Article pubs.acs.org/IECR
Volatile Organic Compounds Removal from Gas Streams by Adsorption onto Activated Carbon Tănase Dobre, Oana C. Pârvulescu,* Gustav Iavorschi, Marta Stroescu, and Anicuta̧ Stoica Chemical and Biochemical Engineering Department, Politehnica University of Bucharest, 1-3 Gheorghe Polizu, 011061, Bucharest, Romania ABSTRACT: This paper’s aim was to model the adsorption of VOC (volatile organic compound) vapor from gas streams onto fixed bed porous adsorbent particles under various operation conditions. A mathematical model based on gas plug flow was developed to predict the process dynamics. An experimental study to highlight the influence of process variables on adsorption dynamics, described by bed saturation curves, was performed in order to identify the more relevant model parameters for the laboratory scale fixed bed adsorption column used. Experimental data concerning the adsorption of n-hexane and 2-propanol from air streams onto fixed bed granular activated carbon emphasized an increase in saturation adsorption capacity of activated carbon at high values of operation temperature and species boiling point as well as at a low level of air superficial velocity. The model predicted well the real conditions, and it could facilitate the design, scale-up, and operation of fixed bed adsorption columns.
■
gas stream onto fixed bed adsorbent particles involves the following steps: (i) mass transfer of species from the gas phase to the adsorbent particle surface (convection, axial dispersion, diffusion through a gas film surrounding the particle), (ii) mass transfer of species inside the particle pores (pore diffusion), and (iii) physical or chemical adsorption/desorption of species on the pore wall surface (surface reaction). When the adsorption and desorption rates became equal, the overall adsorption rate is zero and the equilibrium (saturation) state is achieved. Equilibrium (saturation) adsorption capacity and adsorption process kinetics depend on (i) characteristics of the adsorbent particle (size, shape, density, specific surface area, porosity, pore size distribution) and of the fixed bed particles (void fraction, density, height), (ii) characteristics of the adsorbed species (chemical composition, molar mass, boiling point, vapor pressure, molecular size, initial concentration), and (iii) operation conditions (temperature, gas flow rate, humidity). An increase in saturation adsorption capacity of activated carbon with molar mass and boiling point of VOC species was reported.5,7,8,12 A more rapid saturation of fixed bed activated carbon and an augmentation of saturation adsorption capacity were noticed at larger values of gas flow rate.12 Some studies assessed an increase in the amount of VOC species adsorbed onto activated carbon with operation temperature, indicating that the adsorption phenomenon is dominated by chemosorption.12,18 In order to scale-up the VOCs removal from gas streams by adsorption onto fixed bed adsorbent particles, various models have been developed. Considering that the process is controlled by external diffusion (in the gas film surrounding the particle), internal diffusion (inside the particle pores), and surface reaction
INTRODUCTION Volatile organic compounds (VOCs) are pollutants generated by chemical, pharmaceutical, food, and electronic material industries as well as by motor vehicles, power plants, solvent use, etc. They are usually characterized by boiling temperatures less than or equal to 250 °C measured at a standard atmospheric pressure of 101.325 kPa and by high vapor pressures, exceeding 0.5 kPa at 25 °C.1−4 Due to their large vapor pressure values, they evaporate easily under normal conditions and enter the atmosphere. According to the definition given by WHO (World Health Organization), VOCs are organic compounds whose boiling points range from 50−100 °C to 240−260 °C. These compounds are precursors of photochemical oxidants and some of them are very harmful for both human health and the environment, even at very low concentration. Therefore, the diminishing of VOCs concentration in the ambient air, in order to maintain and improve its quality, is an important task. There are various methods to treat VOCs gaseous streams, e.g., condensation, absorption, adsorption, thermal, catalytic, or photocatalytic oxidation.1−18 Among them, the adsorption has been recognized as an effective and regenerative technique for VOCs emissions control, especially at low concentrations.5,9,10 Likewise, it is nondenaturing, highly selective, energy efficient, and relatively inexpensive. Adsorbents are natural or synthetic materials with amorphous or microcrystalline structure, e.g., activated carbon, zeolites, alumina, silica gel, and synthetic polymers, which are available as granules, extruded pellets, fibers, powder, etc.5,9,17 Activated carbon is the most used adsorbent due to its selectivity, highly developed porous structure, large specific surface area, good mechanical properties, biocompatibility, chemical stability, low cost, and great accessibility. It is produced from a wide variety of carbonaceous precursors such as coal (lignite, anthracite, bituminous coal), wood, and also various agricultural and forest byproducts.4,19−22 Fixed beds of activated carbon are usually used to remove various VOCs species from air flows.1−5,8−10,12,17 Generally, the adsorption process of one or more solutes from a © 2014 American Chemical Society
Received: Revised: Accepted: Published: 3622
August 1, 2013 February 4, 2014 February 11, 2014 February 11, 2014 dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
Industrial & Engineering Chemistry Research
Article
of n-hexane is greater than that of 2-propanol due to boiling point and vapor pressure. Atmospheric air was employed as carrier gas for VOCs vapor.
(on the pore wall surface), characteristic equations of mathematical model are as follows: (i) mass and heat balance for the bulk gas phase in the fixed bed, (ii) mass and heat balance inside the adsorbent particle pores, and (iii) adsorption/ desorption rate.23 These equations contain specific equilibrium and kinetic parameters which should be estimated from suitable correlations by fitting experimental data determined in independent batch and dynamic studies. To avoid time-consuming and expensive experiments as well as to reduce the number of equations, restrictions, and parameters, various simplifying assumptions can be adopted, e.g., ideal plug or perfectly mixed flow of fluid phase, negligible pore diffusion resistance, linear adsorption kinetics, and isothermal conditions.3,6,8−11,13,15,16,23−26 Generally, ideal plug (no axial dispersion) or perfectly mixed flow is considered at low values of fixed bed height and high values of fluid superficial velocity, whereas the pore diffusion resistance is neglected for adsorbent particles containing pores whose diameter is less than the adsorbed molecule size or for fine particles with large pores. This paper has aimed to model the adsorption of n-hexane and 2-propanol from air streams onto granular activated carbon in a laboratory scale fixed bed column. A removal of n-hexane and 2propanol from atmospheric air is of interest due to the adverse health effects of these VOCs at concentration levels larger than the regulated limit values, which can appear during their production, storage, and use. Accordingly, both VOCs species are precursors of photochemical oxidants which are recognized respiratory toxicants.27,28 Moreover, n-hexane is listed as a Hazardous Air Pollutant (HAP) in the Clean Air Act Amendments of 1990, being a suspected reproductive, respiratory, and nervous system toxicant.27,29,30 A mathematical model based on plug flow of the gas phase was developed to predict the process dynamics. An experimental study to highlight the influence of process variables, i.e., air superficial velocity, operation temperature, and VOC species properties, on bed saturation curves was performed in order to identify the model parameters.
Table 2. Physical Properties of VOC Species
EXPERIMENTAL SECTION Materials. Coal-based granular activated carbon, Calgon CPG LF 12 × 40, supplied by Chemviron Carbon Corporation, was used as adsorbent. Prior to each experiment, the adsorbent was dried for 12 h at 120 °C to remove all adsorbed gases and moisture content. The characteristics of activated carbon used for the VOCs removal study are summarized in Table 1. VOCs species consisted of n-hexane and 2-propanol of analytical purity, purchased from Sigma Aldrich Corporation. These were used without further purification, and their physical properties, measured under normal conditions of temperature and pressure, i.e., 20 °C and 101 325 Pa, are listed in Table 2.27,28 The volatility Table 1. Physical Properties of Activated Carbon Granule
melting point (°C)
boiling point (°C)
vapor pressure at 20 °C (kPa)
n-hexane 2-propanol
86.18 60.10
0.660 0.785
−95 −89
69 82
16 4
■
0.14 cm 0.85 g/cm3 0.64 1.1 × 107 cm2/g
RESULTS AND DISCUSSION Characteristic Experimental Data of Adsorption Equilibrium and Dynamics. Experimental saturation curves of adsorbent fixed bed, expressed as adsorption capacity of activated carbon for i species, qi (g/g), i.e., adsorbed species mass divided by adsorbent bed mass, versus time, τ (s), under various operation conditions, are illustrated in Figure 2. All experiments
Fixed Bed mass, mb density, ρb void fraction, εb
density (g/cm3)
Experimental Setup and Procedure. The laboratory setup employed to study VOCs adsorption onto activated carbon is shown in Figure 1. The setup consists of two main sections, i.e., a gas−VOC vapor mixture preparation section and a heating− adsorption section. In the section of gas−VOC vapor mixture preparation, the carrier gas (atmospheric air) was first fed by a compressor (1) into a silica gel column (2) in order to remove the humidity. The air stream, whose flow rate was controlled by a valve (3) and measured by a flow-meter (4), was further bubbled into the liquid VOC species (pure n-hexane or 2-propanol) from a bubbler (5). The heating−adsorption section mainly consisted of the coil heater (6), adsorption column (7), fins heater (8), and fan (9) which were fitted in an enclosure with polystyrene isolating walls (10). The mixture of air and VOC vapor from the bubbler (5) was heated in the coil heater (6), fed into the bottom of adsorption column (7), and up-flowed through the adsorbent fixed bed until the bed was saturated. Air temperature inside the isolating enclosure was raised and maintained at a constant value by the fins heater (8). The fan (9) has intensified the thermal transfer between air and fins, as well as provided a uniform distribution of temperature in the system. The thermal agent in the coil and fins heaters was warm water fed by a centrifugal pump (11) from a thermostatic bath (12). Air temperature inside the enclosure was measured by a digital thermometer (13). Twenty-eight grams of granular activated carbon were packed into the glass column (7), dc = 1.7 cm internal diameter and Hc = 29 cm height, which was set into a plastic support (14) and put on an OHAUS Adventurer Pro AV412 digital balance (15), 0.01 accuracy. The amount of adsorbed VOC was estimated based on the increase of adsorption column mass, which was continuously measured by the digital balance (15) and registered by a computer (16). It was considered that the saturation state was achieved when the column mass did not vary in time for 10 min. The bubbler containing the liquid i VOC species was weighed before and after each adsorption experiment in order to determine the VOC mass which was picked-up by the carrier gas, mi = mi0 − mifin. Experimental Variables. The experimental study was performed for two VOCs species (n-hexane and 2-propanol) at two values of air superficial velocity, w (0.7 and 1.7 cm/s), and three values of operation temperature, t (25 °C, 30 °C, and 40 °C). All experiments were carried out at atmospheric pressure.
■
mean diameter, dP density, ρP porosity, εP specific surface area, σ
i species
molar mass (g/mol)
28 g 0.51 g/cm3 0.40 3623
dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
Industrial & Engineering Chemistry Research
Article
Figure 1. Experimental setup: 1, compressor; 2, silica gel column; 3, valve; 4, flow-meter; 5, bubbler; 6, coil heater; 7, adsorption column; 8, fins heater; 9, fan; 10, polystyrene isolating layer; 11, centrifugal pump; 12, thermostatic bath; 13, digital thermometer; 14, plastic support; 15, digital balance; 16, computer.
Figure 2. Experimental saturation curves of fixed bed granular activated carbon for n-hexane (a) and 2-propanol (b) adsorption: ◆ w = 0.7 cm/s, t = 25 °C; ■ w = 0.7 cm/s, t = 30 °C; ▲ w = 0.7 cm/s, t = 40 °C; ◇ w = 1.7 cm/s, t = 25 °C; □ w = 1.7 cm/s, t = 30 °C; △ w = 1.7 cm/s, t = 40 °C.
Table 3. Saturation Data of VOC Species Adsorption onto Fixed Bed Granular Activated Carbon exp.
i species
1 2 3 4 5 6 7 8 9 10 11 12
n-hexane
2-propanol
air superficial velocity
operation temperature
saturation adsorption capacity for i species
saturation concentration of i species in gas phase
w (cm/s)
t (°C)
qi∞ (g/g)
ci∞ = ci0 (g/cm3)
0.7 1.7 0.7 1.7 0.7 1.7 0.7 1.7 0.7 1.7 0.7 1.7
25
0.244 0.203 0.260 0.217 0.299 0.246 0.357 0.307 0.379 0.315 0.418 0.346
1.310 × 10−3 1.045 × 10−3 1.494 × 10−3 1.075 × 10−3 1.442 × 10−3 1.128 × 10−3 0.248 × 10−3 0.165 × 10−3 0.238 × 10−3 0.159 × 10−3 0.218 × 10−3 0.148 × 10−3
30 40 25 30 40
initial concentration of i species in gas phase, ci∞ = ci0 (g/cm3), evaluated based on VOC mass picked-up by the carrier gas, mi (g), volumetric flow rate of carrier gas, GV (cm3/s), and final operation time, τfin (s):
were replicated three times and values of relative standard dispersion less 5% were obtained, indicating a good reproducibility. Table 3 contains values of saturation (equilibrium) adsorption capacity, qi∞ (g/g), as well as values of saturation/ 3624
dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
Industrial & Engineering Chemistry Research
Article
Table 4. Estimated Values of Model Parameters exp.
i species
1 2 3 4 5 6 7 8 9 10 11 12
n-hexane
2-propanol
ci ∞ = ci0 =
air superficial velocity
operation temperature
adsorption constant
desorption constant
root mean squared error
equilibrium constant
w (cm/s)
t (°C)
kai (s−1)
kdi × 104 (s−1)
δ
Ki × 10−3
0.7 1.7 0.7 1.7 0.7 1.7 0.7 1.7 0.7 1.7 0.7 1.7
25
0.13 0.41 0.15 0.45 0.19 0.52 0.17 0.31 0.21 0.37 0.34 0.50
4.1 12.4 4.5 12.8 4.7 13.9 0.19 0.45 0.21 0.46 0.24 0.47
0.025 0.016 0.018 0.006 0.012 0.014 0.030 0.018 0.037 0.020 0.046 0.020
0.325 0.331 0.333 0.352 0.404 0.374 8.947 6.889 10.000 8.044 14.167 10.638
mi m − mi fin = i0 πd 2 GV τfin wτfin 4c
30 40 25 30 40
(adsorption) rate, Ra (g/cm3s), from backward reaction (desorption) rate, Rd (g/cm3s); • adsorption rate varies linearly with respect to free surface site ratio, 1 − qi/Qi, and species concentration in gas phase, ci (g/cm3), i.e., Ra = kaiεb(1 − qi/Qi)ci, where kai (s−1) is adsorption rate constant and Qi (g/g) maximum adsorption capacity at monolayer level; • desorption rate varies linearly with respect to species adsorption capacity, qi (g/g), i.e., Rd = kdiρbqi, where kdi (s−1) is desorption rate constant; • the system operates under isothermal conditions. Characteristic mathematical model of fixed bed adsorption of i species from gas phase is described by the following system of equations and restrictions: • conservation equation of i species in bulk gas phase in the fixed bed, where x (cm) is axial distance and Dl (cm2/s) axial dispersion coefficient:
(1)
Characteristic saturation data of VOC species adsorption summarized in Table 3 highlight a decrease in initial concentration of i species in the gas phase at high values of air superficial velocity and species boiling point as well as an increase in saturation adsorption capacity for i species at large values of operation temperature and species boiling point and a low level of air superficial velocity. Values of saturation adsorption capacity ranging between 0.20 g/g and 0.42 g/g were obtained, according to data reported in other studies, i.e., from 0.20 g/g to 0.50 g/ g.8,12 The plots in Figure 2 emphasize that the adsorption capacity of i species increases almost linearly in time until the saturation value is attained. The saturation time decreases with gas superficial velocity and increases with species boiling point. Steeper saturation curves and lower values of saturation adsorption capacity are obtained at a high value of air superficial velocity as well as for the compound with a greater volatility and a lower boiling point (n-hexane). The results reveal that, at larger values of operation temperature, the process kinetics is faster and saturation adsorption capacity has higher values, indicating the chemosorption as a prevalent phenomenon. The experimental saturation curves are similar to those reported in the related literature referring to VOCs adsorption from gas streams onto fixed bed activated carbon.12 Modeling of Fixed Bed Adsorption of VOCs Species from Gas Stream. The model selected to predict VOC species adsorption from a gas stream onto fixed bed granular activated carbon has been based on the following simplifying assumptions: • plug flow with axial dispersion of gas phase occurs along the column; • activated carbon granule structure is predominately microporous (σ = 1.1 × 107 cm2/g), consisting of very narrow micropores, which are not accessible to VOC species, as well as of mesopores, whose mean diameter is much larger than VOC molecule size; • a monolayer adsorption of VOC molecules occurs onto the surface of adsorbent particles and mesopore walls; • pore diffusion resistance is negligible (fine adsorbent particles with very narrow micropores and large mesopores); • overall adsorption rate of molecular species depends on the competition between adsorption and desorption; i.e., it is equal to the subtraction of forward reaction
εb
∂q ∂ci ∂c ∂ 2c + w i − εbDl 2i + ρb i = 0 ∂τ ∂τ ∂x ∂x
(2)
• overall adsorption rate equation of i species: ρb
⎛ q ⎞ = kaiεb⎜⎜1 − i ⎟⎟ci − kdiρb qi ∂τ Qi⎠ ⎝
∂qi
(3)
• relationship for mean adsorption capacity at time τ of i species, where H (cm) is total height of fixed bed: qi , mn(τ ) =
1 H
∫0
H
qi(x , τ ) dx
(4)
• initial conditions: τ = 0, 0 < x ≤ H , ci = 0, τ = 0, x = 0,
qi = 0
ci = ci0 , qi = 0
(5)
• boundary conditions: τ > 0, x = 0,
∂ci(0, τ ) w = (ci0 − ci) ∂x Dl
τ > 0, x = H ,
∂ci(H , τ ) =0 ∂x
(6)
The system of eqs 2−6 was numerically solved by an adequate finite difference method. Values of mean adsorption capacity, qi,mn(τ), were calculated depending on values of known 3625
dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
Industrial & Engineering Chemistry Research
Article
Table 5. Estimated Values of Arrhenius Parameters in Equation 10 i species
w (cm/s)
Aai × 10−3 (s−1)
Eai (kcal/mol)
δArr,a
Adi × 103 (s−1)
Edi (kcal/mol)
δArr,d
0.7 1.7 0.7 1.7
0.34 0.06 3.60 0.28
4.66 2.91 8.63 4.02
0.006 0.005 0.013 0.003
9.15 13.83 0.23 0.13
1.84 1.43 2.84 0.30
0.029 0.004 0.009 0.002
n-hexane 2-propanol
parameters, i.e., H = 23.5 cm, εb = 0.51, ρb = 0.4 g/cm3, Dl, ci0, and Qi. Because the axial number Péclet for gases passing through packed beds of spherical particles was found to be about 2, the axial dispersion coefficient was calculated as follows: Dl = (wdP)/ 2 = 0.07w (cm2/s).23 Characteristic values of maximum adsorption capacity at monolayer level, Qn‑hexane = 1.04 g/g and Q2‑propanol = 0.54 g/g, were estimated based on equilibrium experimental data according to the linearized form of the Langmuir isotherm eq 10, where KLi is the Langmuir equilibrium constant. These values of Qi are in a good agreement with those determined in other research concerning the modeling of gas− solid equilibrium using the Langmuir equation, e.g., 0.460 g/g for C6H6, 0.600 g/g for CCl4, and 0.297 g/g for CHCl3.8 ρb 1 1 = + qi ∞ Qi εbci ∞Q iKLi (7)
and the slope of the straight line given by a plot of ln ka/di versus 1/T. Estimated values of Arrhenius parameters in eq 10 depending on VOC species and air superficial velocity are listed in Table 5. Arrhenius parameter values for the adsorption process, determined with a root mean squared error δArr,a ≤ 0.013, decrease with superficial velocity and increase with species boiling point. The values of adsorption activation energy, Eai, are in a good agreement with those estimated in other studies regarding the modeling of VOCs species adsorption from gas streams onto fixed bed activated carbon, e.g., 4.0 kcal/mol for C6H6, 4.8 kcal/mol for CCl4, and 5.2 kcal/mol for CHCl3.8 Arrhenius parameters values for the desorption process regressed with a root mean squared error δArr,d ≤ 0.029 are lower than characteristic values of the adsorption process. The values of desorption activation energy, Edi, decreases with superficial velocity, whereas the values of desorption pre-exponential factor, Adi, are significantly larger for species with a higher boiling point, i.e., 2-propanol. Substituting Arrhenius eq 10 in correlation 9, the equilibrium constant can be written as follows: A K i = ai e−(Eai − Edi)/ RT = Ai e−Ei / RT Adi (11)
Minimizing the root mean squared error, δ, described by relation 8, where qi,exp(j) represents experimental adsorption capacity of activated carbon for i species at dimensionless time j = τ/Δτ (j = 0, 1, ..., M), Δτ = τfin/M (s) is the time step, and M a positive integer, the values of kinetic parameters kai and kdi were identified.
The values of the pre-exponential factor, Ai = Aai/Adi, and activation energy, Ei = Eai − Edi, obtained from the intercept and the slope of the straight line given by a plot of ln Ki versus 1/T, are summarized in Table 6. As can be seen, the values of the
M
δ(kai , kdi) =
∑ j = 0 [qi , mn(j) − qi ,exp(j)]2 M+1
(8)
Estimated values of adsorption rate coefficient, kai, and desorption rate coefficient, kdi, under various operation conditions, as well as of root of mean squared errors, δ, are listed in Table 4. As expected, the values of both kinetic parameters increase with operation temperature. Additionally, results summarized in Table 4 reveal that the values of kai and kdi are higher at the large value of air superficial velocity and the values of kdi are significantly lower than those of kai, indicating that the rate of forward reaction (adsorption) is much faster than that of backward reaction (desorption). Accordingly, the adsorption reaction can be considered irreversible, which corresponds to a chemosorption process. Table 4 contains also the values of equilibrium constant, Ki, defined as rate coefficients ratio, i.e.:
Ki =
kai kdi
Table 6. Estimated Values of Arrhenius Parameters in Equation 11 i species n-hexane 2-propanol
Ai × 10−5
Ei (kcal/mol)
δArr
0.7 1.7 0.7 1.7
0.37 0.04 155.94 20.99
2.82 1.48 5.79 3.73
0.023 0.009 0.022 0.001
Arrhenius parameters, determined with a high accuracy (δArr ≤ 0.023), decrease with superficial velocity and increase with species boiling point. The values of Ei are consistent with those estimated in other research, e.g., 8.2 kcal/mol for C6H6, 10.4 kcal/mol for CCl4, and 10.6 kcal/mol for CHCl3.8 Experimental and predicted saturation curves under various operation conditions, compared in Figure 3, are in a good agreement (δ ≤ 0.02). Deviations between experimental and calculated data are probably due to the assumptions adopted, e.g., negligible pore diffusion resistance, linear adsorption, and desorption kinetics.
(9)
An increasing variation of equilibrium constant, Ki, with operation temperature, t, and boiling point is highlighted by the results listed in Table 4. The rate coefficients can be expressed by the Arrhenius equation, where Aa/di (s−1) is adsorption/ desorption pre-exponential factor, Ea/di (cal/mol) adsorption/ desorption activation energy, T (K) absolute temperature, and R ideal gas constant (R = 1.987 cal/(mol K)): ka / di = Aa / di e−Ea/di / RT
w (cm/s)
■
CONCLUSIONS In this study, a mathematical model based on plug flow of the gas phase, negligible pore diffusion resistance, and isothermal conditions was developed to predict the dynamics of VOCs adsorption from gas streams onto fixed bed adsorbent. An experimental study to assess the influence of process variables on adsorption dynamics, described by bed saturation curves, was
(10)
The values of pre-exponential factors, Aai and Adi, as well as of activation energies, Eai and Edi, were obtained from the intercept 3626
dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
Industrial & Engineering Chemistry Research
Article
Figure 3. Experimental (points) and predicted (lines) saturation curves of fixed bed granular activated carbon for n-hexane (a) and 2-propanol (b) adsorption at w = 1.7 cm/s: ◇ t = 25 °C; □ t = 30 °C; △ t = 40 °C. (6) Bhatia, S.; Abdullah, A. Z.; Wong, C. T. Adsorption of butyl acetate in air over silver-loaded Y and ZSM-5 zeolites: Experimental and modelling studies. J. Hazard. Mater. 2009, 163 (1), 73−81. (7) Cal, M. P.; Larson, S. M.; Rood, M. J. Experimental and modeled results describing the adsorption of acetone and benzene onto activated carbon fibers. Environ. Prog. 1994, 13 (1), 26−30. (8) Chuang, C. L.; Chiang, P. C.; Chang, E. E. Modeling VOCs adsorption onto activated carbon. Chemosphere 2003, 53, 17−27. (9) Das, D.; Gaur, V.; Verma, N. Removal of volatile organic compound by activated carbon fiber. Carbon 2004, 42, 2949−2962. (10) Dwivedi, P.; Gaur, V.; Sharma, A.; Verma, N. Comparative study of removal of volatile organic compounds by cryogenic condensation and adsorption by activated carbon fiber. Sep. Purif. Technol. 2004, 39, 23−37. (11) Joly, A.; Perrard, A. Linear driving force models for dynamic adsorption of volatile organic compound traces by porous adsorbent beds. Math. Comput. Simul. 2009, 79 (12), 3492−3499. (12) Kawasaki, N.; Kinoshita, H.; Oue, T.; Nakamura, T.; Tanada, S. Study on adsorption kinetic of aromatic hydrocarbons onto activated carbon in gaseous flow method. J. Colloid Interface Sci. 2004, 275, 40−43. (13) Kolade, M. A.; Kogelbauer, A.; Alpay, E. Adsorptive reactor technology for VOC abatement. Chem. Eng. Sci. 2009, 64 (6), 1167− 1177. (14) Mo, J.; Zhang, Y.; Xu, Q.; Lamson, J. J.; Zhao, R. Photocatalytic purification of volatile organic compounds in indoor air: A literature review. Atmos. Environ. 2009, 43, 2229−2246. (15) Valdes-Solis, T.; Linders, M. J. G.; Kapteijn, F.; Marban, G.; Fuertes, A. B. Adsorption and breakthrough performance of carboncoated ceramic monoliths at low concentration of n-butane. Chem. Eng. Sci. 2004, 59 (13), 2791−2800. (16) Won, D. Y.; Sander, D. M.; Shaw, C. Y.; Corsi, R. L.; Olson, D. A. Validation of the surface sink model for sorptive interactions between VOC’s and indoor materials. Atmos. Environ. 2001, 35, 4479−4488. (17) Wu, J. Modeling adsorption of organic compounds on activated carbon-a multivariate approach. Ph.D. Thesis, Umea University, Sweden, 2004. (18) Zeinali, F.; Ghoreyshi, A. A.; Najafpour, G. D. Adsorption of dichloromethane from aqueous phase using granular activated carbon: isotherm and breakthrough curve measurements. Middle East J. Sci. Res. 2010, 5 (4), 191−198. (19) Aygun, A.; Yenisoy-Karakas, 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, 189−195. (20) Baquero, M. C.; Giraldo, L.; Moreno, J. C.; Suárez-García, F.; Martínez-Alonso, A.; Tascón, J. M. D. Activated carbons by pyrolysis of coffee bean husks in presence of phosphoric acid. J. Anal. Appl. Pyrol. 2003, 70 (2), 779−784.
conducted in order to identify the model parameters. Adsorption of n-hexane and 2-propanol vapor from air streams onto a commercial coal-based granular activated carbon was investigated. The results emphasized an increase in saturation adsorption capacity of activated carbon for VOC species at high values of operation temperature and species boiling point as well as at a low level of air superficial velocity. An increase in adsorbed species uptake with operation temperature indicates the chemosorption as a prevalent phenomenon. The model predicted well the real conditions and could facilitate the design, scale-up, and operation of fixed bed adsorption columns. It can be applied to a wider range of fixed bed granular adsorbents and VOCs, e.g., (i) adsorbents with a microporous structure and VOCs species with an equivalent spherical diameter larger than the micropore size, (ii) fine granules with large pores and small VOCs molecules, and (iii) nonporous granules along with any VOC species.
■
AUTHOR INFORMATION
Corresponding Author
*(O.C.P.) Tel.: +4021 402 38 10. Fax: +4021 241 58 18. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors are thankful to Mr. Jiang Zhengkun for his help in the experimental part and data processing.
■
REFERENCES
(1) Gupta, K. N.; Rao, N. J.; Agarwal, G. K. Adsorption of toluene on granular activated carbon. Int. J. Chem. Eng. Appl. 2011, 2 (5), 310−313. (2) Gupta, K. N.; Rao, N. J.; Agarwal, G. K. Adsorption of xylene on granular activated carbon in a packed bed. Int. J. Sci. Technol. Res. 2012, 1 (3), 90−93. (3) Gupta, K. N.; Rao, N. J.; Agarwal, G. K. Vapor phase adsorption of xylene on granular activated carbon-experimental and theoretical breakthrough curves. Indian J. Chem. Technol. 2013, 20, 26−32. (4) Ramalingam, S. G. Adsorption of volatile organic compounds and regeneration of activated carbon. Ph.D. Thesis, University Nantes Angers Le Mans, France, 2012. (5) Bhargavi, R.; Kadirvelu, K.; Kumar, N. S. Vapor phase adsorption of homologous aliphatic ketones on activated spherical carbon. Int. J. Environ. Sci. 2011, 1 (5), 938−947. 3627
dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
Industrial & Engineering Chemistry Research
Article
(21) El-Hendawy, A. A. Surface and adsorptive properties of carbons prepared from biomass. Appl. Surf. Sci. 2005, 252, 287−295. (22) Girgis, B. S.; El-Hendawy, A. A. Porosity development in activated carbons obtained from date pits under chemical activation with phosphoric acid. Microporous Mesoporous Mater. 2002, 52, 105−117. (23) Dobre, T.; Sanchez Marcano, J. Chemical engineering - modelling, simulation and similitude; Wiley: New York, 2007. (24) Aris, R. Interpretation of sorption and diffusion data in porous solids. Ind. Eng. Chem. Fundam. 1983, 22, 150−151. (25) Clausse, M.; Bonjour, J.; Meunier, F. Adsorption of gas mixtures in TSA adsorbers under various heat removal conditions. Chem. Eng. Sci. 2004, 59 (17), 3657−3670. (26) Do, D.; Rice, R. G. On the relative importance of pore and surface diffusion in non-equilibrium adsorption rate processes. Chem. Eng. Sci. 1987, 42 (10), 2269−2284. (27) Alberta Environment; Cantox Environmental Inc. Assessment report on hexane for developing an ambient air quality objectives; Alberta Environment: Edmonton, Alberta, Canada, 2004. (28) Alberta Environment; Cantox Environmental Inc. Assessment report on isopropanol for developing an ambient air quality objectives; Alberta Environment: Edmonton, Alberta, Canada, 2004. (29) Agency for Toxic Substances and Disease Registry (ATSDR). Toxicological profile for n-hexane; U.S. Department of Health and Human Services, Public Health Service: Atlanta, GA, 1999. (30) U.S. Environmental Protection Agency (EPA). Health effects notebook for hazardous air pollutants; U.S. EPA, Office of Air Quality Planning and Standards, Air Risk Information Support Center: Durham, NC, 1994.
3628
dx.doi.org/10.1021/ie402504u | Ind. Eng. Chem. Res. 2014, 53, 3622−3628
