Ultrasound processing of coffee silver skin, brewer’s spent grain and potato peel wastes for phenolic compounds and amino acids: a comparative study

Abstract

Awareness towards utilizing food-processing by-products are increasing in health as well as environmental purview. Coffee silver skin (CSS), potato peel (PP) and brewer’s spent grain (BSG) are voluminous by-products in their respective processing industries. The present study compared these three by-products for their prospective utilization in producing polyphenols-rich aqueous extracts by using ultrasound-assisted extractions (UAE). A probe-type sonicator was used for ultrasound treatments. The total phenolic contents in the extracts were assessed by Folin-Ciocalteu assay, while the phenolic profiles of the extract was characterized by LC-Q-TOF mass spectrometry. The microstructure of the samples after UAE was evaluated by scanning electron microscopy (SEM). Ultrasound treatment enhanced the rate of extraction and recovered 2.79, 2.12 and 0.66 mg gallic acid equivalents/g of TPC from CSS, PP and BSG, respectively in 30 min, which correspond to recoveries of 97.6%, 84.5% and 84.6%, respectively, compared to conventional solid–liquid extractions carried out for 24 h. The extraction yield was dependent on the particle size of the raw materials and the highest yield was obtained from the materials with 100–250 µm particle size. The SEM imaging revealed that ultrasound treatment caused prominent tissue damage. Extracts contained mainly hydroxycinnamic acid derivatives of phenolic acids. PP and CSS had the highest amounts of umami free amino acids (0.13 mg/g in each), while BSG contained the highest amount of essential amino acids (92 mg/g). The present work shows that CSS, PP and BSG are good sources of polyphenols and UAE can be employed to enhance the extraction efficiency as means of a green approach.

Electronic supplementary material

The online version of this article (10.1007/s13197-020-04738-2) contains supplementary material, which is available to authorized users.

Introduction

The management of voluminous wastes and by-products generated during the processing is one of the major challenges of food and brewing industries. The extra cost and strict environmental legislation associated with waste treatment before disposal pose burden to the agri-food processing industries. Nevertheless, there is an immense opportunity to convert such wastes or by-products into value-added products, which not only facilitates sustainable management of energy and resources, but also a solution to growing environmental concerns. Several scientific investigations have proven that the food processing wastes could serve as cheap raw materials for the production of various nutrients and food ingredients (Chandrasekaran 2013). Coffee silver skin (CSS), potato peel (PP) and brewer’s spent grain (BSG) are few among such primary by-product available from their respective industry for value addition, which otherwise discarded or used as animal feed or organic fertilizer.

Coffee silver skin (CSS) is a primary by-product generated by the coffee bean roasting industry (Toschi et al. 2014). CSS is a thin outer layer of coffee beans, accounting up to 4.2% of the bean (Polidoro et al. 2018). Recent studies have revealed that it is a valuable source of antioxidant molecules (Jiménez-Zamora et al. 2015), dietary fibre (Behrouzian et al. 2016), bioactive compounds (Fernandez-Gomez et al. 2016), cosmetic ingredients (Rodrigues et al. 2015), and bio-oil (Polidoro et al. 2018). PP is a voluminous waste generated by the potato-based food processing industries (e.g. potato chips manufacturing industry) that accounts roughly about 0.16 tons per ton of raw potato processed (Pathak et al. 2018). The PP waste is prone to rapid microbial spoilage and causes pollution (Galhano dos Santos et al. 2016). It could be used as an alternative source of several bioactive phytochemicals with antimicrobial, antioxidant, anticancer and anti-inflammatory properties (Wu 2016). BSG is a major residue generated after brewing, accounting for 85% of the total wastes produced by the brewing industry (Mussatto et al. 2006). Globally BSG reaches up to 39 million tons annually (Lynch et al. 2016). It is a good source of protein (ca. 20%) and fibre (ca. 70%) and widely used as animal feed (Mussatto 2014; Lynch et al. 2016). Recent investigations have shown that it is a sustainable source for nutrients and bioactive compounds suitable for human consumption (Lynch et al. 2016).

Polyphenols are the major class of bioactive compounds present in CSS, PP and BSG (Regazzoni et al. 2016; Lynch et al. 2016; Kumari et al. 2017). Polyphenols are recognized within the food and pharmaceutical sectors owing to their wide array of health effects (Pandey and Rizvi 2009). Several conventional and non-conventional extraction methods have been established to recover polyphenols from plant sources, nevertheless, conventional solid–liquid extraction is the most commonly employed technique. Previous studies have shown that alcohol, usually methanol, provides superior recovery of polyphenols from plant matrices (Kumari et al. 2017). The present study attempted to use water as an extraction solvent along with ultrasound assisted extraction (UAE) technology to improve extraction yield by accelerating solvent diffusion and mass transfer as apart from its ‘green’, sustainable and cost-effective benefits (Kumari et al. 2017; Savic Gajic et al. 2019; Savic and Savic Gajic 2020). UAE is also considered as an efficient alternative to conventional solid–liquid extraction methods as it is relatively simple, versatile, cost-effective, and easy to use and scale up compared to other non-conventional techniques. In the present study, the efficiency of UAE in rapid extraction of polyphenols from three different food-processing wastes (CSS, PP and BSG) was investigated in comparison to conventional extraction, besides amino acid and polyphenols profile by using chromatographic techniques.

Materials and methods

Samples

Illy S.p.A. (Trieste, Italy) donated dried samples of CSS. Aperitivos Gus S.L. (Riego de la Vega, León, Spain) donated fresh potato peel (PP). Fresh samples were dried and all samples were then homogenized into fine particles using a blender. The powder was separated into two fractions based on particle size such as 100 to 250 µm and 500 and 750 µm.

Chemicals

All solvents used for LC–MS (LC–MS grade), gallic acid, trichloroacetic acid (TCA) and Folin-Ciocalteu’s phenol reagent were obtained from Sigma–Aldrich, Dublin, Ireland. Other solvents and regents (reagent grade) used in the study were obtained from Fisher Chemical, Dublin, Ireland. Milli-Q water (Merck Millipore, MA, USA) was used in all experiments.

Proximate analysis

Proximate analyses of raw materials were carried out based on AOAC official methods (AOAC 2000). Gravimetric method was used to determine moisture content in an oven at 105 °C. Ash content was analysed according to AOAC method 923.03. Soxhlet extraction with n-hexane was used to determine total crude fat. Carbohydrate content was determined by weight difference. All analyses were carried out in triplicates.

Total protein content analysis

Total protein content present in the extract was determined by determination of total nitrogen content using LECO FP628 protein analyser (LECO Corp., MI, USA), according to the procedure described elsewhere (Kadam et al. 2017).

Total extractable polyphenol in water (conventional extraction)

Powdered sample (2.5 g) was added into a 100 mL screw-cap bottle containing 50 mL of Milli Q water and agitated continuously at 1000 rpm for 24 h at room temperature. The extract was centrifuged (1000 g, Sigma Laborzentrifugen GmbH, Germany) and the supernatant was analysed for total polyphenol content (TPC).

Ultrasound assisted extraction (UAE)

The UAE was performed according to Zhang et al. (2018) by taking 10 g of raw material in a jacketed glass reactor containing 200 mL of Milli Q water. The extraction was initiated by sonicating the sample at a power level of 7.8 or 49.5 W by immersing a probe type sonicator having a diameter of 1.2 cm and height of 13.7 cm (Model XL2020, SONICATOR, UK,) at a depth of 2 cm. The temperature inside the reactor was controlled by circulating water (25 °C) through the jacket. Sample aliquots were taken every 2, 5, 10, 20 and 30 min and immediately centrifuged at 14,400g for 3 min. The supernatants were transferred to screw-capped containers and stored at − 20 °C until further analyses.

TPC determination

The TPC in the extract was analysed by Folin-Ciocalteu spectrophotometric assay and expressed as mg gallic acid equivalent (GAE)/g of dry sample material (Ganesan et al. 2008). Briefly, 100 µL of extract was vortexed with Folin-Ciocalteu reagent (100 µL), methanol (100 µL) and 20% sodium carbonate solution (700 µL) in a microtube. The mixture was incubated at room temperature of 23–25 °C in the dark for 20 min followed by centrifugation at 14,400 g for 3 min. The optical density of the supernatant was determined at 735 nm using a UV–Vis spectrophotometer (U-2900, Hitachi, Ireland). The concentration of phenolic compounds was determined based on a linear calibration curve prepared using gallic acid standard solutions.

Amino acid analysis

Prior to free amino acid (FAA) analysis, proteins present in the samples were precipitated by adding 24% w/v TCA (1:1 v/v) followed by centrifugation at 14,400g for 10 min (Sigma Laborzentrifugen GmbH, Germany). The supernatant was mixed with citrate buffer of pH 2.2 and an internal standard solution (norleucine) followed by derivatization with ninhydrin. The ninhydrin-derivatives of amino acids were analysed using an amino acid analyzer (Jeol JLC-500/V, Jeol Ltd., Garden city, Herts, UK). A sodium ion high performance cation exchange column (Jeol Ltd., Herts, UK) was used to separate the analytes according to manufacturer instructions. For total amino acids (TAA) analysis, samples were initially hydrolysed at 110 °C for 23 h using 6 M hydrochloric acid followed by amino acid analysis as described above (Kadam et al. 2017).

HPLC-Q-TOF–MS/MS profiling of polyphenols

The phenolic compounds in the extracts were separated on a reversed phase column (Atlantis T3 C18) connected to a HPLC (Alliance 2695, Waters Corporation, Milford, MA, USA) and identified by using quadrupole time of flight mass spectrometry (Q-TOF Premier™ mass spectrometer, Waters Corporation, Milford, USA) in negative mode as described by Zhang et al. (2018). Phenolic compounds were tentatively identified by evaluating their molecular mass (m/z) and fragmentation pattern in comparison with the available database.

Microstructure evaluation

A FEI Quanta 3D FEG DaulBeam scanning electron microscope (SEM) (FEI Ltd, Hillsboro, USA) was employed to evaluate the microstructure of samples before and after the extraction by sputter coating samples on a double-sided carbon tape (Zhang et al. 2018).

Results and discussion

Proximate composition

Moisture, protein, carbohydrates, fat and ash (%) in CSS, PP and BSG were distinctive (Supplementary datasheet Table S1). The crude protein content (26.18%) in BSG samples was higher than PP (15.8%) and CSS (10.16%). According to Lynch et al. (2016), crude protein content in BSG may vary from 19 to 30% depending on different brewing factors. CSS contained the highest carbohydrates (78.26%), which is similar to the level reported by a previous work (Toschi et al. 2014). The fat content was similar in all the three raw materials (about 4%).

TPC recovery

The total extractable polyphenols in CSS, PP and BSG obtained from solid–liquid extraction was 2.86 ± 0.08, 2.51 ± 0.43 and 0.78 ± 0.04 mg GAE/g, respectively, indicating CSS and PP are good sources of polyphenols. Recent studies have reported that up to 122.9 mg GAE/g and 7.67 mg GAE/g (dry weight basis) of TPC can be recovered from CSS (Regazzoni et al. 2016) and PP (Kumari et al. 2017), respectively. In case of CSS, Wen et al. (2019) showed that ultrasound-assisted aqueous methanol extraction provides slightly better yield of phenolic compounds compared to aqueous extraction (~ 6 compared to ~ 9 mg GAE/g sample). BSG contained relatively low amounts of polyphenols, which is in agreement with a previous work that has reported up to 0.2 mg GAE/g (fresh weight basis) of TPC (Fărcaş et al. 2015). It should be noted that the TPC in extract is dependent on several factors including origin of raw material, previous processing methods and extraction conditions. In the previous studies, aqueous alcohol (methanol or ethanol) were used to extract TPC as this solvent system provides superior recovery of polyphenols (Do et al. 2014).

In the present investigation, the UAE was carried out to enhance the recovery rate of TPC in the aqueous medium. According to results, up to 2.79 ± 0.48, 2.12 ± 0.22 and 0.66 ± 0.14 mg GAE/g of TPC can be recovered from CSS, PP and BSG, respectively (Fig. 1a–c), after ultrasound treatment of 30 min that corresponds to a recovery of 97.6%, 84.5% and 84.6%, respectively, compared to conventional extraction carried out for 24 h. The extraction kinetics revealed that the rate of extraction increased with the extraction time and the maximum yield was obtained after 20 or 30 min of ultrasound treatment, depending on the type of sample, particle size and applied power (Fig. 1).

Fig. 1.

Kinetics of UAE of TPC from a coffee silver skin, b potato peels and c Brewer’s spent grain

In general, highest TPC was recovered from raw materials with smaller particle size (100–250 µm) when compared to the larger ones (500–750 µm) (Fig. 1). During solid–liquid extraction process, smaller particles provide greater surface area that in turn enable enhanced diffusion and mass transfer, leading to increased rate of extraction. Other researchers have also reported similar results while extracting antioxidant polyphenols from other plant matrices (Makanjuola 2017; Shewale and Rathod 2018).

The highest TPC was obtained when the applied power was 49.5 W in comparison to lower level of 7.8 W for all the matrices regardless of extraction time or particle size (Fig. 1). The ultrasound treatment improves extraction efficiency by rupturing the cell membrane and intracellular organelles and by improving the mass transfer of analytes (McDonnell and Tiwari 2017). Besides, ultrasound can reduce the particle size and increase the dissolution of analytes through intense cavitation effects induced by ultrasonic waves (Chemat et al. 2017). These effects would increase with power level, leading to increased extraction yield (Alzorqi et al. 2017).

LC-Q-TOF profiling of polyphenols

The CSS extracts mainly constituted chlorogenic acids and feruloyl-, and coumaroyl-derivatives of quinic acid. Presence of quinic acid in these compounds was characterised with the MS/MS fragment ions at m/z 191, attributed to deprotonated quinic acid in the negative ion mode (Table 1). Though the presence of caffeic acid in CSS was reported by Borrelli et al. (2004), this was not detected in the present study. The PP extract also showed the presence of phenolic acids like p-coumaric acid, chlorogenic acid, caffeic acid and protocatechuic acid as reported by other researchers previously (Al-Weshahy and Rao 2009; Singh and Saldaña 2011). Chlorogenic acid and caffeic acid concentrations in PP have previously been found to range from 0.78 to 2.79 mg/g and from 0.26 to 0.72 mg/g of PP (dry basis) respectively, depending upon the potato varieties, while p-coumaric and ferulic acids was found to occur in trace amounts (Al-Weshahy and Rao 2009). Although previous investigators reported the presence of gallic acid in PP extracts (Singh and Saldaña 2011), it could not be detected in the PP used in the present study. The BSG extract contained flavonoids including ( +) catechin and, hydroxycinnamic acid derivatives including p-coumaric acid, caffeic acid and protocatechuic acid. McCarthy et al. (2013) also reported these phenolic acids along with caffeic and ferulic acid derivatives, which, however, were not detected in the present study. Many phenolic acids including ferulic and p-coumaric acids exists in insoluble bound-form and alkaline hydrolysis is required to obtain them in their free form (Mussatto et al. 2007). Overall, aqueous extraction predominately released phenolic acid from CSS, PP and BSG, possibly due to the polar nature of these compounds.

Table 1.

Free polyphenolic compounds characterised in each of the three extracts

Compound	Retention time (min)	Observed [M–H]− m/z	Major fragments [M–H]− m/z	Molecular formula	References
Coffee silver skin (CSS) extract
5-Feruloylquinic acid	6.50	366.9	191, 173	C17H20O9	Regazzoni et al. (2016)
3-Caffeoylquinic acid	3.60	352.9	191, 179, 172.9	C16H18O9	Clifford et al. (2003)
5-Caffeoylquinic acid	2.10	352.9	191, 179	C16H18O9	Bresciani et al. 2014)
4-Caffeoylquinic acid	3.20	352.9	191, 173	C16H18O9	Bresciani et al. (2014)
3-Coumaroylquinic acid	3.65	337	191, 173, 163	C16H18O8	Clifford et al. (2003)
5-Coumaroylquinic acid	3.65	337	191, 173, 163	C16H18O8	Bresciani et al. (2014)
Potato peel (PP) extract
p-Coumaric acid	6.66	163	119	C9H8O3	Sánchez-Rabaneda et al. (2004)
Caffeic acid	4.75	179	135	C9H8O4	(Sánchez-Rabaneda et al. 2004)
Chlorogenic acid	2.20	353	191	C16H18O9	Oertel et al. (2017)
Protocatechuic acid	2.32	153	109	C7H6O4	Sánchez-Rabaneda et al. (2004)
Brewer’s spent grain (BSG) extract
p-Coumaric acid	2.55	163	119	C9H8O3	Sánchez-Rabaneda et al. (2004)
Caffeic acid	4.75	179	135	C9H8O4	Sánchez-Rabaneda et al. (2004)
Protocatechuic acid	2.26	153	109	C7H6O4	Sánchez-Rabaneda et al. (2004)
(+) Catechin	3.32	289	245	C15H14O6	Carvalho et al. (2015)
Vanillic acid	7.72	167	123	C8H8O4	Carvalho et al. (2015)

Compounds are tentatively identified based on the molecular weight, fragmentation pattern, available database and the literature

Microstructure analysis

The physical changes induced by ultrasound on the raw materials during UAE was evaluated by scanning electron microscopy. The untreated raw material had a rough surface without any apparent cell damage (Figs. 2, 3, 4). The pellet obtained after conventional extraction wrinkles on the surface, possibly due to shrinkage after the mass transfer of intracellular compounds. The samples obtained after UAE had a different morphology than the untreated and conventionally treated samples having a smooth surface with fine cracks indicating cell damage (Figs. 2, 3, 4). Moreover, the surface appeared more compact, possibly due to breakage of intercellular linkages and reorganization of cellular mass. Similar morphological changes by the use of UAE have been observed by others (Altemimi et al. 2016). In principle, ultrasound waves cause cell damage due to cavitation effects, that enables rapid diffusion of solvent and mass transfer (Chemat et al. 2017).

Fig. 2.

Scanning electron microscopy images of coffee silver raw material (a, b), after conventional extraction (24 h) (c, d) and after UAE (30 min) (e, f). a, c and e are samples of 100–250 µm particle size and b, d and f are samples of 500–750 µm particle size

Fig. 3.

SEM images of PP raw material (a, b), after conventional extraction (24 h) (c, d) and after UAE (30 min) (e, f). a, c and e are samples with a particle size of 100–250 µm and b, d and f are samples with a particle size of 500–750 µm

Fig. 4.

SEM images of BSG raw material (a, b), after conventional extraction (24 h) (c, d) and after UAE (30 min) (e, f). a, c and e are samples with a particle size of 100–250 µm and b, d and f are samples with a particle size of 500–750 µm

Amino acid analysis

The concentrations of free amino acids in CSS, PP and BSG were 0.58, 0.39 and 0.18 mg/g, respectively (Table 2). The highest concentration of umami free amino acids (monosodium glutamate-like) were detected in CSS and PP extracts (0.13 mg/g each), where the highest concentration of Glu was detected in PP extracts (0.11 mg/g). CSS, BSG and PP are seldom investigated in the past for their potential utilization as flavour ingredients. The concentrations of total amino acids in CSS, PP and BSG were 85, 42 and 247 mg/g, respectively and the highest concentration of essential amino acids (sum of Lys, Met, His, Leu, Phe, Thr, Val, Ile) were detected in BSG (92 mg/g, Table 2). The remaining essential amino acids, namely Met and Trp were not detected in any samples, likely due to artefactual oxidation during acid hydrolysis employed before the analysis.

Table 2.

Amino acid compositions (mg/g) of various samples

Amino acid (AA)	Coffee silver skin		Potato peels		Brewer’s spent grain
	FAA	TAA	FAA	TAA	FAA	TAA
Taurine	ND	4 ± 0.2	ND	5 ± 0.51	ND	2.7 ± 0.3
Methionine sulfone	ND	1 ± 0.1	ND	0.56 ± 0.05	ND	7 ± 0.5
Aspa	0.068 ± 0.009	9 ± 0.2	0.018 ± 0.003	4 ± 0.34	0.003 ± 0.000	19 ± 0.4
Thr	0.014 ± 0.001	4 ± 0.3	0.006 ± 0.000	2 ± 0.2	ND	8 ± 0.1
Ser	0.012 ± 0.000	5 ± 0.3	0.005 ± 0.000	2 ± 0.2	0.003 ± 0.000	11 ± 1.5
Glua	0.062 ± 0.006	9 ± 0.1	0.111 ± 0.008	4 ± 0.1	0.021 ± 0.003	53 ± 7
Gly	0.012 ± 0.002	6 ± 0.1	0.010 ± 0.001	2 ± 0.3	0.002 ± 0.000	10 ± 0.4
Ala	0.039 ± 0.004	5 ± 0.2	0.025 ± 0.002	2.8 ± 0.03	0.007 ± 0.001	17 ± 0.6
Cys	0.047 ± 0.003	3.4 ± 0.07	0.041 ± 0.002	4.1 ± 0.01	0.016 ± 0.002	2.3 ± 0.04
Val	0.011 ± 0.001	6 ± 0.1	0.016 ± 0.001	3 ± 0.4	0.004 ± 0.000	14 ± 0.6
Met	0.010 ± 0.001	0.01 ± 0.01	0.003 ± 0.000	0.02 ± 0.00	0.001 ± 0.000	0.03 ± 0.00
Ile	0.010 ± 0.000	5 ± 0.2	0.012 ± 0.000	2 ± 0.1	0.001 ± 0.000	11 ± 0.2
Leu	0.001 ± 0.000	6 ± 0.7	0.006 ± 0.000	2 ± 0.2	0.004 ± 0.000	27 ± 1
Tyr	ND	4 ± 0.4	ND	1.9 ± 0.01	ND	10 ± 0.6
Phe	ND	5 ± 0.0	0.011 ± 0.001	1.8 ± 0.04	0.007 ± 0.001	15 ± 1
His	0.104 ± 0.001	4 ± 0.2	0.049 ± 0.002	2.9 ± 0.08	0.042 ± 0.002	8 ± 0.3
Lys	0.010 ± 0.000	3 ± 0.2	0.004 ± 0.000	2 ± 0.19	0.004 ± 0.000	8 ± 0.5
Arg	0.077 ± 0.004	5 ± 0.4	0.021 ± 0.002	2.1 ± 0.04	0.025 ± 0.001	13 ± 0.3
Umami	0.13 ± 0.015	NA	0.13 ± 0.010	NA	0.024 ± 0.003	NA
Sweet	0.077 ± 0.007	NA	0.046 ± 0.004	NA	0.012 ± 0.001	NA
Bitter	0.21 ± 0.008	NA	0.12 ± 0.006	NA	0.084 ± 0.004	NA
Essential AA	NA	34 ± 1.7	NA	15 ± 1	NA	92 ± 4
Conditionally essential AA	NA	17 ± 1	NA	10 ± 0.31	NA	34 ± 1
Nonessential AA	NA	28 ± 1	NA	12 ± 0.73	NA	100 ± 10
Total AA	0.58 ± 0.044	85 ± 4	0.39 ± 0.025	42 ± 3	0.18 ± 0.013	237 ± 16

FAA free amino acids, TAA Total amino acids, FAA are grouped based on their taste characteristics: umami (Asp + Glu); sweet (Thr + Ser + Gly + Ala); bitter (Val + Met + Ile + Leu + Phe + His + Arg + Trp); tasteless (Tyr + Lys + Tau). Essential amino acids: (His + Ile + Leu + Lys + Met + Phe + Thr + Val); Conditionally essential AA: (Arg + Cys + Gly + Tyr); Non-essential amino acids: (Ala + Asp + Glu + Ser). Values are represented as mean ± standard deviation (SD, n = 2). SD of 0.000 indicates an SD below the significant figures shown; ND-not detected, NA-not applicable

^aFor TAA, Asp is expressed as a sum of Asp and Asn, and Glu as sum of Glu and Gln considering the fact of hydrolysis of Asn and Gln to Asp and Glu, respectively, during acid hydrolysis

The amino acid composition of the raw materials and the extracts will provide important information on nutritional quality of the products by revealing content and concentrations of essential and non-essential amino acids. Moreover, free amino acids such as Asp and Glu are well-known in imparting umami flavour in foods (Poojary et al. 2017), which could be useful in employing these phenolic rich extracts as food ingredients. Previous studies have shown that water is an ideal solvent to extract umami free amino acids from plant materials and algae (Poojary et al. 2017, 2019; Hildebrand et al. 2020).

Conclusion

The present work reveals that CSS, PP and BSG are a good source of phenolic compounds and the highest TPC was obtained from CSS. The LC–MS/MS analysis revealed that aqueous extraction released mainly phenolic acids from these raw materials. The UAE enables rapid recovery of TPC (84.5–97.6% in 30 min) than the conventional extraction, possibly due to enhanced cellular destruction caused by ultrasonic treatment as seen by scanning electron microscopy. BSG found to be a good source of essential amino acids. Overall, the present study shows that CSS, PP and BSG extracts could be used a potential source of total polyphenols in the preparation of nutraceuticals and food additives.

Electronic supplementary material

Below is the link to the electronic supplementary material.