Review
Graphene Materials
Graphene Oxide: From Tunable Structures to Diverse
Luminescence Behaviors
Qingsong Mei,* Bianhua Liu, Guangmei Han, Renyong Liu, Ming-Yong Han,
and Zhongping Zhang*
Since the first discovery of luminescent graphene oxide (GO), exponentially
increasing investigations on the tunable structures and surfaces for
modulating its optical properties have struggled to expand applications
in imaging, sensing, biomedical diagnostics, and so on. Here, the latest
works on reconstructing or modifying the structures and surfaces of GO
to achieve diverse luminescence are systematically reviewed, including
fluorescence, electroluminescence, and chemiluminescence. Moreover, the
fundamental difficulties of the investigations and applications of luminescent
GO nanomaterials are clarified to inspire more constructive thoughts for
expanding their application boundaries.
1. Introduction
In 2004, the appearance of atomically thin graphene nanosheets
through peeling off highly oriented pyrolytic graphite com-
pletely reversed scientists’ beliefs that single-layer carbon
atom structure was thermodynamically unstable because it
was prone to form curved structures such as fullerenes and
nanotubes under ambient conditions.[1] Due to its extraordi-
nary physical properties, this revolutionary 2D material was
then awarded the 2010 Nobel Prize in physics, and ignited
tremendous and increasing research interests in the fields of
physics and material science over the last decade.[2] Moreover,
Dr. Q. Mei
School of Food and Biological Engineering
Hefei University of Technology
Hefei, Anhui 230009, China
E-mail: qsmei@hfut.edu.cn
Prof. B. Liu, Dr. G. Han, Dr. R. Liu, Prof. M.-Y. Han, Prof. Z. Zhang
CAS Center for Excellence in Nanoscience
Institute of Intelligent Machines
Hefei, Anhui 230031, China
E-mail: zpzhang@iim.ac.cn
Prof. Z. Zhang
School of Chemistry and Chemical Engineering
Anhui University
Hefei, Anhui 230601, China
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/advs.201900855.
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co.
KGaA, Weinheim. This is an open access article under the terms of the
Creative Commons Attribution License, which permits use, distribution
and reproduction in any medium, provided the original work is properly
cited.
DOI: 10.1002/advs.201900855
Adv. Sci. 2019, 6, 1900855 1900855  (1 of 16)
www.advancedscience.com
graphene also provides chemists with
a huge imaginary space to create novel
carbon atoms–packed materials, 2D gra-
phene derivatives, such as nitrogen- or
boron-doped graphene, graphdiyne,
graphane, and so on.[3] The fascinating
chemical/physical properties of these
materials reinspire researchers with
intense interests in graphene oxide (GO),
a typical graphene derivative that actually
has been discovered by Brodie through
oxidizing flake graphite in nitric acid and
potassium chlorate as early as 1859.[4]
Like graphene, GO possesses a large rigid
π-conjugated planar structure, coupling
with carboxyl groups on the edge and hydroxyl/epoxy groups
on basal plane,[5] which provides various possibilities to recon-
struct the structure and surface of GO for the creation of new
physicochemical properties toward the potential applications in
biological imaging or sensing, drug delivery, composite mate-
rials, and luminescent devices.
Introducing oxygen-containing groups into graphene lattice
makes GO nanosheets exhibit many amazing properties, such
as excellent water dispersibility, surface modifiability, biolog-
ical compatibility, and so on.[6] The most notable is that the
infinite large sp2 domain is disrupted after oxidizing, and the
bandgap can be opened to transform the nonluminescent gra-
phene into luminescent GO. However, the epoxy and carboxyl
groups on GO usually induce nonradiative recombination of
localized electron–hole (e–h) pairs. Thus, the traditional GO
remains to keep its nonemissive property.[7] Moreover, Lu et al.
first reported the highly efficient fluorescence-quenching
ability of GO to a variety of fluorophores by the energy transfer
mechanism,[8,9] which has been widely used to design fluores-
cent sensors for the detection and imaging of analytes.[10–18]
This universal sensing platform should be attributed to its
rigid π-conjugated planar structures for easily anchoring flu-
orescence donors and detection targets, and its widespread
absorbance spanning from visible to near-infrared (NIR)
region.[19,20]
With increasing understanding on GO, various methods
have been developed to appropriately modulate its struc-
tures, and meanwhile, the diverse luminescence behaviors
are also observed by photon or electron excitation. Typically,
chemical modification may break its large π-conjugated net-
work to form numerous isolated sp2 domains on the rigid
plane, and facilitate the radiative recombination of localized
e–h pairs, thus giving out various luminescence. On the other
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

hand, oxidative or reductive cutting of GO nanosheets into 0D

quantum dots (QDs) can further create discrete energy levels. Qingsong Mei received
In the past decade, our group has focused on the investiga- his Ph.D. in 2012 from
tions of structural tunability and luminescent properties of the University of Science
GO. For example, we have proposed an alkylamine modifica- and Technology of China,
tion method to greatly improve GO luminescent quantum yield under the supervision of
for the first time,[7] developed a novel peroxidation strategy Prof. Zhongping Zhang.
to enrich GO luminescent colors,[21] and recently found that After a short research
ultrastrong chemiluminescence (CL) could be initiated by period at the Institute
GO.[22] Additionally, a series of progresses on luminescent GO of Intelligent Machines,
nanomaterials and their sensing applications have been made Chinese Academy of
during these years.[23] Sciences, he moved to Hefei
In early stage of luminescent GO investigations, Loh et al. University of Technology as
published the first review that has highlighted the chemical an associate professor in 2013. His research interests
platform of GO for optical applications.[24] Later, Cao et al. focus on the design and preparation of novel graphene-
have summarized the fluorescent properties of graphene based fluorescent nanomaterials and upconversional
versus other carbon materials.[25] Additionally, the biomed- nanoparticles, as well as their applications in sensing
ical applications of graphene and GO, such as fluorescence and bioimaging.
bioimaging and cancer therapy, have been reviewed.[26]
However, the recent advances on GO urgently need a sys-
tematical summary to account for the relationships between Zhongping Zhang received
tunable structures and diverse luminescence behaviors, his Ph.D. in polymer science
and prospect the future breakthrough in fundamental and and engineering from the
applicable explorations. In this review, we first introduce University of Science and
several classical structural models of GO and the modula- Technology of China. He is
tion methods for its morphology sizes, surface function- currently a full distinguished
alities, as well as its bandgap energies, then summarize the professor at the Anhui
latest progresses on the diverse luminescent behaviors from University and the Institute of
GO and its derivatives, including photoluminescence (PL), Intelligent Machines (Chinese
electroluminescence (EL), and CL, and finally give future Academy of Sciences). His
perspectives, hoping to offer a full-scale insight to help the research interests span from
further exploration of exceptional emissive GO nanomate- molecular engineering of
rials, and expand their application boundaries. nanomaterial surfaces to nanosensors and visual assay/
imaging.

2. Structural Diversities of GO
To preferably modify GO nanostructures, it is of great impor-
tance to give a comprehensive insight into its chemical
structure. However, the structure of GO varies in size, mor-
phology, and surface/lateral groups, which depends on its
preparation protocols. In general, preparation of GO from
graphite mainly uses three typical protocols employing flake
graphite as initial material.[27,28] It can be traced back to the
year of 1859 that Brodie discovered the formation of yellow
graphitic acid after oxidizing flake graphite in nitric acid
and potassium chlorate, and found that the obtained mate-
rial could be dispersed in pure or basic water.[4,29] After that,
Staudenmaier optimized Brodie’s preparation procedure by
use of concentrated H2SO4 and HNO3, minimizing explo-
sion risks caused by the accumulation of ClO2.[30] Hummers
and Offeman developed an alternative oxidation method by
reacting graphite with a mixture of potassium permanganate
and concentrated H2SO4.[31] This procedure can be applied
on a multigram scale and becomes the most frequently
used method to prepare GO. Though others have developed
slightly modified preparation protocols, these three methods
mentioned above comprise the primary routes for synthesis
of GO.
2.1. Chemical Structures of GO
Since the first preparation by oxidizing graphite flakes, studies
on GO structural models have never been terminated.[32] Up
to date, various models have been proposed based on the
molecular dynamics simulations, first-principle calculations,
and experimental characterizations.[33] However, because GO
is not stoichiometric and nearly amorphous, its precise chem-
ical structure is still uncertain and may vary with the synthesis
methods and environment factors. In this review, we briefly
summarize the proposed models for clearly understanding
the chemical treatment protocols of luminescent GO which is
introduced hereinafter.
Originally, Hofmann and Holst proposed that oxygen atoms
were bound to carbon atoms of the hexagon layer planes by
epoxy linkages with an ideal formula of ether bonds.[34] How-
ever, this model does not contain any hydrogen atoms that
are indeed existed in GO nanosheets. Ruess then assumed
that a wrinkled carbon network is composed of trans-linked
cyclohexane chairs, and tetrahedrally coordinated carbon atoms
were bound to axial hydroxyl groups.[35] This is the first model to
account for the hydrogen content of GO nanosheets, while does
not contain CO bonds. Clauss and co-workers then modified

Adv. Sci. 2019, 6, 1900855 1900855  (2 of 16) © 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
this model with carbon double bonds, ketone,
and enolic groups, and explained that the
observed acidity of GO should be ascribed to
enolic hydroxyl species and carboxylic groups
on the GO nanosheets.[36] After stereochem-
ical reconsiderations, this model was revised
by Scholz and Boehm.[37] They assumed a
keto/enol tautomerism in GO, and the enol
form endowed the acidic sites. Szabo et al.
revised the Scholz model, and proposed a
corrugated carbon skeleton consisting of two
kinds of regions, trans-linked cyclohexane
chairs, and ribbons of flat hexagons with
carbon double bonds and functional groups
such as tertiary hydroxyl, 1.3-ether, ketone, Figure 1. A)[38] The Lerf–Klinowski structural model of GO nanosheet. Reproduced with
permission. Copyright 2006, American Chemical Society. B) An updated model proposed by
quinone, and phenol.[38]
Gao et al. Reproduced with permission.[41] Copyright 2009, Springer Nature. C) π-conjugated
Based on nuclear magnetic resonance carbon radical model of GO nanosheet. Reproduced with permission.[22] Copyright 2014,
spectroscopy (NMR) studies, Lerf and co- Wiley-VCH. D) GO nanosheets structural model with surface-bound oxidative debris.
workers depicted a GO layer as a random Reproduced with permission.[44] Copyright 2011, Wiley-VCH.
distribution of flat aromatic regions with
unoxidized benzene rings and wrinkled regions of alicyclic six-
membered rings bearing carbon double bonds, hydroxyl and
ether groups, and the sheets of GO terminated with hydroxyl
and carboxylic groups.[39] This model is named Lerf–Klinowski
model (Figure  1A), and is widely adopted in literatures.
Cai et al. also found that hydroxyl groups bonded to carbon
atom accompanying with epoxy groups bonded to neighboring
carbon atoms through NMR studies.[40] Ajayan and co-workers
considered the Lerf–Klinowski model as base structure, and
further offered experimental support for a peripheral structure
containing five- and six-membered ring lactols and possibly
an occasional 2-hydroxynaphthalic anhydride or 1,3-dihydrox-
yxanthone (Figure 1B).[41] They believed that the presence of
numerous tertiary alcohols in GO allowed for the possibility
of reacting with nearby carboxylic acids either on the same GO
sheet or an adjacent sheet to generate an ester. By use of ultra-
high-resolution transmission electron microscopy images, Zettl
and co-workers found that GO nanosheets were highly inho-
mogeneous. They believed that hydroxyls and epoxies were the
dominant functionalities because only these oxidative function-
alities would result in fully restoring graphitic character once
reduced, as other high energy peroxide functionalities would
induce CC bond breakage.[42] Additionally, based on first-prin-
ciple calculations, Li et al. proposed an unzipping mechanism,
where the epoxy groups were suggested to have a preference
for aligning in a line. The aligned epoxy groups then induced a
rupture of the underlying CC bonds.[43]
The research about chemical structures of GO does not end
there due to many arising phenomena which existing structure
models cannot explain. Recently, our group provided a novel
understanding of the reaction mechanism for synthesis of GO
and new insight into its chemical structure.[22] We believed that
a large number of π-conjugated carbon radicals were formed by
the addition of hydroxyl radicals to the carbon double bonds on
the disrupted π-network plane of GO (Figure 1C). In addition,
another recent model believed that GO consisted of two different
components: one was graphene with a low degree of oxidation,
while the other was oxidative debris (OD) of graphite which
was a polyaromatic oxide material and was produced during
Adv. Sci. 2019, 6, 1900855 1900855  (3 of 16)
www.advancedscience.com
the oxidative transformation of graphite into GO (Figure 1D).[44]
After the purification process, OD was claimed to be only par-
tially removed with about 30% of the GO weight, and the residual
OD was randomly adhered to the surface of GO nanosheets.
2.2. Morphological Diversity and Functionalization of GO
The synthesis strategies not only affect functional groups of GO
nanosheets, but also change their morphologies. Dependent
on the reacting conditions, such as starting graphite mate-
rials, types of oxidants, and sonication times, the obtained GO
could exhibit as large and flat nanosheets with a lateral size of
hundreds of nanometers, or 0D QDs with diameters of sev-
eral nanometers. After transformation of carbon double bonds
in graphite layers to various oxygenated groups, three major
structural features might be presented in the nanosheets: holes
within the basal plane caused by the evolution of carbon atoms
into gaseous CO or CO2, graphitic sp2 regions, and sp2/sp3
mixed regions.[45] Under more harsh synthesis conditions, for
example, utilization of ultrastrong oxidants, sonicating for long
times, and so on, the sp2-hybridized carbon networks will break
down into small fragments.
Furthermore, the strong π-conjugated network and abun-
dant oxygen-containing species on GO nanosheets facilitate the
modifications with various targets. DNA, peptides, conjugated
polymer chains can be adsorbed onto GO surface through non-
covalent π–π interactions or electrostatic interactions. Many
organic molecules are able to covalently conjugate with GO.
For example, carboxyl groups at the edge of GO nanosheets can
react with amino groups or hydroxyl groups through acyl reac-
tions, epoxy groups in GO planes also provide active sites for
ring-opening reactions with amino groups, adjacent dihydroxy
groups react with borate compounds.[5] These structural varia-
tions, including morphological sizes and covalent functionali-
zation of GO nanomaterials, greatly modulate their properties,
such as surface hydrophilicity, organelle targetable ability, and
especially tune the bandgap energies through controlling the
size of sp2 domains on GO nanosheets.
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

2.3. Bandgap Modulation of GO

Although the exact structure of GO remains uncertain, it is

well accepted that the bandgap is highly dependent on their
morphology sizes, oxygen coverage densities of O/C ratio and
surface functionalities, and so on. First, the numbers of con-
jugated aromatic rings increase with the size growth of GO
nanosheets, resulting in the decrease of bandgap energy. Chho-
walla and co-workers reported that the calculated gap between
the highest occupied molecular orbital (HOMO) and the lowest
unoccupied molecular orbital (LUMO) of a single benzene
ring was 7 eV, while it decreased down to 2 eV for a cluster
of 20 aromatic rings (Figure  2A).[46] Thus, the pure graphene
nanosheet is a type of zero-bandgap nanomaterial, but the gap
energy will increase after cutting into ideal graphene quantum
dots or clusters. At the typical sp2 domain size of 2.5–6 nm, the
predicted bandgap in GO can be varied from 0.58 to 0.24 eV.[47]
Second, the oxygen-containing groups, such as hydroxyl, epoxy,
and carboxyl groups, interrupt the conjugated π system of GO
nanosheets and isolate the sp2 carbon into independent gra-
phene-lattice islands. Therefore, the oxygen coverage in GO
nanosheets would influence their bandgap energies. Huang
et al. investigated the stability of GO bandgap for oxygen density
from 6.25% to 50%,[48] and found that the most stable bandgap
for the lowest calculated coverage (6.25%) was 0.109 eV and for
the highest calculated coverage (50%) was 3.004 eV (Figure 2B).
Notably, the bandgap was not a monotonic function of oxygen
density but exhibited local minimal values at O/C = 11.1%
and 25%. Third, for ideal graphene clusters, the gap energy
decreases after conjugation with various functional groups
such as atomic vacancy, hydroxyl, epoxy, and carboxyl groups.
Chien et al. found that these functional groups created distor-
tions in the aromatic rings, and these disorder-induced local-
ized states caused the absorption in the lower energy regions
based on Gaussian and the time-dependent density functional
theory simulations (Figure 2C).[49]
In brief, the precise chemical structures of GO have always
been controversial during these years, and even to this day, no
unambiguous model exists. The greatest divergence between
every model is the types and locations of oxygen-containing
groups on GO nanosheets. However, the nonstoichiometry
chemical structures, inhomogeneous morphologies, and
random distributions of functional groups on GO greatly
inspire researchers’ interests on exploring its novel optical
properties. From the viewpoint of making GO luminesce, it is
necessary to modulate its bandgap through altering its mor-
phology sizes or oxygen coverages according to its chemical
structures. Table  1 summaries the reported strategies for size
modulation or chemical modification of GO nanosheets, which
were always used to synthesize luminescent GO nanomaterials.
Figure 2.  A) Energy gap of π–π* transitions as a function of the fused
aromatic ring numbers. Reproduced with permission.[46] Copyright 2010,
Wiley-VCH. B) The bandgap of reduced GO versus the O/C ratios. The
3. Fluorescence from Diverse GO Nanosheets squares, circles, and triangles corresponded to the most stable, second
and Their Derivatives most stable, and the third most stable configurations of each O/C ratio,
respectively. Reproduced with permission.[48] Copyright 2012, American
In contrast to graphene with infinite sp2 system, GO contains Institute of Physics. C) The energy levels of ideal graphene nanosheets
a number of sp3 carbon atoms covalently bonded to oxygen, and decorated with various functional groups such as an atomic vacancy,
forming the isolated sp2 domains located at the large GO plane, hydroxyl, epoxy, and carboxyl groups. Reproduced with permission.[49]
as discussed above. Therefore, the HOMO and LUMO of GO Copyright 2012, Wiley-VCH.

Adv. Sci. 2019, 6, 1900855 1900855  (4 of 16) © 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Table 1.  Summary of the reported strategies for the size modulation or
chemical modification of GO nanosheets.
are isolated with a definite bandgap in GO nanosheets, and
the bandgap value increases with the oxygen coverage density
and their different arrangements.[49] In theory, a faint original
fluorescence with very low quantum yields can be expected in
common GO nanosheets synthesized by the classical methods,
which is attributed to the π electrons confined in localized
sp2 domains.[50] With in-depth understanding of luminescent
behaviors of GO, a new field and bright future of luminescent
GO nanomaterials have been shed by recent intensive explora-
tions on synthesis, modification, and luminescent mechanism.
3.1. Fluorescence from Pristine GO Nanosheets
Enlightened by the fluorescence of single-wall carbon
nanotubes,[51] the fluorescence of original GO was first reported
by Dai and co-workers in 2008. They developed a density
gradient ultracentrifugation method to separate ultrasmall
nano-GO (NGO) with lateral dimensions down to 10 nm,
and found that the NGO solution gave a PL emission peaked
at 570 nm when excited at 400 nm.[52] After modification
with amine-terminated branched polyethylene glycol (PEG),
the obtained NGO–PEG exhibited an emission maximum
blueshifted to 520 nm. After this initial study, many efforts
have been made to investigate the PL features of small sized
GO, which might be located at red regions. For example,
Zhang et al. have prepared small and uniform sized GO (S-GO)
sheets by using carbon nanohorns as starting materials via an
oxidative exfoliation method.[53] The obtained S-GO possessed
fluorescence peak position at 560 nm when excited at 440 nm,
but was gradually redshifted to 660 nm when excitation wave-
length increased to 600 nm. For the large sized GO nanosheets,
its original fluorescence also has been broadly reported. For
instance, Luo et al. have reported a PL emission centered around
750 nm when excited at 500 nm for both aqueous and solid GO
nanosheets.[54] Chen and Yan also reported that GO colloid gave
a NIR fluorescence with a maximum at 650 nm upon excita-
tion at 450 nm, and believed it was presumably assigned to the
radiative recombination of e–h pairs localized within small sp2
Adv. Sci. 2019, 6, 1900855 1900855  (5 of 16)
www.advancedscience.com
carbon domains embedded in sp3 matrix.[55] This unique NIR
fluorescence exhibited a reversible and sensitive response to
ionic strength and pH values, and also was used to develop a
biosensor for dopamine through photo­induced charge transfer
mechanism.[18c] Likewise, Galande et al. have found strongly
pH-dependent visible fluorescence from aqueous dispersion of
GO.[56] The GO suspension showed a broad peak centered near
668 nm under 440 nm excitation. As pH increased from acidic
toward neutral values, a monotonic decrease of this emission
peak was observed. The emission at 668 nm disappeared and
two relatively sharp peaks near 482 and 506 nm appeared after
increasing pH values.
As to the origin of GO fluorescence, its electronic band struc-
ture and the density of states near Fermi level should be consid-
ered. Around this theory, a number of related hypotheses for
luminescent mechanisms of pristine GO have been proposed.
The theoretically calculation suggested a bandgap ranging from
0 to few electron volts, which was dependent on the ratio of
carbon bonds to oxidation functional groups and their different
arrangements.[57] According to the band structure and density
of state, Eda et al. ascribed the fluorescence of the as-prepared
GO to the optical transitions from structural disorder-induced
localized states in the π–π* gap of sp2 sites.[46,48] Gurzadyan
and co-workers believed that all three kinds of functionalized
groups CO, CO, and OCOH were related to the emis-
sion.[57] They found that the dominant fluorescence was found
to originate from electronic transitions among nonoxidized
carbon regions and boundary of oxidized carbon atom regions.
To explain the strong excitation wavelength dependent
fluorescence of GO, Cushing et al. proposed a “giant red-
edge effect” mechanism.[58a] In their work, GO was chemically
treated in HNO3 and KOH solutions to create carboxyl-rich and
hydroxyl-rich GO, respectively. The carboxyl-rich GO showed
a broad emission band centered at 630 nm, but the hydroxyl-
rich GO exhibited at 500 nm. The fluorescence of primary GO
was observed to be a superposition of these two emissions.
When GO nanosheets were present in a polar solvent, the sol-
vation dynamics were slowed down to the same time scale as
the fluorescence, which broadened the fluorescence peak of
GO and redshifted up to 200 nm with increasing of excitation
wavelengths. This “giant red-edge effect” of GO disappeared in
a nonpolar solvent, leading to a narrow fluorescence peak that
was independent of the excitation wavelength. In order to fur-
ther discover what GO characteristics were responsible for the
presence and extent of the giant red-edge effect, they recently
investigated the excitation wavelength–dependent fluorescence
in a series of nitrogen-, boron-, fluorine-, and oxygen-doped
GO, and found that only the out-of-plane strain was the critical
factor in GO.[58b]
Moreover, a few researchers believed that the fluorescent
properties of GO should be ascribed to some highly oxidized
molecular species, OD, which adsorbed on GO nanosheets.
Rourke and co-workers have found that treatment of GO with
NaOH separated the material into two components: a color-
less but highly fluorescent OD and a darker nonfluorescent
material containing GO-like sheets.[59] The as-produced GO
showed a weak and broad PL, while the OD fluoresced more
intensely. Therefore, they believed that those OD adsorbents
were accountable for the optical properties of GO. However,
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Naumov et al. recently reported that the unique spectroscopic
properties of the combined system were independent on the
presence of OD through analysis of absorption and emission
spectra as well as lifetime measurements, while confirmed that
the optical properties were governed by its internal structure
rather than OD.[60]
3.2. Fluorescence from Functionalized GO Nanosheets
The oxygen-containing groups on GO nanosheets, such as
carboxyl groups or epoxy groups, always induce nonradia-
tive recombination of e–h pairs, making the emission intensi-
ties of pristine GO nanosheets very weak. Therefore, many
efforts have been devoted to improve GO luminescent effi-
ciency through chemical functionalization of GO nanosheets to
remove these sites. Our group found that butylamine-modified
GO nanosheets, through amide reaction of carboxyl groups and
ring-opening reaction of epoxy groups, showed an emission
maximum at 430 nm under the excitation wavelength of 350 nm
with a quantum yield of 13%, which was greatly enhanced
relative to the original GO nanosheets.[7] Similar PL charac-
teristics could also be observed when GO was modified with
other alkylamines such as 1,6-hexylenediamine, octylamine,
dodecylamine, and diamine-terminated poly(ethylene glycol).
Many other strategies have been developed to remove these
nonradiative recombination sites, achieving bright lumines-
cence, which are detailedly discussed in the following section.
3.2.1. Reductive/Oxidative Treatment of GO Nanosheets
Reductive treatment is always used to remove oxygen-
containing groups during fabrication of graphene from
GO nanosheets. Meanwhile, the bandgap structure of GO
nanosheets is modulated, leading to many novel optical
Figure 3.  A) Tunable PL from GO. a) Normalized PL spectra of GO suspensions after photothermal reduction treatment. b) Photographs of GO
suspensions at different reduction times. Reproduced with permission.[49] Copyright 2012, Wiley-VCH. B) Schematic illustration of ultrafast PL migration
of GO after reduction. Reproduced with permission.[61] Copyright 2013, American Chemical Society. C) Schematic illustration of direct observation of
spatially heterogeneous single-layer GO reduction kinetics. Reproduced with permission.[62] Copyright 2013, American Chemical Society.
Adv. Sci. 2019, 6, 1900855 1900855  (6 of 16)
www.advancedscience.com
properties. For instance, Eda et al. observed blue PL cen-
tered around 390 nm for thin-film samples deposited from
thoroughly exfoliated suspensions after exposure to hydra-
zine vapor.[46] By appropriately controlling the concentra-
tion of isolated sp2 clusters within the carbon–oxygen sp3
matrix through reduction treatment, the localizations of e–h
pairs were facilitated to radiative recombination, and the PL
intensity can be increased by tenfold compared to the as-
synthesized material. Moreover, it was found that bandgap
energy of GO nanosheets can be continuously tuned by pre-
cisely controlling reductive extent, thus inducing diverse
luminescent colors. Chen and co-workers studied the PL
evolution from GO to reduced GO by steady-state Xe lamp
irradiation with different exposure times.[49] Incremental
reduction times led to a gradual decrease in the red emis-
sion and a corresponding increase in the blue emission.
The emission peak shifted from 600 to 450 nm (Figure  3A).
Kikkawa and co-workers also found that the PL shifted from
predominately red emission to blue after continuous wave
xenon lamp exposure (Figure 3B).[61] By the aid of characteri-
zation of fluorescence quenching and blueshift of the single
GO sheet, McDonald et al. have directly observed the light-
induced photolysis of GO’s oxygen containing functional
groups (Figure 3C).[62] Infrared irradiative reduction also has
been used to investigate relationship between the modifica-
tion of individual oxygen functional groups and the resultant
optical properties of GO.[63] All of these studies demonstrated
that reductive treatment led to removal of oxygen-containing
groups, and creation of numerous small sp2 clusters within
the sp3 matrix. The defect-assisted localized states in GO due
to oxygen-containing groups induced longer wavelength emis-
sion, while the e–h recombination among these newly formed
sp2 clusters gave out short wavelength emission. Therefore,
the tunable fluorescence during reduction is attributed to the
variation of the relative fluorescence intensity ratios from the
heterogeneous electronic structures of GO.[48,61–63]
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Another interesting strategy is doping new luminescent
center in GO nanosheets during solvothermal reduction pro-
cess to achieve multicolored emission. Gan et al. reported that
Mn2+-bonded GO can exhibit tunable PL in the range of 400–
550 nm.[64] Authors dispersed GO and KMnO4 in deionized
water, and then transferred the supernatant to an autoclave and
heated to form Mn2+-bonded reduced GO. The PL peak posi-
tion monotonically increased from 430 to 550 nm as the excita-
tion wavelength increased from 320 to 490 nm. The intensity of
the long-wavelength PL (480–550 nm) stemmed from the intro-
duction of Mn2+ was enhanced more than the blue emission at
430 nm, which was essentially a quantum confinement effect
related to the size of the sp2 clusters. Since Mn2+ was bonded
to the sp2 clusters, the short distance and energy level overlap
led to effective energy transfer from Mn2+ to the sp2 clusters. As
a result, the radiative recombination rate was increased signifi-
cantly and the PL was consequently enhanced.
Many oxidative protocols also have been reported for tuning
the bandgap structure of GO nanosheets to achieve diverse
fluorescence colors. Our group recently reported a novel
method to synthesis multicolored emissive GO nanosheets.[21]
We found that the emissions of GO can be tuned from dark
brown to cyan with emission peaks shifting from 590 to
490 nm, simply through modulating oxidation times in piranha
solution (Figure 4A). This phenomenon was ascribed to atomic
oxygen generated in piranha solution, which could convert
carbon double bond in aromatic rings of GO nanosheets into
epoxy groups, then hydroxyl groups or carbonyl pairs. There-
fore, with elongating reaction time, the size of sp2 carbon
cluster decreased, leading to the extension of π→π* energy
gap and the emissions blueshift. GO obtained from single
layer graphene sheet by short time treatment with oxygen
plasma could also exhibit bright emission with a broad band
centered around 700 nm (Figure 4B).[65] Intense laser excitation
with power exceeding 1 mW would lead to the attenuation of
emission intensities (position 3 in Figure 4B). For multilayer
graphene sheets, no luminescence could be observed following
this treatment, which because only the topmost layer was
etched by oxygen plasma and the formed emission could be
immediately quenched by subjacent untreated graphene layers.
With the help of redox reaction using a solid electrolyte thin
film, the bandgap of GO can be in situ and nonvolatile tuned
by simply applying direct current (DC) voltage within an all-
solid-state device (Figure 4C).[66] The PL peak wavelength can
be tuned from 393 to 712 nm by adjusting the applied DC
voltage between −3.5 and 2.5 V (Figure 4D), where the polarity
of voltage was defined as positive when GO was oxidized and
negative when reduced. For the pristine GO, a broad emission
peak at 676 nm was observed. This red emission became weak
and another peak appeared at 544 nm upon a DC bias of −2.5 V.
Upon subsequent application of an opposite DC bias voltage
(1.5 V), an electrochemical oxidation from reduced GO to GO
was caused, and the PL peaks can shift from 544 to 690 nm.
3.2.2. Fluorescence from Fluorophore-Conjugated GO Nanosheets
Direct conjugation of fluorophores onto GO nanosheets always
failed to endow it fluorescent ability because of resonance
energy transfer mechanism. However, a few ingenious designs
could anchor the fluorescein on GO nanosheets and effectively
prevent fluorescence quenching. For example, Huang and co-
workers have reported that PEG2000 was covalently grafted to
hydroxyl and carboxylic groups of GO.[67] Fluorescein (Fluo)
was then activated and further conjugated to PEG-modified
GO (Fluo-G). PEG served as a bridge between fluorescein and

Figure 4.  A) Schematic illustration for the preparation of multicolored GO nanosheets by reacting with atomic oxygen. Reproduced with permission.[21]
Copyright 2016, American Chemical Society. B) Confocal PL image for a graphene nanosheet after treating with oxygen plasma for 5 s. Reproduced with
permission.[65] Copyright 2009, American Chemical Society. C) Schematic illustration of GO-based all-solid-state electric double-layer transistor. G, S,
and D denote the gate, source, and drain electrodes. Reproduced with permission.[66a] Copyright 2014, Wiley-VCH. D) Direct current bias dependence of
normalized PL spectra tuned by application of various bias voltages for 1200 s. Reproduced with permission.[66b] Copyright 2015, American Chemical Society.

Adv. Sci. 2019, 6, 1900855 1900855  (7 of 16) © 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

GO sheets, and prevented fluorescence quenching. Compared
with the spectrum of free fluorescein, a redshift of 5 nm was
observed in the spectrum of Fluo-G, suggesting the successful
conjugation. The approach of direct binding of semiconductors
QDs to GO also failed to make it fluoresce.[68] To overcome this
problem, Hu et al. modified GO by the surface adsorption of an
amphiphilic polypeptide, poly(l-lysine), which had functions of
improving dispersion and endowing good adhesiveness to GO
surface.[69a] After overnight incubation, QDs were adsorbed to
the polypeptide–GO. The obtained QD–GO nanocomposites
were highly stable and showed a bright fluorescent image from
a glass slide that contained a spread mixture of three differently
colored QD–GO samples under irradiation from a single UV
source. Additionally, the PL of QD–GO nanocomposites could
be used for in vivo imaging of internal tissues in small animals.
Zhu et al. used NGO as a surface modification agent to encap-
sulate a type of aggregation-induced emission nanoparticle
(NP) (TPE-TPA-FN, TTF).[69b] The surface modification of NGO
not only increased the emission efficiency of TTF nanoparti-
cles, but also improved their water dispersibility. With the help
of high nonlinear optical efficiency of TTF, the obtained TTF–
NGO NPs could emit bright three-photon luminescence, and
were further used to image the architecture of blood vessels in
mice ears and the distribution of nanoparticles in zebrafish.
3.3. Fluorescence from GO Quantum Dots
In order to effectively modulate bandgap structure to achieve
diverse fluorescence of GO, another promising approach is to
cut and exfoliate the nanosheets into smaller and thinner 0D
QDs, because this variation can greatly increase the number
and density of isolated sp2 domains. On the other hand, the
small size and high water dispersibility of GO QDs can also
greatly promote its cellular entrance for biological applications.
The fluorescent GO QDs reviewed herein mainly focused on the
materials obtained from GO nanosheets by means of reduction
cutting or oxidation cutting. Novel luminescent phenomena
different from traditional GO nanosheets are associated with the
quantum confinement and edge effects occurring in GO QDs.
3.3.1. Reductive Cleavage of GO Nanosheets into QDs
To the best of our knowledge, top-down method is widely used
to prepare GO-based QDs, such as solvothermal means, elec-
trochemical cutting methods, and so on. Pan et al. reported a
simple hydrothermal approach for the cutting of micrometer-
sized GO sheets into surface functionalized QDs with the
average diameter of 9.6 nm for the first time.[70] The obtained
QDs emitted bright blue luminescence at 430 nm with a
quantum yield of 6.9% when excited at 320 nm. They proposed
that the epoxy groups and carbonyl groups in GO nanosheets
broke up during the hydrothermal deoxidization process to
form QDs eventually, and generate free zigzag sites with a
carbene-like triplet ground state accounting for the strong blue
fluorescence (Figure 5A). Kim et al. also used hydrothermal cut-
ting method to prepared a series of particular sized GO QDs for
studying relationship between edge shapes and fluorescence
properties.[71] They reported that these QDs appeared as the
circular shape with mixed edges of zigzag and armchair when
its diameter was lower than 17 nm, but as the polygonal shape
mostly with armchair edges for diameters larger than 17 nm.
With varying the average size of QDs from 5 to 35 nm, the vis-
ible PL peak energy showed a minimum when the diameter was
17 nm. This unusual PL behaviors were ascribed to the circular-
to-polygonal-shape and the corresponding edge-state varia-
tions as the QD sizes increased. Besides solvothermal cutting

Figure 5.  A) Hydrothermal reduction of GO nanosheets into blue luminescent graphene QDs. a) Mechanism for the hydrothermal cutting of GO into
GQDs. b) Models of the GQDs in acidic (right) and alkali (left) media. c) Typical electronic transitions of triple carbenes at zigzag sites. Reproduced
with permission.[70] Copyright 2010, Wiley-VCH. B) Optically tunable amino-functionalized graphene QDs. a) Emission images of the obtained GQDs
(upper) and GQD@polymer hybrids (bottom) under irradiation with a 365 nm UV lamp. b) PL and selected UV–vis absorption spectra of the obtained
GQDs. Reproduced with permission.[73] Copyright 2012, Wiley-VCH. C) Scheme for preparation of GQDs and GQDs–NHR from GO. Reproduced with
permission.[74a] Copyright 2013, American Chemical Society.

Adv. Sci. 2019, 6, 1900855 1900855  (8 of 16) © 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
method, photolysis reduction could also cut GO nanosheets
into QDs. Sun et al. used isopropanol as the reducing agent
assisted with UV irradiation to prepare bright blue lumines-
cent GO-based QDs.[72] When irradiated by UV light, (CH3)2C•
was yielded from isopropanol, which could reduce the oxidative
groups on GO nanosheets to realize photochemical reduction
and cutting. These studies demonstrated that the disorder
states induced by oxygen-containing groups decreased and
large number of isolated sp2 domains formed after reductive
cutting, and the e–h recombination among these sp2 cluster-
like states greatly enhanced the fluorescence.
Despite many efforts have been made for the synthesis of
GO-based QDs, most of them exhibit similar blue fluorescence
with the maximum emission peak centered around 420 nm. It
is intriguing if the bandgap energy can be continuously modu-
lated to achieve multicolored emission after cutting into QDs.
Tetsuka et al. recently reported a novel hydrothermal pathway
for graphene-based QDs (GQDs) from GO nanosheets that was
edge terminated by a primary amine, allowing the electronic
structure to be modified through the effective orbital resonance
of amine moieties with graphene core.[73] The fluorescence of
GQDs could be tuned from violet to yellow region simply by
changing the initial concentration of ammonia and the tem-
perature of the amino hydrothermal treatment (Figure 5B).
A primary amine at the edge significantly altered the whole
electronic structure of the GQDs, resulting in narrowing the
optical bandgap. The resonance feature between the delocal-
ized π orbital and the molecular orbital in the amino group
was thus an origin of both the edge treatment dominated
optical tunability and the high quantum efficiency. Many other
groups also found that different amino-groups’ modification
of QDs edge structure would induce different colored fluores-
cence. For instance, Jeon and co-workers have also compared
the fluorescence of GO-based QDs with and without amine
functional groups (Figure 5C).[74] In their work, GO was first
cut by oxidation in a mixture of H2SO4/HNO3, and then further
Figure 6.  A) Scheme for the photo-Fenton reaction of GO sheets. B) The corresponding PL spectra of the GO aqueous suspensions before and
after reactions for different times. Reproduced with permission.[76] Copyright 2012, American Chemical Society. C) Scheme for oxidation cutting of
carbon fiber into GQDs. D) UV–vis spectra and corresponding fluorescent photos of GQDs after reacting at different temperatures. Reproduced with
permission.[77] Copyright 2012, American Chemical Society.
Adv. Sci. 2019, 6, 1900855 1900855  (9 of 16)
www.advancedscience.com
reduced by a N2H4 reduction process. To fabricate the amine
group–functionalized GQDs (GQD–NHR), oxidized GO was
reacted with diamine terminated polyethylene glycol (PEG-
diamine) before N2H4 reduction steps through the ring-opening
reaction of the epoxy groups. The obtained GQDs exhibited a
bluish-green color with the maximum PL position at 500 nm,
whereas the GQD–NHR exhibited a more yellowish color with
emission peak red-shifted to ≈528 nm. Wu et al. recently also
demonstrated that GQDs exhibited PL emission from blue to
yellow after a one-pot hydrothermal treatment of nanosized GO
with ammonia, aliphatic primary amines, and amino-substi-
tuted organelle targetable compounds, respectively.[75]
3.3.2. Oxidative Cutting of GO Nanosheets to QDs
Oxidative cutting is another effective method for preparation of
GO-based QDs. The formed QDs are also expected to exhibit
excellent fluorescence properties due to the quantum confine-
ment and edge effects. Zhang and co-workers demonstrated
an oxidative strategy to cut micrometer-sized GO sheets by
Fenton reagent under an UV irradiation (Figure  6A).[76] The
photo-Fenton reaction of GO was initiated at the carbon atoms
connected with oxygen containing groups, and CC bonds
were broken subsequently. The as-generated QDs have uni-
form crystallinity because the carbon atoms connected with
the hydroxyl and epoxide groups have been removed during
the photo-Fenton reaction, and assumed the strong fluores-
cence centered at 450 nm with an excitation-dependent emis-
sive property (Figure 6B). Peng et al. also have reported an acidic
etching method to cut traditional pitch-based carbon fibers into
QDs in the size range of 1–4 nm (Figure 6C).[77] The fluores-
cence of QDs can be tailored to blue, green, and yellow by var-
ying the size of QDs, which should be ascribed to the variation
in density and nature of sp2 sites (Figure 6D). Tour and co-
workers have developed a facile approach to prepare different
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

nanometer-sized QDs from various coals.[78] The GQDs derived
from bituminous coal named as b-GQDs exhibited diameters
of 2.96 ± 0.96 nm. GQDs from coke (c-GQDs) and anthracite
(a-GQDs) showed a uniform size of 5.8 ± 1.7 and 29 ± 11 nm.
The emission maxima of a-GQD, c-GQD, and b-GQD solutions
were at 530, 480, and 460 nm when excited at 345 nm, respec-
tively. Additionally, the bandgap of the GQDs synthesized from
anthracite could also be controlled by the reaction temperature
of oxidation process, resulting in the synthesis of different sized
GQDs with fluorescence properties covering the visible spec-
trum. As expected, when the GQD size increased from 4.5 to
70 nm, the peak emission was redshifted from 520 to 620 nm,
which was in accordance with the quantum confinement effect.
Li et al. have reported an alternative electrochemical approach
for direct preparation of GO-based QDs with a uniform size of
3–5 nm, which presented a green luminescence and could be
retained stable in water for several months.[79]
3.3.3. Luminescence Difference between GO-Based QDs
and Graphene QDs
The detailed investigation of fluorescence difference between
QDs directly derived from graphene (GQDs) and QDs
obtained from GO (GOQDs) can benefit the in-depth under-
standing of their luminescence mechanism. For example,
Liu et al. have successfully synthesized highly homogeneous
GQDs by shaking graphite nanoparticles in an ethanol/H2O
mixture on a vortex mixer, and GOQDs from the suspension
of GO solution after sonication and centrifugation.[80] Both
of the obtained GQDs and GOQDs exhibited circular shapes
within a 4 nm diameter, however, GOQDs possessed various
oxygenous functional groups, and GQDs had pure sp2 carbon
crystalline structure without oxygenous defects. The spectral
study showed that GQDs and GOQDs exhibited clear blue
(420 nm) and green color (480 nm) emissions, respectively,
and the fluorescence intensity of GQDs was about 3.5 times
higher than that of GOQDs, although its concentration was
about 10 times smaller than that of GOQDs (Figure  7A,B).
Yoon et al. also employed graphite as a starting material to
synthesize GOQDs, reduced GOQDs, and GQDs (Figure 7C),
and then systematically investigated their fluorescence dif-
ference.[81] It was found that the fluorescence intensity was
remarkably quenched and the emission peaks were redshifted
and broadened after increasing the oxygen concentration.
Therefore, these luminescence differences revealed that the
blue luminescence of GQDs was dominated by intrinsic states
in the high-crystalline structure, but the green luminescence

Figure 7.  A) PL spectra and the corresponding fluorescent images of GQDs and GOQDs. B) PL excitation spectra of GQDs and GOQDs with varying
detection emission wavelengths. Reproduced with permission.[80] Copyright 2013, Wiley-VCH. C) Schematic illustration of entire processes for GOQD,
reduced GOQD (RGOQD), and GQD. Reproduced with permission.[81] Copyright 2016, Wiley-VCH.

Adv. Sci. 2019, 6, 1900855 1900855  (10 of 16) © 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
of GOQDs originated from defect states with oxygenous func-
tional groups.[80,81]
4. Unusual Luminescence Behaviors from GO
Although various photoluminescence behaviors have been
revealed on GO nanosheets and QDs, these advances are
mainly based on our current knowledge of the structures and
compositions of GO, as well as the inspirations from other
luminescent material systems. Actually, GO nanosheets also
exhibit many intriguing luminescent properties, such as mul-
tiphoton luminescence, electroluminescence, and chemilu-
minescence. In this part of review, we summarized the latest
discoveries about these unusual luminescence ascribed to the
synergetic interactions of quantum confinement and surface
chemistry effect.
4.1. Multiphoton Fluorescence from GO
Since Li and co-workers reported the upconversion lumi-
nescence (UCL) phenomenon of hydrazine hydrate reduced
GO–PEG,[82] several publications have reported that the UCL
occurred in various GO derivatives. For instance, alkylamine-
modified GQDs exhibited an upconversion emission wave-
length redshift when changing excitation from 600 to 900 nm,
which was deduced to the two or multiphoton active process.[83a]
Li et al. reported a hydrothermal approach for preparation of
N-doped GQDs with a diameter of ≈1–7 nm, which also pos-
sessed excellent UCL properties.[83b] The upconverted emission
peaks shifted from 415 to 516 nm when excitation wavelengths
changed from 560 to 900 nm. The maximum PL intensity
appeared at 425 nm when excitation wavelength was 640 nm.
However, as many literatures reported, an intermediate state is
essential for the upconversion of the carriers and the excitation
light should be coherent photons from a laser. Tan et al. recently
commented that there might be mistakes in interpreting
normal luminescence excited by the second-order diffraction
light of wavelength λ/2 as evidence of UCL.[84] Gan et al. also
observed that the GQD’s emission intensity dependent on tem-
perature and excitation intensity exhibited the same features as
those from normal PL.[85] The time resolved PL confirmed that
the decay of the emission obtained with 640 nm excitation was
almost equal to the emission obtained with 320 nm excitation.
All the results indicated that the so called “UCL” under excita-
tion of a xenon lamp in GQDs was artificial, and should be con-
tributed to the second-order diffraction light of wavelength λ/2
coexisting in the excitation light of wavelength λ. In addition,
they also observed the real UCL from GQDs under excitation
of a pulsed laser.
Li et al. developed GO nanoparticles with the maximum emis-
sion peak located at 590 nm for two-photon luminescence cell
imaging.[86] GO nanoparticles were functionalized with trans-
ferrin (Trf) and PEG molecules for targeting cancer cells and sta-
bilizing the particles in cell culture buffers, respectively. A plot
of the PL intensity versus the incident power on a logarithmic
scale fitted a straight line with a slope close to two, indicating
that the PL was attributed to two-photon excitation. A low power
Adv. Sci. 2019, 6, 1900855 1900855  (11 of 16)
www.advancedscience.com
with 7 mW of an ultrafast pulsed laser was sufficient to induce
strong PL in living cells, while 30 mW was needed for Trf-modi-
fied fluorescein isothiocyanate (FITC) dyes. Almost at the same
time, He and co-workers also observed the two-photon- and
three-photon-induced PL from GO nanoparticles under femto­
second (fs) laser excitation.[87a] A two-photon luminescence
spectrum was observed in the range of 400–700 nm with the
maximum peak located at 550 nm by using a 810 nm fs pulsed
laser. Authors found that the two-photon-induced emission
intensity was proportional to the square of the fs excitation inten-
sity, confirming the two-photon process. Similarly, the three-
photon luminescence was also observed when GO sample was
stimulated by 1260 nm pulse laser. After conjugation with PEG
molecules, the GO nanoparticles were intravenously injected
into mice body from the tail vein, and observed their flow, distri-
butions, and clearance in the blood vessels by utilizing a deep-
penetrating two-photon imaging technique. Moreover, GO–PEG
nanoparticles were also microinjected into the brain of gene-
transfected mice, and showed that the nanoparticles located
at 300 µm depth in the brain could be clearly distinguished.
Recently, they also reported an upconversion luminescence of
GO nanosheets in a GO hybrid waveguide system under exci-
tation by the continuous-wave laser.[87b] Gong and co-workers
also reported that the N-doped GQDs (N-GQDs) prepared by
a facile solvothermal method using dimethylformamide as a
solvent and nitrogen source, possessed two-photon luminescent
properties.[88] A low laser power of 1 mW was sufficient to excite
strong two-photon fluorescence of N-GQD in this work. More
importantly, a large imaging depth of 1800 µm was achieved
in tissue phantom using N-GQD as two-photon probes, which
significantly extended the fundamental two-photon imaging
depth limit.
4.2. Electroluminescence of GO Derivatives
The high carrier mobility and thermal conductivity of graphene
facilitate the spatially localized accumulation of hot electrons in
an electrically biased suspended graphene layer, making gra-
phene an ideal material to serve as a nanoscale thermal radia-
tive emitter. For instance, Freitag et al. demonstrated a thermal
mid-infrared emission from electrically biased graphene on
SiO2 substrates.[89a] However, such thermal radiation was inef-
ficient, and only a small fraction of the applied energy was con-
verted into light radiation. To overcome this problem, Kim et al.
developed an electrically biased suspended graphene device,
and found that the suspended graphene channel began to emit
visible light at its center once the source–drain bias voltage
exceeded a threshold value, and the emission brightness and
area increased with bias voltage.[89b] The thermal radiation effi-
ciency was enhanced to 1000-fold, and the emitted light was
intense to be visible by the naked eye.
Except for thermal radiation, graphene also could act as an
EL material in nanoscale field-effect light-emitting diode (LED)
with tunable emission color.[90a] For example, Ren and co-
workers have demonstrated a bright spectrally tunable EL from
blue (450 nm) to red (750 nm) at the reduced GO-based field-
effect LED (GFLED), as shown in Figure 8.[90b] They explained
that the EL resulted from the recombination of Poole-Frenkel
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Figure 8.  A) Scheme for the charge injection process of GFLED. B) Bright red emission from the GFLED on a flexible polyethylene terephthalate
substrate. C) Schematic of the gate voltage-dependent EL. Inset photos were corresponding emission images from a real device. Reproduced with
permission.[90b] Copyright 2015, Springer Nature.
emission ionized electrons at the localized energy levels arising
from semireduced GO, and holes from the top of π band
(Figure 8A). The GFLED device showed a high brightness of
up to 6000 cd m−2, with efficiency around 1%, and the EL emis-
sions were continuously tuned from light blue to dark red by
electrically or chemically adjusting the Fermi levels (Figure 8C).
Under high field, the localized electrons were excited into the
lowest unoccupied discrete energy level. Radiative recombi-
nation between the thermally emitted electrons and the free
holes resulted in the EL. Additionally, graphene or graphene
quantum dots ranging from 2 to 10 nm also could be used to
prepare LED devices for achieving bright EL.[91]
Intense electrogenerated chemiluminescence (ECL) signals
of GO also have been reported by many groups. For example,
Fan et al. have reported fairly intense ECL which was over than
1.8 × 106 photon counts s−1 cm−2 from a GO aqueous suspen-
sion containing NaClO4, phosphate buffer saline, and tri-n-
propylamine (TPrA).[92] When the electrode potential was
scanned to a positive value, further oxidation of GO directly
took place either on the Pt electrode during collision or via
TPrA radical cations. The deprotonation reaction of TPrA gen-
erated a highly reductive radical intermediate and the radiative
recombination of this radical and GO led to the formation of
excited-state of GO for the generation of ECL. Likewise, Zhu and
co-workers also found the ECL phenomenon from GO-derived
GQDs.[93] They developed a facile one-pot microwave-assisted
approach for the preparation of GQDs with greenish-yellow PL
from GO nanosheets under acid conditions. Furthermore, the
GQDs were moderately reduced with NaBH4 to obtained blue-
luminescent GQDs. Except for well-known PL properties, these
two kinds of GQDs behaved intensive ECL with K2S2O8 as core-
actant. They believed that the ECL mechanism of the GQDs
was similar to that of semiconductor QDs, Si nanoclusters, and
carbon nanoclusters. First, strongly oxidizing SO4•− radicals and
GQDs•− radicals were produced by electrochemical reduction of
S2O82− and GQDs, respectively. Then, SO4•− radicals could react
Adv. Sci. 2019, 6, 1900855 1900855  (12 of 16)
www.advancedscience.com
with GQDs•− via electron-transfer annihilation, producing an
excited state that finally emitted light.
4.3. Chemiluminescence Initiated by GO
Similar as fluorescence quenching, GO also could quench
the CL through resonance energy transfer (CRET) mecha-
nism.[94] For instance, Park and co-workers recently designed
a GO-based CRET platform for homogeneous immunoassay
of C-reactive protein, which was a key marker for human
inflammation and cardiovascular diseases in human serum
samples.[95] In the CRET platform, GO played a key role
as an energy acceptor, while luminol served as a donor.[96]
Nevertheless, our group reported a novel phenomenon that the
freshly prepared GO could directly trigger the CL of luminol
without the need of additional oxidizing agents (Figure  9A).
We believed that excess H2O2 might be decomposed to ·OH
radicals in the presence of KMnO4, and ·OH radicals imme-
diately added to the double bonds at the GO plane, leading to
form π-conjugated carbon radicals (Figure 1C). As the carbon
radicals were located at the π-network plane, the single electron
was likely to conjugate with π electrons at neighboring double
bonds. The single electron was mobile among the conjugated
CC bonds, significantly prompting its oxidizing capability and
the reaction efficiency with the target molecules adsorbed on
the basal plane of GO, such as luminol molecules (Figure 9B).
In contrast to previously reported results, the π-conjugated
carbon radicals directly initiated the long-lasting, visible CL of
luminol, which was even stronger than that triggered by horse-
radish peroxidase.[22] This work provided a new insight into the
structure and properties of freshly prepared GO.
Compared with large sized GO, decreasing its lateral
dimension to less than 20 nm would greatly enhance the CL
signals. For example, Chi and co-workers recently reported
that hydrazide-modified single layer graphene quantum dots
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Figure 9.  A) The strong and long-lasting visible CL produced by addition
of luminol into freshly prepared GO solution. B) The CL mechanism
triggered by the π-conjugated carbon radicals in GO. Reproduced
with permission.[22] Copyright 2014, Wiley-VCH. C) CL mechanism of
hydrazide-modified single-layer GQDs. Reproduced with permission.[97a]
Copyright 2014, Royal Society of Chemistry.
(HM-SGQDs) prepared from GO, which had an abundant of
luminol-like units, accordingly exhibited unique and excel-
lent CL in the absence of any strong oxidant. As shown in
Figure 9C, the hydrazide group of HM-SGQDs was deproto-
nated in basic solution to form an anion.[97a] Subsequently,
the anions were chemically oxidized by the dissolved oxygen
(O2), producing a strong CL signal. The CL intensity was
mainly dependent on pH value and concentration of O2,
implying the potential applications of HM-SGQDs in pH and
O2 sensors. Gao et al. also reported that immobilization of
Adv. Sci. 2019, 6, 1900855 1900855  (13 of 16)
www.advancedscience.com
N-(aminobutyl)-N-(ethylisoluminol) (ABEI) molecules onto the
surface of GQDs demonstrated an excellent CL activity.[97b] This
novel GQD platform showed nearly 7 times higher CL emis-
sion compared to that of ABEI-functionalized GO nanosheets
under the same conditions. Authors believed that the excellent
CL property should be ascribed to the catalytic effect of GQDs,
which facilitated the generation of several radicals to produce
the excited state oxidation product of ABEI, resulting in strong
light emission.
5. Conclusion and Outlook
Recent booming advances on synthesis of GO nanomaterials
greatly promote the understanding of its chemical structures. It
has been widely recognized that GO possesses various oxygen-
containing functional groups, such as hydroxyl and epoxy
groups on the lateral plane, phenol and carboxylic groups at the
edge of nanosheets. Some specific groups, such as ketone, qui-
none, five-/six-membered ring lactols, and even π-conjugated
carbon radicals vary for every study which is highly dependent
on the synthesis conditions. This structural heterogeneity
does not hamper GO applications on materials or chemistry
fields, but endows GO with a multitude of unique properties
including excellent water dispersibility, superior biocompat-
ibility, and superior surface modifiability, particularly, its tun-
able luminescence properties. In this review, we systematically
summarized the latest works about the various luminescence
from GO and its derivatives, including PL, EL, and CL. After
various chemical treatments, such as solvothermal reactions,
reductive or oxidative cutting, surface modification, and so on,
the morphology sizes or surface functionalities of GO would be
kaleidoscopic, resulting in diverse luminescent signals. Upon
summary of these advances, it can be observed that control of
oxygen coverage (or the ratio of sp2 carbon to sp3 carbon) and
the types of oxygen-containing groups on GO nanosheets is
crucial for enhancing the luminescent efficiency. In addition,
the modulation of size and edge states (shapes or functional
groups) is useful for tuning luminescence colors of GO nano-
materials. This principle provides significant guidance for the
future design and synthesis of luminescent GO nanomaterials.
The developments of luminescent GO have been far-
reaching, however, there still remains several key challenges
to improve its luminescent properties and expand their appli-
cation boundaries. First, the nonstoichiometric chemical
structures of GO make it difficult to precisely and accurately
modulate its bandgap for obtaining specific fluorescence emis-
sions. As depicted in the above sections, these luminescent GO
nanomaterials commonly are a mixture of nanosheets with
different sizes, resulting in broad emission bands. The com-
plexity of GO samples tremendously limits their applications
when employing its PL, CL, or ECL as signal donors. There-
fore, achievement of controllably synthesis of uniform GO
nanomaterials with high quantum yields just as semiconductor
quantum dots and silicon nanocrystals, or purification their
compositions through column chromatography, would greatly
improve their luminescent properties. Second, the synthesis
protocols which most literatures reported always lead to gen-
erate blue fluorescent GO nanomaterials, while precisely tuning
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
luminescent colors to green, yellow, or red is rarely reported. A
few latest works on fluorescent carbon dots demonstrated that
the codoping of nitrogen, sulfur, or other heteroatoms in the
luminescent center would extend emissions to near-infrared
region, which may offer a novel avenue to redshift GO lumines-
cence.[98] Third, the in-depth luminescence mechanism should
be further studied by using computational simulation and
some specific photophysical measurements, such as synchro-
tron radiation technique and luminescent lifetime measure-
ments under extreme conditions, which may reveal some novel
luminescent properties, and provide significant guidance on
the synthesis of high-quality luminescent GO nanomaterials.
In addition, this achievement can also suggest some advices
on the preparation of other luminescent 2D nanosheets, such
as black phosphorus, MoS2, etc., and elucidation luminescent
mechanisms for these nanomaterials. Finally, bright EL of GO
has been widely reported, yet seldom applied in biomedical
sensing fields. Along with the rapid developments of flexible
electronics, EL-based wearable GO biomedical devices will
become a promising direction because of its intriguing opto-
electronic properties.
GO nanosheets, as a bridge of inorganic nanomaterials
and organic molecules, are attracting increasing attention due
to their intriguing physicochemical properties. It is antici-
pated that further combination of newly explored GO with
many other functionalized nanomaterials, such as other 2D
nanosheets, metal–organic frameworks, polymer nanoparticles,
upconversion nanoparticles, aggregation-induced emission
nanoparticles, and so on,[99] will provide new insights for con-
structing integrated platforms for biomedical, environmental,
or energy applications.
Acknowledgements
This work was supported by the National Natural Science Foundation of
China (Grant Nos. 21675038, 21335006, 21475135, and 21775001), the
National Basic Research Program of China (Grant No. 2015CB932002),
the Major/Innovative Program of Development Foundation of Hefei
Center for Physical Science and Technology (Grant No. 2017FXZY001),
and the Fundamental Research Funds for the Central Universities (Grant
No. JZ2019HGTB0060).
Conflict of Interest
The authors declare no conflict of interest.
Keywords
chemiluminescence, electroluminescence, fluorescence, graphene oxide,
tunable structures
Received: April 12, 2019
Revised: May 19, 2019
Published online: June 14, 2019
[1] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang,
S. V. Dubonos, I. V. Grigorieva, A. A. Firsov, Science 2004, 306,
666.
Adv. Sci. 2019, 6, 1900855 1900855  (14 of 16)
www.advancedscience.com
[2] a) V. Singh, D. Joung, L. Zhai, S. Das, S. I. Khondaker, S. Seal, Prog.
Mater. Sci. 2011, 56, 1178; b) Y. Xu, G. Shi, X. Duan, Acc. Chem.
Res. 2015, 48, 1666; c) M. Liu, R. Zhang, W. Chen, Chem. Rev. 2014,
114, 5117; d) F. Bonaccorso, Z. Sun, T. Hasan, A. C. Ferrari, Nat.
Photonics 2010, 4, 611; e) M. I. Katsnelson, Mater. Today 2007, 10,
20; f) Y. C. Jiang, Z. Y. Wang, M. Y. Shang, Z. P. Zhang, S. D. Zhang,
Nanoscale 2015, 7, 10950; g) W. Zhang, J. Wei, H. J. Zhu, K. Zhang,
F. Ma, Q. S. Mei, Z. P. Zhang, S. H. Wang, J. Mater. Chem. 2012,
22, 22631; h) S. D. Zhang, Q. H. Tao, Z. Y. Wang, Z. P. Zhang,
J. Mater. Chem. 2012, 22, 20166; i) Y. Y. Liang, Y. G. Li, H. L. Wang,
J. G. Zhou, J. Wang, T. Regier, H. J. Dai, Nat. Mater. 2011, 10, 780.
[3] a) A. Zitolo, V. Goellner, V. Armel, M. T. Sougrati, T. Mineva,
L. Stievano, E. Fonda, F. Jaouen, Nat. Mater. 2015, 14, 937;
b) L. S. Panchakarla, K. S. Subrahmanyam, S. K. Saha,
A. Govindaraj, H. R. Krishnamurthy, U. V. Waghmare, C. N. R. Rao,
Adv. Mater. 2009, 46, 4726; c) G. X. Li, Y. L. Li, H. B. Liu, Y. B. Guo,
Y. J. Li, D. B. Zhu, Chem. Commun. 2010, 46, 3256; d) D. C. Elias,
R. R. Nair, T. M. G. Mohiuddin, S. V. Morozov, P. Blake,
M. P. Halsall, A. C. Ferrari, D. W. Boukhvalov, M. I. Katsnelson,
A. K. Geim, K. S. Novoselov, Science 2009, 323, 610.
[4] B. C. Brodie, Philos. Trans. R. Soc. London 1859, 149, 249.
[5] a) J. Kim, L. J. Cote, J. X. Huang, Acc. Chem. Res. 2012, 45, 1356;
b) D. R. Dreyer, A. D. Todd, C. W. Bielawski, Chem. Soc. Rev. 2014,
43, 5288; c) D. R. Dreyer, S. Park, C. W. Bielawski, R. S. Ruoff, Chem.
Soc. Rev. 2010, 39, 228; d) A. Dimiev, D. V. Kosynkin, L. B. Alemany,
P. Chaguine, J. M. Tour, J. Am. Chem. Soc. 2012, 134, 2815.
[6] C. Cheng, S. Li, A. Thomas, N. A. Kotov, R. Haag, Chem. Rev. 2017,
117, 1826.
[7] Q. S. Mei, K. Zhang, G. J. Guan, B. H. Liu, S. H. Wang, Z. P. Zhang,
Chem. Commun. 2010, 46, 7319.
[8] C. H. Lu, H. H. Yang, C. L. Zhu, X. Chen, G. N. Chen, Angew. Chem.,
Int. Ed. 2009, 48, 4785.
[9] a) R. S. Swathi, K. L. Sebastian, J. Chem. Phys. 2008, 129, 054703;
b) R. S. Swathi, K. L. Sebastian, J. Chem. Phys. 2009, 130, 086101.
[10] H. Pei, J. Li, M. Lv, J. Y. Wang, J. M. Gao, J. X. Lu, Y. P. Li, Q. Huang,
J. Hu, C. H. Fan, J. Am. Chem. Soc. 2012, 134, 13843.
[11] H. F. Dong, W. C. Gao, F. Yan, H. X. Ji, H. X. Ju, Anal. Chem. 2010,
82, 5511.
[12] C. L. Zhang, Y. X. Yuan, S. M. Zhang, Y. H. Wang, Z. H. Liu, Angew.
Chem., Int. Ed. 2011, 50, 6851.
[13] X. Q. Liu, F. Wang, R. Aizen, O. Yehezkeli, I. Willner, J. Am. Chem.
Soc. 2013, 135, 11832.
[14] J. Balapanuru, J. X. Yang, S. Xiao, Q. L. Bao, M. Jahan, L. Polavarapu,
J. Wei, Q. H. Xu, K. P. Loh, Angew. Chem., Int. Ed. 2010, 49,
6549.
[15] Y. Wen, C. Peng, D. Li, L. Zhuo, S. He, L. Wang, Q. Huang,
Q. H. Xu, C. Fan, Chem. Commun. 2011, 47, 6278.
[16] a) B. W. Liu, Z. Y. Sun, X. Zhang, J. W. Liu, Anal. Chem. 2013,
85, 7987; b) H. Zhang, S. S. Jia, M. Lv, J. Y. Shi, X. L. Zuo, S. Su,
L. H. Wang, W. Huang, C. H. Fan, Q. Huang, Anal. Chem. 2014, 86,
4047; c) J. Liu, C. Wang, Y. Jiang, Y. Hu, J. Li, S. Yang, Y. Li, R. Yang,
W. Tan, C. Z. Huang, Anal. Chem. 2013, 85, 1424.
[17] H. Jang, S. R. Ryoo, Y. K. Kim, S. Yoon, H. Kim, S. W. Han,
B. S. Choi, D. E. Kim, D. H. Min, Angew. Chem., Int. Ed. 2013, 52,
2340.
[18] a) J. H. Jung, D. S. Cheon, F. Liu, K. B. Lee, T. S. Seo, Angew. Chem.,
Int. Ed. 2010, 49, 5708; b) C. Z. Zhu, D. Dan, Y. H. Lin, 2D Mater.
2015, 2, 032004; c) J. L. Chen, X. P. Yan, K. Meng, S. F. Wang, Anal.
Chem. 2011, 83, 8787; d) J. Zhou, X. H. Xu, W. Liu, X. Liu, Z. Nie,
M. Qing, L. H. Nie, S. Z. Yao, Anal. Chem. 2013, 85, 5746.
[19] D. Li, M. B. Muller, S. Gilje, R. B. Kaner, G. G. Wallace, Nat.
Nanotechnol. 2008, 3, 101.
[20] a) A. Wisitsoraat, J. P. Mensing, C. Karuwan, C. Sriprachuabwong,
K. Jaruwongrungsee, D. Phokharatkul, T. M. Daniels, C. Liewhiran,
A. Tuantranont, Biosens. Bioelectron. 2017, 87, 7; b) Y. Q. Yang,
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
A. M. Asiri, Z. W. Tang, D. Du, Y. H. Lin, Mater. Today 2013, 16, 365;
c) M. Pumera, Mater. Today 2011, 14, 308.
[21] Q. S. Mei, J. Chen, J. Zhao, L. Yang, B. H. Liu, R. Y. Liu, Z. P. Zhang,
ACS Appl. Mater. Interfaces 2016, 8, 7390.
[22] L. Yang, R. L. Zhang, B. H. Liu, J. P. Wang, S. H. Wang, M. Y. Han,
Z. P. Zhang, Angew. Chem., Int. Ed. 2014, 53, 10109.
[23] a) Q. S. Mei, Z. P. Zhang, Angew. Chem., Int. Ed. 2012, 51, 5602;
b) K. Zhang, L. Yang, H. J. Zhu, F. Ma, Z. P. Zhang, S. H. Wang,
Analyst 2014, 139, 2379; c) X. C. Chen, S. M. Yu, L. Yang, J. P. Wang,
C. L. Jiang, Nanoscale 2016, 8, 13669; d) Q. S. Mei, C. L. Jiang,
G. J. Guan, K. Zhang, B. H. Liu, R. Y. Liu, Z. P. Zhang, Chem.
Commun. 2012, 48, 7468.
[24] K. P. Loh, Q. L. Bao, G. Eda, M. Chhowalla, Nat. Chem. 2010, 2,
1015.
[25] L. Cao, M. J. Meziani, S. Sahu, Y. P. Sun, Acc. Chem. Res. 2013, 46,
171.
[26] a) L. Y. Feng, L. Wu, X. G. Qu, Adv. Mater. 2013, 25, 168;
b) E. Morales-Narvaez, A. Merkoci, Adv. Mater. 2012, 24, 3298;
c) P. T. Yin, S. Shah, M. Chhowalla, K. B. Lee, Chem. Rev. 2015,
115, 2483; d) J. M. Yoo, J. H. Kang, B. H. Hong, Chem. Soc. Rev.
2015, 44, 4835; e) K. Yang, L. Z. Feng, X. Z. Shi, Z. Liu, Chem. Soc.
Rev. 2013, 42, 530; f) Y. Q. Dong, Q. Wang, H. S. Wu, Y. M. Chen,
C. H. Lu, Y. W. Chi, H. H. Yang, Small 2016, 12, 5376; g) L. L. Zhang,
C. L. Jiang, Z. P. Zhang, Nanoscale 2013, 5, 3773; h) J. L. Li, B. Tang,
B. Yuan, L. Sun, X. G. Wang, Biomaterials 2013, 34, 9519.
[27] a) D. Chen, H. B. Feng, J. H. Li, Chem. Rev. 2012, 112, 6027;
b) D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii,
Z. Z. Sun, A. Slesarev, L. B. Alemany, W. Lu, J. M. Tour, ACS Nano
2010, 4, 4806; c) M. J. Allen, V. C. Tung, R. B. Kaner, Chem. Rev.
2010, 110, 132.
[28] C. Botas, P. Alvarez, C. Blanco, R. Santamaria, M. Granda, P. Ares,
F. Rodriguez-Reinoso, R. Menendez, Carbon 2012, 50, 275.
[29] H. K. Jeong, Y. P. Lee, R. J. W. E. Lahaye, M. H. Park, K. H. An,
I. J. Kim, C. W. Yang, C. Y. Park, R. S. Ruoff, Y. H. Lee, J. Am. Chem.
Soc. 2008, 130, 1362.
[30] a) L. Staudenmaier, Ber. Dtsch. Chem. Ges. 1898, 31, 1481;
b) M. J. McAllister, J. L. Li, D. H. Adamson, H. C. Schniepp,
A. A. Abdala, J. Liu, M. Herrera-Alonso, D. L. Milius, R. Car,
R. K. Prud’homme, I. A. Aksay, Chem. Mater. 2007, 19, 4396;
c) J. R. Lomeda, C. D. Doyle, D. V. Kosynkin, W. F. Hwang,
J. M. Tour, J. Am. Chem. Soc. 2008, 130, 16201.
[31] a) W. S. Hummers Jr., R. E. Offeman, J. Am. Chem. Soc. 1958, 80,
1339; b) C. K. Chua, Z. Sofer, M. Pumera, Chem. - Eur. J. 2012, 18,
13453; c) S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas,
A. Kleinhammes, Y. Jia, Y. Wu, S. T. Nguyen, R. S. Ruoff, Carbon
2007, 45, 1558.
[32] S. Mao, H. H. Pu, J. H. Chen, RSC Adv. 2012, 2, 2643.
[33] A. Bagri, C. Mattevi, M. Acik, Y. J. Chabal, M. Chhowalla,
V. B. Shenoy, Nat. Chem. 2010, 2, 581.
[34] U. Hofmann, R. Holst, Ber. Dtsch. Chem. Ges. 1939, 72, 754.
[35] G. Ruess, Monatsch. Chem. 1946, 76, 381.
[36] A. Clauss, R. Plass, H. P. Boehm, U. Hofmann, Z. Anorg. Allg. Chem.
1957, 291, 205.
[37] W. Scholz, H. P. Boehm, Z. Anorg. Allg. Chem. 1969, 369,
327.
[38] T. Szabo, O. Berkesi, P. Forgo, K. Josepovits, Y. Sanakis, D. Petridis,
I. Dekany, Chem. Mater. 2006, 18, 2740.
[39] A. Lerf, H. He, T. Riedl, M. Forster, J. Klinowski, Solid State Ionics
1997, 101, 857.
[40] W. W. Cai, R. D. Piner, F. J. Stadermann, S. Park, M. A. Shaibat,
Y. Ishii, D. X. Yang, A. Velamakanni, S. J. An, M. Stoller, J. H. An,
D. M. Chen, R. S. Ruoff, Science 2008, 321, 1815.
[41] W. Gao, L. B. Alemany, L. J. Ci, P. M. Ajayan, Nat. Chem. 2009, 1,
403.
Adv. Sci. 2019, 6, 1900855 1900855  (15 of 16)
www.advancedscience.com
[42] K. Erickson, R. Erni, Z. Lee, N. Alem, W. Gannett, A. Zettl, Adv.
Mater. 2010, 22, 4467.
[43] Z. Y. Li, W. H. Zhang, Y. Luo, J. L. Yang, J. G. Hou, J. Am. Chem. Soc.
2009, 131, 6320.
[44] J. P. Rourke, P. A. Pandey, J. J. Moore, M. Bates, I. A. Kinloch,
R. J. Young, N. R. Wilson, Angew. Chem., Int. Ed. 2011, 50, 3173.
[45] S. H. Dave, C. C. Gong, A. W. Robertson, J. H. Warner,
J. C. Grossman, ACS Nano 2016, 10, 7515.
[46] G. Eda, Y. Y. Lin, C. Mattevi, H. Yamaguchi, H. A. Chen, I. S. Chen,
C. W. Chen, M. Chhowalla, Adv. Mater. 2010, 22, 505.
[47] C. Mattevi, G. Eda, S. Agnoli, S. Miller, K. A. Mkhoyan, O. Celik,
D. Mastrogiovanni, G. Granozzi, E. Garfunkel, M. Chhowalla, Adv.
Funct. Mater. 2009, 19, 2577.
[48] H. M. Huang, Z. B. Li, J. C. She, W. L. Wang, J. Appl. Phys. 2012,
111, 054317.
[49] C. T. Chien, S. S. Li, W. J. Lai, Y. C. Yeh, H. A. Chen, I. S. Chen,
L. C. Chen, K. H. Chen, T. Nemoto, S. Isoda, M. W. Chen, T. Fujita,
G. Eda, H. Yamaguchi, M. Chhowalla, C. W. Chen, Angew. Chem.,
Int. Ed. 2012, 51, 6662.
[50] a) S. K. Pal, Carbon 2015, 88, 86; b) J. H. Shen, Y. H. Zhu, X. L. Yang,
C. Z. Li, Chem. Commun. 2012, 48, 3686; c) O. Kozak, M. Sudolska,
G. Pramanik, P. Cigler, M. Otyepka, R. Zboril, Chem. Mater. 2016,
28, 4085; d) L. L. Li, G. H. Wu, G. H. Yang, J. Peng, J. W. Zhao,
J. J. Zhu, Nanoscale 2013, 5, 4015; e) S. Vempati, T. Uyar, Phys.
Chem. Chem. Phys. 2014, 16, 21183.
[51] a) J. T. Robinson, G. S. Hong, Y. Y. Liang, B. Zhang,
O. K. Yaghi, H. J. Dai, J. Am. Chem. Soc. 2012, 134, 10664;
b) N. Nakayama-Ratchford, S. Bangsaruntip, X. M. Sun, K. Welsher,
H. J. Dai, J. Am. Chem. Soc. 2007, 129, 2448; c) G. S. Hong,
J. C. Lee, J. T. Robinson, U. Raaz, L. M. Xie, N. F. Huang,
J. P. Cooke, H. J. Dai, Nat. Med. 2012, 18, 1841.
[52] X. M. Sun, Z. Liu, K. Welsher, J. T. Robinson, A. Goodwin, S. Zaric,
H. J. Dai, Nano Res. 2008, 1, 203.
[53] M. F. Zhang, T. Okazaki, Y. Iizumi, E. Miyako, R. Yuge, S. Bandow,
S. Iijima, M. Yudasaka, J. Mater. Chem. B 2016, 4, 121.
[54] Z. T. Luo, P. M. Vora, E. J. Mele, A. T. C. Johnson, J. M. Kikkawa,
Appl. Phys. Lett. 2009, 94, 111909.
[55] J. L. Chen, X. P. Yan, Chem. Commun. 2011, 47, 3135.
[56] C. Galande, A. D. Mohite, A. V. Naumov, W. Gao, L. J. Ci,
A. Ajayan, H. Gao, A. Srivastava, R. B. Weisman, P. M. Ajayan,
Sci. Rep. 2011, 1, 85.
[57] J. Z. Shang, L. Ma, J. W. Li, W. Ai, T. Yu, G. G. Gurzadyan, Sci. Rep.
2012, 2, 792.
[58] a) S. K. Cushing, M. Li, F. Q. Huang, N. Q. Wu, ACS Nano 2014,
8, 1002; b) S. K. Cushing, W. Ding, G. Chen, C. Wang, F. Yang,
F. Huang, N. Wu, Nanoscale 2017, 9, 2240; c) P. Zheng, N. Wu,
Chem. - Asian J. 2017, 12, 2343.
[59] H. R. Thomas, C. Valles, R. J. Young, I. A. Kinloch, N. R. Wilson,
J. P. Rourke, J. Mater. Chem. C 2013, 1, 338.
[60] A. Naumov, F. Grote, M. Overgaard, A. Roth, C. E. Halbig,
K. Norgaard, D. M. Guldi, S. Eigler, J. Am. Chem. Soc. 2016, 138,
11445.
[61] A. L. Exarhos, M. E. Turk, J. M. Kikkawa, Nano Lett. 2013, 13,
344.
[62] M. P. McDonald, A. Eltom, F. Vietmeyer, J. Thapa, Y. V. Morozov,
D. A. Sokolov, J. H. Hodak, K. Vinodgopal, P. V. Kamat, M. Kuno,
Nano Lett. 2013, 13, 5777.
[63] R. Maiti, A. Midya, C. Narayana, S. K. Ray, Nanotechnology 2014, 25,
495704.
[64] Z. X. Gan, S. J. Xiong, X. L. Wu, C. Y. He, J. C. Shen, P. K. Chu, Nano
Lett. 2011, 11, 3951.
[65] T. Gokus, R. R. Nair, A. Bonetti, M. Bohmler, A. Lombardo,
K. S. Novoselov, A. K. Geim, A. C. Ferrari, A. Hartschuh, ACS Nano
2009, 3, 3963.
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
[66] a) T. Tsuchiya, K. Terabe, M. Aono, Adv. Mater. 2014, 26, 1087;
b) T. Tsuchiya, T. Tsuruoka, K. Terabe, M. Aono, ACS Nano 2015, 9,
2102.
[67] C. Peng, W. B. Hu, Y. T. Zhou, C. H. Fan, Q. Huang, Small 2010, 6,
1686.
[68] a) B. Farrow, P. V. Kamat, J. Am. Chem. Soc. 2009, 131, 11124;
b) L. L. Li, K. P. Liu, G. H. Yang, C. M. Wang, J. R. Zhang, J. J. Zhu,
Adv. Funct. Mater. 2011, 21, 869.
[69] a) S. H. Hu, Y. W. Chen, W. T. Hung, I. W. Chen, S. Y. Chen, Adv.
Mater. 2012, 24, 1748; b) Z. F. Zhu, J. Qian, X. Y. Zhao, W. Qin,
R. R. Hu, H. Q. Zhang, D. Y. Li, Z. P. Xu, B. Z. Tang, S. L. He, ACS
Nano 2016, 10, 588.
[70] D. Y. Pan, J. C. Zhang, Z. Li, M. H. Wu, Adv. Mater. 2010, 22, 734.
[71] S. Kim, S. W. Hwang, M. K. Kim, D. Y. Shin, D. H. Shin, C. O. Kim,
S. B. Yang, J. H. Park, E. Hwang, S. H. Choi, G. Ko, S. Sim, C. Sone,
H. J. Choi, S. Bae, B. H. Hong, ACS Nano 2012, 6, 8203.
[72] H. J. Sun, L. Wu, N. Gao, J. S. Ren, X. G. Qu, ACS Appl. Mater.
Interfaces 2013, 5, 1174.
[73] H. Tetsuka, R. Asahi, A. Nagoya, K. Okamoto, I. Tajima, R. Ohta,
A. Okamoto, Adv. Mater. 2012, 24, 5333.
[74] S. H. Jin, D. H. Kim, G. H. Jun, S. H. Hong, S. Jeon, ACS Nano
2013, 7, 1239.
[75] X. Wu, L. Ma, S. Sun, K. Jiang, L. Zhang, Y. Wang, H. Zeng, H. Lin,
Nanoscale 2018, 10, 1532.
[76] X. J. Zhou, Y. Zhang, C. Wang, X. C. Wu, Y. Q. Yang, B. Zheng,
H. X. Wu, S. W. Guo, J. Y. Zhang, ACS Nano 2012, 6, 6592.
[77] J. Peng, W. Gao, B. K. Gupta, Z. Liu, R. Romero-Aburto, L. H. Ge,
L. Song, L. B. Alemany, X. B. Zhan, G. H. Gao, S. A. Vithayathil,
B. A. Kaipparettu, A. A. Marti, T. Hayashi, J. J. Zhu, P. M. Ajayan,
Nano Lett. 2012, 12, 844.
[78] a) R. Q. Ye, C. S. Xiang, J. Lin, Z. W. Peng, K. W. Huang, Z. Yan,
N. P. Cook, E. L. G. Samuel, C. C. Hwang, G. D. Ruan, G. Ceriotti,
A. R. O. Raji, A. A. Marti, J. M. Tour, Nat. Commun. 2013, 4, 2943;
b) R. Q. Ye, Z. W. Peng, A. Metzger, J. Lin, J. K. Mann, K. W. Huang,
C. S. Xiang, X. J. Fan, E. L. G. Samuel, L. B. Alemany, A. A. Marti,
J. M. Tour, ACS Appl. Mater. Interfaces 2015, 7, 7041.
[79] Y. Li, Y. Hu, Y. Zhao, G. Q. Shi, L. E. Deng, Y. B. Hou, L. T. Qu, Adv.
Mater. 2011, 23, 776.
[80] F. Liu, M. H. Jang, H. D. Ha, J. H. Kim, Y. H. Cho, T. S. Seo, Adv.
Mater. 2013, 25, 3657.
[81] H. Yoon, Y. H. Chang, S. H. Song, E. S. Lee, S. H. Jin, C. Park, J. Lee,
B. H. Kim, H. J. Kang, Y. H. Kim, S. Jeon, Adv. Mater. 2016, 28, 5255.
[82] J. H. Shen, Y. H. Zhu, C. Chen, X. L. Yang, C. Z. Li, Chem. Commun.
2011, 47, 2580.
[83] a) S. J. Zhu, J. H. Zhang, S. J. Tang, C. Y. Qiao, L. Wang, H. Y. Wang,
X. Liu, B. Li, Y. F. Li, W. L. Yu, X. F. Wang, H. C. Sun, B. Yang,
Adv. Funct. Mater. 2012, 22, 4732; b) M. Li, W. B. Wu, W. C. Ren,
H. M. Chen, N. J. Tang, W. Zhong, Y. W. Du, Appl. Phys. Lett. 2012,
101, 103107.
Adv. Sci. 2019, 6, 1900855 1900855  (16 of 16)
www.advancedscience.com
[84] D. Z. Tan, S. F. Zhou, J. R. Qiu, ACS Nano 2012, 6,
6530.
[85] Z. X. Gan, X. L. Wu, G. X. Zhou, J. C. Shen, P. K. Chu, Adv. Opt.
Mater. 2013, 1, 554.
[86] J. L. Li, H. C. Bao, X. L. Hou, L. Sun, X. G. Wang, M. Gu, Angew.
Chem., Int. Ed. 2012, 51, 1830.
[87] a) J. Qian, D. Wang, F. H. Cai, W. Xi, L. Peng, Z. F. Zhu, H. He,
M. L. Hu, S. L. He, Angew. Chem., Int. Ed. 2012, 51, 10570; b) Z. Li,
J. Yang, M. Shi, L. Yang, Y. Cheng, X. Hu, X. Jiang, X. Xing, S. He,
J. Phys. Chem. C 2018, 122, 16866.
[88] Q. Liu, B. D. Guo, Z. Y. Rao, B. H. Zhang, J. R. Gong, Nano Lett.
2013, 13, 2436.
[89] a) M. Freitag, H. Y. Chiu, M. Steiner, V. Perebeinos, P. Avouris, Nat.
Nanotechnol. 2010, 5, 497; b) Y. D. Kim, H. Kim, Y. Cho, J. H. Ryoo,
C. H. Park, P. Kim, Y. S. Kim, S. Lee, Y. L. Li, S. N. Park, Y. S. Yoo,
D. Yoon, V. E. Dorgan, E. Pop, T. F. Heinz, J. Hone, S. H. Chun,
H. Cheong, S. W. Lee, M. H. Bae, Y. D. Park, Nat. Nanotechnol.
2015, 10, 676.
[90] a) M. C. Chong, N. Afshar-Imani, F. Scheurer, C. Cardoso,
A. Ferretti, D. Prezzi, G. Schull, Nano Lett. 2018, 18, 175;
b) X. M. Wang, H. Tian, M. A. Mohammad, C. Li, C. Wu, Y. Yang,
T. L. Ren, Nat. Commun. 2015, 6, 7767.
[91] W. Kwon, Y. H. Kim, C. L. Lee, M. Lee, H. C. Choi, T. W. Lee,
S. W. Rhee, Nano Lett. 2014, 14, 1306.
[92] F. R. F. Fan, S. Park, Y. W. Zhu, R. S. Ruoff, A. J. Bard, J. Am. Chem.
Soc. 2009, 131, 937.
[93] L. L. Li, J. Ji, R. Fei, C. Z. Wang, Q. Lu, J. R. Zhang, L. P. Jiang,
J. J. Zhu, Adv. Funct. Mater. 2012, 22, 2971.
[94] a) Y. Y. Su, Y. Lv, RSC Adv. 2014, 4, 29324; b) H. Chen, Q. Gao,
J. Z. Li, J. M. Lin, J. Photochem. Photobiol., C 2016, 27, 54.
[95] J. S. Lee, H. A. Joung, M. G. Kim, C. B. Park, ACS Nano 2012, 6,
2978.
[96] a) W. Shen, Y. Q. Yu, J. N. Shu, H. Cui, Chem. Commun. 2012, 48,
2894; b) S. Bi, T. T. Zhao, B. Y. Luo, Chem. Commun. 2012, 48, 106;
c) Y. He, H. Cui, J. Mater. Chem. 2012, 22, 9086.
[97] a) Y. Q. Dong, R. P. Dai, T. Q. Dong, Y. W. Chi, G. N. Chen,
Nanoscale 2014, 6, 11240; b) L. F. Gao, L. Ju, H. Cui, J. Mater.
Chem. C 2017, 5, 7753.
[98] a) Y. Xu, X. Wang, W. L. Zhang, F. Lv, S. Guo, Chem. Soc. Rev. 2018,
47, 586; b) K. Hola, M. Sudolska, S. Kalytchuk, D. Nachtigallova,
A. L. Rogach, M. Otyepka, R. Zboril, ACS Nano 2017, 11, 12402;
c) M. T. Hasan, R. Gonzalez-Rodriguez, C. Ryan, N. Faerber,
J. L. Coffer, A. V. Naumov, Adv. Funct. Mater. 2018, 28, 1804337.
[99] a) Z. Wang, Q. Jingjing, X. Wang, Z. Zhang, Y. Chen, X. Huang,
W. Huang, Chem. Soc. Rev. 2018, 47, 6128; b) S. Zhu, Y. Song,
J. Wang, H. Wan, Y. Zhang, Y. Ning, B. Yang, Nano Today 2017, 13,
10; c) C. Z. Zhu, D. Du, Y. H. Lin, Biosens. Bioelectron. 2017, 89,
43; d) M. Y. Xiong, Q. M. Rong, H. M. Meng, X. B. Zhang, Biosens.
Bioelectron. 2017, 89, 212.
© 2019 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim