Adsorption of nitrogen-heterocyclic compounds on bamboo charcoal: Kinetics, thermodynamics, and microwave regeneration

Adsorption of nitrogen-heterocyclic compounds on bamboo charcoal: Kinetics, thermodynamics, and microwave regeneration

Journal of Colloid and Interface Science 390 (2013) 189–195 Contents lists available at SciVerse ScienceDirect Journal of Colloid and Interface Scie...

443KB Sizes 1 Downloads 14 Views

Journal of Colloid and Interface Science 390 (2013) 189–195

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Adsorption of nitrogen-heterocyclic compounds on bamboo charcoal: Kinetics, thermodynamics, and microwave regeneration Peng Liao b,c, Songhu Yuan a,b,⇑, Wenjing Xie b, Wenbiao Zhang d, Man Tong b,c, Kun Wang b a

State Key Lab of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin 150090, PR China State Key Lab of Biogeology and Environmental Geology, China University of Geosciences, Wuhan 430074, PR China c Environmental Science Research Institute, Huazhong University of Science and Technology, Wuhan 430074, PR China d Zhejiang A & F University, Lin’an 311300, PR China b

a r t i c l e

i n f o

Article history: Received 6 July 2012 Accepted 17 September 2012 Available online 26 September 2012 Keywords: Bamboo charcoal Kinetics Thermodynamic Microwave regeneration Nitrogen-heterocyclic compounds Wastewater

a b s t r a c t The adsorption kinetics and thermodynamics of nitrogen-heterocyclic compounds (NHCs), pyridine, indole and quinoline, in aqueous solutions on bamboo charcoal (BC), as well as the regeneration of spent BC by microwave radiation, are investigated. BC is produced by incomplete combustion of moso bamboo at high temperature and nitrogen atmosphere. Adsorption kinetics is analyzed using pseudo-first-order and pseudo-second-order as well as Weber–Morris model. The results show that NHC adsorption on BC is predominantly regulated by surface diffusion in initial 1 h followed by intraparticle diffusion in later stage. BC exhibits a strong adsorption affinity to NHCs, and the adsorption isotherms are well described by Freundlich model. Thermodynamic analysis indicates that the adsorption is spontaneous and endothermic. Adsorption site energy analysis illustrates a distribution of adsorption energy, which indicates the heterogeneous sites on BC for NHC adsorption. Furthermore, spent BC with NHC adsorption can be effectively regenerated by MW radiation. The adsorption capacity becomes even higher than that of virgin BC after five times of adsorption–regeneration cycles. This study proves BC is a promising adsorbent for NHC removal in wastewater. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction Nitrogen-heterocyclic compounds (NHCs) are often present in coking, refinery, and pharmaceutical wastewater. They are hazardous in nature with a long-term persistence in the environment because of poor biodegradability [1–3]. Adsorption by porous materials, particularly activated carbon (AC), is an effective method for NHC removal from wastewaters [1,4]. However, AC is mostly made from the non-renewable source of coal with the problem of high cost and difficult regeneration [5]. Therefore, it is necessary to develop cost-effective and renewable adsorbents to remove NHC from wastewater. Bamboo charcoal (BC) is a new environmentally friendly and renewable bioresource with mesoporous structure [5–7]. It is produced from incomplete combustion of moso bamboo at high temperature (800 °C) and nitrogen atmosphere [7,8]. Bamboo is a renewable bioresource due to its fast-growing speed and short growth period. The total species, growing stock, and harvesting

⇑ Corresponding author at: State Key Lab of Biogeology and Environmental Geology, China University of Geosciences, Wuhan 430074, PR China. Fax: +86 27 67848629. E-mail addresses: [email protected], [email protected] (S. Yuan). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2012.09.037

amount of bamboo in China are greatest in the Asia-Pacific region [6]. With the rapid development of bamboo industry, BC increasingly attracts interests in a wide range of applications [9]. The large surface area, large pore volume, and high hydrophobicity endow BC great potential in water purification [5–7,10]. BC is reported to be a superior adsorbent for removing many organic contaminants in wastewaters [5,11–14]. Our recently work revealed the mechanism for NHC adsorption on BCs and modified BCs, showing that the relative contribution dominating NHC adsorption conformed to surface area > hydrophobic interaction > electrostatic interaction > p–p electron-donor–acceptor (EDA) interaction [15]. Nevertheless, the kinetics and thermodynamics for NHC adsorption on BC have not been investigated, which is crucial for the potential applications of BC to remove NHCs from wastewater. The rule of adsorption dependence on time and temperature can provide valuable information of rate-limiting step and energy changes during adsorption [16]. The kinetics and thermodynamics of NHC adsorption on different environmental matrices, including AC, molecular sieve (Ti-HMS), and alumina [1,4,17], have been investigated. The properties of BC are different from those materials, and the adsorption mechanisms are also different. Therefore, it is necessary to elucidate the kinetics and thermodynamics for NHC adsorption on BC. To reduce the cost of operation and waste disposal, regeneration of spent adsorbents is an important issue influencing the

P. Liao et al. / Journal of Colloid and Interface Science 390 (2013) 189–195

applicability of adsorbent in wastewater treatment. Thermal and chemical methods are most commonly applied for the regeneration of exhausted adsorbent [18,19]. However, these methods are time and energy consuming resulting in high cost, and moreover, pore structure of adsorbents is deteriorated to some extent [20,21]. In recent years, regeneration of spent carbon-based adsorbent by microwave (MW) radiation has been proposed as a potentially viable alternative to traditional methods [22,23]. MW regeneration is high speed and energy-saving with the quick destruction or combustion of adsorbed contaminants [21,23]. BC is a good MW absorber, receiving MW energy directly through dipole rotation and ionic conduction [24]. Furthermore, our recent work found that BC modified by MW radiation had much higher adsorption capacity to dyes compared with virgin BC [7]. Hence, it is rational to use MW radiation for the regeneration of spent BC. In this study, the adsorption of three NHCs, pyridine, indole and quinoline, on BC will be investigated with emphasis on adsorption kinetics and thermodynamics as well as the spent BC regeneration by MW radiation. The main objectives are (1) to identify the key process controlling the rate of NHC adsorption on BC, (2) to measure the spontaneity and thermodynamic feasibility of the adsorption process, and (3) to evaluate the feasibility of spent BC regeneration by MW radiation. 2. Materials and methods 2.1. Chemicals and materials Pyridine (P99.5%) was purchased from Tianjin Kermel Chemical Reagent Co. Ltd. Indole (P99.5%) and quinoline (98%) were supplied by Aladdin Chemistry Co. Ltd. The procedure for BC production was provided in previous work [7,15]. Briefly, moso bamboo was evaporated at 100–150 °C for 2–3 d, pre-carbonized at 150–270 °C for 2–3 d, carbonized at 270–450 °C for 1–2 d, calcined at 450–800 °C for 0.5–1 d, and finally kept at 800 °C for a short time. All the processes were carried out under nitrogen atmosphere. The obtained massive BC was ground and sieved through a 100-mesh screen. According to previous characterization [15], the as-produced BC predominant existence of mesopores (2–50 nm) and the main elements are carbon, oxygen, hydrogen, and nitrogen. The BC primarily consists of graphite sheets and functional groups, that is, CAC, CAO, [email protected], COOA, p–p and [email protected] The surface of BC is negatively charged at the pH of 7.0. After washed with DI water, the BC particles were dried at 105 °C for 6 h and stored for use. DI water (18.0 MX cm) was obtained from a Millipore Milli-Q system. All the other reagents were above analytical grade.

2.3. Regeneration of spent BC by MW radiation One gram of BC was added to 200 mL of NHCs solution containing 200 mg L1 in a series of 250-mL iodine bottles. The bottles were sealed and shaken for 24 h at 150 rpm in dark (298 ± 1 K). Preliminary experiments prove the adsorption reached saturation at this condition. Then, NHCs-saturated BC was dried at 105 °C for 1 h, followed by radiation regeneration in an MW oven (2450 MHz, 550 W) for 5 min. Then, the MW treated BC was reused to adsorb NHCs under identical conditions. The adsorption–regeneration process was repeated for five cycles. The loss of BC mass for each cycle was measured to be less than 5%. After radiation, the NHCs on BC were extracted by 20 mL of acetone/dichloromethane (3:2, v/v) mixture and 0.2 ml of 37% HCl at 150 rpm for 2 h, followed by ultrasonication (200 W, 40 kHz) for 30 min. The extraction procedure was repeated three times. 2.4. Analysis and characterization After filtration through 0.45-lm membrane, the concentrations of pyridine, indole, and quinoline in the supernatant were analyzed at 256, 269, and 276 nm, respectively, on a Cary 50 ultraviolet visible spectrophotometer (Varian, USA). Textural characteristics of BC were determined by N2 adsorption at 77 K with an accelerated surface area and porosimeter (ASAP-3000, Micromeritics). Elemental analysis was performed using a Vario Micro-cube elementar analyzer (Elementar Company, Germany). The surface functional groups were measured by a diffuse reflectance infrared Fourier transform spectroscopy (DRIFT, VERTEX 70, Bruker Corporation). X-ray photoelectron spectroscopy (XPS) measurements were car-

(a)

20

15

qe (mg g-1)

190

Pyridine

10

Indole Quinoline 5

0

0

5

10

15

20

25

Time (h)

(b)

third step

20

2.2. Adsorption kinetics and isotherms

second step

qe (mg g-1)

first step

For the adsorption kinetics, a weighed quantity of BC (0.05 g) and 20 mL of aqueous solution containing 50 mg L1 NHCs were added into a 30-mL serum bottle. Preliminary experiments show that adsorption using this dosage gave a straightforward comparison of adsorption. 5 mmol L1 CaCl2 was used as a background electrolyte and 200 mg L1 of NaN3 was added as biocides to inhibit the degradation by incidental bacteria. The bottles were sealed and placed in a shaking table at 150 rpm in dark (298 ± 1 K). At regular time intervals, the bottles were taken out for NHC analysis. With respect to adsorption isotherms, the equilibrium time was set at 24 h with different initial NHC concentrations at 5– 200 mg L1 and different temperatures at 298, 308, and 318 K. The initial pH was set at 7.0, and the variation was less than 0.2 during the whole process. Control experiments without BC result in less than 5% loss of NHCs. All the experiments were conducted in duplicate.

15

10

Pyridine Indole

5

0

Quinoline

0

1

2

3

4

5

t1/2 (h1/2) Fig. 1. (a) Adsorption kinetics of pyridine, indole, and quinoline on BC. Symbols stand for the experimental data and lines stand for the pseudo-second-order fitting, (b) Weber–Morris model plots of pyridine, indole, and quinoline adsorption on BC. Error bars represent standard deviations. The reaction conditions were based on 50 mg L1 initial NHCs concentration, 2.5 g L1 BC, 298 K and pH 7.0.

191

P. Liao et al. / Journal of Colloid and Interface Science 390 (2013) 189–195

ried out on a VG Multilab 2000 apparatus with Al Ka radiation (1486.6 eV) as excitation source.

stage is noticeably lower than that of surface diffusion (Fig. 1b) as the pore volume of BC (0.22 cm3 g1) is too small to accommodate the larger molecular-sized of NHCs (0.81–1.02 cm3 g1). Previous studies suggested that the adsorption of NHCs on other adsorbents, such as AC, is governed by intraparticle diffusion, and the adsorption process mainly occurs in the micropore [1]. In

3. Results and discussion 3.1. Adsorption kinetics

40

(a)

The adsorption of the three NHCs on BC reaches equilibrium within 24 h (Fig. 1a). In order to evaluate the adsorption kinetics mechanisms, the experimental data were fitted by the pseudofirst-order and pseudo-second-order kinetic models as well as the diffusion-based Weber–Morris models [25,26]. It is noted that the pseudo-second-order kinetic model fits the adsorption data better than pseudo-first-order model as indicated by higher R2 values (Table 1). The rate constant (k2) of indole is higher than those of pyridine and quinoline because indole has relatively larger octanol–water distribution coefficient (log Kow) [15]. The adsorption rate (V0) of NHCs on BC derived from pseudo-second-order model decreased with increasing contact time (Table 1), which indicates the gradual occupation of the adsorption sites. Three steps are involved in pseudo-second-order kinetic model: (i) NHC molecules diffuse from liquid phase to liquid–solid interface; (ii) NHC molecules move from liquid–solid interface to solid surfaces; and (iii) NHC molecules diffuse into the particle pores [26,27]. Herein, the first step is not rate-limited because the adsorption was performed at rigorous shaking conditions. To reveal the relative contribution of surface and intraparticle diffusion to the kinetic process, the adsorption kinetics is further fitted using Weber–Morris model (Table 1). Intraparticle diffusion is assumed to be the sole rate-controlling step if the regression of qe versus t1/2 is linear and the plot passes through origin [28]. In this study, qe is linearly correlated with t1/2 with positive intercepts (Table 1), suggesting that intraparticle diffusion is not the ratecontrolling step for the adsorption. The plots of qe and t1/2 exhibit a tri-linearity stage (Fig. 1b) implying three successive adsorption steps. About 57–76% of NHCs are adsorbed by BC in the first stage, that is, initial 1 h, which is attributed to the instantaneous occupation of most available surface sites by electrostatic attraction. The surface of BC is negatively charged at pH 7 [15], and the cationic fraction of pyridine, indole, and quinoline at pH 7 is 14%, 99% and 11%, respectively (calculated by their pKa values. 5.23 for pyridine, 2.40 for indole, and 4.90 for quinoline). Therefore, the k2 value for indole is higher than that of pyridine and quinoline. Moreover, the thickness of the boundary layer (C) for indole adsorption are remarkably larger compared with pyridine and quinoline (Table 1), suggesting that surface diffusion plays a more important role for indole adsorption. In the second stage, that is, 2–14 h, only 21–33% of NHCs is adsorbed on BC. This may be attributed to the slow diffusion of NHCs from surface film into micropore [29]. The rate of intraparticle diffusion in the third

qe (mg g-1)

30

298 K 20

308 K 318 K 10

0

(b)

qe (mg g-1)

40

30

298 K 308 K

20

318 K 10

0 40

(c)

qe (mg g-1)

30

298 K 20

308 K 318 K

10

0

0

20

40

60

80

100

120

Ce (mg L-1) Fig. 2. Adsorption isotherms of (a) pyridine, (b) indole, and (c) quinoline on BC at different temperatures. The reaction conditions were based on 2.5 g L1 BC, pH 7.0, and 24 h reaction time. Symbols stand for experimental data and lines denote the Freundlich fitting. Error bars represent standard deviations.

Table 1 Parameters for pseudo-first-order and pseudo-second-order as well as Weber–Morris models.a Organics

qe,exp (mg g1)

Pseudo-first-order modelb 1

qe,cal (mg g Pyridine Indole Quinoline a

15.8 19.5 17.7

7.5 ± 1.2 7.4 ± 1.2 9.9 ± 1.1

)

1

k1 (h

)

0.11 ± 0.01 0.14 ± 0.01 0.17 ± 0.01

Pseudo-second-order modelc 2

R

qe,cal (mg g

0.895 0.905 0.966

15.4 ± 0.8 19.2 ± 0.6 17.7 ± 0.5

1

)

V0 (mg g 20.24 35.97 20.12

1

h

Weber–Morris modeld 1

)

1

k2 (g mg 0.08 0.10 0.06

h

1

)

2

R

C

kip (mg g1 h1/2)

R2

0.997 0.998 0.996

5.3 ± 1.4 8.5 ± 2.1 5.7 ± 1.5

2.5 ± 0.5 2.7 ± 0.7 2.9 ± 0.5

0.732 0.577 0.767

The reaction conditions were based on 50 mg L1 initial NHCs concentration, 2.5 g L1 BC, 298 K and pH 7.0. The pseudo-first-order model is given by ln (qe  qt) = ln qe  k1 t, where qe and qt are the amounts of adsorbed NHCs at equilibrium and at time t (mg g1), k1 is the equilibrium rate constant of pseudo-first-order kinetics (h1). c The pseudo-second-order model is determined by q1 ¼ k 1q2  1t þ q1 ¼ V10  1t þ q1 , where k2 is the equilibrium rate constant of the pseudo-second-order kinetics (g mg1 h1) t e e 2 e and V0 is the initial adsorption rate (mg g1 h1). d 1/2 The Weber–Morris model is expressed by qt = kip t + C, where kip is the intra-particle diffusion rate constant (mg g1 h1/2) and C (mg g1) is a constant that reflects the thickness of boundary layer. b

192

P. Liao et al. / Journal of Colloid and Interface Science 390 (2013) 189–195

temperature revealed that the adsorption of NHCs by BC may involve not only physical adsorption but also chemical adsorption [36]. Moreover, the enhanced adsorption affinity of NHCs at higher temperature may be due to activation and creation of some new active sites on the surface of BC. This could be due to the fact that

comparison with AC, it is worth noting that the surface diffusion plays a more significant role in NHC adsorption on BC than intraparticle diffusion (Fig. 1b). The reason may be that BC is primarily in mesoporous structure while AC is mainly in micropore structure [15,30]. Additionally, for a given adsorbent, our recent work also revealed that the adsorption kinetics of dyes on BC is attributed to both surface diffusion and intraparticle diffusion [7].

(a) 20

298 K 308 K 318 K

E * kJ mol

-1

3.2. Adsorption isotherms and thermodynamics The adsorption isotherms of NHCs at different temperatures are illustrated in Fig. 2. The Freundlich model [31,32], which was used successfully in fitting adsorption isotherms of organic compounds (e.g., antibiotics, dyes, and NHCs) by BCs in our previous studies [5,7,15], was employed here to fit the experimental data. All isotherms exhibit nonlinear characteristics and are well fitted by Freundlich model. The model parameters are tabulated in Table 2. BC had the higher KF for NHCs (5.82–13.51, Table 2) than reported for AC (1.92–4.85, [1]), implying it may be used as a potential adsorbent for NHCs removal from wastewater. Fig. 2 also demonstrates that the adsorption of NHCs on BC increases with increasing temperature. With the increase in temperature, the rate of molecular diffusion increases and the viscosity of solution decreases [33,34], which facilitates NHC molecules moving across the external boundary layer into the internal pores of BC. In order to reveal the inherent energetic changes during the process, thermodynamic parameters [33], such as standard Gibbs free energy change (DG0), standard enthalpy change (DH0), and standard entropy change (DS0), are calculated (Table 2). The negative values of DG0 at different temperatures indicate the favorable and spontaneous nature of NHC adsorption onto BC. It is worth noting that DG0 became more negative as temperature increased, indicating that the driving force for the adsorption of three NHCs increased with increasing temperature. Similar results were also found in literatures [7,33]. The positive values of DH0 for the adsorption indicate an endothermic process of adsorption. In addition, the values of DH0 for NHCs increased in the order of indole > quinoline > pyridine, which further indicated that the adsorption of indole on BC was stronger than that of pyridine and quinoline. Furthermore, the positive values of DS0 reflect the high affinity of BC to NHCs and the increase in randomness at the solid-solution interface during the adsorption process. It has been reported that the adsorption of NHCs on AC and Ti-HMS is an exothermic, irregularity decreased, and spontaneous physical process [4,35]. On the contrary, the properties of those reported adsorbents are significantly different with BC in present study. The extent of adsorption increased with increasing

16

12

8

(b)

28

298 K 308 K 318 K

E * kJ mol

-1

24 20 16 12 8

(c)

28

298 K 308 K 318 K

E * kJ mol

-1

24 20 16 12 8 5

10

15

20

25

30

35

40

qe (mg g-1) Fig. 3. Adsorption site energy curves for (a) pyridine, (b) indole, and (c) quinoline adsorption on BC at different adsorbate loadings. The data were calculated by Eq. (1) using the data in Table 2.

Table 2 Freundlich model parameters and thermodynamic parameters for adsorption of pyridine, indole, and quinoline on BC at different temperatures.a Organics

Temperature (K)

Freundlich modelb KF (mg g

Pyridine

Indole

Quinoline

a

298 308 318 298 308 318 298 308 318

1

)/(mg L

5.82 ± 0.56 6.12 ± 0.86 6.20 ± 0.92 13.51 ± 1.71 14.90 ± 1.72 15.80 ± 1.09 10.07 ± 0.79 12.09 ± 1.54 14.14 ± 1.24

Thermodynamic parametersc

1 n

)

2

n

R

K0 (L g1)

DG0 (kJ mol1)

DH0 (kJ mol1)

DS0 (J mol1 K1)

0.39 ± 0.02 0.38 ± 0.03 0.37 ± 0.04 0.22 ± 0.03 0.21 ± 0.03 0.20 ± 0.01 0.24 ± 0.02 0.23 ± 0.03 0.19 ± 0.02

0.983 0.963 0.959 0.998 0.921 0.969 0.971 0.926 0.940

1.67 1.75 1.81 12.95 17.25 20.13 4.08 6.03 7.15

1.28 1.43 1.58 6.39 7.19 7.99 3.57 4.43 5.29

3.17

14.93

22.14

86.25

17.45

79.99

The reaction conditions were based on 2.5 g L1 BC, pH 7.0, and 24 h reaction time. The Freundlich model is described by qe ¼ K F C ne , where qe (mg g1) and Ce (mg L1) are the equilibrated concentration of the adsorbate in adsorbent and solution, respectively, KF is the adsorption coefficient and n is a dimensionless parameter related to the surface heterogeneity. 0 c H0 Thermodynamic parameters are determined by the following equations: K 0 ¼ Cqee (1), DG0 = RT ln K0 (2), DG0 = DH0  TDS0 (3), ln K 0 ¼  DRT þ DRS (4), where K0 is the adsorption coefficient, R is gas constant (kJ mol1 K1), and T is the absolute temperature (K), respectively. b

193

P. Liao et al. / Journal of Colloid and Interface Science 390 (2013) 189–195

an increase in temperature augments mobility of NHCs from bulk to the BC surface and degree of penetration within BC structure overcoming the activation energy barrier and therefore enhancing the rate of intraparticle diffusion [37]. The similar results were also found in our previous study for adsorption of dyes on BCs [7]. 3.3. Adsorption site energy distribution To examine energetic characteristics of NHCs and BC interactions, the average adsorption energy (E) and site energy distribution (F(E)) of NHCs on BC at different temperatures are described with the following function [33,38]:

  nE qe ¼ K F  ðC s Þn  exp  RT   K F nðC s Þn nE  exp  RT RT

ð2Þ 3.4. Regeneration of spent BC by MW radiation The regeneration of spent adsorbents is important for actual application of adsorption process [39]. Fig. 5 shows the adsorption of NHCs on virgin BC and MW regenerated BC after five successive cycles. A high efficiency of regeneration is achieved and moreover,

(a) 298 K

-1

F(E *) (g mol kg kJ )

1.2

-1

308 K 318 K 0.8

(a)

0.4

0.0 20

25

30

35

10

298 K

30

2

1

5

10

15

20

(c)

4

3

0 40

30

qe (mg g-1)

-1

298 K 308 K 318 K

20

10

25

(c) -1

0 40

318 K

qe (mg g-1)

-1

(b)

308 K

3

0

F(E *) (g mol kg kJ )

20

4

-1

F(E *) (g mol kg kJ )

(b)

40

30

qe (mg g-1)

FðE Þ ¼

ð1Þ

As illustrated in Figs. 3 and 4, both the E and F(E) for adsorption of NHCs on BC follow the order of 318 > 308 > 298 K. This further proves that the adsorption of NHCs is elevated at higher temperatures. In addition, Fig. 3 demonstrates that E decreased dramatically for BC with the increasing loadings of NHCs. NHC molecules occupied the high energy adsorption sites first at low concentrations and then spread to the low energy adsorption sites at high concentrations. The similar result was also reported by Wang et al. [33]. The affinity of the solute to the adsorbent surface is related to the position of the energy distribution mean on the energy axis [38]. Thus, the higher value of mean energy is in accordance with the higher adsorption affinity. As shown in Fig. 3, indole has the highest E value compared with pyridine and quinoline, implying that BC has more adsorption sites for indole than for pyridine and quinoline. This also agreed with the order of site energy distribution curves (Fig. 4).

2

20

1 10 0

10

15

20

25

E * (kJ mol ) -1

0

0

1

2

3

4

5

Cycles Fig. 4. Adsorption site energy distribution curves for (a) pyridine, (b) indole, and (c) quinoline adsorption on BC at different temperatures. The data were calculated by Eq. (2) using the data in Table 2.

Fig. 5. Effect of regeneration times on (a) pyridine, (b) indole, and (c) quinoline adsorption on BC. Error bars represent standard deviations.

194

P. Liao et al. / Journal of Colloid and Interface Science 390 (2013) 189–195

the adsorption capacity of BC after regeneration is even higher than that of virgin BC. This proves that MW radiation is effective for the regeneration of spent BC. The similar results were also observed for the regeneration of granular activated carbon by Quan and Liu [20,22]. It has been reported that the adsorption capacity of carbonbased adsorbents (i.e., ACs and CNTs) after regeneration by traditional thermal methods decreased gradually with subsequent regeneration cycles [40]. Heating by conventional means occurs by convection, that is, the high temperature gas conducts from the surface to the center of adsorbents. The desorbed molecules produced near the center have to traverse a high temperature surface region. However, it is difficult to prevent the decomposition of desorbed molecules inside the porous domain because the transmission of heat into the inner core is slow by this method. This causes the physical rupture of the pore walls and a greater obstruction of the pore [40]. Unlike the case of conventional heating, MW heats the adsorbents from the inside, creating a thermal temperature gradient that decreases toward the surface of the material [41]. Hence, the desorbed molecules produced in the core of the adsorbent diffuse toward the lower temperature region more quickly. This reduced the formation of coke deposits inside the porous domain, resulting in a better preservation of the textural characteristics of adsorbents [40]. In our case, since desorption occurs at the surface of the BC, the NHCs molecules inside the BC have to migrate to the surface. As the diffusion toward the surface is higher than that toward intraparticle (Fig. 1b), the desorption process is enhanced during MW radiation. Regarding the regeneration of spent BC by MW radiation, several processes may take place in theory. Firstly, gases coming from the desorption of volatiles and the decomposition of organic substances adsorbed are evolved [21,23]. The release of the gases will modify the surface area and porosity, thereby enhancing the adsorption capacity. Secondly, the partially decomposed organic substances will form coke deposits, blocking the porosity, and decreasing the adsorption capacity of BC [23,42]. Thirdly, the surface oxygen-containing functional groups of BC may be dramatically decreased in the presence of high temperature [40]. The p-electron density of the graphite layers increased due to the dispersive interactions between the NHCs and the p-electron

density of the graphite [15,43]. This will be in favor of adsorption by means of chemisorption. To reveal the mechanism for MW regeneration, the regenerated BC regenerated after one, three, and five cycles of adsorption was characterized. Table 3 shows the BC after MW regeneration exhibited an increase in external surface area, average pore diameter, and total pore volume relative to BC, but a decrease in BET surface area. The changes induced in textural properties of regenerated BC are due to the annealing effect on the carbonaceous skeleton at high temperature and the production of coke deposits within the porous structure [40]. The pore-size distribution (Fig. 6a) shows the presence of more mesopores (2–50 nm) in the regenerated BC. The contribution of mesopores to the total surface area of BC is usually smaller relative to micropores [20]. Nevertheless, mesopores play an important role in the adsorption of the large-molecular organics which cannot access to pores of smaller dimensions [44]. Elemental composition analysis suggests that the regenerated BC had higher carbon content and lower oxygen content, which is associated with higher hydrophobicity. The DRIFT spectra and XPS results (Fig. 6b, Table 3) show a substantial increase in graphite and p–p groups and a dramatic decrease in surface oxygen-containing functional groups after MW regeneration. This is because most oxygencontaining functional groups were decomposed under the MWinduced high temperature [45]. Therefore, the mechanism for MW regeneration can be explained from the following aspects: (1) MW radiation increases the external surface area, average pore diameter, and total pore volume, even though the total surface area is decreased in some cases; (2) the hydrophobicity of regenerated BC is increased; (3) MW radiation increases graphitic content and p–p content on BC’s surface, thereby increasing the adsorption via p–p interaction [15,46,47]. 3.5. Implications for practical application This study proves that BC, a new renewable porous bioresource with meso and large pore structure, is efficient in adsorbing NHCs in aqueous solutions. In particular, the high speed of NHCs adsorption in the initial 1 h, which is governed by surface diffusion, is beneficial for practical applications. Meanwhile, the thermodynamic analysis demonstrates the adsorption process of NHCs on

Table 3 Comparison of main characteristics of MW regenerated BC and virgin BC. Adsorbents

Textural properties Elemental compositionb (%)

Virgin BC BC-MW1a BC-MW3a BC-MW5a Adsorbents

Peak (eV) Virgin BC BC-MW1a BC-MW3a BC-MW5a a b c d e f g h i j

SBETc

Smicrd

Sextere

Daverf

Pore volume (cm3 g1)

C

H

O

N

(m2 g1)

(m2 g1)

(m2 g1)

(nm)

Vmicrg

Vmesh

V ti

81.38 81.45 82.42 82.71

2.45 1.77 1.53 1.42

9.46 6.75 6.03 5.68

0.45 1.07 1.09 1.12

255.7 234.2 224.3 229.9

231.5 197.1 176.6 166.8

24.2 37.2 46.6 83.1

2.30 2.52 2.68 2.85

0.107 0.091 0.090 0.087

0.113 0.175 0.246 0.382

0.220 0.266 0.336 0.469

Surface oxygen-containing functional groupsj Graphite

CAC

CAO

[email protected]

COOA

p–p

[email protected]

284.6 66.4 76.1 78.4 79.3

285 13.2 8.8 7.8 7.2

286.5 8.4 4.9 4.5 4.1

288 10.7 7.6 6.4 6.1

288.9 0.5 0.4 0.4 0.3

290.1 0.4 0.9 1.2 1.4

287.3 0.3 1.3 1.5 1.6

BC-MW1, BC-MW3, and BC-MW5 denote the BC regenerated by MW radiation after one, three, and five cycles of adsorption, respectively. Determined by Vario Micro-cube elementar analyzer. SBET: BET surface area calculated in the relative pressure region P/P0 = 0.05–0.35. Smicr: Micropore area is determined by t-plot methods. Sexter: External surface area determined by t-plot methods. Daver: Average pore diameter obtained from BJH equation using N2 isotherms. Vmicro: Micropore volume determined by t-plot methods. Vmes: Mesopore volume calculated by Vt–Vmicro. Vt: Total pore volume determined at P/P0 = 0.97. Obtained from XPS results.

P. Liao et al. / Journal of Colloid and Interface Science 390 (2013) 189–195

Acknowledgments

(a) 0.05 Pore volume (cm3 g-1)

195

0.04

BC BC-MW1

0.03

BC-MW3 BC-MW5

This work was supported by the State Key Lab of Urban Water Resource and Environment (HIT) (No. QA201008), the China Environmental Protection Foundation, and the Fundamental Research Funds for the Central Universities, China University of Geosciences (Wuhan) (No. CUGL110608).

0.02

References 0.01

0.00 0

4

8

12

16

20

Pore size (nm)

Transmittance (%)

(b)

C-C C-H C-O 1580 1441 1258 C=O 1710

BC

-OH -CH

BC-MW1 BC-MW3 BC-MW5

3200

2800

2400

2000

Wavenumber

1600

1200

800

(cm-1)

Fig. 6. Comparison of (a) pore-size distribution and (b) DRIFT spectra of MW regenerated BC and virgin BC. Pore-size distribution was calculated by BJH method. BC-MW1, BC-MW3, and BC-MW5 refer to the BC regenerated by MW radiation after one, three, and five cycles of adsorption, respectively.

BC is spontaneous under ambient conditions. As a comparison, conventional carbonaceous materials such as AC mainly consist of micropores, so adsorption of bulky organic chemicals such as NHC may be difficult because of size-exclusion effect [1,30]. Moreover, in China, the cost of BC ($421–571 per ton, [8]) is lower in comparison with commercial AC ($1200–2000 per ton). And spent BC can be effectively regenerated by MW radiation without loss of adsorption capacity. This process has demonstrated its effectiveness for the removal of NHCs, which highlights an optimistic outlook for the practical application of BC. That is, the pollutants in wastewater can be adsorbed on BC in adsorption tower, and MW radiation can be applied when adsorption saturation is reached. 4. Conclusions This study investigated the adsorption kinetics and thermodynamics of NHCs such as pyridine, indole, and quinoline in aqueous solutions on a renewable bioresource of BC. BC exhibits a strong adsorption affinity to NHCs, and the adsorption process is predominantly regulated by surface diffusion in initial 1 h and by intraparticle diffusion in later stage. The adsorption isotherms are well described by Freundlich model, and the adsorption process is spontaneous and endothermic. Adsorption site energy analysis indicates the heterogeneous sites on BC for NHC adsorption. Spent BC with NHC adsorption can be effectively regenerated by MW radiation, and the adsorption capacity becomes even higher than that of virgin BC after five times of adsorption–regeneration cycles. Therefore, this new biosorbent is expected to have applicability in the removal of NHCs and probably other organic pollutants from wastewater.

[1] D. Mohan, K.P. Singh, S. Sinha, D. Gosh, Carbon 42 (2004) 2409–2421. [2] K.V. Padoley, S.N. Mudliar, R.A. Pandey, Bioresour. Technol. 99 (2008) 4029– 4043. [3] Y. Bai, Q. Sun, R. Sun, D. Wen, X. Tang, Environ. Sci. Technol. 45 (2011) 1940– 1948. [4] H. Zhang, G. Li, Y. Jia, H. Liu, J. Chem. Eng. Data 55 (2010) 173–177. [5] Y. Fan, B. Wang, S.H. Yuan, X.H. Wu, J. Chen, L.L. Wang, Bioresour. Technol. 101 (2010) 7661–7664. [6] Z. Tan, L. Sun, J. Xiang, H. Zeng, Z. Liu, S. Hu, J. Qiu, Carbon 50 (2012) 362–371. [7] P. Liao, M.I. Zainab, W.B. Zhang, S.H. Yuan, M. Tong, Chem. Eng. J. 195–196 (2012) 339–346. [8] Z. Tan, J. Qiu, H. Zeng, H. Liu, J. Xiang, Fuel 90 (2011) 1471–1475. [9] X.Z. Guo, L.J. Zhang, L.Q. Yan, H. Yang, Mater. Lett. 64 (2010) 331–333. [10] S.Y. Wang, M.H. Tsai, S.F. Lo, M.J. Tsai, Bioresour. Technol. 99 (2008) 7027– 7033. [11] K. Mizuta, T. Matsumoto, Y. Hatate, K. Nishihara, T. Nakanishi, Bioresour. Technol. 95 (2004) 255–257. [12] T. Asada, T. Ohkubo, K. Kawata, K. Oikawa, J. Health Sci. 52 (2006) 585–589. [13] D. Zhao, J. Zhang, E. Duan, J. Wang, Appl. Surf. Sci. 254 (2008) 3242–3247. [14] S.F. Li, D. Umereweneza, Sep. Sci. Technol. 47 (2012) 104–111. [15] P. Liao, S.H. Yuan, W.B. Zhang, M. Tong, K. Wang, J. Colloid Interface Sci. 382 (2012) 74–81. [16] Z.J. Wu, H. Joo, K. Lee, Chem. Eng. J. 112 (2005) 227–236. [17] M. Almarri, X.L. Ma, C.S. Song, Ind. Eng. Chem. Res. 48 (2009) 951–960. [18] M.A. Ferro-García, E. Utrera-Hidalgo, J. Rivera-Utrilla, C. Moreno-Castilla, J.P. Joly, Carbon 31 (1993) 857–863. [19] K. Okawa, K. Suzuki, T. Takeshita, K. Nakano, Water Res. 41 (2007) 1045–1051. [20] X. Quan, X.T. Liu, L.L. Bo, S. Chen, Y.Z. Zhao, X.Y. Cui, Water Res. 38 (2004) 4484–4490. [21] C.O. Ania, J.B. Parra, J.A. Menendez, J.J. Pis, Water Res. 41 (2007) 3299–3306. [22] X.T. Liu, X. Quan, L.L. Bo, S. Chen, Y.Z. Zhao, Carbon 42 (2004) 415–422. [23] J. Wang, X. Peng, Z. Luan, C. Zhao, J. Hazard. Mater. 178 (2010) 1125–1127. [24] Q.S. Liu, T. Zheng, N. Li, P. Wang, G. Abulikemu, Appl. Surf. Sci. 256 (2010) 3309–3315. [25] P. Oleszczuk, B. Pan, B.S. Xing, Environ. Sci. Technol. 43 (2009) 9167–9173. [26] X.L. Wang, L. Shu, Y.Q. Wang, B.B. Xu, Y.C. Bai, S. Tao, B.S. Xing, Environ. Sci. Technol. 45 (2011) 9276–9283. [27] Y.S. Ho, G. McKay, Process Biochem. 34 (1999) 451–465. [28] W.J. Weber, J.C. Morris, J. Sanitary Eng, Div. Proceed. Am. Soc. Civil Eng. 89 (1963) 31–59. } Yasemin, A. Mahir, Dyes Pigm. 75 (2007) 701–713. [29] D. Mehmet, O. [30] L.L. Ji, F.L. Liu, Z.Y. Xu, S. Zheng, D.Q. Zhu, Environ. Sci. Technol. 44 (2010) 3116–3122. [31] S.J. Zhang, T. Shao, S.K. Bekaroglu, T. Karanfil, Environ. Sci. Technol. 43 (2009) 5719–5725. [32] G. Limousin, J.P. Gaudet, L. Charlet, S. Szenknect, V. Barthes, M. Krimissa, Appl. Geochem. 22 (2007) 249–275. [33] Z.Y. Wang, X.D. Yu, B. Pan, B.S. Xing, Environ. Sci. Technol. 44 (2010) 978–984. } Demirbas, P. Turan, Chem. Eng. J. 139 (2008) [34] M. Alkan, M. Dogan, Y. Turhan, O. 213–223. [35] J. Wen, X. Han, H.F. Lin, Y. Zheng, W. Chua, Chem. Eng. J. 164 (2010) 29–36. [36] D. Rameshrajaa, V.C. Srivastavaa, J.P. Kushwahab, I.D. Mall, Chem. Eng. J. 181– 182 (2012) 343–351. [37] J.P. Kushwaha, V.C. Srivastava, I.D. Mall, Bioresour. Technol. 101 (2010) 3474– 3483. [38] M.C. Carter, J.E. Kilduff, W.J. Weber, Environ. Sci. Technol. 29 (1995) 1773– 1780. [39] C. Shen, Y. Shen, Y.Z. Wen, H.Y. Wang, W.P. Liu, Water Res. 45 (2011) 5200– 5210. [40] C.O. Ania, J.B. Parra, J.A. Menendez, J.J. Pis, Microporous Mesoporous Mater. 85 (2005) 7–15. [41] M. Miura, H. Kaga, A. Sakurai, T. Kakuchi, K. Takahashi, J. Anal. Appl. Pyrolysis 71 (2004) 187–199. [42] C.O. Ania, J.A. Menéndez, J.B. Parra, J.J. Pis, Carbon 42 (2004) 1383–1387. [43] D.M. Nevskaia, A. Santianes, V. Munoz, A.G. Ruiz, Carbon 37 (1999). 1065– 1047. [44] G.S. Miguel, S.D. Lambert, N.J.D. Graham, Water Res. 35 (2001) 2740–2748. [45] P. Chingombe, B. Saha, R. Wakeman, Carbon 43 (2005) 3132–3143. [46] F. Wang, J. Yao, K. Sun, B.S. Xing, Environ. Sci. Technol. 44 (2010) 6985–6991. [47] X.L. Wang, Y. Liu, S. Tao, B.S. Xing, Carbon 48 (2010) 3721–3728.