Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres

Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres

Accepted Manuscript Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres Jianguo Zhou, Yuef...

1MB Sizes 2 Downloads 55 Views

Accepted Manuscript Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres Jianguo Zhou, Yuefeng Wang, Jitong Wang, Wenming Qiao, Donghui Long, Licheng Ling PII: DOI: Reference:

S0021-9797(15)30247-2 http://dx.doi.org/10.1016/j.jcis.2015.10.001 YJCIS 20787

To appear in:

Journal of Colloid and Interface Science

Received Date: Revised Date: Accepted Date:

8 August 2015 30 September 2015 1 October 2015

Please cite this article as: J. Zhou, Y. Wang, J. Wang, W. Qiao, D. Long, L. Ling, Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres, Journal of Colloid and Interface Science (2015), doi: http://dx.doi.org/10.1016/j.jcis.2015.10.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres Jianguo Zhou, Yuefeng Wang, Jitong Wang, Wenming Qiao, Donghui Long*, Licheng Ling

State Key Laboratory of Chemical Engineering, East China University of Science and Technology, Shanghai 200237, China 

Corresponding authors: Donghui Long

Tel: +86 21 64252924. Fax: +86 21 64252914.

E-mail: [email protected];

1

Abstract: High-surface-area mesoporous carbon microspheres were successfully synthesized by a spraying method with the purpose of removing Cr(VI) from waste water. Various factors influencing the adsorption of Cr(VI), including pH, adsorption temperature, and contact time were studied. As the adsorption process was pH dependent, it showed maximum removal efficiency of Cr(VI) at pH 3.0. Pseudo-second-order model was found to best represent the kinetics of Cr(VI) adsorption. The adsorption parameters were determined using both Langmuir and Freundlich isotherm models, and Qm value was as high as 165.3 mg/g. The thermodynamic parameters including standard Gibb's free energy (ΔG0), standard enthalpy (ΔH0) and standard entropy (ΔS0) were investigated for predicting the nature of adsorption, which suggested the adsorption was an endothermic and a spontaneous thermodynamically process. Furthermore, Fe3O4-loaded MCMs were prepared to rapidly separate the adsorbent from the solution by a simple magnetic process. Fe3O4-loaded MCMs had a high adsorption capacity of 156.3 mg/g, and a good regeneration ability with a capacity of 123.9 mg/g for the fifth adsorption-desorption cycle.

Keywords: Hexavalent chromium removal; Adsorption; Mesoporous carbon microspheres; Magnetic separation

2

1. Introduction Chromium is a primary metal pollutant introduced into water bodies used by modern industries like plastic, pigment, wood preservative, electroplating, leather tanning, cement, mining, dyeing and fertilizer [1]. Chromium usually exists in both trivalent and hexavalent forms, and hexavalent chromium Cr(VI) is highly toxic which can cause severe diseases such as dermatitis, kidney circulation, lung cancer and even death [2]. For this reason, Cr(VI) is considered as a priority hazardous pollutant. According to the recommendation of The World Health Organization (WHO), the maximum allowable limit for Cr(VI) in drinking water is at the level of 0.05 mg/L. Therefore, it is of prime importance to remove Cr(VI) from wastewater before its emission to water bodies. Until now, various methods have been developed to deal with Cr(VI) from aqueous media, including chemical precipitation [3], ion exchange [4], membrane processes [5], electrodialysis [6], photocatalytic degradation [7] and adsorption [8,9]. Among these methods, adsorption is a preferred method due to its high efficiency, flexibility, simplicity of design and operation. What’s more it does not generate secondary pollutant during industrial operations. Activated carbons have been widely used for the adsorption of Cr(VI) in terms of their wide availability, developed physical and chemical properties [10, 11]. The micropores in activated carbons could physically adsorb Cr(VI) from water via the Van de Waals force. However, the small pore size and low pore volume within the activated carbons would lead to a quite slow mass transport rate. Mesoporous carbon is believed to be a promising material to solve these problems, because its relatively larger pore size would accelerate the diffusion of Cr(VI) ions while the developed pore volume could provide enough space room to accommodate more ions. Owing to its merits such as high specific surface area, well-developed mesoporous channel and large pore volume 3

[12], mesoporous carbon should display high adsorption capacity and fast adsorption rate. However, there are only limited reports on the Cr(VI) removal using mesoporous carbon materials [13, 14]. In addition, the mesoporous carbons are usually prepared in the form of powders or irregular particles. The additional control of the morphology may widen the applicability of these materials for practical applications, such as microspheres, because of their high stacking density, low flow resistance, good mechanical properties and facility in recovering [15, 16]. During waste water treatment, the rapid and facile separation of the exhausted adsorbents from the solid-liquid mixture is also desired. Magnetic adsorbents [17] could provide a preferred approach to meet the target since they could be separated by a simple magnetic process. In addition, this method is very cost-effective and environmental friendly. Magnetic adsorbents have been prepared and used for contaminants removal in aqueous media by introducing magnetic nanoparticles (Fe3O4, γ-Fe2O3 or α-Fe2O3) to pristine adsorbents, such as alumina, activated carbon and carbon nanotubes [18]. To date, magnetic mesoporous carbon have not been fully investigated for the removal of the Cr(VI). Recently, we reported a low-cost and high-throughput spray drying method to produce mesoporous carbon microspheres (MCMs) with smooth morphology and developed porous structure [16]. In view of the health problems caused by Cr(VI), the present study was performed to evaluate MCMs as adsorbents for the removal of Cr(VI) from polluted water by systematic evaluation of the parameters involved such as pH, time and temperature. The adsorption kinetics of Cr(VI) onto MCMs were evaluated by using pseudo-first-order and pseudo-second-order. In addition, the Langmuir and Freundlich adsorption isotherms were applied to study the kinetics of adsorption and to calculate isotherm parameters. Thermodynamic parameters such as standard Gibb's free energy (ΔG0), standard enthalpy (ΔH0) and standard entropy (ΔS0) for Cr(VI) adsorption were also 4

estimated. Furthermore, magnetic mesoporous carbon microspheres (M-MCMs) were prepared by Fe3O4 loading to separate the adsorbent from the solution conveniently and rapidly. The cost-effective performance and simple operation indicate the MCMs can be taken as a promising adsorbent for removal of Cr(VI) from wastewater for effluent treatment in industries.

2. Materials and methods 2.1 Preparation of MCMs The MCMs were prepared by a spray drying method using resorcinol-formaldehyde polymer as carbon source and colloidal silica sol as hard template, according to our previous report [16]. In a typical synthesis, the resorcinol and formaldehyde with mole ratio of 1:2 were added into the silica sols with continuous magnetic stirring, after which the sol was further diluted with distilled water to the desired content (%w/v). The mixed sol was stirred for 1 h at 40 °C, and then spray-dried using a spray-dryer (Blon-6000y, Shanghai Bilon Instrument Co., Ltd). The solution was pumped into the nozzle at the rate of 500 mL/h, together with a constant spray air flow. The liquid was atomized into fine droplets at a constant pressure of 0.3 MPa. The inlet temperature was set to 120 °C, while generally the outlet temperature was in the range of 58-65 °C. Fine powder was discharged continuously from the drying chamber and then collected using a cyclone separator. The obtained powders were then carbonized in a nitrogen flow at 800 °C for 3 h with a heating rate of 5 °C/min. Finally, the MCMs were obtained by the dissolution of silica nanoparticles in 2 M NaOH solution at 80 °C, isolated by filtration, washed with distilled water, and dried at 100 °C. Herein, three samples were prepared and named as MCM-1, MCM-2 and MCM-3, responding to the weight ratio of resorcinol-formaldehyde/SiO2 in the precursor solution of 0.5, 0.75 and 1.0, 5

respectively. 2.2 Preparation of Fe3O4-loaded MCMs The Fe3O4 loaded MCMs sample was prepared by an incipient wetness impregnation. In a typical procedure, 1.0 g of MCM-1 were dipped into a solution with 0.58 g of Fe(NO3)3·9H2O and well mixed for 1 h. The mixture were dried at ambient temperature for one night and then at 80 °C for 24 h, finally heat-treated at 400 °C for 4 h in nitrogen flow. The resultant Fe3O4 loaded MCMs were named MCM-1-Fe. 2.3 Characterization The X-ray diffraction (XRD) patterns were acquired on a Rigaku D/max 2550 diffractometer operating at 40 KV and 20 mA using Cu Kα radiation (λ = 1.5406 Å). The thermogravimetric analysis (TA Instrument Q600 Analyzer) was carried out in an air flow rate. The samples were heated to 800 °C at a rate of 10 °C /min. Nitrogen adsorption/desorption isotherms were measured at 77 K with a Quadrasorb SI analyzer. Before the measurements, the samples were degassed in vacuum at 473 K, respectively for at least 12 h. The Brunauer-Emmett-Teller (BET) method was utilized to calculate the specific surface area (SSA). The total pore volume was calculated at relative pressure of 0.985. The average pore size was derived from desorption branch by using the Barrett-Joyner-Halenda (BJH) model. 2.4 Adsorption Experiments A stock solution (100 mg/L) of Cr(VI) was prepared by dissolving analytical grade K2Cr2O7 into distilled water. The test solution of Cr(VI) used in each study was prepared by diluting the stock solution. The pH value for test solution was adjusted by 0.1 M HCl and 0.1 M NaOH solution. In a typical experiment, a definite amount 0.2 g/L (10 mg) of the adsorbents was added into a capped 6

conical flasks with 50 mL of 50 mg/L Cr(VI) solution at pH=3 for adsorption experiments and shaken at 100 rpm using a constant temperature bath oscillator at 25 oC (SHA-C, China). A number of experimental variables such as contact time (0-24 h), pH (1-9), concentration (5-100 mg/L) and temperature (15-45 oC) had been studied to investigate the Cr(VI) removal process. To investigate the regenerated performance of the MCM-Fe, 0.5 M Sodium hydroxide solution was used for the desorption process to regenerate the MCM-Fe with accumulated chromium after a water rinse, and then followed by immersing in a 0.5 M hydrochloric acid solution for 2h. After filtration, rinsing to neutral and drying, the regeneration MCM-Fe was obtained. Concentrations of hexavalent chromium in the filtrate were determined spectrophotometrically using diphenylcarbazide solution at λ = 540 nm [19]. In order to model the adsorption behavior and calculate the adsorption capacity of MCMs, adsorption isotherms were studied. Both Langmuir and Freundlich isotherms [20] were employed for Cr(VI) adsorption. The two models can be expressed as Eqs. (1) and (2), respectively:

where, Ce (mg/L) is the equilibrium concentration of the adsorbate; qm (mg/g) is the maximum adsorption capacity of the adsorbent; b (L/mg) is a constant related to the adsorption energy; qe (mg/g) is the equilibrium absorption capacity of the adsorbent; KF is a constant related to the adsorption capacity of the adsorbent, and n is a constant related to the adsorption intensity. The study of adsorption kinetics describes the solute uptake rate. This latter controls the residence time of adsorbate up-take at the solid-solution interface, including the diffusion process. The mechanism of adsorption depends on the physical and chemical characteristics of the 7

adsorbent as well as on the mass transfer process. The results obtained from the experiments were used to study the kinetics chromium adsorption. The rate kinetics of Cr(VI) adsorption on the MCMs, were analyzed using pseudo-first-order [21] and pseudo-second-order models[22] . The agreement between experimental data and the model predicted values was expressed by correlation coefficients (R2). An attempt was made to fit the obtained data to Eqs. 3 and 4, which are the standard linearized-integral forms of pseudo-first-order and pseudo-second-order kinetic models, respectively, (3) Where qt is amount of Cr(VI) (mg/g of adsorbent) removed at time t (min), qe is the amount of Cr(VI) removed at equilibrium and k1 is the rate of adsorption (min-1).

Where k2 is the pseudo-second-order rate constant of adsorption (g/mg/min). Thermodynamic parameters such as standard Gibb's free energy (ΔG0), standard enthalpy (Δ H0) and standard entropy (ΔS0) for Cr(VI) adsorption were also estimated. The Gibbs free energy changes of the adsorption are determined using Eq. (5) and K0 values are calculated extrapolating the ln (Qe/Ce)-Ce plot [23].

The enthalpy changes (ΔH0) and entropy changes (ΔS0) are calculated using Van’t Hoff equation (Eq. (6)).

8

3. Results and discussion 3.1 Characterization of the MCMs

Fig.1 SEM images (a, b) of MCM-1; TEM image (c) of MCM-1; Nitrogen adsorption–desorption isotherms (d), and the resulting pore size distribution curves (e) of MCMs.

The MCMs were synthesized by a scalable spray-drying technique as we previous reported [16]. The key to our synthesis lies in forming and heating atomized droplets of an aqueous solution containing the resorcinol-formaldehyde precursors and silica sols. Thus polymer/silica microspheres were continuously produced via a spray dryer, which can easily amplify on a large scale by an industrial instrument. After carbonization and silica removal, MCMs were obtained. The pore structure of MCMs was reversely replicated from the colloidal silica nanoparticles, thus the porosity could be adjusted by changing the particle size of colloidal silica or the mass ratio of silica/polymer precursors. Herein, three samples with different mesoporous structures were prepared by changing the weight ratio of polymer/SiO2 (0.5, 0.75 and 1.0) for Cr(VI) removal. 9

Table 1 Pore structure parameters of initial and Fe-loaded MCMs SBETa

Smicb

Vtotalc

Vmicd

DBJHe

(m2/g)

(m2/g)

(cm3/g)

(cm3/g)

(nm)

MCM-1

1221

594

2.7

0.253

9.5

MCM-2

1177

377

1.9

0.168

7.7

MCM-3

1006

382

1.4

0.143

7.7

MCM-1-Fe

1061

353

2.3

0.157

9.5

Sample

a

BET surface area. bMicropore surface area calculated by t-plot method. cTotal pore volume.

d

Micropore volume. eBJH desorption average pore diameter.

As illustrated in Fig.1a, the obtained MCMs show spherical morphology, smooth surface and non-aggregation characteristics with the particle size in the range of 1-8 μm. Unique mesopores could be easily observed on the external surface of MCMs (Fig. 1b). TEM image (Fig. 1c) shows that the MCMs exhibit a three dimensional wormhole-like structure in which the mesopores are linked with each other. The pore structures of MCMs are further analyzed by using N2 adsorption. As shown in Fig.1d, all the MCMs exhibit similar type-IV isotherms with relatively narrow hysteresis loops located at P/P0 of 0.7-0.9, responding to the typical mesoporous structures. Adsorption platforms at low relative pressures (P/P0<0.1) are also observed, suggesting the presence of large amounts of micropores. The BJH pore size distributions in Fig. 1e confirm the MCMs have the similar mean mesopore sizes of ca. 8-10 nm, which is approximately equal to the average diameter of single silica nanoparticle (7 nm). The detailed porosity parameters are given in Table 1. Obviously, increasing the ratio of RF polymer to silica leads to decreases in both BET surface areas and pore volumes. It 10

should be noted that all these microspheres exhibit relatively high BET surface areas (up to1000 m2/g), being close to these of the commercial activated carbons (800-1200 m2/g).

Fig.2 SEM images of MCM-1-Fe (a), (b); Elemental mapping images of MCM-1-Fe (c); XRD of MCM-1 and MCM-1-Fe (d), and TGA curves of MCM-1 and MCM-1-Fe (e).

In order to separate the adsorbent from the solution conveniently and rapidly, magnetite nanoparticle was used to modify the MCMs. SEM images of the Fe3O4-loaded MCMs are shown in Fig.2 (a) and (b). It can be seen the Fe3O4-loaded MCMs remain a completely spherical shape, non-aggregation appearance. Elemental mapping images (Fig. 2c) of MCM-Fe indicate that C and Fe are homogeneously distributed throughout the microspheres. XRD patterns (Fig. 2d) of Fe3O4-loaded 11

MCMs show the presence of six characteristic Fe3O4 peaks (JSPDS85-1436), which are located at 30.16◦ (2 2 0), 35.68◦ (3 1 1), 43.18◦ (4 0 0), 53.8◦ (4 2 2), 57.14◦ (5 1 1) and 62.82◦ (4 4 0) [24]. The magnetization curve of the MCM-Fe was measured and shown in Fig. S1. The porous parameters of the MCM-Fe are obtained using N2 adsorption, as shown in Table 1. After Fe3O4 loading, the pore volume and surface area slightly decrease to 2.3 cm3/g and 1061 m2/g, respectively, due to the pore filling effect. The Fe3O4 content of the sample is determined to be ca 10% using thermogravimetric analysis (TGA) (Fig. 2e), which is in good agreement with the designed Fe3O4 loading.

3.2 Adsorption experiments 3.2.1 Effect of pH on Cr(VI) removal

Fig. 3 Effect of pH on Cr(VI) adsorption by MCM-1with Cr(VI) concentration of 50 mg/L and carbon dosage of 10 mg at 298K.

The pH of the aqueous solution is an important controlling parameter in the heavy metal adsorption process. The role of pH was firstly examined at the same Cr(VI) concentration and carbon 12

dosage and different initial pH values, as shown in Fig.3. The adsorption capacity of Cr(VI) increases from 100 mg/g (pH 1) to 122 mg/g (pH 3), and thereafter it decreases with further increase in pH. The reason of this pH-depended adsorption should be due to the surface chemistry of adsorbent and the forms of the dissolved ions which are greatly influenced by pH [23]. Under strongly acidic condition, the carbon surface should be protonated by the large numbers of H + ions, which is favor the electrostatic attraction between Cr(VI) (in the form of HCrO4-) and the charged surface. When increasing pH values from 3 to 9, the HCrO4- gradually converts to the divalent CrO42-. Meanwhile, the protonation degree of carbon surface is also weakened under weak acidic conditions, which result in increased electrostatic repulsion between more negatively charged ions and carbon surface. Thus the overall Cr(VI) removal by MCM-1 is higher at low pH values and the highest adsorption occurs at the pH value of 3. So all the experiments in the next will take place at pH=3. The similar pH-depended trend was also observed by some other researchers [25, 26]. 3.2.2 Kinetic study

Fig. 4 Kinetics curves of Cr(VI) adsorption by MCM-1, MCM-2 and MCM-3 at 298 K 13

Table 2 Kinetic parameters of Cr(VI) adsorption by MCM-1, MCM-2 and MCM-3 Pseudo-first-order

qe,exp

Pseudo-second-order

Sample mg/g

k1

qe (mg/g)

R2

k2

qe (mg/g)

R2

MCM-1

139.3

0.196

103.9

0.983

0.00496

144.9

0.994

MCM-2

125.1

0.185

89.3

0.976

0.00581

129.9

0.994

MCM-3

115.5

0.169

75.7

0.977

0.00692

119.0

0.995

The effect of contact time on adsorption of Cr(VI) by MCMs at pH of 3.0 was investigated and the results are shown in Fig. 4. All the three samples exhibit quite fast adsorption rate for Cr(VI) in the initial stage, and the adsorbing capacities could reach ca. 60 mg/g at the end of 1 h. This value is significantly higher than these of activated carbons (Table S1) [27-29]. When further prolonging the contact time, the adsorption rates become obviously slow so equilibrium is acquired till the end of 24 hours. In order to study the adsorption kinetics of Cr(VI) onto MCMs, the frequently used models, known as pseudo-first-order and pseudo-second-order, are used to fit the kinetics data. The value of regression coefficients and rate constants calculated from the fitting plots (Fig. S2 and S3) are tabulated in Table 2. The R2 value of the pseudo-second-order model (0.994) suggests that this could describe the adsorption of Cr(VI) better than the pseudo-first-order model with the R2 value of 0.97-0.98. Therefore it can be concluded that the rate limiting step in the adsorption is mainly chemisorption which involves valence forces, occurred possibly due to sharing or exchange of electrons between MCMs and Cr(VI).

14

3.2.3 Adsorption isotherm

Fig. 5 Adsorption isotherm of Cr(VI) adsorption by MCM-1, MCM-2 and MCM-3 at 298 K

Table 3 Fitting parameters of adsorption isotherm Langmuir

Temperature Adsorbents

2

2

b

qm (mg/g)

RL

KF

n

RF

MCM-1

0.831

165.3

0.995

81.7

5.85

0.998

MCM-2

0.839

145.6

0.996

72.6

5.90

0.995

MCM-3

0.873

131.9

0.997

67.6

6.15

0.987

K

298

Freundlich

The adsorption isotherm experiments were conducted at constant temperature (298 K) and pH (3). The results are plotted in Fig.5 and the adsorption capacity follows an order of MCM-1 > MCM-2 > MCM-3. As we all know, surface area and pore volume of adsorbents play both important roles in the adsorption process. The MCM-1 possesses the highest BET surface area and largest total pore 15

volume, which could provide abundant adsorption sites for Cr(VI). The shapes of isotherms suggest that there are high-energy adsorption sites to favor strong adsorption at low equilibrium concentrations for the MCM. To analyses the adsorption data, the Langmuir and Freundlich isotherm equations are used and calculated parameters from the fitting plots (Fig. S4 and S5) are given in Table 3. The results show that both Langmuir and Freundlich models well follow Cr(VI) adsorption by MCMs with all R2 values over 0.98. However, the correlation coefficients of Langmuir model are a little higher, suggesting the Langmuir isotherm correlated better than Freundlich with the experimental data. The Langmuir model suggests homogeneous surface and monolayer adsorption process which indicates active sites are well distributed throughout the MCMs. The comparison of maximum adsorption capacity of the MCMs with that of other various adsorbents is represented in Table S1. As represented in this table the adsorption capacity of MCM for Cr(VI) is generally higher than these of other adsorbents (< 79 mg/g). The usually used adsorbents with low area surface and small pore volume can’t provide enough adsorption activated site, and the narrow pore channel will prevent Cr(VI) diffusion to the adsorption activated site, so few Cr(VI) could be adsorbed. Comparing with other adsorbents including activated carbons, the MCMs have considerable BET area surface but larger pore volume and pore size, which could provide large amount of activated site for Cr(VI) adsorption and the mesopores can offer unobstructed diffusion path so the Cr(VI) ion can diffuse to the activated site through the mesopores without blockage, and a high adsorption capacity is achieved easily. The adsorption capacity varies and depends mainly on the initial Cr(VI) concentration and characteristics of the individual adsorbent. Nevertheless, the current experiments are carried out to confirm the applicability of as-prepared mesoporous carbon adsorbents to effectively remove Cr(VI) from the solution. 16

3.2.4 Thermodynamic study

Fig. 6 Adsorption isotherms of Cr(VI) adsorption by MCM-1 at different temperature

Table 4 Thermodynamic parameters of Cr(VI) adsorption by MCM-1 Temperature

ΔG0

ΔH0

ΔS0

kJ/mol

kJ/mol

J/mol•K

42.5

206.3

ln K0 K 288

7.10

-17.0

298

7.58

-18.1

308

8.30

-19.9

318

8.72

-20.9

The influence of temperature on the adsorption of Cr(VI) by MCM-1 is presented in Fig.6. Experimental results at different temperatures show that Cr(VI) adsorption is slightly increasing with temperature, similarly to the case of Cr(VI) adsorption on the activated carbons [11], but in apparent contradiction with the exothermic character of adsorption phenomena. The overall adsorption 17

process is thus referred as apparently endothermic process. The increase in adsorption with temperature may be attributed to either kinetic effects due to enhanced ion diffusion or to the “activation” of new adsorption sites on carbon surface at higher temperature. This result could be further confirmed by the thermodynamic parameters evaluated of adsorption. The results obtained after regression analysis of the data are shown in Fig. S6 and Table 4. The negative values of ΔG0 suggest the adsorption process is thermodynamically feasible and spontaneous within the temperature range (288 K-318 K). While the positive value of ΔS0 demonstrates the increased randomness at the solid-liquid interface during the adsorption of Cr(VI) onto MCMs. The value of enthalpy change is +42.5 kJ/mol, indicating the adsorption process should belongs to a combination of chemical adsorption and complexation, whose enthalpy change is usually >29 kJ/mol and 8-60 kJ/mol.

18

3.3 Magnetic separation

Fig. 7 Adsorption isotherms of Cr(VI) adsorption by MCM-1-Fe at 298 K (a); Regeneration studies of Cr(VI) adsorption by MCM-1-Fe with five cycles (b); Photographs of MCM-1-Fe dispersed in water (c) and interaction between the magnet and dispersed MCM-1-Fe (d).

The adsorption isotherms at 298 K of Cr(VI) adsorption by Fe3O4 -loaded MCMs are shown in Figure 7a. Compared to the pristine MCM-1, the Fe3O4 loading causes a slightly decrease in adsorption capacity, which can be explained by the decreasing of the pore structure parameters. It should be noted that after loading10 wt. % Fe3O4, the MCMs can still maintain a good performance of 156.3 mg/g for Cr(VI) removal, higher than the most reports results in activated carbons (Table S1). Owing to their good magnetic response (Fig. S1), the exhausted Fe3O4-loaded MCMs could be easily separated from the solution by magnetism. Figure 7(c) and (d) shows the digital images of 19

separation of MCM-1-Fe from aqueous dispersion using an external magnet. These microspheres are readily isolated from their dispersion within 1 min. The isolated microspheres can be regenerated by using 0.5 M NaOH as eluent. As shown in Figure 7b, the Cr(VI) adsorption capacity still remain 123.9 mg/g, with 80% of the initial adsorption capacity after the fifth recycle. These unique properties ensure a rapid switch in the separation-regeneration process that can afford a fast and efficient separation with the aid of a magnetic field.

4. Conclusions In conclusion, mesoporous carbon microspheres with high surface area and large pore volume are prepared by a simple and scalable spray drying method. These obtained carbon microspheres are employed as the adsorbent for Cr(VI) adsorption application. The effect of solution pH is very important for Cr(VI) adsorption. At pH 3, maximum adsorption capacity is reached with the Qe value of 122 mg/g. Adsorption kinetics for Cr(VI) is well fitted by pseudo-second-order kinetic model. The Langmuir isotherm model gives the best fit to equilibrium experimental data and Qm value is obtained at 298K for MCM-1 (165.3 mg/g). The thermodynamic parameters like standard Gibb's free energy (ΔG0), standard enthalpy (ΔH0) and standard entropy (ΔS0) are investigated for predicting the nature of adsorption. The introducing of Fe3O4 particle could separate the adsorbent from solution conveniently and rapidly with only a slightly reducing of adsorption capacity. Easy synthesis of MCMs, high chromium adsorption capacity, separate the adsorbent from the solution conveniently and rapidly, and good recovery make these adsorbents being potential alternatives for removal of toxic chromium ion from wastewater.

20

Acknowledgments This work was partly supported by MOST (2014CB239702) and National Science Foundation of China (No. 51302083, No. 51172071, No.51272077), and Fundamental Research Funds for the Central Universities and Shanghai Raising Star Program.

References 1. H. S. Altundogan, Process Biochemistry 2005, 40, 1443-1452. 2. J. Kota; Z. Stasicka, Environmental Pollution 2000, 107 (3), 263-283. 3. A. Verma; S. Chakraborty; J. K. Basu, Separation and Purification Technology 2006, 50 (3), 336-341. 4. S. Rengaraj; K. H. Yeon; S. H. Moon, Journal of Hazardous Materials 2001, 87 (1-3), 273-287. 5. R. A. Kumbasar, Separation & Purification Technology 2008, 64 (1), 56-62. 6. C. Peng; H. Meng; S. Song; S. Lu; A. L. Valdivieso, Separation Science & Technology 2005, 39 (7), 1501-1517. 7. Q. Cheng; C. Wang; K. Doudrick; C. K. Chan, Applied Catalysis B Environmental 2015, 176-177, 740-748. 8. F. Hussain; G. Sheng; Z. Gaoke, Journal of Colloid & Interface Science 2015, 442, 30-38. 9. J. H. Wang; E. Chang; S. L. Ding; C. L. Zheng, Environmental Science & Technology 2012, 35 (2), 67-72. 10. D. Mohan; K. P. Singh; V. K. Singh, Industrial & Engineering Chemistry Research 2005, 44 (4), 1027-1042. 21

11. F. Di Natale; A. Erto; A. Lancia; D. Musmarra, Journal of Hazardous Materials 2015, 281, 47-55. 12. M. Hartmann; A. Vinu; G. Chandrasekar, Chemistry of Materials 2005, 17 (4), 829-833. 13. H. Chen; T. Yan; F. Jiang, Journal of the Taiwan Institute of Chemical Engineers 2014, 45 (4), 1842-1849. 14. J. Zhu; H. Gu; J. Guo; M. Chen; H. Wei; Z. Luo; H. A. Colorado; N. Yerra; D. Ding; T. C. Ho, Journal of Material Chemistry A 2014, 2 (7), 2256-2265. 15. A. A. Deshmukh; S. D. Mhlanga; N. J. Coville, Materials Science & Engineering R Reports 2010, 70, 1-28. 16. X. Li; J. Zhou; J. Wang; W. Qiao; L. Ling; D. Long, RSC Advances 2014, 4, 62662-62665. 17. P. Xu; G. M. Zeng; L. H. Dan; L. F. Chong; H. Shuang; H. Z. Mei; L. Cui; W. Zhen; H. Chao; X. X. Geng, Science of the Total Environment 2012, 424 (4), 1-10. 18. S. Qu; J. Wang; J. Kong; P. Yang; G. Chen, Talanta 2007, 71 (3), 1096-1102. 19. R. Klimaviciute; J. Bendoraitiene; R. Rutkaite; A. Zemaitaitis, Journal of Hazardous Materials 2010, 181 (1-3), 624-32. 20. Z. Aksu; E. Balibek, Journal of Hazardous Materials 2007, 145 (1-2), 210-220. 21. L. N. Shi; Z. Xin; Z. L. Chen, Water Research 2011, 45 (2), 886-892. 22. P. Miretzky; A. F. Cirelli, Journal of Hazardous Materials 2010, 180 (1), 1-19. 23. D. Mohan; C. U. Pittman, Journal of Hazardous Materials 2006, 137 (2), 762-811. 24. D. H. Chen; M. H. Liao, Journal of Molecular Catalysis B Enzymatic 2002, 16 (9), 283-291. 25. M. Ma; Y. Lu; R. Chen; L. Ma; Y. Wang, Open Journal of Applied Sciences 2014, 4 (5), 275-284. 22

26. A. S. Bayazit; Ouml; Z. Kerkez, Chemical Engineering Research & Design 2014, 92 (11), 2725-2733. 27. L. Khezami; R. Capart, Journal of Hazardous Materials 2005, 123 (1), 223-231. 28. K. Selvi; S. Pattabhi; K. Kadirvelu, Bioresource Technology 2001, 80 (1), 87-89. 29. I. Zinicovscaia; T. Mitina; T. Lupascu; G. Duca; M. V. Frontasyeva; O. A. Culicov, Water, Air, & Soil Pollution 2014, 225 (3), 1-6.

23

Graphical abstract

Effective removal of hexavalent chromium from aqueous solutions by adsorption on mesoporous carbon microspheres

High-surface-area mesoporous carbon microspheres (MCMs) were successfully synthesized by spraying method with the purpose of removing Cr(VI) from waste water. The adsorption capacity could be achieved as high as 165.3 mg/g which is among the highest values for the state-of-the-art adsorbents.

24