CETvol87 DOI: 10.3303/CET2187079 Paper Received: 17 August 2020; Revised: 24 January 2021; Accepted: 3 May 2021 Please cite this article as: Iannone A., Sapone V., Di Paola L., Cicci A., Bravi M., 2021, Extraction of Anthocyanins from Grape ( Vitis Vinifera ) Skins Employing Natural Deep Eutectic Solvents (naDES), Chemical Engineering Transactions, 87, 469-474 DOI:10.3303/CET2187079 CHEMICAL ENGINEERING TRANSACTIONS VOL. 87, 2021 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Laura Piazza, Mauro Moresi, Francesco Donsì Copyright © 2021, AIDIC Servizi S.r.l. ISBN 978-88-95608-85-3; ISSN 2283-9216 Extraction of Anthocyanins from Grape (Vitis vinifera) Skins Employing Natural Deep Eutectic Solvents (NaDES) Annalaura Iannonea,*, Vittoria Saponea, Luisa Di Paolab, Agnese Ciccic, Marco Bravia aDipartimento di Ingegneria Chimica,Materiali e Ambiente, via Eudossiana, 00184, Roma bFaculty of Engineering, Università CAMPUS BioMedico, Via A. del Portillo, 21, 00128 Roma, Italy c Bio-P s.r.l., via di Vannina, 88, 00156, Roma annalaura.iannone@uniroma1.it In the last years, the possibility to exploit the high added value bioactive compounds obtained by the solid residues of the winemaking process in cosmetic, food supply and pharmaceutical field was explored. To enhance the biocompatibility of the final products and reduce the environmental impact of traditional extraction routes, researchers attempted at developing sustainable and green extraction processes. In the present work, the feasibility of employing NaDES (Natural Deep Eutectic Solvents) as a cheap and non-toxic alternative to traditional solvents for the extraction of bioactive compounds from winery wastes of Merlot marc was studied. Furthermore, a two-level factorial design was used to study the single and combined effect of process variables, namely, solid to liquid ratio dry grinding time and extraction time, on the extraction yield. Experimental show that, in the range studied, solid to liquid ratio has a predominant effect and both grinding and extraction times have a slightly appreciable effect on the extraction yield. 1. Introduction Grapes (Vitis vinifera) are among the most cultivated and consumed fruits in the world, especially in the wine industry. Winemaking leads to the production of a very large amount of waste, in particular pulp, skins and shoots. The conservation and disposal of residues represents a significant cost for wine industries. Winemaking wastes are generally disposed of or, alternatively, used for composting. In order to limit the environmental impact of the wastes and the disposal costs, wastes were employed for production of heat, exploiting their high lignocellulosic fraction, or destined to the field of bio-fertilization and animal feed (Rani et al., 2020). In the last years, several studies reported that the solid residues of the winemaking process are particularly rich in high commercial value bioactive compounds, such as phenolic compounds (stilbenes, phenolic acids, flavonoids), pigments, tannins and alcohols (Teixeira et al., 2013; Chen et al., 2018; Šuković et al., 2020). Anthocyanins are the most common and valuable pigments in red grapes and are extensively studied in the field of natural dyes, in order to replace synthetic dyes that are often toxic to humans. Among these, enocyanin is coloured extract of an intense red obtained from grape pomace, which is produced as described by Mazza and Miniati (1993). The most common extraction solvents are hydroalcoholic, such as mixtures with different titers of ethanol or methanol (Librán et al., 2013; Makris et al., 2007). These methods, however, require large amounts of solvent and long extraction times (Wang and Weller, 2006). To reduce extraction times, the use of microwave and / or ultrasound systems was combined with traditional methods to facilitate the release of compounds in the plant matrix (Beres et al., 2017). Other unconventional approaches involve the use of sub- and supercritical CO2, ASE (accelerated solvent extraction), pressurized hot water, overheated hydroalcoholic mixtures (Schieber, 2017). Although many of these techniques allow obtaining high extraction yields, in recent years, sustainable and greener extraction techniques were studied. Researchers attempted at developing extraction processes that would allow an appropriate exploitation of raw materials and optimal consumption of solvents, minimizing energy consumption. 469 The use of NaDESs (Natural Deep Eutectic Solvents) as an alternative to traditional solvents for the extraction of bioactive compounds from agro-industrial wastes was considered in agreement with many of the principles of Green Chemistry (Cicci et al., 2019; Smith et al., 2014). NaDESs have unique physicochemical properties, are easily adaptable to process requirements and are generally composed of low cost, readily available and non-toxic compounds. Among these, there are quaternary ammonium salts (such as choline chloride) with neutral donors of naturally derived hydrogen bonds, such as vitamins, amines, carboxylic acids, alcohols and sugars (Dai et al., 2013). In different studies, NaDES composed of choline chloride, proline, betaine as hydrogen bond acceptors and organic acids (malic, oxalic, citric acids), sugars (glucose, fructose, xylose) and polyols (glycerol) as hydrogen bond donors, in opportune molar ratios, were employed for the extraction of bioactive compounds from wine industry wastes (Bosiljkov et al. 2017; Radošević et al., 2016; Panić et al. 2019). In our previous work (Sapone et al., 2020), choline chloride and L (+) tartaric acid based NaDES was employed for the extraction of phenolic species from Merlot marc. In the present contribution, as a further step forward, a preliminary study of the effect, single and combined, of operating parameters such as solid to liquid ratio, dry grinding time and extraction time on the anthocyanins extraction yield was carried out. To assess the single and combined effect of process parameters, a two-level factorial design was used. 2. Materials and methods 2.1 Chemicals Choline chloride, L (+) tartaric acid, DPPH (2,2-Diphenyl-1-picrylhydrazyl), methanol (purity ≥ 99.5 %) and hydrochloric acid (37 w/v %) were purchased from Merck (Darmstadt, Germany). E163 enocyanin (50 % w/w) was made available as a courtesy by Ruffini Aromi. Demineralized water was used for the experiments. All reagents were used without further purification. 2.2 Biomass preparation Merlot cultivar grape pomace was supplied by the Oenological Institute of Conegliano (Treviso, Veneto) after a macerative vinification. Pomace was pressed, spread out in a thin layer, and stabilised by employing 1 % v/v sulphuric acid and potassium metabisulphite, portioned, vacuum-packed and frozen until use. Before the extraction, the skins were manually separated from the pomace and ground in a coffee grinder. 2.3 NaDES preparation NaDES was obtained by mixing choline chloride and tartaric acid in 3:2 molar ratio inside a jacketed glass vessel at 90 ± 0.1 ° C, until a clear and homogeneous mixture was obtained. The mixture was subsequently cooled down to room temperature and 40 % (w/w %) water was added to obtain a lower solvent viscosity. The NaDES obtained was stored at room temperature in a closed bottle. 2.4 Extraction procedure Preliminarily to the actual extraction, a two-stage total extraction was carried out to assess the maximum anthocyanin content, according to the procedure described by Bonfigli et al. (2017), with slight modifications. Briefly, a small amount of grape skins was ground for 30 s and treated with 70 % methanol-water at pH 0.7 (obtained by adding opportune volume of hydrochloric acid), with a solid to liquid ratio of 0.03 g/mL at 45 °C, under magnetic stirring, for 24 hours. The extraction stirring rate was 250 rpm. Subsequently, the solid was separated from the liquid by filtration and the solid was treated with the same volume of fresh solvent in the same conditions of the first stage for 1 hour. The NaDES extraction of anthocyanins was carried out in 100 mL glass flasks in the solid to liquid ratios 0.13 - 0.44 g/mL, dry grinding times 6 - 24 s, for extraction times of 120 - 390 min, in a horizontal stirrer (170 rpm) at 25°C. The solid-liquid separation was achieved by centrifugation (1165 g) for 10 minutes. The extracts obtained were stored at 4 °C for further analyses. 2.5 Analytical methods The moisture content of the biomass was measured gravimetrically by oven drying the biomass at 105°C to constant weight. The total anthocyanin content was determined spectrophotometrically (Shanghai Mapada Spectrophotometer UV- 1800 PC) by reading the absorbance at 528 nm against a calibration curve obtained with standard solutions of enocyanin. The extraction yield was obtained dividing the anthocyanin content of each sample by the equivalent enocyanin content obtained with the two-stage total extraction, expressed as weight of enocyanin equivalents on dry basis (DB). The antioxidant activity was evaluated by the DPPH assay (Kedare and Singh, 2011) and the results were expressed as trolox equivalents antioxidant capacity (TEAC) on dry basis, using a calibration curve obtained with standard solutions of trolox (6-hydroxy-2,5,7,8- tetramethylchroman-2-carboxylic acid). 470 2.6 Factorial design For the preliminary investigation of the effects of solid to liquid ratio (X1), dry grinding time (X2) and extraction time (X3) on the extraction yield, a two-level factorial design was employed. The factorial design was composed of 8 points and 6 replicated points to ensure accuracy in the estimation of the experimental error. The selection of factor levels, which are reported in Table 1 in both uncoded and coded values, was carried out according to preliminary experiments. Coded values were obtained as reported in Eq(1), where xi represents the coded value of i th variable, Xi is the uncoded value of the i th factor, Xi,0 is the value of the i th factor at the central level and ΔXi is the step size. The experimental design layout is shown in Table 2. The design of experiments and statistical analysis was performed by using Design Expert 7.0 software® (Version 7.1.6, Stat-Ease Inc., Minneapolis, MN, USA). (1) Table 1: Actual and coded levels of the experimental design factors Factor Unity Levels -1 0 1 Solid to liquid ratio (X1) g/mL 0.13 0.29 0.44 Grinding time (X2) Extraction time (X3) s min 6 121 15 255 24 388 Table 2: Experimental design layout, observed response (Yield) and antioxidant activity of the extracts Run X1 X2 X3 Yield (%) TEAC (mg TE/g DB) 1 0.44 24 388 25.67 3.57 2 3 4 5 6 7 8 9 10 11 12 13 14 0.29 0.44 0.13 0.29 0.13 0.13 0.29 0.44 0.29 0.13 0.44 0.29 0.29 15 6 6 15 24 24 15 24 15 6 6 15 15 255 388 121 255 121 388 255 121 255 388 121 255 255 32.16 25.15 42.81 35.14 45.54 53.56 34.96 25.34 34.10 48.33 21.07 34.17 35.80 3.92 3.31 4.20 4.09 5.36 6.08 4.7 3.05 4.76 5.33 2.39 5.06 4.41 3. Results and discussion 3.1 Extract characterization The moisture content measured for the grape skins was 64.4 %. Spectrophotometric analyses for each extract show an intensity peak the wavelength band comprised between 500 and 550 nm, suggesting the presence of a anthocyanins’ residue after maceration. The UV-Vis spectrum of the central point is reported in Figure 1 and shows an intensity peak at 520 nm. The antioxidant activities of the extracts are reported in Table 2. The results show that the antioxidant activity increases according to the extraction yield. 3.2 Extraction modelling The results in terms of extraction yield and antioxidant activity are reported in Table 2. (2) 𝑦𝑦 = 𝛽𝛽0 + �𝛽𝛽𝑖𝑖𝑥𝑥𝑖𝑖 3 𝑖𝑖=1 + � � 𝛽𝛽𝑖𝑖𝑖𝑖 𝑥𝑥𝑖𝑖𝑥𝑥𝑖𝑖 3 𝑖𝑖 =𝑖𝑖+1 3 𝑖𝑖=1 471 The influence of the three factors, namely, solid to liquid ratio (g/mL), grinding time (s) and extraction time (min) was evaluated by means of the factorial design previously described. The model was represented by a second-order polynomial equation, Eq(2), where βi are the coefficients associated to the three main effects, βij are those associated to the cross interactions, xi are the coded independent variables and y is the process response variable. Figure 1: UV-Vis spectrum of the extract obtained employing choline-chloride Figure 2: Parity plot In Figure 2, the parity plot is reported. It can be observed that the mathematical model provides a good approximation of the experimental data, being the points uniformly distributed around the bisector. Furthermore, values of coefficient of determination, R 2 (0.9838), adjusted-R 2 (0.9766), prediction-R 2 (0.9508) witness a good fit of data. The low value of the coefficient of variation (4.13 %) ensured a high reproducibility of the experiments. The lack of fit was not significant and there were no violations of basic analysis of variance. In Table 3, the model βi coefficients, standard errors, F-values and p-values are reported. The three factors were all statistically significant and influenced the extraction yield. The predominant effect on the response variable was exerted by the solid to liquid ratio, as it can be observed by the high negative value of β1 coefficient. In fact, as the solid to liquid ratio increases, the extraction yield decreases. This factor is also involved in a statistically significant interaction with extraction time, as it can also be observed in Figure 3. 472 Table 3: Regression coefficients (βi) of the polynomial model (Eq.2) with their standard errors (SE), F-values (F) and p-values (p). Coeff Term Value Standard Error F-value p-value β0 β1 β2 β3 β13 Intercept X1 X2 X3 X1 X3 +35.27 -11.36 +1.59 +2.24 -1.14 0,47 0,47 0,47 0,47 0,47 432.47 8.13 16.11 4.17 <0.0001 0.0191 0.0030 0.716 Figure 3: 3d plot showing the effect of solid to liquid ratio (g/mL) and extraction time (min) on the extraction yield. Several studies employing factorial designs indicated extraction time as a significant factor in a solid-liquid extraction of bioactive compounds from plant biomass (Lavecchia et al., 2015; Kaur et al., 2008). However, in the present study, the effect of extraction time exerts a slightly positive effect on the extraction yield, according to the positive, but low value of coefficient β2. This behaviour could be attributed to the time range used in the present study, where shorter extraction times were excluded. According to the value of β3 coefficient, dry grinding time exerts a rather negligible effect on the extraction yield. This behaviour could be due to the conditions of spent grape skins, that already underwent a maceration and the previously described downstream operations. Therefore, further mechanical operations might have not substantially affected the biomass. The results in terms of extraction yield obtained are apparently comparable to those obtained with hydroalcoholic solvents (75 % EtOH-H2O v/v) by Librán et al. (2013). However, they are far lower than those obtained in the study of Bosiljkov et al. (2017), with a choline chloride – malic acid based NaDES. This behaviour could be attributed to a higher solubility of anthocyanins and to the different density of the solvent due to the use of malic acid, instead of tartaric acid. Furthermore, the authors employed wine lees and an ultra-sound equipment to enhance the extraction process. 4. Conclusions Wastes from the wine industry are still very rich in bioactive compounds and can represent a promising raw material for cosmetic, nutraceutical and food products. The use of NaDES could represent an effective alternative to traditional solvents for the extraction process, as they can be formulated employing components of natural origin, compatible with the final use of the bioactive compounds. In the present work, a NaDES based on choline chloride and citric acid was used to recover anthocyanins from red grape pomace, obtaining yields up to 54%. A two-level factorial design was employed to evaluate the effects of three factors, namely, solid to liquid ratio, biomass dry grinding time and extraction time. A second order polynomial equation provided a good fit of the experimental data. Liquid to solid ratio appeared to be the factor exerting the highest effect on the extraction yield. 473 The results in terms of anthocyanin content and antioxidant activity of the extracts, encourage further studies to expand extraction time range and explore the combination of the use of NaDES with pretreatments, such as ultrasound, microwaves and enzymatic processes, in order to fully explore the potential of this approach. Acknowledgments This work was partly supported by the WINGED project (WINe GrEen Distillery, POR FESR Lazio 2014-2020. Progetto T0002E0001 - A0206 Avviso Pubblico "Circular Economy e Energia". DE n. G17436/2018). References Beres C., Costa G., Cabezudo I., da Silva-James N., Teles A., Cruz A., Mellinger-Silva C., Tonon R., Cabral L., Freitas S., 2017, Towards integral utilization of grape pomace from winemaking process: A review, Waste Management, 68, 581–594. Bonfgli M., Godoy E.; Reinheimer M. A.; Scenna N. J., 2017, Comparison between conventional and ultrasound-assisted techniques for extraction of anthocyanins from grape pomace. Experimental results and mathematical modelling, Journal of Food Engineering, 207, 56–72. Bosiljkov T., Dujmić F., Cvjetko Bubalo M., Hribar J., Vidrih R., Brnčić M., Zlatic E., Radojčić Redovniković I., Jokić S., 2017, Natural deep eutectic solvents and ultrasound-assisted extraction: Green approaches for extraction of wine lees anthocyanins, Food and Bioproducts Processing, 102, 195–203. Chen Q., Diao L., Song H., Zhu X., 2018, Vitis amurensis Rupr: A review of chemistry and pharmacology, Phytomedicine, 49, 111–122. Cicci A., Bravi M., 2019, Leveraging novel green solvents to drive conceptual and practical biorefinery innovation, Chapter In: Basile A., Centi G., De Falco M., Iaquaniello G. (Eds.), Studies in Surface Science and Catalysis. Elsevier. Dai Y., van Spronsen J., Witkamp G.-J., Verpoorte R., Choi Y.H., 2013, Natural deep eutectic solvents as new potential mediafor green technology, Analytica Chimica Acta, 766, 61–68. Kaur, D., Wani, A.A., Oberoi, D.P.S., Sogi, D.S., 2008. Effect of extraction conditions on lycopene extractions from tomato processing waste skin using response surface methodology, Food Chemistry, 108, 711–718. Kedare S.B., Singh R.P., 2011, Genesis and development of DPPH method of antioxidant assay, Journal of Food Science and Technology, 48, 412–422. Lavecchia R., Medici F., Piga L., Zuorro A., 2015. Factorial design analysis of the recovery of flavonoids from bilberry fruit by-products, International Journal of Applied Engineering Research, 10, 43555–43559. Librán, M.; Mayor, C.; Garcia-Castello, L.M.; Vidal-Brotons, E. 2013, Polyphenol extraction from grape wastes: Solvent and pH effect, Agricultural Sciences, 04, 56–62. Makris D., Boskou, G., Andrikopoulos, N. 2007, Polyphenolic content and in vitro antioxidant characteristics of wine industry and other agri-food solid waste extracts, Journal of Food Composition and Analysis, 20, 125–132. Mazza G., Miniati, E. (Eds), 1993, Anthocyanins in Fruits, Vegetables, and Grains (1st ed.). CRC Press. Panić M., Gunjević V., Cravotto G., Radojčić Redovnikovića I., 2019, Enabling technologies for the extraction of grape-pomace anthocyanins using natural deep eutectic solvents in up-to-half-litre batches extraction of grape-pomace anthocyanins using NADES, Food Chemistry, 300. Radošević K., Ćurko N., Gaurina Srče, V., Cvjetko Bubalo M., Tomašević M., Kovačević Ganić K., Radojčić Redovniković I., 2016, Natural deep eutectic solvents as beneficial extractants for enhancement of plant extracts bioactivity, LWT - Food Science and Technology, 73, 45–51. Rani, J., Indrajeet, Rautela, A., Kumar, S., 2019. Biovalorization of winery industry waste to produce value- added products, Chapter In: Rathinam, N.K., Sani, R. (Eds.), Biovalorisation of Wastes to Renewable Chemicals and Biofuels. Elsevier. Sapone V., Cicci A., Franceschi D., Vincenzi S., Bravi M., 2020, Antioxidant Extraction and Bioactivity Preservation from Winery By-products by Natural Deep Eutectic Solvents (NaDES), Chemical Engineering Transactions, 79, 157-162. Schieber A., 2017, Side Streams of Plant Food Processing As a Source of Valuable Compounds: Selected Examples, Annual Review of Food Science and Technology, 8, 97–112. Smith E., Abbott A., Ryder K., 2014, Deep Eutectic Solvents (DESs) and Their Applications, Chemical Reviews, 114, 11060–11082. Šuković D., Knežević B., Gašić U., Sredojević M., Ćirić I., Todić S., Mutić J., Tešić Ž., 2020, Phenolic profiles of leaves, grapes and wine of grapevine variety vranac (Vitis vinifera L.) from Montenegro, Foods, 9. Teixeira A., Eiras-Dias J., Castellarin S.D., Gerós H., 2013, Berry phenolics of grape-vine under challenging environments, International Journal of Molecular Sciences, 14, 18711–18739. 474