The authors present an electrochemically controlled, drug releasing neural interface composed of a glassy carbon (GC) microelectrode array combined with a multilayer poly(3,4-ethylenedioxythiophene) (PEDOT) coating. The system integrates the high stability of the GC electrode substrate, ideal for electrical stimulation and electrochemical detection of neurotransmitters, with the on-demand drug-releasing capabilities of PEDOT-dexamethasone compound, through a mechanically stable interlayer of PEDOT-polystyrene sulfonate (PSS)-carbon nanotubes (CNT). The authors demonstrate that such interlayer improves both the mechanical and electrochemical properties of the neural interface, when compared with a single PEDOT-dexamethasone coating. Moreover, the multilayer coating is able to withstand 10 × 106 biphasic pulses and delamination test with negligible change to the impedance spectra. Cross-section scanning electron microscopy images support that the PEDOT-PSS-CNT interlayer significantly improves the adhesion between the GC substrate and PEDOT-dexamethasone coating, showing no discontinuities between the three well-interconnected layers. Furthermore, the multilayer coating has superior electrochemical properties, in terms of impedance and charge transfer capabilities as compared to a single layer of either PEDOT coating or the GC substrate alone. The authors verified the drug releasing capabilities of the PEDOT-dexamethasone layer when integrated into the multilayer interface through repeated stimulation protocols in vitro, and found a pharmacologically relevant release of dexamethasone.

Neural interfaces represent an important tool for enabling better fundamental understanding of the central and peripheral nervous systems, as well as for restoring damaged or lost sensorimotor functions.1–4 While advances in microfabrication technology and polymer science have led to miniaturized microelectrode arrays with better impedance and stiffness matching with the host tissue,5–9 long-term implantation of such neural interfaces continues to face undesired immune reactions.10–12 These reactions are typically manifested by inflammation of the tissue surrounding the device and a subsequent glial scar formation, which eventually leads to the encapsulation of the implant, worsening of the signal quality, and deterioration of the electrodes.10,13,14 This sequence of deleterious events is the main reason why obtaining stable and durable neural electrode-tissue interfaces continues to be a significant challenge.

One way to improve the long-term performance of neural implants is to coat the electrodes with biocompatible conductive polymers, such as poly(3,4-ethylenedioxythiophene) (PEDOT), thereby lowering the impedance of the electrode–electrolyte interface and promoting neural interactions.15–17 PEDOT has been shown to exhibit superior thermal and chemical stability compared to other conductive polymers,18 although it can delaminate if deposited in the form of thick layers or if the affinity with the substrate is not sufficient to withstand chronic electrical stimulation.19–21 

Further, PEDOT coatings can be integrated with bioactive molecules with anti-inflammatory properties, like dexamethasone and anionic dexamethasone phosphate (DEX), which could significantly help to preserve a stable chronic neural interface22–25 by modulating the inflammatory response to the implant. Specifically, DEX can be electrochemically released from the embedding conductive polymer into the tissue in a controlled manner and potentially inhibit the inflammatory tissue response in acute and chronic neural applications.23,24 Furthermore, a recently published study validated the potentiality of DEX functionalized PEDOT over 12 weeks in vivo.25 However, the incorporation of DEX as a dopant of PEDOT matrices has been reported to negatively affect both the electrochemical properties and the stability of the polymer coatings, making them more prone to brittle failure and delamination.26–28 Recently, it has been reported that electrochemical deposition of PEDOT carried out in an aqueous solution containing DEX-loaded carbon nanotubes (CNT) provides an interesting approach to increase the stability of the coating as well as to optimize the drug release.29,30 However, this method provided significant decrease of impedance only for CNTs doped PEDOT, whereas for DEX-CNTs the impedance was not substantially lower than for uncoated electrodes (see Table S1 in the supplementary material for a comparison).49 

The purpose of this study is to report on a new strategy to achieve DEX release from PEDOT matrices while minimizing the aforementioned side effects due to the additional doping of free DEX. Glassy carbon (GC) was chosen as an ideal substrate, thanks to its well-known excellent electrochemical properties.31 Moreover, we recently reported the superior electrochemical properties, in vivo performance, and long term stability under electrical stimulation of a new electrode material fabricated from lithographically patterned glassy carbon.32 In the same study, we also found that PEDOT-polystyrene sulfonate (PSS) adhered significantly better to GC than to Pt leading to high stable coatings. The strategy involves the optimization of a multilayer PEDOT electrodeposition technique for the surface functionalization of GC microelectrodes, consisting of an initial layer of PEDOT-PSS-CNT (PC), that provides superior charge–transport, mechanical and stability properties, followed by a second layer of PEDOT-DEX (PD) which supplies an electrochemically controlled drug-releasing mechanism, to form the two-layered coating PC//PD (PCD). Through the aforementioned electrodeposition method, we achieved the formation of highly stable, multilayer, low-impedance microelectrodes with superior electrical and electrochemical properties while introducing the possibility to release an anti-inflammatory drug (DEX) to combat on-site inflammatory reactions. The addition of a separate layer of PD on top of PC allows for the electrically controlled drug release without weakening the electrical and electromechanical performance of the device. Specifically, by electrodepositing the first layer of PC onto GC electrodes, we show that their electrochemical properties increases significantly, even when compared to the impedance drop obtained by electrodepositing a single layer of PD of similar thickness on similar GC substrates. This is not so unexpected as it is well understood that the addition of CNTs into PEDOT-PSS matrices improves the electrochemical properties, sturdiness and stability of the polymer.33,34 In addition, we demonstrate that the layers are well adhered to each other and to the GC substrate, forming a highly stable compound, and validate the electrochemically controlled release of DEX from the multilayer PCD coating through the in vitro detection and measurement of DEX in 0.9% NaCl aqueous solution.

1. Glassy carbon commercial electrodes

Commercial GC electrodes (3 mm diameter, IJ Cambria Scientific, Ltd.) where used, taking advantage of their large size and reusability, as benchmark electrodes for the optimization of the multilayer deposition technique, for scotch tape test, and for in vitro drug-release characterization. The optimized multilayer deposition technique was then adopted for experimental work on GC microelectrode arrays.

2. Glassy carbon microelectrode array fabrication

The fabrication of the thin-film devices used for this study is described in detail elsewhere.6 In summary, the glassy carbon electrodes were fabricated using the negative photoresist SU-8, which was patterned and pyrolyzed at 1000 °C in inert atmosphere. Subsequently, a layer of photosensitive polyimide (HD Microsystem) was spun and patterned onto the electrodes, as a substrate for the subsequent layers. Metals (Cr and Au) were then deposited on the substrate to create the conductive traces, and finally, a second layer of polyimide was spun on the traces to electrically insulate them. An image of a 32-electrode array (180 μm diameter electrodes) made of GC, PC, and PD on a flexible polyimide substrate is reported in Fig. 1(a).

Fig. 1.

(a) Image of a 32-electrode array (180 µm diameter electrodes) made of GC, PC, and PD on a flexible polyimide substrate. (b) Schematic of a multilayer PCD electrode cross section showing the three layers mentioned earlier.

Fig. 1.

(a) Image of a 32-electrode array (180 µm diameter electrodes) made of GC, PC, and PD on a flexible polyimide substrate. (b) Schematic of a multilayer PCD electrode cross section showing the three layers mentioned earlier.

Close modal

PC were coelectrodeposited from an aqueous mixture containing 0.5 M 3,4-ethylenedioxythiophene (EDOT, Sigma-Aldrich) 1 mg/ml of suspended carboxylated multi walled carbon nanotubes (MWCNT, NC 3151, <4% of COOH functional groups, Nanocyl S.A., Belgium) and 0.6 wt. % of poly(sodium 4-styrenesulfonate) (PSS, Sigma-Aldrich). The composite electrodeposition solution was prepared by suspending MWCNTs in ultrapure water (Milli-Q, Millipore) by horn sonication (Vibra-Cell VCX130, Sonics and Materials) for 30 min while cooling the suspension with an ice bath (6 s at 66% duty cycle pulses, 4 W/ml). PSS and the monomer were added to the suspension immediately after. The electrochemical deposition was carried out in potentiodynamic mode, sweeping the working electrode potential between 0 and 0.95 V with a scan rate of 100 mV/s for a total of 10 cycles (see Fig. S2 in the supplementary material for representative deposition plot).

Depositions were carried out using a potentiostat/galvanostat (Reference 600, Gamry Instruments) connected to a three-electrode electrochemical cell with a platinum wire counter electrode and a Ag/saturated AgCl reference electrode.

PD coatings were electrodeposited from a 0.05 M EDOT (Sigma-Aldrich) aqueous solution containing 0.05 M of predissolved dexamethasone 21-phosphate disodium salt (Sigma-Aldrich). The electrochemical deposition was carried out in potentiodynamic mode, sweeping the working electrode potential between 0 and 1.2 V, with a scan rate of 100 mV/s for a total of 30 cycles. Depositions were performed using a potentiostat/galvanostat (Reference 600, Gamry Instruments) connected to a three-electrode electrochemical cell with a platinum counter electrode and a Ag/saturated AgCl reference electrode.

PCD was prepared following the procedure described for PD. In this case, the working electrode is represented by the previously prepared PC electrodes, and the number of cycles is decreased from 30 to 10.

The electrochemical behavior of the microelectrodes was studied in a 0.9% w/w NaCl aqueous solution, by cyclic voltammetry (CV) to quantify the capacitive charging, and by electrochemical impedance spectroscopy (EIS) to determine the electrical properties of the system over a large range of frequencies. During the CV tests, the working electrode potential was swept between 0.5 and −0.5 V, maintaining a scan rate of 100 mV/s. During the EIS measurements, a sine wave (10 mV RMS amplitude) was superimposed at the potential of 0 V while varying the frequency from 1 to 105 Hz. EIS and CV were carried out using a potentiostat/galvanostat (Reference 600, Gamry Instruments) connected to a three-electrode electrochemical cell with a platinum wire as counter electrode and a Ag/saturated AgCl reference electrode. Thus, all potentials are referred to Ag/saturated AgCl (+0.197 V vs normal hydrogen electrode). The software zsimpwin v 3.2 (EChem Software) was used for equivalent circuit modeling of EIS data.

Commercial GC electrodes (3 mm diameter, IJ Cambria Scientific, Ltd.) and the microfabricated multilayer electrodes were routinely examined via optical microscopy using a Leica Zoom APO 16 and a Leica M205 FA, both equipped with a Leica DFC290 digital camera (Leica Microsystems, Germany). High resolution scanning electron microscope (HRSEM) analyses were performed on the microelectrodes to better investigate their morphology. The HRSEM of all the devices was carried out using a Jeol JSM-7500FA microscope (Jeol, Japan) equipped with a cold field emission gun (FEG) operating at different acceleration voltages, depending on the type of coating. In the case of the PD coated microelectrodes, a carbon coating of the surface was necessary to perform the HRSEM imaging due to its relatively low conductivity. In particular, a 10 nm-thick layer of carbon was deposited on those devices using an Emitech K950X high vacuum turbosystem (Quorum Technologies, Ltd., UK).

X-ray photoelectron spectroscopy (XPS) was carried out using a Kratos Axis UltraDLD spectrometer (Kratos Analytical Ltd., UK) on multilayer PCD and PD coated microelectrodes. XPS spectra were acquired using a monochromatic Al Kα source operated at 20 mA and 15 kV. The analyses were carried out on 300 × 700 μm areas centered on the microelectrodes. High resolution spectra were collected at pass energy of 20 eV and energy step of 0.1 eV, and the Kratos charge neutralizer system was used on all specimens. Spectra have been charge corrected to the main line of the C 1s spectrum set to 284.8 eV and analysed with casaxps software (Casa Software, Ltd., version 2.3.17). In the present report, XPS analysis mainly focused on the energy regions typical for F, S, and P peaks, chosen as markers for the presence of dexamethasone 21-phosphate (F and P peaks) in the PEDOT layer (S peaks). XPS peaks areas have been then calculated to obtain the relative F, S, and P atomic content in the PCD and PD coatings after normalization to the relative sensitivity factors (closely related to the cross-section of the x-ray induced photoemission process).

In order to test the electrochemical stability of the materials, the ability of PCD and PD coatings to resist intense and prolonged current stimulation patterns was verified by repeatedly applying a series of 1 × 106 pulses. After each pulse series, the impedance was measured, and the integrity of coatings was checked through SEM morphological analyses. The stimulation patterns consist of 1 × 106 series of cathodic-first charge balanced biphasic current pulses with 300 μA current amplitude, 200 μs cathodic half-phase period, and a frequency of 1 kHz in 0.9% w/w NaCl aqueous solution. Voltage transient responses during stimulation were acquired using a potentiostat/galvanostat (PARSTAT 2273, Princeton Applied Research) to simultaneously inject the stimulation current pulses and record the corresponding voltage excursions between working and counter electrodes.

Delamination was tested using a modified ASTM (American Society for Testing Materials) tape adhesion test (D3359-02).27 This test involves applying an adherent force (tape) to determine coating adhesion using standard Scotch 3M “low-medium adhesion masking tape for delicate surfaces” (#2080). The tape was allowed to sit for 5 min, after which it was slowly peeled back off the electrode surface maintaining a 180° angle between the tape and the electrode surface. A Leica M205 FA stereoscope equipped with a Leica DFC290 digital camera (Leica Microsystems, Germany) was used to image both the PCD and PD coatings before and after delamination.

To quantify the release of DEX, PCD was coated on a commercial GC electrode (3 mm diameter) and subjected to the releasing protocol in a water solution of NaCl (0.9% w/w). This consisted of a multiple stimulus trigger in a two-electrode system, with the coated GC acting as the working electrode and a Pt wire acting as the opposite electrode. Pulses consisted of −1 V for 200 μs followed by 0 V for 200 μs to each microelectrode.

The release of DEX was estimated by the absorbance at 242 nm of the releasing solution in a standard UV Quartz Spectrometer Cuvette Cell (Path Length 1 cm, Volume 0.7 ml). Analyses were performed with a Cary 300 UV-Vis Spectrophotometer (Agilent Technologies). The amount of released DEX was calculated per stimulation cycle assuming the molar extinction coefficients at 242 nm in the order of 1.33 × 104 M−1 cm−1 as evaluated by the calibration curve (see Fig. S1 in the supplementary material) and in accordance with literature data.35,36

A potentiodynamic electrodeposition method was used to prepare the PEDOT coated electrodes as it has been previously shown that this procedure produces PEDOT coatings with increased stability, morphological control, and reproducibility, along with improved electrochemical properties when compared to other electrodeposition methods, i.e., galvanostatic and potentiostatic methods.37 In Fig. 2(a), we report EIS plots recorded for PEDOT coated microelectrodes (diameter in the order of 180 μm). It can be seen that for all materials, i.e., PC, PCD, and PD, the |Z| impedance is frequency independent ranging from 105 to 40 Hz of the spectra for both PC and PD coated electrodes. Surprisingly, the subsequent insertion of a second layer of PD on the top of PC produced a significant increase of this near ohmic behavior, up to frequencies lower than 10 Hz. More importantly, the lowest impedance magnitude was obtained for the multilayered PCD and was significantly reduced when compared to PD. The value of |Z| at the frequencies of 10 Hz and 1 kHz for PC, PD, and PCD is reported in Table I. The PD composite, for similar thickness [see Figs. 3(b) and 3(c)], exhibited higher electrochemical impedance in the whole frequency domain, indicating a less efficient charge transport behavior in accordance with the literature.29 This difference in charge transport behavior was also evident during SEM imaging: to avoid charging effects, the PD composite had to be carbon coated to obtain high resolution images, while thanks to less charging effects, this was not necessary for the PCD material.

Fig. 2.

(a) EIS and (b) CV of PC, PCD, and PD microelectrodes.

Fig. 2.

(a) EIS and (b) CV of PC, PCD, and PD microelectrodes.

Close modal
Table I.

Magnitude impedance and CTC values of bare GC and for PC, PD, and PCD coated microelectrodes.

CoatingZ at 10 HzZ at 1 kHzCTC
GC 885.5 ± 218.0 kΩ 18.5 ± 3.7 kΩ 12.2 ± 3.6 mC/cm2 
PC 4.4 ± 1.4 kΩ 2.2 ± 0.5 kΩ 24.6 ± 2.3 mC/cm2 
PD 20.5 ± 0.4 kΩ 10.3 ± 1.6 kΩ 18.1 ± 0.4 mC/cm2 
PCD 2.7 ± 0.5 kΩ 1.9 ± 0.3 kΩ 168.7 ± 5.0 mC/cm2 
CoatingZ at 10 HzZ at 1 kHzCTC
GC 885.5 ± 218.0 kΩ 18.5 ± 3.7 kΩ 12.2 ± 3.6 mC/cm2 
PC 4.4 ± 1.4 kΩ 2.2 ± 0.5 kΩ 24.6 ± 2.3 mC/cm2 
PD 20.5 ± 0.4 kΩ 10.3 ± 1.6 kΩ 18.1 ± 0.4 mC/cm2 
PCD 2.7 ± 0.5 kΩ 1.9 ± 0.3 kΩ 168.7 ± 5.0 mC/cm2 
Fig. 3.

(a) and (b) Representative SEM images of multilayer PCD coated microelectrode: (a) top view and (b) cross section. (c) Representative SEM image of PD cross section. (d)–(i) Higher magnification SEM images of the surface morphology of (d), (g) PC, (e), (h) multilayer PCD and (f), (i) PD coatings obtained by Jeol JSM-7500FA FEG-SEM.

Fig. 3.

(a) and (b) Representative SEM images of multilayer PCD coated microelectrode: (a) top view and (b) cross section. (c) Representative SEM image of PD cross section. (d)–(i) Higher magnification SEM images of the surface morphology of (d), (g) PC, (e), (h) multilayer PCD and (f), (i) PD coatings obtained by Jeol JSM-7500FA FEG-SEM.

Close modal

The charge transfer capability (CTC), calculated as the time integral of an entire CV cycle between 0.5 and −0.5 V, for bare GC and for PC, PCD, and PD coatings, is reported in Table I. An example of CVs of PC, PD, and multilayer PCD coated microelectrodes is shown in Fig. 1(b). SEM images of the PC, multilayer PCD and PD coated GC microelectrodes are reported in Fig. 3. PC [Figs. 3(d) and 3(g)] show a 3D spongelike morphology due to the fine nanoscale CNTs scaffold structure, in accordance with our previously reported works regarding the deposition of PC on metal substrates.38,39

The second layer of PD, deposited on top of the PC coating [Figs. 3(e) and 3(h)], preserves this fine structure even though the PD layer itself exhibits a less porous and more compact morphology and, importantly, maintains an optimal adhesion between these two different PEDOT based layers, as shown by the absence of discontinuities between the two films in the cross-section image [Fig. 3(b)]. The highest CTC calculated from CV analysis is exhibited by the PCD coated microelectrodes and is related to its higher active surface area due to its very porous structure. Kayinamura et al.40 demonstrated how low, frequency-independent impedance over a wide frequency range (from ∼10 Hz to 100 kHz) originates from a two-layer homogeneous morphology of the PEDOT film, and accordingly, our results suggest an optimal interaction between the two layers, as confirmed by EIS measurements. The PD composite directly electrodeposited on bare GC electrodes presents a very compact morphology for almost the entire thickness (about 1.1 μm), while the exposed surface consists of a more porous, micro-globular-like morphology (about 1 μm diameter) [Figs. 3(f) and 3(i)]. In this case, the adhesion of the conductive polymer layer with the GC surface is very weak and presents a lack of continuity between the two surfaces that can be observed by the cross section view reported in Fig. 3(c).

Equivalent circuit modeling was applied to fit the experimental EIS in order to gain insight into charge transfer dynamic properties, including charge transfer rates at the interface between PEDOT coating and GC underlayer. The model consisted (see Fig. S3 in the supplementary material) of a solution resistance (Rs), a double layer capacitance (Cdl) in parallel with a charge transfer resistance (RCT), a bulk capacitance (Cd) represented by a constant phase element (CPE) due to the coating roughness, and a finite-length Warburg diffusion impedance (ZD). This model was first proposed by Danielsson et al.41 and by Asplund and coworkers later on,26 except for the use a pure capacitor instead of the CPE. The CPE impedance is given by 1/ZCPE = Q0(jω)n, where Q0 represents the admittance at = 1 rad/s: the CPE resembles a pure capacitor when the exponent n = 1.41 The Warburg impedance ZD is composed by a diffusional pseudocapacitance (CD), a diffusional time constant (τD), a diffusion resistance (RD = τD/CD) and can be defined by Eq. (1),42 

ZD=RDcoth(jωτD)1/2(jωτD)1/2.
(1)

Relevant EIS parameters obtained by the circuit modeling of experimental data for two representative samples of PCD and PD coatings, respectively, are summarized in Table II, while representative fitting curves are shown in Fig. S5 in the supplementary material.

Table II.

EIS parameters extracted by the fitting.

Q0 × 10−6 (S s1/2)nRCT (Ω)Cdl × 10−9 (F)τCT (μs)CD (μF)RD (Ω)τD (μs)CCV (μF)
PCD 223 0.89 506 1.25 0.6 14 1336 18.6 19.5 
PD 1.7 0.93 9556 0.165 11.1 2.88 6585 18.9 2.0 
Q0 × 10−6 (S s1/2)nRCT (Ω)Cdl × 10−9 (F)τCT (μs)CD (μF)RD (Ω)τD (μs)CCV (μF)
PCD 223 0.89 506 1.25 0.6 14 1336 18.6 19.5 
PD 1.7 0.93 9556 0.165 11.1 2.88 6585 18.9 2.0 

Analyzing Nyquist plots reported in Fig. 4, in the high frequency domain (>100 kHz), a slight semicircle (inset), which is more pronounced in the case of PD, can be observed. The presence of this high-frequency-semicircle has been ascribed to fast charge transfer processes at the electrode|polymer,41,43–45 polymer|solution45 interfaces, as well as to rapid charge transport dynamic in the polymeric film.45 As emerged from the SEM imaging, in particular, from the cross section shown in Fig. 3(c), the PD coating is poorly adhered to the GC surface and therefore it is not surprising that the charge transfer at the GC|PD interface is much slower when compared with the GC|PCD interface, with a time constant of 11 and 0.6 μs and a charge transfer resistance of 9500 and 500 Ω, respectively. Moreover the higher porosity observed for the PCD layer (Fig. 3) is confirmed by the higher capacitance, that can be calculated by the formula CCV = I/v where I = current (average of cathodic and anodic) and v = potential scan rate. A higher capacitance can be also related to a higher content of electroactive material, as a consequence of a more efficient incorporation of DEX for PCD coatings, in accordance with XPS results (see Table III, discussed later). We observed that the diffusional impedance ZD is strongly dominated by the diffusional resistance (RD = τD/CD) [see Eq. (1)] which is almost five times higher in the case of PD coatings, suggesting faster ion transport in PCD film.

Fig. 4.

Representative Nyquist plots for (a) PCD and (b) PD coatings (inset representing a magnification of the high frequency domain).

Fig. 4.

Representative Nyquist plots for (a) PCD and (b) PD coatings (inset representing a magnification of the high frequency domain).

Close modal
Table III.

XPS determination of DEX incorporation into PEDOT films.

Atomic elementPCDPD
S (at. %) 76 ± 1 89 ± 1 
F (at. %) 13 ± 1 6 ± 1 
P (at. %) 11 ± 1 5 ± 1 
Atomic elementPCDPD
S (at. %) 76 ± 1 89 ± 1 
F (at. %) 13 ± 1 6 ± 1 
P (at. %) 11 ± 1 5 ± 1 

In order to test the electrochemical stability of the materials, the ability of PCD and PD coatings to resist intense and prolonged current stimulation patterns was verified by repeatedly applying a series of one million pulses, as previously described in the Experimental Methods section. Each pulse corresponds to a cathodic charge density of 0.23 mC/cm2, verified to be into the range of the charge injection limit, defined as the maximum quantity of charge an electrode can inject before reaching faradaic processes. Voltage transients were collected in order to evaluate the charge injection properties of the materials (see Fig. S4 in the supplementary material). Voltage excursions for PC and PCD are much smaller in amplitude than for bare GC and are very ohmic in their shape. In particular, it was found that both anodic and cathodic voltage excursions are within in the safe limit of 0.5 V for PCD coatings. The PCD coating was able to withstand 10 million pulses with negligible change in EIS [Fig. 5(a)], while the PD coating shows a large increase in impedance after 5 × 106 pulses [Fig. 5(f)]. SEM analysis validates the electrochemical results, showing that the PCD composite presents similar morphology before [Figs. 5(b) and 5(d)] and after [Figs. 5(c) and 5(e)] 10 × 106 pulses, while the PD coating shows a flattened surface and numerous cracks after 5 × 106 million pulses [Figs. 5(h) and 5(j)].

Fig. 5.

(a) Representative impedance spectra of multilayer PCD coated microelectrodes before and after electrochemical stability experiments. (b)–(e) Representative SEM images of the morphology (different magnifications) of PCD coated microelectrode before (b), (d) and after (c), (e) stability experiments. (f) Representative impedance spectra of PD coated microelectrodes before and after stability experiments. (g)–(j). Representative SEM images of the morphology of PD coated microelectrode before (g), (i) and after (h), (j) stability experiments.

Fig. 5.

(a) Representative impedance spectra of multilayer PCD coated microelectrodes before and after electrochemical stability experiments. (b)–(e) Representative SEM images of the morphology (different magnifications) of PCD coated microelectrode before (b), (d) and after (c), (e) stability experiments. (f) Representative impedance spectra of PD coated microelectrodes before and after stability experiments. (g)–(j). Representative SEM images of the morphology of PD coated microelectrode before (g), (i) and after (h), (j) stability experiments.

Close modal

To evaluate any possible delamination due to mechanical stress, we performed a modified scotch test tape28 on both PCD and PD electrodeposited onto relatively large commercial GC (diameter = 3 mm) electrodes, that were chosen for their ease of use during the test and following optical imaging. Electrochemical impedance was collected before and after the adhesion test on both modified GC electrodes. In this case, the impedances of the different materials are more similar, probably due to the larger surface area of the electrodes as compared to the microelectrodes discussed earlier (diameter = 180 μm). The optical images of the coatings before and after the adhesion stress are reported in Fig. 6. PCD exhibits minimal delamination as confirmed by the low increase of |Z| within the overall frequency interval [Fig. 6(a)]. Bode impedance analysis of PD shows a change in its shape after the adhesion test, and the new spectra resemble the EIS profile of the pristine GC electrode [see Fig. S6(a) in the supplementary material], consisting of lower |Z| values in the higher frequencies (105–103 Hz) followed by a rapid increase of |Z|, as reported in Fig. 6(d). This suggests the exposure of free GC islands as a consequence of the mechanical stress performed on PD, as also confirmed by optical microscope imaging that clearly shows the exposure of the underlying black colored glassy carbon substrate and the disappearance of the blue colored PD coating following the adhesion test [Figs. 6(e), 6(f), and S6(b)]. Performing the same test on the PCD multilayer, we find that the surface texture and color is preserved [Figs. 6(b), 6(c), and S6(b)].

Fig. 6.

(a) Representative impedance spectra of multilayer PCD coated GC electrode before and after scotch tape test. (b), (c) optical images of PCD coating before (b) and after (c) scotch tape test. (d) Representative impedance spectra of PD coated GC electrode before and after scotch tape test. (e), (f) optical images of PD coating before (e) and after (f) scotch tape test.

Fig. 6.

(a) Representative impedance spectra of multilayer PCD coated GC electrode before and after scotch tape test. (b), (c) optical images of PCD coating before (b) and after (c) scotch tape test. (d) Representative impedance spectra of PD coated GC electrode before and after scotch tape test. (e), (f) optical images of PD coating before (e) and after (f) scotch tape test.

Close modal

The effective loading of DEX in PCD and PD coatings was determined via XPS analysis on microelectrode arrays, and the results are reported in Table III, The results show that, as expected for dexamethasone 21-phosphate, the F:P atomic ratio is 1 (within the experimental error), supporting the reliability of the experimental method used. Moreover, the XPS results indicate that the PC layer plays an important role also in determining the amount of DEX incorporated by the conductive polymer backbone with respect to the monomer EDOT (the doping ratio), since the ratio between either F or P (DEX markers) and S moves from 0.06 in the PD layer to 0.16 in the PCD case. It has been reported that the typical doping ratios for PEDOT coatings incorporating DEX or other biomolecules fall in the range of 0.2–0.4.28,46 Thus, the extremely low DEX content observed for PD films are consistent with a poor dopant efficiency, reflecting a less optimized coating, in accordance with optical and electrochemical characterization.

After proving that our strategy to optimize a multilayer PEDOT electrodeposition technique for the surface functionalization of GC microelectrodes allows for the realization of highly stable, low-impedance, DEX loaded devices with superior electrochemical and mechanical properties, we quantified in vitro the release of DEX. For this purpose PCD was deposited onto commercial GC electrodes with a nominal area of 0.07 cm2 (3 mm in diameter), in order to guarantee a detectable amount of drug in the solution, taking into consideration the linear correlation range between absorbance and concentration of DEX, that has been estimated in the order of 0.5 μg/ml.47 For the first 100 pulses, the stimulated drug release profile shows a quasilinear behavior, with a total (average) of 91.8 μg/cm2 DEX released. The following release trend becomes slower, reaching a total amount of released DEX of 140 μg/cm2 after 400 pulses, as shown in Fig. 7.

Fig. 7.

Cumulative in vitro DEX release from PCD coated onto GC electrode (3 mm diameter).

Fig. 7.

Cumulative in vitro DEX release from PCD coated onto GC electrode (3 mm diameter).

Close modal

These values are amply sufficient for decreasing the inflammation of the neural tissue after an implant as reported in literature.22,29,48 By controlling and regulating the number of the electrical pulses, there is the possibility to significantly lower the postimplantation brain tissue inflammation, preventing further immune reactions and the eventual encapsulation of the implanted device.

In this study, we introduce a novel approach to fabricate a highly stable neural interface with superior electrical and electrochemical properties and built-in drug-delivery system. The electrodes are made from a glassy carbon substrate, on which two different PEDOT layers are electrodeposited to form PCD coating. We demonstrated that this experimental approach is beneficial for lowering the total impedance as well as imparting mechanical and electrochemical stability, as compared to PD coated on native GC electrodes. The electrochemically controlled release of DEX was performed and estimated in vitro by measuring the absorbance of the solution at a fixed wavelength (242 nm) and confirmed the functionality of the hybrid system. To the best of our knowledge, this is the first time that DEX based microelectrodes provide together optimal electrochemical and mechanical properties as well as furnishing electrochemically controlled drug release. Future studies will focus on in vivo performance of the PCD coated glassy carbon electrodes in controlling inflammatory processes in animal models and further optimizing the DEX-release-protocol. These studies, coupled with in depth in vitro biocompatibility essays, will contribute to make a further step ahead toward the use of conductive polymer based materials for neural implant applications.

This work was supported by National Science Foundation [Grant No. EEC-1028725] under the ERC program. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

1.
S. J.
Bensmaia
and
L. E.
Miller
,
Nat. Rev. Neurosci.
15
,
313
(
2014
).
2.
J. J.
Wazen
,
J. B.
Spitzer
,
S. N.
Ghossaini
,
J. N.
Fayad
,
J. K.
Niparko
,
K.
Cox
,
D. E.
Brackmann
, and
S. D.
Soli
,
Otolaryngol.-Head Neck
129
,
248
(
2003
).
3.
A. Y.
Chow
and
V. Y.
Chow
,
Neurosci. Lett.
225
,
13
(
1997
).
4.
L. G.
Cohen
 et al,
Nature
389
,
180
(
1997
).
5.
K.
Wang
,
H. A.
Fishman
,
H.
Dai
, and
J. S.
Harris
,
Nano Lett.
6
,
2043
(
2006
).
6.
M.
Vomero
 et al,
J. Micromech. Microeng.
26
,
025018
(
2016
).
7.
I. R.
Minev
 et al,
Science
347
,
159
(
2015
).
8.
V.
Castagnola
,
E.
Descamps
,
A.
Lecestre
,
L.
Dahan
,
J.
Remaud
,
L. G.
Nowak
, and
C.
Bergaud
,
Biosens. Bioelectron.
67
,
450
(
2015
).
9.
Y.
Whulanza
,
E.
Battini
,
L.
Vannozzi
,
M.
Vomero
,
A.
Ahluwalia
, and
G.
Vozzi
,
J. Nanosci. Nanotechno.
13
,
188
(
2013
).
10.
V. S.
Polikov
,
P. A.
Tresco
, and
W. M.
Reichert
,
J. Neurosci. Methods
148
,
1
(
2005
).
11.
M.
Vomero
,
I.
Dryg
,
T.
Maxfield
,
W.
Shain
,
S.
Perlmutter
, and
S.
Kassegne
,
ECS Trans.
72
,
91
(
2016
).
12.
A. A.
Schendel
 et al,
J. Neural Eng.
11
,
046011
(
2014
).
14.
C.
Marin
and
E.
Fernández
,
Front. Neuroeng.
3
,
8
(
2010
).
15.
X. T.
Cui
and
D. C.
Marin
,
Sens. Actuator, B
89
,
92
(
2003
).
16.
M. R.
Abidian
and
D. C.
Martin
,
Biomaterials
29
,
1273
(
2008
).
17.
X. T.
Cui
and
D. D.
Zhou
,
IEEE Trans. Neural Syst. Rehabil.
15
,
502
(
2007
).
18.
H.
Yamato
,
M.
Ohwa
, and
W.
Wernet
,
J. Electroanal. Chem.
397
,
163
(
1995
).
19.
20.
B.
Wei
,
J.
Liu
,
L.
Ouyang
,
C.
Kuo
, and
D. C.
Martin
,
ACS Appl. Mater. Interfaces
7
,
15388
(
2015
).
21.
L.
Ouyang
,
B.
Wei
,
C.
Kuo
,
S.
Pathak
,
B.
Farrell
, and
D. C.
Martin
,
Sci. Adv.
3
,
e1600448
(
2017
).
22.
Y.
Zhong
and
R. V.
Bellamkonda
,
Brain Res.
1148
,
15
(
2007
).
23.
D.
Svirskis
,
J.
Travas-Sejdic
,
A.
Rodgers
, and
S.
Garg
,
J. Controlled Release
146
,
6
(
2010
).
24.
R.
Wadhwa
,
C. F.
Lagenaur
, and
X. T.
Cui
,
J. Controlled Release
110
,
531
(
2006
).
25.
C.
Boehler
,
C.
Kleber
,
N.
Martini
,
Y.
Xie
,
I.
Dryg
,
T.
Stieglitz
,
U. G.
Hofmann
, and
M.
Asplund
,
Biomaterials
129
,
176
(
2017
).
26.
M.
Asplund
,
H.
von Holst
, and
O.
Inganäs
,
Biointerphases
3
,
83
(
2008
).
27.
S.
Baek
,
R. A.
Green
, and
L. A.
Poole-Warren
,
J. Biomed. Mater. Res., A
102
,
2743
(
2014
).
28.
J. A.
Goding
,
A. D.
Gilmour
,
J. P.
Martens
,
L. A.
Poole-Warren
, and
R. A.
Green
,
J. Mater. Chem. B
3
,
5058
(
2015
).
29.
C. L.
Kolarcik
 et al,
J. Neural Eng.
12
,
016008
(
2015
).
30.
N. A.
Alba
,
Z. J.
Du
,
K. A.
Catt
,
T. D. Y.
Kozai
, and
X. T.
Cui
,
Biosensors
5
,
618
(
2015
).
31.
R. L.
McCreery
,
Chem. Rev.
108
,
2646
(
2008
).
32.
M.
Vomero
 et al,
Sci. Rep.
7
,
40332
(
2017
).
33.
T. D.
Kozai
 et al,
IEEE Trans. Biomed. Eng.
63
,
111
(
2016
).
34.
R.
Gerwig
,
K.
Fuchsberger
,
B.
Schroeppel
,
G. S.
Link
,
G.
Heusel
,
U.
Kraushaar
,
W.
Shuhmann
,
A.
Stett
, and
M.
Stelzle
,
Front. Neuroeng.
5
,
8
(
2012
).
35.
R. A.
Colbert
and
D. A.
Young
,
Proc. Natl. Acad. Sci. U. S. A.
83
,
72
(
1986
).
36.
H. C.
Horner
,
A.
Munck
, and
G. E.
Lienhard
,
J. Biol. Chem.
262
,
17696
(
1987
).
37.
V.
Castagnola
,
C.
Bayon
,
E.
Descamps
, and
C.
Bergaud
,
Synth. Met.
189
,
7
(
2014
).
38.
E.
Castagnola
,
E.
Maggiolini
,
L.
Ceseracciu
,
F.
Ciarpella
,
E.
Zucchini
,
S.
De Faveri
,
L.
Fadiga
, and
D.
Ricci
,
Front. Neurosci.
10
,
151
(
2016
).
39.
E.
Castagnola
,
A.
Ansaldo
,
E.
Maggiolini
,
G. N.
Angotzi
,
M.
Skrap
,
D.
Ricci
, and
L.
Fadiga
,
ACS Nano
7
,
3887
(
2013
).
40.
Y. P.
Kayinamura
,
M.
Ovadia
,
D.
Zavitz
, and
J. F.
Rubinson
,
ACS Appl. Mater. Interfaces
2
,
2653
(
2010
).
41.
P.
Danielsson
,
J.
Bobacka
, and
A.
Ivaska
,
J. Solid State Electrochem.
8
,
809
(
2004
).
42.
E.
Barsoukov
and
J. R.
Macdonald
,
Impedance Spectroscopy: Theory, Experiment, and Applications
, 2nd ed. (
John Wiley & Sons, Inc.
,
Hoboken, NJ
,
2005
).
43.
S.
Carli
,
E.
Busatto
,
S.
Caramori
,
R.
Boaretto
,
R.
Argazzi
,
C. J.
Timpson
, and
C. A.
Bignozzi
,
J. Phys. Chem. C
117
,
5142
(
2013
).
44.
H.
Tian
,
Z.
Yu
,
A.
Hagfedlt
,
L.
Kloo
, and
L.
Sun
,
J. Am. Chem. Soc.
133
,
9413
(
2011
).
45.
J.
Bobacka
,
A.
Lewenstam
, and
A.
Ivaska
,
J. Electroanal. Chem.
489
,
17
(
2000
).
46.
R. A.
Green
,
N. H.
Lovell
, and
L. A.
Poole-Warren
,
Biomaterials
30
,
3637
(
2009
).
47.
C.
Boehler
and
M.
Asplund
,
J. Biomed. Mater. Res., A
103
,
1200
(
2015
).
48.
W.
Shain
,
L.
Spataro
,
J.
Dilgen
,
K.
Haverstick
,
S.
Retterer
,
M.
Isaacson
,
M.
Saltzman
, and
J. N.
Turner
,
IEEE Trans. Neural Syst. Rehabil. Eng.
11
,
186
(
2003
).
49.
See supplementary material at http://dx.doi.org/10.1116/1.4993140 for a table comparing values of most relevant parameters between this study and literature, calibration curve for DEX, PEDOT-PSS-CNT electrodeposition plot, equivalent circuit, voltage transients and EIS fitted curves.

Supplementary Material