Environmental Pollution 251 (2019) 344e353

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Nitrogen and sulfur codoped graphene aerogels as absorbents and

visible light-active photocatalysts for environmental remediation
applications+
Yiqun Jiang a, 1, Shamik Chowdhury b, 1, Rajasekhar Balasubramanian a, *
a
Department of Civil & Environmental Engineering, National University of Singapore, 1 Engineering Drive 2, Singapore 117576, Singapore
b
School of Environmental Science and Engineering, Indian Institute of Technology Kharagpur, West Bengal 721302, India

a r t i c l e i n f o a b s t r a c t

Article history: Graphene aerogels (GAs) are increasingly being recognized as high performance multifunctional mate-
Received 1 March 2019 rials to tackle our current and emerging environmental concerns. In order to extend the application
Received in revised form potential of GAs, herein we have successfully synthesized nitrogen (N) and sulfur (S) codoped GAs
28 April 2019
(NSGAs) via a simple, scalable, and inexpensive approach. Owing to their large specific surface area (up to
Accepted 29 April 2019
Available online 30 April 2019
132 m2 g
1), profound porosity, superior mechanical properties, and coexistence of N and S atoms with
tunable atomic content and bonding configurations, the as-prepared NSGAs demonstrated exceptional
absorption capacity toward a broad spectrum of oils and organic solvents, with an average absorption
Keywords:
Graphene aerogels
rate many folds higher than conventional absorbents. Further, the NSGAs exhibited excellent photo-
Nitrogen and sulfur codoping catalytic activity for the decomposition of recalcitrant organic compounds under visible light illumina-
Oil absorption tion due to pronounced synergistic coupling effect between the heteroatoms. Specifically, after 5 h of
Photocatalysis exposure to visible light, a degradation efficiency of over 99% was observed and more than 84% of the
Organic pollutants total organic carbon was eliminated. Radical trapping experiments revealed that superoxide anion
Visible light radicals are the predominant oxygen reactive species driving the photocatalytic reactions. More
importantly, the mineralization byproducts did not pose any significant antibacterial activity, illustrating
the environmentally benign nature of these macroscale photocatalysts.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction
In recent years, three-dimensional (3D) graphene-based mac-
rostructures (GBMs), such as sponges, foams, hydrogels, aerogels
and xerogels, have been the subject of intense research and
development (R&D) efforts to address our pressing environmental
issues in a sustainable manner (Shen et al., 2014). The structural
integrity, mechanical robustness, and interconnected porosity of 3D
GBMs make them extremely suitable for a wide range of environ-
mental remediation applications (Yousefi et al., 2019). Freestanding
graphene aerogels (GAs), in particular, hold great promise for a
range of water treatment and air purification applications (Maleki,
2016; Gorgolis and Galiotis, 2017). The large specific surface area of
+
This paper has been recommended for acceptance by Eddy Y. Zeng.
* Corresponding author.
E-mail address: ceerbala@nus.edu.sg (R. Balasubramanian).
1
These authors contributed equally to this work.
GAs combined with their versatile surface chemistry, fast electron
conduction, and rapid mass transport enables efficient removal of a
broad spectrum of waterborne pollutants such as synthetic dyes,
potentially toxic elements, pesticides, herbicides, organic solvents,
salts, and pathogens as well as greenhouse gases such as CO2
(Chowdhury et al., 2019; Wu et al., 2019; Yousefi et al., 2019).
In order to enhance the environmental performance of GAs,
doping heteroatoms such as nitrogen (N) into their nanoscale
building blocks has attracted enormous research attention in recent
years (Fang et al., 2015). The incorporation of N leads to polarization
in the sp2 conjugated network and creates a band gap around the
Dirac point by repressing the neighboring density of states (Wang
et al., 2012a, 2014). This local electronic rearrangement in turn
improves the surface reactivity, electrical conductivity, and optical
absorption properties of GAs for a wide array of promising envi-
ronmental cleanup applications (Zhao et al., 2012; Sui et al., 2015;
Du et al., 2016; Jiang et al., 2017, 2019). For example, N-doped GAs
(NGAs) have shown remarkable performance as absorbents for a
multitude of organic compounds (Yang et al., 2016), with a

https://doi.org/10.1016/j.envpol.2019.04.132
0269-7491/© 2019 Elsevier Ltd. All rights reserved.
 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353
maximum absorption uptake of up to 210 g g
1 (Rahmani et al.,
2018), which is over four times greater than that of pristine GA
(Wang et al., 2012b). Moreover, it has recently been reported that
the dye removal efficiency of NGAs prepared using a hydrothermal
method is enhanced by about 68% over that of undoped GA (Jiang
et al., 2017). Very recently, the impressive photocatalytic activity
of NGAs has been demonstrated for degradation of recalcitrant
waterborne pollutants (Jiang et al., 2019; Wang et al., 2019). Usu-
ally, NGAs have been particularly designed and fabricated for spe-
cific environmental remediation applications with different N
atomic concentrations and molecular configurations. In order to
advance their practical applications, the fabrication of multifunc-
tional NGAs is highly desirable (Kotal et al., 2016).
In this regard, the incorporation of sulfur (S) together with N
into the hexagonal lattice of GAs could yield advanced and versatile
material systems with enhanced sorption capacities, desired pho-
tocatalytic properties and novel functionalities because of the
synergistic coupling effect between the heteroatoms (Su et al.,
2013; Huang et al., 2016; Saana Amiinu et al., 2016; Tian et al.,
2016; Ma et al., 2017). Specifically, S doping would likely result in
a localized structural distortion due to the large atomic size of S
compared with carbon (C) and the long SeC bond (Serp and
Machado, 2015), activating the adjacent sp2-C into coordinatively
unsaturated sp3-C. Further, the difference between the outermost
orbital energies of S and C atoms would probably induce a
nonuniform spin density distribution (Wang et al., 2014), convert-
ing the S atom into a catalytic center (Chen et al., 2016). This feature
of S-doped graphene together with the enhanced electron donation
capability of nearby C atoms due to N doping is thought to syner-
gistically boost the surface reactivity of GAs for a myriad of envi-
ronmental applications (Kicin  ski et al., 2014). However, to the best
of our knowledge, the effect of simultaneously introducing N and S
dopants on the environmental application of GAs remains relatively
unexplored.
To fill this knowledge gap, herein we have prepared N and S
codoped GAs (NSGAs) via a simple, scalable, and ecofriendly
method. The resulting lightweight N/S dual-doped carbonaceous
aerogels exhibit hierarchically interconnected meso/macroporosity
with large specific surface area and superior mechanical properties.
Such unique textural features, along with the coexistence of N and S
dopants with different types of bonding configurations, not only
offer a large number of accessible sites for molecular binding, but
also provide abundant reactive regions for catalytic reactions.
Attributing to these beneficial structural features, our versatile
NSGAs demonstrate tremendous potential for two selected envi-
ronmental remediation applications: (1) absorption of oils and
organic solvents, and (2) visible light-induced photodecomposition
of recalcitrant organic pollutants from aqueous phase.
2. Experimental
2.1. Synthesis of nitrogen/sulfur codoped graphene aerogels
NSGAs were fabricated by a simple one-step hydrothermal
approach. In a typical procedure, a weighed amount of ammonium
thiocyanate (NH4SCN, 99 wt%; Sigma-Aldrich) was added into
40 mL aqueous dispersion of graphene oxide (GO, 2 mg mL
1
dispersion in water; Sigma-Aldrich) and continuously stirred for
20 min at room temperature, followed by ultrasonic treatment at
30  C for 30 min in an ultrasonic bath operating at 37 kHz (Elma-
sonic S 60 H, Elma Schmidbauer GmbH, Germany). The as-obtained
homogenous mixture was then sealed in a stainless steel Teflon-
coated autoclave (100 mL) and treated hydrothermally at 180  C
for 12 h. The autoclave was then allowed to cool naturally to
ambient temperature. The resultant hydrogel was rinsed
345
repeatedly with copious amounts of ultrapure water to remove all
leftover chemicals, and ultimately freeze dried at
50  C for 24 h to
obtain the aerogel. By varying the mass of NH4SCN in 80 mg GO, a
series of NSGAs were prepared. The aerogels with different GO to
NH4SCN mass ratio of 1:0.5, 1:0.33, 1:0.25 and 1:0.20 were labeled
as NSGA-I, NSGA-II, NSGA-III and NSGA-IV, respectively. For com-
parison, pristine graphene aerogel (GA) (without adding any
doping agent) and N-doped graphene aerogel (NGA) (using 8 mL of
ammonium hydroxide (28e30% ammonia basis; Sigma-Aldrich) as
the N elemental precursor) was also prepared employing the same
procedure.
2.2. Instrumental characterization
A broad range of state-of-the-art microscopy (field emission
scanning electron microscopy, FESEM; transmission electron mi-
croscopy, TEM), spectroscopy (Raman spectroscopy; energy-
dispersive X-ray spectroscopy, EDX; X-ray photoelectron spectros-
copy, XPS; diffuse reflectance spectroscopy) and diffraction (X-ray
diffraction, XRD) techniques as well as classical surface-specific
methods, such as basic measurements of zeta potential, surface
area, average pore volume and pore size, were used to examine the
bulk and surface properties of the developed materials. For details
on these characterization procedures, please see the Supplemen-
tary information.
2.3. Environmental applications
In order to illustrate the multifunctional properties of the as-
prepared monolithic materials, the NSGAs were evaluated for two
selected environmental remediation applications: (1) absorption of
oils (diesel oil, engine oil, pump oil, and soybean oil), petroleum
products (kerosene), and organic solvents (acetone, n-hexane,
phenol, and toluene), and (2) visible light-induced photodecom-
position of a recalcitrant dye pollutant (acridine orange). The
detailed description of the corresponding experimental procedures
is provided in the Supplementary information.
3. Results and discussion
3.1. Materials characterization
In this work, a facile one-pot strategy, involving hydrothermal
processing of a homogeneous aqueous mixture of GO and NH4SCN,
was adopted to fabricate mechanically strong and lightweight GAs
doped with both N and S atoms (Fig. S1). Specifically, upon hy-
drothermal treatment at 180  C, NH4SCN decomposed into highly
reactive nitrogen and sulfur compounds (such as ammonia,
hydrogen sulfide, carbon disulfide, etc.), which subsequently
reacted with the defective sites and oxygenic functionalities (such
as carboxyl, carbonyl, epoxy, etc.) of GO (Su et al., 2013). As such,
some of the C atoms in the graphitic lattice were replaced with
heteroatoms (N and S). Additionally, the GO nanosheets underwent
reduction and eventually coalesced via supramolecular interactions
(such as pep stacking and hydrogen bonding), thus forming N/S
codoped 3D GBMs (Fig. S2). The microstructure of the as-obtained
NSGAs was characterized via FESEM and TEM imaging (Fig. 1,
Figs. S3 and S4). Similar to the morphology observed for pristine
monolithic GA (Fig. S3a), the NSGAs possessed an extensive
network of circular and/or elliptical shaped pores (Fig. 1a and
Fig. S3b‒d), interconnected through wrinkled graphene nanosheets
(Fig. 1b and Fig. S4). Moreover, the vast majority of the pores were
in the meso/macropore range, as inferred from the corresponding
BJH pore size distribution plots (Fig. S5), with the average pore
diameter spanning from 3.78 to 11.56 nm. The surface area and the
346 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353
Fig. 1. (a) FESEM and (b) TEM image of NSGA-IV. Scale bars in (a) and (b) represent 20 mm and 200 nm, respectively. (c) Wide-angle XRD profile of the as-synthesized GA, NGA and
NSGAs. The interlayer distance (d) between the 2D graphene building blocks in the aerogels, calculated using the Bragg's equation (Xing et al., 2016; Liu et al., 2017), is listed in the
inset. (d) Raman spectroscopic signatures of the as-prepared GA, NGA and NSGAs.
total pore volume of the NSGAs are listed in Table S1. Interestingly,
there was a dramatic decline in the surface area and pore volume of
the bulk materials at higher NH4SCN concentrations which could be
attributed to greater doping-induced defects. The defects doubled
up as interlocking sites for the flexible graphene sheets during the
self-assembling process to form a porous network with higher
interconnectivity.
Next, the effect of incorporating N and S dopants into the hon-
eycomb lattice of the fabricated 3D graphene-based macro-
architectures was examined through a quantitative analysis of their
wide-angle diffraction data. Irrespective of their doping level, the
NSGAs exhibited exemplary scattering patterns of semicrystalline
materials (Fig. 1c). The broad peak situated in the 2q range of
21 e28 , with maxima centred at around 2q ¼ 24 , can be indexed
to (0 0 2) facets of graphitic carbon, while the minor hump-like
feature appearing at about 2q ¼ 43 can be assigned to the
graphitic (1 0 0) crystal plane (Chowdhury et al., 2018). Moreover,
the relatively weak intensities of these peak indicate that the as-
prepared NSGAs were made up of highly disordered and
randomly oriented graphene sheets, with the interlayer distance
increased by almost 10% compared to their singly N-doped coun-
terpart (Fig. 1c). This can be rationalized by the fact that the cova-
lent radius of S (102 pm) is much larger than those of C (77 pm) and
N (75 pm) (Qie et al., 2015). As such, the introduction of S atoms can
considerably enlarge the interlayer spacing between graphene
sheets by steric occupancy (Li et al., 2017). Such an expanded
interlayer spacing has been reported to endow 2D-based materials
with beneficial structural and electronic modulations for high
performance catalytic applications (Gao et al., 2015).
Additionally, Raman spectroscopic characterization was carried
out to probe the lattice distortion and electronic charge localization
induced by N and S codoping. The comparative Raman scattering
spectra of the various synthesized samples is shown in Fig. 1d. The
G band at around 1570 cm
1 is a characteristic feature of in-plane
stretching mode of all sp2 bonded pairs (including CeC, CeN and
CeS bonds), while the D band at about 1330 cm
1 is mainly asso-
ciated with defective/disordered graphitic structures, including in-
plane heteroatom substitutions, vacancies, and/or grain
boundaries/edges (May et al., 2013). In comparison with undoped
and single doped samples prepared under similar conditions, the
peak positions of G band were downshifted by 6e13 cm
1, probably
because of the tensile (stretching) strain on the CeC bonds
resulting from the incorporation of heteroatoms into the sp2 con-
jugated backbone of our as-developed N/S dual-doped GAs. This is
conceivable since the bond lengths of CeC (1.42 Å) and CeN (1.41 Å)
are quite similar, whereas the CeS (1.74 Å) bond distance is usually
about 25% longer than that of the CeC bond (Chen et al., 2016).
However, there was no direct correlation between the ID/IG values
(i.e., the intensity ratio of the D to G band) and the bulk doping level
of the aerogels. Moreover, the ID/IG ratio computed for NSGA-III was
the lowest amongst all the samples, implying that both the dopants
preferred to occupy the edge planes of this material.
In order to explore the elemental composition of our chemically
modified graphene monoliths, EDX chemical mapping was per-
formed with different colors representing the distribution of
various elements in the sample (Fig. S6). As can be seen, the as-
prepared NSGAs were primarily composed of C and O. In addi-
tion, trace amounts of N and S were detected in all the samples,
confirming that N and S atoms were successfully codoped into the
constituent graphene sheets of the aerogels. However, the
elemental distribution of N was higher than S in all the samples
(Table S2). Specifically, the N doping level increased according to
the sequence: NSGA-IV (1.44 wt%) < NSGA-II (1.73 wt%) < NSGA-III
(2.38 wt%) < NSGA-I (2.42 wt%), while the total S content decreased
in the following order: NSGA-I (1.69 wt%) > NSGA-III (1.49 wt
%) > NSGA-IV (1.37 wt%) > NSGA-II (1.24 wt%). These trends suggest
that the atomic fraction of both N and S in the as-prepared samples
was not directly proportional to the amounts of NH4SCN added to
GO during the hydrothermal reaction.
Further, the precise bonding state of N and S atoms in our dual
heteroatom-doped graphene monoliths was established through
interpretation of their XPS spectra. The survey scan generated the
characteristic C1s, O1s, N1s and S2p photoelectron peaks at binding
energies (B.E.) of about 284, 532, 400 and 166 eV, respectively
(Fig. S7). Employing the iterative Shirley method, the high resolu-
tion N1s peak can be deconvoluted into two prominent signals
 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353
centred at ~398 and ~400 eV (Fig. 2a and Fig. S8), which can be
assigned to the well documented pyridinic (two CeN bonds in a
hexagon), and pyrrolic (two CeN bonds in a pentagon) type N
structural defects, respectively (Wang et al., 2012a). Surprisingly,
the emergence of a new peak at ~396 eV possibly indicates the
presence of substitutional N (Kusano et al., 2017); however, its exact
chemical identity is unknown. Fig. 2c compiles the fitting results for
the N1s spectra of the as-prepared N/S dual-doped monolithic
aerogels. The pyridinic N species was invariably present in all the
samples and was also the dominant N-bonding configuration in the
graphene core of all the materials, except NSGA-II wherein the
spectrum consisted primarily of pyrrolic N. Interestingly, no
decomposition peak representative of pyrrolic N was found in
NSGA-III; instead, substitutional N was the dominant defect
structure. Meanwhile, the high resolution S2p spectra can be fitted
into several distinct subpeaks corresponding to the different
chemical environments of the embedded S atoms in the hexagonal
carbon skeleton of the aerogels (Fig. 2b and Fig. S9). The major peak
situated at ~166 eV for all doping levels originates from oxidized
sulfur forms, principally sulfoxide (Upare et al., 2013). The
appearance of the rather weak but broad peak at the lower B.E. of
160.5 eV implies that some of the sulfoxide functionalities in NSGA-
III were reduced to sulfides in the high temperature-pressure hy-
drothermal environment (Fantauzzi et al., 2015). In addition, the
peak located at around 162 eV is indicative of the presence of
organosulfur (thiol) ligands (Wang et al., 2015). However, the thiol
(eSH) signal was not detected in the S2p XPS spectrum of NSGA-III
(Fig. 2d), suggesting that the S atoms mainly existed at the edge
plane and defect sites of the graphene building blocks of NSGA-III in
Fig. 2. High resolution (a) N1s core level XPS spectrum and (b) S2p core level XPS spectrum of NSGA-IV. Relative distribution of the different (c) N- and (d) S-structural motifs in the
NSGAs.
347
the form of thiophene-like structures. This was corroborated by the
peak-fitting resulting from the spineorbit coupling position of
S2p1/2 of the CeSeC covalent bond for thiophene S at 164.8 eV
(Yang et al., 2015) (Fig. S9). Last but not the least, the deconvolution
of an intense but relatively smaller photoelectric peak at 168.2 eV
can be attributed to the bonding of oxidized sulfur moieties
(CeSOXeC) such as sulfonate groups (‒SO‒3) to the 3D graphene
framework of the NSGA-IV monolith (Calvo et al., 2009; Yang et al.,
2012). In order to confirm the above findings, the O1s peak of each
material was also deconvoluted into its various components, as
shown in Fig. S10. These curve fitting results for the O1s spectrum
provide further evidence that S atoms were embedded in the form
of oxidized sulfur species (such as sulfonic acid and sulfonate
functional groups) on the surface of the 2D nanoscale building
blocks of the aerogels. A schematic demonstrating the various to-
pological defects in our bulk NSGA samples, as comprehended on
the basis of XPS core-level analysis, is given in Fig. 3.
Additionally, the light absorption properties of the NSGAs were
investigated through UVeVis diffuse reflectance spectroscopy. As
shown in Fig. S11, the absorption spectra of the NSGAs did not vary
significantly and presented high reflectance across the full visible
light portion of the electromagnetic spectrum. Interestingly, pris-
tine GA, even without the presence of a sizeable energy gap, dis-
played sufficient absorption in the visible region. However, after
incorporation of N and S heteroatoms, the intensity of the ab-
sorption increased by several folds. Applying the KubelkaeMunk
(KeM) function for direct allowed transition (Lόpez and Gόmez,
2012; Ntholeng et al., 2016), i.e., [F(R)hv]2 ¼ A (hv e Eg) where
F(R) is proportional to the absorption coefficient (a), hv is the
348 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353

Fig. 3. The various types of N and S functionalities observed in our as-fabricated NSGA materials: (1a) pyridinic N, (1b) pyridinic N with monovacancy, (1c) triple vacancy pyridinic
N, (2) pyrrolic N, (3) substitutional N, (4) sulfide, (5) thiol, (6) sulfoxide, (7) thiophenic and (8) sulfonic.

photon energy, Eg is the energy of the optical transition corre-

sponding to the optical energy band gap, and A is the proportion-
ality constant, the band gap energy could be calculated from the
intercept of the linear portion of the plot of [F(R)hv]2 versus hv
(Fig. S12). The Eg values estimated by this method were 1.51, 1.51,
1.52 and 1.42 eV for NSGA-I, NSGA-II, NSGA-III and NSGA-IV,
respectively.

3.2. Absorption of oils and organic solvents

With frequent occurrence of oil spill incidents and accidental

leakage of hydrophobic organic solvents (such as benzene, toluene,
cyclohexane, dichloromethane, etc.), the development of highly
efficient, lightweight, and mechanically strong absorbents has
become a top priority (Ge et al., 2016; Chowdhury et al., 2019). As
noted in the preceding section, our fabricated NSGAs have a highly
interconnected meso/macroporous structure with enlarged pores
and large pore volumes, and these properties endow aerogels with
good absorption properties. Furthermore, the aerogels developed
in this study have densities as low as 44.71 mg cm
3. Assuming that
the carbon density is 2 g cm
3, up to 97.76% of the aerogels can be
filled with air, which is extremely beneficial for absorption appli-
cations. Indeed, when our NSGA-IV aerogel was brought into con-
tact with soybean oil, the oil layer immediately began to shrink and
disappeared completely within a minute (inset of Fig. 4). More
importantly, the adsorbed oil could be removed from the
thermally-stable NSGA-IV absorbent through direct combustion in
air for repeated use (inset of Fig. 4). To further evaluate the ab-
sorption potential of the NSGA-IV aerogel, we investigated its ab-
sorption capacities for several commercial petroleum products
(e.g., diesel oil, engine oil, pump oil and kerosene) and common
organic solvents (e.g., acetone, n-hexane, phenol and toluene). The
absorption efficiency was assessed by percentage weight gain, i.e.,
the ratio of the mass of the organic liquid absorbed to that of the
dry aerogel.
As illustrated in Fig. 4, the freestanding NSGA-IV monolith
Fig. 4. Absorption capacity of NSGA-IV for various oils and organic solvents. The digital
images in the inset show the absorption of soybean oil by NSGA-IV over time.
displayed excellent absorption ability for various classes of organic
compounds. In particular, depending on the density and surface
tension of the hydrocarbons, the NSGA-IV material could absorb
oils and organic solvents up to 22e63 times its own weight. This
performance is superior to that of other contemporary absorbents
such as activated carbon (<1 times) (Lillo-Ro  denas et al., 2005),
marshmallow-like macroporous gels (6e14 times) (Hayase et al.,
2013), polyacrylonitrile foam (<18 times) (Zhang et al., 2018), gra-
phene aerogel (19e26 times) (Ren et al., 2017), spongy graphene
aerogel (29e54 times) (Luo et al., 2017), conjugated microporous
polymers (<33 times) (Li et al., 2011), polyurethane sponge (15e20
times) (Zhu et al., 2011), graphene foam (10e37 times) (Niu et al.,
2012), graphene/a-FeOOH aerogel (13e27 times) (Cong et al.,
2012), and nanocellulose aerogels (20e40 times) (Korhonen et al.,
2011). These results clearly demonstrate the application potential
of our synthesized multiple heteroatom doped graphene-based
 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353 349

aerogels for oil spill response and cleanup operations. We believe
that the organic compounds permeate via capillary action into the
interconnected void spaces of NSGA-IV, and are subsequently
entrapped within its bimodal meso/macroporous matrix, contrib-
uting to the high absorption capacity. Furthermore, it is interesting
to note that the absorption ability of NSGA-IV for phenol and
toluene was higher than that for acetone and n-hexane. This is
likely because aromatic hydrocarbons have greater affinity for
graphene-based materials and wet them better than aliphatic
compounds via pep bonding, as has been reported previously by
Qian et al. (2014).
3.3. Photocatalytic degradation of textile dyes
The direct discharge of dye-containing effluents from textile
mills and leather processing industries is detrimental to aquatic
environments. Heterogeneous photocatalysis hold enormous po-
tential to transform a broad spectrum of water borne pollutants
into relatively innocuous end products (e.g., CO2 and H2O). We,
therefore, also studied systematically the dye degradation ability of
our NSGA samples under visible light exposure. Acridine orange
(AO) was selected as a model dye contaminant due to its frequent
occurrence in industrial wastewaters and mutagenic activities in
humans (Chowdhury et al., 2018). Since photocatalysis is essentially
a surface phenomenon, occurring predominantly on the
photocatalyst surface rather than in the bulk, the dye/photocatalyst
suspension was first stirred for 1 h without any light exposure in
order to establish adsorption‒desorption equilibrium between the
dye and the photocatalyst surface (‘Light off’ period of Fig. 5a).
Compared to their undoped and single doped counterparts, the
adsorption capacity of the NSGAs toward AO appeared to be sub-
stantially higher, which was expected to provide excellent photo-
catalytic performance. The ‘Light on’ period in Fig. 5a presents the
visible light driven photocatalytic degradation of AO over the as-
prepared graphene-based macroscale structures as a function of
irradiation time. Even in the absence of any photocatalyst, the dye
molecules underwent a photodegradation of about 34% upon
exposure to visible light, which is a typical case of direct photolytic
transformation. Meanwhile, the bulk GA and NGA samples, both
exhibited fairly weak photocatalytic activity, affording only ~49%
and ~70% degradation of AO under visible light illumination for 5 h,
respectively. In contrast, almost complete decolorization
(95.29e99.23%) was achieved within 4 h over the freestanding
NSGAs. These results clearly ascertain that codoping with two
different elements is an appealing approach to boost the photo-
catalytic efficiency and thus the environmental performance of
GAs, due to the redistribution of charge (spin) density and newly
created active sites induced by the synergistic interactions between
the heteroatoms. Nevertheless, in order to confirm that the dye
removal using our as-developed NSGAs is based predominantly on

Fig. 5. (a) Visible light-induced photocatalytic degradation of AO by GA, NGA and NSGAs (experimental conditions: C0 ¼ 20 mg L
1; catalyst dose ¼ 0.33 g L
1; temperature ¼ 25  C).
(b) Percentage removal of TOC during photocatalytic degradation of AO over the NSGAs. (c) Effect of initial dye concentration on the photocatalytic degradation of AO by NSGA-IV
under visible light irradiation (experimental conditions: catalyst dose ¼ 0.33 g L
1; temperature ¼ 25  C). (d) Effect of pH on the photocatalytic degradation of AO by NSGA-IV under
visible light irradiation (experimental conditions: C0 ¼ 20 mg L
1; catalyst dose ¼ 0.33 g L
1; temperature ¼ 25  C).
350 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353
the photocatalytic degradation mechanism, we also studied the
adsorption and photocatalytic decolorization of AO over NSGA-IV
under identical experimental conditions (Fig. S13). The use of
NSGA-IV in the dark resulted in a dye removal level of approxi-
mately 38%. On the other hand, the use of NSGA-IV under visible
light illumination led to a dye elimination efficiency of close to
100%. These findings confirm that photocatalysis was the dominant
mechanism governing the purification process.
It should be noted, however, that the apparent disappearance of
the dye from the reaction medium does not necessarily imply a
complete degradation of the pollutant but rather the formation of
transformation products. These transformation products can
sometimes be even more persistent and/or toxic than the parent
compound. Therefore, the complete mineralization of the dye prior
to wastewater discharge to the sewer is of significant interest (Yu
et al., 2012). To this end, the total organic carbon (TOC) is an
important water quality parameter and its reduction is highly
desirable. Fig. 5b compares the rate of TOC disappearance during
the photocatalytic degradation of AO with the as-prepared NSGAs
under visible light irradiation. It can be observed that the highest
experimentally obtained TOC elimination rate was ~84%, reflecting
the superior dye photomineralization potential of our N/S codoped
hierarchically porous aerogels. Further, the TOC removal percent-
ages decreased according to the following sequence: NSGA-IV
(84.30%) > NSGA-III (76.40%) > NSGA-II (68.82%) > NSGA-I
(34.60%). In spite of its relatively low doping level amongst all the
prepared samples, the excellent photocatalytic effect of NSGA-IV
can be ascribed to the following factors. First, the highest specific
surface area and the largest pore size of NSGA-IV (Fig. S5 and
Table S1) provide more active sites for dye adsorption/activation. As
a prerequisite for photocatalyzed reactions, the enhanced adsorp-
tivity augment the photocatalytic activity. Second, the smallest
band gap energy of the NSGA-IV sample facilitates a more efficient
utilization of visible light, increasing the number of both electrons
and holes taking part in the photocatalytic reaction. This in turn
builds up the concentration of primary oxidants (viz., hydroxyl
radicals and superoxide anions) on the photocatalyst surface, thus
enhancing the photocatalytic performance. Last but not the least,
unlike the other bulk materials, up to 30% of the S atoms exists
exclusively as sulfonic acid groups within the host lattice of NSGA-
IV (Fig. 2d). Like pyridinic N, these oxidized sulfur clusters, situated
primarily at the edges or defect sites, serve as a preferential site for
absorbing oxygen (key electron acceptor in aqueous photocatalysis)
due to their large charge (or spin) density (Zhang et al., 2014).
Moreover, when N and S atoms are simultaneously incorporated
into the sp2 network of graphene, the corresponding charge (or
spin) densities are further improved (Su et al., 2013), and conse-
quently promote the generation of more primary oxidants. As a
result, NSGA-IV even with a fairly low amount of N and S exhibited
the best photocatalytic performance. Subsequently, the NSGA-IV
sample was more thoroughly evaluated for its practical applica-
tions in photocatalysis, as described in the following paragraphs.
Since the initial concentration of the target recalcitrant organic
compound (AO) can have a profound influence on the reaction
dynamics, the photocatalytic ability of NSGA-IV toward degrada-
tion of AO was explored over the concentration range of
5e25 mg L
1. Fig. 5c presents the time-dependent photo-
decolorization curves of AO as a function of initial dye concentra-
tions. Clearly, when the initial AO concentration was amplified from
5 mg L
1 to 15 mg L
1, the AO photodecomposition rate increased
sharply and it took less time to achieve the same photocatalytic
performance. This is likely because enhancement of the concen-
tration gradient augments the net driving force to effectively sup-
press the overall mass transfer resistance of the dye molecules
between the aqueous and solid phases, thus facilitating rapid
diffusion of AO within the 3D porous structure of the NSGA-IV
material, and ultimately resulting in faster photocatalytic degra-
dation of AO. Upon further increasing the initial mass of the dye in
the reaction medium up to 25 mg L
1, the degradation curves
became shallower and it took longer to completely degrade AO for
the same amount of light absorbance. It could be explained that the
net flux of light reaching the photocatalyst surface decreases at a
higher initial concentration of AO. This in turn partially offsets the
number of electronehole pairs participating in the photocatalytic
reaction, thus slowing down the visible light-induced degradation
reaction to some extent. Moreover, at higher initial dye concen-
trations, the concentrations of the intermediates generated during
the photocatalytic process are also expected to be higher. These
intermediates would compete with dye molecules for interaction
with primary oxidants, which eventually tend to lower the AO
degradation rate.
From an application point of view, it is equally important to
study the effect of operating pH on the photocatalytic process, since
the pH of the reaction medium dictates the surface charge char-
acteristics of the photocatalyst; and hence the overall rate of
oxidative degradation. Fig. 5d shows the photodecomposition
profile of AO as a function of irradiation time under acidic and
alkaline conditions. Evidently, increasing the initial pH of the dye
solution from pH 4 increased the removal efficiency of AO up to pH
7 and did not show any significant changes thereafter. The observed
trend can be attributed to the protonation of the pyridinic N moi-
eties and the oxidized sulfur species (sulfoxides and sulfonate) on
NSGA-IV upon exposure to acidic environments (Su et al., 2015;
Rauf et al., 2016). As a consequence, the adsorption of AO on the
photocatalyst surface is hindered by electrostatic repulsion, thus
potentially compromising the photocatalytic activity of our mate-
rial. Conversely, with rising pH, deprotonation of the concerned N
and S functionalities allows the cationic pollutants to be readily
adsorbed onto the positively charged surface of NSGA-IV through
electrostatic attraction, resulting in higher photocatalytic perfor-
mance. The plausibility of this explanation is corroborated by
measuring the surface zeta potentials as a function of pH (Fig. S14).
The zeta potential of NSGA-IV became more negative with
increasing pH, confirming that higher pH is more favorable for
adsorption of the cationic AO dye and its subsequent
photodegradation.
It is well documented that various reactive oxygen species
(ROS), such as singlet molecular oxygen (1O2), hydroxyl radicals
(,OH) and superoxide anions (O,
2 ), are formed when the photo-
induced electrons (e‒) and holes (hþ) interact with surface adsor-
bed O2, and these active species in turn play a crucial role in the
actual oxidation mechanism of photocatalytic reactions (Nosaka
and Nosaka, 2017). Therefore, in order to assess the individual
contributions of hþ, 1O2, O,
,
2 , and OH to the AO photodegradation
process, their respective quenchers or scavengers were separately
spiked into the reaction medium. The results thus obtained are
summarized in Fig. S15. When the hole sacrificial agent potassium
iodide (KI) was added, the AO photodecolorization efficiency was
virtually unaffected, confirming that holes were not the main active
agent in our system. Upon introducing the selective 1O2 radical
scavenger sodium azide (NaN3), the degradation percentage only
declined slightly. On the other hand, the removal rate of AO drop-
ped to 93.36% in the presence of isopropyl alcohol (IPA), a well-
known specific quencher for ,OH radicals. These findings suggest
that the aqueous phase 1O2 and ,OH radicals were not the key
factors affecting the visible light-induced photocatalytic decom-
position of AO. In the meantime, the photodegradation of AO was
greatly suppressed by the addition of the O,
2 sacrificial reagent
benzoquinone (BQ) into the reaction system, implying that O,
2
anions were the primary oxidant governing the photocatalytic
 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353
oxidation of AO over NSGA-IV under visible light irradiation. On the
basis of these observations, the plausible pathways for the photo-
degradation of AO can be proposed as follows:
AO þ hv / AO*
AO* þ NSGA
IV / AOþ þ NSGA
IV ðe

CB Þ
NSGA
IV þ hv / NSGA
IV ðhþ
VB þ e

CB Þ


O2 þ e

CB / O2
O

2 þ hþ
VB / 1
O2
H2 O þ hþ
VB / OH


þ Hþ
hþ þ OH
/ 
OH
VB
O
 1
2 ; OH; O2 þ AOþ / CO2 þ H2 O
þ Mineralization Products
Initially, the surface adsorbed dye molecules are promoted to an
electronically excited state upon exposure to visible light (eq. (1)).
Immediately after photoexcitation, the electrons (e‒) in excited
states are quickly injected into the conduction band (CB) of NSGA-
IV (eq. (2)). Concurrently, by absorbing photons having energies
equal to or greater than the band gap of NSGA-IV, some electrons
(e‒) from the valence band (VB) are photoexcited to the CB (eq. (3)).
This direct interband transition leaves an equal number of holes
(hþ) in the VB (eq. (3)). Thereafter, the photoinduced electrons at
the CB reduce the surface adsorbed oxygen molecules to O,
2 rad-
icals (eq. (4)). Likewise, the photogenerated holes oxidize the sur-
face bound water molecules, or OH
groups to ,OH radicals (eqs. (6)
and (7)). Alongside, a fraction of the produced O,
2 is converted to
1
O2 through direct oxidization by VB holes (as in eq. (5)). Eventu-
ally, due to their strong oxidizing ability, these ROS (O,
,
2 , OH, and
1
O2) cleave the dye molecules, disintegrate them into a range of
transient aromatic intermediates, and finally mineralize them into
H2O, CO2 and other inorganic anions (eq. (8)). However, the precise
estimation of the lifetime of the photogenerated radicals as well as
identification of the transient reaction intermediates and end
products is imperative in order to validate the proposed
mechanism.
Furthermore, the toxicity of the photodegradation products of
AO was evaluated by monitoring the viability of the Gram negative
bacterium Escherichia coli (E. coli) using the conventional plate
counting technique. The untreated dye solution (25 mg L
1)
induced obvious toxicity to E. coli, as evident from the very low cell
viability (~18%) depicted in Fig. S16. On the contrary, the survival
rate of E. coli remained almost unaltered (more than 95%) in the
presence of the photodegradation end products (Fig. S16). There-
fore, it is reasonable to infer that our as-developed NSGA-IV could
effectively transform recalcitrant organic pollutants into harmless
compounds, and hence warrants further consideration for the
treatment of dye bearing wastewaters.
Apart from exceptional photocatalytic activity, the easy recovery
and repeated use of the spent photocatalyst without substantial
loss in catalytic ability are essential for its prolonged cyclic opera-
tion, which in turn has a direct influence on the total cost of the
treatment process. We, therefore, performed three sequential
photocatalytic runs toward the degradation of AO under identical
experimental conditions to explore the recyclability of our NSGA-IV
351
material. As can be seen in Fig. S17, the AO degradation efficiency
reached over 99% after the first run. During the second run, the
same photocatalyst could give as much as 92% decolorization,
whereas the third run resulted in almost 88% AO removal. The
(1) observed drop in photocatalytic efficiency can be attributed to the
accumulation of oxidation byproducts on the surface of the catalyst
with time. Nevertheless, the NSGA-IV photocatalyst could still
(2)
retain at least 87% of its initial activity even after three catalytic
cycles, demonstrating its long-term durability and good recycla-
(3) bility for continuous operation.
(4)
4. Conclusion
(5) In summary, extremely lightweight, yet mechanically resilient,
N/S codoped GAs have been developed through a facile and scalable
(6) one-pot hydrothermal method. The as-prepared monolithic NSGAs
possessed continuously interconnected meso/macroporous struc-
tures with large specific surface area (107e132 m2 g
1) and abun-
(7)
dant pore space (0.19e0.25 cm3 g
1). Due to these beneficial
structural features, our freestanding NSGAs exhibited outstanding
absorption properties and could absorb hydrocarbons (e.g., engine
(8) oil, diesel oil, pump oil, soybean oil and kerosene) and different
types of organic solvents (such as acetone, n-hexane, phenol and
toluene) up to several times their own weight, and may therefore
be considered for industrial chemical and oil spill cleanup.
Furthermore, the NSGAs displayed remarkable visible light-driven
photocatalytic activities. Particularly, the aerogel with relatively
low concentrations of N and S atoms in its graphene core (i.e.,
NSGA-IV) exhibited the highest photocatalytic efficiency for the
degradation of recalcitrant organic pollutants such as AO. Nearly
complete mineralization of AO with over 84% TOC removal was
observed following exposure to visible light for 5 h. Moreover,
compared to the parent compound, the oxidation byproducts were
fairly innocuous, as inferred from toxicity assessment using E. coli
as a model organism. The superior photocatalytic performance
could be ascribed to the synergistic effect produced by N/S
codoping and 3D porous networks of interlinked 2D graphene
nanosheets. Additionally, the photocatalytic mechanism was clari-
fied by employing scavengers or quenchers of reactive oxygen and/
or radical species, which revealed that superoxide anions played a
key role in the surface-mediated photocatalytic reactions. Ulti-
mately, our preliminary findings explicitly demonstrate that the
coupling of N and S dopants with different geometrical configura-
tions in the graphene lattice synergistically enhance the application
potential of GAs for addressing anthropogenic environmental
degradation.
Acknowledgements
This work was financially supported by the National Research
Foundation (NRF), Prime Minister's Office, Republic of Singapore
through its Campus for Research Excellence and Technological
Enterprise (CREATE) programme (Grant No. R-706-002-101-281) as
well as the National University of Singapore (Grant No. R-302-000-
209-133). Ms. Yiqun Jiang further acknowledges the financial sup-
port provided by the China Scholarship Council, Ministry of Edu-
cation of the People's Republic of China for her Doctoral study.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.envpol.2019.04.132.
352 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353

References graphene oxide foams. Adv. Mater. 24, 4144e4150.

Nosaka, Y., Nosaka, A.Y., 2017. Generation and detection of reactive oxygen species
in photocatalysis. Chem. Rev. 117, 11302e11336.
Calvo, A., Yameen, B., Williams, F.J., Azzaroni, O., Soler-Illia, G.J.A.A., 2009. Facile
Ntholeng, N., Mojela, B., Gqoba, S., Airo, M., Govindraju, S., Moloto, M.J., Van Wyk, J.,
molecular design of hybrid functional assemblies with controllable transport
Moloto, N., 2016. Colloidal synthesis of pure CuInTe2 crystallites based on the
properties: mesoporous films meet polyelectrolyte brushes. Chem. Commun.
HSAB theory. New J. Chem. 40, 10259e10266.
2553e2555.
Qian, Y., Ismail, I.M., Stein, A., 2014. Ultralight, high-surface-area, multifunctional
Chen, J.-F., Mao, Y., Wang, H.-F., Hu, P., 2016. Theoretical study of heteroatom doping
graphene-based aerogels from self-assembly of graphene oxide and resol.
in tuning the catalytic activity of graphene for triiodide reduction. ACS Catal. 6,
Carbon 68, 221e231.
6804e6813.
Qie, L., Chen, W., Xiong, X., Hu, C., Zou, F., Hu, P., Huang, Y., 2015. Sulfur-doped
Chowdhury, S., Jiang, Y., Muthukaruppan, S., Balasubramanian, R., 2018. Effect of
carbon with enlarged interlayer distance as a high-performance anode material
boron doping level on the photocatalytic activity of graphene aerogels. Carbon
for sodium-ion batteries. Adv. Sci. 2, 1500195.
128, 237e248.
Rahmani, Z., Rashidi, A.M., Kazemi, A., Samadi, M.T., Rahmani, A.R., 2018. N-doped
Chowdhury, S., Pan, S., Balasubramanian, R., Das, P., 2019. Three-dimensional
reduced graphene oxide aerogel for the selective adsorption of oil pollutants
graphene-based macroscopic assemblies as super-absorbents for oils and
from water: isotherm and kinetic study. J. Ind. Eng. Chem. 61, 416e426.
organic solvents. In: Naushad, M. (Ed.), A New Generation Material Graphene:
Rauf, M., Zhao, Y.-D., Wang, Y.-C., Zheng, Y.-P., Chen, C., Yang, X.-D., Zhou, Z.-Y.,
Applications in Water Technology. Springer, Cham, pp. 43e68.
Sun, S.-G., 2016. Insight into the different ORR catalytic activity of Fe/N/C be-
Cong, H.-P., Ren, X.-C., Wang, P., Yu, S.-H., 2012. Macroscopic multifunctional
tween acidic and alkaline media: protonation of pyridinic nitrogen. Electro-
graphene-based hydrogels and aerogels by a metal ion induced self-assembly
chem. Commun. 72, 71e74.
process. ACS Nano 6, 2693e2703.
Ren, R.-P., Li, W., Lv, Y.-K., 2017. A robust, superhydrophobic graphene aerogel as a
Du, X., Liu, H.-Y., Mai, Y.-W., 2016. Ultrafast synthesis of multifunctional N-doped
recyclable sorbent for oils and organic solvents at various temperatures.
graphene foam in an ethanol flame. ACS Nano 10, 453e462.
J. Colloid Interface Sci. 500, 63e68.
Fang, Q., Shen, Y., Chen, B., 2015. Synthesis, decoration and properties of three-
Saana Amiinu, I., Zhang, J., Kou, Z., Liu, X., Asare, O.K., Zhou, H., Cheng, K., Zhang, H.,
dimensional graphene-based macrostructures: a review. Chem. Eng. J. 264,
Mai, L., Pan, M., Mu, S., 2016. Self-organized 3D porous graphene dual-doped
753e771.
with biomass-sponsored nitrogen and sulfur for oxygen reduction and evolu-
Fantauzzi, M., Elsener, B., Atzei, D., Rigoldi, A., Rossi, A., 2015. Exploiting XPS for the
tion. ACS Appl. Mater. Interfaces 8, 29408e29418.
identification of sulfides and polysulfides. RSC Adv. 5, 75953e75963.
Serp, P., Machado, B., 2015. Nanostructured Carbon Materials for Catalysis. The
Gao, M.-R., Chan, M.K.Y., Sun, Y., 2015. Edge-terminated molybdenum disulfide with
Royal Society of Chemistry, Cambridge.
a 9.4-Å interlayer spacing for electrochemical hydrogen production. Nat.
Shen, Y., Fang, Q., Chen, B., 2014. Environmental applications of three-dimensional
Commun. 6, 7493.
graphene-based macrostructures: adsorption, transformation, and detection.
Ge, J., Zhao, H.-Y., Zhu, H.-W., Huang, J., Shi, L.-A., Yu, S.-H., 2016. Advanced sorbents
Environ. Sci. Technol. 49, 67e84.
for oil-spill cleanup: recent advances and future perspectives. Adv. Mater. 28,
Su, Y., Zhang, Y., Zhuang, X., Li, S., Wu, D., Zhang, F., Feng, X., 2013. Low-temperature
10459e10490.
synthesis of nitrogen/sulfur co-doped three-dimensional graphene frameworks
Gorgolis, G., Galiotis, C., 2017. Graphene aerogels: a review. 2D Mater. 4, 032001.
as efficient metal-free electrocatalyst for oxygen reduction reaction. Carbon 62,
Hayase, G., Kanamori, K., Fukuchi, M., Kaji, H., Nakanishi, K., 2013. Facile synthesis of
296e301.
marshmallow-like macroporous gels usable under harsh conditions for the
Su, S., Wei, J., Zhang, K., Qiu, J., Wang, S., 2015. Thermo- and pH-responsive fluo-
separation of oil and water. Angew. Chem. Int. Ed. 52, 1986e1989.
rescence behaviors of sulfur-functionalized detonation nanodiamond-poly(N-
Huang, H., Zhu, J., Zhang, W., Tiwary, C.S., Zhang, J., Zhamg, X., Jiang, Q., He, H.,
isopropylacrylamide). Colloid Polym. Sci. 293, 1299e1305.
Wu, Y., Huang, W., Ajayan, P.M., Yan, Q., 2016. Controllable codoping of nitrogen
Sui, Z.-Y., Meng, Y.-N., Xiao, P.-W., Zhao, Z.-Q., Wei, Z.-X., Han, B.-H., 2015. Nitrogen-
and sulphur in graphene for highly efficient Li-oxygen batteries and direct
doped graphene aerogels as efficient supercapacitor electrodes and gas adsor-
methanol fuel cells. Chem. Mater. 28, 1737e1745.
bents. ACS Appl. Mater. Interfaces 7, 1431e1438.
Jiang, Y., Chowdhury, S., Balasubramanian, R., 2017. Nitrogen-doped graphene
Tian, W., Zhang, H., Duan, X., Sun, H., Tade, M.O., Ang, H.M., Wang, S., 2016. Ni-
hydrogels as potential adsorbents and photocatalysts for environmental
trogen- and sulfur-codoped hierarchically porous carbon for adsorptive and
remediation. Chem. Eng. J. 327, 751e763.
oxidative removal of pharmaceutical contaminants. ACS Appl. Mater. Interfaces
Jiang, Y., Chowdhury, S., Balasubramanian, R., 2019. New insights into the role of
8, 7184e7193.
nitrogen-bonding configurations in enhancing the photocatalytic activity of
Upare, P.P., Yoon, J.-W., Kim, M.Y., Kang, H.-Y., Hwang, D.W., Hwang, Y.K., Kung, H.H.,
nitrogen-doped graphene aerogels. J. Colloid Interface Sci. 534, 574e585.
 ski, W., Szala, M., Bystrzejewski, M., 2014. Sulfur-doped porous carbons: syn- Chang, J.-S., 2013. Chemical conversion of biomass-derived hexose sugars to
Kicin
levulinic acid over sulfonic acid-functionalized graphene oxide catalysts. Green
thesis and applications. Carbon 68, 1e32.
Chem. 15, 2935e2943.
Korhonen, J.T., Kettunen, M., Ras, R.H.A., Ikkala, O., 2011. Hydrophobic nanocellulose
Wang, H., Maiyalagan, T., Wang, X., 2012a. Review on recent progress in nitrogen-
aerogels as floating, sustainable, reusable, and recyclable oil absorbents. ACS
doped graphene: synthesis, characterization, and its potential applications.
Appl. Mater. Interfaces 3, 1813e1816.
ACS Catal. 2, 781e794.
Kotal, M., Kim, J., Oh, J., Oh, I.-K., 2016. Recent progress in multifunctional graphene
Wang, J., Shi, Z., Fan, J., Ge, Y., Yin, J., Hu, G., 2012b. Self-assembly of graphene into
aerogels. Front. Mater. 3, 29.
three-dimensional structures promoted by natural phenolic acids. J. Mater.
Kusano, D., Emori, M., Sakama, H., 2017. Influence of electronic structure on visible
Chem. 22, 22459e22466.
light photocatalytic activity of nitrogen-doped TiO2. RSC Adv. 7, 1887e1898.
Wang, X., Sun, G., Routh, P., Kim, D.-H., Huang, W., Chen, P., 2014. Heteroatom-
Li, A., Sun, H.-X., Tan, D.-Z., Fan, W.-J., Wen, S.-H., Qing, X.-J., Li, G.-X., Li, S.-Y.,
doped graphene materials: syntheses, properties and applications. Chem. Soc.
Deng, W.-Q., 2011. Superhydrophobic conjugated microporous polymers for
Rev. 43, 7067e7098.
separation and adsorption. Energy Environ. Sci. 4, 2062e2065.
Wang, Y., Zhang, B., Xu, M., He, X., 2015. Tunable ternary (P, S, N)-doped graphene as
Li, Z., Bommier, C., Chong, Z.S., Jian, Z., Surta, T.W., Wang, X., Xing, Z., Neuefeind, J.C.,
an efficient electrocatalyst for oxygen reduction reaction in an alkaline medium.
Stickle, W.F., Dolgos, M., Greaney, P.A., Ji, X., 2017. Mechanism of Na-ion storage
RSC Adv. 5, 86746e86753.
in hard carbon anodes revealed by heteroatom doping. Adv. Energy Mater. 7,
Wang, J., Duan, X., Dong, Q., Meng, F., Tan, X., Liu, S., Wang, S., 2019. Facile synthesis
1602894.
denas, M.A., Cazorla-Amoro  s, D., Linares-Solano, A., 2005. Behaviour of of N-doped 3D graphene aerogel and its excellent performance in catalytic
Lillo-Ro
degradation of antibiotic contaminants in water. Carbon 144, 781e790.
activated carbons with different pore size distributions and surface oxygen
Wu, Y., Zhu, J., Huang, L., 2019. A review of three-dimensional graphene-based
groups for benzene and toluene adsorption at low concentrations. Carbon 43,
materials: synthesis and applications to energy conversion/storage and envi-
1758e1767.
ronment. Carbon 143, 610e640.
Liu, Z., Li, D., Dai, H., Huang, H., 2017. Enhanced properties of tea residue cellulose
Xing, L.-B., Xi, K., Li, Q., Su, Z., Lai, C., Zhao, X., Vasant Kumar, R., 2016. Nitrogen,
hydrogels by addition of graphene oxide. J. Mol. Liq. 244, 110e116.
sulfur-codoped graphene sponge as electroactive carbon interlayer for high-
Lόpez, R., Gόmez, R., 2012. Band-gap energy estimation from diffuse reflectance
energy and -power lithiumesulfur batteries. J. Power Sources 303, 22e28.
measurements on solegel and commercial TiO2: a comparative study. J. Sol. Gel
Yang, Z., Yao, Z., Guifa, L., Fang, G., Nie, H., Liu, Z., Zhou, X., Chen, X., Huang, S., 2012.
Sci. Technol. 61, 1e7.
Sulfur-doped graphene as an efficient metal-free cathode catalyst for oxygen
Luo, Y., Jiang, S., Xiao, Q., Chen, C., Li, B., 2017. Highly reusable and superhydrophobic
reduction. ACS Nano 6, 205e211.
spongy graphene aerogels for efficient oil/water separation. Sci. Rep. 7, 7162.
Yang, Z., Dai, Y., Wang, S., Cheng, H., Yu, J., 2015. In situ incorporation of a S, N doped
Ma, X.-X., Dai, X.-H., He, X.-Q., 2017. Co9S8-modified N, S, and P ternary-doped 3D
carbon/sulfur composite for lithium sulfur batteries. RSC Adv. 5, 78107e78025.
graphene aerogels as high-performance electrocatalyst for both the oxygen
Yang, W., Gao, H., Zhao, Y., Bi, K., Li, X., 2016. Facile preparation of nitrogen-doped
reduction reaction and oxygen evolution reaction. ACS Sustain. Chem. Eng. 5,
graphene sponge as a highly efficient oil absorption material. Mater. Lett. 178,
9848e9857.
95e99.
Maleki, H., 2016. Recent advances in aerogels for environmental remediation ap-
Yousefi, N., Lu, X., Elimelech, M., Tufenkji, 2019. Environmental performance of
plications: a review. Chem. Eng. J. 300, 98e118.
graphene-based 3D macrostructures. Nat. Nanotechnol. https://doi.org/10.1038/
May, P., Lazzeri, M., Venezuela, P., Herziger, F., Callsen, G., Reparaz, J.S., Hoffmann, A.,
s41565-018-0325-6.
Mauri, F., Maultzsch, J., 2013. Signature of the two-dimensional phonon
Yu, K., Yang, S., Liu, C., Chen, H., Li, H., Sun, C., Boyd, S.A., 2012. Degradation of
dispersion in graphene probed by double-resonant Raman scattering. Phys. Rev.
organic dyes via bismuth silver oxide initiated direct oxidation coupled with
B 87, 075402.
sodium bismuthate based visible light photocatalysis. Environ. Sci. Technol. 46,
Niu, Z., Chen, J., Hng, H.H., Ma, J., Chen, X., 2012. A leavening strategy to prepare
7318e7326.
 Y. Jiang et al. / Environmental Pollution 251 (2019) 344e353 353

Zhang, L., Niu, J., Li, M., Xia, Z., 2014. Catalytic mechanisms of sulfur-doped graphene
as efficient oxygen reduction reaction catalysts for fuel cells. J. Phys. Chem. C
118, 3545e3553.
Zhang, Q., Wang, H., Wang, L., Zhuang, Y., Li, W., Zhou, Y., Gu, S., Wang, X., Yang, H.,
Xu, W., 2018. Wood-inspired fabrication of polyacrylonitrile solid foam with
superfast and high absorption capacity for liquid without selectivity. ACS Appl.
Mater. Interfaces 10, 41871e41877.
Zhao, Y., Hu, C., Hu, Y., Cheng, H., Shi, G., Qu, L., 2012. A versatile, ultralight, nitrogen-
doped graphene framework. Angew. Chem. Int. Ed. 51, 11371e11375.
Zhu, Q., Pan, Q., Liu, F., 2011. Facile removal and collection of oils from water sur-
faces through superhydrophobic and superoleophilic sponges. J. Phys. Chem. C
115, 17464e17470.