Introduction

A large amount of synthetic dyes are used in different industries such as textile (Patil et al. 2010), printing (Seema et al. 2018), cosmetics, food (Guerra et al. 2018), paper making (Blus et al. 2014), etc. Anionic dyes are applied mostly because of their excellent color fastness, bright colors, etc. However, dyes are usually toxic and harmful to humans and environment and need to be removed from the effluents. Various treatments are applied for dye removal such as coagulation-flocculation (Panswed and Wongchaisuwan 1986), oxidation (Malik and Saha 2003), membrane separation (Ciardelli et al. 2001), electrochemical processes (de Paiva et al. 2018), etc. Among these, adsorption is a widely-used method because of its low operating cost, capability to operate at very low concentrations, simple design and easy technical access (Cheng et al. 2016; Abousalman-Rezvani et al. 2019a). Activated carbon (Mezohegyi et al. 2012), inorganic-organic hybrid materials (García et al. 2014), industrial byproducts and wastes (Ahmad et al. 2009), and hydrogels (Soleimani et al. 2018) are the most used adsorbents for removal of different dyes.

It is well-known that ionic hydrogel adsorbents can be used for removal of oppositely charged dyes, ions, and heavy metals (Kaşgöz and Durmus 2008; Bekiari et al. 2008; Atta et al. 2012; Lu et al. 2015; Abousalman-Rezvani et al. 2019b). Hydrogels adsorb materials from aqueous solutions and keep them dissolved and swell up by water absorption (Abdollahi et al. 2016; Nikravan et al. 2018) by making hydrogen bonding with water (Fallahi-Samberan et al. 2019) or electrostatic interactions (Modarresi-Saryazdi et al. 2018; Dehghani et al. 2019). The amount of absorbed water can be controlled by variation of pH or temperature (Liu et al. 2010; Balea et al. 2017).

Introduction of natural materials into structure of hydrogels affects swelling properties and therefore adsorption capacity. Among them, those based on nontoxic and biodegradable polysaccharides are one of the most attractive ones due to their unique properties like hydrophilicity, ecofriendly behavior, and biodegradation (Abdeen and Mohammad 2014; Sharma et al. 2015). Cellulose as naturally renewable and the most available polysaccharide has unique mechanical and physical properties (Khadivi et al. 2019a,b), and due to its hydroxyl groups, it can be modified by different compounds (Bayramoglu et al. 2012; Zhou et al. 2014) to improve water absorption and dye adsorption (Ibrahim et al. 2007). Pei et al. (2013) reported fabrication of surface-quaternized cellulose nanofibrils to be used for removal of Congo red as anionic dye. Stanciu and Nichifor (2019) synthesized dextran hydrogel with quaternary ammonium pendent groups and used it in removal of methyl orange (MO) as anionic dyes. Haque et al. (2011) prepared an iron terephthalate as metal-organic frameworks (MOFs) and applied it for the adsorption of MO and methylene blue (MB) as anionic and cationic dyes respectively. Fang et al. (2016) synthesized a high-capacity cationic hydrogel adsorbent (CHA) for adsorption of Acid Black. Zhao et al. (2012) produced semiIPN hydrogel composites by photopolymerization of poly(ethylene glycol) and acrylamide (AAm) in presence of chitosan (CS) and used them for dye adsorption studies. Chatterjee et al. (2016) synthesized chitosan-based hydrogel beads and studied the effect of alkali treatment on morphology, adsorption ability, and surface chemistry. Balea et al. (2019) combined cellulose nanofibers with chitosan for a successfully removal of flexographic inks (copper phthalocyanine blue, carbon black, and diarylide yellow), and Sanchez-Salvador et al. (2018) showed the dye removal effect of nanocellulose-graft-chitosan polymers. Zhang et al. (2017) synthesized a magnetic cellulose/Fe2O3 hydrogels by sol-gel transition method and used it as adsorbent of MO. Chen et al. (2017) prepared nitrogen-doped TiO2/Q-NFC hybrid cryogels by one pot hydrothermal treatment method and their MO adsorption/photodegradation properties were characterized. Li et al. (2017) prepared a cationic absorbent poly-epichlorohydrin-ethylenediamine hydrogel for removal of dyes and insoluble oil from wastewater. Karthika and Vishalakshi (2015) studied the grafting of DMAEMA on Gellan gum and its adsorption capacity for MO. Liu et al. (2015a, b) synthesized poly(DMAEMA-co-AA) hydrogels as adsorbent of anionic and cationic dyes. Salama et al. prepared an adsorbent based on carboxymethyl cellulose (CMC) for MO removal from aqueous solution. For this aim, DMAEMA was grafted onto backbone of CMC. Then, the effect of solution pH, contact time, and dye concentration on the adsorption behavior was studied (Salama et al. 2015). Zheng et al. (2018) synthesized a composite film using chitosan and dialdehyde microfibrillated cellulose nano fibrils by solvent-casting method for adsorbtion of Congo red from aqueous solution. Lin et al. (2016) prepared an adsorbent based on cross-linked CMC by grafting of dimethyldiallylammonium chloride (DMDAAC), and its adsorption properties on MO and MB dyes were investigated. Kono et al. (2016) synthesized a series of cationic cellulose hydrogels (CCGs) by crosslinking reaction with PEGDE as a crosslinking agent, and their adsorption capacity for three types of anionic dyes, AR13, AB92, and AR112 was investigated. Oladipo et al. (2014) prepared a semiIPN composite hydrogel from cellulose-graft-polyacrylamide and HAP where adsorption behavior for RB2 was investigated. Tu et al. (2017) prepared the chitosan/REC/cellulose composite hydrogels for removal of Congo red from aqueous solution in different adsorption conditions. Kono (2015) fabricated a cellulose polyampholyte hydrogels and the adsorption capacity of the hydrogels was investigated for three anionic dyes AR9, AR13, and AB92.

In current work, NCC is modified by (3-amino)propyltriethoxysilane (APTES) and hexadecyltrimethoxysilane (HMTS) to prepare ANCC and HNCC respectively to investigate the effect of surface groups on swelling and dye removal behaviors of nanocomposite hydrogels. Crosslinked PDMAEMA/NCC hydrogels containing various amounts of (modified) NCC are fabricated by in situ polymerization and used as MO adsorbent. Due to pH-sensitive swelling properties of the hydrogels, dye adsorption properties are studied at different pH values. Finally, adsorption kinetics of MO is investigated.

Experimental methods

Materials

2-(dimethylamino)ethyl methacrylate (DMAEMA, Aldrich, 98%) was distilled and stored at 0 °C. Microcrystalline cellulose (MCC, Sigma-Aldrich, 98%), sulfuric acid (H2SO4, Merck, 98%), (3-aminopropyl)triethoxysilane (APTES, Sigma-Aldrich, 99%), hexadecyltrimethoxysilane (HDTMS, Sigma-Aldrich, ≥ 85%), N,N'-methylenebis(acrylamide) (MBA, Sigma-Aldrich, 99%), potassium persulfate (KPS, Merck, 99%), azobisisobutyronitrile (AIBN, Aldrich, 98%), methyl orange (MO, Merck, %), sodium hydroxide (NaOH, Merck, ≥ 97), and ethanol (Kimia Alcohol Zanjan, 96%) were used as received.

Preparation and modification of NCC

NCC rods (100–200 nm in length and 10–20 nm in diameter, Fig. S1) were prepared by acid hydrolysis of MCC (10–50 μm in length and 10–25 μm in diameter, Fig. S1) as reported in literature (Golshan et al. 2017; Safavi-Mirmahalleh et al. 2019). In modification step, NCC (0.5 g) was sonicated (Hielscher UP400, 30 min at room temperature) in ethanol (50 mL) and after that, modifier was added dropwise (2 mL, 8.5 mmol APTES, and 5.13 mmol HDTMS). Reaction was continued for 3 h at 60 °C under nitrogen atmosphere. Obtained suspension was centrifuged (Hettich Universal 320) at 10000 rpm and washed with water/ethanol three times. The final product (ANCC for APTES and HNCC for HDTMS) was then freeze-dried (Dena Vacuum FD-5010-BT) overnight.

Synthesis of PDMAEMA-based nanocomposite hydrogels

To synthesize nanocomposite hydrogels, DMAEMA was crosslinked by MBA in presence of different amount of (modified) NCCs (1.0–3.0 wt. %). To this end, NCC was ultrasonically (Hielscher UP400, 30 min at room temperature) dispersed in water, and then, DMAEMA was added to the mixture where a homogenous media was obtained using a magnetic stirrer. After that MBA (0.26 g, 1.7 mmol) in 5-mL water was added. After rising temperature to 60 °C, KPS (0.06 g, 0.2 mmol) was added, and reaction was completed after 24 h under nitrogen atmosphere. A same procedure was performed for modified NCCs (ANCC and HNCC). The samples were washed with distilled water for 3 days via dialysis method, and then final product was dried in vacuum at 50 °C. HNCC containing hydrogels were prepared in ethanol due to dispersibility problems of HNCC in water. Nanocomposite hydrogels were named HNCx, HACx, and HHCx for NCC, ANCC, and HNCC containing samples, respectively where x was the weight percentage of nanorods.

Investigation of swelling behavior

Gravimetry was used to study swelling behavior of nanocomposite hydrogels. After swelling of dried hydrogels at different pH values, sampling was performed at predetermined time intervals, and weighting was done after removing excess water on surface by Kleenex. The SR for each sample was calculated based on dry and swollen weights (Wd and WS) via Eq. (1) (Anirudhan and Rejeena 2012):

$$ \mathrm{SR}=\frac{Ws- Wd}{Wd} $$
(1)

Investigation of adsorption behaviors

Dye uptake capacity of hydrogels was measured in aqueous solutions of MO at two pH values (acidic condition of pH = 1 where the hydrogels are fully-protonated and basic condition of pH = 10 where the hydrogels are deprotonated) at ambient temperature. Hydrogel (0.05 g) was soaked in MO solution (5 mg/L, 100 mL) and samples were taken out (1 mL) and replaced with fresh water during 6 days. Concentration of MO was evaluated at 468 nm for pH = 1 and 508 nm for pH = 10. The adsorption capacity (qe, mg dye/g hydrogel) was obtained via Eq. (2) (Akkaya et al. 2009).

$$ {q}_e=\frac{\left( Ci\hbox{--} Ce\right)\ V}{m} $$
(2)

Ci and Ce were initial and instantaneous concentrations of MO respectively considering V liter of solution and m g of hydrogel.

Characterization

Fourier transform infrared (FT-IR) spectroscopy was performed by means of a Bruker Tensor 27 FT-IR-spectrophotometer in the range between 500 and 4000 cm−1 with a resolution of 4 cm−1. An average of 24 scans has been carried out for each sample. Particle size, particle size distribution, and zeta potential were measured by Malvern Instruments dynamic light scattering (DLS). The morphology of nanoparticles and nanocomposite hydrogels was observed by HITACHI S-4160 field emission scanning electron microscope (FE-SEM). The samples were coated with gold before FE-SEM characterization.

Results and discussion

Herein, we have fabricated crosslinked PDMAEMA/NCCs hydrogels with various contents of nanoparticles. Then, they have been used as adsorbents of MO dye from water. The schematic of fabrication process and adsorption behavior of MO by hydrogels are presented in Scheme 1.

Scheme 1
scheme 1

Fabrication route and MO dye absorption of nanocomposite hydrogels

Full size image

Synthesis and characterization of NCCs and nanocomposite hydrogels

FT-IR spectra of NCC, ANCC, and HNCC are shown in Fig. 1. Peak at 3450 cm-1 was due to hydroxyl groups participating in intracellular hydrogen bonding of cellulose (Haqani et al. 2017; Riazi et al. 2018) and peak at 2850–2950 cm-1 was ascribed to C–H stretching vibrations (Panahian et al. 2014a, b; Bidsorkhi et al. 2016). Peaks between 1310 and 1370 cm-1 originated from bending vibration of C–H and C–O groups in the polysaccharide rings of cellulose (Kargarzadeh et al. 2012) and peak at 1050–1110 cm-1 was ascribed to bending vibration of C–O–C. After modification, ANCC showed new peaks at 1564, 3290, and 3320 cm-1 due to stretching and bending vibrations of primary amines (Gao et al. 2009; Sharifzadeh et al. 2016). However, absorption peaks of Si–O–Si bonds were coincided on C–O bending modes of cellulose. HNCC showed symmetric and asymmetric vibrations of CH2 groups at 2925 and 2854 cm-1 (Sarsabili et al. 2013; Luo et al. 2017). It is reported that size of nanoparticles could be affected after modification depending on both modifier length and interactions between modified nanoparticles (Banaei and Salami-Kalajahi 2015; Mahmoud et al. 2020). For example, change in surface plasmon resonance of gold nanoparticles results in increasing the particles size (Mazloomi-Rezvani et al. 2018) whereas there are some reports on reduction of size of different particles through modification (Pyatenko et al. 2009; Mahmoud et al. 2020). Thus, unmodified NCC and modified NCCs were analyzed by DLS at room temperature (Fig. 1). Z average particle size was reported 304.7 (polydispersity index (PDI) = 0.070), 45.7 (PDI = 0.01), and 121 nm (PDI = 0.007) for NCC, ANCC, and HNCC respectively. Modified nanoparticles showed smaller size than NCC because the breakdown of hydrogen bonding in NCC structure which prevents the agglomeration of NCCs. Larger particle size of HNCC than ANCC originated from hydrophobic interactions of HNCC as driving force of aggregation. Also, all samples showed narrow size distribution. Zeta potential of NCC, ANCC, and HNCC was obtained − 19.6, + 2.2, and − 2.1 mV respectively. Replacing hydroxyl groups with alkyl chain of HDTMS in HNCC resulted in smaller zeta potential whereas ANCC had a positive zeta potential originated from amines of APTES (Yamada et al. 2006).

Fig. 1
figure 1

FTIR, DLS, and zeta potential results of NCC, ANCC, and HNCC nanoparticles

Full size image

FE-SEM images of HNC, HAC, and HHC hydrogels are shown in Fig. 2. This analysis was carried out on dried hydrogels due to difficulty of FE-SEM method in presence of water in the structure of hydrogel nanocomposites. According to results, all NCCs were aggregated in PDMAEMA matrix. The aggregation of nanoparticles might be related to a nanoparticles content, which led to interaction of neighbor nanoparticles in nanocomposite hydrogels (Ma et al. 2017). Meanwhile, the surface of nanocomposite hydrogels was rough because water diffused to structure of hydrogels and resulted in uneven surface morphology (Luo et al. 2018). Moreover, micrographs clearly illustrated that hydrogels morphology was dependent on the nanoparticles content and increasing NCC led to filling of structure and resulted in lower water penetration to structure (Bashir et al. 2017). However, HHC nanocomposite hydrogels had the most aggregated structure due to hydrophobic nature of HNCC surface, and dispersive forces were the most important driving force to form big agglomerations.

Fig. 2
figure 2

FE-SEM images of PDMAEMA/(modified) NCC nanocomposite hydrogels

Full size image

Swelling behavior of PDMAEMA/(modified) NCC nanocomposite hydrogels

Ability of swelling is one of the most important parameters to evaluate the properties of hydrogels. Higher swelling ratio means the ability of water absorption onto structure of a hydrogel. Herein, to study the effect of surface treatment of NCC on swelling ratio, different nanoparticles in neat or modified state are incorporated into the structure of PDMAEMA hydrogels. Figure 3 shows the swelling behavior of hydrogel nanocomposites containing 2 wt. % of (modified) NCCs. Generally, PDMAEMA-based hydrogels had very hydrophilic characteristics with high water adsorption capacity in acidic media. However, swelling ratio decreased continuously as pH increased. This was related to the protonation degree of dimethylamino (DMA) groups in structure of PDMAEMA. At acidic media, DMA groups were protonated which resulted in electrostatic repulsions between them (Dehghani et al. 2020). This caused swelling of hydrogel structure and diffusion of water molecules into the structure. However, by increasing pH value, DMA groups turned to half-protonated and deprotonated states which resulted in deswelling of hydrogels and decreasing water absorption (Noein et al. 2017). In hydrogel nanocomposites, HHC2 was the least swollen one because of HNCC’s hydrophobic nature that led to less water absorption onto structure of hydrogels. Except pH = 10, HAC2 was the highest swollen sample because of its hydrophilic nature due to presence of primary amines on the surface. However, at pH = 10, HNC2 had a higher swelling ratio than HAC2. This might be due to deprotonation of amine groups on the surface of ANCC and electrostatic repulsions between negative charges on the surface of NCC which led to expansion of nanocomposite hydrogels and easy diffusion of water into structure of hydrogel.

Fig. 3
figure 3

SR of HNC2, HAC2, and HHC2 hydrogels at different pHs

Full size image

Figure 4 represents the effect of content of NCCs on SR at two pH values (1 and 10). At pH = 1, PDMAEMA had strong hydrophilic characteristics, and NCC and ANCC as hydrophilic nanoparticles had no positive effect on nature of hydrogel hydrophilicity. Besides, they limited the expansion of 3D network by exertion of physical confinement on swelling of and filling hydrogel structure. As a result, swelling ratio was decreased by adding NCC and ANCC nanoparticles. In HHC hydrogels, adding HNCCs led to dramatic decrease of SR due to hydrophobic structure of HNCC and structure filling of hydrogel with nanoparticles. This led to formation of more rigid structure and difficult water diffusion into network of hydrogels (Lim et al. 2017). At pH = 10, PDMAEMA electrostatic interactions were suppressed (Fallahi-Sambaran et al. 2018), and the effect of added nanoparticles became more predominant. Also, hydrogen bonding might exist among the free amino groups those limited the mobility of network chains and resulted in a dense hydrogel network; thus, a lower water absorption was exhibited. As a result, NCC and ANCC induced improvement in water absorption capacity of nanocomposite hydrogels due to their hydrophilic nature. Also, in the case of ANCC nanoparticle, free electron pair of the amino groups of ANCC nanoparticle could conjugate with carbonyl moieties in DMAEMA units. However, this interaction was weakened dramatically in acidic media due to protonation of tertiary amines (Hou et al. 2015). In HHC samples, even at pH = 10, higher amount of HNCC resulted in lower swelling ratio due to its hydrophobic surface nature that limited absorption of water molecules into hydrogel 3D structure.

Fig. 4
figure 4

Swelling ratio versus time for HNC, HAC, and HHC hydrogel nanocomposites at different pH values

Full size image

Removal of MO by hydrogel nanocomposites

Dye removal by hydrogels is influenced by ionization degree of adsorptive molecule, charge of adsorbent, and dissociation of groups on functions of adsorbent (Tu et al. 2017). Conclusively, pH of dye solution is one of the most important parameters in adsorption process. At low pH, DMA groups of PDMAEMA are become ionized where electrostatic repulsions between positive DMA groups results in expansion of hydrogel. Besides, electrostatic interactions between positive DMA groups and negatively charged MO dye are become dominated. Thus, we chose these conditions according to pKa value of PDMAEMA (~ 8) where at pH = 1 all DMA moieties are protonated and positively charged (Mohammadi et al. 2017). However, at pH = 10 DMA groups become deprotonated and electrostatic interactions are disappeared. Thus, for more comparison, we reported the highest and lowest adsorption capacity. Also, according to swelling ratios, significant difference between swelling behaviors of hydrogels was observed between pH values of 1 and 10. Moreover, higher pH values were not tested because of hydrolysis of ester bonds in the structure of (meth)acrylates (Panahian et al. 2014b). Figure 5 represent the effect of pH of dye solution on adsorption capacity of hydrogels containing 2 wt. % NCCs where adsorption capacity of all samples was decreased at higher pH values. This originated from electrostatic interactions between deprotonated hydrogel structure at pHs higher than pKa of PDMAEMA and MO. Beside this, deprotonation of DMA moieties at pH values higher than pKa of DMAEMA resulted in shrinkage of hydrogel structure. Conclusively, MO molecules diffused more difficult into structure of hydrogel and decreased dye adsorption capacity was observed (Zhao et al. 2012). Modification of NCC also effected adsorption capacity and although HAC2 showed the highest adsorption capacity, no significant variation of adsorption capacity was obtained for different samples at pH = 10. This originated from low SR of hydrogels and difficult diffusion of MO into three-dimensional network. Also, at pH = 1, HAC2 and HHC2 nanocomposite showed the highest and the lowest adsorption capacity respectively. This could be explained by swollen structure of hydrogel nanocomposites as discussed in Figs. 3 and 4. Also, at pH = 1, ANCC had positive surface charge that could interact with anionic MO dye whereas HNCC and NCC had negative surface charge which discouraged dye adsorption into structure of hydrogel nanocomposites.

Fig. 5
figure 5

Adsorption of MO for HNC2, HAC2, and HHC2 hydrogels vs. time at different pH of dye solution

Full size image

Figure 6 shows that introduction of (modified) NCC and increasing its content resulted in decreasing the adsorption capacity for all sample in acidic and alkali conditions. This originated from filling of structure and difficulty of MO diffusion from solution into hydrogel. Also, by increasing pH from 1 to 10, adsorption capacity decreased from 1.64 to 0.95 mg/g for neat PDMAEMA hydrogel. This might be ascribed to the lower swelling ratio and also deprotonation of DMA groups at pH = 10 which results in less electrostatic interactions between PDMAEMA and MO as anionic dye. This coincides with the results of Salama et al. (2015) where by increasing pH from 2 to 6, the adsorption capacity was decreased from 193 to 135 mg/g due to more ionized groups at lower pH and electrostatic repulsions with adjacent groups. Also, incorporation of ANCC resulted in the least negative effect on adsorption capacity. ANCC had primary amine groups on its surface those could help adsorption of anionic MO dye; however, it exerted limitations on swelling of hydrogel during adsorption process. Lin et al. (2016) prepared a microsphere adsorbent based on CMC and DMDAAC for adsorption of anionic dyes where by increasing the amount of cationic monomer of DMDAAC, the adsorption capacity of microspheres was increased. NCC and HNCC affected adsorption capacity significantly because their surface charge was negative and electrostatic repulsions between them and dye resulted in lower adsorption of dye beside exertion of limitations on swelling of hydrogel. In this field, Liu et al. (2015a, b) synthesized the cellulose-based bioadsorbent that showed negative zeta potentials. Thus, the adsorption capacity was decreased due to augmentation electrostatic repulsions that coincides our result.

Fig. 6
figure 6

Adsorption of MO by HNC, HAC, and HHC hydrogels versus time at different conditions

Full size image

To investigate the adsorption kinetics of MO through different hydrogel nanocomposites, we used well-known mathematical adsorption models namely pseudo-first-order model (Eq. (3)), pseudo-second-order model (Eq. (4)), and intraparticle diffusion model (Eq. (5)) (Chen and Wang 2009; Melo et al. 2018; Bulut and Karaer 2015).

$$ \ln \left({q}_e-{q}_t\right)=\ln {q}_e-{k}_1t $$
(3)
$$ \frac{t}{q_t}=\frac{1}{q_e^2{k}_{2\kern0.5em }}+\frac{t}{q_e} $$
(4)
$$ {q}_t={k}_{id}{t}^{\frac{1}{2}}+c $$
(5)

where qe and qt are the extent of adsorbed dye on hydrogel at equilibrium state and any time respectively, and k1, k2, and kid are constants. Calculated k1, k2, and kid and also regression correlation coefficient (R2) values are shown in Table 1. From the kinetic data, it could be concluded that pseudo-first-order kinetics model was the best model among data fitting at both pH = 1 and 10.

Table 1 Adsorption kinetic parameters of different models (k1 (h−1), k2 (g mg−1 h−1), and kid (mg g−1 h−1/2) for MO adsorption into hydrogel nanocomposites
Full size table

Conclusions

Herein, we prepared hydrogel nanocomposites with different modified and amount of NCCs to investigate the effect of surface chemistry of different modified NCCs on dye adsorption. Then, swelling and MO adsorption behaviors of hydrogel nanocomposites were studied at different pH values. HAC nanocomposites were the most swollen ones whereas HHC hydrogels had the least swelling ratio at pH values lower than pKa of DMAEMA. However, HNC samples showed the highest SR at pH = 10 regarding filling of hydrogel structure and surface charge of nanoparticles. Moreover, lower pH resulted in protonation of DMA moieties of PDMAEMA, and their electrostatic repulsions caused higher SR of hydrogels. MO adsorption of hydrogel nanocomposites clarified that MO removal was influenced by the surface chemistry of NCCs, pH value of dye solution, and content of NCCs. As a main result, higher pH value resulted in lower adsorption capacities because DMAEMA DMA groups stayed on deprotonated form and could not do interaction with MO dye. However, HHC nanocomposite hydrogels had the lowest adsorption capacity due to lack of interacting functional groups on surface of HNCC and its negative surface charge. Also, kinetics studies showed that pseudo-first-order kinetics was the best model for predicting MO adsorption through all types of hydrogel nanocomposites.