Novel Mesoporous Lignin-Calcium for Efficiently Scavenging Cationic Dyes from Dyestuff Effluent.

A novel adsorbent lignin-calcium was fabricated by a simple flocculation-sedimentation approach to remove methylene blue. The structure and morphology of the well-prepared sample were analyzed by multiple characterization methods. Lignin-calcium microspheres demonstrated a mesoporous and inserted layer structure with a coarse surface. Methylene blue (MB) adsorption by lignin-calcium complied with the Langmuir model, showing a maximum adsorption amount of 803.9 mg/g, exceeding that reported in the literature by 3-22-fold. The adsorption kinetics matched the pseudo-second-order model well. The pore volume diffusion model was technically applied to evaluate the mass transfer mechanisms. The effective pore volume diffusion coefficient was 6.28 × 10-12 m2/s. Furthermore, lignin-calcium exhibited excellent adsorbability for methylene blue across a pH range from 3 to 11 and could be regenerated by hydrochloric acid with an elution efficiency of 62.44%. Multiple mechanisms may support the adsorption. Altogether, the tailor-made lignin-calcium is promising as an efficient and sustainable adsorbent for scavenging cationic dyes from dyestuff effluent.

Introduction

Dyes have been extensively used in spinning, printing, papermaking, foodstuff, maquillage, and other aspects.[1] Discharging dye effluent untreated previously will threaten the ecosystem and cause toxic and carcinogenic health problems.[2,3] Dyeing contaminant ordinarily leads to an expansive area pollution. Currently, it has been always a tough nut to crack due to its weak adhesion and degradation resistance. In decades, a considerable amount of research has focused on removing dyes by multitudinous methods.[4,5] Chemical and biological methods are effective. Meanwhile, they usually require specialized equipment and high energy consumption accompanied with large amounts of by-products.[6] Adsorption prevails over other means in dyeing sewage treatment, relying on its gentle operation, high cost-effectiveness, and nonderived contamination.[7−10] As a traditional adsorbent, activated carbon (AC) is commonly used in various wastewater treatment. Nevertheless, it is also well known for high cost and difficulty of regeneration. Constantly emerging adsorbents derived from biomass have attracted keen interest in virtue of their rich resource and green marketing concept.[11] Activated carbon nanofiber was fabricated from coconut shell charcoal through electrospinning and iodine treatment for methylene blue (MB) adsorption.[12] As reported by Adolfsson et al., cellulose was carbonized, oxidized, and connected with filters for removal of MB.[13] A hierarchical 3D porous carbon was developed by Liu et al. from lotus leaves for capturing rhodamine B.[14] However, many of them were yet below the desired requirement, inevitably gotten drenched by small surface area, sluggish adsorption rate, and poor reusability.[15,16] As a result, it is very essential to screen out a desirable candidate.

It is common knowledge that lignin is viewed as the second-most resourceful macromolecule in nature.[17] Currently, the majority of lignin, as the industrial by-product, is combusted for power, which brings about huge waste of resources, along with serious air pollution. In view of affluent oxygen functional groups as aliphatic and phenolic hydroxyl, carboxyl, and carbonyl, lignin is identified as an available natural adsorbent for dyestuff. However, raw material lignin does not possess competent adsorbability owing to its inactivated structure. For instance, organosolv lignin (OL) obtained from rice straw was used to eliminate methylene blue, with an adsorption capability of 40 mg/g at 20 °C.[18] There are immense potentials of lignin modification resorting to these functional groups above and responsive ortho-para H atoms of phenolic −OH. Lignin and lignin-based adsorbents for dyeing have always stirred up researchers’ enthusiasm, which brought massive relevant reports. Zhang et al.[19] designed a composite trimethyl quaternary ammonium salt of lignin-sodium alginate (QL-SA) from modified lignin and sodium alginate for trapping methylene blue and acid black ATT. The optimal pH values for the decolorization of two dyes were 8 and 3, respectively. Wang et al.[20] developed an aminated lignin from hexane-diamine and lignin under ultrasonication for eliminating Congo red and Eriochrome blue black R (EBBR). Feng et al.[21] used acetic acid lignin to clear methylene blue, a manufacturing process of which contained multifarious operation: deacetylation in an alkaline system, acidification, precipitation, extraction with methanol, evaporation, solventing out, and vacuum drying. Overwhelming majority of them involved varieties of potentially toxic materials, complicated preparation process, poor adsorbability, and narrow pH applicable range, which greatly hindered their development in practical application.[22,23] So, it is urgent to devise a useful bug-free lignin-based adsorbent. Currently, lignin incorporated with metal for dye wastewater treatment was less reviewed. Carboxy-methylated lignin-aluminum (CML-Al), carboxy-methylated lignin-manganese (CML-Mn), and carboxy-methylated lignin-ferrum (CML-Fe) derived from carboxyl-methylated lignin and Al3+/Mn2+/Fe3+ were developed for capturing Procion Blue MX-R and Brilliant Red 2BE.[24,25] Typically, methylene blue (MB) was regularly recommended as an objective pollutant here. Although there are many preparation methods for lignin-based adsorbents, most of them involve expensive and toxic modifying agents, harsh reaction conditions, and complex preparation steps. A new method of flocculation–sedimentation is adopted due to the advantages of cheap and readily available materials, facile reaction condition, and convenient operation.

In this text, the raw material lignin was capable of being directly used without any treatment by expensive modification agents. Lignin-calcium (LC) was obtained conveniently by the flocculation–sedimentation method for capturing MB under a mimetic system for the first time. The preparation process involved a negatively charged dissociated phenoxy ion chelating with free ionic calcium Ca2+ in the alkaline aqueous solution (reaction formula as depicted in Figure S1), which broke the stable state of the system and led to the formation of microflocs of lignin-calcium (LC). The aggregates grew in size, became compact, and precipitated at the bottom of the flask under the increasing gravity ultimately. Instead of the shortcomings mentioned above, considerable adsorption capacity, favorable stability, and broad-spectrum applicability were yearned for LC. Diversified characterizations were studied particularly, involving Fourier transform infrared spectroscopy (FT-IR), X-ray diffraction (XRD), Brunauer–Emmett–Teller (BET), thermogravimetric analysis (TGA), X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and energy dispersive spectroscopy (EDS). Subsequently, multiple factors on MB adsorption were comprehensively investigated, consisting of solid–liquid ratio, pH, initial concentration, ionic strength and regeneration studies, and adsorption thermodynamics. A pore volume diffusion model (PVDM) was applied to describe the diffusional mass transfer process.

Results and Discussion

Material Characterizations

Fourier transform infrared (FT-IR) spectra of lignin are presented in Figure a; the stretching vibration of the hydroxyl group was pointed to peak at 3402 cm–1. Peaks at 2921 and 2852 cm–1 fell to methyl and methylene C–H stretching vibrations, respectively. The peak at 1706 cm–1 was ascribed to that of carbonyl group. Peaks at 1602, 1511, and 1464 cm–1 were for aromatic rings. All those peaks related to lignin were preserved in LC with a slight shift and abated intensity. The peak at 1358 cm–1 belonged to phenolic −OH, which diminished in lignin-calcium. This may be due to the interaction of dissociated phenolic hydroxyl −O– and Ca2+, which indicated the successful construction of lignin-calcium through a chelation reaction. The spectra of the adsorbent lignin-calcium before and after MB adsorption (lignin-Ca + MB) were also compared. For lignin-Ca + MB, peaks at 1597 and 1125 cm–1 were for C=N and C–N stretching vibrations, respectively, which were from the MB molecule. The difference could reflect the successful adherence of MB onto the adsorbent.

Figure 1

(a) FT-IR spectrum, (b) XRD patterns, (c) N2 adsorption–desorption isotherm, and (d) pore size distribution.

As shown in Figure b, the X-ray diffraction (XRD) pattern of lignin maintained a characteristic diffraction peak located at 22°, affirming an amorphous structure. A weakened intensity peak was clearly seen in LC, indicating a destructed crystalline structure compared to lignin. It meant that the phenolate ion (−O–) on lignin chelating with free Ca2+ ions generated lignin-calcium along with an aneretic degree of crystallinity, which may provide accessibility of MB diffusion into its inner structure.[26]

N2 adsorption–desorption isotherms are exhibited in Figure c. Hysteresis curves of LC suggested a mesoporous structure. Based on the Brunauer–Emmett–Teller (BET) method, the specific surface areas of LC were 2.92 and 0.14 m2/g for lignin. Cumulative pore volumes were 0.143 and 0.053 cm3/g for LC and lignin, respectively. In Figure d, the pore diameter distribution of LC was comparably uniform with an average value of 11.8 nm. Academically, these differences would reinforce the adsorption performance of LC, facilitating MB molecules voluntarily permeating into interior pores. As a result, the incorporation of inorganic Ca2+ into lignin modified the original macromolecule structure with a distinct enhancement of specific surface area and cumulative pore volumes, as well as constituting an affluent mesoporous structure, which would improve the adsorption characteristic of the adsorbent.

Thermogravimetric curves viewed from Figure a presented two main weight loss regions. Apparently, LC manifested much lower total weight loss (70.9%) compared with that of lignin (99.37%). In general, the initial weight loss signified evaporation of physically adsorbed water within 25–160 °C. Polymer structure decomposition was for loss at a range of 240–660 °C. Also, a horizontal line between 660 and 800 °C reflected that almost nothing was left in lignin. For LC, apart from water loss, the rapid decomposition of organic constituent initiated at temperature above 270 °C. The improvement of thermal stability was possibly caused by the modification of oxygen-containing functional groups (phenolate ion −O–) chelating with Ca2+.

Figure 2

(a) TGA curves, (b) XPS spectra of lignin-calcium with and without MB, and (c) O 1s and (d) Ca 2p spectra of lignin-calcium.

Meanwhile, from the X-ray photoelectron spectroscopy (XPS) spectra displayed in Figure b, new peaks with binding energies of 164.5 eV (S 2p) and 399 eV (N 1s) were observed in LC after MB adsorption (LC + MB) referred to LC, proving that MB adsorbed onto LC. In Figure c,d, the peak at 532.3 eV of O 1s was assigned to C–O–H. Apparently, the peak for LC + MB slightly migrated after MB was trapped, verifying its possible association with MB+.[27] Meanwhile, peaks at 347.4 and 350.9 eV assigned to Ca 2p of two samples were entirely overlapped, suggesting good stability.

Optical images were taken on a microscope (YS100, Nikon, Japan) with 100× magnification. Lignin particles were inclined to agglomeration, as pictured in Figure S2. In essence, lignin particle agglomeration generated by intermolecular hydrogen bonding was not conducive to the adsorption process.[28] LC, however, displayed translucent sheets, which was quite different from lignin. As discovered from scanning electron microscopy (SEM) images in Figure a,b, it was clearly found that the lignin particles agglomerated, while the LC microsphere appeared to be a fluffy globule with enriched porosity. From the insets of transmission electron microscopy (TEM) images in Figure c,d, LC possessed more distinct flaky apertures after the lignin macromolecular matrix was interspersed with Ca2+. We could infer that the roughness and flaky texture morphology of LC may accommodate MB conveniently.[29]

Figure 3

SEM and TEM images of (a, c) lignin and (b, d) lignin-calcium.

To confirm the elemental composition of the sample, the energy dispersive spectroscopy (EDS) spectra are described in Figure S3. C (72.16 at. %), O (27.27 at. %), Na (0.57 at. %); C (63.43 at. %), O (33.45 at. %), Na (0.00 at. %), and Ca (3.12 at. %) were for lignin (in Figure S3a) and LC (in Figure S3b). The appearance of the Ca element indicated that lignin went through a successful modification process by binding with Ca2+. Results of LC + MB were as follows: C (65.81 at. %), O (26.48 at. %), Ca (2.42 at. %), N (4.89 at. %), S (0.36 at. %), and Cl (0.04 at. %). Similarly, compared with LC, MB adsorption by LC (LC + MB) was well validated by the emergence of N, S, and Cl components.

Impact of Adsorbent Dosage

The influence of LC dosage on MB (400 mg/L) adsorption was optimized from a future economic perspective. As demonstrated in Figure a, R (%) of MB increased fast originally with a supplement of dosage due to numerous active sites provided. Then, R gradually relaxed and approached equilibrium (R = 99.5%) at 50 mg/20 mL with q of 168 mg/g. Conversely, q declined with increasing dosage gradually, which could be ascribed to increased unsaturated sites and the reduced total surface area of LC as well as a lengthened diffusion path owing to particle aggregation. Taking into consideration the invested cost and intended result, 0.05 g/20 mL was chosen as a compromise.

Figure 4

Effect of (a) adsorbent dosage and (b) pH.

Impact of pH

In most cases, adsorption behavior was greatly impressionable by pH. The effect of pH was estimated over a pH range of 2–12. Scientifically, adsorbate concentration variation of solution generated by pH should be taken into consideration, as well as pH changing by addition of adsorbent.[30,31] Hence, in this study, both the initial concentration of MB solution with different pH values and the equilibrium pH value of the solution were also assessed. We have found that the addition of adsorbent LC caused a minor pH change, which could be ignored. It was conducted with 50 mg/20 mL dosage in 400 mg/L MB solution at 333 K. As discovered in Figure b, for pH of 2, LC had an extremely poor adsorption magnitude (22.8 mg/g) and removal (15.24%) toward MB. The zeta potential of LC was 3.2 mV, being slightly positively charged. Under acid conditions, intense protonation of hydroxyl groups endowed massive positive charges to LC. Hence, electrostatic repulsion between LC+ and MB+ largely inhibited the adsorption process. When pH ascended to 3, the zeta potential plunged to −22.4 mV, accompanied by q and R jumping prominently. Satisfactorily, q showed negligible change across the pH region from 3 to 11 (162–168 mg/g), and R could reach 98%. Theoretically, electrostatic force was susceptible to pH variation. So, it was deduced that there were diversified adsorption mechanisms instead of just electrostatic attraction, which would be profoundly elucidated in the follow-up section.

Adsorption Isotherms

Adsorption isotherms were explored at 10 mg/20 mL dosage in MB solution with different concentrations. As shown in Figure a, experimental data were simulated under Langmuir and Freundlich models separately, which were enumerated as eqs and 2.[32,33] The former elucidated single-layer adsorption on the symmetrical surface, differentiating from the latter with multilayer adsorption on the idiosyncratic one

Figure 5

(a) Langmuir and Freundlich equilibrium adsorption isotherms and (b) effect of initial concentration on adsorption kinetics.

q (mg/g) is the maximal scavenging magnitude, and K (L/mg) and K (mg/g) are Langmuir equilibrium and Freundlich adsorption constants accordingly. 1/n identifies the adsorption intensity. Additionally, the average relative error (ARE) was taken to further verify the model-fitting degree according to eq .

y is defined as the experimental value of the dependent variable, while y is that of simulated one, and n points out the number of trials.

As listed in Table , the determination coefficients (R2) of the Langmuir model were higher than that of the Freundlich model at various temperatures, agreeing with ARE analysis. The values of ARE in the Langmuir model were all smaller than those in the Freundlich model at four different temperatures, which revealed that experimental data possessed a better goodness of fit with the Langmuir model. It was presented that MB adsorption onto LC coincided with the Langmuir model with q of 803.9 mg/g at 333 K. Higher temperature triggered higher degrees of freedom and more effective attachment between LC and MB. As given as eq , values of splitting factor R were distributed from 0.021 to 0.617, drawn in Figure S4, indicating a preferential adsorption. Values of 1/n in the Freundlich model fell between 0.1 and 0.5, also implying a beneficial adsorption.[34]

Table 1

Isotherm Parameters for MB Adsorption on LC

model	parameter	303 K	313 K	323 K	333 K
Langmuir	qm (mg/g)	389.0	580.7	653.2	803.9
	KL (L/mg)	0.035	0.040	0.037	0.042
	R2	0.982	0.979	0.976	0.970
	ARE	3.94	4.19	4.28	4.80
Freundlich	KF (L/g)	97.7	130.4	136.4	161.2
	1/nF	0.23	0.25	0.26	0.28
	R2	0.968	0.957	0.951	0.943
	ARE	9.98	11.2	13.2	14.23

Essentially, adsorption capability is a prerequisite for the adsorbent when being accepted. As concluded in Table , LC unveiled significantly better adsorbability toward MB over a broader pH range than other adsorbents. Furthermore, LC improved the utilization value of lignin with enhanced adsorption capacity compared to the raw lignin. It was perceived as an adequate and efficient adsorbent for cationic dye sewage treatment.

Table 2

Comparison of MB Adsorption Quantity of Different Adsorbents toward MB

adsorbent	qm (mg/g)	applicable pH range	reference
lignin-chitosan composite	36.3	6–9	(9)
organosolv lignin	40.0	5–9	(18)
acetic acid lignin (AAL)	63.3	5–7	(21)
Fe3O4@lignosulfonate/phenolic microsphere	283.6	10–12	(23)
Meranti sawdust	120.5	8–10	(35)
alkali-extracted lignin	121.2	5–7	(36)
lignin magnetic nanoparticles (LMNPs)	211.4	10–12	(37)
magnetic xanthate-modified chitosan	197.8	5–11	(38)
80% insoluble subdivision of EHL (corn stalk)	431.1	6–9	(39)
lignin-calcium (LC)	803.9	3–11	current study

Thermodynamics

Abstractly, the adsorption thermodynamic study could explain a series of thermodynamic conception, such as enthalpy change (ΔH0) (kJ/mol), Gibbs free energy change (ΔG0) (kJ/mol), and entropy change (ΔS0) (kJ/(mol·K)). These values could illuminate temperature influencing the adsorption course and mirror adsorption behavior. They were associated by following eqs –7[39,40]

R (8.314 J/(mol·K)) is the universal gas constant and T (K) means the systematic absolute temperature. ΔS0 and ΔH0 are determined from the Van’t Hoff equation in Figure S5. All these parameters are available in Table . Generally, the negative values of ΔG0 declared a voluntary capture. The higher the temperature, the less the ΔG0 values, which signified that rising temperature was conducive to MB attachment on LC. The value of heat from physisorption fell within the scope of 2.1–20.9 and 80–200 kJ/mol for chemisorption. Thus, the value of ΔH0 (28.7 kJ/mol) revealed an endothermic and physical–chemical adsorption process.[41] Moreover, ΔS0 (147.4 J/(mol·K)) showed an entropy production-accompanied randomness collision at the solid–liquid interface.

Table 3

Thermodynamic Parameters of Adsorption

adsorbate	T (K)	ΔG0 (kJ/mol)	ΔH0 (kJ/mol)	ΔS0 (J/(mol·K))
MB	303	–15.9	28.7	147.4
	313	–17.4
	323	–18.9
	333	–20.4

Adsorption Kinetics

Adsorption kinetics of MB onto LC was executed at various initial concentrations with a 50 mg/20 mL dosage at 303 K. As shown in Figure b, higher q was accompanied by a higher C0. q rose perpendicularly with increasing time at first and then decelerated gradually and a subsequent flat stage arrived within 30 min, indicating the appearance of adsorption equilibrium. The fast uptake happening at an early stage could be sustained by sufficient attachment sites available for MB adsorption and a high mass transfer driving force formed by concentration gradient pressure between the MB solution and LC particles, beneficial to overcoming the external resistance. Then, it tended to decelerate and keep balance because of a few sites available. The influence of temperature on adsorption kinetics was as well assessed. Adsorption magnitude (q) rose up from 133 mg/g (303 K) to 168 mg/g (333 K).

As discerned in Figure b, kinetic experimental data were guided into pseudo-first-order and pseudo-second-order patterns, the nonlinearized form of which could be described as eqs and 9.[42,43]

Academically, eqs and 11 were the linearized forms of the pseudo-first-order and pseudo-second-order models, respectively.

t is the contact time (min), and k1 (min–1) and k2 (g/(mg·min)) are referred to as rate constants of pseudo-first-order and pseudo-second-order kinetic models accordingly. The associated parameters are set out in Table . q (mg/g) is the adsorption quantity of the experimental value, and q (mg/g) is that of the calculated one. Compared to the pseudo-first-order model, the determination coefficient (R2 > 0.98) in the pseudo-second order was higher and the calculated adsorption uptake q is closer to the experimental data q at any initial concentration. In addition, from the ARE analysis in Table , experimental data matched better under the pseudo-second-order model with smaller ARE at any initial concentration of MB solution. Comparably, the counterpart of temperature on adsorption kinetics under the two models of linear form was also provided with an accordant outcome, which is demonstrated in Figure S6. Consequently, MB adsorption by LC obeyed the pseudo-second-order model.

Table 4

Kinetic Parameters for MB Adsorption by LC at Various Initial Concentrations (303 K)

		pseudo-first order				pseudo-second order
C0 (mg/L)	qe,exp (mg/g)	k1 (min–1)	qe,cal (mg/g)	R2	ARE	k2 (g/(mg·min))	qe,cal (mg/g)	R2	ARE
50	20.9	0.11	19.2	0.980	6.26	0.007	20.9	0.992	2.41
100	44.6	0.38	41.3	0.973	5.27	0.013	43.3	0.984	3.23
200	75.7	0.48	73.1	0.967	7.57	0.009	76.3	0.989	1.80
300	105.0	0.51	100.3	0.969	6.04	0.007	104.5	0.983	2.44
400	134.0	0.75	128.9	0.964	5.73	0.009	133.4	0.980	2.29

To deeply discuss the adsorption kinetics and diffusional mass transfer mechanism, the pore volume diffusion model (PVDM) was appropriately executed in this study, which correctly described the MB molecule transport in MB solution inside of the LC particles.[44] Governing equations of this model were defined as follows:[45]

Mass balance of the boundary layer surrounding the LC particle:

Mass balance in the LC particle:

Initial conditions and boundary conditions:where K (m/s) is the average film mass transfer coefficient with a value of 1.63 × 10–2 m/s. In this research, C and C (mg/L) are defined as the MB concentration of bulk solution and that of the external surface of LC, respectively. ε (%) is the porosity of LC particle. ρ (g/mL) signifies the apparent density of LC microspheres. D (m2/s) is the effective pore volume diffusion coefficient (6.28 × 10–12 m2/s). C (mg/L) is the concentration of MB at an instantaneous position and time.

Distinctly, it was observed in Figure that the PVDM matched the experimental data reasonably well at concentrations of 50, 100, and 200 mg/L with a lower value of ARE, with an inferior fitting degree for higher concentration. The intraparticle diffusion was solely determined by effective pore volume diffusion (D ≠ 0, D = 0), which dominated the adsorption process. The related parameters of this diffusional mass transfer model are detailed in Table S1. Additionally, the result of kinetics with temperature variation under this model was also surveyed. As exhibited in Figure S7, when adsorption occurred at higher temperatures (323 and 333 K), a better fitting degree was obtained compared to that at 303 and 313 K. It was concluded that the PVDM model predicted the kinetic experimental data well (especially for low concentration and high temperature) and shed light on the mass transfer process of MB on the LC adsorbent.

Figure 6

Diffusional mass transfer model for adsorption at various initial concentrations.

Impact of Ionic Strength

As a frequently used additive, inorganic salts were involved in textile wastewater. Herein, the effect of ionic strength on adsorption was surveyed in the presence of LiCl, NaCl, and KCl ranging from 0 to 0.25 mol/L. As presented in Figure a, the adsorption amount (q) first increased rapidly and then decreased with increasing salt addition. The peak values were 0.04, 0.08, and 0.1 mol/L for LiCl, NaCl, and KCl concentrations. MB aggregation was sensitive to trace amount of salt, which brought about an enhanced adsorption capacity at the first stage.[46] Then, q decreased with more concentrated salts, interpreting the increasingly pronounced competitive adsorption between hydrated ions and MB+, inhibiting the attachment of MB to LC.[47] Normally, K+ had a less hydrated radius compared to Li+ and Na+; it would be more liable to seize active sites.[9,48] Therefore, K+ exhibited a more pronounced q reduction. In spite of their nuance, similar tendency curves were observed.[49,50] As reported by Oladoja et al., an electrolyte could form outer-sphere complexes through electrostatic attraction or inner-sphere complexes via covalent bond formation.[51,52] The adsorbed anions through the former were sensitive to the changes in ionic strength and the adsorption capacity was decreased due to competition adsorption with weakly adsorbing anions. Conversely, the uptaken anions by the latter were insusceptible to ionic strength or responded to intensified ionic strength with promoted adsorption quantity. Overall, salt within the scope of tested concentration was verified to have an inconspicuous influence on adsorption. So, it was deduced that MB adsorption occurred via inner-sphere complexation. Therefore, LC possessed good anti-salt interference and showed promising potential in practical use.

Figure 7

(a) Effect of ionic strength and (b) reusability of lignin-calcium for MB adsorption.

Regeneration

For industrialization scale, good regenerability signifies less cost. Furthermore, Oladoja et al. thought that desorption behavior also manifested the adsorption mechanism.[53] Different eluent may involve different interaction between the adsorbent and adsorbate. Dye detachment from the adsorbent by water, strong acid or base, and organic acid could explain the acting force of weak bond, ion exchange, and chemisorption between the adsorbent and dye correspondingly. On this occasion, four different kinds of eluting solvents, deionized water, 0.1 mol/L hydrochloric acid (HCl), 0.1 mol/L sodium hydroxide (NaOH) aqueous solution, and 0.1 mol/L acetic acid (CH3COOH), were separately used to catch MB that adheres to LC.[54,55] Until adsorption equilibrium was reached, the adsorbent with active sites occupied was removed from the solution by filtration and dried at 60 °C for 12 h. As recorded in Figure b, desorption in hydrochloric acid (HCl) and acetic acid (CH3COOH) presented considerable elution efficiency (E) values of 62.44 and 54.80%, respectively, which could be attributed to ion exchange and chemisorption existing in the capture of MB by LC. For the elution efficiency (E) of 62.44% in hydrochloric acid (HCl), it was manifested that ion exchange was involved in the process of MB uptake by the adsorbent lignin-calcium (LC). During the association of dissociative negative hydroxyl group −O– from LC with cationic dye MB+, the hydrogen ions H+ that originated from the hydroxyl group −OH on LC got released into the liquid phase synchronously, and Cl– was discharged from MB+. With regard to the E value of 54.80% in an organic acid, such as acetic acid (CH3COOH), it was demonstrated that MB entrapment onto the adsorbent LC was related to the chemisorption process.[56] However, desorption in deionized water and NaOH aqueous solution was very inefficient (E = 8.62 and 3.46%, respectively). It was further confirmed that the binding force between the adsorbent LC and the adsorbate MB was very strong instead of a weak bond, which explained indirectly the reinforced adsorption capability of LC. The uncompleted desorption was due to the complexation of MB molecules and active sites on LC. Legitimately, LC demonstrated an ideal renewable candidate and might provide a preponderance for large-scale utilization.

Adsorption Mechanisms

On account of lignin-calcium competent for purging MB across acidic, neutral, and alkaline regions, one single force may be insufficient to support the excellent adsorption behavior. There were different mechanisms involved in the adsorption process. In view of the isoelectric point of LC being within the pH range between 2 and 3, when the pH value was greater than that, the surface of the adsorbent would present positive charges. As schematically illuminated in Figure , dissociated hydroxyl −O– of LC and =N+ in cationic dye MB would generate electrostatic force, which was beneficial for the adsorption process. From the desorption experimental result, we could conclude that ion exchange (capable of being effectively desorbed using a strong acid or base, such as HCl) existed between the adsorbent LC and MB molecule and MB adsorption onto LC was a chemisorption behavior (capable of being effectively desorbed using an organic acid, such as CH3COOH). So, the adsorbent LC with deprotonated hydroxyl −O– would connect with MB+, along with released H+ from LC, and Cl– remained in the MB molecule, which occurred during the ion exchange process. In addition, there were three other constructive microforces: hydrogen bonding, π–π interaction, and mesopore filling. According to the XPS result, the O 1s (532.3 eV) peak migrated after MB was trapped, shedding light on the notion that the O atom was associated with the MB molecule possibly, contributing to MB adsorption. That could be put down to hydrogen bonding, which came into being voluntarily by H from −OH on LC connected with the N atom and benzene rings in MB. π–π interaction existed between massive π systems from LC and aromatic rings of MB dye, which promoted the linkage between LC and MB. Thanks to the morphology and structure characterization, the rich mesopore and interlayer structure enabled MB molecule diffusion into the inner space of LC. Mesopore filling availed MB uptake by LC. Moreover, a good fitting degree was obtained from the diffusion mass transfer process under the PVDM model, which implied that intraparticle resistance governed the MB adsorption process. The outstanding performance of LC for MB adsorption was encouraged, which indicated adequately a bright prospect as an adsorbent in factual dye wastewater treatment.

Figure 8

Proposed mechanisms of MB adsorption onto lignin-calcium.

Conclusions

In this work, a mild and facile flocculation–sedimentation approach was adopted to fabricate mesoporous lignin-calcium (LC) microspheres for eliminating MB. LC possessed favorable and steady adsorption performance in the pH range of 3–11. MB adsorption by LC accorded with pseudo-second-order and Langmuir models with a maximum adsorption capacity q of 803.9 mg/g, substantiated to be an autogenous, thermonegative, entropy augment behavior, which was substantially commanded by intraparticle resistance under the pore volume diffusion model (PVDM). Additionally, LC could be efficiently regenerated by HCl and CH3COOH with desorption efficiency values of 62.44 and 54.80%, respectively. All these results demonstrated that LC would place hopes on clearing cationic dyes from dyestuff effluent.

Materials and Methods

Materials

Lignin was afforded by Shandong Longlive Bio-Technology Co., Ltd. Hydrochloric acid (HCl), sodium hydroxide (NaOH), calcium hydroxide (Ca(OH)2), methylene blue (MB, C16H18ClN3S), acetic acid (CH3COOH) were of analytical grade and were obtained from Sinopharm Group. Real methylene blue wastewater was provided by Jiangsu Shengji Chemical Co., Ltd. Deionized water was used throughout all the experiments.

Fabrication of Lignin-Calcium

Lignin-calcium (LC) was simply fabricated via flocculation–sedimentation. Two grams of lignin powder was dissolved in 50 mL of 0.1 mol/L NaOH solution. A transparent Ca2+ solution was obtained by filtrating a mixture containing 0.4 g of calcium hydroxide and 250 mL of deionized water. Then, both were poured into a 500 mL three-necked flask and agitated at a speed of 250 rpm. The system was heated up to 60 °C and maintained for 2 h at this temperature, following settlement for 6 h without any operation. The brown flocculent precipitate was obtained by filtration and rinsed with water to reach neutral pH. Finally, LC was obtained by vacuum drying at 60 °C for 12 h.

Characterizations

Fourier transform infrared (FT-IR) spectra of lignin and lignin-calcium were recorded by a Thermo Scientific iS5 spectrometer (Nicolet Co., USA) using the attenuated total reflection (ATR) approach, which were scanned across 400–4000 cm–1 by 16 scans. X-ray diffraction (XRD) patterns of lignin and lignin-calcium were obtained across 10° to 80° (2θ) in 5°/min on a Bruker D8 Advance X-ray diffractometer (Bruker Inc., Germany). N2 adsorption–desorption isotherms of the samples were conducted with a TriStar II3020 apparatus (Micromeritics Instrument Co., USA) for surface areas, and the pore diameter was obtained by Brunauer–Emmett–Teller (BET) analysis. Thermogravimetric analysis (TGA) was conducted via a thermal analyzer system (TGA55, TA Instruments, USA), which was scanned from room temperature to 800 °C in 10 °C/min under air atmosphere with a gas flow rate of 100 mL/min. The X-ray photoelectron spectroscopy (XPS) spectra of LC before and after MB adsorption were obtained using a photoelectron apparatus (Ulvac PHI 5300 ESCA, Japan) within 0–1100 eV binding energy by an Al K-alpha source gun. Scanning electron microscopy (SEM) images of energy dispersive spectroscopy (EDS) analysis were acquired through a NoVa Nano SEM 250 at 20 kV (FEI Co., USA). SEM images were explored using a JSM-6390LV (JEOL, Japan) at 3.0 kV. Transmission electron microscopy (TEM) pictures were surveyed by a FEI Tecna G220 STWIN microscope (FEI Co., USA) with a LaB6 electron gun at 200 kV. Zeta potentials of samples with concentration around 0.1 g/L within pH from 2 to 12 were measured on a ZS90 Malvern Zetasizer (Malvern Instruments Ltd., UK). The pH value of dye solution was modulated by 0.1 mol/L HCl or NaOH solution and monitored by a pH meter (Hamilton Bonaduz AG, Switzerland).

Adsorption Experiments

MB solution (1 g/L) was accurately compounded in 1 L volumetric flask at room temperature in advance. Then, a series of various concentration of MB solution were diluted to a desired concentration. Isotherm study was investigated in a concentration of 20–1000 mg/L MB solution at 303, 313, 323, and 333 K. Also, effects of adsorbent dosage, pH, and ionic strength on adsorption and kinetics under different temperatures were all evaluated in 400 mg/L MB solution, as well as the regeneration test. For adsorption kinetics experiment, the effect of initial concentration and time was determined under 50, 100, 200, 300, and 400 mg/L MB solutions. Additionally, the adsorption behavior of the adsorbent in actual MB wastewater was also assessed.

Adsorption experiments were launched in an orbital shaker (MQD-A2, Shanghai Minquan Instrument Co., Ltd., China) at 180 rpm. The experimental system included a certain amount of LC dispersed in 20 mL of MB solution. Once adsorption equilibrium was reached, the supernatant was sampled and analyzed at 664 nm using a UV–vis spectrophotometer (A560, AOE, Shanghai). Effects of various factors were investigated by bulk adsorption tests. The pH of solution was regulated by 0.1 mol/L NaOH and HCl. All tests proceeded in triplicate for taking the average value.

The adsorption amounts adsorbed per unit mass of adsorbent at equilibrium and time t (mg/g), q and q, were counted according to eqs and 19where C0 (mg/L) represents the initial concentration of dye, and C and C refer to the concentrations of dye at time t (min) and equilibrium time. V (mL) and m (mg) are the volume of dye solution and the mass of adsorbent, respectively.

Removal of dye was calculated as eq

Batch Desorption

Regeneration was also evaluated considering industrial-scale probability. Fifty milligrams of LC was dispersed in 20 mL of MB solution. As adsorption balance was achieved, the supernatant was sampled and evaluated to obtain mad. The adsorbent LC loaded with MB was collected by filtration and vacuum-dried at 60 °C for 12 h. Then, it was dispersed in a 50 mL conifer flask containing 20 mL of appointed eluting solvent (0.1 mol/L hydrochloric acid, 0.1 mol/L sodium hydroxide aqueous solution, and 0.1 mol/L acetic acid) and underwent rotating vibration for 2 h for detachment. Afterward, the supernatant was sampled and evaluated to obtain md once more. The elution efficiency (E) was obtained by eq where md (mg) is the mass desorbed of the adsorbate, and mad (mg) stands for the mass desorbed of the adsorbate bound to the adsorbent.