Electrochemical Platform Based on Modified Ti Electrodes to Test a Food Allergen Presence

Celiac disease (CD) is an autoimmune disorder. For these patients, the only treatment is a gluten free diet. In this work we showed that there can be an easy and cheap method to test gluten presence. We used several modified Ti electrodes. All modified electrodes were prepared using electrochemical methods. Modified electrodes were tested using differential pulse voltammetry in solutions with or without gluten. Selected modified Ti electrode was tested using different known gluten concentrations and extracts from different aliments.

electrode (18). DPV was used with reduced graphene oxide modified electrodes with antigliadin antibody. At least 4 hours are needed to prepare the modified electrode (19). A research study was using both impedance and DPV. They were using modified electrodes with antibody based on the use of dithiols (20).
A simpler method was using a graphite mine and rapid differential pulse voltammetry (5).
In the present research, we used modified Titanium (Ti) electrodes and DPV. We used different surface modifications: nanotubes obtained by anodization, nanofibers and graphene oxide. We tested different food products with the modified electrodes.
First, Ti round samples, with 8 mm diameter (99.99% purity, Alpha Aesar) were polished and cleaned with ultrasounds. Distiled water, ethanol and acetone were used (15 minutes each). Electrode was named E1.
The nanotubes (TiNT) were prepared at room temperature in a viscous electrolyte: EG: 0.5 wt% NH4F + 2 vol.% H2O at 50 V, 2 h. E1 was used as working electrode and a Pt wire as counter electrode. A Matrix MPS-7163 was used as power supply. After preparation, they were rinsed with distilled water, and subjected to ultrasonic treatment 30 seconds in distilled water. Electrode was named E2.
TiO2 nanofibers (TiNF) were prepared at high voltages using a previous described method (21) with some modifications. Briefly, electrolyte was: 4 wt.% TBut + 8 wt.% PVP in DMF/isopropanol (mass ratio 1/1) + 2 wt.% acetic acid. Prior to electrodeposition, the samples were coated with 100 μl solution using a spin-coater Laurell WS-650-23 (Laurell Technologies). The dynamic mode was used, with 3000 rpm, 30 seconds. For nanofibers formation, solution was fed at a rate of 0.5 ml/h using a Legato180 pump (Kd Scientific) from a plastic syringe with a blunt tip needle. 20 kV were applied between the needle tip and Cu collector plate (at 10 cm distance) for 30 min. A high-power supply was used -PS/EJ30P20 (Glasmann). Samples were dried overnight and treated 4 h at 450°C. E1 was used as working electrode. Resulted electrode was named E3.
Gluten presence evaluation was made using DPV between 0 V and 1.2 V, with 50 mV amplitude at 20 mV scan rate. Electrolyte used was DMSO:acetate buffer (pH 4.8) 1:4. Acetate buffer was prepared from natrium acetate and glacial acetic acid; with or without gluten.
Gluten extraction: flour samples (1g) were mixed with distilled water (10 mL) for 1 h under ultrasound stirring. After centrifugation at 8000 rpm for 10 min a pellet was formed. This pellet was immersed into ethanol (60%, 50 mL) for 1 h, centrifuged (30 min, 8000 rpm). 5 mL of the supernatant was used for analysis dissolved in 25 mL buffer pH 4.8 Equipment: scanning electron microscopy (SEM) images for all samples were recorded with Thermo Scientific FEI Quanta 650 FEG (Hillsboro, OR, USA) high-performance scanning electron microscope. All electrochemical data were recorded with a potentiostat/galvanostat from Metrohm Autolab (PGSTAT 302N). Electrochemical cell was of three electrodes in a single compartment.

Results and discussions
Recorded curves during sample preparation and their corresponding SEM images for modified electrodes are presented in Figure 1. For GO deposition on Ti using CV (30 cycles) the recorded curve is seen in Figure 1 a. It is a typical CV curve for a redox process. In the SEM image (Figure 1b) it can be seen the graphene agglomerations. For sample Ti/NT, the curve recorded during anodization is presented in Figure 1 c. It is typical anodization curve with revealing the mechanism of nanotubes formation previously described (22). In the SEM image (Figure 1d), very ordered nanotubes are visible, having diameters around 100 nm. Using high voltage process, random aligned TiO2 nanofibers were prepared (Figure 1 e). The CV curve for GO deposition on top of nanotubes (TiNTGO), the currents recorded during CV (Figure 1 f) were higher compared with the ones recorded in TiGO case (Figure 1a). It is visible that the current was increasing with each cycle. In the SEM image (Figure 1  g), it is visible that the nanotubes were covered with a film. Their diameters are smaller compared with Ti/NT sample. Most of the nanotubes remained open after GO deposition. The Ti surface modification influence is visible once again in the curve recorded during GO electrodeposition on top of nanofibers ( Figure 1h). This curve is very different compared with the ones recorded for TiGO30 sample and TiNTGO sample. GO electrodeposition on top of nanofibers is clearly seen in Figure 1 i. Nanofibers diameters were changed after GO electodeposition ( Figure 1 e and i). To see which of the proposed electrodes it is desirable to evaluate the presence of gluten, we tested six electrodes using DPV. One electrode was E1 (polished and cleaned Ti) and the others were modified Ti electrodes. From the Figure 2, Figure 3. Only five concentrations were presented. At sixth concentration, probably Ti was oxidized and the signal was not proportional with the gluten concentration anymore. Gluten concentration was plotted against peak height. We calculated the limit of detection (LOD) and the limit of quantification (LOQ) in each case, using relations described in literature (23), with Excel program. In first case we obtained LOD 36 ppm and LOQ 119 ppm. In the second case, results were: LOD 66 ppm and LOQ 220 ppm. It can be seen that increasing the cycles number during GO electrodeposition did not increased the GO film conductivity, so the results were not improved. To see if the E4 30 TiGO electrode can be used to analyze real samples, we prepared three kind of flour samples extracts. From Figure 4, it can be seen that with this modified Ti electrode, we were able to distinguish between flour samples. The higher signal was obtained for wheat flour. This is known to have gluten. A smaller signal compared with the one obtained in DMSO:acetate buffer with 66 ppm gluten was obtained for corn flour , labeled by manufacturer "may contain traces of gluten". A signal compared with the one obtained in DMSO:acetate buffer without gluten was obtained for flour labeled by manufacturer "gluten free".

Figure 4. Flour samples testing with E4 30 TiGO
To verify the reproducibility and repeatability, we freshly prepared E4 30 TiGO at different times. We recorded the signal in DMSO:acetate buffer with 100 ppm gluten. From results presented in Figure  5 a. The result is similar. We also performed five consecutive scans in DMSO:acetate buffer with 100 ppm gluten. We obtained a good reproducibility (Figure 5b).

Conclusions
Different nanostructures (nanotubes, nanochannels, nanofibers) were prepared for a potential gluten detection application. GO was also used as titanium or nanostructured titanium surface modification method. Their structure was analyzed using SEM.
Modified electrodes were tested for gluten presence evaluation using an electrochemical method (DPV). Ti modified with graphene oxide and nanofibers gave a signal in DPV in solutions containing https://doi.org/10. 37358/Rev. Chim.1949 Rev. Chim., 71 (5) known gluten concentrations. Increasing graphene oxide layer thickness (by increasing cycles number during electrodeposition), does not increased the conductivity. We tested different kinds of flour with Ti modified with graphene oxide: wheat flour (known to have gluten), corn flour (labeled "may contain traces of gluten") and a mixed fluor labeled "gluten free". We were able to distinguish between them.
Future experiments need to be performed to try to lowed LOD and LOQ, but up to this point, the proposed electrode (Ti with graphene oxide) and proposed method (DPV) can be used to detect gluten from aliments which were contaminated with gluten during manufacturing process.