Oct 24, 2016 - ABSTRACT: A novel biopolymer-based aerogel was developed by ... aerogel with the dye resulted in time-dependent dye adsorption, with ...
Article pubs.acs.org/Langmuir
Reusable Green Aerogels from Cross-Linked Hairy Nanocrystalline Cellulose and Modified Chitosan for Dye Removal Han Yang, Amir Sheikhi, and Theo G. M. van de Ven* Department of Chemistry, Centre for Self-Assembled Chemical Structures, Pulp and Paper Research Centre, McGill University, Montreal H3A 2A7, QC, Canada S Supporting Information *
ABSTRACT: A novel biopolymer-based aerogel was developed by freeze-drying a hydrogel, synthesized from cross-linking bifunctional hairy nanocrystalline cellulose and carboxymethylated chitosan through a Schiff base reaction. The nanocelluloses, bearing aldehyde and carboxylic acid groups, facilitated the cross-linking with chitosan through imine bond formation while providing negatively charged functional groups, and chitosan was modified to accommodate carboxylic acid. The potential of this bioaerogel in environmental remediation was examined in a model system comprising methylene blue, a cationic dye. Electrostatic complexation between acidic groups on the anionic aerogel with the dye resulted in time-dependent dye adsorption, with long-time equilibrium dye concentration fitting well to the Langmuir isotherm, yielding a maximum adsorption capacity of ∼785 mg g−1 and equilibrium constant K ∼ 0.0089 at room temperature. Dynamics of adsorption was modeled by numerically solving the unsteady-state diffusion−adsorption mass balance in a 1D spherical coordinate, which attested to a diffusion-controlled process with a Langmuir adsorption time constant τads ∼ 7.6 s. To the best of our knowledge, this bioaerogel exhibits the highest removal capacity as yet for any reusable adsorbents prepared from biopolymers. Successful adsorption−regeneration cycles proved an excellent reusability, and the adsorption capacity remained constant over a wide pH range (e.g., pH > 7). This work may pave the way toward ultralight green functional materials.
■
INTRODUCTION Synthetic dyes have complex structures, are difficult to eliminate from contaminated water, and have harmful impacts, including teratogenetic, carcinogenic, and mutagenic effects on human health.1 Dyes are widely used in many industries, such as textile, leather, paper, plastics, printing, and cosmetics.2 The rapid development of global industrialization has further increased the dye pollution problems in water, which is the most important and necessary natural resource for human beings and other living creatures. Thus, it is an important and challenging task to eliminate dyes in industrial effluents before they are discharged into the environment.3 Various methods have been used for removing dyes from industrial wastewater, including photocatalytic oxidation, electrochemical destruction, membrane filtration, and adsorption, among which adsorption has been recognized as an economic treatment method as a result of its easy operation and relatively low cost.4 Adsorption is a physicochemical process in which molecules are attached to the surface of an adsorbent by physical forces (e.g., van der Waals forces and/or an electrostatic attraction between oppositely charged adsorbate molecules and an adsorbent surface) or chemical reactions (e.g., covalent bonding). To achieve an efficient adsorption, it is important to select a suitable adsorbent according to the charge or functional groups carried by target dyes. Besides efficient adsorption, reducing the toxicity of the adsorbent is also important to avoid secondary © 2016 American Chemical Society
pollution. Thus, an environmentally friendly and efficient adsorbent is always highly desired. Accordingly, many biopolymer-based adsorbents have been developed from agricultural waste and plants, such as rice husk,5 jute fiber carbon,6 wheat bran,7 sunflower seed shells,8 sugar cane bagasse,9 chitin,10 chitosan,11 and modified cellulose.12 Among naturally available biopolymers, cellulose is the most abundant, environmentally friendly, renewable, and biodegradable material on earth, which suggest cellulose as one of the potential resources for fabricating green adsorbents. Recently, nanomaterials, benefiting from high specific surface area and active site density for interaction with target molecules, have secured an important place among adsorbents, owing to their high adsorption capacity. Nanocrystalline cellulose (NCC), also know as cellulose nanocrystals (CNC), is a major type of bionanomaterial prepared from cellulose. NCC can be produced from wood pulp by sulfuric acid hydrolysis (introducing sulfate half-ester groups),13 ammonium persulfate treatment (introducing carboxyl groups),14 or periodate−chlorite oxidation (yielding electrosterically stabilized15 NCC (ENCC) with a high content of carboxyl groups).16 NCC has negatively charged functional groups, which can interact with cationic dyes by electrostatic Received: August 18, 2016 Revised: October 23, 2016 Published: October 24, 2016 11771
DOI: 10.1021/acs.langmuir.6b03084 Langmuir 2016, 32, 11771−11779
Article
Langmuir attraction.17 The content of carboxyl groups on NCC can be increased through 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) radical mediated oxidation to improve the removal capacity of NCC.18 NCC can also be modified with primary amines for adsorption of anionic dyes.19 ENCC is a recent novel type of NCC,20 which has not yet been used in the study of dye removal; however, it has a significant capacity in sequestering of heavy metal ions, such as copper ion removal from aqueous media.21 However, at low ion concentration, NCC or ENCC adsorbents maintain colloidal stability and often suffer from separation and regeneration difficulties. Chitosan is also a biodegradable and renewable biopolymer, derived from deacetylation of the biopolymer chitin, which is the third most abundant polysaccharide in the world (after cellulose and hemicellulose).22 Traditionally, when chitosan is directly used in water purification, it is mostly effective in adsorbing negatively charged dyes23 and heavy anions24 in acidic conditions as a result of protonated amine groups. In this work, we use chitosan as a green cross-linker by taking advantage of its amine groups as functional moieties participating in the formation of covalent bonds. Aerogels are highly porous solids made from wet gels in which the liquid phase in the gels has been replaced by a gas (usually air).25 Aerogels exhibit many unique properties, such as low density, high porosity, large surface area, ultralow thermal conductivity, ultralow refractive index, and ultralow dielectric constant.26 The most investigated aerogels are traditionally prepared from silica (such as silica dioxide) and several kinds of non-silica inorganic oxides (e.g., titanium, tin, and aluminum). Recently, aerogels prepared from natural polymers (e.g., starch, chitosan, and cellulose) have been proposed due to their renewability, biodegradability, and biocompatibility.27 Cellulose derivatives, especially nanocellulose, including nanofibrillar cellulose (NFC)28−30 and nanocrystalline cellulose (NCC),31,32 show particular promise in the preparation of flexible and environmentally friendly aerogels. In this work, we aim at preparing nanocrystalline cellulose with bifunctional moieties (carboxyl and aldehyde groups) by sequential periodate and partial chlorite oxidations followed by a hot water treatment. We refer to these NCCs as bifunctional hairy nanocrystalline cellulose or simply bifunctional NCC (BNCC). Chitosan (CT) was modified in advance to carboxymethylated CT (CMCT), increasing the carboxyl group content. The amine groups on chitosan and aldehyde groups on NCC are able to form covalent imine bonds through a Schiff base reaction, cross-linking the BNCC particles and yielding a hydrogel network. The adsorbent was readily obtained by freeze-drying the hydrogel to yield an all-natural aerogel (BNCC−CMCT). Characterization of BNCC−CMCT was done by solid carbon-13 NMR, atomic force microscopy (AFM), and scanning electron microscopy (SEM). Methylene blue (MB), a cationic dye (structure shown in Figure 1, molar mass 319.5 g mol−1), is used as a model target molecule in this
dye adsorption study. The adsorption isotherm and dynamics of MB removal by BNCC−CMCT, the pH effect, and aerogel reusability were investigated in this work.
■
EXPERIMENTAL SECTION
Materials. Softwood kraft pulp sheets (provided by FPinnovations) were used as the starting cellulose material. Chemicals for reactions, including sodium (meta) periodate, ethylene glycol, hydroxylamine hydrochloride, sodium chlorite, chloroacetic acid, chitosan (from crab shell, brookfield viscosity >200), and methylene blue hydrate, were purchased form Sigma-Aldrich. Propanol was provided by Fisher, hydrogen chloride (0.1 and 1 M) and sodium hydroxide (10 mM, 0.1 and 1 M) were from Fluka, anhydrous ethyl alcohol was from commercial alcohols, and sodium chloride was from ACP Chemicals Inc. All chemicals were used as received. Milli-Q water (18.2 MΩ cm, Millipore Milli-Q purification system) was used in all experiments. Preparation of Bifunctional NCC (BNCC). One gram of softwood kraft pulp was soaked in water, well dispersed by a disintegrator (Noram Quality Control and Research Equipment Limited), and then filtered to remove extra water from the pulp. Next, 0.98 g of NaIO4 and the wet pulp were added to 67 mL of water, including the moisture from the wet pulp. The reaction beaker was wrapped with aluminum foil to block light. The pulp was stirred at room temperature for 96 h, and then 1 mL of ethylene glycol was added to this mixture to stop the reaction by quenching the residual periodate. The dialdehyde modified cellulose (DAMC) was washed thoroughly with water by filtration. The aldehyde content of DAMC was determined by the hydroxylamine hydrochloride method previously reported.33 A visual demonstration of this procedure may be found elsewhere.34 BNCC was prepared by converting part of aldehyde groups on DAMC to carboxyl groups by chlorite oxidation, followed by a hot water treatment.35,36 Briefly, 1 g of never-dried DAMC and 0.54 g of NaClO2, 2.9 g of NaCl, and 0.54 g of H2O2 were dispersed in 50 mL of 0.5 M acetic buffer solution (pH = 5), and the slurry was stirred for 24 h. Subsequently, the oxidized pulp was rinsed with water, followed by four times washing with 70% ethanol, and then dried in an oven at 50 °C. The bifunctionally modified cellulose fiber (0.2 g) and 20 mL of water were added to a 50 mL flask, and the suspension was stirred at 80 °C in an oil bath for 1 h and subsequently centrifuged at 8000 rpm for 10 min (Aventi J-E centrifuge from Beckman Coulter) to remove the unfibrillated fibers (a negligible amount). The supernatant (containing BNCC) was precipitated by adding propanol (the weight of propanol is 1.5 times of supernatant), collected, and stored at 4 °C for further usage. Preparation of Carboxymethylated Chitosan. For the carboxymethylation of chitosan, we follow a previous method, with some modifications.37 Sodium hydroxide (1.35 g) was first dissolved in a propanol/water mixture with a volume ratio of 8:2. One gram of chitosan was added to this alkaline solution, which was stirred and allowed to swell at room temperature for 1 h. Subsequently, 1.5 g of chloroacetic acid was dissolved in 2 mL of propanol and added to the chitosan slurry in five equal portions in a period of 30 min. The mixture continued to react for another 4 h at room temperature. The reaction was stopped by adding 50 mL of 70% ethanol and filtered through a nylon cloth. The white solid was washed with 80% ethanol for four times and then with anhydrous ethanol. Finally, the powder was dried in an oven at 50 °C to obtain carboxymethylated chitosan (CMCT). Preparation of Aerogel (BNCC−CMCT). A BNCC suspension (1 wt %) and a CMCT solution (1 wt %) were heated by stirring in an oil bath at 60 °C for 1 h. Then, these two sets of samples were mixed by a homogenizer (Polytron PT 2500E) at 8000 rpm for 1 min to form a hydrogel. The hydrogel was frozen at −80 °C for 12 h and then freezedried by a freeze-dryer (Thermo ModulyoD). Dye Adsorption. The MB adsorption process was conducted in batch experiments. MB solutions with desired concentrations were prepared by successive dilution of a stock MB solution with water. To obtain the calibration curve of MB, the absorbance of MB solutions
Figure 1. Molecular structure of methylene blue (MB). 11772
DOI: 10.1021/acs.langmuir.6b03084 Langmuir 2016, 32, 11771−11779
Article
Langmuir with predetermined concentrations at λmax = 664 nm was detected by a UV−vis spectrophotometer (Cary 5000 UV−vis−NIR spectrophotometer). This calibration curve was used to determine the concentration of MB solution after absorption in the ongoing adsorption experiments. Equilibrium Experiments. Batch adsorption measurements were performed to obtain the maximum adsorption of MB by the BNCC− CMCT aerogel. Five milliliters of MB solutions with known initial concentrations and 1 mg of the adsorbent were placed in a 10 mL vial and agitated using a magnetic stirrer at 120 rpm for 24 h to ensure the adsorption process has reached equilibrium. The equilibrium concentration (Ce) of MB was then determined using a UV−vis spectrometer at 664 nm. Equation 1 was used to calculate the adsorbed amount of dye per gram of adsorbents (mg g−1) at equilibrium. Γe =
C0 − Ce V mads
sidebands were suppressed by the TOSS sequence.39 Typically, 8000 scans were acquired. Atomic Force Microscopy (AFM) Imaging. The morphology of BNCC particles was obtained by AFM (Nanoscope IIIa MultiMode with Extender (Veeco Metrology Group, Santa Barbara, CA)). A drop of BNCC suspension was placed on a freshly cleaved mica surface for 10 min, followed by rinsing off the excess liquid. The experiments were conducted in tapping mode using silicon cantilevers (ACTA model, AppNano) with a nominal spring constant ∼37 N m−1, nominal resonant frequency ∼300 kHz, and nominal tip radius ∼6 nm. Nanoscope Analysis 1.4 was used to process the AFM images. Scanning Electron Microscopy (SEM) Imaging. The aerogel sample was mounted on a specimen pin by a double-sided carbon tape and coated with a 3 nm thick layer of Pt by a high-vacuum coater (Leica EM ACE600). The aerogel microstructure was observed by SEM imaging (FEI Inspect F-50 FE-SEM). The images were taken at an accelerating voltage of 10 kV. Porosity Measurement. The porosity of BNCC−CMCT aerogel was determined by the ethanol displacement method.40 Ethanol is able to penetrate into the pores easily and is not expected to change the geometrical volume of the aerogel. A piece of BNCC−CMCT (w1) was immersed in 40 mL of anhydrous ethanol (ρ = 0.789 g mL−1) and then placed in a desiccator under a reduced pressure for 8 min to remove air bubbles inside the aerogel. The aerogel was taken out, and the ethanol on its surface was gently removed by a piece of filter paper. The aerogel was reweighed (w2) immediately. The porosity ε was calculated by the equation
(1)
where Γe (mg g−1) is the equilibrium amount of dye adsorbed by 1 g of adsorbent, C0 (mg L−1) and Ce (mg L−1) are the initial and equilibrium concentrations of MB, V (L) is the volume of MB solution, and mads (g) is the adsorbent weight. The volume of aerogel expands about 10 times. Kinetic Experiments. For kinetic adsorption studies, 1 mg of BNCC−CMCT adsorbent was mixed with 5 mL of MB solution (240 mg L−1) and stirred at 120 rpm. The experiments were repeated for various desired times. The bulk MB concentration (C) at various times was measured by a UV−vis spectrophotometer. The adsorbed dye amounts were calculated from eq 2.
Γ=
C0 − C V mads
ε=
(w2 − w1)/ρ Vads
(3)
where Vads is volume of the aerogel calculated from its geometrical dimensions. An average value was taken from three replicates.
■
(2)
RESULTS AND DISCUSSION Characterizations of BNCC and Aerogel. The morphology of BNCC by AFM is shown in Figure 2. BNCC is a rod-
where Γ (mg g−1) is the amount of dye adsorbed for 1 g of adsorbent at time t and C0 (mg L−1) and C (mg L−1) are the initial concentration and concentration of MB at time t. Effect of pH on Adsorption. Various MB solutions with initial pH ∼ 2−12 were investigated to determine the influence of pH on the efficiency of MB removal by BNCC−CMCT. The pH of a MB solution was adjusted by adding 1 mol L−1 HCl or 1 mol L−1 NaOH solutions. The initial MB concentration is 100 mg L−1, and the experiments were performed for 4 h with stirring at 120 rpm; then the concentration of MB was measured with a UV−vis spectrophotometer. The removed amount of MB at each pH value was calculated using eq 1. Reusability of Adsorbent. The reusability of adsorbent was also investigated. One milligram of BNCC−CMCT was placed in 5 mL of MB (50 mg L−1) solution and stirred at 120 rpm for 1 h; then the concentration of MB was measured by UV−vis. The dye on BNCC− CMCT was desorbed by soaking the aerogel in 10 mL of 0.1 M HCl and agitating at 80 rpm with a shaker for 10 min, followed by washing with 0.1 M NaOH, rinsing with water, and finally rinsing with ethanol and drying under air. The adsorption−desorption procedure was repeated six times. The MB removal at each step was calculated using eq 1. Characterizations. Conductometric Titration. The carboxyl group contents of BNCC and CMCT were determined with a Metrohm 836 Titrando instrument according to a previously reported method.38 A certain amount of sample (with a solid content of ∼20 mg) and 2 mL of NaCl solution (20 mmol L−1) were added to 140 mL of Milli-Q water, and 0.1 M HCl was added to adjust the pH to ∼3. Then, the suspension was titrated by a 10 mM NaOH solution at a rate of 0.1 mL min−1 until pH ∼ 11. The part of the titration curve which represents a weak acid provides the carboxyl content. Solid-State Carbon-13 NMR Measurements. Solid-state carbon-13 NMR spectra were acquired on a Varian VNMRS400 NMR spectrometer operating at 100.5 MHz. Cross-polarization spectra were obtained with a contact time of 1.5 ms and a recycle delay of 2 s. The sample was spun in a 7.5 mm rotor at 5000 Hz, and spinning
Figure 2. AFM image of BNCC nanoparticles.
like nanoparticle, obtained from a hot water treatment of periodate and partially chlorite oxidized cellulose fibers. The major advantage of this process is that it does not require a strong acid hydrolysis and/or extensive mechanical treatment (which usually involves a high energy consumption or specifically designed instruments), and no additional postpurification is required after the formation of the nanocellulose particles. Furthermore, in this way, one is able to easily obtain nanocellulose particles with desired charge content by adjusting the chlorite oxidation level on aldehyde groups prior to a hot water treatment. As shown in Figure 2, BNCC from this work 11773
DOI: 10.1021/acs.langmuir.6b03084 Langmuir 2016, 32, 11771−11779
Article
Langmuir
terial; the flexible CMCT polymer chains connect BNCC nanorods, forming a porous network. Furthermore, BNCC and CMCT not only function as a hydrogel construction material but also provide negatively charged carboxyl groups, which can increase the available sites for binding cationic dye molecules by electrostatic attraction, improving the capability of dye adsorption. (The characterization of chemical groups on BNCC and CMCT by solid-state C-13 NMR is shown in the Supporting Information.) The hydrogel was freeze-dried to form an aerogel, which was able to easily stand on the tips of the fine awns of a green foxtail without bending them (Figure 3b). This biopolymer-based aerogel is highly porous and lightweight; the porosity of the aerogel determined by solvent exchange was 98.8 ± 0.3% (since ethanol may not completely fill the pores, the porosity is likely underestimated). The microstructure of dry aerogel was investigated by SEM. The image shows that the aerogel has an open porous geometry with pore sizes in the range of 35−70 μm (Figure 3c), which are separated by “walls”, sheet-like and ultrathin structures revealed by the enlargement shown in the inset (Figure 3c). These large pores are beneficial for easy and quick mass penetration during adsorption. Adsorption Isotherm. An adsorption isotherm is an important tool for the description of how adsorbate molecules interact with an adsorbent surface. To investigate the relationship between the aerogel and MB molecules at equilibrium, and to obtain the maximum adsorption capacity of the aerogel, Langmuir and Freundlich isotherms were fitted to the equilibrium adsorption data. For Langmuir isotherms, it is assumed that each adsorbate molecule adsorbing onto the adsorbent surface has the same adsorption activation energy; thus, the adsorption process results in a monolayer coverage over a homogeneous adsorbent surface (provided there is sufficient adsorbate). Also, no adsorbate migrates after adsorption, and the adsorption is reversible. The Langmuir isotherm can be expressed by the following equation:41,42
has a length of about 110−150 nm and width of about 8 nm. The yield of BNCC in this study is about 40%. BNCC particles have two types of functional groups, namely aldehyde and carboxyl groups, besides the original hydroxyl groups. These two functional groups on one particle make BNCC potentially a more versatile nanocellulose than traditional NCC. The charge contents of BNCC and CMCT are ∼3 and ∼3.4 mmol g−1, respectively, measured by conductometric titration. Mixing a BNCC suspension with a CMCT solution, a transparent hydrogel is formed instantly, which remains stable even when the vial is upside down (Figure 3a). The content of
Figure 3. (a) Photograph of a transparent BNCC−CMCT hydrogel in an upside down vial, (b) photograph of a piece of aerogel standing on the tips of the fine awns of a green foxtail, and (c) SEM image of the aerogel, with the inset showing an enlargement of the “walls” formed in the aerogel.
carboxyl groups in the hydrogel is about 3.2 mmol g−1. Crosslinking has occurred between the amine groups on CMCT and the aldehyde groups on BNCC, by imine bond formation, as shown in Figure 4a. This cross-linking reaction occurs without adding any other chemicals, no hazardous byproducts are formed in this reaction, and there is no requirement for any postpurification treatment. All these advantages make this cross-linking process an environmentally friendly process. The cartoon in Figure 4b shows the formation of the hydrogel network. BNCC works as the supporting nanoma-
1 K 1 = + Γe Ce Γm
(4)
with the equilibrium constant K given by
Figure 4. (a) Schematic of cross-linking reaction between BNCC and CMCT and (b) cartoon for the hydrogel formation. 11774
DOI: 10.1021/acs.langmuir.6b03084 Langmuir 2016, 32, 11771−11779
Article
Langmuir K=
τads τdes
(5)
where Γe is the adsorption capacity at equilibrium, Ce the equilibrium concentration of MB in solution and Γm the maximum adsorption capacity. τads and τdes represent the characteristic times of adsorption and desorption, which equal the reciprocals of their corresponding rate constants. Langmuir plots for adsorption of MB in zero salt and 0.1 M NaCl are shown in Figure 5. Γm and K were calculated from the intercept
Figure 6. Experimental data and the Freundlich isotherm fit for MB adsorption by the BNCC−CMCT aerogel in solutions with zero salt or 0.1 M NaCl at pH ∼ 7.5 and T ∼ 22 °C.
Table 1. Isotherm Parameters for MB Adsorption in Solutions with 0 and 0.1 M NaCl at 22 °C Langmuir model
Figure 5. Langmuir isotherm plots for MB adsorption by the BNCC− CMCT aerogel in solutions with zero salt or 0.1 M NaCl (pH ∼ 7.5 and T ∼ 22 °C).
Freundlich model
NaCl (M)
Γm (mg g )
K
R
KF
n
R2
0 0.1
784.8 272.2
0.0089 0.046
0.986 0.990
493.7 130.9
9.1 5.2
0.796 0.917
2
Table 2. Comparison of the Maximum MB Adsorption by Various Adsorbents
and slope of the linear fitting, respectively. Increasing the ionic strength by NaCl addition may result in dye aggregation and a consequent diffusion coefficient reduction,43 increasing τads and thus increasing K. Moreover, the salt-mediated aggregation of MB may result in colloid formation and deviation from MB ionic behavior, decreasing the maximum adsorption capacity. Also, elevated ionic strengths decrease the swelling ratio of the anionic aerogel, affecting the gel pore size. A Freundlich isotherm, which can describe heterogeneous adsorption systems, and which is not restricted to the formation of monolayer coverage, can be represented as44 Γe = KFCe1/ n
−1
adsorbent
pH
Γm (mg g−1)
rice husk sugar cane bagasse9 NCC17 NCC modified by TEMPO reaction18 commercial activated carbon45 cellulose nanofibrils46 cellulose nanofibrils aerogel47 banana pith carbon48 chitosan/bentonite composite49 BNCC−CMCT aerogel (this work)
7 7 7.5 6.5 7.4 9
312.0 99.6 101.2 769.0 980.3 122.2 3.70 233.4 142.9 785
5
4 5.1 7.5
CMCT is also comparable to commercial activated carbon (980.3 mg g−1),45 although its capacity is about 20% lower, but activated carbon is very costly to prepare and regenerate. The maximum adsorption (in the absence of salt) is about 86% of the amount (909 mg g−1) calculated from charge stoichiometry (the content of carboxyl groups is 3.2 mmol g−1 and the molar mass of a MB+ ion is 284 g mol−1). The reason that Γm is less than the stoichiometric value can be attributed to the solution pH (7.5), slightly lower than the second pKa of dicarboxylic acid groups (∼8.0, obtained from a pH titration). Almost half of the aerogel negative charge originates form the C6-conjugated carboxylic acid on modified chitosan, which is all deprotonated at pH ∼ 7.5. At this pH, BNCC provides two adjacent carboxylic acid groups on C2 and C3, one of which being almost fully deprotonated (pKa ∼ 4.6) while for the other one (pKa ∼ 8.0) 32% carboxylic acid groups are deprotonated. In total, BNCC−CMCT provide ∼50% (CMCT) + 25% (fully dissociated carboxyl groups on BNCC) + 8% = 83% of maximum possible COO−, a value close to the
(6)
where Γe is the adsorption capacity at equilibrium, Ce is the equilibrium concentration of MB in solution, and KF and n are constants. The Freundlich adsorption isotherm fit to the equilibrium data in the absence or presence of NaCl is shown in Figure 6. Compared with the Freundlich adsorption isotherm, the Langmuir adsorption isotherm can describe this adsorption process more accurately, since the Langmuir isotherm fit has a much higher correlation coefficient (Table 1). The negatively charged carboxyl groups on the aerogel are mainly responsible for binding MB to the aerogel through electrostatic attraction. The maximum adsorption capacity of BNCC−CMCT is 785 mg g−1, which is much higher than that of adsorbents made from other nature-based materials (listed in Table 2). NCC, modified by TEMPO-mediated oxidation, shows a capacity close to that of BNCC−CMCT,18 but this NCC adsorbent is difficult to recycle and regenerate. The adsorption of BNCC− 11775
DOI: 10.1021/acs.langmuir.6b03084 Langmuir 2016, 32, 11771−11779
Article
Langmuir experimental maximum removal capacity. Note that the steric hindrance of an adsorbed MB molecule on one of the two adjacent carboxyl groups may be unfavorable for an otherwise approaching MB, decreasing the probability of one-to-one MBCOO− binding. Theoretical Considerations of Adsorption Dynamics. Here, we theoretically consider the dynamic diffusion− adsorption of dye molecules from a well-stirred bulk fluid to the swollen aerogel, shown in Figure 7.
Table 3. Physicochemical Characteristics of the BNCC− CMCT Aerogel-Assisted Dye Removal Process Used in Solving the Mathematical Model A = 4πR2 C C = C/C0 C0 D kads kdes K = kdes/ kads m mads n0 = C0V/ (madsΓm) r r = r/R R Δr = R/m t t = t(D/ R2) V ε θ = Γ/Γm
Figure 7. Schematic of a spherical aerogel, equally discretized to m points in the radial direction, simplifying the governing adsorption− diffusion partial differential equation (eq 8) and corresponding boundary conditions to m ordinary differential equations (eqs 11−13), which were solved numerically. The aerogel porosity is denoted by ε.
ρ Γ = V(C0 − C)/ mads Γm
The bulk dye concentration C at a desired time t is at a timedependent equilibrium with the aerogel in a system comprising a spherical aerogel with swollen radius R. Upon introducing the aerogel to the solution, the trapped air inside the aerogel pores is nearly instantaneously replaced by the solution, providing a uniform dye concentration inside and outside the adsorbent. The solution is well stirred, and the mass transfer resistance from the bulk to the outer aerogel surface (after a concentration gradient formation as a result of dye adsorption) is negligible. The dye diffusion inside the aerogel is considered to be similar to the bulk, because the pore size O(nanocellulose length, ∼102 nm) is considerably larger than the dye molecule size ( 0), the time change in the bulk dye concentration is equal to the dye diffusion from aerogel surface to the center (boundary condition 1), and as a result of symmetry at the aerogel center, ∂C/∂r = 0 (boundary condition 2). The governing PDE (eq 8) can be nondimensionalized and converted to a set of ordinary differential equations50 using central finite differences for spatial derivatives (methods of lines, MOL). The set of eqs 11−13 provides m ordinary differential equations with the initial condition C = 1 at t = 0, describing the time change of dye concentration at m radial positions from the aerogel surface (n = 1) to the center (n = m), which were solved numerically in Matlab.
(7)
C(n + 1) − 2C(n) + C(n − 1) dC(n) = dt (Δr)2
where the left-hand-side term is the time change of bulk concentration and adsorbed dye, balanced by the right-handside 1D diffusion in a spherical geometry. Equation 7 can be extended as ⎡ 2 ∂C ρ(1 − ε) ∂θ ∂C ∂ 2C ⎤ = D⎢ + 2⎥− Γm ∂t ε ∂t ⎣ r ∂r ∂r ⎦
swollen aerogel outer surface area (m2) bulk dye concentration (kg m−3) dimensionless bulk dye concentration initial bulk dye concentration (0.24 kg m−3) dye bulk diffusion coefficient (∼5 m2 s−1)51 adsorption rate constant (0.1319 s−1) desorption rate constant (1.17 × 10−3 s−1) Langmuir isotherm equilibrium constant (0.0089)
+
ρR2(1 − ε) C(n + 1) − C(n − 1) 2 − εDC0 1 − (n − 1)Δr 2Δr
× Γmkads[(n0 − θ(n))(1 − θ(n)) − Kθ(n)],
(8)
2≤n≤m−1
(11)
with dθ = kads(n0 − θ )(1 − θ ) − kdesθ dt
with discretized boundary conditions (9)
dC(1) −A C(1) − C(2) = εR dt V Δr
where 11776
(12) DOI: 10.1021/acs.langmuir.6b03084 Langmuir 2016, 32, 11771−11779
Article
Langmuir and
C(m) = C(m − 1)
(13)
where dimensionless C = C/C0, r = r/R, θ ranges from 0 to 1, A = 4πR2, and t = t(D/R2). At the beginning of the process (t = 0), C = 1. Note that in this model the aerogel tortuosity is considered ∼1, because the nanoparticles constructing the adsorbent have at least one large dimension (length L ≫ dye size). Figure 8 presents the time change in adsorbent coverage, fitting the experimental data by adjusting kads. If the mass
Figure 9. Adsorption isotherm dependency on pH for the BNCC− CMCT aerogel−MB system at T ∼ 22 °C.
Figure 8. Fractional surface coverage of the BNCC−CMCT aerogel by MB versus dimensionless time (t = tD/R2, symbols) and the best theoretical prediction by solving eqs 11−13 using kads ∼ 0.1319 s−1 (solid line, R2 ∼ 0.95). The dashed line presents a similar best fit when diffusion is neglected, which yields an implausibly small kads ∼ 6.25 × 10−4 s−1 (R2 ∼ 0.94). Note that the initial MB concentration is 240 mg L−1, corresponding to n0 = 1.53, D ∼ 5 m2 s−1, R ∼ 5.4 mm, pH ∼ 7.5, and T ∼ 22 °C.
transfer is not considered (dashed line), the apparent adsorption rate constant kads ∼ 6.25 × 10−4 s−1, corresponding to an adsorption time constant τads ∼ 26.7 min, an implausibly long time for the adsorption of the positively charged MB on the negatively charged functional groups. When the diffusion is taken into account, the best fit (solid line, Figure 8) furnishes kads ∼ 0.1319 s−1, corresponding to an adsorption time constant τads ∼ 7.6 s, attesting to a diffusion-controlled adsorption process. Note that K = kdes/kads ∼ 0.0089 remains constant in all cases, because it is obtained from long-time equilibrium removal data (shown in Figure 5). Effects of pH. Solution acidity influences the charge density of the adsorbent, which in turn affects the adsorption behavior of the adsorbent. Figure 9 shows the results for the adsorption of MB by BNCC−CMCT aerogel at various pHs. The plateau value increases with increasing pH. When the solution pH is increased, carboxyl groups are deprotonated, and the negative charge density of the aerogel increases, increasing the number of adsorption sites for MB+. Langmuir plots of MB adsorption at different pHs are shown in Figure 10, and the Langmuir isotherm parameters are listed in Table 4. At pH ∼ 3, the maximum adsorption is about 192 mg g−1, which is about 25% of the maximum adsorption at pH ∼ 7.5. The adsorption capacity is decreased at low pH as a result of the protonation of carboxylic acid groups. It was noticed that at pH ∼ 2 almost no
Figure 10. Langmuir plots for MB adsorption by the BNCC−CMCT aerogel at various pH and T ∼ 22 °C.
Table 4. Langmuir Isotherm Parameters for MB Adsorption by the BNCC−CMCT Aerogel at Various pH and T ∼ 22 °C pH
Γm (mg g−1)
K
R2
3 4 7.5
192 419 785
0.0316 0.0147 0.0089
0.990 0.995 0.986
dye was adsorbed to the aerogel, indicating that MB may be desorbed at low pH (
