Biosensors and Bioelectronics 86 (2016) 623–629

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

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

Handheld analyzer with on-chip molecularly-imprinted biosensors for

electrical detection of propofol in plasma samples
Chien-Chong Hong a,n, Chih-Chung Lin b, Chian-Lang Hong c,d,nn,1, Zi-Xiang Lin a,
Meng-Hua Chung a, Pei-Wen Hsieh b,e
a
BioMEMS and Nanobiosystems Laboratory, Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu, Taiwan
b
Department of Anesthesiology, Chang Gung Memorial Hospital, Taoyuan, Taiwan
c
Department of Anesthesiology, Chang Gung Memorial Hospital, Chiayi, Taiwan
d
Chang Gung University of Science and Technology, Chiayi, Taiwan
e
Graduate Institute of Natural Products, School of Traditional Chinese Medicine, and Graduate Institute of Biomedical Sciences, College of Medicine, Chang
Gung University, Taoyuan, Taiwan

art ic l e i nf o a b s t r a c t

Article history: This paper proposes a novel handheld analyzer with disposable lab-on-a-chip technology for the elec-
Received 10 May 2016 trical detection of the anesthetic propofol in human plasma samples for clinical diagnoses. The developed
Received in revised form on-chip biosensors are based on the conduction of molecularly imprinted polymers (MIPs) that employ
29 June 2016
label-free electrical detection techniques. Propofol in total intravenous anesthesia is widely used with a
Accepted 9 July 2016
target-controlled infusion system. At present, the methods employed for detecting blood propofol con-
Available online 11 July 2016
centrations in hospitals comprise high-performance liquid chromatography and ion mobility spectro-
Keywords: metry. These conventional instruments are bulky, expensive, and difficult to access. In this study, we
Propofol developed a novel plastic microfluidic biochip with an on-chip anesthetic biosensor that was char-
Microfluidic chip
acterized for the rapid detection of propofol concentrations. The experimental results revealed that the
Molecularly imprinted polymer
response time of the developed propofol biosensors was 25 s. The specific binding of an MIP to a non-
Biosensor
Label-free electrical detection imprinted polymer (NIP) reached up to 560%. Moreover, the detection limit of the biosensors was
0.1 μg/mL, with a linear detection range of 0.1–30 μg/mL. The proposed disposable microfluidic biochip
with an on-chip anesthetic biosensor using MIPs exhibited excellent performance in the separation and
sensing of propofol molecules in the human plasma samples. Compared with large-scale conventional
instruments, the developed microfluidic biochips with on-chip MIP biosensors present the advantages of
a compact size, high selectivity, low cost, rapid response, and single-step detection.
& 2016 Elsevier B.V. All rights reserved.

1. Introduction controlled pharmacokinetic models (e.g., the Marsh model and the
Schnider model). However, patient pharmacokinetics and phar-
The intravenous anesthetic agent propofol (2.6-diisopropyl- macodynamics vary with age, cardiac output, coexisting diseases,
phenol) is a short-acting general anesthetic that is widely used for concurrent drug administration, body temperature, and body-
induction and maintenance through continuous IV infusion and weight (Coetzee et al., 1995). These factors play a critical role in
rapid recovery in clinical practice. In total intravenous anesthesia, determining target concentrations of propofol. Propofol has se-
propofol has recently become a popular and practical anesthetic dative and hypnotic effects, and the anesthetic depth is de-
termined for the effect site in brain. However, it is infeasible in
because of the availability of a propofol target-controlled infusion
clinical practice to determine the propofol concentration in the
(TCI) pump with a built-in TCI system, which enables the user to
brain. Even if we could directly measure the concentration of
set the propofol injection concentration by using computer-
propofol in the brain, it would be necessary to know the accurate
propofol concentrations in the plasma. Currently, a method for
n
Correspondence to: Present address: 101, Sec. 2, Kung Fu Rd., Engineering measuring or detecting the real-time anesthetic propofol agent
Building I, Rm. 629, Hsinchu, Taiwan. concentration, confirming that the drug has reached the effect site,
nn
Corresponding author at: Department of Anesthesiology, Chang Gung Mem- is unavailable.
orial Hospital, Chiayi, Taiwan.
E-mail addresses: [email protected] (C.-C. Hong),
Several methods are used for monitoring the depth of an-
[email protected] (C.-L. Hong). esthesia, including monitoring the bispectral index, which reflects
1
Present address: 6. West Sec. Chia Pu Road, Putzu City, ChiaYi Hsien, Taiwan. a signal-processed electroencephalographic parameter (Ouchi,

http://dx.doi.org/10.1016/j.bios.2016.07.032
0956-5663/& 2016 Elsevier B.V. All rights reserved.

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i
624 C.-C. Hong et al. / Biosensors and Bioelectronics 86 (2016) 623–629

2015). However, at present, an ideal, highly reliable monitor is or electrochemical detection compared with nonconducting MIPs.
unavailable. In one study, the electroencephalography patterns The challenge presented by nonconductive MIPs in electrical bio-
showed high variation during propofol treatment (Sigalovsky, sensors can be addressed by conducting a precise, thin conformal
2003). Patient awareness under general anesthesia remains deposition (approximately 10 nm) of nonconducting MIPs on na-
unpreventable (Khan et al., 2014). Several studies have reported notubes to enhance the performance of electrical MIP biosensors
abnormal blood propofol concentrations during infusion in obese (Cai et al., 2010).
patients (Tachibana et al., 2014) and epidural blockade-dosed pa- In a past study, we devised microfluidic biochips with on-chip
tients (Sitsen et al., 2016), as well as hypothermia effects in in- MIP films on the basis of the colorimetric principle, with Gibb's
tensive care unit patients (Bjelland et al., 2013) and scoliosis-op- reagent for propofol detection (Hong et al., 2010). Compared with
erated children (Panchatsharam et al., 2014). Numerous past stu- the optical sensing mechanism, electrical sensing provides a more
dies have compared their results in using various traditional in- effective and direct approach with compact structures for the
struments for the detection of plasma propofol concentrations, transduction of analyte recognition. In addition, electrical sensing
such as high-performance liquid chromatography and ion mobility can easily be directly integrated with handheld, embedded elec-
spectrometry (Yarbrough et al., 2012; Zhou et al., 2015). These tronic systems. Furthermore, label-free detection, which does not
conventional instruments are bulky, expensive, and difficult to require color reagents or label tags, can be applied during elec-
access. In addition, only people who have received proper training trical detection with conducting MIPs. In this study, we developed
can operate them; furthermore, these instruments cannot be used and characterized a handheld analyzer with molecularly imprinted
to detect real-time propofol concentrations in human plasma biosensors for electrical detection of anesthetic propofol in plasma
samples. To achieve effective blood propofol concentrations and samples. An alternative simplified detection method is used for a
avoid adverse effects resulting from excessive or insufficient pro- cost-effective handheld analyzer. The proposed analyzer presents
pofol administration, we require a more convenient clinical means the following advantages: compact design, full integration, ease of
of monitoring blood propofol concentrations. use, high specificity, cost effectiveness, rapid response, small
Certain studies have investigated the possibility of propofol sample volume, and single-step detection.
sensing for breath gases (Harrison et al., 2003; Laurila et al., 2011).
It's unclear understanding to relate blood concentrations in breath
samples. Past studies have investigated propofol-sensing on the 2. Materials and methods
basis of optical detection with molecularly imprinted polymers
(MIPs) (Hong, et al., 2010; Li et al., 2012), optical detection with solid The electrical propofol detection system includes a handheld
phase extraction (Cowley et al., 2012) and electrochemical detection analyzer and a disposable plastic lab on a chip with conducting
with PVC membrane-coated sensors (Kivlehan et al., 2015). molecularly imprinted biosensors. As shown in Scheme 1a, the
The demand for point-of-care testing in clinical diagnostics is handheld platform consists of Java-programmed controlled ARM-
increasing. Point-of-care assays can considerably improve the 11-based embedded systems, a 4-inch touchscreen, charging and
quality of medical care. In the past decade, technologies for clinical discharging circuit, battery, electronic chip socket, and disposable
diagnostics have been widely developed and successfully applied microfluidic biochip with on-chip biosensors. The general-purpose
in several fields (Abel, 2015). Among these technologies, in which input and output (GPIO) interface on the embedded board is used
lab-on-a-chip applications and microfluidics play a critical role to trigger the charging and discharging circuit. The analog-to-di-
(Chin et al., 2007; Mohammed and Desmulliez, 2011), it is most gital (ADC) module on the embedded board is employed for signal
crucial to achieve biosensors with highly specific recognition and data acquisition from the biosensors to the embedded system for
highly sensitive transduction. Molecular imprinting provides an further data processing. Scheme 1b shows a schematic of the
approach to highly specific binding that has been adapted from microfluidic biochips with on-chip molecularly imprinted bio-
biological antigen and antibody binding. The molecular imprinting sensors for propofol sensing; the microfluidic biochip was de-
method is relatively simple and easy to perform in a suitable signed for the precise transport of samples to the biosensor mi-
fashion, requiring only functional monomers, templates, solvents, crochambers by capillary force. The microfluidic chip has a three-
and cross-linking agents. In molecular imprinting, polymerization layer structure. The fluidic inlets and outlets are on the top layer.
is followed by the removal of the template. During these pro- Patterned double-sided stickers are positioned between the top
cesses, a number of functional monomers are assembled in an and bottom layers to form a microfluidic channel. In addition, the
orderly manner, with their functional groups placed at the desired conducting MIP electrical biosensors are situated on the bottom
sites within cavities of the desired size. Therefore, MIPs with layer. Compared to non-conducting MIP electrodes, conducting
binding sites, having a specific shape and functional group re- MIP electrodes have much lower impedance, which are higher
cognition, can be used for performing high-specificity and high- sensitivity and better accuracy in tiny surface change. The im-
selectivity sensing on the target molecular compound (Uzun and printed nanocavities on the MIPs are used for the specific binding
Turner, 2016). A morphological study of nanocavities on MIPs re- of propofol molecules. The working principle of the molecularly
ported an enhancement in MIP performance (Hong et al., 2012). imprinted biosensors is depicted in Scheme 1c. When propofol
MIP materials have exhibited good performance in separating or molecules are separated from human plasma samples and cap-
extracting target proteins from blood plasma samples (Hong et al., tured by MIP biosensors, the surface electrical properties on the
2013; Karimian et al., 2013; Chen et al., 2016; Gao et al., 2015). In MIP biosensors are altered because of the binding of the propofol
another study, MIPs were realized on ITO electrodes for electro- molecules, which reduces the conductivity of the MIPs. The bio-
chemical sensing (Yoshimi et al., 2013). sensors are used to directly detect electrical changes. The binding
Conducting polymers are widely used in gas and chemical of propofol molecules on the MIPs affects the time constant of the
sensors (Janata and Josowicz, 2003). Pyrrole is the most ex- discharging curves. The conducting MIP, which provides highly
tensively employed conducting polymer in electrochemical bio- specific binding to the analyte, is used for separating and sensing
sensors (Ramanavicius et al., 2006). Furthermore, certain studies propofol, functioning as a plastic antibody for capturing propofol
have applied conducting MIPs on Au and Pt electrodes for elec- in human plasma samples.
trochemical sensing (Lakshmi et al., 2009; Ramanaviciene and The electrical equivalent circuits of the conducting MIP IDA
Ramanavicius, 2004). Because of the considerably smaller im- biosensor in the microfluidic channel are shown in Scheme 2a. Ccp
pedance, conducting MIPs have higher sensitivity with electrical and Rcp represent the surface capacitance and resistance of the

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i
 C.-C. Hong et al. / Biosensors and Bioelectronics 86 (2016) 623–629 625

Scheme 1. Schematic of our developed propofol sensing system based on molecularly-imprinted biosensors. (a) handheld analyzer, (b) propofol capture and sensing on the
molecularly imprinted biosensor, and (c) specific binding of propofol molecules in imprinted nanocavities.

Scheme 2. Electrical equivalent circuit of the single conducting MIP electrode before and after propofol molecule binding.

biosensor, respectively. The electrical equivalent circuits of a single copolymers (COCs) (Topas COC 6015, Tg: 150 °C) were chosen as
conducting MIP before and after propofol molecular binding are the chip substrates. Because COCs exhibit high resistance against
shown in Scheme 2b. Propofol is a nonconducting molecule; after several chemical solvents, they are compatible with most pro-
analyte binding on MIPs, propofol molecules in imprinted nano- cesses in biomedical microelectromechanical systems.
cavities contribute surface capacitance and also block some con-
ductivity of MIPs, thereby increasing surface impedance. The
changes in surface impedance could be measured by a costly LCR 3. Fabrication and integration
meter. However, to realize propofol sensing with a cost-effective
handheld analyzer, the discharging detection mechanism was The fabrication steps comprised the fabrication of the micro-
employed to measure the propofol concentration, rather than fluidic biochips and the MIP biosensors as well as the encapsula-
measuring the surface impedance directly. The time constant, tion of the device. A plastic material, namely COC, was chosen as
which was derived from the discharging curve of the MIP bio- the substrate. In this study, 4-inch COC wafers with a thickness of
sensor, was used to sense changes in surface impedance. 1.4 mm were fabricated using plastic injection molding techniques,
For this study, pyrrole and propofol were purchased from Sig- and MIP biosensors were fabricated on the COC substrate. The
ma-Aldrich for the electropolymerization process. Cyclic olefin fabrication process for the MIP membranes involved three steps:

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i
626 C.-C. Hong et al. / Biosensors and Bioelectronics 86 (2016) 623–629

Fig. 1. Photographs of the fabricated propofol sensing system. (a) Microfluidic stickers, (b) handheld system with plastic microfluidic chip, (c) plastic microfluidic biochip
with on-chip conducting molecularly-imprinted biosensors, and (d) electropolymerized conducting molecularly-imprinted polymer on gold IDA electrodes.

combination, electropolymerization, and extraction. A monomer, thick double-sided tape. The microfluidic channel was formed
namely pyrrole, was mixed with template molecules in a me- by sandwich-bonding the patterned double-sided stickers be-
thanol solvent. Furthermore, the MIP was composed of propo- tween two blank COC chips. Afterward, the MIP biosensor was
fol–pyrrole, with a molar ratio of 1:5. A blank 4-inch COC plastic integrated with a plastic microfluidic chamber. Fig. 1b displays
wafer was prepared for electron beam deposition of a 100-nm the handheld analyzer with the fabricated COC plastic micro-
gold film. Photolithography was conducted to pattern four bio- fluidic chip. The fabricated plastic microfluidic chip and the on-
chips on each COC wafer and four IDA electrodes on each COC chip IDA conducting MIP biosensor are shown in Fig. 1c and d,
biochip. Subsequently, the biochips were immersed in the MIP respectively.
solution. A LabVIEW-controlled program with power sources
was connected to the electrodes. A current density of 25 μA/
mm2 was applied for 30 s for electropolymerization at 25 °C. 4. Results and discussion
After the electropolymerization of the conducting polymer
electrodes, the imprinted sites were formed by releasing the Fig. 2a–d present the SEM cross-sectional images of the fabri-
template molecules from the MIP electrodes with the methanol cated MIP for various polymerization durations. The thickness of
solvent for 24 h. Then, the MIP film was patterned on gold IDA the MIP increased linearly with the duration of polymerization, as
electrodes. Fig. 1a shows the microfluidic stickers, which were shown in Fig. 2e. The MIP thickness ranged from 133 to 739 nm.
composed by pressing a CNC-machined knife mold on 120-μm- According to the SEM observations, the polymer film had a slightly

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i
 C.-C. Hong et al. / Biosensors and Bioelectronics 86 (2016) 623–629 627

Fig. 2. Electropolymerization results. (a)-(d) SEM pictures, and (e) electropolymerized MIP thickness with durations.

nonuniform surface formation for a short deposition duration. In This study was approved by the institutional ethics committee
this study, polymerization was set for 30 s to fabricate the MIP. of Chang Gung Memorial Hospital. Written informed consent was
Compared with the optically-polymerized MIP film (Hong et al., obtained from patients undergoing elective surgery. Propofol-free
2012), the electropolymerized MIP film had an uneven surface blood samples were obtained before the surgery and were cen-
morphology, as shown in Fig. S1. trifuged for 20min for plasma separation. The measurements

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i
628 C.-C. Hong et al. / Biosensors and Bioelectronics 86 (2016) 623–629

curves under various propofol concentrations that were measured

between 0 and 30μg/mL. Afterward, the time constants of the
discharging curves for the various propofol concentrations could
be calculated from the curves. In this study, we tested the dis-
charging curve of the bare gold IDA electrodes in the blood plasma
samples with a propofol concentration of 20μg/mL, finding a slight
increase in voltage, as shown in Fig. S2. The discharging slopes
were calculated as a calibration curve (Fig. 4c). The variations of
the developed biosensors were tested with different propofol
concentrations (Fig. 4d). The coefficient of the variations for the
propofol biosensors was 6.87%. The proposed electronic propofol
biosensors with MIP electrodes for the electrical detection of
propofol exhibited high performance in separating and sensing
anesthetic propofol molecules.
The new-generation propofol analyzer is based on the label-
free electrical detection method with a conducting MIP. We
measured the propofol concentrations of patient B and compared
the results by performing HPLC, as shown in Fig. S3a. The relative
standard deviation was 3.7%. Although HPLC cannot be used for
determining precise propofol concentrations likely caused by the
difficulties of performing column separation with propofol, the
results revealed a linear trend. Fig. S3b shows the measured pro-
pofol concentrations in the plasma sample of patient B that were
determined using the developed biosensors on the handheld
analyzer; these concentrations were calculated using the calibra-
Fig. 3. Dynamic measurements during analyte binding on MIP biosensors. tion curve shown in Fig. 4c. The relative standard deviation
(a) Capacitance measurements and (b) resistance measurements.
was 8.7%.
In this study, the response time of the developed biosensors
indicated that patient A had 139.5mEg/L of Na þ , 106.7mEq/L of was 25 s. The specific binding tests were performed using the
Cl  , and 4.62g/dL of albumin. By contrast, patient B had 134.6mEg/ characterization of the absorption ratio of propofol molecules on
L of Na þ , 107.7mEq/L of Cl  , and 3.00g/dL of albumin. The plasma the MIP and NIP surfaces. The specificity (i.e., imprinting factor) of
sample from patient A was used to establish a calibration curve, the MIP films was 560%. Because of its high specificity, we believe
and that from patient B was employed for comparison. In the that the biosensor can be employed for accurate, real-time de-
experiment, 2% propofol injection samples were spiked into the tection of propofol concentration; furthermore, the biosensor can
plasma samples to obtain samples with different propofol con- aid in monitoring the anesthetic level and the pharmacokinetics of
centrations. The microfluidic chip was inserted into the socket on propofol under general anesthesia in the future.
the battery-powered handheld analyzer for detection. Then, 4-μL When whole blood testing is performed, materials (e.g., blood
plasma samples with known propofol concentrations were drop- cells) that affect surface electrical properties should be filtered out.
ped at the inlet of the microfluidic biochips. The plasma samples Centrifugal plasma separation from whole blood is currently per-
required approximately 6–8s to self-fill the microchannels by ca- formed in hospital laboratories. The developed electrical bio-
pillary force. Afterward, the detection procedure was performed sensors with on-chip microfilters, in addition to a vacuum module
by manipulating the developed Java program in the ARM-em- for plasma separation from whole blood—both of which are under
bedded system to communicate through the GPIO and ADC in- investigation—would further simplify the detection process for
terface for discharging trigger, signal processing, data acquisition, on-site and point-of-care testing.
and storage. Fig. 3 shows the dynamic propofol binding results
obtained with an electrochemical impedance spectroscope. From
the binding curve for capacitance measurements obtained using 5. Conclusions
an electrical impedance measurement instrument, the binding
reached saturation at 20s. After the analyte binding, the capaci- In this study, we successfully developed a novel plastic mi-
tance decreased around 28% and the resistance increased crofluidic biochip with an on-chip anesthetic biosensor that was
around 16%. characterized for the electrical detection of propofol in human
Propofol molecules, which are nonconducting, mainly con- plasma samples. The handheld analyzer with on-chip molecularly
tribute to changes in surface impedance after binding to the MIP. imprinted biosensors using label-free electrical detection methods
According to the equivalent circuit of the biosensors, the analyte demonstrated excellent performance in separating and sensing
concentration can be measured by detecting changes in surface anesthetic propofol molecules. The experimental findings show
impedance, which is related to the captured analyte. The dis- that the developed microfluidic system with propofol MIP bio-
charging detection mechanism was employed to measure the sensors can successfully detect propofol samples with concentra-
propofol concentration, rather than measuring the surface im- tions of 0.1–30 μg/mL, whereas the specific binding of an MIP was
pedance directly. The time constant, which was derived from the found to reach up to 560%. The limit of detection and the response
discharging curve of the MIP biosensor, was used to sense changes time of the developed biosensors were 0.1 μg/mL and 25 s, re-
in surface impedance. The voltage bias of the biosensor is set to spectively. Compared with conventional large-scale instruments,
0.30V. The developed biosensor began the measurement at 20s. the developed microfluidic system with MIP biosensors presented
The discharging for 5s is enough for calculating the time constant. the advantages of a compact design, full integration, ease of use,
According to the measured discharging curves for the NIP and MIP, high specificity, cost effectiveness, rapid response, small sample
shown respectively in Fig. 4a and b, the specific binding of the volume, and single-step detection. In the future, the developed
biosensor reached up to 560%. Fig. 4b shows the discharging label-free electrical biosensors with MIP recognition technique can

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i
 C.-C. Hong et al. / Biosensors and Bioelectronics 86 (2016) 623–629 629

Fig. 4. Single-step propofol detection in human serum samples by using the developed handheld analyzer. (a) Dynamic measurements of propofol using NIP IDA electrodes,
(b) dynamic measurements of propofol using conducting MIP IDA biosensors (c) RC value for different propofol concentrations as the calibration curve, and (d) sensor-to-
sensor variation for three different propofol concentrations using the developed handheld analyzer.

be further applied in disposable lab-on-chip applications, such as Gao, R., Zhao, S., Hao, Y., Zhang, L., Cui, X., Liu, D., Zhang, M., Tang, Y., 2015. J. Sep.
point-of-care protein analysis and on-site food safety analysis, Sci. 38, 3914–3920.
Harrison, G.R., Critchley, A.D.J., Mayhew, C.A., 2003. Br. J. Anaesth. 91 (6), 797–799.
which are under investigation. Hong, C.-C., Chang, P.-H., Lin, C.-C., Hong, C.-L., 2010. Biosens. Bioelectron. 25,
2058–2064.
Hong, C.-C., Lin, C.-C., Hong, C.-L., Chang, P.-H., 2012. Biomed. Microdevices 14 (3),
Acknowledgement 435–441.
Hong, C.-C., Chen, C.-P., Horng, J.-C., Chen, S.-Y., 2013. Biosens. Bioelectron. 50,
425–430.
This research was supported by the National Science Council of Janata, J., Josowicz, M., 2003. Nat. Mater. 2, 19–24.
Taiwan, Taiwan (MOST 102–2221-E-007–013-MY2) and Chang Karimian, N., Vagin, M., Zavar, M.H.A., Chamsaz, M., Turner, A.P.F., Tiwari, A., 2013.
Biosens. Bioelectron. 50, 492–498.
Gung Memorial Hospital-National Tsing Hua University Joint Re- Khan, M.S., Zetterlund, E.-L., Green, H., Oscarsson, A., Zackrisson, A.-L., Svanborg, E.,
search Project (CMRPG390131 & CMRPG3A0051). Lindholm, M.-L., Persson, H., Eintrei, C., 2014. Basic Clin. Pharmacol. Toxicol. 115,
565–570.
Kivlehan, F., Chaum, E., Lindner, E., 2015. Analyst 140, 98.
Laurila, T., Sorvajarvi, T., Saarela, J., Toivonen, J., Wheeler, D.W., Ciaffoni, L., Ritchie,
Appendix A. Supplementary material G.A.D., Kaminski, C.F., 2011. Anal. Chem. 83, 3963–3967.
Lakshmi, D., Bossi, A., Whitcombe, M.J., Chianella, I., Fowler, S.A., Subrahmanyam, S.,
Supplementary data associated with this article can be found in Piletska, E.V., Piletsky, S.A., 2009. Anal. Chem. 81, 3576–3584.
Li, L., Ding, H., Di, B., Li, W., Chen, J., 2012. Analyst 137, 5632.
the online version at http://dx.doi.org/10.1016/j.bios.2016.07.032. Mohammed, M.-I., Desmulliez, M.P.Y., 2011. Lab. Chip 11, 569–595.
Ouchi, K., 2015. Glob. J. Anesth. 2 (2), 054–056.
Panchatsharam, S., Callaghan, M., Day, R., Sury, M.R.J., 2014. Anesth. Analg. 119,
1150–1157.
References
Ramanavicius, A., Ramanaviciene, A., Malinauskas, A., 2006. Electrochim. Acta 51,
6025–6037.
Abel, G., 2015. Expert Rev. Mol. Diagn. 15 (7), 853–855. Ramanaviciene, A., Ramanavicius, A., 2004. Biosens. Bioelectron. 20, 1076–1082.
Bjelland, T.W., Klepstad, P., Haugen, B.O., Nilsen, T., Dale, O., 2013. Drug Metab. Sigalovsky, N., 2003. AANA J. 71 (5), 373–379.
Dispos. 41, 214–223. Sitsen, E., Olofsen, E., Lesman, A., Dahan, A., Vuyk, J., 2016. Anesth. Analg. 122 (5),
Cai, D., Ren, L., Zhao, H., Xu, C., Zhang, L., Yu, Y., Wang, H., Lan, Y., Roberts, M.F., 1341–1349.
Chuang, J.H., Naughton, M.J., Ren, Z., Chiles, T.C., 2010. Nat. Nanotechnol. 5, Tachibana, N., Niiyama, Y., Yamakage, M., 2014. Eur. J. Anaesthesiol. 31, 701–707.
597–601. Yarbrough, J., Harvey, R., Cox, S., 2012. J. Chromatogr. Sci. 50, 162–166.
Chen, F., Zhao, W., Zhang, J., Kong, J., 2016. Phys. Chem. Chem. Phys. 18, 718–725. Yoshimi, Y., Sato, K., Ohshima, M., Piletska, E., 2013. Analyst 138, 5121–5128.
Chin, C.D., Linder, V., Sia, S.K., 2007. Lab. Chip 7 (41), 41–57. Zhou, Q., Li, E., Wang, Z., Gong, Y., Wang, C., Guo, L., Li, H., 2015. J. Breath Res. 9,
Coetzee, J.F., Glen, J.B., Wium, C.A., Boshoff, L., 1995. Anesthesiology 82, 1328–1345. 016002.
Cowley, N.J., Laitenberger, P., Liu, B., Jarvis, J., Clutton-Brock, T.H., 2012. Anaesthesia Uzun, L., Turner, A.P.F., 2016. Biosens. Bioelectron. 76, 131–144.
67, 870–874.

Please cite this article as: Hong, C.-C., et al., Biosensors and Bioelectronics (2016), http://dx.doi.org/10.1016/j.bios.2016.07.032i