Surface-Grafted Biocompatible Polymer Conductors for Stable and Compliant Electrodes for Brain Interfaces

Abstract

Durable and conductive interfaces that enable chronic and high-resolution recording of neural activity are essential for understanding and treating neurodegenerative disorders. These chronic implants require long-term stability and small contact areas. Consequently, they are often coated with a blend of conductive polymers and are crosslinked to enhance durability despite the potentially deleterious effect of crosslinking on the mechanical and electrical properties. Here the grafting of the poly(3,4 ethylenedioxythiophene) scaffold, poly(styrenesulfonate)-b-poly(poly(ethylene glycol) methyl ether methacrylate block copolymer brush to gold, in a controlled and tunable manner, by surface-initiated atom-transfer radical polymerization (SI-ATRP) is described. This "block-brush" provides high volumetric capacitance (120 F cm^─3), strong adhesion to the metal (4 h ultrasonication), improved surface hydrophilicity, and stability against 10 000 charge-discharge voltage sweeps on a multiarray neural electrode. In addition, the block-brush film showed 33% improved stability against current pulsing. This approach can open numerous avenues for exploring specialized polymer brushes for bioelectronics research and application.

Keywords: PEDOT; SI‐ATRP; neural interface; polymer brushes; self‐assembly.

Figure 1

PSS‐b‐PPEGMEMA brushes tethered to gold surface as a backbone for PEDOT polymerization enable stable, conductive, conformal, and uniform coverage of the surface. a) Schematic Illustration showing the ECoG microelectrode connected to a flat flex cable (FFC), placed on the brain tissue. Right: The flexible electrode with zoom‐in of the block‐brush tethered to the gold contact's surface. b) The desired properties for the interface between the metal electrode and the brain tissue for long‐term and efficient charge transport during recording or stimulating brain activity, which is facilitated by strong adhesion, electronic and ionic transport, and c) high density of the brushes. The block copolymer brushes PEDOT:PSS‐b‐PPEGMEMA, are represented by the following colors, where PSS is orange, PPEGMEMA is gray, and PEDOT is blue. d) Schematic representation and the molecular structure of the PEDOT complexed with PSS‐b‐PPEGMEMA brushes, composed of the polyelectrolyte PSS and the ionic conducting elastomer PPEGMEMA. The adhesion to the surface is enabled by the Au‐S bond. e) SEM images of the cross‐section of the gold surfaces. The cross‐section verifies dense, film‐like brushes on the gold surface. The scale bar is 200 nm.

Figure 2

Electrical properties and stability tests of the adhesive films. a) The conductivity of the films including, pristine, SpinG, SpingGA, and block‐brush (n = 3). The insets indicate the water contact angle measurements of the different PEDOT‐based films on gold substrates. The block‐brush PEDOT demonstrated the lowest water contact angle among the stable films. We note that the pristine PEDOT:PSS film is not considered a stable film as it dissolves upon contact with water. b) Schematic illustration of EIS/CV three‐electrode setup. The inset shows EDL in the block‐brush polymer film. c) EIS and d) CV curves of bare gold and block‐brush film on gold before and during multiple CV cycling. The block‐brush films are stable during 3500 cycles of CV stressing (0.4 to −0.4 V) e) Characterization of the CSC of the block‐brush film versus the SpinGA film before incubation in PBS (n = 3, P = 2.85 × 10⁻⁶, 0.8 to −0.4 V). ****P < 0.0001 using one‐tailed t‐test. f) The block‐brush films show a slower decrease in CSC over 31 d of incubation in PBS at 50°. The SpinGA films show a faster decrease to 80% of the initial CSC (CSC₀) only 12 d after incubation and to 55% after 31 d of incubation.

Figure 3

Mechanical stability of the bulk films. a) Schematic illustration of the weak adhesion of spin‐coated pristine PEDOT:PSS versus the strong adhesion of block‐brush PEDOT. b) Corresponding optical microscope images of the films before (left) and after (right) ultrasonication tests. The scale bars are 100 µm. c) Corresponding Raman spectra before (blue) and after (cyan) ultrasonication, show complete PEDOT removal for the pristine sample after 2 min ultrasonication, versus the negligible difference in PEDOT spectrum for the brushes after 4 h sonication. d) 90° Peel (glass/Cr/Au/PEDOT‐based film/PI tape) test for the PEDOT‐based films. e) Average shear strength between PDMS (1:50) and PEDOT‐based films on Au/Cr/glass. Data are shown for pristine (92 ± 3 nm), SpinGA (167 ± 20 nm), brush (115 ± 7 nm), and block‐brush (180 ± 46 nm), (n = 3). f) High‐resolution optical images displaying excellent conformability of the block‐brush grown of PDMS on a soft substrate (top). The SpinGA on PDMS does not conform to the soft substrate and presents an air gap (bottom). The scale bar is 0.5 mm. g) Schematic Illustration of the AFM tip indenting block‐brush film for nanomechanical characterization. h) The Young's modulus of 1.7 ± 0.6 MPa (in water) was calculated via the Dimitriadis model, using force deformation curves and a deformation map (Figure S21, Supporting Information). i) Comparison of previously reported conductive materials‐based films with our work in terms of Young's modulus. Such conventional implantable electrical probes include silicon electrodes,^{[ 77 ]} tetrode,^{[ 78 ]} planar polyimide probes^{[ 79 ]} and flexible Au–PET` cuff electrodes^{[ 43 ]} or PEDOT formulations, which include acid‐treated PEDOT:PSS hydrogel,^{[ 51 ]} electrodeposited PEDOT,^{[ 16} , ^{17 ]} and spin‐coated PEDOT:PSS with 1% GOPS.^{[ 52 ]}

Figure 4

Electrochemical characterization of multidiameter microelectrode array and films long‐term stability. a) Image of the 32‐channel electrode array containing 100, 200, 400, and 1000 µm diameter contacts. b) Impedance spectrum and c) water window comparison between the three PEDOT‐based thin films for the 1000 µm diameter contacts. d) CIC for the three PEDOT‐based thin films for the 400 µm diameter contact. e) Breakdown of the 400 µm electrode contacts during biphasic current pulse stressing. Pulses were delivered at 50 Hz. Top left inset: Example biphasic pulse delivered during current stressing. f) Left: Focused ion beam (FIB) image of the block‐brush contact before and after 5000 CV cycles, showing no change to the film morphology. Right: Microscope images of the 1000 µm diameter contacts before (top) and after (bottom) 5000 CV cycles.

Figure 5

Block‐brush PEDOT film on microelectrode array record somatotopic functional cortical columns. (a) Schematic of the rat brain implanted with a 16‐channel, 4.8 mm‐by‐4.8 mm array, and the air puff stimulation of individual whiskers. b) Baseline recording of the brain activity, with baseline RMS values for the three materials. c) Magnified microscope image of the electrode on the rat barrel cortex. d) Spectral analysis of the mean trial‐averaged response across low‐impedance channels to whisker air puff stimulation for the block‐brush (left), electrodeposited control (middle), and SpinGA control (right) samples. Responses from the three materials showed similar spectral profiles, with onset time ≈20 ms poststimulus and high power in frequency range < 80 Hz. e) Trial‐averaged responses from six individual low‐impedance channels from each array, block‐brush (left), electrodeposited control (middle), and SpinGA control (right). The dashed line indicates the time of air puff stimulation. Responses from the three materials are similar in shape and amplitude, indicating that the block‐brush film can capture neural activity.