Science of the Total Environment 574 (2017) 13791388

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Review

Novel carbon nanotube (CNT)-based ultrasensitive sensors for trace

mercury(II) detection in water: A review
Lok R. Pokhrel a,, Nicholas Ettore a, Zachary L. Jacobs a, Asha Zarr a, Mark H. Weir a, Phillip R. Scheuerman b,
Sushil R. Kanel c, Brajesh Dubey d
a
Division of Environmental Health, Department of Epidemiology and Biostatistics, College of Public Health, Temple University, 1301 Cecil B. Moore Avenue, Ritter Annex, Philadelphia, PA 19122,
USA
b
Department of Environmental Health, College of Public Health, East Tennessee State University, Johnson City, TN 37604, USA
c
Department of Systems Engineering and Management, Air Force Institute of Technology, 2950 Hobson Way, Wright-Patterson AFB, OH 45433, USA
d
Indian Institute of Technology Kharagpur, Kharagpur, West Bengal 721320, India

H I G H L I G H T S G R A P H I C A L A B S T R A C T

Hg pollution via natural and man-made

sources is of global public health con-
cern.
CNTs' potential to detect trace level of
solvated Hg2 + in water is critically
reviewed.
Five major types of CNT-based sensors
are identied and discussed.
Sensor designs and testing protocols
described here shall inform future re-
search.
Low-cost, facile and highly sensitive
sensors are anticipated in the future.

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

Article history: Infamous for Mad hatter syndrome and Minamata disease, mercury (Hg) is ranked high on the Agency for
Received 22 May 2016 Toxic Substances and Disease Registry's priority list of hazardous substances for its potent neurologic, renal,
Received in revised form 7 August 2016 and developmental toxicities. Most typical exposures are via contaminated water and food. Although regulations
Accepted 7 August 2016
and advisories are exercised at various levels, Hg pollution from both natural and anthropogenic sources has
Available online 16 August 2016
remained a major public health and safety concern. Rapid detection of solvated aqueous Hg2+ ions at low levels
Editor: D. Barcelo is critical for immediate response and protection of those who are vulnerable (young children, pregnant and
breast-feeding women) to acute and chronic exposures to Hg2+. Various types of sensors capable of detecting
Keywords: Hg in water have been developed. In particular, the novel use of engineered carbon nanotubes (CNTs) has gar-
Carbon nanotubes nered attention due to their specicity and sensitivity towards Hg2+ detection in solution. In this focused review,
Mercury we describe the sensitivity, selectivity and mechanisms of Hg2+ ion sensing at trace levels by employing CNT-
Sensor based various sensor designs, and appraise the open literature on the currently applied and proof-of-concept
Potentiometric methods. Five different types of CNT-based sensor systems are described: potentiometric, DNA-based uores-
Colorimetric
cence, surface plasmon resonance (SPR), colorimetric, and stripping voltammetric assays. In addition, the

Corresponding author.
E-mail address: tuf80999@temple.edu (L.R. Pokhrel).

http://dx.doi.org/10.1016/j.scitotenv.2016.08.055
0048-9697/ 2016 Elsevier B.V. All rights reserved.
1380 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388

DNA-uorescence recognized merits and shortcomings for each type of electrochemical sensors are discussed. The knowledge from
Surface plasmon resonance this succinct review shall guide the development of the next generation CNT-based biochemical sensors for rapid
Voltammetric Hg2+ detection in the environment, which is a signicant rst step towards human health risk analysis of this
legacy toxicant.
2016 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1380
1.1. Carbon nanotubes (CNTs) as a novel sensor material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1380
1.2. Environmental Hg contamination and its detection using CNTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1380
2. Literature search and synthesis of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
3. CNT-based nanosensors for aqueous Hg2 + detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
3.1. CNT-based potentiometric sensor for aqueous Hg2 + detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
3.1.1. Sensing mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
3.1.2. Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
3.1.3. Shortcomings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
3.2. CNT-based DNA hybrid uorescent sensor for aqueous Hg2 + detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
3.2.1. Sensing mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
3.2.1.1. Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
3.2.2. Shortcomings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.3. CNT-based colorimetric sensor for aqueous Hg2 + detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.3.1. Sensing mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.3.2. Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.3.3. Shortcomings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.4. CNT-based surface plasmon resonance (SPR) sensor for aqueous Hg2 + detection . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.4.1. Sensing mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.4.2. Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
3.4.3. Shortcomings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
3.5. CNT-based stripping voltammetric sensor for aqueous Hg2 + detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
3.5.1. Sensing mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
3.5.2. Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
3.5.3. Shortcomings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
4. Summary and future outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387

1. Introduction
1.1. Carbon nanotubes (CNTs) as a novel sensor material
Considered an allotropic form of carbon, carbon nanotubes (CNTs)
generally appear in two forms: those with a single layer of sp2 carbon
sheet rolled into a tube referred as single-wall carbon nanotubes
(SWCNTs), and those with multiple layers of same or different chiral-
ities stacked together forming a tube called multi-wall carbon nano-
tubes (MWCNTs) (see Fig. 1; Balasubramanian and Burghard, 2005;
Gao et al., 2012). Because of their various unique physicochemical prop-
erties, including electrical, optical, thermal, and mechanical properties,
CNTs have found wide applications in catalyst, automotive (stronger
and lighter body parts for cars, aircrafts and spaceships), semiconductor
(nanocircuitry, exible display, transistor) and energy harvesting indus-
tries (solar cells and fuel cells as highly conductive lms, lithium ion
batteries) (Shatkin, 2008; De Volder et al., 2013). Mixing CNTs with
polymers or precursor resins allows for engineering stiff and tough
products without compromising useful mechanical properties and in-
creasing material damping at the same time. Resin mixes have been suc-
cessfully applied to enhance product performance, for example, in
sporting goods (e.g., baseball bats, tennis rackets) and ber composites
such as in lightweight wind turbine blades. CNTs can be added to organ-
ic precursors to improve strength and stiffness in carbon bers, thus
allowing for applications such as lightning-strike protection, deicing,
and structural integrity monitoring for aircraft, among others (De
Volder et al., 2013). CNTs can also be used as a ame-retardant for plas-
tics and could serve as a good alternative to halogenated ame retar-
dants, which are restricted due to their known hazards in the
environment (De Volder et al., 2013). CNTs are also used as a multi-pur-
pose coating; for example, when ship hulls are painted with CNT-laden
paints they inhibit algal and barnacle attachments. In addition, coated
metals could promote stiffness and strength and provide an electronic
pathway for cathodic protection while discouraging corrosion (De
Volder et al., 2013). Owing to high adsorption and/or reactivity, CNTs
have shown potential for selective removal of contaminants from the
environment (air, water), including salt removal during desalination
or reverse osmosis (Zhao et al., 2014; Inukai et al., 2015). More recently,
CNTs have found potential novel applications in sensors for specic de-
tection of trace levels of environmental contaminants including heavy
metals such as lead (Pb), mercury (Hg), and cadmium (Cd) (Shatkin,
2008; Kim et al., 2009).
1.2. Environmental Hg contamination and its detection using CNTs
Hg is a ubiquitous non-essential trace metal known for its toxicity to
plants, animals and humans (USEPA, 2004; WHO, 2005). Sources of Hg
in the environment are both natural and anthropogenic (USGS, 2000).
Worldwide while natural sources tend to make up to 40% of the total
Hg in air, anthropogenic sources contribute to the remaining 60%
(TVA, 2002). Despite present as vapor in atmosphere, the dominant
source of mercury toxicity has been through the consumption of sh
 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388 1381

Fig. 1. Different types of carbon nanotubes (CNTs) (e.g., zigzag, armchair and chiral) that can be engineered from rolling-up a graphene sheet (a), and a multiwalled CNT made from three
layers of CNTs each with unique chirality (b). Images are reprinted with permission from Balasubramanian and Burghard, 2005, copyright @ The Wiley-VCH Verlag GmbH &Co. KGaA,
Weinheim, 2005.

and shellsh grown in Hg contaminated waters in different parts of the
world (USEPA, 2004). In the United States alone, Hg contamination of
fresh water sources is cited for over 77% of the sh advisories in the con-
tiguous 46 states (USEPA, 2004).
Water contamination of Hga highly toxic heavy metalis a legacy
problem worldwide (USEPA, 2004; WHO, 2005). Although regulations
and efforts are exercised at various levels, Hg pollution from both natu-
ral and anthropogenic sources has remained a major public health and
safety concerns (ATSDR, 1999; NRC, 2000; WHO, 2005). Because solvat-
ed aqueous mercuric ions (Hg2+) are environmentally more persistent
and can be transformed to methylated forms (e.g., CH3-Hg, [CH3]2-Hg)
that are more bioaccumulative (ATSDR, 1999; WHO, 2005), their rapid
detection at low levels is critical for rapid response and protection of
the public, in particular, sensitive populations such as children, preg-
nant and breast-feeding women.
Techniques often used for Hg2+ detection in aqueous phase are lim-
ited for many reasons:
(1) The need to hire trained personnel capable of operating the
specialized systems (e.g., Graphite furnace-atomic absorption
spectroscopy (GF-AAS), Inductively coupled-plasma-mass
spectrometry (ICP-MS));
(2) Lengthy and cumbersome protocols;
(3) Longer time needed for sample preparations;
(4) Inherent uncertainty in calibration standards as supplied by
the vendors;
(5) Ensuing calibration curves which often times are curvilinear
(versus linear as what is typically expected) and therefore suf-
fer from errors when extrapolations are made outside the cal-
ibration range.
(6) Method and/or instrument detection limit (DL) further limits
the measurement of trace levels of Hg in water as instrument
such as GF-AAS allows for measuring Hg as low as ppb (g/L)
levels; and
(7) Instrument sensitivity also changes with each new runs.
1382 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388
Therefore, this longstanding issue of low level Hg2 + detection in
water has been recognized as a signicant knowledge gap (Barry et
al., 2009; Kim et al., 2009; Zhang et al., 2010; Guo et al., 2011a, b;
Bagheria et al., 2013; Deng et al., 2010; Mashhadizadeh et al., 2015).
Fast and sensitive alternatives to the conventional techniques for Hg
measurement are available, such as Direct Mercury Analyzer (DMA)
or Atomic Fluorescence Spectroscopy (AFM), but these are yet to be
widely adopted.
Researchers have developed novel approaches that combine high
selectivity, sensitivity and simplicity to precisely detect and quantify
trace levels of aqueous Hg2 +. One such method is the use of electro-
chemical nano-enabled sensors, called nanosensors, being developed
using nanomaterials of Au, Ag, Si, and CNTs, including their hybrid nano-
composites (Darbha et al., 2007; Fan et al., 2010; Guo et al., 2011a, b;
Wu et al., 2011; Fu et al., 2012; Zhang et al., 2012; Gong et al., 2013;
Rajabi et al., 2013; Hadavifar et al., 2014; Mohamed and Salam, 2014;
Wei et al., 2014; Sadrolhosseini et al., 2014; Koedrith et al., 2015). In-
spired by the aforementioned importance of CNTs coupled with their in-
creasing emergent properties, we focus our review on the current
understanding of the mechanisms associated with aqueous Hg2+ ions
adsorption and/or binding to CNTs. We present an analysis of the
open literature on the currently applied and proof-of-concept
methods documented for rapid on-site detection of aqueous Hg2+
ions using CNT-based novel detection systems. Methods include poten-
tiometric, DNA-based uorescence, surface plasmon resonance (SPR),
colorimetric, and stripping voltammetric assays. In addition to the
mechanisms for target analyte detection (in this case, Hg2+), the recog-
nized merits and shortcomings for each of the electrochemical sensors
are discussed. The results of this analysis shall guide the development
of novel CNT-based sensors for rapid Hg2+ detection in the environ-
ment, which is a signicant rst step towards the goal of human health
risk analysis of this legacy contaminant.
2. Literature search and synthesis of data
Using Scopus as a primary search engine, we conducted extensive
literature search using key words: carbon nanotube, mercury, sen-
sor, nanosensor, carbon nanotube AND mercury, carbon nanotube
AND sensor, sensor AND mercury, nanosensor AND mercury, car-
bon nanotube AND mercury AND sensor, or carbon nanotube AND
mercury AND nanosensor. These key terms used alone or in combina-
tion enabled us to synthesize useful and targeted information from a set
of relevant articles. Only peer-reviewed articles published in English be-
tween 2011 and January 2016 were downloaded, of which abstracts
were rst reviewed and were subjected for critical analysis if deemed
useful to this work. To further ascertain that no relevant papers were
missed, we used Google Scholar (https://scholar.google.com/) and
Elsevier's ScienceDirect (http://www.sciencedirect.com/) for electronic
searches which provided us relevant articles as old as 2003 and were in-
cluded in this review. Use of carbon nanotube AND mercury AND sen-
sor as key words yielded the most relevant articles (n N 40) that were
critically analyzed in this review. Review of the bibliographies of the re-
trieved articles also provided pertinent references that were included
herein.
3. CNT-based nanosensors for aqueous Hg2+ detection
3.1. CNT-based potentiometric sensor for aqueous Hg2+ detection
3.1.1. Sensing mechanism
Electrochemical (EC) methods for the detection of trace metals have
proven applicable, economical, and efcient. Khani et al. (2010) de-
scribed doping of carbon paste ion-selective electrodes with MWCNTs
for Hg ion detection using potentiometry, a technique that involves
measuring changes in electrical potential of a reference electrode
when immersed in a sample solution. The change in potential with
concentration can be quantied by the Nernst equation (Stock and
Orna, 1989; Feiner and McEvoy, 1994):
E E0 0:592=nlogC 1
where E denotes solution potential, E0 denotes standard electrode po-
tential, n is the number of electrons transferred in the reaction, C repre-
sents concentration of ions in solution, 0.592 is the calculated value
from the Gibbs energy change under non-standard conditions using
equation RT/F, where R denotes gas constant, T denotes temperature
in Kelvin and F denotes Faraday constant (Stock and Orna, 1989;
Feiner and McEvoy, 1994).
Chemically modied carbon paste electrodes (CMCPE) are a combi-
nation of graphite and a pasting liquid binder. Previous construction uti-
lized nonconductive mineral or parafn oil, which proved ineffective
due to limited electrochemical sensing capability (Khani et al., 2010).
The high electrical conductivity, lightweight, and exceptional mechani-
cal strength of MWCNTs thus offer an attractive option for increased
carbon paste stability and electrode sensitivity. Potentiometric mea-
surements were conducted employing CMCPEs coupled with pH/mV
(as indicator electrode) and Ag/AgCl meters (as reference electrode)
for the detection of Hg (@ pH = 3.0) in aqueous samples. Peak electrode
performance was measured at 20% MWCNTs concentration (Khani et
al., 2010). To enhance sensitivity of the potentiometric sensor, 1-(2-
ethoxyphenyl)-3-(3-nitrophenyl) triazene (ENTZ) was added as an ion-
ophore to the aqueous mixture of Hg2+ ions (Khani et al., 2010). The use
of 1-n-butyl-3-methylimidazolium tetrauoroborate (BMIM.BF4) also
enhanced the potentiometric detection of the sensor by mediating lipo-
philic action between ions (Khani et al., 2010). As Hg2+ ions become
more mobilized, it improves the sensitivity of the potentiometric sensor
for Hg2+ ions detection. Khani et al. (2010) reported satisfactory detec-
tion of ionic Hg2+ in the range 98.8103.3% using the proposed sensor.
Mashhadizadeh et al. (2015) recently described similar
methods for Hg2 + ion detection, combining the use of 1,3-bis(2-
methoxybenzene)triazene (BEPT) as a sensitive ionophore.
MWCNTs that were applied had diameter in the range 815 nm.
Upon application of 3.5 mg BEPT to the carbon paste structure, a
prominent increase in the detection of Hg 2 + ions by CMCPEs
was observed. It is suggested that the interfaces of confounding
metal cations are suppressed when BEPT is used as an ionophore,
thus enabling selective detection of Hg 2 + ions in aqueous form
(Mashhadizadeh et al., 2015).
3.1.2. Advantages
MWCNT-ionic liquid-carbon paste electrode is documented as an ef-
fective tool for quantifying low Hg2+ levels (low ppb) in solution. This
detection method is considered sustainable as the electrodes can me-
chanically be polished as desired (Khani et al., 2010). Only by the sev-
enth week of the experiment, apparent wear and tear of the sensors
was observed (Khani et al., 2010). Consistent with the earlier work,
Mashhadizadeh et al. (2015) also found no change in the detection
level even by the end of four-week long experiment. One of the salient
characteristics of MWCNTs is that they increase the transduction of the
target chemical (i.e., Hg2+) to the electrical signal, thereby multiplying
response time and working concentration range (Khani et al., 2010).
3.1.3. Shortcomings
The nature of potentiometric measurements requires an addition of
a chemical ionophore (e.g., ENTZ, BMIM.BF4, BEPT) to facilitate detec-
tion of Hg2+ ions quickly and accurately. Use of MWCNTs alone, howev-
er, may result in lower sensitivity. Excessive MWCNT addition could
lessen the uniformity and adhesion of carbon paste, so careful analysis
is advised (Mashhadizadeh et al., 2015).
 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388 1383

3.2. CNT-based DNA hybrid uorescent sensor for aqueous Hg2+ detection
3.2.1. Sensing mechanism
Both labeled and label-free Hg2 + ion-specic oligonucleotide
uorophores are described in the literature (Chiang et al., 2008; Wang
et al., 2008a, 2008b; Wang and Liu, 2008; Ye and Yin, 2008; Guo et al.,
2009; Zhang et al., 2010; Guo et al., 2011a, b). In recent years, progress
in the development and application of CNT-based DNA hybrid uores-
cent sensor has received much attention. Among such advances is the
use of thymine-thymine (T-Ts) mismatches, commonly found within
the duplex DNA, which can be applied as an additive to enhance detect-
ability of the biochemical sensors (see Fig. 2; Zhang et al., 2010). Such
oligonucleotide pairs are the key components of this sensor owing to
their strong afnity to Hg2+ ion binding, unlike with other metal ions,
thus enabling formation of T-Hg2+-T base pairs. Accurate detection of
Hg2+ ions via the application of T-Ts has now been possible. Once the
T-Ts are coupled with the oligonucleotide strands, the detection of
Hg2+ ions within the aqueous solutions is reportedly rapid (Zhang et
al., 2010; Guo et al., 2011a, b).
In a recent experiment, Zhang et al. (2010) employed SWCNTs and a
uoropore (6-carboxyuorescein) labeled T-rich ssDNA (single strand-
ed DNA) to enhance accuracy of CNT-based DNA hybrid sensors.
Whereas the dye-labeled ssDNA would wrap around SWCNTs in the ab-
sence of Hg2+ in the system, thereby transforming into a stable ssDNA-
SWCNTs complex; the presence of Hg2 + in the system, on the other
hand, enhanced the uorescence (emission wavelength = 520 nm) as
determined by uorescence microscopy, thereby conrming the pres-
ence of aqueous Hg2+ in solution (Zhang et al., 2010). Moreover, addi-
tion of aqueous 4.0 M Hg2 + led to 2.7 times greater uorescence
intensity compared to that from ssDNA-SWCNT sample alone (without
Hg2+ ions). At higher Hg2+ ion concentrations (8.0 M), the uores-
cence intensity however plateaued when the equilibrium was reached
(Zhang et al., 2010).
It is shown that the uorescence intensity (em = 520 nm) is log-lin-
early correlated with the soluble Hg2+ ions in the system (Fig. 4B). The
detection range for the DNA hybrid sensor is reported to be in the range
0.058.0 M Hg2+ ions (Zhang et al., 2010). This sensor was also evalu-
ated for selectivity by testing the sensor in the presence of polyvinylpyr-
rolidone (PVP) a polymer with known interference potential for
sensors of this type. Zhang et al. (2010) found that even in the presence
of PVP sensitivity was not adversely affected, thus justifying the selec-
tivity of this sensor.
In another experiment, DNA-intercalating Sybr Green I (SG I) uo-
rescent dye (as a label) was used to reduce background noise and to en-
hance rapid detection of Hg2+ ions (Guo et al., 2011a, b). T-rich ssDNA
adsorbing to SWCNT sidewalls would act as a conjoining uorescent
compound quencher. In the co-presence of additional T-rich ssDNA
with Hg2+ ions, the mismatch interactions between T-T pairs would re-
sult in a double-stranded DNA (dsDNA) (Guo et al., 2011a, b). During
the experiment, the use of a label-free SG I orescent dye improved
the detection of Hg2+ ions and allowed for construction of the cost ef-
fective sensor (Guo et al., 2011a, b).

Fig. 2. Schema depicting mechanism underpinning SWCNTs-based DNA uorescent sensor to specically detect Hg2+ ions in aqueous solution (A). Variation in uorescence emission
spectra of ssDNA/SWCNTs upon addition of increasing concentrations of Hg2+ ions in tris-HCl buffer (B). Differences in uorescence intensity upon addition of different cations (at
50 M) including Hg2+ ions (4 M; C). For more details on the fabrication of this sensor design, please refer to Zhang et al., 2010. Reprinted with permission from Zhang et al., 2010,
Copyright @ The Royal Society of Chemistry 2010.
1384 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388

In a recent experiment, Gong et al. (2013) combined a label-free

DNA with SWCNTs and developed a chemiresistive biosensor for Hg2+
ion detection based on structure- switching DNA. The authors hybrid-
ized amino-PolyT oligonucleotide of 15-bases (5-TTT TTT TTT TTT
TTT-3) with polyA (5-AAA AAA AAA AAA AAA-3) forming a duplex
of polyT-polyA. The polyT-polyA duplex was found to be dehybridized
upon exposure to aqueous Hg2+ ion, inferred from the change in resis-
tance/conductance of the device. At 250 nM aqueous Hg2+ concentra-
tion, the sensor recorded increased source and drain current
suggesting decreased resistance. The authors explained this as the
dehybridization of polyT:polyA duplex resulting the formation of T-
Hg2-T duplex and the release of polyA from SWNTs surface (Gong
et al., 2013). The sensor design is depicted in Fig. 5.

3.2.1.1. Advantages
Generally, this approach shows three excellent features: rst, ssDNA
may or may not need labeled dye to demonstrate high efciency for
uorescence quenching by CNTs. Second, the interaction of ssDNA
with SWNTs and the high afnity binding of thymine to Hg2+ improve
the specicity for Hg2+ detection (Zhang et al., 2010). Third, the use of
SG I (as a dye) holding orescent molecules with strong afnity towards
dsDNA improves Hg2+ detection (Guo et al., 2011a, b). Furthermore, SG
I offers a quantum yield of ~ 0.80 once it connects to dsDNA, which is
100 times as much an yield obtainable from the use of routine sole
dyes (e.g., EtBr) and also incurs no any changes in the DNA strand
(Guo et al., 2011a, b). Label-free structure-switching DNA-SWCNTs sen-
sor is a latest development in the nanosensor design for aqueous Hg2+
detection (Fig. 3; Gong et al., 2013) with appreciable sensitivity (detec-
tion limit 100 nM) and specicity (Fig. 4). Specicity of the sensor was
evaluated by including Ca+2, Mg+2, and Mn+2 along with Hg2+ in the
Fig. 4. Responses of SWNTs chemiresistive label-free biosensor for different metal ions.
The concentrations of the metal ions were 1 M and incubation period was 30 min. Each
data point is an average of measurements from four independent sensors at different
point of time and the error bars represent 1 standard deviation. Reprinted with
permission from Gong et al., 2013. Copyright 2013, AIP Publishing LLC.
same solution. As can be seen in Fig. 4, there was no clear overlap in
the response from the sensors and therefore Gong et al. (2013) conclud-
ed that the sensors were specic to Hg2+ ions. This is a limited specic-
ity trial but does demonstrate the specicity of the sensor to Hg2+ ions.
In view of its simplistic design and ease of construction, these CNT-
based DNA hybrid sensors might be used in the eld to monitor Hg pol-
lution in the natural waters.

Fig. 3. Schematic depiction of the steps involved in SWNT-based label-free biosensor fabrication through covalent functionalization of SWNTs with amino-labeled polyT followed by
hybridization with polyA. For more details on the fabrication of this sensor design, please refer to Gong et al., 2013. Reprinted with permission from Gong et al., 2013. Copyright 2013,
AIP Publishing LLC.
 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388
3.2.2. Shortcomings
Akin to colorimetric sensors, CNT-based DNA hybrid sensors are re-
portedly less sensitive to binding and sensing of heavy metal ions other
than Hg2+ as shown in Fig. 2C & 4.
3.3. CNT-based colorimetric sensor for aqueous Hg2+ detection
3.3.1. Sensing mechanism
Upon addition of Hg2+ ions to the aqueous suspension containing
CNTs, a signicant drop in the absorptive bands of the colorimetric sen-
sor (wavelength N 1000 nm) was observed. The absorptive bands mea-
sured prior to addition of Hg2+ ions showed only subtle deviation in the
spectroscopic measurements (Kim et al., 2009). The product of the stan-
dard potential difference (E0 = E0metal ion E0CNT) between CNTs and
heavy metal ions provides a rationale for the drop in absorption (spec-
tral measurement) during the subsequent addition of Hg2+ ions (Kim et
al., 2009). The band gap coupled with large surface area facilitate spon-
taneous redox chemistry between CNTs and Hg2+ ions making this col-
orimetric sensor a valuable tool for on-site sensing applications (see Fig.
5; Kim et al., 2009). Sensitive measurable detection of Hg2+ was docu-
mented in a range of 10 nM1 mM concentrations (Kim et al., 2009).
High specicity of CNTs to bind with trace levels of Hg2+ (10 nM)
has been demonstrated and veried in tandem using methods such as
eld effect transistors (FETs), eld emission-scanning electron micros-
copy (FE-SEM), and UVvis near infrared (NIR) spectroscopy (Kim et
al., 2009). Control experiments using the mixture containing Ca2 +,
Hg2+ and Na+ reveal high selectivity of CNTs to Hg2+ ions. To ensure
that confounding effects from environmental variables did not inuence
the performance of the colorimetric sensor, FET gating effects were
measured for selective Hg2+ detection (Kim et al., 2009). The FET gating
effects selectively produce specic conductivity in the presence of Hg2+
ions regardless of the co-presence of other heavy metal ions (e.g., Ca2+
and Na2+) in the mixture (Kim et al., 2009).
3.3.2. Advantages
CNTs have the ability to remain stable while sensing target analytes
and provide selective charge transport from adsorbents. Particularly,
SWCNTs show higher afnity for chemical markers, largely due to the
band gap energies stemming from surrounding reactive (i.e., local di-
electric or redox) environments thus facilitating Hg2 + detection in
aqueous solution (Fig. 6). The study results are promising in that
SWCNTs-based colorimetric sensors can effectively detect Hg2+ ions
Fig. 5. Selective detection of Hg2+ ions with CNT-based sensors fabricated according to
Kim et al. (2009). Octadecyltrichlorosilane (OTS) monolayer patterned on glass surface
on which CNT suspension (100 mg/L) was placed; the contact electrodes were coated
with AZ5214 positive photoresist while leaving a 4 m gap as active nanotube channel.
For more details on the fabrication of this sensor design, please refer to Kim et al., 2009.
Reprinted with permission from Kim et al., 2009 (adapted from Kim et al., 2009; Images
reprinted with permission from The American Chemical Society, 2009).
1385
Fig. 6. Plausible mechanisms for the detection of Hg2+ (A) and Pb2+ (B) ions by CNT-
based sensors according to Kim et al. (2009) (adapted from Kim et al., 2009; Images
reprinted with permission from The American Chemical Society, 2009).
at the maximum allowable level (10 nM) in drinking water supply
(Kim et al., 2009).
3.3.3. Shortcomings
As for other heavy metals examined, the negative products as
displayed (Ca2+ = 2.924 V, Pb2+ = 0.126 V vs normal hydrogen
electrode, NHE) make them unfavorable chemical targets for colorimet-
ric CNT sensors (Kim et al., 2009). Owing to the irreversible nature of the
redox reaction occurring between the CNTs and Hg2+ ions, this CNT-
based Hg sensor seems nonreusable.
3.4. CNT-based surface plasmon resonance (SPR) sensor for aqueous Hg2+
detection
3.4.1. Sensing mechanism
A surface plasmon resonance (SPR) is a form of electromagnetic
wave travelling in the direction parallel to the planar metal surface.
Any chemical changes occurring on the metal surface (e.g., adsorption
of molecules to the conducting surface) will alter the refractive index
(RI) of the medium adjacent to the metal boundary, thereby resulting
in resonance angle shift (Fig. 7A; Raether, 1988).
Polypyrrole (PPy) is a polymer used in conjunction with nanoparti-
cles to mediate effective biosensors. Identied as an excellent conduc-
tor, PPy-MWCNT sensor was developed and tested by Sadrolhosseini
et al. (2014) for the detection of Hg, Pb, and Fe in aqueous solutions. Sur-
face plasmon resonance (SPR) is highly feasible technique as it enables
higher sensitivity, stability, and reproducibility of metal ion detection
(Sadrolhosseini et al., 2014). CNT solution synthesized by employing
electrochemical polymerization was deposited onto a thin gold-coated
glass slide (Sadrolhosseini et al., 2014). Increased MWCNTs percentage
facilitates a reduction in the resonance angle (56.971 to 56.293) and a
reectance shift from 0.244 to 0.294 for the SPR test (Sadrolhosseini et
al., 2014). The sensing layers detect through binding to ionic particles.
3.4.2. Advantages
Conductive polymers are favored as they are environmentally
friendly since they do not leach any chemical into the monitored
water. Use of SPR technique combined with MWCNTs would increase
detection sensitivity for Hg ions as compared to other metals such as
Pb and Fe analyzed (Sadrolhosseini et al., 2014). Stability of PPy-
MWCNT has been proven as solvents such as oil and fuel do not dissolve
and/or disintegrate the composite (Sadrolhosseini et al., 2014). Al-
though analysis using ICP-MS can offer higher accuracy compared to
1386 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388

Fig. 7. Schematic showing a SPR setup with a He-Ne laser, a polarizer, a photodiode, a ow cell with the target analyte (Hg[II]), a rotation stage, and a high refractive index prism (A); Angle
shift test for PPy-MWCNT sensor layer vs PPy control layer (B); SPR signal and sensogram related to mixture of: Pb and Fe ions at saturation value (C); Hg, Pb and Fe ions at saturation value
(D); and sensogram for detection of ions in the mixture (E) (adapted from Sadrolhosseini et al., 2014).

the SPR sensor, the latter is cheaper, simpler, and more compact than
ICP and other spectroscopic methods (Sadrolhosseini et al., 2014).
3.4.3. Shortcomings
The enhanced accuracy of SPR sensors is limited to singular trace
metal detection in aqueous solution. Except for Hg ions, the sensitivity
of SPR sensor towards detection of metals such as Pb and Fe is consider-
ably low (detection limit = 10 ppb) (Sadrolhosseini et al., 2014).
3.5. CNT-based stripping voltammetric sensor for aqueous Hg2+ detection
3.5.1. Sensing mechanism
Voltammetry is a method of electrochemical determination of vari-
ous solution phase ionic species at trace levels by measuring current
as a function of voltage applied (Brett et al., 1994). Square wave anodic
stripping voltammetry (SWASV) has been established as an effective
technique for Hg2+ detection, specically when mediated by gold sub-
strate electrodes and CNTs. Compounds such as thiophenol (Ph-SH) and
thiol functionalized chitosan (CS-SH) have been used to minimize con-
founding interactions from other metals in solution (Wei et al., 2014;
Deng et al., 2010). A strong positive correlation (R = 0.9993) between
the concentrations of Hg2+ ions (linear range: 590 nM with a detec-
tion limit of 3 nM) and stripping peak current was attained using
SWASV (Wei et al., 2014). Deng et al. (2010) reported the response of
SWASV to Hg2 + in the range 10140 nM with a detection limit of
3 nM. The sensor exhibited 97103% recovery for Hg2+ ions, suggesting
its satisfactory performance (Wei et al., 2014).
3.5.2. Advantages
The SWASV sensor offers high sensitivity for the detection of trace
ions in solution phase and is equally cost-effective as it allows for reduc-
ing sample analysis time while increasing the sample throughput and
does not require the removal of dissolved oxygen from the sample
(Wei et al., 2014). Strong adsorption of thiol-rich CS-SH to Hg2+ and in-
creased conductivity of the nanocomposite CS-SH-MWCNTs lm ren-
dered high sensor performance towards soluble Hg2+ ions detection
(Deng et al., 2010). Comparative performance and detection limits for
different types of voltammetry-based sensors are presented in Table 1.
3.5.3. Shortcomings
Addition of CNTs is necessary to prevent interference between other
metal ions (Wei et al., 2014). At high negative potentials (N 0.3 V)
other chemicals in solution could also be reduced and may interfere
with the detection of Hg2+. The authors recommend applying the depo-
sition potential of 0.3 V and the use of 1 M HCl for optimal sensitivity
of the sensor device (Deng et al., 2010).
4. Summary and future outlook
Development of rapid, sustainable and yet highly sensitive CNT-
based biochemical nanosensors to specically detect low levels of
Hg2+ in natural water samples is highly desired. Such sensors would
enable limiting environmental exposures to vulnerable populations
from the toxic Hg ions, thereby protecting them from known Hg-in-
duced multi-organ toxicities (ATSDR, 1999). Conventional testing sys-
tems are considered less accurate as they suffer from method/
instrument detection limits and are also cumbersome to operate. This
deciency has provided motivation for this work to assess the efcacy
of recently documented CNT-based sensors for aqueous Hg2+ detection
(Koedrith et al., 2015).
The adsorption or binding of aqueous Hg2+ ions to CNTs allows for
facile detection and could be a more desirable measurement options.
Adoption of the sensors outlined in this mini-review coupled with fu-
ture research aiming to simplify the documented sensor designs and/
or testing protocols might inform cost reduction and improve detection
limits for the sensors. Despite its non-reusable characteristic, the high
 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388
Table 1
Comparison of CNT-based sensors performance with other types of sensors for Hg2+ detection in solution phase; CNT-based sensors are found more sensitive towards trace level (as low as
3 nM) of Hg2+ detection (adapted from Wei et al., 2014).
Types Of sensors used Types of voltammetry used
CS-SH/MWCNT/GCEa ASVh
SWCNT-PhSH/Aub ASV
Graphene-CA/Auc ASV
Silica-SH/GCEd ASV
SAMMS-SH/CPEe ASV
Metallothionein/Auf DPVi
MPS/Aug DPV
a
Glassy carbon electrode modied with thiol-functionalized chitosan-multiwalled carbon nanotubes.
b
Gold electrode modied with single walled carbon nanotubes-thiophenol.
c
Gold electrode modied with Graphene-cysteamine.
d
Glassy carbon electrode modied with thiol-functionalized silica lm.
e
Carbon paste electrode modied with thiol-terminated self-assembled monolayer on mesoporous silica.
f
Gold electrode modied with Metallothionein.
g
Gold electrode modied with Mercaptopropyl functionalized silica.
h
Anodic stripping voltammetry.
i
Differential pulse voltammetry.
specicity of colorimetric sensor makes it a desirable candidate. A drop
in absorption wavelength due to the introduction of CNTs in solution
would enable solvated Hg2 + ions detection despite the presence of
other ions in the mixture (Kim et al., 2009). The success of CNT-based
DNA hybrid sensors for detecting aqueous Hg2 + is ascribed to the
high afnity of Hg2+ ions to form T-Hg2+-T base pairs when exposed
to oligonucleotide pairs within the DNA duplex. Various methods of
DNA addition described herein could offer insights into improving spec-
icity through measurable orescence both with and without labeling
dyes. Conversely, SPR detects any shift in electromagnetic wave related
to change in metal surface characteristics. Coupling a conductive PPy
polymer to MWCNTs, one could detect Hg in aqueous solutions in the
form of a visible angle shift with the SPR test. CNT-based potentiometric
sensors are also known to differentiate electrochemical changes in a so-
lution. Adding CNTs to a liquid-carbon paste electrode is promising
owing to its high afnity binding to detect low Hg2 + levels in water.
Voltammetric sensor using SWASV is found to be highly sensitive
when coupled with gold electrode and CNTs. Measuring current as a
function of voltage applied, the sensor could signal a higher stripping
peak current in a higher Hg2+ ion concentration solution. A limitation
in the use of these as well as other advanced sensors requires an evalu-
ation of their long-term functionality in non-laboratory conditions.
Should such sensors be used in water utility, industry, commercial or
for point of use assessment, these like other conventional sensors
need to be tested for performance for real world conditions and uses.
However, these inventive approaches and progress highlighted in this
review is an important step in sensor development leading towards
inline detection and risk analysis.
In summary, this review assessed the state-of-the-science of CNT-
based Hg sensor technology. Nanosensors capable of precisely detecting
multiple heavy metals including other toxic chemicals within a minis-
cule volume of environmental media and/or in the eld environment
might be the next step desirable for analytical chemists, toxicologists
and risk assessors.
Acknowledgments
LRP gratefully acknowledges and thank Daniel R. Strongin (Temple
University) for the helpful review of the earlier draft. LRP also acknowl-
edges the support from Temple University's College of Public Health
(Grant number 142848).
References
Agency for Toxic Substances and Disease Registry (ATSDR), 1999. Toxicological Prole for
Mercury. U.S. Department of Health and Human Services Available at: http://atsdr1.
atsdr.cdc.gov:8080.
1387
Detection limit (nM) Measurement range (nM) References
3 10140 Deng et al. (2010)
3 590 Wei et al. (2014)
3 540 Zhou et al. (2012)
4.3 10100 Cesarino and Cavalheiro (2008)
15 1008000 Yantasee et al. (2003)
80 150300 Ju and Leech (2000)
100 1001000 Walcarius and Delacote (2005)
Bagheria, H., Afkhami, A., Khoshsafar, H., Rezaei, M., Shirzadmehr, A., 2013. Simultaneous
electrochemical determination of heavy metals using a triphenylphosphine/
MWCNTs composite carbon ionic liquid electrode. Sensors Actuators B 186, 451460.
Balasubramanian, K., Burghard, M., 2005. Chemically functionalized carbon nanotubes.
Small 1, 180192.
Barry, R.C., Lin, Y., Wang, J., Liu, G., Timchalk, C.A., 2009. Nanotechnology-based electro-
chemical sensors for biomonitoring chemical exposures. J. Expo. Sci. Environ.
Epidemiol. 19 (1), 118.
Brett, C.M.A., Lima, J.L.F.C., Garcia, M.B.Q., 1994. Square-wave anodic stripping voltammetry
in stationary and owing solution: a comparative study. Analyst 119 (6), 12291234.
Cesarino, I., Cavalheiro, E.T.G., 2008. Thiol-functionalized silica thin lm modied electrode
in determination of mercury ions in natural water. Electroanalysis 20 (21), 23012309.
Chiang, C.K., Huang, C.C., Liu, C.W., Chang, H.T., 2008. Oligonucleotide-based uorescence
probe for sensitive and selective detection of mercury(II) in aqueous solution. Anal.
Chem. 80, 37163721.
Darbha, G.K., Ray, A., Ray, P.C., 2007. Gold nanoparticle-based miniaturized nanomaterial
surface energy transfer probe for rapid and ultrasensitive detection of mercury in soil,
water, and sh. ACS Nano 1 (3).
De Volder, M.F.L., Tawck, S.H., Baughman, R.H., Hart, J.A., 2013. Carbon nanotubes: pres-
ent and future applications. Science 339 (6119), 535539.
Deng, W.F., Tan, Y., Li, Y., Wen, Y., Su, Z., Huang, Z., Huang, S., Meng, Y., Xie, Q., Luo, Y., Yao,
S., 2010. Square wave voltammetric determination of Hg(II) using thiol functional-
ized chitosan-multiwalled carbon nanotubes nanocomposite lm electrode.
Microchim. Acta 169, 367373.
Fan, A., Ling, Y., Lau, C., Lu, J., 2010. Direct colorimetric visualization of mercury (Hg2+)
based on the formation of gold nanoparticles. Talanta 82 (2), 687692.
Feiner, A.S., McEvoy, A.J., 1994. The Nernst equation. J. Chem. Educ. 71, 493.
Fu, X.C., Wu, J., Nie, L., Xie, C.G., Liu, J.H., Huang, X.J., 2012. Electropolymerized surface ion
imprinting lms on a gold nanoparticles/single-wall carbon nanotube nanohybrids
modied glassy carbon electrode for electrochemical detection of trace mercury (II)
in water. Anal. Chim. Acta 720, 2937.
Gao, C., Guo, Z., Liu, J.-H., Huang, X.-J., 2012. The new age of carbon nanotubes: an updated
review of functionalized carbon nanotubes in electrochemical sensors. Nanoscale 4,
19481963.
Gong, J.L., Sarkar, T., Badhulika, S., Mulchandani, A., 2013. Label-free chemiresistive bio-
sensor for mercury (II) based on single-walled carbon nanotubes and structure-
switching DNA. Appl. Phys. Lett. 102 (1), 013701.
Guo, L., Hu, H., Sun, R., Chen, G., 2009. Highly sensitive uorescent sensor for mercury ion
based on photoinduced charge transfer between uorophore and pi-stacked T-
Hg(II)-T base pairs. Talanta 79, 775779.
Guo, L., Yin, N., Nie, D., Gan, J., Li, M., Fu, F., Chen, G., 2011a. Label-free uorescent sensor
for mercury (II) ion by using carbon nanotubes to reduce background signal. Analyst
136 (8), 16321636.
Guo, Y., Wang, Z., Qu, W., Shao, H., Jiang, X., 2011b. Colorimetric detection of mercury,
lead and copper ions simultaneously using protein-functionalized gold nanoparticles.
Biosens. Bioelectron. 26 (10), 40644069.
Hadavifar, M., Bahramifar, N., Younesi, H., Li, Q., 2014. Adsorption of mercury ions from
synthetic and real wastewater aqueous solution by functionalized multi-walled car-
bon nanotube with both amino and thiolated groups. Chem. Eng. J. 237, 217228.
Inukai, S., Cruz-Silval, R., Ortiz-Medina, R., Morelos-Gomez, A., Takeuchi, K., Hayashi, T.,
Tanioka, A., Araki, T., Tejima, S., Noguchi, T., Terrones, M., Endo, M., 2015. High-perfor-
mance multifunctional reverse osmosis membranes obtained by carbon nanotube-
polyamide nanocomposite. Sci. Rep. 5, 13562.
Ju, H., Leech, D., 2000. Electrochemical study of a metallothionein modied gold disk elec-
trode and its action on Hg2+ cations. J. Electroanal. Chem. 484 (2), 150156.
Khani, H., Rofouei, M.K., Arab, P., Gupta, V.K., Vafaei, Z., 2010. Multi-walled carbon nano-
tubes-ionic liquid-carbon paste electrode as a super selectivity sensor: application to
potentiometric monitoring of mercury ion (II). J. Hazard. Mater. 183 (1), 402409.
Kim, H.T., Lee, J., Hong, S., 2009. Highly selective environmental nanosensors based on
anomalous response of carbon nanotube conductance to mercury ions. ACS 113
(45), 1939319396.
1388 L.R. Pokhrel et al. / Science of the Total Environment 574 (2017) 13791388
Koedrith, P., Thasiphu, T., Weon, J., Boonprasert, R., Tuitemwong, K., Tuitemwong, P.,
2015. Recent trends in rapid environmental monitoring of pathogens and toxicants:
potential of nanoparticle-based biosensor and applications. Sci. World J. 112.
Mashhadizadeh, M.H., Ramezani, S., Rofouei, M.K., 2015. Development of a novel
MWCNTs-triazene-modied carbon paste electrode for potentiometric assessment
of Hg (II) in the aquatic environments. Mater. Sci. Eng. C 47, 273280.
Mohamed, R.M., Salam, M.A., 2014. Photocatalytic reduction of aqueous mercury (II)
using multi-walled carbon nanotubes/Pd-ZnO nanocomposite. Mater. Res. Bull. 50,
8590.
NRC (National Research Council) Committee on the Toxicological Effects of
Methylmercury, 2000. Toxicological Effects of Methylmercury. National Academy
Press, Washington, DC.
Raether, H., 1988. Surface plasmons on smooth and rough surfaces and on gratings.
Springer Tracts Mod. Phys. 111, 1117.
Rajabi, H.R., Roushani, M., Shamsipur, M., 2013. Development of a highly selective
voltammetric sensor for nanomolar detection of mercury ions using glassy carbon
electrode modied with a novel ion imprinted polymeric nanobeads and multi-
wall carbon nanotubes. J. Electroanal. Chem. 693, 1622.
Sadrolhosseini, A.R., Noor, A.S.M., Bahrami, A., Lim, N.H., Talib, A.Z., Mahdi, A.M., 2014. Ap-
plication of polypyrrole multi-walled carbon nanotube composite layer for detection
of mercury, lead and iron ions using surface plasmon resonance technique. PLoS One
9 (4), e93962.
Shatkin, J.A., 2008. Nanotechnology: health and environmental risks. Perspect. Nanotech.
1194.
Stock, J.T., Orna, M.V., 1989. Electrochemistry, past and present. ACS 390.
Tennessee Valley Authority (TVA), 2002. On the Air: Mercury Emissions. Available at:
http://www.tva.gov/environment/air/ontheair/merc_emis.htm.
United States Environmental Protection Agency (USEPA), 2004. Fish Advisories. Available
at: http://www.epa.gov/ost/sh/states.htm.
United States Geological Survey (USGS), 2000. Mercury in the Environment: Fact Sheet
146-00. Available at: http://www.usgs.gov/themes/factsheet/146-00/.
Walcarius, A., Delacote, C., 2005. Mercury(II) binding to thiol-functionalized mesoporous
silicas: critical effect of pH and sorbent properties on capacity and selectivity. Anal.
Chim. Acta 547, 313.
Wang, J., Liu, B., 2008. Highly sensitive and selective detection of Hg2+ in aqueous solu-
tion with mercury-specic DNA and Sybr Green I. Chem. Commun. 39, 47594761.
Wang, Z., Lee, J.H., Lu, Y., 2008a. Highly sensitive turn-on uorescent sensor for Hg2+ in
aqueous solution based on structure-switching DNA. Chem. Commun. 44,
60056007.
Wang, H., Wang, Y., Jin, J., Yang, R., 2008b. Gold nanoparticles-based colorimetric and
turn-on uorescent probe for mercury (II) ions in aqueous solution. Anal. Chem.
80 (23), 90219028.
Wei, J., Yang, D., Chen, H., Gao, Y., Li, H., 2014. Stripping voltammetric determination of
mercury (II) based on SWCNT-PhSH modied gold electrode. Sensors Actuators B
Chem. 190, 968974.
World Health Organization (WHO), 2005. Mercury in Drinking-water: Background Docu-
ment for Development of WHO Guidelines for Drinking-water QUALITY.
Wu, J., Li, L., Zhu, D., He, P., Fang, Y., Cheng, G., 2011. Colorimetric assay for mercury (II)
based on mercury-specic deoxyribonucleic acid-functionalized gold nanoparticles.
Anal. Chim. Acta 694 (1), 115119.
Yantasee, W., Lin, Y., Zemanian, T.S., Fryxell, G.E., 2003. Voltammetric detection of lead(II)
and mercury(II) using a carbon paste electrode modied with thiol self-assembled
monolayer on mesoporous silica (SAMMS). Analyst 128, 467472.
Ye, B.C., Yin, B.C., 2008. Highly sensitive detection of mercury (II) ions by uorescence po-
larization enhanced by gold nanoparticles. Angew. Chem. Int. Ed. 47 (44),
83868389.
Zhang, L., Li, T., Li, B., Li, J., Wang, E., 2010. Carbon nanotubeDNA hybrid uorescent sen-
sor for sensitive and selective detection of mercury(II) ion. Chem. Commun. 46,
14761478.
Zhang, Y., Yuan, Q., Chen, T., Zhang, X., Chen, Y., Tan, W., 2012. DNA-capped mesoporous
silica nanoparticles as an ion-responsive release system to determine the presence of
mercury in aqueous solutions. Anal. Chem. 84 (4), 19561962.
Zhao, H., Qiu, S., Wu, L., Zhang, L., Chen, H., Gao, C., 2014. Improving the performance of
polyamide reverse osmosis membrane by incorporation of modied multi-walled
carbon nanotubes. J. Membr. Sci. 450, 249256.
Zhou, H., Wang, X., Yu, P., Chen, X., Mao, L., 2012. Sensitive and selective voltammetric
measurement of Hg2+ by rational covalent functionalization of graphene oxide
with cysteamine. Analyst 137, 305308.