Formation and Electrochemical Evaluation of Polyaniline and Polypyrrole Nanocomposites Based on Glucose Oxidase and Gold Nanostructures

Nanocomposites based on two conducting polymers, polyaniline (PANI) and polypyrrole (Ppy), with embedded glucose oxidase (GOx) and 6 nm size gold nanoparticles (AuNPs(6nm)) or gold-nanoclusters formed from chloroaurate ions (AuCl4−), were synthesized by enzyme-assisted polymerization. Charge (electron) transfer in systems based on PANI/AuNPs(6nm)-GOx, PANI/AuNPs(AuCl4−)-GOx, Ppy/AuNPs(6nm)-GOx and Ppy/AuNPs(AuCl4−)-GOx nanocomposites was investigated. Cyclic voltammetry (CV)-based investigations showed that the reported polymer nanocomposites are able to facilitate electron transfer from enzyme to the graphite rod (GR) electrode. Significantly higher anodic current and well-defined red-ox peaks were observed at a scan rate of 0.10 V s−1. Logarithmic function of anodic current (log Ipa), which was determined by CV-based experiments performed with glucose, was proportional to the logarithmic function of a scan rate (log v) in the range of 0.699–2.48 mV s−1, and it indicates that diffusion-controlled electrochemical processes were limiting the kinetics of the analytical signal. The most efficient nanocomposite structure for the design of the reported glucose biosensor was based on two-day formed Ppy/AuNPs(AuCl4−)-GOx nanocomposites. GR/Ppy/AuNPs(AuCl4−)-GOx was characterized by the linear dependence of the analytical signal on glucose concentration in the range from 0.1 to 0.70 mmol L−1, the sensitivity of 4.31 mA mM cm−2, the limit of detection of 0.10 mmol L−1 and the half-life period of 19 days.


Characteristics of Cyclic Voltammograms Registered Using GR/PANI/AuNPs(6nm)-GOx, GR/PANI/AuNPs(AuCl4 − )-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx Electrodes
Novel nanocomposites based on AuNPs and polymers are able to improve various useful characteristics of biosensors [44]. The main aim of the present study was the investigation of a glucose biosensor based on enzyme-assisted-formed PANI/GOx and Ppy/GOx nanocomposites with incorporated gold nanoparticles. Enzyme-assisted polymerization/oligomerization depends on the concentration of polymerizable monomers in polymerization-bulk solution, the activity of the applied enzyme and the duration of the polymerization procedure [37,39].
Investigations of the shape of CV and the influence of scan rate on a shift of red-ox peaks provide some important information on the mechanism of electrochemical reaction and the rate of electron transfer from electrochemically active species [39]. The performance of the bare GR electrode and the same electrode, which was modified by PANI/AuNPs(6nm)-GOx, PANI/AuNPs(AuCl4 − )-GOx, Ppy/AuNPs(6nm)-GOx and Ppy/AuNPs(AuCl4 − )-GOx nanocomposites, was investigated by CV in 0.05 mol L −1 SA buffer, pH 6.0, in the presence of a soluble red-ox mediator-PMS. The reduction of the pyrrole monomer and the oxidation of chloroaurate ions proceeded simultaneously with the formation of Ppy and Au 0 [43]. The influence of scan rate has been evaluated and is presented in Figure 2 and Table 1. No red-ox peaks were observed on the surface of the bare GR electrode when potential was swept in the range from −0.70 to 0.80 V. Potential diapason below 0.80 V was selected due the oxidation of gold at 1.10 V [5]. As is seen from Figure 2, well-defined anodic peaks were obtained on GR electrodes modified by PANI/AuNPs(6nm)-GOx, Ppy/AuNPs(6nm)-GOx and Ppy/AuNPs(AuCl4 − )-GOx nanocomposites, instead of PANI/AuNPs(AuCl4 − )-GOx-based nanocomposites. After the deposition of the nanocomposite's layer on the working electrode, the form of registered  The human serum sample was diluted (1:10) in 0.05 mol L −1 SA, pH 6.0, and centrifuged at 14.6 × 10 3 g. All investigations were performed using the GR electrode modified by Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites. The electrochemical measurements were done in 10 times diluted human serum with 10 mmol L −1 of glucose before and after the addition of 1 mmol L −1 fructose, mannose, galactose, xylose, or saccharose. To evaluate the influence of ascorbic and uric acids on the developed biosensor, CV measurements were applied in 10 times diluted human serum with 10 mmol L −1 of glucose, 10 mmol L −1 of glucose and 0.01, 0.05 or 0.1 mmol L −1 of ascorbic acid (AA), and 10 mmol L −1 of glucose and 0.01 or 0.05 mmol L −1 of uric acid (UA). Novel nanocomposites based on AuNPs and polymers are able to improve various useful characteristics of biosensors [44]. The main aim of the present study was the investigation of a glucose biosensor based on enzyme-assisted-formed PANI/GOx and Ppy/GOx nanocomposites with incorporated gold nanoparticles. Enzyme-assisted polymerization/oligomerization depends on the concentration of polymerizable monomers in polymerization-bulk solution, the activity of the applied enzyme and the duration of the polymerization procedure [37,39].

Characteristics of Cyclic Voltammograms Registered Using GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4
Investigations of the shape of CV and the influence of scan rate on a shift of red-ox peaks provide some important information on the mechanism of electrochemical reaction and the rate of electron transfer from electrochemically active species [39]. The performance of the bare GR electrode and the same electrode, which was modified by PANI/AuNPs (6nm) -GOx, PANI/AuNPs (AuCl4 Ppy/AuNPs (6nm) -GOx and Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites, was investigated by CV in 0.05 mol L −1 SA buffer, pH 6.0, in the presence of a soluble red-ox mediator-PMS. The reduction of the pyrrole monomer and the oxidation of chloroaurate ions proceeded simultaneously with the formation of Ppy and Au 0 [43]. The influence of scan rate has been evaluated and is presented in Figure 2 and Table 1. No red-ox peaks were observed on the surface of the bare GR electrode when potential was swept in the range from −0.70 to 0.80 V. Potential diapason below 0.80 V was selected due the oxidation of gold at 1.10 V [5]. As is seen from Figure 2, well-defined anodic peaks were obtained on GR electrodes modified by PANI/AuNPs (6nm) -GOx, Ppy/AuNPs (6nm) -GOx and Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites, instead of PANI/AuNPs (AuCl4 − ) -GOx-based nanocomposites. After the deposition of the nanocomposite's layer on the working electrode, the form of registered cyclic voltammograms changed and became much wider than that before the modification. This effect illustrates a significant increase of the electrical capacitance of the modified electrode, and this fact is in line with previously published research on the evaluation of Ppy layers of different thickness and/or morphology [46].
Polymers 2020, 12, x FOR PEER REVIEW 6 of 20 cyclic voltammograms changed and became much wider than that before the modification. This effect illustrates a significant increase of the electrical capacitance of the modified electrode, and this fact is in line with previously published research on the evaluation of Ppy layers of different thickness and/or morphology [46]. It was determined that the diffusion barrier of GR electrodes, additionally modified by PANI or Ppy using enzyme-assisted polymerization, increased. It was observed that for GR/PANI/AuNPs(6nm)-GOx, GR/PANI/AuNPs(AuCl4 − )-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes, Epa has been shifted to more positive values, from −0.002 to 0.23 V, from 0.096 to 0.26 V, from −0.009 to 0.24 V and from −0.007 to 0.21 V respectively, when the scan rate increased from 0.005 to 0.30 V s −1 . These red-ox peaks are attributed to the oxidation of glucose to gluconolactone, as it was reported previously [17], and indicates that the GR electrode modified by PNC with embedded GOx in the presence of AuNPs shows good electrocatalytic activity for electrochemical oxidation of glucose. The electrochemical process is quasi-reversible and one cathodic peak has appeared on the GR electrode modified by PNC embedded with GOx and AuNPs on the reverse scan. It is seen that for   Figure 2.

The Kind of Polymer Nanocomposites on GR
It was determined that the diffusion barrier of GR electrodes, additionally modified by PANI or Ppy using enzyme-assisted polymerization, increased. It was observed that for GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes, E pa has been shifted to more positive values, from −0.002 to 0.23 V, from 0.096 to 0.26 V, from −0.009 to 0.24 V and from −0.007 to 0.21 V respectively, when the scan rate increased from 0.005 to 0.30 V s −1 . These red-ox peaks are attributed to the oxidation of glucose to gluconolactone, as it was reported previously [17], and indicates that the GR electrode modified by PNC with embedded GOx in the presence of AuNPs shows good electrocatalytic activity for electrochemical oxidation of glucose.
The electrochemical process is quasi-reversible and one cathodic peak has appeared on the GR electrode modified by PNC embedded with GOx and AuNPs on the reverse scan. It is seen that for GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4  (Table 1). A diffusion-controlled, quasi-reversible electrochemical process was observed and the average value of the anodic and cathodic potential was calculated according to the methodology presented in Reference [10]. The formal potential (E 0 ) at 0.30 V s −1 for GR/PANI/AuNPs (6nm) -GOx was −0.13 V, for GR/PANI/AuNPs (AuCl4 − ) -GOx was −0.11 V, for GR/Ppy/AuNPs (6nm) -GOx was −0.10 V and for GR/Ppy/AuNPs (AuCl4 − ) -GOx was −0.11 V. It indicates that the GR electrode modified by PNC nanocomposites containing GOx and AuNPs was characterized by a rather high degree of the reversibility and sufficient charge transfer rate. It was determined that the scan rate in the range from 0.10 to 0.30 V s −1 was the most optimal for accurate determination of red-ox peaks in cyclic voltammograms.

) -GOx Electrodes
Scan rate influences the electrochemical behavior of electrodes: a rather thick diffusion layer is formed when a slow scan rate is applied, and on the contrary, a thinner layer is formed when a fast scan rate is applied [47]. Anodic current (I pa ) is proportional to the applied scan rate and the relationship between log I pa and log v is linear when the "classic" electrochemical system is under investigation [39,47,48]. The influence of scan rate on the anodic current obtained on the surface of GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes in the absence and in the presence of glucose is presented in Figures 3A and 4A, respectively. As is seen from the presented results, the value of anodic current for investigated systems has increased with the increasing scan rate.  (Figures 3A and 4A). The value of the anodic current for GR/PANI/AuNPs(6nm)-GOx and GR/PANI/AuNPs(AuCl4 − )-GOx electrodes ( Figure 3A) at the 0.30 V s −1 scan rate in the presence of glucose was 1.27 and 1.18 times higher in comparison with the anodic current registered for the same electrodes in the absence of glucose. Anodic currents for GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes ( Figure 4A) at the 0.30 V s −1 scan rate in the presence of glucose was 1.79 and 2.82 times higher than for the same electrodes without glucose. When the voltage sweep is faster, the red-ox reactions on the electrode do not have enough time to undergo completely [39]. Therefore, the scan rate of 0.10 V s −1 has been determined as the most optimal to achieve well-defined peaks in cyclic voltammograms and sufficiently high differences in red-ox peaks of cyclic voltammograms in the presence and absence of glucose.  The relationship between log Ipa and log v evaluated for GR/PANI/AuNPs(6nm)-GOx, GR/PANI/AuNPs(AuCl4 − )-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes in the presence of glucose is presented in Figures 3B and 4B, respectively. It is seen that in all cases, the logarithmic function of the anodic current was proportional to log v in the range of 0.699-2.48 mV s −1 , which is in the agreement with data published in another paper [48]. The correlation coefficient of lines registered by GR/PANI/AuNPs(6nm)-GOx, GR/PANI/AuNPs(AuCl4 − )-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes ( Figure 4B Figure 4A) at the 0.30 V s −1 scan rate in the presence of glucose was 1.79 and 2.82 times higher than for the same electrodes without glucose. When the voltage sweep is faster, the red-ox reactions on the electrode do not have enough time to undergo completely [39]. Therefore, the scan rate of 0.10 V s −1 has been determined as the most optimal to achieve well-defined peaks in cyclic voltammograms and sufficiently high differences in red-ox peaks of cyclic voltammograms in the presence and absence of glucose.
The relationship between log I pa and log v evaluated for GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes in the presence of glucose is presented in Figures 3B and 4B, respectively. It is seen that in all cases, the logarithmic function of the anodic current was proportional to log v in the range of 0.699-2.48 mV s −1 , which is in the agreement with data published in another paper [48]. The correlation coefficient of lines registered by GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 ) -GOx electrodes is diffusion-controlled and, therefore, such electrode can be applied for electrochemical biosensing of glucose in a rather broad concentration range. Mathematically, the slope of line (a) is determined by the equation of linear dependence (y = ax + b, where a = ∆y/∆x = (y 2 − y 1 )/(x 2 − x 1 ) = tan(a)). As is seen from Figure 3B, the slope of the line determined for the GR/PANI/AuNPs (6nm) -GOx electrode (tan(a) = 0.754) is 1.25 times steeper than that calculated for the GR/PANI/AuNPs (AuCl4 − ) -GOx electrode (tan(a) = 0.601). The same difference of the slope was determined for GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes. As it is seen from Figure 4B, the slope of the line characterized for the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode (tan(a) = 0.781) is 1.12 times steeper in comparison with the results calculated for the GR/Ppy/AuNPs (6nm) -GOx electrode (tan(a) = 0.700). The steeper slope presents the increase of electron rate on the surface of GR/PANI/AuNPs (6nm) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes. It is in agreement with a statement that AuNPs could enhance the sensitivity and stability of a glucose biosensor by catalyzing oxidation of H 2 O 2 , which is formed during enzymatic reaction, and effectively facilitate the electron transfer through a nanocomposite matrix due to good conductivity [2,4]. During enzyme-assisted polymerization, low molecular weight polymers remain in aqueous solutions while large molecular weight polymers are adsorbed on the electrode surface [49]. The efficiency of enzyme-assisted polymerization, the size of synthesized PNC and the conductivity strongly depend on the duration of polymerization [37]. Some researchers are reporting that with the increase of size from 39 to 1080 nm of Ppy nanotubes, which were decorated with AuNPs, the conductivity decreased from 75 to 0.24 S cm −1 [40]. However, the formation of larger PNC structures decreases the sensitivity of developed biosensors, due to the lower conductivity of Ppy nanotubes, which were decorated with gold particles [40]. In our previous research, we have reported that the optimal duration of enzyme-assisted polymerization of PNC was shorter than 4.5 days [44]. It was determined that by the increase of enzymatic synthesis duration from 1 to 4.5 days, the hydrodynamic diameter of PANI/AuNPs (6nm) -GOx, PANI/AuNPs (AuCl4 − ) -GOx, Ppy/AuNPs (6nm) -GOx and Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites increased from 679 to 1128, from 570 to 659, from 512 to 594 and from 366 to 388 nm, respectively [50]. We predicted that the sensitivity of glucose biosensors developed in this paper based on GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes will be decreased when too-long polymerization duration is applied. Gold nanoparticles promote the electron transfer between glucose oxidase and the working electrode [27,51]. To evaluate the influence of enzyme-assisted polymerization on CV peak currents registered using GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes, the enzyme-assisted polymerization was performed for periods between 2 and 4 days in the dark at room temperature. To evaluate the anodic current of CV correctly, the difference of anodic peaks in the absence and the presence of 48 mmol L −1 glucose was calculated, and the obtained results are presented in Figure 5. During enzyme-assisted polymerization, low molecular weight polymers remain in aqueous solutions while large molecular weight polymers are adsorbed on the electrode surface [49]. The efficiency of enzyme-assisted polymerization, the size of synthesized PNC and the conductivity strongly depend on the duration of polymerization [37]. Some researchers are reporting that with the increase of size from 39 to 1080 nm of Ppy nanotubes, which were decorated with AuNPs, the conductivity decreased from 75 to 0.24 S cm −1 [40]. However, the formation of larger PNC structures decreases the sensitivity of developed biosensors, due to the lower conductivity of Ppy nanotubes, which were decorated with gold particles [40]. In our previous research, we have reported that the optimal duration of enzyme-assisted polymerization of PNC was shorter than 4.5 days [44]. It was determined that by the increase of enzymatic synthesis duration from 1 to 4.5 days, the hydrodynamic diameter of PANI/AuNPs(6nm)-GOx, PANI/AuNPs(AuCl4 − )-GOx, Ppy/AuNPs(6nm)-GOx and Ppy/AuNPs(AuCl4 − )-GOx nanocomposites increased from 679 to 1128, from 570 to 659, from 512 to 594 and from 366 to 388 nm, respectively [50]. We predicted that the sensitivity of glucose biosensors developed in this paper based on GR/PANI/AuNPs(6nm)-GOx, GR/PANI/AuNPs(AuCl4 − )-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes will be decreased when too-long polymerization duration is applied. Gold nanoparticles promote the electron transfer between glucose oxidase and the working electrode [27,51]. To evaluate the influence of enzyme-assisted polymerization on CV peak currents registered using GR/PANI/AuNPs(6nm)-GOx, GR/PANI/AuNPs(AuCl4 − )-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes, the enzyme-assisted polymerization was performed for periods between 2 and 4 days in the dark at room temperature. To evaluate the anodic current of CV correctly, the difference of anodic peaks in the absence and the presence of 48 mmol L −1 glucose was calculated, and the obtained results are presented in Figure 5. As it is seen from Figure 5   As it is seen from Figure 5, the value of the anodic current for GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 glucose biosensor can be designed when enzyme-assisted polymerization lasting 2 days is applied for the formation of a nanocomposite-based layer. The sensitivity of GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes calculated from data presented in Figure 5 was 0.0361, 0.0071, 0.0684 and 0.0772 mA mM cm −2 , respectively. It was determined that after 2 days of enzyme-assisted polymerization, the sensitivity of the GR/PANI/AuNPs (6nm) -GOx electrode was 5.08 times higher, if compared with the GR/PANI/AuNPs (AuCl4 − ) -GOx electrode. Meanwhile, the sensitivity of the GR/Ppy/AuNPs (6nm) -GOx electrode was 1.13 times lower than that obtained on the GR/Ppy/AuNPs (AuCl4 To evaluate the role of AuNPs on the sensitivity in glucose sensing, the results obtained in the presence and absence of AuNPs in the structure of PANI and Ppy nanocomposites after 2 days of polymerization were compared. As can be seen from the results presented in Figure 5, the value of the anodic current for GR/PANI/GOx and GR/Ppy/GOx electrodes was 0.090 and 0.107 mA, respectively. The sensitivity of the GR/PANI/GOx (0.0264 mA mM cm −2 ) electrode was 1.37 times lower if compared with the GR/PANI/AuNPs (6nm) -GOx electrode. Meanwhile, the sensitivity of the GR/PANI/GOx electrode was 3.72 times higher in a comparison with the GR/PANI/AuNPs (AuCl4 − ) -GOx electrode, which were characterized as presented above. The sensitivity of the GR/Ppy/GOx (0.0314 mA mM cm −2 ) electrode was 2.18 and 2.46 times lower than that of GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes, respectively. It indicates the advantage of polymeric nanocomposites with AuNPs. In any case, 2 days of enzyme-assisted polymerization is the most suitable duration for the development of GR/PANI/AuNPs (6nm) -GOx, GR/PANI/AuNPs (AuCl4 − ) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes. The higher value of the anodic current calculated from differences of anodic peaks of CV in the presence and absence of glucose for GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes is based on the conducting nature and increased 'active surface area' of nanocomposites, which offers more freedom regarding the orientation of entrapped GOx [10].

The Evaluation of the Stability and Analytical Characteristics of GR Electrodes Modified by PNC
The next stage of investigations was the evaluation of the stability of glucose biosensors based on GR electrodes modified by the developed PNC. For this purpose, PANI/AuNPs (6nm) -GOx, Ppy/AuNPs (6nm) -GOx and Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites, which were formed by 2 days of enzyme-assisted polymerization, were evaluated. The application of PANI/AuNPs (AuCl4 − ) -GOx nanocomposites was refused due to the irregular form of cyclic voltammograms. The stability of the developed glucose biosensors was investigated within 6 or 22 days by the registration of CVs. GR electrodes modified by PANI/AuNPs (6nm) -GOx, Ppy/AuNPs (6nm) -GOx and Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites (as described in Section 2.3) were hanged over 0.05 mol L −1 SA buffer, pH 6.0, solution at 4 • C between measurements. Registered anodic currents of prepared GR/PANI/AuNPs (6nm) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes were equated to 100% and are presented in Figure 8. As it is seen from Figure 7A,B (line 1), any red-ox peaks were observed on the surface of the bare GR electrode. However, in the case of the GR/PANI/AuNPs(6nm)-GOx electrode ( Figure 7A, line 2), two oxidation peaks were registered at the 0.332 and 0.601 V values of potential, while for the GR/PANI/AuNPs(6nm)-GOx electrode ( Figure 7A, line 3), only one oxidation peak at 0.349 V was observed. The conversion of PANI leucoemeraldine form to emeraldine salt showed the oxidation peaks at 0.332 and 0.349 V and it is in agreement with the results of other authors [53,54]. A 0.601 V potential peak in the case of the GR/PANI/AuNPs(6nm)-GOx electrode could be associated with the oxidation of emeraldine into pernigraniline form [53]. The CV obtained using GR/Ppy/AuNPs(6nm)-GOx ( Figure 7B, line 4) and GR/Ppy/AuNPs(AuCl4 − )-GOx ( Figure 7B, line 5) electrodes showed the peaks at 0.425 and 0.545 V respectively, and may be associated with cation polaron's transition into cation bipolaron's state of polypyrrole [54].

The Evaluation of the Stability and Analytical Characteristics of GR Electrodes Modified by PNC
The next stage of investigations was the evaluation of the stability of glucose biosensors based on GR electrodes modified by the developed PNC. For this purpose, PANI/AuNPs(6nm)-GOx, Ppy/AuNPs(6nm)-GOx and Ppy/AuNPs(AuCl4 − )-GOx nanocomposites, which were formed by 2 days of enzyme-assisted polymerization, were evaluated. The application of PANI/AuNPs(AuCl4 − )-GOx nanocomposites was refused due to the irregular form of cyclic voltammograms. The stability of the developed glucose biosensors was investigated within 6 or 22 days by the registration of CVs. GR electrodes modified by PANI/AuNPs(6nm)-GOx, Ppy/AuNPs(6nm)-GOx and Ppy/AuNPs(AuCl4 − )-GOx nanocomposites (as described in Section 2.3) were hanged over 0.05 mol L −1 SA buffer, pH 6.0, solution at 4 °C between measurements. Registered anodic currents of prepared GR/PANI/AuNPs(6nm)-GOx, GR/Ppy/AuNPs(6nm)-GOx and GR/Ppy/AuNPs(AuCl4 − )-GOx electrodes were equated to 100% and are presented in Figure 8. The analytical response of GR/PANI/AuNPs (6nm) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes towards glucose after storage for 6 days decreased down to 16.3%, 10.1% and 78.1%, respectively, of their initial value. It means that the value of the anodic current obtained on GR/PANI/AuNPs (6nm) -GOx, GR/Ppy/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (AuCl4 − ) -GOx electrodes after 6 days was 6.13, 9.90 and 1.28 times lower than that immediately after the design of electrodes. It is seen that electrochemical biosensors based on the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode were 4.79 and 7.73 times more stable than those based on GR/PANI/AuNPs (6nm) -GOx and GR/Ppy/AuNPs (6nm) -GOx electrodes. The further investigation was performed with 2 days of enzyme-assisted synthesized Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites due to their high stability and the simplicity of surface modification of the GR electrode. The τ 1/2 (50% of initial response) for the developed GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode was 19 days. Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites formed on the surface of the GR electrode improved the efficiency of charge transfer from immobilized GOx and maintained their bio-catalytic activity. During electrochemical measurements, electrons are transferred toward the positively charged surface of the GR electrode and current is registered. In the presence of glucose and dissolved oxygen in the solution of SA buffer, immobilized on the GR electrode, GOx contained in polymeric nanocomposites generates hydrogen peroxide and gluconolactone, which is hydrolyzed to gluconic acid ( Figure 1). PMS is able to re-oxidase the red-ox active center of GOx and the electrons are transferred via a reduced form of the mediator PMSH 2 , in two ways: (i) directly to the GR electrode and (ii) through AuNPs [26]. AuNPs are able to facilitate an electron transfer between protein, the red-ox mediator and the GR electrode, and to improve the sensitivity of the developed analytical system [26,51]. To evaluate the analytical characteristics of the developed electrochemical biosensor based on the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode, anodic current responses were measured by CV at the glucose concentration range from 0.10 to 48.4 mmol L −1 . The anodic current of the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode increased due to the GOx-catalyzed reaction and the increased concentration of oxidizable products ( Figure 9A). As it is seen from the presented CVs ( Figure 9A), the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode is also involved in the electrochemical reduction of O 2 and H 2 O 2 and it is in line with findings presented by some other researchers, which evaluated graphene/nano-Au/GOx systems immobilized on the GC electrode [19]. The anodic peak registered for the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode appeared at 0.138 V, while the cathodic peak appeared at −0.343 V.
The GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode can be used for glucose determination due to the increase of the anodic current of CVs ( Figure 9A). The amount of glucose being oxidized by GOx immobilized on the electrode is proportional to the concentration of glucose present in sample solution [18,19]. Hyperbolic dependence, which is presented in Figure 9B, is in agreement with Michaelis-Menten kinetics. The maximal current generated during the enzyme-catalyzed reaction (I max ) and the apparent Michaelis constant (K M(apparent) ) were correspondingly a and b parameters of hyperbolic function y = ax/(b + x), which has been used for the approximation of results. The reported glucose biosensor based on the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode was characterized by ∆I max = 0.292 mA and K M(apparent) = 0.348 mmol L −1 . A low value of K M(apparent) indicates high affinity of glucose towards immobilized glucose oxidase [27].
the GR/Ppy/AuNPs(AuCl4 − )-GOx electrode is also involved in the electrochemical reduction of O2 and H2O2 and it is in line with findings presented by some other researchers, which evaluated graphene/nano-Au/GOx systems immobilized on the GC electrode [19]. The anodic peak registered for the GR/Ppy/AuNPs(AuCl4 − )-GOx electrode appeared at 0.138 V, while the cathodic peak appeared at −0.343 V. -GOx electrode at a low concentration of glucose was higher than that at a high concentration. Lower sensitivity (0.0772 mA mM cm −2 ) at high concentrations (48.4 mmol L −1 ) of glucose than that at low concentrations could be related to two factors: (i) the slower diffusion of glucose, because at higher glucose concentrations the viscosity of glucose solution increases, and (ii) adsorption/desorption dynamics of glucose at high concentrations [19]. The evaluated sensitivity was higher than that determined by other researchers on the surfaces of the GC electrode modified by GOx/P-L-Arg/f-MWCNTs composites (48.86 µA mM cm −2 ) [10] and of the GC electrode modified by GOx-poly(L-lysine)/reduced graphene oxide-ZrO 2 composites (11.65 µA mM cm −2 ) [15]. The relative standard deviation of the analytical signal at glucose concentration of 0.70 mmol L −1 on the GR/Ppy/AuNPs (AuCl4 − ) -GOx electrode was 13.9%. The value of the LOD for the developed glucose biosensor was determined as 0.10 mmol L −1 . The calculated LOD is in the same range as that determined by some other researchers for the GC electrode modified by GOx/P-L -Arg/f-MWCNTs composites (0.10 mmol L −1 ) [10], and lower than that calculated for the GC electrode modified by GOx-poly(L-lysine)/reduced graphene oxide-ZrO 2 composites (0.13 mmol L −1 ) [15]. According to the statement by Vilian et al. [15], the increase of the electron transfer kinetics for glucose biosensing and electrocatalytic activity can be determined by high surface area of gold nanostructures embedded within polymers.
The high sensitivity and the low limit of detection are significant advantages of the GR electrode modified by Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites, even if the linear range of the developed glucose biosensor was lower than the physiological level of glucose in biological samples, in the range from 3 to 8 mmol L −1 [55,56] (in the blood serum for healthy persons: 3.9-5.6 mmol L −1 ).

) -GOx Nanocomposites for Glucose Determination in the Sample of Human Serum
The addition of the interfering species, such as 1.0 mmol L −1 of fructose, mannose, xylose or saccharose, in the solution of 10.0 mmol L −1 glucose do not show the influence on the registered current of glucose ( Figure 10A), while the addition of 1.0 mmol L −1 galactose in the solution of 10.0 mmol L −1 glucose decreased the registered current of glucose by 5.81% in comparison with results obtained for the solution, which contained only glucose without galactose. The influence of ascorbic and uric acids on the determination of glucose is presented in Figure 10B. It is seen that AA and UA have negligible effects on the registered current: after the addition of 10   The suitability of the developed biosensor based on the GR electrode modified by Ppy/AuNPs(AuCl4 − )-GOx nanocomposites for glucose determination in the samples of human serum was performed. For this purpose, human serum 10 times diluted with 0.05 mol L −1 SA buffer solution, pH 6.0, with known initial glucose concentrations, was investigated. Then, various amounts of glucose were added in order to imitate samples with increasing glucose concentrations in the linear detection range of the analyte. Each sample of human blood serum was assessed at least three or four times, and the achieved results were expressed by the average values and are presented in Table 2. It was found that recoveries of developed glucose biosensor in the sample of human serum were in the range of 93.6-94.8%. The suitability of the developed biosensor based on the GR electrode modified by Ppy/AuNPs (AuCl4 − ) -GOx nanocomposites for glucose determination in the samples of human serum was performed. For this purpose, human serum 10 times diluted with 0.05 mol L −1 SA buffer solution, pH 6.0, with known initial glucose concentrations, was investigated. Then, various amounts of glucose were added in order to imitate samples with increasing glucose concentrations in the linear detection range of the analyte. Each sample of human blood serum was assessed at least three or four times, and the achieved results were expressed by the average values and are presented in Table 2. It was found that recoveries of developed glucose biosensor in the sample of human serum were in the range of 93.6-94.8%. The response of CV was registered in human blood serum 10 times diluted with 0.05 mol L −1 SA buffer, pH 6.0, sample containing 6 mmol L −1 PMS.
The advantages of the developed biosensor based on the graphite rod electrode modified by Ppy/AuNPs (AuCl4 − ) -GO x nanocomposites are: (i) the high sensitivity (4.31 mA mM cm −2 ) and rather good stability (19 days), (ii) the low limit of detection (0.10 mmol L −1 ), (iii) the applicability in multiple analyses, (iv) the low price of a single analysis, (v) the short duration of a single measurement (20 s) and (vi) the high resistance to interfering materials, which allows to apply modified GR electrodes for glucose biosensing in clinical practice. The development of a glucose biosensor without a red-ox mediator is expected to be the next step of our investigations in this research area.

Conclusions
PANI and Ppy nanocomposites with GOx and AuNPs were formed by enzyme-assisted synthesis of PANI and Ppy and were deposited on the surface of GR electrodes. The relationship between log I pa and log v registered from CV anodic peaks in the presence of glucose was linear and the electrochemical reaction was controlled by the diffusion. The glucose biosensor based on GR modified by Ppy/AuNPs (AuCl4