an activation time of 60 min, the increase in temperature from 700 to 850 °C (see S-700-60 and S-850-60) leads to greater N2 adsorption capacities
7478
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009
Figure 8. Micrograph of activated carbon S-850-30. Magnification: (a) 1000 and (b) 10 000.
of the ACs, giving rise to a wider pore size distribution. A further increase in temperature (S-900-60) resulted in the decrease in the N2 adsorption capacity because the action of steam over such a long time could be causing the external burning of the sample. The micropore widening effect induced by the increase in the burnoff degree observed in the three series depended on the temperature. It can be seen from Table 4 that increasing the activation time from 60 to 120 min at 700 °C (S-700-60 and S-700-120) had no effect on the micropore volume and only little effect on the mesopore volume (resulting in a decrease in % SINT from 95 to 79%). At 850 °C, increasing the activation time from 30 to 45 and 60 min (S-850-30, S-850-45, and S-85060) gave rise to a major increase in the N2 adsorption capacity (699, 966, and 1361 m2 g-1, respectively). At 900 °C, the increase in activation time had less effect than at 850 °C because, as can be seen from Table 4, the pore volumes increased to a lesser extent, especially in the case of mesopores. The optimal steam activation conditions to get microporous ACs would be high temperatures and short activation times, to attain degrees of burnoff of around 35-40%, as in S-850-30 and S-900-30. This is consistent with the results reported by other workers4-6,13 that in steam activation microporosity development reaches a maximum at a degree of burnoff of
Figure 9. Micrograph of activated carbon C-850-480. Magnification: (a) 1000 and (b) 10 000. Table 5. Textural Characteristics of the Char and Selected Activated Carbons Prepared with Steam and CO2 as Determined by CO2 Adsorption at 0 °C sample
E0 (kJ mol-1)
req (nm)
W0 (cm3 g-1)
char C-850-120 C-850-480 S-850-30
11.1 10.4 7.6 9.5
0.71 0.72 0.80 0.75
0.46 0.69 0.44 0.67
Table 6. Textural Characteristics of Selected Activated Carbons from Hg and He Measurements
C-850-120 C-850-480 S-850-30
VmeP (cm3 g-1)
VmaP (cm3 g-1)
FHg (g cm-3)
FHe (g cm-3)
Vt (cm3 g-1)
0.13 0.18 0.20
0.74 1.77 1.03
0.45 0.32 0.47
1.98 2.20 1.99
1.16 2.31 1.59
approximately 50%. Above that value, further activation converts micropores to mesopores. To produce mesoporous carbons, the optimal activation conditions would be moderate temperatures (around 700 °C) and long activation times (about 120 min). Finally, under analogous experimental conditions, steam activation produces ACs with broader pore size distributions and greater SEXT than CO2 activation. This is due to its greater reactivity and faster diffusion through the porous carbon network. See, for example, the great difference in SEXT between
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009
7479
Figure 10. FTIR spectra of char and selected samples.
C-850-60 and S-850-60 (49 and 203 m2 g-1, respectively). This suggests that although both agents develop porosity of all sizes, CO2 is more selective toward the creation of internal area, whereas steam enlarges the micropores more effectively. From the textural characteristics given in Table 4, it can be concluded that for low degrees of burnoff (around 20-30%) the two activations result in similar porosity development. For example, one may compare C-850-60 with an SBET of 542 m2 g-1, Vmi of 0.30 cm3 g-1, and degree of burnoff of 20.6% and S-700-60 with a burnoff of 27.2%, SBET of 542 m2 g-1, and Vmi of 0.30 cm3 g-1. However, for greater degrees of burnoff, steam seems to be more effective than CO2, yielding ACs with higher pore volumes. Tomko´w et al.,10 Molina-Sabio et al.,3 and Ku¨hl et al.42 studied the steam and CO2 activation of coke at similar degrees of burnoff and concluded that the former led to higher adsorption capacities and wider pore size distributions. Ryu et al.43 also obtained broader pore size distributions with steam, although they reached greater micropore volumes with CO2 activation. One infers from these and the present study that, in general, steam yields ACs with a greater mesopore contribution than CO2. Such a difference is due to the way in which these agents react with the active sites of the pore structure: steam attacks the active sites at the center and on the walls of the pores simultaneously, whereas CO2 primarily reacts with the active sites at the center of the pores and then attacks their walls, therefore needing larger activation times to develop new microporosity and broaden the existing micropores.44 3.2.2. CO2 Adsorption. From the CO2 adsorption isotherms at 0 °C, the micropore volume (W0) was calculated and compared with that obtained by N2 adsorption at -196 °C (Vmi). As an example, the values corresponding to the char and some ACs are given in Table 5. Garrido et al.37 compared exhaustively the N2 and CO2 adsorption of a group of ACs, analyzing the three possibilities (Vmi < W0, Vmi ≈ W0, Vmi > W0). They found that the situation Vmi< W0, as in the char, C-850-120 and S-850-30, is typical of chars and very microporous ACs, with narrow micropores where N2 adsorption has diffusional limitations. The situation Vmi > W0, found in C-850-480 corresponds to ACs with larger and more heterogeneous microporosity. These facts can be inferred from the micropore size distributions in Figure 4, as well as from the mean equivalent radius (req) and the characteristic energy (E0) given in Table 5. From Table 5, one concludes that the char and the samples C-850-120 and S-850-30 have very
similar mean equivalent radii, whereas that of C-850-480 is shifted toward larger micropores. One also deduces from Figure 4 that the pore size distribution of C-850-480 is broader. Comparing samples C-850-120 and C-850-480 in Table 5, one can see that increasing activation gave rise to an increase in the mean equivalent radius (from 0.72 to 0.80 nm) and the characteristic energy decreased (from 10.4 to 7.6 kJ mol-1). Comparison of the pore size distributions of S-850-30 and C-850-120 confirms the arguments given above about the widening effect caused by steam from the early stages of activation: 30 min were enough to induce in S-850-30 a mean equivalent radius larger than that found in C-850-120, which has a very similar degree of burnoff. 3.2.3. Hg Porosimetry and He Stereopyknometry. Figure 5 includes, as an example, the cumulative pore volume plots for samples char, S-850-30, C-850-120, and C-850-480. The textural characteristics obtained from these analyses are listed in Table 6. It can be seen that, although the volume of macropores was greater than that of mesopores in all the samples, the pore size distribution upon activation was very different. First, CO2 activation for 120 min resulted in the increase of the macropore volume. Nevertheless, the trend found for the mesopore volume has to be interpreted with caution: according to the Hg intrusion, CO2 activation caused a decrease in the mesopore volume (VmeP(char) ) 0.19 cm3 g-1 and VmeP(C850-120) ) 0.13 cm3 g-1). However, we noted higher mesopore volume in C-850-120 than in the char by N2 adsorption (0.09 and 0.05 cm3 g-1, respectively). This difference may be due to the lower detection limit of the porosimeter used (3 nm), which makes this technique less sensitive to the narrower mesopore range that might be significant in this sample. Some compaction effects or even mesopore destruction could also occur at the high pressures applied. Also noticeable is the shape of the intrusion curve of C-850-120, which exhibits negligible intrusion volumes at ranges of pore radii above 2000 nm. CO2 activation for 480 min (sample C-850-480) produced major meso- and especially macropore development, as observed in Figure 5 and Table 6. The increase in mercury intrusion volume was very marked in the macropore interval (as can be seen from its slope). The intrusion curves of C-850-120 and S-850-30 for the interval of radii up to 100 nm are very similar and present a feature typical of a material with heterogeneous pore size distribution: the Hg volumes increase over the entire range of pore radii, indicating that both meso- and macropores were developed.
7480
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009
3.2.4. SEM Characterization. Figures 6-9 show the micrographs taken on the following samples: char, C-850-120, C-850-480, and S-850-30. The micrographs of the char in Figures 6a and b show the existence of a porous structure in which the skeleton has a smooth surface and contains many craters. These features have been observed in almond shell chars by Marcilla et al.,39 who proposed that the craters are in fact the chimneys formed as the hot gases produced during the carbonization escaped suddenly from the sample. As can be seen in Figures 7 and 8, the activation causes various topological changes in the char, regardless of whether the agent was CO2 or steam: (i) the chimneys disappear leaving behind only the holes and (ii) new pores are formed in the previously smooth skeleton. Also, whereas CO2 activation for 120 min (Figure 7a and b) gave rise to a significant development of small pores, further activation up to 480 min (Figure 9a and b) resulted in the enlargement of these small pores. Large interconnected channels are evident in Figure 9a, and signs of a significant volume of micropores are also observed (see the spotted walls in Figure 9b). Finally, the activating agent influenced the topographical characteristics of the carbon surfaces. Although the values of equivalent micropore radius (req) were very similar for samples of C-850-120 and S-850-30 (Table 5), the former exhibited a hollower pore structure, while the latter showed a thicker structure. This is consistent with the activation mechanism of these agents, as discussed above: CO2 is more likely to seek the active sites inside the pores, thus opening new channels toward the inner part of the particle. 3.2.5. Fourier Transform Infrared (FTIR) Analysis. Figure 10 shows the FTIR spectra of the char and samples C-850120, C-850-480, and S-850-30. The spectra obtained show that the surface chemistry of the char and activated carbons does not exhibit significant differences, independently of the activating agent used. Only slight differences on the intensity of the bands were detected. Among the surface functional groups, which were identified according to the information found in the bibliography,45,46 the following ones can be highlighted: The wide bands at 3400 cm-1 can be related to vibrations ν(OsH) in water molecules. The weak band found around 2900 cm-1 could be due to carboxylic acids or aliphatic groups. Also, the band at 2350 cm-1 is generally associated with vibrations in aliphatic bonds. The bands in the region 1400-1800 cm-1 can be attributed to CdO bonds, related to the presence of carbonyl groups or aromatic rings. Finally, the weak band detected around 1123 cm-1 can be related to ν(CsO) bonds in lactones, epoxides, and ether structures. 4. Conclusions Steam and CO2 activations of walnut shells allow one to obtain ACs with a high porosity development. Both agents mainly develop microporosity in the char at low degrees of burnoff. The activation time plays a critical role in the pore size distribution of the ACs produced by both activating agents; higher burnoffs cause the progressive widening of the pores in the carbon. However, because of steam’s greater reactivity in comparison with CO2 and the way in which it interacts with the active sites located anywhere in the pore structure, this widening of the carbon porosity occurs more sharply in steam activation. In summary, ACs with high pore volumes and a variety of pore size distributions can be produced from walnut shells by tailored CO2 or steam activation.
Acknowledgment The authors express their gratitude to the “Junta de Extremadura-Consejerı´a de Economı´a, Comercio e Innovacio´n” and “Almaraz Nuclear Plant” for the financial support through projects 2PRO4B016/PRI07A088 and 169/06, respectively. S.R. thanks the “Junta de Extremadura” for her research grant. J.M.V.N. is grateful to the Fundac¸a˜o para a Cieˆncia e Tecnologia (Portugal) and the European Regional Development Fund (FEDER) for financial support through Project PTDC/CTM/ 66552/2006. Literature Cited (1) Hu, Z.; Vansant, E. F. Synthesis and Characterization of a ControlledMicropore-Size Carbonaceous Adsorbent Produced from Walnut Shell. Micropor. Mat. 1995, 3, 603. (2) Arriagada, R.; Garcı´a, R.; Molina-Sabio, M.; Rodrı´guez-Reinoso, F. Effect of Steam Activation on the Porosity and Chemical Nature of Microporous Carbons. Carbon 1989, 27, 23. (3) Molina-Sabio, M.; Gonza´lez, M. T.; Rodrı´guez-Reinoso, F.; Sepu´lvedaEscribano, A. Effect of Steam and Carbon Dioxide Activation in the Micropore Size Distribution of Activated Carbon. Carbon 1996, 34, 505. (4) Gonza´lez, M. T.; Rodrı´guez-Reinoso, F.; Garcı´a, A. N.; Marcilla, A. CO2 Activation of Olive Stones Carbonized under Different Experimental Conditions. Carbon 1997, 35, 159. (5) Roma´n, S.; Gonza´lez, J. F.; Gonza´lez-Garcı´a, C. M.; Zamora, F. Control of Pore Development during CO2 and Steam Activation of Olive Stones. Fuel Process. Technol. 2008, 89, 715. (6) Rodrı´guez-Reinoso, F.; Molina-Sabio, M.; Gonza´lez, M. T. The Use of Steam and CO2 as Activating Agents in the Preparation of Activated Carbons. Carbon 1995, 33, 15. (7) Toles, C. A.; Marshall, W. E.; Wartelle, L. H.; McAloon, A. Steamor Carbon Dioxide-Activated Carbons from Almond Shells: Physical, Chemical and Adsorptive Properties and Estimated Cost of Production. Bioresour. Technol. 2000, 75, 197. (8) Rodrı´guez-Reinoso, F.; Lo´pez-Gonza´lez, J. D.; Berenguer, C. Activated Carbons from Almond Shells-I. Preparation and Characterization by Nitrogen Adsorption. Carbon 1982, 20, 513. (9) Nabais, J. M. V.; Nunes, P.; Carrott, P. J. M.; Ribeiro-Carrott, M. M. L.; Macı´as-Garcı´a, A.; Dı´az-Dı´ez, M. A. Production of Activated Carbons from Coffee Endocarp by CO2 and Steam Activation. Fuel Process. Technol. 2008, 89, 262. (10) Tomko´w, K.; Siemieniewska, T.; Czechowski, F.; Jankowska, A. Formation of Porous Structures in Activated Brown-Coal Chars Using O2, CO2 and H2O as Activating Agents. Fuel 1977, 56, 121. (11) Walker, P. L., Jr. Production of Activated Carbons: Use of CO2 vs. H2O as Activating Agent. Carbon 1996, 34, 1297. (12) Gonza´lez, J. F.; Encinar, J. M.; Gonza´lez-Garcı´a, C. M.; Sabio, E.; Ramiro, A.; Canito, J. L.; Gan˜an, J. Preparation of Activated Carbons from Used Tyres by Gasification with Steam and Carbon Dioxide. Appl. Surf. Sci. 2006, 252, 5999. (13) Wigmans, T. Industrial Aspects of Production and Use of Activated Carbons. Carbon 1989, 27, 13. (14) Aygu¨n, A.; Yenisov-Karaka¸, S.; Duman, I. Production of Granular Activated Carbon from Fruit Stones and Nutshells and Evaluation of their Physical, Chemical and Adsorption Properties. Microporous Mesoporous Mater. 2003, 66, 189. (15) Kim, J. W.; Sohn, M. H.; Kim, D. S.; Sohn, S. M.; Kwon, Y. S. Production of Granular Activated Carbon from Waste Walnut Shell and its Adsorption Characteristics for Cu2+ ion. J. Hazard. Mater. B 2001, 85, 301. (16) Martı´nez, M. L.; Torres, M. M.; Guzma´n, C. A.; Maestri, D. M. Preparation and Characteristics of Activated Carbon from Olive Stones and Walnut Shells. Ind. Crop. Prod. 2006, 23, 23. (17) Hayashi, J.; Horikawa, T.; Takeda, I.; Muroyama, K.; Nasir-Ani, F. Preparing Activated Carbons from Various Nutshells by Chemical Activation with K2CO3. Carbon 2002, 40, 2381. (18) Ahmedna, M.; Marshall, W. E.; Husseiny, A. A.; Rao, R. M.; Goktepe, I. The Use of Nutshell Carbons in Drinking Water Filters for Removal of Trace Metals. Water Res. 2004, 38, 1062. (19) Schro¨der, E.; Thomauske, K.; Weber, C.; Hornung, A.; Tumiatti, V. J. Experiments on the Generation of Activated Carbon from Biomass. J. Anal. Appl. Pyrol. 2007, 79, 106. (20) Kutics, K.; Kotsis, L.; Argyela´n, J.; Szolcsa´nyi, P. Study of the Adsorption Characteristics and Pore Structure of Activated Carbons. Surf. Coat. Technol. 1985, 25, 87.
Ind. Eng. Chem. Res., Vol. 48, No. 16, 2009 (21) Hulla y Antracita: Determinacio´n de la Humedad Total; Spanish Norm UNE 32001-81, 1981. (22) Combustibles Minerales So´lidos: Determinacio´n del Contenido en Materias Vola´tiles; Spanish Norm UNE 32019-84, 1984. (23) Combustibles Minerales So´lidos: Determinacio´n de Cenizas. Spanish Norm UNE 32004-84, 1984. (24) Robertson, J. B.; Van Soest, P. J. The Analysis of Dietary Fibre in Food; Marcel Dekker: New York, 1981. (25) Daud, W. M. A. W.; Ali, W. S. W. Comparison on Pore Development of Activated Carbon Produced from Palm Shell and Coconut Shell. Bioresour. Technol. 2004, 93, 63. (26) Gergova, K.; Petrov, N.; Eser, S. Adsorption Properties and Microstructure of Activated Carbons Produced from Agricultural byProducts by Steam Pyrolysis. Carbon 1994, 32, 693. (27) Gan˜a´n, J. Aprovechamiento Energe´tico mediante Combustio´n, Pirolisis y Gasificacio´n de los Residuos del Almendro. Ph D. Dissertation, University of Extremadura, Spain, 2004. (28) Brunauer, S.; Emmett, P.; Teller, E. Adsorption of Gases in Multimolecular Layers. J. Am. Chem. Soc. 1938, 60, 309. (29) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Siemieinewska, T. Reporting Physisorption Data for Gas/ Solid Systems with Special Reference to the Determination of Surface Area and Porosity (Recommendations 1984). Pure Appl. Chem. 1985, 57, 603. (30) Rodrı´guez-Reinoso, F.; Martı´n-Martı´nez, J. M.; Prado-Burguete, C.; McEnaney, B. A. Standard Adsorption Isotherm for the Characterization of Activated Carbons. J. Phys. Chem. 1987, 91, 515. (31) Dubinin, M. M. Chemistry and Physics of Carbon; Walker, Jr. P. L., Eds.; Marcel Dekker: New York, 1986. (32) Gregg, S. G.; Sing, K. S. W. Adsorption, Surface Area and Porosity; Academic Press: London, 1992. (33) Medek, J. Possibility of Micropore Analysis of Coal and Coke from the Carbon Dioxide Isotherm. Fuel 1977, 56, 131. (34) Shields, J. E.; Lowell, S. A Method for the Determination of Ambient Temperature Adsorption of Gases on Porous Materials. J. Colloid Interface Sci. 1985, 103, 226.
7481
(35) Leo´n y Leo´n, C. A. New Perspectives in Mercury Porosimetry. AdV. Colloid Interfac. Sci. 1998, 76, 341. (36) Rouquerol, J.; Avnir, D; Fairbridge, C. W.; Everet, D. H.; Haynes, J. H.; Pernicone, N.; Ramsay, J. D. F.; Sing, K. S. W.; Unger, K. K. Recommendations for the Characterization of Porous Solids. Pure Appl. Chem. 1994, 66 (8), 1739. (37) Garrido, J.; Linares-Solano, A.; Martı´n-Martı´nez, J. M.; MolinaSabio, M.; Rodrı´guez-Reinoso, F.; Torregrosa, R. Use of N2 and CO2 in the Characterization of Activated Carbons. Langmuir 1987, 3, 76. (38) Iley, M.; Marsh, H.; Rodrı´guez-Reinoso, F. The Adsorptive Properties of Carbonised Olive Stones. Carbon 1973, 11, 633. (39) Marcilla, A.; Garcı´a-Garcı´a, S.; Asensio, M.; Conesa, J. A. Influence of Thermal Treatment Regime on the Density and Reactivity of Activated Carbons from Almond Shells. Carbon 2000, 38, 429. (40) Carrot, P. J. M.; Freeman, J. J. Evolution of Micropore Structure of Activated Charcoal Cloth. Carbon 1991, 29, 499. (41) Lo´pez-Gonza´lez, J. D.; Martı´nez-Vı´lchez, F.; Rodrı´guez-Reinoso, F. Preparation and Characterization of Active Carbons from Olive Stones. Carbon 1980, 18 (6), 413. (42) Ku¨hl, H.; Kashani-Motlagh, M. M.; Mu¨hlen, H. J.; van Heek, K. H. Controlled Gasification of Different Carbon Materials and Development of Pore Structure. Fuel 1992, 71, 879. (43) Ryu, S. K.; Jin, H.; Gondy, D.; Pusset, N.; Ehrburger, P. E. Activation of Carbon Fibers by Steam and Carbon Dioxide. Carbon 1993, 331, 841. (44) Feng, B.; Bathia, S. K. Variation of the Pore Structure of Coal Chars during Gasification. Carbon 2003, 41, 507. (45) Zawadki, J. Chemistry and Physics of Carbon; Marcel Dekker: New York, 1989. (46) Socrates, G. Infrared Characteristic Group Frequencies; John Wiley & Sons: Chichester, 1980.
ReceiVed for reView December 2, 2008 ReVised manuscript receiVed July 13, 2009 Accepted July 13, 2009 IE801848X
