Removal of Naphthalene, Phenanthrene, and Pyrene by Sorbents

Sorbents from Hot Gas. A. M. MASTRAL,* , † ... It is the first time that the removal of polycyclic aromatic ... (Py)] from combustion hot gas has be...
0 downloads 0 Views 80KB Size
Download PDF
Environ. Sci. Technol. 2001, 35, 2395-2400

Removal of Naphthalene, Phenanthrene, and Pyrene by Sorbents from Hot Gas A . M . M A S T R A L , * , † T . G A R C IÄ A , † M . S . C A L L EÄ N , † M . V . N A V A R R O , † A N D J . G A L B AÄ N ‡ Instituto de Carboquı´mica, CSIC, P.O. Box 589, 50080-Zaragoza, Spain, and Department of Analytical Chemistry, Zaragoza University, 50007-Zaragoza, Spain

It is the first time that the removal of polycyclic aromatic hydrocarbons (PAH) containing different aromatic rings number [naphthalene (Np), phenanthrene (Phe), and pyrene (Py)] from combustion hot gas has been carried out. The aim was to relate the sorbents textural characteristics with the adsorption capacity of these 2-4-ring PAH at the conditions emitted at energy generation. The sorbents textural parameters [total micropore volume (VN2), narrow micropore volume (VCO2), mesopore volume (VBJH), and the free active sites] were analyzed trying to correlate them with their Np, Phe, and Py adsorption capacities. To get this aim, single and multiple linear regressions (MLR) were applied to the three PAH. A principal component analysis was performed to generate new and uncorrelated variables. It enabled us to show that the relations between the textural parameters were analyzed using a principal components regression (PCR). The PCR analysis had a good statistical quality, but neither did it allow differentiating free active site types nor did VN2 and VCO2. The correlations were thus set up applying a MLR to the original variables The regression statistical quality was similar to the PCR analysis, and it could give an easier explanation of the parameters that affected the adsorption. In Np adsorption, the 87% data variance was explained, and the adsorption was positively correlated to VCO2 and the micropore mean diameter (L). In the Phe regression there was 98% variance explained, and its adsorption was positively correlated to the VN2 and the micropore distribution, n. Finally, in the Py adsorption, the 96% data variance was explained, and this adsorption was positively correlated to VN2 and VBJH. These dependencies were according to the molecular parameters of these compounds (molecular diameter and volatility) because the higher the number of aromatic rings of the PAH, the more favored the adsorbateadsorbate interactions. Besides, the higher the mean diameter micropores, the lower the diffusional problems showed by Np, Phe, and Py.

Introduction Currently, the main aim in energy production is centered on the maximum energy generation with the lowest cost fulfilling * Corresponding author e-mail: [email protected]; phone: +34-976-733977; fax: +34-976 733318. † CSIC. ‡ Zaragoza University. 10.1021/es000152u CCC: $20.00 Published on Web 05/04/2001

 2001 American Chemical Society

the emissions regulation. One of the newest technologies used to generate energy is fluidized beds combustion (FBC). It is known that the main advantage of energy production in FBC is associated with inorganic emissions abatement due to the combustion conditions being allowed to control and decrease SOx, NOx, and COx emissions. Nevertheless, organic emissions also constitute an important source of pollution whose concern is based in their harmful character (1). This is the reason these compounds are acquiring a notable concern. However, nowadays there is a lack of information about organic emissions from the new energy systems. Although the organic compounds show a wide variety, a special group constituted by the polycyclic aromatic hydrocarbons (PAHs) is notable. These PAHs are able to modify the normal metabolic functions of cells, originating mutagenic and carcinogenic effects (2-4). These compounds are mainly generated in combustion processes of fossil fuels (5, 6) through mechanisms that can be classified in two processes: pyrolysis (7) and pyrosynthesis (8-11). By heating, the fuels structure is partially cracked (pyrolysis) and devolatilized into smaller, unstable and highly reactive fragments (radicals) with a very short average life. By recombination, these radicals could generate more stable compounds such as PAH (pyrosynthesis). Because of the PAH high volatility and reactivity (12), PAHs can be released and not only supported onto the particulate matter (PM) but also in the gas phase (6, 13). While the more volatile compounds, compounds of 2 and 3 aromatic rings, are mainly released in gas phase, the compounds containing 3 or more aromatic rings in their molecule could be the majority associated with the PM. Therefore, the PAH gas/solid partitioning is associated to the liquid vapor pressure, the size and surface area of the suspended particulate matter, the PM chemical composition, and the ambient temperature (14, 15). These characteristics along with the PAH volatile character will determine the way in which they are emitted from a combustion process. The main advantage of the PAH supported onto the PM is that they can be trapped using the proper systems to collect the PM as cyclones, electrostatic precipitators, scrubbers, etc. On the contrary, the most volatile compounds can be released to the atmosphere by a chimney, and their control is more difficult because of the use of sorbents necessary to capture them. Published work on PAH trapping using activated lignite cokes in waste incineration plants shows that PAH are emitted in the flue gas below the detection limit (16), but the study was performed qualitatively from waste materials incineration instead of quantitatively as it is done in this work. In a previous work (17), the authors demonstrated that the Phe adsorption capacity in different carbon materials depends on the material textural characteristics and shows a good correlation to the total micropore volume (VN2) from the N2 isotherm data calculated by the Dubinin-Radusckevich equation (DR). In this work, the different behavior of Np, Phe, and Py, compounds of 2-4 aromatic rings, and their adsorption capacity onto different adsorbents are studied in concentration intervals normally emitted at energy generation. With the aim of studying the influence of the textural characteristics of carbon materials on their adsorption, a single and multiple variable analysis are carried out in order to try to explain the maximum of the data variance. VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2395

TABLE 1. Active Carbons Textural Parameters DR equation

DA equation

TPD

BJH method

sample

origin

VN2 (cm3/g)

VCO2 (cm3/g)

n

L (nm)

CO (µmol/g)

CO2 (µmol/g)

VBJH (cm3/g)

CA-1 CA-2 CA-3 CA-4 CA-5 CA-6 CA-7 CA-8 CA-9 CA-10 CA-11 CA-12 CA-13 CA-14 CA-15 CA-16

tire tire coal tire cherry grape coal apricot coal coal coconut coal coal apricot coal apricot

0.06 0.04 0.11 0.15 0.13 0.19 0.21 0.46 0.30 0.36 0.57 0.41 0.52 0.52 0.48 0.62

0.03 0.01 0.12 0.08 0.18 0.21 0.19 0.25 0.28 0.28 0.42 0.22 0.27 0.32 0.36 0.29

2.39 2.23 2.06 1.61 2.40 2.32 1.98 2.00 2.04 1.93 2.32 1.54 1.63 1.64 1.72 1.52

0.9 1.1 1.3 1.4 0.7 0.8 1.1 1.6 1.0 1.1 1.1 1.5 1.4 1.4 1.3 1.6

864 277 499 944 1860 1476 1405 839 669 678 681 1134 661 705 612 884

373 121 542 295 874 1009 578 592 353 392 465 344 357 161 321 482

0.130 0.390 0.110 0.390 0.004 0.070 0.060 0.220 0.110 0.260 0.100 0.250 0.430 0.570 0.140 0.510

Experimental Section Sixteen carbon materials of different origin were used to study their Np, Phe, and Py adsorption capacity. CA-1, CA-2, and CA-4 were carbon black from tire pyrolysis. CA-1 and CA-2 were not subjected to any activation process, but CA-4 was activated in order to increase its micropore volume. CA-3 was a coke from German lignite. CA-5, CA-6, and CA-11 were active carbons from cherry stones, grape seeds, and coconut shells, respectively. CA-8, CA-14, and CA-16 were different active carbons from apricot stones. Finally, CA-7, CA-9, CA10, CA-12, CA-13, and CA-15 were different chars from Spanish lignite. As the textural characteristics of these sorbents were the interest of this paper and not their activation protocol, the activation procedures are not commented on. The active carbons used in this work were deeply characterized through their textural parameters, and the data obtained are compiled in Table 1. The 16 porous carbons were characterized by N2 and CO2 adsorption at 77 and 273 K, respectively, using an ASAP 2000 (Micromeritics). Their micropore volumes were obtained applying the DR equation and mesopore volume (VBJH) by applying the BJH method (18) to the N2 isotherm data. Their corresponding micropore distribution and diameter mean size were calculated by applying the Dubinin-Astakhov (DA) equation to the CO2 isotherm data. The surface oxygen groups characterization was carried out on a Pulse Chemisorb 2700 (Micromeritics) by temperature-programmed desorption (TPD). The carrier flow rate was adjusted with an electronic mass flow controller. The gases evolved by thermal decomposition from the surface groups were measured by gas chromatography with Porapak N-120 and molecular sieve columns. The temperature program was as follows: 35 °C isothermal for 4 min and from 35 to 180 °C at 35 °C/min. Experiments were carried out using the laboratory-scale rig shown in Figure 1. The adsorbing bed, composed of 1550 mg of adsorbent (100-200 µm average particle size diameter) depending on its surface, was mixed with 1.0 g of sand, inert material with the same particle size, to provide enough bed length (11 cm) assessing in this way a uniform flow throughout the reactor. Blank tests were carried out to check the inert material adsorption capacity. Before each experiment was started, it was necessary to reach a constant and known concentration of the specific PAH studied in the bed reactor inlet stream by passing the saturator outlet gas stream to the flame ionization detector (FID) until this concentration was obtained. When this was reached, the gas stream was passed through the reactor 2396

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001

FIGURE 1. Experimental system for Phe adsorption: A) thermocouple; B) gas flow (He); C) phenantrene bed; D) purge; E) ceramic furnace; F) thermostatic oven; G) adsorbent bed; H) bypass and I) FID detector. starting the reaction time, which lasted until saturation was reached (C0 ) C). Adsorption capacities, w* (mg of PAH/mg of AC), were calculated as tbC0Q/W, where tb was the breakthrough time for 2% of penetration time (min); C0 was the Phe inlet stream concentration (mg/mL), Q was the flow (mL/min), and W was the adsorbent weight (mg). The PAH concentration (C) in the outlet gas stream was directly measured by a previously calibrated FID.

Results The aim of this work was to look for any correlation between the sorbents textural parameters and their adsorption capacity at the conditions that Np, Phe, and Py are emitted at the chimney from power stations. Therefore, a deep textural characterization of the 16 carbon materials used was performed. Their characteristics are shown in Table 1, whereVN2 is the total micropore volume (pore size < 2 nm), calculated from the DR equation for the N2 isotherm data;

TABLE 2. Molecular Parameters for the Three PAH Studied compd

Vw (cm3/mol)

σ (nm)

Np Phe Py

73.97 107.80 132.33

0.62 0.70 0.75

FIGURE 3. Breakthrough curves for the three polyaromatic compounds for the CA-3 sample: (A) Np, (B) Phe, and (C) Py.

TABLE 3. PAH Adsorption Capacities

FIGURE 2. Np (9), Phe (O) and Py (4) adsorption isotherms at 150 °C. VCO2 (pore size < 0.7 nm) is the narrow micropore volume, calculated from the DR equation for the CO2 isotherm data and the mesopore volume; VBJH is the pore size 2-50 nm, which was calculated from the BJH method for the N2 isotherm data; and the n exponent was obtained from applying the DA equation to the CO2 isotherm data. For most of the carbon adsorbents, n ranged from 1 to 4. An n value higher than 2 means a highly homogeneous, small micropore carbon adsorbent and an n value lower than 2 means a strongly activated and heterogeneous carbon. The L value was obtained at the maximum value of the distribution curve using the Stoeckli equation when the characteristic adsorption energy, from the DA equation to the CO2 isotherm data, ranges from 42 to 20 kJ/mol and using the DR equation when the E0 values were lower. Table 1 also shows the decomposition products from the surface oxygen-containing groups (CO and CO2) that were determined by TPD. The CO2 was released at low temperatures, as a result of the decomposition of acid surface groups. CO has their origin from weakly acidic, neutral, and basic groups (19), which are thermally more stable, and therefore was released at higher temperatures. Np, Phe, and Py molecular parameters (molecular volume and mean diameter) were calculated according to Bondi (20) and are shown in Table 2. The CA-3 sample was chosen to study the behavior of the three PAH at different concentrations on all sorbents adsorption capacities (see Figure 2). The election of the CA-3 sample was due to its lower total micropore volume, which allowed shortening the run time. CA-1 and CA-2 samples were discarded in spite of them showing a lower total micropore volume than CA-3, but their breakthrough point was produced close to the blank, which could lead to erroneous results. In all studied concentration ranges, it could be observed, as shown in Figure 2, that the Np adsorption was lower than the Phe one and that the Phe adsorption was lower than the Py one. This was due to the fact that the three PAH adsorption into the CA-3 sample did not present diffusional problems, according to its n and L values (see Table 1) because they showed a wide micropore distribution with a mean micropore size appropriate to the adsorption of three polyaromatic hydrocarbons studied. Figure 2 reflects that the Np presented a type I isotherm while Ph and Py showed type II isotherms (21) without reaching asymptotic behavior due to capillary condensation in micropores. In Py

sample

Np

Phe

Py

sample

Np

Phe

Py

CA-1 CA-2 CA-3 CA-4 CA-5 CA-6 CA-7 CA-8

1.5 1.5 25.2 21.4 1.9 35.5 45.6 64.7

12.9 8.0 61.7 64.2 11.8 73.8 101.8 199.0

21.0 41.7 71.8 90.5 13.3 61.1 122.8 198.7

CA-9 CA-10 CA-11 CA-12 CA-13 CA-14 CA-15 CA-16

71.2 67.2 106.6 54.4 57.7 63.2 113.6 109.4

161.4 188.8 271.5 233.8 260.4 285.2 274.4 332.2

157.6 260.0 288.3 230.7 332.3 330.3 253.8 384.3

TABLE 4. Relationship between Pore Volume and PAH Adsorption Capacities for the Carbonous Materials regression coeff, r (% explained variance) pore vol

Np

Phe

Py

VN 2 VCO2 VBJH

0.892 (79) 0.893 (80) 0.173 (3)

0.984 (97) 0.848 (71) 0.460 (21)

0.959 (92) 0.785 (61) 0.573 (33)

adsorption, the capillary condensation is more favored, probably due to its lower volatility. Once the CA-3 sample behavior for the three PAH adsorptions was known, a PAH inlet concentration ≈1.5 ppm was put into the bed reactor in order to study the PAH adsorption capacity of the 16 carbon materials. The PAH adsorption capacity was obtained from the breakthrough curves, Figure 3. The high slope for each PAH indicated that the three PAH adsorptions happened very quickly, with minimum mass transfer effect and very fast kinetics (22). However, the different adsorption capacities for each studied hydrocarbon were dependent on their molecular size and volatility. The adsorption capacities of the 16 carbon materials for the three studied PAH are showed in Table 3, where the w*/C0 ratio is given. Because of the experimental design, it was not possible to keep constant the three PAH inlet concentrations in all experiences run, so w* could not be compared. Later on, these values will be compared with the textural characteristics of the carbon materials.

Discussion A single variable analysis was carried out to evaluate the influence of each textural parameter (Table 1) in the three PAH adsorption capacities. In Table 4, the influence of VN2, VCO2, and VBJH on the adsorption capacity of each compound, the regression coefficient parameters (r), and, the explained variance percentages are shown. It was observed that the statistical quality of this regression was only acceptable in VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2397

TABLE 5. Relation between PAH Adsorbed and Surface Groups Evolved during Heat Treatmenta sample

Np/CO

Np/CO2

Np/O*

Phe/CO

Phe/CO2

Phe/O*

Pyr/CO

Pyr/CO2

Pyr/O*

CA-1 CA-2 CA-3 CA-4 CA-5 CA-6 CA-7 CA-8 CA-9 CA-10 CA-11 CA-12 CA-13 CA-14 CA-15 CA-16

0.0 0.1 0.8 0.6 0.0 0.6 0.6 1.7 2.3 2.3 3.8 1.2 3.1 2.4 4.2 2.5

0.1 0.2 0.7 1.8 0.0 0.8 1.6 2.5 4.3 3.9 5.6 3.9 5.7 10.5 8.0 4.7

0.0 0.0 0.2 0.4 0.0 0.2 0.4 0.7 1.1 1.1 1.6 0.7 1.5 1.6 2.1 1.2

0.1 0.3 1.0 0.8 0.0 0.4 0.7 2.4 2.2 2.7 4.7 1.4 3.4 3.1 4.1 4.0

0.2 0.7 0.9 2.5 0.1 0.6 1.6 3.4 4.1 4.6 6.9 4.6 6.3 13.5 7.8 7.4

0.0 0.2 0.3 0.5 0.0 0.2 0.4 1.0 1.0 1.2 2.0 0.9 1.6 2.1 2.0 1.9

0.1 1.0 1.3 1.0 0.1 0.4 0.8 3.1 2.3 3.7 4.5 2.1 5.3 5.5 4.2 5.1

0.3 2.3 1.2 3.1 0.1 0.5 2.0 4.4 4.3 6.4 6.6 6.9 9.9 23.9 8.1 9.4

0.1 0.5 0.4 0.6 0.0 0.1 0.5 1.3 1.1 1.7 1.9 1.3 2.6 3.7 2.1 2.5

a

O*, total oxygen evolved calculated as CO + 2CO2.

TABLE 6. Correlation Coefficients between the Normalized Textural Parameters and the PCs component variable

PC1

CO O CO2 VCO2 VN 2 L n VBJH

0.951 0.925 0.796

PC2

PC3

PC4

0.989 0.890 0.895 -0.764 -0.845

the cases of Phe and Py in relation to VN2, but these values were not high enough, and a multivariable analysis was recommended to increase the model statistical quality. Previous to this analysis, the effect of surface chemistry upon the Np, Phe, and Py retention was studied. The 16 samples studied in this work cover a wide range of surface chemistry because of knowing the influence of surface groups. However, hydrocarbons show low polarity, and the active sites in the sorbents should not have relevant influence in the PAH adsorption. If active sites are considered as those generated in the carbon after the desorption of oxygen groups such as CO and CO2, the results should show that there is not any correlation between the amount of each PAH adsorbed and the amount of CO or CO2 groups desorbed. Results in Table 5 show that the ratios PAH/CO, PAH/CO2, and PAH/O* differed from 1, corroborating that there was poor correlation between the three PAHs adsorbed and the site formed after CO or CO2 desorption. There was also poor correlation to the total number of free sites (CO + 2CO2). Therefore, although the samples showed great differences in their surface chemistry (Table 1), their behavior regarding the three studied PAH adsorptions showed that it was mainly porous texturedependent. To improve the quality of the model, a multivariate analysis to the data obtained was performed (23). A principal component analysis was carried out to try to reduce the number of independent variables by generating new and uncorrelated variables, the principal components (PCs). The analysis was developed using SPSS software, and the eight normalized variables were VN2, VCO2, n, L, CO, CO2, VBJH, and O*. The correlation coefficients (Table 6) between the normalized variables and the PCs were calculated as the product of the square root eigenvalues and scores. The chosen 2398

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001

PCs number was a total of four, which explained more than 95% of the data variance. In Table 7, the eigenvalues and the proportion variance explained for each one of them are compiled. To know the PCs chemical meanings, an equamax rotation to the generated autovector space was carried out. In this new space, PC1 was positively correlated to the total active sites, CO and CO2 desorbed groups, and total number of free sites (CO + 2CO2). PC2 was strongly correlated to VN2 and VCO2, which means that it indicates the total volume microporosity, and PC3 was positively correlated to L and negatively correlated to n. Therefore, PC3 showed the micropore distribution, and finally, PC4 was negatively correlated to the mesopore volume. From this analysis, it was possible to differentiate between the micropore distribution (PC3) and the micropores volume influence (PC2). However, the model did not allow differentiating between the influence of the different superficial oxygenated groups, acid and basic, and the VCO2 and VN2 for the three studied PAH. Once the PCs were known, a multiple linear regression (PCR) was carried out using a stepwise procedure with SPSS software. The best results obtained for each PAH are shown in the following equations:

Np ) 52.54 + 32.73(8.8) × PC2 + 10.93(2.9) × PC3 (1) CA ) 16 r ) 0.932

sE ) 14.4 SNp ) 37.0 % explained variance ) 87

Phe ) 158.18 + 94.21(19.8) × PC2 + 45.44(9.6) × PC3 - 27.97(-5.9) × PC4 - 14.8(-3.1) × PC1 (2) CA ) 16 r ) 0.990

sE ) 18.4 SPhe ) 110/4 % explained variance ) 98

Py ) 178.62 + 97.75(13.32) × PC2 + 47.42(6.5) × PC3 - 45.78(-6.2) × PC4 - 20.61(-2.8) × PC1 (3) CA ) 16 r ) 0.980

sE ) 28.4

SPy)122.1 % explained variance

where CA was the carbon materials number; r was the regression coefficient, which was used to compare the different sized models fit quality (24); sE was the regression standard deviation; and SPAH was the standard deviation associated to the dependent variable. The numbers in parentheses beside the independent variable coefficients represented the calculated Student t test for each coefficient from their standard deviations. The Student t parameters

associated with each variable coefficient were all higher than 2, indicating a significant relation between PCs and PAH adsorption capacities. It could be observed that the % explained variance was higher than the one obtained through the single variable analysis. The next step consisted of the chemical interpretation related to the adsorption process. The Np adsorption was independent of the free active sites according to its hydrophobic nature and to the VBJH value, because at 150 °C the volatility of the molecule avoided the capillary condensation in the mesopores. On the other hand, the Np adsorption was positively correlated to the micropore volume and to the wide micropore distribution. The Phe adsorption was dependent on the four PCs. As could be expected, Phe was positively correlated to the micropore volume, to the mesopore volume, and to the wide micropore distribution with a high mean diameter. However, a negative correlation to the free active sites was observed. It means that the lower the free active sites number, the higher the micropore volume and, therefore, the higher the amount of Phe adsorbed. The Py adsorption showed the same dependence than Phe, but for Py the relative importance of each variable was different. Results where the relative importance of each variable was calculated from the change produced in the r2 value, % variance explained, when a new variable was introduced to the regression are shown in Table 8. From this table, it appears that the micropore volume was the most important variable for the three PAH adsorptions. The micropore distribution also contributed to the Np, Phe, and Py adsorption capacities. A wide distribution with a high mean micropore diameter favored these three PAH adsorptions, avoiding diffusional problems. Moreover, it was observed that the mesopore volume influenced the Phe and Py adsorption. The relative importance of this variable was higher for Py according to its lower volatility. The Phe and Py adsorption were also correlated to free active sites, but their relative importance was very low. The sign of the coefficient was negative, and according to the hydrophobic nature of these molecules, the adsorption was not being produced on the free active sites. The lower the free active sites, the higher the micropore volume in the studied adsorbents. As the influence of some variables, like VCO2 and VN2 or different free active sites, could not be evaluated separately from the PCR analysis, a multiple linear regression (MLR) of normalized variables was carried out in order to study the influence of each one. The results obtained are shown in the following equations:

Np ) 52.54 + 30.71(7.9) × VCO2 + 9.93(2.6) × L (4) CA ) 16

r ) 0.931

sE ) 14.56 SNp ) 37.0 % explained variance ) 87

Phe ) 158.18 + 99.84(19.3) × VN2 - 15.38(-3.0) × n (5) CA ) 16

r ) 0.990

sE ) 16.4 SPhe ) 110.4 % explained variance ) 98

Py ) 178.62 + 106.3(13.7) × VN2 + 26.1(3.37) × VBJH (6) CA ) 16 r ) 0.978

sE ) 27.3 SPy ) 122.1 % explained variance ) 96

The statistical quality of regressions, variance percentages explained from standard deviations and Student t parameters were similar to those introduced in the PCR analysis, but the chemical information obtained from the correlation was more direct.

TABLE 7. Eigenvalues and Associated Variances of the PCs eigenvalue % variance % accumulated

PC1

PC2

PC3

PC4

2.532 31.65 31.65

1.885 23.56 51.21

1.762 22.02 77.24

1.443 18.04 95.27

TABLE 8. Relative Importance of Each Variable in PCR Analysis relative importance (%) variable

Np

Phe

Py

PC1 PC2 PC3 PC4

0.0 90.0 10.0 0.0

1.8 74.3 17.2 6.5

2.9 66.8 15.7 14.7

TABLE 9. Comparison of the Predictive Ability of PCR and MLR for Three PAH observed values

predicted values by PCR

predicted values by MLR

sample

Np

Phe

Py

Np

Phe

Py

Np

Phe

Py

CA-1 CA-2 CA-3 CA-4 CA-5 CA-6 CA-7 CA-8 CA-9 CA-10 CA-11 CA-12 CA-13 CA-14 CA-15 CA-16

2 2 25 21 2 36 46 65 71 67 107 54 58 63 114 109

13 8 62 64 12 74 102 199 161 189 272 234 260 285 274 332

21 42 72 91 13 61 123 199 158 260 288 231 332 330 254 384

-4 -11 28 17 26 40 41 77 59 63 103 65 77 78 92 91

-18 9 64 95 28 68 98 215 164 194 273 210 263 293 268 307

-4 46 56 128 29 66 103 217 181 223 286 232 304 356 285 350

-9 -7 29 22 24 35 41 75 61 65 102 63 73 86 94 86

4 1 46 88 39 74 101 227 144 180 268 224 276 276 251 333

21 49 45 109 40 83 92 252 149 204 294 230 317 337 251 383

The Np adsorption was positively correlated to the VCO2 and L,although the explained variance percentage was still low. This could be due to the Np high volatility at 150 °C, and its adsorption would be produced in pores of molecular dimensions in the range dp ) σ-2σ. Therefore the adsorption was carried out partially at the VN2 and partially at the VCO2 micropores. About the L value, the Np adsorption was favored by high values, which did not imply diffusional problems. The Phe adsorption was positively correlated to the VN2 and negatively to the n value. The regression quality was good because the VN2 includes the all micropore range where the adsorption was produced. This agreed with the Phe molecular diameter and its volatility, which allowed the cooperative process in the wider micropores (dp ) 2σ-5σ) and involved the adsorbate-adsorbate interactions. With respect to the n value, the appropriate carbon materials for the Phe adsorption were those with a wide micropore distribution and strongly activated carbons. The Py adsorption was positively correlated to the VN2 and VBJH, and it was a consequence of the Py diameter and its low volatility. In addition to the micropore filling and the cooperative process in the wider micropores, the capillary condensation into the mesopores was also produced. Table 9 shows the experimental results as compared to the predictable values by the model. In most cases, there was a good concordance between them, except for CA-1, CA-2 (only for Np), and CA-5 samples. In the case of low microporosity samples (CA-1 and CA-2), the values obtained with the model were lower than the expected ones. For the VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2399

narrow microporosity sample (CA-5), the values were overestimated with the model. Anyway, CA-1, CA-2, and CA-5 samples did not see to be proper for these three PAH abatements from hot gas because their adsorption was close to the blank. The elimination of these samples from the model was assessed, but the results obtained implied a decrease in the regression quality, and for this reason, the model obtained is not showed in this paper. Summarizing, the Np, Phe, and Py removal from hot gas by sorbents has been undertaken at the concentration interval they could be emitted from energy generation. The Np, Phe, and Py adsorption capacities of the studied sorbents have been related with their textural parameters by single and multiple linear regressions. The principal components regression analysis was applied and, despite its good statistical quality, did not allow a simple results interpretation. The multiple linear regression, with similar statistical quality to the former, showed that: (i) The Np adsorption is directly related to the adsorbent narrow micropore volume and to the diameter mean size (L); a proper carbon material for Np adsorption will be the one showing a high micropore volume with a micropore diameter large enough to avoid diffusional effects. (ii) The Phe adsorption is directly related to the adsorbent total micropore volume and the micropore distribution (n), so that the Phe adsorption will be favored by high micropore volume sorbents with a wide micropore distribution in order to show a cooperative effect between Phe molecules, promoting the Phe elimination. (iii) The Py adsorption is directly related to the adsorbent total micropore volume and to the volume contributed by the 2-50 nm sized pores (VBJH). Therefore, a proper carbon material aimed to Py elimination should show both a high micropore volume and a high mesopore volume where the capilar condensation could happen. In general, it could be concluded that, for the three studied PAHs, the higher the micropore volume of the carbon material, the higher their adsorption capacities if the micropore size and distribution were adequate. Besides, the higher the number of the aromatic rings in the polyaromatic hydrocarbon molecule, the more favored the adsorbateadsorbate interactions.

Acknowledgments Authors would like to thank the European Community for its partial support and to the DGA (Arago´n, Spain) for its fellowship (T.G.). Also we would like to thank Prof. Dr. Linares Solano’s group for the CA-7, CA-9, CA-10, CA-11, CA-12, C-13, and CA-15 samples supply.

Literature Cited (1) Adams, D. F. Separation sampling for aerosol and gaseous semivolatile organics. An overview and examle prceedings for the

2400

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001

(2)

(3)

(4)

(5) (6) (7) (8) (9) (10) (11)

(12)

(13) (14) (15) (16) (17) (18) (19) (20) (21) (22)

(23) (24)

1987 EPA/APCA Symposium on measurement of toxic and reated ar pllutants; Research Triangle Park: NC, 1987. 2. Polynuclear Aromatic Compounds, Part 1; International Agency for Research on Cancer (IARC), Chemical Environmental and Experimental Data: Lyon, France, 1983. 3.Evaluation and Estimation of Potential Carcinogenic Risks of PAH: Carcinogen Assessment Group; Office of Health and Environmental Assessment, Office of Research and Development, U.S. EPA: Washington, DC, 1985. 4.Scientific Criteria Document for Multimedia Standards Development PAH, Part 1: Hazard Identification and DoseResponse; Ministry of Environment and Energy: Ontario, 1997. Mastral, A. M.; Calle´n, M. S.; Murillo, R.; Garcı´a, T. Environ. Sci. Technol. 1999, 33 (23), 4155-4158. Mastral, A. M.; Calle´n, M. S.; Murillo, R.; Garcı´a, T. Environ. Sci. Technol. 1999, 33, 3177-3184. Mastral, A. M.; Calle´n, M. S.; Murillo, R. Fuel 1996, 75 (13), 1533. Bonfanti, L.; DeMichelle, G.; Riccardi, J.; Lopez-Doriga, E. Combust. Sci. Technol. 1994, 101, 505. Williams, P. T.; Andrews, G. E.; Bartle, K. D. Fuel 1986, 65, 1150. Barbella, R.; Bertoli, C.; Ciajolo, A.; D’Anna, A. Combust. Flame 1990, 82, 191. Bjorseth, A.; Ramdahl, T. Handbook of PAH, Vol. 2: Emission Sources and Recent Progress in Analytical Chemistry; Marcel Dekker: New York, 1985. Sloss, L. L.; Gardner, C. A. Sampling and analysis of trace emissions from coal-fired power stations; IEA Coal Research, IEACR/77; London, 1995; p 48. Grace Lee, W.-M.; Tong, H.-C.; Yeh, S.-Y. J. Environ. Sci. Health 1993, A28 (3), 563-583. Butler, J. D.; Crossley, P. Atmos. Environ. 1981, 15, 91-94. Yamasaki, H.; Kuwate, K.; Miyamoto, H. Environ. Sci. Technol. 1982, 16, 189-194. Grodten, T.; Schmidt, D.; Dannecker, W. Gefahrstoffe-Reinhalt. Luft 1998, 58, 205-209. Mastral, A. M.; Garcı´a, T.; Calle´n, M. S.; Navarro, M. V. Energy Fuels 2001, 15, 1-7. Barret, E. P.; Joyner, L. G:; Halenda, P. P. J. Am. Chem. Soc. 1951, 73, 373-380. Allen, J. O.; Paschkewitz, J. S.; Plummer, E. F.; Lafleur, A. L.; Sarofim, A. F.; Smith, K. A. Aerosol Sci. Technol. 1999, 30, 16-29. Bondi, A. J. Phys. Chem. 1964, 68, 441-451. Brunauer, S. The Physical Adsorption of Gases and Vapors; Clarendon Press: Oxford, England, 1945. Vermeulen, T.; LeVan, M. D.; Hiester, N. K.; Klein, G. In Perry’s Chemical Engineers Handbook; Perry, R. H., Green, D. W., Maloney, J. A., Eds.; McGraw-Hill: Tokyo, 1984. Brasquet, B.; Bourges, B.; Le Cloirec, P. Environ. Sci. Technol. 1999, 33, 4226-4231. Nirmalakhandan, N. N.; Speece, R. E. Environ. Sci. Technol. 1990, 24, 575-580.

Received for review July 10, 2000. Revised manuscript received January 19, 2001. Accepted March 12, 2001. ES000152U