Flexible sensor with electrophoretic polymerized graphene oxide/PEDOT:PSS composite for voltammetric determination of dopamine concentration

We demonstrate a novel, flexible sensor with graphene oxide/PEDOT:PSS (GO/PEDOT:PSS) composite for voltammetric determination of selective low levels of dopamine. The well-distributed GO and EDOT:PSS suspension in water were deposited simply and polymerized. Consequently, the EDOT:PSS provided a strong interaction between GO and PEDOT:PSS, and it also had well-tailored interfacial properties that allowed the highly selective and sensitive determination of DA. Since the interfacial net charge is well-constructed, the sensor satisfies both the requirements of selectivity and the highly sensitive detection of low amounts of DA. In the results, the sensor with the GO/PEDOT:PSS composite exhibited a low interfacial impedance of about 281.46 ± 30.95 Ω at 100 Hz and a high charge storage capacity (53.94 ± 1.08 µC/cm2) for the detection of dopamine. In addition, the interference from ascorbic acid was reduced effectively to a minimum by electrostatic charge repelling of the AA and the distinct difference for the oxidation peak of the UA. Due to the fact that the GO/PEDOT:PSS composite had a net negative charge and, enhanced interfacial properties, the sensor showed a dopamine detection limit of 0.008 μM and a sensitivity of 69.3 µA/µMcm2.

ure 1a show the conceptual drawing for the process of fabricating the GO/PEDOT:PSS on a thin Au working electrode of the fabricated flexible sensor and the configuration of the fabricated sensor with the GO/PEDOT:PSS.
The PEDOT:PSS as a doping agent for increasing the interfacial properties of the GO layer can be obtained from electro-polymerization of the charge-balanced EDOT monomer with PSS 27 . The conductivity of PEDOT can be enhanced by doping with hydrophilic segments of PSS, which can stabilize the dispersion of EDOT:PSS in aqueous solution.
When the GO and EDOT:PSS were mixed, the PSS and EDOT chains were separated by the weakened coulombic attraction between PSS and EDOT, which was caused by addition of various functional groups (such as -COOH and -OH) of the GO domain 28 . These separated EDOT chains can interact with the GO nanosheets to extend the conductive network, which means the interaction between the GO sheets and PEDOT can help to form more conductive pathways 29 .
In order to find an optimal condition for the GO/PEDOT:PSS composite, the GO/PEDOT:PSS composites with various composite ratios (i.e., GO : EDOT:PSS = 1:1, 2:1, 5:1, and 10:1) and polymerization times (i.e., 50, 150, 300, and 600 s) were fabricated, and their impedance and CV properties were compared. Supplementary  Fig. S1 shows that the interfacial impedance of the GO/PEDOT:PSS composite was increased sharply as the decrements of EDOT:PSS were added (107.62 ± 0.33, 117.89 ± 0.38, 281.46 ± 30.95, and 2429.61 ± 98.74 Ω), while the charge storage capacity (CSC) of the GO/PEDOT:PSS composite, i.e., the real activation area, was decreased as the decrements of EDOT:PSS were added (70.81 ± 0.14, 63.97 ± 0.67, 53.94 ± 1.08, and 20.36 ± 0.01 μC/cm 2 ). As expected, the PEDOT:PSS, as a representative conductive polymer, affected the electrochemical properties of the GO/PEDOT:PSS composite. However, the DPV peak current responses to the various DA concentrations of the GO/EDOT:PSS composite that was prepared (i.e., 1:1, 2:1, 5:1, and 10:1) showed that the GO/EDOT:PSS compo-site with the 5:1 condition had the lowest limit of detection (0.01 μM) and linearity (R 2 = 0.9636) in the DA concentration range up to 0.7 μM (Supplementary Fig. S2). This might be caused that the remained oxygencontaining groups of GO after polymerization (most of the negatively charged functional groups of GO are interacted with the PEDOT backbone) impart to the electrostatic interaction between composite electrode and the DA molecules 30 . In addition, the 300 s polymerization time had the lowest interfacial impedance and the highest CSC value among the other polymerization time conditions, as shown in Supplementary Fig. S3. This was caused by the increased surface activation area with time by the polymerized EDOT:PSS. In the case of a polymerization www.nature.com/scientificreports/ time of 600 s, the interface impedance and CSC of the GO/PEDOT:PSS composite was increased and decreased more than the 300 s condition, which might be caused by the thickly formed GO layer, which must be degraded by the interfacial properties that induce slow adsorption and electron transfer kinetics 31 . Therefore, we selected the 5:1 mixture with the 300 s electropolymerization time for the growth of the GO/PEDOT:PSS composite.  46 ± 30.95 Ω, respectively. In the measured CV curve, the CSC value which means accumulated charges, was expanded gradually in the order of Au < GO < GO/PEDOT:PSS < PEDOT:PSS that were 8.16 ± 0.03, 33.20 ± 0.3, 53.94 ± 1.08, and 240.98 ± 5 μC/cm 2 , respectively. Although the interfacial impedance of the GO:PEDOT/PSS composite electrode was relatively lower than the GO sheet and had a more enlarged CSC than the GO sheet due to the addition of PEDOT:PSS, their interfacial properties fell short of the lowest interfacial impedance and the largest CV curve of the PEDOT:PSS. The interfacial impedance and CSC of the GO/PEDOT:PSS composite can be controlled by adding more EDOT:PSS. The PEDOT:PSS was not appropriate for the DPV based DA determination due to brittle adhesion and low detection property (data was not shown). Figure 2 clearly shows the surface morphologies of the fabricated Au, GO, PEDOT:PSS, and GO/PEDOT:PSS (mixture ratio of 5:1, polymerization time of 300 s). The electrodeposited GO layer onto the thin Au electrode showed the typical wrinkling structure, while the pristine PEDOT:PSS layer that was polymerized onto the thin Au electrode exhibited a homogeneous distribution of the nanoparticles like grain of sand. In the surface morphology of the GO/PEDOT:PSS composite, it seemed that the PEDOT:PSS nanoparticles were well distributed along the ridges formed by the GO layer. These results must have a high correlation with the measured interfacial impedance and CV in Fig. 1b Supplementary Table S2). Typical carbon spectra (C1s) of GO can be fitted to the three peaks at 284.6, 286.6, and 287.9 eV. These components can be assigned to the C-C, C-O, and C=O bands, respectively 35 . It is obvious that the peak intensities of C-O and C=O were strong in GO, which should be compared with the weak intensities of PEDOT:PSS. After the growth of the GO/PEDOT:PSS composite, the C-C peak had a strong intensity due to the aromatic rings of the GO/PEDOT:PSS composite. In  Supplementary Fig. S7. The peak currents of DA was increased gradually up to pH 7.4 and the peak potential was also shifted negative potential as increment of pH value. The Supplementary Fig. S7b showed the relationship between the peak current and peak potential of DA to the different pH values. It can be found that the E pa values shift negatively with increasing pH value from 5.0 to 9.0, indicating that the redox reaction of DA on GO/PEDOT:PSS were accompanied by proton transfer 40 . The linear regression equation for peak potentials and pH could be expressed as E pa (V) = − 0.069 pH + 0.733 (R 2 = 0.9990). The slope − 69 mVpH −1 for DA was close to a theoretical value of − 59 mVpH −1 that given by the Nernstian equation for equal number of two electrons and two proton transfer process 41,42 . The reaction kinetics of DA at GO/PEDOT:PSS composite to the scan rate variation was investigated by cyclic voltammetry. Figure 4a shows the cyclic voltammograms of the sensor with the GO/PEDOT:PSS composite at various scan rates (10-100 mVs −1 ) in PBS (pH = 7.4) that contained 1 mM DA. At the scan rate of 10 mVs −1 , we measured and calculated a pair of redox peaks, the ratio of anodic and cathodic peak currents (I pa /I pc ), i.e., approximately 1.05, and a peak-to-peak separation (ΔE P ) of about 60.42 mV 41 . The anodic peak potential of DA shifted positively and the cathodic peak potential of DA shifted negatively as the scan rate increased. These results imply that the electrochemical oxidation/reduction of DA is fully reversible 44 . I pa and I pc as functions of scan rate were plotted in Fig. 4b. Two linear regression equations were obtained, i.e., I pa (μA) = 0.2716ν 1/2 ((mVs −1 ) 1/2 ) − 0.06766 (R 2 = 0.9486) and I pc (μA) = − 0.2987ν 1/2 ((mVs −1 ) 1/2 ) − 0.7718 (R 2 = 0.9317). To further investigate the electrochemical oxidation, the sensor with the GO/PEDOT:PSS composite was tested in the presence of 5 μM DA, 1 mM AA, and 50 μM UA at different scan rates. As shown in Fig. 4c, the oxidation current peaks of DA and UA clearly were separated based on the increment of scan rate with a peak separation (ΔE P ) of 292 mV. The negatively charged AA (pKa = 4.10) 16 reaction was not observed, since the surface of the GO/PEDOT:PSS has been shown to have a net negative charge due to the presence of oxygen-containing functional groups on the edges of the GO subunits 45 . It is obvious that the oxidation currents of UA and AA do not affect the determination or detection of DA. In addition, the slopes of the oxidation / reduction current peaks to the 5 μM DA were I pa (μA) = 0.0910ν 1/2 ((mVs −1 ) 1/2 ) + 0.1406 (R 2 = 0.9891) and I pc (μA) = − 0.0089ν 1/2 ((mVs −1 ) 1/2 ) − 0.2202 (R 2 = 0.9478), respectively (Fig. 4d). As the scan rate was increased, the oxidation and reduction current peaks (I pa and I pc ) increased, and anodic and cathodic peak potentials were shifted positively and negatively, respectively.  To further investigate the selective oxidation of DA without being affected by interfering species, the sensor with the GO/PEDOT:PSS composite was tested in various concentrations of DA that contained physiological concentrations of AA and UA. As illustrated in Fig. 5c, d, the oxidation current peaks showed linear responses to the change of DA concentrations (range from 0.01 to 100 μΜ) in the presence of 1 mM AA and 50 μΜ UA. The linear regression equations were expressed as I pa (μA) = 0.4768C DA (μΜ) + 0.5409 (R 2 = 0.9790) and I pa (μA) = 0.0812C DA (μΜ) + 4.7784 (R 2 = 0.9972) from 0.01 to 10 μΜ (10.5 μA/μMcm 2 ) and from 10 to 100 μΜ (1.8 μA/μMcm 2 ) (n = 3), respectively. The potential region where the DA oxidation current peak appears had shifted slightly positively, unlike when DA alone was oxidized, which might be caused by the affect of the change in the interfacial charge during the repelling of AA. An oxidation current peak by AA was not observed, and, despite the increment of the DA concentration, the oxidation current peaks by UA underwent negligible changes in a separated potential region. The performance of the fabricated flexible sensor with the GO/PEDOT:PSS composite was compared with other previous works that are shown in Supplementary Table S3. As shown in Supplementary Table S3, the sensor with GO/PEDOT:PSS composite exhibited comparable performance with reported high sensitive DA detection electrodes for the clinical level DA detection. Unlike the previous reported works, this sensor is favourable for various flexible sensor application based on high sensitive and selective DA sensor (especially free from AA oxidation) in a wide detection range.

Reproducibility and stability of the sensor with the GO/PEDOT:PSS composite. The sensor
with the GO/PEDOT:PSS composite and the GO sheet were prepared, and their responses were compared to the periodic DPV measurement, as shown in Supplementary Fig. S8. In a Supplementary Fig. S8a, it was found that the current response to 10 μM DA of the sensor with the GO sheet was decreased sharply in 10 cycles of DPV scanning (current change of 57.1%), while the sensor with the GO/PEDOT:PSS composite exhibited remarkable stability (current change of 9.02%). This might be caused by the oxygen-containing groups of the GO sheets, which allow easy swelling and disperse in water and some other solvents 46,47 . These drawbacks of the GO sheets can be overcome by combining them with PEDOT:PSS. The well-dispersed EDOT and PSS molecules in aqueous solution and the functional groups of GO may lead to strong interaction with the in situ polymerized PEDOT from EDOT. The chemical interaction between the negatively-charged, oxygen-containing groups, such as epoxides, hydroxides, and, the carboxyl groups of GO and positively charged PEDOT chains, plays a crucial role in the stability of the GO/PEDOT:PSS composites.  Fig. S8, sensor with Au and electrode exhibited low linearity and relatively higher detection limit than that of GO/PEDOT:PSS, while the sensor with GO/PEDOT:PSS showed high linearity and low detection limit. In the case of the sensor with PEDOT:PSS, although it showed high sensitivity, it's not good enough for low level of DA.
Real sample. In order to prove its practical feasibility, the fabricated sensor with the GO/PEDOT:PSS composite was investigated to determine the DA serum sample. All of the serum samples were diluted 50 times with PBS (pH 7.4) before measurement. No other pretreatment process was performed. Table 1 provides a comparison of the DPV peak currents to the various concentrations of DA in real samples. The recovery was in the range of 88 to 130% with a relative standard deviation (RSD, n = 3) of less than 6.3%. From this result, the flexible sensor with the GO/PEDOT:PSS composite exhibited sufficient performance to be used immediately for low-level DA determination applications.

Conclusions
The simple electropolymerized GO/PEDOT:PSS composite on a flexible sensor provides a fast and simple approach for the detection of DA by allowing the efficient fabrication of a highly sensitive and selective electrode interface. In this study, we have shown that the electropolymerization of a mixture of GO and EDOT:PSS provides a facile and effective sensor with a GO/PEDOT:PSS composite for the detection of dopamine that contains the interference species of AA and UA. The PEDOT:PSS presented a well-distributed morphology along the ridges  www.nature.com/scientificreports/ formed by the GO layer. The PEDOT:PSS provided a strong combination for the GO layer as well as providing enhancement of the real activation area and roughness larger than the GO sheet, along with excellent electrochemical characteristics. The flexible sensor with GO/PEDOT:PSS showed a significantly improved capability for the sensitive, selective, and stable determination of dopamine (sensitivity of 69.3 µA/µMcm 2 , and a low detection limit of 0.008 μΜ), compared to the GO sheet, in terms of the DPV peak potential separation and the current peaks. The simultaneous detection of DA in the presence of AA and UA also was achieved with high selectivity, a sensitivity, and a low detection limit. In addition, through the serum sample test, the sensor with the GO/ PEDOT:PSS composite proved to be relatively well matched and feasible for the practical determination of DA. This work is the first to report a simple preparation of a flexible sensor with an electropolymerized GO/ PEDOT:PSS composite, which is expected to show great promise for in-vivo and in-vitro sensing of neurotransmitters, as well as in other integrated wearable and in vivo bioelectronics.
Fabrication of a flexible sensor with a thin Au electrode. Supplementary Fig. S9 shows the conceptual drawings for the process of fabricating the flexible sensor with Au working, counter, the reference electrode, and a photograph of the sensor that was fabricated. Briefly, the first polyimide (PI, thickness of 20 µm) as a substrate layer was spin-coated on a 4-inch SiO 2 /Si wafer. After curing in a convection oven for 10 min at 90 °C, 10 min at 110 °C, and 60 min at 210 °C, a negative photoresist (DNR-L300-30) was spin-coated on top of the PI layer for the lift-off process. After patterning using a mask aligner (MA6, Karl Suss, Garching, Germany), Cr/Au (10/100 nm) were deposited using an e-beam evaporator. After the lift-off process using acetone, the second PI was spin-coated and cured for the insulation layer (with thickness of 3 μm). The positive photoresist (AZ 9260) was coated on the second PI layer to open the electrode site and the connector pad. After patterning, the exposed PI patterns were etched by reactive ion etching (Plasma Therm, St. Petersburg, FL, USA). A laser dicing machine (M-2000, Exitech, Oxford, UK) was used to cut the perimeter of the defined sensor. Then, the flexible sensor was detached easily from the Si wafer.
Preparation of the GO/PEDOT:PSS composite on an Au working electrode. Figure 1a shows that 0.01 M EDOT and 0.1 M PSS were mixed in deionized water. Then, the prepared EDOT:PSS solution and the GO suspension (4 mg/ml in water) were well mixed with various ratios, i.e., 1:1, 2:1, 5:1, and 10:1. The GO/ EDOT:PSS mixture that was prepared was simply and selectively polymerized on the thin Au working electrode by the EPP method (4 μA current for 50, 150, 300, and 600 s). PSS typically is the dopant material that is used for PEDOT because it reinforces the structures by the bonding coulomb interaction 48 . During the polymerization from GO/EDOT:PSS to the GO/PEDOT:PSS composite, the PSS moderates the molecular entanglement into PEDOT, which might induce an improved interaction between GO and the polymerized EDOT (PEDOT). The Pt wire and Ag/AgCl electrode as counter and reference electrodes, respectively, were used for the growth of the GO/PEDOT:PSS composite. After polymerization, the composite was dried for 5 h at room temperature.
Characterization. The electrochemical performances of the sensors were evaluated by an Autolab (PGSTAT 302 N, NOVA software, Ecochemie, Utrecht, The Netherlands) at room temperature. Three electrode configurations were used for CV, EIS, and DPV with an Au reference, Au counter, and Au (2.4 mm in diameter) or modified electrode (GO, PEDOT:PSS, GO/PEDOT:PSS) as a working electrode.
The CV with potential limits of − 0.2 and 0.8 V was performed with a scan rate of 100 mVs −1 , and the frequency range of EIS was from 1 to 10 5 Hz. The parameters of the DPV measurements were set as follows, i.e., the scan rate was 50 mVs −1 , the pulse width was 0.06 s, and the amplitude was 30 mV. All solutions were prepared freshly every day and kept in the dark at 4 ℃ to avoid the oxidation of DA. All of the experiments were conducted at ambient temperature.
The surface morphologies and elemental analyses of the electrode were evaluated respectively by scanning electron microscopy (SEM, Regulus8230), Fourier transform infrared spectroscopy (FT-IR, Thermo fisher (is10) instrument), and X-ray photoelectron spectroscopy (XPS, Ulvac, Japan) with a monochromatic Al Kα X-ray source.