Introduction

With the mass production and wide application of dyes, the printing and dyeing wastewater has become one of the main pollution sources, which seriously endangers the lives of animals and plants and eventually leads to great harm to the entire food chain [1]. Among various dyes, methylene blue (MB) (Scheme 1a) as a typical and widely used cationic dye could lead to serious harm to human health [2]. Treatment technologies, including biological treatment [3], flocculation precipitation [4], membrane separation [5], redox [6], photocatalytic degradation [7], ion exchange [8] and adsorption, have been utilized for the treatment of the dye wastewater. Among them, adsorption is known as a popular approach due to its high efficiency, simple operation and low cost [9, 10]. Therefore, various novel adsorbents have been developed and used on the adsorption removal of dyes from the printing and dyeing wastewater, such as clays, zeolites and their composites, biosorbents, agricultural solid wastes, metal organic frameworks and carbonaceous nanomaterials [11].

Scheme 1
scheme 1

Structure of methylene blue (a) and repeated disaccharide unit of \({\iota}\)-carrageenan (b)

Full size image

In the past decades, graphene oxide (GO) as an important derivative of graphene has attracted increasing attention on the adsorption [11, 12]. The existence of abundant functional groups, such as carboxyl (–COOH) and hydroxyl (–OH), endowed GO the ability to interact with water molecules through hydrogen bonding, showing high hydrophilicity [13]. Moreover, the large surface area and abundant oxygen-containing functional groups have endowed GO excellent adsorption ability to remove organic dyes from water [14]. However, the high colloidal stability of GO sheets makes it hard to collect them after the adsorption [15], hindering potential practical applications. To address this issue, magnetic separation as an excellent separation technology has been utilized, where magnetic adsorbents can be facilely separated using an external magnetic field within several seconds or minutes [16]. However, the preparation process for magnetic adsorbents was usually complicated and the magnetism gradually decreased after the repeated use. Besides, secondary pollution could be caused through leaching metal species into water under acid conditions.

Recently, three-dimensional GO-based materials, such as anionic polyacrylamide/GO aerogels [17], agar/GO [18] and lignosulfonate/GO composite aerogels [19], have been constructed through incorporation of polymers to GO sheets with the assistance of cross-linking agents, which were used on the adsorption removal of dyes from water. These GO-based aerogels have high pore volumes, large surface areas and abundant oxygen-containing functional groups and thus generally show large adsorption capacities for dyes. In addition, they could be easily separated from water after use [20]. For the preparation of GO-based aerogels, polysaccharide biopolymer is often used.

Polysaccharide biopolymer materials have attracted widespread attention due to their unique renewable, non-toxic, low cost, biocompatible and biodegradable properties. Carrageenan is a family of linear sulfated polysaccharides that are obtained by alkaline extraction from some species of red seaweeds, which has been widely used as gelling agent, thickener and stabilizer in many fields. Among them, \({\iota}\)-carrageenan (Scheme 1b) is a hydrophilic colloid with a long-chain structure and high stability [21]. Each disaccharide unit of \({\iota}\)-carrageenan has two sulfonate groups, which can interact with cationic dyes [22]. The ability of carrageenan to interact with cationic dyes [23] and its gelling property make it a promising participant for preparing GO-based aerogels, which are used for the dye adsorption [24].

Therefore, in the present study, natural polysaccharide of \({\iota}\)-carrageenan and GO are employed to prepare a kind of novel GO-based aerogel, which is subsequently used for the dye adsorption. The integration of polysaccharide long-chains and GO nanosheets resulted in a porous aerogel composite, which could be facilely separated from water after the adsorption. A gentle and simple synthesis approach is applied to avoid severe changes of the oxygen-containing functional groups on the surface of GO, and the whole process is free of any toxic polymer or cross-linking agent. In this study, MB was chosen as a typical organic dye in water and its adsorption behavior on the GO/\({\iota}\)-Car aerogel was investigated in detail.

Experimental

Chemicals

GO was purchased from Chengdu Organic Chemistry Co., Ltd., and its detailed properties are listed in Table S1 of Supplementary Materials. \({\iota}\)-carrageenan and methylene blue (MB) were obtained from J&K Scientific Ltd. Ethanol, HCl (36.0–38.0%), NaOH (≥ 96.0%) and alkyl glycoside (APG) were supplied from Tianjin Damao Chemical Reagent Co., Ltd. Deionized water was used throughout the experiments.

Preparation and characterization of GO/\({\iota}\)-Car aerogel

In a typical synthesis, a given amount of APG (1% v/v) was added dropwise into 20 mL of dispersed GO solution (4 mg mL−1) with continuous stirring for 30 min. Here, APG is a natural green non-ionic surfactant prepared by dehydration condensation of starch or its hydrolyzed sugar with fatty alcohol. The mixture was then heated at 80 °C while adding 16 mg of \({\iota}\)-carrageenan and stirred for another 1 h until completely dissolved. After heating, the mixture was cooled at room temperature to form a hydrogel. Then, the as-prepared material was put in a refrigerator at -20 °C for 4 h and freeze-dried for 24 h to obtain the GO/\({\iota}\)-Car aerogel. Actually, it was found that increasing the GO/\({\iota}\)-Car ratio could enhance the removal of MB, but the aerogel prepared at a ratio larger than 5:1 was not stable enough and would collapse after the adsorption. Therefore, the mass ratio of GO to \({\iota}\)-carrageenan was fixed at 5:1. The GO/\({\iota}\)-Car aerogel was washed with deionized water before its use for the adsorption.

The aerogel was characterized by scanning electron microscopy (SEM), Fourier transform infrared spectrometer (FTIR), thermogravimetric analysis (TGA), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and zeta potential measurements. Details for the methods of characterization are provided in Text S1 of Supplementary Materials.

Adsorption of single MB

Batch experiments were conducted for MB adsorption on GO/\({\iota}\)-Car aerogel. For comparison, the adsorption performance of raw \({\iota}\)-carrageenan was also tested and is discussed in Text S2. The remaining MB concentration after the adsorption was determined at 664 nm with a UV–Vis spectrophotometer (Sunny Hengping Scientific Instrument Co., Ltd., China). Kinetic studies were performed by fitting the data to the pseudo-first-order kinetic model [25], pseudo-second-order kinetic model [26] and intra-particle diffusion model [27]. Adsorption isotherms were fitted with the Langmuir model [28] and Freundlich model [29]. Thermodynamic analysis was conducted by calculating Gibbs free energy change (ΔG0), enthalpy change (ΔH0) and entropy change (ΔS0) [30, 31]. More details are provided in Text S3 of Supplementary Materials.

Selective adsorption for cationic and anionic dyes

In order to test the selective removal ability of the aerogel, 5 mg of GO/\({\iota}\)-Car was soaked into 20 mL of MB/MO mixed solution (CMB = CMO = 10 mg L−1). The mixed solution was prepared by physical mixing 20 mg L−1 MB solution with 20 mg L−1 MO solution in equal volume. The mixed solutions were sampled at different time intervals after the adsorption, and absorbances were measured at 664 and 464 nm by using the UV–Vis spectrometer. Concentrations of MB and MO in the mixed solutions were calculated by solving the corresponding binary linear equations, which are provided in Text S4 of Supplementary Materials.

Reuse of GO/\({\iota}\)-Car aerogel

In order to investigate the reusability of the GO/\({\iota}\)-Car aerogel, 25 mg of GO/\({\iota}\)-Car aerogel was firstly added into a conical flask containing 25 mL of MB aqueous solution (100 mg L−1). After the adsorption for 480 min, the GO/\({\iota}\)-Car aerogel was separated from the aqueous solution and was regenerated with 0.1 M HCl for several times. Subsequently, the desorption and adsorption processes were repeated three cycles and changes of the removal rate were analyzed.

Results and discussion

Characterization of GO/\({\iota}\)-Car

The morphologies of GO and GO/\({\iota}\)-Car were analyzed by SEM (Fig. 1). SEM images of GO (Fig. 1a, b) exhibit a typical two-dimensional structure with many wrinkles [32]. The GO/\({\iota}\)-Car aerogel composite (Fig. 1c, d) exhibited a three-dimensional porous structure, which resulted from the connection of \({\iota}\)-carrageenan long chains and GO.

Figure 1
figure 1

SEM images of GO (a, b) and the GO/\({\iota}\)-Car aerogel (c, d) at different magnifications

Full size image

GO, \({\iota}\)-carrageenan and the GO/\({\iota}\)-Car aerogel were further analyzed by FTIR (Fig. 2). The broad band at 3200–3500 cm−1 can be found in all spectra due to the abundant hydroxyls present in all three samples. In the spectrum of GO, bands appear at 1731, 1621, 1055 and 1386 cm−1, which are attributed to C = O, C = C, C–O stretching and C–H deformation vibrations [33]. For \({\iota}\)-carrageenan, bands for C–H stretching vibration at 2933 cm−1 and C–O–C stretching vibration at 1159, 1075 and 930 cm−1 were observed [34, 35]. In addition, the bands at 1029 and 1267 cm−1 were associated with glycosidic linkages [36] and O = S = O stretching [34, 37], respectively. GO/\({\iota}\)-Car exhibited both bands of GO and \({\iota}\)-carrageenan with slightly shifted wavenumbers, which was possibly due to structural changes between free and GO-bonded \({\iota}\)-carrageenan. The spectrum of GO/\({\iota}\)-Car + MB was also analyzed for the sample of GO/\({\iota}\)-Car aerogel after the MB adsorption. Bands in the spectrum at 1598, 1489, 1338 and 885 cm−1 indicated that MB had been adsorbed by the aerogel successfully [38].

Figure 2
figure 2

FTIR spectra of GO, \({\iota}\)-carrageenan, GO/\({\iota}\)-Car and GO/\({\iota}\)-Car-MB

Full size image

The behaviors of GO, \({\iota}\)-carrageenan and GO/\({\iota}\)-Car aerogel under the heat treatment were also analyzed. As shown in Fig. 3, the weight of carrageenan decreased drastically below 185 °C, losing 23.1% of its weight. This was due to the weight loss caused by the volatilization of water, and the residual weight at 800 °C was 29.0% (losing 71.0%). TGA curves of GO and GO/\({\iota}\)-Car aerogel were similar and could be divided into three main stages. The first stage was the weight loss of adsorbed water below 190 °C [39]. Percentages of the weight loss of GO and GO/\({\iota}\)-Car aerogel were 16.9% and 16.0%, respectively. The second stage was a rapid weight-losing stage between 190 °C and 235 °C, during which the functional groups decomposed [40]. Percentages of the weight loss of GO and GO/\({\iota}\)-Car aerogel were 19.1% and 18.0%, respectively. The third stage was ascribed to the decomposition of more stable functional groups, and the final residual weights were 42.9% and 43.5% for GO and GO/\({\iota}\)-Car aerogel, respectively. As GO accounts for a much larger proportion than \({\iota}\)-carrageenan in the composite aerogel, the TGA curve of GO/\({\iota}\)-Car was more similar to that of GO [18].

Figure 3
figure 3

TGA curves of GO, \({\iota}\)-carrageenan and GO/\({\iota}\)-Car

Full size image

Figure 4 displays XRD patterns of GO, \({\iota}\)-carrageenan and the GO/\({\iota}\)-Car aerogel. For GO, a strong diffraction peak appeared at 2θ = 10.64°, which was the characteristic diffraction peak of GO. The layer spacing (d) of GO was calculated to be 0.83 nm according to the Bragg’s law [41]. The diffraction peaks observed in the pattern of \({\iota}\)-carrageenan were attributed to KCl impurity (JCPDS 41-1476) [42]. For the GO/\({\iota}\)-car aerogel, the diffraction peak of GO completely disappeared, indicating that GO sheets were well exfoliated in the composite aerogel [43]. There were no diffraction peaks existed as those appeared for \({\iota}\)-carrageenan, which might be related to the washing away of KCl impurity during the preparation of the GO/\({\iota}\)-Car aerogel.

Figure 4
figure 4

XRD patterns of GO, \({\iota}\)-carrageenan and GO/\({\iota}\)-Car aerogel

Full size image

XPS spectra were used to characterize their surface functional groups (Fig. 5). Survey scans of GO and GO/\({\iota}\)-Car (Fig. 5a, d) possessed obvious C1s and O1s peaks. In the C1s spectra (Fig. 5b, e), three peaks were ascribed to carbon in C = O, C–O and C–C & C = C bonds [44]. In the O1s spectra (Fig. 5c, f), oxygen in C–O and C = O bonds were indicated as two peaks [45]. As shown in Table S2, the contribution of C–O in C1s spectrum of GO/\({\iota}\)-Car increased significantly, which was due to the introduction of a large number of hydroxyl groups by the polysaccharide molecules. The reduction of graphene oxide during the preparation process might occur, but the result here indicated that its contribution was not enough to offset the increase of hydroxyl groups. The increased contribution from C–O was also observed from the O1s spectra (Fig. 5c, f and Table S2), supporting the above conclusion.

Figure 5
figure 5

XPS spectra of GO (a, b and c) and GO/\({\iota}\)-Car (d, e and f)

Full size image

Effects of pH and adsorbent dosage on MB adsorption

The effect of pH on the MB adsorption by GO/\({\iota}\)-Car aerogel is illustrated in Fig. 6a. The removal rate rose with increasing pH values. Similar pH effects were observed for the MB removal using other adsorbents such as poly (c-glutamic acid) [46] and alginate [47]. The increased MB adsorption amounts at high pH values were due to less competition between H+ and MB and the more significant deprotonation of functional groups on the surface of GO/\({\iota}\)-Car [48]. As shown in Table S3, negative zeta potentials were observed for the GO/\({\iota}\)-Car aerogel, and the absolute values increased with increasing pH. The interaction between the aerogel and MB molecules is non-covalent interaction, i.e., electrostatic and π–π interaction [49]. At higher pH, stronger electrostatic attraction between the negatively charged carboxyl, hydroxyl and sulfonic acid groups and the positively charged MB cations resulted in a larger adsorption amount.

Figure 6
figure 6

Effects of pH (a) and GO/\({\iota}\)-Car aerogel dosage (b) on the MB adsorption (adsorption conditions: dosage = 0.4 g L−1, C0 = 100 mg L−1, T = 298 K, t = 240 min (a); C0 = 100 mg L−1, T = 298 K, pH = 7, t = 240 min (b))

Full size image

Figure 6b shows the effect of GO/\({\iota}\)-Car aerogel dosage on the MB adsorption. With increasing dosages, the removal rate of MB increased first and then remained almost unchanged. At a higher aerogel dosage, the amount of available adsorption sites increased, leading to a higher removal rate until almost complete removal of MB was achieved. From Fig. 6b, the optimal dosage was determined to be 1.0 g L−1.

Adsorption kinetics

The effect of adsorption time at initial MB concentrations of 100, 200 and 400 mg L−1 is shown in Fig. 7a. The adsorption curve showed a rapid adsorption in initial 60 min, which was ascribed to a large number of adsorption sites on the surface of GO/\({\iota}\)-Car aerogel [50]. The decrease of adsorption rate with prolonged time was due to the reduction of the MB concentration gradient between the two phases.

Figure 7
figure 7

Effect of contact time on MB adsorption capacity (a), pseudo-first-order kinetic model (b), pseudo-second-order kinetic model (c) and intra-particle diffusion model (d) for MB adsorption on GO/\({\iota}\)-Car aerogel (adsorption conditions: dosage = 1.0 g L−1, T = 298 K, pH = 7, t = 0 ~ 240 min)

Full size image

The kinetic parameters obtained by nonlinear fitting to the pseudo-first-order kinetic model (Fig. 7b) and pseudo-second-order kinetic model (Fig. 7c) models are shown in Table 1. Coefficients of determination (R2) of the latter model were higher, and the calculated adsorption amounts were closer to the corresponding experimental values (qexp). Therefore, the pseudo-second-order kinetic model performed better for the description of MB adsorption behavior on the GO/\({\iota}\)-Car aerogel compared to the pseudo-first-order kinetic model.

Table 1 The parameters of pseudo-first-order kinetic model and pseudo-second-order kinetic model for MB adsorption on GO/\({\iota}\)-Car aerogel
Full size table

The intra-particle diffusion model was also used (Fig. 7d and Table 2). The adsorption curve can be divided into three linear segments, indicating three adsorption stages. At the initial stage, there were abundant adsorption sites available on the surface of GO/\({\iota}\)-Car aerogel, and the adsorption proceeded quickly. With the adsorption going on, the amount of adsorption sites gradually decreased, and MB molecules gradually formed a boundary layer, resulting in declined adsorption rate [51]. The final stage involved diffusion of MB molecules into pores and pore-wall surfaces of the GO/\({\iota}\)-Car aerogel until the saturation stage reached.

Table 2 The parameters of intra-particle diffusion model for MB adsorption on GO/\({\iota}\)-Car aerogel
Full size table

Adsorption isotherms

The nonlinear fitting results of Langmuir and Freundlich isotherms are shown in Fig. 8, and the corresponding parameters are listed in Table 3. The results showed that high R2 was obtained for these two models, indicating both models could describe the adsorption process. As R2 of the Freundlich model was higher than that of the Langmuir model, the adsorption might follow the multilayer adsorption and heterogeneity distribution mechanism [29].

Figure 8
figure 8

Langmuir isotherm (a) and Freundlich isotherm (b) for the adsorption of MB onto the GO/\({\iota}\)-Car aerogel (adsorption conditions: dosage = 1.0 g L−1, t = 480 min, pH = 7, C0 = 100 ~ 1000 mg L−1)

Full size image
Table 3 The parameters of Langmuir and Freundlich isotherm for MB adsorption on GO/\({\iota}\)-Car aerogel
Full size table

The maximum adsorption capacity (qm) of MB on the GO/\({\iota}\)-Car aerogel was calculated to be 245.28 mg g−1 from Langmuir model. As shown in Table 4, the GO/\({\iota}\)-Car aerogel showed higher or comparable adsorption capacity of MB than reported values, including graphene aerogel [52], Fe3O4/RGO aerogel [53] and CMC/chitosan/GO [54]. Although such comparisons may not be sufficient enough due to different experimental conditions such as pH value and temperature, the result here suggested that the GO/\({\iota}\)-Car aerogel prepared may be an economical and efficient green adsorbent for removal of cationic dyes such as MB.

Table 4 The adsorption capacity of MB on various adsorbents
Full size table

Adsorption thermodynamics

Thermodynamic parameters were obtained from the Van't Hoff diagram (Fig. S1). As shown in Table 5, the value of ΔG0 was negative at all temperatures, indicating that the adsorption of MB on the GO/\({\iota}\)-Car aerogel was spontaneous. As the temperature increased, the absolute value of ΔG0 increased as well. The value of ΔH0 was negative, indicating that the adsorption process of MB on the GO/\({\iota}\)-Car aerogel was an exothermic process, which was in accordance with the experimental result that the adsorption amount of MB decreases with increasing temperature. The adsorption of MB molecules led to freedom loss and thus decreased entropy, while desorption of water molecules, which occurred at the same time [59] and change the adsorbent surface, resulted in increased entropy [60]. The positive ΔS0 indicated the net disorder increase in the whole adsorption process.

Table 5 The parameters of adsorption thermodynamics for MB adsorption on the GO/\({\iota}\)-Car aerogel
Full size table

Selective adsorption

To investigate the adsorption selectivity of GO/\({\iota}\)-Car aerogel for cationic dyes, MO as a typical anionic dye was mixed with MB and the adsorption of MB from mixed dye solutions was investigated (Fig. 9). At pH = 7, the removal rate of MB approached 90% at 240 min, while MO was only removed by about 6%. The inserted picture in Fig. 9 shows the color change from green (the color of the mixed solution of MB and MO) to orange after the adsorption, which also verified the selective removal of MB. This result was mainly attributed to the electrostatic attraction between cationic dye MB and the negatively charged functional groups on the surface of GO/\({\iota}\)-Car aerogel. In contrast, the anionic dye MO was expected to suffer electrostatic repulsion, resulting in a very low removal rate. Therefore, the GO/\({\iota}\)-Car aerogel prepared in this study could selectively adsorb cationic rather than anionic dyes, showing some potential for practical applications on the dye separation.

Figure 9
figure 9

The corresponding removal rate of MB/MO mixture by the GO/\({\iota}\)-Car aerogel (adsorption conditions: dosage = 0.25 g L−1, CMB = CMO = 10 mg L−1, T = 298 K, pH = 7)

Full size image

In order to better understand the adsorption mechanism, schematic diagrams are provided in Fig. 10 based on previous reports [12, 61, 62] and the results of this study. Generally, electrostatic interaction, π–π stacking and hydrophobic interaction were involved for the adsorption. Particularly, the positively charged MB cation tended to interact with the GO/\({\iota}\)-Car aerogel which possessed negatively charged carboxyl, hydroxyl and sulfonic acid groups mainly through electrostatic attraction (Fig. 10a). For the adsorption of MO (Fig. 10b), the negatively charge on MO molecules would cause electrostatic repulsion from the GO/\({\iota}\)-Car aerogel, resulting in limited adsorption amount. Consequently, the GO/\({\iota}\)-Car aerogel was more favorable for the adsorption of MB than MO.

Figure 10
figure 10

Adsorption mechanisms of MB (a) and MO (b) on the GO/\({\iota}\)-Car aerogel

Full size image

Desorption and reuse of the adsorbent

The reusability is of great importance for an adsorbent. After the adsorption at initial MB concentration of 100 mg L−1, adsorbent dosage of 1 g L−1 and adsorption time of 480 min, the GO/\({\iota}\)-Car aerogel was separated from the solution and was regenerated with 0.1 M HCl for several times. Images of the color change of the solutions during the adsorption–desorption process are shown in Fig. S3, indicating the efficient desorption of MB from the GO/\({\iota}\)-Car aerogel used. In addition, from Fig. 11, it can be seen that the removal rate remained at a satisfactory level and could still reach 91.9% after three desorption–adsorption cycles. Therefore, GO/\({\iota}\)-Car aerogel could be a reusable and effective adsorbent for the cationic dye removal.

Figure 11
figure 11

Removal rates of MB by the GO/\({\iota}\)-Car aerogel in different cycles

Full size image

Conclusion

A novel three-dimensional GO/\({\iota}\)-Car composite aerogel was prepared in a facile and green approach and then employed as an adsorbent for MB removal from water. Kinetic studies showed that the adsorption process followed the pseudo-second-order kinetic model. According to isotherm studies, the maximum adsorption capacity of MB on the GO/\({\iota}\)-Car at 293 K was calculated to be 245.28 mg g−1, which was higher or comparable to those values of reported adsorbents. Thermodynamic calculations proved that the adsorption was a spontaneous, exothermic process with the enthalpy change of -10.36 kJ mol−1 and the entropy change of 54.64 J mol−1 K−1. More importantly, the aerogel could selectively adsorb MB in the MB/MO mixed solution, which was attributed to the electrostatic attraction between the negatively charged aerogel and the positively charged MB. The MB removal rate by the GO/ \({\iota}\)-Car aerogel could still reach 91.9% after three desorption–adsorption cycles. Therefore, the GO-based aerogel prepared in this study has potential applications for the cationic dye adsorption and dye separation from water.