Heterogeneity of activated carbons with different surface chemistry in adsorption of phenol from aqueous solutions

Heterogeneity of activated carbons with different surface chemistry in adsorption of phenol from aqueous solutions

Applied Surface Science 252 (2006) 5752–5762 www.elsevier.com/locate/apsusc Heterogeneity of activated carbons with different surface chemistry in ad...

482KB Sizes 2 Downloads 41 Views

Recommend Documents

No documents
Applied Surface Science 252 (2006) 5752–5762 www.elsevier.com/locate/apsusc

Heterogeneity of activated carbons with different surface chemistry in adsorption of phenol from aqueous solutions K. La´szlo´ a,*, P. Podkos´cielny b,1, A. Da˛browski b,2 a

Department of Physical Chemistry, Budapest University of Technology and Economics, Budafoki, H-1521 Budapest, Hungary b Faculty of Chemistry, Department of Theoretical Chemistry, Maria Curie-Skłodowska University, pl. M. Curie-Skłodowskiej 3, 20-031 Lublin, Poland Received 12 April 2005; received in revised form 30 June 2005; accepted 25 July 2005 Available online 8 September 2005

Abstract The heterogeneity of activated carbons is investigated on the basis of adsorption isotherms of phenol from dilute aqueous solutions at different values of pH in the solution. The original carbon studied was prepared from polyethyleneterephtalate (PET). Its various oxygen surface functionalities were systematically changed by additional nitric acid and heat treatments. The Dubinin–Astakhov adsorption-isotherm equation was used to evaluate the parameters characterizing the adsorption of phenol from dilute water solutions on activated carbon surfaces. Adsorption energy distribution functions were calculated by the INTEG algorithm, based on a regularization method. Analysis of distribution functions for activated carbons provides significant comparative information about their energetic heterogeneity. Moreover, a comparison of the resulting energies obtained from the distributions can be made with enthalpic data. # 2005 Elsevier B.V. All rights reserved. Keywords: Liquid-phase adsorption; Activated carbon; Phenol; Heterogeneity; Porosity

1. Introduction

* Corresponding author. Tel.: +36 1 463 1893; fax: +36 1 463 3767. E-mail addresses: [email protected] (K. La´szlo´), [email protected] (P. Podkos´cielny), [email protected] (A. Da˛browski). 1 Tel.: +48 81 5375518; fax: +48 81 5375685. 2 Tel.: +48 81.5375605; fax: +48 81 5375685.

Phenol belongs to a group of common environmental contaminants. This toxic weak acid causes an unpleasant taste and odor even at low concentrations in water. Selected properties of phenol are shown in Table 1. For pKa > pH phenol can be adsorbed ‘‘in toto’’—the situation is similar to that for molecular adsorption. Different methods are designed to remove phenol and its derivatives. Adsorption by activated

0169-4332/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2005.07.027

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762 Table 1 Selected properties of phenol Molecular weight pKa Solubility (g/dm3) Cross sectional area (nm2 molecule1) [36]

94.11 9.89 82 0.3–0.42

carbons (ACs) is the best and most frequently used method, because ACs possess perfect adsorption ability for relatively low-molecular-weight organic compounds as phenols. Other methods include, for example, aerobic and anaerobic biodegradation, oxidation by ozone, and uptake by ion exchange resins, etc. It has been established that the adsorption capacity of activated carbon (AC) for the liquid adsorption of aromatic compounds depends on a number of factors [1]: the nature of the same adsorbent—its pore structure, functional groups, ash content; the nature of the adsorbate, its pKa, functional groups present, polarity, aqueous solubility, molecular size and weight; the solution conditions such as pH, ionic strength, the adsorbate concentration. Such factors as the type of precursor of AC, the aqueous solubility of the compound and oxygen availability in the solution, i.e., ‘‘oxic or anoxic conditions of adsorption’’ [2] are also important. It is known that ACs have strongly heterogeneous surfaces. The heterogeneity of the AC surface stems from two sources, namely geometrical and chemical. The geometrical heterogeneity (porosity) is the result of differences in the size and shape of pores, as well as pits, vacancies and steps. Chemical heterogeneity is associated with different functional groups at a surface, and with various surface contaminants. Both the chemical and geometrical heterogeneities contribute to the unique sorption properties of ACs. However, from adsorption data measured for a definite system, we can obtain information concerning only the ‘‘relative heterogeneity’’ of the adsorbent. It provides information about active centers, which can be detected by the molecules of a given adsorbate [3]. The micropore volume of ACs limits phenol adsorption, and the actual position beneath this limit is determined by their acid–base characteristics [4]. In the range of adsorption in micropores, it is assumed that competition exists between micropore filling of the smallest micropores and adsorption on

5753

active sites located in larger micropores. In the case of adsorption in the smallest micropores, the ‘‘p–p’’ interactions are the strongest—primary micropore filling occurs. However, it can be supposed that surface carbonyl and basic groups that take part in ‘‘donor–acceptor complex’’ formation are located mainly in larger micropores, and the increase in their concentration favors adsorption in those pores [5]. The mechanism of phenol adsorption is also determined by the so-called ‘‘solvent effect’’ [6]. Adsorption of water by ACs together with the change in the energy of phenol–water interactions with the rise in temperature greatly changes the mechanism of phenol adsorption. This ability seems to be the leading factor balancing the equilibrium between the ‘‘p–p interactions’’ and the ‘‘donor–acceptor complex’’ formation. The effect of carbon surface chemical composition on the adsorption of phenol decreases with rising temperature and geometrical heterogeneity determines adsorption of phenol at higher temperatures [6]; at ambient temperature, however, the influence of surface functionalities is significant. The AC studied in this paper was prepared from polyethyleneterephtalate (PET) by a two-step physical activation method. Its various oxygen surface functionalities were systematically changed by additional nitric acid and heat treatment. The surface groups and the delocalized electrons of the graphitic structure determine the apparent acid/base character of the AC surface and cause the surface properties of carbons to depend on the pH in aqueous solutions. The distribution of the surface functionalities is thus of fundamental importance in AC-based water treatment processes. Our main object in this paper is to investigate the heterogeneity effects of the original APET and modified ACs upon the adsorption of phenol from water, at different pH values of the solution. The heterogeneous properties of solid adsorbents can essentially be described by their so-called adsorptionenergy distribution function. To obtain this function, the Fredholm integral equation of the first kind must be solved. This equation, however, is numerically illposed [7,8], so the exclusive use of the least squares method for solving it can lead to distorted results. A regularization method was therefore used, which significantly improves the solution [7–14].

5754

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

2. Experimental

Table 2 The porosity characteristic of activated carbons studied

2.1. Preparation of the APET carbon

Activated carbon

SBET (m2/g)

Vt (cm3/g)

wo (cm3/g)

wads (nm)

wDR (nm)

AC samples were prepared from polyethyleneterephtalate (PET) by a two-step physical activation [15]. The PET was carbonized at 750 8C for 30 min in a steel reactor, flushed with N2 (50 dm3/h). After cooling, pyrolyzed material was ground and sieved. Particles (0.8–2.0 mm) were activated in a steam flow (18 g/h at 900 8C) in a rotary quartz reactor. The steam was diluted with N2 at the molar ratio 1:1. The time of activation, optimized to a burn-off of approximately 50%, was 90 min [16].

APET APET3 APET6 APET3H APET6H

1214 1320 1165 1382 1488

0.52 0.57 0.50 0.60 0.64

0.50 0.53 0.47

0.87 0.86 0.86 0.87 0.86

0.86 0.83 0.82

2.2. Preparation of the modified carbons Part of the samples was treated with concentrated HNO3 at room temperature for 3 h (APET3) and 6 h (APET6) to achieve changes of surface functionalization. Carbons were then washed in distilled water and extracted in a Soxhlet apparatus until neutral pH was obtained. An additional heat treatment at 700 8C for 30 min in N2 flow was applied to part of samples APET3 and APET6 to modify the concentration and the distribution of the surface functional groups. Carbons obtained are designated as APET3H and APET6H, respectively. 2.3. Porosity of the carbons Low temperature nitrogen adsorption isotherms were used to characterize the texture of the ACs. The isotherms, reported earlier [15], belong to type I, showing that all carbons studied are strongly microporous. Neither the reaction with nitric acid nor the additional heat treatment in inert atmosphere results in a significant change of the surface morphology (see Table 2). All the carbon samples exhibit very similar BET surface area and porosity characteristics. The narrow average pore size (0.86–0.87 nm) suggests that the introductory meso- (and macropores) are practically absent. The above findings were also confirmed by SAXS measurements [17,18]. 2.4. X-ray photoelectron spectroscopy (XPS) The surface chemical composition of the samples was determined by XPS [19]. The distributions of the

Notes: wads = (2Vtot/SBET); wo: micropore volume; wDR = 2k/Eo, where k = 9 kJ nm/mol was adopted (Dubinin–Radushkevich (DR) equation).

carbon and oxygen structures (at.%, 1%) derived from the XP C1s and XP O1s spectra, respectively, were published recently in tabular form [15]. The XPS results showed that the surface of the APET carbon contains 93 at.% carbon (of which 52.7 at.% of surface carbon atoms are present in a graphitic form) and 7 at.% oxygen, which leads to an O/C ratio of 0.075. The acid treatment slightly increases the O/C ratios for the granular samples APET3. However, the longer treatment (APET6) does not further increase the overall surface oxygen content. Additional heat treatment results in a further slight increase in the O/C ratio (APET3H, APET6H). Among the oxygencontaining surface functionalities (peaks of II–IV types), species contributing to peak II (i.e. carbon present in phenolic, alcohol or ether groups) are the most abundant. The HNO3 treatment increases the ratio of the oxygen related peaks (APET3). The longer acid treatment results in a more balanced distribution of peaks (APET6). After the heat treatment of this carbon sample, however, groups belonging to peak II again become dominant (APET6H). 2.5. Acid/base properties The carbons studied are characterized by the pH of their aqueous solution detected under standardized conditions [20]. The measured values differ significantly from the pH of nitrogen flushed doubly distilled water in which the carbon samples were immersed (pH  6.8), due to protolytic processes taking place during the immersion of the carbon samples. The original APET carbon surface reveals a basic surface character (the pH value of the carbon suspension was found to be pHslurry 7.5), but other carbons possess an acidic surface character. The decreased pH in the

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

carbon suspensions indicates proton release from the acid treated samples, originating from dissociation of acidic groups formed on the carbon surface during oxidative treatment. The pHPZC values of the carbons were measured by the so-called drift method [21]. The pHPZC is in the basic range in the case of the APET carbon. The acidic treatment shifts its value to the acidic range, while the combined treatment (HNO3 + heating) results in a practically neutral pHPZC value. The Boehm titration method was used to determine the number of oxygenated acidic and basic surface groups. In this aim, the carbon samples were titrated with 0.05 M NaOH and HCl solutions, respectively [22]. The comprehensive acid/base characterization of the carbons studied is included in Table 3. The pHslurry and the total acidity show the same trend. The longer acid treatment results in a higher concentration of the acidic groups, but does not affect significantly the pHslurry, as the amount of basic groups decreases proportionally (%). Additional heat treatment results in a decomposition of some of the acidic groups and an increase of the basic groups (APET3H, APET6H). The pHslurry increased by 2 units. It is obvious that the various surface treatments yield carbon surfaces of various compositions. As XPS characterizes the surface only to a depth of a few nanometers and in dry conditions, the information obtained by ‘‘wet methods’’ is more important. It should be mentioned that, although the surfaces studied exhibit different functional group distributions in the XPS table, the heat-treated samples behave very similarly in aqueous conditions. 2.6. Adsorption from dilute phenol aqueous solution The equilibrium phenol isotherms were measured at various pH values by the batch method from their

5755

solutions prepared with doubly distilled water and the appropriate buffer solutions at ambient temperature [23]. Initial and equilibrium concentrations were determined by detecting the UV absorption of the solutes. The contact times needed to reach equilibrium were deduced from preliminary kinetic measurements [24].

3. Theoretical Among different expressions, the Dubinin–Astakhov (DA) adsorption-isotherm equation is appropriate for describing adsorption from dilute aqueous solutions on heterogeneous surfaces of ACs and is probably the most universal [3,6,25–27]. It has the following form: ut ðce Þ ¼

     cs =ceq n na ¼ exp  RT ln nam bs Eos

(1)

where ut(ce) is the fractional coverage of the adsorbent surface, na the adsorption capacity, nam the maximum adsorption in the micropores, R and T the gas constant and temperature, respectively, bs the affinity coefficient (for phenol equal to 1), Eos the characteristic energy of phenol adsorption, ceq and cs are the equilibrium phenol concentration and the saturation concentration at temperature T, respectively. It is accepted that the empirical parameter n of the DA equation is a structural parameter of carbon [28,29] and it is determined by the standard deviation of micropore sizes [30,31]. However, for carbon adsorbents a direct connection exists between the values of the exponent n and the type of adsorbent [32,33]. For carbons, the values of this parameter lie usually in the range 1–4: n > 2 is frequently found for adsorbents with narrow micropores of small size

Table 3 The acid/base characteristic of activated carbons studied Activated carbon

pHslurry

pHPZC

Total acidity (mequiv./g)

Total basicity (mequiv./g)

% of the basic groups

Basic/acidic

Total number of titrated groups/100 nm2

APET APET3 APET6 APET3H APET6H

7.5 4.0 3.8 5.8 5.7

9.25 4.45 3.81 7.12 7.13

80 640 725 442 380

420 196 193 319 328

84.0 23.4 21.0 41.9 46.3

5.25 0.31 0.27 0.72 0.86

24.8 38.1 47.5 33.2 28.7

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

5756

range, while n < 2 is found for adsorbents with a wide range of pore sizes. The parameter n is also responsible for the adsorption mechanism in micropores. Lowering n leads to a change in this mechanism from primary to simultaneous primary and secondary micropore filling [27]. On the other hand, a decrease in the characteristic adsorption energy, Eos, also changes the shapes of the isotherms. The change is connected with the change in the micropore filling mechanism, and secondary micropore-filling predominates as Eos decreases [6]. Fig. 1A and B show the influence of the parameters n and Eos, respectively, on the shape of the model D–A adsorption isotherm for phenol adsorption from the aqueous solution. The effect of characteristic energy of phenol adsorption Eos on the shape of adsorption isotherms for the phenol adsorption from aqueous solution has been shown previously [6,34]. In our study, like many authors [6,26,34–38], we established a value of the parameter n equal to 4. Values of nam, as well as of the parameter Eos were calculated by curvilinear fitting of the theoretical isotherm to the experimental one (‘‘Minuit’’ procedure [39]). The corresponding error function is defined in the following way: Qmin ¼

k X ðna ðexpÞ  na ðtheorÞÞ2

(2)

i¼1

where na(exp) denotes experimental values of the amount of adsorbed phenol and na(theor) are the theoretical values from the DA equation. 3.1. Calculation of the adsorption-energy distribution function For adsorption of two-component liquid mixtures on a heterogeneous solid surface, each adsorption site is characterized by the energy difference E12 = E1  E2 of the two components, i.e., phenol (1) and water (2). For dilute solutions, the total fractional coverage of solute, ut(ceq) in the surface phase may be expressed as follows [8–10]: Z E12max xV expðE12 =RTÞ FðE12 Þ dE12 ut ðceq Þ ¼ 1 þ xV expðE12 =RTÞ E12min (3)

Fig. 1. (A) Influence of the parameter n on the shape of model D–A adsorption isotherm for the phenol adsorption from aqueous solution (Eos = 20 kJ/mol, nam = 3 mmol/g and cs = 871.321 mmol/dm3). (B) Influence of the parameter Eos on the shape of D–A adsorption isotherm for the phenol adsorption from aqueous solution (n = 4, nam = 3 mmol/g and cs = 871.321 mmol/dm3).

where F(E12) is the normalized distribution function, which characterizes the adsorbent heterogeneity in sol terms of E12 = E1  E2; x ¼ ceq =csol eq , where ceq is the solubility of the solute, R the gas constant and T is the absolute temperature. V = V(ceq, ut) is a model-dependent function [8,9]. If a lattice model is used to describe molecular interactions in both phases, V can be represented by the ratio of the molecular partition functions for the

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

5757

surface and bulk solutions [8]. In our study, this function was assumed to be equal to unity [10]. The regularization method [7] was used for inverting Eq. (3) with respect to the energy distribution function, F(E12). From a mathematical point of view, Eq. (3) is a linear Fredholm integral equation of the first kind: Z b Kðz; yÞ f ðzÞ dz (4) gðyÞ ¼ a

where g(y) is a known function, estimated on the basis of the experimental adsorption isotherm; the integral kernel, K(z, y), represents the local isotherm of Eq. (3) and f(z) denotes the unknown adsorption-energy distribution function. The regularization method requires a discretization of the integral equation by a quadrature method. Eq. (4) must be transformed into a system of linear equations of the type, g = Af, where the onedimensional matrices g and f represent the functions g and f, respectively, but A is a two-dimensional matrix that represents the kernel. In simple words, regularization consists in replacing the problem of minimizing the functional, jjAf  gjj2, by another one that prevents strong oscillations of the distribution function [7]. This can be done by adding an additional term, gjjCfjj2, to the minimization functional S: S ¼ jjAf  gjj2 þ gjjCfjj2

(5)

The quantity g is the so-called regularization parameter, which is a measure for weighting both terms in Eq. (5) [7–14]. The most popular choice for C is the unit matrix applied already by Tikhonov [40]. It is should be emphasized that the regularization method makes no assumption about the shape of the energy distribution function.

4. Results and discussion To examine the influence of the pH on the adsorption of phenol, this phenomenon was investigated on APET carbon at room temperature with aqueous solutions of different pH. The isotherms obtained are shown in Fig. 2A. The symbols denote experimental points, while the curves are the theoretical isotherms estimated using the (D–A)

Fig. 2. (A) Adsorption isotherms from aqueous phenol solutions of various pH on APET carbon at room temperature. The symbols are the measured values of isotherm and the lines are the theoretical isotherms calculated from the D–A equation. (B) Adsorption-energy distribution functions for phenol on the carbon studied.

Eq. (1). To investigate the influence of carbon surface chemistry, the adsorption was studied on modified carbons from pH 3 and unbuffered solutions (pH  6.5). Figs. 3A–6A present the measured adsorption isotherms. In phenol adsorption on APET carbon (Table 4), the values of nam increase monotonically from pH 3 to the unbuffered solution. As for the values of the characteristic energy parameter Eos, the order is the same. This means that the strongest interaction of phenol with the surface occurs for the unbuffered solution, and the

5758

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

Fig. 3. (A) Adsorption isotherms from aqueous phenol solutions from pH 3 and unbuffered solutions on APET carbon (room temperature). The symbols are the measured values, the lines are the theoretical isotherms calculated from the D–A Eq. (1). (B) Adsorption-energy distribution functions for phenol on the APET carbon.

Fig. 4. (A) Adsorption isotherms from aqueous phenol solutions from pH 3 and unbuffered solutions on APET3 carbon (room temperature). The symbols are the measured values, the lines are the theoretical isotherms calculated from the D–A Eq. (1). (B) Adsorption-energy distribution functions for phenol on the APET3 carbon.

weakest for solution with pH 3. In phenol adsorption on modified carbons (Table 5), the order of the nam values, as well as of the Eos parameters for each carbon, is maintained: pH 3 < unbuffered. It is evident that the uptake of phenol is always larger for the unbuffered solution than that at lower pH. The data of Tables 4 and 5 suggest that the quantity of phenol, nam, adsorbed by the carbons corresponds to a smaller equivalent volume than required to fill the micropores. Using the molar volume of solid phenol

at 293 K, Vm = 89 cm3 [26], yields the equivalent volumes Vequiv. = namVm, which are always smaller than the micropore volumes (see Table 2). At low pH, increased adsorption of protons takes place. This leads to an increase in water adsorption (associated water complexes) and to blocking of some of the most active phenol-adsorbing sites or blocking of some pores by water molecules. The same effect is caused by undissociated hydrophilic surface acidic groups. This interpretation is supported by the values

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

Fig. 5. (A) Adsorption isotherms from aqueous phenol solutions from pH 3 and unbuffered solutions on APET3H carbon (room temperature). The symbols are the measured values, the lines are the theoretical isotherms calculated from the D–A Eq. (1). (B) Adsorption-energy distribution functions for phenol on the APET3H carbon.

of the surface area, vDA, available for a single phenol molecule too (see Tables 4 and 5). Cross-sectional areas reported for phenol vary between 0.3 and 0.42 nm2 [41]. At pH 3, the interaction between the carbon surface and phenol is weak. From the comparison of the number of functional groups (Table 3) and the population of the admolecules (Tables 4 and 5) it is clear that dispersion is the dominant interaction: the number of adsorbed phenols is several times greater than the number of titrated sites.

5759

Fig. 6. (A) Adsorption isotherms from aqueous phenol solutions from pH 3 and unbuffered solutions on APET6H carbon (room temperature). The symbols are the measured values, the lines are the theoretical isotherms calculated from the D–A Eq. (1). (B) Adsorption-energy distribution functions for phenol on the APET6H carbon.

In the direction from pH 3 to the pH of the unbuffered solution, the reduction of surface protonation effect takes place and the resultant surface charge decreases. Increasing the pH gradually improves the strength of the interaction, and the adsorption performance. When the adsorption takes place in unbuffered solutions (pH  6.5), the carbon surfaces contain both protonated and deprotonated sites. The surface is more densely populated in all cases due to the combined acidic and basic interactions acting under these conditions. Competition with

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

5760

Table 4 Parameters characterizing adsorption of phenol on APET carbon Medium

nam (mmol/g)

Eos

QDA

Vequiv.

vDAa (nm2)

Number of adsorbed molecules/100 nm2

pH 3 pH 4 pH 5 Unbuffered

2.83 2.90 2.94 2.97

20.16 21.77 22.56 23.30

0.1760 0.1975 1.7987 0.6035

0.252 0.258 0.262 0.264

0.712 0.695 0.686 0.679

140 143 145 147

a

vDA = SBET/(namNA) and NA is Avogadro’s number.

water molecules also takes place, as indicated by the high value of the molecular areas available for one adsorbate molecule, vDA. An interesting fact is the increase in carbon alkalinity upon outgassing (see Tables 3 and 5). The APET3H and APET6H carbons have a slightly higher capacity, nam, than the APET3 and APET6 carbons, respectively, which is probably due to the elimination of acidic groups. The loss of these hydrophilic groups may reduce pore blockage by their associated water complexes [42]. Furthermore, the destruction of these groups reduces the acidity of the carbon surface and eliminates the neutralization of the basic groups caused by their presence [43]. Outgassing the activated carbon as early as at 500 8C removes more acidic groups, while e.g. the number of carbonyl-type groups on the carbon surface increases, which may be due to the fact that carbonyl groups in the proximity of hydroxyl and carboxyl groups can condense to form lactones or lactols [44]. Higher outgassing temperatures (700 8C in our experiment) favor the formation of additional basic sites. Pyrone-type structures are generated by thermal decomposition of oxygenated acidic groups, formation of active sites capable of

fixing oxygen in ether form, and rearrangement with existing carbonyl groups that resist pyrolysis [43]. The adsorption-energy distribution functions obtained by the inversion of Eq. (3) are shown in Figs. 2B–6B for the ACs studied. Numerically stable functions were obtained with the value of the regularization parameter, g = 0.1. Fig. 2B shows the calculated values of the adsorption-energy distribution functions for phenol on APET carbon in different values of pH. The single peak of the energydistribution function for (pH 3) shows a maximum max at about E12 ¼ E1  E2 ¼ 19:09 kJ=mol. That for pH 4 is lower, broader and shifted towards a slightly max higher energy (E12 ¼ 20:45 kJ=mol) compared to the pH 3 peak. The lowest and the broadest peak, covering the widest range of adsorption energies E12 is that for the unbuffered solution. Figs. 3B–6B show the adsorption-energy distribution functions for modified carbons in acidic (pH 3) and unbuffered aqueous solutions. In all cases the peak for the unbuffered solution is lower, broader and shifted towards higher max energy (greater E12 ) compared to the (pH 3) peak. The direction of these changes correlates well with the order of parameters, Eos.

Table 5 Parameters characterizing adsorption of phenol on the ACs studied Medium

nam (mmol/g)

Eos

QDA

Vequiv.

vDAa (nm2)

Number of adsorbed molecules/100 nm2

APET

pH 3 Unbuffered

2.83 2.97

20.16 23.30

0.1760 0.6035

0.252 0.264

0.712 0.679

140 147

APET3

pH 3 Unbuffered

2.12 2.29

21.62 22.06

0.2213 0.1132

0.189 0.204

1.034 0.957

96 104

APET3H

pH 3 Unbuffered

2.72 2.97

21.39 22.49

0.1935 0.6671

0.242 0.264

0.844 0.773

118 129

APET6H

pH 3 Unbuffered

2.76 3.22

22.19 22.35

0.1721 0.7684

0.246 0.287

0.895 0.767

111 130

a

vDA = SBET/(namNA) and NA is Avogadro’s number.

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762 Table 6 Characterization of adsorption-energy distribution functions for phenol on APET carbon Medium

max E12 (kJ/mol)

Peak height (mol/kJ)

Peak location (kJ/mol)

pH 3 pH 4 pH 5 Unbuffered

19.09 20.45 21.06 22.41

0.111 0.098 0.093 0.083

0–27.66 0–29.20 0–32.35 0–33.19

The characteristics of the adsorption-energy distribution functions for ACs studied are given in Tables 6 and 7, respectively. The tables list the values max of energies corresponding to peak maximum, E12 , peak height values and the ranges of peak locations. The direction of changes of the above parameter values correlates well with the order of parameters, Eos, in the D–A Eq. (1). It is evident that the larger the value of the parameter Eos the larger are the values of max E12 (strength of the interactions with the surface increases), also the broader is the adsorptiondistribution function, i.e., it spreads over a wider range of possible adsorption energies. The calculated F(E12) distributions suggest that the adsorption energy of phenol on APET carbon is stronger by more than 22 kJ/mol than that of water for the unbuffered solution. A comparison of the resulting energies E12 obtained from distributions can be made with calorimetric data from the literature, i.e., for the max E12 values in unbuffered phenol solutions and for available molar enthalpies in phenol/water systems on non-modified carbons. The enthalpy values are in the range 22–30 kJ/mol depending on the nature of the adsorbent [6,26,45,46]. Relatively, the most similar Table 7 Characterization of adsorption-energy distribution functions for phenol on ACs studied Medium

max E12 (kJ/mol)

Peak height (mol/kJ)

Peak location (kJ/mol)

APET

pH 3 Unbuffered

19.09 22.41

0.111 0.083

0–27.66 0–33.19

APET3

pH 3 Unbuffered

20.51 20.72

0.097 0.091

0–29.78 0–33.33

APET3H

pH 3 Unbuffered

20.08 21.03

0.097 0.092

0–29.39 0–33.90

APET6H

pH 3 Unbuffered

21.02 21.45

0.092 0.091

0–30.20 0–31.14

5761

data were obtained by Barton et al. [45]. They found the molar enthalpy change to be equal to 22 kJ/mol for the phenol/water system on AR-BPL porous carbon (SBET = 1088 m2/g). In that paper, only the total number of acidic sites and basic sites was reported. About 75% of the functional groups were basic in nature. For the APET carbon of the present study, SBET was 1214 m2/g, and 84% of the total functional groups were basic. Unfortunately, this comparison is incomplete—only a few characteristics are available. The max reason for the discrepancies between these E12 values and the enthalpy values from the literature are obvious, i.e., they arise from differences in porosity as well as from a different surface chemistry.

5. Conclusions The Dubinin–Astakhov equation with a selected value of n = 4 is appropriate for describing phenol adsorption from aqueous solution in the narrow micropores of the activated carbons studied. It should be stressed that the (D–A) equation with n = 4 is very useful not only for studying adsorption of phenol on activated carbons [6,26,34–38] but also for studying adsorption of many other organics, such as paracetamol, acetanilide, aniline, from aqueous solutions on initial AC and modified ACs at different temperatures and pH levels [37,47]. The values of the maximum adsorption nam in the micropores can be correlated both with the concentration of surface functional groups of carbons and with the physicochemical properties of the solute molecules [37]. On the basis of results obtained in this way, the adsorption mechanism of selected organics can be explained. The present paper shows that increasing the pH gradually improves the strength of the interaction and adsorption performance. The uptake of phenol is always larger for the unbuffered solution than that for a lower pH. The sequence of the energy-distribution peaks F(E12) obtained by the regularization method, in the order of increasing energy E12 correlates well with that of the parameter, Eos, of the D–A equation. The strongest interactions of phenol with the surface occur for the unbuffered solution, and the weakest for the solution with pH 3. Furthermore, the peaks are ‘‘lower’’ and ‘‘broader’’ for the unbuffered solution,

5762

K. La´szlo´ et al. / Applied Surface Science 252 (2006) 5752–5762

which indicates the largest spread of interactions between the carbon surface and phenol. The calculated F(E12) distribution suggests that the adsorption energy of phenol on APET carbon exceeds that of water for the unbuffered solution by more than 22 kJ/mol. The enthalpy values from the literature confirm the correctness of the adsorption energy values obtained in our paper.

Acknowledgements This research was supported by the Hungarian National Research Fund (OTKA, grant No. T046532). The author thanks Katalin Josepovits for XPS measurements. The technical assistance of Emese Fu¨lo¨p and Gyo¨rgy Bosznai is gratefully acknowledged.

References [1] F. Haghseresht, S. Nouri, J.J. Finnerty, G.Q. Lu, J. Phys. Chem. B 106 (2002) 10935. [2] R.D. Vidic, M.T. Suidan, Environ. Sci. Technol. 25 (1991) 1612. [3] M. Jaroniec, R. Madey, Physical Adsorption on Heterogeneous Solids, Elsevier, Amsterdam, 1988. [4] H. Jankowska, A. S´wia˛tkowski, J. Choma, Active Carbon, Ellis Harwood, New York, NY, 1991. [5] L.R. Radovic, C. Moreno-Castilla, J. Rivera-Utrilla, Carbon materials as adsorbents in aqueous solutions, in: L.R. Radovic (Ed.), Chemistry and Physics of Carbon, A Series of Advances, vol. 27, 2001, pp. 227–405. [6] A.P. Terzyk, J. Colloid Interface Sci. 268 (2003) 301. [7] M.V. Szombathely, P. Bra¨uer, M. Jaroniec, J. Comp. Chem. 13 (1992) 17. [8] M. Heuchel, P. Bra¨uer, M. von Szombathely, U. Messow, W.D. Einicke, M. Jaroniec, Langmuir 9 (1993) 2547. [9] M. Heuchel, M. Jaroniec, R.K. Gilpin, P. Bra¨uer, M. von Szombathely, Langmuir 9 (1993) 2537. [10] M. Heuchel, M. Jaroniec, Langmuir 11 (1995) 1297. [11] P. Podkos´cielny, A. Da˛browski, M. Bu¨low, Appl. Surf. Sci. 196 (2002) 312. [12] A. Da˛browski, P. Podkos´cielny, M. Bu¨low, Colloid Surf. A 212 (2003) 109. [13] P. Podkos´cielny, A. Da˛browski, O.V. Marijuk, Appl. Surf. Sci. 205 (2003) 297. [14] K. La´szlo´, P. Podkos´cielny, A. Da˛browski, Langmuir 19 (2003) 5287.

[15] K. La´szlo´, Micropor. Mesopor. Mat. 80 (2005) 205. [16] K. La´szlo´, A. Bota, L.G. Nagy, Carbon 35 (1997) 593. [17] K. La´szlo´, K. Marthi, C. Rochas, F. Ehrburger-Dolle, F. Livet, E. Geissler, Langmuir 20 (2004) 1321. [18] K. La´szlo´, O. Czakkel, K. Josepovits, C. Rochas, E. Geissler, Langmuir 21 (2005) 8443. [19] K. La´szlo´, E. Tomba´cz, K. Josepovits, Carbon 39 (2001) 1217. [20] C.H. Tessmer, R.D. Vidic, L.J. Uranowski, Environ. Sci. Technol. 31 (1997) 1872. [21] G. Newcombe, R. Hayes, M. Drikas, Colloids Surf. A 78 (1993) 65. [22] H.P. Boehm, E. Diehl, W.R. Heck, Angew. Chem. Int. Ed. Engl. 3 (1964) 669. [23] A. Szu˝cs, Surface properties of polymer based carbons, Budapest, Hungary, Budapest University of Technology and Economics, M.Sc. Thesis, 1999. [24] A. Bota, K. La´szlo´, L.G. Nagy, G. Subklew, H. Schlimper, M.J. Schwuger, Adsorption 2 (1996) 81. [25] A. Deryło-Marczewska, M. Jaroniec, Adsorption of organic solutes from dilute solutions on solids, in: E. Matijevic´ (Ed.), Surface and Colloid Science, vol. 14, Plenum Press, New York, 1987, pp. 301–379. [26] F. Stoeckli, M.V. Lo´pez-Ramo´n, C. Moreno-Castilla, Langmuir 17 (2001) 3301. [27] A.P. Terzyk, P.A. Gauden, P. Kowalczyk, Carbon 40 (2002) 2879. [28] S.G. Chen, R.T. Yang, Langmuir 10 (1994) 4244. [29] S.G. Chen, R.T. Yang, J. Colloid Interface Sci. 177 (1996) 298. [30] V.K. Dobruskin, Langmuir 14 (1998) 3840. [31] J. Jagiełło, J.A. Schwarz, J. Colloid Interface Sci. 154 (1992) 225. [32] R.C. Bansal, J.B. Donnet, H.F. Stoeckli (Eds.), Active Carbon, Marcel Dekker, New York, Basel, 1988. [33] S. Błaz˙ewicz, A. S´wia˛tkowski, B.J. Trznadel, Carbon 37 (1999) 693. [34] A.P. Terzyk, Adsorpt. Sci. Technol. 21 (2003) 539. [35] F. Stoeckli, D. Hugi-Cleary, Russ. Chem. Bull. 50 (2001) 2060. [36] F. Stoeckli, Russ. Chem. Bull. 50 (2001) 2265. [37] A.P. Terzyk, J. Colloid Interface Sci. 275 (2004) 9. [38] E. Fernandez, S. Gotovac, D. Hugi-Cleary, V. Lopez-Ramon, F. Stoeckli, Chimia 57 (2003) 616. [39] F. James, M. Roos, Comp. Phys. Commun. 10 (1975) 343. [40] A.N. Tikhonov, Dokl. Akad. Nauk USSR 39 (1943) 195. [41] K. La´szlo´, A. Szu˝cs, Carbon 39 (2001) 1945. [42] R.W. Coughlin, F.S. Ezra, Environ. Sci. Technol. 2 (1968) 291. [43] E. Papirer, S. Li, J.B. Donnet, Carbon 25 (1987) 243. [44] H.P. Boehm, Carbon 32 (1994) 759. [45] S.S. Barton, M.J.B. Evans, J.A.F. MacDonald, Polish J. Chem. 71 (1997) 651. [46] F. Salvador, M.D. Mercha´n, Carbon 34 (1996) 1543. [47] A.P. Terzyk, J. Colloid Interface Sci. 272 (2004) 59.