The Energetics of the Hydrogenation of a Single-Walled Carbon Nanotube
SINGLE WALLED CARBON NANOTUBE-BASED MULTI-JUNCTION … · 2015-10-15 · single walled carbon...
Transcript of SINGLE WALLED CARBON NANOTUBE-BASED MULTI-JUNCTION … · 2015-10-15 · single walled carbon...
SINGLE WALLED CARBON NANOTUBE-BASED MULTI-JUNCTION
BIOSENSOR FOR DETECTION OF FOODBORNE PATHOGENS
A THESIS SUBMITTED TO THE GRADUATE DIVISION OF THE
UNIVERSITY OF HAWAI‘I AT MᾹNOA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
IN
FOOD SCIENCE
AUGUST 2014
By
Kara Yamada
Thesis Committee:
Soojin Jun, Chairperson
Alvin Huang
Yong Li
ii
© 2014, Kara Yamada
iii
ACKNOWLEDGEMENTS
I would like to express my deepest appreciation and gratitude to my committee
chair, Dr. Soojin Jun, who has continuously supported me throughout my education and
research. Thank you for encouraging me to purse a Master’s degree and for giving me the
opportunity to be a part of your Food Processing lab. I truly value and appreciate your
research advice and mentorship. Without your guidance, this thesis would not have been
possible.
I would also like to acknowledge my committee members, Dr. Yong Li and Dr.
Alvin Huang, as well as Dr. Wayne Iwaoka, for their teaching insights and guidance.
Especially Dr. Li for igniting my interest in food microbiology, kindly allowing me
access to your Food Microbiology lab, and providing the bacterial strains used in this
research. Also many thanks to Dr. Jaehyun Chung and colleagues for their research
collaboration and Tina Carvalho for the use of the biological electron microscope lab.
Thank you to my lab members for your friendship, support, assistance, and
insightful advice. Particularly Dr. Won Choi for the tremendous help, advice, and
mentorship throughout my thesis research and Natthakan Rungraeng for your technical
help and microscope assistance. I will always remember our time spent together in room
118.
Last but not least, I would like to thank my family and friends for their continuous
love and support throughout my education.
iv
ABSTRACT
Rapid identification of bacterial pathogens in food is urgently needed to ensure
food safety. Current detection methods do not meet industrial demands in terms of
performance, time, cost, and simplicity. In spite of significant progresses, the
development of a sensitive biosensor for practical applications remains a challenge. In
this study, a single walled carbon nanotube- (SWCNT) based junction sensor was
designed as an alternative detection method for foodborne pathogens. Gold tungsten
wires (Ø: 50 μm) coated with polyethylenimine (PEI) and SWCNTs were aligned to form
a crossbar junction functionalized with streptavidin and biotinylated antibodies. By
coating the wires’ cross section with bio-nano materials, a sandwich of layered SWCNTs
and biomolecules creates a bio-nano junction when targeted bacteria bind and form
immune complex reactions. The parallel SWCNT platform layers convert the molecular
binding events at the junction into measurable electrical current signals. As a result,
changes in electrical current (∆I) after bioaffinity reactions between bacterial cells and
antibodies on the SWCNT surface were monitored to evaluate the sensor’s performance.
Escherichia coli K-12 and Staphylococcus aureus were used as target microorganisms for
single and multi-analyte detection. The SWCNT-based sensing platform generated a ∆I
signal response seven-folds higher in a high concentration of E. coli (108 CFU/mL), than
compared to a junction sensor without SWCNTs. Thereby, an improvement in sensing
magnitude was achieved with SWCNTs. Electrical current measurements from the single
junction sensor demonstrated a linear relationship (R2 = 0.973) between the changes in
current and concentrations of E. coli in range of 102-105 CFU/mL with a detection limit
of 103 CFU/mL and a detection time of 2 min. The design of a portable 2 x 2 multi-
junction sensing array demonstrated an improved sensitivity with a 102 CFU/mL limit of
detection for E. coli (R2 = 0.978) and S. aureus (R2 = 0.992). Microbial cocktail samples
of E. coli and S. aureus showed similar measurement trends for multiplexed detection in
10 µL and 100 µL batch samples. Therefore, the developed label-free SWCNT-based
multi-junction biosensor shows potential as a sensitive and simple device with portable
and multiplexed applications.
v
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ............................................................................................... iii
ABSTRACT ....................................................................................................................... iv
LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
LIST OF ABBREVIATIONS ........................................................................................... xii
1. INTRODUCTION .......................................................................................................... 1
2. LITERATURE REVIEW ............................................................................................... 4
2.1. Introduction ............................................................................................................. 4
2.2. Literature Review ................................................................................................... 4
2.2.1. Foodborne pathogens and illnesses ............................................................. 4
2.2.2. Traditional detection methods and drawbacks ............................................ 5
2.2.3. Biosensor technology for foodborne pathogen detection ............................ 7
2.2.4. Single walled carbon nanotubes for biosensor technology ....................... 10
2.2.5. SWCNT mechanisms for detection of bio-analytes .................................. 14
2.2.6. SWCNT-based biosensors for foodborne pathogen detection .................. 15
2.3. Conclusion overview ............................................................................................. 16
3. MATERIALS & METHODS ....................................................................................... 18
3.1. Sensor design & fabrication materials .................................................................. 18
3.2. SWCNT coating technique .................................................................................... 19
3.3. Antibody immobilization and microbial preparation ............................................ 21
3.4. SWCNT-based biosensor: single junction ............................................................ 22
3.4.1. Device fabrication process ........................................................................ 22
3.4.2. Signal measurements ................................................................................. 23
3.4.3. Sensitivity and specificity testing .............................................................. 25
3.4.4 FESEM imaging ......................................................................................... 25
3.4.5. Data analysis ............................................................................................. 25
3.5. SWCNT-based biosensor: multi-junction ............................................................. 26
vi
3.5.1. Device fabrication process ........................................................................ 26
3.5.2. Multiplex circuit design and system scale down ....................................... 28
3.5.3. Multiplexing sensitivity tests .................................................................... 31
3.5.4. Data analysis ............................................................................................. 33
4. RESULTS & DISCUSSION ........................................................................................ 35
4.1. SWCNT-based single junction biosensor ............................................................. 35
4.1.1. SWCNT dip coating surface morphology ................................................. 35
4.1.2. Characterization of the bio-nano functionalization layers ........................ 39
4.1.3. Sensitivity tests for Escherichia coli K-12 ................................................ 43
4.1.4. Specificity tests against Staphylococcus aureus ....................................... 44
4.2. SWCNT-based multi-junction biosensor .............................................................. 45
4.2.2. Sensitivity tests for E. coli K-12 and S. aureus ......................................... 46
4.2.3. Simultaneous detection of E. coli and S. aureus ....................................... 49
4.2.4. Mathematical circuit for multi-junction resistance calculation ................. 50
5. CONCLUSION ............................................................................................................ 52
REFERENCES ................................................................................................................. 53
vii
LIST OF TABLES
Table 2.1. Main characteristics of some culture-based and rapid detection methods.
.............................................................................................................................7
Table 2.2. Commercial biosensor technologies for foodborne pathogen
detection. ..........................................................................................................10
Table 4.1. Multi-junction sensing array resistance calculated using the
developed mathematical circuit. .......................................................................51
viii
LIST OF FIGURES
Fig. 2.1. Schematic illustration of the main components in a biosensor. .......................... 8
Fig. 2.2. Classification of biosensors based on biorecognition and transducing elements. 9
Fig. 2.3. (a) Schematic of a single graphene sheet rolled up to form a SWCNT. (b)
Graphene sheet illustrating lattice vectors, a1 and a2, and the chiral vector, Ch =
na1 + ma2. The achiral, limiting cases of zigzag (n, 0) and armchair (n, n) are
indicated with thick, dashed lines, and the chiral (θ) angle is measured from the
zigzag direction. The light, dashed parallel lines define the unrolled, infinite
SWCNT (Modified from Odom et al., 2002). (c) SWCNT armchair, zigzag, and
chiral forms (Modified from Iijima, 2002). ......................................................... 12
Fig. 2.4. (a) Structure of graphene’s honeycomb lattice of carbon atoms. (b) Illustration
of graphene bands and its conducting states as a function of the electron wave
vector k. The black hexagon defines the first Brilluoin zone of graphene. There
are no conducting states except along special directions where cones of states
exist. The centers of the cones are defined as the graphene k points. Depending on
the way the graphene vector is rolled up, SWCNTs can either be classified as (c)
a metal, slice passes through the center of a cone, k, or (d) a semiconductor, with a
gap between the filled hole states and the empty electron states (McEuen et al.,
2002). ................................................................................................................... 13
Fig. 2.5. Calculated I-Vlg curves before (black) and after (red) protein adsorption for (a)
electrostatic gating effect corresponding to a shift of the semiconducting bands
downward and (b) Schottky barrier effect that corresponds to a change of the
difference between metal and SWCNT work functions. Insets illustrate the
corresponding changes in the band diagrams for hole and electron doping
respectively (Heller et al., 2008). ......................................................................... 15
Fig. 3.1. SWCNT dip-coating experimental set-up. The stepping motor provided
controlled insertion and withdrawal velocity (vw) for uniform nano coatings. A
maximum of four microwires were coated at a time. .......................................... 20
Fig. 3.2. An illustration of the SWCNT dip-coating process. A constant vw generates a
capillary force between the SWCNT-DMF solution and the microwire. Suspended
ix
SWCNTs flow into the meniscus due to capillary force and the SWCNTs adhere
to the PEI coated wire surface. ............................................................................ 21
Fig. 3.3. Antibody immobilization process (Schematic is not drawn to scale). A micro
volume of streptavidin is applied to the wire junction. Thereafter, biotin
conjugated antibodies are applied and bind to streptavidin. ................................ 21
Fig. 3.4. Schematic of the single junction sensor device. ................................................ 23
Fig. 3.5. (a) Schematic of the single junction sensing system. (b) Single junction circuit.
............................................................................................................................... 23
Fig. 3.6. Antigen-antibody reactions on a bio-nano functionalized junction sensor. (Not
drawn to scale). .................................................................................................... 24
Fig. 3.7. (a) Un-etched copper chip. (b) Etched sensor chip with a PDMS sample well
and copper connector pads. .................................................................................. 26
Fig. 3.8. Schematic of wire soldering (a) set-up and (b) procedure. ............................... 27
Fig. 3.9. Multi-junction sensor chip assembly (a) Two SWCNT coated microwires are
soldered onto the connector pads. (b) Mica spacers are placed on the chip. (c)
Two more coated wires are soldered on top of the mica spacers perpendicular to
the first two wires. (d) Image of assembled sensor chip. ..................................... 28
Fig. 3.10. Switching module circuit design. (a) Single relay circuit made up of a ground
(GND), negative-positive-negative (NPN) transistor, and voltage drain source
(Vdd) is used to direct current (I) into the SDPT relay where it is switched between
normally open (NO) and normally closed (NC) contact, depending on the signal
of I/O Pin. The SDPT controls the connection to the junction biosensor for current
measurement. (b) Multiplexing multi-junction circuit composed of four relays.
Circuits are designed in Auto CAD (2014, San Rafael, CA). .............................. 29
Fig. 3.11. DC power supply circuit. Alternating current (AC) is transformed by the digital
variac and converted into direct current (DC) by the rectifier. Capacitors (C) are
used to control the fluctuation of voltages to maintain a constant VDC for power
supply. .................................................................................................................. 30
Fig. 3.12. Amperemeter circuit. ....................................................................................... 30
x
Fig. 3.13. Multi-junction sensor system. (a) 3-D printed multiplexing circuit module and
sensing platform. Close up image of the opened sensing platform (b) before a
multi-junction sensor chip is added and (c) after sensor is secured. .................... 31
Fig. 3.14. Multi-junction sensor functionalization set-up with (a) anti-E. coli antibodies
and (b) anti-S. aureus antibodies tested with pure E. coli and S. aureus samples.
............................................................................................................................... 32
Fig. 3.15. Schematic illustration of batch-type multiplexing tests using (a) 10 μL
microbial cocktail samples of E. coli and S. aureus and (b) 100 μL samples. .... 33
Fig. 3.16. Mathematical circuit developed to determine junction resistance. (a) Electric
current pathway and (b) circuit equivalent at a single junction. (c) Equations used
to calculate junction resistance. ........................................................................... 34
Fig. 4.1. SEM images of a microwire electrode. (a) Electrode surface before PEI coating,
(b) after 50%, and (c) 1% PEI coating. ................................................................ 36
Fig. 4.2. SEM images of SWCNTs adhered to (a) 50%, (b) 1%, and (c) 0.1% PEI coated
microwires. ........................................................................................................... 37
Fig. 4.3. SEM images corresponding to the number of SWCNT dip-coats. (a, b) One 38
SWCNT dip-coat. (c, d) Two dip-coats. (e, f) Three dip-coats. Figures a, c, and e
in the first column are captured at the 5 μm scale; whereas figures b, d, and f in
the second column are taken at 500 nm scale. ..................................................... 38
Fig. 4.4. I-V curve from 0 to 1 VDC corresponding to individual sensor modification
layers. ................................................................................................................... 41
Fig. 4.5. Effect of SWCNTs on signal response during step-wise surface modification
and E. coli K-12 detection at 1 VDC. .................................................................... 42
Fig. 4.6. Averaged change in current in response to captured E. coli on a junction sensor
with and without SWCNTs. Inset: SEM image of E. coli captured on the bio-nano
junction sensor surface through immune complex reactions with biotinylated anti-
E. coli antibodies. Significant differences between signal measurements are
indicated by the different superscripts at a 95% confidence level (probability <
0.05). .................................................................................................................... 43
Fig. 4.7. Relationship between changes in current and concentrations of E. coli K-12
(101-105 CFU/ mL) bound to the E. coli functionalized junction sensor.
xi
Significant differences between signal measurements and bacteria concentration
are indicated by the different superscripts at a 95% confidence level (probability <
0.05). .................................................................................................................... 44
Fig. 4.8. E. coli junction sensor signal response to S. aureus samples (103-105 CFU/ mL)
in comparison to E. coli detection signal responses. Significant differences
between signal measurements are indicated by the different superscripts at a 95%
confidence level (probability < 0.05). .................................................................. 45
Fig. 4.9. (a) Electrical current calibration curve for an anti-E. coli antibody functionalized
multi-junction sensor tested with a negative control (100 CFU/mL) and E. coli in
the range of 101-105 CFU/mL. (b) SEM image of E. coli cells bound to sensor
surface. ................................................................................................................. 46
Fig. 4.10. (a) Electrical current calibration curve for an anti-S. aureus antibody
functionalized multi-junction sensor tested with a negative control (100 CFU/mL)
and S. aureus in the range of 101-105 CFU/mL. (b) SEM image of S. aureus cells
bound to sensor surface. ....................................................................................... 47
Fig. 4.11. Relationship between changes in current and concentrations of E. coli and S.
aureus from 101-105 CFU/mL. Average signals (current drop) with different
superscripts are significantly different at 95% confidence level (probability <
0.05). *E. coli and S. aureus ∆I values were analyzed separately. ...................... 48
Fig. 4.12. Current measurements for simultaneous detection of E. coli and S. aureus in 10
and 100 µL samples in comparison to calibration measurements. ...................... 49
xii
LIST OF ABBREVIATIONS
∆I change in current (A)
1-D one-dimensional
2-D two-dimensional
3-D three-dimensional
A area (in2)
AC alternating current
ANOVA analysis of variance
BSA bovine serum albumin
C capacitance
CDC U.S. Center for Disease Control and Prevention
CFU colony forming units
Ch graphene chiral vector
CNT carbon nanotube
CPU central processing unit
DAC digital to analog convertor
DAQ data acquisition unit
DI deionized water
DC direct current
DMF N-N-dimethylformamide
DNA Deoxyribonucleic acid
ELISA enzyme-linked immunosorbent assay
GRD ground
F force
FESEM field emission scanning electron microscope
FET field effect transistor
GND ground
I electric current (A)
I-V current-voltage
k electron wave vector
LOD limit of detection (CFU/mL)
MNP magnetic nanoparticle
MPN most probable number
n electron transport
NC normally closed
NO normally open
NPN negative-positive-negative transistor
xiii
NR nanorod
NW nanowire
Ø diameter (m)
OP operation amplifier
p hole transport
PCA plate count agar
PC print circuit board
PCR polymerase chain reaction
PDMS polydimethylsiloxane
PEI polyethylenimine
QD quantum dot
Q-PCR quantitative real-time-PCR
R resistance (Ω)
RAM random access memory
RT-PCR reverse transcriptase-PCR
SDPT single pole double throw relay
SEM scanning electron microscopy
SPR surface plasmon resonance
SSR solid state relay
STEC shiga toxin–producing Escherichia coli
SWCNT single-walled carbon nanotube
TSB tryptic soy broth
VDC applied direct current voltage (V)
Vdd voltage drain (V)
Vlg voltage liquid gate potential (V)
vw withdrawal velocity (mm/s)
VR variable resistance (Ω)
WHO World Health Organization
θ chiral angle
σ allowable stress (PSI)
Ω ohm
1
1. INTRODUCTION
Food safety monitoring is a key aspect within the food industry. The security and
safety of our food depends on the ability to identify, detect, and trace food pathogens.
Many efforts have been made by food regulatory agencies and manufacturers to minimize
the risks for foodborne illnesses, such as implementing good agricultural practices, good
manufacturing practices, and hazard analysis and critical control point programs
(Velusamy et al., 2010). Yet, reducing the occurrence of microbial contamination
remains a challenge. As production of minimally processed foods and globalization of the
food supply expand, the occurrence of foodborne illness outbreaks continue to increase,
threatening our health and economy (Scallan et al, 2011). Therefore, detection methods
play a significant role in aiding to prevent and identify foodborne pathogens.
Sensitive and rapid detection of foodborne pathogens from a food sample is
difficult to achieve, as most detection methods are either time consuming, labor intensive,
laboratory-based, or expensive. An identification method that is simple and affordable,
with adaptability to detect multiple pathogens, specificity to distinguish between different
bacteria, and sensitivity to detect bacteria in food samples without the need for pre-
enrichment is the key challenge in the field of pathogen detection.
Despite greater biological understanding and innovative technological
developments, current detection methods have significant drawbacks. Traditional plate
counting, though accurate and affordable, is time-consuming and requires sample pre-
enrichment. DNA amplification methods offer a faster detection time with good
sensitivity, but are laborious and expensive; and magnetic-based approaches are
applicable to complex food samples, but require lengthy sample preparation, costly
reagents, and limited sensitivity (Kim et al., 2013). Furthermore, most detection methods
require bench-top instruments in stationary laboratories that can only be operated by
skilled personnel.
Of the various techniques, biosensors, originating from the integration of
molecular biology and information technology, show high promise due to its potential for
portable, rapid, and sensitive detection (Mello and Kubota, 2002). Particularly, nano-
2
based immunosensors are gaining interest for food applications, as nanomaterials can be
used as catalytic tools, immobilization platforms, or as optical labels, exhibiting
improved sensitivity, stability, and response time. Among the nanomaterials, single
walled carbon nanotubes (SWCNTs) have emerged as building blocks for nanosensor
platforms (Grunner, 2006) due to their extraordinary structural, mechanical, and electrical
properties (Zhou et al., 2002). Enhanced sensing performance from the integration of
SWCNTs in biosensors is attributable to its bio and size compatibility (Allen et al.,
2007), structural flexibility (Katz and Willner, 2004), low capacitance, and axial
electrical conductivity (Kang et al., 2006). SWCNTs have been observed to amplify the
electrochemical reactivity of biomolecules (Wang, 2005; Vashist et al., 2011), as it is
sensitive towards minute variations in its surrounding environment. SWCNTs have also
been integrated into electrochemical immunosensors based on field effect transistor
(FET) designs (Besteman et al., 2003; Bousssaad et al., 2003; Artyukhin et al., 2006) and
for electrode surface modification as a means to improve electron transfer rates and
working surface area (Okuno et al., 2007; Zhao et al., 2011). Studies have also used
nanotubes to construct molecular junctions based on its ability to control the energy gap
of electrons (Forzani et al., 2004; Aguilar et al., 2005; Maruccio et al., 2007). Despite
potential applications, to our knowledge, bio-nano based junctions for detection of
foodborne pathogens are not represented in literature.
Hereby, the goal of this research was to explore the effects of a SWCNT
nanomaterial sensing platform in a SWCNT-based, label-free microwire junction sensor
for detection of foodborne pathogens. The developed biosensor operates by optimizing a
bio-nano modified biorecognition surface to convert molecular binding events at the
junction between target antigens and antibodies into measurable electrical signals. The
following parameters were studied: SWCNT surface morphology, electrical current
characterization of bio-nano functionalization layers, and single and multi-analyte
detection capabilities.
The overall objective of this thesis was to incorporate SWCNTs into a micro-scale
biosensing device for enhanced bacterial sensing performance with potential as a rapid
sensing unit for potential portable multiplexed applications. Specific objectives leading to
this goal were:
3
Objective 1: Develop a protocol for SWCNT functionalization on a microwire electrode
surface.
Objective 2: Design a SWCNT-based single junction sensor for single analyte detection
and explore the effects of SWCNTs on its sensing performance.
Objective 3: Apply the SWCNT junction format into a multi-junction sensing array for
simultaneous bacterial detection in liquid samples.
4
2. LITERATURE REVIEW2
2.1. Introduction
In this section, current detection techniques as well as novel nano-based strategies
that have been developed and aimed to provide food processing operators and food safety
authorities with the ability to rapidly detect foodborne pathogens will be reviewed. In
addition, the structural and electrical properties of SWCNTs and its detection
mechanisms will be discussed.
2.2. Literature Review
2.2.1. Foodborne pathogens and illnesses
Though the safety of our food supply has improved throughout the years, the
prevalence of foodborne illness outbreaks are difficult to overcome. The World Health
Organization (WHO) defines foodborne illnesses as diseases, usually infectious or toxic,
caused by agents that enter the body through the ingestion of foods (Velusamy et al.,
2010). According to the U.S. Center for Disease, Control and Prevention (CDC), every
year, an estimated 48 million Americans are sickened from consumption of contaminated
foods, of those, 128,000 are hospitalized, and 3,000 die of foodborne diseases (2011).
Additionally, $51 billion (90% Cl, $31.2- $76.1 billion) is lost annually due to medical
costs, productivity losses, and illness-related mortality caused by foodborne pathogens
(Scharff, 2012).
Bacteria account for 91% of total foodborne illness outbreaks (Yang and Bashir,
2008; CDC, 2011). Salmonella (non-typhodial), Clostridium perfringens, Campylobacter
spp., Staphylococcus aureus, Escherichia coli (STEC) O157:H7, and Listeria
monocytogenes are the main pathogenic bacterial contributors to domestically acquired
foodborne illnesses (CDC, 2011; Scallon et al., 2011). In 2006, E. coli O157:H7,
confirmed to have originated from contaminated fresh baby spinach leaves, spread
nationwide resulting in 205 confirmed illnesses, including 31 cases of hemolytic uremic
syndrome, 104 hospitalizations, and four deaths (Food and Drug Administration [FDA],
2007). In 2008, a multistate outbreak of Salmonella Typhimurium linked to peanut butter
sickened 714 Americans and resulted in recalls of 3,913 different products made by 361
5
companies (CDC, 2010). And, recently in 2011, a Listeria monocytogenes outbreak
traced back to cantaloupes led to 147 illnesses, 33 deaths, and one miscarriage (CDC,
2012). Hence, foodborne pathogens pose a serious risk to food safety and a threat to our
global food supply chain (Radke and Alocilja, 2005).
2.2.2. Traditional detection methods and drawbacks
Food safety screening for the detection and identification of foodborne pathogens
continue to rely on culture-based, immunology-based, and nucleic acid-based methods
(Velusamy et al., 2010).
Conventional culture-based detection, a standardized microbiological technique,
isolates and enumerates viable bacterial cells on selective media for biochemical
confirmation. Though sensitive and specific, this method is labor intensive and time
consuming (Ivnitski et al., 1999; Yang et al., 2008; Yang and Bashir, 2008). It involves a
complex series of tests, beginning with pre-enrichment, followed by selective
enrichment, biochemical screening, and serological confirmation of food samples, which
can take up to 10 days for conformation (Velusamy et al., 2010). Therefore, it is not
suitable for food industrial applications, where a timely response to possible health risks
is crucial. In addition, enumeration of bacterial cells on selective growth agar can yield
false negative results if cells are viable-but non-culturable (Yang et al., 2008).
Immunology-based methods using target specific antibodies for detection of
foodborne pathogens provides improved specificity and sensitivity of target bacterial
cells (Song and Vo-Dinh, 2004). A wide range of immunoassays exist from homogenous
immunoassays, in which antigen-antibody complexes are directly visible or measurable,
i.e. agglutination reactions, to heterogeneous immunoassays, where unbound antibodies
are separated and the remaining bound conjugated antibodies are detected (Boer and
Beumer 1999). Among the many immunological methods, enzyme-linked
immunosorbent assays (ELISA), including direct, sandwich, and competitive ELISAs,
have become established screening tools due to their simplicity and affordability (Yang et
al., 2008). The main limitations of ELISA methods are their relatively high detection
limits, normally 104- 105 CFU/mL (Mandal et al., 2011) (Table 2.1), and demand for pre-
analytical enrichment procedures. As an effort to improve sensing abilities, multiplex
6
immunoassays were developed (Karoonuthaisiri et al., 2009; Zhao et al., 2009). Park et
al., (2010) developed an ELISA method on an immunochromatographic strip using a
horse radish peroxidase (HRP) solution for optical detection of S. Typhimurium, S.
aureus, and E. coli O157:H7. The limit of detection (LOD) was in range of 103-105
CFU/mL with an assay time of 20 min. E. coli and S. Enteritidis were detected using a
sandwich immunomagnetic separation method with quantum dot fluorescence markers
for fluorescence measurement to achieve a 102 CFU/mL detection in 2 h (Dudak and
Boyaci, 2009). However, these assays rely on the use of fluorophore labels and
chemiluminescence detection, requiring many reagents, large amounts of antibodies, and
separation and optical equipment.
Nucleic acid-based detection methods, primarily polymerase chain reaction (PCR)
techniques, are widely employed (Lampel et al., 2000; Burtscher and Wuertz 2003;
Malorny et al., 2004). Conventional PCR with detection limits ranging from 102-103
CFU/mL (Matthews and Montville, 2008), relies on the amplification of target genes in a
thermocycler, separation of PCR products by gel electrophoresis, followed by
visualization and analysis of the resulting electrophoretic patterns (Lopez-Campos et al.,
2012). Many PCR tests, including conventional, real-time, multiplex, and reverse
transcriptase-PCR (RT-PCR) methods, have been validated and commercialized to make
PCR a standard tool used by food microbiology laboratories (Jasson et al. 2010).
Quantitative real-time PCR (Q-PCR) has greatly increased the speed and sensitivity of
PCR-based detection methods, as results can be obtained in an hour or less. Kawasaki et
al., (2009) evaluated a multiplex PCR system for simultaneous detection of Salmonella
spp., L. monocytogenes, and E. coli O157:H7 in meat, poultry, and dairy products. A 5
CFU/25 g limit of detection was achieved with 20 h enrichment time. Suo et al., (2010)
investigated the use of a TaqMan based multiplex Q-PCR assay for fluorescent detection
(LOD: 18 CFU/10 g) of pathogens in ground beef after 20 h enrichment. Commercially
available Bax System Real-time PCR assay and MicroSEQ Detection Kit were also tested
with various foodborne pathogens and achieved detection limits of 104 CFU/mL and 1-3
CFU/25 g, with enrichment times ranging from 20-31 h (Roda et al., 2012). Despite low
detection limits and commercial applications of multiplex Q-PCR assays, these methods
7
require skilled personal in laboratory settings, lengthy enrichment preparation steps,
multiple primers, fluorophore labeling, and expensive equipment.
Table 2.1. Main characteristics of some culture-based and rapid detection methods
Quantitative test method Sensitivitya Specificity Assay time (h)
Culture-based (MPN) <10-100 MPN of bacteria per gram Good 24-48
Culture-based (Viable counts) >10-100 Good 24-72
Bioluminescence 104 No <1-3
Immunology-based (ELISA) 104-105 Moderate/good <1-3
Nucleic acid-based (Q-PCR) <102 Excellent <1-3
a CFU/g or mL, unless otherwise stated
(Modified from Mandal et al., 2011)
Due to the limitations of current detection methods for foodborne pathogens, there
is a need for rapid and accurate detection methods that provide sensitive and specific
results with minimal cost and labor. The expansion of our food industry and the growing
concern for food safety has motivated researchers in the biosensor field to develop new
technologies that show promise for future commercial industrial applications.
2.2.3. Biosensor technology for foodborne pathogen detection
Biosensor technology is emerging as an alternative to microbiological tests that
are centralized in stationary laboratories, requiring complex instruments and well-trained
technicians. Intensive research continues in an effort to develop portable and sensitive
biosensors for real-time detection of foodborne pathogens. A biosensor is an analytical
device composed of a biological recognition element that recognizes the target analyte
and a transducer, which converts the corresponding biological responses into measurable
electrical signals (Fig. 2.1).
8
Fig. 2.1. Schematic illustration of the main components in a biosensor.
Biosensors are classified according to the biorecognition element and transducer
used, which is selected based on the properties of each target analyte and the type of
physical magnitude to be measured. Despite the large variety of bioreceptors from tissues
to biomimics, there is an increased interest in the development of immunosensors due to
antigen/antibody complex stability, adjustable affinity, and variable specificity (Hall et
al., 2002; Mello and Kubota 2002; Heo and Hua, 2009). In addition, antibodies can be
conjugated with avidin-biotin complexes, enzymes, fluorescent compounds, and
electrochemically active substances to enhance signal amplification and sensitivity (Hall
et al., 2002).
Transducers are divided into categories based on the method of signal
transduction, such as optical, electrochemical, and mass-based (Fig. 2.2). Optical
biosensors measure the absorption, reflection, refraction, or dispersion of luminescent,
fluorescent, or colorimetric signals produced by the interaction of target analytes and
bioreceptors. For example, surface plasmon resonance (SPR) sensors measure changes in
the refractive index of an aqueous layer when pathogens bind to the receptors
immobilized on the transducer surface (Bhunia et al., 2004; Subramanian et al., 2006;
Taylor et al., 2006). On the other hand, electrochemical biosensors, sub-classified into
amperometric, potentiometric, impedimetric, and conductometric categories, measure
parameters such as current, potential, impedance, and conductance, and are promising for
low cost miniaturized manufacture (Ruan et al., 2002; Chemburu et al., 2005; Pal et al.,
Target analyte
9
2008). Mass-based sensors measure small changes in mass through the generation and
transmission of acoustic waves with an oscillating crystal (Pathirana et al., 2000).
Fig. 2.2. Classification of biosensors based on biorecognition and transducing elements.
Optical and electrochemical transduction systems are the most common groups of
biosensors to be investigated for foodborne pathogen detection. Taylor et al., (2006)
detected E. coli O157:H7, S. Typhimurium, L. monocytogenes, and C. jejuni by means of
a sandwich immunoassay label-free SPR biosensor with LOD values ranging from 103-
105 CFU/mL. Fluorescence resonance energy and evanescent excitation were also applied
to detect E. coli O157:H7 and S. aureus in similar ranges (Ligler et al., 2007). Despite its
multiplexing abilities, optical methods are generally less suitable for routine detection,
owing to their higher cost and complexity (Roda et al., 2012). Even though current
electrochemical techniques do not meet all industrial sensing demands, they have
attracted more attention due to their rapidity, low cost, and suitability to be integrated in
miniaturized automated assays (Yang et al., 2008). However, electrochemical biosensor
research has been well studied for single analyte detection (Ruan et al., 2002; Radke and
10
Alocilja, 2005) with limited advances in multiplexed pathogen detection (Palchetti and
Mascini, 2008).
Though there is an immense amount of biosensor research for foodborne
pathogens, the number of commercial biosensors available for purchase is limited (Table
2.2). Factors that prevent pilot-scale biosensors from being manufactured for commercial
use include accuracy, reproducibility, and the ability to detect a wide range of bacteria.
In addition, current commercial biosensors do not meet industrial food safety
requirements for cost and sensitivity, as the desirable detection limit in food samples is
less than 1 cell per 25 g of food (Velusamy et al., 2010).
Table 2.2. Commercial biosensor technologies for foodborne pathogen detection.
Company Biosensor Transducer Measured
property
Limit of detection
(CFU/mL)
Analysis
time
BIA-core BIACORE Optical SPR 104 < 2 h
Reichert Reichert SR
7000
Optical SPR 103 < 1 h
Research
International
Analyte 2000 Optical Total internal
reflection
103 8- 24 h
Biotrace Unilite Optical Reflection 105 15 min
Bactomatic
Inc.
Bactometer Electrochemical Impedimetric 105 3-8 h
Malthus
Instruments
Malthus 2000 Electrochemical Potentiometric,
conductometric
105 8- 24 h
Biosensor SpA Midas Pro Electrochemical Amperiometric 106 20 min
Universal
Sensors
PZ106
Immunosensor
Piezoelectric Resonance
frequency
106 40 min
(Modified from Mello and Kubota 2002; Subramanian et al., 2006; Valadez et al., 2009;
Ohk et al., 2010; Arora et al., 2011; Wang et al., 2011).
Despite recent technological advances, there still exist many challenges and
opportunities to improve current biosensors to achieve simple, rapid, versatile, and
inexpensive detection of food contaminants.
2.2.4. Single walled carbon nanotubes for biosensor technology
Nanotechnology is emerging as a promising way to obtain fast, reliable, and
precise information on the safety of our food products. The integration of one-
dimensional (1-D) nanomaterials into biosensing platforms as labels or transducer
modifiers offers substantial advantages for the detection of bacterial analytes (Allen et al.,
11
2007). Engineered nanomaterials, such as magnetic nanoparticles (MNPs) (Varshney and
Li, 2007; Ravindranath et al., 2009), carbon nanotubes (CNTs) (Chunglok et al., 2011;
Zhao et al., 2011), nanorods (NRs) (Wang and Irudayaraj, 2008), quantum dots (QDs)
(Zhao et al., 2009; Vinayaka and Thakur, 2010), and nanowires (NWs) (Wang et al.,
2008) have been used in biosensor research for improved sensitivity, fast response and
recovery, and potential for integration in large scale array systems.
Amongst the many nanomaterials, CNTs have been intensively studied and have
held a fundamental role in the field of nanotechnology due to their electrical, structural,
and mechanical properties. In particular, single-walled carbon nanotubes (SWCNTs),
discovered by Iijima in the early 1990s (Iijima, 2002), is becoming a key focus of
biosensor research. SWCNTs are seamless nanometer-diameter cylinders with tube
lengths in the micrometer order, each comprised of a single layer of graphene (Fig. 2.3
(a)). The diameter and chirality (θ) of a SWCNT are defined by its graphene chiral
vector, Ch = na1 + ma2 ≡ (n, m), where a1 and a2 represent the lattice vectors and n and m
are integers that represent the number of unit vectors along two directions of the graphene
lattice (McEuen et al., 2002; Odom et al., 2002) (Fig. 2.3 (b)). Depending on the
curvature and angles in which a graphene sheet is rolled, SWCNTs can be classified into
zigzag, armchair, and chiral forms (Karousis and Tagmatarachis, 2010) (Fig. 2.3 (c)).
12
Fig. 2.3. (a) Schematic of a single graphene sheet rolled up to form a SWCNT. (b)
Graphene sheet illustrating lattice vectors, a1 and a2, and the chiral vector, Ch = na1 + ma2.
The achiral, limiting cases of zigzag (n, 0) and armchair (n, n) are indicated with thick,
dashed lines, and the chiral (θ) angle is measured from the zigzag direction. The light,
dashed parallel lines define the unrolled, infinite SWCNT (Modified from Odom et al.,
2002). (c) SWCNT armchair, zigzag, and chiral forms (Modified from Iijima, 2002).
The unique electrical properties of SWCNTs stem from the electronic structure of
the two-dimensional (2-D) graphene sheet, composed of a single atomic layer of graphite
with a honeycomb lattice of sp2 bonded carbon atoms (McEuen et al., 2002) (Fig. 2.4 (a)).
Graphite is a semimetal or zero-gap semiconductor whose valence and conduction bands
touch and are degenerate at the six wave vector k points, indicated at the center of the
cones at the corners of the first Brillouin zone (Odom et al., 2002) (Fig. 2.4 (b)). Thus, in
a SWCNT, the momentum of the electrons moving around the circumference of the tube
is reduced to slices through the band structure. If one of these slices pass through a k
point, the SWCNT will be metallic, in which (n, m) indices satisfy the condition that (n
−m)/3 is an integer (Fig. 2.4 (c)); otherwise, it will be semi-conducting (McEuen et al.,
2002) (Fig. 2.4 (d)). Armchair SWCNTs are metallic, while zigzag and chiral forms can
be either metallic or semiconducting (Karousis and Tagmatarchis, 2010).
13
Fig. 2.4. (a) Structure of graphene’s honeycomb lattice of carbon atoms. (b) Illustration
of graphene bands and its conducting states as a function of the electron wave vector k.
The black hexagon defines the first Brilluoin zone of graphene. There are no conducting
states except along special directions where cones of states exist. The centers of the cones
are defined as the graphene k points. Depending on the way the graphene vector is rolled
up, SWCNTs can either be classified as (c) a metal, slice passes through the center of a
cone, k, or (d) a semiconductor, with a gap between the filled hole states and the empty
electron states (McEuen et al., 2002).
SWCNTs have become basic building blocks for molecule-based electrical
circuits. Aside from its unique electronic properties, SWCNTs offer high strength,
flexibility, stability, and biomolecule compatibility. SWCNTs are size compatible to
single biomolecules, as they have the smallest diameter of 1 nm (Chen et al., 2003;
Cherukuri et al., 2004; Barone et al., 2005). Since all carbon atoms are in direct contact
with its environment, SWCNTs also provide maximum interaction with adjacent
biomolecules (Allen et al., 2007; Maroto et al., 2007; Heller et al., 2008). Also, the low
charge carrier density of SWCNTs is comparable to the surface charge density of
proteins, which makes SWCNTs well suited for electronic detection of target
biomolecules (Heller et al., 2006). Therefore, in comparison to 2-D thin films where
14
binding to the surface leads to depletion or accumulation of charge carriers on the surface
of a planar device, the charge accumulation or depletion in the 1-D nanostructure takes
place in the “bulk” of the structure, creating large electrical property changes that can
potentially enable the detection of a single molecule (Wanekaya et al., 2006).
2.2.5. SWCNT mechanisms for detection of bio-analytes
When protein is adsorbed onto SWCNTs, a measurable change in electrical
properties occurs that can be exploited for detection of bacterial cells (Allen et al., 2007).
Physical mechanisms underlying electronic detection of biomolecules include
electrostatic gating, changes in gate coupling, carrier mobility changes, and Schottky
barrier effects (Heller et al., 2008).
Choi and Hong (2012) explored the sensing mechanism of a network of SWCNTs
and observed changes in electrical properties when sensitive electromechanical coupling
of SWCNTs occurred. The low capacitance of SWCNTs shortened the response time of
the electrical resistance changes induced by the mechanical deformation. Two possible
mechanisms underlying the electrical resistance change of the SWCNT film in response
to mechanical loading is (1) change in morphology of the network caused by a
disconnection of electrical pathways among SWCNTs and (2) band gap change in
SWCNTs due to lattice strain. Star et al. (2003) studied the charge transfer reaction
between streptavidin and SWCNTs. The streptavidin-biotin binding induced geometric
deformations, leading to scattering sites on the nanotube and a reduction in conductance.
Chen et al. (2004) also observed a conductance change as proteins adsorbed to the
surface of semi-conducting SWCNT devices. Measurable changes in the electrical
conductance were believed to be induced by gating effects or charge transfer to
nanotubes. It was also noted that the adsorbed proteins modulated the band alignment and
Schottky barrier of the device’s metal contact-nanotube region.
Heller et al. (2008) studied the gate effects of protein adsorption on liquid gate
potential dependence of device conductance. I (source drain current) –Vlg (liquid gate
potential) curves were used as tools to identify SWCNT electronic modulation. When
charged protein molecules adsorbed on a SWCNT by electrostatic gating, a screening
charge (doping) effect shifted the I-Vlg curve along the voltage axis due to partial charge
15
transfer (Fig. 2.5 (a)). In the case of Schottky barrier effects (Fig. 2.5 (b)), adsorbed
biomolecules on a SWCNT modulated the band alignment. Because the Schottky barrier
height changes in opposite directions for hole (p) and electron (n) transport, an
asymmetric conductance change for p- and n-branches of I-Vlg was observed. Therefore,
from extensive protein-adsorption experiments on SWCNT transistor devices,
electrostatic gating and Schottky barrier effects dominate sensing mechanisms (Heller et
al., 2008).
Fig. 2.5. Calculated I-Vlg curves before (black) and after (red) protein adsorption for (a)
electrostatic gating effect corresponding to a shift of the semiconducting bands
downward and (b) Schottky barrier effect that corresponds to a change of the difference
between metal and SWCNT work functions. Insets illustrate the corresponding changes
in the band diagrams for hole and electron doping respectively (Heller et al., 2008).
2.2.6. SWCNT-based biosensors for foodborne pathogen detection
Semiconducting SWCNT-based sensors have been fabricated based on field effect
transistor (FET) designs, in which, either individual or networks of SWCNTs serve as
electron channels between source and drain electrodes (Allen et al., 2007). SWCNT-
FETs have been extensively studied for detection of proteins, carbohydrates, DNA, and
alcohols (Besteman et al., 2003; Boussaad et al., 2003; Wang, 2005; Artyukhin et al.,
2006; Allen et al., 2007); however, only recently have SWCNT-based sensors been
exploited to detect food pathogens. Functionalization of antibodies, aptamers, proteins, or
enzymes onto the surface of SWCNTs are implemented as sensing layers together with
the capability of SWCNTs to transduce the charge transfer by using FETs. Huang et al.
(2004) adsorbed anti-Salmonella antibodies on SWCNTs and exposed it to 108 CFU/mL
of Salmonella Typhimurium for 1 h. The bacteria linked to the CNTs were observed with
16
scanning electron microscopy (SEM). SWCNTs were functionalized with bovine serum
albumin (BSA) and anti-E. coli antibodies to demonstrate SWCNTs potential to detect
pathogenic Escherichia coli O157:H7 (Lin et al., 2006). An aptamer functionalized
SWCNT-FET array was evaluated for the detection of E. coli cells within the range of
105-107 CFU/mL (So et al., 2008). A network of SWCNTs was also used as conductor
channels by Villamizar et al., (2008) to selectively detect Salmonella Infantis at 102
CFU/mL within 1 h. Chunglok et al., (2011) integrated SWCNTs as an ELISA labeling
platform to detect 103 CFU/mL of Salmonella Typhimurium within a 2 h incubation time.
SWCNTs exhibit potential advantages to achieve low bacterial detection limits in a
reduced amount of time. However, the intricate sensor designs and elaborate fabrication
processes are preventing SWCNT biosensors from evolving into practical sensing tools
for industrial applications. Therefore, a SWCNT-based biosensor with simple fabrication
and minimal sensing procedures will offer an important step toward development of
selective biorecognition devices with enhanced sensitivity for foodborne pathogen
detection.
2.3. Conclusion and thesis overview
In spite of significant analytical detection progress, a biosensor with simple
operation, real-time sensitive measurement, affordable costs, and in-field applications has
not been successfully developed. As an effort to take advantage of SWCNTs’ unique
electrical and structural properties, SWCNTs were used as a coating to amplify
transducing effects of microwire sensing electrodes.
This research was intended to investigate the use of SWCNTs for single and multi-
analyte detection. The first part of this thesis was aimed to study the effects of SWCNTs
on electrical current measurements of a single junction biosensor. SWCNT surface
morphology was explored through current-voltage measurements. Sensing performance
of the single junction sensor was evaluated for detection of Escherichia coli K-12, as the
model microorganism. Based on the research findings, the second part of this thesis
focused on the development of a multi-junction sensing array for multi-analyte detection
using E. coli and Staphylococcus aureus. Results showed that the network of SWCNTs
17
significantly contributed to sensing measurements at the functionalized junctions,
suggesting potential application of junction arrays for bacterial sensing.
18
3. MATERIALS & METHODS
The nanomaterial SWCNT has been identified as a potential biosensing platform
material to improve biorecognition signals during bacterial detection. In this research, an
electrical biosensor composed of SWCNT-coated microwire electrodes assembled to
form crossbar junctions with immobilized antibodies was fabricated. A bio-nano coating
procedure was developed for electrode surface modification with polyethylenimine,
SWCNTs, streptavidin, and biotin-conjugated polyclonal antibodies. A bench-top single
junction sensing system was designed. In order to determine the effects of SWCNT
coating on the sensor, step-wise characterization of the functionalization layers were
observed through current-voltage measurements. The sensor’s single analyte detection
performance was evaluated through sensitivity and specificity tests. Next, multi-analyte
detection was investigated. The bench-top system used for the single junction sensor was
scaled down into a portable sensing unit and a disposable multi-junction array sensor chip
was fabricated and evaluated for multiplexing capabilities.
3.1. Sensor design & fabrication materials
The following materials and instruments were used during sensor design and
fabrication:
7% gold-tungsten plated wire (Ø: 50 μm, Lot # Q11254) was manufactured from
ESPI Metals (Ashland, OR). Ultem® polyetherimide, mica sheets, stainless steel flat
head slotted machine screws (6-32 x ¾), and nuts were supplied by McMaster-Carr
(Santa Fe Springs, CA). Polydimethylsiloxane (PDMS; Sylgard 184 silicone elastomer
curing agent and base, Lot # 0007590060) was ordered through Dow Corning (Midland,
MI). Copper clad printed circuit boards (160.78 x 114.30 x 1.52x mm) and etchant
protectant sheets were purchased from Radio Shack (Honolulu, HI). SWCNTs (>95%
purity, Ø: 15 ± 5 nm, 1-5 μm lengths, Lot # 20121217) were purchased from NanoLab,
Inc. (Waltham, MA).
Etchant solution was purchased from Radio Shack (Honolulu, HI). Alcohol (95%,
Cat # BDH1158), BD BactoTM peptone (Ref # 211677), BD BBLTM trypticase soy broth
(Ref # 211768), and BD DifcoTM plate count agar (Ref # 247910) were procured from
19
VWR (West Chester, PA). N,N-dimethylformamide (DMF, Product # 227056),
polyethylenimine (PEI, branched, average Mw ~25,000, Product # 408727), and
streptavidin from Streptomyces avidinii (affinity purified, lyophilized from 10 mM
potassium phosphate, ≥13 U/mg protein, Product # 85878) were supplied from Sigma
Aldrich (St. Louis, MO). Biotinylated polyclonal Escherichia coli (Product # PA1-
73031) and Staphylococcus aureus (Product # PA1-73174) antibodies and OXOID
MacConkey agar (Lot # 1434592) were purchased from Thermo Fisher Scientific
(Waltham, MA).
Microwire sanitization and SWCNT dispersion were performed using a digital
sonifier (450, Branson, Danbury, CT). An automated XYZ stage (Franklin Mechanical &
Control Inc., Gilroy, CA,) controlled by the COSMOS program (Velmex, Inc.,
Bloomfield, NY), furnace (Thermolyne, Thermo Scientific, Waltham, MA), and solder
kit (Radio Shack, Honolulu, HI) were used during SWCNT coating and wire assembly. A
desktop 3D printer was purchased from Lulz Bot (TAZ 4, Loveland, CO) for sensor
device fabrication. For the sensing measurements, a function generator (33220A, Agilent,
Santa Clara, CA) and picoammeter (6485, Keithley, Cleveland, Ohio) were integrated
into the sensing system.
3.2. SWCNT coating technique
CNT networks can be formed by several approaches including spin coating and
spray coating (Jang et al., 2008). Spin coating is a simple method for forming SWCNT
networks, but is limited to small planar areas and large amounts of SWCNT colloid
solution is lost. Spray coating, though simple and applicable to large surface areas, is not
useful for obtaining uniform networks. To overcome these difficulties, a dip-coating
method was used to coat microwire electrodes.
To create a SWCNT suspension, SWCNTs were dispersed in DMF at
concentrations of 0.1 g/L by sonicating the solution for 6 h (Rouse et al., 2004). After the
initial 6 h of sonication, the dispersion was further sonicated for 2 h before use each day.
Prior to SWCNT coating, microwire electrodes were cut to a length of 31 mm and
sonicated in DI water, followed by 70% alcohol for 5 min each. Sanitized wires were
dried in a furnace at 175°C for 10 min and mounted onto the automated XYZ stage for
20
step-wise surface modification (Fig. 3.1). A computer was used to initiate the XYZ motor
control program. Wires were immersed into glass vials containing 9 mL of 1% PEI
solution for 5 min and withdrawn at a constant withdrawal velocity (vw) of 6 mm/min.
PEI coated wires were baked at 175C for 1 h (Cairns, 2013). Subsequently, wires were
re-mounted onto the XYZ stage and immersed into the SWCNT-DMF suspension for 5
min. The PEI coated wires were withdrawn at the same constant velocity (vw= 6
mm/min) to generate a high capillary force, which creates a large influx of SWCNT
colloids onto the wire (Jang et al., 2008) (Fig. 3.2). The modified wire’s surface was
observed after each dip-coat using FESEM imaging to determine the appropriate number
of coats needed to achieve a uniform PEI-SWCNT network.
Fig. 3.1. SWCNT dip-coating experimental set-up. The stepping motor provided
controlled insertion and withdrawal velocity (Vw) for uniform nano coatings. A maximum
of four microwires were coated at a time.
21
Fig. 3.2. An illustration of the SWCNT dip-coating process. A constant vw generates a
capillary force between the SWCNT-DMF solution and the microwire. Suspended
SWCNTs flow into the meniscus due to capillary force and the SWCNTs adhere to the
PEI coated wire surface.
3.3. Antibody immobilization and microbial preparation
After SWCNT wire coating and junction assembly, antibodies were immobilized
onto each junction to enhance sensor specificity. Self-assembled monolayers (SAM) of
streptavidin and biotinylated antibodies were formed on the SWCNT coated microwires
(Fig. 3.3) (Lu and Jun, 2012). 5 μL of streptavidin was pipetted onto each junction. After
5 min, the droplet of streptavidin was removed. Subsequently, 5 μL of biotinylated
antibodies (anti-E. coli or anti-S. aureus) was added to the junction for another 5 min, to
form avidin-biotin complexes. Then, the droplet of biotinylated antibodies was removed
and the functionalized bio-nano junction sensor was dried and ready for cell detection.
Fig. 3.3. Antibody immobilization process (Schematic is not drawn to scale). A micro
volume of streptavidin is applied to the wire junction. Thereafter, biotin conjugated
antibodies are applied and bind to streptavidin.
22
Frozen stock cultures of Escherichia coli K-12 and Staphylococcus aureus were
obtained from the Food Microbiology collection, University of Hawaii. All experiments
were conducted in a certified Biosafety Level II laboratory. 100 μL of each isolate was
cultured separately in 10 mL of tryptic soy broth (TSB; pH 7.3) and incubated for 24 h at
37°C to make a stock culture of each organism.
Test samples were prepared from serial dilutions of the stock cultures using 0.1%
peptone water (pH 7.2). The initial concentrations of E. coli and S. aureus stock cultures
were obtained by plate counting methods on plate count agar (PCA). For microbial
cocktails, a mixture of E. coli K-12 and S. aureus was prepared by centrifuging 1 mL of
each stock culture at 2000 rpm for 15 min (Muhammad-Tahir and Alocilja, 2003). The
resulting pellets were suspended in 0.1% peptone water to make a 10 mL solution. The
microbial cocktail suspension was serially diluted in peptone water.
3.4. SWCNT-based biosensor: single junction
3.4.1. Device fabrication process
The single junction sensor device was fabricated using two Ultem squares (26 x 26
x 5 mm) and four stainless steel screws with nuts. One of the Ultem squares was used to
make the base holder. The center of the base was drilled 2.5 mm deep to create a sample
well. The well was filled with PDMS (10:1 ratio of base to curing agent) and cured at
60°C for 1 h. The second Ultem square with a 13 mm diameter hole in the center was
used as a cover to secure the wire junction. Two SWCNT-coated wires were assembled
in an orthogonal fashion and fixed on opposite sides of a mica frame, creating a 10 μm
gap, respectively, in between the two wires. The mica frame with attached wires was
placed on the base holder, aligning the junction directly above the sample well. Then, the
cover was placed over the mica frame to secure the junction from moving during sample
application. The base and cover were held together with the screws and nuts (Fig. 3.4).
23
Fig. 3.4. Schematic of the single junction sensor device.
3.4.2. Signal measurements
To obtain sensor readings, the fabricated biosensor was connected to a function
generator for voltage input, a picoammeter for electrical current readings, and a computer
for data logging (Fig. 3.5).
Fig. 3.5. (a) Schematic of the single junction sensing system. (b) Single junction circuit.
To begin detection, 10 μL of DI water was placed on the functionalized junction and
current measurements (Iantibody) were taken as the control reading. Then a 10 μL aliquot of
serial diluted E. coli K-12 was added to the junction sensor for 2 min to allow antibody-
antigen reactions to occur (Figure 3.6).
(a) (b)
Switch
Amperemeter
Single junction
24
Fig. 3.6. Antigen-antibody reactions on a bio-nano functionalized junction sensor. (Not
drawn to scale).
After the reaction, samples were washed with DI water to minimize non-specific
binding effects and the current of the sample (Iantibody-bacteria) was measured in 10 μL of DI
water. DI water was used as the medium instead of a conventional electrochemical buffer
to eliminate the need for a reference electrode (Kim et al., 2013). Though,
electrochemical buffers are a common media for electric measurements, it is typically
used in three-electrode systems because electrochemical reactions are well characterized
when a reference electrode is present. However, precise control of the distance between
the bio-nano modified microelectrodes and reference electrode would be a challenge. In
addition, ionic concentration of a buffer solution can vary depending on temperature and
humidity, thus requires calibration by a reference electrode.
A drop in electrical current was determined from the difference between the
current output of the control (Iantibody) and the current output of the inoculated samples
(Iantibody-bacteria). The difference in measured current (Δ I ) was the reduction in current due
to the sensor’s electrical property changes as a response to immune complex formations
(Eq. 3.1). The resulting signal (current drop) was proportional to the cell concentration.
ΔI= I antibody - I antibody-bacteria (3.1)
25
3.4.3. Sensitivity and specificity testing
Sensitivity testing was conducted to determine the biosensor’s limit of detection.
For the sensitivity testing, serial dilutions of E. coli K-12 (101-105 CFU/mL) were tested
with the anti-E. coli antibody functionalized junction sensor. Specificity testing was
performed to determine the ability of the sensor to detect target antigen in a non-specific
culture. For the specificity testing, serial dilutions of S. aureus (103-105 CFU/mL) were
tested with the junction sensor having E. coli specific antibodies.
3.4.4. FESEM imaging
A field emission scanning electron microscope (FESEM) (Pacific Biosciences
Research Center, University of Hawaii) was used to visualize the sensor’s surface, before
and after the capture of E. coli cells. Surface coating of the wires was required before
loading the samples into the FESEM due to the nonconductive properties of bacteria.
Each junction sample was attached to a conductive carbon tape adhered to an aluminum
stub and pretreated in a Hummer 6.2 sputter coater for 45 s to achieve a thin layer of
gold/palladium. Coated junction samples were then observed with a Hitachi S-4800
FESEM.
3.4.5. Data analysis
Three replications were performed for each experiment (n=3). The ΔI for these
three signals were averaged and plotted against the corresponding cell concentration.
Standard deviations of the current changes were expressed as error bars in the
corresponding graphs. Statistical analysis was conducted using a single factor analysis of
variance (ANOVA) in Statistical Analysis Software (SAS, Cary, NC). All biosensors
were assumed to have the same physical properties. For the purpose of this study, the
Duncan’s multiple range test was conducted to determine if the biosensor responses
corresponding to the bacterial concentrations were statistically different when the
probability was less than 0.05 (95% confidence level).
26
3.5. SWCNT-based biosensor: multi-junction
3.5.1. Device fabrication process
Disposable sensor chips were fabricated to create multi-junction arrays. Copper
clad printed circuit (PC) boards were cut into square chips (26 x 26 x 1.5 mm) (Fig. 3.7
(a)). A 13 mm diameter hole was drilled into the center of each chip. After drilling,
etching protectant film was applied to the chip to designate connector pad locations.
Copper chips were then submerged into etchant solution to remove all exposed copper
areas. After 1 h, the etched chips were removed from the solution and rinsed with water.
The etching protectant film was removed with acetone or an abrasive metal sponge. Each
chip was sanded to achieve smooth, even surfaces and angled edges. The hole was filled
with PDMS to create a sample well (Fig. 3.7 (b)).
Fig. 3.7. (a) Un-etched copper chip. (b) Etched sensor chip with a PDMS sample well
and copper connector pads.
Each sensor chip was composed of four SWCNT coated wires assembled to create a 2 x 2
junction array. The wires were placed orthogonally and fixed onto the connector pads via
soldering. Two 100 g weights were used to adjust the tension of the wire during soldering
(Fig. 3.8 (a)). The weight used to adjust wire tension was determined using the allowable
stress ( ) equation (Eq. 3.2).
(3.2)
Where represents allowable stress (PSI), F represents force (lb), and A is area (in2).
27
A force of 182 g was calculated as the maximum force allowed for tensile recovery of the
microwire based on a of 116000 PSI and an area of 3.14 x 10-6 in2. Therefore, 100 g
weights were within the allowable stress range, maintaining the wire’s tensile strength.
After soldering, the wire was cut to remove excess wire hanging on the sides of the chip
(Fig. 3.8 (b)).
Fig. 3.8. Schematic of wire soldering (a) set-up and (b) procedure.
First, two SWCNT coated wires were soldered onto the chip, parallel to each other (Fig.
3.9 (a)). After, two mica spacers were placed on opposite sides of the chip, in between the
connector pads and PDMS well (Fig. 3.9 (b)). Subsequently, the remaining two wires
were aligned over the mica spacers and soldered onto the connector pads, perpendicular
to the first two wires (Fig. 3.9 (c)), and ready for bacterial detection (Fig. 3.9 (d)).
(b) (a)
28
Fig. 3.9. Multi-junction sensor chip assembly (a) Two SWCNT coated microwires are
soldered onto the connector pads. (b) Mica spacers are placed on the chip. (c) Two more
coated wires are soldered on top of the mica spacers perpendicular to the first two wires.
(d) Image of assembled sensor chip.
3.5.2. Multiplex circuit design and system scale down
A multiplexing circuit was built to measure current at the four bio-nano junctions
in real-time. Since our target is to design a portable, simple and cheap biosensor, the
device should be equipped with a fast central processing unit (CPU) and memory to solve
second order non-linear simultaneous equations induced from the relationship between
the four junctions. Therefore, the basic stamp 2 module (Parallax Inc.) with 20 MHz CPU
speed and 32 bytes RAM (random access memory) was selected to plant numerical codes
using its own language (PBASIC) to automatically calculate the resistance of each
junction within five seconds. The Jacobian matrix method was selected as the numerical
method to solve the non-linear simultaneous equations. The fabricated circuit was
composed of three modules: a switch, power source, and picoammeter module.
A switching module was designed with four relays, one relay (Fig. 3.10 (a)) per
junction. Relays can be divided into two types of switches, mechanical and solid state
relay (SSR). Originally, four single pole double throw (SDPT) micro mechanical relay
switches, operated at 5 VDC, were used. However, due to its slower switching time over a
few milliseconds (ms) and larger size, SDPT relays were replaced with SSR switches
(Fig. 3.10 (b)). SSR switches offered faster switching time of a few nanoseconds (ns) and
smaller dimensions to aide in sensor scale down and portability.
29
Fig. 3.10. Switching module circuit design. (a) Single relay circuit made up of a ground
(GND), negative-positive-negative (NPN) transistor, and voltage drain source (Vdd) is
used to direct current (I) into the SDPT relay where it is switched between normally open
(NO) and normally closed (NC) contact, depending on the signal of I/O Pin. The SDPT
controls the connection to the junction biosensor for current measurement. (b)
Multiplexing multi-junction circuit composed of four relays. Circuits are designed in
Auto CAD (2014, San Rafael, CA).
A variable DC power supply for the biosensor was built by applying the field
effect transistors (FETs) and a digital to analog converter (DAC) embedded into the basic
stamp 2 (Fig. 3.11). The voltage with double effective digits under zero can be changed
from 0 to 5 V. In the experiment, a voltage of 1V was decided empirically to achieve the
micro and nano ranges of current suitable to the device.
(a) (b)
30
Fig. 3.11. DC power supply circuit. Alternating current (AC) is transformed by the digital
variac and converted into direct current (DC) by the rectifier. Capacitors (C) are used to
control the fluctuation of voltages to maintain a constant VDC for power supply.
To measure the micro and nano ranges of current, the proper feedback resistors
were installed. An operation amplifier (OP) was used to overcome burden voltages. The
OP Amp was selected due to its small input offset voltage, negligible input bias current,
and low power consumption. A diode and fuse in parallel with the current meter were
installed to protect the circuit against high currents that may occur due to improper range
selection (Fig. 3.12). The device was manually calibrated using variable resistors.
Fig. 3.12. Amperemeter circuit.
The multiplexed circuit module was housed in a 3D printed box (120 x 92 x 52
mm) with an on/off switch (Fig. 3.13 (a)). A sensing platform (80 x 80 x 30 mm) with an
31
acrylic observation window (34 x 34 mm) was also built using the 3-D printer to create
an isolated sensing environment during electrical current measurements.
Fig. 3.13. Multi-junction sensor system. (a) 3-D printed multiplexing circuit module and
sensing platform. Close up image of the opened sensing platform (b) before a multi-
junction sensor chip is added and (c) after sensor is secured.
The base of the platform was designed with a 26 x 26 x 2 mm well in the center for the
multi-junction sensor chip and eight magnetic slots, two on each side (Fig. 3.13 (b)), to
secure device connection (Fig. 3.13 (c)).
3.5.3. Multiplexing sensitivity tests
Each experiment was conducted in triplicate with 10 μL bacterial samples applied
to each junction. The sensitivity of the multi-junction sensor was evaluated using pure
serially diluted E. coli and S. aureus cultures. Anti- E. coli antibody functionalized multi-
junction sensors were tested with 101-105 CFU/mL E. coli K-12 solutions (Fig. 3.14).
While, anti- S. aureus antibody functionalized sensors were tested with 101-105 CFU/mL
S. aureus. Linear regression curves of current measurements were graphed as calibration
curves for the multi-junction arrays.
Multiplexing circuit module
Isolated sensing
platform
Magnetic connectors
(a) (b)
(c)
32
Fig. 3.14. Multi-junction sensor functionalization set-up with (a) anti-E. coli antibodies
and (b) anti-S. aureus antibodies tested with pure E. coli and S. aureus samples.
To evaluate the sensors’ ability to simultaneously detect E. coli and S. aureus,
each sensor chip was functionalized with E. coli and S. aureus antibodies. The two top
junctions of each sensor represented by R1.1 and R1.2 were functionalized for E. coli
detection and the bottom two junctions, R2.1 and R2.2, were functionalized for S. aureus.
As a preliminary study for simultaneous detection, 10 μL samples of a mixed culture of
E. coli and S. aureus (102-105 CFU/mL) were applied to each junction and current values
were measured (Fig. 3.15 (a)). The multi-junction’s batch sensing performance in larger
volumes (i.e. 100 μL microbial cocktail) was also investigated (Fig. 3.15 (b)). Current
measurements were compared to the calibration curves generated from the sensitivity
tests.
(a) (b)
33
Fig. 3.15. Schematic illustration of batch-type multiplexing tests using (a) 10 μL
microbial cocktail samples of E. coli and S. aureus and (b) 100 μL samples.
3.5.4. Data analysis
Three replications were conducted per experiment. To determine the sensitivity of
the multi-junction array and calibrate the sensor for pure cultures of E. coli and S. aureus,
the current (I) was measured at each of the four junctions and averaged together to obtain
a representative I measurement for the sensing chip. The ∆I for E. coli and S. aureus
detection were also calculated to clearly observe the change in current signal to compare
with the control.
To determine the multiplexing capability of the sensor to simultaneously detect E.
coli and S. aureus in a mixed sample, I measurements from the anti-E. coli functionalized
junctions, R1.1 and R1.2, and anti-S. aureus junctions, R2.1 and R2.2, were averaged.
A mathematical circuit was also built to calculate the resistance at each junction
(R1.1, R1.2, R2.1, and R2.2), based on the assumption that as bacterial cells bind to each
junction, all four junctions become interrelated. For example, the electric current that
passes through one junction, R1.1, may also pass through the other junctions, R1.2, R2.1,
and R2.2 (Fig. 3.16 (a)). Thereby, R1.1 is parallel to the series, R1.2, R2.1, and R2.2 (Fig. 3.16
(b)). Thus, the following circuit equations were set-up based on the circuit theory (Fig.
3.16 (c)). The second order, non-linear simultaneous equations were solved by
Mathematica software (Wolfram Research, Champaign, IL). The experimental current
values were used in the equations to calculate the resistance of each junction. The
(a) (b)
34
calculated resistances of the sensor tested with pure and microbial cocktails were
compared.
Fig. 3.16. Mathematical circuit developed to determine junction resistance. (a) Electric
current pathway and (b) circuit equivalent at a single junction. (c) Equations used to
calculate junction resistance.
(a) (b) (c)
35
4. RESULTS & DISCUSSION
4.1. SWCNT-based single junction biosensor
A disposable, label-free bio-nano functionalized junction sensor was designed, in
which two SWCNT-coated microwires were aligned in a cross bar formation. The wires’
point of intersection was functionalized with antibodies to act as a bio-molecular sensing
probe. By coating the wires’ cross section with bio-nano materials, a sandwich of layered
SWCNTs and biomolecules existed at the point of intersection between the microwires,
thus creating a bio-nano junction when target analytes bind and form immune complex
reactions.
The fabricated sensor was evaluated to detect Escherichia coli K-12, as the
model microorganism. Gold tungsten wires (Ø = 50 µm) dip-coated with 1% PEI solution
and SWCNTs were assembled to form a crossbar junction. The junction was
functionalized with streptavidin and polyclonal biotinylated anti-E. coli antibodies for
specificity. Changes in electrical current (∆I) after bioaffinity reactions between bacterial
cells (E. coli K-12) and antibodies on the SWCNT surface were monitored to evaluate the
sensor’s performance. Current-voltage (I-V) curves demonstrated SWCNTs’ signal
amplifier role in the sensing platform, as the averaged ∆I increased from 33 nA to 290 nA
with the presence of SWCNTs in a 108 CFU/mL concentration of E. coli. Sensitivity tests
determined a linear relationship (R2 = 0.973) between the changes in current and
concentrations of E. coli in range of 102-105 CFU/mL. Current decreased as cell
concentrations increased, due to increased immune complex reactions formed on the bio-
nano modified surface. The detection limit of the developed sensor was 103 CFU/mL
with a detection time of 2 min.
4.1.1. SWCNT dip coating surface morphology
Surface morphology of the junction sensor is an important factor affecting the
signal response, particularly, the morphology of the SWCNT coating, since networks of
SWCNTs form electrical pathways for electrical measurement. A FESEM was used to
observe the surface topographies of the modified wire electrodes.
An initial dip-coat of PEI was applied to modify the surface polarity of the wire.
36
PEI promotes SWCNT adhesion by creating a charged surface with its amine groups for
high binding affinity for SWCNTs (Rouse et al., 2004). A preliminary SEM study was
conducted to determine the appropriate concentration of PEI needed for uniform SWCNT
adhesion. Originally, a 50% PEI solution was used as purchased (Lu and Jun, 2012).
When compared to the bare wire electrode surface (Fig. 4.1 (a)), the 50% PEI coating
appeared to have formed a thick polymer layer (Fig. 4.1 (b)). Whereas, a highly diluted
1% PEI, as suggested by Cairns, 2013, produced a thin polymer coating on the wire’s
surface (Fig. 4.1 (c)). A 0.1% PEI solution was also considered (Cairns, 2013). However,
since the 1% solution showed minimal differences between the control (bare wire), 0.1%
solution was not photographed.
The 50%, 1%, and 0.1% PEI coated wires were dip-coated into the SWCNT-DMF
solution (Fig. 4.2). The SWCNT networks on the 50% PEI-wire were fully embedded
into the polymer layer, with uneven distribution (Fig. 4.2 (a)). SWCNTs colloids on the
1% PEI layer were evenly dispersed and protruded from the surface; suitable for further
SWCNT sidewall functionalization (Fig. 4.2 (b)). SWCNTs were also observed on the
0.1% PEI surface, but networks were scarce (Fig. 4.2 (c)). Therefore, based on
preliminary experiments, microwires were coated with 1% PEI solution before being dip-
coated into the SWCNT-DMF colloidal solution.
Fig. 4.1. SEM images of a microwire electrode. (a) Electrode surface before PEI coating,
(b) after 50%, and (c) 1% PEI coating.
37
Fig. 4.2. SEM images of SWCNTs adhered to (a) 50%, (b) 1%, and (c) 0.1% PEI coated
microwires.
SEM images were also used to determine the appropriate number of dip-coats on
the 1% polymer layer. Surface morphology was observed after one, two, and three
SWCNT dip-coats. SWCNTs can be seen on the PEI-modified wire surface after one
SWCNT dip-coat (Fig. 4.3 (a, b)). The vapor/solid/liquid interfacial energy created a
concaved meniscus of the SWCNT solution during wire withdrawal, enabling SWCNT
colloids to flow into the meniscus by capillary force and adhere to the wire surface (Jang
et al., 2008). Though SWCNTs appeared to be distributed on the surface, only a thin
SWCNT network was formed.
38
Fig. 4.3. SEM images corresponding to the number of SWCNT dip-coats. (a, b) One
SWCNT dip-coat. (c, d) Two dip-coats. (e, f) Three dip-coats. Figures a, c, and e in the
first column are captured at the 5 μm scale; whereas figures b, d, and f in the second
column are taken at 500 nm scale.
The SWCNT network density increased after a second dip-coat (Fig. 4.3 (c, d)). The self-
assembly of SWCNTs into a dense network is due to the intermolecular van der waal
force between SWCNTs (Jang et al., 2008). A third dip-coat produced a slight increase in
surface morphology at the 500 nm scale (Fig. 4.2 (f)). Jang et al., (2008) and Ng et al.,
(2008) observed similar trends, in which large numbers of SWCNTs were coated on the
substrate in the first and second dip-coatings, while the numbers of SWCNTs coated on
the substrate decreased in the subsequent coatings. This can be explained by the substrate
being gradually covered by SWCNT networks, resulting in a wider meniscus angle and a
39
decrease of solution influx. Although the third dip-coat appeared to have a dense
SWCNT network, similar to the second-dip coat surface morphology, large amounts of
SWCNT bundles adhered to the three dip-coated surface, implying a non-uniform surface
(Fig. 4.2 (e)).
SEM images of PEI-SWCNT morphology demonstrated that the dip-coating
method allowed for easy control of SWCNT networks onto the 1% PEI layer and that a
two-dip repetition was suitable to achieve uniform SWCNT networks.
4.1.2. Characterization of the bio-nano functionalization layers
Surface chemistry is critical to the physical properties of SWCNTs, as every atom
is on the surface. SWCNT sidewall functionalization is important for self-assembly on
surfaces. In order to preserve the sp2 nanotube structure and thus their electronic
characteristics, it is imperative to functionalize the sidewalls in a non-covalent manner
(Chen et al., 2001). The technique based on molecular self-assembly was used for
antibody immobilization onto the junction surface. The self-assembled monolayers were
composed of PEI, SWCNTs, streptavidin, and biotinylated antibodies for detection of
target bacteria.
Due to the cationic nature of PEI and hydrophobic nature of SWCNTs, negatively
charged streptavidin readily adsorbed onto the PEI-SWCNT modified surface via
electrostatic and hydrophobic interactions (Chen et al., 2001; Lu et al., 2011). Then,
biotinylated antibodies non-covalently bind to streptavidin, as a result of streptavidin’s
tetravalency for biotin (Pei et al., 2001; So et al., 2008). The dynamics of charge transfer
at the electrode interface are strongly influenced by the nature of the electrode surface
(Pei et al., 2001). In that, the adsorption of insulating materials on the SWCNT-modified
electrode junction is anticipated to alter the electron transfer at the microwire’s surface.
Thus, changes in electrode behavior after each modification step were investigated
through current-voltage (I-V) measurements.
Figure 4.4 shows the I-V responses for the bio-nano junction sensor throughout the
step-wise functionalization and detection process. An input voltage was swept from 0 to 1
VDC to study the electrical response of the functionalized layers (Kim et al., 2013). An
electrical measurement was conducted for each layer.
40
The average current at 1 VDC for a bare, non-coated junction was 0.001 µA.
Followed by an increase to 5.7 µA after PEI-SWCNTs surface modification. The
dramatic increase in current demonstrated that the SWCNT colloidal layer functioned as
a conductor for electron transfer (Grunner et al., 2006). Once the junction was
functionalized with streptavidin, the current decreased to 1.5 µA. Current further dropped
to 0.6 µA when the biotinylated antibodies were applied to the streptavidin layer, and 0.3
µA after exposure to a high concentration of E. coli K-12 (108 CFU/mL). Since the
electronic properties of SWCNTs are a strong function of its atomic structure, surface
modification may induce changes in the electrical conductance of the nanotubes (Kang et
al., 2006). Particularly, protein adsorption on SWCNTs have been observed to cause a
drop in conductance due to the small electrostatic disturbances from the biomolecules as
they modulate the local work-function and band alignment of the SWCNT networks
(Boussaad et al., 2003; Chen et al., 2004; Heller et al., 2008). Semi-conducting SWCNTs
have also been studied to exhibit an electron scattering mechanism in response to target
analyte binding, as seen from an overall drop in current measurements (Allen et al.,
2007).
41
Fig. 4.4. I-V curve from 0 to 1 VDC corresponding to individual sensor modification
layers.
The effect of SWCNTs on the sensing magnitude was also evaluated. Figure 4.5
demonstrates a greater suppression of current at 1 VDC during the functionalization
process and bacterial detection when SWCNTs were integrated into the sensor. A larger
Δ I of 290 nA was measured using a SWCNT-modified junction sensor with 108
CFU/mL E. coli solution (Fig. 4.6). Whereas, I of 33 nA was measured from a junction
sensor without SWCNTs and applied to the same E. coli concentration. The results
indicated that the network of SWCNTs enhanced the signal response by seven-folds. The
greater ΔI may be attributed to the change in electronic structure of SWCNTs occurring
as a result of functionalization and E. coli cell loading. In addition, SWCNTs may have
enlarged the sensing platform surface area for an increased amount of bio-receptors,
thereby imparting a higher sensitivity towards E. coli (Zhao et al., 2011).
42
Fig. 4.5. Effect of SWCNTs on signal response during step-wise surface modification
and E. coli K-12 detection at 1 VDC.
Sensor functionalization with streptavidin and biotinylated anti-E. coli antibodies was
validated by SEM imaging, as E. coli cells successfully immune-reacted on the
biocompatible surface (Fig. 4.6 inset).
43
Fig. 4.6. Averaged change in current in response to captured E. coli on a junction sensor
with and without SWCNTs. Inset: SEM image of E. coli captured on the bio-nano
junction sensor surface through immune complex reactions with biotinylated anti-E. coli
antibodies. Significant differences between signal measurements are indicated by the
different superscripts at a 95% confidence level (probability < 0.05).
4.1.3. Sensitivity tests for Escherichia coli K-12
Sensitivity test were conducted with 101-105 CFU/mL concentrations of E. coli K-
12. Current measurements decreased as cell concentrations increased, hence, an increase
in change in electrical current (∆I) in relationship to the increased bacterial loadings on
the bio-nano modified surface. The cause for this finding is not yet fully understood, but
the phenomenon in which binding sites of antibodies on the junction become saturated
with antigen concentrations, inducing a greater degree of change in the SWCNT
electrical properties, may be partly responsible. ∆I values were analyzed in a linear
regression model. A goodness-of-fit with R2 = 0.951 was achieved between the ∆I
measurements and concentrations of E. coli suspension in the range of 101-105 CFU/mL.
However, the use of a 10 µL sample volume limits the representation of 101 CFU/mL E.
coli cells, thus the ∆I value corresponding to the 101 CFU/mL sample may be due to
background noise. Therefore, only E. coli samples starting from 102-105 CFU/mL were
44
used to determine the linear regression for detection (R2 = 0.973) (Fig. 4.7). Based on the
significant differences between concentrations, the detection limit for the E. coli specific
junction sensor was 103 CFU/mL with a detection time of 2 min.
Fig. 4.7. Relationship between changes in current and concentrations of E. coli K-12
(101-105 CFU/ mL) bound to the E. coli functionalized junction sensor. Significant
differences between signal measurements and bacteria concentration are indicated by the
different superscripts at a 95% confidence level (SAS, probability < 0.05).
4.1.4. Specificity tests against Staphylococcus aureus
Specificity of the sensor towards E. coli was measured against serial diluted
samples of a pure culture of S. aureus (103-105 CFU/mL) and compared to E. coli ∆I
values measured from the sensitivity experiments. After S. aureus was applied to the anti-
E. coli antibody functionalized sensor, the ∆I varied from 7-10 nA, respectively, which
may be attributed to the sensor’s background noise or non-specific binding of S. aureus
on the junction surface (Fig. 4.8). In comparison, a ∆I ranging from 36-106 nA was
observed with pure E. coli samples in the same concentration range, indicating the
specificity of the functionalized sensor towards E. coli.
45
Fig. 4.8. E. coli junction sensor signal response to S. aureus samples (103-105 CFU/ mL)
in comparison to E. coli detection signal responses. Significant differences between
signal measurements are indicated by the different superscripts at a 95% confidence level
(probability < 0.05).
4.2. SWCNT-based multi-junction biosensor
Based on the single SWCNT junction biosensor design and sensing performance,
a multi-junction sensor was investigated for sensitive and multiplex capabilities. In
addition, the sensing system was scaled down for portable detection.
PEI-SWCNT coated microwires were aligned to form a 2 x 2 junction sensing
array and evaluated for multiplexed detection of E. coli K-12 and S. aureus. 10 µL
bacterial samples were applied to each junction. Serial diluted samples of E. coli and S.
aureus were used to create a calibration curve to be compared with sensing
measurements from mixed culture samples. Electric current measurements in response to
bacterial detection at the individual junctions were averaged to determine the ∆I values to
represent the 2 x 2 array. A linear regression was observed for both the E. coli and S.
aureus functionalized array sensors, R2 = 0.978 and R2 = 0.992, in range of 102-105
46
CFU/mL. Mixed samples of E. coli and S. aureus showed similar measurement trends for
multiplexed detection in 10 µL and 100 µL batch samples.
4.2.2. Sensitivity tests for E. coli K-12 and S. aureus
The multi-junction array sensor was functionalized with anti-E. coli antibodies
and tested with pure E. coli K-12 concentrations from 101-105 CFU/mL. The averaged
current values of the sensor were used to create an E. coli calibration curve (Fig. 4.9 (a)).
As observed during single junction measurements, the current decreased as the E. coli
concentration increased. When no bacteria were present (control, 100 CFU/mL) the
current value was 757 ± 13 nA, respectively. Once E. coli was added to the junctions,
antigen-antibody reactions occurred on the bio-nano modified junction surface (Fig. 4.9
(b)), changing the SWCNT morphology, thus altering its electrical properties. The current
was reduced down to 384 ± 36 nA in a high concentration of E. coli, 105 CFU/mL. The
control measurements confirmed that the responses were caused exclusively by the
binding between E. coli and the antibody and the subsequent transduction of the SWCNT
layer.
Fig. 4.9. (a) Electrical current calibration curve for an anti-E. coli antibody functionalized
multi-junction sensor tested with a negative control (100 CFU/mL) and E. coli in the
range of 101-105 CFU/mL. (b) SEM image of E. coli cells bound to sensor surface.
(a) (b)
47
The same trend was observed with the anti- S. aureus functionalized sensor tested
in pure S. aureus concentrations, but on a different current scale. A drop in current was
measured from 620 ± 15 nA (control) to 131 ± 18 nA, respectively, as S. aureus cells
(101-105 CFU/mL) were captured on the sensor (Fig 4.10). This suggested that the multi-
junction sensor is capable of detecting E. coli and S. aureus simultaneously in a mixed
bacterial solution, since the current signals vary depending on the target analyte and
antibody. For example, a 103 CFU/mL concentration of E. coli detected on E. coli
specific junctions had an averaged current signal of 532 ± 22 nA, while the same
concentration of S. aureus detected with S. aureus specific junctions generated a current
of 332 ± 28 nA. A similar current trend was observed in Fernandes et al., (2014) study, in
which E. coli O157:H7 cells detected on a nanoparticle-based DNA electrochemical
biosensor exhibited higher current values than S. aureus of the same concentration.
Fig. 4.10. (a) Electrical current calibration curve for an anti-S. aureus antibody
functionalized multi-junction sensor tested with a negative control (100 CFU/mL) and S.
aureus in the range of 101-105 CFU/mL. (b) SEM image of S. aureus cells bound to
sensor surface.
(a) (b)
48
To highlight the sensor’s ability to transduce the biorecognition reactions between
antibody and antigen into measurable signals, the ∆I was calculated for E. coli and S.
aureus sensors in range of 101-105 CFU/mL (Fig. 4.11). As bacterial concentrations
increased, a greater change in current was measured proportional to the drop in current.
The E. coli sensor’s ∆I ranged from 24 to 373 nA, whereas, the S. aureus sensor’s ∆I
varied from 40 to 489 nA. A linear regression was observed for both the E. coli and S.
aureus functionalized array sensors, R2 = 0.978 and R2 = 0.992, in range of 102-105
CFU/mL.
Fig. 4.11. Relationship between changes in current and concentrations of E. coli and S.
aureus from 101-105 CFU/mL. Average signals (current drop) with different superscripts
are significantly different at 95% confidence level (probability < 0.05). *E. coli and S.
aureus ∆I values were analyzed separately
In comparison to the single junction sensor’s ∆I relationship with E. coli (Fig.
4.7), the multi-junction sensor’s ∆I values for the 102 CFU/mL samples for both E. coli
and S. aureus were significantly different from the values at 101 CFU/mL. The design of
the 2 x 2 junction array on the sensing chip may have contributed to the sensor’s
49
sensitivity due to the improved stability and uniformity of the sensor and the use of four
junctions to measure a sample. In addition, the magnitude of the ∆I differed between the
two bacterial strains, which makes simultaneous detection in samples with various
pathogens possible.
4.2.3. Simultaneous detection of E. coli and S. aureus
To explore the response of the multi-junction sensor for simultaneous bacterial
detection, mixed samples of E. coli and S. aureus were tested. Current responses from
batch mixed samples, when 10 µL volume samples were pipetted on each junction and
100 µL samples covered all four junctions, were analyzed. Current measurement values
corresponding to E. coli and S. aureus detection followed the same decreasing trend seen
in the pure microbial samples (Fig. 4.12).
Fig. 4.12. Current measurements for simultaneous detection of E. coli and S. aureus in 10
and 100 µL samples in comparison to calibration measurements.
Data suggests that there is potential for multiplexed detection of different pathogens
when selective antibodies are used and tested in batch samples. Individual 10 µL samples
50
applied to each junction show a closer current trend to the calibration values; however,
even a larger sample volume can be measured for multiplexed detection.
4.2.4. Mathematical circuit for multi-junction resistance calculation
A mathematical circuit equation was designed to calculate the resistance at each
junction. Based on the assumption that all four junctions are connected to each other
when bacterial samples are applied, the current measurement at one junction may be
affected by the biorecognition events occurring at the other three junctions. Therefore, the
equivalent circuit equations were used to determine the resistance of the specific junction
without the bias from the other junctions. Experimental current values measured at the
four junctions were plugged into the second order non-linear simultaneous equations. The
resistance corresponding to each junction was obtained from the numerical method. The
calculated resistance values of top (R1.1 and R1.2) and bottom (R2.1 and R2.2) junctions
were averaged to represent the resistance for E. coli and S. aureus, respectively (Table
4.1). The averaged resistance for E. coli and S.aureus detected in pure and mixed (10 µL)
samples were compared.
The resistances obtained for the multi-junction sensor tested in the pure bacterial
samples had a good agreement with the values calculated from the microbial cocktail
samples, as there was no significant difference between the samples in range of 102-104
CFU/mL for E. coli detection and 102-105 CFU/mL for S. aureus detection. A significant
difference was observed between the single and mixed E. coli 105 CFU/mL
measurements, which may have been due to non-specific binding (Radke and Alocilja,
2005). In addition, the resistance values increased from 102-105 CFU/mL and
corresponded to the experimental current data trend, as resistance is inversely related to
current. Increased bacteria cell binding at the junctions serve as an insulating material,
thus altering the interfacial electron transfer at the junctions and modulating the electrical
properties of the SWCNT sensing platform.
51
Table 4.1. Multi-junction sensing array resistance calculated using the developed
mathematical circuit
Concentration (CFU/mL) Analyte detection Resistance (10-3 Ω)
E. coli:
Averaged R1.1 and R1.2
S. aureus:
Averaged R2.1 and R2.2
102 Single 2.15 ± 0.07a 3.56 ± 0.13a*
Mixed 2.14 ± 0.04a 3.70 ± 0.06a*
103 Single 2.36 ± 0.08b 4.10 ± 0.10b*
Mixed 2.35 ± 0.05b 4.21 ± 0.09b*
104 Single 2.59 ± 0.06c 5.19 ± 0.08c*
Mixed 2.73 ± 0.21c 5.17 ± 0.06c*
105 Single 2.80 ± 0.05c 7.04 ± 0.06d*
Mixed 3.44 ± 0.02d 6.96 ± 0.10d*
52
5. CONCLUSION
In this study, the incorporation of immobilized antibodies and a SWCNT-
modified sensing platform into a disposable, label-free bio-nano functionalized junction
sensor shows great promise for the detection of bacteria in serially diluted cultures. The
sensing signal of the fabricated single junction SWCNT-based biosensor was amplified
seven-folds with the presence of SWCNTs. Testing the biosensor in E. coli K-12
concentrations from 102-105 CFU/mL demonstrated a linear relationship (R2 = 0.973)
between E. coli concentrations and changes in electric current. Specificity tests against S.
aureus demonstrated minimal non-specific binding on the E. coli functionalized sensor.
A detection limit of 103 CFU/mL with a detection time of 2 min was feasible. To improve
the single junction sensing performance, a multi-junction sensor was designed with
improved fabrication uniformity and stability. The sensing system was scaled down to a
485 g portable and affordable system composed of a multiplexing circuit module, isolated
sensing platform, and multi-junction array chip. Testing the 2 x 2 multi-junction array
with pure and mixed E. coli and S. aureus concentrations showed potential for batch-type
multiplexed detection down to 102 CFU/mL. A linear regression with an R2 = 0.978 was
observed for E. coli and R2 = 0.992 for S. aureus functionalized array sensors, in range of
102-105 CFU/mL. Microbial cocktail samples of E. coli and S. aureus showed similar
measurement trends for multiplexed detection in 10 µL and 100 µL batch samples.
The SWCNT biosensor developed represents the first step toward developing a
bio-nano junction sensing array for simultaneous detection of multiple pathogens. By
offering simplicity, portability, and rapid detection, this biosensor shows potential as a
field-deployable system for food and agricultural applications. Future studies will
consider the evaluation of the developed sensor with food samples as well as bacterial
cell concentration methods to further improve the limit of detection.
53
REFERENCES
Aguilar, A.D., Foranzi, E.S., Li, X., Tao, N., Nagahara, L.A., Amlani, I., Tsui, R. 2005.
Chemical sensors using peptide-functionalized conducting polymer nanojunction
arrays. Appl Phys Lett. 87, 193108-1-193108-3.
Allen, B.L., Kichambare, P.D., Star, A. 2007. Carbon nanotube field-effect-transistor-
based biosensors. Adv Mater. 19, 1439-1451.
Arora, P., Sindhu, A., Dilbaghi, N., Chaudhury, A. 2011. Biosensors as innovative tools
for the detection of food borne pathogens. Biosens Bioelectrons. 28, 1-12.
Artyukhin, A.B., Stagermann, M., Friddle, R.W., Stroeve, P., Bakajin, O., Noy, A. 2006.
Controlled electrostatic gating of carbon nanotube FET devices. Nano Lett. 6,
2080-2085.
Barone, P.W., Baik, S., Heller, D.A., Strano, M.S. 2005. Near-infrared optical sensors
based on single-walled carbon nanotubes. Nature Materials. 4, 86-U16.
Besteman, K., Lee, J., Wiertz, F.G.M., Heering, H.A., Dekker, C. 2003. Enzyme-coated
carbon nanotubes as single-molecule biosensors. Nano Lett. 3, 727-730.
Bhunia, A.K., Geng, T., Lathrop, A., Valadez, A., Morgan, M.T. 2004. Optical
immunosensors for detection of Listeria monocytogenes and Salmonella
enteritidis from food. Proceedings of SPIE - The International Society for Optical
Engineering. 5271, 1-6.
Boer, E., Beumer, R.R. 1999. Methodology for detection and typing of foodborne
microorganisms. Int J Food Microbiol. 50, 119-30.
Boussaad, S., Tap, N.J., Zhang, R., Hopson, T., Nagahara, L.A. 2003. In situ detection of
cytochrome c adsorption with single walled carbon nanotube device. Chem
Commun. 13, 1502-1503.
Burtscher, C., Wuertz, S. 2003. Evaluation of the use of PCR and reverse transcriptase
PCR for detection of pathogenic bacteria in biosolids from anaerobic digestors
and aerobic composters. Appl Environ Microbiol. 69, 4618-27.
Cairns, A. 2013. Optimization study of polyethyleneimine surface coating for microtip
DNA Purification (Doctoral dissertation, University of Washington).
54
Centers for Disease Control and Prevention. 2010. Multistate outbreak of Salmonella
Typhimurium infections linked to peanut butter, 2008-2009 (Final updated).
Retrieved from http://www.cdc.gov/salmonella/typhimurium/update.html.
Centers for Disease Control and Prevention. 2011. Estimates of foodborne illness in the
United States. Retrieved from http://www.cdc.gov/foodborneburden/2011-
foodborne-estimates.html.
Centers for Disease Control and Prevention. 2012. Multistate Outbreak of Listeriosis
Linked to Whole Cantaloupes from Jensen Farms, Colorado. Retrieved from
http://www.cdc.gov/listeria/outbreaks/cantaloupes-jensen-
farms/index.html?s_cid=cs_654.
Chemburu, S., Wilkins, E., Abdel-Hamid, I. 2005. Detection of pathogenic bacteria in
food samples using highly-dispersed carbon particles. Biosens Bioelectron. 21,
491–9.
Chen, R.J., Bangsaruntip, S., Drouvalakis, K.A., Kam, N.W.S., Shim, M., Li, Y.M., Kim,
W., Utz, P.J., Dai, H. 2003. Noncovalent functionalization of carbon nanotubes
for highly specific electronic biosensors. Proceedings of the National Academy of
Sciences of the United States of America. 100, 4984-4989.
Chen, R.J., Choi, H.C., Bangsaruntip, S., Yenilmez, E., Tang, X., Wang, Q., Chang, Y.,
Dai, H. 2004. An investigation of the mechanisms of electronic sensing of protein
adsorption on carbon nanotube devices. J Am Chem Soc. 126, 1563-1568.
Chen, R.J., Zhang, Y., Wang, D., Dai, H. 2001. Noncovalent sidewall functionalization
of single-walled carbon nanotubes for protein immobilization. Am Chem Soc.
123, 3838-3839.
Cherukuri, P., Bachilo, S.M., Litovsky, S.H., Weisman, R.B. 2004. Near-infrared
fluorescence microscopy of single-walled carbon nanotubes in phagocytic cells. J
Am Chem Soc. 126, 15638-15639.
Choi, W., Hong, J. 2012. Rapid electomechanical transduction on a single-walled carbon
nanotube film: sensing fast mechanical loading via detection of electrical signal
change. J Am Chem Soc. 51, 14714-14721.
55
Chunglok, W., Wuragil, D.K., Oaew, S., Somasundrum, M., Surareungchai, W. 2011.
Immunoassay based on carbon nanotubes-enhanced ELISA for Salmonella
enterica serovar Typhimurium. Biosens Bioelectron. 26, 3584-3589.
Dudak, F.C., Boyaci, I.H. 2009. Multiplex detection of Escherichia coli and Salmonella
enteritidis by using quantum dot-labeled antibodies. J Rapid Meth Automat
Microbol. 17, 315-327.
Fernandes, A.M., Zhang, F., Sun, X. 2014. A multiplex nanoparticles-based DNA
electrochemical biosensor for the simultaneous detection of Escherichia coli
O157:H7 and Staphylococcus aureus. Int J Curr Microbiol App Sci. 3, 750-759.
Food and Drug Administration. 2007. FDA finalizes report on 2006 spinach outbreak.
Retrieved from http://www.fda.gov/newsevents/newsroom/pressannouncements/
2007/ucm108873.htm.
Forzani, E.S., Zhang, H.Q., Nagahara, L.A., Amlani, I., Tsui, R., Tao, N. 2004. A
conducting polymer nanojunction sensor for glucose detection. Nano Lett. 4,
1785-1788.
Grunner, G. 2006. Carbon nanotube transistors for biosensing applications. Anal Bioanal
Chem. 384, 322-335.
Hall, R.H. 2002. Biosensor technologies for detecting microbiological foodborne hazards.
Microbes Infect. 4, 425-32.
Heller, I., Janssens, A.M., Mannik, J., Minot, E.D., Lemay, S.G., Dekker, C. 2008.
Identifying the mechanism of biosensing with carbon nanotube transistors. Nano
Lett. 8, 591-595.
Heller, I., Kong, J., Williams, K.A., Dekker, C., Lemay, S.G. 2006. Electrochemistry at
single-walled carbon nanotubes: The role of band structure and quantum
capacitance. J Am Chem Soc. 128, 7353-7359.
Heo, J., Hua, S.Z. 2009. An overview of recent strategies in pathogen sensing. Sensors. 9,
4483-4502.
Huang, T.S., Tzeng, Y., Liu, Y.K., Chen, Y.C., Walker, K.R., Guntupalli, R., Liu, C.
2004. Immobilization of antibodies and bacterial binding on nanodiamond and
56
carbon nanotubes for biosensor applications. Diamond Relat. Mater. 13, 1098-
1102.
Iijima, S. 2002. Carbon nanotubes: past, present, and future. Physica B. 323, 1-5.
Ivnitski, D., Abdel-Hamid, I., Atanasov, P., Wilkins, E. 1999. Biosensors for detection of
pathogenic bacteria. Biosens Bioelectron. 14, 599-624.
Jang, E.Y., Kang, T.J., Im, H.W., Kim, D.W., Kim, Y.H. 2008. Single-walled carbon-
nanotube networks on large-area glass substrate by the dip-coating method. Small.
4, 2255-2261.
Jasson, V., Jacxsens, L., Luning, P., Andreja, R., Uyttendaele, M. 2010. Alternative
microbial methods: an overview and selection criteria. Food Microbiolgy. 27,
710-730.
Kang, I., Heung, Y.Y., Kim, J.H., Lee, J.W., Gollapudi, R., Subramaniam, S.,
Narasimhadevara, S., Hurd, D., Kirikera, G.R., Shanov, V., Schulz, M.J., Shi, D.,
Boerio, J., Mall, S., Ruggles-Wren, M. 2006. Introduction to carbon nanotube and
nanofiber smart materials. Compos Part B-Eng. 37, 382-394.
Karoonuthaisiri, N., Charlermroj, R., Uawisetwathana, U., Luxanamil, P., Kirtikara, K.,
Gajanandana, O. 2009. Development of antibody array for simultaneous detection
of foodborne pathogens. Biosens Bioelectron. 24, 1641-1648.
Karousis, N., Tagmatarchis, N. 2010. Current progress on the chemical modification of
carbon nanotubes. J. Am. Chem. Soc. 110, 5366-5397.
Katz, E., Willner, I. 2004. Biomolecule-functionalized carbon nanotubes: applications in
nanobioelectronics. ChemPhysChem. 5, 1084-1104.
Kawasaki, S., Fratamico, P.M., Horikosi, N., Okada, Y., Takeshita, K., Sameshima, T.,
Kawamoto, S. 2009. Evaluation of a multiplex PCR system for simultaneous
detection of Salmonella spp., Listeria monocytogenes, and Escherichia coli
O157:H7 in foods and in food subjected to freezing. Foodborne Pathog Dis. 6, 81-
89.
Kim, J-H., Hiraiwa, M., Lee, H-B., Lee, K-H., Cangelosi, G.A., Chung, J-H. 2013.
Electrolyte-free amperometric immunosensor using a dendritic nanotip. RSC Adv.
13, 4281-4287.
Lampel, K.A., Orlandi, P.A., Kornegay, L. 2000. Improved template preparation for
57
PCR-based assays for detection of food-borne bacterial pathogens. Appl Environ
Microbiol. 66, 4539-4542.
Ligler, F.S., Sapsford, K.E., Golden, J.P., Shriver-Lake, L.C., Taitt, C.R., Dyer, M.A.,
Barone, S., Myatt, C.J. 2007. The array biosensor: portable, automated systems.
Anal Sci. 23, 5-10.
Lin, Y., Elkin, T., Taylor, S., Gu, L., Chen, B., Veca, L.M., Zhou, B., Yang, H., Brown,
J., Joseph, R., Jones, E., Jiang, X., Sun, Y-P. 2006. Preparation, characterization,
and evaluation of immune carbon nanotubes. Micochim Acta. 152, 249-254.
Lopez-Campos, G., Martinez-Suarez, J.V., Aguado-Urda, M., Alonso, V.L. 2012.
Microarray detection and characterization of bacterial foodborne pathogens.
SpringerBriefs in Food, Health, and Nutrition. 9, 126-158.
Lu, L., Chee, G., Yamada, K., Jun, S. 2013. Electrochemical impedance spectroscopic
technique with a functionalized microwire sensor for rapid detection of foodborne
pathogens. Biosens Bioelectron. 42, 492-495.
Lu, L., Jun, S. 2012. Evaluation of a microwire sensor functionalized to detect
Escherichia coli bacterial cells. Biosens Bioelectron. 36, 257-261.
Malorny, B., Paccassoni, E., Fach, P., Bunge, C., Martin, A., Helmuth, R. 2004.
Diagnostic Real-Time PCR for Detection of Salmonella in Food. Appl Environ
Microbiol. 70, 7046-7052.
Mandal, P.K., Biswas, A.K., Choi, K., Pal, U.K. 2011. Methods for rapid detection of
foodborne pathogens: an overview. Am J Food Technol. 6, 87-102.
Maroto, A., Balasubramanian, K., Burghard, M., Kern, K. 2007. Functionalized metallic
carbon nanotube devices for pH sensing. Chemphyschem. 8, 220-223.
Maruccio, G., Marzo, P., Krahne, R., Passaseo. A., Cingolani, R., Rinaldi, R. 2007.
Protein conduction and negative differential resistance in large-scale nanojunction
arrays. Small. 3, 1184-1188.
Matthews, K.R., Montville, T.J. 2008. Food microbiology-an introduction. 2nd ed. New
York: American Society for Microbiology.
McEuen, P.L., Fuhrer, M., Park, H. 2002. Single-walled carbon nanotube electronics.
IEEE T Nanotechnol. 1, 78-85.
58
Mello, L.D., Kubota, L.T. 2002. Review of the use of biosensors as analytical tools in the
food and drink industries. Food Chem. 77, 237-256.
Muhammad-Tahir, Z., Alocilja, E.C. 2003. A conductometric biosensor for biosecurity.
Biosens Bioelectron. 18, 813-819.
Ng, M.H.A., Hartadi, L.T., Tan, H., Poa, C.H.P. 2008. Efficient coating of transparent
and conductive carbon nanotube thin films on plastic substrates. Nanotechnology.
19, 1-5.
Odom, T.W., Huang, J-L., Mieber, C.M. 2002. Single-walled carbon nanotubes from
fundamental studies to new device concepts. Annals of the New York Academy
of Sciences. 1, 203-215.
Ohk, S.H., Koo, O.K., Sen, T., Yamamoto, C.M., Bhunia, A.K. 2010. Antibody–aptamer
functionalized fibre-optic biosensor for specific detection of Listeria
monocytogenes from food. J Appl Microbiol. 109, 808-817.
Okuno, J., Maehashi, K., Kerman, K., Takamura, Y., Matsumoto, K., Tamiya, E. 2007.
Label-free immunosensor for prostate-specific antigen based on single-walled
carbon nanotube array-modified microelectrodes. Biosens Bioelectron. 22, 2377-
2381.
Pal, S., Ying, W., Alocija, E.C., Downes, F.P. 2008. Sensitivity and specificity
performance of a direct- charge transfer biosensor for detecting Bacillus cereus in
selected food matrices. Biosys Eng. 99, 461-468.
Palchetti, I., Mascini, Marco. 2008. Electoanalytical biosensors and their potential for
food pathogen and toxin detection. Anal Bioanal Chem. 391, 455-471.
Park, J., Park, S., Kim, Y-K. 2010. Multiplex detection of pathogens using an
immunochromatographic assay strip. Biochip J. 4, 305-312.
Pathirana, S.T., Barbaree, J., Chin, B.A., Hartell, M.G., Neely, W.C., Vodyanoy, V.
2000. Rapid and sensitive biosensor for Salmonella. Biosens Bioelecton. 15, 135-
141.
Pei, R., Cheng, Z., Wang, E., Yang, X. 2001. Amplification of antigen-antibody
interactions based on biotin labeled protein-streptavidin network complex using
impedance spectroscopy. Biosens Bioelectron. 16, 355-361.
59
Radke, S.M., Alocilja, E.C. 2005. A microfabricated biosensor for detecting foodborne
bioterrism agents. IEEE Sens J. 5, 744-750.
Ravindranath, S.P., Mauer, L.J., Deb-Roy, C., Irudayaraj, J. 2009. Biofunctionalized
magnetic nanoparticle integrated mid-infrared pathogen sensor for food matrixes.
Anal Chem. 81, 2840-2846.
Roda, A., Mirasoli, M., Roda, B., Bonvicini, F., Colliva, C., Reschiglian, P. 2012. Recent
developments in rapid multiplexed bioanalytical methods for foodborne
pathogenic bacteria detection. Micochim Acta. 178, 7-28.
Rouse, J.H., Lillehei, P.T., Sanderson, J., Siochi, E.J. 2004. Polymer/single-walled
carbon nanotube films assembled via donor-acceptor inter actions and their use as
scaffolds for silica deposition. Chem Mater. 16, 3904-3910.
Ruan, C., Yang, L., Li, Y. 2002. Immunobiosensor chips for detection of Escherichia coli
O157:H7 using electrochemical impedance spectroscopy. Anal Chem. 74, 4814-
4820.
Scallan, E., Hoekstra, R.M., Angulo, F.J., Tauxe, R.V., Hoekstra, R.M. 2011. Foodborne
illness acquired in the United States- major pathogens. Emerg Infect Dis. 17, 7-
16.
Scharff, R.L. 2012. Economic burden from health losses due to foodborne illness in the
United States. Food Prot. 75, 123-131.
So, H., Park, D-W., Jeon, E-K., Kim, Y-H., Kim, B.S., Lee, C-K., Choi, S.Y., Kim, S.C.,
Chang, H., Lee, J-O. 2008. Detection and titer estimation of Escherichia coli
using aptamer-functionalized single-walled carbon-nanotube field-effect
transistors. Small. 4, 197-201.
Song, J.M., Vo-Dinh, T. 2004. Miniature biochip system for detection of Escherichia coli
O157:H7 based on antibody-immobilized capillary reactors and enzyme-linked
immunosorbent assay. Analytica Chimica Acta. 507, 115-121.
Star, A., Gabriel, J.P., Bradley, K., Gru1ner, G. 2003. Electronic detection of specific
protein binding using nanotube FET devices. Nano Lett. 3, 459-463.
Subramanian, A., Irudayaraj, J., Ryan, T. 2006. A mixed self-assembled monolayer-
based surface plasmon immunosensor for detection of E-coli O157: H7. Biosens
Bioelectron. 21, 998-1006.
60
Suo, B., He, Y., Tu, S.I., Shi, X. 2010. A multiplex real-time polymerase chain reaction
for simultaneous detection of Salmonella spp., Escherichia coli O157:H7, and
Listeria monocytogenes in meat products. Foodborne Pathog Dis. 76, 619-628.
Taylor, A.D., Ladd, J., Yu, Q.M., Chen, S.F., Homola, J., Jiang, S.Y. 2006. Quantitative
and simultaneous detection of four foodborne bacterial pathogens with a multi-
channel SPR sensor. Biosens Bioelectron. 22, 752–758.
Valadez, A.M., Lana, C.A., Tu, S.I., Morgan, M.T., Bhunia, A.K. 2009. Evanescent
Wave Fiber Optic Biosensor for Salmonella Detection in Food. Sensors. 9, 5810-
5824.
Varshney, M., Li, Y. 2007. Intedigitated array microelectrode based impedance biosensor
coupled with magnetic nanoparticle-antibody conjugates for detection of
Escherichia coli O157:H7 in food samples. Biosens Bioelectron. 22, 2408-2414.
Vashist, S.K., Zheng, D., Al-Rubeaan, K., Luong, J.H.T., Sheu, F. 2011. Advances in
carbon nanotube based electrochemical sensors for bioanalytical applications.
Biotechnol Adv. 29, 169-188.
Velusamy, V., Arshak, K., Korostynska, O., Oliwa, Kamila., Adley, C. 2010. An
overview of foodborne pathogen detection: in the perspective of biosensors.
Biotechnol Adv. 28, 232-254.
Villamizar, R.A., Maroto, A., Rius, F.X., Inza, I., Figueras, M.J. 2008. Fast detection of
Salmonella Infantis with carbon nanotube field effect transistors. Biosens
Bioelectron. 24, 279–283.
Vinayaka, A.C., Thakur, M.S. 2010. Focus on quantum dots as potential fluorescent
probes for monitoring food toxicants and foodborne pathogens. Anal Bioanal
Chem. 397, 1445-1455.
Wanekaya, A.K., Chen, W., Myung, N.V., Mulchandani, A. 2006. Nanowire-based
electrochemical biosensors. Electroanalysis. 18, 533-550.
Wang, C., Irudayaraj, J. 2008. Gold nanorod probes for the detection of multiple
pathogens. Small. 4, 2204-2208.
Wang, J. 2005. Carbon-nanotube based electrochemical biosensors: a review.
Electroanalysis. 17, 7-14.
Wang, R., Ruan, C., Kanayeva, D., Lassiter, K., Li, Y. 2008. TiO2 nanowire bundle
61
microelectrode based impedance immunosensor for rapid and sensitive detection
of Listeria monocytogenes. Nano Lett. 8, 2625-2631.
Wang, Y., Ye, Z., Si, C., Ying, Y. 2011. Subtractive Inhibition Assay for the Detection of
E. coli O157:H7 Using Surface Plasmon Resonance. Sensors, 11, 2728-2739.
Yang, L. 2008. Dielectrophoresis assisted immuno-capture and detection of foodborne
pathogenic bacteria in biochips. Talanta. 80, 551-558.
Yang, L., Bashir, R. 2008. Electrical/ electrochemical impedance for rapid detection of
foodborne pathogenic bacteria. Biotechnol Adv. 26, 135-150.
Zhao, G., Zhan, Z., Dou, W. 2011. A disposable immunosensor for Shigella flexneri
based on multiwalled carbon nanotube/sodium alginate composite electrode. Anal
Biochem. 408, 53-58.
Zhao, Y., Ye, M., Chao, Q., Jia, N., Ge, Y., Shen, H. 2009. Simultaneous detection of
multifood-borne pathogenic bacteria based on functionalized quantum dots
coupled with immunomagnetic separation in food samples. J Agric Food Chem.
57, 517-524.