- Email: [email protected]
Development of a conducting polymer cell impedance sensor
Sensors and Actuators B 176 (2013) 667–674
Contents lists available at SciVerse ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Development of a conducting polymer cell impedance sensor Affar S. Karimullah a,∗∗ , David R.S. Cumming b , Mathis Riehle c , Nikolaj Gadegaard a,∗ a
Division of Biomedical Engineering, University of Glasgow, Glasgow G12 8LT, UK Division of Electronics and Nanoscale Engineering, University of Glasgow, Glasgow G12 8LT, UK c Centre for Cell Engineering, University of Glasgow, Glasgow G12 8QQ, UK b
a r t i c l e
i n f o
Article history: Received 15 June 2012 Received in revised form 7 September 2012 Accepted 22 September 2012 Available online 5 October 2012 Keywords: PEDOT Conducting polymer electrodes MDCK cell culture Cell impedance spectroscopy
a b s t r a c t Research in to label free methods for biological analysis have brought interesting developments. Cell impedance spectroscopy has been one of the promising outcomes. Here we show the development of an 8-well impedance measurement setup and studied the use of conducting polymers as electrode material in cell impedance spectroscopy. We have developed devices using PEDOT:PSS electrodes and shown its advantages (lower impedance and faster to reach electrochemical equilibrium) over conventional materials, such as gold. It is observed through electrochemical analysis that the lower interfacial impedance is due to the low charge transfer resistance of PEDOT:PSS. MDCK cell proliferation experiments were performed using both types of electrode materials to provide a comparative result. We applied electrical modeling methods to understand the cell–substrate interactions and shown its applications in cell impedance spectroscopy. This study presents the development and advantages of cell impedance spectroscopy using conducting polymer electrodes. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The need for label free and real time analysis in cell biology is constantly increasing. The nature of cell biological research requires parallel experiments in large numbers that yield cumulative statistical results. Research to meet these demands has led to many technological advances in fields such as lab-on-chip devices and instrumentation that process large work intensive experiments in a factory line style. Most lab-on-chip devices still require cells to be labeled before analysis [1]. Novel ways of detection along with parallel processing are required which do not interfere with the cells in culture. Cell impedance spectroscopy (CIS), a method pioneered by Giaever and Keese [2], provides such a possibility. CIS has been an evolving label-free tool for over two decades. It allows cells to be monitored in vitro and data can be collected for multiple experiments in parallel. Early research showed that the method could not only be applied for cell proliferation and motility measurements, but also be used to assess morphological characteristics (confluence of the cells) and behavioral aspects (such as metastasis) of the cells [3,4]. The applications and research for CIS have been growing rapidly including in field toxicological studies [5], drug testing [6], measuring mesenchymal stem cell
∗ Corresponding author. Tel.: +44 141 3305243. ∗∗ Corresponding author. Tel.: +44 141 3306691. E-mail addresses: [email protected] [email protected] (N. Gadegaard).
(A.S.
Karimullah),
0925-4005/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2012.09.075
differentiation [7–9], the measurement of cell substrate separation [10] and the monitoring of relaxation and contractility of muscle cells [11]. Electrodes for CIS have predominantly been based on noble metals. Conducting polymers have been researched thoroughly for antistatic coatings, electrodes (electrochemical capacitors [12], electrochromic devices, ion sensors [13]) and have been instrumental in new forms of biosensors [14]. Conducting polymers were first reported in the mid 20th century [15–17] but it was the discovery by Shirakawa et al. in 1977 that is considered the highlight for their discovery of polyacetylene [18] and earned him, along with Heegar and MacDiarmid, the Nobel Prize in chemistry in 2000. The application of conducting polymers in biological research are numerous and they are being extensively used for tissue engineering, neural probes, biosensors, drug delivery and actuators [19]. Commercially available conducting polymers, such as poly(3,4-ethylenedioxythiophene) (PEDOT), are commonly used as conductive coatings and intermediate layers in organic electronics [20]. PEDOT has the advantage of being transparent and biocompatible [21] while being considered the most stable (in the atmosphere and oxidizing environments) conducting polymer currently available [22–24]. It allows for simple and flexible methods for patterning, and also has the ability to be blended with other polymer compounds to create novel materials with unique properties [25,26]. PEDOT has been proven to have good electrochemical stability in phosphate buffer solutions, even when polarized [27,28]. It has been studied as a neural electrode coating and is considered one of the promising new materials with higher charge injection limits
668
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
and better signal to noise ratio in measurements of neural activity [29–31]. PEDOT electrodes tend to delaminate over a period of a few weeks when in a saline solution [28], however Reza et al. found PEDOT nano tube based coatings to have improved adhesion of the coating to the substrate and the lower impedance provides better neural signal recordings [32]. They also found that the material improved the neural attachment and neurite outgrowth [33] and by coating a electrospun biodegradable polymer (loaded with drugs) with PEDOT, they were able to use electrical stimulations to release these drugs [34]. Most CIS setups work at frequencies where the higher impedance of nobel metal electrodes is of little concern. However lower frequencies can contain information useful for curve fit analysis and possibly distinguish the cell type. Not only can PEDOT as an electrode material for CIS provide a lower interfacial impedance but due to its transparency cells are accessible to high resolution microscopy, which makes this a uniquely suitable material for this application. This paper presents development of a cell impedance measurement device that utilizes PEDOT:PSS (solution processable PEDOT with poly(styrenesulfonate) (PSS) as doping agent) as the electrode material. The devices were fabricated using photolithographic methods. A complete impedance measurement setup was implemented in Labview which provides more flexibility in the device design and data analysis. A multiplexer was also designed and implemented for a multi well device that allows to measure eight experiments in parallel. As the design has the same electrode connection lengths, ensuring, that results for all the wells are not impacted by variations in inductance and series resistance of the material. The design shows the possibilities of creating simple and cost effective polymer based cell impedance measurement electrode devices. measurements with 4-(2-hydroxyethyl)-1Impedance piperazineethanesulfonic acid (HEPES) saline solution showed a lower interfacial impedance of PEDOT:PSS electrodes in comparison to similar sized and shaped Au electrodes, highlighting the electrical advantage of PEDOT:PSS as an electrode material. Through electrochemical impedance analysis and curve fitting of the results to Randle’s theoretical model [35] of an electrode–electrolyte interface, we determined that the lower charge transfer resistance of PEDOT:PSS as compared to Au is the reason for the reduced interfacial impedance. Experiments with the epithelial Madin-Darby Canine Kidney (MDCK) cell line showed that the PEDOT:PSS electrodes were fully capable of being used for CIS. Due to the fast charge transport capabilities of PEDOT:PSS, the electrodes quickly reach their equilibrium state. Curve fitting of the experimental data to typical electrical models of biological cells in series with the electrode equivalent circuit, show the versatility of CIS as a method to observe not only cell growth but also the cell–substrate interaction.
2. Materials and methods Commercially available poly (3,4 ethylenedioxythiophene) poly(styrenesulfonate) (Orgacon S305 plus, AGFA) was used. It is a 0.54% by weight aqueous solution of PEDOT:PSS. The S305 plus includes binders to help adhesion to substrates and stabilizers for improved environmental stability. SU8-3005 (Microchem, Newton, MA, USA) and Microposit S-1818 (Shipley, Coventry, UK) photo resists were used for fabrication. EC solvent (Ethyl Lactate, Microposit, Shipley, Coventry, UK) was used to develop SU-8, and MicroDev (Microposit, Shipley, Coventry, UK) was used for S-1818 development after a 1:1 dilution with RO water. The glass substrates used were 50 mm × 75 mm microscope slides (Corning, Amsterdam, The Netherlands). Acetone, methanol and iso-propanol (Sigma Aldrich, Dorset, UK) were used for cleaning the
substrates and devices at various steps. Wells were created using -slide 8-well culture dishes (Ibidi, Munich, Germany). MDCK II cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with Medium 100, fetal bovine serum (FBS), sodium pyruvate and antibiotic all acquired from GIBCO, Invitrogen. Prior to cell seeding, the device wells were sterilized with 70% ethanol and rinsed with HEPES saline solution before being coated with poly-L-lysine (PLL Mol. Wt. 150,000–300,000, cell culture tested, Sigma, Poole, UK). Incubation during the cell culture and measurements was done at 37 ◦ C in a 5% CO2 environment. In certain experiments cell culture inserts (Ibidi, Munich, Germany) were used to seed cells in a specific area in contrast to a full coverage of the culture well areas. 2.1. Instrumentation and software Impedance measurements for the electrode comparison were taken using a QuadTech 1420 LCR meter (from 20 Hz to 1 MHz; Quadtech, Marlborough, MA, USA). Cell impedance measurements were done using an Agilent 4294A impedance analyzer (Agilent, Berkshire, UK). The impedance measurement equipment was connected to the wells through a multiplexer. The multiplexer was designed with four 1:16 channel switches (two ADG726, Analog Devices). By keeping the multiplexers separate for each of the four current and voltage probe lines, the configuration further negates any effects due to the multiplexer. The ADG726 package has an “ON” resistance of only 4 and does not mitigate the accuracy of the system. A NI-DAQ 6211 (National Instruments, Berkshire, UK) provided the digital signals to the multiplexers and a single power supply was used to provide positive and negative voltage to linear voltage regulators, which feed the rest of the circuitry. The multiplexer board holds the device and both sit inside an incubator. The setup is controlled using software designed on Labview (National Instruments, Berkshire, UK). The amplitude of the signals used for the electrochemical comparison was 20 mV, which is low enough to work in the linear region of the current and over-potential relationship [36,37]. The amplitude for CIS measurements was kept higher (80 mV) which was set after it was observed to be free of any noise when measured using both types of electrodes. It is small enough to be free of harmonic distortion and neither caused any adverse effect to the cell culture. 2.2. Device fabrication A single device had eight wells, arranged in four columns and two rows. Each well had a pair of electrodes: a working electrode and a reference electrode. The design of the electrodes was similar to those available from Applied Biophysics ECIS systems with a large reference electrode and a circular working electrode whose dimensions and geometry are defined by the windows created in the insulation layer. The devices (Fig. 1) were fabricated using photolithography on a glass substrate. The glass substrates were cleaned using acetone, methanol and propane-2-ol, 5 min in an ultrasonic bath, followed by oxygen plasma cleaning (GaLa Instrumente, Bad Schwalbach, Germany) for 3 min at 160 W. An alignment layer was deposited on the substrate via a lift off process involving S-1818 lithography and aluminum deposition by evaporation using a Plassys metal evaporator (Plassys, Marolles En Hurepoix, France). Fig. 1 shows the fabrication process for the PEDOT:PSS electrodes and the SU-8 insulation layer. PEDOT:PSS was spin coated at 1000 RPM for 30 s onto the substrate. This was done twice to achieve approximately 100 nm thickness. The sample was baked at 95 ◦ C for 1 min between the two layers and for 5 min after the final coating. S-1818 photo resist was then spin coated and baked at 95 ◦ C for 5 min. After lithography and development (90 s) of S-1818, PEDOT:PSS was exposed to 5% sodium
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
669
Fig. 1. (Left) Schematic of a single row of the CIS device. Each row has two pairs of electrodes (one pair is shown inside a well). The working electrode has a window in the nearly transparent SU-8 insulation layer. (Right) Schematic of the fabrication of PEDOT:PSS devices. (A) Substrate with PEDOT:PSS layer. (B) S-1818 layer on top. (C) S-1818 is patterned and developed. (D) PEDOT:PSS electrode formed by oxidizing with bleach, rinsing in RO water and removing S1818 with acetone. (E) SU-8 coating and (F) Window created in SU-8 over electrodes and points of contact.
hypochlorite solution for 10 s and rinsed in running RO water till the oxidized PEDOT:PSS was completely removed. The sample was cleaned in acetone, propan-2-ol, and an oxygen plasma (1 min at 60 W) then spin coated with SU8-3005 and baked at 95 ◦ C for 10 min. The SU-8 was patterned using photolithography and given a post exposure bake before development in EC solvent for less than 1 min. A final bake at 150 ◦ C for 5 min was performed before an oxygen-plasma cleaning (1 min at 80 W). Wells were attached using silicone sealant and silver paint was applied to the PEDOT:PSS electrodes at the end where the connectors would interface with them. This provided some mechanical strength so that the PEDOT:PSS layer was not scratched by the metal connector pins. The Au electrode devices were made by photolithography using S-1818 (inverted pattern to the PEDOT:PSS electrode mask). After development, 2 nm of Ti was deposited for adhesion purposes followed by 50 nm of Au deposited using a Plassys metal electron beam evaporator. The S-1818 was removed with acetone and SU-8 was spun onto the device and patterned similar to the PEDOT:PSS devices. A final bake at 150 ◦ C for 5 min and plasma cleaning was performed as well. 3. Results and discussion 3.1. Electrode material comparison We investigated the differences between using PEDOT:PSS and Au as electrode materials for CIS. The basic differences between the two in their physical characteristics are already understood, but the electrical behavior of the two electrode materials under cell culture conditions needed to be studied further. Earlier research had shown the benefits of PEDOT:PSS for its charge injection capabilities and its electrochemical behavior. Here we would like to assimilate those studies and use them to compare the electrochemical characteristics of the two materials. Impedance spectroscopy results of Au and PEDOT:PSS electrodes in HEPES saline solution were used to evaluate their electrical impedance model values. The electrodes were of the same dimensions, with a working electrode area of 4.15 × 10−4 cm2 (230 m in diameter) and the reference electrode area being 6 × 10−2 cm2 , which is sufficiently large to have minimal contribution on the impedance values measured. The equivalent circuit model considered is essentially Randles’ circuit (with a Warburg element) which is considered to be the simplest model that provides a good interpretation of an electrochemical cell catering for both the double layer capacitance and the Faradaic components [35]. The model we used varied slightly as it had a frequency dependent double layer capacitance that is modeled mathematically as a constant phase element [38] and provided the best curve fits. This constant phase element behavior could rise due to surface
roughness of PEDOT:PSS, distribution of reaction rates with the many chemical compounds present in HEPES saline, and nonuniform current distribution paths as the electrodes are not directly facing each other [36,39]. Complex non-linear least square (CNLS) curve fitting was used to find the values for the double layer capacitance ‘Q’ and ‘n’ (modeled as a constant phase element, where Cdl = Q(jω)n−1 and Zcpe = 1/Q(jω)n ), charge transfer resistance ‘Rt ’, series resistance of the bulk solution and electrodes ‘Rs ’ and the Warburg impedance coefficient ‘Aw ’. Two values for Rs are shown, one that is per unit area and the other which is the actual value. This is because in the case of PEDOT:PSS, the majority of the series resistance will be due to the electrode to connector length and not necessarily relevant to the understanding of the impedance model of the electrode or bulk of the solution. It can also be argued that due to the added resistivity of the PEDOT:PSS, the actual potential drop across the electrode–electrolyte interface is not the same anymore. However, given the fact that we are already in the linear region of the current and over-potential relationship, the charge transfer resistance value will hold true even for smaller voltages. Also, given that the series resistance in the case of PEDOT:PSS is less than half the value of Rt , the potential drop across the interfacial impedance will still be larger than 10 mV and the difference between the two materials is still far too great. Eq. (1) shows the total impedance for the electrode, where ω is the angular frequency. Fig. 2 shows the comparison between absolute impedance |Z| values measured with HEPES saline using Au and PEDOT:PSS electrodes respectively. The table in Fig. 3 shows the results of curve fit to the model shown in Fig. 2 (mathematically represented in Eq. (1)). The blue lines in the plots represent impedance predicted by the theoretical model using the results of the fit. Each frequency point is marked and the graphs show a tight fit to the measured impedance values (the red dots). The results show that the Au electrodes are more polarizable and have high values for the Faradaic impedance components of the circuit. PEDOT:PSS electrodes show low Faradaic impedance and a larger double layer capacitance with lower value of ‘n’, meaning that its characteristics are further from that of an ideal capacitor. Zelectrods = Rs +
1 √ √ Q (jω) + (1/(Rt + (Aw / ω) + (Aw /j ω)))
(1)
Bobacka et al. have shown that PEDOT:PSS has fast charge transport from electrode to the electrolyte and that the polymer has an excess of supporting electrolyte which provides good charge transfer capability. Hernandez-Labrado et al. have discussed diffusion of ions into the polymer layer due to its porous nature causing a frequency dependant Warburg behavior [40,41]. The results here are in close agreement to those cases, however we must consider that the PEDOT:PSS layer here, since it has been
670
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
Fig. 2. Absolute impedance (|Z|) results of PEDOT:PSS and Au electrodes (230 m diameter) with HEPES saline solution and the interfacial impedance equivalent circuit model. Zcpe is the impedance of the frequency dependant double layer capacitance (a constant phase element). Zw is the Warburg impedance.
Fig. 3. (Left) Table and (Right) Nyquist plots of curve fit results using the model shown in Fig. 2 for the impedance of both types of electrodes. Z is the real value and Z is the imaginary value of impedance (Z) measured from 20 Hz to 1 MHz.
spin coated, is not as porous or thick as what they have polymerized electrochemically in their experiments. The possibility of excess solvent in the PEDOT:PSS electrodes is also supported by the signs of adhesion loss after prolonged exposure to aqueous solvents during our experiments. The requirements of CIS however, do not necessitate any further inquiry into this electrochemical research.
3.2. Cell experiments MDCK proliferation was measured using both Au and PEDOT:PSS electrodes. The electrodes for these experiments had an approximate area of 2 × 10−3 cm2 , that is a 500 m diameter window. Fig. 4 shows results of the CIS setup measuring all 8 wells using both PEDOT:PSS and Au electrodes. Wells 1 and 2
Fig. 4. Results of MDCK experiments using PEDOT:PSS and Au electrodes. Data at 1 kHz of all 8-wells is presented. Values shown are the changes from measurement taken at 0 h. The black vertical line shows the time point where media was changed.
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
671
Fig. 5. Results of MDCK using PEDOT:PSS electrodes (Left) and Au electrodes (Right). The graphs are the differences in absolute impedance |Z| and the phase from the initial values. The electrode windows in the SU-8 are indicated by the white rings and the dotted black vertical lines show when the media was changed. Bottom diagram shows how the culture insert is placed at a distance from the electrodes and cells are seeded inside that insert. Once the cells have adhered to the surface the insert can be removed to allow the cells to grow in an outward direction. The cells will eventually start covering the working electrode and reach confluence.
(both black lines) are control wells in each case and the rest of the wells were seeded with approximately 8 × 105 cells. The results show real time impedance changes due to the cell coverage on the electrode surface. The plots show the difference from the initial impedance values, that is |Z| and (phase angle). Initially, the cells attach to the substrate and start covering the working electrode which causes a considerable change in the phase whereas the total change to |Z| is smaller. After approximately 20 h (23 h for Au electrode results), the cell growth increases due to proliferation and cells begin to achieve confluence which causes |Z| to increase. In the case of PEDOT:PSS the phase changes, at this frequency, is increasingly negative where as the experiment with Au electrodes shows increasingly positive phase changes. This is a rather interesting variation. |Z| for both devices are constantly decreasing initially and show sudden changes right after media is change. The change after the media imitates the initial change meaning it is related to chemical changes in the media over time. This could be due to pH/chemical changes over time. Judging from observation of the control wells it is likely to be due to changes in concentration of dissolved CO2 .
Fig. 5 shows results of experiments comparing where cells were seeded in a region above the working electrodes using cell culture inserts (Fig. 5, bottom). After allowing the cells to settle, spread and migrate for 24 h, the inserts were removed and the measurements were started. This allows us to measure cell coverage of an electrode by the mean of migration in a particular direction instead of proliferation over the entire well, somewhat similar to a wound healing experiment. It also insures that there is no contribution to the impedance measurements due to cell growth on the reference electrode. Media was changed each day and the micrographs marked (a–d) were taken at the specified time points. The experiments were stopped when the cell confluence was observed by optical microscopy inspection. The inserts were not placed at precisely the same distances from the working electrode, thus the times for the seeded cells growing and reaching the electrodes are not exactly equal for both the experiments. The observed changes in the measured impedance for both electrode systems correspond to the cells covering the electrode area. Changing the media every 24 h reduced the effects of variation in pH. However, each time media was changed (indicated by the vertical dotted lines), the
672
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
Fig. 6. (a) Impedance difference over frequency spectrum. (b) Curve fit results for the cell model curve fit from 25 h onwards (when the model becomes valid) and also shows results for R s . (c) Biological cell electrical model (red) in series with the electrode impedance (blue). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
electrochemical balance of the electrode was momentarily affected. PEDOT:PSS electrodes recovered faster from these effects (as seen in Fig. 5) than Au electrodes. This is most probably due to the fast charge transfer at the electrode interface. At lower frequencies the electrochemical variations affect Au electrodes more than what is observed with the PEDOT:PSS electrodes and can reduce the reliability of the measurements. Results of Fig. 5(a) show, that after approximately 30 h we find that the real part of the impedance at the lower frequency (40 Hz) keep increasing but the phase part settles. This shows that at these frequencies the change is mostly resistive. At higher frequencies the real part settles but the phase part keeps changing which indicates a reactive behavior. The variation of changes between the two frequencies shows how the impedance at different frequencies provides different type of information. Fig. 6(a) shows how the impedance spectrum changed due to cell coverage in the PEDOT:PSS electrode experiment shown in Fig. 5. The impedance and phase values shown are not the actual values measured but the difference from the measurements taken after 1 h of beginning the experiment. This helps to negate any changes due to initial electrochemical variations. By taking the difference from a time point where the electrodes have no cell coverage we assume that the impedance due to cells is in series to the electrode impedance electrical model. Qiu et al. [10] have used a similar approach to measuring the cell–substrate distance for cardiomyocytes, where the cells were modeled by a simple resistor in parallel to a capacitor. The disadvantage of such an assumption is that the model is not valid for partial electrode coverage and holds true once the cells have covered the majority of the electrode. X. Huang et al. [42] have discussed a model that considers the cell coverage of the electrode but their model does not consider the effects of phase changes. Our approach to this was to keep the model simple and use the theory of electrical elements similar to Qiu et al. but with a certain addition. Fig. 6(a) shows that at 31 h the |Z| spectrum is similar to the impedance spectrum of a parallel resistor and capacitor circuit (RC circuit). However an inspection of the shows that the phase does not match a parallel RC circuit. Instead the results show a closer match to a resistor in parallel to a capacitor with both in series to another resistance, Rs . This series resistance can be theoretically explained as a constriction of current in the small gap under the cell similar to what Giaever and Keese have discussed
[2]. Rcell can be attributed to the resistance to current flow between gaps of adjacent cells and Ccell is due to the capacitive effect of cells adhering to the surface. We have used this simple parallel RC circuit in series with a resistance to model the cell (as shown in Fig. 6(c) and mathematically in Eqs. (2) and (3)), and further analyze the results using curve fitting techniques similar to that applied for the electrode analysis. Zcell = Rs +
1 1/Rcell + (2ωCcell )
ZTotal = Zelectrode + Zcell
(2) (3)
Fig. 6(b) shows the results of the curve fit for changes after 25 h. Rs can be found directly as |Z| at higher frequencies. This is because at higher frequencies the majority of the current will flow through Ccell which will provide the path of least impedance and thus the impedance seen at these frequencies is due to Rs alone. We have then used this Rs value in the curve fit for the cell model. Initially the parallel RC model does not exist as the cell coverage of the electrode does not exist or is not enough to have a significant effect on the curve fit. As the impedance increases, the curve fit results in values for Rcell , which follows changes in the impedance results shown in Fig. 5(a) for 40 Hz. Values for Ccell can be estimated at this point as well and after approximately 30 h the value of Ccell reaches a value around 2.5 nF and is very slowly increasing toward 3 nF. Rs has a constant increase and at just before 30 h, it starts increasing with a faster rate once the cells begin covering the electrode. Rcell keeps increasing and this would show that the cell–cell gap is tightening (increasing confluency). This shows that once the electrode has been fully covered and the cells create a confluent layer, Rcell has a more significant contribution to the impedance changes. The increase in Rcell and Rs can be portrayed as a decrease in cell–substrate distance [10,43] and tighter confluent layers. The sensitivity of the system can be defined by Eq. (4) where Z is the change in impedance due to cells. If ZElectrode Z then the sensitivity of the system is minute. Sensitivity = 1 +
Zelectrods Zelectrods + Z
(4)
From the observations above we can assume a maximum Z of 40 k at 40 Hz. Assuming a single cell is a circular disk with radius 15 m (an area of approximately 7 × 10−6 cm2 ) means that for an
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
electrode with an area of approximately 2 × 10−3 cm2 , to be fully covered would require approximately 285 cells. Thus, every two cells contribute 280.7 to the change in impedance. For the experiments shown in Figure 5 the PEDOT:PSS electrodes have |Z| of 23 k and the gold electrodes have |Z| of 170 k at 40 Hz which can be considered as Zelectrode for each case. Thus, for two cells covering the electrode the sensitivity would be 0.012 and 0.00164 respectively. This shows the improvement in sensitivity of PEDOT:PSS electrodes due to the significantly lower interfacial impedance which would also allow the impedance measurement instrument to work well above its SNR threshold. The added improvement of reaching electrochemical stability faster than gold electrodes would also mean reduction in errors in the measured impedance. 4. Conclusions We have developed a method for fabricating devices for cell impedance spectroscopy with conducting polymer (PEDOT:PSS) electrodes using all solution based materials and photolithographic techniques. The lower interfacial impedance of PEDOT:PSS electrodes improves the sensitivity of measurements and allows the electrodes to reach electrochemical stability faster. This provides reliable measurements at the lower frequencies of the spectrum and smaller electrode dimensions with low impedance. We have also modeled both PEDOT:PSS and Au electrodes and compared the values of the elements which are consistent with the known electrochemical properties of the materials. By modeling the changes due to cell growth we have shown better understanding of the cell culture and capabilities of cell impedance spectroscopy. The loss of adhesion of PEDOT:PSS to the substrate, due to water uptake, is one of the drawbacks but can be negated by careful design (using an insulation layer in our case) or by varying the properties of the PEDOT layer, such as doping used or the coating method (thus varying the physical nature of the material). How this affects the overall characteristics as an electrode would need to be studied. Due to the simple and easy aqueous based processing of the PEDOT:PSS, there are possibilities of developing cost effective CIS devices that benefit from higher transparency. Lab-on-chip devices can also be upgraded with the use of conducting polymer coatings on their electrodes, fabricated using techniques such as ink jet printing. The successful application of a conducting polymer for cell impedance spectroscopy as shown here means that it is worth studying further for its useful properties and new possibilities in biosensors. Acknowledgments We would like to thank the Glasgow Research Partnership in Engineering (GRPe) for financial support. We also wish to thank AGFA and DKSH for kindly providing samples of the AGFA Orgacon product line. References [1] C.D. Chin, V. Linder, S.K. Sia, Commercialization of microfluidic point-of-care diagnostic devices, Lab on a Chip 12 (12) (2012) 2118–2134. [2] I. Giaever, C.R. Keese, Micromotion of mammalian cells measured electrically, Proceedings of the National Academy of Sciences of the United States of America 88 (1991) 7896–7900. [3] C.R. Keese, I. Giaever, A biosensor that monitors cell morphology with electrical fields, Engineering in Medicine and Biology Magazine, IEEE 13 (1994) 402–408. [4] C.R. Keese, K. Bhawe, J. Wegener, I. Giaever, Real-time impedance assay to follow the invasive activities of metastatic cells in culture, BioTechniques 33 (2002) 842–844, 846, 848–850. [5] T.M. Curtis, M.W. Widder, L.M. Brennan, S.J. Schwager, W.H. van der Schalie, J. Fey, N. Salazar, A portable cell-based impedance sensor for toxicity testing of drinking water, Lab on a Chip 9 (2009) 2176–2183. [6] F. Xie, Y. Xu, L. Wang, K. Mitchelson, W. Xing, J. Cheng, Use of cellular electrical impedance sensing to assess in vitro cytotoxicity of anticancer drugs in a human kidney cell nephrotoxicity model, Analyst 137 (2012) 1343–1350.
673
[7] M. Angstmann, I. Brinkmann, K. Bieback, D. Breitkreutz, C. Maercker, Monitoring human mesenchymal stromal cell differentiation by electrochemical impedance sensing, Cytotherapy 13 (2011) 1074–1089. [8] C. Hildebrandt, H. Büth, S. Cho, Impidjati, H. Thielecke, Detection of the osteogenic differentiation of mesenchymal stem cells in 2D and 3D cultures by electrochemical impedance spectroscopy, Journal of Biotechnology 148 (2010) 83–90. [9] S. Reitinger, J. Wissenwasser, W. Kapferer, R. Heer, G. Lepperdinger, Electric impedance sensing in cell–substrates for rapid and selective multipotential differentiation capacity monitoring of human mesenchymal stem cells, Biosensors and Bioelectronics 34 (2012) 63–69. [10] Y. Qiu, R. Liao, X. Zhang, Real-time monitoring primary cardiomyocyte adhesion based on electrochemical impedance spectroscopy and electrical cell–substrate impedance sensing, Analytical Chemistry 80 (2008) 990–996. [11] S. Haas, H.-G. Jahnke, M. Glass, R. Azendorf, S. Schmidt, A.A. Robitzki, Real-time monitoring of relaxation and contractility of smooth muscle cells on a novel biohybrid chip, Lab on a Chip 10 (2010). [12] G.A. Snook, P. Kao, A.S. Best, Conducting-polymer-based supercapacitor devices and electrodes, Journal of Power Sources 196 (2011) 1–12. [13] J. Bobacka, Conducting polymer-based solid-state ion-selective electrodes, Electroanalysis 18 (2006) 7–18. [14] M. Gerard, A. Chaubey, B.D. Malhotra, Application of conducting polymers to biosensors, Biosensors and Bioelectronics 17 (2002) 345–359. [15] R. McNeill, D. Weiss, D. Willis, Electronic conduction in polymers. IV. Polymers from imidazole and pyridine, Australian Journal of Chemistry 18 (1965) 477–486. [16] R.M.a.D.W.B.A. Bolto, Electronic conduction in polymers. III. Electronic properties of polypyrrole, Australian Journal of Chemistry 16 (1963) 1090. [17] G. Inzelt, Conducting Polymers A New Era in Electrochemistry, Springer Verlag, Berlin, 2008. [18] C.K. Chiang, C.R. Fincher Jr., Y.W. Park, A.J. Heeger, H. Shirakawa, E.J. Louis, S.C. Gau, A.G. MacDiarmid, Electrical conductivity in doped polyacetylene, Physical Review Letters 39 (1977) 1098–1101. [19] N.K. Guimard, N. Gomez, C.E. Schmidt, Conducting polymers in biomedical engineering, Progress in Polymer Science 32 (2007) 876–921. [20] A. van Dijken, A. Perro, E.A. Meulenkamp, K. Brunner, The influence of a PEDOT:PSS layer on the efficiency of a polymer light-emitting diode, Organic Electronics 4 (2003) 131–141. [21] L.J. del Valle, F. Estrany, E. Armelin, R. Oliver, C. Alemán, Cellular adhesion, proliferation and viability on conducting polymer substrates, Macromolecular Bioscience 8 (2008) 1144–1151. [22] L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J.R. Reynolds, Poly(3,4ethylenedioxythiophene) and its derivatives: past, present, and future, Advanced Materials 12 (2000) 481–494. [23] M. Dietrich, J. Heinze, G. Heywang, F. Jonas, Electrochemical and spectroscopic characterization of polyalkylenedioxythiophenes, Journal of Electroanalytical Chemistry 369 (1994) 87–92. [24] F. Jonas, L. Schrader, Conductive modifications of polymers with polypyrroles and polythiophenes, Synthetic Metals 41 (1991) 831–836. [25] T.S. Hansen, K. West, O. Hassager, N.B. Larsen, Highly Stretchable, Conductive polymer material made from poly(3,4-ethylenedioxythiophene) and polyurethane elastomers, Advanced Functional Materials 17 (2007) 3069–3073. [26] T.S. Hansen, K. West, O. Hassager, N.B. Larsen, Integration of conducting polymer network in non-conductive polymer substrates, Synthetic Metals 156 (2006) 1203–1207. [27] H. Yamato, M. Ohwa, W. Wernet, Stability of polypyrrole and poly(3,4ethylenedioxythiophene) for biosensor application, Journal of Electroanalytical Chemistry 397 (1995) 163–170. [28] M.T. Elin, M.L.M. Asplund, T.A. Nyberg, O.W. Inganäs, H.v. Holst, Stability of poly(3,4-ethylene dioxythiophene) materials intended for implants, Journal of Biomedical Materials Research, Part B: Applied Biomaterials 9999 (2010), NA. [29] S.F. Cogan, Neural stimulation and recording electrodes, Annual Review of Biomedical Engineering 10 (2008) 275–309. [30] S. Venkatraman, J. Hendricks, Z.A. King, A.J. Sereno, S. Richardson-Burns, D. Martin, J.M. Carmena, In vitro and in vivo evaluation of PEDOT microelectrodes for neural stimulation and recording, IEEE Transactions on Neural Systems and Rehabilitation Engineering 19 (2011) 307–316. [31] R.A. Green, N.H. Lovell, G.G. Wallace, L.A. Poole-Warren, Conducting polymers for neural interfaces: challenges in developing an effective long-term implant, Biomaterials 29 (2008) 3393–3399. [32] M.A. Reza, J.M. Corey, D.R. Kipke, D.C. Martin, et al., Conducting-polymer nanotubes improve electrical properties, mechanical adhesion, neural attachment, and neurite outgrowth of neural electrodes, Small (2010), NA. [33] M.A. Reza, K.A. Ludwig, T.C. Marzullo, D.C. Martin, D.R. Kipke, et al., Interfacing conducting polymer nanotubes with the central nervous system: chronic neural recording using poly(3,4-ethylenedioxythiophene) nanotubes, Advanced Materials (2009), NA. [34] M.R. Abidian, D.H. Kim, D.C. Martin, Conducting-polymer nanotubes for controlled drug release, Advanced Materials 18 (2006) 405–409. [35] J.E.B. Randles, Kinetics of rapid electrode reactions, Discussions of the Faraday Society 1 (1947) 11–19. [36] A.J. Bard, L.R. Faulkner (Eds.), Electrochemical Methods:Fundamentals and Applications., John Wiley & Sons, New York, 2001. [37] D.A. Borkholder, Cell Based Biosensors Using Microelectrodes, Department of Electrical Engineering, Stanford University, 1998.
674
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
[38] J.-B. Jorcin, M.E. Orazem, N. Pébère, B. Tribollet, CPE analysis by local electrochemical impedance spectroscopy, Electrochimica Acta 51 (2006) 1473–1479. [39] E.T. McAdams, A. Lackermeier, J.A. McLaughlin, D. Macken, J. Jossinet, The linear and non-linear electrical properties of the electrode–electrolyte interface, Biosensors and Bioelectronics 10 (1995) 67–74. [40] J. Bobacka, A. Lewenstam, A. Ivaska, Electrochemical impedance spectroscopy of oxidized poly(3,4-ethylenedioxythiophene) film electrodes in aqueous solutions, Journal of Electroanalytical Chemistry 489 (2000) 17–27. [41] G.R. Hernández-Labrado, R.E. Contreras-Donayre, J.E. Collazos-Castro, J.L. Polo, Subdiffusion behavior in poly(3,4-ethylenedioxythiophene): Polystyrene sulfonate (PEDOT:PSS) evidenced by electrochemical impedance spectroscopy, Journal of Electroanalytical Chemistry 659 (2011) 201–204. [42] H. Xiaoqiu, N. Duc, D.W. Greve, M.M. Domach, Simulation of microelectrode impedance changes due to cell growth, IEEE Sensors Journal 4 (2004) 576–583. [43] S. Cho, H. Thielecke, Electrical characterization of human mesenchymal stem cell growth on microelectrode, Microelectronic Engineering 85 (2008) 1272–1274.
Biographies Affar Karimullah received his bachelor of engineering degree in electronics from National Institute of Science and Engineering (NUST), Pakistan in 2007. He received
his masters of sciences in electrical and electronics engineering from University of Glasgow in 2009 and is currently pursuing a doctoral degree at the same institute. His research focus is on application of organic electronic material in bioelectronics. David R.S. Cumming (SMIEEE, FIET, B.Eng., Glasgow and Ph.D. Cambridge), holds the Chair of Electronic Systems at the University of Glasgow. He is Director of the Electronics Design Centre for Heterogeneous Systems and leads the Microsystems Technology Group in the School of Engineering. His research has led to numerous industry partnerships and commercialization of CMOS sensor arrays for genome sequencing. Mathis Riehle has been the Director of the Centre for Cell Engineering since 2004. He did his PhD in Biology at the J W Goethe University Frankfurt in the group for Cinematic Cell Research with Prof. Bereiter-Hahn, followed by postdoctoral work in the interdisciplinary group of Prof. Adam Curtis and Prof Chris Wilkinson, which in 1997 became the Centre for Cell Engineering. In 2000 he took on a lectureship in Cell Engineering at the University of Glasgow and became Reader in 2006. Nikolaj Gadegaard graduated from the University of Copenhagen with a BSc in chemistry (1995), MSc in physics (1998) and a PhD in biophysics (2002). He took up a PDRA position in the Centre for Cell Engineering (University of Glasgow, 2002) and held a personal research fellowship from the Royal Society of Edinburgh (2003–2006) and is currently a Reader at University of Glasgow. He is an academic member of the Centre for Cell Engineering and is heading the fabrication activities within the group. His current research is focused on nanostructured surfaces by electron beam lithography for cell and tissue engineering applications.
Contents lists available at SciVerse ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Development of a conducting polymer cell impedance sensor Affar S. Karimullah a,∗∗ , David R.S. Cumming b , Mathis Riehle c , Nikolaj Gadegaard a,∗ a
Division of Biomedical Engineering, University of Glasgow, Glasgow G12 8LT, UK Division of Electronics and Nanoscale Engineering, University of Glasgow, Glasgow G12 8LT, UK c Centre for Cell Engineering, University of Glasgow, Glasgow G12 8QQ, UK b
a r t i c l e
i n f o
Article history: Received 15 June 2012 Received in revised form 7 September 2012 Accepted 22 September 2012 Available online 5 October 2012 Keywords: PEDOT Conducting polymer electrodes MDCK cell culture Cell impedance spectroscopy
a b s t r a c t Research in to label free methods for biological analysis have brought interesting developments. Cell impedance spectroscopy has been one of the promising outcomes. Here we show the development of an 8-well impedance measurement setup and studied the use of conducting polymers as electrode material in cell impedance spectroscopy. We have developed devices using PEDOT:PSS electrodes and shown its advantages (lower impedance and faster to reach electrochemical equilibrium) over conventional materials, such as gold. It is observed through electrochemical analysis that the lower interfacial impedance is due to the low charge transfer resistance of PEDOT:PSS. MDCK cell proliferation experiments were performed using both types of electrode materials to provide a comparative result. We applied electrical modeling methods to understand the cell–substrate interactions and shown its applications in cell impedance spectroscopy. This study presents the development and advantages of cell impedance spectroscopy using conducting polymer electrodes. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The need for label free and real time analysis in cell biology is constantly increasing. The nature of cell biological research requires parallel experiments in large numbers that yield cumulative statistical results. Research to meet these demands has led to many technological advances in fields such as lab-on-chip devices and instrumentation that process large work intensive experiments in a factory line style. Most lab-on-chip devices still require cells to be labeled before analysis [1]. Novel ways of detection along with parallel processing are required which do not interfere with the cells in culture. Cell impedance spectroscopy (CIS), a method pioneered by Giaever and Keese [2], provides such a possibility. CIS has been an evolving label-free tool for over two decades. It allows cells to be monitored in vitro and data can be collected for multiple experiments in parallel. Early research showed that the method could not only be applied for cell proliferation and motility measurements, but also be used to assess morphological characteristics (confluence of the cells) and behavioral aspects (such as metastasis) of the cells [3,4]. The applications and research for CIS have been growing rapidly including in field toxicological studies [5], drug testing [6], measuring mesenchymal stem cell
∗ Corresponding author. Tel.: +44 141 3305243. ∗∗ Corresponding author. Tel.: +44 141 3306691. E-mail addresses: [email protected] [email protected] (N. Gadegaard).
(A.S.
Karimullah),
0925-4005/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2012.09.075
differentiation [7–9], the measurement of cell substrate separation [10] and the monitoring of relaxation and contractility of muscle cells [11]. Electrodes for CIS have predominantly been based on noble metals. Conducting polymers have been researched thoroughly for antistatic coatings, electrodes (electrochemical capacitors [12], electrochromic devices, ion sensors [13]) and have been instrumental in new forms of biosensors [14]. Conducting polymers were first reported in the mid 20th century [15–17] but it was the discovery by Shirakawa et al. in 1977 that is considered the highlight for their discovery of polyacetylene [18] and earned him, along with Heegar and MacDiarmid, the Nobel Prize in chemistry in 2000. The application of conducting polymers in biological research are numerous and they are being extensively used for tissue engineering, neural probes, biosensors, drug delivery and actuators [19]. Commercially available conducting polymers, such as poly(3,4-ethylenedioxythiophene) (PEDOT), are commonly used as conductive coatings and intermediate layers in organic electronics [20]. PEDOT has the advantage of being transparent and biocompatible [21] while being considered the most stable (in the atmosphere and oxidizing environments) conducting polymer currently available [22–24]. It allows for simple and flexible methods for patterning, and also has the ability to be blended with other polymer compounds to create novel materials with unique properties [25,26]. PEDOT has been proven to have good electrochemical stability in phosphate buffer solutions, even when polarized [27,28]. It has been studied as a neural electrode coating and is considered one of the promising new materials with higher charge injection limits
668
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
and better signal to noise ratio in measurements of neural activity [29–31]. PEDOT electrodes tend to delaminate over a period of a few weeks when in a saline solution [28], however Reza et al. found PEDOT nano tube based coatings to have improved adhesion of the coating to the substrate and the lower impedance provides better neural signal recordings [32]. They also found that the material improved the neural attachment and neurite outgrowth [33] and by coating a electrospun biodegradable polymer (loaded with drugs) with PEDOT, they were able to use electrical stimulations to release these drugs [34]. Most CIS setups work at frequencies where the higher impedance of nobel metal electrodes is of little concern. However lower frequencies can contain information useful for curve fit analysis and possibly distinguish the cell type. Not only can PEDOT as an electrode material for CIS provide a lower interfacial impedance but due to its transparency cells are accessible to high resolution microscopy, which makes this a uniquely suitable material for this application. This paper presents development of a cell impedance measurement device that utilizes PEDOT:PSS (solution processable PEDOT with poly(styrenesulfonate) (PSS) as doping agent) as the electrode material. The devices were fabricated using photolithographic methods. A complete impedance measurement setup was implemented in Labview which provides more flexibility in the device design and data analysis. A multiplexer was also designed and implemented for a multi well device that allows to measure eight experiments in parallel. As the design has the same electrode connection lengths, ensuring, that results for all the wells are not impacted by variations in inductance and series resistance of the material. The design shows the possibilities of creating simple and cost effective polymer based cell impedance measurement electrode devices. measurements with 4-(2-hydroxyethyl)-1Impedance piperazineethanesulfonic acid (HEPES) saline solution showed a lower interfacial impedance of PEDOT:PSS electrodes in comparison to similar sized and shaped Au electrodes, highlighting the electrical advantage of PEDOT:PSS as an electrode material. Through electrochemical impedance analysis and curve fitting of the results to Randle’s theoretical model [35] of an electrode–electrolyte interface, we determined that the lower charge transfer resistance of PEDOT:PSS as compared to Au is the reason for the reduced interfacial impedance. Experiments with the epithelial Madin-Darby Canine Kidney (MDCK) cell line showed that the PEDOT:PSS electrodes were fully capable of being used for CIS. Due to the fast charge transport capabilities of PEDOT:PSS, the electrodes quickly reach their equilibrium state. Curve fitting of the experimental data to typical electrical models of biological cells in series with the electrode equivalent circuit, show the versatility of CIS as a method to observe not only cell growth but also the cell–substrate interaction.
2. Materials and methods Commercially available poly (3,4 ethylenedioxythiophene) poly(styrenesulfonate) (Orgacon S305 plus, AGFA) was used. It is a 0.54% by weight aqueous solution of PEDOT:PSS. The S305 plus includes binders to help adhesion to substrates and stabilizers for improved environmental stability. SU8-3005 (Microchem, Newton, MA, USA) and Microposit S-1818 (Shipley, Coventry, UK) photo resists were used for fabrication. EC solvent (Ethyl Lactate, Microposit, Shipley, Coventry, UK) was used to develop SU-8, and MicroDev (Microposit, Shipley, Coventry, UK) was used for S-1818 development after a 1:1 dilution with RO water. The glass substrates used were 50 mm × 75 mm microscope slides (Corning, Amsterdam, The Netherlands). Acetone, methanol and iso-propanol (Sigma Aldrich, Dorset, UK) were used for cleaning the
substrates and devices at various steps. Wells were created using -slide 8-well culture dishes (Ibidi, Munich, Germany). MDCK II cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with Medium 100, fetal bovine serum (FBS), sodium pyruvate and antibiotic all acquired from GIBCO, Invitrogen. Prior to cell seeding, the device wells were sterilized with 70% ethanol and rinsed with HEPES saline solution before being coated with poly-L-lysine (PLL Mol. Wt. 150,000–300,000, cell culture tested, Sigma, Poole, UK). Incubation during the cell culture and measurements was done at 37 ◦ C in a 5% CO2 environment. In certain experiments cell culture inserts (Ibidi, Munich, Germany) were used to seed cells in a specific area in contrast to a full coverage of the culture well areas. 2.1. Instrumentation and software Impedance measurements for the electrode comparison were taken using a QuadTech 1420 LCR meter (from 20 Hz to 1 MHz; Quadtech, Marlborough, MA, USA). Cell impedance measurements were done using an Agilent 4294A impedance analyzer (Agilent, Berkshire, UK). The impedance measurement equipment was connected to the wells through a multiplexer. The multiplexer was designed with four 1:16 channel switches (two ADG726, Analog Devices). By keeping the multiplexers separate for each of the four current and voltage probe lines, the configuration further negates any effects due to the multiplexer. The ADG726 package has an “ON” resistance of only 4 and does not mitigate the accuracy of the system. A NI-DAQ 6211 (National Instruments, Berkshire, UK) provided the digital signals to the multiplexers and a single power supply was used to provide positive and negative voltage to linear voltage regulators, which feed the rest of the circuitry. The multiplexer board holds the device and both sit inside an incubator. The setup is controlled using software designed on Labview (National Instruments, Berkshire, UK). The amplitude of the signals used for the electrochemical comparison was 20 mV, which is low enough to work in the linear region of the current and over-potential relationship [36,37]. The amplitude for CIS measurements was kept higher (80 mV) which was set after it was observed to be free of any noise when measured using both types of electrodes. It is small enough to be free of harmonic distortion and neither caused any adverse effect to the cell culture. 2.2. Device fabrication A single device had eight wells, arranged in four columns and two rows. Each well had a pair of electrodes: a working electrode and a reference electrode. The design of the electrodes was similar to those available from Applied Biophysics ECIS systems with a large reference electrode and a circular working electrode whose dimensions and geometry are defined by the windows created in the insulation layer. The devices (Fig. 1) were fabricated using photolithography on a glass substrate. The glass substrates were cleaned using acetone, methanol and propane-2-ol, 5 min in an ultrasonic bath, followed by oxygen plasma cleaning (GaLa Instrumente, Bad Schwalbach, Germany) for 3 min at 160 W. An alignment layer was deposited on the substrate via a lift off process involving S-1818 lithography and aluminum deposition by evaporation using a Plassys metal evaporator (Plassys, Marolles En Hurepoix, France). Fig. 1 shows the fabrication process for the PEDOT:PSS electrodes and the SU-8 insulation layer. PEDOT:PSS was spin coated at 1000 RPM for 30 s onto the substrate. This was done twice to achieve approximately 100 nm thickness. The sample was baked at 95 ◦ C for 1 min between the two layers and for 5 min after the final coating. S-1818 photo resist was then spin coated and baked at 95 ◦ C for 5 min. After lithography and development (90 s) of S-1818, PEDOT:PSS was exposed to 5% sodium
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
669
Fig. 1. (Left) Schematic of a single row of the CIS device. Each row has two pairs of electrodes (one pair is shown inside a well). The working electrode has a window in the nearly transparent SU-8 insulation layer. (Right) Schematic of the fabrication of PEDOT:PSS devices. (A) Substrate with PEDOT:PSS layer. (B) S-1818 layer on top. (C) S-1818 is patterned and developed. (D) PEDOT:PSS electrode formed by oxidizing with bleach, rinsing in RO water and removing S1818 with acetone. (E) SU-8 coating and (F) Window created in SU-8 over electrodes and points of contact.
hypochlorite solution for 10 s and rinsed in running RO water till the oxidized PEDOT:PSS was completely removed. The sample was cleaned in acetone, propan-2-ol, and an oxygen plasma (1 min at 60 W) then spin coated with SU8-3005 and baked at 95 ◦ C for 10 min. The SU-8 was patterned using photolithography and given a post exposure bake before development in EC solvent for less than 1 min. A final bake at 150 ◦ C for 5 min was performed before an oxygen-plasma cleaning (1 min at 80 W). Wells were attached using silicone sealant and silver paint was applied to the PEDOT:PSS electrodes at the end where the connectors would interface with them. This provided some mechanical strength so that the PEDOT:PSS layer was not scratched by the metal connector pins. The Au electrode devices were made by photolithography using S-1818 (inverted pattern to the PEDOT:PSS electrode mask). After development, 2 nm of Ti was deposited for adhesion purposes followed by 50 nm of Au deposited using a Plassys metal electron beam evaporator. The S-1818 was removed with acetone and SU-8 was spun onto the device and patterned similar to the PEDOT:PSS devices. A final bake at 150 ◦ C for 5 min and plasma cleaning was performed as well. 3. Results and discussion 3.1. Electrode material comparison We investigated the differences between using PEDOT:PSS and Au as electrode materials for CIS. The basic differences between the two in their physical characteristics are already understood, but the electrical behavior of the two electrode materials under cell culture conditions needed to be studied further. Earlier research had shown the benefits of PEDOT:PSS for its charge injection capabilities and its electrochemical behavior. Here we would like to assimilate those studies and use them to compare the electrochemical characteristics of the two materials. Impedance spectroscopy results of Au and PEDOT:PSS electrodes in HEPES saline solution were used to evaluate their electrical impedance model values. The electrodes were of the same dimensions, with a working electrode area of 4.15 × 10−4 cm2 (230 m in diameter) and the reference electrode area being 6 × 10−2 cm2 , which is sufficiently large to have minimal contribution on the impedance values measured. The equivalent circuit model considered is essentially Randles’ circuit (with a Warburg element) which is considered to be the simplest model that provides a good interpretation of an electrochemical cell catering for both the double layer capacitance and the Faradaic components [35]. The model we used varied slightly as it had a frequency dependent double layer capacitance that is modeled mathematically as a constant phase element [38] and provided the best curve fits. This constant phase element behavior could rise due to surface
roughness of PEDOT:PSS, distribution of reaction rates with the many chemical compounds present in HEPES saline, and nonuniform current distribution paths as the electrodes are not directly facing each other [36,39]. Complex non-linear least square (CNLS) curve fitting was used to find the values for the double layer capacitance ‘Q’ and ‘n’ (modeled as a constant phase element, where Cdl = Q(jω)n−1 and Zcpe = 1/Q(jω)n ), charge transfer resistance ‘Rt ’, series resistance of the bulk solution and electrodes ‘Rs ’ and the Warburg impedance coefficient ‘Aw ’. Two values for Rs are shown, one that is per unit area and the other which is the actual value. This is because in the case of PEDOT:PSS, the majority of the series resistance will be due to the electrode to connector length and not necessarily relevant to the understanding of the impedance model of the electrode or bulk of the solution. It can also be argued that due to the added resistivity of the PEDOT:PSS, the actual potential drop across the electrode–electrolyte interface is not the same anymore. However, given the fact that we are already in the linear region of the current and over-potential relationship, the charge transfer resistance value will hold true even for smaller voltages. Also, given that the series resistance in the case of PEDOT:PSS is less than half the value of Rt , the potential drop across the interfacial impedance will still be larger than 10 mV and the difference between the two materials is still far too great. Eq. (1) shows the total impedance for the electrode, where ω is the angular frequency. Fig. 2 shows the comparison between absolute impedance |Z| values measured with HEPES saline using Au and PEDOT:PSS electrodes respectively. The table in Fig. 3 shows the results of curve fit to the model shown in Fig. 2 (mathematically represented in Eq. (1)). The blue lines in the plots represent impedance predicted by the theoretical model using the results of the fit. Each frequency point is marked and the graphs show a tight fit to the measured impedance values (the red dots). The results show that the Au electrodes are more polarizable and have high values for the Faradaic impedance components of the circuit. PEDOT:PSS electrodes show low Faradaic impedance and a larger double layer capacitance with lower value of ‘n’, meaning that its characteristics are further from that of an ideal capacitor. Zelectrods = Rs +
1 √ √ Q (jω) + (1/(Rt + (Aw / ω) + (Aw /j ω)))
(1)
Bobacka et al. have shown that PEDOT:PSS has fast charge transport from electrode to the electrolyte and that the polymer has an excess of supporting electrolyte which provides good charge transfer capability. Hernandez-Labrado et al. have discussed diffusion of ions into the polymer layer due to its porous nature causing a frequency dependant Warburg behavior [40,41]. The results here are in close agreement to those cases, however we must consider that the PEDOT:PSS layer here, since it has been
670
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
Fig. 2. Absolute impedance (|Z|) results of PEDOT:PSS and Au electrodes (230 m diameter) with HEPES saline solution and the interfacial impedance equivalent circuit model. Zcpe is the impedance of the frequency dependant double layer capacitance (a constant phase element). Zw is the Warburg impedance.
Fig. 3. (Left) Table and (Right) Nyquist plots of curve fit results using the model shown in Fig. 2 for the impedance of both types of electrodes. Z is the real value and Z is the imaginary value of impedance (Z) measured from 20 Hz to 1 MHz.
spin coated, is not as porous or thick as what they have polymerized electrochemically in their experiments. The possibility of excess solvent in the PEDOT:PSS electrodes is also supported by the signs of adhesion loss after prolonged exposure to aqueous solvents during our experiments. The requirements of CIS however, do not necessitate any further inquiry into this electrochemical research.
3.2. Cell experiments MDCK proliferation was measured using both Au and PEDOT:PSS electrodes. The electrodes for these experiments had an approximate area of 2 × 10−3 cm2 , that is a 500 m diameter window. Fig. 4 shows results of the CIS setup measuring all 8 wells using both PEDOT:PSS and Au electrodes. Wells 1 and 2
Fig. 4. Results of MDCK experiments using PEDOT:PSS and Au electrodes. Data at 1 kHz of all 8-wells is presented. Values shown are the changes from measurement taken at 0 h. The black vertical line shows the time point where media was changed.
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
671
Fig. 5. Results of MDCK using PEDOT:PSS electrodes (Left) and Au electrodes (Right). The graphs are the differences in absolute impedance |Z| and the phase from the initial values. The electrode windows in the SU-8 are indicated by the white rings and the dotted black vertical lines show when the media was changed. Bottom diagram shows how the culture insert is placed at a distance from the electrodes and cells are seeded inside that insert. Once the cells have adhered to the surface the insert can be removed to allow the cells to grow in an outward direction. The cells will eventually start covering the working electrode and reach confluence.
(both black lines) are control wells in each case and the rest of the wells were seeded with approximately 8 × 105 cells. The results show real time impedance changes due to the cell coverage on the electrode surface. The plots show the difference from the initial impedance values, that is |Z| and (phase angle). Initially, the cells attach to the substrate and start covering the working electrode which causes a considerable change in the phase whereas the total change to |Z| is smaller. After approximately 20 h (23 h for Au electrode results), the cell growth increases due to proliferation and cells begin to achieve confluence which causes |Z| to increase. In the case of PEDOT:PSS the phase changes, at this frequency, is increasingly negative where as the experiment with Au electrodes shows increasingly positive phase changes. This is a rather interesting variation. |Z| for both devices are constantly decreasing initially and show sudden changes right after media is change. The change after the media imitates the initial change meaning it is related to chemical changes in the media over time. This could be due to pH/chemical changes over time. Judging from observation of the control wells it is likely to be due to changes in concentration of dissolved CO2 .
Fig. 5 shows results of experiments comparing where cells were seeded in a region above the working electrodes using cell culture inserts (Fig. 5, bottom). After allowing the cells to settle, spread and migrate for 24 h, the inserts were removed and the measurements were started. This allows us to measure cell coverage of an electrode by the mean of migration in a particular direction instead of proliferation over the entire well, somewhat similar to a wound healing experiment. It also insures that there is no contribution to the impedance measurements due to cell growth on the reference electrode. Media was changed each day and the micrographs marked (a–d) were taken at the specified time points. The experiments were stopped when the cell confluence was observed by optical microscopy inspection. The inserts were not placed at precisely the same distances from the working electrode, thus the times for the seeded cells growing and reaching the electrodes are not exactly equal for both the experiments. The observed changes in the measured impedance for both electrode systems correspond to the cells covering the electrode area. Changing the media every 24 h reduced the effects of variation in pH. However, each time media was changed (indicated by the vertical dotted lines), the
672
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
Fig. 6. (a) Impedance difference over frequency spectrum. (b) Curve fit results for the cell model curve fit from 25 h onwards (when the model becomes valid) and also shows results for R s . (c) Biological cell electrical model (red) in series with the electrode impedance (blue). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
electrochemical balance of the electrode was momentarily affected. PEDOT:PSS electrodes recovered faster from these effects (as seen in Fig. 5) than Au electrodes. This is most probably due to the fast charge transfer at the electrode interface. At lower frequencies the electrochemical variations affect Au electrodes more than what is observed with the PEDOT:PSS electrodes and can reduce the reliability of the measurements. Results of Fig. 5(a) show, that after approximately 30 h we find that the real part of the impedance at the lower frequency (40 Hz) keep increasing but the phase part settles. This shows that at these frequencies the change is mostly resistive. At higher frequencies the real part settles but the phase part keeps changing which indicates a reactive behavior. The variation of changes between the two frequencies shows how the impedance at different frequencies provides different type of information. Fig. 6(a) shows how the impedance spectrum changed due to cell coverage in the PEDOT:PSS electrode experiment shown in Fig. 5. The impedance and phase values shown are not the actual values measured but the difference from the measurements taken after 1 h of beginning the experiment. This helps to negate any changes due to initial electrochemical variations. By taking the difference from a time point where the electrodes have no cell coverage we assume that the impedance due to cells is in series to the electrode impedance electrical model. Qiu et al. [10] have used a similar approach to measuring the cell–substrate distance for cardiomyocytes, where the cells were modeled by a simple resistor in parallel to a capacitor. The disadvantage of such an assumption is that the model is not valid for partial electrode coverage and holds true once the cells have covered the majority of the electrode. X. Huang et al. [42] have discussed a model that considers the cell coverage of the electrode but their model does not consider the effects of phase changes. Our approach to this was to keep the model simple and use the theory of electrical elements similar to Qiu et al. but with a certain addition. Fig. 6(a) shows that at 31 h the |Z| spectrum is similar to the impedance spectrum of a parallel resistor and capacitor circuit (RC circuit). However an inspection of the shows that the phase does not match a parallel RC circuit. Instead the results show a closer match to a resistor in parallel to a capacitor with both in series to another resistance, Rs . This series resistance can be theoretically explained as a constriction of current in the small gap under the cell similar to what Giaever and Keese have discussed
[2]. Rcell can be attributed to the resistance to current flow between gaps of adjacent cells and Ccell is due to the capacitive effect of cells adhering to the surface. We have used this simple parallel RC circuit in series with a resistance to model the cell (as shown in Fig. 6(c) and mathematically in Eqs. (2) and (3)), and further analyze the results using curve fitting techniques similar to that applied for the electrode analysis. Zcell = Rs +
1 1/Rcell + (2ωCcell )
ZTotal = Zelectrode + Zcell
(2) (3)
Fig. 6(b) shows the results of the curve fit for changes after 25 h. Rs can be found directly as |Z| at higher frequencies. This is because at higher frequencies the majority of the current will flow through Ccell which will provide the path of least impedance and thus the impedance seen at these frequencies is due to Rs alone. We have then used this Rs value in the curve fit for the cell model. Initially the parallel RC model does not exist as the cell coverage of the electrode does not exist or is not enough to have a significant effect on the curve fit. As the impedance increases, the curve fit results in values for Rcell , which follows changes in the impedance results shown in Fig. 5(a) for 40 Hz. Values for Ccell can be estimated at this point as well and after approximately 30 h the value of Ccell reaches a value around 2.5 nF and is very slowly increasing toward 3 nF. Rs has a constant increase and at just before 30 h, it starts increasing with a faster rate once the cells begin covering the electrode. Rcell keeps increasing and this would show that the cell–cell gap is tightening (increasing confluency). This shows that once the electrode has been fully covered and the cells create a confluent layer, Rcell has a more significant contribution to the impedance changes. The increase in Rcell and Rs can be portrayed as a decrease in cell–substrate distance [10,43] and tighter confluent layers. The sensitivity of the system can be defined by Eq. (4) where Z is the change in impedance due to cells. If ZElectrode Z then the sensitivity of the system is minute. Sensitivity = 1 +
Zelectrods Zelectrods + Z
(4)
From the observations above we can assume a maximum Z of 40 k at 40 Hz. Assuming a single cell is a circular disk with radius 15 m (an area of approximately 7 × 10−6 cm2 ) means that for an
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
electrode with an area of approximately 2 × 10−3 cm2 , to be fully covered would require approximately 285 cells. Thus, every two cells contribute 280.7 to the change in impedance. For the experiments shown in Figure 5 the PEDOT:PSS electrodes have |Z| of 23 k and the gold electrodes have |Z| of 170 k at 40 Hz which can be considered as Zelectrode for each case. Thus, for two cells covering the electrode the sensitivity would be 0.012 and 0.00164 respectively. This shows the improvement in sensitivity of PEDOT:PSS electrodes due to the significantly lower interfacial impedance which would also allow the impedance measurement instrument to work well above its SNR threshold. The added improvement of reaching electrochemical stability faster than gold electrodes would also mean reduction in errors in the measured impedance. 4. Conclusions We have developed a method for fabricating devices for cell impedance spectroscopy with conducting polymer (PEDOT:PSS) electrodes using all solution based materials and photolithographic techniques. The lower interfacial impedance of PEDOT:PSS electrodes improves the sensitivity of measurements and allows the electrodes to reach electrochemical stability faster. This provides reliable measurements at the lower frequencies of the spectrum and smaller electrode dimensions with low impedance. We have also modeled both PEDOT:PSS and Au electrodes and compared the values of the elements which are consistent with the known electrochemical properties of the materials. By modeling the changes due to cell growth we have shown better understanding of the cell culture and capabilities of cell impedance spectroscopy. The loss of adhesion of PEDOT:PSS to the substrate, due to water uptake, is one of the drawbacks but can be negated by careful design (using an insulation layer in our case) or by varying the properties of the PEDOT layer, such as doping used or the coating method (thus varying the physical nature of the material). How this affects the overall characteristics as an electrode would need to be studied. Due to the simple and easy aqueous based processing of the PEDOT:PSS, there are possibilities of developing cost effective CIS devices that benefit from higher transparency. Lab-on-chip devices can also be upgraded with the use of conducting polymer coatings on their electrodes, fabricated using techniques such as ink jet printing. The successful application of a conducting polymer for cell impedance spectroscopy as shown here means that it is worth studying further for its useful properties and new possibilities in biosensors. Acknowledgments We would like to thank the Glasgow Research Partnership in Engineering (GRPe) for financial support. We also wish to thank AGFA and DKSH for kindly providing samples of the AGFA Orgacon product line. References [1] C.D. Chin, V. Linder, S.K. Sia, Commercialization of microfluidic point-of-care diagnostic devices, Lab on a Chip 12 (12) (2012) 2118–2134. [2] I. Giaever, C.R. Keese, Micromotion of mammalian cells measured electrically, Proceedings of the National Academy of Sciences of the United States of America 88 (1991) 7896–7900. [3] C.R. Keese, I. Giaever, A biosensor that monitors cell morphology with electrical fields, Engineering in Medicine and Biology Magazine, IEEE 13 (1994) 402–408. [4] C.R. Keese, K. Bhawe, J. Wegener, I. Giaever, Real-time impedance assay to follow the invasive activities of metastatic cells in culture, BioTechniques 33 (2002) 842–844, 846, 848–850. [5] T.M. Curtis, M.W. Widder, L.M. Brennan, S.J. Schwager, W.H. van der Schalie, J. Fey, N. Salazar, A portable cell-based impedance sensor for toxicity testing of drinking water, Lab on a Chip 9 (2009) 2176–2183. [6] F. Xie, Y. Xu, L. Wang, K. Mitchelson, W. Xing, J. Cheng, Use of cellular electrical impedance sensing to assess in vitro cytotoxicity of anticancer drugs in a human kidney cell nephrotoxicity model, Analyst 137 (2012) 1343–1350.
673
[7] M. Angstmann, I. Brinkmann, K. Bieback, D. Breitkreutz, C. Maercker, Monitoring human mesenchymal stromal cell differentiation by electrochemical impedance sensing, Cytotherapy 13 (2011) 1074–1089. [8] C. Hildebrandt, H. Büth, S. Cho, Impidjati, H. Thielecke, Detection of the osteogenic differentiation of mesenchymal stem cells in 2D and 3D cultures by electrochemical impedance spectroscopy, Journal of Biotechnology 148 (2010) 83–90. [9] S. Reitinger, J. Wissenwasser, W. Kapferer, R. Heer, G. Lepperdinger, Electric impedance sensing in cell–substrates for rapid and selective multipotential differentiation capacity monitoring of human mesenchymal stem cells, Biosensors and Bioelectronics 34 (2012) 63–69. [10] Y. Qiu, R. Liao, X. Zhang, Real-time monitoring primary cardiomyocyte adhesion based on electrochemical impedance spectroscopy and electrical cell–substrate impedance sensing, Analytical Chemistry 80 (2008) 990–996. [11] S. Haas, H.-G. Jahnke, M. Glass, R. Azendorf, S. Schmidt, A.A. Robitzki, Real-time monitoring of relaxation and contractility of smooth muscle cells on a novel biohybrid chip, Lab on a Chip 10 (2010). [12] G.A. Snook, P. Kao, A.S. Best, Conducting-polymer-based supercapacitor devices and electrodes, Journal of Power Sources 196 (2011) 1–12. [13] J. Bobacka, Conducting polymer-based solid-state ion-selective electrodes, Electroanalysis 18 (2006) 7–18. [14] M. Gerard, A. Chaubey, B.D. Malhotra, Application of conducting polymers to biosensors, Biosensors and Bioelectronics 17 (2002) 345–359. [15] R. McNeill, D. Weiss, D. Willis, Electronic conduction in polymers. IV. Polymers from imidazole and pyridine, Australian Journal of Chemistry 18 (1965) 477–486. [16] R.M.a.D.W.B.A. Bolto, Electronic conduction in polymers. III. Electronic properties of polypyrrole, Australian Journal of Chemistry 16 (1963) 1090. [17] G. Inzelt, Conducting Polymers A New Era in Electrochemistry, Springer Verlag, Berlin, 2008. [18] C.K. Chiang, C.R. Fincher Jr., Y.W. Park, A.J. Heeger, H. Shirakawa, E.J. Louis, S.C. Gau, A.G. MacDiarmid, Electrical conductivity in doped polyacetylene, Physical Review Letters 39 (1977) 1098–1101. [19] N.K. Guimard, N. Gomez, C.E. Schmidt, Conducting polymers in biomedical engineering, Progress in Polymer Science 32 (2007) 876–921. [20] A. van Dijken, A. Perro, E.A. Meulenkamp, K. Brunner, The influence of a PEDOT:PSS layer on the efficiency of a polymer light-emitting diode, Organic Electronics 4 (2003) 131–141. [21] L.J. del Valle, F. Estrany, E. Armelin, R. Oliver, C. Alemán, Cellular adhesion, proliferation and viability on conducting polymer substrates, Macromolecular Bioscience 8 (2008) 1144–1151. [22] L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J.R. Reynolds, Poly(3,4ethylenedioxythiophene) and its derivatives: past, present, and future, Advanced Materials 12 (2000) 481–494. [23] M. Dietrich, J. Heinze, G. Heywang, F. Jonas, Electrochemical and spectroscopic characterization of polyalkylenedioxythiophenes, Journal of Electroanalytical Chemistry 369 (1994) 87–92. [24] F. Jonas, L. Schrader, Conductive modifications of polymers with polypyrroles and polythiophenes, Synthetic Metals 41 (1991) 831–836. [25] T.S. Hansen, K. West, O. Hassager, N.B. Larsen, Highly Stretchable, Conductive polymer material made from poly(3,4-ethylenedioxythiophene) and polyurethane elastomers, Advanced Functional Materials 17 (2007) 3069–3073. [26] T.S. Hansen, K. West, O. Hassager, N.B. Larsen, Integration of conducting polymer network in non-conductive polymer substrates, Synthetic Metals 156 (2006) 1203–1207. [27] H. Yamato, M. Ohwa, W. Wernet, Stability of polypyrrole and poly(3,4ethylenedioxythiophene) for biosensor application, Journal of Electroanalytical Chemistry 397 (1995) 163–170. [28] M.T. Elin, M.L.M. Asplund, T.A. Nyberg, O.W. Inganäs, H.v. Holst, Stability of poly(3,4-ethylene dioxythiophene) materials intended for implants, Journal of Biomedical Materials Research, Part B: Applied Biomaterials 9999 (2010), NA. [29] S.F. Cogan, Neural stimulation and recording electrodes, Annual Review of Biomedical Engineering 10 (2008) 275–309. [30] S. Venkatraman, J. Hendricks, Z.A. King, A.J. Sereno, S. Richardson-Burns, D. Martin, J.M. Carmena, In vitro and in vivo evaluation of PEDOT microelectrodes for neural stimulation and recording, IEEE Transactions on Neural Systems and Rehabilitation Engineering 19 (2011) 307–316. [31] R.A. Green, N.H. Lovell, G.G. Wallace, L.A. Poole-Warren, Conducting polymers for neural interfaces: challenges in developing an effective long-term implant, Biomaterials 29 (2008) 3393–3399. [32] M.A. Reza, J.M. Corey, D.R. Kipke, D.C. Martin, et al., Conducting-polymer nanotubes improve electrical properties, mechanical adhesion, neural attachment, and neurite outgrowth of neural electrodes, Small (2010), NA. [33] M.A. Reza, K.A. Ludwig, T.C. Marzullo, D.C. Martin, D.R. Kipke, et al., Interfacing conducting polymer nanotubes with the central nervous system: chronic neural recording using poly(3,4-ethylenedioxythiophene) nanotubes, Advanced Materials (2009), NA. [34] M.R. Abidian, D.H. Kim, D.C. Martin, Conducting-polymer nanotubes for controlled drug release, Advanced Materials 18 (2006) 405–409. [35] J.E.B. Randles, Kinetics of rapid electrode reactions, Discussions of the Faraday Society 1 (1947) 11–19. [36] A.J. Bard, L.R. Faulkner (Eds.), Electrochemical Methods:Fundamentals and Applications., John Wiley & Sons, New York, 2001. [37] D.A. Borkholder, Cell Based Biosensors Using Microelectrodes, Department of Electrical Engineering, Stanford University, 1998.
674
A.S. Karimullah et al. / Sensors and Actuators B 176 (2013) 667–674
[38] J.-B. Jorcin, M.E. Orazem, N. Pébère, B. Tribollet, CPE analysis by local electrochemical impedance spectroscopy, Electrochimica Acta 51 (2006) 1473–1479. [39] E.T. McAdams, A. Lackermeier, J.A. McLaughlin, D. Macken, J. Jossinet, The linear and non-linear electrical properties of the electrode–electrolyte interface, Biosensors and Bioelectronics 10 (1995) 67–74. [40] J. Bobacka, A. Lewenstam, A. Ivaska, Electrochemical impedance spectroscopy of oxidized poly(3,4-ethylenedioxythiophene) film electrodes in aqueous solutions, Journal of Electroanalytical Chemistry 489 (2000) 17–27. [41] G.R. Hernández-Labrado, R.E. Contreras-Donayre, J.E. Collazos-Castro, J.L. Polo, Subdiffusion behavior in poly(3,4-ethylenedioxythiophene): Polystyrene sulfonate (PEDOT:PSS) evidenced by electrochemical impedance spectroscopy, Journal of Electroanalytical Chemistry 659 (2011) 201–204. [42] H. Xiaoqiu, N. Duc, D.W. Greve, M.M. Domach, Simulation of microelectrode impedance changes due to cell growth, IEEE Sensors Journal 4 (2004) 576–583. [43] S. Cho, H. Thielecke, Electrical characterization of human mesenchymal stem cell growth on microelectrode, Microelectronic Engineering 85 (2008) 1272–1274.
Biographies Affar Karimullah received his bachelor of engineering degree in electronics from National Institute of Science and Engineering (NUST), Pakistan in 2007. He received
his masters of sciences in electrical and electronics engineering from University of Glasgow in 2009 and is currently pursuing a doctoral degree at the same institute. His research focus is on application of organic electronic material in bioelectronics. David R.S. Cumming (SMIEEE, FIET, B.Eng., Glasgow and Ph.D. Cambridge), holds the Chair of Electronic Systems at the University of Glasgow. He is Director of the Electronics Design Centre for Heterogeneous Systems and leads the Microsystems Technology Group in the School of Engineering. His research has led to numerous industry partnerships and commercialization of CMOS sensor arrays for genome sequencing. Mathis Riehle has been the Director of the Centre for Cell Engineering since 2004. He did his PhD in Biology at the J W Goethe University Frankfurt in the group for Cinematic Cell Research with Prof. Bereiter-Hahn, followed by postdoctoral work in the interdisciplinary group of Prof. Adam Curtis and Prof Chris Wilkinson, which in 1997 became the Centre for Cell Engineering. In 2000 he took on a lectureship in Cell Engineering at the University of Glasgow and became Reader in 2006. Nikolaj Gadegaard graduated from the University of Copenhagen with a BSc in chemistry (1995), MSc in physics (1998) and a PhD in biophysics (2002). He took up a PDRA position in the Centre for Cell Engineering (University of Glasgow, 2002) and held a personal research fellowship from the Royal Society of Edinburgh (2003–2006) and is currently a Reader at University of Glasgow. He is an academic member of the Centre for Cell Engineering and is heading the fabrication activities within the group. His current research is focused on nanostructured surfaces by electron beam lithography for cell and tissue engineering applications.