Commercial synthetic hectorite Laponite RD (monovalent cation exchange capacity CEC=0.74 mmol/g) was obtained from Laportes Industries (Luton, UK). Horseradish peroxidase lyophilized powder (EC 1.11.1.7, 193 U/mg), citrate buffer-stabilized spherical gold nanoparticles (particle size (5±1) nm), hydroquinone, caffeine, chlorogenic, ascorbic, citric, glutamic, tartaric acids and tetrachloroauric(III) acid trihydrate, 20% (m/V) in water polydiallyldimethylammonium chloride, sodium hydroxide, sodium carbonate, and Folin-Ciocalteu reagent were purchased from Sigma-Aldrich S.A, Merck (St. Louis, MO, USA). Multiwall carbon nanotubes (outer diameter (30±15) nm, length 1-5 μm) were purchased from NanoLab Inc (Waltham, MA, USA). Sulfuric, nitric and tetrachloride acetic acids, 2,2'-dipyridyl and iron(III) chloride were from Biopack (Buenos Aires, Argentina). Phosphoric acid and absolute ethanol were obtained from Dorwill (Buenos Aires, Argentina). Reagent grade buffer phosphate sodium salts NaH₂PO₄ and Na₂HPO₄, hydrogen peroxide (30%) and glucose were obtained from J.T.Baker (Phillipsburg, NJ, USA). Water-soluble polystyrene copolymers of vinylbenzylthymine (VBT) and vinylbenzyl triethylammonium chloride (VBA) ({[(VBT)(VBA)₄]⁴⁺}₂₅) or VBT and vinylphenyl sulphonate (VPS), ({[(VBT)(VPS)₁₆]^16-}₂₅) were synthesized as described before (33, 34). In all cases, the solutions were prepared using triple distilled water.

Eggplants (Solanum melongena L. cv. Black Nite), carrots (Daucus carota L. cv. Chantenay) and green leafy vegetables rocket (Eruca Sativa Mill.), lettuce (Lactuca sativa cv. Capitata) and chard (Beta vulgaris cv. Cicla) were purchased from a local producer (Santiago del Estero, Argentina). After harvest, the products were stored at (4±1) °C (Briket Master 5000; Buenos Aires, Argentina) until used. Before extract preparation, vegetables were washed in chlorate water and air dried.

The PPhC determination by electrochemical and spectrophotometric methods was carried out with extracts obtained by homogenizing fresh eggplants and carrots (1 and 5 g, respectively) with 20 mL of methanol.

The extracts for AA electrochemical determination were prepared using 8 g of fresh eggplants and 12 mL of 0.1 M phosphate buffer solution, pH=2, and 5 g of green leafy vegetables and 10 mL of the same buffer solution. For spectrophotometric determination, eggplant and green leafy vegetables were prepared following the same procedure but using 6% (m/V) trichloroacetic acid (TCA) as solvent.

In all cases, after homogenizing the vegetable tissues with the solvent, samples were filtered (Whatman^® grade 589/2, Ф=100 mm) and centrifuged at 10 000×g for 15 min at 4 °C (Ependorf 5430; Thermo Fisher, Madrid, Spain).

Fresh-cut eggplants were exposed to different UV-C light intensities for 300 s, using a home-made reflective stainless steel chamber (1.70 m×1.00 m×0.10 m) containing 12 fluorescent germicidal lamps (254.7 nm, TUV 36W/G36; Philips, Amsterdam, The Netherlands) distributed equally in the upper and lower part of the steel chamber (35). Samples were divided into three groups according to the UV-C light intensity treatment: (i) untreated used as control, (ii) 1.0, and (iii) 10.0 kJ/m². Radiation intensity was monitored with a portable digital radiometer (Cole-Parmer Instrument Company, Vernon Hills, IL, USA). Samples were packed in polypropylene (PP) trays (17.4 cm×13.8 cm×4.8 cm; Cellpack SA, Santa Fe, Argentina), heat-sealed with 35 μm thick PP film (O₂ permeability: 2.58·10^-6 mol/(m²·s·Pa) and CO₂ permeability: 9.30·10^-6 mol/(m²·s·Pa) at 25 °C and 90% RH (data provided by National Institute of Industrial Technology, Santiago del Estero, Argentina), and stored at (4±1) °C in the fridge (Briket Master 5000). The vegetable extracts were prepared as described in Extract preparation, and the changes in the PPhC and AA contents were tested by electrochemical and spectrophotometric assays on days 0, 2, 4 and 6 of storage, considering that fresh-cut fruits and vegetables are available in the market for a maximum of 6-7 days.

Measurements were carried out at room temperature in a three-electrode electrochemical cell with 0.1 M phosphate buffer as supporting electrolyte. A platinum wire, an Ag|AgCl|Cl^- (3 M) and different modified glassy carbon electrodes (GCE; CH Instruments, Bee Cave, TX, USA) were used as counter, reference, and working electrodes, respectively. Amperometry was performed with a potentiostat/galvanostat (model Teq4; nanoTeq, Buenos Aires, Argentina) under convective conditions, and the working potential was chosen in order to reach stationary state conditions.

For PPhC determination, the GCE surface was modified with a hydrogel composed of Laponite RD, citrate buffer-stabilized spherical gold nanoparticles, copolymers of vinylbenzylthymine (VBT) and vinylbenzyl triethylammonium chloride (VBA) ({[(VBT)(VBA)₄]⁴⁺}₂₅), and horseradish peroxidase enzyme (HRP), as described in our previous work (31). PPhC determination is based on the cyclic enzymatic reaction of HRP. The enzyme is oxidized by the action of hydrogen peroxide. The oxidized enzyme takes electrons from the PPhC, which are ultimately reduced at the electrode surface at -200 mV (working potential), resulting in a reduction current proportional to the PPhC concentration. The high specificity towards PPhC and catalytic nature of the immobilized HRP allows obtaining an amplified electrical signal proportional to analyte concentration, without the interference of ascorbic, tartaric or citric acids (31), potential interferences in food analysis.

On the other hand, functionalized multiwall carbon nanotubes (fMWCNT), poly(diallyldimethylammonium chloride) capped gold nanoparticles (AuNP@PDDA, particle size (12±2) nm), and copolymers of vinylbenzylthymine (VBT) and vinylphenylsulphonate (VPS) ({[(VBT)(VPS)₁₆]^16-}₂₅) were used to modify the GCE surface. A working potential of 50 mV was applied to determine the AA mass fraction. Although in this case a biorecognition element was not immobilized on the working electrode, the inclusion of nanomaterials favours complete oxidation of AA at a low positive potential with a high sensitivity (32).

Mass fractions of both analytes were determined by the standard addition method using the extract supernatant. Before the sample addition, two successive aliquots of a standard solution were aggregated to verify proportionality between the current signal and the concentration (31, 32). To evaluate the possible food matrix effects, the apparent recovery (R_A) of the signal produced by a third aliquot of the standard solution was measured (36). The standard solutions used to carry out the determination of PPhC and AA were chlorogenic acid (ChlA) and AA, respectively.

Spectrophotometric method

A calibration curve was performed according to Singleton and Rossi (22). Different volumes of ChlA standard solution were mixed with 50 μL Folin-Ciocalteu reagent (1 M). Next, 100 μL Na₂CO₃ (20%, m/V) in 0.1 M NaOH were added, and after sonication in darkness at 25 °C for 60 min, the absorbance was measured at 765 nm in a UV-Vis spectrophotometer (model 8453; Agilent Technologies, Santa Clara, CA, USA). The PPhC concentration was determined according to the sample absorbance and the Lambert-Beer law.

The procedure described by Kampfenkel (23) was used to carry out the calibration plot for AA determination. Different concentrations of an AA standard solution were mixed with 0.2 M phosphate buffer at pH=7, 1 mL of 10% (m/V) TCA, 800 μL of 42% (V/V) H₃PO₄, 800 μL of 4% 2,2'-dipyridyl (in 70% V/V ethanol), and 400 μL of 3% (m/V) FeCl₃ were added. The mixture was incubated at 42 °C for 45 min, and the absorbance at 525 nm was measured (model 8453; Agilent). Using the Lambert-Beer law, the AA concentration was obtained by reading the absorbance intensity. Results for PPhC were expressed as ChlA equivalents in mg per kg of fresh mass and for AA in mg per 100 g of fresh mass.

We selected these methods because they are widely used for vegetable matrices, and are validated by the Association of Official Analytical Chemists (AOAC) (37).

Data analyses were performed with OriginPro v. 2019^TM software (38). All experiments were performed using the Statgraphics Centurion XV v. 15.2.06 (39), and the results were represented as mean value±standard deviation (S.D.). The mean values of specific differences were analyzed by least significant difference (LSD) test, followed by the analysis of variance (ANOVA). The significance level of 0.05 was considered.

The enzymatic biosensor was used to determine the PPhC content in fresh-cut eggplant and carrot extracts. Fig. 1a shows the bioelectrode current response profiles at working potential of -200 mV as a function of time in a cell containing 600 μM H₂O₂ in 0.1 M phosphate buffer at pH=5 (supporting electrolyte), after the addition of 0.85 μM ChlA standard solution and 10 and 20 μL fresh-cut eggplant and carrot samples prepared in triplicate, respectively, as indicated in the figure.

Before the quantification of the PPhC mass fraction in the fresh-cut eggplant and carrot extracts, the effects of the apparent resistivity (R_A) on the possible interferences in the matrix were analyzed, which was calculated with the following equation (36):

where x(S+I) is the measured response for the sample spiked with interference, x(S) is the measured value for the unspiked sample, and x(I) is the value of interference. Matrix effect is related to the possible interference of any component of the sample other than the analyte. The R_A expressed as percentage was 93 and 90% for eggplant and carrot extracts respectively. These R_A values indicate that any interference from any component of the sample (AA, for example) could be negligible.

Thereby, the electrochemical determination of the PPhC content was carried out in triplicate, obtaining the average values of (761±15) and (53±3) mg/kg for eggplant and carrot extracts, respectively.

To compare the results obtained with the electrochemical method, PPhC determination was also carried out with conventional spectrophotometric Folin-Ciocalteu method which measures the total antioxidant activity (22). Measurements were performed using 350 μL of the vegetable extracts, and absorbance was monitored in the 190-1100 nm range. The maximum absorbance at 765 nm allowed to determine the PPhC mass fraction (N=3) on fresh mass basis (782±33) and (55±4) mg/kg in eggplant and carrot extracts, respectively.

Fig. 1b shows the PPhC mass fraction obtained with electrochemical and spectrophotometric methods. Results showed that the determination of the PPhC content carried out by both methods was not significantly different, with a confidence level of 95%, even when working with different fruit and vegetable matrices. The convenience of the electrochemical biosensor to evaluate the PPhC content was demonstrated even when using a smaller sample volume.

Fig. 2a shows the current time profiles for 1 µM AA standard solution and 90 and 50 µL eggplant and green leafy vegetable samples, respectively, prepared in triplicate. Measurements (N=3) were performed in 0.1 M phosphate buffer (pH=7) at 50 mV as working potential. The AA quantification was carried out in a similar way to that described for PPhC using the standard addition method, and evaluating the R_A. The AA mass fraction on fresh mass basis was obtained as follows: (5.52±0.06), (290.00±0.06), (38.00±0.04) and (6.20±0.05) mg/100 g in eggplant, rocket, lettuce and chard, respectively. Furthermore, the apparent recoveries were 90% (eggplant), 92% (rocket), and 91% (lettuce and chard), indicating that any component in the matrix of these fruit and vegetables, other than the AA, does not significantly interfere with the signal.

The obtained values (Fig. 2b) were in agreement with those obtained with the spectrophotometric method described by Kampfenkel (23). AA spectrophotometric determination was carried out in triplicate; 200 μL fresh-cut eggplant and green leafy vegetable extracts were used and the absorbance intensity was monitored at 525 nm. The obtained values for AA mass fraction on fresh mass basis were: (5.8±0.9), (305±1), (41.0±0.9), and (6.7±0.7) mg/100 g for eggplant, rocket, lettuce and chard, respectively. As can be seen, there were no statistically significant differences between the results obtained by both methods, in a confidence interval of 95%. Therefore, the results showed the ability of the electrochemical sensor to determine the mass fraction of AA.

To verify the versatility of the electrochemical method to detect small changes in the AA and PPhC contents, experiments similar to those described in the previous sections were performed. For this purpose, eggplants were previously irradiated with different UV-C light intensities, stored for different time periods, and prepared in triplicate according to the descriptions in Extract preparation (40).