Prototyping all-polymer bioelectrical signal transducers A. Blau1*§, A. Murr2§, S. Trellenkamp3, C. Dautermann3, S. Wolff3 M. Heuschkel4, J. Wuesten5, C. Ziegler2 and F. Benfenati1 1

The Italian Institute of Technology, Dept. of Neuroscience and Brain Technologies, Via Morego 30, 16163 Genoa, Italy, www.iit.it University of Kaiserslautern, Dept. of Physics, Erwin-Schroedinger-Str. 46, 67663 Kaiserslautern, Germany, www.uni-kl.de/physik 3 Nano+Bio Center at the University of Kaiserslautern, Erwin-Schroedinger-Str. 13, 67663 Kaiserslautern, Germany, www.nbc.uni-kl.de 4 Ayanda Biosystems SA, PSE Parc Scientifique, Building C, EPFL, 1015 Lausanne, Switzerland, www.ayanda-biosys.com 5 Institute for Microtechnique Mainz GmbH, Carl-Zeiss-Str. 18-20, 55129 Mainz, Germany, www.imm-mainz.de * Corresponding author. E-mail address: [email protected]. §These two authors contributed equally to the presented work. 2

Abstract — For historical reasons, the signal transduction interface of bioelectronic devices is commonly based on metals or inorganic (semi-)conductors. This also applies to application areas where artificial components such as biomedical screening devices, in vitro microelectrode arrays and in vivo neuroprosthetics come into direct contact with biological tissue. In a proof-ofconcept microelectrode array design study, we present an alternative all-polymer approach for the low-cost fabrication of bioelectrical signal transduction devices with adjustable flexibility, electrical impedance and transparency. The fabrication process entailed three steps. Firstly, by means of a replica-moulding strategy, different types of transparent polymers were microstructured by two-level SU-8 masters to create vias for contact pads and electrodes, and indentations for interconnecting microchannels. Secondly, recesses in the insulating polymer sheets were filled with conductive polymer composites based on quasi-transparent polystyrenesulfonate doped poly(3,4-ethylenedioxythiophene) (PEDOT:PSS). In a last step, the passive microelectrode arrays were backside-insulated by a second layer of a transparent polymer. The electrical properties of the resulting polymer microelectrode arrays were characterized by impedance spectroscopy, baseline noise measurements and recordings of bioelectrical signals from acute preparations of chicken cardiomyocytes. Biocompatibility was tested with in vitro cultures of cortical neurons derived from embryonic chicken. Keywords — Neural interfaces, microelectrode arrays, replica moulding, PDMS, PEDOT:PSS electroconductive polymer.

I. INTRODUCTION Most bioelectrical transducers are based on metallic (platinum, palladium, gold, iridium) or inorganic (indium-doped tin oxide (ITO), titanium nitride (TiN), iridium oxide, carbon allotropes (nanotubes, graphene, glassy carbon, diamond), silicon) conductors or combinations thereof. Commonly, conductor traces and electrodes are generated and structured by processes such as photolithography, evaporation, etching, screen printing, micromachining or electroplating. Most of these techniques not only require access to special equipment and clean-room infrastructure, but also demand for carrier substrates that withstand the respective processing conditions. These boundary conditions impose limits onto device biocompatibility, bio-stability, flexibility, transparency and exploitation pathways for their mass fabrication [11]. Particularly in the context of designing cochlear, retinal or deep brain implants, various strategies have been suggested to overcome some of these limitations [3, 5, 8]. In addition, there is a steady trend to create more flexible devices because biological tissue is soft [7]. While insulating polymers are already accepted and widely used as elastic electrode carriers, electro-conductive polymers have not yet replaced conventional conductor materials. Although there is proof that they significantly enhance the electrical characteristics of conventional metal-based bioelectric signal transducers [1, 6, 10], there is currently no report on their use as the only transduction element in neuroprosthetics. However, recent advances in polymer material development and processing technologies propose a variety of new design and fabrication possibilities for generating flexible, all-polymeric microelectrode arrays. It can be foreseen that - similar to the recent shift from silicon to organic electronics in consumer devices - conductive polymers will complement if not substitute classical conductor materials in future biomedical and neuroprosthetic transducers.

O. Dössel and W.C. Schlegel (Eds.): WC 2009, IFMBE Proceedings 25/IX, pp. 327–330, 2009. www.springerlink.com

328

A. Blau et al.

II. MATIERIALS AND METHODS

B. Replica moulding

The individual device fabrication steps for generating a transparent, all-polymer microelectrode array are depicted in figure 1.

Degassed poly(dimethylsiloxane) (PDMS) (Dow Corning, Sylgard 184; Wacker, Elastosil RT601) was poured onto the microstructured SU-8, leveled to the highest SU-8 geometries either by spinning (30 s, 1000 – 4000 rpm) or by a laser printer transparency sheet, and allowed to cure for 6 h at 60°C or overnight at room temperature. Thus, the pad and electrode features penetrated the entire PDMS slab, while the interconnection channels between pads and electrodes were only imprinted half-height. With the help of cold ethanol, the microstructured PDMS slab was peeled from the wafer and placed upside-down onto a polytetrafluorethylene (PTFE) tray, thereby giving access to all of its cavities. C. Cavity filling with electroconductive polymers

Fig. 1 Fabrication of a moulding master (1-7), replica-moulding of an insulating polymer skeleton (8-9), definition of electroconductive polymer traces (10) and device finalization through backside insulation with a nonconductive polymer (11-12).

A. Master fabrication Photopolymers such as SU-8 allow the fabrication of high-aspect ratio, multi-level, large-area replica-moulding templates [4]. Based on SU-8 technology, one type of moulding masters was generated in the following way: the negative photoresin SU-8 100 (Microchem) was spun (Delta 80 Gyrset, Suess) onto a 4” silicon wafer with an SU-8 layer thickness of up to 250 µm and soft-baked at 65°C for 20 min, and at 95°C for 55 min. After cool-down, it was exposed to UV (120 s, 15 mW/cm2, EVG 620 mask aligner) through the first of two repro-printed photolithography masks (2480 dpi) to define pad, interconnection lead and electrode geometries of a 60-microelectrode array. The SU8 layer was then post-baked at 65°C for 10 min, followed by 95°C for 25 min, and thereafter allowed to cool down to room temperature. A second SU-8 100 layer was spun onto the first layer with a thickness of up to 250 µm, soft-baked, exposed to UV light through the second mask - this time to define pad and electrode geometries only - and again postbaked using the same parameters as for the creation of the first layer. In a subsequent development step, the nonexposed photoresin was washed away to yield a two-level microstructure with a total height of up to 500 µm.

The hydrophobic PDMS was optionally hydrophilized by O2 plasma exposure (30 s, 50 watt, PICO-UHP, Diener Electronic). The solvent of a conductive polymer nanoparticle dispersion, polystyrenesulfonate doped poly(3,4-ethylenedioxythiophene) (PEDOT:PSS) (Baytron CPP105, H.C. Starck), was partly removed in a desiccator vacuum to increase its viscosity. The cavities in the PDMS slab were filled with the thickened dispersion. Excess dispersion was wiped off from non-cavity areas using the edge of a rubber wiper blade. The dispersion was then tempered at 80-160°C for up to 60 min until completely dried as a thin, transparent, slightly bluish electroconductive film on the cavity walls and the PTFE tray. To increase the thickness and final conductivity of the resulting film, this coating procedure could be repeated several times. D. Backside insulation In a last step, the back side of the microstructured PDMS slab with the conductive polymer film in its cavities was electrically insulated by a second PDMS layer. After curing, the sandwich was peeled from the PTFE tray. Due to the “non-stick” properties of PTFE, the electroconductive polymer film of the pads and central microelectrodes was completely transferred onto the PDMS. The rigidity of such soft (” shore A 50) all-polymer microelectrode array with a thickness below 1.5 mm could be increased by bonding it to a stiff polymer carrier (e.g. poly(methyl methacrylate) (PMMA)). E. Electrical characterization Comparative impedance measurements (100 mHz - 5 kHz) were performed for individual electrode traces of a polymer microelectrode array in phosphate buffered saline (PBS) using a potentiostat (BES, PowerSuite, EG&G Applied Research). An Ag/AgCl wire

IFMBE Proceedings Vol. 25

Prototyping All-Polymer Bioelectrical Signal Transducers

served as the reference electrode, a Pt sheet (8 mm2) as the counter electrode. Noise levels on individual electrodes were evaluated using a 60-channel recording station (MEA60, Multi Channel Systems). F. Testing device performance and biocompatibility The coordinated firing of cardiac myocytes was recorded from acute preparations of embryonic chicken hearts (E8) in Hanks’ Balanced Salt Solution (HBSS). Biocompatibility and biostability of the polymer electrode arrays were evaluated over a period of two weeks on cultures of dissociated cortical chicken neurons (E10) in serum-free medium (NBM/B27, Gibco) using standard cell dissociation and cell culture protocols. III. RESULTS Figure 2 depicts a replica-molded PDMS electrode array.

329

For different electrodes, RMS noise varied between ±9 µV to ±50 µV. It is hypothesized that differences in impedance and noise may be attributed to the nonuniformity of the different conducting path geometries and to artefacts resulting from manual fabrication of the prototypes. Electrical characteristics were not affected by normal use (autoclavation, several weeks of cell culturing under standard cell culturing conditions: 5% CO2 in a humidified incubator at 37°C). However, the conductivity of a PEDOT:PSS film was completely lost when exposed to bleach (hypochlorite). While the scratch and bending stability of the thin conductive PEDOT:PSS films was high, focal pressure exerted by the amplifier contact pins onto the contact pads ruptured them. This was due to the softness of the underlying PDMS insulation layer. Pad film stability could be increased by either substituting the flexible PDMS backside insulation for a more rigid epoxy (UHU plus endfest 300) or a UVA-polymerizable acrylate-based gel (1Phasen-Gel, nail-discount-24), or by reinforcing the pads with electroconductive silver-epoxy (ITW Chemtronics) in a post-processing step. The almost constant impedance over a wide frequency spectrum allowed the recording of low-frequency components in myocardiograms from acute heart preparations; the bioelectric signals (P, Q, R, S and T wave deflections) were clearly visible (Fig. 4).

Fig. 2 Prototype of a replica-moulded PDMS microelectrode array (49 x 49 mm2) with a multilayer film of bluish, still somewhat transparent conductive polymer inlays.

The absolute impedance value of the polymer traces with electrode diameters of about 80 µm were almost constant and below 1 MOhm over the range of 0.1 Hz to 5 kHz with an average of 316 ± 176 k: (Fig. 3).

Fig. 3 Comparison of the averaged absolute impedance of Ø 80 µm PEDOT:PSS electrodes (red triangles; n = 8) with that of Ø 30 µm TiNcoated Au electrodes (green circles n = 20) on commercial microelectrode arrays (30/200iR, Multi Channel Systems) measured in phosphate buffered saline (PBS) (Ag/AgCl reference, Pt counter electrode).

Fig. 4 Acute myocardiogram (chicken E8) picked up on 19 of 60 electrodes of a polymer microelectrode array (MEA60 recording station & software, Multi Channel Systems): Low-frequency components (P, Q, R, S and T deflections) are clearly visible.

IFMBE Proceedings Vol. 25

330

A. Blau et al.

IV. CONCLUSIONS

ACKNOWLEDGMENT

In this proof-of-concept design and validation study, we presented an all-polymer approach for the low-cost fabrication of passive bioelectrical signal transduction devices with adjustable geometry, flexibility, transparency and electrical characteristics. Exemplarily, an electrode array made of a microstructured, electrically insulating PDMS carrier with embedded PEDOT:PSS conductor lines and electrodes was described. The utilized materials were biocompatible [2, 9, 10], optically transparent and thus suitable for concurrent morphology studies by inverse light microscopy. They furthermore tolerated standard sterilization procedures (e.g. autoclavation, 70% ethanol) without noticeable deterioration. Electrode array geometries were defined by a microstructured SU-8 moulding master. Its two-level design ensured that electrode and pad features within the insulating polymer substrate penetrated through the entire top layer of the array while the interconnection channels for the leads stayed covered by an insulating skin. Conceptually, the design of such moulding masters allows the pattern transfer to a variety of polymers, either along the depicted replica-moulding pathway or by hot-embossing into thermoformable polymer sheets. Furthermore, a given master can be replicated by a two-step pattern transfer to e.g. epoxies without requiring access to clean-room or microstructure processing facilities. The presented concept allows for variations in the choice of materials and the general production design. Not only can masters with different mechanical properties be fabricated (e.g. by using other microstructuring processes such as high density plasma Advanced Silicon Etching (ASE®)). Without much effort, the in vitro construct may also be translated to fabricating soft, curvature-adaptive in vivo probes (e.g. in the context of cortical surface field potential recordings, pacemakers, retinal implants). Array properties such as stiffness, transparency, hydrophilicity and surface chemistry can be tuned through combinations of matrix materials and array post-processing steps. And finally, impedance and capacitive properties of the electrodes can be further enhanced by the addition of other conductor materials (such as graphite, graphene or carbon nanotubes) to the PEDOT:PSS dispersion, however, at the cost of compromising electrode transparency. In follow-up studies, electrode dimensions will be scaled down to diameters around 10-40 µm to better compare their performance with that of commercially available electrode arrays. While the passive recording capabilities of such devices have already been proven in vitro, their performance in different voltage or current stimulation scenarios still needs to be evaluated. Taking into account the critical findings with respect to the electroactive and mechanical long-term stability of conductive polymers under physiological conditions [6], principle parameters such as the chargeinjection capacity of a PEDOT:PSS electrode, its change and its degradation over time have to be investigated.

The authors would like to thank the team at IMM for their support in converting CAD drawings to photolithography mask layouts. Special thanks to Tanja Neumann, Simone Riedel, Francesca Succol and Marina Nanni for their expert advice and assistance in tissue preparation.

REFERENCES 1.

Abidian MR, Martin DC (2008) Experimental and theoretical characterization of implantable neural microelectrodes modified with conducting polymer nanotubes. Biomaterials 29:1273-1283

2.

Blau A, Neumann T, Ziegler C, Benfenati F (2009) Replica-molded poly(dimethylsiloxane) culture vessel lids attenuate osmotic drift in long-term cell culturing. J. Biosci. 34(1)

3.

Briggs R, Tykocinski M, Saunders E, Hellier W, Dahm M, Pyman B, Clark G (2001) Surgical implications of perimodiolar cochlear implant electrode design: avoiding intracochlear damage and scala vestibuli insertion. Cochlear Implants Int 2:135-149

4.

Campo Ad, Greiner C (2007) SU-8: a photoresist for high-aspect-ratio and 3D submicron lithography. Journal of Micromechanics and Microengineering 17:R81-R95

5.

Cheung KC (2007) Implantable microscale neural interfaces. Biomed Microdevices 9:923-938

6.

Green RA, Lovell NH, Wallace GG, Poole-Warren LA (2008) Conducting polymers for neural interfaces: challenges in developing an effective long-term implant. Biomaterials 29:3393-3399

7.

Hosp JA, Molina-Luna K, Hertler B, Atiemo CO, Stett A, Luft AR (2008) Thin-film epidural microelectrode arrays for somatosensory and motor cortex mapping in rat. J Neurosci Methods 172:255-262

8.

Kim ET, Seo JM, Zhou JA, Jung H, Kirn SJ (2004) A retinal implant technology based on flexible polymer electrode and optical/electrical stimulation. 2004 IEEE International Workshop on Biomedical Circuits and Systems, pp S1/8- 12-15

9.

Peterson SL, McDonald A, Gourley PL, Sasaki DY (2005) Poly(dimethylsiloxane) thin films as biocompatible coatings for microfluidic devices: cell culture and flow studies with glial cells. J Biomed Mater Res A 72:10-18

10. Richardson-Burns SM, Hendricks JL, Foster B, Povlich LK, Kim DH, Martin DC (2007) Polymerization of the conducting polymer poly(3,4ethylenedioxythiophene) (PEDOT) around living neural cells. Biomaterials 28:1539-1552 11. Stieglitz T (2007) Neural prostheses in clinical practice: biomedical microsystems in neurological rehabilitation. Acta Neurochir Suppl 97:411-418 Author: Axel Blau Institute: The Italian Institute of Technology (IIT) Dept. of Neuroscience and Brain Technologies (NBT) Street: Via Morego 30 City: 16163 Genoa Country: Italy Email: [email protected]

IFMBE Proceedings Vol. 25