Nitrogen-doped graphene hydrogels as potential adsorbents and photocatalysts for environmental remediation

Accepted Manuscript Nitrogen-doped graphene hydrogels as potential adsorbents and photocatalysts for environmental remediation Yiqun Jiang, Shamik Chowdhury, Rajasekhar Balasubramanian PII: DOI: Reference:

S1385-8947(17)31108-7 http://dx.doi.org/10.1016/j.cej.2017.06.156 CEJ 17237

To appear in:

Chemical Engineering Journal

Received Date: Revised Date: Accepted Date:

24 April 2017 25 June 2017 26 June 2017

Please cite this article as: Y. Jiang, S. Chowdhury, R. Balasubramanian, Nitrogen-doped graphene hydrogels as potential adsorbents and photocatalysts for environmental remediation, Chemical Engineering Journal (2017), doi: http://dx.doi.org/10.1016/j.cej.2017.06.156

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Nitrogen-doped graphene hydrogels as potential adsorbents and photocatalysts for environmental remediation †

‡

Yiqun Jiang, Shamik Chowdhury, Rajasekhar Balasubramanian †

†*

Department of Civil & Environmental Engineering, National University of Singapore,

1 Engineering Drive 2, Singapore 117576, Republic of Singapore ‡

Centre for Advanced 2D Materials, National University of Singapore,

6 Science Drive 2, Singapore 117546, Republic of Singapore

____________________________ * Corresponding author. Tel.: +65 65165135; fax: +65 67744202. E-mail address: [email protected] (R. Balasubramanian).

ABSTRACT The chemical modification of self-assembled graphene hydrogels is a topic of emerging interest to harness the excellent physicochemical properties of two-dimensional (2D) graphene for macroscopic applications. We synthesized a series of mechanically strong and lightweight nitrogen (N)-doped graphene hydrogels (NGHs), with different doping concentrations, through a simple one-pot hydrothermal reaction and systematically evaluated their performance as both adsorbents and photocatalysts for environmental remediation. Acridine orange (AO) was chosen as a model pollutant. The successful incorporation of N atoms into the carbon lattice of the macroscale 3D graphene-based materials was verified by Raman spectroscopy, Fourier transform infrared spectroscopy (FTIR) and X-ray photoelectron spectroscopy (XPS).

Although the N content of the

graphene macroassemblies varied inversely with doping density, a conspicuous increase in specific surface area was observed at all doping levels, resulting in a higher adsorption capacity and surface reactivity than undoped hydrogels. The adsorption equilibrium was best represented by the Langmuir isotherm (with maximum monolayer coverage of 124 mg g−1 at 25 °C) while the adsorption kinetics followed both the pseudo-first and pseudo-second order rate expressions. Further, the NGHs could effectively photodegrade 20 mg L−1 AO solution by almost 70% within 5 h of visible light irradiation. The strong photooxidative ability of the NGHs originates from the synergistic effect of N functionalization and 3D interconnected mesoporous network structure, leading to greater uptake of AO, better absorption of visible light and rapid spatial separation of photogenerated electron–hole pairs.

Keywords: Graphene hydrogels, N-doping, Acridine orange, Adsorption, Photocatalysis, Visible light.

1. Introduction Two-dimensional (2D) graphene, by virtue of its many interesting properties, has proved itself as an exciting material with a wide range of opportunities for new scientific knowledge generation and technological innovations [1‒3]. However, 2D graphene tends to form irreversible agglomerates or even restack to form graphite, causing a significant decrease in the ultrahigh surface area of individual graphene sheets [4]. In order to overcome this restacking issue and explore the macroscopic applications of graphene, the integration of 2D graphene macromolecule sheets into 3D macroscopic assemblies and ultimately into a functional system has recently emerged as an increasingly important approach [5‒7]. Consequently, many exotic 3D graphene-based macrostructures (GBMs), such as aerogels, hydrogels, sponges, foams, frameworks with periodic structures, honeycomb-like structures, porous films, vertical sheets, etc., have been intensively developed during the past few years [8,9]. These rationally designed macroscopic architectures exhibit low mass density, large accessible surface area, abundant nanopores, continuously interconnected networks and channels, excellent electrical conductivity, high electrochemical stability and superior mechanical flexibility [8‒10].

As a result, 3D GBMs can (1) support rapid electron

transport in 3D space, (2) ensure faster ion diffusion, (3) provide adequate space for molecular adsorption and (4) promote efficient mass transfer [9].

In addition, 3D GBMs have large internal pores and can therefore serve as ideal scaffolds for functionalization with heteroatoms, functional polymers, inorganic nanostructures, as well as many other carbon nanomaterials, leading to new material systems with unique properties and novel functionalities [8, 11‒13]. In particular, doping with heteroatoms such as nitrogen (N) offers many novel possibilities for tailoring the structural and electronic properties of graphene [14]. The incorporation of N with higher electronegativity than

carbon (C) creates polarization in the sp2 hybridized network, which in turn opens a band gap close to the Dirac point by suppressing the nearby density of states and bestows graphene with semiconducting properties [15]. Moreover, the manipulated local electronic structures greatly enhance the binding affinity of the carbon matrix [15]. In view of these considerations, N-doping of 3D GBMs represents an exciting and promising new research direction for development of next-generation of robust adsorbents and photocatalysts for environmental remediation.

However, the macroscopic performance of N-doped graphene materials (NGMs) depends mainly on their N content and the type of N bonding configuration within the carbon lattice, including pyridinic-N, pyrrolic-N and quaternary-N (or graphitic-N) [16]. Among these N species, pyridinic-N and quaternary-N are sp2 hybridized whereas pyrrolic-N is sp3 hybridized [16]. Consequently, the relative proportion of these three N bonding moieties play a key role in achieving the maximum functionality of NGMs, but were seldom investigated in previous studies. Modifying the structure and surface properties of graphene satisfactorily and synchronously through the specific type of N dopant may provide new insights to fabricate 3D GBMs with pre-programmed conformations and custom-made properties. However, to the best of our knowledge, no systematic experimental study has been conducted yet on the modulation of the physicochemical properties of 3D GBMs for environmental remediation applications by controlling the N-doping concentration and configuration.

Herein, we therefore prepared mechanically strong and lightweight N-doped graphene hydrogels (NGHs) with different doping density through a facile one-pot hydrothermal process using ammonium hydroxide as N precursor and rigorously evaluated their potential

application as both an adsorbent and a photocatalyst for removal of organic pollutants from aqueous media. Acridine orange (AO) was selected as the model contaminant since it is frequently reported to be present as one of the major recalcitrant organic pollutants in urban wastewaters and is also a highly potent mutagen [17,18]. For instance, the dye can easily intercalate between the nitrogenous base pairs of deoxyribonucleic acid (DNA) because of its planar polycyclic molecular structure (Fig. 1), resulting in elongation and unwinding of the double helix [19‒21]. These structural modifications can induce functional changes, often leading to inhibition of DNA transcription, replication and repair [20]. Benefitting from the synergistic effect of N-doping and highly conductive 3D interconnected porous graphene skeleton, the as-fabricated NGH materials could afford high surface areas, short ion diffusion lengths and rapid electron/ion transport channels, thereby demonstrating excellent adsorption capacity and superior photocatalytic activity under visible light irradiation toward AO. In particular, the maximum adsorption uptake as well as the highest photodecolorization rate was observed for the hydrogel with a N loading of 6.78 at.%. Overall, our results represent one of the first studies to demonstrate that tuning the various physicochemical characteristics of heteroatom-doped graphene macroassemblies, such as the doping amount and the type of dopant bonding in the carbon matrix, is a prerequisite to realize their full potential as adsorbents/photocatalysts for environment-related applications.

2. Experimental 2.1. Materials Graphene oxide (GO, 4 mg mL−1 dispersion in water; Sigma-Aldrich), ammonium hydroxide (NH4OH, 28‒30% ammonia basis; Sigma-Aldrich), acridine orange (AO, C.I. 46005; Sigma-Aldrich), ethanol (C2H5OH, absolute, ≥99.8%; Sigma-Aldrich), hydrochloric acid (HCl, fuming, ≥37%; Fluka) nitric acid (HNO3, 69%; Honeywell) and sodium hydroxide

(NaOH, pellets, ≥98%; Sigma-Aldrich) were used as obtained from the supplier without any further purification. Deionized water was used throughout the experiments.

2.2. Synthesis of NGHs NGHs were synthesized through a facile one-pot hydrothermal method.

In a typical

procedure, a desired amount of NH4OH was added into 40 mL GO aqueous dispersion and stirred for 10 min at room temperature. The resulting homogenous mixture was sealed in a 100 mL Teflon-lined stainless steel autoclave and hydrothermally treated at 180 °C for 12 h. The autoclave was then naturally cooled to ambient temperature.

The as-obtained NGH

was washed repeatedly with water to remove any residual chemical and finally freeze-dried for 24 h. By altering the volume of NH4OH in the reaction mixture (VGO/VNH4OH = 2.5, 5, 7.5, 10), four different NGHs with varying doping levels were developed, referred to as “NGH-1”, “NGH-2”, “NGH-3”, and “NGH-4”, respectively.

For comparison, pristine

graphene hydrogel (denoted as “GH”) was also prepared using the same procedure without adding any NH4OH.

2.3. Instrumental characterization Field emission scanning electron microscopy (FESEM) imaging was conducted on a FEI Verios 460 (FEI Ltd., USA) field emission microscope. Wide angle X-ray diffraction (XRD) patterns were collected on a Bruker D8 ADVANCE (Bruker Co., Germany) X-ray diffractometer equipped with Ni-filtered Cu Kα radiation source (λ = 0.15 nm). X-ray photoelectron spectroscopy (XPS) measurements were performed through a VG ESCA 220i-XL imaging system (Thermo VG Scientific Ltd., UK). Monochromatic Al Kα X-ray (hν = 1486 eV) was employed for analysis with a photoelectron take-off angle of 90o to the surface plane.

Fourier transform infrared (FTIR) spectra were recorded on a Varian

Excalibur 3100 FTIR spectrometer (Varian Inc., USA) by the KBr pellet method. Raman spectra were acquired at room temperature using an Alpha300 R confocal Raman imaging microscope (WITec GmbH, Germany) with laser excitation at 532 nm.

The textural

characteristics were quantified by measuring the N2 adsorption/desorption isotherms at ‒196 °C in a Micromeritics ASAP 2020 surface area and porosity analyzer (Micromeritics Instrument Co., USA). All samples were outgassed at 100 °C under vacuum for 12 h prior to the N2 adsorption measurements.

The specific surface area (Ssp) was determined

employing the Brunauer–Emmett–Teller (BET) model to the N2 adsorption data in the relative pressure (P/P0) range of 0.05–0.20 while the pore size distribution (PSD) analyses were performed by applying the Barrett–Joyner–Halenda (BJH) formalism to the desorption branch of the isotherms. The total pore volume (Vtotal) was estimated from the amount of N2 adsorbed at P/P0 = 0.99. The micropore volume (Vmic) was evaluated by the Lippens and Boer t-plot method.

2.4. Preparation of dye solutions Dye stock solution (100 mg L−1) was prepared by dissolving 100 mg of AO in 1 L distilled water. An experimental dye solution of the desired concentration was prepared by further dilution of the stock solution with suitable volume of distilled water. The initial solution pH was adjusted with 0.1 (M) HCl and 0.1 (M) NaOH solutions using a digital pH meter (pH 700, Eutech Instruments Pte. Ltd., Singapore) calibrated with standard buffer solutions.

2.5. Batch adsorption experiments A series of batch adsorption tests were conducted to evaluate the effect of different process parameters, such as pH (3–9, step size: 1) and temperature (20–30 °C, step size: 5 °C), on the uptake of AO by NGHs. In each experiment, 30 mL dye solution (5 mg L−1) was taken

in a 250 mL glass-stoppered Erlenmeyer flask and a weighed amount of the as-prepared NGH (6 mg) was added to the solution. The flask was then agitated at a constant speed (200 rpm) and temperature in an incubator shaker (LM-575RD, Yihder Technology Co. Ltd., Taiwan). At the end of the adsorption period, the free-standing NGH was directly removed from the solution and the residual liquid-phase dye concentration was measured using a Hitachi U2800 UV/Vis spectrophotometer (Hitachi Co. Ltd., Japan) at the maximum absorbance wavelength (λmax) of 490 nm.

The kinetics of the adsorption process was

determined by analyzing the adsorptive uptake of the dye from the aqueous solution at predetermined time intervals.

Finally, isotherm studies were carried out at room

temperature (25 °C) by agitating 30 mL dye solution of different concentrations (5, 10, 15, 20, and 25 mg L−1) with 6 mg of NGH for 24 h.

2.6. Photocatalytic measurements The photocatalytic efficiency of the as-synthesized NGH materials was examined for degradation of AO at room temperature (25 °C) in an indigenously built in-house apparatus under visible light irradiation. Specifically, the effect of initial dye concentration (5–25 mg L−1, step size: 5 mg L−1) on the photodegradation rate of AO was investigated. In a typical measurement, 6 mg of NGH was suspended into 18 mL dye solution. The suspension was then magnetically stirred without visible light exposure for 1 h to attain the adsorption/desorption equilibration. Following this, the reaction mixture was illuminated using four 11 W Xe lamps fitted with UV cut-off filter (λ < 420 nm). During irradiation, aliquots were withdrawn at pre-fixed time intervals and the residual AO concentration was quantified using a Hitachi U2800 UV/Vis spectrophotometer (Hitachi Co. Ltd., Japan) at λmax = 490 nm.

2.7. Theory All calculations (viz., adsorption capacity, percentage dye removal, photodegradation efficiency) and details of kinetics, isotherm and thermodynamic modeling are described in the Supplementary information.

3. Results and discussion 3.1. Materials characterization Scheme 1 illustrates the synthesis of NGHs via a facile one-step hydrothermal reaction using GO and NH4OH as C and N sources, respectively. By controlling the doping level, the size of the graphene hydrogels can be easily adjusted (Fig. 2a). In addition, the as-prepared NGHs are extremely lightweight and can support at least 4000 times their own weight, as demonstrated in Fig. 2b and c, respectively. FESEM imaging reveals that the NGHs are formed by partial overlapping or coalescing of flexible graphene sheets (Fig. 3), through either hydrophobic or π–π stacking interactions in 3D [22]. Regardless of the doping concentration, each of the NGHs possesses a well-defined interconnected porous network structure with pore sizes ranging from submicrometers to several micrometers. Although pristine GO suspension can also be assembled into 3D interconnected macroscopic framework upon hydrothermal reduction (Fig. 3a), graphene sheets appear to form relatively strong association amongst themselves upon introduction of N-related defects, yielding more densely cross-linked channels (Fig. 3b–e). Consequently, the BET specific surface area of the NGHs (up to 151 m2 g−1 for NGH-2) is much higher than that of the undoped hydrogel (75 m2 g−1) (Table 1). An excess amount of NH4OH in the reaction mixture, however, dramatically reduces the surface area by about 17.88% for NGH-1. In contrast, no specific trend is observed for the evolution of pore volume with N doping (see Table 1) and is consistent with the findings of previous studies [23,24]. The pore size distribution obtained

by the BJH model shows that much of the pore space (0.25 cm3 g−1) lies in the 2–55 nm range (Fig. 4), with average pore diameter less than those of GH (Table 1).

Wide-angle XRD patterns of the bulk hydrogels present two prominent diffraction peaks centred at around 2θ of 24o and 43o (Fig. 5a), conforming to the graphitic (0 0 2) and (1 0 1) crystal planes [25]. According to Bragg’s equation [26], the interlayer spacing (d) between graphene sheets in the freeze-dried hydrogels is 0.366 (2θ = 24.3o), 0.337 (2θ = 25.7o), 0.341 (2θ = 25.4o), 0.344 (2θ = 25.2o) and 0.348 nm (2θ = 24.9o) for GH, NGH-1, NGH-2, NGH-3 and NGH-4, respectively. These d values are much lower than that of the GO precursor (d = 0.818 nm) but slightly higher than that of natural graphite (d = 0.335 nm) [27]. The aforementioned results indicate the presence of π–π stacking interaction between graphene sheets in the NGHs owing to considerable recovery of the graphitic crystalline structure, since the two common N bonding configurations, i.e., quaternary-N and pyridinicN, are sp2 hybridized within the carbon lattice [16,28]. In addition, the broadness of the peaks and their weak intensity imply that the freeze-dried NGHs are made of highly disordered few graphene layers that are randomly placed on top of each other, which could be due to incomplete deoxygenation of the reduced GO sheets in the hydrogels [25]. These residual oxygen functionalities can entrap ample water molecules in the carbon matrix during the self-assembly of the 2D nanoscale building blocks in the hydrothermal environment, which together with the π–π stacking of graphene layers results in the successful formation of the hydrogels [22].

Moreover, as the amount of N precursor

increases, the sp2 conjugated domains of the reduced GO sheets also increases, which in turn enhances the strength of interaction and degree of cross-links among the graphene macromolecules, leading to the observed decrease in the surface area of the hydrogels. This is also evident from the smallest d value of NGH-1 among all the monoliths.

The microstructure of the as-prepared NGHs was further probed by Raman spectroscopy, as depicted in Fig. 5b. Evidently, all the samples exhibit two distinct bands at about 1350 and 1595 cm−1 which correspond to the well-documented D (disorder or defect) and G (graphitic carbon) bands, respectively [29,30]. Moreover, the Raman intensity ratio of the D band to G band (ID/IG) for the NGHs appear to be lower than that of GH due to the recovery of the sp2 conjugated system to a certain extent, which is in accordance with the XRD measurements. In addition, the defect levels of NGH-1 (ID/IG = 0.99), NGH-2 (ID/IG = 0.97) and NGH-3 (ID/IG = 1.05) are essentially similar. This could be due to the different types of N-induced defects (i.e., in-plane substitutions, vacancies, or grain boundaries/edges) as well as their positions and arrangements in the sp2 hybridized carbon backbone of the materials. Nevertheless, a higher ID/IG ratio of NGH-4 (1.13) compared to GH (1.10) may be attributed to the greater substitution of C atoms by N atoms within the honeycomb lattice. This finding is also ascertained by the appearance of a very sharp bandwidth at around 3520– 3320 cm−1 (corresponding to N─H stretching vibration) in the FTIR spectrum of NGH-4 (Fig. 5c).

In order to get a more in-depth understanding of the N doping effect in the 3D graphene macroassemblies, the stoichiometry of the as-synthesized hydrogels was determined using XPS. As expected, the wide survey scan of GH produces the characteristic C1s and O1s peaks at binding energy (B.E.) of about 284 and 532 eV, respectively (Supplementary Fig. S1). The appearance of a new peak at approximately 400 eV for NGHs suggests the successful incorporation of N atoms into the graphene basal plane during the hydrothermal process (Supplementary Fig. S1). Further, the N1s spectra can be deconvoluted into three major components, which correspond to the different bonding states of N in the 2D graphene building blocks (Fig. 6a): pyridinic-N (B.E. 398.1–399.3 eV), pyrrolic-N (B.E. 399.8–401.2

eV), and quaternary- or graphitic-N (B.E. 401.1–402.7 eV) [16].

A schematic

demonstrating the various N bonding configurations is given in Fig. S2 (see Supplementary information). Similarly, the different chemical environment of the C atoms within the NGHs was identified through deconvolution of the high-resolution C1s XPS scan (Supplementary Fig. S3). The major peak situated at 284.5 eV is attributed to the presence of graphite-like sp2 C–C bonds [31].

The two additional components centered at 285.2–

285.7 eV and 286.2–286.9 eV can be assigned to the N–sp2–C and N–sp3–C bonds, respectively [31,32]. Additionally, the total N content (atomic percentage) of the NGHs was estimated from the ratio of the integrated peak areas of the N1s and C1s core levels, and was found to be ~5.54%, ~6.78%, ~6.98% and ~7.56% for NGH-1, NGH-2, NGH-3 and NGH-4, respectively. The inverse relationship between the amount of N precursor and the number of N atoms embedded in the graphene lattice can be attributed to the high concentration of ammonia that inhibits the formation of graphitic-N bonds [33]. Moreover, the high density of N radical in the reaction mixture may also cause instability of the chemisorbed N on C [34], resulting in the lowest N doping of NGH-1 in our experiment. These explanations are confirmed by enumerating the percentage concentration of each bonding state per N content. As summarized in Fig. 6b, NGH-1 has a relatively low amount of graphitic-N (~1.31 at.%) in comparison with the other samples: NGH-2 (~2.08 at.%), NGH-3 (~1.88 at.%), NGH-4 (~1.82 at.%). In addition, the presence of only sp2 hybridized N atoms in its graphene core further clarifies the narrow interlayer distance and smallest surface area of NGH-1 among the synthesized N-doped hydrogels.

3.2. Adsorption of AO Preliminary adsorption experiments were conducted employing a batch experimental setup to assess the suitability of the as-prepared NGHs as adsorbents for removal of AO from

aqueous environment. In comparison to GH, the dye uptake potential of the NGHs is significantly higher (Fig. 7a). These results clearly show that incorporation of N atoms in the 2D nanoscale building blocks of 3D graphene macroassemblies has a positive influence on the adsorption capacity by improving the surface area and enhancing the adsorption strength. Meanwhile, NGH-2 displays the highest dye adsorption efficiency (up to 97% at 5 mg L−1 and 20 °C), and exhibits an excellent adsorption performance even at high concentration levels (as much as 100.82 mg g−1 at around 25 mg L−1) (Fig. 7b). The impressive performance of NGH-2 is most likely due to its relatively large specific surface coupled with greater density of electrochemically active N species (i.e., pyridinic-N (~1.31 at.%) and pyrrolic-N (~3.39 at.%)) in its interconnected graphene networks [35]. Consequently, NGH-2 was chosen as the best material for adsorption of AO, and was more rigorously evaluated for its practical applications, as discussed in the following sections.

3.2.1. Effect of pH Solution pH is perhaps the most important parameter governing an adsorption process, since it not only determines the surface charge of the adsorbent but also the degree of ionization and speciation of the adsorbate [36]. The effect of pH on the adsorption of AO by NGH-2 is presented in Fig. 8. The dye removal capacity increases sharply from 54.62% at pH 3 to 98.31% at pH 5, which can be attributed to electrostatic forces of attraction between the positively charged N moieties of AO and the electronegative N atoms in the graphene matrix of NGH-2. However, at the test pH of 6, the dye adsorption is significantly hindered (only 68.69% removal), which could possibly be due to protonation of the pyridinic-N atoms [37], resulting in repulsive electrostatic interaction. Conversely, further increase in pH facilitates deprotonation of the acidic functional groups (i.e., the residual oxygen functionalities and the protonated pyridinic-N) on the surface of NGH-2 [37], thus

promoting adsorption of AO to almost 94.53% at pH 9. Based on these observations, it is reasonable to infer that the unique surface chemistry of N-doped graphene monoliths can lead to different adsorption properties compared with those of undoped hydrogels.

3.2.2. Effect of temperature Temperature is another key variable influencing the adsorption process. Fig. 9 depicts the dye adsorption profile of NGH-2 as a function of contact time at three different temperatures.

Clearly, the dye removal capacity increases with rising temperature,

indicating that the uptake of AO on NGH-2 is kinetically controlled by an endothermic process. As temperature increases, the mobility of the AO molecules increases whereas the retarding forces acting on the dye decreases, which in turn enhances their uptake. Additionally, the rapid rate of adsorption during the first 60 min of the dye–adsorbent contact is invariably due to greater availability of free binding sites on the adsorbent surface. The rate of adsorption, however, soon levels off, eventually reaching equilibrium because of complete saturation of the adsorbent. Particularly, after 5 h, the amount of dye adsorbed does not present any relative change over time.

The fairly rapid establishment of

equilibrium suggests the application potential of NGH-2 in industrial wastewater treatment units, where large volumes of effluent will have to be treated while operating with short adsorption cycle times.

3.2.3. Adsorption kinetics A quantitative evaluation of the dye uptake kinetics of NGH-2 was subsequently carried out by directly fitting the experimental kinetic data to the pseudo-first order (Supplementary Eq. (S5)) and pseudo-second order rate equations (Supplementary Eq. (S6)). The model constants obtained by nonlinear regression using the “Nonlinear Curve Fit” function of

Origin Pro 8.0 (OriginLab, Northampton, MA) along with the error function values are compiled in Table 2. The exceptionally high R2 values and relatively low χ2 values confirm that the pseudo-second order model can reasonably define the adsorption of AO on NGH-2 at low temperatures (Fig. 9b). However, as the temperatures rises, the adsorption becomes diffusion controlled and the process follows the pseudo-first order rate expression (Fig. 9a). Additionally, k1 increases with rising temperature (Table 2), implying that diffusion through the pore networks is faster at higher temperatures.

Consequently, the possibility of intraparticle diffusion resistance controlling the adsorption of AO on NGH-2 was explored by analyzing the isothermal kinetic data with the WeberMorris model (Supplementary Eq. (S7)). The intraparticle diffusion plots are multilinear over the entire time range (Supplementary Fig. S4), attesting that both film diffusion (i.e., migration of dye from the bulk of the solution across the liquid film surrounding the adsorbent) and intraparticle diffusion (i.e., diffusion through pore fluids and/or along the pore walls) regulate the dye adsorption process [38,39].

3.2.4. Adsorption isotherms While isotherms are a thermodynamic basis of any adsorption process, equilibrium isotherm models can offer useful information about the nature of interactions between the adsorbate and the adsorbent, facilitating an in-depth understanding of the underlying adsorption mechanism. For that reason, adsorption isotherm data were collected at 25 °C and fitted to the classical Langmuir and Freundlich model equations (Supplementary Eqs. (S8) and (S9)). The fitting of the models to the experimental equilibrium data was accomplished by nonlinear regression (Supplementary Fig. S5). The calculated model parameters as well as the corresponding error estimates are listed in Table 2. Clearly, the Langmuir isotherm

provided a better fit to the dye adsorption data than the Freundlich model. The goodness-offit of the Langmuir model indicates that the adsorption of AO occurs at energetically equivalent, non-interacting sites on the surface of NGH-2, and that once a molecule occupies a site no further adsorption can take place at that site, thereby forming a monolayer. In addition, the maximum dye adsorption capacity of NGH-2 is comparable to, or even higher than many other solid adsorbents (Table 3), and hence merits further consideration for treatment of textile effluents.

3.2.5. Activation energy and isosteric heat of adsorption The activation energy (Ea, kJ mol−1) is an important parameter providing fundamental information on the adsorption mechanism. Low activation energies (5–40 kJ mol−1) are characteristic of physical adsorption, while high activation energies (40–800 kJ mol−1) represent chemical adsorption [48]. The Ea for adsorption of AO on NGH-2 was, therefore, computed according to the Arrhenius equation (Eq. (S10)). An Ea value of ~23 kJ mol−1, obtained from the slope of the linear plot of ln k2 vs. 1/T (Supplementary Fig. S6), suggests that the adsorption mode is predominantly of physical nature, and may be ascribed to van der Waals forces and electrostatic attractions between the electronegative adsorbent and the positively charged dye molecules. The low Ea value (< 25–30 kJ mol−1) also corroborates the fact that the adsorption phenomenon is intraparticle diffusion controlled [49]. This is because pore diffusion usually has weak temperature dependence [49,50].

Knowledge of the isosteric heat of adsorption (∆HX, kJ mol−1) is also essential when designing a practical adsorption system (including equipment and ancillary components). Hence, in the present study, the ∆HX of AO on NGH-2 was determined at a constant surface coverage (i.e., qe = 2, 4, 6, 8, and 10 mg g−1) using the Clausius–Clapeyron equation

(Supplementary Eqs. (S11) and (S12)). The values obtained are tabulated in Table 4, and plotted explicitly in Fig. S7 (see Supplementary information). As can be seen, the ∆HX increases linearly with surface loading, which is likely due to pronounced lateral interactions between neighboring dye molecules at higher loadings rather than adsorption on the adsorbent surface. This hypothesis is further verified by fitting the experimental equilibrium data at 25 °C to the Fowler–Guggenheim model (Eq. 1) [51]:

2W  C (1   )  ln  e   ln K FG    RT  

(1)

where θ is the fractional coverage (= qe/qm), KFG (L mg−1) is the Fowler-Guggenhiem equilibrium constant, W (kJ mol−1) is the interaction energy between adsorbed molecules, R is the universal gas constant (8.314 J mol−1 K−1), and T (K) is the temperature. If the interaction between the adsorbed molecules is attractive (i.e., W is positive), the heat of adsorption increases with surface loading due to increased interactions between the adsorbed molecules. Conversely, if the interaction among adsorbed molecules is repulsive (i.e., W is negative), the heat of adsorption decreases with surface loading.

When the adsorbed

molecules do not interact with each other, W tends to be zero [52,53].

The interaction

energy, deduced from the slope of the linear plot of ln [Ce(1–θ)/θ] vs. θ (Supplementary Fig. S8), turned out to be positive (W = 21.51 kJ mol−1), confirming the existence of attractive intermolecular forces between the adsorbed dye molecules.

Hence, with more AO

molecules present at higher surface loadings, multilayer adsorption takes place and adsorbate‒adsorbate interactions, which are less energetic, become more dominant. It, therefore, appears that the basic assumptions of the Langmuir isotherm are probably not fulfilled at high concentrations. In such cases, we should then focus on the predictive ability of the isotherm model rather than on the meaning of the model parameters [54].

3.2.6. Regeneration and reuse Regeneration of the spent adsorbent is critically important for the practical application of any solid adsorbent, since it governs the techno-economic feasibility of an adsorption process. Desorption experiments were, therefore, performed by agitating the dye-loaded hydrogel with 30 mL of eluent at 30 °C for 24 h. Three different eluting agents were tested, namely HCl (0.1 M), HNO3 (0.1 M) and ethanol (absolute). Near complete regeneration was obtained in all cases, confirming that the adsorption of AO on NGH-2 is reversible and of physical nature. More importantly, the regenerated material could be used repeatedly for at least three consecutive cycles without any significant loss in adsorption capacity (Supplementary Fig. S9), except when using HCl as an eluent which might be due to partial collapse of the 3D interconnected network structure, as observed through FESEM (Supplementary Fig. S10).

3.3. Photodegradation of AO The photocatalytic activity of the as-prepared NGHs toward AO was tested under visible light irradiation. After 5 h of exposure, the photocatalytic efficiency followed the order: NGH-2 > NGH-1 > NGH-4 > NGH-3 (Fig. 10a). Although there is no direct correlation between the total N content and the photodecolorization performance of the NGH materials, the maximum removal of approximately 70% is significantly higher than several other competing photocatalyst systems reported in the literature, as evident from Table 5. It can be explained that upon irradiation with visible light, the electrons (e‒) in the valence band (VB) of NGHs, composed of a filled band of π orbitals, O 2p orbitals, and N 2p orbitals, are excited to the empty π* orbitals of the conduction band (CB), thereby leaving holes (h+) in the VB. These photoinduced holes either can directly oxidize the surface adsorbed dye, or may be trapped by surrounding water molecules, producing hydroxyl radicals (•OH).

Meanwhile, the photogenerated electrons reduce the dissolved oxygen to generate superoxide radicals (•O2−). The •O2− then reacts with H+ to form H2O2, which in turn is rapidly decomposed to •OH. Ultimately, both •OH and •O2−, being very strong oxidizing agents, significantly promotes the oxidation of AO into CO2, H2O, mineral acids, etc., and thus efficiently enhance the overall photocatalytic efficiency. Based on this discussion, we propose a plausible mechanism (Scheme 2) to explain the degradation of AO by NGHs under visible light irradiation. The fundamental photophysical and photochemical processes involved during the heterogeneous photocatalytic decolorization may be represented by the following chain reactions:

Subsequently, the photocatalytic decomposition of AO was investigated quantitatively by fitting the experimental data to the Langmuir–Hinshelwood model [64].

R= 

kr KC dC  kr  dt 1  KC 

(9)

where, R is the reaction rate, kr is the reaction rate constant, θ is the surface coverage, K is the adsorption coefficient of the reactant, and C is the reactant concentration. Since C is in millimolar levels in the present study, the product KC is negligible with respect to unity

[65], and the above equation can be simplified to the following pseudo first-order rate expression [66]:

 Ct    kapp t  C0 

ln 

(10)

where C0 and Ct are the reactant concentrations at times t = 0 and t = t, respectively, and kapp (min−1) is the apparent reaction rate constant determined by plotting ‒ln(Ct/C0) against the reaction time (t). As can be seen in Fig. 10b, the excellent linearity between ‒ln(Ct/C0) and t corroborates that the photocatalytic reduction of AO can be reasonably approximated by the pseudo first-order rate model. The calculated kapp values are 3.33  10−3, 3.76  10−3, 3.18  10−3 and 3.35  10−3 min−1 for NGH-1, NGH-2, NGH-3 and NGH-4, respectively and are consistent with their photodecolorization efficiencies (Fig. 10a). These results demonstrate convincingly that the as-prepared NGHs hold great promise for visible light responsive photocatalytic decomposition of AO. In particular, NGH-2 displays the highest apparent reaction rate constant, which can be attributed to its largest surface area amongst all the four hydrogels examined. Specifically, the high specific surface area results in greater adsorption of the substrate to the catalyst surface, leading to faster photodegradation. In addition, the large specific surface area provides more active sites, which in turn enable greater absorption of visible light energy and substantially increases the catalytic efficiency. Moreover, the 3D interconnected mesoporous structure (i) allows multiple reflection and scattering of incident light [67], thereby maximizing the light harvesting capability, (ii) increases the population of photogenerated electrons and holes participating in the photocatalytic reaction by effectively promoting the separation of photogenerated electron– hole pairs [68], and (iii) facilitates rapid electron transfer and fast ion diffusion in 3D space [9], which synergistically leads to the best photocatalytic activity.

Consequently, the visible light-induced photoactivity of NGH-2 was further examined for different initial concentrations of AO and the corresponding pseudo first-order kinetic model plots are presented in Fig. 11. The photodegradation rate constant for 5 mg L−1, 10 mg L−1, 15 mg L−1, 20 mg L−1 and 25 mg L−1 of AO are 8.97  10−3 min−1, 1.22  10−2 min−1, 6.17  10−3 min−1, 3.76  10−3 min−1 and 2.96  10−3 min−1, respectively.

Evidently, the

photooxidation rate of AO increases linearly upon increasing the initial AO concentration from 5 mg L−1 to 10 mg L−1. This increase is probably due to adequate availability of active sites, which brings about effective adsorption and rapid photodegradation. However, the photodecolorization rate constant decreases gradually with further increase in initial dye concentration, which can be ascribed mainly to the limited number of accessible binding sites on the NGH-2 photocatalyst. As a consequence, fewer catalytic sites are available to the contaminant at higher concentrations, leading to reduced photocatalytic activity. Nevertheless, the observed trend is reasonably similar to other studies on photodegradation of organic contaminants in aqueous media [69,70].

4. Conclusion In summary, lightweight and mechanically strong N-doped monolithic graphene hydrogels, with varying doping levels, were successfully developed through a facile, scalable, one-step hydrothermal reaction using GO and NH4OH as C and N sources, respectively. The N contents of the graphene macroassemblies present an inverse relation with doping density. However, regardless of the doping level, the NGHs display much higher surface area and more conductive electrical pathways than their pristine counterpart, and hence show significantly improved adsorption capacity and photocatalytic activity toward AO, a common industrial pollutant. In particular, the NGH-2 sample with a N content of 6.78 at.% (1.31 at.% pyridinic N, 3.39 at.% pyrrolic N and 2.08 at.% quaternary N) and a SBET of 151

m2 g−1 exhibits far superior adsorption properties (a maximum uptake of 124 mg g−1 at 25 °C) than many other competing solid adsorbents.

Although the adsorption kinetics

correlates well with the pseudo-second order rate law, the overall reaction is controlled by a number of processes, such as film diffusion and intraparticle diffusion. In addition, the adsorption equilibrium is best simulated by the Langmuir isotherm. Thermodynamic investigations suggest that the adsorption mode is spontaneous, endothermic and physical in nature.

The isosteric heat of adsorption has an unusual dependence on surface loading,

which is most likely due to attractive intermolecular forces among the neighboring dye molecules. Additionally, the photocatalytic efficiency of NGH-2 for degradation of AO is considerably higher than that of many other photocatalytic materials under visible light irradiation.

The excellent photooxidative performance is attributed to the synergistic

interplay between N functionalization and 3D cross-linked mesoporous structure, which induces enhanced light absorption, reduced recombination of electron–hole pairs and rapid photoinduced electron transfer in 3D space. However, the precise pathways involved in the photodecolorization reaction as well as the time-dependent evolution of reaction intermediates and end products remain largely unknown and should be elucidated in detail. This insight into reaction pathways would in turn lead to an atomic-scale understanding of the photooxidation properties of these promising macroscale 3D graphene-based materials and provide a rational guidance for further optimizing their photocatalytic performance. Additionally, since industrial waste streams contain different types of pollutants and other undesirable substances, considerable efforts must also be devoted to evaluating and establishing the dye removal potential from real effluents. The anticipated results would provide the ultimate laboratory demonstration of the separation performance prior to scaleup. Overall, our preliminary findings demonstrate that NGH-2 holds tremendous potential

in advancing state-of-the-art water treatment technologies, and may be further considered for various other applications, including batteries, fuel cells, electrochemical capacitors, etc.

Acknowledgements This work was financially supported by the National Research Foundation (NRF), Prime Minister’s Office, Republic of Singapore through Centre for Advanced 2D Materials (CA2DM) at National University of Singapore. We are sincerely thankful to CA2DM for the seed grant as well as the in-kind support. Ms. Yiqun Jiang further acknowledges the financial support provided by the Chinese Scholarship Council, Ministry of Education of the People’s Republic of China for her Doctoral study.

Appendix A. Supplementary information Supplementary information associated with this article can be found in the online version.

References [1]

K. Geim, K. S. Novoselov, The rise of graphene, Nat. Mater. 6 (2007) 183‒191.

[2]

V. Singh, D. Joung, L. Zhai, S. Das, S. I. Khondaker, S. Seal, Graphene based materials: past, present and future, Prog. Mater. Sci. 56 (2011) 1178‒1271.

[3]

K. S. Novoselov, V. I. Fal’ko, L. Colombo, P. R. Gellert, M. G. Schwab, K. Kim, A roadmap for graphene, Nature 490 (2012) 192–200.

[4]

L. Jiang, Z. Fan, Design of advanced porous graphene materials: from graphene nanomesh to 3D architectures. Nanoscale 6 (2014) 1922-1945.

[5]

S. Yin, Z. Niu, X. Chen, Assembly of graphene sheets into 3D macroscopic structures, Small 8 (2012) 2458‒2463.

[6]

H.-P. Cong, J.-F. Chen, S.-H Yu, Graphene-based macroscopic assemblies and architectures: an emerging material system, Chem. Soc. Rev. 43 (2014) 7295‒7325.

[7]

Y. Xu, G. Shi, X. Duan, Self-assembled three-dimensional graphene macrostructures: synthesis and applications in supercapacitors, Acc. Chem. Res. 48 (2015) 1666‒1675.

[8]

Z.-S. Wu, Y. Sun, Y.-Z. Tan, S. Yang, X. Feng, K. M llen, Three-dimensional graphene-based

macro-

and

mesoporous

frameworks

for

high-performance

electrochemical capacitive energy storage, J. Am. Chem. Soc. 134 (2012) 19532‒19535. [9]

M. F. El-Kady, Y. Shao, R. B. Kaner, Graphene for batteries, supercapacitors and beyond, Nat. Rev. Mater. 1 (2016) 16033.

[10] Q. Shi, Y. Cha, Y. Song, J.-I. Lee, C. Zhu, X. Li, M.-K. Song, D. Du, Y. Lin, 3D graphene-based hybrid materials: synthesis and applications in energy storage and conversion, Nanoscale 8 (2016) 15414‒15447. [11] Q. Fang, Y. Shen, B. Chen, Synthesis, decoration and properties of three-dimensional graphene-based macrostructures: a review, Chem. Eng. J. 264 (2015) 753‒771. [12] S. Mao, G. Lu, J. Chen, Three-dimensional graphene-based composites for energy applications, Nanoscale 7 (2015) 6924‒6943. [13] Z. Yan, W. Yao, L. Hu, D. Liu, C. Wang, C.-S. Lee, Progress in the preparation and application of three-dimensional graphene-based porous nanocomposites, Nanoscale 7 (2015) 5563‒5577. [14] X. Wang, G. Sun, P. Routh, D.-H. Kim, W. Huang, P. Chen, Heteroatom-doped graphene materials: syntheses, properties and applications, Chem. Soc. Rev. 43 (2014) 7067‒7098. [15] Z. Xing, Z. Ju, Y. Zhao, J. Wan, Y. Zhu, Y. Qiang, Y. Qian, One-pot hydrothermal synthesis of nitrogen-doped graphene as high-performance anode materials for lithium ion batteries. Sci. Rep. 6 (2016) 26146. [16] H. Wang, T. Maiyalagan, X. Wang, Review on recent progress in nitrogen-doped graphene: synthesis, characterization, and its potential applications, ACS Catal. 2 (2012) 781‒794.

[17] G. K. Parshetti, S. Chowdhury, R. Balasubramanian, Hydrothermal conversion of urban food waste to chars for removal of textile dyes from contaminated waters, Bioreour. Technol. 161 (2014) 310‒319. [18] R. Arshad, S. Farooq, N. Iqbal, S. S. Ali, Mutagenic effect of acridine orange on the expression of penicillin G acylase and β-lactamase in Escherichia coli, Lett. Appl. Microbiol. 42 (2006) 94–101. [19] F. Ito, T. Kakiuchi, T. Nagamura, Excitation energy migration of acridine orange intercalated into deoxyribonucleic acid thin films, J. Phys. Chem. 111 (2007) 6983‒6988. [20] H.-K. Liu, P. J. Sadler, Metal complexes as DNA intercalators, Acc. Chem. Res. 44 (2011) 349‒359. [21] N. P. E. Barry, P. J. Sadler, Exploration of the medical periodic table: towards new targets. Chem. Commun. 49 (2013) 5106‒5131. [22] Y. Xu, K. Sheng, C. Li, G. Shi, Self-assembled graphene hydrogels via a one-step hydrothermal process, ACS Nano 4 (2010) 4324–4330. [23] B. Adeniran, R. Mokaya, Is N-doping in porous carbons beneficial for CO2 storage? Experimental demonstration of the relative effects of pore size and N-doping, Chem. Mater. 28 (2016) 994–1001. [24] T. Panja, D. Bhattachrjya, J.-S. Yu, Nitrogen and phosphorous co-doped cubic ordered mesoporous carbon as a supercapacitor electrode material with extraordinary cyclic stability, J. Mater. Chem. A 3 (2015) 18001–18009. [25] S. Chowdhury, R. Balasubramanian, Highly efficient, rapid and selective CO2 capture by thermally treated graphene nanosheets, J. CO2 Util. 13 (2016) 50–60. [26] V. Deerattrakul, P. Dittanet, M. Sawangphruk, P. Kongkachuichay, CO2 hydrogenation to methanol using Cu-Zn catalyst supported on reduced graphene oxide nanosheets, J. CO2 Util. 16 (2016) 104–113. [27] S. Chowdhury, G. K. Parshetti, R. Balasubramanian, Post-combustion CO2 capture using mesoporous TiO2/graphene oxide nanocomposites, Chem. Eng. J. 263 (2015) 374–384.

[28] W.-D. Xue, R. Zhao, A simple approach towards nitrogen-doped graphene and metal/graphene by solid-state pyrolysis of metal phthalocyanine, New J. Chem. 38 (2014) 2993‒2998. [29] M. S. Dresselhaus, A. Jorio, M. Hofmann, G. Dresselhaus, R. Saito, Perspectives on carbon nanotubes and graphene Raman spectroscopy, Nano Lett. 10 (2010) 751–758. [30] X. Mi, G. Huang, W. Xie, W. Wang, Y. Liu, J. Gao, Preparation of graphene oxide aerogel and its adsorption for Cu2+ ions, Carbon 50 (2012) 4856–4864. [31] B. J. Matsoso, K. Ranganathan, B. K. Mutuma, T. Lerotholi, G. Jones, N. J. Coville, Time-dependent evolution of the nitrogen configurations in N-doped graphene films, RSC Adv. 6 (2016) 106914–106920. [32] D. Wei, Y. Liu, Y. Wang, H. Zhang, L. Huang, G. Yu, Synthesis of N-doped graphene by chemical vapor deposition and its electrical properties, Nano Lett. 9 (2009) 1752–1758. [33] J. Li, Z. Ren, Y. Zhou, X. Wu, X. Xu, M. Qi, W. Li, J. Bai, L. Wang, Scalable synthesis of pyrrolic N-doped graphene by atmospheric pressure chemical vapor deposition and its terahertz response, Carbon 62 (2013) 330‒336. [34] S. Lim, R. Li, W. Ji, J. Lin, Effects of nitrogenation on single-walled carbon nanotubes within density functional theory, Phys. Rev. B 76 (2007) 195406. [35] L. Lai, J. R. Potts, D. Zhan, L. Wang, C. K. Poh, C. Tang, H. Gong, Z. Shen, J. Lin, R. S. Ruoff, Exploration of the active center structure of nitrogen-doped graphenebased catalysts for oxygen reduction reaction. Energy Environ. Sci. 5 (2012) 7936– 7942. [36] S. Chowdhury, P. Saha, Sea shell powder as a new adsorbent to remove basic green 4 (malachite green) from aqueous solutions: equilibrium, kinetic and thermodynamic studies, Chem. Eng. J. 164 (2010) 168–177. [37] L. Zuo, Y. Guo, X. Li, H. Fu, X. Qu, S. Zheng, C. Gu, D. Zhu, P. J. J. Alvarez, Enhanced adsorption of hydroxyl- and amino-substituted aromatic chemicals to nitrogen-doped multiwall carbon nanotubes: a combined batch and theoretical calculation study, Environ. Sci. Technol. 50 (2016) 899‒905.

[38] G. Crini, P.-M. Badot, Application of chitosan, a natural aminopolysaccharide, for dye removal from aqueous solutions by adsorption processes using batch studies: a review of recent literature, Prog. Polym. Sci. 33 (2008) 399–447. [39] S. Banerjee, R. K. Gautam, A. Jaiswal, M. C. Chattopadhyaya, Y. C. Sharma, Rapid scavenging of methylene blue dye from a liquid phase by adsorption on alumina nanoparticles, RSC Adv. 5 (2015) 14425–14440. [40] D. Sidiras, F. Batzias, E. Schroeder, R. Ranjan, M. Tsapatsis, Dye adsorption on autohydrolyzed pine sawdust in batch and fixed-bed systems, Chem. Eng. J. 171 (2011) 883–896. [41] S. Qadri, A. Ganoe, Y. Haik, Removal and recovery of acridine orange from solutions by use of magnetic nanoparticles, J. Hazard. Mater. 169 (2009) 318–323. [42] I. Safarik, L. F. T. Rego, M. Borovska, E. Mosiniewicz-Szablewska, F. Weyda, M. Safarikova, New magnetically responsive yeast-based biosorbent for the efficient removal of water-soluble dyes, Enzyme Microb. Technol. 40 (2007) 1551–1556. [43] S. K. H. Nejad-Darzi, A. Samadi-Maybodi, M. Ghobakhluo, Synthesis and characterization of modified ZSM-5 nanozeolite and their applications in adsorption of acridine orange dye from aqueous solution, J. Porous Mater. 20 (2013) 909–916. [44] H. Wang, Y. Chen, Y. Wei, A novel magnetic calcium silicate/graphene oxide composite material for selective adsorption of acridine orange from aqueous solutions, RSC Adv. 6 (2016) 34770–34781. [45] R. Gong, J. Sun, D. Zhang, K. Zhong, G. Zhu, Kinetics and thermodynamics of basic dye sorption on phosphoric acid esterifying soybean hull with solid phase preparation technique, Bioresour. Technol. 99 (2008) 4510‒4514. [46] F. B. Shahri, A. Niazi, Synthesis of modified maghemite nanoparticles and its application for removal of acridine orange from aqueous solutions by using BoxBehnken design, J. Magn. Magn. Mater. 396 (2015) 318–326. [47] L. Sun, B. Fugetsu, Graphene oxide captured for green use: influence on the structures of calcium alginate and macroporous alginic beads and their application to aqueous removal of acridine orange, Chem. Eng. J. 240 (2014) 565–573.

[48] A. A. Jalil, S. Triwahyono, M. R. Yaakob, Z. Z. Z. Azmi, N. Sapawe, N. H. N. Kamarudin, H. D. Setiabudi, N. F. Jaafar, S. M. Sidik, S. H. Adam, B. H. Hameed, Utilization of bivalve shell-treated Zea mays L. (maize) husk leaf as a low-cost biosorbent for enhanced adsorption of malachite green, Bioresour. Technol. 120 (2012) 218–224. [49] M. Al-Ghouti, M. A. M. Khraisheh, M. N. M. Ahmad, S. Allen, Thermodynamic behaviour and the effect of temperature on the removal of dyes from aqueous solution using modified diatomite: a kinetic study, J. Colloid Interface Sci. 287 (2005) 6–13. [50] M. Doğan, H. Abak, M. Alkan, Adsorption of methylene blue onto hazelnut shell: kinetics, mechanism and activation parameters, J. Hazard. Mater. 164 (2009) 172– 181. [51] O. Hamdaoui, E. Naffrechoux, Modeling of adsorption isotherms of phenol and chlorophenols onto granular activated carbon part I. Two-parameter models and equations allowing determination of thermodynamic parameters, J. Hazard. Mater. 147 (2007) 381–394. [52] S. Rangabhashiyam, N. Anu, M. S. Giri Nandagopal, N. Selvaraju, Relevance of isotherm models in biosorption of pollutants by agricultural byproducts, J. Environ. Chem. Eng. 2 (2014) 398–414. [53] H. Tahermansouri, Z. Dehghan, F. Kiani, Phenol adsorption from aqueous solutions by functionalized multiwalled carbon nanotubes with a pyrazoline derivative in the presence of ultrasound, RSC Adv. 5 (2015) 44263–44273. [54] F. Delage, P. Pre, P. Le Cloirec, Mass transfer and warming during adsorption of high concentrations of VOCs on an activated carbon bed: experimental and theoretical analysis, Environ. Sci. Technol. 34 (2000) 4816‒4821. [55] S. B. Khan, M. Faisal, M. M. Rahman, A. Jamal, Exploration of CeO2 nanoparticles as a chemi-sensor and photo-catalyst for environmental applications, Sci. Total Environ. 409 (2011) 2987–2992. [56] S. H. Kim, A. Umar, R. Kumar, A. A. Ibrahim, G. Kumar, Facile synthesis and photocatalytic activity of cocoon-shaped CuO nanostructures, Mater. Lett. 156 (2015) 138–141.

[57] R. Selvaggi, L. Tarpani, A. Santuari, S. Giovagnoli, L. Latterini, Silica nanoparticles assisted photodegradation of acridine orange in aqueous suspensions, Appl. Catal., B. 168–169 (2015) 363–369. [58] S. B. Khan, M. Faisal, M. M. Rahman, A. Jamal, Low-temperature growth of ZnO nanoparticles: photocatalyst and acetone sensor, Talanta 85 (2011) 943–949. [59] M. Faisal, S. B. Khan, M. M. Rahman, A. Jamal, K. Akhtar, M. M. Abdullah, Role of ZnO–CeO2 nanostructures as a photocatalyst and chemi-sensor, J. Mater. Sci. Technol. 27 (2011) 594–600. [60] I. Ahmad, S. B. Khan, T. Kamal, A. M. Asiri, Visible light activated degradation of organic pollutants using zinc–iron selenide, J. Mol. Liq. 229 (2017) 429–435. [61] Y. Wang, L. Sun, B. Fugetsu, Morphology-controlled synthesis of sunlight-driven plasmonic photocatalysts Ag@AgX (X = Cl, Br) with graphene oxide template, J. Mater. Chem. A 1 (2013) 12536–12544. [62] R. A. Rakkesh, D. Durgalakshmi, S. Balakumar, Nanostructuring of a GNS–V2O5– TiO2 core–shell photocatalyst for water remediation applications under sun-light irradiation, RSC Adv. 5 (2015) 18633–18641. [63] R. A. Rakkesh, D. Durgalakshmi, S. Balakumar, Graphene based nanoassembly for simultaneous detection and degradation of harmful organic contaminants from aqueous solution, RSC Adv. 6 (2016) 34342–34349. [64] J. Xu, W. Meng, Y. Zhang, L. Li, C. Guo, Photocatalytic degradation of tetrabromobisphenol A by mesoporous BiOBr: efficacy, products and pathway, Appl. Catal., B. 107 (2011) 355–362. [65] M. Ge, J. W. Li, L. Liu, Z. Zhou, Template-free synthesis and photocatalytic application of rutile TiO2 hierarchical nanostructures, Ind. Eng. Chem. Res. 50 (2011) 6681–6687. [66] X. Xiao, R. Hao, M. Liang, X. Zuo, J. Nan, L. Li, W. Zhang, One-pot solvothermal synthesis of three-dimensional (3D) BiOI/BiOCl composites with enhanced visiblelight photocatalytic activities for the degradation of bisphenol-A, J. Hazard. Mater. 233–234 (2012) 122–130.

[67] G. Li, Q. Chen, J. Lan, Facile synthesis, metastable phase induced morphological evolution and crystal ripening, and structure-dependent photocatalytic properties of 3D hierarchical anatase superstructures, ACS Appl. Mater. Interfaces 6 (2014) 22561– 22568. [68] H. Li, J. Han, N. Guo, H. Yu, The design of 3D artificial leaves with spatially separated active sites for H2 and O2 generation and their application to water splitting, Chem. Commun. 52 (2016) 4080–4083. [69] L. Luo, Y. Yang, A. Zhang, M. Wang, Y. Liu, L. Bian, F. Jiang, X. Pan, Hydrothermal synthesis of fluorinated anatase TiO2/reduced graphene oxide nanocomposites and their photocatalytic degradation of bisphenol A, Appl. Surf. Sci. 353 (2015) 469‒479. [70] R.-a. Doong, C.-Y. Liao, Enhanced visible-light-responsive photodegradation of bisphenol A by Cu, N-codoped titanate nanotubes prepared by microwave-assisted hydrothermal method, J. Hazard. Mater. 322 (2017) 254‒262.

FIGURE CAPTIONS

Figure 1. Chemical structure of acridine orange (AO). Color legend: C (silver), H (white), N (cyan).

Scheme 1. Key steps involved in the synthesis of N-doped graphene hydrogels (NGHs).

Figure 2. (a) Digital photographs of the as-synthesized NGH bulk materials with different N-doping levels. (b) NGH-2 propped up on the stamen of a flower (Lillium sp.). (c) NGH-2 (height  diameter = 1.45 cm  0.95 cm) supporting a 200 g knob weight, more than 4000 times its own weight of 50 mg.

Figure 3. Low-magnification FESEM images of (a) GH (scale bar = 50 μm), (b) NGH-1 (scale bar = 30 μm), (c, d) NGH-2 (scale bars = 20 and 10 μm, respectively), (e) NGH-3 (scale bar = 10 μm) and (f) NGH-4 (scale bar = 30 μm).

Figure 4. Pore size distribution curves of the as-prepared NGHs derived from the N2 desorption branch by applying the BJH formulism.

Figure 5. (a) Wide-angle XRD profiles, (b) Raman spectra and (c) FTIR spectra of the synthesized NGH samples.

Figure 6. (a) Deconvoluted high-resolution N1s spectrum of NGH-2. (b) Distribution of various N bonding configurations in the as-synthesized NGH materials.

Figure 7. (a) Dye adsorption efficiency of N-doped and undoped graphene hydrogels as determined from preliminary batch adsorption experiments (experimental conditions: C0 = 5 mg L−1; m/V = 6 mg/0.03 L; temperature = 20 °C; agitation speed = 200 rpm). (b) Digital

photographs illustrating the adsorption potential of NGH-2 for different initial concentrations of AO (experimental conditions: m/V = 6 mg/0.03 L; temperature = 20 °C; agitation speed = 200 rpm)

Figure 8. Effect of solution pH on the percentage removal of AO by NGH-2 (experimental conditions: C0 = 5 mg L−1; m/V = 6 mg/0.03 L; temperature = 20 °C; agitation speed = 200 rpm).

Figure 9. Effect of temperature on the dye uptake capacity of NGH-2 as a function of contact time (experimental conditions: C0 = 5 mg L−1; pH = 5; m/V = 6 mg/0.03 L; agitation speed = 200 rpm) and (a) pseudo-first order and (b) pseudo-second order kinetic model fits to the experimental data.

Figure 10. Photocatalytic decolorization of AO under visible light irradiation by the assynthesized NGHs as a function of reaction time (experimental conditions: C0 = 20 mg L−1; catalyst dose = 0.33 g L−1; temperature = 25 °C) (a) and the corresponding pseudo first-order kinetic model plots (b).

Figure 11. Effect of initial dye concentration on the photodegradation kinetics of AO by NGH-2 under visible light irradiation (experimental conditions: catalyst dose = 0.33 g L−1; pH = 5; temperature = 25 °C).

Scheme 2. Plausible reaction mechanism for degradation of AO over the N-doped graphene monoliths upon irradiation with visible light.

Figure 1

Scheme 1

(a) (c)

(b) Figure 2

(a)

(b)

(c)

(d)

(e)

(f)

Figure 3

Differential Pore Volume (cm3 g-1)

GH

NGH-1

NGH-2

NGH-3

NGH-4

0

5

10

15

20

25

30

35

Pore Width (nm)

Figure 4

40

45

50

55

GH NGH-1 NGH-2 NGH-3 NGH-4

(0 0 2)

Intensity (a.u.)

(a)

(1 0 1)

10

15

20

25

30

35

40

45

50

2(degree)

(b)

GH NGH-1 NGH-2 NGH-3 NGH-4

D

Intensity (a.u.)

G

1000

1200

1400

1600

1800

2000

Raman Shift (cm-1) -NH2

Transmittance (%)

NGH-4

-COO- -C-NH2

NGH-3 NGH-2 NGH-1 GH

(c) 4000

3500

3000

2500

2000

1500

Wavelength (cm-1)

Figure 5

1000

500

(a)

pyrrolic-N

Intensity (a.u.)

pyridinic-N

quanternary-N

394

396

398

400

402

404

406

408

Binding Energy (eV)

Nitrogen Concentration (at%)

10 9 8

(b)

7 6 5 4 3 2 1 0

Figure 6

Pyridinic N Pyrrolic N Quaternary N

NGH-1

NGH-2

NGH-3

NGH-4

GH NGH-1 NGH-2 NGH-3 NGH-4

Dye Removal (%)

100 80

(b)

60 40 20 0

(a) 0

50

100

Time (min)

Figure 7

150

200

100

Dye Removal (%)

90 80 70 60 50

3

4

5

6

7

Initial solution pH

Figure 8

8

9

35

(a)

30

qt (mg g-1)

25 20 15 10 20C 25C 30 C

5 0 0

100

200

300

400

Time (min) 35

(b)

30

qt (mg g-1)

25 20 15 10 20C 25C 30 C

5 0 0

100

200

Time (min) Figure 9

300

400

20 NGH-1 NGH-2 NGH-3 NGH-4

Ct (mg L-1)

16 12 8 4

(a)

Dark Light On

0

-50

0

50

100 150 200 250 300 350

Time (min) 1.2

NGH-1 (R2 = 0.989) NGH-2 (R2 = 0.977) NGH-3 (R2 = 0.986) NGH-4 (R2 = 0.983)

1.0

ln(Ct/C0)

0.8 0.6 0.4 0.2

(b)

0.0 0

100

200

t (min)

Figure 10

300

CO2 + H2O •O2 e‒

e‒

e‒

e‒

e‒

e‒

e‒

O2

hv > Eg Valence Band

‒ h+

•OH

Scheme 2

AO

h+

h+

h+

h+

CO2 + H2O

h+

‒

e‒

Conduction Band

OH

AO

h+

h+

N 2p

4

5 mg L-1 (R2 = 0.986) 10 mg L-1 (R2 = 0.958) 15 mg L-1 (R2 = 0.993) 20 mg L-1 (R2 = 0.977) 25 mg L-1 (R2 = 0.994)

ln(Ct/C0)

3

2

1

0 0

100

200

t (min) Figure 11

300

TABLES Table 1. Textural properties of the N-doped and undoped monolithic graphene hydrogels as determined from N2 adsorption/desorption isotherms measured at –196 °C.

a

Sample

SBET (m2 g−1)a

Vt (cm3 g−1)b

Dp (nm)c

GH

75

0.23

12.22

NGH-1

124

0.12

3.94

NGH-2

151

0.23

6.04

NGH-3

135

0.25

7.37

NGH-4

128

0.11

3.57

BET specific surface area, bTotal pore volume, cAverage pore diameter

Table 2. Isotherm parameters and kinetic constants for adsorption of AO on NGH-2.

Isotherm parameters Langmuir

Freundlich

qm (mg g−1)

KL (L mg−1)

R2

χ2

KF (mg g−1) (L mg−1)1/n

n

R2

χ2

124.24

3.95

0.98

11.67

91.32

6.56

0.63

470.87

Kinetic constants Temp. (°C )

20 25 30

qe,exp (mg g−1)

25.10 9 27.15 27.42

Pseudo-first-order

Pseudo-second-order

qe,cal (mg g−1)

k1 (min−1)

R2

χ2

qe,cal (mg g−1)

k2 (g mg−1 min−1)

24.49 26.96 27.57

3.89  10−2 4.80  10−2 5.88  10−2

0.97 0.99 0.99

1.66 0.33 0.11

27.12 29.54 30.15

2.06  10−3 2.42  10−3 2.81  10−3

R2 0.99 0.98 0.96

χ2 0.41 1.19 2.72

Table 3. Comparison of the maximum AO uptake capacity of NGH-2 with those of other recently reported adsorbents.

a

Adsorbent

pH

Temp. (°C)

Dye conc. (mg L-1)

qm (mg g-1)a

Reference

Pine sawdust (autohydroloyzed) γ-Fe2O3 nanoparticles Magnetically modified fodder yeast cells Fe-ZSM-5 nanozeolite Food waste hydrochar Calcium silicate/GO Esterified soybean hull SDS-coated γ-Fe2O3 nanoparticles Calcium alginate/GO N-doped graphene hydrogel

7.8‒8 4.5±0.3 – 9 8 6 6 5.1 – 5

23 24 Room temp. 25 40 25 – 25 – 25

1.8‒9 3.7–184 1000‒2000 15–35 10‒100 10–140 50‒500 10–500 – 5–25

19 59 62 63 79 193 238 286 836 124

Sidiras et al. [40] Qadri et al. [41] Safarik et al. [42] Nejad-Darzi et al. [43] Parshetti et al. [17] Wang et al. [44] Gong et al. [45] Shahri and Niazi. [46] Sun and Fugetsu. [47] This study

Values are rounded off to the nearest whole number wherever applicable.

Table 4. Isosteric heat of adsorption for different amounts of AO adsorbed on NGH-2. qe (mg g−1)

‒∆HX (kJ mol−1)

R2

2 4 6 8 10

47.96 48.68 49.42 50.20 51.02

0.945 0.945 0.945 0.945 0.945

Table 5. Comparison of the AO photodegradation efficiency of NGH-2 with other recently reported photocatalyst systems.

Photocatalyst

Dye conc. (mol L−1)

Light source

Irradiation time (min)

Degradatio (%)

CeO2 nanoparticles CuO nanococoons Si nanoparticles ZnO nanoparticles ZnO–CeO2 ZnSe–FeSe Ag@AgCl@graphene Graphene–V2O5–TiO2 ZnO–Ag–graphene N-doped graphene hydrogel

3.0 × 10−5 3.7 × 10−5 1.0 × 10−5 3.0 × 10−5 3.0 × 10−5 3.0 × 10−5 7.5 × 10−5 2.5 × 10−5 3.0 × 10−5 7.5 × 10−5

UV UV Visible UV Visible Visible Sunlight Sunlight Sunlight Visible

170 80 50 80 170 180 25 20 20 300

38 60 58 96 85 86 95 100 99 70

50

GRAPHICAL ABSTRACT

100 Adsorption Photocatalysis

Efficiency (%)

80 60 40 20 0 NGH-1

NGH-2

NGH-3

Increasing N content

51

NGH-4

RESEARCH HIGHLIGHTS



N-doped graphene hydrogels developed by facile one-step hydrothermal reaction



N content inversely proportional to doping amount



Brunauer-Emmett-Teller specific surface area increases at all doping levels



Maximum dye adsorption and highest photodecolorization at N loading of 6.78 at.%



Better photoreduction due to synergistic effect of N-doping and 3D network structure

52