Graphene-based biosensor using transport properties

Abstract

The potential of graphene nanoribbons (GNR's) as molecular-scale sensors is investigated by calculating the electronic properties of the ribbon and the organic molecule ensemble. The organic molecule is assumed to be absorbed at the edge of a zigzag GNR. These nanostructures are described using a single-band tight-binding Hamiltonian. Their transport spectrum and density of states are calculated using the nonequilibrium Green's function formalism. The results show a significant suppression of the density of states (DOS), with a distinct response for the molecule. This may be promising for the prospect of GNR-based single-molecule sensors that might depend on the DOS (e. g., devices that respond to changes in either conductance or electroluminescence). Further, we have investigated the effect of doping on the transport properties of the system. The substitutional boron and nitrogen atoms are located at the center and edge of GNR's. These dopant elements have significant influence on the transport characteristics of the system, particularly doping at the GNR edge.

Full text

Page 1

PHYSICAL REVIEW B 83, 045401 (2011)
Graphene-based biosensor using transport properties
R. Chowdhury,*S. Adhikari,†P. Rees, and S. P. Wilks
Swansea University, Singleton Park, Swansea SA2 8PP, UK
F. Scarpa
University of Bristol, Bristol BS8 1TR, UK
(Received 6 June 2010; revised manuscript received 7 October 2010; published 3 January 2011)
The potential of graphene nanoribbons (GNR’s) as molecular-scale sensors is investigated by calculating the
electronic properties of the ribbon and the organic molecule ensemble. The organic molecule is assumed to be
absorbed at the edge of a zigzag GNR. These nanostructures are described using a single-band tight-binding
Hamiltonian. Their transport spectrum and density of states are calculated using the nonequilibrium Green’s
function formalism. The results show a significant suppression of the density of states (DOS), with a
distinct response for the molecule. This may be promising for the prospect of GNR-based single-molecule
sensors that might depend on the DOS (e.g., devices that respond to changes in either conductance or
electroluminescence). Further, we have investigated the effect of doping on the transport properties of the
system. The substitutional boron and nitrogen atoms are located at the center and edge of GNR’s. These
dopant elements have significant influence on the transport characteristics of the system, particularly doping at
the GNR edge.
DOI: 10.1103/PhysRevB.83.045401PACS number(s): 73.22.−f, 73.40.−c
I. INTRODUCTION
Carbon nanotubes (CNT’s) have been shown to exhibit sig-
nificantchangesinelectronictransportphenomenonduetothe
presence of absorbed biomolecules.1–5Previous experimental
workdemonstratingthepossibilityofusingthesepropertiesfor
biosensing applications3,4was performed using nonmetallic
CNT’s in a field-effect transistor (FET) configuration.4The
gate of the FET can be either a metallic back gate or a liquid
gate. The back-gated configuration is usually complicated to
fabricate, whereas precise controlling of the gate potential
is required for liquid-gated devices. However, the detailed
mechanism of the interaction is not yet fully understood.
Severalsuggestionshavebeenproposedincludingelectrostatic
gating liquid gating case4and electron donation by the
detected species.1
Growing research interest in the application of carbon
nanostructure emerged in 2004, led by the experimental
discovery of the stable form of graphene.6–8Graphene is made
up of a single layer of carbon atoms packed into a two-
dimensional honeycomb lattice. It has attracted tremendous
attention in both its two-dimensional and one-dimensional
forms, the latter being obtained by patterning the layer into
a strip or ribbon.9Scanning probe microscopy of graphene
ribbons10revealed bright stripes along its edges, suggesting a
large density of states at the edge near the Fermi level. The
electronic properties of graphene nanoribbons (GNR’s)11,12
definedbytheirquasi-one-dimensionalelectronicconfinement
and the shape of the ribbon ends,13indicates remarkable
applications in graphene-based devices.8GNR’s have simi-
larities in many properties of CNT’s.14However, due to their
planar structure, some of the properties seem to be easier to
manipulate than CNT’s.15
As for the case of nanotubes, electron transport
properties16,17and conductance18are expected to be found in
graphene structures.19,20In particular, different quantization
rules have been predicted for pure GNR’s with zigzag
(ZGNR’s)21and armchair20,22(AGNR’s) edge shaped. Edge
states present in zigzag ribbons provide a single channel for
electron conduction which is not the case for the armchair
configuration. The nature and robustness of these states near
the zigzag edges of ZGNR’s, for different edge shapes and
chemical edge modifications, have been extensively discussed
before.23
In this paper we study the effect of organic molecule
adsorption on the transport of single-layer GNR’s in a two-
probe configuration. In particular, ZGNR’s with a quasi-
one-dimensional organic fragment are considered here. We
choose linear polyaromatic hydrocarbons such as anthracene
(see Fig. 1). Understanding their interaction with GNR’s
will help to understand how full organic or biofragments or
proteins (which are too complex to be presently simulated
by the state-of-the-art ab initio techniques used here) interact
with ZGNR’s. These molecules could be useful to simulate
the effects on the electronic transport of ZGNR’s for the
development of graphene-based sensor devices. Further, we
investigate the effect of doping24on the transport properties of
the system. The substitutional boron and nitrogen atoms are
located at the center and edge of GNR’s (see Fig. 2). These
dopant elements have a significant influence on the transport
characteristics of the system, particularly doping at the GNR
edge.
II. MODELING AND SIMULATION
The electronic structures and geometry relaxations of the
ribbon and the hybrid system of a ribbon with an attached
organic molecule are calculated by using density-functional-
theory (DFT)-based nonequilibrium Greens function (NEGF)
formalism within the TRANSIESTA25,26framework, which
is implemented in the ATOMISTIX TOOLKIT (ATK) package
(version 2008.02).25,27In our transport calculations, the
exchange-correlationpotentialisdescribedbythelocaldensity
approximation (LDA).26,28The Troullier-Martins nonlocal
1098-0121/2011/83(4)/045401(8)045401-1©2011 The American Physical Society

Page 2

CHOWDHURY, ADHIKARI, REES, WILKS, AND SCARPA PHYSICAL REVIEW B 83, 045401 (2011)
FIG. 1. (Color online) (a) Geometric configuration of bare
ZGNR.N denotesthenumberofrepetitionunitcellsalongthelength
and M denotes the number of atoms along the width. Increasing
N and M increases the length and width of ZGNR, respectively.
A similar procedure is adopted for the system of ZGNR with an
attached organic fragment. (b) Schematic configuration of a two-
probe system for a zigzag nanoribbon of ten unit cells in length with
attachedmolecule.Theconfigurationisdividedintothreeregions:left
electrode, right electrode, and central scattering region. The attached
organic fragment is anthracene (C14H10), which is essentially an
organic semiconductor. It is a solid polycyclic aromatic hydrocarbon
consisting of three fused benzene rings and used as a scintillator for
detectors of high-energy photons, electrons, and alpha particles.
pseudopotentials29are used to model core electrons, and
valence electrons are expanded in a SIESTA30localized basis
set. The C-C and C-H bond lengths are set to be 1.42 and
1.1˚ A, respectively. All atomic positions are fully relaxed with
a force tolerance of 0.001 eV/˚ A. The system is analyzed with
1 × 1 × 300 uniformly spaced k points (300 k points in the
transport direction). We performed a convergence study with
respect to k-points sampling. For this purpose, the setup is
increased by increasing the k points in the transport direction
up to 500; however, there is no change in the transmission
spectrum. The self-consistent calculations are performed with
a mixing rate set to 0.01 and the convergent criterion for total
energy is 10−5eV.
The use of the basis set in the ab initio simulations has
a significant effect on the results, as demonstrated in.31,32
FIG. 2. (Color online) Schematic doped configuration of a two-
probe system for a zigzag nanoribbon of ten unit cells in length with
attached molecule.
Due to this fact, first we performed bare ZNR’s, where we
repeated the simulations using the double-ζ and single-ζ basis
sets. It was found that the difference between the relative
current changes predicted by the double-ζ and the single-ζ
basis sets is not significant in the range of the bias used. In
our calculations, the single-ζ polarized basis set is used and
the mesh cutoff is set to be 150 Ry to save computational
time. Moreover, ATK uses periodic boundary conditions in the
directions transverse to the transport direction. To assure that
nosignificantinteractionoccursbetweentheactualsimulation
box and its repeated images, as was previously pointed out,5
we used the vacuum pad of >10 ˚ A in the x and y directions.
It is also noted that vacuum padding is needed to allow
electrostatic interactions to decay for systems. The size of
the simulation box is increased in the transverse direction till
the relative changes in the calculated current are observed as
∼1%. The transport mechanism27of the system is studied
045401-2

Page 3

GRAPHENE-BASED BIOSENSOR USING TRANSPORT ...
PHYSICAL REVIEW B 83, 045401 (2011)
using DFT-based NEGF formalism. Using NEGF theory, the
transmission coefficients can be obtained as
T(E) = tr(ϒRGCϒLG+
C).
(1)
Here the subscripts C, L, and R are used to denote central
scattering, left electrodes, and right electrodes, respectively.
GCand ϒL(R)denote the corresponding Greens functions and
imaginary parts of the self-energies, respectively. The current
passing through the scattering central region is calculated by
the Landauer formula33
?μR
Here μLand μR are the chemical potential of the left and
right electrodes, kB is the Boltzmann constant, Ttempis the
temperature, f(E − μ) =
distribution function, and T(E) is the transmission function.
The following are program parameters used in the present
analyses:
(i) Iteration mixing parameter: Algorithm = Pulay and
Diagonal mixing parameter = 0.1;
(ii) Basissetparameters:Type=single-ζ polarized,Radial
sampling=0.001Bohr,Energyshift=0.001Ry,Deltarinn=
0.8, v0= 40.0 Ry and Split norm = 0.15;
(iii) Iteration control parameters: Tolerance = 10−5, Crite-
rion = Total Energy and Max steps = 500;
(iv) Two probe algorithm parameters: Electrode con-
straint = Off and Initial density type = Equivalent Bulk;
(v) Energycontourintegralparameters:Circlepoints=30,
Integral lower bound = 3 Ry, Fermi line points = 10, Fermi
function poles = 4, Real axis infinitesimal = 0.01 eV and Real
axis point density = 0.02 eV;
(vi) Two center integral parameters: Cutoff = 2500.0 Ry
and Points = 1024.
In the present study, we choose linear polyaromatic hydro-
carbons anthracene (C14H10) (see Fig. 1), which is essentially
a molecular organic semiconductor. It is a solid polycyclic
aromatic hydrocarbon consisting of three fused benzene rings
and can be used as a scintillator for detectors of high-energy
photons, electrons, and alpha particles.
I(V) =2e
h
μL
[f(E − μL) − f(E − μR)]T(E)dE.
(2)
1
1+exp?E−
μ
kBTtemp?is the Fermi-Dirac
III. RESULTS AND DISCUSSION
A. System of zigzag graphene ribbon
We calculated the system properties of bare ZGNR and
ZGNR with attached organic molecules. The band-structure
energy (BSE)34(i.e., the sum of the energies of all states
weighted with their respective occupation for bare ZGNR)
is found as −1839.04 eV, whereas for ZGNR with attached
organic objects, this increases to −2222.71 eV. Figure 3(a)
compares the estimated transmission coefficient (T) for the
bare ZGNR with organic object attached in ZGNR. All
energies are relative to the Fermi energy. It is clear that the
ZGNR exhibits metallic behavior: There is no band gap,
as observed from other tight-binding calculations.9,35,36The
Fermi level (Ef= 0) lies at the midgap, and the states
contributing to the conduction (contributing states) have a
−2 −1.5 −1−0.50 0.51 1.52
0
0.5
1
1.5
2
2.5
3
Transmission
E−Ef (eV)
Bare ZGNR
ZGNR + bio−object
(a) Transmission for bare ZGNR (6,0) of ten unit cells and the
same ZGNR with attached organic-fragment
−1 −0.8−0.6−0.4−0.2 00.20.4 0.6 0.81
0
500
1000
1500
Density of state (1/eV)
E−Ef (eV)
Bare ZGNR
ZGNR + bio−object
(b) Density of states
FIG. 3. (Color online) Comparison of (a) the transmission for
bare ZGNR (6,0) of ten unit cells and the same ZGNR with attached
organic object as a function of energy at equilibrium. The calculation
was performed at ambient temperature T = 300 K. The BSE for bare
ZGNR is found as −1839.04 eV, whereas for ZGNR with attached
organic molecule, this increases to −2222.71 eV. At the energy range
of about ±0.2 eV, there is one conducting channel which results
in a unit transmission coefficient. However, the transmission drops
down about 35% due to the attachment of the organic object at the
Fermi level. (b) DOS for the bare (6,0) ZGNR and the same ZGNR
with attached polyaromatic hydrocarbons. It can be observed that
there is a large reduction of states near to the Fermi level due to the
attachment of hydrocarbons. This results in the reduction of the NDR
phenomenon of ZGNR.
range of energy of about ±0.2 eV. A decrease in T in the
energies corresponding to the gap states is observed when
the polyaromatic hydrocarbon is attached to the bare ZGNR.
This may be explained by the hydrocarbon competing with
the ZGNR for the states injected from the electrodes. The
coupling of ZGNR with the organic object also reduces the
gap state, which can be further demonstrated through Fig. 4.
045401-3

Page 4

CHOWDHURY, ADHIKARI, REES, WILKS, AND SCARPAPHYSICAL REVIEW B 83, 045401 (2011)
0 0.10.2 0.30.40.50.6 0.7 0.8 0.91
0
1
2
3
4
5
6
7
8
9x 10−5
Current (A)
Voltage (V)
Bare ZGNR
ZGNR + bio−object
00.05 0.10.150.2
0
0.5
1
1.5
2x 10−5
FIG. 4. (Color online) I-V characteristics for the bare ZGNR
(6,0) of ten unit cells in length and the same ZGNR with attached
polyaromatic hydrocarbons. The inset figures show the NDR of
ZGNR and ZGNR with attached organic molecules. The origin of
NDR in this system is due to the large reduction in the transmission
when the bias exceeds about 0.13 V. It is also found that the
attachmentofhydrocarbonsreducestheNDRofZGNR.Theoriginof
thisreductioncanbetracedbackfromthetransmission[seeFig.3(a)]
and the DOS [see Fig. 3(b)].
The current (I)-voltage (V) characteristics of the (6,0) ribbon
are shown in Fig. 4 for the bare ribbon and for the same ribbon
with an attached organic object. It is clear that in both cases
the current increases with bias, then virtually saturates until
0.2V.WhenV exceeds0.2,thecurrentstartsagaintoincrease
with the applied bias. This indicates the onset of a new carrier
transport mechanism. It is noticed that the attaching of the
aromatic molecule results in a small reduction in current when
V ? 0.2, while it has a significant effect for higher bias levels.
It induces a larger decrease in the current for V ? 0.2, but
also causes a clear increase in the current when V exceeds
0.2. To understand the presented result, we now explain the
large transmission coefficients presented in the figures. The
transmission probability for a single incoming electron wave
is always between 0 and 1. The transmission spectrum, plotted
in this study, is not a spectrum of the probability of a single
electron being transmitted through the two-probe. It is the sum
of the probability of all the single modes being transmitted
through the two-probe, so if there, at a given energy, are three
blockwavescalledA,B,andC,thenthetransmissionspectrum
is the sum of the probability of A,B, and C added together.
A similar characteristic for bare graphene was reported in a
recentstudy.37Nowweexplainthemechanismofthetransport
phenomenon for an explanation for the features of the (I-V)
characteristics.
When the applied bias is lower than the band gap of the
system (note that band gap refers to the electronic structure
of the ZGNR) in a scattering region, electrons can move from
the left electrode to the right electrode in the free states of the
conduction band creating a left-to-right current (LRC). In a
similar manner, electrons can move into the free states in the
left electrode from the conduction band of the right electrode
creating a right-to-left current (RLC). However, the current
due to the former mechanism will be greater than the latter
mechanism. This is caused mainly by the energy barrier. As
the bias increases, the energy barrier increases and therefore
RLC decreases. Eventually, LRC gets saturated. Electrons in
the valence band of the left electrode cannot tunnel to the
right electrode until the applied bias is less than the band
gap of the system in the scattering region. This is due to the
unavailability of states at the corresponding energies on the
right electrode. As the applied bias is increased and exceeds
thebandgap,freestatesoftherightelectrodebecomeavailable
andthenstartcontributingtotheelectrode-electrodetunneling.
This mechanism makes a sudden increase in the current when
V ? 0.3.
The inset figures of Fig. 4 show the negative differential
resistance (NDR) phenomenon9,38–40in the I-V curves of
ZGNR and ZGNR with attached organic molecules. To
understand the origin of NDR in this system, in Fig. 4 we
compare the transmission coefficients at the current peak and
valley voltages. We observe NDR in the I-V characteristics
as has been previously observed in nanotube and ribbon
structures with a zigzag chain configuration.9,38–41The origin
of this effect is due to the need for parity conservation of
the incoming and outgoing electronic wave functions. It can
also be argued that the existence of the negative differential
resistance region is due to the gap in the transmission
coefficient.41As one can see from the figure, there is a large
reduction in the transmission when the bias exceeds about
0.13 V. It is also found that the attachment of hydrocarbons
reduces the NDR of ZGNR. The origin of this reduction can
be traced back from the transmission [see Fig. 3(a)] and the
density of states (DOS) [see Fig. 3(b)]. The DOS shown in
Fig. 3(b) is integrated over the contributing states energies
described previously. The state significantly drops down due
to the coupling of hydrocarbon to the ZGNR. Moreover, the
calculated charge of bare ZGNR is about 207.96q (where q
is the electronic charge), compared to a charge of 252.46q in
the case of ZGNR with attached organic objects. Further study
is needed to see the effect of the attachment of an aromatic
molecule with a semiconducting ribbon (e.g., AGNR), which
is the focus of future work.
To study the variation of the DOS and transport properties
with the length and width of the ZGNR system, we have
calculated the DOS and transport by varying the lengths and
widths of the ribbon. Figures 5 and 6 present the variation of
differentpropertiesfordifferentlengthsandwidths.Thestrong
dependence of the transmission and DOS on the ribbon can be
observedfromthepresentedfigures.Thismaybeexplainedby
thefactthatthebandgapofthebaregrapheneribbondecreases
with increasing ribbon width.37In a recent CNT-based sensor
study,5it was reported that the current of the bare tube is
larger (at the same bias) for larger-radius tubes due to their
smallerbandgap.Thecurrentiscalculatedastheintegralofthe
transmission. Therefore, following the similar argument made
for the CNT-based sensor, the current in the ribbon increases
with ribbon width, which, in turn, demonstrates the increasing
ofthetransmissionwithincreasingribbonwidth.Furthermore,
the strong rise in the transmission upon increasing the widths
of the ribbons were reported in very recent studies.37,42,43
Next, we consider the effect of the concentration of adsorbed
045401-4

Page 5

GRAPHENE-BASED BIOSENSOR USING TRANSPORT ...
PHYSICAL REVIEW B 83, 045401 (2011)
−1−0.8−0.6 −0.4−0.20 0.2 0.40.6 0.81
0.5
1
1.5
2
2.5
3
Transmission
E−Ef (eV)
N = 12
N = 18
N = 24
(a) Transmission spectrum
−1−0.8 −0.6 −0.4−0.20 0.20.40.6 0.81
0
2000
4000
6000
8000
10000
12000
Density of state (1/eV)
E−Ef (eV)
N = 12
N = 18
N = 24
(b) Density of states
FIG. 5. (Color online) Comparison of (a) the transmission for
ZGNR, as a function of energy at equilibrium and (b) DOS. The
width of the ribbon [see Fig. 1(a)] is kept as M = 6, while the length
is varying 4 [e.g., N is increasing as in Fig. 1(a)].
molecules on the conductance. Figure 7 presents the variation
of conductance and the DOS on absorbed molecules. Note
that the first three molecules are attached sequentially in one
side edge of the graphene and the other three molecules are
attached sequentially at the other edge of the graphene. It can
be observed that, for the first three molecules, there is very
little change in transmission. However, due to the attachment
of molecules at the other edges, transmission/conductance
drops down significantly. This can be explained as follows.
In the structure there are two conductance paths for electronic
transfer on either side of the ribbon. Due to the absorption
of molecules, the local density of electronic states along the
conductance path available for conduction decreases. In this
study we have considered the first three molecules to be
absorbed on one side of the ribbon, while the remaining three
are added to the other side. The first three molecules block
conduction in one conductance path, however, the other path
−1 −0.8 −0.6 −0.4−0.20 0.20.4 0.60.81
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Transmission
E−Ef (eV)
M = 6
M = 8
M = 10
M = 12
(a) Transmission spectrum
−1 −0.8 −0.6−0.4−0.20 0.2 0.4 0.6 0.81
0
2000
4000
6000
8000
10000
12000
Density of state (1/eV)
E−Ef (eV)
M = 6
M = 8
M = 10
M = 12
(b) Density of states
FIG. 6. (Coloronline)Comparisonof(a)transmissionforZGNR,
as a function of energy at equilibrium and (b) DOS. The length of the
ribbon [see Fig. 1(a)] is kept as N = 12, while the width is varying
[e.g., M is increasing as in Fig. 1(a)].
is available for conduction so the transmission is only slightly
reduced.Whenthemoleculesattachtothesecondconductance
path, waves are blocked in this conductance path. This causes
the transmission to reduce significantly. Due to this fact, we
observedastrikinglydifferentbehaviorinthisgraphene-based
sensorstudy.Clearly,whenactingasasensor,ingeneral,there
is no control over to which edge the molecules might attach.
Forthisreason,westudiedattachedmoleculesonbothedgesof
the graphene ribbon.
B. Effect of doping
In this section we examine the effect of doping on ZGNR
with attached organic objects. The BSE for boron doped at
the center is found as −2183.06 eV, whereas for doping
at the edge, it is about −2192.50 eV. Clearly there is no
significant difference in band energies. However, compared
045401-5

Page 6

CHOWDHURY, ADHIKARI, REES, WILKS, AND SCARPAPHYSICAL REVIEW B 83, 045401 (2011)
−1 −0.8 −0.6−0.4−0.20 0.2 0.4 0.6 0.81
0
0.5
1
1.5
2
2.5
3
Transmission
E−Ef (eV)
Mass = 1
Mass = 2
Mass = 3
Mass = 4
Mass = 5
Mass = 6
(a) Transmission spectrum
−1 −0.8−0.6−0.4 −0.20 0.20.40.6 0.81
0
2000
4000
6000
8000
10000
12000
Density of state (1/eV)
E−Ef (eV)
Mass = 1
Mass = 2
Mass = 3
Mass = 4
Mass = 5
Mass = 6
(b) Density of states
FIG. 7. (Color online) (a) Effect of concentration of adsorbed
molecules on the conductance. The attached molecules are arranged
alonglength[M = 18asinFig.1(a)].Inthefigure,Mass=6denotes
that six molecules are attached to the ribbon. The calculated total
energies for different configurations are as follows: −20378.48 eV,
−22044.70 eV, −23723.85 eV, −25356.74 eV, −26928.01 eV, and
−28528.93 eV for Mass = 1,2,3,4,5, and 6, respectively. Note that
the first three molecules are attached sequentially in one edge of
the graphene and the other three molecules are attached sequentially
at the other edge of the graphene. It can be observed that, for the
first three molecules, there is very little change in transmission.
However, due to the attachment of molecules at the other edges,
transmission/conductance drops down significantly. (b) Density of
states.
to the undoped system the band energy of the doped system
reduces to about 30 eV. Similarly, for nitrogen doping at
the center, the BSE is found as −2264.61 eV, whereas for
doping at the edge, it is about −2250.89 eV. This clearly
demonstrates that nitrogen doping increases the band energy
by about 42 eV. Figure 8 shows the I-V characteristics of the
(6,0) ribbon with an attached organic fragment, both boron
and nitrogen doping. We observe the same trend of the current
0 0.10.2 0.30.40.50.6 0.7 0.80.91
0
1
2
3
4
5
6
7
8
9x 10−5
Current (A)
Voltage (V)
Central doping
Edge doping
Undoped
0 0.050.1 0.15 0.2
0
0.5
1
1.5
2x 10−5
(a) Doping with Boron
0 0.1 0.2 0.30.4 0.50.6 0.70.80.91
0
1
2
3
4
5
6
7
8
9x 10−5
Current (A)
Voltage (V)
Central doping
Edge doping
Undoped
00.05 0.10.15 0.2
0
0.5
1
1.5
2x 10−5
(b) Doping with Nitrogen
FIG. 8. (Color online) I-V characteristics (a) for boron and
(b) nitrogen doped at the center and edge of ZGNR with attached
polyaromatic hydrocarbon. The inset figures shows the NDR of
ZGNRwithattachedorganicmoleculesalongwiththecorresponding
doped system. We notice that the effect of the edge doping has a
strictly increasing pattern, whereas for central doping the current
is virtually bias independent from V = 0.1 to V = 0.2 and NDR
phenomenon can be observed. It can also be observed that current
transmission due to edge doping is much greater compared to the
undoped system at V ? 0.2. In contrast, central doping makes a
reductionincurrenttransmission.However,thesmallappliedvoltage
transmission of current in a doped system is always smaller than the
undoped system.
with increasing bias as mentioned previously. We notice that
the effect of the edge doping has a strictly increasing pattern,
whereasthecentraldopingcurrentisvirtuallybiasindependent
from V = 0.1 to V = 0.2 and NDR phenomenon can be
observed. It can also be observed that current transmission
due to edge doping is much greater compared to the undoped
system at V ? 0.2. In contrast, central doping makes a
045401-6

Page 7

GRAPHENE-BASED BIOSENSOR USING TRANSPORT ...
PHYSICAL REVIEW B 83, 045401 (2011)
reduction in current transmission. For small applied voltage,
however, the transmission of current in the doped system
is always lower than the undoped system. We also noticed
that the doping of nitrogen at the ribbon edge is significantly
higher than others. This may be attributed to the fact that
edge nitrogen is able to inject more states into the system.
The calculated charge of central and edge doping due to
boron is about 251.59q and 251.48q, respectively, whereas
due to nitrogen doping it is about 253.39q and 253.44q,
respectively.
IV. CONCLUSION
In this work we demonstrated that GNR shows a clear
change in conductance in response to the attachment of an
aromatic molecule. We also studied the effect of doping on
the energy states as well as the transport properties of GNR
with attached organic objects. There exist different transport
mechanisms depending on the applied bias. These type of
structures seem to be useful to describe, qualitatively, the
effects on the transport properties of ZGNR when organic
molecules are attached to the ribbon edges. The energy states
and transmission of the ZGNR suggests that ZGNR can
be used as a spectrograph sensor device. Additionally, the
significant effect of doping on these quasi-one-dimensional
systems can be observed in the transmission spectrum. Based
on these results, one may propose an extended and more
detailed study of these nanostructures acting as nanosensor
devices. An interesting task would be to investigate the
effect of a large number of molecules randomly distributed
along the ribbon edges on transport properties of these
hybrid systems. A systematic analysis following this line
may be useful to determine the type and concentration
of foreign entities which could be detected with these
kinds of structures.
ACKNOWLEDGMENTS
RC acknowledges the support of the Royal Society through
the award of aNewton International Fellowship. SAgratefully
acknowledges the support of the Leverhulme Trust for the
award of the Philip Leverhulme Prize.
*r.chowdhury@swansea.ac.uk
†s.adhikari@swansea.ac.uk
1G. Gruner, Anal. Bioanal. Chem. 384, 322 (2006).
2K. Balasubramanian and M. Burghard, Anal. Bioanal. Chem. 385,
452 (2006).
3J. Wang, Electroanalysis 17, 7 (2005).
4I. Heller, A. M. Janssens, J. Mannik, E. D. Minot, S. G. Lemay, and
C. Dekker, Nano Lett. 8, 591 (2008).
5G. B. Abadir, K. Walus, and D. L. Pulfrey, Nanotechnology 21,
015202 (2010).
6M. Ezawa, Phys. Rev. B 73, 045432 (2006).
7A. Hirsch, Angew. Chem., Int. Ed. 48, 6594 (2009).
8H. Santos, L. Chico, and L. Brey, Phys. Rev. Lett. 103, 086801
(2009).
9V. N. Do and P. Dollfus, J. Appl. Phys. 107, 063705
(2010).
10B. Biel, X. Blase, F. Triozon, and S. Roche, Phys. Rev. Lett. 102,
096803 (2009).
11Y. Li, Z. Zhou, P. Shen, and Z. Chen, J. Phys. Chem. C 113, 15043
(2009).
12E.-J. Kan, Z. Li, J. Yang, and J. G. Hou, J. Am. Chem. Soc. 130,
4224 (2008).
13L.Rosales,M.Pacheco, Z.Barticevic,A.Latg´ e,andP.A.Orellana,
Nanotechnology 19, 065402 (2008).
14D. V. Kosynkin, A. L. Higginbotham, A. Sinitskii, J. R. Lomeda,
A. Dimiev, B. K. Price, and J. M. Tour, Nature (London) 458, 872
(2009).
15E. Prodan and R. Car, Phys. Rev. B 80, 035124 (2009).
16Y.-Y. Zhang, J.-P. Hu, X. C. Xie, and W. M. Liu, Physica
B-Condensed Matter 404, 2259 (2009).
17N. M. R. Peres, J. Phys.: Condens. Matter 21, 323201 (2009).
18X. Du, I. Skachko, A. Barker, and E. Y. Andrei, Nature Nanotech-
nology 3, 491 (2008).
19J. A. Rogers, Nature Nanotechnology 3, 254 (2008).
20H. B. Heersche, P. Jarillo-Herrero, J. B. Oostinga, L. M. K.
Vandersypen, and A. F. Morpurgo, Nature (London) 446, 56
(2007).
21J. Park, H. Yang, K. S. Park, and E.-K. Lee, J. Chem. Phys. 130,
214103 (2009).
22S. K. Maiti, Solid State Commun. 149, 973 (2009).
23K. Nakada, M. Fujita, G. Dresselhaus, and M. Dresselhaus, Phys.
Rev. B 54, 17954 (1996).
24T. B. Martins, R. H. Miwa, A. J. R. da Silva, and A. Fazzio, Phys.
Rev. Lett. 98, 196803 (2007).
25M. Brandbyge, J. Mozos, P. Ordejon, J. Taylor, and K. Stokbro,
Phys. Rev. B 65, 165401 (2002).
26A. Grigoriev, N. V. Skorodumova, S. I. Simak, G. Wendin,
B. Johansson, and R. Ahuja, Phys. Rev. Lett. 97, 236807 (2006).
27J. Taylor, H. Guo, and J. Wang, Phys. Rev. B 63, 245407 (2001).
28J. Perdew and A. Zunger, Phys. Rev. B 23, 5048 (1981).
29N.Troullier and J.Martins,
(1991).
30J. Soler, E. Artacho, J. Gale, A. Garcia, J. Junquera, P. Ordejon,
and D. Sanchez-Portal, J. Phys.: Condens. Matter 14, 2745
(2002).
31G. B. Abadir, K. Walus, and D. L. Pulfrey, Appl. Phys. Lett. 94,
176101 (2009).
32D. Michalska, L. J. Schaad, P. Carsky, B. A. Hess, and C. S. Ewig,
J. Comput. Chem. 9, 495 (1988).
33Y. Meir and N. S. Wingreen, Phys. Rev. Lett. 68, 2512 (1992).
34J. Taylor, H. Guo, and J. Wang, Phys. Rev. B 63, 245407
(2001).
35S. Reich, J. Maultzsch, C. Thomsen, and P. Ordejon, Phys. Rev. B
66, 035412 (2002).
36B. Gharekhanlou, M. Alavi, and S. Khorasani, Semicond. Sci.
Technol. 23, 075026 (2008).
37U. Treske, F. Ortmann, B. Oetzel, K. Hannewald, and F. Bechstedt,
Phys. Status Solidi A 207, 304 (2010).
Phys. Rev.B
43,1993
045401-7

Page 8

CHOWDHURY, ADHIKARI, REES, WILKS, AND SCARPAPHYSICAL REVIEW B 83, 045401 (2011)
38H. Cheraghchi and K. Esfarjani, Phys. Rev. B 78, 085123 (2008).
39H. Ren, Q.-X. Li, Y. Luo, and J. Yang, Appl. Phys. Lett. 94,
173110 (2009).
40H. Cheraghchi and H. Esmailzade, Nanotechnology 21, 205306
(2010).
41D. Dragoman and M. Dragoman, Appl. Phys. Lett. 90, 143111
(2007).
42D. Basu, M. J. Gilbert, L. F. Register, S. K. Banerjee, and
A. H. MacDonald, Appl. Phys. Lett. 92, 042114 (2008).
43F. Schwierz, Nature Nanotechnology 5, 487 (2010).
045401-8