A Mobile Analytical Device for On-Site Quantitation of Anthocyanins in Fruit Beverages

Anthocyanins are antioxidant and anti-inflammatory ingredients in various fruit beverages, for which their conservation and quantitation are important for the food industry. In this paper, we report a simple, portable device for accurate on-site determination of total monomeric anthocyanins in fruit beverages employing a Wi-Fi scanner coupled with a flexible microchip and a free mobile app. The detection principle is based on the pH-induced colorimetric reactions of anthocyanins performed in a specially designed microchip and validated with standard spectrophotometric measurements. The microchip with multiple testing vials was prepared with the benchtop molding method with a common PDMS elastomer and a transparency film; the photo of the scanned microchip is wirelessly sent to a smartphone and the RGB values of individual reaction vials in the microchip are analyzed with a free mobile app to determine the corresponding concentrations. It was demonstrated that the quantitation performance of this POCT device is comparable with conventional spectrophotometry in the determination of total anthocyanins in both standard solutions and fruit beverages.

S-2 Scheme S1. Detailed procedure to prepare the PDMS microchip.
Firstly, 18 stainless steel cylinders (5 mm in diameter and 3 mm thick) were placed in a petri dish (16 mm apart from each other in two rows), and "immobilized" by attaching disc magnets (5 mm in diameter and 3 mm thick) underneath the petri dish. Subsequently, the petri dish was filled with PDMS precursors; particularly the silicone elastomer base and the curing agent were mixed at 10:1 ratio with 4 % (w) alumina powder added. The mold was placed in a vacuum chamber for degassing for one hour and then heated at 80 C in an oven for 4 h. Once the PDMS microchip was removed from the "mold", each of the 18 mini-vial was punched to have two inlets for sample injection using a flat-end needle (1 mm in diameter, as shown in the main text, Scheme 1B).

S-3
Figure S1. Reversibility test of the pH dependent structural change of anthocyanidins.
UV-Vis absorption spectra of the "fermented Rosso" sample upon changing the pH from 1.0 to 4.5 and vice versa. The numbers in brackets on the legend indicates the number of cycles. The result is in agreement with previous reports; as example, see Tang Table S1. Three replicate UV-Vis measurements (absorbance at 520 nm) of Cy3G standard solutions.
The average absorbance (Avg) and standard deviation (SD) were determined accordingly and shown in Figure 3a of the main text.    Figure 4 of the main text. (C) ∆R, ∆G, and ∆B (differences between the values at pH 1.0 and pH 4.5) vs. the Cy3G concentration, where the solid lines show the best linear fits to the experimental data. All data points are color-coded correspondingly, i.e., red for R, blue for B, and green for G. The linearity of these fits is not comparable with that of Figure 5B (∆R/RGB vs. [Cy3G]; R 2 = 0.999).
[Cy3G] (g/mL) For each trial, Cy3G standard solutions of different concentration were adjusted to two different pH values, followed by scanning with the portable scanner, sending photos to a smartphone, and analyzing using the Color Grab app. Once the RGB values for each mini-vial were obtained from the app (Figure 5a), a calibration curve to correlate R/RGB and the concentration of Cy3G was then established (Figure 5b); these replicates were to determine the standard deviations shown as error bars (see main text for details).  (a) Picture of the fermented wine and raw juice samples provided by the Bayou Brewing Club, which is a sub-division of Laca Biotech Inc., and (b) the natural colors of each beverage as received. Among the nine sample beverages, the "raw" ones are the original grape juice; the "fermented" ones are the wines made from their corresponding raw juices. We can see that the raw juices are generally darker than the fermented ones, indicative of the higher abundance of total anthocyanidins.

Figure S5. Three replicate experiments for determining the concentrations of Cy3G in different beverage samples with the microchip device.
According to the established calibration curve with standard Cy3G solutions ( Figure 5), the concentration of Cy3G in each sample tested were determined; the standard deviations were obtained from these three repeated trials, as presented in Figure 6 (see main text for details).