Sisal fiber-based solid amine adsorbent and its kinetic adsorption behaviors for CO₂

Abstract

A solid amine adsorbent based on sisal fiber was prepared by grafting copolymerization, in which amine was covalently bound onto sisal fiber. In the absence of moisture, the adsorption capacity of the triethylenetetramine-aminated solid amine fiber (SF-AM-TETA) for CO₂ could be 0.75 mmol g⁻¹ through physical adsorption of pores inherited from the bio-texture of sisal fibers. The adsorption capacity of SF-AM-TETA was greatly promoted by the presence of moisture, which could reach 4.20 mmol g⁻¹ at room temperature. After 7 adsorption–desorption cycles, the regenerated fiber showed a adsorption capacity of 3.92 mmol g⁻¹, 93% regeneration efficiency, which confirmed that SF-AM-TETA was capable of keeping its stability and CO₂ adsorption capacity even after numbers of regeneration cycles. Furthermore, in order to better interpret the adsorption behavior and adsorption kinetics of the solid amine fiber SF-AM-TETA, two models were applied to fit CO₂ adsorption experimental data at different temperatures. The results indicated that the solid amine fiber tended to absorb CO₂ by a physical process in the absence of moisture while the chemical absorption was the main process occurring in the presence of moisture.

---

1. Introduction

Carbon dioxide Capture and Storage (CCS) is the most effective method among numerous CO₂ emission reduction methods.^1–3 The adsorption method using solid amine as adsorbent is recognized as the most promising technique for the separation of CO₂ after combustion.^4–6 Solid amine adsorbents have been developed through immobilizing amines onto the solid supports like active carbon,^7,8 zeolite,^9,10 and porous silica via physical or chemical methods,^11–14 and have been proved to be good CO₂ adsorbents. Compared to granular supports, the fibrous adsorbents enjoy many advantages, such as large external surface, short transit distance, low pressure drops, high chemical stability and desirable flexibility, which make them efficient adsorbents for CO₂. In our previous works, various fibrous solid amine fibers have been successfully prepared, in which amine were coated or grafted onto fibers including glass fiber,^15,16 polyacrylonitrile fiber,^17,18 polypropylene fiber.^19–22 Recently, viscose fiber (VF)^23,24 was chosen as matrix to prepared solid amine fiber, and the viscose fiber-based solid amine fiber showed a better adsorption performance due to its hydrophilicity. Lin et al. reported that VF-based solid amine fibers²³ showed a better adsorption performance compared with PP-based solid amine fiber, Wu et al.²⁴ further confirmed that well hydrophilicity of the matrix had a positive effect on the CO₂ adsorption through comparative analysis of hygroscopic rate. Xu et al.²⁵ concluded hydroxyl groups of adsorbents would not absorb CO₂ themselves, but could promote CO₂ chemical adsorptions with less activated energy.

As a vegetable fiber, sisal fiber is a kind of renewable resource. It not only possesses the advantage of viscose fiber, such as plentiful hydroxyl groups, good hydrophilicity, it also inherit organic texture of vegetable, such as plentiful pores and large external surface, all of which would be of benefit to the adsorption of CO₂. It will be a sustainable way to prepare an adsorption material for CO₂ capture by using low cost and renewable sisal fiber as raw materials.

The aim of the present study is to investigate the application of sisal fiber in the preparation of the adsorbent for carbon dioxide capture, evaluating the effect of amino density and surface chemistry on the CO₂ adsorption properties. Despite the experimental studies carried on so far, direct modification of nature vegetable fibers, including sisal fiber to prepare solid amine fiber are seldom reported in the application of CO₂ adsorption.²⁶ Most of the studies reported in the literatures focused on application of this biomass as bio-adsorbent for the removal of dye,^27–29 and heavy metal ions.^30–32 In this paper, a sisal-based solid amine fiber was prepared by grafting acrylamide onto sisal fiber, and then followed by aminating the grafted sisal fiber with amine reagents. The surface chemistry of fiber was tailored. The synergistic effect of surface chemistry and bio-texture of fibers on the CO₂ adsorption capacity and kinetics were evaluated.

2. Experimental

2.1 Materials and reagents

Sisal fiber (SF) was purchased by Dongfang Sisal Group Co. Ltd. Guangdong Province, and cut into 2.5–3 cm, soaked with sodium hydroxide (NaOH) at room temperature to break the lignin seal and disrupt the crystalline structure of cellulose. Acrylamide (AM), ethylenediamine (EDA), diethylenetriamine (DETA), triethylenetetramine (TETA), tetrathylenepentamine (TEPA), ferrous ammonium sulfate (FAS), hydrogen peroxide (H₂O₂, 30%) and aluminium chloride hexahydrate (AlCl₃·6H₂O) with analytic reagent grade were purchased from Guangzhou Reagent Company. Deionized water was used to prepare all solutions in the study.

2.2 Preparation of solid amine adsorbent

1 g alkali-treated sisal fiber was immersed in 50 mL deionized water, and stirred for 30 min while N₂ was bubbled to remove oxygen in the solution. After that, 10 mL 6% (w/w) FAS and 0.5 mL 2% (v/v) H₂O₂ was added as initiator, and 10 min later, 10 mL 7% (w/v) AM monomer was added into the reaction solution. The graft polymerization was performed at 70 °C for 4 h. The resulting fiber, polyacrylamide-grafted sisal fiber (SF-AM), was washed with deionized water at 80 °C to remove the residual monomers and homopolymers, and then dried at 60 °C under vacuum. Graft yield (G%) and graft efficiency (E%) were calculated according to the following eqn (1) and (2).where W₀ and W_g are the weights of SF and SF-AM, respectively. And m is the mass of AM.

1 g SF-AM were added into 50 mL 80% (v/v) TETA solution, 4% (w/v) AlCl₃·6H₂O was then added as the catalyst of aminating reaction. The mixture was reacted at 110 °C for 8 h. Then, the amine modified fiber was washed thoroughly with deionized water and dried at 60 °C under vacuum. The obtained fiber, called as solid amine fiber, was labelled as SF-AM-TETA. Similar processes were taken when using other amine agent for amination, such as EDA, DETA, and TEPA. And the solid amine fiber were defined as SF-AM-X (X represents EDA, DETA, and TEPA). Amination rate Da (%) of the resulting solid amine adsorbent was calculated according to the following eqn (3).

where W_g and W_a are weights of the SF-AM and SF-AM-X, respectively. And G is the grafting degree of SF-AM. 71 and 17 are the molecular mass (g mol⁻¹) of AM and NH₃, respectively. M_a is the molecular mass (g mol⁻¹) of amine agent.

2.3 Structure characterization

Nitrogen, hydrogen, and carbon content of the fiber samples were determined by Perkin-Elmer Elemental Analyzer Vario EL (Germany). FT-IR spectra were collected from 4000 to 400 cm⁻¹ on a Nicolet/Avatar 330 FT-IR spectrometer. Thermal stability of all samples was evaluated with a TGA analyzer (Netzsch TG-20), in which samples were heated under a nitrogen atmosphere from ambient temperature to 600 °C with a heating rate of 10 °C min⁻¹. Morphology of fibers, sputter coated with Au, were observed with a field emission scanning electron microscope (Hitachi S-4800, Japan).

2.4 CO₂ adsorption procedures

1.0 g solid amine adsorbent was placed in an adsorption column (Φ = 1.3 cm). After air in the column and pre-adsorbed species were thoroughly removed by a dry N₂ flow with a flow rate of 30 mL min⁻¹, the CO₂/N₂ mixture gas was introduced into the fixed bed to perform the adsorption. The inlet and outlet concentration of CO₂ was determined by a gas chromatograph (D7900, Techcomp, China) with a thermal-conductivity detector (TCD). The adsorption temperature was adjusted by a water bath. The inlet concentration of CO₂ was adjusted to 10% in volume ratio.where Q is the adsorption capacity of solid amine adsorbent (mmol CO₂ per g), t is the adsorption time (min), and C_in and C_eff were the influent and effluent flow rate of CO₂ (mL min⁻¹), respectively. V is the total flow rate, 30 mL min⁻¹; W and 22.4 are the weight of solid amine adsorbent (g) and molar volume of gas (mmol mL⁻¹), respectively. After adsorption, pure nitrogen gas at a flow rate of 30 mL min⁻¹ was introduced through the column at 90 °C for 30 or 60 min to regenerate the spent adsorbent sample.

3. Results and discussion

3.1 Chemical characterization of SF and modified fibers

The chemical compositions and amino group contents of the fibers were measured using elemental analysis. As shown in Table 1, the nitrogen content of SF-AM reached to 6.30 wt% through grafting polymerization with AM. Assuming all nitrogen on SF-AM was derived from grafted AM, the amide group content of SF-AM was calculated to be 4.50 mmol g⁻¹. After amination, total nitrogen content of the solid amine fiber SF-AM-TETA had further increased to 10.68 wt%. Since the increment of nitrogen content of SF-AM-TETA was stemmed from alkyl amino groups of TETA, the acquired alkyl amino content of SF-AM-TETA based on elemental analysis result could be 4.66 mmol g⁻¹ according to the calculation.

Table 1 Elemental analysis of SF and modified fibers

Fibers	Element content (wt%)			Group content (mmol g^−1)
	C	H	N	Amide	Alkyl amino
SF	41.51	6.58	0.00
SF-AM	42.31	7.02	6.30	4.50	—
SF-AM-TETA	42.97	8.18	10.68	2.97	4.66

---

The chemical structures before and after modification of SF were characterized by using FT-IR. As shown in Fig. 1, the peaks of 898, 1028, and 1167 cm⁻¹ in the spectrum of SF were related to bending vibration of C–H bond and asymmetric stretching vibrations of –C–O bond, which are the typical value of cellulose. The peak at 1738 cm⁻¹ was related to the vibrations of –COO formed between the carboxylic acid group in hemicellulose and the phenolic hydroxyl group in lignin.³³ The broad peak of 3418 cm⁻¹ was related to the stretching vibrations of –O–H bond. No obvious change of IR spectrum of alkali treated SF (identified as ASF in Fig. 1) was observed, which indicated that the property of the cellulose was remained after it was treated with NaOH. In the spectrum of SF-AM, the intensity of the peak of 1652 cm⁻¹ which was attributed to the stretching vibrations of –C=O was enhanced, what's more, the intensity of the peak of 3418 cm⁻¹ was also enhanced, which was attributed to the overlapping of the vibrations of –O–H bond and –N–H. The peak of 1678 cm⁻¹ and 1612 cm⁻¹ was related the amide I and II regions. These characteristic peaks indicated that AM was grafted onto sisal fibers. The peaks of 3418 cm⁻¹, 1678 cm⁻¹ and 1612 cm⁻¹ in the spectrum of SF-AM-TETA was stronger than those in the spectrum of SF-AM, which were contributed to the vibrations of –N–H. Combined with the result of elemental analysis, the successful preparation of SF-AM-TETA was further verified.^18,23,34

Fig. 1 FT-IR spectra of SF and its modified fibers.

As shown in Fig. 2, there was no obvious mass loss of the four samples till they were heated to 280 °C, which indicated that SF and the modified fibers remained good thermal stabilities when the temperature was lower than 280 °C. There were two mass loss platforms for the curve of SF in the region of 300–380 °C, which was contributed to the decomposition of pectin, hemicellulose and cellulose.³⁵ After alkali treatment, there was only one mass loss platform in the curve of alkali treated SF (identified as ASF in Fig. 2) in that region. It indicated that the impurities had been removed from ASF. In the region of 380–500 °C, both the curves of SF and ASF showed a gradual mass loss because of the decomposition of lignin. Grafting with AM and amination seemed to improve the thermal stability of SF-AM and SF-AM-TETA, which was beneficial for the usage of CO₂ adsorption at high temperature.

Fig. 2 TGA curves of SF and its modified fibers.

The SEM images of SF and its modified fibers in Fig. 3 indicated that surface of sisal fiber was clean after alkali treatment, and pectin, ash content and other alkali-soluble impurities were removed. Compared with ASF, the surface of SF-AM became much rougher, and was covered a thin membrane. The surface of SF-AM-TETA was wrapped more tightly. The grooves were filled by the aminates.

Fig. 3 SEM morphology of SF and its modified fibers.

3.2 Adsorption behaviours of the solid amine adsorbent for CO₂

As shown in Table 2, the adsorption capacity of SF-AM-X (X = EDA, DETA, TETA or TEPA) in the absence of moisture was much lower than that in the moist condition. CO₂ was difficult to diffuse into the inner part of the fiber when it was not swollen, and could react only with amine groups on the surface of SF-AM-X. However, in the presence of moisture, H₂O played an important role in the reaction between CO₂ and amine groups of the adsorbent, and had a substantial promoting effect on CO₂ adsorption.^36–38 As shown in the follow eqn (5) to (8), in the presence of moisture. 1 mol RNH₂ or R₁R₂NH could adsorb 1 mol CO₂ to form bicarbonate (eqn (5) and (6)), while 1 mol RNH₂ or R₁R₂NH could only adsorb 0.5 mol CO₂ to form carbamate in the absence of H₂O (eqn (7) and (8)).

CO₂ + RNH₂ + H₂O ⇌ RNH₃⁺ + HCO₃⁻ (5)

CO₂ + R₁R₂NH + H₂O ⇌ R₁R₂NH₂⁺ + HCO₃⁻ (6)

CO₂ + 2RNH₂ ⇌ RNH₃⁺ + RNHCOO⁻ (7)

CO₂ + 2R₁R₂NH ⇌ R₁R₂NH₂⁺ + R₁R₂NCOO⁻ (8)

Table 2 Adsorption capacity and amine efficiency of SF-AM-X

Amination agent	EDA		DETA		TETA		TEPA
Alkyl amino content (mmol g^−1)	5.56		5.76		6.14		6.33
Adsorption conditions	Dry	Moist	Dry	Moist	Dry	Moist	Dry	Moist
Adsorption capacity (mmol g^−1)	0.16	1.59	0.20	2.91	0.75	4.20	0.85	4.13
Amine efficiency (%)	2.87	28.6	3.47	50.52	12.21	68.34	13.42	65.24

---

Comparison of adsorption capacities and amine efficiencies of SF-AM-X indicated that SF-AM-TETA and SF-AM-TEPA showed better adsorption performance than SF-AM-EDA and SF-AM-DETA, though they had close alkyl amino content. SF-AM-TETA showed the highest adsorption capacity (4.20 mmol g⁻¹) and amine efficiency (68.34%). The longer chain segment of TETA and TEPA would be benefit to the contacting with CO₂, thus, their modified fibers showed higher adsorption capacity.

The breakthrough curves of SF-AM-TETA under different temperature were illustrated in Fig. 4. In the column graph of Fig. 4, the trends indicated that adsorption amount of SF-AM-TETA for CO₂ decreased with the increase of adsorption temperature, the decrease became more obvious in the presence of moisture. The adsorption amount of SF-AM-TETA was 2.08 mmol g⁻¹ at 60 °C, only 46% of that at 10 °C. The reaction between alkyl amino groups and CO₂ molecules is an exothermic reaction, the rising temperature would be unfavourable to the CO₂ chemical adsorption process, it would also accelerate the decomposition of ammonium carbonate,²¹ resulting in a decrease of adsorption amount under high temperature. It was also been proved by the comparison of the breakthrough curves in Fig. 4.

Fig. 4 CO₂ adsorption breakthrough curves of SF-AM-TETA under different temperature (SF-AM-TETA: 1 g, T = 10, 25, 40, 60 °C, C (CO₂) = 10%).

The spent SF-AM-TETA was regenerated by heating it at 90 °C in pure nitrogen atmosphere. The regenerability of SF-AM-TETA was evaluated by its adsorption capacity of the regenerated fiber. The adsorption breakthrough curves and capacities of regenerated SF-AM-TETA were shown in Fig. 5. The adsorption capacity of fresh fiber was 4.20 mmol g⁻¹. After 1^st regeneration, the adsorption capacity of the regenerated adsorbent was 3.96 mmol g⁻¹, and the regeneration efficiency was 94%. After 7^th regeneration, the breakthrough curves is almost the same with the 1^st regenerated adsorbent, and the adsorption capacity still remained 3.92 mmol g⁻¹, the regeneration efficiency was 93%. All these results indicated good regenerability of SF-AM-TETA.

Fig. 5 Adsorption performance of regenerated SF-AM-TETA, (SF-AM-TETA: 1 g, T = 25 °C, C (CO₂) = 10%).

3.3 Adsorption kinetics of SF-AM-TETA

In order to better interpret the adsorption behaviour and adsorption kinetics of the solid amine fiber SF-AM-TETA, two such models, pseudo-first order and pseudo-second order^39–41 were applied to fit CO₂ adsorption experimental data at different temperatures, namely. The equations associated with the kinetic models explored in this work are given as follows.

The pseudo-first order kinetic equation:

q_t = q_e[1 − exp(−k_ft)] (9)

The pseudo-second order kinetic equation:

where t is the time elapsed from the beginning of the adsorption process, q_t is the amount adsorbed at a given point in time, and q_e represents the amount adsorbed at equilibrium. k_f (min⁻¹) and k_s (g mmol⁻¹ min⁻¹) are rate constants, respectively.

Fig. 6 showed the CO₂ adsorption capacity vs. time at 10 °C, 25 °C, 40 °C, 60 °C, and the corresponding profiles as predicted by the different kinetic models. The values of kinetic model parameters for CO₂ adsorption on SF-AM-TETA were presented in Table 3. Combine with the data of the pseudo-first order and the pseudo-second order kinetic models, it was found that the pseudo-second order presented some limitations with respect to the prediction of CO₂ adsorption on dry SF-AM-TETA, and deviated significantly from the experimental data. At the same time, the pseudo-first kinetic model was clearly consistent with the plots of experimental data. It indicated that solid amine fiber SF-AM-TETA adsorbed CO₂ mainly through physical adsorption process in the absence of moisture.

Fig. 6 Experimental CO₂ capacities on SF-AM-TETA under different temperature and corresponding fit to kinetic models.

Table 3 Kinetic model parameters for CO₂ adsorption on SF-AM-TETA under different temperature

Kinetic model	Parameter	Dry SF-AM-TETA (T/°C)				Moist SF-AM-TETA (T/°C)
		10	25	40	60	10	25	40	60
Experimental	qe(exp)	0.700	0.746	0.518	0.479	4.527	4.200	2.559	1.990
Pseudo-first order model	qe(fit)	0.678	0.748	0.510	0.446	4.654	4.366	2.659	1.839
	kf	0.221	0.197	0.479	0.479	0.019	0.024	0.033	0.051
	R^2	0.998	0.999	0.995	0.995	0.990	0.998	0.998	0.996
Pseudo-second order model	qe(fit)	0.774	0.869	0.547	0.479	6.267	5.755	3.470	2.277
	ks	0.386	0.288	1.579	1.805	0.003	0.004	0.009	0.023
	R^2	0.991	0.991	0.984	0.984	0.996	0.999	0.993	0.988

---

Different from the adsorption process of SF-AM-TETA in the absence of moisture, the pseudo-second order kinetic model was consistent with the measured value of moist SF-AM-TETA very well (as shown in Fig. 6(b)). In the presence of moisture, H₂O is participated with the chemical adsorption process (the chemical equations are presented in eqn (5) and (6)), which has deeply promoted CO₂ chemical adsorption process. Thus, chemical adsorption process became the predominant adsorption process in the presence of moisture, which is agreement with the conclusion of the analysis of kinetic models. Increasing with the temperature, fast movement of the molecular CO₂ tends to desorb from the adsorbents, and the adsorption capacity decreased.

3.4 Adsorption activation energy of SF-AM-TETA

From the previously shown Table 3, it can be inferred that an increase in adsorption temperature results in an increase in k with fitting-well models. This temperature effect also can be demonstrated through the Arrhenius equation, which is the common method to express the adsorption activation energy (E_a):

K = Ae^−E_a/RT (11)

where K is the overall mass transfer coefficient, which is k_f, or k_s displayed in Table 3. A is the frequency factor or collision factor, E_a is activation energy. R is the molar gas constant, and T is the temperature. Eqn (11) can also be rearranged as:

The activation energy (E_a) can be analyzed directly from the slope of a plot between ln K and reciprocal of temperature (1/T).^39,42

As the results of Fig. 6 and Table 3, the pseudo-first order model was considered adequate in the absence of moisture, we used the corresponding kinetic coefficients at various temperatures (10, 25, 40, 60 °C) to calculate the parameters of an Arrhenius linearized form (eqn (12)), and displayed the results in Fig. 7. Similarly, some results in the presence of moisture also displayed which were calculated based on the pseudo-second order model.

Fig. 7 Arrhenius plots for the kinetic constants obtained by the different kinetic models.

Fig. 7 showed the plots of the linearized Arrhenius equation, along with the corresponding values of A and E_a. As seen, the calculated E_a for pseudo-second order models in the presence of moisture was larger than that for pseudo-first order models in the absence of moisture. Obviously, the chemical adsorption needed more activation energy than the physical adsorption. In fact, the adsorption is a complex process, the calculated parameters of these kinetic models was not the actual value in a strict sense.^39,43 It must be treated carefully the relationship between the calculated value and the actual activation energy. Even so, the Arrhenius equation was useful to predict K at different temperatures, while the adequacy of these semi-empirical models to determine thermodynamic parameters is still debatable.

In the presence of moisture, the solid amine fiber required more activation energy to overcome the potential energy barrier to cause adsorption activity on the solid surface than that in the absence of moisture. There were two purposes, one was for the adsorption, and the other was for the reaction. At the same time, solid amine fiber in the absence of moisture offered the energy for the adsorption. On the whole, the interaction between the amine of SF-AM-TETA and CO₂ was relatively weak, which would be advantageous to desorption and regeneration of the sorbents.

3.5 The comparison of adsorption capacities with other sorbents

Hydroxyl groups on amine adsorbents, though not absorb CO₂ themselves, have been proved to promote CO₂–amine reaction by stabilizing carbamate formation through hydrogen bonding.^23,25 The well hydrophilicity of VF substrate also had a positive effect on the CO₂ adsorption.²⁴ In this study, adsorption capacities of amine immobilized on various solid supports like viscose fiber, PP fiber, which have different hydrophilicity, were compared in Table 4. In the table, adsorption data of amines supported on silica particles or polymeric particles are also listed. The results in Table 4 show that amine immobilized on porous supports usually has a remarkable adsorption amount in dry condition due to the physical adsorption by the porous structure. However, amine adsorbents of fibrous substrates showed excellent adsorption performance in moist condition. Compared with other solid amine adsorbents using viscose, PP and glass fibers as substrates, SF-AM-TETA showed a relatively high adsorption capacity whether in the absence or presence of moisture. The unique advantages of plentiful pores and large external surface of the sisal fiber-based adsorbent are beneficial to the physical adsorption in the absence of moisture. While the synergistic effects between these hydroxyl groups and amine groups on the solid amine fiber also greatly promote the chemical adsorption in the presence of moisture.

Table 4 Comparison of CO₂ adsorption capacity of solid amine sorbents

Substrate	Amine	Temperature (°C)	pCO2 (atm)	Adsorption capacity (mmol g^−1)		Ref.
				Dry	Moist
SF	TETA	25	0.10	0.75	4.20	This work
VF	TETA	30	0.10		2.36	23
PP	TETA	30	0.10		2.03	23
Glass fiber	PEI	30	0.24		4.12	16
PMMA	PEI	45	0.10		3.53	44
Y-Type zeolite	TEPA	60	0.15	2.59	4.31	45
SBA-15	APTES	60	0.15	0.66		46
Fumed silica	PEG + PEI	50	1	3.41		47

---

4. Conclusions

Amine modified sisal fiber was prepared by grafting acrylamide and then aminating. The prepared solid amine adsorbent SF-AM-TETA showed good adsorption performance for CO₂. Due to pores and large external surface inherited from the organic texture of vegetable, this SF-based solid amine fiber could adsorb 0.75 mmol CO₂ per g in the absence of moisture. The adsorption capacity of SF-AM-TETA could reach 4.20 mmol g⁻¹ in the presence of moisture due to the synergistic effects between hydroxyl groups and amine groups on amine sisal fiber. The solid amine fiber showed well regenerability, the regeneration efficiency could keep at 93% adsorption capacity of the fresh adsorbent after numbers of regeneration cycles. The pseudo-second order kinetic model was consistent with the measured value of moist SF-AM-TETA very well, indicating that the predominant adsorption process of SF-AM-TETA in the presence of moisture was mainly chemical adsorption. Further calculation of the activation energy by Arrhenius equation indicated that the synthesized solid amine fiber required less activation energy to adsorb CO₂, which was beneficial to desorption and regeneration of the sorbents. The low-cost and renewable vegetable fibers will be a sustainable resource for the preparation of solid amine adsorbents for CO₂ capture.

Acknowledgements

The authors gratefully acknowledge the financial support provided by the National Natural Science Foundation of China (Grant No. 51473187), National Natural Science Foundation of China (Grant No. 51303030), and Natural Science Foundation of Guangdong Province (2014A030313192).

References

1. D. M. D'Alessandro, B. Smit and J. R. Long, Angew. Chem., Int. Ed., 2010, 49, 6058–6082.
2. C. W. Jones, Annu. Rev. Chem. Biomol. Eng., 2011, 2, 31–52.
3. D. Y. C. Leung, G. Caramanna and M. M. Maroto-Valer, Renewable Sustainable Energy Rev., 2014, 39, 426–443.
4. A. S. Bhown and B. C. Freeman, Environ. Sci. Technol., 2011, 45, 8624–8632.
5. E. Blomen, C. Hendriks and F. Neele, Energy Procedia, 2009, 1, 1505–1512.
6. J. Gibbins and H. Chalmers, Energy Policy, 2008, 36, 4317–4322.
7. C. Lin, H. Y. Zhang, X. Y. Lin and Y. F. Feng, Adv. Mater. Res., 2013, 634–638, 746–750.
8. M. Keramati and A. A. Ghoreyshi, Phys. E, 2014, 57, 161–168.
9. F. Su, C. Lu, S.-C. Kuo and W. Zeng, Energy Fuels, 2010, 24, 1441–1448.
10. D. P. Bezerra, R. S. Oliveira, R. S. Vieira, C. L. Cavalcante and D. C. S. Azevedo, Adsorption, 2011, 17, 235–246.
11. J. Tanthana and S. S. Chuang, ChemSusChem, 2010, 3, 957–964.
12. W. Klinthong, K. J. Chao and C. S. Tan, Ind. Eng. Chem. Res., 2013, 52, 9834–9842.
13. Y. Le, D. Guo, B. Cheng and J. Yu, J. Colloid Interface Sci., 2013, 408, 173–180.
14. X. Y. Zhang, H. Y. Qin, X. X. Zheng and W. Wu, Mater. Res. Bull., 2013, 48, 3981–3986.
15. Y. Du, Z. Du, W. Zou, H. Li, J. Mi and C. Zhang, J. Colloid Interface Sci., 2013, 409, 123–128.
16. P. Li, B. Ge, S. Zhang, S. Chen, Q. Zhang and Y. Zhao, Langmuir, 2008, 24, 6567–6574.
17. P. Y. Li, S. J. Zhang, S. X. Chen, Q. K. Zhang, J. J. Pan and B. Q. Ge, J. Appl. Polym. Sci., 2008, 108, 3851–3858.
18. Y. Yang, H. Li, S. Chen, Y. Zhao and Q. Li, Langmuir, 2010, 26, 13897–13902.
19. Q. Wu, S. Chen and H. Liu, RSC Adv., 2014, 4, 27176–27183.
20. T. Xu, Q. H. Wu, S. X. Chen and M. W. Deng, RSC Adv., 2015, 5, 32902–32908.
21. L. Z. Zhuang, S. X. Chen, R. J. Lin and X. Z. Xu, J. Mater. Res., 2013, 28, 2881–2889.
22. T. Xu, Q. Wu, S. Chen and M. Deng, RSC Adv., 2015, 5, 32902–32908.
23. R. Lin, L. Zhuang, X. Xu and S. Chen, J. Colloid Interface Sci., 2013, 407, 425–431.
24. Q. Wu, S. Chen, S. Luo and T. Xu, J. Appl. Polym. Sci., 2015, 133, 42840–42846.
25. X. Xu, C. Song, B. G. Miller and A. W. Scaroni, Ind. Eng. Chem. Res., 2005, 44, 8113–8119.
26. T. T. N. Bachelor and P. Toochinda, Environ. Technol., 2012, 33, 2645–2651.
27. J. S. Cheng, J. Du and W. J. Zhu, Carbohydr. Polym., 2012, 88, 61–67.
28. A. A. Kadam, H. S. Lade, S. M. Patil and S. P. Govindwar, Bioresour. Technol., 2013, 132, 276–284.
29. M. A. Adebayo, L. D. Prola, E. C. Lima, M. J. Puchana-Rosero, R. Cataluna, C. Saucier, C. S. Umpierres, J. C. Vaghetti, L. G. da Silva and R. Ruggiero, J. Hazard. Mater., 2014, 268, 43–50.
30. O. Karnitz, L. V. A. Gurgel and L. F. Gil, Carbohydr. Polym., 2010, 79, 184–191.
31. P. L. Homagai, K. N. Ghimire and K. Inoue, Bioresour. Technol., 2010, 101, 2067–2069.
32. C. Liu, H. H. Ngo, W. Guo and K. L. Tung, Bioresour. Technol., 2012, 119, 349–354.
33. Z. Y. Zhang, D. W. Rackemann, W. O. S. Doherty and I. M. O'Hara, Biotechnol. Biofuels, 2013, 6, 986–991.
34. X. Liu, R. Lin, S. Chen, N. Ma and Y. Huang, J. Appl. Polym. Sci., 2012, 126, E345–E353.
35. S. Kalia and S. Vashistha, J. Polym. Environ., 2012, 20, 142–151.
36. A. Sayari and Y. Belmabkhout, J. Am. Chem. Soc., 2010, 132, 6312–6314.
37. A. Sayari, Y. Belmabkhout and E. Da'na, Langmuir, 2012, 28, 4241–4247.
38. F. Su, C. Lu and H. S. Chen, Langmuir, 2011, 27, 8090–8098.
39. R. Serna-Guerrero and A. Sayari, Chem. Eng. J., 2010, 161, 182–190.
40. X. Wang, L. L. Chen and Q. J. Guo, Chem. Eng. J., 2015, 260, 573–581.
41. X. Wang, Q. J. Guo and T. T. Kong, Chem. Eng. J., 2015, 273, 472–480.
42. Q. Liu, J. J. Shi, S. D. Zheng, M. N. Tao, Y. He and Y. Shi, Ind. Eng. Chem. Res., 2014, 53, 11677–11683.
43. S. Azizian, J. Colloid Interface Sci., 2004, 276, 47–52.
44. M. L. Gray, J. S. Hoffman, D. C. Hreha, D. J. Fauth, S. W. Hedges, K. J. Champagne and H. W. Pennline, Energy Fuels, 2009, 23, 4840–4844.
45. F. Su, C. Lu, S.-C. Kuo and W. Zeng, Energy Fuels, 2010, 24, 1441–1448.
46. N. Hiyoshi, K. Yogo and T. Yashima, Microporous Mesoporous Mater., 2005, 84, 357–365.
47. S. Meth, A. Goeppert, G. K. S. Prakash and G. A. Olah, Energy Fuels, 2012, 26, 3082–3090.

---