sensors

Article
A Biosensor-CMOS Platform and Integrated Readout
Circuit in 0.18-µm CMOS Technology for Cancer
Biomarker Detection
Abdulaziz Alhoshany 1, *, Shilpa Sivashankar 2 , Yousof Mashraei 1 , Hesham Omran 3 and
Khaled N. Salama 1
1 Computer, Electrical and Mathematical Science and Engineering Division (CEMSE), King Abdullah
University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia;
yousof.mashraei@kaust.edu.sa (Y.M.); khaled.salama@kaust.edu.sa (K.N.S.)
2 Department of Biomedical Engineering, University of Chapel Hill/North Carolina State University, Raleigh,
NC 27695, USA; ssivash@ncsu.edu
3 The Integrated Circuits Lab, Faculty of Engineering, Ain Shams University, Cairo 11535, Egypt;
hesham.omran@eng.asu.edu.eg
* Correspondence: abdulaziz.alhoshany@kaust.edu.sa

Received: 27 July 2017; Accepted: 21 August 2017; Published: 23 August 2017

Abstract: This paper presents a biosensor-CMOS platform for measuring the capacitive coupling
of biorecognition elements. The biosensor is designed, fabricated, and tested for the detection and
quantification of a protein that reveals the presence of early-stage cancer. For the first time, the
spermidine/spermine N1 acetyltransferase (SSAT) enzyme has been screened and quantified on the
surface of a capacitive sensor. The sensor surface is treated to immobilize antibodies, and the baseline
capacitance of the biosensor is reduced by connecting an array of capacitors in series for fixed exposure
area to the analyte. A large sensing area with small baseline capacitance is implemented to achieve a
high sensitivity to SSAT enzyme concentrations. The sensed capacitance value is digitized by using a
12-bit highly digital successive-approximation capacitance-to-digital converter that is implemented
in a 0.18 µm CMOS technology. The readout circuit operates in the near-subthreshold regime and
provides power and area efficient operation. The capacitance range is 16.137 pF with a 4.5 fF absolute
resolution, which adequately covers the concentrations of 10 mg/L, 5 mg/L, 2.5 mg/L, and 1.25 mg/L
of the SSAT enzyme. The concentrations were selected as a pilot study, and the platform was shown
to demonstrate high sensitivity for SSAT enzymes on the surface of the capacitive sensor. The tested
prototype demonstrated 42.5 µS of measurement time and a total power consumption of 2.1 µW.

Keywords: biosensor; interdigitated electrodes; SSAT enzyme diagnostic platform; capacitive sensor
interface; early cancer detection; CMOS

1. Introduction
Limited access to health-care systems and less healthy lifestyles in third world countries emphasize
the need for low-cost point-of-care diagnostic systems [1]. Many drugs available in the market are
designed to kill, attack, or at least slow the growth of cancer cells. In short, they are poisons with
accompanying side effects. To avoid these side effects, many other cancer therapies have been
pursued. One of the latest therapies for cancer includes metabolomics–based tests. Metabolomics,
one of the “omic” sciences in systems biology, is the global assessment and validation of endogenous
small-molecule biochemicals (metabolites) within a biologic system [2]. The metabolomics-based
test consists of a single, one-shot application of amantadine, a drug which is already approved in
both the US and Canada, after which a urine sample is collected two to four hours later and then

Sensors 2017, 17, 1942; doi:10.3390/s17091942 www.mdpi.com/journal/sensors

Sensors 2017, 17, 1942 2 of 12

analyzed. Acetyl amantadine (AcAm) is recognized as an exogenous cancer biomarker because it is

the product of a metabolic process that is known to be significantly upregulated in cancerous cells.
After ingestion, the anti-Parkinson and anti-viral drug amantadine is acetylated in the body by the
enzyme spermidine/spermine N1 acetyltransferase (SSAT) to give AcAm, which can be detected in
patient urine. Enzymes have been quantified using solid phase extraction (SPE) and tandem liquid
chromatography with mass spectrometry (LCMS) from urine samples [3]. Raman spectroscopy is
employed in medical diagnostics due to its ability to characterize and identify molecules with an
inherent specificity [4]. Further, sensitivity in quantification is improved by Surface Enhanced Raman
Scattering (SERS) [5]. However, techniques previously used to quantify AcAm in urine, such as liquid
chromatography-mass spectrometry (LC-MS), are undesirable for clinical adoption due to high costs
and long run times [6]. Therefore, there is a need to develop low-cost, fast, and portable diagnostic
platforms to screen for the SSAT enzyme.
The SSAT enzyme production varies from person to person depending on the diet and lifestyle of
the person. If the concentration of SSAT enzyme in the blood is higher than the normal production of
theses enzymes then further diagnostics are required to precisely detect cancer progress. Almost all
cells produce polyamines that include spermidine and spermine, but the increased amounts are
prevailed in rapidly multiplying cells [7]. To show a dynamic range and sensitivity, a study on a
population of different regions needs to be carried out to provide statistical analysis. Our aim is to
provide an early-stage cancer detection system rather than monitoring the progression of cancer. As a
pilot study, various concentrations were screened from 10 to 1.25 mg/L of the SSAT enzyme.
Recently, diagnostic platforms have been widely used to measure various types of analytes.
These platforms include DNA analysis [8–10], protein quantification [11–15], and bacteria/cell growth
monitoring [16–18]. Beyond these, CMOS biosensors are promising, as they can provide outstanding
performance in sensing element and signal processing.
In in vitro diagnosis, different transducing mechanisms have been introduced for protein
quantification. Optical sensing is reported in [13]. The photodetectors are used to transduce the
optical properties from the fluorescence labels into an electrical signal. However, an excitation light
source, filtering, and sophisticated sample pre-processing are required which increase the complexity
of the biosensor and the power consumption. Another technique is to use an electrode or impedance
sensor for enzyme/protein detection [19,20]. However, a potentiostat or network/impedance analyzer
is utilized for interfacing the biosensor which is high-cost, large-sized, and complicated. Another
approach that uses magnetism sensing has been proposed [12,21]. Hall sensors have been employed to
transduce the magnetic field to voltage/current signals, and giant magnetoresistive (GMR) sensors are
used in resistive sensing of the magnetic field. However, this type of sensing needs magnetic-particle
labeling and magnetically sensitive transducers for which the GMR sensors are not available in
standard CMOS technology. Moreover, signal amplification and conditioning are needed for the sensed
signals, which increase the power consumption of the biosensor. Capacitive biosensors are promising
as point-of-care devices due to their simple, low-cost structure, rapid detection, and portability.
Moreover, capacitive sensors do not consume static power, and can be co-integrated with CMOS
circuitry. Therefore, they are very attractive for use in battery-powered systems. Capacitive biosensors
for protein detection are proposed in [22,23], where an LCR meter has been used for capacitance
measurements. The LCR meter is a bulky, complicated, and expensive instrument. However, energy
efficient integrated capacitance-to-digital converter circuitry can digitize the capacitance measurements
with very low power consumption [24–29].
In this paper, we introduce an in vitro diagnostic biosensor-CMOS platform to screen the SSAT
enzyme for low-power, rapid detection, and low cost. The sensing element is implemented based
on capacitive sensing. It has a large sensing area with small baseline capacitance (femtofarad
range) by connecting the capacitors in series, which increases the sensitivity and reduces the power
consumption on the side of the readout circuit. The readout circuit is realized by a capacitance-domain
successive-approximation (SAR) technique. It uses an inverter-based circuit that allows supply-voltage
Sensors 2017, 17, 1942 3 of 12

scaling2017,
Sensors for17,low-power
1942 requirement. Moreover, the readout circuit uses a clocked inverter-based 3 of 12
quantizer for cascaded
Sensors 2017, 17, 1942 amplification by turning cascaded quantizers off when they are not in use.
3 of 12
are not in use. Thus, the average power of the quantizer is reduced. A binary-weighted
Thus, the average power of the quantizer is reduced. A binary-weighted capacitive digital-to-analog capacitive
digital-to-analog
converter
are not in(DAC) converter
is
use. Thus, implemented (DAC)
the average by isusing
powerimplemented
a coarse-fine
of by using
the quantizer a coarse-fine
architecture
is reduced. to cover architecture
the wide dynamic
A binary-weighted to capacitive
cover the
range
wide
of the dynamic
SSAT range
enzyme of the SSAT
concentrations enzyme
while concentrations
occupying a smallwhile occupying
chip area.
digital-to-analog converter (DAC) is implemented by using a coarse-fine architecture to cover the a small chip area.
wideThe remainder
dynamic rangeof
remainder ofofthis
thepaper
this SSATisis
paper organized
organized
enzyme asasfollows.
follows.
concentrations Section
while2occupying
Section provides
2 provides aadetailed
asmall description
detailed
chip area. of the
description of
materials
the materialsandand
The remainderthethe
capacitive
ofcapacitivesensing
this paper sensing element. asSection
element.
is organized Section
follows.3 Section
3presents
presents the
thereadout
2 providesreadout circuit architecture
architectureofand
circuitdescription
a detailed the
operation.
operation.
materials andMeasurement
Measurement
the capacitive results
resultsare
arediscussed
sensing element.inSection
discussed Section
in Section34, presents
and thethe
4, and conclusion
theconclusion
readoutis drawn in
is drawn
circuit Section 5. and
in Section
architecture 5.
operation. Measurement results are discussed in Section 4, and the conclusion is drawn in Section 5.
2. Materials and Capacitive Sensing Element
2. Materials and Capacitive Sensing Element
2.1. Capacitive
2.1. Capacitive Biosensor
Biosensor
2.1. Capacitive
The Biosensor
The fabrication
fabrication process of
process the capacitive
of the capacitive sensing
sensing element
element is is shown
shown in
in Figure
Figure 1.
1. AA glass
glass substrate
substrate
of 500
of 500 µm thickness
μmfabrication is submersed
thickness process
is submersed in a hot piranha
in a hot piranha solution, then washed with distilled water and
The of the capacitive sensingsolution,
element then washed
is shown with distilled
in Figure 1. A glasswater and
substrate
dried
dried in a nitrogen drier. Ten nanometer (10 nm) titanium (Ti) and 400 nm gold (Au) metal layers are
of 500inμma nitrogen
thicknessdrier. Ten nanometer
is submersed in a hot(10 piranha
nm) titanium (Ti) then
solution, and 400 nm gold
washed with(Au) metalwater
distilled layersand
are
deposited
deposited by sputtering
by sputtering at 5 mTorr vacuum with 400 Wdc energy. The photolithography process is
dried in a nitrogen drier.at 5 mTorr
Ten vacuum
nanometer withtitanium
(10 nm) 400 Wdc(Ti)energy. Thenm
and 400 photolithography
gold (Au) metal process is
layers are
used to
used to pattern
pattern the
the layers.
layers. Next,
Next, aa dry
dry etch
etch process utilizing
utilizing argon
argon (Ar)
(Ar) gas
gas at
at 300
300 W W plasma power
deposited by sputtering at 5 mTorr vacuumprocess
with 400 Wdc energy. The photolithography plasma power
process is
is
is used
used to etch the exposed metal to form the electrodes, then the photoresist is removed by using
used to to etch the
pattern theexposed metala to
layers. Next, dryformetchthe electrodes,
process then
utilizing the photoresist
argon is removed
(Ar) gas at 300 W plasma bypower
using
acetone and
acetone and Isopropyl
Isopropyl Alcohol
Alcohol (IPA).
(IPA).
is used to etch the exposed metal to form the electrodes, then the photoresist is removed by using
acetone and Isopropyl Alcohol (IPA).

(a) (b) (c)

(a) (b) (c)

(d) (e) (f)

(d)
Figure 1. Fabrication of capacitive sensing element:(e)(a) Clean glass substrate; (b) Sputtering
(f) of 10 nm
Figure 1. Fabrication of capacitive sensing element: (a) Clean glass substrate; (b) Sputtering of 10 nm
Ti
Ti and
Figure 400 nm Au; (c)
1. Fabrication
and 400 nm Au; (c)of Photoresist coating;
capacitive coating;
Photoresist (d)
sensing (d) Patterning
element: of
(a) Clean
Patterning photoresist; (e)
glass substrate;
of photoresist; Photoresist develop,
(b) Sputtering
(e) Photoresist of 10 dry
develop, nm
dry
etching,
etching, and
Ti and 400 photoresist
andnm
photoresist removal
Au; (c) Photoresist(f) Dielectric
coating;
removal (f) deposition
(d) deposition
Dielectric S i3N4 (50 nm)/Parylene C (300 nm).
Patterning Sof Nphotoresist; (e) Photoresist
(50 nm)/Parylene C (300develop,
nm). dry
i3 4
etching, and photoresist removal (f) Dielectric deposition Si3N4 (50 nm)/Parylene C (300 nm).

Figure 2. The sensor micrograph.

Figure 2. The sensor micrograph.
The capacitive sensor is a 9-segment array, as shown in Figure 2f. Each segment was realized
with The
interdigitated
capacitiveelectrodes
sensor is afabricated
9-segment onarray,
a glass
assubstrate.
shown inThe thickness
Figure of the
2f. Each metalwas
segment finger is 410
realized
nm
withwith 50 μm ofelectrodes
interdigitated inner spacing between
fabricated on athe fingers.
glass TheThe
substrate. dielectric
thicknesslayer is then
of the metaldeposited by
finger is 410
chemical vapor deposition (CVD) to form a 50 nm silicon nitride layer. After that,
nm with 50 μm of inner spacing between the fingers. The dielectric layer is then deposited by the capacitive
sensor is coated
chemical vapor with a 300 nm
deposition thinto
(CVD) film of Parylene
form a 50 nm C to enhance
silicon nitridethe isolation
layer. Afterandthat,tothe
keep it in the
capacitive
dynamic
sensor is range
coatedofwith
the readout
a 300 nmcircuit afterof
thin film depositing
Parylene the
C toanalyte.
enhanceThe
themeasured
isolation baseline
and to keepcapacitance
it in the
dynamic range of the readout circuit after depositing the analyte. The measured baseline capacitance
Sensors 2017, 17, 1942 4 of 12

The capacitive sensor is a 9-segment array, as shown in Figure 2f. Each segment was realized with
interdigitated electrodes fabricated on a glass substrate. The thickness of the metal finger is 410 nm
with 50 µm of inner spacing between the fingers. The dielectric layer is then deposited by chemical
vapor deposition (CVD) to form a 50 nm silicon nitride layer. After that, the capacitive sensor is coated
Sensorsa2017,
with 300 17,
nm thin film of Parylene C to enhance the isolation and to keep it in the dynamic range
1942 4 of 12
of the readout circuit after depositing the analyte. The measured baseline capacitance of the sensor,
of the sensor,
which which
has a 2.92 mm has
× 3 amm
2.92area
mmto×cover
3 mmthearea to cover
droplet thebiology
of the dropletsample,
of the biology
is 26 fF. sample, is 26 fF.
The micrograph
The micrograph of the sensor
of the sensor is shown in Figure 2. is shown in Figure 2.

2.2. Sample
Sample Sites
The following
followingmaterials
materialswere
were used
used forfor surface
surface modification
modification ofsensor
of the the sensor and experiments.
and experiments. SSAT
SSAT antibody
antibody was purchased
was purchased from
from Santa Santa
Cruz Cruz Biotechnology,
Biotechnology, Inc.TX, (Dallas,
Inc. (Dallas, TX,Spermidine-
USA) [30]. USA) [30].
Spermidine-spermine+N1-Acetyltransferase+1+SAT1
spermine+N1-Acetyltransferase+1+SAT1 was purchased wasfrom
purchased
Insightfrom Insight Biotechnology.
Biotechnology. Tween-20,
Tween-20, glutaraldehyde, and bovine serum albumin (BSA) were purchased
glutaraldehyde, and bovine serum albumin (BSA) were purchased from Sigma. from Sigma.

2.3. Surface
2.3. Surface Modification
Modification and
and Antibody
Antibody immobilization
immobilization
The detailed
The detailed procedure
procedure ofof the
the antibodies
antibodies immobilization
immobilization on on Parylene
Parylene CC isis shown
shown in in Figure
Figure 3.
3.
The cleaned surfaces of the capacitive sensors were incubated in 5% glutaraldehyde
The cleaned surfaces of the capacitive sensors were incubated in 5% glutaraldehyde solution (cross- solution
(cross-linking
linking agent)
agent) in 1× PBS 1× PBS (phosphate
in (phosphate buffer
buffer saline) saline)(pH
solution solution (pHkept
7.4) and 7.4)for
and2 hkept
in anfor 2 hargon
inert in an
inert argon atmosphere inside an enclosed chamber. Argon gas aids in absorbing
atmosphere inside an enclosed chamber. Argon gas aids in absorbing active agents on the treated active agents on the
treated and
surface surface and covalently
covalently couples couples the
the active active
agent to agent to theofsurface
the surface of the substrate
the substrate [31]. The [31].
mostThe most
common
common group present on the exterior of a protein is lysine with amino groups. Glutaraldehyde
group present on the exterior of a protein is lysine with amino groups. Glutaraldehyde helps in the helps
in the crosslinking
crosslinking of proteins
of proteins to proteins
to proteins or proteins
or proteins to a surface-immobilized
to a surface-immobilized substratesubstrate
due to thedue to the
presence
presence
of its twoof its two
amine amine functional
functional groups [32].
groups [32].

(a) (b) (c)

(d) (e) (f)

Figure
Figure 3.3.Procedure
Procedureof of antibodies
antibodies immobilization:
immobilization: (a) Coated
(a) Coated with 10with 10 μL glutaraldehyde
µL glutaraldehyde and
and incubated
incubated with Argon Gas for 2 h; (b) Coated with 2 μL antibody for 1 h; (c) Immobilized
with Argon Gas for 2 h; (b) Coated with 2 µL antibody for 1 h; (c) Immobilized antibodies; (d) 2 µL of antibodies;
(d)
2% 2BSA
μL of 2%added
was BSA was
and added and for
incubated incubated
30 min;for
(e) 30 min; antigen
Added (e) Added(10,antigen (10, 5,
5, 2.5, 1.25 2.5, 1.25
mg/L); (f) mg/L); (f)
Antibody
Antibody and antigen
and antigen interaction.interaction.

After 2 h of incubation, the surfaces were thoroughly washed with PBS Tween-20/deionized (DI)
After 2 h of incubation, the surfaces were thoroughly washed with PBS Tween-20/deionized
water. A drop-cast method was used to apply antibody (2 μL) on the surfaces with a fine micropipette
(DI) water. A drop-cast method was used to apply antibody (2 µL) on the surfaces with a fine
followed by incubation for 1 h at room temperature. The surfaces of the sensors were washed again
micropipette followed by incubation for 1 h at room temperature. The surfaces of the sensors were
in a similar manner and immersed in a 2% BSA solution as a blocking reaction for 30 min at room
washed again in a similar manner and immersed in a 2% BSA solution as a blocking reaction for
temperature to avoid any non-specific binding. After blocking, surfaces were again washed with the
30 min at room temperature to avoid any non-specific binding. After blocking, surfaces were again
PBS Tween-20 buffer, which removes most of the unbound antibodies from the surfaces.
washed with the PBS Tween-20 buffer, which removes most of the unbound antibodies from the
Subsequently, the samples were incubated in the dark with a 2 μL solution of the SSAT enzyme. Both
antibody and antigen conjugation were detected by the change in capacitance. The difference in
capacitance corresponds to the concentration of the SSAT added after immobilization of the antibody.
The capacitive biosensor can be reused with extensive cleaning and recalibration of the sensors.
However, they get attached to the readout circuit via a simple connector and can be made disposable
due to the reduced cost of manufacturing them [22–34], while the electronic readout system can be
Sensors 2017, 17, 1942 5 of 12

surfaces. Subsequently, the samples were incubated in the dark with a 2 µL solution of the SSAT
enzyme. Both antibody and antigen conjugation were detected by the change in capacitance. The
difference in capacitance corresponds to the concentration of the SSAT added after immobilization of
the antibody. The capacitive biosensor can be reused with extensive cleaning and recalibration of the
sensors. However, they get attached to the readout circuit via a simple connector and can be made
disposable due to the reduced cost of manufacturing them [22–34], while the electronic readout system
can be reused over and over. Moreover, the biosensors are designed to give quick readouts of the
Sensors
sample. 2017, 17, 1942 once the antibodies are immobilized on the surfaces, the stability is not of concern,
Therefore, 5 of 12
as the antibodies do not degrade within a few hours. However, if the sensors have to be used for
detection, then they
onsite detection, thenhave
theytohave
be shipped using Peltier
to be shipped using coolers. AnotherAnother
Peltier coolers. approach is to eliminate
approach certain
is to eliminate
factors that decrease
certain factors the stability
that decrease of biosensors
the stability [35].[35].
of biosensors Factors that
Factors affect
that the
affect thelong-term
long-termstability
stability of
of
biosensors include (1) fouling and contamination of the samples; (2) the lifetime of
biosensors include (1) fouling and contamination of the samples; (2) the lifetime of the immobilizedthe immobilized
enzymes;
enzymes; and and (3) strength of
(3) strength of immobilization.
immobilization.

3.
3. Successive
Successive Approximation
Approximation Readout
Readout Circuit
Circuit
There
There are
are several
severaltechniques
techniquesfor forimplementing
implementingthe thereadout
readoutcircuit
circuitforfor
capacitive
capacitivesensors
sensorssuch as
such
incremental sigma-delta, period modulation, successive approximation, and
as incremental sigma-delta, period modulation, successive approximation, and dual slope [25–29]. dual slope [25–29]. The
successive
The successiveapproximation
approximation readout
readoutcircuit is adopted
circuit is adopted to tointerface
interfacethe thefabricated
fabricatedbiosensor,
biosensor, as as it
it
achieves
achieves better energy efficiency.
better energy efficiency.
The schematicofof
The schematic thethe readout
readout circuit
circuit is shown
is shown in Figure in4.Figure 4. It
It is based onis based on a successive-
a successive-approximation
approximation
data converter thatdatauses
converter
chargethat uses charge to
redistribution redistribution to produce
produce a digital word.aItdigital word.
consists It consists
of the unknown of
the unknown
capacitive capacitive
sensor (CSENSsensor (CSENS), a binary-weighted
), a binary-weighted DAC (CREFDAC ), an (C REF), an amplifier
amplifier with feedback
with feedback loop, a
loop, a single-ended
single-ended comparator,
comparator, andSAR
and control control SAR
logic. Thelogic. The single-ended
single-ended comparator comparator
consists ofconsists of two
two cascaded
cascaded inverters
inverters with clockedwith
latchclocked
based onlatch
NAND based
gates.onThe NAND gates.
amplifier andThe amplifier
comparator and comparator
(cascaded inverters)
(cascaded inverters) are implemented by using a clocked inverter. The circuit
are implemented by using a clocked inverter. The circuit operation is divided into two phases: samplingoperation is divided
into two phases: sampling phase and conversion phase. Two standard non-overlapping
phase and conversion phase. Two standard non-overlapping clock phases, P1 and P2, are required clock phases,
P1 andwith
along P2, are required
a SAR clock. along with a SAR clock.

P1

CP SAR
VOUT VCMP
Control
CSENS ... ...
20Cunit 2n-1Cunit Logic
P2 P1 CREF (DAC) P1 P2

VREF VREF VREF P1Clk

P2

P1 P2 P1 P2
VCMP

P1 Sampling Sampling

P2 Conversion

Clk φ0 φ1 φ2
... ... φ12 φ0 φ1

VCMP
... ...
Figure 4. Readout circuit with associated timing diagram.
Figure 4. Readout circuit with associated timing diagram.

To illustrate the readout circuit operation, Figure 5a shows the sampling phase of the readout
circuit. When clock phase P1 is high, the amplifier is configured as a unity-gain amplifier by turning
on the feedback switch. Thus, it employs the switching threshold voltage (VM1) to charge the binary-
weighted DAC, parasitic capacitance ( ), and the capacitive sensor (CSENS). The switches of the
binary-weighted DAC are implemented by inverters. They are used to drive the bottom plates of the
Sensors 2017, 17, 1942 6 of 12

To illustrate the readout circuit operation, Figure 5a shows the sampling phase of the readout
circuit. When clock phase P1 is high, the amplifier is configured as a unity-gain amplifier by turning
on the feedback switch. Thus, it employs the switching threshold voltage (VM1 ) to charge the
binary-weighted DAC, parasitic capacitance (CP ), and the capacitive sensor (CSENS ). The switches of
the binary-weighted DAC are implemented by inverters. They are used to drive the bottom plates
of the binary-weighted DAC to ground (GND) while the bottom plate of the sensing capacitor is
connected to reference voltage (VREF ). The single-ended comparator for cascaded amplification is
Sensors
turned2017, 17, 1942 this phase to reduce the average power consumption.
off during 6 of 12

Figure 5. (a) Schematic and operation of sampling phase; (b) Schematic and operation of
Figure 5. (a) Schematic and operation of sampling phase; (b) Schematic and operation of conversion phase.
conversion phase.

The second step is the conversion phase of the readout circuit, as shown in Figure 5b. In this
phase,Thethesecond
clock step
P1 isislow
the and
conversion phase switch
the feedback of the readout
is turnedcircuit, as shown
off. Thus, in Figureamplifier
an open-loop 5b. In thisis
phase,
formed.the clock P1 the
Moreover, is low and plates
bottom the feedback switch is turned DAC
of the binary-weighted off. Thus, an open-loop
are either connectedamplifier
to VREF oris
formed.depending
GND, Moreover, theon bottom
the N-bitplates
SARof the binary-weighted
logic output. The DAC are either
equivalent connectedof
capacitance to V
the or GND,
REFelements
depending on the N-bit SAR logic output. The equivalent capacitance of the elements
connected to VREF and GND are CREF,ON and CREF,OFF, respectively, while the bottom plate of the sensing connected to
V REF and GND
capacitor are CREF,ON
is connected to GND. CREF,OFF
and The , respectively,
single-ended while the
comparator forbottom
cascadedplate of the sensing
amplification capacitor
is turned on
is connected to GND. The single-ended comparator for cascaded amplification
during this phase. When the law of charge conservation is applied, the output voltage of the amplifier is turned on during
this be
can phase. When as
expressed the law of charge conservation is applied, the output voltage of the amplifier can be
expressed as
( − , )
VOUT = V = ++ ΔVo==V ++ (CSENS − CREF,ON ) VREF A
∆Vo (1)
(1)
M1 M1
CT
where A isisthe
where thefinite
finitegain
gainofofthe
theinverter,
inverter,∆V is the
ΔVOO is the change
change in in the
the output
output voltage,
voltage, and
and CT isis(C(CSENS
SENS ++
C + C ). Thus, the output voltage from the comparator
CREF + CP ). Thus, the output voltage from the comparator is given by
REF P is given by

=( 0, ΔVo < 0
(2)
1, ∆Vo><00
0, ΔVo
VCMP = (2)
The successive-approximation algorithm keeps 1, ∆Vo >0
changing until it matches with
,
an error less than 1 least-significant bit (LSB). The output voltage of the comparator is determined by
The successive-approximation algorithm keeps changing C until it matches CSENS with
the sign of (CSENS − CREF,ON), regardless of its absolute value (ΔVOREF,ON
). Thus, the circuit is insensitive to
an error less than 1 least-significant bit (LSB). The output voltage of the comparator is determined by
VREF and CP variation. When the mismatch between the amplifier and the comparator is considered
the sign of (CSENS − CREF,ON ), regardless of its absolute value (∆VO ). Thus, the circuit is insensitive to
in the analysis, the input voltage to the comparator is given by
VREF and CP variation. When the mismatch between the amplifier and the comparator is considered in
the analysis, the input voltage ( is given
− by, )
= to the−comparator
= + (3)
(CSENS − CREF,ON ) VREF
where is the switching VOUT − Vvoltage
VI N = threshold M2 = of the comparator, and A +isVOS
the offset voltage (VM1(3)
−
CT
VM2). Thus, the minimum achievable resolution is given by
where VM2 is the switching threshold voltage of the comparator, and VOS is the offset voltage
=
(VM1 − VM2 ). Thus, the minimum achievable resolution is given by (4)

VOS selected to limit the conversion error. For the

The LSB, gain, and full-scale range are therefore
LSB = CT (4)
A Voffset
implemented readout circuit, the gain is 98, and the REF voltage is 3.3 mV.
The binary-weighted capacitive DAC is used to compare the digital output with the capacitive
input. The least-significant bit (LSB) capacitor is selected based on matching properties of the
capacitors. Fringing capacitance between metal wires on the same layer is used to implement the fine
binary-weighted DAC. Therefore, an 8-bit fine DAC with absolute resolution 4.5 fF is implemented.
The total fine DAC capacitance (composed of 255 units) is 1.147 pF. The metal-insulator-metal (MIM)
capacitor is used to implement the unit capacitor of the coarse binary-weighted DAC. It is chosen
such that the DNL of the coarse DAC is less than 1/2 LSB of the fine DAC and the full scale of the fine
Sensors 2017, 17, 1942 7 of 12

The LSB, gain, and full-scale range are therefore selected to limit the conversion error. For the
implemented readout circuit, the gain is 98, and the offset voltage is 3.3 mV.
The binary-weighted capacitive DAC is used to compare the digital output with the capacitive
input. The least-significant bit (LSB) capacitor is selected based on matching properties of the
capacitors. Fringing capacitance between metal wires on the same layer is used to implement the fine
binary-weighted DAC. Therefore, an 8-bit fine DAC with absolute resolution 4.5 fF is implemented.
The total fine DAC capacitance (composed of 255 units) is 1.147 pF. The metal-insulator-metal (MIM)
capacitor is used to implement the unit capacitor of the coarse binary-weighted DAC. It is chosen such
that the DNL of the coarse DAC is less than 1/2 LSB of the fine DAC and the full scale of the fine
DAC covers the unit capacitor of the coarse DAC. Therefore, a 4-bit coarse DAC with unit capacitor 1
pF is implemented and combined with fine DAC, which results in a 16.137 pF capacitance range in a
Sensors 2017,
compact 17, 1942
area. 7 of 12

4. Measurement Results
Sensors 2017, 17, 1942 7 of 12

The prototype readout circuit was implemented in a standard 0.18 µm μm CMOS process. Figure 6
4. Measurement Results
photograph of the
shows the chip photograph the fabricated
fabricated prototype.
prototype. The active area is 0.06 mm22 including the
0.06 mm
The prototype
binary-weighted DAC,readout
analogcircuit was implemented
circuitry, in a standard
and SAR control 0.18 μm
logic blocks. A CMOS
registerprocess. Figureto6 read
file is used
shows the chip photograph of
signals through the fabricated
through a 4-wire prototype.
4-wire serial The active area
serial peripheral-interface is 0.06 mm 2 including the
peripheral-interface (SPI) bus. The interface bus
and write all digital signals
binary-weighted DAC, analog circuitry, and SAR control logic blocks. A register file is used to read
minimizes the bond pads and package pins. A printed circuit board was designed and fabricated to
and write all digital signals through a 4-wire serial peripheral-interface (SPI) bus. The interface bus
test the
thechip
chipand
and the
the sensor.
sensor. The chip was mounted on the board
circuit board and connected to the
minimizes the bond padsThe
andchip was pins.
package mounted on the
A printed circuit
circuit board wasand connected
designed to the capacitive
and fabricated to
capacitive
biosensor, biosensor,
as
test the shown
chip andinas shown
Figure
the 7.inThe
sensor. Figure
chip7.was mounted on the circuit board and connected to the
capacitive biosensor, as shown in Figure 7.

Figure 6. Chip
Figure photo
6. Chip ofofthe
photo theprototype
prototype readout circuitinin0.18
readout circuit 0.18
μmμm CMOS
CMOS
µm process.
process.
process.

Figure
Figure 7. Test
7. Test boardfor
board fortesting
testing the
the packaged
packagedchip
chipand
andthethe
biosensor.
biosensor.

An on-chipFigure
dummy 7. capacitive
Test boardsensor is used
for testing the to test the performance
packaged and the dynamic range of
chip and the biosensor.
the readout circuit. The measured transfer function of the readout circuit is shown in Figure 8. The
curve
An shows dummy
on-chip a linear relation between
capacitive sensortheisoutput
used to codes
test and
the input capacitances
performance and when the dummy
the dynamic range of
capacitive sensor varies between 0 and 16.137 pF in 4.5 fF steps. The circuit is quantization-noise
the readout circuit. The measured transfer function of the readout circuit is shown in Figure 8. The
limited, and the noise is less than 1 LSB: i.e., it is limited by the smallest unit capacitor, as shown in
Sensors 2017, 17, 1942 8 of 12

An on-chip dummy capacitive sensor is used to test the performance and the dynamic range of the
readout circuit. The measured transfer function of the readout circuit is shown in Figure 8. The curve
shows a linear relation between the output codes and input capacitances when the dummy capacitive
sensor varies between 0 and 16.137 pF in 4.5 fF steps. The circuit is quantization-noise limited, and the
noise is less than 1 LSB: i.e., it is limited by the smallest unit capacitor, as shown in Figure 9. It shows
the standard
Sensors
Sensors 2017, deviation
2017,
17, 17,
19421942 of the measured readout-circuit output as a function of sensor capacitance.
8 of8 12
of 12

Figure
Figure
Figure 8. Measured
8.8. Measured
Measured transfer
transfer
transfer function
function of readout
of of
function readout circuit.
circuit.
readout circuit.

Figure 9. Measured noise output as a function of capacitance.

Figure
Figure 9. Measured
9. Measured noise
noise output
output as as a function
a function of capacitance.
of capacitance.

Acetylation
Acetylation is is
Acetylation performed
is performed
performed bybybythe
thetheenzyme
enzyme
enzyme SSAT,
SSAT,
SSAT, which
which
which hashas
has
beenbeen
been proven
proven
proven to tobeto
be be present
present
present in in in elevated
elevated
elevated
levels in many
levels
levels in in
many cancers
many cancers
cancers including
including
including lung
lunglung [36],
[36],
[36], breastbreast
breast [37],
[37],
[37], prostate,
prostate,
prostate, melanoma,
melanoma,
melanoma, and and and gastrointestinal
gastrointestinal
gastrointestinal tract
tract
tractcancers
cancers
cancers [38].
[38].
[38]. TheThe The
enzyme enzyme
enzyme is is
alwaysis always
always present
present present
in inallall in all mammalian
mammalian
mammalian cells;
cells; thethe cells; theisdifference
difference
difference is
that that is that
cancer
cancer
patients
patients have
have a a higher
higher proportion
proportion of ofthethe enzyme
enzyme than
than
cancer patients have a higher proportion of the enzyme than healthy subjects. Therefore, various healthy
healthy subjects.
subjects. Therefore,
Therefore, various
various
concentrations
concentrations
concentrations of oftheof the antigen
theantigen
antigen from
from from 1010
10 toto
to 1.251.25
1.25 mg/Lmg/L
mg/L havehave
havebeenbeen screened.
screened.
been The
screened. The change
change
The in in
change capacitance
capacitance of of
in capacitance
of the sensor is shown in Figure 10. The baseline capacitance of the sensors was aboutthe
thethe sensor
sensor is is shown
shown in in Figure
Figure 10. 10.
The The baseline
baseline capacitance
capacitance of of
the the sensors
sensors was was about
about 26 26
fF fF while,
while, the 26 fF
capacitance
capacitance change
change was was substantial
substantial (about
(about 300300fF)fF)
uponupon adding
adding antibodies
antibodies and and after
after immobilization.
immobilization.
while, the capacitance change was substantial (about 300 fF) upon adding antibodies and after
Further,
Further, upon
upon adding
adding BSA,
BSA, thethe capacitance
capacitance was was unchanged
unchanged or or
thethe change
change was was negligible,
negligible, showing
showing
immobilization. Further, upon adding BSA, the capacitance was unchanged or the change was
thethe specificity
specificity of ofthethe immobilized
immobilized antibodies.
antibodies. Figure
Figure 11 11 reports
reports thethe sensor
sensor response
response to tothethe different
different
negligible, showingofthe
concentrations
concentrations ofthespecificity
theSSAT
SSATenzyme of the immobilized
enzyme withwithrespect
respect antibodies. Figure
to tothethebaseline
baseline 11 reportsafter
capacitance
capacitance the sensor
after response
antibody
antibody
to the different
immobilization.
immobilization. concentrations
The The measurements of
measurements the SSAT
showshow enzyme
thatthatthethe with respect
capacitance
capacitance of ofto
thethethe baseline
sensor
sensor capacitance
increases
increases withwith thetheafter
antibody
decreaseimmobilization.
decrease in in concentration
concentration The measurements
of ofthethe enzyme.
enzyme. show
The The that
error
error the capacitance
bars
bars showshow thethe of the sensor
variation
variation in in increases
thethe change
change with
of of the
decrease in concentration
capacitance
capacitance forfordifferent of the
different enzyme.
samples.
samples. The The
The mean error
mean and bars
and show
thethe the variation
standard
standard deviation
deviation inare
the
are change
shown
shown offor
for capacitance
every
every
concentration.
concentration.
for different samples.The Theplatform
platform
The mean wakeswakes
andup upintermittently
the intermittently
standard every20 20
every
deviation sshown
s to
are toperform
perform
forand anddigitize
every digitize thethe
concentration.
The platform wakes up intermittently every 20 s to perform and digitize the measurementThe
measurement
measurement of of
the the biosensor
biosensor and and thenthen returns
returns to to sleep
sleep mode
mode during
during the the testing
testing of of
the the sample.
sample. The of the
sensor shows thethe response
sensor
biosensor shows
and then response
returns toand
and sleep stabilization
stabilization
mode within within
during 1 the1 minute,
minute, which
which
testing indicates
of indicates
the sample. a high
a high sensitivity
sensitivity
The sensortoshowsto
SSAT
SSAT enzyme
enzyme concentrations.
concentrations. The The average
average change
change in in capacitance
capacitance forfor
thethe highest
highest and and thethe lowest
lowest
the response and stabilization within 1 min, which indicates a high sensitivity to SSAT enzyme
concentrationis is5 5folds.
concentration folds.Figure
Figure12 12shows showsthetheratio ratioof ofincrease
increasein inthethecapacitance
capacitance(CAntigen (CAntigen− −
concentrations. The average change in capacitance for the highest and the lowest concentration is
CAntibody/CAntibody) with concentration. The ratio of capacitance increase was around 35 times for the
CAntibody/CAntibody) with concentration. The ratio of capacitance increase was around 35 times for the
concentration
concentration 1.25
1.25 mg/L
mg/L and
and decreased
decreased to to 6 times
6 times forfor
thethe concentration
concentration 10 10 mg/L.
mg/L. This
This clearly
clearly indicates
indicates
that
that a wide
a wide range
range of of enzyme
enzyme concentrations
concentrations could
could bebe quantified
quantified using
using thethe diagnostic
diagnostic platform.
platform. The
The
Sensors 2017, 17, 1942 9 of 12

5 folds. Figure 12 shows the ratio of increase in the capacitance (CAntigen − CAntibody /CAntibody ) with
concentration. The ratio of capacitance increase was around 35 times for the concentration 1.25 mg/L
and decreased to 6 times for the concentration 10 mg/L. This clearly indicates that a wide range of
enzymeSensors
concentrations
Sensors 2017, 17, 1942 could be quantified using the diagnostic platform. The platform can
2017, 17, 1942
quantify
9 of 12
9 of 12
and digitize the given sample as low as 1.25 mg/L to 10 mg/L in its dynamic range with sufficient
platform can
platform can quantify
quantify and
and digitize the
the given
given sample
sample as low as 1.25 mg/L to 10 mg/L in its dynamic
margin for any offset or sensor digitize
capacitance variation. as low as 1.25 mg/L to 10 mg/L in its dynamic
range with sufficient margin for any offset
range with sufficient margin for any offset or sensor or sensor capacitance
capacitance variation.
variation.
Table 1 lists
Table the
1 performance
lists the of the of
performance readout
the circuit.
readout It shows
circuit. It showsthe measurement
the measurement
measurement time
time andandthetheaverage
Table 1 lists the performance of the readout circuit. It shows the time and the
power for analog
average and digital
power for analogcircuits.
analog At full-scale
and digital
digital circuits. input capacitance,
At full-scale
full-scale the total power
input capacitance,
capacitance, consumption
the total
total power of
average power for and circuits. At input the power
the readout circuit
consumption isof2.1
the µW, of
readout which
circuit the
is analog
2.1 μW, ofrepresents
which the 85.7%
analog and the
represents
consumption of the readout circuit is 2.1 μW, of which the analog represents 85.7% and the digital digital
85.7% represents
and the digital 14.3%.
represents
Table 14.3%.
2 summarizes
represents 14.3%. performance comparison with previously published Biosensor-CMOS
Table 22 summarizes
summarizes performance
performance comparison
comparison with with previously
previously published
Table
platforms. The introduced Biosensor-CMOS platform shows the ability published Biosensor-CMOS
to screen Biosensor-CMOS
and quantify the SSAT
platforms. The
platforms. The introduced
introduced Biosensor-CMOS
Biosensor-CMOS platformplatform shows
shows the the ability
ability to
to screen
screen and
and quantify
quantify thethe
enzyme SSAT
on theenzyme
surface on
of the
the
capacitive
surface of the
sensor.
capacitive
It is based
sensor. It is
on
based
capacitive
on
sensing,
capacitive
which
sensing, which
is is
low-cost
low-
and
SSAT enzyme on the surface of the capacitive sensor. It is based on capacitive sensing, which is low-
label-free. Moreover,
cost and
and label-free. the platform
label-free. Moreover,
Moreover, the achieves
the platform an
platform achieves ultra-low
achieves an power
an ultra-low
ultra-low power consumption
power consumption
consumption of of 2.1
of 2.1
2.1 μWµW
μW among among
among the
cost
publishedtheplatforms,
the published
which iswhich
published platforms,
platforms, fourisisorders
which of magnitude
four orders
four orders of magnitude
of magnitude lesslesspower
less
than
power than
power
thestate-of-the-art.
than the
the state-of-the-art.
state-of-the-art.

500
500
400
400
(fF)
Capacitance(fF)

300
300
Capacitance

200
200
100
100

00
Baseline
Baseline Antibody SSAT
Antibody SSAT BSA
BSA

Figure
FigureFigure 10. Measured
10. Measured
10. Measured baseline,
baseline, antibody, and
antibody,
baseline, antibody, and BSA
and BSA capacitances
BSAcapacitances of the
capacitances
of the sensor.
ofsensor.
the sensor.

Figure 11.
Figure
Figure Sensor
11. response
11. Sensor
Sensor to different
response
response differentconcentrations
to different
to ofantigens:
concentrations of
concentrations of antigens:
antigens: (a) (a)
(a) 10 mg/L;
10 mg/L;
10 mg/L; (b) 55 (b)
(b) 5 mg/L;
mg/L;
mg/L;
(c)
(c) 2.5 mg/L;2.5 mg/L; (d)
(d) 1.25
(c) 2.5 mg/L; 1.25
(d) mg/L.mg/L.
1.25 mg/L.
Sensors 2017, 17, 1942 10 of 12
Sensors 2017, 17, 1942 10 of 12

Figure 12. The ratio of increase in capacitance with various concentration of the SSAT.
Figure 12. The ratio of increase in capacitance with various concentration of the SSAT.
Table 1. Readout circuit performance summary.
Table 1. Readout circuit performance summary.
Technology CMOS 180 nm
Power Supply (Analog/Digital) 0.9 V/1 V
TechnologyAbsolute Resolution 4.5 fF CMOS 180 nm
Capacitance Range
Power Supply (Analog/Digital) 16.137 pF 0.9 V/1 V
Noise (rms)
Absolute Resolution <1.4 fF 4.5 fF
Capacitance Range (Analog)
Power 1.8 μW 16.137 pF
Noise (rms) Power (Digital) 0.3 μW <1.4 fF
Power (Analog) Resolution 12 Bits 1.8 µW
Measurement Time
Power (Digital) 42.5 μs 0.3 µW
Resolution 12 Bits
Table 2. Comparison to some of the pre-existing biosensor-CMOS platforms.
Measurement Time 42.5 µs
Reference BIOCAS 2010 [11] JSSC 2013 [12] VLSI 2015 [13] JSSC 2013 [21] This Work
Target Protein Protein Protein Protein Protein
Table 2. Comparison
Technology (nm) 350
to some of the 180
pre-existing 65biosensor-CMOS 180
platforms.180
Sensing
Reference Parameter ImpedanceJSSC 2013Magnetism
BIOCAS 2010 [11] [12] Fluorescence
VLSI 2015 [13] Magnetism Capacitive This Work
JSSC 2013 [21]
Target Protein Human serum
Protein Protein SLPI cancer
Protein SSAT Protein
Target Protein Protein G Streptavidin
Technology (nm) 350 180 albumin 65 marker 180 enzyme 180
Au
TransducerImpedanceAu electrodesMagnetism
Sensing Parameter Hall sensor Fluorescence
Photodetector Magnetism
GMR sensor Capacitive
electrodes
Target Protein Protein G Human serum albumin Streptavidin SLPI cancer marker SSAT enzyme
Magnetic Qdot 800 Magnetic
Transducer LabelingAu electrodesLabel-free Hall sensorparticle Photodetector
fluorophore GMR sensor
particle
Label-freeAu electrodes

LabelingImmobilization
Label-free Yes Magnetic particle
Yes Qdot 800Yes
fluorophore Yes
Magnetic particleYes Label-free
Immobilization Power Yes 84.8 mW Yes 300 mW 66 Yes
mW 50.4 mW Yes 2.1 μW Yes
Power 84.8 mW 300 mW 66 mW 50.4 mW 2.1 µW
5. Conclusions
An in vitro diagnostic, biosensor-CMOS platform was developed to screen and quantify SSAT
5. Conclusions
enzyme for rapid detection that is low power and inexpensive. It will benefit early detection of cancer,
An inwhich
vitrodramatically
diagnostic,improves the survival rate
biosensor-CMOS of cancerwas
platform patients. A large sensing
developed to screenarea and
with quantify
small SSAT
baseline capacitance was introduced to implement a capacitive biosensor that shows a high
enzyme for rapid detection that is low power and inexpensive. It will benefit early detection of
sensitivity to SSAT enzyme concentrations. A successive readout circuit incorporating a clocked
cancer, which dramatically
inverter-based improveswas
SAR architecture the survivaltorate
introduced read of
andcancer
processpatients. A large
the information from sensing
the area
with smallbiosensor.
baseline By capacitance
using a clockedwas introduced
digital to implement
inverter as a quantizer, a capacitive
the readout biosensor
circuit is able that shows
to achieve an
ultra-low power
a high sensitivity to SSATconsumption.
enzyme The diagnostic platform
concentrations. can be miniaturized
A successive readout into a complete
circuit incorporating a
clocked inverter-based SAR architecture was introduced to read and process the information from
the biosensor. By using a clocked digital inverter as a quantizer, the readout circuit is able to achieve
an ultra-low power consumption. The diagnostic platform can be miniaturized into a complete
microsystem in a compact size by integrating the sensing electrode and the associated readout circuit
in a single tiny package. Therefore, the miniaturized device can be produced in low-cost, small size,
and mass-production technology.

Acknowledgments: The research reported in this publication was supported by funding from King Abdullah
University of Science and Technology (KAUST).
Sensors 2017, 17, 1942 11 of 12

Author Contributions: Abdulaziz Alhoshany designed and developed the platform, performed the experiments
and testing, drafted and revised the manuscript. Shilpa Sivashankar designed experiments, drafted and revised
manuscript. Yousof Mashraei assisted in the fabrication of the sensor, and Hesham Omran assisted in the tapeout
of the readout circuit. Khaled N. Salama directed and supervised the research.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Okeke, I.N. Diagnostic insufficiency in Africa. Clin. Infect. Dis. 2006, 42, 1501–1503. [CrossRef] [PubMed]
2. Serkova, N.J.; Glunde, K. Metabolomics of cancer. In Tumor Biomarker Discovery: Methods and Protocols;
Tainsky, M.A., Ed.; Humana Press: Totowa, NJ, USA, 2009; pp. 273–295.
3. Sitar, D.S.; Bras, A.P.M. Method for Assaying Non-Spermine/Spermidine Activity of Spermidine/Spermine
N1 -acetyltransferase (SSAT). U.S. Patent 6811967 B2, 2 November 2004.
4. Vo-Dinh, T.; Allain, L.R.; Stokes, D.L. Cancer gene detection using surface-enhanced Raman scattering
(SERS). J. Raman Spectrosc. 2002, 33, 511–516. [CrossRef]
5. Yu, W.W.; White, I.M. A simple filter-based approach to surface enhanced Raman spectroscopy for trace
chemical detection. Analyst 2012, 137, 1168–1173. [CrossRef] [PubMed]
6. Cao, G.; Hajisalem, G.; Li, W.; Hof, F.; Gordon, R. Quantification of an exogenous cancer biomarker in
urinalysis by Raman Spectroscopy. Analyst 2014, 139, 5375–5378. [CrossRef] [PubMed]
7. Soda, K. The mechanisms by which polyamines accelerate tumor spread. J. Exp. Clin. Cancer Res. 2011, 30,
95. [CrossRef] [PubMed]
8. Lee, K.H.; Choi, S.; Lee, J.O.; Yoon, J.B.; Cho, G.H. CMOS capacitive biosensor with enhanced sensitivity for
label-free DNA detection. In Proceedings of the 2012 IEEE International Solid-State Circuits Conference, San
Francisco, CA, USA, 19–23 February 2012.
9. Pai, A.; Khachaturian, A.; Chapman, S.; Hu, A.; Wang, H.; Hajimiri, A. A handheld magnetic sensing
platform for antigen and nucleic acid detection. Analyst 2014, 139, 1403–1411. [CrossRef] [PubMed]
10. Norian, H.; Field, R.M.; Kymissis, I.; Shepard, K.L. An integrated CMOS
quantitative-polymerase-chain-reaction lab-on-chip for point-of-care diagnostics. Lab Chip 2014, 14,
4076–4084. [CrossRef] [PubMed]
11. Manickam, A.; Chevalier, A.; McDermott, M.; Ellington, A.D.; Hassibi, A. A CMOS Electrochemical
Impedance Spectroscopy (EIS) biosensor array. IEEE Trans. Biomed. Circuits Syst. 2010, 4, 379–390. [CrossRef]
[PubMed]
12. Gambini, S.; Skucha, K.; Liu, P.P.; Kim, J.; Krigel, R. A 10 kPixel CMOS hall sensor array with baseline
suppression and parallel readout for immunoassays. IEEE J. Solid State Circuits 2013, 48, 302–317. [CrossRef]
13. Hong, L.; McManus, S.; Yang, H.; Sengupta, K. A fully integrated CMOS fluorescence biosensor with on-chip
nanophotonic filter. In Proceedings of the 2015 Symposium on VLSI Circuits (VLSI Circuits), Kyoto, Japan,
16–19 June 2015.
14. Berggren, C.; Bjarnason, B.; Johansson, G. An immunological Interleukine-6 capacitive biosensor using
perturbation with a potentiostatic step. Biosens. Bioelectron. 1998, 13, 1061–1068. [CrossRef]
15. Carrara, S.; Bhalla, V.; Stagni, C.; Benini, L.; Ferretti, A.; Valle, F.; Samorì, B. Label-free cancer markers
detection by capacitance biochip. Sens. Actuators B Chem. 2009, 136, 163–172. [CrossRef]
16. Prakash, S.B.; Abshire, P. Tracking cancer cell proliferation on a CMOS capacitance sensor chip.
Biosens. Bioelectron. 2008, 23, 1449–1457. [CrossRef] [PubMed]
17. Wang, H.; Mahdavi, A.; Tirrell, D.A.; Hajimiri, A. A magnetic cell-based sensor. Lab Chip 2012, 12, 4465–4471.
[CrossRef] [PubMed]
18. Dupont, E.P.; Labonne, E.; Maruyama, Y.; Vandevyver, C.; Lehmann, U.; Gijs, M.A.; Charbon, E. Fluorescent
magnetic bead and cell differentiation/counting using a CMOS SPAD matrix. Sens. Actuators B Chem. 2012,
174, 609–615. [CrossRef]
19. Baker, K.L.; Bolger, F.B.; Lowry, J.P. Development of a microelectrochemical biosensor for the real-time
detection of choline. Sens. Actuators B Chem. 2017, 243, 412–420. [CrossRef]
20. Oseev, A.; Schmidt, M.P.; Hirsch, S.; Brose, A.; Schmidt, B. Two-component dielectric dispersion impedance
biosensor for in-line protein monitoring. Sens. Actuators B Chem. 2017, 239, 1213–1220. [CrossRef]
Sensors 2017, 17, 1942 12 of 12

21. Hall, D.A.; Gaster, R.S.; Makinwa, K.A.A.; Wang, S.X.; Murmann, B. A 256 pixel magnetoresistive biosensor
microarray in 0.18 µm CMOS. IEEE J. Solid State Circuits 2013, 48, 1290–1301. [CrossRef] [PubMed]
22. Sivashankar, S.; Sapsanis, C.; Buttner, U.; Salama, K.N. Flexible low-cost cardiovascular risk marker biosensor
for point-of-care applications. Electron. Lett. 2015, 51, 1746–1748. [CrossRef]
23. Mashraei, Y.; Sivashankar, S.; Buttner, U.; Salama, K.N. Integration of fractal biosensor in a digital microfluidic
platform. IEEE Sens. J. 2016, 16, 8775–8783. [CrossRef]
24. Jung, W.; Jeong, S.; Oh, S.; Sylvester, D.; Blaauw, D. 27.6 A 0.7pF-to-10nF fully digital capacitance-to-digital
converter using iterative delay-chain discharge. In Proceedings of the 2015 IEEE International Solid-State
Circuits Conference—(ISSCC), San Francisco, CA, USA, 22–26 February 2015.
25. Oh, S.; Jung, W.; Yang, K.; Blaauw, D.; Sylvester, D. 15.4b Incremental Sigma-Delta Capacitance-to-Digital
Converter with Zoom-in 9b Asynchronous SAR; VLSI Circuits Digest of Technical Papers; Institute of Electrical
and Electronics Engineers Inc.: Piscataway, NJ, USA, 2014; pp. 1–2.
26. Yuming, H.; Zu-yao, C.; Pakula, L.; Shalmany, S.H.; Pertijs, M. 27.7 A 0.05 mm2 1V capacitance-to-digital
converter based on period modulation. In Proceedings of the 2015 IEEE International Solid-State Circuits
Conference—(ISSCC), San Francisco, CA, USA, 22–26 February 2015.
27. Alhoshany, A.; Omran, H.; Salama, K.N. A 45.8 fJ/Step, energy-efficient, differential SAR
capacitance-to-digital converter for capacitive pressure sensing. Sens. Actuators A Phys. 2016, 245, 10–18.
[CrossRef]
28. Omran, H.; Alhoshany, A.; Alahmadi, H.; Salama, K.N. A 33fJ/Step SAR capacitance-to-digital converter
using a chain of inverter-based amplifiers. IEEE Trans. Circuits Syst. I Regul. Papers 2017, 64, 310–321.
[CrossRef]
29. Sanjurjo, J.P.; Prefasi, E.; Buffa, C.; Gaggl, R. An energy-efficient 17-bit noise-shaping Dual-Slope
Capacitance-to-Digital Converter for MEMS sensors. In Proceedings of the ESSCIRC Conference 2016:
42nd European Solid-State Circuits Conference, Lausanne, Switzerland, 12–15 September 2016.
30. Santa Cruz Biotechnology, Inc. Available online: https://www.scbt.com (accessed on 23 August 2017).
31. Terlingen, J.G.A.; Jan, F.; Hoffman, A.S. Immobilization of surface active compounds on polymer supports
using a gas discharge process. J. Biomater. Sci. Polym. Ed. 1993, 4, 31–33. [CrossRef]
32. Kim, D.; Herr, A.E. Protein immobilization techniques for microfluidic assays. Biomicrofluidics 2013, 7.
[CrossRef] [PubMed]
33. Sapsanis, C.; Sivashankar, S.; Omran, H.; Buttner, U.; Salama, K.N. Capacitive immunosensor for C-reactive
protein quantification. In Proceedings of the 2015 IEEE 58th International Midwest Symposium on Circuits
and Systems (MWSCAS), Fort Collins, CO, USA, 2–5 August 2015.
34. Sapsanis, C.; Buttner, U.; Omran, H.; Belmabkhout, Y.; Shekhah, O.; Eddaoudi, M.; Salama, K.N. A
nafion coated capacitive humidity sensor on a flexible PET substrate. In Proceedings of the 2016
IEEE 59th International Midwest Symposium on Circuits and Systems (MWSCAS), Abu Dhabi, UAE,
16–19 October 2016.
35. Yabuki, S. How to lengthen the long-term stability of enzyme membranes: Trends and strategies. Catalysts
2017, 7, 36. [CrossRef]
36. Mank-Seymour, A.R.; Murray, T.R.; Berkey, K.A.; Xiao, L.; Kern, S.; Casero, R.A. Two active copies of the
X-linked gene spermidine/spermine N-1 acetyltransferase (SSAT) in a female lung cancer cell line are
associated with an increase in sensitivity to an antitumor polyamine analogue. Clin. Cancer Res. 1998, 4,
2003–2008. [PubMed]
37. Pledgie-Tracy, A.; Billam, M.; Hacker, A.; Sobolewski, M.D.; Woster, P.M.; Zhang, Z.; Casero, R.A.;
Davidson, N.E. The role of the polyamine catabolic enzymes SSAT and SMO in the synergistic effects
of standard chemotherapeutic agents with a polyamine analogue in human breast cancer cell lines.
Cancer Chemother. Pharmacol. 2010, 65, 1067–1081. [CrossRef] [PubMed]
38. Babbar, N.; Gerner, E.W.; Casero, R.A., Jr. Induction of spermidine/spermine N1 -acetyltransferase (SSAT) by
aspirin in Caco-2 colon cancer cells. Biochem. J. 2006, 394, 317–324. [CrossRef] [PubMed]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).