Ultrasound-assisted adsorption of copper(II) ions on hazelnut shell activated carbon

Ultrasound-assisted adsorption of copper(II) ions on hazelnut shell activated carbon

Ultrasonics Sonochemistry 16 (2009) 557–563 Contents lists available at ScienceDirect Ultrasonics Sonochemistry journal homepage: www.elsevier.com/l...

266KB Sizes 0 Downloads 16 Views

Ultrasonics Sonochemistry 16 (2009) 557–563

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultsonch

Ultrasound-assisted adsorption of copper(II) ions on hazelnut shell activated carbon D.D. Milenkovic´ a, P.V. Dašic´ b, V.B. Veljkovic´ c,* a b c

High Chemical Technological School, Kosancˇic´eva 36, 37000 Kruševac, Serbia High Mechanical Technical School, Radoja Krstic´a 19, 37240 Trstenik, Serbia Faculty of Technology, University of Niš, Bulevar Oslobodenja 124, 16000 Leskovac, Serbia

a r t i c l e

i n f o

Article history: Received 27 August 2008 Received in revised form 5 November 2008 Accepted 1 December 2008 Available online 11 December 2008 PACS: 43.35.+d 68.43.Mn 82.20.Pm Keywords: Activated carbon Adsorption Equilibrium Hazelnut shell Modeling Kinetics Ultrasound Ultrasound-assisted adsorption

a b s t r a c t The present study was aimed to removal of Cu(II) ions from aqueous solution by ultrasound-assisted adsorption onto the granular activated carbon obtained from hazelnut shells. The attention was focused on modeling the equilibrium and kinetics of Cu(II) adsorption onto the granular activated carbon. The granular activated carbon was prepared from ground dried hazelnut shells by simultaneous carbonization and activation by water steam at 950 °C for 2 h. Adsorption isotherm data were better fitted by the Langmuir model than the Freundlich model in both the absence and the presence of ultrasound. The maximum adsorption capacity of the adsorbent for Cu(II), calculated from the Langmuir isotherms, in the presence of ultrasound (3.77 mmol/g) is greater than that in the absence of ultrasound (3.14 mmol/g). The adsorption process in the absence and the presence of ultrasound obeyed to the pseudo second-order kinetics. The removal of Cu(II) ions was higher in the presence of ultrasound than in its absence, but ultrasound reduced the rate constant. The intraparticular diffusion model indicated that adsorption of Cu(II) ions on the granular activated carbon was diffusion controlled as well as that ultrasound promoted intraparticular diffusion. Ó 2008 Elsevier B.V. All rights reserved.

1. Introduction Adsorption has demonstrated its efficiency and economic feasibility for the removal of ions, colors, bacteria and organic compounds from industrial processes and waste effluents, compared to other physical and chemical separation and purification processes. Different porous solid materials have gained importance in industrial applications of adsorption processes. In recent years, raw low-cost and abundant natural materials such as agricultural wastes, known as biosorbents, have been widely studied for removal of the above mentioned pollutants from water. These studies include peat, pine bark, banana pith and peel, rice bran, soybean and cottonseed hulls, peanut shells and hulls, hazelnut shells, rice husk, sawdust, wool fibres, orange peel, saffron corm, coirpith, cocoa shells, etc. Recognizing a great potential of hazelnut shells for their applications as a biosorbent, a number of researchers have already performed some investigations in Turkey, which is the biggest * Corresponding author. Tel.: +381 16 247 203; fax: +381 16 242 859. E-mail addresses: [email protected] (D.D. Milenkovic´), dasicp@ yahoo.com (P.V. Dašic´), [email protected] (V.B. Veljkovic´). 1350-4177/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ultsonch.2008.12.002

hazelnut producer in the world, and Italy. So far, hazelnut shells have been used for adsorption studies of metal ions such as Cu(II) [1], Ni(II), Cd(II) and Pb(II) [2] as well as dye [3]. The equilibrium and kinetics of Cu(II) removal by hazelnut shells from aqueous solution are described by the Langmuir isotherm model and the pseudo second-order kinetics, respectively [1]. An often disadvantage of biosorbents is the relatively low adsorbing capacity. In some cases, the adsorption capacity can be improved by preparing activated carbon from plant materials. The suitability of agricultural by-products for the manufacture of granular activated carbon has been discussed by Heschel and Klose [4]. Researchers have used different procedures for activated carbon preparation from plant materials. The improvement process usually includes washing, drying and comminution of plant materials, followed by impregnation, carbonization and activation. Activated carbons have been already prepared from hazelnut shells and used as adsorbents for removal of toxic metals from aqueous solutions. Some researchers impregnated ground hazelnut shells by concentrated H2SO4 [5–7], KOH under ultrasound irradiation [8] or ZnCl2 (30%) [9], followed by carbonization and activation at elevated temperatures. The other researchers carbonized ground hazelnut shells at 800 °C for 2 h [10] or carbonized crushed raw

D.D. Milenkovic´ et al. / Ultrasonics Sonochemistry 16 (2009) 557–563

558

Nomenclature aL bF C CI C0 k1 k2 KF KL

Langmuir constant (l/mmol) Freundlich exponent (dimensionless) liquid phase concentration (mmol/l) constant of the intraparticular diffusion model (mmol/g) initial concentration of solute in aqueous phase (mmol/l) rate constant of pseudo first-order sorption (min1) rate constant of pseudo second-order sorption (g/mmol min) Freundlich constant (ðmmol=gÞ=ðmmol=lÞbF ) Langmuir equilibrium constant (l/g)

material at 650 °C and then kept the produced charcoal at 900 °C for 1 h under nitrogen atmosphere, followed by steam activation at 800 °C [11]. The Freundlich isotherm model is more accurate to describe the isotherm of Cu(II) adsorption on the active carbon prepared by alkaline impregnation of hazelnut shells under ultrasound irradiation than the Langmuir isotherm model [8]. There is no study in the literature on the kinetics of Cu(II) removal from aqueous solution by activated carbons from hazelnut shells. Activated carbons, both granular or powdered, have demonstrated higher adsorbing effects than corresponding raw plant materials because of their extended surface area, microporous structure and special surface reactivity for removal of toxic metal ions from aqueous solutions [5–8,10], as well as for gold cyanide adsorption [11]. For instance, the maximum adsorption capacity of the activated carbon from alkaline impregnated hazelnut shells under ultrasound for removal of Cu(II) ions (40 mg/g) is approximately twice greater than that of the origin plant material (22 mg/g), which is probably due to the higher specific area of the former adsorbent (BET surface areas 10.1 and 0.188 m2/g of the activated carbon and the raw material, respectively) [8]. The use of activated carbons is often suffered from the relatively low adsorption rate because of their microporous structures and long diffusion path through solid particles. Ultrasound has been proven to exhibit several effects in solid–liquid systems such as the enhancement of mass transfer rate, the increase of the surface area by forming many micro-cracks on the solid surface and the clean-up of solid particle surfaces [12]. The role of ultrasound on the adsorption/desorption processes has been recently studied and controversial effects have been found. For instance, Rege and coworkers [13] found that the desorption rate of activated carbon significantly increased under sonication conditions. Also, Li et al. [14] reported that the adsorption capability of phenol in the presence of ultrasound is less than that in its absence. However, Schueller and Yang [15] reported that ultrasound improved the mass transfer coefficient through cavitation and acoustic streaming, which could be the reason for the enhancement of the adsorption kinetics. The shear forces generated during the cavitation are mostly responsible for the enhancement of Pb(II) and Cu(II) removal by saffron corm in the presence of ultrasound [16]. There is a relative lack of information in the literature about the removal of metal ions from aqueous solution by adsorption combined with ultrasound (so called ultrasound-assisted adsorption). No specific study has been done on the equilibrium and kinetics of Cu(II) ions from aqueous solution onto the granular activated carbon prepared from hazelnut shells in the presence of ultrasound. Therefore, the overall objective of the present study was aimed to removal of Cu(II) ions from an aqueous solution by ultrasound-assisted adsorption onto the granular activated carbon obtained from hazelnut shells. The main goals were to evaluate the

kp mGAC q qe qm R V t

intraparticular diffusion rate constant (mmol/g min0.5) mass of sorbent (g) solid phase concentration at specified time (mmol/g) solid phase concentration at equilibrium (mmol/g) monolayer solid phase maximum adsorption capacity (mmol/g) coefficient of linear correlation volume of copper(II) acetate solution (l) time (min)

effects of ultrasound on the equilibrium and kinetics of Cu(II) adsorption onto the granular activated carbon and to choose optimal equilibrium and kinetic adsorption models. 2. Experimental section 2.1. Materials Hazelnut (Corylus avellana L.; the variety is Tonda Istriana) shells were supplied from the central part of Serbia. Fresh hazelnut shells were washed several times with distilled water to remove surface impurities, dried at 100 °C overnight, crushed by a hammer mill and simultaneously carbonized and activated by water steam in an oven at 950 °C for 2 h. Afterwards, the granular activated carbon was washed three times with distilled water, dried at 110 °C for 24 h and stored in a desiccator. Copper(II) acetate, analytical reagent grade, was purchased from Merck Co. Distilled water was used to prepare aqueous solutions of copper(II) acetate. 2.2. Batch adsorption experiments 2.2.1. Set-up The set-up consisted of an ultrasonic cleaner (EI, Niš, Serbia; total nominal power: 2  50 W; and internal dimensions: 300  220  155 mm), operating at 40 kHz frequency. The cleaner was filled with distilled water up to 1/3 of its volume (about 3.5 l). All experiments were carried out at 25 °C (±0.2 °C). The temperature was controlled and maintained by water circulating from a thermostated bath by means of a pump. An erlenmayer flask (250 ml) used as adsorption vessel was fixed on a swinger (90 oscillations per minute). In both equilibrium and kinetic experiments, the copper(II) acetate solution (100 ml) and the adsorbent (1.0 g) were put into the flask fixed on the swinger. The flask was swung through water at a certain distance from the bottom of the cleaner. In the silent adsorption experiments, the ultrasound generator was switched off. 2.2.2. Equilibrium experiments In these experiments, the initial copper(II) acetate concentration was in the range between 0.013 and 0.260 mol/l. In a preliminary equilibrium test under silent conditions applying the smallest (0.013 mol/l) and the highest (0.260 mol/l) copper(II) acetate concentration, it was established that the equilibrium was reached after 5 h. For sure, all equilibrium experiments lasted 6 h. After establishing equilibrium, a sample was taken from the flask and was centrifuged (1500 rpm for 5 min) for the removal of adsorbent particles. The Cu(II) concentration in the supernatant was measured by AAS

D.D. Milenkovic´ et al. / Ultrasonics Sonochemistry 16 (2009) 557–563

559

(Perkin–Elmer 1100B). The amount of Cu(II) ions adsorbed was calculated from the following mass balance equation:



ðC 0  CÞ  V mGAC

ð1Þ

where q is the amount of Cu(II) ions adsorbed at time t; C0 and C are the initial Cu(II) concentration and the Cu(II) concentration at time t, respectively; V is the volume of copper(II) acetate solution (=100 ml); and mGAC is the amount of adsorbent (=1.0 g). For each sample, the Cu(II) concentration was measured in triplicate and the mean value was used as the equilibrium one. 2.2.3. Kinetic experiments The initial copper(II) acetate concentration was 134 mmol/l. The actual Cu(II) concentration in the liquid phase was periodically measured during the adsorption by an ion selective electrode (type GPE 201.801, IHTM, Belgrade, Serbia). The amount of Cu(II) ions adsorbed was calculated by Eq. (1). 3. Results and discussion Fig. 1. Adsorption isotherms of Cu(II) ions on granular activated carbon from hazelnut shells in the absence (s) and in the presence of ultrasound (d) at 25 °C.

3.1. Characterization of granular activated carbon The activated carbon prepared from hazelnut shells by simultaneous steam carbonization and activation at 950 °C for 2 h was characterized by standard methods (Table 1). According to its particle size, the prepared active carbon is granular. The specific surface area, measured as BET, of 1651 m2/g, and the iodine number of 1646 mg/g are the highest values so far reported for activated carbons from hazelnut shells. Alkali [8], acid [6] and salt (ZnCl2 [9]) impregnated active carbons had the specific surface area of 10.1, 441 and 793 m2/g, respectively, while the specific surface area of those not impregnated activated carbons were 825 m2/g [11] and 786 m2/g [10]. Thus, the prolonged carbonization/activation process at elevated temperature favored to the formation of porous activated carbon. 3.2. Adsorption isotherms The adsorption isotherms of Cu(II) ions on the granular activated carbon from hazelnut shells at 25 °C in the absence of ultrasonic field as well as in the presence of ultrasonic field are shown in Fig. 1. As can be seen from this figure, the amount of Cu(II) ions adsorbed on the granular activated carbon in the presence of ultrasound is greater than the one in the absence of ultrasound. The

Table 1 The properties of granular active carbon prepared from hazelnut shells. Characteristics

Method

Value

Specific surface area, m2/g Iodine number, mg/g Methylene blue index, cm3 pH Value Ash, % Damp, % Bulk density, kg/l

BET, N2

1651

ASTM D 4607

1646

CEFIC (Test methods for activated carbon, Method 2.4) ASTM D 3838 ASTM D 2866 ASTM D 2867 ASTM D 2854

20

Granular structure, %

Pore volume, cm3 g1

DIN 4188

9.1 6.5 8.1 0.416 >1.6 mm 1–1.6 mm 0.51.0 mm <0.5 mm

3.1 70.0 26.1 0.9

Micropore Mesopore Macropore

0.681 0.701 0.223

same effect of ultrasound on the sorption capacity of saffron corm for removal of Cu(II) ions was explained by the acoustic cavitation (formation, growth and collapse of the cavity) under sonication [16]. In addition, as seen from Fig. 1, both adsorption isotherms are non-linear and seem to approach to the maximum values. The shape of the curves indicated that several well-known models reported in the literature could be successfully employed to describe the adsorption isotherms. The Langmuir and Freundlich isotherms are the most frequently used models (Table 2). The former model is applicable to homogeneous sorption, where the sorption of each molecule has an equal sorption activation energy, and the latter one is empirical in nature. In the present work, both adsorption isotherms were used to model the relationships between the amount of Cu(II) ions adsorbed and its equilibrium concentration in solution in the absence and the presence of ultrasound for 6 h at 25 °C. Fig. 2 illustrates that the two experimental equilibrium curves that were obtained in the presence of ultrasound and its absence are well represented by both equilibrium models applied in this study. However, when the Langmuir isotherm model was applied to our experimental data, better fits were obtained both in the presence of ultrasound and its absence than with the Freundlich isotherm model, as can be seen in Table 3 when comparing the corresponding linear correlation coefficients and the normalized root mean square (RMS). The higher R-value for the Langmuir isotherm model than for the Freundlich isotherm model might be due to homogeneous distribution of active sites on shell surface as it was explained in the case of Cu(II) ions removal by fresh hazelnut shells [1]. The Langmuir constants KL, aL and qm as well as the Freundlich constants KF and bF are also displayed in Table 3. The Langmuir adsorption constant aL defines the ratio of adsorption and desorption rate constants and is related to the free energy of adsorption. Its value represents the affinity of Cu(II) ions to the adsorbent. By

Table 2 Adsorption isotherms. Isotherm

Integral form

Linear form

Langmuir

KL Ce qe ¼ 1þa L Ce

Ce qe

Freundlich

qe ¼

K F C be F

¼ K1L þ KaLL C e

ln qe ¼ ln K F þ bF ln C e

D.D. Milenkovic´ et al. / Ultrasonics Sonochemistry 16 (2009) 557–563

560

pared to the silent adsorption. Also, the monolayer saturation capacity at equilibrium qm in the presence of ultrasound was greater than that in the absence of ultrasound. The adsorption capacity of Cu(II) ions by corm of saffron was also greater in the presence of ultrasound than in the silent conditions [16]. This was attributed to cavitation effects which increased capability of the porous particle structure for Cu(II) ions adsorption and/or the appearance of new sites of sorption by disruption of sorbent particles. Maximum adsorption capacities for removal of Cu(II) ions from aqueous solution by the granular activated carbon achieved in silent and ultrasound-assisted adsorption are 3.14 and 3.77 mmol/ g (calculated from the Langmuir isotherm model), respectively, and are much higher than that (0.62 mmol/g) reported for the activated carbon obtained by alkaline impregnation of hazelnut shells under sonication [8]. This was attributed to greater specific surface area of the granular activated carbon used in the present study. 3.3. Adsorption kinetics

Fig. 2. Linear forms of adsorption isotherms of Cu(II) ions on granular activated carbon from hazelnut shells in the absence (s) and the presence of ultrasound (d) at 25 °C: (a) the Langmuir model and (b) the Freundlich model.

Table 3 Parameters of adsorption isotherms, linear correlation coefficient and standard deviation. Isotherm

Parameter

Silent adsorption

Ultrasound-assisted adsorption

Langmuir

K L ; l=g aL ; l=mmol qm ; mmol=g R Normalized RMS, %

0.0446 0.0146 3.05 0.995 ±5.7

0.0799 0.0212 3.77 0.998 ±5.1

Freundlich

K F ; ðmmol=gÞ=ðmmol=lÞbF bF ; l R Normalized RMS, %

0.123 0.562 0.970 ±11.9

0.241 0.489 0.993 ±6.1

comparing the values of aL for the silent and the ultrasound-assisted adsorption, one can conclude that ultrasound positively affected the affinity of Cu(II) ions to the granular activated carbon as it was previously noticed for the adsorption of Cu(II) ions by corm of saffron [16]. The same conclusion was withdrawn from values of the Freundlich constant KF, related to the adsorption capacity, which was doubled in the presence of ultrasound, com-

The variations of the amount of Cu(II) ions adsorbed on the granular activated carbon obtained from hazelnut shells with the progress of silent and ultrasound-assisted adsorption at 25 °C are shown in Fig. 3. The adsorption duration time of 120 min was selected based on the earlier reported results using hazelnut shells [2]. This figure illustrates the effects of ultrasound on the kinetics of Cu(II) ions adsorption on the granular activated carbon obtained from hazelnut shells. In the beginning of the adsorption process, Cu(II) ions were rapidly adsorbed, then the adsorption rate was slowed down and finally the equilibrium was gradually reached. The highest rates of Cu(II) ions removal at the beginning was probably due to the larger surface area of hazelnut shells available for adsorption and probably the strong interaction between the Cu(II) ions and the surface of adsorbent. In the later periods, the surface adsorption sites become exhausted and the removal rate was controlled by the rate of Cu(II) ions transportation from the exterior to the interior sites of the adsorbent particles. The removal of Cu(II) ions was higher in the presence of ultrasound than in its absence, due to the cavitation process which increased the diffusion process by the micro-jet and streaming produced in the collapse of the cavity. Ultrasound affected the distribution of the sites of energy (tends towards a homogeneous distribution) in addition to the effect of cavitation. The rate of a sorption process is influenced by combined actions of several important factors related to the adsorbent, the adsorbate

Fig. 3. Adsorption kinetics of Cu(II) ions on granular activated carbon from hazelnut shells in the silent adsorption (s) and the ultrasound-assisted adsorption (d) at 25 °C.

D.D. Milenkovic´ et al. / Ultrasonics Sonochemistry 16 (2009) 557–563

561

and the process conditions. Each sorption process involves several steps such as diffusion of the solute from the solution to the film surrounding the sorbent particles (bulk diffusion), diffusion of the solute through the film to the particle surface (external diffusion), diffusion of the solute from the particle surface though pores to the internal active sites (internal or pore diffusion) and uptake of the solute on active sites by different mechanisms (physico-chemical sorption, ion exchange, etc.). The rate of each step can generally control the overall sorption rate. To examine the controlling mechanism of adsorption process studied, three kinetic models, namely pseudo first-order, pseudo second-order and intraparticular diffusion were used to test the experimental data. Their differential, integral and linear forms are displayed in Table 4. Values of the parameters of the kinetic models applied, the linear correlation coefficient and the normalized RMS are shown in Table 5. When the pseudo first-order kinetics was tested, the values of the equilibrium amount of Cu(II) ions adsorbed, qe, was calculated by the non-linear regression assuming the first-order exponential growth of the amount of Cu(II) ions adsorbed with time (qe ¼ 1:20 and 1:53 mmol=g or silent and ultrasound-assisted adsorption, respectively). From the linear forms of the kinetic curves (shown in Fig. 4), the linear correlation coefficient and the normalized RMS (given in Table 5), one can conclude that all the kinetic models applied fitted well the experimental data. All the linear correlation coefficients obtained are higher than 0.990, with only one exception R = 0.986 for the second portion of the intraparticular diffusion model. Among the three kinetic models, the pseudo second-order model generated the best-fit to the experimental data of both the silent and the ultrasound-assisted adsorption systems, having the highest linear correlation coefficient and the lowest normalized RMS (Fig. 2b). This model is based on the assumption that the rate-limiting step might be a chemical reaction between the adsor-

Table 4 Kinetic models. Kinetic model

Differential form

Pseudo first-order

dq dt

Pseudo second-order

dq dt

¼ k2 ðqe  qÞ

Intraparticular diffusion

@q @t

¼ Deff

¼ k1 ðqe  qÞ 2

@2 q @t2

Integral form q ¼ qe ð1  e q¼

k2 q2e t 1þk2 qe t

Linear form

k1 t

pffiffi q ¼ kp t þ C I

Þ

ln qeqq ¼ k1 t e

t q

¼ k 1q2 þ q1 t e 2 e pffiffi q ¼ kp t þ C I

Table 5 Parameters of kinetic models, linear correlation coefficient and standard deviation. Kinetic model

Model parameters

Silent adsorption

Ultrasound-assisted adsorption

Pseudo first-ordera

k1 , min1 R Normalized RMS, % k2 , g/mmol min qe , mmol/g R Normalized RMS, %

0.0452 0.990 ±11.3 0.0457 1.41 1.000 ±1.7

0.0447 0.996 ±0.7 0.0336 1.81 1.000 ±1.7

kp , mmol/g min0.5 R Normalized RMS, % kp , mmol/g min0.5 C I , mmol/g R Normalized RMS, %

0.173 0.998 ±4.6 0.0557 0.656 0.993 ±1.2

0.214 0.999 ±2.5 0.0776 0.776 0.986 ±1.8

Pseudo second-order

Intraparticular diffusion –First stage portionb

–Second stage portion

a qe was calculated by the non-linear regression assuming the first-order exponential growth of the amount of Cu(II) ions adsorbed with time (qe = 1.20 and 1.53 mmol/g for silent and ultrasound-assisted adsorption, respectively). b The intercept C I ¼ 0 for both silent and ultrasound-assisted adsorption.

Fig. 4. Kinetic models for the removal of Cu(II) ions on granular activated carbon from hazelnut shells in the absence of ultrasonic field (s) and in the presence ultrasonic field (d) at 25 °C: (a) pseudo first-order; (b) pseudo second-order; and (c) intraparticular diffusion.

bent and the adsorbate. Thus, the pseudo second-order model is potentially a generalized kinetic model for the adsorption system studied. The applicability of this kinetic model and the second-

D.D. Milenkovic´ et al. / Ultrasonics Sonochemistry 16 (2009) 557–563

562 Table 6 Activated carbons from agricultural wastes. Agriculture waste/ (BET surface area)

Temperature (°C)

pH

Mixing

Cu(II) concentration (mg/g)

qm (mg/g)

Initial Cu(II) concentration (g/l)

k2 (g/mmol min)

Reference

Hazelnut shells, steam activated (1651 m2/g)

25

5.0

820–16,500

200

8.5

0.0457

Present study

Hazelnut shells (10.1 m2/g) Hazelnut husks Peanut hulls (208.0 m2/g) Peanut shells (725 m2/g) Ceiba pentandra hulls (521 m2/g) Rubber wood dust, H3PO4 activated (1674 m2/g)

18 18 30 na 30 30

naa 5.7 5.0 4.8 6.0 6.0

Swinging, 90 oscillations per minute Ultrasound, 40 kHz, 100 W Shaking, 170 rpm Shaking, 200 rpm Shaking, 180 rpm Stirring, 300 rpm Shaking, na Shaking, 180 rpm

na na 20–80 818–3177 1–28 5–40

240 39.54 6.645 65.57 50.4 20.8 5.73

– 200 10–20 mg/l – 40–100 10–40

0.0336 nsb ns Pseudo first-order ns na 0.0937

[8] [22] [18] [23] [24] [25]

a b

na – Not available. ns – Not studied.

order nature of the adsorption process of Cu(II) ions on fresh hazelnut shells [1,2] and corm of saffron [16] have been already reported. In the present study, ultrasound positively affected the equilibrium constant independently of the reaction order assumed, but either did not influenced the pseudo first-order rate constant (Fig. 4a) or reduced the pseudo second-order rate constant (Fig. 4b). These observances disagreed with the earlier reported positive effects of ultrasound on the rate constants in the case of the above mentioned natural adsorbents [1,16]. In the case of Reactive Black 5 adsorption from aqueous solution by limestone, ultrasound increased significantly the adsorption capacity, but the increase of k2 was very small (0.0583 and 0.0580 g/mmol min for ultrasound-assisted and silent adsorption, respectively) [17]. Because both the pseudo first-order and the pseudo second-order models cannot identify the diffusion mechanism, the intraparticular diffusion model was also tested in the present study. Fig. 4c indicates that the plots in the absence and the presence of ultrasound presented a dual nature, implying that two steps occurred. Since the two best-fit straights of the first initial stage passed through the origin, there was not an initial boundary layer resistance both in the presence of ultrasound and its absence. This might be explained by intensive mixing of the suspension by both the swinging alone and the ultrasound-assisted swinging, eliminating the external diffusion resistance. The results presented in Fig. 4c also indicated that adsorption of Cu(II) ions on the granular activated carbons was diffusion controlled. However, the adsorption rate of heavy metal ions on fresh hazelnut shells was not diffusion controlled [2]. The sharper first-stage portion was attributed to the gradual adsorption region, where the intraparticular diffusion was rate-limited, and the second portion was the final equilibrium stage where intraparticular diffusion started to slow down [18–20]. During these two stages, the Cu(II) ions were slowly transported via intraparticular diffusion in the particles and were finally retained in the pores. Ultrasound promoted intraparticular diffusion during both stages, as the rate constants characterizing the two regions increased in the ultrasound-assisted adsorption, compared to the silent condition. The increase of the intraparticular diffusion rates in the two stages in the presence of ultrasound was approximately for 24% and 40%, respectively, compared to silent condition. The intercept of the second stage portion, CI, is proportional to the boundary layer thickness, which gives an insight into the tendency of the metal ions to adsorb to the adsorbent or remain in solution [21]. The values of CI for the silent and the ultrasound-assisted adsorptions are also given in Table 5. The intercept was greater in the presence of ultrasound than in its absence, depicting the higher ultrasound-assisted adsorption capacity of the granular active carbon, compared to silent condition [21].

3.4. Comparison of activated carbons from different agricultural adsorbents The adsorption capacities (calculated from the Langmuir isotherm model) and the pseudo second-order rate constants of the granular activated carbon obtained from hazelnut shells and other agricultural adsorbents are compared in Table 6. It can be concluded that the granular activated carbon from hazelnut shells adsorbs Cu(II) ions from aqueous solution more than other activated carbons obtained from agricultural wastes. 4. Conclusion The granular activated carbon produced from hazelnut shells has high specific surface area and highly developed microporous structure. It has the highest specific surface area among the active carbons produced from hazelnut shell so far, regardless impregnation was applied or not. This indicates that the prolonged simultaneous carbonization and steam activation processes (2 h) at elevated temperatures (950 °C) favored the formation of microporous structure. Adsorption isotherm data for removal of Cu(II) from aqueous solution fit well both Langmuir and Freundlich models for both silent and ultrasound-assisted adsorption, but higher values of R and smaller values of normalized RMS are obtained from the former model, probably due to homogeneous distribution of active sites on shell surface. The kinetics of Cu(II) removal on the granular activated carbon from hazelnut shells follows the pseudo second-order model, indicating that the rate-limiting step might be a chemical reaction between the adsorbent and the adsorbate. As the intraparticular diffusion model fits well the kinetic data, adsorption of Cu(II) ions on the granular activated carbons is also diffusion controlled. The primary benefits of sonication is the increase of the adsorption capacity of the granular activated carbon from hazelnut shells and the promotion of intraparticular diffusion, while its drawback is the reduction of the rate of pseudo second-order reaction. The maximum adsorption capacity for removal of Cu(II) ions from aqueous solution is higher for about 20% for ultrasound-assisted adsorption than for silent adsorption. The rate constant of intraparticular diffusion enhances for about 24%, while the pseudo secondorder rate constant is reduced for about 26% in the presence of ultrasound. The prepared granular activated carbon can be used as an effective adsorbent for batch adsorption of Cu(II) from aqueous solution. Although it is prepared from the agricultural waste with no value by a relatively simple production process, the conclusion about its application for removal Cu(II) and other toxic metal ions

D.D. Milenkovic´ et al. / Ultrasonics Sonochemistry 16 (2009) 557–563

should be withdrawn only after a detailed techno-economic analysis of the complete removal process. References [1] Ö. Demirbasß, A. Karadag˘, M. Alkan, M. Dog˘an, Y. Bulut, Z. Tez, J. Hazard. Mater. 153 (2008) 677–684. [2] Y. Bulut, Z. Tez, J. Hazard. Mater. 149 (2007) 35–41. [3] F. Ferrero, J. Hazard. Mater. 142 (2007) 144–152. [4] W. Heschel, E. Klose, Fuel 74 (1995) 1786–1791. [5] G. Cimino, A. Passerani, G. Toscano, Water Res. 34 (2000) 2955–2962. [6] E. Demirbasß, M. Kobya, S. Öncel, S. S ß encan, Biores. Technol. 84 (2002) 291–293. [7] M. Kobya, Biores. Technol. 91 (2004) 317–321. [8] E. S ß ayan, Chem. Eng. J. 115 (2006) 213–218. [9] A. Aygün, S. Yenisoy-Karakasß, I. Duman, Micropor. Mesopor. 66 (2003) 189– 195. [10] M. Kazemipour, M. Ansari, S. Tajrobehkar, M. Majdzadeh, H.R. Kermani, J. Hazard. Mater. 150 (2008) 322–327. [11] M. Yalcin, A.I. Arol, Hydrometallurgy 63 (2002) 201–206.

[12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]

563

L.H. Thompson, L.K. Doraiswamy, Ind. Eng. Chem. Res. 38 (1999) 1215–1248. S.U. Rege, R.T. Yang, C.A. Cai, AIChE J. 44 (1998) 1519–1528. Z. Li, X. Li, H. Xi, B. Hua, Chem. Eng. J. 86 (2002) 375–379. B.S. Schueller, R.T. Yang, Ind. Eng. Chem. Res. 40 (2001) 4912–4918. M.H. Entezari, T. Soltani, J. Hazard. Mater. 160 (2008) 88–93. M.H. Entezari, Z. Sharif Al-Hoseini, N. Ashraf, Ultrason. Sonochem. 15 (2008) 433–437. S. Goswami, U.C. Ghosh, Water SA 31 (2005) 597–602. S. Karaca, A. Gürses, M. Ejder, M. Açikyildiz, J. Colloid Interf. Sci. 277 (2004) 257–263. Í.A. Sßengil, M. Özacar, J. Hazard. Mater. 157 (2008) 277–285. J.C. Igwe, A.A. Abia, C.A. Ibeh, Int. J. Environ. Sci. Technol. 5 (2008) 1735–1772. M. Imamoglu, O. Tekir, Desalination 228 (2008) 108–113. K. Wilson, H. Yang, C.W. Seo, W.E. Marshal, Biores. Technol. 97 (2006) 2266– 2270. M. Madhava Rao, A. Ramesh, G. Purna Chandra Rao, K. Seshaiah, J. Hazard. Mater. B 129 (2006) 123–129. M.H. Kalavathy, T. Karthikeyan, S. Rajgopal, L.R. Miranda, J. Colloid, Interf. Sci. 292 (2005) 354–362.