Graphene-based biosensors

Sabine Szunerits and Rabah Boukherroub

Université de Lille, CNRS, Centrale Lille, ISEN, Université de Valenciennes, UMR 8520-IEMN, 59000 Lille, France
rsfs.royalsocietypublishing.org
SS, 0000-0002-1567-4943

Review
Cite this article: Szunerits S, Boukherroub R.
2018 Graphene-based biosensors. Interface
Focus 8: 20160132.
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024
Reliable data obtained from analysis of DNA, proteins, bacteria and other
disease-related molecules or organisms in biological samples have become
a fundamental and crucial part of human health diagnostics and therapy.
The development of non-invasive tests that are rapid, sensitive, specific
and simple would allow patient discomfort to be prevented, delays in diag-
nosis to be avoided and the status of a disease to be followed up. Bioanalysis
is thus a progressive discipline for which the future holds many exciting
opportunities. The use of biosensors for the early diagnosis of diseases has

http://dx.doi.org/10.1098/rsfs.2016.0132 become widely accepted as a point-of-care diagnosis with appropriate speci-

Accepted: 31 January 2018
One contribution of 13 to a theme issue
‘The biomedical applications of graphene’.
Subject Areas:
nanotechnology, biomaterials
Keywords:
biosensors, graphene, diagnostics
Author for correspondence:
Sabine Szunerits
e-mail: sabine.szunerits@univ-lille1.fr
ficity in a short time. To allow a reliable diagnosis of a disease at an early
stage, highly sensitive biosensors are required as the corresponding bio-
markers are generally expressed at very low concentrations. In the past 50
years, various biosensors have been researched and developed encompass-
ing a wide range of applications. This contrasts the limited number of
commercially available biosensors. When it comes to sensing of biomarkers
with the required picomolar ( pM) sensitivity for real-time sensing of biologi-
cal samples, only a handful of sensing systems have been proposed, and
these are often rather complex and costly. Lately, graphene-based materials
have been considered as superior over other nanomaterials for the develop-
ment of sensitive biosensors. The advantages of graphene-based sensor
interfaces are numerous, including enhanced surface loading of the desired
ligand due to the high surface-to-volume ratio, excellent conductivity and a
small band gap that is beneficial for sensitive electrical and electrochemical
read-outs, as well as tunable optical properties for optical read-outs such as
fluorescence and plasmonics. In this paper, we review the advances made in
recent years on graphene-based biosensors in the field of medical diagnosis.

1. Introduction
Biosensors are analytical devices that convert a biochemical/biological reaction
into a measurable physico-chemical signal, which is proportional to the analyte
concentration. A typical biosensor thus consists of two elements: a surface-
linked biological component that interacts selectively with the analyte of
interest in blood or serum and a transducer for the detection of the analyte-
binding event on the surface. The major advantage of using biosensors
compared with other conventional biochemical assays such as immunoassays
and polymerase chain reaction based strategies is the fast response time
(normally several minutes) along with high specificity. The first biosensor
reported dates back to the work by Leland C. Clark, who is considered to be
the father of biosensors. Based on his experience with the oxygen electrode
[1], he proposed making electrochemical sensors more intelligent by entrapping
enzymes such as glucose oxidase onto the oxygen electrode using a dialysis
membrane [2]. This glucose analyser became commercially available in 1975
in the form of an amperometric sensor. The idea of immobilizing antibodies
rather than enzymes on the sensor transducer emerged in the early 1980s
with the work by Lieberg et al. [3]. This work paved the way for the commercial
success of surface plasmon resonance (SPR) biosensors (Biacore Technology,
launched in 1990); these devices rely on monitoring the change in the plasmonic
signal upon antibody–antigen affinity reaction in real time. Since then various
biosensors have been developed encompassing a wide range of applications.

& 2018 The Author(s) Published by the Royal Society. All rights reserved.
 2
(a) production methods
mechanical exfoliation, liquid-based exfoliation, CVD growth

rsfs.royalsocietypublishing.org
(b) structure of graphene-based materials

graphene graphene oxide (GO) reduced graphene oxide (rGO) porous rGO

(c) electrode preparation

graphene transfer, drop-casting, electrostatic interaction,

Interface Focus 8: 20160132

electrophoretic deposition (EDP), electrochemical reduction (ER), etc.

(d) some selected SEM images of graphene-coated interfaces

Au-graphene GC-prGO Au/PDDA/GO Au-rGO GC-prGO

Downloaded from https://royalsocietypublishing.org/ on 05 January 2024

(transfer) (drop-casting) (electrostatic) (EPD) (ER)

1 mm 1 mm 1 mm 1 mm 1 mm

Figure 1. Construction of graphene-based biosensor interfaces: (a) current preparation methods; (b) chemical structures of different graphene derivatives widely used for
biosensing; (c) methods for the transfer of graphene-based materials to solid substrates; (b) scanning electron micrograph (SEM) images of graphene-coated interfaces using
different deposition methods and different graphene precursors. GC, glassy carbon electrode; PDDA, poly(diallyldimethylamonium). (Online version in colour.)

One of the many remaining challenges in biosensor
development concerns the efficient and selective capture of
the biological recognition event. In order to achieve the often
requested picomolar ( pM) limit of detection (LOD) for biologi-
cal analytes, nanomaterials have been intensively investigated
as candidates for transducer coatings. The possibility of enhan-
cing the amount of bioreceptor immobilization with
consequently improved signal read-out makes this biosensor
approach greatly appealing. Among the different nanomaterials
considered [4], graphene and its various forms such as graphene
oxide (GO), reduced graphene oxide (rGO), graphene nanorib-
bons (GNRs) and so on have received worldwide attention for
the development of biosensors. Different sensing mechanisms
including optical, electrochemical or electrical can be employed
with graphene-based biosensors, in the following noted as
G-biosensors. In the case of electrochemical (amperometric, vol-
tammetric, impedimetric) G-biosensors and electrical sensing
concepts (graphene-based field effect transistors (G-FETs)),
the high electron transfer rates, high charge-carrier mobility
and low electrical noise levels are of utmost importance for
highly sensitive detection of biomarkers and other biological
analytes in serum and blood samples [5,6]. Furthermore, chemi-
cally derived graphene derivatives exhibit a high density of
edge-plane-like defect sites, providing many active sites for elec-
tron transfer to chemical and biological species [5]. Also, the
high optical transparency of graphene monolayers makes
them ideal materials for optical-based G-biosensors and
highly beneficial to improve the sensing performance of plas-
monic sensors [7]. The fluorescence quenching ability of GO
resulted in the development of several fluorescence resonance
energy transfer (FRET)-based G-biosensors [8]. Graphene has
also been shown to be an emerging material as a surface-
enhanced Raman substrate (SERS) due to its ability to generate
strong chemical enhancement [9]. Independent of the detection
method adopted, the presence of hydrophobic domains or p-
systems on graphene-based transducers renders them excellent
supporting layers for immobilization of biomolecules. How-
ever, the control of non-specific interactions becomes even
more important when compared with other sensing surfaces,
especially when analysing biological liquids [10]. Improvement
in the fabrication of non-fouling graphene transducers is one of
the essential steps in the development of high-performance
G-biosensors [10].
To obtain a general overview of the results achieved in this
field, some of the key works around the development of point-
of-care sensing in biological fluids using G-biosensors will be
highlighted here. As the examples of G-biosensors are count-
less, the focus will be on the direct detection of small
molecules such as glucose and dopamine, DNA and protein
biomarkers (e.g. folic acid protein, lysozyme, prostate-specific
antigen) using immunosensing, and also on pathogen detec-
tion. It is hoped that this brief overview of G-biosensors and
their interest for the biomedical field will stimulate further
research activities together with commercialization of some
of the proposed concepts to help advance personalized care.
2. Preparation of graphene-based biosensors
A number of different approaches for the synthesis of graphene
and its derivatives such as GO, rGO, porous-reduced graphene
oxide (prGO) and GNRs are available in the literature (figure 1).
Large area single and few layer high-quality graphene
nanosheets can be produced by chemical vapour deposition
(CVD) methods on nickel or copper and are commercially acces-
sible. Such graphene sheets are nowadays routinely transferred
to any transducer interface using mainly polydimethylsiloxane
supported transfer processes [11]. The high quality of CVD gra-
phene and the possibility of obtaining mono- and bilayer
modified electrical interfaces makes such electrodes advan-
tageous for G-FETs and plasmonic biosensing. The use of
chemically derived GO and rGO nanosheets, obtained from a
 Table 1. Selected examples of most performing graphene-based biosensors for different analytes. 3D, three-dimensional; EC, electrochemistry; FET, field effect 3
transistor; FRET, fluorescence resonance energy transfer; GO, graphene oxide; rGO, reduced graphene oxide; GQDs, graphene quantum dots; GOx, glucose oxidase;

rsfs.royalsocietypublishing.org
Au NPs, gold nanoparticles; DPV, differential pulse voltammetry; LOD, limit of detection; PDDA, poly(diallyldimethylamonium); PSA, prostate-specific antigen;
SERS, surface-enhanced Raman substrate; SPR, surface plasmon resonance; ssDNA, single-stranded DNA.

analyte sensor design detection LOD ref

glucose 3D graphene foam – Co3O4 nanowires EC 20 nM [13]

glucose graphene þ GOx FET 0.1 mM [6]
glucose GQDs – bipyridine boronic acid fluorescence 1 mM [14]
dopamine rGO– polyvinylpyrrolidone EC 0.2 nM [15]

Interface Focus 8: 20160132

DNA GO and GQD– ssDNA FRET 75 pM [16]
DNA graphene – Au NPs – ssDNA SPR 500 aM [11]
DNA GO nanowalls DPV 9.6 zM [17]
DNA graphene FET 10 fM [18]
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024

lysozyme Au/PDDA – GO –Micrococcus lysodeikticus SPR 3.4 nM [19]

folic acid Au – rGO DPV 1 pM [10]
folic acid graphene SPR 5 fM [20]
b-amyloid magnetic/plasmonic GO SERS 100 fg ml21 [21]
PSA rGO FET 1 fM [22]
21
Escherichia coli graphene – anti-E. coli FET 10 cfu ml [23]

graphite precursor through solution-based exfoliation aiming at
weakening the van der Waals forces between the graphene
layers, is the most commonly used synthetic approach for the
construction of G-biosensors. It is a relatively cheap method
for obtaining GO/rGO on a large scale, with the additional
benefit of possible modulation of the morphology and porosity
of the nanosheets. Furthermore, doping with non-metallic
elements such as nitrogen, sulfur or boron allows the electronic
structure of these materials to be modulated and leads, in gen-
eral, to improved electrical and electro-catalytic properties.
Reduction of the GO-flake size results in better dispersible
structures of 3–20 nm in size consisting of no more than five
layers; these structures exhibit a high surface area and are
termed graphene quantum dots (GQDs) [12].
Different techniques such as drop-casting, spin-coating, elec-
trostatic interaction between positively charged interfaces and
the negatively charged GO/rGO nanosheets, electrophoretic
deposition (EPD) and electrochemical reduction of GO can be
employed to coat electrical as well as inert surfaces with the
chemically derived graphene materials. The method of choice
depends on the use after application and the employed transdu-
cer element. Table 1 gives a short list of selected biological
analytes of interest together with the method employed for
their detection and the LOD which can be reached, most of
them being discussed in more detail in this review.
3. G-biosensors for glucose and dopamine
One challenging and important molecule to monitor is glucose.
An increase in glucose levels is critical for human health as
hyperglycaemia, defining diabetes, leads to premature death
caused by macrovascular and microvascular complications.
Close monitoring of the blood glucose concentration can largely
help to manage diabetes. Tremendous efforts have been put into
the development of efficient and reliable methods for glucose
sensing. Graphene-based glucose sensors are generally built
by immobilizing glucose oxidase (GOx) onto the graphene sur-
face, such as the graphene–FET proposed by Huang et al. [24]
(figure 2a). In this work, GOx was covalently linked via its
amine groups to 1-pyrenebutanoic acid succinimidyl ester,
where the pyrene end is firmly attached to graphene through
p–p stacking interactions. Measuring a change in
conductance allowed glucose detection down to 0.1 mM.
Although the use of GOx allows for high selective detection of
glucose, non-enzymatic glucose sensors based on the inte-
gration of electro-catalytic sites for glucose, often in the form
of nanoparticles (NPs), onto graphene have been pursued
[13,25,26]. While these sensors are less selective to glucose and
need to be for the most part operated in an alkaline medium,
they display several advantages such as better stability than
GOx-based interfaces and often improved sensitivity with
detection limits in the low micromolar to nanomolar range. In
the case of N-doped prGO loaded with Cu NPs (N-prGO-Cu
NPs) the improved sensitivity to glucose is attributed to the elec-
tro-catalytic behaviour of the CuO/CuOOH couple (figure 2b).
In the presence of glucose, the redox peak of Cu(0)/Cu(I) stays
unchanged, while the peak of the Cu(I)/Cu(II) transition is
decreased, reflecting the formation of a Cu(I)–glucose complex.
The band at þ0.4 V is strongly increased, in line with the activity
of Cu(III) in basic medium (figure 2b) [27].
A different non-enzymatic glucose-sensing approach based
on the use of GQDs modified with boronic acid-substituted
bipyridine ligands, which serve as a fluorescence quencher
upon electrostatic interaction with GQDs, was proposed [14]
(figure 2c). When glucose is added to the system, the boronic
acid moieties are converted to tetrahedral anion glucoboronate
esters, which neutralize the net charge of the bipyridinium,
thus greatly diminishing the quenching effect and recovering
the fluoresence intensity of the GQDs (figure 2c).
Another small biomolecule of interest is dopamine. It is
one of the crucial catecholamine neurotransmitters widely
found in serum samples at concentrations between 10 nM and
1 mM. Abnormal levels can result in a variety of diseases
 (c) 4
(a) enzymatic-based G-FET for glucose
fluoresence-based glucose sensor using

rsfs.royalsocietypublishing.org
graphene
boronic acid-modified GQDs
glucose
linker low fluorescence high fluorescence

LOD = 0.1 mM
Vds = 100 mV
fluorescent GQD

GOx
D-glucose + O2 + H2O D-glucono-1,5-lactone + H2O2 LOD = 1 mM

Interface Focus 8: 20160132

(b)
selective dopamine sensing (d)
non-enzymatic amperometric glucose sensor
C Cu ascorbic acid (AA) serotonin (ST)
6
no glucose 10 mA
j (mA cm–2)

4 8 mA
+ glucose
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024

2
0
0.6 mm –2 –0.1 0 0.1 0.2 0.3 –0.2 0 0.2 0.4 0.6 0.8 1.0
N-prGO-CuO NPs on a –4
–1 –0.5 0 0.5 1 dopamine (DA) DA
gold electrode
E/V versus Ag/AgCl AA ST
15 mA

Cu(II)O + OH– Cu(III)OOH+ e– LOD = 250 nM

+ e–
Cu(III) + glucose Cu(II)+ + D-glucono-1, 5-lactone + H2O2 0.1 0.2 0.3 0.4 0.5 0.6 –0.2 0 0.2 0.4 0.6 0.8 1.0
E/V versus Ag/AgCl E/V versus Ag/AgCl

Figure 2. G-based sensors of small molecules such as glucose and dopamine: (a) CVD graphene modified with glucose oxidase (GOx) using a bifunctional pyrene linker
for the construction of a G-FET for glucose (reprinted with permission from Huang et al. [23]); (b) non-enzymatic glucose sensor operating under basic conditions based
on N-doped porous-reduced graphene oxide loaded with CuO NPs (N-prGO-Cu NPs) (reprinted with permission from Maaoui et al. [27]); (c) graphene quantum dots
(GQDs) modified with boronic acid-substituted bipyridine ligands for non-enzymatic glucose sensing under physiological conditions (reprinted with permission from Li
et al. [14]); (d ) differential pulse voltammetry of ascorbic acid (1 mM), serotonin (1 mM) and dopamine (1 mM) and a mixture of all three (0.1 M) on glassy carbon
electrodes modified with hydrazine-reduced GO (reprinted with permission from Alwarappan et al. [28]). (Online version in colour.)

(Huntington’s disease, Parkinson’s disease, schizophrenia) with
high levels increasing the risk of high blood pressure-related dis-
eases. Rapid and accurate detection of dopamine at low cost has
thus become increasingly important in clinical diagnostics. The
electrochemical activity of dopamine makes its electrochemical
detection rather appealing. However, uric acid (UA), ascorbic
acid (AA) and serotonin (ST) coexist with dopamine in the extra-
cellular fluids of the central nervous system in mammals at even
higher concentrations and can be oxidized at a potential close to
that of dopamine. The excellent electrochemical properties of
graphene for dopamine together with chemical enhancement
via p–p stacking interactions have enabled the development
of dopamine-specific G-sensors even in the presence of high
concentrations of AA, UA and ST [28,29] (figure 2d). Extremely
low LODs could be achieved using electrochemically reduced
GO in combination with polyvinylpyrrolidone capable of
detecting 0.2 nM dopamine in the presence of 1 mM AA [15].
4. DNA sensing with G-biosensors
The need for rapid and sensitive DNA analysis is an important
issue in clinical diagnosis. Major studies have focused on the
sequence-specific recognition of ssDNA and on the detection
of single nucleotide polymorphisms (SNPs). SNPs are a
common form of genomic variation occurring in every 100–
300 bp and are related to many major diseases and disorders,
such as Parkinson’s disease, Alzheimer’s disease, diabetes
and various cancers. The development of analytical
approaches for selective DNA sensing has consequently been
strongly pursued with the belief that low-cost systems suitable
for DNA analysis could revolutionize modern healthcare. Elec-
trodes modified with graphene oxide nanowalls (GONs) with
preferred vertical orientation [17] or graphene nanoplatelets
[30] are capable of catalytically oxidizing the four DNA bases
simultaneously, resulting in DNA sensors with an LOD as
low as 9.4 zM [17] (figure 3a).
The fluorescence quenching ability of GO was exploited by
several research groups for the detection of hybridization
events. When a dye-labelled ssDNA is immobilized via non-
covalent binding onto GO, the fluorescence is quenched; this
non-covalent interaction is reversible. ssDNA interaction with
GO occurs thus via p 2 p stacking, hydrophobic interactions
and hydrogen bonding. Even though both GO and ssDNA
are negatively charged, DNA can be adsorbed on GO in buffers
containing a high concentration of salts to screen electrostatic
repulsion [31]. In double-stranded DNA (dsDNA), the nucleo-
tide bases are hidden in the helical structures, preventing their
effective interaction with the GO surface in contrast to ssDNA.
In the presence of a complementary DNA (cDNA) target, a
duplex is formed, disturbing the GO–ssDNA interaction and
resulting in the release of the formed dsDNA, at which point
fluorescence is restored (figure 3b). This approach has been
recently used in combination with ssDNA-modified GQDs
for FRET-based DNA sensing [16] (figure 3b).
This concept was further applied to SPR-based DNA
sensing [11] (figure 3c). Gold nanostars modified with
ssDNA were integrated into commercially available
 (c) SPR-based DNA sensing 5
(a) electrochemical DNA sensing with graphene
oxide nanowalls

rsfs.royalsocietypublishing.org
0.2
RGNW
GONW
12 G
RGNS
GONS
cDNA

I (mA)
graphite
GC
10 500 nm
C Au-graphene/Au NPs–ssDNA
8 0
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 LOD = 1 zM
I (mA)

E/V (versus Ag/AgCI)

DSPR signal
6 A
G
4
T
2 compl. DNA
3-mismatched DNA 10 nm

Interface Focus 8: 20160132

0 0 5 10 15 20 25
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 LOD = 500 aM
E/V (versus Ag/AgCl) time (min)

(b) FRET between ss-DNA-modified GQDs and GO

mechanism hybridization
cDNA on GO, then
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024

DNA interaction with GO desorption

adsorption

FRET
tDNA
LOD = 75 pM
mDNA
dsDNA–rGQDs ssDNA–rGQDs/GO

Figure 3. DNA sensing with G-biosensors: (a) single DNA electrochemical biosensing using graphene nanowalls (GNWs): SEM image of GNWs formed by electrophoretic
deposition onto graphite rod, differential pulse voltammograms of dsDNA (0.1 mM) in phosphate-buffered saline (0.1 M, pH 7) on different interfaces (reprinted with per-
mission from Akhavan et al. [17]); (b) mechanism of DNA interaction with GO (reprinted with permission from Liu et al. [31]) and FRET-based DNA sensing using GQDs and GO
(reprinted with permission from Qian et al. [16]); (c) graphene–SPR-based DNA sensing: transmission electron micrograph (TEM) image of a gold nanostructure together with
the change in SPR signal upon incubation with cDNA and mismatched DNA (reprinted with permission from Zagorodko et al. [11]). (Online version in colour.)

graphene-coated SPR chips. As dsDNA has a lower affinity prevailing analytical protocols (e.g. enzyme-linked immuno-
to graphene, hybridized targets are released from the graphene sorbent assays) for biomarker detection lag far behind the
surface, resulting in a negative shift of the SPR signal; requirements for clinical utility and research. Graphene-
this allows easy discrimination between mismatched based immunoassay platforms, where specific antibodies
DNA and offers a label-free approach for DNA detection are immobilized onto graphene to capture selectively the bio-
with an LOD of 500 aM over a linear dynamic range up to marker analyte, have shown on the other hand excellent
10 nM [11]. sensitivity [33] (figure 4). For example, we have recently
The feasibility of G-FET-based sensors for nucleic acid- demonstrated that post-functionalization of rGO-modified
based analytes has been shown by Zheng et al. [18]. Peptide electrodes by simple immersion into a solution of folic acid
nucleic acid (PNA) was non-covalently immobilized to the allowed for the development of an electrochemical-based
graphene channel and hybridization with target DNA pro- sensor for folic acid protein with an LOD of 1 pM [10]
duced a left-shift in the Dirac point with an LOD of 10 fM (figure 4a) and a plasmonic sensor with a 5 fM LOD [20].
[18]. Dontschuk et al. [32] have also shown the usefulness of Levels of folic acid protein in serum can increase up to
G-FETs for DNA sequencing. They demonstrated experimen- 22 pM in metastic diseases. Given that human serum is free
tally that G-FETs are able to measure distinct coverage- of folic acid proteins, detection of this protein in serum
dependent conductance signatures upon adsorption of the serves as an early stage cancer diagnostic step.
four different DNA nucleobases—a result that was attributed A rGO-based FET modified with prostate-specific antigen-
to the formation of an interface dipole field. a1-antichymotrypsin (PSA-ACT) was used by Kim et al. [22] to
detect fM levels of PSA with a dynamic range over six orders of
magnitude (figure 4b). An SPR-based read-out was used by
Cosnier and co-workers [34] for the detection of cholera toxin
5. G-based sensors for protein biomarkers on graphene-coated gold chips modified with pyrene-
Protein biomarkers are specific molecules existing in blood or nitrilotriacetic acid (NTA) with an LOD of 5 pg ml21. We
tissues, whose measurement or identification is very critical showed recently the suitability of Micrococcus lysodeikticus-
and efficient in the prediction, diagnosis and monitoring of modified GO-coated SPR interfaces to sense serum lysozyme
cancer and many other diseases. The clinical utility of bio- levels with an LOD of 3.4 nM [19] (figure 4c). Lysozyme, an
markers to diagnose disease requires the capability to enzyme found in biological fluids, is upregulated in leukaemia,
measure femto- to picomolar concentrations of these, which renal diseases as well as a number of inflammatory diseases.
is also important to understand cellular processes and to While lysozyme concentration in the serum of healthy people
search for new protein biomarkers. The LODs of the ranges from 27 to 301 nM, patients with inflammatory bowel
 folic acid protein on Au–rGO/folic acid (a) 6
PSA on rGO/PSA antibody (b)
differential pulse voltammetry field effect transistor

rsfs.royalsocietypublishing.org
a c
LOD = 1 pM PSA mAb
1.6 100 pH 7.4
100 fg ml–1 150

conductivity, s (mS)
1 pg ml–1
10 pg ml–1 100
0 pM

DVg,min (mV)
90 pH 6.2

j (mA cm–2)
100 pg ml–1
1 ng ml–1
1.2 30 pM 80
10 ng ml–1
100 ng ml–1
50
0 pl(PSA-ACT)6.8
1 ug ml–1

0.8 100 pM 70
–50
–100
pH 7.4

500 pM 60
smin
Vg,min –150
0.4 1 nM –0.5 0 0.5 1.0 1.5
102 103 104 105 106 107 108 109
Au–rGO/FA Vg (V)
PSA-ACT concentration, (C) (fg ml–1)

0 LOD = 1 fM
–0.4 –0.3 –0.2 –0.1 0 100
lysozyme (100 mM) (e)

normalized signal (%)

E/V versus SCE rGO
80
graphene

Interface Focus 8: 20160132

60
PEG

pyrene-PEG
40
BSA
(c) lysozyme on Au–PDDA/GO–M. 20 (d)
serum
lysodeikticus 0
surface plasmon resonance
Alzheimer’s disease marker on
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024

GO/plasmonic–magnetic platform
0 mm 5 10 15 20
surface-enhanced Raman spectroscopy
0 1.31 mm 100 3000
1.20

Raman intensity
SPR signal

5 1.00 2000
0
10 0.80 1000

0.60
15
0.40
–100 0
1000 1500
wavenumber (Cm–1)
20
0-0.2-1-10-20-40 mg ml–1
1
0
–200 LOD = 100 fg ml–1
0 2 4 6 8 10
LOD = 3.4 nM time (min)

Figure 4. G-biosensors for protein sensing using: (a) an electrochemical sensor for folic acid protein: DPV upon addition of increasing concentrations of folic acid
proteins (reprinted with permission from He et al. [10]); (b) G-FET for the analysis of prostate biomarkers (reprinted with permission from Kim et al. [22]); (c)
graphene – SPR-based sensing of lysozyme: atomic force microscopy image of an Au – (PDDA/GO)2 interface modified with Micrococcus lysodeiktikus and the
change in SPR responses to different concentrations of lysozyme added to fetal bovine serum (reprinted with permission from Vasilescu et al. [19]); (d ) concen-
tration-dependent SERS spectra from a tau protein-conjugated nanoplatform after magnetic separation together with an SEM image of a core-shell NP-attached
hybrid GO and HR-TEM image (reprinted with permission from Demeritte et al. [21]). BSA, bovine serum albumin; PEG, polyethylene glycol. (Online version in colour.)

disease show micromolar levels of lysozyme. Recently, a multi- as low as 10 cfu ml21 to be detected [23]. Graphene oxide in
functional nanoplatform based on magnetic–plasmonic NPs combination with E. coli O157:H7 antibody-conjugated quan-
attached to GO allowed for the sensitive detection of Alzhei- tum dots was used as a pathogen-revealing agent by
mer’s disease biomarkers (b-amyloid, tau proteins) down to exploiting the universal highly efficient long-range quenching
100 fg ml21 [21] (figure 2d). These examples highlight the properties of GO; an LOD of 3.8  103 cfu ml21 was achieved
efficient use of G-biosensors for protein analysis. [36]. Graphene printed onto water-soluble silk and modified
One main hurdle of all these sensors when performing tests with antimicrobial peptides allowed bioselective detection of
in human serum samples, often not evoked in the literature, is bacteria at single-cell levels remotely [37].
linked to the high non-specific interaction between the graphene
surface and serum proteins. We have compared a number of
different strategies to reduce non-specific binding of clinical
serum samples spiked with lysozyme (100 mM) on rGO [20] 7. Conclusion and perspectives
(figure 4e). While simple immersion into serum decreased
We tried to review the most recent advances in graphene-based
strongly the anti-fouling properties of graphene, rGO modifi-
biosensors by selectively highlighting a variety of different
cation with pyrene–polyethylene glycol (PEG) units has been
examples for the detection of some molecules of biomedical
shown to result in the best non-fouling interface [20,35].
interest. Routinely, such sensors achieve a picomolar LOD,
with some even reaching the low femtomolar concentration
range. The possibility that a large range of different detection
methods can be employed with graphene-based sensors is of
6. Bacteria and viruses high advantage, as depending on the looked after final appli-
The specific and sensitive detection of pathogenic micro- cation, sensor size and read-out can be customized at will.
organisms remains a big scientific challenge and a practical There is, however, still an urgent need for moving beyond
problem of enormous significance. Pathogen diagnosis is cur- research by developing new concepts for achieving even
rently based on culturing the microorganism on agar plates better sensitivity and selectivity, in order to bring some of the
with the disadvantage of being long (minimum of 24 h) and current sensors into real biomedical applications. Even
ignoring viable but non-culturable cells. G-FET biosensors though a large number of sensors reported in the literature
have been successfully applied for the detection of bacteria exhibit good storage stability and repeatability which are
and their metabolic activities. CVD graphene modified with important for complex sensors involving nanomaterials and
anti-Escherichia coli antibodies allowed E. coli concentrations manual step preparations, the performance in real biological
 samples is often not reported. As non-specific interactions are biosensors as point-of-care alternatives for patients. The success 7
of primary concern in graphene-based interfaces, this lack of of any new biosensor material lies in addition in its reproduci-

information together with the large-scale reproducibility of
the fabrication of graphene biosensor interfaces are probably
two of the most crucial limiting factors for current commercia-
lization. When it comes to in vivo application of some of the
sensing concepts, graphene-based biosensors are still in their
infancy. Toxicity and biocompatibility issues still need to be
addressed carefully to avoid any undesired secondary health
effects. Current in vivo and in vitro assessments of the biostabil-
ity of the sensors are encouraging and promising for further
technological transfer.
rsfs.royalsocietypublishing.org
bility and possible industrial-scale production. The emergence
of a number of companies providing mono- and bilayered
graphene nanosheets on several interfaces, GO, rGO and even
modified matrices, has been an additional motivation for
using graphene for biosensor applications.
Data accessibility. This article has no additional data.
Competing interests. We declare we have no competing interests.
Funding. Financial support from the Centre National de la Recherche

Interface Focus 8: 20160132

Different challenges are still to be overcome. The collabor-
ation between material scientists, chemists, physicists as well
as engineers and medical personnel is of fundamental impor-
tance to drive this field further and to propose graphene-based
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024
Scientifique (CNRS), the University Lille 1, the Hauts-de-France
region, the CPER ‘Photonics for Society’, the Agence National de la
Recherche (ANR) through FLAG-ERA JTC 2015-Graphitivity, and the
EU through the Marie Sklodowska-Curie action (H2020-MSCA-RISE-
2015, PANG-690836) and Graphitivity is acknowledged.

References
1. Clark LCJ. 1956 Monitor and control of blood and
tissue oxygen tensions. Trans. Am. Soc. Artif. Intern.
Organs 2, 41 –48.
2. Clark LCJ, Lyons C. 1962 Electrode systems for
continuous monitoring in cardiovascular surgery.
Ann. N.Y. Acad. Sci. 102, 29 –45. (doi:10.1111/
j.1749-6632.1962.tb13623.x)
3. Lieberg B, Nylander C, Lundström I. 1983 Surface
plasmon resonance for gas detection and
biosensing. Sens. Actuators 4, 299–304. (doi:10.
1016/0250-6874(83)85036-7)
4. Holzinger M, Le Goff A, Cosnier S. 2014
Nanomaterials for biosensing applications: a review.
Front. Chem. 2, 63. (doi:10.3389/fchem.2014.00063)
5. Ambrosi A, Chua CK, Bonanni A, Pumera M. 2014
Electrochemistry of graphene and related materials.
Chem. Rev. 114, 7150–7188. (doi:10.1021/
cr500023c)
6. Viswanathan S et al. 2015 Graphene –protein field
effect biosensors: glucose sensing. Mater. Today 18,
513–522. (doi:10.1016/j.mattod.2015.04.003)
7. Szunerits S, Maalouli N, Wijaya E, Vilcot JP,
Boukherroub R. 2013 Recent advances in the
development of graphene-based surface plasmon
resonance (SPR) interfaces. Anal. Bioanal. Chem.
405, 1435–1443. (doi:10.1007/s00216-012-6624-0)
8. Kim J, Cote LJ, Kim F, Huang J. 2010 Visualizing
graphene based sheets by fluorescence quenching
microscopy. J. Am. Chem. Soc. 132, 260–267.
(doi:10.1021/ja906730d)
9. Sharma B, Frontiera RR, Henry A-I, Ringe E, Van
Duyne RP. 2012 SERS: materials, applications, and
the future. Mater. Today 15, 16– 25. (doi:10.1016/
S1369-7021(12)70017-2)
10. He L, Wang Q, Mandler D, Li M, Boukherroub R,
Szunerits S. 2015 Detection of folic acid protein
in human serum using reduced graphene oxide
electrodes modified by folic-acid. Biosens.
Bioelectron. 75, 389–395. (doi:10.1016/j.bios.
2015.08.060)
11. Zagorodko O, Spadavecchia J, Yanguas Serrano A,
Larroulet I, Pesquera A, Zurutuza A, Boukherroub R,
Szunerits S. 2014 Highly sensitive detection of DNA
hybridization on commercialized graphene-coated
surface plasmon resonance interfaces. Anal. Chem.
86, 11 211–11 216. (doi:10.1021/ac502705n)
12. Sun H, Wu L, Wei W, Qu X. 2013 Recent advances in
graphene quantum dots for sensing. Mater. Today
16, 433– 442. (doi:10.1016/j.mattod.2013.10.020)
13. Dong XC, Xu H, Wang XW, Huang YX, Chan-Park
MB, Zhang H, Wang LH, Huang W, Chen P. 2012 3D
Graphene –cobalt oxide electrode for high-
performance supercapacitor and enzymeless glucose
detection. ACS Nano 6, 3206–3213. (doi:10.1021/
nn300097q)
14. Li Y-H, Zhang L, Huang J, Liang R-P, Qiu J-D. 2013
Fluorescent graphene quantum dots with a boronic
acid appended bipyridinium salt to sense
monosaccharides in aqueous solution. Chem.
Commun. 49, 5180. (doi:10.1039/c3cc40652k)
15. Liu Q, Zhu X, Huo Z, He X, Liang Y, Xu M. 2012
Electrochemical detection of dopamine in the
presence of ascorbic acid using PVP/graphene
modified electrodes. Talanta 97, 557–562. (doi:10.
1016/j.talanta.2012.05.013)
16. Qian ZS, Shan XY, Chai LJ, Ma JJ, Chen JR, Feng H.
2014 A universal fluorescence sensing strategy
based on biocompatible graphene quantum dots
and graphene oxide for the detection of DNA.
Nanoscale 6, 5671–5674. (doi:10.1039/
C3NR06583A)
17. Akhavan O, Ghaderi E, Rahighi R. 2012 Toward
single-DNA electrochemical biosensing by graphene
nanowalls. ACS Nano 6, 2904–2916. (doi:10.1021/
nn300261t)
18. Zheng C, Huang L, Zhang H, Sun Z, Zhang Z, Zhang
G-J. 2015 Fabrication of ultrasensitive field-effect
transistor DNA biosensors by a directional transfer
technique based on CVD-grown graphene. ACS Appl.
Mater. Interfaces 7, 16 953–16 959. (doi:10.1021/
acsami.5b03941)
19. Vasilescu A et al. 2017 Surface plasmon resonance
based sensing of lysozyme in serum on Micrococcus
lysodeikticus-modified graphene oxide surfaces.
Biosens. Bioelectron. 89, 525 –531. (doi:10.1016/j.
bios.2016.03.040)
20. He L, Pagneux Q, Larroulet I, Yanguas Serrano A,
Pesquera A, Zurutuza A, Mandler D, Boukherroub R,
Szunerits S. 2017 Label-free femtomolar cancer
biomarker detection in human serum using
graphene-coated surface plasmon resonance chips.
Biosens. Bioelectron. 89, 606 –611. (doi:10.1016/
j.bios.2016.01.076)
21. Demeritte T, Nellore BPV, Kanchanapally R, Sinha
SS, Pramanik A, Chavva SR, Ray CP. 2015 Hybrid
graphene oxide based plasmonic-magnetic
multifunctional nanoplatform for selective
separation and label-free identification of
Alzheimer’s disease biomarkers. ACS Appl. Mater.
Interfaces 7, 13693. (doi:10.1021/acsami.5b03619)
22. Kim DJ, Sohn IY, Jung JH, Yoon OJ, Lee NE, Park JS.
2013 Reduced graphene oxide field-effect transistor
for label-free femtomolar protein detection. Biosens.
Bioelectron. 41, 621–626. (doi:10.1016/j.bios.2012.
09.040)
23. Huang YX, Dong XC, Liu Y, Li L-J, Chen P. 2011
Graphene-based biosensors for detection of bacteria
and their metabolic activities. J. Mater. Chem. 21,
12358. (doi:10.1039/c1jm11436k)
24. Huang YX, Dong XC, Shi Y, Li CM, Li L-J, Chen P.
2010 Nanoelectronic biosensors based on CVD
grown graphene. Nanoscale 2, 1485–1488. (doi:10.
1039/c0nr00142b)
25. Subramanian P, Niedziolka-Jonsson J, Lesniewski A,
Wang Q, Li M, Boukherroub R, Szunerits S. 2014
Preparation of reduced graphene oxide–Ni(OH)2
composites by electrophoretic deposition:
application for non-enzymatic glucose sensing.
J. Mater. Chem. A 2, 5525–5533. (doi:10.1039/
C4TA00123K)
26. Wang Q, Wang Q, Li M, Szunerits S, Boukherroub R.
2015 Preparation of reduced graphene oxide/Cu
nanoparticle composites through electrophoretic
deposition: application for nonenzymatic glucose
sensing. RSC Adv. 5, 15 861– 15 869. (doi:10.1039/
C4RA14132F)
 27. Maaoui H, Singh SK, Teodoresu F, Coffinier Y, Barras
A, Chtourou R, Kurungot S, Szunerits S,
Boukherroub R. 2017 Copper oxide supported on
three-dimensional ammonia-doped porous reduced
graphene oxide prepared through electrophoretic
deposition for non-enzymatic glucose sensing.
Electrochim. Acta 224, 346 –354. (doi:10.1016/j.
electacta.2016.12.078)
28. Alwarappan S, Erdem A, Liu C, Li C-Z. 2009 Probing
the electrochemical properties of graphene
nanosheets for biosensing applications. J. Phys.
Chem. C 113, 8853 –8857. (doi:10.1021/jp9010313)
29. Shang NG, Papakonstantinou P, McMullan M, Chu
M, Stamboulis A, Potenza A, Dhesi SS, Marchetto H.
2008 Catalyst-free efficient growth, orientation and
biosensing properties of multilayer graphene
nanoflake films with sharp edge planes. Adv. Funct.
Downloaded from https://royalsocietypublishing.org/ on 05 January 2024
Mater. 18, 3506 –3514. (doi:10.1002/adfm.
200800951)
30. Bonanni A, Chu CK, Zhao GJ, Sofer Z, Pumera M.
2012 Inherently electroactive graphene oxide
nanoplatelets as labels for single nucleotide
polymorphism detection. ACS Nano 6, 8546. (doi:10.
1021/nn301359y)
31. Liu B, Sun Z, Zhang X, Liu J. 2013 Mechanisms of
DNA sensing on graphene oxide. Anal. Chem. 85,
7987 –7993. (doi:10.1021/ac401845p)
32. Dontschuk N et al. 2015 A graphene field-effect
transistor as a molecule-specific probe of DNA
nucleobases. Nat. Commun. 6, 6563. (doi:10.1038/
ncomms7563)
33. Bollella P, Fusco G, Tortolini C, Sanzò G, Favero G,
Gorton L, Antiochia R. 2017 Beyond graphene:
electrochemical sensors and biosensors for
biomarkers detection. Biosens. Bioelectron. 89,
152 –166. (doi:10.1016/j.bios.2016.03.068)
34. Singh M, Holzinger M, Tabrizian M, Winters S,
Berner NC, Cosnier S, Duesberg GS. 2015
Noncovalently functionalized monolayer graphene 8
for sensitivity enhancement of surface plasmon
rsfs.royalsocietypublishing.org
resonance immunosensors. J. Am. Chem. Soc. 137,
2800– 2803. (doi:10.1021/ja511512m)
35. Halouane F et al. 2017 Selective isolation and
eradication of E. coli associated with urinary tract
infections using anti-fimbrial modified magnetic
reduced graphene oxide nanoheaters. J. Mater.
Chem. B 5, 8133– 8142. (doi:10.1039/C7TB01890H)
36. Morales-Narvaez E, Hassan A-R, Merkoci A. 2013
Graphene oxide as a pathogen-revealing agent:
sensing with a digital-like response. Angew. Chem.
Interface Focus 8: 20160132
Int. Ed. 52, 13 779 –13 783. (doi:10.1002/anie.
201307740)
37. Mannoor MS, Tao H, Clayton JD, Sengupta A,
Kaplan DL, Naik RR, Verma N, Omenetto FG,
McAlpine MCM. 2012 Graphene-based wireless
bacteria detection on tooth enamel. Nat. Commun.
3, 763. (doi:10.1038/ncomms1767)