**1. Introduction**

Conductive polymers are a new generation of smart materials extensively used in organic bioelectronics, mostly in the development of neural implants, biosensors, and active controlled release systems [1–4]. Poly (3,4-ethylenedioxythiophene) (PEDOT) is a conductive polymer synthetized from 3,4-ethylenedioxythiophene (EDOT), used as a coating in diverse types of sensors due to its biocompatibility, conductivity, processing versatility, and stability [5,6]. Moreover, PEDOT is reported as a promising material for the immobilization of enzymes and other biologically active molecules [2,7,8]. The incorporation of charged molecules into the PEDOT backbone is described through an electrostatic mechanism due to the formation of charge carriers and the doping process during the electropolymerization process [9]. The subsequent release of the charged compounds was

reported to be dependent on the polymer thickness and charge applied during the electrochemical stimulus [10–13].

Diverse implants and sca ffolds are developed in regenerative medicine to serve as extracellular matrices for cell colonization [14–16]. Many of them are loaded with bioactive agents to improve the therapeutic e fficacy and safety of the drugs, playing important roles in treatment of several chronic diseases, damaged tissues, and providing a potential stimulation of di fferent types of cells [17–19].

Although diverse engineering groups established di fferent types of implants for a broad range of applications, those implants can elicit body responses involving inflammatory processes, which may result in the formation of glial scars due to neural devices specifically [12,13,20,21]. One strategy to avoid immune responses consists of releasing an anti-inflammatory biomolecule (i.e., dexamethasone) in the vicinity of the implant [11,13,22,23]. Dexamethasone (Dx) is a synthetic glucocorticoid that reduces inflammation in the central nervous system, acting through glucocorticoid receptors found in most neurons and glial cells. Due to being locally delivered, the specificity and e fficiency of dexamethasone means that only small amounts of the drug are required [13,22,24–28].

κ-Carrageenan (κC) is a sulfonated polysaccharide recently used in aqueous micellar dispersions for the polymerization of EDOT, since it provides an appropriate environment for the monomer dispersion while acting as a doping agen<sup>t</sup> in the conductive layer [29–31]. According to the previous work, the electrochemical properties of PEDOT are retained when κC is used as a doping agen<sup>t</sup> [29,30], avoiding a potential delamination during the reduction-oxidation process needed during the active delivering process. Biocompatibility of PEDOT:κC composite has been demonstrated in previous studies [2,29].

In this work, we induce the loading of dexamethasone phosphate during the deposition of the electroactive composite onto a bare gold electrode by changing the amount of drug in the dispersion prior the polymerization. κC was incorporated to maintain the electrochemical stability and biocompatibility of the PEDOT matrix and the subsequent drug release using electrical stimulation. The presence of κC and Dx inside the conductive film was confirmed by μ-Raman spectroscopy and their e ffect in the topography was studied using profilometry. Dexamethasone release was evaluated by cyclic voltammetry and High-Resolution (HR) mass spectrometry. Therapeutic doses of dexamethasone were achieved during the electrical stimulation of the bioelectronic device.

### **2. Results and Discussion**

### *2.1. Evaluation of the Stability and Size of the Dispersion Systems*

The dispersions used to electrodeposit the monomer and the Dx on the electrode were evaluated by their ζ-potential values and particle size distribution in order to determine its stability in aqueous medium. ζ-potential data was obtained for the six prepared dispersions, and they are shown in Table 1. It is possible to observe that EDOT:κC:Dx has an appropriate stability (−48.70 mV), which is dominated for the κC micellar system (−43.30 mV). Values of ζ-potential over −30 mV are considered stable assuming that an electrostatic charge is the main stabilization mechanism and the colloidal system is in the range of hundreds [32,33]. The anionic nature of the κC and Dx avoids aggregation due to the negative values obtained in the ζ-potential analysis, which are comparable with previously reported results for these molecules [30,34,35]. A stable dispersion prevents aggregation or deposition of the particles that carried the monomer during the electrochemical deposition. Additionally, the stable system may allow a homogeneous dispersion of κC and dexamethasone in the electrodeposited film as seen by Raman spectroscopy.


**Table 1.** ζ-potential values of dispersions used in the fixation of the drug on the electrode.

Particle size measurements of the main three dispersions were performed to determine the dimension of their aggregates after the sonication process. Figure 1a shows the size distribution for the κC 0.2% *w*/*v* solution, it is possible to observe a single population for the surfactant. Some authors have reported previously that κC solutions are polydisperse (two or more populations), because it increases the gel behavior due to its polysaccharide nature [36,37]. Nevertheless, they emphasized that the main signal for the κC aggregates has an average size in the range of 800 to 1000 nm [37], which agrees with our results. The intensive sonication process before the measures and the low concentration of κC used in the analysis may explain why only one population were observed in the κC size distribution, similar to a previous report [30].

**Figure 1.** Size distribution (d. nm) of (**a**) κC; (**b**) κC:Dx; and (**c**) EDOT:κC:Dx dispersions, measured by dynamic light scattering (DLS) method.

On the other hand, once the Dx was added to the dispersion, a polydisperse behavior was found in the κC:Dx system and two populations were detected (Figure 1b,c). Dexamethasone solutions are characterized by a single population with a particle size average of 100 nm [38] and was consistent with our results. Eventually, it is possible to observe that the stability of the system has remained when the monomer was added (Figure 1c). The stability of the dispersions depends mainly on the used surfactant and it has an important influence in the physical and electrochemical properties of the electrodeposited films [39].

### *2.2. Analysis of the Topography and Composition of PEDOT:*κ*C:Dx Coating by* μ*-Raman Spectroscopy and Profilometry Methods*

The PEDOT:κC:Dx composite was obtained from a EDOT:κC:Dx dispersion by electrochemical deposition under galvanostatic conditions (Figure S1), as it was established in a previous work [2,30]. Then, the topography of the PEDOT:κC:Dx coating was characterized before (Sa: 0.270 ± 0.005 μm, surface area: 1361 mm2, negative volume 0.1562 mm3, and volume 1.695 mm3) and after (Sa: 0.250 ± 0.005 μm, surface area: 1337 mm2, negative volume 0.1707 mm3, and volume 1.690 mm3) releasing the Dx from the conductive coating. The roughness data of both surfaces did not show significant di fferences between them (see Figure 2a,b). The volume ratio between peaks and valleys describes the symmetry in the surface topography. A negative value is indicative of more distinct valleys and positive of more distinct peaks about the average plane. Our samples were dominated by peaks and low negative volume (around ten times) and those values are consistent with a previous report for PEDOT:κC coatings [30]. It is suggested that rough surfaces in comparison with smooth surfaces improve cell attachment due to the formation of specific surface-cell contacts by increasing the expression of di fferent integrins subunits [40,41]. Although, diverse authors have reported that surface roughness values higher than 0.5 μm are desirable to ensure the maximum attachment and proliferation of cells, large rough surfaces also stimulate more anti-inflammatory responses because the activation of M2 macrophages and the subsequent release of anti-inflammatory cytokines [42]. The PEDOT:κC:Dx surface roughness value and the lack of their significative variation during the delivery of dexamethasone may indicate the reliability of electroactive composite for cell culture studies, since no additional mechanism may be seemed due to the topography changes.

**Figure 2.** Profilometry images obtained for PEDOT:κC:Dx films (**a**) before and (**b**) after 160 cycles of cyclic voltammetry in a 0.10 M ammonium acetate solution.

The qualitative composition of the conductive film was determined using confocal μ-Raman spectroscopy before (Figure 3a,c) and after (Figure 3b,d) 160 sweeps of electrical stimulation in a 4 μm<sup>2</sup> area and 5 μm depth inside the composite. The analysis was performed in order to determine the presence of PEDOT, dexamethasone, and κ-carrageenan inside the electroactive composite. The signal was obtained and plotted in a 2D image that allows the association of the signal (counts) to the presence of the corresponding functional groups for each component.

**Figure 3.** 2D confocal Raman map of the 1430 cm<sup>−</sup><sup>1</sup> band (**a**) before release process and (**b**) after 160 release cycles. Raman mapping of the 1625 cm<sup>−</sup><sup>1</sup> band intensity (**c**) before release process and (**d**) after 160 release cycles at 0.5 μm depth inside the conductive layer. The yellow areas are related to the presence of PEDOT and κC/Dx, respectively.

PEDOT shows a strong signal in the spectral range of 1421–1442 cm<sup>−</sup>1, associated to the thiophene symmetric Cα = Cβ stretching [2,30,43] and its oxidation state. The corresponding signal was obtained from the composite before and after 160 cycles of electrical stimulation (Figure S2) and it was mapped at 1430 ± 25 cm<sup>−</sup><sup>1</sup> (Figure 3a,b), where bright yellow dots corresponded to presence of PEDOT. A homogeneous distribution of the conductive polymer was detected in both samples.

Additionally, a relative intense band at 1625 ± 30 cm<sup>−</sup><sup>1</sup> was detected, corroborating the qualitative existence of Dx and κC in the conductive film (Figure 3c,d). This signal, in the 2D, is distributed through the conductive matrix. The result is similar to previous studies [13,26], which reported the characteristic spectral signals of dexamethasone in the ranges of 3200–3500 cm<sup>−</sup>1, 2850–3000 cm<sup>−</sup>1, and near to 1650 cm<sup>−</sup>1, as is verified in Figure S3, corresponding to hydroxyl, methyl, and carbonyl groups, respectively. Dexamethasone and κC act as doping agents, so there is a consistent association of the respective signal for both molecules and the PEDOT band. The identification of the band at 1625 cm<sup>−</sup><sup>1</sup> overlapping with PEDOT signal, confirmed the presence of the doping agen<sup>t</sup> before and even after electrochemical stimulation, as is shown in Figure S2a,b, respectively. Adding κC in the formulation provides a proper doping agen<sup>t</sup> during the release of the Dx, reducing the degradation by overoxidation and eventually delamination as is shown in Figure S4 [30].

### *2.3. Dexamethasone Release Experiments from the PEDOT:*κ*C:Dx Coating*

Drug loading into the conducting polymers films is based on the fact that these kinds of polymers are electrically oxidized during the polymerization processes, generating charge carriers [9,44,45]. The doping agen<sup>t</sup> (e.g., Dx and κC) is incorporated to the oxidized polymer [46] to maintain charge neutrality. In this work, dexamethasone 21 phosphate and κC are used as doping agents, the presence of sulfate and phosphate groups imparts negative charges in the polysaccharide and the drug, respectively.

The electrochemical controlled release studies from PEDOT:κC:Dx coating were performed within a potential range of −600 to 1000 mV to evaluate intrinsic redox processes of the film [13,35,45]. Figure 4 shows the characteristic oxidation and reduction potential signal ranges at 0 to 500 mV and −100 to −400 mV, respectively, after a different number of voltammetry scans. According to some authors, the voltammetric behavior of dexamethasone shows a reduction signal at the potential of −350 mV [13,45], which indicates the release of the drug from a stimulated electrode. The corresponding CV signals are shown in Figure 4, this signal gradually decreased according to the sweep

number, disappearing completely after 160 cycles of electrical stimulation. Electrochemical reduction of a conducting polymer results in the migration of small doping molecules from the conducting composite to maintain the electro neutrality of the matrix [44,46]. Thus, the application of alternating positive and negative potentials during cyclic voltammetry analysis caused the release of the Dx from the PEDOT coating.

**Figure 4.** Cyclic voltammograms for the PEDOT:κC:Dx recorded at 25 mV·s<sup>−</sup><sup>1</sup> after 10, 60, and 160 cycles of electrical stimulation in ammonium acetate 0.10 M.

Spontaneous release of the dopant from the PEDOT structure is an instant process, but the Dx release is slow, since it is driven by diffusion from the inner film to the surface. κC is a large molecule, this type of dopant is more attached into the polymer coating and it is not leached out during the electrical stimulation, granting to the polymer greater electrochemical stability [13,46,47], as confirmed by Raman spectroscopy.

The release profile of the Dx was investigated under passive conditions (unstimulated) and active electrically stimulation using an ammonium acetate 0.10 M solution as supporting electrolyte. The surface area of the electrode is associated with promoting larger amounts of passive drug release according to the second Fick's law of diffusion [48,49], yet, in our case, the electrode surface and total area are maintained virtually constant. The quantification of Dx from the PEDOT:κC:Dx modified electrodes was achieved using HR-mass spectrometry (Figure 5).

The active release profile was performed with a total of 76 CV sweeps in five release events, taking around 300 min to be completed. Accordingly, the passive release profile from unstimulated electrodes were evaluated over the same period of 300 min.

Figure 5a shows the passive release profile of Dx as a function of square root of time according to the Higuchi model for the drug release from a polymer film [27,50], where pure Fickian diffusion is the dominant phenomena [48]. The low diffusion value, in the beginning of the process, may depend on the slow penetration of supportive electrolyte into the polymeric film [49]. The pattern changed after 80 min and a higher diffusion value reflects the diffusivity of the passive Dx release process. The three systems (1 mM, 5 mM, and 10 mM) showed analogous Fickian diffusion behavior.

**Figure 5.** (**a**) The passive release profile of Dx as a function of square root of time, over 300 min from unstimulated electrodes. The active electrically controlled delivery process by stimulation events (columns) compared to the passive release profile (line) using: (**b**) 1 mM, (**c**) 5 mM, and (**d**) 10 mM of Dx in the initial formulation.

On the other hand, Figure 5b–d showed a remarkable dependency of the released Dx concentration during the electrical stimulated events (bars) compared to a passive unstimulated electrode (line). Some authors have studied controlled drug release systems using conductive polymers such as polypyrrole and PEDOT, where the anionic molecule is used as doping agen<sup>t</sup> and their subsequent release is mainly determined via diffusion [11,13,44,45,51]. Nevertheless, for a controllable release system, it is desirable to have a high active release and low diffusion relationship [11,12], as shown by our system (see Figure 5). For instance, the initial concentration of 10 mM released in the passive process ca. 2% of the delivered Dx in stimulated process. This is probably associated with the use of κC as second doping in the matrix, which grants the film stability and integrity during stimulation cycles [30,46].

The therapeutic dosages of Dx in mesenchymal stem cell cultures are effective at levels of 100–1000 nM to promote their differentiation to osteoblast or in order to be used during anti-inflammatory treatment [52–54]. In this work, the accumulative concentration of the released Dx using 1 mM and 5 mM initial formulations (Figure 5b,c) were 300 nM (0.66 <sup>μ</sup>g·cm<sup>−</sup>2) and 600 nM (1.60 <sup>μ</sup>g·cm<sup>−</sup>2), respectively. Even though, these values are at therapeutically relevant levels, they are in part determined by the Dx amount release via diffusion.

Instead, when 10 mM of the drug was poured in the initial formulation, a total of 3700 nM (8.89 <sup>μ</sup>g·cm<sup>−</sup>2) of cumulative Dx was detected. This concentration range far in excess of the quantity of dexamethasone released from similar systems using an identical initial concentration of the drug for the coating preparation, for which values are even lower than 5.03 <sup>μ</sup>g·cm<sup>−</sup><sup>2</sup> [11,12,51]. Such concentrations

surpass the amount of the drug needed in cell cultures and it is not recommended to apply in biological systems. Nonetheless, using a specific electrochemical stimulation profile may be allowed to provide an adequate quantity of the drug for di fferent biological applications.

### **3. Materials and Methods**
