This is the accepted version of the following article: Kudr J., Zhao L., Nguyen E.P., Arola H., Nevanen T.K., Adam V., Zitka O., Merkoçi A.. Inkjet-printed electrochemically reduced graphene oxide microelectrode as a platform for HT-2 mycotoxin immunoenzymatic biosensing. Biosensors and Bioelectronics, (2020). 156. 112109: - . 10.1016/j.bios.2020.112109, which has been published in final form at https://dx.doi.org/10.1016/j.bios.2020.112109 © https://dx.doi.org/10.1016/j.bios.2020.112109. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 1 Inkjet-printed electrochemically reduced graphene oxide microelectrode 2 as a platform for HT-2 mycotoxin immunoenzymatic biosensing 3 Jiri Kudr a, b, Lei Zhao a, c, Emily P. Nguyen a, Henri Arola d, Tarja K. Nevanen d, Vojtech Adam b, e, 4 Ondrej Zitka b, e, Arben Merkoçi a, f,* a 5 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and The Barcelona Institute of 6 Science and Technology, Bellaterra, ES-08193, Spain b 7 Department of Chemistry and Biochemistry, Mendel University in Brno, Zemedelska 1, Brno, CZ-613 8 00, Czech Republic c 9 Department of Chemical Engineering, School of Engineering, UAB, Bellaterra, ES-08193, Spain d 10 VTT Technical Research Centre of Finland, Espoo, FI-02150, Finland e 11 Central European Institute of Technology, Brno University of Technology, Brno, CZ-616 00, Czech 12 Republic 13 f ICREA – Institucio Catalana de Recerca i Estudis Avançats, Barcelona, ES-08010, Spain 14 *Corresponding author. E-mail: arben.merkoci@icn2.cat (Arben Merkoçi) 15 ABSTRACT 16 The design and application of an inkjet-printed electrochemically reduced graphene oxide 17 microelectrode for HT-2 mycotoxin immunoenzymatic biosensing is reported. A water-based graphene 18 oxide ink was first formulated and single-drop line working microelectrodes were inkjet-printed onto 19 poly(ethylene 2,6-naphthalate) substrates, with dimensions of 78 μm in width and 30 nm in height after 20 solvent evaporation. The printed graphene oxide microelectrodes were electrochemically reduced and 21 characterized by Raman and X-ray photoelectron spectroscopy spectroscopies in addition to 22 microscopies. Through optimization of the electrochemical reduction parameters, differential pulse 23 voltammetry were performed to examine the sensing of 1-naphthol (1-N), where it was revealed that 2 24 reduction times had significant effects on electrode performance. The developed microelectrodes were 25 then used as an immunoenzymatic biosensor for the detection of HT-2 mycotoxin based on carbodiimide 26 linking of the microelectrode surface and HT-2 toxin antigen binding fragment of antibody (anti-HT2 27 (10) Fab). The HT-2 toxin and anti-HT2 (10) Fab reaction was reported by anti-HT2 immune complex 28 single-chain variable fragment of antibody fused with alkaline phosphatase (anti-IC-HT2 scFv-ALP) 29 which is able to produce an electroactive reporter – 1-N. The biosensor showed detection limits of 1.6 30 ng ∙ mL−1 and a linear dynamic range of 6.3 – 100.0 ng ∙ mL−1 within a 5 minute incubation with 1- 31 naphthyl phosphate (1-NP) substrate. 32 33 1. Introduction 34 Graphene is a carbon allotrope comprising of a single layer of carbon atoms arranged in a honeycomb 35 lattice structure. The oxygen content in graphene-based materials play an important role in the field of 36 electrochemistry, as it significantly decreases the material´s conductivity (Benchirouf et al. 2016; Chen 37 et al. 2011). Reduced graphene oxide (RGO) represents the transition material – possesses good 38 heterogeneous electron transfer rate but still enough oxygen groups that can facilitate functionalization 39 (Ren et al. 2017). Electrochemical reduction of graphene oxide is a fast, simple and economic alternative 40 approach to other conventional chemical reduction methods, which typically involve the use of toxic 41 reducing agents and can lead to contamination of the resulting material. Furthermore, electrochemical 42 reduction methods can be easily adapted for applications where thermal reduction is not possible, for 43 example reduction of structures on plastic substrates (Toh et al. 2014). 44 Traditionally, electronic devices fabricated by photolithography suffer from several disadvantages such 45 as complicated assembly processes, expensive instrumentation and use of harmful chemicals. Inkjet 46 printing (IJP) is a promising alternative technique for the production of advanced future electronic 47 devices as it allows for direct (no template) and non-contact printing onto flexible substrates such as 48 plastics, nanocellulose, and wearable e-textiles (Hermerschmidt et al. 2018; Hoeng et al. 2017; Karim et 49 al. 2017). Depending on the capabilities of the printer, even deposition of functional materials with 50 extraordinary precision can be achieved (Bachmann et al. 2017). Due to these outstanding properties, 51 IJP is not only a suitable choice for graphics, but also for printing of semi-conductive and/or insulating 52 materials for thin-film transistors, electrochemical sensors, and wearable electronics (da Silva et al. 53 2014; Medina-Sanchez et al. 2014; Zhan et al. 2017). Furthermore, IJP has the added advantage for ease 54 of mass production of small, disposable, low cost electrochemical biosensors. Depending on the 55 intended application, the selection of conductive material(s) is mainly determined by the desired 56 conductivity, optical transparency and bending stability (especially important for flexible electronics) 57 (He and Derby 2017; Kamyshny and Magdassi 2014). Typically, metal-based inks, such as Au or Ag 58 nanoparticle solutions, are used as standard materials for IJP to fabricate conductive tracks (Su et al. 59 2019; Zikulnig et al. 2019). Recently, the incorporation of 2D materials for IJP has been considered as 60 an alternative for some applications, because metallic nanoparticles based inks are expensive, tend to 61 oxidize and require high-temperature post-printing sintering (Bonaccorso et al. 2016; Park et al. 2007; 62 Singh et al. 2010). Moreover, Ag and Cu nanoparticles used for the conductive ink formulations are not 63 stable in the most common solvents such as water, ethanol or acetone (Dearden et al. 2005; Park et al. 64 2007). Although the characteristics of IJP graphene-based platforms cannot compete with the high 65 conductivity, deposition control and optical transparency of chemically deposited graphene, it has 66 nonetheless attracted attention of researchers worldwide (Bonaccorso et al. 2016; Singh et al. 2010). 67 One of the primary challenges of printing with pristine graphene is the tendency of the material to 68 aggregate, leading to low ink stability and printhead nozzles clogging (Michel et al. 2017). To overcome 69 this, one approach is to benefit from exfoliated GO as a source material, since oxygen-containing groups 70 of GO enables its dispersion in water-based inks. These inks can create continuous films over desired 71 substrates and a conductive film can be obtained using post-printing reduction (Pei and Li 2017). 4 72 With respect to agro-economics and the impact on global agriculture, the HT-2 toxin, produced by 73 different Fusarium species, is a member of the group of the most relevant mycotoxins (Hussein and 74 Brasel 2001). HT-2 can contaminate small grain cereals and maize in the field, but also during wet 75 storage and inappropriate transport. It has been shown that HT-2 toxin is only partially degraded during 76 thermal food processing (Kuchenbuch et al. 2018). Causing apoptosis and inhibiting protein synthesis, 77 HT-2 is contaminant of concern to human and animal health (Yang et al. 2017). According to a 78 Recommendation of The European Commission, in cases of analytical screening techniques the limit of 79 detection should not exceed 25 μg ∙ kg−1 for the sum of HT-2 and T-2 toxin (EC 2013). Common levels 80 of toxins in contaminated grains for human consumption and breakfast cereals were reaching from tens 81 to hundreds of μg ∙ kg−1 (Kaukoranta et al. 2019). 82 In this paper, we present the development of an ink, based on GO, and demonstrate its compatibility for 83 inkjet printing by fabricating electrodes where the working microelectrodes were formed via inkjet 84 printing. After optimization of the reduction parameters, we obtained electrochemically reduced 85 graphene oxide (ERGO) microelectrodes and the performance was evaluated by differential pulse 86 voltammetry (DPV) for the sensing of 1-naphthol (1-N). The results indicated that the ERGO 87 microelectrodes were suitable for biosensing and their application was demonstrated as an enzymatic 88 immunosensor for the detection of HT-2 mycotoxin. 89 2. Materials and methods 90 2.1. Chemicals 91 Ethylene glycol (EG), sodium dodecyl sulfate (SDS), 1-NP monosodium salt, 1-N, diethanolamine 92 (DEA), n-hydroxysuccinimide (NHS), n-(3-dimethylaminopropyl)-n-ethylcarbodiimide hydrochloride 93 (EDC), 2-(N-Morpholino)ethanesulfonic acid (MES) and other chemicals were purchased from Sigma- 94 Aldrich (St. Louis, MO, USA) in ACS purity and were used as received. All water-based solutions were 95 prepared using milli-Q water (resistivity higher than 18.2 MΩ ∙ cm−1 at 25 °C) from Millipore 96 (Burlington, MA, USA). 97 2.2. Fabrication of electrode 98 Poly(ethylene 2,6-naphthalate) (PEN, 0.075 mm thickness, Goodfellow, Huntingdon, UK) substrate was 99 cleaned using atmospheric-pressure oxygen plasma for 30 min (Harrick Plasma, Ithaca, NY, USA), 100 followed by soaking in 1M NaOH bath for 30 min at 25 °C, then washed with water and dried with 101 compressed air flow. The counter electrodes and graphite tracks of working electrodes were screen- 102 printed using carbon sensor paste (C2030519P4, Gwent Group, Pontypool, UK). Pseudoreference 103 electrodes were printed using Ag/AgCl paste (C61003P7, Gwent Group, Pontypool, UK). The electrodes 104 were cured at 120 °C for 30 min in an oven (JP Selecta, Barcelona, Spain). 105 40 mL of single layer GO water solution (N002-PS-1.0, Angstron Materials, Dayton, OH, USA) of 106 concentration 1 wt% was sonicated for 2 h (80 % amplitude, 1 s pulse, 1 s pause, cooling temperature 107 15 °C, protecting temperature 40 °C) using an ultrasonicator equipped with a 1000 W sonic head and a 108 water cooling system (Col-Int TECH, Irmo, SC, USA). To create the ink, 600 mg of the GO solution 109 was added to 3.6 mL of milli-Q water, 6 mg of SDS and 1.8 mL of EG and mixed using a Q125 probe 110 sonicator (Qsonica, Melville, NY, USA) with the following settings: time 300 s, amplitude 100 %, pulse 111 2 s, pause 1 s. The solution was cooled using an ice bath during sonication. The ink was obtained by 112 filtering with cellulose acetate syringe filter (0.45 μm cut-off, VWR, Radnor, PA, USA). 113 The GO working electrodes were printed using a Dimatix Materials Printer DMP-2850 (Fujifilm, 114 Minato, Japan) with a DMC-11610 printhead dispensing 10 pL nominal drop volume (nozzle diameter 115 of 21 μm). The printhead and substrate plate were set to 35 °C. A single nozzle with continuous jetting 116 was used to print 10 mm-long working electrodes. Each line was printed 30-times with 30 s pauses 117 between individual layers and drop-spacing of 30 μm. The dispersion was printed at a voltage of 11 V 6 118 with a frequency of 5 kHz. A custom waveform profile was used for the GO ink printing. The distance 119 between the printhead and substrate was set to 500 μm. 120 Prior to reduction of the GO, the electrodes were treated at 120 °C for 2 hours. Then, 50 μL of KCl 121 solution in milli-Q water was pipetted onto the electrode to cover the pseudoreference electrode, counter 122 electrode and the whole GO strip, including the undefined part of the graphite carbon working electrode 123 track. The reduction was performed using chronopotentiometry at −1.5 V. After the reduction, the 124 electrodes were washed with milli-Q water, dried with compressed air and the insulating line (Dielectric 125 ink D2070423P5, Gwent Group, Pontypool, UK) was added to prevent solutions from having contact 126 with the graphite part of the working electrode. Finally, the electrodes were treated at 200 °C for 5 min. 127 2.3. Biosensor fabrication 128 Anti-HT2 (10) Fab was attached to the ERGO microelectrode via EDC/NHS linker chemistry. In detail, 129 after the electrodes were briefly washed with MES buffer (pH 6), a 30 μL drop of EDC/NHS solution 130 (0.1 M, in MES buffer) was added onto the printed area and left to incubate at room temperature for 131 30 min. The electrodes were then thoroughly washed with MES buffer to remove any excess EDC/NHS 132 and 10 μL of anti-HT2 (10) Fab (typically 10 μg ∙ mL−1, in phosphate buffered saline (PBS) pH 6.5) was 133 added onto the electrode and incubated for another 30 min. The drop was then removed and 30 μL of 134 Tris buffer (pH 8) was added onto the electrode for 30 min. Subsequently, 30 μL of blocking buffer (2 % 135 bovine serum albumine in PBS pH 7.4 with 0.05 % tween-20) was pipetted onto the electrode and 136 incubated for 30 min. The electrodes were washed twice with washing buffer (PBS pH 7.4 with 0.05 % 137 tween-20) and samples containing the mycotoxin (30 μL total volume, diluted in blocking buffer in 1:1 138 ratio, 30 min incubation) were added. The electrodes were then washed thoroughly with washing buffer 139 and dried at 25 °C. Anti-IC-HT2 scFv-ALP was then added for 30 min and then removed with washing 140 buffer. Analysis was performed by addition of 5 mM 1-NP dissolved in 0.1 M DEA buffer (pH 9.8 141 containing 1 mM MgCl2). After 5 min of incubation with 1-NP, DPV electrochemical analysis was 142 performed using the following settings: initial potential −0.8 V, end potential 0.7 V, speed of 143 polarization 0.5 mV ∙ s−1, modulation amplitude 175 mV, modulation time 10 ms, interval time 1 s, no 144 deposition or conditioning potential was set. 145 2.4. Biological materials 146 Anti-HT2 (10) Fab and anti-IC-HT2 ScFv-ALP fusion antibody were provided by VTT (Espoo, Finland) 147 (Arola et al. 2017). HT-2 toxin (15-Acetoxy-3α,4β-dihydroxy-8α-(3-methylbutyryloxy)-12,13- 148 epoxytrichothec-9-ene) was bought from Sigma-Aldrich (St. Louis, MO, USA). The content of a whole 149 vial of the mycotoxin was dissolved in dimethyl sulfoxide (1 mg ∙ mL−1) to create the stock solution. 150 Individual HT-2 concentrations were obtained by dilution. Gibco One Shot Fetal bovine serum (Thermo 151 Fisher Scientific, Waltham, MA, USA) was used as a matrix for real sample analysis. 152 2.5. Electrode effective surface calculation 153 In order to determine the electroactive area of the ERGO microelectrode, cyclic voltammograms (CV) 154 were obtained using 1 mM [Fe(CN)6]3−/[Fe(CN)6]4− redox probe in 0.1 M KCl. The electroactive surface 155 was calculated according to Randles-Sevcik equation for reversible electrode process: 156 Ip = 2.69 ∙ 105 ∙ A ∙ D1/2 ∙ n3/2 ∙ ν1/2 ∙ C Eq. (1) 157 where Ip is anodic current peak (A), A is the electroactive area (cm2), D is the diffusion coefficient of 158 [Fe(CN)6]4− in solution and is 6.1 ∙ 10−6 cm2 · s −1, n is the number of electrons transferred in half- 159 reaction (1 in case of [Fe(CN)6]4−), ν is scan rate (0.1 V ∙ s−1 was chosen) and C is 160 [Fe(CN)6]4− concentration (mol ∙ L−1). 161 2.6. Instrumentation 162 For characterization of the materials and the electrodes, Magellan 400 L field emission scanning electron 163 microscope (SEM, FEI, Hillsboro, OR, USA), EasyDrop Contact Angle Measuring instrument (Kruss, 8 164 Hamburg, Germany), optical microscope (Olympus1X71, Olympus, Tokyo, Japan) were used. Asylum 165 MFP-3D atomic force microscope (AFM) (Oxford Instruments, Abingdon, UK) was used for electrode 166 characterization. Bruker Dimension FastScan atomic force microscope (Bruker Nano Surface, Santa 167 Barbara, CA, USA) operated in tapping mode was used for GO characterization. X-ray photoelectron 168 spectroscopic (XPS) measurements were performed at room temperature with a SPECS PHOIBOS 150 169 hemispherical analyzer (SPECS GmbH, Berlin, Germany) in a base pressure of 5 ∙ 10−10 mbar using 170 monochromatic Al K-alpha radiation (1486.74 eV) as an excitation source. XPS spectra were then 171 evaluated using Fityk sofware. Raman spectra were obtained using a confocal Raman microscope 172 alpha300r equipped with a 488 nm laser (WITec, Ulm, Germany) and parameters of 1.5 mW laser power, 173 grating of 600 g ∙ nm−1, objective 50x, exposure of 10 s and 3 accumulations for every single spectra. 174 For electrochemical measurements, an Autolab PGSTAT 302 (Metrohm, Herissau, Switzerland) 175 equipped with frequency response analyzer PGST30 was used. CV and DPV were evaluated using GPES 176 electrochemical software. EIS spectra, including equivalent circuit elements calculations, were evaluated 177 using NOVA (both Metrohm, Herissau, Switzerland). 178 3. Results and discussion 179 Commercial GO was utilized as the main component within the ink. Prior to ink formulation the GO 180 stock solution, containing 1 wt% of GO, was thoroughly sonicated. The dense black GO solution was 181 then mixed and sonicated with EG, water and SDS to obtain the GO water-based ink containing 182 1 mg ∙ mL−1 GO (0.1 wt%), 30 % EG and 1 mg ∙ mL−1 SDS. It is recommended for materials sizes to be 183 less than 1/100th the size of nozzle, to avoid obstruction of the printhead nozzle and guarantee successful 184 printing. To ensure that the GO sheets were of suitable dimensions, the GO sheets were initially treated 185 to intense continuous sonication and then filtered to remove the undesired sizes. Figure 1(a) is an SEM 186 micrograph of the ink and indicates that the GO collected after the formulation procedure was within the 187 lateral ranges of 226 – 875 nm. Compared to source GO, the lateral diameter of treated GO significantly 188 decreased after sonication (see Fig. S1 for comparison). It is well-known that sonication induces 189 fragmentation of the GO sheets and can create crystal structure defects, which are important for 190 heterogeneous electron transfer (HET) in electrochemistry (Ye and Feng 2016; Zhong et al. 2014). 191 Further, the presence of single-layer GO sheets within stock solution provided by manufacturer was 192 confirmed by atomic force microscopy (AFM). The obtained data showed the graphene oxide thickness 193 of about 1 nm (Fig. S2). In accordance with previously published articles, the used graphene oxide can 194 be interpreted as single-layer sheets (Bartlam et al. 2018; Paredes et al. 2009; Stankovich et al. 2006). 195 As the viscosity of water is too low and the surface tension is too high to be used for IJP, EG was added 196 as a rheology modifier and humectant, while SDS was added to adjust the ink ’s surface tension (Li et 197 al. 2018; Vaseem et al. 2012). Printing resolution and homogeneity of printed design are influenced by 198 droplet spread and are attributed to the wettability of the ink with respect to the substrate (Stringer and 199 Derby 2009). Both surface chemistry and roughness contribute to the wettability effects of substrate. 200 Since the droplets produced in IJP are very small (10 pL in our case) there are high demands for substrate 201 quality. As a substrate for electrode printing, we have selected transparent and flexible PEN suitable for 202 development of wearable electrode sensors (Shiwaku et al. 2018). The substrate was treated with 203 atmospheric-pressure oxygen plasma in combination with chemical hydrolysis prior to printing. As 204 reported by Gonzales et al. (Gonzalez et al. 2008), plasma treatment of PEN can induce oxidation of the 205 aromatic rings of the polymer chains. After the plasma treatment, the substrate was rinsed with NaOH 206 to increase the substrate roughness (Cammarano et al. 2013). Figure 1(b) and 1(c) shows a decrease of 207 the contact angle of the droplet upon plasma and NaOH treatment of the substrate, indicating that the 208 wettability of the substrate was improved after such treatments. 209 The printing of the electrochemical sensor was performed using various types of printing techniques. 210 The counter electrodes, pseudoreference electrodes and the working electrodes tracks were printed using 211 traditional screen printing techniques. The active area of the working electrode was inkjet-printed using 212 the developed GO ink, ensuring that a small section of the area was overlapping with the screen printed 10 213 track to enable its electrochemical activity. An SEM micrograph of the whole electrode can be seen in 214 Fig. S2. Following the reduction step, an insulating layer was added to cover the entire working electrode 215 graphitic parts and to define the active surface of electrodes. Scheme of ERGO microelectrodes main 216 fabrication steps can be seen in Fig. 1(d). 217 218 Fig. 1. (a) SEM micrograph of GO flakes contained in ink. (b) Contact angle for droplet of IJP ink on 219 non-treated and (c) treated PEN substrate. (d) Scheme of ERGO microelectrode fabrication. 220 3.1. ERGO microelectrode characterization 221 An optical microscope image of the working electrode track and the GO printed active area interface is 222 presented in Fig. 2(a). The large dark feature in the upper left is the end of the screen-printed working 223 electrode (WE) track, while a continuous strip of GO can be seen as a narrow grey line. The printed GO 224 strip exhibits well-defined edges with a width of 78 μm. As thermal reduction of the GO strip was not 225 possible, due to potential damages to the PEN substrate, an electrochemical approach was adopted for 226 the reduction of the GO active area to ERGO. The electrochemical reduction was performed in 3 mM 227 KCl solution at a potential of −1.5 V vs. Ag/AgCl. As reported elsewhere, successful reduction was 228 visually observed by a color change of the GO from light brown to light black (Fig. 1(b)) (Kudr et al. 229 2016). The reduction of GO to ERGO was also confirmed using Raman spectroscopy by observation of 230 the two characteristic D and G bands of graphene-based materials. It has been reported that the D band 231 is closely related to the degree of defects and disorder within the hexagonal lattice and should not be 232 observed in the case of pristine graphene, like that grown by chemical vapor deposition (Juang et al. 233 2010). Conversely, the G band corresponds to the in-plane vibrations of sp2 carbon atoms of graphene 234 lattice and thus increases with reduction of C-O/C=O. As such, in the case of ERGO, the ratio of these 235 peak intensities (ID/IG) are often used to survey the level of oxidation of graphene-based materials 236 (Ramesha and Sampath 2009). In both our samples of GO and ERGO, the D and G bands were observed 237 at 1354 cm−1 and 1578 cm−1, respectively (Fig. 1(c)) and without any shifts upon comparison of the 238 before and after spectra. However, the intensities of these peaks show obvious variations and suggest 239 successful electrochemical reduction of the GO to ERGO as the ID/IG ratio increased from 0.90 to 1.25. 240 Furthermore, this observation also suggests an increase of lattice-defects during the reduction process 241 and is in good agreement with previous studies (Kuang et al. 2018). Since the procedure of electrodes 242 fabrication included thermal treatment, Raman spectra of a thermally treated working electrode were 243 obtained as a control (see Fig. S3 for comparison). The ERGO electrodes were also characterized using 244 XPS to determine the chemical states of carbon and oxygen (Fig. 1(d)). Shirley background subtraction 245 and fitting with Voigt function (convolution of Gaussian and Lorentzian peak shapes) were performed. 246 Upon evaluation of the C1s spectrum, the carbon peak was deconvoluted to 5 individual peaks: C-C sp2 247 (284.5 eV), C-OH (285.5 eV), C-O (286.4 eV), C=O (287.6 eV) and O-C=O (289.9 eV) (Casero et al. 248 2014; Casero et al. 2013). Four distinguishable oxygen containing peaks were observed in the O1s 249 spectrum (C=O at 530.7 eV, O-C=O at 532.2 eV, OH at 533.1 eV and C-O-C at 533.7 eV). It 250 demonstrates that even after electrochemical reduction, carboxyl functional groups which will facilitate 251 covalent attachment of antibodies to electrode were still presented within ERGO electrodes (Raj and 252 John 2013). AFM height profile suggests that the ERGO strip electrode goes 30 nm above the PEN 253 substrate (Fig. 1(e)). Given that the intrinsic thickness of GO is 1 nm we can conclude that the electrode 254 strip was created by around 30 layered sheets of ERGO (Zhang et al. 2010). An AFM phase image of 255 the electrode and substrate borderline can be seen in Fig. S4. 12 256 257 Fig. 2. (a) Optical microscope image of IJP GO strip electrode before reduction and (b) during reduction. 258 (c) Raman spectra of GO electrode before reduction step (blue line) and after electrochemical reduction 259 with thermal treatment (red line). (d) Deconvoluted XPS C 1s and O 1s (inset) spectrum of ERGO. (e) 260 Height profile of ERGO strip electrode obtained by AFM. 261 3.2. ERGO microelectrode optimization for 1-N sensing 262 The ERGO microelectrodes were characterized with cyclic voltammetry at varying speeds of 263 polarization. As shown in Fig. 3(a), the ferri/ferrocyanide redox reaction displayed well-defined 264 oxidation and reduction peaks where it can be observed that there is an increase of the current with each 265 increased scan rate. The anion oxidation current of ferrocyanide was plotted against the square root of 266 the scan rates and revealed a linear relationship (Fig. 3(b)), suggesting that there is a diffusion controlled 267 mechanism for the oxidation/reduction of ferro/ferricyanide on the ERGO electrode. However, large 268 shifts of the faradaic peaks which suggested slow electrode kinetics towards oxidation/reduction of the 269 used probe were observed. Brownson et al. reported similar behavior for highly ordered pyrolytic 270 graphite electrodes with a single layer of pristine graphene (Brownson et al. 2011). This can be also 271 caused by the oxygen content in ERGO as the ferri/ferrocyanide redox probe is highly sensitive to the 272 oxygen content within the electrode material (Tan et al. 2014). Using Randles-Sevcik equation (Eq. (1)) 273 for reversible electrode process, the electroactive surface of electrode was determined to be 5.2 mm2 274 which was by 0.2 mm2 bigger than the geometrical surface. The influence of KCl concentration in the 275 reduction solution and reduction time was determined based on the sensitivity of the ERGO strip 276 electrode towards sensing of 1-N. These experiments were performed with electrodes fabricated using 277 the same procedure as described for the microelectrodes, but with a 10-times larger geometric surface 278 (1.0 × 5 mm ̴ 5 mm2) of working electrode used to observe the enhanced differences. As is evident from 279 Fig. 3(c), the reduction time of such fabricated GO electrodes influences 1-N sensing. It is consistent 280 with the findings of Feng et al., where they described the electrochemical reduction of graphene oxide 281 as a multi-step process and that the abundance of oxygen-containing groups depended on the reduction 282 time (Feng et al. 2016). It is worth noting that even after 300 s of reduction, at a constant potential of 283 −1.5 V vs. Ag/AgCl, the whole printed GO layer was reduced, as indicated by a color change from brown 284 to black. Since the reduction time of 900 s increased the electrochemical signal by 45 ± 5 % (n = 3), we 285 can conclude that the reduction time plays an important role for the sensing of 1-N. Further increase of 286 reduction time to 1200 s enhanced electrode performance negligibly. Regarding GO microelectrodes 287 reduction, a time of 1000 s was normally used but from our experience the reduction process is highly 288 dependent on the quality (homogeneity) of the printed GO strip (drop spacing, number of printed layers 289 and even substrate pretreatment procedure). No significant changes of 1-N signals were observed with 290 varying concentrations of KCl. This suggests that the electrolyte solution had no effect on the 291 performance of the ERGO microelectrode (Fig. 3(d)) and consequently a concentration of 3 mM KCl 292 was chosen for reduction of the GO electrodes. 293 The dependence of 1-N DPV signal on the modulation amplitude, modulation time and speed of 294 polarization were assessed. The optimal results of these measurements are indicated by a red diamond 295 in Fig. 3(e), (f) and (g). Thus for the DPV measurements, the modulation amplitude was set to 175 mV, 296 modulation time 10 ms and a scan rate of 5 mV ∙ s−1 was used. As a last step, the dependence of 1-N 297 oxidation signal at +0.05 V vs its concentration in DEA buffer was evaluated. The linear range of the 14 298 calibration curve was observed between 31.25 and 0.24 μM concentration with satisfactory factor of 299 determination, R2 = 0.994. It showed that an electrode with settings as described above enabled reliable 300 and sensitive detection of 1-N (Fig. 3(h)). 301 302 Fig. 3. (a) CV of 1 mM [Fe(CN)6]3−/[Fe(CN)6]4− in 0.1 M KCl (5 – 200 mV ∙ s−1) recorded using ERGO 303 microelectrode. (b) The dependence of 1 mM [Fe(CN)6]3−/[Fe(CN)6]4− anodic peak heights on the square 304 root of scan rate. (c) The dependence of 1-N peak on reduction time (300 – 900 s) of GO and (d) 305 concentration of KCl (3 – 100 mM) in reduction solution (red diamond indicates condition selected as 306 optimal; typical DPV voltammograms are shown within insets; arrow is showing trend corresponding to 307 x axis parameter increase). (e) Dependence of 1-N peak on modulation amplitude (5 – 200 mV), (f) 308 modulation time (10 – 100 ms) and (g) speed of polarization (5 – 50 mV ∙ s−1). (h) Dependence of signal 309 on 1-N concentration (0.24 – 31.25 µM). 310 311 3.3. ERGO strip microelectrode biosensor 312 Biosensors based on electrochemical enzymatic activity are promising for numerous diagnostic fields, 313 such as medical screening, food analysis and environmental monitoring (Zhang et al. 2018a; Zhang et 314 al. 2018b). The detection system we used as a proof-of-concept is well documented and is still frequently 315 used as an ELISA (enzyme-linked immunosorbent assay). The proposed detection mechanism is based 316 on the attachment of immobilization antibody fragment (anti-HT2 (10) Fab) onto the electrode. The 317 coupling was performed via carbodiimide chemistry, using EDC/NHS, where the residual carboxylic 318 groups found on the surface of ERGO were crosslinked to the primary amines of the Fab fragment. The 319 anti-IC-HT2 scFv genetically fused with alkaline phosphatase (ALP) is able to recognize the 320 immunocomplex of HT-2 toxin and anti-HT2 (10) Fab. Anti-HT2 (10) Fab is not able to specifically 321 recognize HT-2 from T-2 toxin thus the specificity of analysis was provided by anti-IC-HT2 scFv which 322 shows no cross-reactivity with T-2 and the related molecules (T-2-triol, T-2-tetraol, diacetoxyscirpenol, 323 15-acetyldeoxynivalenol, 3- acetyldeoxynivalenol, deoxynivalenol, deoxynivalenol-3-glucoside, 324 nivalenol, and neosolaniol) (Arola et al. 2016). ALP is an enzyme which cleaves the phosphate from 325 different molecules with higher efficiency in alkaline conditions. Here non-electroactive 1-NP was used 326 as an ALP substrate. Electroactive product 1-N was produced from 1-NP by ALP (Fig. 4(a)). 327 Biosensors should be cost-effective and allow for scalable manufacturing. IJP and the use of recombinant 328 fusion antibodies definitely meet these demands. The proposed method of graphene-related material IJP 329 in combination with inkjet-printed tracks (screen-printed here) can dramatically decrease the price of 330 such sensors. The electrochemical reduction seems to be inappropriate since it takes several minutes 331 depending on printed morphology. However the use of a multichannel potentiostat can make this process 332 more time-effective. 16 333 334 Fig. 4. (a) Scheme of suggested ERGO microelectrode HT-2 toxin biosensor. (b) EIS nyquist plot of 335 ERGO microelectrode modified with different amount of anti-HT-2 Fab (0 – 66.7 ng ∙ mm−2) in 1 mM 336 [Fe(CN)6]3−/[Fe(CN)6]4− in 0.1 M KCl and equivalent circuit used for data evaluation in inset. (c) 337 Dependence of R component of EIS equivalent circuit resistance on amount applied anti-HT-2 Fab 338 (0 – 66.7 ng ∙ mm−2). (d) Dependence of suggested biosensor background 1-N signal on amount of 339 applied anti-HT-2 IC scFv-ALP (5 – 600 ng). (e) Dependence of DPV 1-N oxidation signal on 340 concentration of applied HT-2 toxin (6.3 – 100.0 ng ∙ mL−1). Baseline-corrected response (peak) of 341 biosensor to HT-2 toxin at concentration corresponding to calibration curve (inset). 342 3.4. Application of the designed biosensor 343 To demonstrate the capabilities of our developed biosensor for possible real-world applications, we have 344 chosen to target the detection of the HT-2 mycotoxin. HT-2 mycotoxins are highly toxic trichothecene 345 mycotoxins produced by Fusarium species and have adverse effects on plant products such as wheat, 346 oats and barley, which are all prevalent crops in the European union (Bryla et al. 2018). The harmful 347 effects of HT-2 are manifested not only by inhibition of protein synthesis and cell proliferation in plants, 348 but can also cause acute or chronic intoxication of humans and animals (Rocha et al. 2005). Therefore a 349 rapid and low-cost detection platform would prove to be advantageous for the health and food safety 350 industry. 351 The successful attachment of anti-HT2 (10) Fab onto the ERGO working electrode was monitored by 352 electrochemical impedance spectroscopy (EIS). The amount of applied anti-HT2 (10) Fab was 353 transformed into nanogram per square millimeter of working electrodes (ng ∙ mm−2). As can be seen in 354 Fig. 4(b), the increasing concentration of attached anti-HT2 (10) Fab resulted in incremental change of 355 the depressed semi-circles in the Nyquist diagram. The impedance measurements were fitted using an 356 equivalent circuit (Fig. 4(b) inset) and the corresponding electronic circuit elements were calculated. 357 The resistive element, labelled in equivalent circuit R, showed increasing trends with increasing amount 358 of applied anti-HT2 (10) Fab (Fig. 4(c)). The amount of applied anti-IC-HT2 scFv-ALP was optimized 359 in order to obtain minimal background response. The same corresponding concentrations of anti-IC- 360 HT2 scFv-ALP were incubated with the anti-HT2 (10) Fab modified electrodes anti-HT2 (10) Fab 361 without introduction of the HT-2 toxin. In our workflow of washing steps, 30 μL of anti-IC-HT2 scFv- 362 ALP (2.5 μg ∙ mL−1) ̴ 75 ng was selected as the optimal amount of anti-IC-HT2 scFv-ALP added onto 363 the electrode (Fig. 4(d)). Background signal in case of higher anti-IC-HT2 scFv-ALP concentrations 364 substantially increased which is not desirable. 365 Using optimized conditions, the biosensor was used to determine different concentrations of HT-2 toxin. 366 The response of the biosensor showed increasing DPV 1-N signals with increasing concentrations of 367 HT-2 toxin and showed good linear dependence (Fig. 4(e)) with a limit of detection (LOD) of 368 1.6 ng ∙ mL−1. The low LOD observed is comparable with other methods such as ELISA, surface 369 plasmon resonance biosensors or high-performance liquid chromatography with fluorescence detection 370 (Lippolis et al. 2008; Meneely et al. 2010; Yoshizawa et al. 2004). Khan et al. (Khan et al. 2018) reported 371 on a fluorescence resonance energy transfer amplified aptasensor with T-2 toxin (T-2 toxin is rapidly 372 hydrolyzed to HT-2 toxin in vivo) with a LOD of 0.93 pg ∙ mL−1. Detection limit of 0.38 ng ∙ mL−1 was 373 reported for non-competitive immunoassay based on utilization of time-resolved fluorescence resonance 18 374 energy transfer (Arola et al. 2016). However, in our case the sensitivity, as well as other detection 375 parameters, can be further tuned by prolonging the incubation times of the substrate with the 376 immobilized ALP (during longer incubation time anti-IC-HT2 scFv-ALP will produce more 1-N). The 377 inset of Fig. 4(e) shows a representative response of the biosensor to 6.3 ng ∙ mL−1 of HT-2 toxin. 378 Noticeably, the peak of 1-N shifted from +0.05 V, in the case of 1-N sensing of the bare ERGO 379 microelectrode, to −0.08 V. This shift can be attributed to a change of the pseudoreference electrode 380 potential due to multiple incubation and washing steps. The analytical parameters of HT-2 detection 381 with the proposed system are shown in Tab. 1. Nowadays, ordinary smartphones can be used as complete 382 electrochemical workstations and miniaturized commercial potentiostats are available. Consequently the 383 reported electrode can be used as a part of affordable and portable point-of-need analytical devices 384 beneficial for food quality control and smart farming (da Silva et al. 2017; Vashist et al. 2015). The 385 reported fabrication procedure enables to manufacture the working electrode arrays attractive for 386 analysis multiplexing. Fabrication of such sensors on transparent plastic materials can be beneficial for 387 integration of additional sensing devices with optical readout. 388 As matrix effects can significantly influence the sensitivity of analysis, an assessment of the matrix 389 effects on the detection of HT-2 were tested using spiked fetal bovine serum (FBS). To determine 390 analytical recovery, one volume of HT-2 standard solution (1 μg ∙ mL−1) was mixed with 9 volumes of 391 FBS and a blocking buffer in ratio of 1:1. Recovery of the standard ranged from 87.8 to 108.5 % (n = 5). 392 393 394 Tab. 1. Analytical parameters of electrochemical detection of HT-2 toxin. Linear Working Regression LODb) Substance dynamic range R2 a) electrode equation (ng ∙ mL−1) (ng ∙ mL−1) y = 1.240x – HT-2 toxin ERGO 6.3– 100.0 0.999 1.6 3.583 395 a) Regression coefficient; b) LOD (S/N=3). 396 4. Conclusion 397 In this study we focused on the fabrication of reduced graphene oxide working electrodes, where reduced 398 graphene was not only the modifier of highly conductive electrode material but also the electrode 399 material itself. We designed and formulated a water-based commercial GO ink that was then used to 400 inkjet print graphene GO, measuring 78 μm in width, onto plasma-treated and chemically etched 401 poly(ethylene 2,6-naphthalate) substrates with screen-printed Ag/AgCl pseudoreference electrode, 402 graphite counter electrode and a graphite working electrode track. The electrochemical reduction of the 403 graphene oxide strip was performed in a KCl solution, where the reduction parameters and electrolyte 404 concentrations were optimized. The successful reduction of the GO was confirmed by Raman 405 spectroscopy, where an increase of D and G peak ratio was observed. The DPV sensing of 1-naphthol 406 with the ERGO microelectrode was optimized with the intention to create an enzymatic electrochemical 407 immunosensor of HT-2 mycotoxin. More precisely, the residual carboxylic groups in ERGO revealed 408 by XPS facilitated the covalent attachment of anti-HT2 (10) Fab via carbodiimide coupling. The 409 presence of analyte in solution was reported by anti-immunocomplex scFv antibody labelled with 410 alkaline phosphatase which produced electroactive 1-naphthol from non-electroactive substrate. The 20 411 proposed biosensor reached biologically relevant analytical parameters and satisfactory recovery in 412 biological matrix. 413 Acknowledgements 414 We acknowledge financial support from EU Graphene Flagship Core 2 Project (No. 785219). This work 415 is also funded by the CERCA Program/Generalitat de Catalunya. ICN2 acknowledges the support of the 416 Spanish MINECO for the Project MAT2017-87202-P and through the Severo Ochoa Centers of 417 Excellence Program under Grant SEV2201320295. Financial support from CEITEC 2020 (LQ1601) is 418 acknowledged. The AFM measurements (GO characterization) performed at Nanobiotechnology Core 419 Facility has been financially supported by Ministry of Education, Youth and Sports of the Czech 420 Republic under the CIISB research infrastructure project LM2015043. 421 The authors wish to express their thanks to Marcos Rosado Iglesias, Elena Del Corro García, Guillaume 422 Sauthier (all Catalan Institute of Nanoscience and Nanotechnology, CSIC and The Barcelona Institute 423 of Science and Technology, Barcelona, Spain) and Jan Přibyl for their expert assistance with SEM, 424 Raman spectroscopy, XPS and AFM analysis, respectively. 425 References 426 Arola, H.O., Tullila, A., Kiljunen, H., Campbell, K., Siitari, H., Nevanen, T.K., 2016. Specific Noncompetitive Immunoassay for HT-2 Mycotoxin Detection. Anal. Chem. 88(4), 427 2446-2452. 428 Arola, H.O., Tullila, A., Nathanail, A.V., Nevanen, T.K., 2017. A Simple and Specific Noncompetitive ELISA Method for HT-2 Toxin Detection. Toxins 9(4). 429 Bachmann, B., Adly, N., Schnitker, J., Yakushenko, A., Rinklin, P., Offenhausser, A., Wolfrum, B., 2017. All-inkjet-printed gold microelectrode arrays for extracellular recording 430 of action potentials. Flex. Print. Electron. 2(3). 431 Bartlam, C., Morsch, S., Heard, K.W.J., Quayle, P., Yeates, S.G., Vijayaraghavan, A., 2018. Nanoscale infrared identification and mapping of chemical functional groups on 432 graphene. Carbon 139, 317-324. 433 Benchirouf, A., Muller, C., Kanoun, O., 2016. Electromechanical Behavior of Chemically Reduced Graphene Oxide and Multi-walled Carbon Nanotube Hybrid Material. 434 Nanoscale Res. Lett. 11. 435 Bonaccorso, F., Bartolotta, A., Coleman, J.N., Backes, C., 2016. 2D-Crystal-Based Functional Inks. Adv. Mater. 28(29), 6136-6166. 436 Brownson, D.A.C., Munro, L.J., Kampouris, D.K., Banks, C.E., 2011. Electrochemistry of graphene: not such a beneficial electrode material? RSC Adv. 1(6), 978-988. 437 Bryla, M., Waskiewicz, A., Ksieniewicz-Wozniak, E., Szymczyk, K., Jedrzejczak, R., 2018. Modified Fusarium Mycotoxins in Cereals and Their Products-Metabolism, 438 Occurrence, and Toxicity: An Updated Review. Molecules 23(4). 439 Cammarano, A., De Luca, G., Amendola, E., 2013. Surface modification and adhesion improvement of polyester films. Cent. Eur. J. Chem 11(1), 35-45. 440 Casero, E., Alonso, C., Petit-Dominguez, M.D., Vazquez, L., Parra-Alfambra, A.M., Merino, P., Alvarez-Garcia, S., de Andres, A., Suarez, E., Pariente, F., Lorenzo, E., 2014. 441 Lactate biosensor based on a bionanocomposite composed of titanium oxide nanoparticles, photocatalytically reduced graphene, and lactate oxidase. Microchim. Acta 181(1-2), 442 79-87. 443 Casero, E., Alonso, C., Vazquez, L., Petit-Dominguez, M.D., Parra-Alfambra, A.M., de la Fuente, M., Merino, P., Alvarez-Garcia, S., de Andres, A., Pariente, F., Lorenzo, E., 444 2013. Comparative Response of Biosensing Platforms Based on Synthesized Graphene Oxide and Electrochemically Reduced Graphene. Electroanalysis 25(1), 154-165. 445 da Silva, E., Miserere, S., Kubota, L.T., Merkoci, A., 2014. Simple On-Plastic/Paper Inkjet-Printed Solid-State Ag/AgCI Pseudoreference Electrode. Anal. Chem. 86(21), 10531- 446 10534. 447 da Silva, E., Souto, D.E.P., Barragan, J.T.C., Giarola, J.D., de Moraes, A.C.M., Kubota, L.T., 2017. Electrochemical Biosensors in Point-of-Care Devices: Recent Advances and 448 Future Trends. ChemElectroChem 4(4), 778-794. 449 Dearden, A.L., Smith, P.J., Shin, D.Y., Reis, N., Derby, B., O'Brien, P., 2005. A low Curing temperature silver ink for use in ink-jet printing and subsequent production of 450 conductive tracks. Macromol. Rapid Commun. 26(4), 315-318. 451 EC, 2013. Commission Recommendation 2013/165/EU of 27 March 2013 on the presence of T-2 and HT-2 toxin in cereals and cereal products. Off. J. Eur. Commun.(91), 12- 452 15. 453 Feng, X.Y., Chen, W.F., Yan, L.F., 2016. Electrochemical reduction of bulk graphene oxide materials. RSC Adv. 6(83), 80106-80113. 454 Gonzalez, E., Barankin, M.D., Guschl, P.C., Hicks, R.F., 2008. Remote Atmospheric-Pressure Plasma Activation of the Surfaces of Polyethylene Terephthalate and Polyethylene 455 Naphthalate. Langmuir 24(21), 12636-12643. 456 He, P., Derby, B., 2017. Inkjet printing ultra-large graphene oxide flakes. 2D Mater. 4(2). 457 Hermerschmidt, F., Burmeister, D., Ligorio, G., Pozov, S.M., Ward, R., Choulis, S.A., List-Kratochvil, E.J.W., 2018. Truly Low Temperature Sintering of Printed Copper Ink 458 Using Formic Acid. Adv. Mater. Technol. 3(12). 459 Hoeng, F., Bras, J., Gicquel, E., Krosnicki, G., Denneulin, A., 2017. Inkjet printing of nanocellulose-silver ink onto nanocellulose coated cardboard. RSC Adv. 7(25), 15372- 460 15381. 461 Hussein, H.S., Brasel, J.M., 2001. Toxicity, metabolism, and impact of mycotoxins on humans and animals. Toxicology 167(2), 101-134. 462 Chen, Y., Zhang, X.O., Zhang, D.C., Yu, P., Ma, Y.W., 2011. High performance supercapacitors based on reduced graphene oxide in aqueous and ionic liquid electrolytes. Carbon 463 49(2), 573-580. 464 Juang, Z.Y., Wu, C.Y., Lu, A.Y., Su, C.Y., Leou, K.C., Chen, F.R., Tsai, C.H., 2010. Graphene synthesis by chemical vapor deposition and transfer by a roll-to-roll process. 465 Carbon 48(11), 3169-3174. 466 Kamyshny, A., Magdassi, S., 2014. Conductive Nanomaterials for Printed Electronics. Small 10(17), 3515-3535. 467 Karim, N., Afroj, S., Malandraki, A., Butterworth, S., Beach, C., Rigout, M., Novoselov, K.S., Casson, A.J., Yeates, S.G., 2017. All inkjet-printed graphene-based conductive 468 patterns for wearable e-textile applications. J. Mater. Chem. C 5(44), 11640-11648. 469 Kaukoranta, T., Hietaniemi, V., Ramo, S., Koivisto, T., Parikka, P., 2019. Contrasting responses of T-2, HT-2 and DON mycotoxins and Fusarium species in oat to climate, 470 weather, tillage and cereal intensity. Eur. J. Plant Pathol. 155(1), 93-110. 471 Khan, I.M., Zhao, S., Niazi, S., Mohsin, A., Shoaib, M., Duan, N., Wu, S.J., Wang, Z.P., 2018. Silver nanoclusters based FRET aptasensor for sensitive and selective fluorescent 472 detection of T-2 toxin. Sens. Actuator B-Chem. 277, 328-335. 473 Kuang, B., Song, W.L., Ning, M.Q., Li, J.B., Zhao, Z.J., Guo, D.Y., Cao, M.S., Jin, H.B., 2018. Chemical reduction dependent dielectric properties and dielectric loss mechanism 474 of reduced graphene oxide. Carbon 127, 209-217. 475 Kudr, J., Richtera, L., Nejdl, L., Xhaxhiu, K., Vitek, P., Rutkay-Nedecky, B., Hynek, D., Kopel, P., Adam, V., Kizek, R., 2016. Improved Electrochemical Detection of Zinc Ions 476 Using Electrode Modified with Electrochemically Reduced Graphene Oxide. Materials 9(1). 477 Kuchenbuch, H.S., Becker, S., Schulz, M., Cramer, B., Humpf, H.U., 2018. Thermal stability of T-2 and HT-2 toxins during biscuit- and crunchy muesli-making and roasting. 478 Food Addit. Contam. Part A-Chem. 35(11), 2158-2167. 479 Li, P.W., Tao, C.A., Wang, B.Y., Huang, J., Li, T.H., Wang, J.F., 2018. Preparation of Graphene Oxide-Based Ink for Inkjet Printing. J. Nanosci. Nanotechnol. 18(1), 713-718. 480 Lippolis, V., Pascale, M., Maragos, C.M., Visconti, A., 2008. Improvement of detection sensitivity of T-2 and HT-2 toxins using different fluorescent labeling reagents by high- 481 performance liquid chromatography. Talanta 74(5), 1476-1483. 482 Medina-Sanchez, M., Martinez-Domingo, C., Ramon, E., Merkoci, A., 2014. An Inkjet-Printed Field-Effect Transistor for Label-Free Biosensing. Adv. Funct. Mater. 24(40), 483 6291-6302. 484 Meneely, J.P., Sulyok, M., Baumgartner, S., Krska, R., Elliott, C.T., 2010. A rapid optical immunoassay for the screening of T-2 and HT-2 toxin in cereals and maize-based baby 485 food. Talanta 81(1-2), 630-636. 486 Michel, M., Biswas, C., Kaul, A.B., 2017. High-performance ink-jet printed graphene resistors formed with environmentally-friendly surfactant-free inks for extreme thermal 487 environments. Appl. Mater. Today 6, 16-21. 488 Paredes, J.I., Villar-Rodil, S., Solis-Fernandez, P., Martinez-Alonso, A., Tascon, J.M.D., 2009. Atomic Force and Scanning Tunneling Microscopy Imaging of Graphene 489 Nanosheets Derived from Graphite Oxide. Langmuir 25(10), 5957-5968. 490 Park, B.K., Kim, D., Jeong, S., Moon, J., Kim, J.S., 2007. Direct writing of copper conductive patterns by ink-jet printing. Thin Solid Films 515(19), 7706-7711. 491 Pei, L.M., Li, Y.F., 2017. Rapid and efficient intense pulsed light reduction of graphene oxide inks for flexible printed electronics. RSC Adv. 7(81), 51711-51720. 492 Raj, M.A., John, S.A., 2013. Fabrication of Electrochemically Reduced Graphene Oxide Films on Glassy Carbon Electrode by Self-Assembly Method and Their Electrocatalytic 493 Application. J. Phys. Chem. C 117(8), 4326-4335. 494 Ramesha, G.K., Sampath, S., 2009. Electrochemical Reduction of Oriented Graphene Oxide Films: An in Situ Raman Spectroelectrochemical Study. J. Phys. Chem. C 113(19), 495 7985-7989. 496 Ren, X., Ma, H.M., Zhang, T., Zhang, Y., Yan, T., Du, B., Wei, Q., 2017. Sulfur-Doped Graphene-Based Immunological Biosensing Platform for Multianalysis of Cancer 497 Biomarkers. ACS applied materials & interfaces 9(43), 37637-37644. 498 Rocha, O., Ansari, K., Doohan, F.M., 2005. Effects of trichothecene mycotoxins on eukaryotic cells: A review. Food Addit. Contam. Part A-Chem. 22(4), 369-378. 499 Shiwaku, R., Matsui, H., Nagamine, K., Uematsu, M., Mano, T., Maruyama, Y., Nomura, A., Tsuchiya, K., Hayasaka, K., Takeda, Y., Fukuda, T., Kumaki, D., Tokito, S., 2018. 500 A Printed Organic Circuit System for Wearable Amperometric Electrochemical Sensors. Sci Rep 8. 501 Singh, M., Haverinen, H.M., Dhagat, P., Jabbour, G.E., 2010. Inkjet Printing-Process and Its Applications. Adv. Mater. 22(6), 673-685. 22 502 Stankovich, S., Dikin, D.A., Dommett, G.H.B., Kohlhaas, K.M., Zimney, E.J., Stach, E.A., Piner, R.D., Nguyen, S.T., Ruoff, R.S., 2006. Graphene-based composite materials. 503 Nature 442(7100), 282-286. 504 Stringer, J., Derby, B., 2009. Limits to feature size and resolution in inkjet printing. J. Eur. Ceram. Soc. 29(5), 913-918. 505 Su, C.H., Chiu, H.L., Chen, Y.C., Yesilmen, M., Schulz, F., Ketelsen, B., Vossmeyer, T., Liao, Y.C., 2019. Highly Responsive PEG/Gold Nanoparticle Thin-Film Humidity 506 Sensor via Inkjet Printing Technology. Langmuir 35(9), 3256-3264. 507 Tan, S.M., Ambrosi, A., Chua, C.K., Pumera, M., 2014. Electron transfer properties of chemically reduced graphene materials with different oxygen contents. J. Mater. Chem. A 508 2(27), 10668-10675. 509 Toh, S.Y., Loh, K.S., Kamarudin, S.K., Daud, W.R.W., 2014. Graphene production via electrochemical reduction of graphene oxide: Synthesis and characterisation. Chem. Eng. 510 J. 251, 422-434. 511 Vaseem, M., Lee, K.M., Hong, A.R., Hahn, Y.B., 2012. Inkjet Printed Fractal-Connected Electrodes with Silver Nanoparticle Ink. ACS applied materials & interfaces 4(6), 3300- 512 3307. 513 Vashist, S.K., Luppa, P.B., Yeo, L.Y., Ozcan, A., Luong, J.H.T., 2015. Emerging Technologies for Next-Generation Point-of-Care Testing. Trends Biotechnol. 33(11), 692-705. 514 Yang, L.C., Tu, D., Zhao, Z.Y., Cui, J., 2017. Cytotoxicity and apoptosis induced by mixed mycotoxins (T-2 and HT-2 toxin) on primary hepatocytes of broilers in vitro. Toxicon 515 129, 1-10. 516 Ye, S.B., Feng, J.C., 2016. The effect of sonication treatment of graphene oxide on the mechanical properties of the assembled films. RSC Adv. 6(46), 39681-39687. 517 Yoshizawa, T., Kohno, H., Ikeda, K., Shinoda, T., Yokohama, H., Morita, K., Kusada, O., Kobayashi, Y., 2004. A practical method for measuring deoxynivalenol, nivalenol and 518 T-2+HT-2 toxin in foods by an enzyme-linked immunosorbent assay using monoclonal antibodies. Biosci. Biotechnol. Biochem. 68(10), 2076-2085. 519 Zhan, Z.Y., An, J.N., Wei, Y.F., Tran, V.T., Du, H.J., 2017. Inkjet-printed optoelectronics. Nanoscale 9(3), 965-993. 520 Zhang, J.L., Yang, H.J., Shen, G.X., Cheng, P., Zhang, J.Y., Guo, S.W., 2010. Reduction of graphene oxide via L-ascorbic acid. Chem. Commun. 46(7), 1112-1114. 521 Zhang, T., Ma, N., Ali, A., Wei, Q., Wu, D., Ren, X., 2018a. Electrochemical ultrasensitive detection of cardiac troponin I using covalent organic frameworks for signal 522 amplification. Biosens. Bioelectron. 119, 176-181. 523 Zhang, T., Xing, B., Han, Q.Z., Lei, Y.F., Wu, D., Ren, X., Wei, Q., 2018b. Electrochemical immunosensor for ochratoxin A detection based on Au octahedron plasmonic 524 colloidosomes. Anal. Chim. Acta 1032, 114-121. 525 Zhong, J.H., Zhang, J., Jin, X., Liu, J.Y., Li, Q.Y., Li, M.H., Cai, W.W., Wu, D.Y., Zhan, D.P., Ren, B., 2014. Quantitative Correlation between Defect Density and Heterogeneous 526 Electron Transfer Rate of Single Layer Graphene. J. Am. Chem. Soc. 136(47), 16609-16617. 527 Zikulnig, J., Hirschl, C., Rauter, L., Krivec, M., Lammer, H., Riemelmoser, F., Roshanghias, A., 2019. Inkjet printing and characterisation of a resistive temperature sensor on 528 paper substrate. Flex. Print. Electron. 4(1). 529 530