Electronically Conductive Hydrogels by in Situ Polymerization of a Water-Soluble EDOT-Derived Monomer

Electronically-conductive hydrogels have gained popularity in bioelectronic interfaces because their mechanical properties are similar to biological tissues, potentially preventing scaring in implanted electronics. Hydrogels have low elastic moduli, due to their high water content, which facilitates their integration with biological tissues. To achieve electronically-conductive hydrogels, however, requires the integration of conducting polymers or nanoparticles. These ‘hard’ components increase the elastic modulus of the hydrogel, removing their desirable compatibility with biological tissues, or lead to heterogeneous distribution of the conductive material in the hydrogel scaffold. A general strategy to transform hydrogels to electronically-conductive hydrogels without affecting the mechanical properties of the parent hydrogel is still lacking. Herein, we report a two-step method for imparting conductivity to a range of different hydrogels by in-situ polymerization of a water-soluble and neutral conducting polymer precursor: 3,4-ethylenedioxythiophene diethylene glycol (EDOT-DEG). The resulting conductive hydrogels are homogenous, have conductivities around 0.3 S m⁻¹, low impedance, and maintain an elastic modulus of 5–15 kPa, which is similar to the pre-formed hydrogel. The simple preparation and desirable properties of the conductive hydrogels are likely to lead to new materials and applications in tissue engineering, neural interfaces, biosensors, and electrostimulation.

1. Introduction

Biomedical electronic devices can be interfaced with the human body to measure physiological signals^1,2 or to provide electrical stimulation for treatment.^3,4 Electronic materials that can be seamlessly interfaced with human tissues or cells are therefore essential to study or stimulate the nervous system, or to serve as neural tissue scaffolds. But, materials challenges remain to be addressed to bridge the gap between ‘soft’ biological tissues and ‘hard’ electronics. Mechanical mismatch between biological tissues and electronics can lead to scar formation between the electronics and the target tissue, which significantly hampers the performance and efficacy of bioelectronic stimulation and recording.^3,5,6 Conductive hydrogels are a promising strategy towards interfacing electronic materials to biological tissues.^6,7 Due to their high water content (70–99 wt%), hydrogels can be as soft and flexible as biological tissues, including skin, muscle, heart, spinal cord and brain (E < 100 kPa).⁶ To impart electrical conductivity, conductive nanoparticles, such as metal nanowires (NWs)^8,9 carbon nanotubes (CNTs),⁷ and conductive polymers^5,7,10 can be incorporated inside the hydrogel. NWs and CNTs, while effective at achieving electronically-conductive hydrogels, were shown to lead to heterogeneities in the network and hydrogels with a high elastic modulus; overall leading to a poor interface with biological entities.⁸ Compared to these nanocomposite hydrogels, conductive hydrogels made from conducting polymers (CPs), including poly(3,4-ethylenedioxythiophene) (PEDOT),^11,12 polypyrrole (PPy),^13–15 or polyaniline (PANI)^15,16 offer a higher compatibility with biological systems, by virtue of their flexibility and potential for ionic—as well as electronic—conductivity.

Several methods have been envisioned for the preparation of conductive hydrogels from CPs.^17,18 The conducting polyelectrolyte complex of PEDOT with poly(styrene sulfonate) (PEDOT:PSS) can be gelled directly from its aqueous solution by increasing the ionic strength with bivalent ions Mg²⁺, Ca²⁺, or multivalent ions Fe³⁺, Ce⁴⁺.^19,20 These conductive hydrogels, however, are only weakly crosslinked leading to poor strength and contain residual ions, which could lead to potential inflammation and cell toxicity.²¹ Thus, the purification process requires large amount of distilled water for at least a week to wash out excess ions.²⁰ Another method is the use of secondary dopants, including sulfuric acid and dimethyl sulfoxide, as both crosslinkers and dopants for PEDOT:PSS.^22,23 However, two major drawbacks limit the applicability of this approach. First, the resulting hydrogels have a high elastic modulus (~ 2 MPa to 6.5 GPa) that is difficult to tune. Second, the preparation of hydrogels requires extra fabrication steps such as dry-annealing and rehydration of PEDOT:PSS films.^22,23 A simpler approach towards conductive hydrogels involves using pre-formed hydrogels as scaffolds for the polymerization of conducting polymers. This approach has the advantage that the mechanical properties can be more precisely controlled by tuning the degree of crosslinking of the pre-formed hydrogel. Ding et al. reports the use of biologically derived heparins hydrogel as a scaffold material for the in-situ polymerization of PANI.²⁴ While the materials have a low storage moduli (900 ± 100 Pa), they exhibit a high impedance. Conductive hydrogels prepared with PPy also present challenges. The formation of a percolating conductive network within the hydrogels is limited by the diffusion of the pyrrole monomer and PPy into hydrogel. Most hydrogels prepared with PPy are therefore comprised of relatively large aggregated and crystalline particles of PPy.^25,26 These aggregates are often mostly restricted to the surface of the hydrogel rather than homogeneously distributed in the bulk, leading to lower bulk conductivity. Alternatively, PEDOT-based hydrogels typically show amongst the highest conductivity and lowest impedance, but are difficult to prepare from a pre-formed hydrogel due to the poor solubility of the EDOT monomer in water.¹¹ To address this solubility problem, Rivnay et al. have integrated the water-soluble, self-doped alkoxysulfonate-functionalized PEDOT (PEDOT-S) into collagen to form composites with comparable physical properties to pristine collagen scaffolds.²⁷ However, the electronic conductivity of the PEDOT-S hydrogels was not reported, possibly due to the low percolation of PEDOT-S.

Herein, we instead report the use of a water-soluble and neutral EDOT derivative bearing a diethylene glycol methyl ether side chain (EDOT-DEG) (Figure 1a). Similarly to previously described approaches, EDOT-DEG can be polymerized in pre-formed hydrogels (Figure 1) to achieve electronically and ionically conductive hydrogels. Compared with previous monomers used in this simple two-step approach, we found that the mechanical properties of the hydrogels were not significantly affected by the introduction of the conductive polymer. Hydrogels containing PEDOT-DEG therefore have low elastic moduli (5–15 kPa), and are ionically and electronically conductive (~ 0.3 S m⁻¹ without secondary dopants). We demonstrated that this strategy is generalizable to a range of hydrogels including anionic PSS hydrogels (Figure 1b), a naturally-occurring polysaccharide, agarose (Figure 1c), and a neutral poly(ethylene glycol) diacrylate (PEGDA)-based hydrogel (Figure 1d). This simple and general approach to the formation of conductive hydrogels with PEDOT-DEG is likely to inform the design of materials for new applications in bioelectronics.

Figure 1. Preparation of electronically-conductive hydrogels.

a) Chemical structure of the water-soluble EDOT-DEG monomer. b) Chemical structure of the precursors to a PSS-based hydrogel (H1). c) Chemical structure of agarose. d) Chemical structure of the precursor to a PEGDA hydrogel. e) Schematic overview of the preparation of the conductive hydrogels. Step i) Loading of EDOT-DEG monomer in the pre-formed hydrogel. Step ii) Oxidative polymerization of PEDOT-DEG.

2. Results and Discussion

To achieve a simple preparation of conductive hydrogels, we hypothesized that EDOT-DEG would be a good candidate for its in-situ polymerization in hydrogels. Due to its ethylene glycol side chain, EDOT-DEG should be water-soluble, allowing for the homogeneous distribution of the electronically conductive polymer inside the hydrogel upon polymerization. In addition, ethylene glycol sides chains have been previously shown to lower the elastic modulus³² and increase the ionic conductivity³³ when compared to PEDOT while maintaining its stability in water and oxygen, and electronic mobility. Following a similar procedure as Roncali et al.,^28,29 we first synthesized EDOT-DEG from the commercially available hydroxymethyl EDOT (Scheme 1). The structure and purity of the monomer was confirmed by nuclear magnetic resonance (¹H NMR and ¹³C NMR, Figure S14) and mass spectrometry. We studied the solubility of EDOT-DEG monomers compared to that of EDOT monomers by ¹H NMR (Figure S15). The water-soluble derivative, EDOT-DEG is approximately four times more soluble than EDOT, 2.1 g/L versus 9.1 g/L (Figure S15).³⁴ The higher solubility in water should ensure a more homogeneous distribution within the hydrogel network.

Scheme 1.

Synthesis of the EDOT-DEG monomer: a water-soluble and neutral derivative of EDOT.

To choose an appropriate oxidant for the chemical oxidative polymerization, we studied the electrochemical behavior of EDOT-DEG, compared to that of EDOT, by cyclic voltammetry (CV). Among their CV scans, the polymerization of EDOT and EDOT-DEG was clearly noticed as the current surged around 0.65 V (Figure S2). The polymerization potential of EDOT-DEG was then determined around 0.72 V which is lower than the oxidation potential of Iron (III), around 0.77 V. Thus, iron(III) chloride, was chosen for the chemical oxidative polymerization of EDOT-DEG in pre-formed hydrogels. To establish the feasibility of our approach, we first chose to focus on PSS-containing hydrogels, as the presence of anionic sulfonate groups could help stabilize the doped form of the p-type conducting polymer.³⁵ The PSS hydrogels were synthesized under standard free radical polymerization conditions using sodium 4-styrenesulfonate (SSNa), ammonium persulfate (APS) as an initiator, and poly(ethylene glycol) diacrylate (PEGDA, M_n = 700 Da) as a crosslinker.³⁰ For our initial studies, we chose a 52:1 molar ratio of NaSS to PEGDA to achieve a hydrogel with an elastic modulus in the 1 – 9 kPa range (control studies with a 26:1 and 13:1 molar ratio gave moduli of 31 kPa and 29 kPa, respectively). Upon polymerization and crosslinking, the resulting hydrogel (H1) assumed the shape of the mold. EDOT-DEG monomers were diffused into the hydrogel by soaking gels in an aqueous monomer solution (0.7 wt%) overnight (Figure 1e). To prove that the hydrogels are homogeneously immersed with monomers, electropolymerization of EDOT-DEG in the hydrogel was conducted (Figure S4). The emergence of a gradient blue color at the working electrode (Figure S5) indicates the homogeneous polymerization of PEDOT-DEG. Polymerizing EDOT-DEG throughout the entire thickness of the hydrogel (~1 cm) even after switching the working electrode, however, proved difficult by electropolymerization as the PEDOT-DEG is primarily located within the first few millimeters of the electrode (Figure S5). In addition, as an electrochemical polymerization set-up is not readily accessible in every laboratory, we rather focused on the chemical oxidative polymerization of EDOT-DEG. Thus, we polymerized EDOT-DEG with iron(III) chloride (FeCl₃) as an oxidant (Figure 1e) to achieve electronically-conductive hydrogels. To simplify the notation of these hydrogels, we will refer to them as C-H1 to indicate a conductive (C) hydrogel derived from H1. We initially chose a 1.5:1 molar ratio of oxidant to monomer, similar to that used in the synthesis of the aqueous PEDOT:PSS.³⁶ As we varied the molar ratio of oxidant to monomer, the electronic properties, as measured by electrochemical impedance spectroscopy (EIS), did not show significant difference (Entries 1, 2, 3 in Table S1, Table S2, and Figure S6a). We also probed the effect of the concentration of EDOT-DEG monomer by doubling the concentration to 1.4 wt%. The conductive hydrogel showed similar conductivity and impedance (Entries 2 and 4 in Table S1, Table S2, and Figure S6b). The results suggest that increasing the concentration of monomer does not lead to higher loading and/or percolation of conducting polymer, and monomer saturation in the hydrogel is already reached at 0.7 wt%.

Next, the microstructure of the hydrogels was studied by scanning electron microscopy (SEM) on freeze-dried cross-sections. The SEM images of both H1 (Figure 2a) and C-H1 (Figure 2b) revealed a 3D interconnected porous structure. Before polymerization, H1 exhibits pore sizes ranging from 50 μm to over 100 μm. After the polymerization of EDOT-DEG, a cross-section SEM of C-H1 (Figure 2b) shows an homogeneous network, with a higher network density than H1. The average width of these pores was found to be approximately 20 μm. Overall, H1 possesses a highly porous structure which facilitated the incorporation of water-soluble EDOT-DEG monomers. Both the pictures of the hydrogels (Figure 2 insets) and the SEM images show that the conductive polymer is homogeneously distributed within the hydrogel scaffold. Importantly, unlike for hydrogels containing PPy, no crystalline aggregates were observed and the porous structure was maintained despite a reduction in the pore size.

Figure 2. Microstructure and appearance of the hydrogels before and after polymerization of EDOT-DEG.

a) SEM image of the cross-sections of freeze-dried H1 and b) C-H1. The insets show photographs of the hydrogel slabs.

In addition to the obvious change in color upon polymerization, the presence of PEDOT-DEG was confirmed by X-ray photoelectron spectroscopy (XPS) of freeze-dried samples (Figure 3). The S(2p) electrons of PEDOT-DEG and PSS have different binding energies, so the ratio of PSS/PEDOT-DEG can be analyzed by XPS.³⁷ The XPS spectra of H1 shows the S (2p) peaks at a binding energy of 169 eV corresponding to the sulfur in the sulfonate groups of PSS. In the XPS spectra of C-H1, new S(2p) signals are observed at 165 eV, corresponding to the sulfur in the thiophene fragment of PEDOT-DEG. The area ratio of the S(2p) peaks can therefore be used to estimate the relative composition of PSS to PEDOT-DEG of samples. The ratio of PSS to PEDOT-DEG of the conductive hydrogels is 1.56:1. This ratio is comparable to that of films of PEDOT:PSS after treatment with methanol.³⁸ The amount of PEDOT-DEG is approximately 2.8 wt% with respect to the hydrogel (93.9 wt% of water).

Figure 3.

S (2p) XPS spectra of the hydrogel H1 and the conductive hydrogel C-H1.

Next, we established the electronic properties of C-H1 compared with H1. We first measured the resistance of the hydrogel between two parallel plate electrodes to determine their bulk conductivity (Table 1). The conductivity, is nearly 55 times higher for the C-H1 conductive hydrogels (0.32 S m⁻¹) than the H1 hydrogels (5.85 ×10⁻³ S m⁻¹).

Table 1.

Electronic and mechanical properties of H1 and C-H1.

Entry	Sample	Conductivitya) [S m−1]	Rea) [Ω]	Ria) [Ω]	χ2	Elastic modulusb) [kPa]	G’c) [kPa]	G”c) [kPa]
1	H1	0.006 ± 0.001	~105	187.3 ± 84.3	0.04	7.1 ± 1.0	0.13	0.02
2	C-H1	0.320 ± 0.012	123.3 ± 30.1	20.0 ± 3.9	0.04	8.4 ± 1.3	0.42	0.01

^a)Average of 3 separate samples of approximately 0.4 cm³;

^b)Measured by indentation tests. Average of 5 measurements.

^c)Measured by rheology.

To gain better insights into the relative ionic and electronic conductivities, we performed electrochemical impedance spectroscopy (EIS). The Bode plots (Figure 4a) showed that, compared to H1, conductive hydrogels C-H1 displayed significantly lower impedance (over two orders of magnitude), especially in the low frequency range. At 0.1 Hz, the impedance magnitude of CH-1 is 159 ± 29 Ω. Nyquist plots (Figure 4b) of the hydrogels were analyzed and fitted to the equivalent circuit model shown in the inset.³⁹ The small χ² value (0.04, Table 1) indicates a good fit for the model. In addition, previous reports of conductive hydrogels derived from conducting polymers used the same equivalent circuit models under a similar cell geometry.^40–42 The values of R_c, R_e, and R_i are respectively the resistive contributions from the assembled cell used for the test, and the electronic and ionic resistance from the conductive hydrogel. C_g is the ideal geometric capacitance of conductive hydrogels. Q_dl is the constant phase element describing the non-ideal double-layer capacitance. Values of all the electrical components are tabulated in Table S3, Entry 2. The average values for R_e and R_i over three independently prepared samples are also tabulated in Table 1. We found that both R_e and R_i of C-H1 are lower than those of H1, with the most dramatic decrease (two orders of magnitude) observed for the electronic resistance. This decrease in electronic resistance is consistent with the incorporation of an interconnected conducting polymer (doped PEDOT-DEG). The electronic conductivity, calculated from the resistance and dimension of hydrogels, is 0.32 S m⁻¹, which agrees with the conductivity measured from the 2-point parallel plate measurements,⁹ while the ionic conductivity is 2 S m⁻¹. Furthermore, our resistance values are comparable with other PEDOT-based hydrogels.^40,41 It should be noted, however, that no secondary additives to enhance the conductivity were used in this procedure. The absence of potentially toxic small molecules reduces the risk of cytotoxicity for applications in bioelectronics.

Figure 4.

Electrochemical impedance spectroscopy studies. a. Bode plots of H1 and C-H1 hydrogels. b. Nyquist plot obtained for the conductive hydrogel C-H1. The inset shows the equivalent circuit model.

As detailed above, the elastic modulus is an important parameter to consider for applications in implanted electronics. The elastic modulus of the hydrogels was measured by indentation testing (Figure 5a). Given a sample thickness of 6 mm and indenter radius of 9 mm, the elastic modulus is calculated by $E=\frac{4}{3F}\sqrt{{\delta }^{3}R}$ for the contact between a rigid indenter and an elastic half-space,⁴³ where F is the measured force and δ is the indentation depth. The indenter was moved relatively slowly to minimize viscoelasticity (0.04 mm/s) so the elastic modulus can be extracted from the slopes of those lines (Figure 5a).⁴³ The elastic modulus of the pre-formed hydrogel H1 was 7.14 kPa (Table 1). After the PEDOT-DEG polymerization, the resulting conductive hydrogels C-H1 maintains an elastic modulus of 8.44 kPa, which is close to its pre-formed hydrogel. This elastic modulus is amongst the lowest ever reported for electronically conductive hydrogels.^23,40,44 By using oscillatory rheological measurements, we quantified the storage modulus (G′) and loss modulus (G″) as a function of frequency of the hydrogels before and after the PEDOT-DEG polymerization (Table 1 and Figure S10). These two moduli quantitatively reveal the elastic and viscous components of the samples. Both H1 and C-H1 exhibit G′ greater than their G″ demonstrating solid-like gels. Additionally, the values of G′ and G′′ are nearly constant across the frequency sweep which means the samples are nearly independent of frequency, even after oxidative polymerization. These characteristics are consistent with strong association and a well-structured gel due to a high degree of crosslinking. As seen in Table 1, the storage modulus of C-H1 is slightly higher than that of H1, 0.42 vs 0.13 kPa, but remains within the same order of magnitude, and the low modulus of the hydrogel is preserved upon EDOT-DEG polymerization. The loss modulus slightly decreased from 0.02 kPa to 0.01 kPa. These results are consistent with the introduction of a solid-like polymer in the hydrogel, but overall the introduction of PEDOT-DEG did not significantly affect the rheological properties of the hydrogel. We also characterized the change in resistance (R/R₀) under mechanical deformation by measuring the resistance of C-H1 while cycling back and forth between 0 and 10% compressive strain for 10 cycles (Figure 5b). We did not pursue these tests over longer periods of time to minimize water loss. Upon compression, the resistance of the conductive hydrogel slightly decreases (3.5%) consistent with a reduction in the sample thickness. The electrical properties of the gel were relatively sustained within this region, with a low gauge factor of 0.35, determined by the ratio between the evolution of the resistance (ΔR/R₀) and the compressive strain. We also observed a slight “break-in” behavior whereby the resistance does not go back to its initial value upon decompression (1.5% change). This behavior could be explained by a strain-induced alignment of the conductive polymer chains within the hydrogel or Mullins effect, or dehydration of the conductive hydrogel, as shown to occur within minutes under ambient conditions during drying studies (Figure S17).

Figure 5. Mechanical properties of H1 and C-H1.

a) Representative example of mechanical indentation tests on H1 and C-H1 samples. b) Change in resistance, expressed as a ratio between resistance (R) and initial resistance (R₀), of C-H1 as it is cycled reversibly between 0 and 10% compressive strain for 10 cycles. Sample had an initial thickness of 6 mm.

Next, we investigated the applicability of our synthetic approach on other hydrogels. We chose hydrogels commonly used in tissue engineering applications: neutral poly(ethylene glycol) diacrylate (PEGDA)⁴⁵ and naturally-occurring agarose.³¹ Similar to the PSS-based hydrogels, we observed a large increase in conductivity in both C-PEGDA and C-agarose when compared to the hydrogels without PEDOT-DEG. The conductivity is nearly 50 times higher for the C-PEGDA conductive hydrogels (0.24 S m⁻¹, Table 2, Entry 2) than the PEGDA hydrogels (4.91 ×10⁻³ S m⁻¹, Table 2, Entry 1). Similarly, the conductivity of C-Agarose hydrogel (0.34 S m⁻¹, Table 2, Entry 4) is 75 times higher than that of pure agarose hydrogels (4.53 ×10⁻³ S m⁻¹, Table 2, Entry 3). One potential concern was that the absence of sulfonate groups in these commonly used hydrogels may prevent the stabilization of the doped PEDOT-DEG and lead to low conductivity. However, the conductivity remained close to that of C-H1, suggesting that the doped form of PEDOT-DEG is sufficiently stabilized by mobile ions in the hydrogels (from PBS or chloride anions from the FeCl₃ oxidant).

Table 2.

Electronic and mechanical properties of the PEGDA and Agarose hydrogels and their corresponding conductive hydrogels.

Entry	Sample	Conductivitya) [S m−1]	Rea) [Ω]	Ria) [Ω]	χ2	Elastic modulusb) [kPa]	G’c) [kPa]	G”c) [kPa]
1	PEGDA	0.005 ± 0.001	~105	268.0 ± 15.2	0.04	14.4 ± 1.1	0.93	0.12
2	C-PEGDA	0.236 ± 0.004	1029.4 ± 31.1	130.6 ±11.5	0.04	14.8 ± 0.8	1.52	0.03
3	Agarose	0.005 ± 0.001	~104	57.7 ± 15.4	0.06	7.1 ± 1.4	1.57	0.05
4	C-Agarose	0.340 ± 0.004	293.2 ± 52.0	27.1 ± 6.3	0.04	5.7 ± 1.2	2.44	0.06

^a)Average of 3 separate samples of approximately 0.4 cm³;

^b)Measured by indentation tests. Average of 5 measurements.

^c)Measured by rheology.

To determine the relative contribution of the electronic and ionic resistance, we again turned to EIS. The Bode plots showed that, compared to pre-formed hydrogels, conductive hydrogels consisting of PEDOT-DEG displayed significantly lower impedance, especially in the low frequency range than the parent hydrogel (Figure S7). Nyquist plots (Figure S8) of all samples were analyzed and fitted to the same equivalent circuit model shown in the inset of Figure 4b. The small χ² values (Table 2) shows that this model still fits with these hydrogels. The electronic (R_e) and ionic resistance (R_i) of each hydrogel were averaged over three independently prepared samples and tabulated in Table 2. The electronic conductivity of C-PEGDA and C-Agarose, calculated from the resistance and dimension of hydrogels, are 0.038 S m⁻¹ and 0.136 S m⁻¹ respectively. The ionic conductivity of C-PEGDA and C-Agarose are 0.306 S m⁻¹ and 1.476 S m⁻¹. The values of R_e significantly decreased with the hydrogels containing PEDOT-DEG. The ionic resistance dropped by approximately half, showing that the addition of the conductive polymer does not prevent ionic transport but rather slightly enhances it. Indentation testing before and after the polymerization of PEDOT-DEG showed that the elastic moduli of the conductive hydrogels are essentially the same as their pre-formed hydrogels (Table 2 and Figure S9). The elastic modulus of C-PEGDA is 14.41 kPa, compared to PEGDA hydrogel’s modulus of 14.81 kPa. C-Agarose has a slightly lower elastic modulus than Agarose hydrogel of 5.69 versus 7.05 kPa. Oscillatory rheology shows similar trends as for H1 and C-H1 (Figures S11 and S12), with a slight increase in the storage modulus, and a decreased or almost identical loss modulus. These experiments confirm the introduction of a stiffer polymer within the hydrogel, but the overall order of magnitude in mechanical properties is similar and the low modulus properties of the hydrogel are preserved after EDOT-DEG polymerization. As observed with C-H1, the microstructure of the hydrogels remains similar before and after polymerization of PEDOT-DEG (Figure S13) which could explain the minimal effect on the mechanical properties. The porous structure is maintained, with what appears to be a thin coating of amorphous conductive polymer which only slightly decreases the pore size.

Lastly, we found that the polymerization of the conductive polymer inside the hydrogel does not affect the macroscopic size and shape of the hydrogel. As seen in Figure 6, hydrogels prepared by mold casting maintain the shape and size within 2% after incorporation and polymerization of PEDOT-DEG.

Figure 6. Retention of the shape of the hydrogel after polymerization of PEDOT-DEG.

a) Picture of the PEGDA hydrogel obtained from mold casting. b) Picture of the conductive hydrogel C-PEGDA.

3. Conclusion

We have reported a simple and rapid method that involves an in-situ oxidative polymerization of a water-soluble EDOT-DEG into a range of hydrogels including crosslinked poly(styrene sulfonate), agarose, and poly(ethylene glycol) diacrylate (PEGDA). By increasing the water solubility of the conducting polymer precursor using ethylene glycol side chains, electrically-conducting hydrogels were obtained with good electronic properties while maintaining the mechanical properties of the parent hydrogel. The conductive hydrogels can be easily molded into different shapes, and retain their geometry and porosity after introduction of the conducting polymer. Our straightforward approach to conducting hydrogels could lead to new materials for tissue engineering, and implantable electronics.

4. Experimental Section/Methods

Materials:

Sodium 4-styrenesulfonate (NaSS), poly(ethylene glycol) diacrylate (PEGDA, M_n = 700 Da), agarose (genetic analysis grade, powder), ammonium persulfate (APS), N,N,N′,N′-tetramethylenediamine (TEMED), iron(III) chloride (FeCl₃), phosphate buffered saline (1x PBS), 1-bromo-2-(2-methoxyethoxy)ethane, hydrochloric acid (HCl), sodium bicarbonate (NaHCO₃), magnesium sulfate (MgSO₄), and sodium chloride (NaCl) were purchased from Sigma-Aldrich and used without further purification. Hydroxymethyl EDOT was purchased from Sarchem Laboratories, Inc and used without further purification.

C223BT screen-printed electrodes (SPEs) were purchased from Metrohm. These electrodes are composed of a 1.60 mm diameter circular gold working electrode, surrounded by the gold auxiliary and pseudo-silver reference electrode (+0.066 V vs. SHE). Polystyrene cuvettes were purchased from Thermo Fisher Scientific with 3 ml solution capacity and 10 mm path length. 12.70 mm wide copper foil tape was purchased from Zehhe.

Synthesis of EDOT-DEG:

The procedure was adapted from the previously reported synthesis of EDOT-DEG by Roncali et al.^28,29 Hydroxymethyl 3,4-ethylenedioxythiophene (1.033 g, 6 mmol) and sodium hydride (0.288 g, 12 mmol) were dissolved in dry THF (10 mL) under N₂ for 3 hours at room temperature. Then, 1-bromo-2-(2-methoxyethoxy)ethane (1.2 mL, 9 mmol) was added dropwise to the mixture at 0 °C. The reaction was stirred at room temperature for 24 hours. The reaction mixture was washed with HCl, NaHCO₃ and saturated NaCl, before drying over MgSO₄. EDOT-DEG was purified by silica gel chromatography, using a mixture of CH₂Cl₂ and MeOH (9:1) as the eluent. EDOT-DEG was obtained as a yellow liquid in 75% yield. ¹H NMR (400 MHz, CDCl₃, 27 °C) δ[ppm]: 6.24 (q, 2H), 4.25 (m, 1H), 4.18 (dd, 1H), 4.01 (q, 1H), 3.69 (q, 1H), 3.61 (m, 7H), 3.48 (m, 2H), 3.31 (s, 3H). ¹³C NMR (400 MHz, CDCl₃, 27 °C) δ[ppm]: 99.67, 99.57, 77.22, 77.01, 76.80, 72.65, 71.95, 71.24, 70.63, 69.68, 66.61, 59.05. MS (ESI): 275 (M⁺ 100%).

Preparation of the PSS hydrogels:³⁰

NaSS (0.96 g, 4.66 mmol) was dissolved in 7 mL of deionized water and heated at 40 °C until all solids were dissolved. APS (0.009 g, 0.04 mmol) was added as a radical initiator with PEGDA (0.064 g, 0.09 mmol) crosslinker, and the mixture was purged with N₂ for 30 min. Finally, an aqueous solution of TEMED (a total volume of 100 μL) was added as a catalyst. The solution was casted into a rectangular polycarbonate mold, for 3 hours to form the hydrogels. The resulting gels were washed with DI water multiple times to ensure residual monomers were washed out.

Synthesis of the agarose hydrogels:³¹

Agarose (0.05 g, 0.16 mmol) was dissolved in 0.15 M phosphate buffered saline (PBS) for the final concentration of 0.5% weight by volume.³¹ The solution was mixed and heated at 120 °C for 25 mins. The solution was then poured into the mold and gelled within 10 mins at room temperature.

Synthesis of the PEGDA hydrogels:

PEGDA (M_n = 700 Da) (0.324 g, 0.46 mmol) and APS (0.009 g, 0.04 mmol) were dissolved in 5 mL of deionized water and heated at 40 °C. The mixture was purged with N₂ for 30 min. Finally, an aqueous solution of TEMED (a total volume of 100 μL) was added as a catalyst. Upon polymerization and crosslinking, the resulting hydrogel assumed the shape of the mold.

General procedure for the synthesis of the conductive hydrogels:

The pre-formed hydrogel was soaked in 10 mL of DI water along with 70 μL EDOT-DEG for 24 hours to ensure monomer diffusion into the hydrogel networks. Then, the hydrogel soaked EDOT-DEG was submerged in another 10 mL solution of 0.1 M FeCl₃. The solution surrounding the hydrogel was stirred at room temperature for 24 hours. The conductive hydrogel was washed throughly by submerging in 3× 20 mL deionized water and 3× 20 mL 1M HCl over 3 hours. Lastly, the resulting conductive hydrogels were submerged in a phosphate buffered saline (1× PBS) solution overnight before measuring the electronic and mechanical properties. The removal of the iron oxidant was confirmed by X-ray photoelectron spectroscopy (XPS) of freeze-dried samples in the range of 700 eV to 740 eV, corresponding to the Fe2p region (Figure S13).