Optimization of the subcritical water treatment from sunflower by-product for producing protein and sugar extracts

Vol.:(0123456789) 1 3 Biomass Conversion and Biorefinery https://doi.org/10.1007/s13399-022-02380-w ORIGINAL ARTICLE Optimization ofthesubcritical water treatment fromsunflower by‑product forproducing protein andsugar extracts GrazielleNáthia‑Neves1 · EstherAlonso1 Received: 4 November 2021 / Revised: 18 January 2022 / Accepted: 20 January 2022 © The Author(s) 2022 Abstract This study proposes the valorization of sunflower by-product (SbP) using subcritical water treatment. For this purpose, a response surface methodology in a 3 × 3 Box–Behnken design was employed to study the effect of temperature (100– 150°C), extraction time (5–25min), and solvent-to-feed ratio (10–20g bidistilled water/g SbP) on the water-soluble compounds. The highest amount of glucose (5.7-mg/g raw material) and xylose (13.5-mg/g raw material) was obtained at 150°C and solvent-to-feed ratio of 15 after 25min of extraction. The highest amount of protein (133.5-mg/g raw material) was obtained at 150°C and solvent-to-feed ratio of 20 after 15min of extraction. The selected temperature conditions did not promote the formation of final degradation products such as furfural and 5-hydroxymethylfurfural (HMF). The extracts obtained in this study showed simultaneous recovery of free monomeric sugars that could be used for bioenergy or biochemical, proteins that could be used for food and pharmaceutical industries and phenolic compounds that could be used for food and nutraceutical applications. Keywords Fermentable sugars· Response surface methodology· Protein· Amino acids 1 Introduction Sunflower cake/meal is the by-product of the edible oil industry. Currently, sunflower is produced in large quantities in the world (53.5 million tons) [1]. In this sense, large quantities of waste materials are generated from the sunflower oil extraction process. In 2019, the Ukraine and Russian federation were the main producers of sunflower by-product (SbP) (6.7 and 5.7 million tons, respectively) followed by the European Union (EU-27) with 4.7 million tons [2]. Despite being produced in large quantities, this agro-industrial residue has very limited reuse. After oil extraction, the residue of sunflower contains a high amount of protein (> 27%) with 4.37% of essential amino acids [3]. The residue resulting from the oil extraction process also contains lignin (20.4–26.62%), cellulose (32.93–39.1%), hemicelluloses (18.4–30.9%) minerals (151.35mg/kg), and polyphenols (2–5%), mainly chlorogenic acid (CGA) (which corresponds to more than 70% of all phenolic present in the SbP) [4–7]. Despite the high nutritional value and its potential uses as gelling, foaming, emulsifying, and polymer in chemical, pharmacological, and food industries [4], the SbP biomass is almost exclusively employed for animal feeding [5]. This is mainly because the use of the SbP in industrial processes has been scarcely reported in the current literature. Therefore, more studies are needed to promote the reuse of the SbP as a renewable and cheap resource in order to add value to the sunflower production chain, incorporating the biorefinery context to this biomass. The use of subcritical water has proven in recent years to be an environmentally friendly green and rapid method to obtain water-soluble compounds. This method can be used for the extraction of bioactive compounds or as a substitute to other methods that use acidic, alkaline, and enzymatic media for the hydrolysis of macromolecules (proteins into peptides and amino acids and polysaccharides to fermentable sugars) [8–10]. To achieve the subcritical state, water is heated at temperatures between its boiling point (100°C) and its critical temperature (374°C) * Grazielle Náthia-Neves graziellenathia.nev[email protected] 1 BioEcoUva, Research Institute On Bioeconomy, High Pressure Processes Group, Department ofChemical Engineering andEnvironmental Technology, University ofValladolid, Prado de la Magdalena 5, 47011Valladolid, Spain Biomass Conversion and Biorefinery 1 3 and pressure is applied in order to ensure that the water will remain in a liquid state throughout the process [11, 12]. The use of this technology has been widely used to extract sugars and phenolic and antioxidants compounds from several by-products such as asparagus [13], peach palm [14], coffee residue [15], and rice straw [16]. For the best of our knowledge, only studies involving the treatment of fresh sunflower using subcritical solvent was reported in the literature. For instance, Sutanto etal. [17] studied the production of fatty acid methyl esters (FAMEs) (yield of 88–93%) from kernels with a subcritical solvent mixture of methanol and acetic acid (200–250°C). Ravber etal. [18] evaluated the subcritical water extraction from sunflower seeds (60–160°C) for the recovery of oil and water-soluble compounds. In this study, the highest amount of oil (44%) was obtained at 130°C at a solvent/solid ratio of 20 and an extraction time of 30min. The effect of subcritical water in the hydrothermal degradation of some compounds present in SbP was also evaluated by the same group [19]. In this study, the authors observed that with temperatures higher than 220°C, no more carbohydrates were found in the extract (only inhibitors) and that the content of amino acid and amine did not change over the reaction time at temperature of 130°C. Therefore, the aim of the present work was the valorization of the SbP using subcritical water to obtain fermentable sugars (glucose and xylose), proteins, amino acids, and CGA. The liquid solution was also analyzed in terms of total phenolic content and antioxidant capacity. The solid co-product obtained after the subcritical water treatment was analyzed in terms of protein content and amino acid profile. 2 Materials andmethods 2.1 Raw material preparation andcharacterization The SbP (a mixture of kernel and shells) was supplied by a local agricultural cooperative located in Olmedo, Spain, in February 2020. The SbP was milled into small particles (particle size of 0.42 ± 0.02mm) in a commercial mill (Cueffer, B07DVN18PS, Taiwan, China). After the milling process, the SbP was characterized in terms of moisture (NREL/TP-510–42,621), ash (NREL/ TP-510–42,622), extractives (NREL/TP-510–42,619), protein (NREL/TP-510–42,625; using a conversion factor of 6.25), soluble lignin (NREL/TP-510–42,618), and insoluble lignin (NREL/TP-510–42,618) by using methodologies recommended by the National Renewable Energy Laboratory described in the technical reports [20–24]. 2.2 Sunflower by‑product extraction using subcritical water The extraction was carried out in a 145mL stainless steel reactor (with a length of 77mm, an internal diameter of 49mm, and a wall thickness of 11mm). The extractor was heated by an electric system, the temperature was measured with a thermocouple and controlled by a PID controller, and the pressure inside the reactor was measured by a manometer (0–25bar). A schematic diagram of the equipment is shown in Fig.1. For each extraction assay, an amount of SbP (accuracy of 0.001g) was loaded in the reactor, and subsequently, Fig. 1 Extraction equipment 300RPM Stainless steel reactor (145 mL) Thermocouple PID controller Manometer (0–25 bar) Magnetic stirrer (300 rpm) Magnet Mixture of SbP and bidistilled wate r Ele ctric heating system Biomass Conversion and Biorefinery 1 3 100g of the bidistilled water was added to obtain an appropriate solvent-to-feed solid ratio (S/F; g bidistilled water/g SbP). The studied values of S/F were 10 (100-g bidistilled water/10g SbP), 15 (100g bidistilled water/6.66g SbP), and 20 (100-g bidistilled water/5g SbP), which were defined by preliminary experiments. Afterward, the system was closed and heated up to the desired temperature. The suspension was stirred continuously at a constant rotational speed of 300rpm during each process to increase mass and heat transfer and prevent local overheat on the inner walls of the extractor. The experimental temperatures were 100, 125, and 150°C. The pressure was the water saturation pressure at the reaction temperature (0.10MPa for 100°C, 0.23MPa for 125°C, and 0.47MPa for 150°C). Extraction times were 5, 15, and 25min. The extraction temperature and time were chosen according to Ravber etal. [19] in order to avoid the thermal degradation of sugars and proteins. The highest severity factor used in this work was log R0 equal to 2.8, calculated by Eq. (1), that is rather below the value of 3.7 to avoid the production of inhibitors such as furfural and 5-HMF (hydroxymethylfurfural) according to previous results of our research group [25]. At the end of the experiments, the reactor was immediately cooled down in an ice bath and the suspension was filtered under vacuum. The liquid phase was filtered through a 0.22-μm cellulose acetate and stored at 4°C for further analysis and the remaining solid residue from each assay of extraction was collected and dried in an oven at 105°C until constant weight for its characterization. being t time in minutes, T temperature in °C, 100°C a reference temperature, and 14.75 an arbitrary constant related with activation energy. 2.3 Analyses ofextracts 2.3.1 pH determination The pH of the extracts was measured after each experimental run using a digital pH meter (Jenway Digital Meter, 3505, England). 2.3.2 Quantification ofsugars, organic acids, andinhibitors byhigh‑performance liquid chromatography (HPLC) The extracts were analyzed using a Water Breeze 1525 HPLC system (Waters, Milford, USA) with an integrated column heater and 717 plus autosampler according to the method described by Adamovic etal. [26]. The separation was carried out on a Shodex SH-1011 (1) R 0= t ∙ exp( T−100 14.75 ) column (SH1011, 7μm, 8mm × 300mm, Waters) using a mobile phase of water acidified with sulfuric acid (0.01N). The temperature, flow rate, and injection volume were 50°C, 0.8mL/min, and 25 µL, respectively. The identification of the sugars and acids was performed by comparison with the standards through retention time using Waters RI Detector 2414. The identification of the inhibitors was performed by comparison with the standards through retention time, at 254nm using a Waters UV–Vis detector. The quantification was performed using an external calibration curve of the analytical standards. The calibration curves of the glucose (R2 = 0.9999), xylose (R2 = 1), and arabinose (R2 = 0.9999) were obtained in the range of 14–1053μg/ mL and the calibration curves of the acetic (R2 = 0.9999) and formic (R2 = 0.9999) acids were obtained in the range of 14–835μg/mL. The calibration curves of the 5-HMF (R2 = 1) and furfural (R2 = 0.9999) were obtained in the range of 22–427μg/mL. Sugars were expressed as mg of sugar per g of SbP. 2.3.3 Quantification oftotal andfree amino acids byHPLC Total and free amino acids were quantified by HPLC with a variable wavelength UV detector (Agilent Technologies 1200 Series, Germany) in the extracts obtained at 150°C, 15min, and S/F of 20. The total and free amino acids were separated in a Zorbax Eclipse AAA C18 column (4.6mm × 150mm, i.d., 3.5μm) at 40°C using an injection volume of 36 µL and a mobile phase of buffer NaH2PO4.H2O (pH = 7.8) (A) and acetonitrile:methanol:water (45:45:10) (B) in a gradient method described by Náthia-Neves and Alonso [3]. For the total amino acid quantification, the extract (approximately 30mg) was hydrolyzed in a microwave system following the protocol described by Náthia-Neves and Alonso [3]. For the free amino acid quantification, the extract was filtered through a microfilter (0.22μm) into a glass vial (2mL) and direct analyzed in the same HPLC system (without microwave pretreatment). The quantification of each amino acid was performed from external calibration curves of the standards purchased from Sigma–Aldrich. The amino acid profile was expressed as mg of amino acid per g of SbP. 2.3.4 Quantification ofCGA byHPLC CGA content in the extracts was quantified by an HPLC–PDA (Waters, Alliance E2695, Milford, USA) system. The separation of the CGA was performed in a C18 column according to the method described by Meinhart etal. [27]. The quantification of the CGA was performed from the external calibration curve of Biomass Conversion and Biorefinery 1 3 the analytical standard purchased from Sigma–Aldrich and the CGA recovery was expressed as mg of CGA per g of SbP. 2.3.5 Total phenolic compounds (TPC) andantioxidant capacity TPC was analyzed using the protocol reported by Singleton etal. [28]. The extracts were diluted in bidistilled water. The extract (40 μL), water (3mL), and Folin–Ciocalteu (200 μL) reagent were mixed and incubated for 3min at room temperature. Then, saturated sodium carbonate (600 μL) was added, and the mixture was incubated for 30min at 40°C. The absorbance was recorded at 765nm and the quantification of TPC was performed from the external calibration curve of the analytical gallic acid (53 to 852μg/mL) standard purchased from Sigma–Aldrich. TPC content was expressed as mg gallic acid equivalent (GAE) per g of SbP. The antioxidant capacity was analyzed by the oxygen radical absorbance capacity method (ORAC). A total of 10-mM potassium phosphate buffer (PBS) at pH 7.4 was used as blank and the Trolox (12.5–200μM) solution was used as a standard. For each analysis, 25 μL of the PBS, 25 μL of the standard dissolution (12.5, 25, 50, 100, and 200μM), 25 μL of the extract, and 150 μL of fluorescein (100nM) were inserted into wells of an opaque fluorescent plate and incubated for 30min at 37°C. Then, 25 μL of AAPH (2,2′-azobis(2-amidino-propane) dihydrochloride) solution (240mM) was added. The fluorescence decrease (excitation wavelength was set at 485nm; emission wavelength at 520nm) was recorded for 130min using a FLUOstar OPTIMA (BMG Labtech, Offenburg, Germany) equipment at 37°C. The antioxidant capacity was expressed as μmol of Trolox equivalent (TE) per g of SbP. 2.4 Analyses ofsolid residue 2.4.1 Determination ofthesolid amount After the subcritical water treatment, the suspension was filtered, and the remaining solid residue was transferred to a porcelain crucible and dried in an oven at 105°C until constant weight. The difference between the initial amount of SbP (on a dry basis) and the final mass remaining in the reactor (on a dry basis) after each experimental run was used to calculate the amount of solid residue. 2.4.2 Protein andtotal amino acids profile byHPLC Protein content was determined following the method proposed by NREL (NREL/TP-510–42,625; using a conversion factor of 6.25). Total amino acids were quantified both in the raw material and in the residual solid after subcritical water treatment. For this purpose, the samples were hydrolyzed in a microwave with acid (6N HCl) and an inert atmosphere. The sample treatment and the separation of the amino acid by HPLC were carried out following the same protocol described in Section2.3.3. 2.5 Experimental design andoptimization Response surface methodology (RSM) was applied to evaluate the effects of extraction parameters and optimize the process conditions. The Box–Behnken experimental design with three numeric factors on three levels was used. In the experimental design, the temperature (x1), extraction time (x2), and S/F (x3) were selected as independent variables (Table1). Fifteen randomized runs with three replicates at the central point were performed. The response variables were fitted to the second-order polynomial model (Eq.2). The experimental design was evaluated using Minitab 16® software (Minitab Inc., State College, PA, USA). The results were statistically tested by analysis of variance (ANOVA) with the significance levels of 5%. where Y represents the response variable, Xi and Xj are the independent variables affecting the response, and β0, βi, βii, and βij are the regression coefficients for intercept, linear, quadratic, and interaction terms, respectively. 3 Results anddiscussion 3.1 Characterization oftheraw material Table2 shows the SbP characterization. The moisture, ash, and protein content were similar to values reported by Ramachandran etal. [29] that found 9% moisture, 6.6% ash, and 34% protein. Lipid content and total extractives soluble in water were similar to those reported by GeneauSbartaï etal. [30] (2% of lipids and 22.8% of water-soluble (2) Y=β 0+ 3 ∑ i= 1 βiXi+ 3 ∑ i= 1 βiiX2 i+∑ 3 ∑ i < j= 1 βijXiX j Table 1 Experimental range and levels of the independent variables SbP sunflower by-product Variables Factor Range and level − 1 0 1 Temperature (°C) x1100 125 150 Time (min) x25 15 25 S/F (g water/g SbP) x310 15 20 Biomass Conversion and Biorefinery 1 3 components). The values of lignin (insoluble lignin + soluble lignin) are in agreement with those found in literature, which range from 8 to 20% [7, 30, 31]. As expected, characterization analysis establishes that SbP is a promising candidate for protein-rich extract production and for the generation of simple sugars. 3.2 Influence ofprocess variables onsugars andprotein recovery The Box–Behnken experimental design with three replicates at the center point was used to investigate the effect of temperature, time, and S/F on the glucose, xylose, protein, TPC, CGA, and antioxidant capacity of the obtained extracts. The experimental results are shown in Table3. Results of ANOVA for the effects of variables on the glucose, xylose, and protein are shown in Table4 (that considers only the significant variables at 5%), with the corresponding coefficients of multiple determinations (R2) and adjusted determination coefficients (adj R2). The regression models were highly significant with satisfactory values of determination coefficients (R2 = 0.939–0.963) and adjusted determination coefficients (adj R2 = 0.915–0.949) (Table4). These results show that the polynomial regression model can be used to navigate the design space for glucose, xylose, and protein responses. The results in Table4 indicated that the linear coefficients of extraction temperature and extraction time were significant on the extraction of glucose, xylose, and protein. In addition, quadratic terms of extraction temperature and the interaction between temperature and time were also significant (P < 0.05) on these responses. The threedimensional response surface and two-dimensional contour plot for glucose, xylose, and protein are presented in Fig.2. The second-order polynomial equations for glucose (mg/g SbP), xylose (mg/g SbP), and protein (mg/g SbP) extraction yields are shown in Eqs. (3), (4) and (5), respectively. The glucose extraction yield ranged from 2.1 to 5.7mg/g SbP and the xylose content ranged from 5.3 to 13.9mg/g SbP for the studied combinations. Traces of arabinose were also identified in the extracts (ranging from 0.1–0.9mg/g Table 2 Sunflower by-product characterization (% w/w) *Results expressed on a dry basis NREL methods Results (wt.%) Moisture 9.2 ± 0.1 Ash 5.78 ± 0.03* Lipids 1.9 ± 0.4* Protein 31.8 ± 0.1* Extractives in water 14.2 ± 0.4* Extractives in ethanol 5.3 ± 0.1* Insoluble lignin 14 ± 2* Soluble lignin 8.2 ± 0.3* Table 3 Experimental results obtained by the extraction using subcritical water of the sunflower by-product SbP sunflower by-product. Results are mean ± standard deviation on a dry basis Run order Independent variables Investigated responses X1X2X3 Temperature (°C) Time (min) S/F (g water/ gSbP) Glucose (mg/g SbP) Xylose (mg/g SbP) CGA (mg/g SbP) TPC (mg/g SbP) ORAC (µmol/g SbP) Protein (mg/g SbP) 1 125 (0) 15 (0) 15 (0) 2.7 ± 0.4 5.6 ± 0.7 4.7 ± 0.2 21.9 ± 0.3 295.3 ± 5.4 44.5 ± 0.3 2 125 (0) 25 (+ 1) 10 (-1) 2.1 ± 0.4 5.8 ± 0.7 3.68 ± 0.04 20.2 ± 0.6 227.4 ± 7.1 42.6 ± 0.7 3150 (+ 1) 15 (0) 20 (+ 1) 5.04 ± 0.04 11.3 ± 0.7 2.50 ± 0.01 26.1 ± 1.3 219.1 ± 14.5 133.5 ± 0.5 4100 (− 1) 15 (0) 10 (− 1) 2.3 ± 0.4 5.4 ± 0.7 7.2 ± 0.2 18.0 ± 0.8 258.2 ± 8.2 33.1 ± 0.5 5100 (− 1) 25 (+ 1) 15 (0) 2.2 ± 0.4 5.8 ± 0.7 6.3 ± 0.1 20.9 ± 0.6 259.1 ± 15.2 38.3 ± 0.1 6150 (+ 1) 15 (0) 10 (− 1) 4.7 ± 0.4 13.9 ± 0.7 2.2 ± 0.4 20.2 ± 1.1 213.3 ± 8.1 115.6 ± 0.1 7 125 (0) 5 (− 1) 10 (− 1) 2.5 ± 0.4 5.8 ± 0.7 6.30 ± 0.01 18.2 ± 0.6 231.4 ± 8.5 40.5 ± 6.2 8 125 (0) 15 (0) 15 (0) 2.7 ± 0.4 6.2 ± 0.7 4.9 ± 0.3 22.1 ± 0.4 288.2 ± 9.2 43.7 ± 0.1 9100 (− 1) 15 (0) 20 (+ 1) 2.2 ± 0.4 5.4 ± 0.7 8.9 ± 0.6 19.7 ± 0.3 284.1 ± 33.1 46.4 ± 0.6 10 150 (+ 1) 5 (− 1) 15 (0) 3.5 ± 0.4 8.4 ± 0.7 5.1 ± 0.2 20.8 ± 0.8 296.2 ± 6.3 93.9 ± 0.3 11 100 (− 1) 5 (− 1) 15 (0) 2.1 ± 0.4 4.6 ± 0.7 8.8 ± 0.1 17.2 ± 0.6 288.5 ± 21.2 37.2 ± 2.4 12 125 (0) 5 (− 1) 20 (+ 1) 2.2 ± 0.4 5.8 ± 0.7 6.5 ± 0.02 21.5 ± 0.2 269.1 ± 8.2 49.4 ± 1.8 13 150 (+ 1) 25 (+ 1) 15 (0) 5.7 ± 0.4 13.5 ± 0.7 2.4 ± 0.2 26.4 ± 2.1 228.4 ± 8.5 130.2 ± 0.7 14 125 (0) 25 (+ 1) 20 (+ 1) 2.5 ± 0.4 6.6 ± 0.7 2.8 ± 0.2 22.8 ± 0.2 275.2 ± 25.1 56.2 ± 9.3 15 125 (0) 15 (0) 15 (0) 2.6 ± 0.4 5.3 ± 0.7 4.8 ± 0.9 22.1 ± 0.2 290.3 ± 5.2 42.1 ± 0.8 Biomass Conversion and Biorefinery 1 3 Table 4 Analysis of variance regression model of the investigated responses for the extraction of sunflower by-product using subcritical water SbP sunflower by-product Source DF Glucose (mg/g SbP) Xylose (mg/g SbP) Protein (mg/g SbP) Coefficient Sum of squares F-value P value Coefficient Sum of squares F-value P value Coefficient Sum of squares F-value P value Model 4 18.3527 38.72 0.001 133.251 54.49 0.001 17,349.7 66.08 0.001 Intercept 24.5129 68.8696 732.152 Linear X11 − 0.376793 12.8857 108.76 0.001 − 1.10497 84.462 138.15 0.001 − 12.1924 12,656.7 192.83 0.001 X21 − 0.248471 0.6165 5.2 0.046 − 0.428441 12.053 19.71 0.001 − 3.81583 12,656.7 192.83 0.001 Quadratic X11 0.00157764 3.6297 30.64 0.001 0.0046752 31.875 52.14 0.001 0.0530166 4099 62.45 0.001 Interaction X1X21 0.00220984 1.2208 10.3 0.009 0.0044095 4.861 7.95 0.018 0.0352834 311.2 4.74 0.05 Residual Error 10 1.1848 6.114 656.4 Lack of fit 4 1.0212 9.36 0.009 3.471 1.97 0.218 270.1 1.05 0.456 Pure error 6 0.1637 2.643 386.3 Total 14 19.5375 139.365 18,006.1 R293.94% 95.61% 96.35% Adjusted-R291.51% 93.86% 94.90% Biomass Conversion and Biorefinery 1 3 Temperature (°C) Time (min) 150140130120110100 25 20 15 10 5 > – – – – < 6 68 810 10 12 12 14 14 Xylose Temperature (°C) Time (min) 150140130120110100 25 20 15 10 5 > – – – < 2 23 34 45 5 Glucose 2 3 4 100100 125 5 12 5 150 18 12 25 18 Glucose (mg/gSbP) Time (min) Te mperature (°C) a) b) c) d) e) f) 50 75 100 100100 125 100 125 12 5 150 18 12 25 Protein (mg/gSbP) Time (min) Te mperature (°C) Temperature (°C) Time (min) 150140130120110100 25 20 15 10 5 > – – – – < 40 40 60 60 80 80 100 100120 120 Protein 5.0 7.5 10.0 100100 125 12.5 12 5 150 18 12 25 Xylose (mg/gSbP) Time (min) Te mperature (°C) Fig. 2 Three-dimensional response surface and two-dimensional contour plot. a Three-dimensional response surface for glucose; b twodimensional contour plot for glucose; c three-dimensional response surface for xylose; d two-dimensional contour plot for xylose; e three-dimensional response surface for protein; f two-dimensional contour plot for protein Biomass Conversion and Biorefinery 1 3 SbP). To the best of our knowledge, there are no studies in the literature reporting the obtention and quantification of sugars from SbP using subcritical water. Ravber etal. [18] reported the extraction of oil and water-soluble phase from fresh sunflower seeds with subcritical water. However, these authors did not quantify the sugar content present in the extracts. Alenezi, Leeke, Santos, and Khan [32] also studied the extraction of sunflower oil using subcritical water without analyzing the sugar content. Temperature showed the strongest impact on total sugars recovery. The linear, quadratic, and the interaction between temperature and time effects had the strongest influence on total sugars extraction. As shown in Fig.2, the increase in temperature with the increase in extraction time had a positive effect on the recovery of xylose and glucose. These results prove that the selected temperature conditions did not favor the formation of final degradation products such as furfural and 5-HMF that can occur at temperatures above 150°C [33]. The pH of the extract ranged from 5.5 to 6.0. As shown in Fig.3, the pH has a slight decrease as the temperature rises from 100 to 150°C, which indicates the presence of acidic materials. In order to verify this hypothesis, an HPLC analysis was carried out, which revealed the presence of traces of organic formic and acetic acids in extracts obtained at 150°C (less than 0.2%). According to literature, the SbP presents a low content of free monomers, 11% glucose, 3.5% xylose, 1.47% arabinose, 1.25% rhamnose, 1.13% mannose, and 1.23% galactose [34]. Glucose is mostly originated from cellulose hydrolysis under temperature between 200 and 300°C [35]. Thus, the recovery of glucose could be improved by increasing the temperature. However, it is worth mentioning that this rise in temperature may lead to the formation of degradation products, limiting the use of the obtained extract in fermentation processes [36]. The content of glucose and xylose in the extracts of SbP are comparable to those observed in extracts of coffee industry residues (glucose = 3.3mg/g coffee powder and xylose = 9.8mg/g defatted coffee cake, obtained at 175°C) [15] and in the extracts of brewer’s spent grains (glucose = 1.8mg/g grain and xylose = 9.0mg/ g grain, obtained at 160°C) [37]. The protein content in the obtained extracts ranged from 33.1 to 133.5mg/g SbP depending on operational conditions. The highest recovery of protein (133.5mg/g SbP) was obtained at 150°C and 15min using S/F of 20 (Run order 3). The amount of protein recovered at 150°C was almost twice the amount recovered at 125°C, indicating the positive effect of increasing temperature in the recovery of these compounds. The results could be due to the higher solubility of proteins at high temperatures. At 150°C, the dielectric constant of the subcritical water is approximately 43.96, and at 100°C, it is approximately 55.43, which seems to decrease the ability of the medium to extract the protein phase. On the other hand, higher temperatures would result in more protein structure rupture, which is associated with the formation of amino acids and peptides. The soluble products presented a toasty aroma when the temperature increased. The interactive effect 012345678910 11 12 13 14 15 2 3 4 5 6 150 °C 125 °C 100 °C pH Run Order Fig. 3 pH of the extracts obtained from sunflower by-product using subcritical water Table 5 Amino acid profile in the extract obtained from sunflower by-product using subcritical water at 150°C, 15min, and S/F of 20 SbP sunflower by-product Amino acids Total amino acids in the extract from sunflower by-product (mg/g extract) (mg/g SbP) Aspartate 29.3 ± 2.7 10.3 ± 10.9 Glutamate 78.4 ± 2.8 27.5 ± 0.9 Asparagine - - Serine 14.9 ± 0.6 5.2 ± 0.2 Glutamine - - Histidine 9.3 ± 0.6 3.3 ± 0.2 Glycine 24.1 ± 0.9 8.5 ± 0.3 Threonine 11.6 ± 0.9 4.1 ± 0.3 Arginine 22.4 ± 0.6 7.9 ± 0.2 Alanine 13.7 ± 0.6 4.8 ± 0.2 Tyrosine 6.8 ± 0.6 2.4 ± 0.2 Cysteine - - Valine 9.5 ± 0.7 3.4 ± 0.2 Methionine 6.8 ± 0.6 2.4 ± 0.2 Tryptophan - - Phenylalanine 12.2 ± 0.4 4.2 ± 0.1 Isoleucine 9.7 ± 0.4 3.4 ± 0.1 Leucine 15.6 ± 0.6 5.5 ± 0.2 Lysine 8.7 ± 0.6 3.1 ± 0.2 Proline 7.2 ± 0.9 2.6 ± 0.3 Total 280.7 ± 3.2 98.5 ± 1.1 Biomass Conversion and Biorefinery 1 3 of temperature and extraction time on protein recovery is shown in Fig.2e and f. The amount of protein in the sunflower raw material was almost 32% (Table2), and the extract obtained at 150°C and 15min using S/F of 20 presented 43% of the protein in the soluble product. This result is comparable to the 59.3% protein yield obtained by Lu etal. [38] when studying the protein extraction from the HPSI using enzyme-assisted subcritical water extraction (120°C). Although the process studied has provided a high yield of proteins, a scale-up and an economic analysis are necessary to determine the economic and technical viability of the developed process on an industrial scale. The positive effect of temperature on the total protein content in the extracts obtained from deoiled rice bran by subcritical water was also reported by Sereewatthanawut etal. [39]. In this study, the protein content extracted ranged from ~ 48mg/g rice bran (100°C and 30min) to 219mg/g rice bran (200°C and 30min). The yields of protein extraction from extruded soybean flakes and non-extruded soybean flakes using subcritical water showed different behaviors. For the extruded soybean flakes, the amount of protein extracted increased with increasing temperature and the solid-to-liquid ratio decreased, while using non-extruded soybean flakes, higher protein yields were observed at low temperatures and low solid-to-liquid ratio [40]. The protein content in the extracts obtained from extruded soybean flakes ranged from 266mg/g extruded soybeans (66°C and 30min) to 731mg/g extruded soybeans (234°C and 30min), and the protein content extracted from soybean flakes ranged from 308mg/g soybeans flakes (100°C and 20min) to 719mg/g soybeans flakes (100°C and 40min) [40]. To verify the presence of amino acids in the extract obtained at 150°C and 15min using S/F of 20 (Run 3), 123456789101112131415 0 10 20 30 40 50 60 70 80 90 100 lairetamwarlaitinifog001repdilosg Run Order Fig. 4 Solid co-products after the subcritical water treatment of the sunflower by-product under different experimental conditions. The mean and standard deviation of the measurements obtained at the central point were 79 ± 3g solid per 100g of initial SbP Fig. 5 Amino acid profile in the sunflower by-product before and after the subcritical water treatment. ASP, aspartate; GLU, glutamate; ASN, asparagine; SER, serine; GLN, glutamine; HIS, histidine; GLY, glycine; THR, threonine; ARG, arginine; ALA, alanine; TYR, tyrosine; CYS, cysteine; VAL, valine; MET, methionine; TRP, tryptophan; PHE, phenylalanine; ILE, isoleucine; LEU, leucine; LYS, lysine; PRO, proline ASP GLU SER HISGLY THRARG ALATYR CYSVAL METPHE ILELEU LYSPRO 0 5 10 15 20 25 30 35 40 45 50 )PbSg/gm(sdicaonimA Amino acid profile Fresh sunflower by-product Treated sunflower by-product