Adsorption of Polycyclic Aromatic Hydrocarbons from Heavy


Adsorption of Polycyclic Aromatic Hydrocarbons from Heavy...

7 downloads 163 Views 1MB Size

Article pubs.acs.org/IECR

Adsorption of Polycyclic Aromatic Hydrocarbons from Heavy Naphthenic Oil Using Commercial Activated Carbons. 1. FluidParticle Studies F. Murilo T. Luna, A. Nilson Oliveira Filho, Caio C. B. Araújo,# Diana C. S. Azevedo, and Celio L. Cavalcante, Jr.* Departamento de Engenharia Química, Grupo de Pesquisa em Separações por Adsorçaõ , Núcleo de Pesquisas em Lubrificantes, Universidade Federal do Ceará, Campus do Pici, Bl. 709, Fortaleza, CE 60.455-900, Brazil S Supporting Information *

ABSTRACT: In this study, the adsorption of polyaromatic hydrocarbons (PAHs) from heavy naphthenic oils (HNOs) on commercial activated carbons was evaluated. Among the aromatics, polycyclic aromatic hydrocarbons (PAHs) are usually present in heavy oils and are contaminants that have potential carcinogenic and mutagenic characteristics. Oils with of PAH concentration less than 3 wt % are labeled as noncarcinogenic. The HNO samples used in this study had a high content of PAHs (ca. 8 wt %). Batch experiments were carried out to evaluate the adsorption capacity of activated carbons for PAHs (>160 mg/g), and to estimate the mass transfer parameters of the adsorption process. The pore diffusion model and the pore−surface diffusion model were applied to interpret and validate the kinetic experimental results. The results indicated the efficiency of activated carbons as potential adsorbent to adjust the polycyclic aromatics content of HNO.

1. INTRODUCTION Aromatic compounds are present in several types of hydrocarbon streams, and their concentration is highly dependent on the original crude oil and its refinery processing. These streams may be used as feedstocks or solvents, according to their aromatics nature and concentration. Polycyclic aromatic hydrocarbons (PAH) present in these heavy streams may cause environmental problems since they have proven carcinogenic and mutagenic effects.1−5 These characteristics led to the study of alternatives for mitigating the concentration level of these compounds, particularly in streams derived from oil feedstocks that tend to cause environmental impacts. PAH compounds are known to resist biological degradation and are difficult to be removed from feedstock streams by conventional physicochemical methods, such as coagulation, flocculation, sedimentation, filtration, or ozonation.6,7 However, adsorption processes may be effective for removing persistent organic pollutants, particularly with activated carbon as sorbent.8−13 Activated carbons can be used to adsorb almost any organic compound. Some benefits, like low energetic demand, easy regeneration, their wide availability and possible use both in liquid or gas phase adsorption made these materials an interesting field of research.10 In general, the adsorption rate and capacity are a function of the nature of the molecules to be © XXXX American Chemical Society

adsorbed and the specific carbon material. Nonpolar compounds are normally retained due to London dispersion forces; on the other hand, polar compounds adsorption is normally related to particular interactions through oxygenate groups in the surface of the material. Several studies about the adsorption of organic compounds present in diluted aqueous solutions have been performed using activated carbons. In particular, the adsorption of phenol is one of the most studied systems due to its industrial and environmental relevance.10,14−24 Despite being largely used for environmental remediation, mechanisms of adsorption in activation carbons are yet not well-defined.10 The textural characteristics of activated carbons may elucidate the different adsorption capacity of these materials. Because of the numerous sources and treatments, activated carbons can be prepared so that they present adequate pores and surface chemical structure and may be specifically designed for separating PAHs. The adsorption of naphthalene in different carbon based materials has been studied by Ania et al.10 showing strong dependence with pore size distribution, microporosity, and their hydrophobic nature. Received: March 17, 2016 Revised: June 22, 2016 Accepted: June 28, 2016

A

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

2.3.2. Total Aromatics by FTIR. Measurements of total aromatics concentrations were performed using an infrared spectrophotometer Fourier transform FX-3000 (BIORAD, USA), equipped with a DTGS detector and a KBr beam splitter. Details on this method may be found in Luna et al.33 2.3.3. Polyaromatics Content. The determination of the mass percentage of polycyclic aromatic hydrocarbons in the samples was performed following the method IP 346.34 This method is normally used for measurements of PAH content. It is very arduous needing a high consumption of solvents in a gravimetric procedure wherein there is a dilution of the oil sample in cyclohexane followed by an extraction with DMSO. Successive extractions with solvents (cyclohexane and DMSO) are needed to separate PAH from the oil samples. After several extraction steps, the solvent is evaporated and the residue, mainly composed of polyaromatics, is quantified in relation to the initially used mass of oil. 2.3.4. Polyaromatics Content Using DMSO Extraction Followed by FTIR. For a rapid evaluation of the PAHs content during the adsorption experiments, a recently developed method33 was used. This method was proposed to reduce the measurement time and the amount of samples and solvents. It consists of initial extraction stages, followed by FTIR measurements. Samples of approximately 0.2 g of oil were submitted to an extraction step with DMSO pre-equilibrated with cyclohexane and subsequently analyzed in the FTIR equipment for PAHs quantification. Details of the method may be found in Luna et al.33 2.4. Adsorption Equilibrium and Kinetic Studies. Adsorption equilibrium experiments were carried out in batch by placing a fixed amount of oil (10 g) in contact with different amounts of adsorbent (0.2, 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 g) in a thermostatic bath at 30 °C under stirring. After reaching equilibrium, the oil was separated by vacuum filtration and the total aromatic and polyaromatic contents were determined. The equilibration time was estimated from kinetic experiments. The same procedure was used to obtain the kinetic curves (10 g of HNO and 3.0 g of adsorbent at 30 °C) by sampling the liquid phase every 10 min. Assuming that nonaromatics are negligibly adsorbed on activated carbons, the concentration in the solid phase can be calculated through a mass balance (eq 1), expressed in (mg) of sorbate/(g) of adsorbent.

To predict the performance of commercial adsorption processes, several models based on intraparticle mass transfer resistances have been reported. These include the pore diffusion model,25,26 the surface diffusion model26−28 and dual-resistance (pore and surface) model.26,29,30 In fact, adsorption processes may involve several simultaneous mass transfer resistances. In this study, the removal of PAHs from samples of heavy naphthenic oils (HNO) was evaluated using commercial activated carbons. That oil feedstock is a vacuum industrial distillate with viscosities varying between 380 and 420 cSt (at 40 °C), obtained from different types of Brazilian crude petroleum and mainly applied to special waxes and lubricants formulation. Without proper treatment, HNO presents a high PAHs concentration. Hence, it is important to reduce this concentration in order to encompass regulatory requisitions for commercial applications.

2. EXPERIMENT AND MODELING 2.1. Materials. The heavy naphthenic oil (HNO) sample was kindly provided by Petrobras (Brazil). Its physicochemical properties are shown in Table S1 (Supporting Information). For PAHs extraction, sodium chloride (J.T. Baker, USA), cyclohexane (>99 wt % Merck, USA), and dimethyl sulfoxide, DMSO, (>99 wt %, J.T. Baker, USA) were used. Deionized water was obtained using a Milli-Q system (Millipore, USA). Naphthalene solutions (>98 wt %, Acros Organics, USA) were used at different concentrations in cyclohexane for the calibration of the total aromatics method. Two granular activated carbons used in this study were 830W (AC1) and 1240 Plus (AC2), both provided by Norit (Netherlands). Prior to use in the experiments, the activated carbons were washed with deionized water and then thermally regenerated, first at 40 °C for 1 h, and then heated to 120 °C at 30 °C/h. After 2 h at 120 °C, the samples were cooled down to 25 °C under vacuum and immediately used in the experiments. 2.2. Textural Properties of Activated Carbons. Nitrogen isotherms at 77 K were measured using an Autosorb-1 MP (Quantachrome, USA). The BET methodology was employed to calculate the specific surface area, and the Dubinin− Radushkevich (DR) equation was used to determine the micropore volume, following procedures reported in Rouquerol et al.31 The total pore volume was estimated from the adsorbed volume of nitrogen at relative pressure of 0.95. The mesopore volume was calculated by the difference between the total pore volume and the micropore volume. The pore size distribution was obtained using the Density Functional Theory (DFT) method. 2.3. Aromatics Content in HNO Samples. 2.3.1. Carbon Distribution Method n-d-M-ASTM D 3238.32 This method estimates the distribution of paraffinic, naphthenic, and aromatic carbons in mineral oils, from refractive index measurements, specific gravity, and estimated molecular weight. Molecular weight can be estimated taking into account measurements of kinematic viscosities at various temperatures, as described in ASTM D2502.32 These results are then used for estimating the percentage distribution of aromatic, paraffinic, and naphthenic carbons (% CA, % CP and % CN, respectively) in a model molecule. If the samples show a significant amount of sulfur (>0.1 wt %), correlations with correction for total sulfur content using X-ray fluorescence must be used as presented in method ASTM D3238.32

q* =

Msol(Ci − Ceq) Mads(1 − Ceq)

(1)

where Ci is the initial concentration of aromatics (mg/g), Ceq is the final concentration of aromatics (mg/g), Msol is the mass of HNO (g), Mads is the mass of adsorbent (g), and q* is the concentration of adsorbate on the solid phase (mg/g of ads.). Langmuir and Toth eqs (eqs 2a and 2b, respectively) were used to fit the equilibrium experimental data for latter calculations in modeling the batch kinetics experiments. kCeq q* = qm 1 + kCeq

(2a)

Ceq q* = qm (b + Ceq υ)1/ υ

(2b)

where q* is the concentration of aromatics in the solid phase, Ceq is the concentration of adsorbate in the fluid in equilibrium, B

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research k is the Langmuir constant, qm is the maximum adsorption capacity, b is a parameter related to the affinity with the adsorbent, and υ is a parameter related to the degree of heterogeneity of the surface. 2.5. Diffusion Models and Parameters Estimations. The estimation of the mass transfer parameters in the batch adsorption kinetics was carried out using the Pore Diffusion Model (PDM) and the Pore-Surface Diffusion Model (PSDM).26 Both models assume a spherical particle of adsorbent in a finite bath containing a component to be adsorbed, where Cb is bulk concentration, Cp is the intraparticle liquid phase concentration, and q is a function of Cp calculated from the isotherm equation. The PDM equations may be described as shown in eqs 3 and 6 (solid phase and liquid phase balances), using appropriate initial and boundary conditions (eqs 4, 5, and 7): Mass balance in particle εp

∂Cp ∂t

+

⎛ ∂ 2C ∂Cp ⎞ ∂q p ⎟ ρp = εpDp⎜⎜ 2 + 2 r ∂r ⎟⎠ ∂t ⎝ ∂r

Mass balance in liquid phase ∂C b 3M =− k f,b(C − Cp|r = R p ) ∂t R p·ρp ·V

Initial condition

t = 0,

Cp = 0,

C = C0

q = f (Cp)

(13)

A tortuosity factor (τ), correlating the molecular bulk diffusivity with the pore diffusivity obtained from the experiments, was estimated using eq 14.26 τ=

Dmεp Dp

(14) 2

where Dm is the molecular diffusion coefficient (cm /min) estimated using the Wilke−Chang equation,35 as shown in eq 15.

(3)

q=0

(4)

Dm = 4.44·10−6

Boundary conditions r = 0, r = R p,

∂Cp ∂r

=0

εpDp,b

∂Cp ∂r

(5b)

Mass balance in liquid phase ∂C b 3M =− k f,b(C − Cp|r = R p ) ∂t R p·ρp ·V

(6)

Initial conditions

t = 0,

C = C0

(7)

where εp is the porosity of the particles, Dp is the pore diffusion coefficient, kf,b is the film mass transfer coefficient, t is the time, and r is the radial coordinate. The Pore-Surface Diffusion Model may be described as shown in eqs 8 and 11, using appropriate initial and boundary conditions (eqs 9, 10 and 12), and Ds is the surface diffusion coefficient: Mass balance in particle ρp ∂ ⎡ ∂Cp ⎤ ∂Cp ∂q ⎤ ∂q 1 ∂⎡ + εp ρp = 2 ⎢Dp ·r 2 ⎥ + 2 ⎢Ds ·r 2 ⎥ ⎣ ∂r ⎦ r ∂r ∂r ⎦ ∂t ∂t r ∂r ⎣

RMSE =

Initial conditions Cp = 0,

q=0

(9)

r = R p,

∂Cp ∂r

Dp

=0

∂Cp ∂r

+ ρp Ds

(10a)

∂q = k f,b(C − Cp) ∂r

1 . N

N

∑ (Csim − Cexp)2 i=1

(16)

3. RESULTS AND DISCUSSION 3.1. Aromatics Content in the HNO Sample. Carbon distribution using ASTM D3238 is presented in Table 1, in terms of aromatic, naphthenic, and paraffinic carbon content. For this analysis, correlations for sulfur content (see Supporting Information, Table S1) were used as described in the ASTM D3238 method. Being an empirical method, based on typical

Boundary conditions r = 0,

(15)

where Csim is the concentration estimated from the model, Cexp is the experimental concentration, and N is the number of measured values. A RMSE = 0 indicates a perfect model prediction of the experimental data.

(8)

t = 0,

(ϕ MM)1/2 T ηV b0.6

where MM is the molar mass of HNO, η is the oil viscosity, T is the temperature, Vb is the molar volume of the diffusing molecule at the boiling temperature, and ϕ is an association coefficient, assumed to be 1.0 for aromatics. The value of Vb (213.8 cm3/mol) was obtained by LeBas volumes as described in Poling et al.,35 using pyrene as a model molecule. An estimation of the external mass transfer coefficient (kf,b) was used as proposed by Furusawa and Smith25 for liquid batch adsorption. The system of algebraic and partial differential equations (PDM and PSDM), with the respective initial and boundary conditions was implemented according to the standard gPROMS syntax. The domains were discretized using the method of orthogonal collocation on finite elements (OCFEM) with six sections and three placing spots per section. The estimative of pore and surface diffusion coefficients were performed using the heteroscedastic estimation method included in the computational package gPROMS.36 The root mean squared error (RMSE) was used to evaluate the fitting of the model to the experimental data, as shown in eq 16.

(5a)

= k f,b(C − Cp)

(12)

For both models, instantaneous equilibrium is assumed between the concentration in the fluid phase within the particles (Cp) and the concentration in the solid adsorbent (q),

Initial condition t = 0,

(11)

(10b) C

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Table 1. Carbon Distribution for HNO According to ASTM D3238 type of carbon

distribution (%)

aromatic naphthenic paraffinic

20.7 45.8 33.5

structures and hypothetical assumptions, these values are merely a broad indication of the composition of the oil in terms of aromatic carbon. For further estimation of the total aromatics content in the HNO samples, a FTIR method was applied. The total aromatics content of the HNO samples was thus calculated as 21.7 wt % using this method, which is broadly consistent with the obtained value using method ASTM D3238 for aromatic carbon distribution. To assess the polyaromatics content, analyses were performed according to the IP 346 method34 and PAH extraction with DMSO followed by quantification using FTIR.33 The average results from three independent measurements for these two methods were 8.2% and 7.3%, respectively. As seen, the values obtained with these two methods are within less than 10% difference and hence the DMSO/FTIR method, as a quicker analytical measurement, was selected to evaluate the subsequent kinetic experiments. The European Union37 recommends using the IP 346 for rating the oil streams because the PAH content relates directly to the products carcinogenicity. Oils with PAH contents lower than 3 wt % are rated as noncarcinogenic. As reported, the sample presented polyaromatic content above the established, strongly suggesting the need for using treatment processes to obtain products with low content of PAHs and therefore less prone to cause environmental impacts. 3.2. Textural Properties of Activated Carbons. Nitrogen isotherms at 77 K for our samples are shown in Figure 1a. Differences in porosity can be clearly noted between the two samples. The obtained isotherms can be classified as type I of the BDDT classification, typical of microporous carbon materials.38 The textural parameters of these materials are reported in Table S2 (Supporting Information). AC1 presents higher values of surface area and pore volumes if compared to AC2. Differences in micropore sizes can be more detailed using the DFT method (Figure 1b). Both activated carbon samples present wide pore size distributions (considerable volume of pores between 1.2 and 3.5 nm). A predominant unimodal distribution is observed for AC1 with the majority of pores between 1.2 and 2.4 nm. In this case, PAH retention could be favored by the overlapping of the adsorption potentials, as previously reported.39 On the other hand, a predominant bimodal distribution is observed for AC2, with peaks in 1.5 and 3.5 nm pore sizes. 3.3. Equilibrium Studies. The equilibrium experimental data for the batch adsorption experiments with activated carbons, in terms of polycyclic aromatics, are shown in Figure 2. Equilibrium curves were fitted for each system according to the Langmuir (Figure 2a) and Toth (Figure 2b) equations, with parameters listed in Table 2. The results indicate that the Toth equation shows better agreement with the experimental data of the equilibrium behavior of HPA onto the activated carbons. The Toth

Figure 1. (a) N2 adsorption (solid symbols)/desorption (open symbols) isotherms at 77 K: (□) AC1; (○) AC2. (b) Pore size distribution obtained by the DFT method. () AC1; (---) AC2.

Figure 2. Equilibrium adsorption isotherms of PAH from HNO samples on activated carbons at 30 °C. (○) AC1; (□) AC2; (a) Langmuir fit; (b) Toth fit.

equation is empirically modified from the Langmuir equation and normally best suited to multilayer adsorption.40 D

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Table 2. Parameters of the Langmuir and Toth Equations for PAH Adsorption on Activated Carbons at 30 °C Langmuir equation

Toth equation

adsorbents

k

qm (mg/g)

R2

b

qm (mg/g)

υ

R2

AC1 AC2

0.092 0.062

186.2 165.5

0.95 0.93

1.957 3.255

220.3 203.4

0.59 0.63

0.98 0.97

To compare the behavior of the two samples of activated carbon used in this study, an adsorption selectivity (αPAH/A) between polyaromatics (PAH) and single aromatics (A) was defined as αPAH/A =

Figure 4. Kinetic adsorption experiments of HNO on activated carbons at 30 °C in terms of PAH concentration decay (C0 = 73 mg/ g): (○) AC1; (□) AC2; pore diffusion model (PDM) using Langmuir equilibrium equation (dotted lines) and Toth equilibrium equation (full line).

qPAH /(qTA − qPAH) C PAH/(C TA − C PAH)

(17)

where αPAH/A is the selectivity of PAH with respect to single aromatics compounds, qPAH is the concentration of PAHs in the solid phase, calculated using a mass balance, as shown in eq 1, for the PAH concentrations in the liquid phase (CPAH); qTA is the concentration of total aromatics in the solid phase; CPAH and CTA are the concentrations of PAHs and total aromatics in the fluid in equilibrium, respectively. The values obtained for this selectivity are plotted in Figure 3 versus adsorption loading.

Figure 5. Kinetic adsorption experiments of HNO on activated carbons at 30 °C in terms of PAH concentration decay (C0 = 73 mg/ g). (○) AC1; (□) AC2; Lines are representation of the Pore-Surface diffusion model (PSDM) using Toth equilibrium equation.

Table 3. Estimated Mass-Transfer Parameters Using PDM

Figure 3. Equilibrium selectivities between polyaromatics (PAH) and single aromatics (A) for activated carbons at 30 °C: (○) AC1; (□) AC2.

It may be seen that, in general, selectivities decrease with increasing loading for both carbon samples. Also, it may be noted higher selectivities for AC1 in all loadings when compared to AC2 (10.1 and 3.7, respectively, at the higher loading). 3.4. Adsorption Kinetics Studies. Batch kinetics experiments, showing the concentration decay (C/Co) versus time, are presented in Figures 4 and 5. The Pore Diffusion Model was first used to estimate pore diffusion coefficients and a tortuosity factor (eq 15) was calculated as shown in Table 3. Both equilibrium equations (Langmuir and Toth) were evaluated in this model, and again lower values of RMSE (eq 16) were observed when using the Toth equation. It may be seen in Table 3 that very high tortuosities were estimated for both activated carbons (≫ 10). Normally tortuosity factors of 2 to 10 should be found in systems in porous adsorbents.26,41 Despite the apparent good agreement observed between the experimental values and the pore diffusion model simulation (Figure 4), it is clearly physically inconsistent and should not be used for real systems calculations. Such high calculated

adsorbents

equilibrium model

kf,b (cm/min)

Dp (cm2/min)

tortuosity factor

AC1

Langmuir

1.38

8.41·10−8

158

AC2

Langmuir

1.62

3.50·10−8

341

AC1

Toth

1.38

8.10·10−8

164

AC2

Toth

1.62

3.32·10−8

359

RMSE 2.5· 10−3 1.7· 10−3 1.1· 10−3 0.9· 10−3

tortuosity values may only be explained if an additional mass transfer resistance is taking place at the same time during the adsorption process. With this observation, a surface diffusion step was added to the simulation model, as described in the modeling section. To simulate the experimental data with the PSDM, a tortuosity factor of 5 was used as input for the pore diffusion step, and thus the surface diffusitvity (Ds) could be estimated using the model described in eqs 8−15 to fit the experimental data (see Figure 5 and Table S3, Supporting Information). A comparison between our estimates for the surface diffusivities and previously reported values for pure aromatic compounds (benzene, naphthalene, phenantrene, and pyrene) in synthetic mixtures is depicted in Figure 6. It may be seen that diffusivities decrease generally with increasing number of aromatic rings in the sorbates. The diffusivity values estimated E

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

on activated carbons. Furthermore, an evaluation of the diffusivity values reported in the open literature for synthetic mixtures versus the number of aromatic rings in the pure components indicates that the HNO sample used in this study would probably present PAHs with predominantly 3−4 aromatic rings. Finally, the oil treatment showed a significant decrease in the PAH content (from 8.2 to 3 wt % using AC1).



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.6b01059. Physicochemical properties of HNO; textural properties of activated carbons; surface diffusion coefficients estimated from PSDM compared to literature data for similar conditions (PDF)

Figure 6. Surface diffusion coefficients estimated from PSDM compared to literature data for similar conditions: (○) Furusawa and Smith;25 (△) Luna et al.;42 (□) Ahn et al.43



in this study, for a real mixture of PAHs in a complex medium (HNO), follow nicely the observed trend, especially in the range between 3 and 4 aromatic rings. 3.5. Evaluation of Product Following the Batch Adsorption Treatment. The evaluated physicochemical properties of the HNO samples after the adsorption treatment with activated carbons showed no significant changes, as shown in Table 4. However, the PAH content (IP346 method) and after treatment for both materials, was significantly reduced (50−60 wt %), when compared to the untreated oil sample.

Corresponding Author

*Tel.: +55-85-3366-9611. Fax: +55-85-3366-9601. E-mail: [email protected]. Present Address

# C.C.B.A.: IFP Training Middle East, Manama Centre, Entrance 4, Office 506, Diplomatic Area, Manama, PO Box 10813, Kingdom of Bahrain.

Notes

The authors declare no competing financial interest.



Table 4. Physicochemical Properties and PAH Content of HNO after Treatment

ACKNOWLEDGMENTS The authors wish to thank financial and logistic support provided by PETROBRAS (Petróleo Brasileiro S.A.) and CNPq (Conselho Nacional de Pesquisa e Desenvolvimento ́ Cientifico). The authors are also grateful to NORIT Activated Carbon (Netherlands) for providing samples.

HNO treated properties

HNO fresh

AC1

AC2

specific gravity at 20 °C, g/cm3 kinematic viscosity at 40 °C, cSt refractive index at 20 °C sulfur content (FRX), wt % PAH content (IP-346), wt %

0.940 410.3 1.521 0.75 8.2

0.938 404.9 1.516 0.18 3.0

0.940 407.0 1.518 0.23 4.2

AUTHOR INFORMATION



4. CONCLUSIONS A sample of heavy naphthenic oil (HNO) containing ca. 8 wt % of polyaromatic hydrocarbons (PAHs) was treated in an experimental batch adsorption system in liquid phase with two samples of commercial activated carbons. The equilibrium adsorption results indicated adsorption capacities between 160 and 220 mg PAH/g of adsorbent. An equilibrium selectivity for the PAHs adsorption with respect to single aromatics showed decreasing values with increasing adsorption loadings and varied between 10 and 16 for sample AC1, and 3−6 for sample AC2. Subsequently, kinetics experiments were performed and evaluated using the Pore Diffusion Model (PDM). Despite the good agreement that was observed between experimental data and the fitting of the model, the estimated pore tortuosities were physically inconsistent (≫10), with estimates well above the usual values observed for macropore diffusion controlled systems. So a dual pore diffusion model (PSDM, pore surface diffusion model) was applied to evaluate the kinetic data fitting the experimental values using the surface diffusivity as adjustable parameter. The estimated surface diffusivities (ca. 10−9 cm2/min) fell within the range of previously reported values for pure polyaromatic compounds in synthetic mixtures F

SYMBOLS b = Toth equilibrium constant C = bulk liquid phase concentration (mg·g−1) C0 = initial concentration (mg·g−1) Ceq = equilibrium concentration (mg·g−1) Cexp = experimental concentration (mg·g−1) Cp = intraparticle liquid phase concentration (mg·g−1) CPAH = concentration of PAHs in equilibrium (mg·g−1) Csim = simulated concentration (mg·g−1) CTA = concentration of total aromatics in equilibrium (mg· g−1) Dm = molecular diffusion coefficient (cm2·min−1) Dp = Pore diffusion coefficient (cm2·min−1) DS = surface diffusion coefficient (cm2·min−1) k = Langmuir equation parameter (g·mg−1) kf,b = film mass transfer coefficient (cm·min−1) MM = molecular weight (g·mol−1) q = solid phase concentration (mg·g−1) qm = maximum adsorption capacity (mg·g−1) qPAH = concentration of PAHs in the solid phase (mg·g−1) qTA = concentration of total aromatics in the solid phase (mg·g−1) r = radial coordinate (cm) Rp = average particle radius (cm) t = time (min) T = absolute temperature (K) DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research u = superficial velocity (cm·min−1) Vb = molar volume at normal boiling point (cm3·g mol−1)

(17) Fierro, V.; Torné-Fernández, V.; Montané, D.; Celzard, A. Adsorption of phenol onto activated carbons having different textural and surface properties. Microporous Mesoporous Mater. 2008, 111, 276. (18) Terzyk, A. P. Further insights into the role of carbon surface functionalities in the mechanism of phenol adsorption. J. Colloid Interface Sci. 2003, 268, 301. (19) Girods, P.; Dufour, A.; Fierro, V.; Rogaume, Y.; Rogaume, C.; Zoulalian, A.; Celzard, A. Activated carbons prepared from wood particleboard wastes: Characterisation and phenol adsorption capacities. J. Hazard. Mater. 2009, 166, 491. (20) Singh, K. P.; Malik, A.; Sinha, S.; Ojha, P. Liquid-phase adsorption of phenols using activated carbons derived from agricultural waste material. J. Hazard. Mater. 2008, 150, 626. (21) Tancredi, N.; Medero, N.; Möller, F.; Píriz, J.; Plada, C.; Cordero, T. Phenol adsorption onto powdered and granular activated carbon, prepared from Eucalyptus wood. J. Colloid Interface Sci. 2004, 279, 357. (22) Park, K. H.; Balathanigaimani, M. S.; Shim, W. G.; Lee, J. W.; Moon, H. Adsorption characteristics of phenol on novel corn grainbased activated carbons. Microporous Mesoporous Mater. 2010, 127, 1. (23) Mohd Din, A. T.; Hameed, B. H.; Ahmad, A. L. Batch adsorption of phenol onto physiochemical-activated coconut shell. J. Hazard. Mater. 2009, 161, 1522. (24) Jung, M. W.; Ahn, K. H.; Lee, Y.; Kim, K. P.; Rhee, J. S.; Park, J. T.; Paeng, K. J. Adsorption characteristics of phenol and chlorophenols on granular activated carbons (GAC). Microchem. J. 2001, 70, 123. (25) Furusawa, T.; Smith, J. M. Fluid-particle and intraparticle mass transport rates in slurries. Ind. Eng. Chem. Fundam. 1973, 12, 197. (26) Ruthven, D. M. Principles of Adsorption and Adsorption Processes; John Wiley & Sons: New York, 1984. (27) Hand, D. W.; Crittenden, J. C.; Thacker, W. E. User orientated batch reactor solutions to the homogeneous surface diffusion model. J. Environ. Eng. 1983, 109, 82. (28) Traegner, U. K.; Suidan, M. T. Evaluation of surface and film diffusion coefficients for carbon adsorption. Water Res. 1988, 23, 267. (29) Do, D. D.; Rice, R. G. On the relative importance of pore and surface diffusion in nonequilibrium adsorption rate processes. Chem. Eng. Sci. 1987, 42, 2269. (30) Hu, X.; Do, D. D.; Rao, G. N. Experimental concentration dependence of surface diffusivity of hydrocarbons in activated carbon. Chem. Eng. Sci. 1994, 49, 2145. (31) Rouquerol, F.; Rouquerol, J.; Sing, K. Adsorption by Powders & Porous Solids; Academic Press: San Diego, 1999. (32) ASTM International. Petroleum products, lubricants and fossil fuels, Section 5; ASTM: Philadelphia, PA, 2015. (33) Luna, F. M. T.; Pontes Filho, A. A.; Trindade, E. D.; Cavalcante, C. L., Jr. Rapid assessment of total and polycyclic aromatic contents in heavy oils. Environ. Monit. Assess. 2016, 188, 215. (34) IP 346. Polycyclic aromatics in petroleum fractions by dimethyl sulfoxide−refractive index method; Test method designation IP 346/80; IP Standards for Petroleum and its Products. Part I: Methods of Analysis and Testing; Vol. 2. 3rd ed. Institute of Petroleum: London, 1993. (35) Poling, B. E.; Prausnitz, J. M.; O’Connell, J. P. The Properties of Gases and Liquids; 5th ed.; McGraw-Hill: New York, 2001. (36) gPROMS User Guide. Advanced Process Modelling and Simulation; Process Systems Enterprise: London, 2006. (37) Commission Regulation, Official Journal of the European Union, N. 552, 2009. (38) Do, D. D. Adsorption Analysis: Equilibria and Kinetics, Series on Chemical Engineering, Vol. 2, Imperial College Press: London, 1998. (39) Mastral, A. M.; García, T.; Callen, M. S.; Murillo, R.; Navarro, M. V.; Lopez, J. M. Sorbent characteristics influence on the adsorption of PAC: I. PAH adsorption with the same number of rings. Fuel Process. Technol. 2002, 77, 373. (40) Toth, J. Adsorption: Theory, Modeling, and Analysis; Marcel Dekker: New York, 2002.

Greek Symbols

αPAH/A = selectivity of PAH with respect to single aromatics compounds εp = particle void fraction ραp = particle apparent density (g·cm−3) ϕ = association coefficient η = dynamic viscosity (cP) τ = tortuosity factor υ = dimensionless parameter related to the surface heterogeneity in Toth equation



REFERENCES

(1) Harvey, R. G. In Polycyclic Hydrocarbons and Carcinogenesis; American Chemical Society: WA, 1991. (2) Boffetta, P.; Jourenkova, N.; Gustavsson, P. Cancer risk from occupational and environmental exposure to polycyclic aromatic hydrocarbons. Cancer Cause Control 1997, 8, 444. (3) Richter, H.; Howard, J. B. Formation of polycyclic aromatic hydrocarbons and their growth to soot - a review of chemical reaction pathways. Prog. Energy Combust. Sci. 2000, 26, 565. (4) Mackerer, C. R.; Griffis, L. C.; Grabowski, J. S., Jr; Reitman, F. A. Petroleum mineral oil refining and evaluation of cancer hazard. Appl. Occup. Environ. Hyg. 2003, 18, 890. (5) Chen, S.; Liao, C. Health risk assessment on human exposed to environmental polycyclic aromatic hydrocarbons pollution sources. Sci. Total Environ. 2006, 366, 112. (6) Guieysse, B.; Viklund, G.; Toes, A. C.; Mattiasson, B. Combined UV-biological degradation of PAHs. Chemosphere 2004, 55, 1493. (7) Gan, S.; Lau, E. V.; Ng, H. K. Remediation of soils contaminated with polycyclic aromatic hydrocarbons (PAHs). J. Hazard. Mater. 2009, 172, 532. (8) Mastral, A. M.; Garcia, T.; Murillo, R.; Callen, M. S.; Lopez, J. M.; Navarro, M. V. PAH mixture removal from hot gas by porous carbons. From model compounds to real conditions. Ind. Eng. Chem. Res. 2003, 42, 5280. (9) Garcıa, T.; Murillo, R.; Cazorla-Amoros, D.; Mastral, A. M.; Linares-Solano, A. Role of the activated carbon surface chemistry in the adsorption of phenanthrene. Carbon 2004, 42, 1683. (10) Ania, C. O.; Cabal, B.; Pevida, C.; Arenillas, A.; Parra, J. B.; Rubiera, F.; Pis, J. J. Removal of naphthalene from aqueous solution on chemically modified activated carbons. Water Res. 2007, 41, 333. (11) Luna, F. M. T.; Pontes-Filho, A. A.; Trindade, E. D.; Silva, I. J.; Azevedo, D. C. S.; Cavalcante, C. L., Jr. Removal of aromatic compounds from mineral naphthenic oil by adsorption. Ind. Eng. Chem. Res. 2008, 47, 3207. (12) Bu, J.; Loh, G.; Gwie, C. G.; Dewiyanti, S.; Tasrif, M.; Borgna, A. Desulfurization of diesel fuels by selective adsorption on activated carbons: Competitive adsorption of polycyclic aromatic sulfur heterocycles and polycyclic aromatic hydrocarbons. Chem. Eng. J. 2011, 166, 207. (13) Kalmykova, Y.; Moona, N.; Strömvall, A. M.; Björklund, K. Sorption and degradation of petroleum hydrocarbons, polycyclic aromatic hydrocarbons, alkylphenols, bisphenol A and phthalates in landfill leachate using sand, activated carbon and peat filters. Water Res. 2014, 56, 246. (14) Hameed, B. H.; Rahman, A. A. Removal of phenol from aqueous solutions by adsorption onto activated carbon prepared from biomass material. J. Hazard. Mater. 2008, 160, 576. (15) Dursun, G.; Cicek, H.; Dursun, A. Y. Adsorption of phenol from aqueous solution by using carbonised beet pulp. J. Hazard. Mater. 2005, 125, 175. (16) Kennedy, L. J.; Vijaya, J. J.; Kayalvizhi, K.; Sekaran, G. Adsorption of phenol from aqueous solutions using mesoporous carbon prepared by two-stage process. Chem. Eng. J. 2007, 132, 279. G

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research (41) Leyva-Ramos, R.; Geankoplis, C. J. Diffusion in liquid-filled pores of activated carbon. I. Pore volume diffusion. Can. J. Chem. Eng. 1994, 72, 262. (42) Luna, F. M. T.; Araújo, C. C. B.; Veloso, C. B.; Silva, I. J., Jr.; Azevedo, D. C. S.; Cavalcante, C. L., Jr. Adsorption of naphthalene and pyrene from isooctane solutions on commercial activated carbons. Adsorption 2011, 17, 937. (43) Ahn, S.; Werner, D.; Karapanagioti, H. K.; McGlothlin, D. R.; Zare, R. N.; Luthy, R. Phenanthrene and pyrene sorption and intraparticle diffusion in polyoxymethylene, coke, and activated carbon. Environ. Sci. Technol. 2005, 39, 6516.

H

DOI: 10.1021/acs.iecr.6b01059 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX