Engineering the ABIO-BIO interface of neurostimulation electrodes using polypyrrole and bioactive hydrogels

Materials

Chemicals and materials were purchased from Sigma Aldrich Co. (St. Louis, MO, USA) as described in earlier publications [54]. Briefly, the monomers 2-hydroxyethyl methacrylate (HEMA) (40.75 mol%), N-(2-hydroxypropyl) methacrylamide (HPMA) (40.75 mol%), poly(ethylene glycol)(360)methacrylate (PEG(360)MA) (5 mol%), N-[tris(hydroxymethyl) methyl]acrylamide (HMMA, 93%) (5 mol%), 2-Methacryloyloxyethyl phosphorylcholine (MPC) (5 mol%), the cross-linker tetra(ethylene glycol) diacrylate (TEGDA, technical grade) (1.0 mol%), the biocompatible viscosity modifier polyvinylpyrrolidone (pNVP, MW ~1 300 000) (2.0 mol%, as monomer repeat unit) and the photo-initiator 2,2-dimethoxy-2-phenylacetophenone (DMPA, 99+%) (0.5 mol%) were prepared as a cocktail. Prior to mixing, methacrylate and diacrylate reagents were passed through an activated alumina column (306312, Sigma-Aldrich Co., St. Louis, MO, USA) in order to remove the polymerization inhibitors hydroquinone and monomethyl ether hydroquinone. A Milli-Q^® plus (Millipore Inc., Bedford, MA, USA) ultrapure water system was used to prepare deionized water. The 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid sodium salt (HEPES, 99.5+%) buffer was prepared to 25 mM and the pH was adjusted to 7.4 using dropwise addition of 10 M NaOH [54]. Dulbecco’s Modified Eagle Medium (DMEM), with high glucose (4500 mg/L glucose) was used where indicated.

The conjugation reagents 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and sulfo-N-hydroxysuccinimide (sulfo-NHS) were purchased from Pierce. The polypyrrole dopant, sulfopropyl methacrylate (SPMA), was purchased from Sigma Aldrich. A Milli-Q^® plus (Millipore Inc., Bedford, MA, USA) ultrapure water system supplied deionized water. All other common chemicals and solvents were purchased from Sigma Aldrich Co. (St. Louis, MO, USA) and were used as received, unless otherwise stated. Polyimide, 8 mil thick Cirlex^® laminate were purchased as 23.5″×23.5″ sheets (Fralock, Valencia, CA, USA) and cut into 6″×6″ sheets, cleaned, plasma treated and sputtered deposited with 100 Å TiW and 1000 Å Au on one side (Thin Films Industries, Inc., Morrisville, PA, USA). These served as a source of gold-coated polyimide electrodes used for in vitro studies.

Electrode fabrication and cleaning

Electrodes, shown in Fig. 1, were rectangular, gold-coated polyimide (10×15 mm) masked with adhesive-backed polyimide tape to create a circular (φ=0.7 mm) electrodeposition window, Fig. 1a. Although polypyrrole was successfully electrodeposited onto this design, use of polyimide tape in this way produced variations between discrete electrodes. Furthermore, because the three characterization methods were inherently destructive and altered the electrode surface, three separate deposition areas per charge density performed in triplicate were required. An additional design, created in SolidWorks^® and manufactured with an LS100 Gravograph CO₂ laser at 5% power (Gravotech Inc., Duluth, GA) started with a single rectangular piece of gold-coated polyimide (7.5×20 mm) and produced three isolated deposition areas as shown in Fig. 1b. A single rectangular piece of polyimide tape (7.5×8 mm) was used as a mask to confine electrodeposition within the intended areas. With this design, polypyrrole could be deposited across all three sections at the same time, with the same charge density, then the sections could each be cut apart for use with one of the three characterization methods. Electrodes were cleaned first with DI water to remove residual dust from the laser, then with ethanol to remove organic compounds from the gold surface. Electrodes were cleaned with UV-ozone for 5 min (UV-ozone Cleaner, Boekel Industries Inc., Feasterville, PA, USA) to remove adsorbed organic debris and rinsed in isopropyl alcohol. Finally, electrodes were made the working electrode of a three-electrode setup, suspended in PBS 7.4, and processed through 20 cycles of cathodic cleaning (scan rate 0.1 V/s, potential range 0 to −1.2 V) to remove surface oxides [1].

Fig. 1:

(a) i) Design and construction of electrodes for in vitro assessments of electropolymerized charge densities of 0–1000 mC/cm². (ii) The constructed electrode after electropolymerization with a charge density of 0.1 mC/cm². (b) Schematic illustration of the multilayer design of the abio-bio interface. (c) Design of flexible, circular gold coated polyimide electrode of φ=7 mm used in animal studies.

Preparation and electrodeposition of polypyrrole

A unique pyrrole solution was formulated for electrodeposition. To aid in the crosslinking of the hydrogel layer to the electrodeposited polypyrrole film, amino ethyl methacrylate (AEMA, 0.25 M) was conjugated to pyrrolyl butyric acid (PyBA) via carbodiimide chemistry. The 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) bonds covalently to PyBA under acidic conditions, forming an unstable o-acylisourea intermediate. Then, N-hydroxysuccinimide (NHS) displaces EDC to form an ester intermediate with increased stability and efficiency for reaction with the primary amine group of AEMA. The chemistry for the conjugation of PyBA-conj-AEMA using EDC-NHS is depicted in Scheme 1. This PyBA-conj-AEMA, combined with pyrrole (Py, 0.5 M), pyrrolyl butyric acid (PyBA, 0.5 M), and the dopant sulfopropyl methacrylate (SPMA, 0.25 M) were incubated in the dark and pH adjusted to 5.0 in order to initiate electropolymerization from the pyrrole solution to yield poly(Py-co-PyBA-co-PyBA-conj-AEMA):SPMA which shall be referred to as PPy:SPMA. Electrodeposition of polypyrrole was conducted using a VerstaSTAT 4 Potentiostat/Galvanostat (Princeton Applied Research, AMETEK, Inc., Oak Ridge, TN, USA) and analyzed with VersaStudio software (10 cycles, 0 to +0.75 V). Laser fabricated gold electrodes were made the working electrode of a three-electrode setup, with a large area platinum mesh as the counter electrode and Ag/AgCl (3 M KCl) as the reference electrode (RE803, ABTECH Scientific, Inc., Richmond, VA, USA). A film of polypyrrole was deposited at seven charge densities (1, 5, 10, 30, 50, 100, 1000 mC/cm²).

Scheme 1:

Schematic illustration of the conjugation of PyBA with AEMA to yield PyBA-conj-AEMA using EDC-NHS (Adapted from [46]).

Preparation and crosslinking of hydrogel

Fluorescent light bulbs within the laboratory space were fitted with UV filtering sleeves (TG-T8TG-UV, lightbulbsurplus.com) to maintain a UV-free environment for monomer handling. The monomers HEMA, TEGDA, and PEGMA were filtered through an alumina column to remove polymerization inhibitors added during manufacturing. The following solid monomers were added to a solution of the three filtered liquid monomers in deionized water: HPMA, HMMA, MPC, pNVP (biocompatible viscosity modifier), and DMPA (photoinitiator). This cocktail was degassed with nitrogen, stirred on a stir plate, and placed in an ultrasonicator (Branson 1510, Gaithersburg, MD, USA) to ensure complete dissolution of solid monomers in the liquid solution. The hydrogel cocktail was then pipetted onto the PPy:SPMA-modified gold electrodes and crosslinked for 5 min under UV (CX-2000, UVP, Upland, CA, USA). Hydrogels were hydrated in the DMEM (Dulbecco’s Modified Eagle Medium, high glucose with 4500 mg/L glucose, Sigma-Aldrich Co., St. Louis, MO, USA)) solution which was used for impedance measurements.

Characterization of PPy:SPMA|Hydrogel electrodes

Electrodes were characterized by scanning electron microscopy (SEM), multiple scan rate cyclic voltammetry (MSRCV) and AC electrical impedance spectroscopy (EIS). To compare differences in surface morphology across the specified range of charge densities, electrodes were sputter-coated with a gold layer using a Cressington 108 Sputter Coater (Cressington Scientific Instruments, UK) and placed on a viewing mount using carbon tape, with additional carbon tape across the top of the electrodeposited area to dissipate charge while imaging. Electrodes were viewed with a JEOL JSM-7500F FE-SEM at 2.0 kV and a magnification of ×20 k. SEM images were analyzed using ImageJ software [55]. Prior to hydrogel layer attachment, PPy:SPMA-modified electrodes were characterized by multiple scan rate cyclic voltammetry (MSRCV) [56] in 50 mM potassium ferrocyanide/ferricyanide [(K₄Fe(CN)₆)/(K₃Fe(CN)₆)] in a background of 0.1 M KCl. The Au|PPy:SPMA-modified gold electrodes were made the working electrode of a three-electrode setup, with a large area platinum mesh as the counter electrode and Ag/AgCl (3 M KCl) as the reference electrode (RE803, ABTECH Scientific, Inc.). MSRCV measurements were taken across a voltage range of 0–0.75 V at variable scan rates (10, 25, 50, 75, 100, 125, 150, and 200 mV/s). Two cycles were applied to allow the system to reach a steady state, with the first cycle discarded and the second set aside for analysis using the Randles-Sevcik equation. Electrical impedance spectra were obtained with a VerstaSTAT 4 Potentiostat/Galvanostat (Princeton Applied Research, AMETEK, Inc., Oak Ridge, TN, USA) and analyzed with VersaStudio software (VersaStudio version 1.51, AMETEK, Inc., Oak Ridge, TN, USA). PPy:SPMA-modified gold electrodes were made the working electrode of a two-electrode setup, with a large surface area platinum mesh acting as both the counter and reference. Measurements were taken across a frequency range of 0.01 Hz–1 MHz at an RMS amplitude of 7.07 mV with 5 points per decade. Impedance parameters such as membrane and charge transfer resistance were obtained in ZSimpWin software version 3.60 (AMETEK, Inc., Oak Ridge, TN, USA) through equivalent circuit analysis using the Randles R[QR] model. The Au|PPy:SPMA|Hydrogel electrodes were similarly characterized. In order to most closely simulate internal body conditions, electrodes were immersed in DMEM and incubated at 37°C using a VWR2310 Water-Jacketed CO₂ Incubator (Marshall Scientific, Hampton, NH, USA) while carrying out these experiments.

Statistical analysis of impedance parameters

Impedance values for cleaned and pristine electrodes, electrodes modified following electropolymerization with polypyrrole:sulfopropyl methacrylate (PPy:SPMA), and electrodes modified by PPy:SPMA and coated with a UV-crosslinked poly(HEMA-co-HPMA-co-MPC)) hydrogel are reported as average values with n=5. Pearson’s correlation coefficient compared with charge density is reported for electroactive area, percent fracture area, charge transfer resistance, and membrane resistance, with ±0.8 representing a strong correlation.

Evolving surface morphology of polypyrrole films

Potentiostatic electropolymerization of Py, PyBA and PyBA-conj-AEMA to various charge densities (0–100 mC/cm²) produced an inherently conductive polymer layer, [P(Py-co-PyBA-conj-AEMA):SPMA], on the gold electrode. Figure 2a shows SEM image analysis of polypyrrole across seven electrodeposition charge densities (1, 5, 10, 30, 50, 100 and 1000 mC/cm²), compared to a bare gold surface. Isolated deposits of polypyrrole were observed at 1 mC/cm², with a semi-continuous film forming at 5 mC/cm². To fully exploit the electroconductive properties of polypyrrole, a uniform, homogenous film with consistent thickness and minimal surface features was anticipated. The density of such features gradually decrease with electrodeposition charge density up to 50 mC/cm². Above this charge density, a second film appears to begin to deposit on top of the first, introducing larger surface features such as those seen at 100 mC/cm². At 1000 mC/cm² saturation was observed. The plot for percent area with surface features is shown in Fig. 2b, showing the clear minimum at 50 mC/cm².

Fig. 2:

(a) SEM images of PPy:SPMA electrodeposited onto gold-sputtered polyimide via potentiostatic electropolymerization at charge densities at 0, 1, 5, 10, 30, 50, 100, and 1000 mC/cm², each taken at acceleration voltage 2.0 kV and ×20 k magnification. Surface features are identified as relief features within the polypyrrole film and appear as the darkest regions of the images in image J analysis. Scale bar 1 μm. (b) Instance of percent area with surface features within the polypyrrole (Au|PPy:SPMA) film as a function of electrodeposition charge density (mC/cm²). Features achieve a minimum at 50 mC/cm². Insert shows the structure of the PPy:SPMA polymer film.

Determination of effective electroactive area from multiple-scan rate cyclic voltammetry

Multiple-scan rate cyclic voltammetry in Fe(II)/Fe(III) serves to delineate the ability of the substrate electrode to support the one-electron charge transfer reaction associated with the reversible Fe(II)–Fe(III) couple. In this regard, performance at bare the gold electrode serves as a reference or control condition to which PPy:SPMA-modified electrodes may be compared. The anodic peak current, i _pa, was determined at each scan rate for each PPy:SPMA modified gold electrode. Figure 3a is a plot of anodic peak current, i _pa, as a function of the square root of scan rate, ν^1/2 (V/s)^1/2, in the familiar form of the Randles–Sevcik equation (Eq. 1) [57]. The plot following electrodeposition to 1 mC/cm² demonstrates a similar linear relationship with scan rate when compared with bare gold with only a modest change in slope consistent with SEM observations at that charge density. The addition of 5, 10 and 30 mC/cm² reveals considerably reduced anodic current and independence of increasing scan rate. This behavior is consistent with a diffusion free, surface confined or mediated reaction. The linear relationship once again emerges at and above 50 mC/cm², with only a slight increase in linearity at 100 mC/cm² and a further larger increase at 1000 mC/cm ² . These patterns confirm a return to the active diffusion-controlled redox reaction at an electrode-active surface. The slopes of each line corresponding to the various electrodeposition charge densities was used to calculate the effective electroactive area via the Randles-Sevcik equation (Eq. 1).

Fig. 3:

(a) Plots of anodic peak current, i _pa, versus the square root of eight scan rates (10, 25, 50, 75, 100, 125, 150, 200 mV) for seven charge densities of electrodeposited PPy:SPMA and control (0 mC/cm²). (b) Effective electroactive area of Au|PPy:SPMA electrodes across charge densities. Below 50 mC/cm², polypyrrole acts as a barrier against diffusion related charge transfer between gold and Fe(II)/Fe(III) in solution. At and above electrodeposition charge of 50 mC/cm², polypyrrole becomes an electrode-active film and supports charge transfer with Fe(II)/Fe(III) in solution.

where i _pa is the peak anodic current (μA), n is the number of electrons involved in the redox reaction, A _eff is the effective electroactive area (cm²), D _o is the diffusion coefficient for ferri/ferrocyanide, C _o is the concentration of ferri/ferrocyanide (mM) in solution, and ν is the scan rate (V/s). Here electroactive area is defined as the portion of the electrode’s surface actively engaged in charge transfer with the ferri/ferrocyanide solution. Given that the test electrodes were comprised of two conductive materials, Au and PPy:SPMA, this could refer to charge transfer between gold and solution, or between polypyrrole and solution. A charge density of 1 mC/cm² produced electrodes with an electroactive area of 8.83 mm², 79% of the actual reference electroactive area (11.22 mm²). Given that SEM revealed only a few isolated deposits of polypyrrole at this charge density, electroactive area most likely corresponds with exposed bare gold in this case. Similar to the observed peak current trends, 5, 10, and 30 mC/cm² exhibit a substantial decrease in electroactive area (0.45, 0.36, and 0.38 mm² respectively), with the polypyrrole films formed at these charge densities acting as an barrier against charge transfer between the electrode and Fe(II)/Fe(III) in solution. Electroactive area increases to 2.96 mm² at 50 mC/cm², suggesting a threshold at or near this charge density at which polypyrrole begins exhibiting its own electrode-like behavior. Electroactive area further increases only slightly at 100 mC/cm² (3.46 mm²), and increases appreciably at 1000 mC/cm ² (5.88 mm² or 52% of bare gold). As shown in Fig. 3b, polypyrrole initially acts as a barrier to charge transfer until 50 mC/cm ² . Above 50 mC/cm ² , it acts like an electrode material [58] and supports diffusion linked charge transfer of Fe(II)/Fe(III) from solution. It is noteworthy that the electrode-activity of Au|PPy:SPMA, as manifiest in the effective surface area, follows a similar trend as the SEM feature analysis, with 50 mC/cm² emerging as a charge density of interest.

Changes in charge transfer resistance by EIS

EIS was used to investigate the electrode characteristics of Au|PPy:SPMA as well as the impact of the hydrogel layer on the impedance characteristics of Au|PPy:SPMA|Hydrogel. Figure 4a shows the Nyquist plots for control and the seven charge densities studied in DMEM at 37°C and 5% CO₂. Figure 4b depicts the plot of the charge transfer resistance as a function of the electrodeposition charge of PPy:SPMA films. Electrical impedance spectroscopy (EIS) revealed that on average, gold electrodes electrodeposited with polypyrrole and hydrogel exhibited an 82% decrease in charge transfer resistance, R_CT, compared with bare gold electrodes. Polypyrrole is known to contribute its characteristic redox chemistry with E_OX and E_RED corresponding to ~0.2 V [59] and −3.6 V [59], [60], respectively. Table 1 shows the impedance parameters extracted from R(QR) model fitting, including R _M (Ω) and R _CT (Ω) for the control and the seven charge densities. Technical triplicates of PPy:SPMA films formed from 50 mC/cm² experimentally showed modest in vitro variability. A Pearson’s correlation coefficients [61] analysis between the electrodeposition charge density and the various parameters extracted from model fitting of the EIS data established a strong anti-correlation between electrodeposition charge density and charge transfer resistance, R_CT, of Pearson’s correlation coefficient, r=−0.83. All other parameters were uncorrelated or very weakly correlated (Due to membrane resistance, R_M, the Pearson’s correlation coefficient, r=0.11).

Fig. 4:

(a) Nyquist plots showing real and imaginary components of the complex impedance, where, for a simple Randles R(QR) equivalent circuit, the rightmost x-intercept of each curve represents the sum of membrane resistance and charge transfer resistance. (b) Plot of R_CT as a function of the electrodeposition charge density of PPy:SPMA films. Insert shows the structure of the two layer polymer film.

Interlayer attachment

Films were incubated in DMEM at 37°C and 5% CO₂ and monitored for delamination. Up to 72 h films maintained their integrity. After 72 h hydrogel layers showed some evidence of lifting, but not complete delamination. Electrodeposited PPy:SPMA films that did not contain the PyBA-conj-AEMA (control) were similarly prepared and incubated. These films produced interlayer delamination within 30–45 min and upon hydration. Electrode assemblies were tested in the DMEM solution at current intensities from 20 mA to 100 mA (biphasic pulses, 1.0 Hz), with maximum current limited to 30 mA. Current was passed through these electrodes to test the coating stability. The stability test for the electrodes subjected to neurostimulation currents limited to 30 mA showed a stability of 0.5 h. There is a biocompatible hydrogel layer on top of the electrically engineered gold interface. The hydrogel layer is chemically attached, via covalent bonding using the ω-acryloyl moiety of PyBA-conj-AEMA, to the electroactive layer electrodeposited on the gold thin film electrode. This appears to delaminate after 30-min of stimulation under the above conditions, as shown by the stability test. The addition of the layer phase shifted the signal to 43°. Future work will focus on improving the interlayer adhesion, measurement of the in vitro biocompatibility and the conduct of animal studies.