Acid Hydrothermal Amendment of Grape Wine Pomace: Enhancement of Phenol and Carbohydrate Co-solubilization

The byproduct from the winery industry has many bioactive compounds which are considered high value-added compounds. In addition, white grape pomace (GP) is rich in carbohydrates, which consequently can be used as raw material for subsequent bioprocesses. The present study carried out low-temperature hydrothermal treatments using different operational conditions: temperature (65–95°C), operational time (120–240 min), sulfuric acid concentration addition (5–15% v/v). The results show that using 65°C, 120 minutes and 15% (v/v) of H 2 SO 4 it was possible to obtain a liquid phase rich in phenols and carbohydrates. Reaching a total of phenol compounds of 2113 ± 140 mg gallic acid/kg GP, composed mainly of 3-hydroxytyrosol (1330 ± 22 mg/kg GP). On the other hand, the carbohydrate solubilization reached 245 g glucose/kg GP. These results demonstrate the potential of hydrothermally treated grape pomace as raw material for biore�nery processes.


Introduction
Grape is one of the most widely cultivated plant crops worldwide [1], It is worth noting that most of the grapes produced globally are destined for wine production, with a global production of more than 27 million tons in 2019 (FAO. 2022). This huge volume is of strong economic importance, reaching an export value of more than US$35 billion in 2019 (FAO. 2022). Among the main producers, Spain, France and China stand out as the countries with the largest vineyards in area, followed to a lesser extent by Italy, Turkey and the US [2].
The moderate consumption of grape wine has been traditionally associated with some health bene ts. Although the alcohol content makes it impossible to consider wine a healthy beverage, many bioactive compounds with antioxidant, antimicrobial and/or anti-in ammatory properties have been determined in grape wine, such as phenolic acids, avonoids or proanthocyanins (Gerardi et al., 2021;Peixoto et al., 2018). These high addedvalue compounds in the grape are also retained in a byproduct generated during grape wine preparation and so-called grape pomace (GP), which accounts for 17-25% of the initial grape mass [6]. GP is mainly composed of the retained grape seeds and skins, although it can also contain pulp remains and stem residues [3,7]. Indeed, GP can content a polyphenol concentration between 1 and 5% in dry matter [3,8]. Likewise, white GP is an outstanding source of sugars, with a content between 55 and 78% in dry matter of the GP [9]. The high sugar content in the white GP makes it a promising biomass as feedstock for different valorization bioprocesses, such as bioethanol production [10], anaerobic digestion [11], or polymer production [12,13]. However, the presence of bioactive compounds can limit the implementation of bioprocesses for valorizing the GP due to its inhibitory properties [11,14]. Therefore, the recovery of the bioactive compounds from the GP would carry a double bene t, i.e., the recovery of addedvalue compounds with a high market interest, and, at the same time, the detoxi cation of the remaining biomass, facilitating its further valorization through bioprocesses [7].
The removal or recovery of bioactive compounds, mainly phenolic acids, has been evaluated through many different methods for GP and similar substrates. For example, Siles et. [15] proposed a short ozonation pre-treatment (15 min), which allowed a more than 50% reduction of the initial concentration of phenols in vinasses from alcohol production. Other authors proposed the extraction of the phenolic compounds using different solvent extractions, such as acetone [16], methanol/water extraction [12,17], or ethanol acidi ed with HCl (Filippi et al., 2022). Several physical methods have been also proposed, such as nano ltration membranes, or ultrasound-and microwave-assisted extraction [19,20]. Indeed, the use of nano ltration membranes allowed the recovery of up to 73% and 92% of the total phenolic content and proanthocyanidins in GP, respectively [19]. However, many authors have stated the need to implement optimized processes that maximize the extraction of the bioactive compounds [7,21]. Indeed, the grape seeds concentrate most of the phenolic compounds in the GP, advocating for its solubilization to achieve a higher bioactive compound recovery [2,22]. Hydrothermal treatments have been widely proposed as a cost-effective option for the solubilization of lignocellulosic substrates [7,23]. However, the application of excessively severe conditions during the hydrothermal treatment would entail the degradation of a fraction of the bioactive compounds and sugars in the substrate, as well as the appearance of undesirable compounds, such as furfural or hydroxymethylfurfural (HMF) [24]. These compounds have been reported to be biological inhibitors and would struggle the further valorization of the substrate after the recovery of the bioactive compounds [25]. As an alternative, low-temperature hydrothermal treatments have been shown to promote protein solubilization, to increase particulate carbohydrate removal, and increase the soluble phenol and soluble sugar contents [26,27] . Therefore, the present research aims to optimize the solubilization of bioactive compounds and carbohydrates from GP through a low-temperature hydrothermal treatment. For this, different treatment condition ranges were assayed, i.e., operational temperature (65-95ºC), treatment duration (120-240 min), and the addition of acid (5-15% v/v H 2 SO 4 ). This paper is especially relevant given that it focuses on a widely generated substrate, GP, proposing a valorization method that aims to maximize the recovery of bioactive compounds and, at the same time, to amend the treated substrate as a suitable feedstock for further bioprocesses by optimizing the soluble sugar content.

Substrate
The white grape pomace (GP) was obtained from the "Viña Aynco" vineyard located in Galvarino, Region of La Araucanía, Chile. Once collected, the GP was immediately stored at -20°C to avoid the uncontrolled fermentation of the substrate.

Hydrothermal treatment
The experiments were carried out in 1-liter glass reactors. In all tests, the GP/acid solution ratio of 15/85 (w/w) was used, 50g of GP (wet base) and 275g of acid solution were added, stirring was kept constant at 300rpm. The reaction temperature was evaluated in a range from 65 to 95°C, the reaction time between 120 and 240 minutes and the acid concentration from 5 to 15% (v/v). The imposed conditions during the hydrothermal treatment were organized with surface response methodology using the Design Expert 11 software. A face-centered central composite design of three factors, with six central points, and three replicates per experiment were applied ( Table 1). The responses studied were the solubilization of the total carbohydrates (considered as a sum of xylose, glucose, and arabinose) and phenolic compounds. ANOVA analyses were performed for each response. In all cases, the model signi cance and lack-of-t tests were performed to check that models were relevant and t the experimental data. All statistical analyses were performed considering α = 0.05 (Supplementary material, Table S1). After each hydrothermal treatment, the samples were centrifuged for 10 minutes at 10,000 rpm, the solid phase was characterized according to lignin, cellulose, total phenolic compound content, volatile and total solids. The liquid phase was ltered using 0.22µm pirinola lters. total sugars, xylose, arabinose, glucose, furfural, HMF, soluble chemical oxygen demand (sCOD), total phenolic compounds and their characterization were measured in the ltered sample.

Chemical analysis
Different parameters were characterized for untreated grape pomace and treated grape pomace, including analysis for liquid and solid phase obtained after hydrothermal treatment. Total solids (TS), volatile solids (VS), and soluble chemical oxygen demand (sCOD) were measured according to standard methods (30). In addition, lignin content was measured for OMSW, using the Klason lignin method (TAPPI T 222 om-02) [28]. Particularly, the acetic acid, furfural and hydroxymethylfurfural (HMF) and the individual carbohydrates such as xylose, glucose, and arabinose obtained in the liquid phase after hydrothermal treatment were quanti ed with a YL9100 high-performance liquid chromatograph (HPLC) equipped with a Bio-Rad Aminex HPX-87H 300 column. For carbohydrates and organic acids a refractive index detector (RID) was used, while furfural and HMF were detected with a diode array detector (DAD) [29]. Total phenol compounds in the liquid and untreated GP (UGP) were quanti ed by spectrophotometry with a gallic acid (GA) calibration curve, using the Folin-Ciocalteu method, expressing the results as mg of GA/Kg GP. For the determination of total phenol content in UGP, previous extraction was carried out using 1 g of OMSW and 1 mL of methanol/water (80:20 v/v). The mixture was stirred for 1 min in a vortex apparatus and centrifuged at 1200 G for 10 min. The methanol layer was separated, and the extraction was repeated four times [30]. To determine the antioxidant capacity in the liquid phase, the antiradical activity (2,2-diphenyl-1-picrylhydrazyl (DPPH) was measured, the DPPH analysis was done measuring the variation in absorbance at 515 nm after 30 min of reaction with the radical DPPH, using Thermo Scienti c's UV-Vis Orion AquaMate 8000 spectrophotometer. The results were expressed in grams of Trolox equivalent per kilogram of grape pomace. Finally, the characterization of the individual phenolic compounds was performed according to what was described previously in Romero-Roman et al. [31] by high performance liquid chromatography-diode array detection electrospray ionization/mass spectrometry (HPLC-DAD-ESI/MSn) using an Agilent Technologies 1220 In nity Liquid Chromatograph equipped with an autoinjector (G1313, Agilent Technologies, Santa Clara, CA, USA) coupled with a diode array detector (1260, Agilent Technologies, Santa Clara, CA, USA) and a Luna 5-µm C18, 100-Å column (250-4.6 mm). The analytical procedure was based on calibration curves using the following standards: p-hydroxybenzoic acid, gallic acid hexoside, protocatechuic acid hexoside, vanillic acid, coutaric acid, caffeic acid, caftaric acid, p-coumaric acid, ferulic acid, chlorogenic acid, ellagic acid, 3,4-dihydroxyphenyl glycol, 3-hydroxytyrosol, p-tyrosol, vanillin. 4-methyl catechol, syringol, 3,4-dimethyl benzyl alcohol, catechin, epicatechin, quercetin-3-rutinoside, quercetin-3-glucoside, quercetin-3-glucuronide, myricetin and quercetin. All were expressed in mg/ Kg GP. Table 2 show the characterization of the solid phases obtained after treatment of grape pomace and for untreated grape pomace. As shown in Table 2, most of the solid phases obtained after the centrifugation of the hydrothermally treated GP presented total solid concentrations higher than 100 g/kg ( Table 2). By contrast, treatments 1, 2, 6, and 11 resulted only in solid phases with total solid concentrations between 60.9 and 82.2 g/kg. These treatments were the hydrothermal treatments with a lower concentration of acid addition ( Table 1). The lower total solid concentration observed in these treatments could be due to higher acid addition and could favor the separation of the solid and liquid phases during the centrifugation step. The enhanced separation with the higher acid addition might be related to a higher degradation of the hemicellulose, the solubilization of which is favored by acidic conditions [32]. The content of lignin and holocellulose in the untreated GP and after each treatment is shown in Table 2. It is observed that for all the treatments applied, the lignin content was similar with untreated grape pomace. The maximum lignin solubilization was only 16.1% (w/w), determined after treatment 6, i.e., decrease from 35.6 to 29.5% (w/w). The low lignin removal was expected due to the hydrolysis conditions applied, where in all cases the working temperature was not higher than 95°C. Some authors have reported that the compact structure of the lignin can require temperatures above 180°C, or even the combination of high temperature and fast decompression processes, such as in steam-explosion technology, to effectively break down the lignin bers [27]. By contrast, the holocellulose content was strongly impacted by the different applied treatments, in all cases the holocellulose solubilization was greater than 62% (w/w). Speci cally, treatment 8 showed the greatest decrease in holocellulose content, starting from 59.6 for the UGP up to 4.7% (w/w), followed by treatment 12 and 7 in which the nal content was 9.9 and 9.5% (w/w) ( Table 2). These results were also corroborated by a confocal laser scanning microscopy analysis (Supplementary material, Figure S1). Table 3 shows the characterization in the liquid phase for UGP and treated GP after different conditions. Particularly, the sCOD in all treatments were higher than the value determined for UGP, i.e., 594 mg/kg of sCOD, reaching values even higher than 900 mg/kg for treatments 3, 4, 7, 8 y 12. These treatments corresponded to the higher doses of sulfuric acid, i.e., 15% w/w of H 2 SO 4 , indicating a positive relation between the acid addition and the solubilization of organic matter. On the other hand, secondary compounds generated during the hydrothermal treatments, such as furfural, hydroxymethylfurfural (HMF) and acetic acid are also reported in Table 3. Furfural and HMF were generated in low quantities reaching a maximum of 0.4 ± 0.01 and 0.05 ± 0.005 mg/kg, respectively (Table 3). These values were markedly lower than the inhibition limit described for some bioprocesses such as anaerobic digestion, where concentrations of around 2.0 g/L of furfural and HMF are required for the full inhibition of microbial activity [33].

Solubilization of organic matter
The low concentrations of furfural and HMF in the present research would be expected since their generation is clearly related to the severity of the applied treatment, especially at temperatures above 120ºC (Cubero-Cardoso et al., 2020).  140 mg GA/kg GP, around 7 times higher than UGP, i.e., 292 ± 25 mg GA/kg GP (Fig. 1). Figure 2 presents the effect of the studied factors (temperature, time and acid addition) on the studied responses. When the temperature was maintained at 65°C, increasing both the acid addition and the reaction time resulted in higher total soluble phenol concentrations ( Fig. 2.a). By contrast, when the temperature was 95°C, the treatment duration had a low impact on the solubilization of phenols (Fig. 2.b). When the acid concentration was 15% (v/v), it was possible to solubilize similar concentrations of phenols, reducing the operational time from 240 to 120 minutes, as long as the temperature increases from 65 to 95°C (Fig. 2.c). The higher release of phenolic compounds due to the acid addition would be related to the enhancement of the solubilization, as previously described for the degradation of the hemicelluloses [32]. Similarly, (Rodríguez-Gutiérrez et al., 2019) reported a 100% enhancement by adding ethanol during the hydrothermal treatment of strawberry extrudate at 90ºC compared to the same hydrothermal treatment without acid addition. The operational temperature had a strong in uence on the solubilization of phenols (Fig. 2.c and 2.d).
Indeed, for a reaction time of 240 minutes and a sulfuric acid concentration of 15% (v/v), the temperature was not a relevant factor in the assessed range (Fig. 2.c). Conversely, for an acid concentration of 5% (v/v), the phenols increased from 554 ± 42 to 1621 ± 102 mg GA/kg GP as the operational temperature increased from 65 and 95°C respectively. A similar behavior was described by (Cubero-Cardoso et al., 2020), which reported an increase of around 53% in the total phenol content in the liquid fraction obtained after the hydrothermal treatment of raspberry extrudate by increasing the operational temperature from 60 and 90ºC due to the release of the phenolic compounds retained in the lignocellulosic structures.
The high concentration of total soluble phenols was directly related to the values of antioxidant power observed in each hydrothermal treatment.
Although the variations in the values of the antioxidant power were lower than for the concentration of total soluble phenols, a positive relation between the acid addition and the increase in the antioxidant power was observed (Fig. 3). In all treatments, the antioxidant power was higher than 17 ± 2 g eq. Trolox/kg (848 ± 1 mg eq. Trolox/L). Indeed, maximum antioxidant power was determined in run 4 (65ºC, 240 min, 15% (v/v) H 2 SO 4 ). i.e., 27 ± 2 g eq. Trolox/kg (1330 ± 96 mg eq. Trolox/L). As a comparison, the liquid fraction obtained after the hydrothermal treatment of the GP in the present study achieved more than double antioxidant power than that reported for the hydrothermally treated raspberry extrudate at 90ºC for 180 min, i.e., 167 ± 1 mg eq. Trolox/L. A high antioxidant power has been reported to be bene cial for the use of the GP extract as the basis for the preparation of UV lters [2,36].

Characterization of individual bioactive compounds
The concentration and composition of the individual phenol compounds in the liquid phases strongly varied in comparison with the UGP due to the hydrothermal treatment conditions ( Table 4). The concentration of several phenolic acids and phenolic derivatives increased, whereas the effect on the concentration of avonols was more limited. Particularly, the highest increase in phenolic acids was determined for the gallic acid hexoside and phydroxybenzoic acid. The concentrations of these increased more than one order of magnitude compared to the UGP ( (v/v) H 2 SO 4 , 95ºC) for p-hydroxybenzoic acid. The increase in the concentration of these compounds can be very attractive due to their properties, e.g.
gallic acid has been shown to exhibit antioxidant, antiviral, antibacterial, antifungal properties and is a positive modulator of germination [37,38]. Quercetin-3glucuronide 65ºC), while the highest concentration of DHPG was 264 ± 5 using treatment 7 (240 min, 15% (v/v) H 2 SO 4 , 95°C). The high concentration of HT and DHPG reported in the present study is relevant due their pharmacological, anti-in ammatory and antioxidant capacities [39,40]. These interesting properties have led to wide use in industry, resulting in a high market price, i.e., €520/kg of 10% phenol extracts [23,41]. The concentrations of HT observed in the present study were markedly higher than average contents reported for other well-known sources of HT such as olives, olive oil and wine, i.e., 629, 5 and 2 mg/kg, respectively (Gallardo-Fernández et al 2022).

Carbohydrates Solubilization
The solubilization of the carbohydrates was evaluated through the measurement of the total soluble carbohydrate concentration, as well as the content of the main individual carbohydrates, i.e., glucose, xylose, and arabinose. As shown in Fig. 4, the total soluble carbohydrate concentration was widely increased several times compared to the UGP after each hydrothermal treatment, regardless of the xed operational conditions. After the hydrothermal treatments, most of the sugars were displaced to the liquid fraction in all cases. i.e., more than 18 times of the initial concentration in the UGP, due to their high solubility [23]. This increase was in line with the reduction in holocellulose observed in the solid phases obtained after the hydrothermal treatments (Table 2), since the degradation of these bers releases simpler soluble carbohydrates [42].
The operational condition with the greatest impact on the total soluble carbohydrate concentrations was the acid addition, achieving the highest values at 15% v/v of H 2 SO 4 , i.e., treatments 3, 4, 7, 8, and 12 (Fig. 4). Indeed, the mean values obtained with the acid addition of 5%, 10%, and 15% v/v of H 2 SO 4 were 173 ± 12, 185 ± 15, and 245 ± 9 g glucose/kg GP, respectively. To a lesser extent, the duration of the hydrothermal treatment was also positively related to an increase in the total soluble carbohydrate, obtaining the maximum concentrations at 240 min. i.e., treatment number 4 and 8 ( Fig. 4). In addition, as can be seen in Fig. 4, the effect of the studied conditions on the sum of the main individual carbohydrates (glucose, xylose and arabinose) saw a similar trend to the one described for total soluble carbohydrates. Figure 5 shows the effect of the studied conditions on the individual carbohydrate solubilization as the sum of glucose, xylose, and arabinose. Speci cally, Fig. 5.a shows that when the acid concentration was 5% (v/v), and the temperature increased from 65 to 95°C, the solubilization of individual carbohydrates increased more than 100%, regardless of the reaction time. However, when the reaction temperature was highest (95°C), the increase in the acid concentration resulted in a 33% increase in solubilization (Fig. 5.d). Note that the acid effect can be in uenced by the working temperature, whereas the reaction time did not show as signi cant an effect (Supplementary material, Table S1). The low in uence of the operational temperature at lower acid concentrations on the individual carbohydrates compounds may be related to the selected temperature range, the values of which are, in every case, lower than that required for an effective solubilization of part of the bers [27]. Indeed, some authors have reported that temperatures above 150ºC-180ºC would be necessary for the solubilization of the hemicelluloses and, minimally, the lignin [43]. The high e ciency observed by increasing the acid addition could be related to an enhancement in the degradation of the hemicellulose, which can be effectively converted into soluble sugars by dilute acid hydrolysis [32]. This result was in line with the report by [35] for the hydrothermal treatment of strawberry extrudate.
Those authors reported an enhancement in the total carbohydrate concentration from 40 g/kg of fresh raw material to 65 g/kg of fresh raw material by adding 0.5% glacial acetic acid at a hydrothermal treatment at 90ºC and 90 min.
The in uence of the operational conditions was also observed on the composition and concentration of the individual carbohydrate compounds (Fig. 6). Thus, Fig. 6 shows that the sum of the concentrations of glucose, xylose and arabinose represented more than 89% of the total carbohydrate concentrations in all hydrothermal treatments. Among them, glucose was the main carbohydrate, representing a percentage of the total carbohydrates between 55% and 74%. Glucose has been previously de ned as the major carbohydrate in white grape pomace [44], being very attractive as feedstock for subsequent fermentation processes [44,45]. In this sense, according to Table 3, it is also interesting considering that microorganism inhibition will not be an issue since the hydrolysate has a poor concentration of inhibitory compound of metabolic activity such as furfural and HMF [33]. Therefore, the selected range of conditions for the hydrothermal treatment of GP would allow for the solubilization of carbohydrates without accumulating undesirable furans that limit the further valorization of the treated biomass.
The concentrations of xylose and arabinose increased when the severity of the treatment conditions increased (Fig. 6). For the hydrothermal treatments at 65ºC, the variation in the H 2 SO 4 addition from 5 to 15% v/v resulted in average increases of around 669% and 45% for xylose and arabinose, respectively, for both hydrothermal treatment durations. Similarly. the increase in the operational temperature of the treatment led to an increase in the xylose concentration, although it had little impact on the arabinose concentrations (Fig. 6). The duration of the hydrothermal treatment also showed a poor relation with the variations in the concentration of both xylose and arabinose. The higher in uence of the addition of H 2 SO 4 on the solubilization of xylose and arabinose is related to the capacity of the dilute acid hydrolysis to degrade the hemicelluloses [32]. Both xylose and arabinose are sugar monomers generated by breaking down hemicelluloses [44,46]. These compounds are also very promising as substrates for biore nery processes, such as obtaining succinic acid [46]. The marked increase in the concentration of carbohydrates compared to the UGP through the proposed acid-assisted hydrothermal treatments will enable the effective amendment of this substrate as feedstock for different fermentation processes [47][48][49][50]. Also, the treatment conditions are varied to adjust the individual carbohydrate pro le to the process requirements.

Conclusions
Proposed treatments allowed the co-solubilization of phenols and carbohydrates, reaching around 90% of holocellulose solubilization. It was observed that the concentration of the acid addition had the strongest effect on the solubilization of both phenols and carbohydrates. Nevertheless, the increase in temperature leads to a lower solubilization of phenols, while the carbohydrate solubilization remains constant. Speci cally, hydroxytyrosol, p-tyrosol and DHPG were the main solubilized phenols. Glucose and xylose were the main carbohydrates, reaching up to 245 g glucose/kg GP. These results demonstrate the potential of the grape pomace as raw material for biore nery processes by applying a suitable amendment.
Declarations Figure 1 Total soluble phenolic compounds for untreated grape pomace (UGP) and treated grape pomace at different conditions.  Antioxidant for untreated grape pomace (UGP) and treated grape pomace at different hydrothermal treatments conditions.

Figure 4
Total carbohydrates and the sum of individual carbohydrates (glucose, xylose, and arabinose) for untreated grape pomace (UGP) and grape pomace after different hydrothermal treatments.