Full Paper 1 DOI: 10.1002/elan.201800483 2 3 4 Comparison of Insulin Determination on NiNPs/chitosan- 5 6 MWCNTs and NiONPs/chitosan-MWCNTs Modified 7 8 Pencil Graphite Electrode 9 10 Ivana Šišoláková,[a] Jana Hovancová,[a] Renáta Oriňaková,*[a] Andrej Oriňak,[a] Daniel Rueda Garcia,[b] 11 Oleg Shylenko,[c] and Jozef Radoňák[d] 12 13 14 Abstract: The rising amount of patients suffering for redox reaction mechanism of insulin and detection 15 diabetes mellitus increases the requirements for effective characteristics. The electrode modified by Ni nanopar- 16 insulin sensors. Carbon materials are a suitable choice for ticles displays linear range of 1 mM–5 mM (R2 0.80), limit 17 the development of insulin sensors due to their electro- of detection (LOD) of 4.34 mM and sensitivity of 0.12 mA/ 18 chemical characteristics. Pencil graphite electrodes (PGE) mM. On the other hand, the electrode modified by NiO 19 represent the trade-off between price and excellent nanoparticles displays enhanced electrochemical charac- 20 conductive properties. The modification of PGE by NiO teristics such as linear range of 0.05 mM–5 mM (R2 0.99), 21 and Ni nanoparticles fixed by chitosan results in surface limit of detection of 260 nM and sensitivity of 0.64 mA/ 22 area enlargement and improved electrocatalytic proper- mM. These properties make the NiO nanoparticles 23 ties. This paper is focused on the comparison of different modified PGE the appropriate candidate for insulin 24 properties of Ni and NiO nanoparticles and their effect on determination. 25 Keywords: insulin · electrochemical determination · pencil graphite electrodes · nickel nanoparticles · nickel oxide nanoparticles 26 27 28 1 Introduction 29 comings of the other methods mentioned above [18]. The 30 Insulin is a polypeptide hormone that consists of 51 amino main advantages of electrochemical methods are low limit 31 acids divided into A and B chain linked by two disulphide of detection, wide linear range, low cost instruments and 32 bridges. A and B chains are composed of 21 and 30 high sensitivity [19]. 33 residues, respectively. It is produced by b-cells located in The most commonly used electrode material for 34 the pancreatic arrays called islets of Langerhans [1] and insulin determination is carbon [20]. Different types of 35 the normal fasting insulin level in blood reaches maximum carbon electrodes such as glassy carbon electrodes [21], 36 value of 25 mIU/l (0.86 mM) [2]. Insulin ensures transport screen printed carbon electrodes [22], pencil graphite 37 of blood glucose to the cells and controls glucose level in electrodes [23] and carbon paste electrodes [24] have been 38 blood [3, 4] The dysfunction of insulin hormone produc- used for insulin determination. In general, PGE consists 39 tion causes the very common disease in the world called of graphite and clay. PGEs have several advantages in 40 diabetes mellitus [5, 6]. comparison to other carbon-based electrodes, including, 41 Therefore, it is necessary to focus on development of 42 cheap, fast and exact insulin sensor [7]. There is a variety 43 of analytical methods used for insulin determination [a] I. Šišolkov, J. Hovancov, R. Oriňakov, A. Oriňak 44 which can be classified into two main groups, immune and Department of Physical Chemistry 45 non-immune [8]. The most frequently used immune University of P.J. Šafrik in Košice 46 methods are radioimmunoassay (RIA) [9], enzyme immu- Moyzesova 11, 040 01 Košice, Slovakia 47 noassay (EIA) [10] and luminescent immunoassay (LIA) Phone number: + 421-55-2342324 Fax: + 421-55-6222124 48 [11]. The main disadvantages of these methods are long- E-mail: renata.orinakova@upjs.sk 49 lasting analysis and low sensitivity [12], which can be [b] D. Rueda Garcia 50 increased by derivatization of insulin with isotopes and Catalan Institute of Nanoscience and Nanotechnology (ICN2) 51 florigenic labels [13]. Non-immune methods such as high Campus UAB, E-08193 Bellaterra, Barcelona, Spain 52 performance liquid chromatography (HPLC) [14, 15] and [c] O. Shylenko 53 capillary electrophoresis (CE) [16] require expensive Institute of Physics 54 instrumentation, long lasting analysis and complex pre- University of P.J. Šafrik in Košice Park Angelinum 9, 040 01 Košice, Slovakia 55 treatment steps [17]. Among the various methods that [d] J. Radoňk 56 have been used for insulin detection, electrochemical Faculty of medicine 57 assays are still considered the best methods for insulin University of P.J. Šafrik in Košice 58 determination, because they could overcome the short- Trieda SNP 1, 040 01 Košice, Slovakia www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 103 Full Paper 1 very low cost, simple modification, widespread availability (TEM), scanning transmission electron microscopy 2 [25] and well defined surface area [11]. Despite carbon (STEM) and EDX analysis. Both of prepared electrodes 3 being considered the most suitable material for insulin showed good electrocatalytic activity towards insulin 4 determination, the bare unmodified carbon electrodes oxidation. The NiONPs/chitosan-MWCNTs/PGE had bet- 5 have some drawbacks which can be eliminated by ter electrochemical characteristics and wider linear range 6 modification with different nanostructures [26]. Therefore, when compared to electrochemical characteristics and 7 various materials for electrode modification such as multi- linear range of NiNPs/chitosan-MWCNTs/PGE, what 8 walled carbon nanotubes (MWCNTs), metal nanopar- makes it a suitable candidate for electrochemical insulin 9 ticles, metal oxide nanoparticles or combination of these sensor. 10 materials have been studied. MWCNTs are frequently 11 used in electrochemical determination of insulin for 12 electrode modifications due to their mechanical stability 2 Experimental 13 and fast electron transport. MWCNTs are also known as 2.1 Chemicals and Materials 14 an excellent option to enlarge electrode surface area, 15 create more active sites for insulin oxidation, increase The multiwalled carbon nanotubes (diameter 3 nm–10 nm 16 voltammetric current and decrease overpotential [27]. It is and 1 mm–10 mm length) were purchased from BOC 17 also well known that large surface area of MWCNTs SCIENCES Creative Dynamics Inc. (USA, NY). Insulin 18 combined with metal nanoparticles or metal oxide nano- Human Recombinant (5807.6 g/mol) was purchased from 19 particles can improve the performance of the final MP Biomedicals (France, Illkrich). Phosphate buffered 20 electrode material [28, 29]. saline (PBS D8662, sterile filtered), sodium hydroxide 21 For this reason, different combinations of MWCNTs (NaOH, 99 %), nickel nitrate hexahydrate 22 and different types of metal nanoparticles like silver (Ni(NO)3 · 6H2O, 99.9 %), potassium hexacyanoferrate 23 nanoparticles [30], silica nanoparticles [8, 12, 31], ruthe- trihydrate (K4[Fe(CN)6] · 3H2O, 99.95 %), nitric acid 24 nium oxide nanoparticles [32], nickel nanoparticles (HNO3, 65 %) and sulfuric acid (H2SO4, 96 %) were 25 (NiNPs) [33] and nickel oxide nanoparticles (NiONPs) purchased from Sigma Aldrich (USA, Missouri). Insulin 26 have been studied [29]. Nickel based nanomaterials have solutions were freshly prepared by dissolving powdered 27 become an intensively studied material not only because insulin in 0.1 M NaOH in PBS before every electro- 28 of their low cost in comparison with Ag, Co, Pd or Pt chemical measurement. All measurements were per- 29 nanoparticles but also because they possess some desir- formed at room temperature and atmospheric pressure. 30 able characteristics such as high electrical conductivity 31 and good electrocatalytic activity [34]. It is known that the 2.2 Instruments 32 chemical and physical properties of the NiONPs and 33 NiNPs are quite different from bulk Ni and NiO crystals. All electrochemical experiments were performed by using 34 Particularly, large surface area and excellent magnetic AUTOLAB type PGSTAT302N (Metrohm, Switzerland) 35 properties are the greatest advantages of synthetized with a three-electrode setup using the bare or modified 36 nanoparticles [27]. In addition, it has been revealed that PGE (Herlitz, Germany) as the working electrode, Ag/ 37 many nickel-based materials exhibit excellent electro- AgCl (saturated KCl) as the reference electrode and 38 catalytic activity towards the oxidation of a wide range of platinum electrode as the counter electrode. The structure 39 small compounds, like insulin or glucose in alkaline media and surface morphology of the electrodes were charac- 40 [35]. The utilisation of polymer membrane such as Nafion, terized by SEM (Jeol JSM 7000N, Japan) with EDX 41 chitosan, polyethylene glycol prevents the fast occupation analysis, TEM and STEM. 42 of active sites at higher insulin concentration and 43 occupation of active sites by chloride ions in phosphate 2.3 Preparation of the Modified Electrode 44 buffered saline (under physiological condition) [13]. These 45 polymers are also used because of their ability to fix The surface of carbon PGE was polished with soft 46 nanoparticles on the electrode surface during electro- sandpaper (400 grit silicon carbide paper) and ultrasoni- 47 chemical measurements [36]. cally cleaned in ethanol for 1 h. 0.5 g of MWCNTs were 48 In this paper, we have studied the electrocatalytic suspended in 80 ml of solution containing nitric and 49 activity of the NiONPs/chitosan-MWCNTs/PGE and sulfuric acid (3 : 1). 10 ml of ethanol was slowly dropped 50 NiNPs/chitosan-MWCNTs/PGE prepared by pulsed elec- into the stirred suspension. The mixture was then refluxed 51 trochemical deposition of NiONPs or NiNPs on chitosan- for 120 min. The MWCNTs were filtered off and dried at 52 MWCNTs modified PGE surface towards insulin oxida- 70 8C for 5 h [37]. 10 mg of activated MWCNTs were 53 tion. The electrochemical properties of modified PGEs dispersed in 5 ml of chitosan and 5 ml of deionized water 54 were studied by cyclic voltammetry, chronoamperometry and ultrasonicated for 1 h to get homogenous suspension. 55 and electrochemical impedance spectroscopy (EIS) in 10 ml of the suspension were dropped onto the polished 56 phosphate buffered saline solution simulating human surface of carbon pencil electrode and dried at room 57 body conditions. The surface of prepared electrodes was temperature for 60 min. 58 characterized by transmission electron microscopy www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 104 Full Paper 1 The deposition conditions for NiONPs were chosen 2 according to B. Rafiee et al. [29]. The deposition of 3 NiONPs was performed from a 40 mM Ni(NO3)2 · 6H2O 4 solution with pH = 2. The pH of the solution was adjusted 5 using sulfuric acid. The pulsed electrodeposition of 6 NiONPs on the PGE was carried out using an optimized 7 double-pulse sequence of potentials: E1 = 0.4 V (vs. Ag/ 8 AgCl/KCl) for t1 = 0.3 s and E2 = 0.0 V (vs. Ag/AgCl/KCl) 9 for t2 = 2.5 s. Electrochemical deposition of NiNPs was 10 performed by applying constant potential E1 for time t1. 11 Prepared electrodes were activated in 0.1 M NaOH 12 solution using cyclic voltammetry method with potential 13 cycling between 100 mV and 700 mV for 10 cycles at the 14 scan rate of 100 mV/s. This step was applied for maximum 15 activation of electrode surface towards insulin oxidation 16 [29]. 17 18 2.4 Electrochemical Determination of Active Surface of 19 Bare PGE and Modified Electrodes 20 21 The active surface area of bare PGE, chitosan-MWCNTs/ 22 PGE, NiNPs/chitosan-MWCNTs/PGE and NiONPs/chito- Fig. 1. TEM image of chitosan-MWCNTs/PGE. 23 san-MWCNTs/PGE was determined using double poten- 24 tial step chronoamperometry method in a solution consist- 25 ing of 1 mM K4[Fe(CN)6] and 0.1 M NaCl. The applied 26 potential steps were E1 = open circuit potential for each the grid surface (Figure 2 A, D). TEM images (Figure 2 27 electrode (176 mV for bare PGE, 132 mV for chitosan- A, D) show that the surface of MWCNTs became coarser 28 MWCNTs/PGE, 54 mV for NiNPs/chitosan-MWCNTs/ after loading with NiNPs and NiONPs in comparison with 29 PGE and 42 mV for NiONPs/chitosan-MWCNTs/PGE) chitosan-MWCNTs surface (Figure 1). Also, detailed 30 and E2 = 700 mV. STEM images of NiNPs/chitosan-MWCNTs/PGE (Fig- 31 ure 2 B, C) and NiONPs/chitosan-MWCNTs/PGE (Fig- 32 ure 2 E, F) show that NiNPs and NiONP were deposited 33 3 Results and Discussion on the surface of chitosan-MWCNTs. NiNPs and NiONPs 34 are mainly distributed on the ruggedness of the chitosan- 3.1 Morphological Characterization of 35 MWCNTs surface. The size and shape of deposited NiNPs NiNPs/chitosan-MWCNTs/PGE and 36 and NiONPs is very similar. The spherical NiNPs and NiONPs/chitosan-MWCNTs/PGE Surface 37 NiONPs are 2 nm–5 nm in diameter without generation of 38 The surfaces of unmodified chitosan-MWCNTs/PGE and aggregates. The chemical composition of the NiNPs/ 39 chitosan-MWCNTs/PGE modified with NiNPs and chitosan-MWCNTs/PGE and NiONPs/chitosan- 40 NiONPs were examined by TEM and STEM at different MWCNTs/PGE surface was determined by EDX analyses. 41 magnifications to get sufficient information about appear- The EDX spectrum confirmed the nickel nanoparticles 42 ance, surface morphology and distribution of metallic deposition and the presence of magnesium (Mg), calcium 43 nanoparticles. Moreover, the mapping EDX analysis was (Ca), silicon (Si), phosphorus (P) and iron (Fe) was caused 44 used to characterise the surface chemical composition of by pencil substrate composition. Figure 2 shows the 45 both modified electrodes. Figure 1 shows the TEM image electrochemically deposited NiONPs and NiNPs particles 46 of chitosan-MWCNTs/PGE without NiNPs or NiONPs. at the chosen potentials. The amount of Ni deposited on 47 The MWCNTs with diameter of 10 nm–20 nm and several the electrode surface was 16.64 % and 20.92 % for NiNPs/ 48 micrometres in length can be seen in Figure 1. The surface chitosan-MWCNTs/PGE and NiONPs/chitosan- 49 of unmodified MWCNTs covered by a chitosan layer is MWCNTs/PGE, respectively. 50 smooth. 51 Figure 2 shows the TEM (A, D) and STEM (B, C, E, 3.2 Electrocatalytic Oxidation of Insulin 52 F) images of modified electrodes together with EDX 53 elemental maps of carbon and nickel (G, H) expressing Electrochemical behaviour of various PGE surfaces towards 54 the elemental composition of the surface of NiNPs/ insulin oxidation in 0.1 M NaOH and PBS was studied using 55 chitosan-MWCNTs/PGE (A, B, C, G) and NiONPs/ cycling voltammetry method (CV) in a potential window 56 chitosan-MWCNTs/PGE (D, E, F, H). Before TEM ranging from 1 V to 2 V with a scan rate of 100 mV/s and 57 images were obtained all samples were placed on copper all measured plots are without iR correction. Figure 3 shows 58 grids which caused unequal distribution of MWCNTs on cyclic voltammograms recorded in 0.1 M NaOH and PBS www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 105 Full Paper 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Fig. 3. Cyclic voltammograms for different electrodes obtained in 0.1 M NaOH and PBS in potential range from 1 V to 2 V, with 18 the scan rate of 100 mV/s, in the absence of insulin (a) and in the 19 presence of 2 mM insulin (b–e): bare PGE (b), chitosan- 20 MWCNTs/PGE (c) NiONPs/chitosan-MWCNTs/PGE (d) and 21 NiNPs/chitosan-MWCNTs/PGE (e). 22 23 24 for bare PGE in absence of insulin (a) and in the presence 25 of 2 mM insulin for bare PGE (b) chitosan-MWCNTs/PGE 26 (c), NiONPs/chitosan-MWCNTs/PGE (d), and NiNPs/chi- 27 tosan-MWCNTs/PGE (e). 28 As shown in these voltammograms, current response 29 towards insulin oxidation on bare PGE (Figure 3 b) is 30 significantly lower in comparison to current response 31 towards insulin oxidation on chitosan-MWCNTs/PGE 32 (Figure 3 c), NiONPs/chitosan-MWCNTs/PGE (Figure 3 33 d) and NiNPs/chitosan-MWCNTs/PGE (Figure 3 e). 34 There is also no apparent peak found for the bare PGE 35 (Figure 3 b), confirming that the bare PGE has no 36 significant electrocatalytic activity towards insulin in 37 aqueous alkaline solutions. Two oxidation peaks were 38 observed in cyclic voltammograms for NiONPs/chitosan- 39 MWCNTs/PGE (d), and NiNPs/chitosan-MWCNTs/PGE 40 (e) because of the two-step oxidation process on the 41 modified electrode. Moreover, reduction peak was de- 42 tected for both electrodes, at potential 0.85 V for NiONPs 43 and at potential 0.35 V for NiNPs. The difference in the 44 values of oxidation and reduction peaks potentials 45 indicates the different mechanism of oxidation process. 46 Firstly, the current response of NiONPs/chitosan- 47 MWCNTs/PGE is significantly lower in comparison with 48 this of the NiNPs/chitosan-MWCNTs/PGE. The mecha- 49 nism of insulin oxidation on NiONPs/chitosan-MWCNTs/ 50 Fig. 2. TEM image of NiNPs/chitosan-MWCNTs/PGE (A), de- PGE can be shown as follows: 51 tailed STEM image of NiNPs/chitosan-MWCNTs/PGE (B, C), 52 EDX maps of Ni and C, with X-ray spectrum of the NiNPs/ NiO þ OH ! NiOðOHÞ þ e ð1Þ 53 chitosan-MWCNTs/PGE surface (G), TEM image of NiONPs/ 54 chitosan-MWCNTs/PGE (D), detailed STEM image of NiONPs/ NiOðOHÞ þ insulin ! NiðOHÞ2 þ product ½33 ð2Þ chitosan-MWCNTs/PGE (E, F) and EDX maps of Ni and C with 55 X-ray spectrum of the NiONPs/chitosan-MWCNTs/PGE surface The oxidative current peaks at 1.2 V and 1.45 V (Fig- 56 (H). 57 ure 3 d) correspond to the reactions expressed by 58 Equations 1 and 2, respectively. www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 106 Full Paper 1 Because of extremely sensitivity of Ni(OH)2 to elec- 2 tron exposure, redox peak appear at potential 0.85 V, 3 which corresponds to the reaction described by Equa- 4 tion 3: 5 6 NiðOHÞ2 þ OH ! NiO þ H2 O ½40 ð3Þ 7 8 Secondly, NiNPs/chitosan-MWCNTs/PGE displays en- 9 hanced current response and oxidation potentials shifted 10 to lower values (0.6 V and 1.1 V), as a consequence of 11 different electroactive species. These oxidative current 12 peaks at 0.6 V and 1.1 V (Figure 3 e) correspond to the 13 reactions expressed by Equations 4–6: 14 15 Ni þ 2OH ! NiðOHÞ2 þ 2e ð4Þ 16 NiðOHÞ2 þ OH ! NiOðOHÞ þ H2 O þ e ð5Þ 17 Fig. 4. Cyclic voltammograms for NiONPs/chitosan-MWCNTs/ 18 NiOðOHÞ þ insulin ! NiðOHÞ2 þ product ½33 ð6Þ PGE (A) and NiNPs/chitosan-MWCNTs/PGE (B) obtained in 19 0.1 M NaOH and PBS in potential range from 1 V to 2 V, with 20 Also potential of redox peak shifted to the more the different scan rates of 25, 50, 75 and 100 mV/s. The plot of 21 oxidation peak current versus scan rate for NiONPs/chitosan- negative values (0.35 V) in comparison to redox peak for 22 MWCNTs/PGE (C) and NiNPs/chitosan-MWCNTs/PGE (D). electrode modified by NiONPs (0.85 V). 23 24 The NiO(OH) species present during the electro- I pa ðNiNPs=chitosan  MWCNTs=PGEÞ ¼ 25 chemical oxidation of insulin on both electrode surfaces, ð8Þ 26 catalyse the oxidation of insulin and enhance current 2:0292:105 vðmV=sÞ þ 1:84:104 27 response when compared to bare electrode [29]. The 28 distance between oxidative and redox peaks in both cases According these results, the electron transfer between 29 indicates irreversible electrochemical oxidation of insulin. the redox sites of both electrode surfaces is a typical 30 Results of electrochemical measurements were not cor- surface-controlled process. The rate determining step is 31 rected for iR drop so the oxidation potentials were shifted the charge transfer on electrode surface. 32 to more positive values. Enlargement of the electrode surface area due to the 33 These measurements confirm better electrocatalytic deposition of nanoparticles increases active surface area 34 activity of NiONPs/chitosan-MWCNTs/PGE and NiNPs/ and number of active sites. The active surface area of 35 chitosan-MWCNTs/PGE towards insulin oxidation com- electrode was measured by chronoamperometry and 36 pared to bare PGE or chitosan-MWCNTs/PGE. Through calculated according to the Cottrell equation: 37 the obtained voltammograms the mechanism of insulin 38 oxidation on NiONPs/chitosan-MWCNTs/PGE and nFAD1=2 C ð9Þ i¼ 39 NiNPs/chitosan-MWCNTs/PGE was described (Equa- p1=2 t1=2 40 tions 1–2 and 4–6). 41 where i is measured current, n is number of exchanged 42 electrons, F is a Faraday constant, A is a surface area of 3.3 Effect of Scan Rate and Active Surface Area of working electrode, C a concentration of K4[Fe(CN)6] and t 43 Working Electrode is the time in seconds. Chronoamperograms for bare PGE 44 45 The effect of scan rate on the electrochemical behaviour (a) chitosan-MWCNTs/PGE (b), NiONPs/chitosan- 46 was studied in 0.1 M NaOH and PBS for NiONPs/ MWCNTs/PGE (c), and NiNPs/chitosan-MWCNTs/PGE 47 chitosan-MWCNTs/PGE (Figure 4 A, C) and NiNPs/ (d) were obtained in 1 mM K4[Fe(CN)6] and 0.1 M NaCl 48 chitosan-MWCNTs/PGE (Figure 4 B, D) in the presence and the plots of oxidation peak current versus t1=2 were 49 of 2 mM insulin. In both cases the anodic peak current fitted by linear function (Figure 5 inset). The active 50 exhibited linear increase with the rising scan rate, with surface areas were 0.33 mm2 for bare electrode, 1.33 mm2 51 correlation coefficient R2 0.99 and 0.95 for NiONPs/ for chitosan-MWCNTs/PGE, 1.34 mm2 for NiONPs/chito- 52 chitosan-MWCNTs/PGE and NiNPs/chitosan-MWCNTs/ san-MWCNTs/PGE and 1.39 mm2 for NiNPs/chitosan- 53 PGE, respectively. The linear regression equations are: MWCNTs/PGE. The active surface area increased more 54 than 4 times after modification, what represents rapid 55 I pa ðNiONPs=chitosan  MWCNTs=PGEÞ ¼ enlargement of active surface area of modified electrodes. ð7Þ The value of the diffusion coefficient of insulin was 56 0:00303vðmV=sÞ þ 0:37 57 calculated on the basis of Cottrell equation and measure- 58 ment was realized for 2 mM insulin and 0.1 M NaOH in www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 107 Full Paper 1 solution resistance, Cdl is double-layer capacitance and Rct 2 is charge transfer reaction resistance. The increase in 3 semicircle diameter in the impedance spectrum represents 4 the decrease in electron-transfer resistance indicating the 5 enhanced electrochemical redox reaction on the surface 6 of the electrode. 7 As shown in Figure 6, the largest semicircle was 8 observed for bare PGE (a), which correspond to the 9 surface with the lowest conductivity. Higher conductivity 10 was achieved by modification of PGE surface with 11 MWCNTs (Figure 6 inset c), while the conductivity 12 decreased after addition of chitosan (Figure 6 inset b). 13 Opposite trend was observed in case of comparison of 14 chitosan-MWCNTs/NiNPs (Figure 6 inset e) with 15 MWCNTs/NiNPs (Figure 6 inset d), where chitosan 16 application caused the decrease in Rct and improves the 17 Fig. 5. Current transients for bare PGE (a) chitosan-MWCNTs/ charge transfer efficiency at electrolyte-electrode inter- 18 PGE (b), NiONPs/chitosan-MWCNTs/PGE (c), and NiNPs/ face. This effect is associated with the stabilization of 19 chitosan-MWCNTs/PGE (d) obtained in 1 mM K4[Fe(CN)6] and NiNPs on MWCNTs PGE surface with chitosan resulting 20 0.1 M NaCl. Inset: the plots of oxidation peak current versus t 1=2. in increased conductivity. The same phenomenon was 21 observed for NiONPs/MWCNTs/PGE (Figure 6 inset f) 22 and NiONPs/chitosan-MWCNTs/PGE (Figure 6 inset g). 23 PBS. The surface area was 0.33 mm2 and n = 1. The In conclusion, the electrode modified by NiONPs and 24 calculated diffusion coefficient was 0.08.106 cm2/s. NiNPs fixed with chitosan enhanced conduction pathway 25 with higher electron transfer. Among them, the NiONPs 26 exhibited higher conductivity and electrocatalytic activity. 3.4 Electrochemical Impedance Spectroscopy 27 To evaluate the effect of insulin concentration on EIS 28 EIS is important analytical tool which was used to study measurement, the Nyquist plots at different insulin 29 the catalytic procedure of insulin electrooxidation on concentrations (1 mM–5 mM) in 0.1 M NaOH in PBS on 30 different electrode surfaces. Figure 6 shows Nyquist plots NiNPs/chitosan-MWCNTs/PGE and NiONPs/chitosan- 31 recorded at 1.45 V in frequency range of 0.1 Hz to 100 MWCNTs/PGE surface were obtained. 32 000 Hz in presence of 2 mM insulin and 0.1 M NaOH in The decrease in semicircles diameter with rising insulin 33 PBS. The most suitable circuit that describes electro- concentration can be seen from the Nyquist plots of both 34 chemical system is depicted in Inset in Figure 6. RS is the NiNPs/chitosan-MWCNTs/PGE (Figure 7 A) and 35 NiONPs/chitosan-MWCNTs/PGE (Figure 7 B) recorded 36 at 1.45 V. Insets in Figure 7 A and B show the insulin 37 concentration dependence on Rct fitted by linear function. 38 The value of Rct decreased with rising insulin concen- 39 tration. Also, the Rs slightly move to the lower values 40 because the solution conductivity/resistivity is directly 41 connected with the concentration of electroactive species 42 and increases/decreases with the rising concentration. 43 Correlation coefficients were 0.89 and 0.95 for NiNPs/ 44 chitosan-MWCNTs/PGE and NiONPs/chitosan- 45 MWCNTs/PGE surfaces, respectively. It follows from 46 these results, that the electrochemical response of 47 NiONPs/chitosan-MWCNTs/PGE is more linear in com- 48 parison with NiNPs/chitosan-MWCNTs/PGE. 49 50 3.5 Effect of Different Insulin Concentration 51 52 On the basis of the voltammetric results described above, Fig. 6. Nyquist plots of different modification of PGE surface: 53 the NiONPs/chitosan-MWCNTs/PGE and NiNPs/chito- unmodified (a), chitosan-MWCNTs/PGE (b), MWCNTs/PGE 54 (c), NiNPs/MWCNTs/PGE (d), NiNPs/chitosan-MWCNTs/PGE san-MWCNTs/PGE appears to be a suitable sensor for 55 (e) NiONPs/MWCNTs/PGE (f) and NiONPs/chitosan-MWCNTs/ the sensitive determination of insulin. Figure 8 shows the 56 PGE (g) in the presence of 2 mM insulin and 0.1 M NaOH in PBS. effect of the various concentrations of insulin (0.05 mM– 57 Applied potential 1.45 V and frequency range of 0.1 Hz to 100 5 mM) on anodic peak current (1.45 V) at the NiONPs/ 58 000 Hz. Inset: equivalent circuit used for data fitting. chitosan-MWCNTs/PGE in 0.1 M NaOH and PBS solu- www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 108 Full Paper 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Fig. 7. Nyquist plots of NiNPs/chitosan-MWCNTs/PGE (A) NiONPs/chitosan-MWCNTs/PGE (B) in the presence of 1 mM (a), 2 mM 20 (b), 3 mM (c), 4 mM (d) and 5 mM (e) insulin in 0.1 M NaOH and PBS. Inset: dependency of Rct on insulin concentration. 21 22 23 san-MWCNTs/PGE are comparable with those obtained 24 by common modified electrodes (Table 1). 25 Also correlation coefficient was similar like correlation 26 coefficients obtained by common modified electrode 27 mentioned in Table 1 [25, 38, 39]. 28 Oxidation of insulin (0.05 mM–5 mM) in 0.1 M NaOH 29 and PBS on the NiNPs/chitosan-MWCNTs/PGE surface 30 was also studied. As shown in cyclic voltammograms of 31 insulin on NiNPs/chitosan-MWCNTs/PGE surface (Fig- 32 ure 9), the anodic peak current (1.1 V) was slightly shifted 33 to lower values, compared to anodic peak current of 34 35 36 37 38 39 Fig. 8. Cyclic voltammograms of 0.1 M NaOH with PBS in the absence (a) and in the presence of 0.05 mM (b), 0.1 mM (c), 40 0.5 mM (d), 0.75 mM (e), 1 mM (f), 2 mM (g), 3 mM (h), 4 mM (i)and 41 5 mM (j) insulin on NiONPs/chitosan-MWCNTs/PGE in potential 42 range 1 V to 2 V with the scan rate of 100 mV/s. Inset: the linear 43 plot of insulin oxidation peak current versus insulin concentra- 44 tion, with correlation coefficient and the straight line equation. 45 46 47 tion. The anodic peak current increased linearly by rising 48 of the insulin concentration, what was confirmed by 49 determination of the correlation coefficient (R2 0.99). The 50 plot of current response versus insulin concentration is 51 linear over the wide concentration range of 0.05 mM–5 mM 52 (Figure 8, inset). The sensitivity of modified electrode Fig. 9. Cyclic voltammograms of 0.1 M NaOH with PBS in the absence (a) and in the presence of 0.05 mM (b), 0.1 mM (c), 53 emerged from the calibration plot with the correlation 0.5 mM (d), 0.75 mM (e), 1 mM (f), 2 mM (g), 3 mM (h), 4 mM (i) 54 coefficient of 0.99 was 0.64 mA/mM. The limit of detection, and 5 mM (j) insulin on NiNPs/chitosan-MWCNTs/PGE in 55 evaluated at a signal to noise ratio of 3 : 2, was found to be potential range 1 V to 2 V with the scan rate of 100 mV/s. Inset: 56 260 nM. The linear calibration range, sensitivity and limit the linear plot of insulin oxidation peak current versus insulin 57 of detection for insulin determination at NiONPs/chito- concentration, with correlation coefficient and the straight line 58 equation. www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 109 Full Paper 1 Table 1. Comparision of analytical parameters for insulin determination with several modified electrodes. 2 Electrode Method LOD Sensitivity Dynamic range Reference 3 4 NiONPs/chitosan-MWCNTs/PGE [a] Amperometry 260 nM 0.64 mA/mM 0.05 mM–5 mM This work 5 DHP [b]/MWCNTs/GCE [c] Amperometry 1 mM 1.33 nA/mM 0.8 mM–2.5 mM [38] 6 CNTs/GCE Amperometry 0.15 mM – 1.5 mM–2.5 mM [1] 7 MWCNTs/DMF [d]/CE [e] Amperometry 250 nM – 250 nM–1.6 mM [39] RuO/RuCN/CME [f] Amperometry 500 nM 441 mA/mM – [32] 8 MWCNTs-Nafion/NiONPs/SPCE [g] Amperometry 6.1 nM 1.83 mA/mM 20 nM–260 nM [29] 9 NiNP/CNTs-modified carbon fiber Amperometry 270 nM 1.11 nA/mM 2 mM–20 mM [35] 10 microelectrode 11 [a] Pencil graphite electrode, [b] Dihydropyran, [c] Glassy carbon electrode, [d] Dimetylformamide, [e] Carbon electrode, [f] Carbon 12 microelectrode, [g] Screen printed carbon electrode 13 14 15 insulin oxidation on NiONPs/chitosan-MWCNTs/PGE 3.6 Stability of NiONPs/chitosan-MWCNTs/PGE 16 (1.45 V). The plot of current response versus insulin 17 concentration was linear only over the concentration Stability test for NiONPs/chitosan-MWCNTs/PGE and 18 range of 1 mM 5 mM (Figure 9, inset), which was lower NiNPs/chitosan-MWCNTs/PGE was performed in 1 mM 19 concentration range in comparison with linear concen- K4[Fe(CN)6] and 0.1 M NaCl solution in potential range 20 tration range obtained for NiONPs/chitosan-MWCNTs/ 0.1 V–0.6 V. Obtained cyclic voltemmograms are shown 21 PGE (0.05 mM–5 mM) (Figure 8). Moreover, the sensitivity in Figure 10. After 50 cycles, current response decreased 22 (0.12 mA/mM) was lower, the correlation coefficient was by only less than 1 % for NiONPs/chitosan-MWCNTs/ 23 insufficient (R2 0.78) and linearity of calibration plot PGE (Figure 10 A) and 5 % for NiNPs/chitosan- 24 deteriorated. The limit of detection, evaluated at a signal MWCNTs/PGE (Figure 10 B). The results indicate a 25 to noise ratio of 3 : 2, was found to be 4.34 mM (Figure 9, better stability for NiONPs in comparison to NiNPs on 26 inset). Further this characteristic of NiNPs/chitosan- the chitosan-MWCNTs electrode surface. 27 MWCNTs/PGE surface was worse as compared to charac- Even, insulin with concentration 2 mM was determined 28 teristic obtained for NiONPs/chitosanMWCNTs/PGE. on five different NiONPs/chitosan-MWCNTs/PGEs and a 29 According to these results, it can be concluded, that reproducible current response with a great relative 30 NiONPs particles possess better electrocatalytic proper- standard deviation (RSD) 4.25 % was obtained (Fig- 31 ties and are most suitable material for insulin determi- ure 11). 32 nation. 33 Table 1 provides the comparison of limits of detection, 4 Conclusion 34 sensitivity and linear dynamic range towards insulin 35 oxidation for different modified carbon electrodes. For all A sensitive electrochemical sensor composed of electro- 36 mentioned electrodes the amperometry method of insulin deposited NiONPs nanoparticles on the PGE surface 37 determination was used. As can be seen in Table 1 the modified with chitosan-MWCNTs was fabricated. Chito- 38 prepared NiONPs/chitosan-MWCNTs/PGE possessed the san as a polymer material successfully stabilized the 39 widest dynamic range in comparison with other men- NiONPs on the modified PGE electrode surface. This fact 40 tioned electrodes (0.05 mM–5 mM). NiONPs/chitosan- was confirmed by TEM and STEM results and EDX 41 MWCNTs modification of PGE also resulted in an analysis, which indicated that the NiONPs with sizes of 42 excellent limit of detection (260 nm), which was signifi- about 4 nm were immobilized on the chitosan-MWCNTs 43 cantly lower than in the case of glassy carbon electrode surface. The small size of the NiONPs in combination 44 modified with MWCNTs and dihydropyran (1 mM) [38] with MWCNTs greatly improved the specific surface area 45 and carbon microelectrode modified with combination of of the PGE and enhanced electrocatalytic activity of PGE 46 RuO and RuCN (500 nM) [32]. Limit of detection of towards insulin oxidation. The electroanalytic properties 47 prepared NiONPs/chitosan-MWCNTs/PGE was also com- of NiONPs modified electrode were compared with 48 parable with other mentioned electrodes, such as characteristics of NiNPs/chitosan-MWCNTs/PGE. The 49 MWCNTs/dimetylformamide/carbon electrode (250 nM) NiONPs display much more suitable properties for insulin 50 [39] and NiNP/CNTs-modified carbon fibermicroelec- determination because of wider linear range, improved 51 trode (270 nM) [35]. Prepared NiONPs/chitosan- sensitivity and lower limit of detection. All measurements 52 MWCNTs/PGE had widest linear dynamic range were performed in PBS, which represents the physiolog- 53 (0.05 mM–50 mM) than mentioned electrode. NiONPs/ ical conditions in human body. All these characteristic 54 chitosan-MWCNTs/PGE had also good sensitivity denote the NiONPs/chitosan-MWCNTs/PGE as an effec- 55 (0.64 mA/mM) in comparison with other mentioned elec- tive and stable sensor for insulin detection in physiological 56 trodes mentioned in Table 1. condition. 57 58 www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 110 Full Paper 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Fig. 10. Cyclic voltammograms of 1st (a) and 50th (b) cycle of determination of 1 mM K4[Fe(CN)6] and 0.1 M NaCl on NiONPs/ 17 chitosan-MWCNTs/PGE (A) and NiNPs/chitosan-MWCNTs/PGE (B) in potential range 0.1 V to 0.6 V with the scan rate of 100 mV/ 18 s. 19 20 21 [5] N. D. D. Group, Diabetes 1995, 28, 1039. 22 [6] K. G. M. M. Alberti, P. Z. Zimmet, Diabetic Med. 1998, 15, 23 539–553. [7] J. Wang, M. Musameh, Anal. Chim. Acta 2004, 511, 33–36. 24 [8] A. Salimi, L. Mohamadi, R. Hallaj, S. Soltanian, Electro- 25 chem. Commun. 2009, 11, 1116–1119. 26 [9] M. Frasconi, C. Tortolini, F. Botre, F. Mazzei, F. Botrè, Anal. 27 Chem. 2010, 82, 7335–7342. 28 [10] W. Bürgi, M. Briner, N. Franken, A. C. Kessler, Clin. 29 Biochem. 1988, 21, 311–314. 30 [11] T. Tanaka, T. Matsunaga, Anal. Chem. 2000, 72, 3518–3522. 31 [12] M. Jaafariasl, E. Shams, M. K. Amini, Electrochim. Acta 2011, 56, 4390–4395. 32 [13] J. Hovancová, I. Šišoláková, R. Oriňaková, A. Oriňak, J. 33 Solid State Electrochem. 2017, 21, 2147–2166. 34 [14] Y. Luo, K. Huang, H. Xu, Anal. Chim. Acta 2005, 64. 35 [15] A. Hvass, B. Skelbaek-Pedersen, J. Pharm. Biomed. Anal. 36 2005, 37, 551–557. 37 [16] K. Ortner, W. Buchberger, M. Himmelsbach, J. Chromatogr. Fig. 11. Cyclic voltammogram of determination of 2 mM insulin in 38 A 2009, 1216, 2953–2957. 0.1 M NaOH and PBS on five different NiONPs/chitosan- [17] N. F. C. Visser, M. Van Harmelen, H. Lingeman, H. Irth, J. 39 MWCNTs/PGEs. Pharm. Biomed. Anal. 2003, 33, 451–462. 40 [18] M. Pikulski, W. Gorski, Anal. Chem. 2000, 72, 2696–2702. 41 [19] B. D. Arya, S. K., Singh, P. S. Malhotra, Electrochemical 42 Acknowledgments Techniques in Biosensors 2007. 43 [20] M. Zhang, C. Mullens, W. Gorski, Anal. Chem. 2005, 77, 44 This work was supported by the Projects APVV-16-0029 6396–6401. 45 of the Slovak Research and Development Agency, VEGA [21] L. Cheng, G. E. Pacey, J. A. Cox, Anal. Chem. 2001, 73, 46 1/0074/17 of the Slovak Scientific Grant Agency, VVGS- 5607–5610. [22] J. P. Hart, A. Crew, E. Crouch, K. C. Honeychurch, R. M. 47 PF-2018-795 and VVGS-PF-2018-794 of Pavol Jozef Pemberton, Anal. Lett. 2004, 37, 789. 48 Šafárik University in Košice. [23] S. Poorahong, C. Thammakhet, P. Thavarungkul, W. Limbut, 49 A. Numnuam, P. Kanatharana, Microchim. Acta 2012, 176, 50 91. 51 [24] J. Wang, G. Rivas, X. Cai, M. Chicharro, P. A. M. Farias, E. References Palecek, Electroanalysis 1996, 8, 902–906. 52 53 [1] Y. Wang, J. Li, Anal. Chim. Acta 2009, 650, 49–53. [25] M. R. Akanda, M. Sohail, M. A. Aziz, A. N. Kawde, Electro- [2] A. K. Yagati, Y. Choi, J. Park, J. W. Choi, H. S. Jun, S. Cho, analysis 2016, 28, 408–424. 54 Biosens. Bioelectron. 2016, 80, 307–314. [26] B. B. Prasad, R. Madhuri, M. P. Tiwari, P. S. Sharma, Electro- 55 chim. Acta 2010, 55, 9146–9156. [3] Y. Pu, Z. Zhu, D. Han, H. Liu, J. Liu, J. Liao, K. Zhang, W. 56 Tan, Analyst. 2011, 136, 4138–4140. [27] E. Martı́nez-Periñán, M. Revenga-Parra, M. Gennari, F. 57 [4] Y. Li, L. Tian, L. Liu, M. S. Khan, G. Zhao, D. Fan, W. Cao, Pariente, R. Mas-Ballesté, F. Zamora, E. Lorenzo, Sensors 58 Q. Wei, Talanta 2018, 179, 420–425. Actuators, B Chem. 2016, 222, 331–338. www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 111 Full Paper 1 [28] A. J. S. Ahammad, J.-J. Lee, M. A. Rahman, Sensors 2009, 9, [36] A. I. Gopalan, K. P. Lee, K. M. Manesh, P. Santhosh, J. H. 2 2289–2319. Kim, J. S. Kang, Talanta 2007, 71, 1774–1781. 3 [29] B. Rafiee, A. R. Fakhari, Biosens. Bioelectron. 2013, 46, 130– [37] B. Erdelyi, A. Oriňak, R. Oriňaková, J. Lorinčı́k, M. 135. Jerigová, D. Velič, M. Mičušı́k, M. Omastová, R. M. Smith, 4 [30] A. Arvinte, A. C. Westermann, A. M. Sesay, V. Virtanen, V. Girman, Appl. Surf. Sci. 2017, 396, 574–581. 5 Sensors Actuators, B Chem. 2010, 150, 756–763. [38] R. M. Snider, M. Ciobanu, A. E. Rue, D. E. Cliffel, Anal. 6 [31] N. Amini, M. B. Gholivand, M. Shamsipur, J. Electroanal. Chim. Acta 2008, 609, 44–52. 7 Chem. 2014, 714–715, 70–75. [39] P. Businova, J. Prasek, J. Chomoucka, J. Drbohlavova, J. 8 [32] J. Wang, T. Tangkuaram, S. Loyprasert, T. Vazquez-Alvarez, Pekarek, R. Hrdy, J. Hubalek, Procedia Eng. 2012, 47, 1235– 9 W. Veerasai, P. Kanatharana, P. Thavarungkul, Anal. Chim. 1238. 10 Acta 2007, 581, 1–6. [40] M. Hasanzadeh, R. Sabzi, Curr. Chem. Letters 2015, 42, 45– [33] Y. Yu, M. Guo, M. Yuan, W. Liu, J. Hu, Biosens. Bioelectron. 54. 11 2016, 77, 215–219. 12 [34] Y. Lin, L. Hu, L. Li, K. Wang, RSC Adv. 2014, 4, 46208– 13 46213. Received: July 13, 2018 14 [35] L. Lu, L. Liang, Y. Xie, K. Tang, Z. Wan, S. Chen, J. Solid Accepted: October 19, 2018 15 State Electrochem. 2018, 22, 825–833. Published online on November 13, 2018 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 www.electroanalysis.wiley-vch.de  2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2019, 31, 103 – 112 112