Sequential Hydrolysis of Chicken Feathers Composed of Ultrasound and Enzymatic Steps: An Enhanced Protein Source with Bioactive Peptides

: Chicken feather is a massive by-product. Its incorrect disposal can lead to serious environmental impacts. However, chicken feather is a promising low-cost keratin source. Keratin products have a wide application in the food and pharmaceutical industry. Mostly, chicken feathers are hydrolyzed by hydrothermal processes, and then applied into animal feed formulations. Despite the low cost, the hydrothermal hydrolysis leads to uncontrolled and low hydrolysis yield. Therefore, the aim of this work was to develop and optimize a sequential strategy of chicken feathers hydrolysis composed of ultrasound and enzymatic hydrolysis (savinase ® ) steps. In the ﬁrst research step an experimental design was built and the optimum hydrolysis condition was obtained at 50 ◦ C and 12.5% (enzyme/chicken feather), using three integrated rectors containing enzyme/substrate and sodium disulﬁte. Then, the ultrasound probe was added in the experimental apparatus in order to investigate the enzymatic hydrolysis assisted by ultrasound treatment. The enzymatic hydrolysis assisted by ultrasound treatment led to high concentrations of peptides, including a dipeptide (245.1868 m / z ). Thus, the sequential hydrolysis strategy composed by two green technologies proposed in this study, enhanced the degree of hydrolysis of chicken feathers, producing bioactive peptides that can be used as ingredients in food products and other sectors.


Introduction
The incorrect disposal of keratinous residues can lead to serious environmental impacts, since >60% of keratinous residues are disposed in landfills, dumpsites, and incinerators (very slow degradation rate, landscape deterioration, pollution, transmission of disease pathogens, among others) [1].
These residues are, inherently, generated at global scale, for instance ≈1.5 Mt of sheep wool, ≈9.5 Mt of poultry feathers, ≈5 Mt of human hairs, and ≈1 Mt of horns [1]. Thus, poultry feathers are the most abundant keratinous residues (≈40%), of which China, India, the USA, and Brazil are the most representative countries.
According to the USDA, in 2021, the total number of broilers produced was 9.13 billion. Only in Brazil, has chicken production increased ≈4%, reaching a record of 14.4 Mt. This represents 1.17 Mt of chicken feathers (CFs) as a by-product [2].
CFs are composed, essentially, of proteins (≥75), in which ≈90% it is keratin [3]. Keratin, a crystalline protein, can be classified into two general classes based on amino acid composition, distribution, and function: type I (the acidic keratins) and type II (the basic keratins). The acidic keratins subdivide into acidic hard and acidic soft keratins [1]. The average molecular weight of feather proteins is ≈10,000 g·mol −1 (10 kDa) [4].
In order to remove lipids, CFs were washed (tap water), and then immersed in an aqueous solution containing 1 g·L −1 household detergent (linear alkyl benzene sodium sulfonate, among others) for 30 min at 45 • C. Then, CFs were rinsed three times with tap water at 45 • C, and two times with distilled water [17]. CFs were oven-dried at 45 • C for 12 h (SP-400, SPlabor, Presidente Prudente-SP/Brazil) and processed in a knife mill (TE-648, Tecnal, Piracicaba-SP/Brazil). The ground CFs were packaged in a hermetically sealed low-density polyethylene plastic bag and stored at −16 • C.

Alkaline Dissolution of Ground Chicken Feathers
Preliminary experiments with CF dissolution were performed in an integrated system composed of three reactors (300 mL) and thermostatic bath (0214M2, Quimis, Diadema-SP/Brazil) under a magnetic stirrer (1000 rpm) ( Figure 1). Ground CFs ranging from 1.32 to 4.68% (m/v) and sodium sulfite (antioxidant) from 0.05 to 0.11 M were added into reactors, with 100 mL of 0.08 M NaOH solution (pH 12.9) and then, into the mixture for 12 h. The degree of hydrolysis was measured every two hours for 12 h [19,20].

Alkaline Dissolution of Ground Chicken Feathers
Preliminary experiments with CF dissolution were performed in an integrated system composed of three reactors (300 mL) and thermostatic bath (0214M2, Quimis, Diadema-SP/Brazil) under a magnetic stirrer (1000 rpm) ( Figure 1). Ground CFs ranging from 1.32 to 4.68% (m/v) and sodium sulfite (antioxidant) from 0.05 to 0.11 M were added into reactors, with 100 mL of 0.08 M NaOH solution (pH 12.9) and then, into the mixture for 12 h. The degree of hydrolysis was measured every two hours for 12 h [19,20].

Optimization of Enzymatic Hydrolysis
After the alkaline dissolution of ground chicken feathers, an enzymatic hydrolysis was carried out in the same integrated system composed of three reactors (300 mL) and a thermostatic bath (0214M2, Quimis, Diadema-SP/Brazil SP/Brazil) under magnetic stirring (60 rpm) ( Figure 1). The enzyme manufacturer's recommendations (Novozymes) detail the working temperature range from 55 to 75 °C and pH range between 7 and 11. The preliminary experiments with ground chicken feathers achieved a higher hydrolysis yield at 50 °C and pH 10.6. Thus, pH was adjusted to 10.6 with HCl (3N). Alkaline dissolution results were considered for CF concentration (4%).
The enzymatic hydrolysis was monitored by the GH every two hours for 10 h according to Sorensen [18]. The protease savinase ® was thermally inactivated (80 °C for 10 min).

Sequential Hydrolysis Composed of Ultrasound and Enzymatic Steps
It was evaluated for the ultrasound effect on chicken feather hydrolysis, prior to the enzymatic step (optimal condition-previously identified). Thus, after the alkaline dissolution of ground chicken feathers, 50 mL of solution were transferred to a plastic beaker and cooled in an ice bath. The optimal sonication condition was determined by design of experiments 2 2 (Table 5) using a probe sonicator (QR500, Eco-sonics, Indaiatuba-SP/Brazil) at 20 kHz and ultrasonic power between 200 and 442 W, for 5 to 23.1 min. The 4 mm diameter titanium micro-point probe was placed centrally, at 0.5 cm depth into the solution.

Optimization of Enzymatic Hydrolysis
After the alkaline dissolution of ground chicken feathers, an enzymatic hydrolysis was carried out in the same integrated system composed of three reactors (300 mL) and a thermostatic bath (0214M2, Quimis, Diadema-SP/Brazil SP/Brazil) under magnetic stirring (60 rpm) ( Figure 1). The enzyme manufacturer's recommendations (Novozymes) detail the working temperature range from 55 to 75 • C and pH range between 7 and 11. The preliminary experiments with ground chicken feathers achieved a higher hydrolysis yield at 50 • C and pH 10.6. Thus, pH was adjusted to 10.6 with HCl (3N). Alkaline dissolution results were considered for CF concentration (4%).
The enzymatic hydrolysis was monitored by the GH every two hours for 10 h according to Sorensen [18]. The protease savinase ® was thermally inactivated (80 • C for 10 min).

Sequential Hydrolysis Composed of Ultrasound and Enzymatic Steps
It was evaluated for the ultrasound effect on chicken feather hydrolysis, prior to the enzymatic step (optimal condition-previously identified). Thus, after the alkaline dissolution of ground chicken feathers, 50 mL of solution were transferred to a plastic beaker and cooled in an ice bath. The optimal sonication condition was determined by design of experiments 2 2 (Table 5) using a probe sonicator (QR500, Eco-sonics, Indaiatuba-SP/Brazil) at 20 kHz and ultrasonic power between 200 and 442 W, for 5 to 23.1 min. The 4 mm diameter titanium micro-point probe was placed centrally, at 0.5 cm depth into the solution.

Microfiltration of Amino Acids and Peptides
In order to purify amino acids and peptides, the CF hydrolysates were neutralized to pH 7.5 with HCl (5N), and then microfiltrated at 28 • C and 2 bar by using a stainless steel cell under magnetic stirring and 0.2 µm polyvinylidene difluoride membrane (MV020, Microdyn Nadir, São Paulo-SP/Brazil) with area of 1.02 × 10 −3 m 2 , previously hydrated with distilled water for 20 min.

Degree of Hydrolysis
The degree of hydrolysis was measured according to the method of Sorensen [18]. Firstly, the 50 mL of formaldehyde-phenolphthalein solution was prepared, in which 1 mL of 0.05% phenolphthalein (50% ethanol solvent) was added to formaldehyde (CH 2 O 37%). The pH was adjusted to 8.5 with BaOH 2 0.2 N and HCL 0.2 N. Then, 10 mL of formaldehydephenolphthalein reaction solution was mixed with 10 mL of feather hydrolysates also adjusted to pH 8.5. Blank samples were prepared with an equivalent volume of distilled water and neutralized formalin solution. The α-amino nitrogen was measured by Equation (1): The degree of hydrolysis was measured by Equation (2): Degree of hydrolysis (%) = (α-amino nitrogen/total nitrogen) × 100 Total nitrogen was determined by Kjeldahl method, 928.08 [18].

Ultrasound Effect on Particle Size Distribution of Dissolved Feathers
The ultrasound (50 mL; 450 W; 12.5 min) effect on the particle size was analyzed by Dispersion Analyser Lumisizer ® (LS 611, L.U.M. GmbH, Jiangsu province/China) 25 • C, 4000 rpm.

Peptide Profile
The keratin hydrolysates peptide profile (molecular mass, distribution of peptides: di-, tri-and oligopeptides) was determined by hydrophobic interaction liquid chromatography coupled to mass ionization by nebulization and qTOF analyzer (HILIC-ESI-qTOF-MS), detailed below: We used an ACQUITY UPLC system class H (Waters, Milford, MA, USA) equipped with a photodiode array (PDA) detector, sample manager, and a quaternary solvent manager. A HILIC column (100 mm, 2.1 mm, particle size 2.7 µm (Supelco, Sigma-aldrich, Jurubatuba-SP/Brazil) was used for the separation. The column and the sample tray were maintained at temperatures of 40 • C and 20 • C, respectively. The sample (3 µL) was injected and separated with a gradient condition at a flow rate of 0.4 mL/min. Solvent A was composed of acetonitrile (ACN) and H 2 O (90:10) and 0.1% of formic acid (v/v); B was composed of H 2 O and ACN (90:10). The elution was performed using the following gradient condition: 0-2 min, 95% of A; 2-10 min, 55% of A; 10-15 min, 5% of A; 15-20 min, 95% of A. A Xevo G2-SQTof (Waters, Wexford/Ireland) bearing an electrospray ionization (ESI) probe operating in positive and negative ionization modes coupled to the UPLC device was used to detect the chemical components of each sample. Nebulizer gas: nitrogen; cone gas flow: 100 L/h; desolvation gas flow: 900 L/h; sampling cone 40 V; source offset 80 V; collision gas: argon; lockspray reference sample was leucine encephalin with reference masses at m/z 554.2615 (ESI−). The desolvation and the ionization source were maintained during the analyses at 250 • C and 90 • C, respectively, while the capillary voltage was 3 kV. A range of 25 to 35 eV was used as the collision energy. Data were acquired in a range of 100-1500 Da, at a scan time of 1.0 s over 20 min and were processed with MassLynx V4.1 (Waters, Wexford/Ireland). Molecular formulas were determined by calculation using MassLynx's elemental composition tool. The molecular formula was restricted by a tolerance of 5 ppm between the calculated and the measured mass values.

Biological Properties of Chicken Feathers Hydrolysate Digestibility
The digestibility of CF hydrolysates in vitro was carried out according to Tiwary and Gupta [5]. Briefly, hydrolysates were diluted into HCl 0.1 N, then pepsin was also added (2 mg/mL). It was incubated at 37 • C for 2 h. After the acid digestion, it was alkalized with NaOH 2 N, then trypsin was added (2 mg/mL) and incubated at 37 • C for 16 h. All samples were centrifuged 6000× g and the supernatants measured at 660 nm (bovine serum albumin curve).

Antioxidant Activity
Antioxidant activities of hydrolysates were measured by capturing free radical methods (DDPH and ABTS) [21,22].

Statistical Analyses
Statistical analyses were carried out using design of experiments (DOE), according to Rodrigues and Lemma [23], at 10% significance level by using Statistica 7.0 software.

Proximate Composition of Feathers
Proximate composition of feathers is presented in Table 1, where it can be observed that the feathers are composed mainly of proteins, around 84%. The centesimal composition of CFs (Table 1) was similar to that reported by Fakhfakh et al. [24] who obtained 85.31% ± 0.43 of protein. It is noteworthy that CF protein content is remarkably high ≥75%, of which ≈90% out of these proteins are keratin [25].

Alkaline Dissolution of Ground Chicken Feathers
The results of preliminary dissolution experiments showed that the highest concentration of ground chicken feathers (4% m/v) led to higher degrees of hydrolysis. On the other hand, sodium sulfite tended to be more efficient at lower concentrations (0.06 M). The disulfide bonds, widely found in chicken feather proteins, can be cleaved by sodium sulfite [19]. Thus, it is very likely that sodium sulfite affects savinase ® activity, which aligns with research published by Adler et al. [26], who investigated the hydrolysis of CFs by thermal treatment (2 bar, 121 • C, 15 min) with sodium sulfite and/or sodium hydroxide. The authors concluded that sodium hydroxide, sodium sulfite, and thermal treatment improved the hydrolysis of CFs. It should be noted that these molecules have complex interactions and different effects, which can, in principle, counteract each other, such as changed bonding, increased/decreased crosslinking, changes in polarity, and network disturbance by additives. In addition, as expected, high temperatures (80 to 97 • C) also produced higher degrees of hydrolysis. However, thermal treatments induce racemization and amino acid oxidation-poor nutritional quality [2,14,[27][28][29][30].
Therefore, the following condition of alkaline dissolution was chosen for the enzymatic hydrolysis (4% substrate, 70 • C, and 0.06 M sodium sulfite for 8 h).

Optimization of Enzymatic Hydrolysis of Chicken Feathers
Design of experiments 2 2 was used to investigate the optimal condition of enzymatic hydrolysis ( Table 2). It is worth noting that the degree of hydrolysis was measured every two hours. Thus, five dependent variables were obtained. The regression coefficients and analysis of variance (ANOVA) for response variables are described in Table 3. Regarding the 10 h statistical model, it presented significance and predictability. In addition, it showed the highest R 2 . Hence, based on statistical analysis (significance, predictability, and R 2 ) the response surface was plotted-10 h (Figure 2). Then, a second order coded model was elaborated by using the Statistica 7.0 software, Equation (3) Clearly, the optimal condition of enzymatic hydrolysis was reached at center points: 12.5% enzyme/substrate ratio and 50 • C, and a very low standard deviation was observed (s = 0.58%). The Kinect analyses showed that at 6 h, the degree of hydrolysis reached a stationary hydrolysis. Hence, the optimal condition of enzymatic hydrolysis can be carried out faster, for 6 h instead of 10 h.
According to Nyo and Nguyen [31] and Uluko et al. [32], a hydrolysis degree around 8% is quite desirable, since higher hydrolyzation can produce free amino acids instead of peptides, for instance 14.4% salmon by-products [33]; 5.07% for milk protein concentrate [32]; 12.5% for peanut protein [31]. Therefore, the optimal condition of enzymatic hydrolysis is related to center points (design of experiments), which leads to a suitable hydrolysis degree (≈8%) after 6 h.
Biomass 2022, 2, FOR PEER REVIEW 7 Hence, based on statistical analysis (significance, predictability, and R 2 ) the response surface was plotted-10 h (Figure 2). Then, a second order coded model was elaborated by using the Statistica 7.0 software, Equation (3) Clearly, the optimal condition of enzymatic hydrolysis was reached at center points: 12.5% enzyme/substrate ratio and 50 °C, and a very low standard deviation was observed (s = 0.58%). The Kinect analyses showed that at 6 h, the degree of hydrolysis reached a stationary hydrolysis. Hence, the optimal condition of enzymatic hydrolysis can be carried out faster, for 6 h instead of 10 h.
According to Nyo and Nguyen [31] and Uluko et al. [32], a hydrolysis degree around 8% is quite desirable, since higher hydrolyzation can produce free amino acids instead of peptides, for instance 14.4% salmon by-products [33]; 5.07% for milk protein concentrate [32]; 12.5% for peanut protein [31]. Therefore, the optimal condition of enzymatic hydrolysis is related to center points (design of experiments), which leads to a suitable hydrolysis degree (≈8%) after 6 h.

Sequential Hydrolysis Composed of Ultrasound and Enzymatic Steps
In order to enhance the hydrolysis of CFs, a sequential hydrolysis of CFs composed by ultrasound and enzymatic steps was performed. The application of ultrasound (cavitation) on proteins, such as keratin, affects their hydration, molecular size, hydrophobicity, and conformation [15,16]. Thus, the ultrasound treatment, prior to enzymatic hydrolysis, can both increase exposure of protein groups and partially hydrolyze proteins. Thus, obviously, these effects favor the action of the enzyme. These effects are particularly desirable, since savinase (savinase ® ) is an endoprotease, that is, a slight exposition of keratin can significantly improve enzymatic activity. In this sense, the analysis of particle size (Table 4) indicated that ultrasound treatment reduced the particle size. The control samples showed an average particle size of 195 nm, whereas the sample treated with ultrasound presented lower average particle size (≈155 nm). Thus, the ultrasound treatment hydrolyzed partially, and/or destabilized protein aggregates, and/or changed the protein conformation.

Sequential Hydrolysis Composed of Ultrasound and Enzymatic Steps
In order to enhance the hydrolysis of CFs, a sequential hydrolysis of CFs composed by ultrasound and enzymatic steps was performed. The application of ultrasound (cavitation) on proteins, such as keratin, affects their hydration, molecular size, hydrophobicity, and conformation [15,16]. Thus, the ultrasound treatment, prior to enzymatic hydrolysis, can both increase exposure of protein groups and partially hydrolyze proteins. Thus, obviously, these effects favor the action of the enzyme. These effects are particularly desirable, since savinase (savinase ® ) is an endoprotease, that is, a slight exposition of keratin can significantly improve enzymatic activity. In this sense, the analysis of particle size (Table 4) indicated that ultrasound treatment reduced the particle size. The control samples showed an average particle size of 195 nm, whereas the sample treated with ultrasound presented lower average particle size (≈155 nm). Thus, the ultrasound treatment hydrolyzed partially, and/or destabilized protein aggregates, and/or changed the protein conformation. Thus, very likely, the sequential hydrolysis of chicken feathers using ultrasound followed by enzymatic hydrolysis can be applied to reach a higher degree of hydrolysis associated to the peptides production. Then, a design of experiments 2 2 (Table 5) was used to evaluate the optimal condition of sequential hydrolysis of chicken feathers using ultrasound and enzymatic hydrolysis.
It is worth noting that the degree of hydrolysis was measured every two hours. Thus, four dependent variables were obtained. The regression coefficients and analysis of variance (ANOVA) for response variables are described in Table 6, in which all dependent variables were significant and predictive models. Nevertheless, the 8 h statistical model showed the highest R 2 . Hence, based on statistical analysis (significance, predictability and R 2 ) the response surface was plotted-8 h (Figure 3). Then, a second order coded model was elaborated by using the Statistica 7.0 software Equation (4) Therefore, the sequential hydrolysis with ultrasound and enzyme reached a higher degree of hydrolysis (when compared to enzymatic hydrolysis without ultrasound). The eco-friendly ultrasound treatment should be integrated to the enzymatic hydrolysis of chicken feathers, since a higher degree of hydrolysis can be reached.
Hence, based on statistical analysis (significance, predictability and R 2 ) the respon surface was plotted-8 h (Figure 3). Then, a second order coded model was elaborated using the Statistica 7.0 software Equation (4)  Therefore, the sequential hydrolysis with ultrasound and enzyme reached a high degree of hydrolysis (when compared to enzymatic hydrolysis without ultrasound). T eco-friendly ultrasound treatment should be integrated to the enzymatic hydrolysis chicken feathers, since a higher degree of hydrolysis can be reached.

Hydrolyzate Microfiltration
For hydrolyzate microfiltration on 0.2 µm membranes, a total flow of ≈2.93 L·h·m was obtained. Figure 4 shows the enzymatic hydrolyzate before and after microfiltrati The hydrolyzate was composed of ultrasound and enzymatic steps and had the sa physical appearance. This operation was successful, obtaining purified hydrolyzates th were used in the analyses, including the peptide profile.

Hydrolyzate Microfiltration
For hydrolyzate microfiltration on 0.2 µm membranes, a total flow of ≈2.93 L·h·m −2 was obtained. Figure 4 shows the enzymatic hydrolyzate before and after microfiltration. The hydrolyzate was composed of ultrasound and enzymatic steps and had the same physical appearance. This operation was successful, obtaining purified hydrolyzates that were used in the analyses, including the peptide profile. Therefore, the sequential hydrolysis with ultrasound and enzyme reached a higher degree of hydrolysis (when compared to enzymatic hydrolysis without ultrasound). The eco-friendly ultrasound treatment should be integrated to the enzymatic hydrolysis of chicken feathers, since a higher degree of hydrolysis can be reached.

Hydrolyzate Microfiltration
For hydrolyzate microfiltration on 0.2 µm membranes, a total flow of ≈2.93 L·h·m −2 was obtained. Figure 4 shows the enzymatic hydrolyzate before and after microfiltration. The hydrolyzate was composed of ultrasound and enzymatic steps and had the same physical appearance. This operation was successful, obtaining purified hydrolyzates that were used in the analyses, including the peptide profile.

Peptide Profile
Peptides are a diverse group of oligomeric structures, composed of short amino acid sequences, usually 2-20 residues. Peptides have remarkable biological properties, such as hormonal regulation, redox homeostasis, neuronal signal, cell signaling, transduction, growth, and immune response [34,35].
In mass spectra ( Figures 5 and 6), the peaks with good resolution were selected, seven main ones for enzymatic hydrolyzate and six for hydrolyzate composed of ultrasound and enzymatic steps.
Peptides are a diverse group of oligomeric structures, composed of short amino acid sequences, usually 2-20 residues. Peptides have remarkable biological properties, such as hormonal regulation, redox homeostasis, neuronal signal, cell signaling, transduction, growth, and immune response [34,35].
In mass spectra ( Figures 5 and 6), the peaks with good resolution were selected, seven main ones for enzymatic hydrolyzate and six for hydrolyzate composed of ultrasound and enzymatic steps.   growth, and immune response [34,35].
In mass spectra ( Figures 5 and 6), the peaks with good resolution were selected, seven main ones for enzymatic hydrolyzate and six for hydrolyzate composed of ultrasound and enzymatic steps.    Table 7 presents the peptide sizes that were obtained from chicken feather hydrolyzates by enzymatic hydrolysis and sequential hydrolysis composed of ultrasound and enzymatic steps. It is worth noting that the analytical method HILIC-ESI-qTOF-MS is quite sensitive to peptides composed of up to 20 amino acids [36]. Both hydrolyzates showed a similar peptide profile, however, the sequential hydrolysis composed of ultrasound and enzymatic steps showed higher intensity, which is related to the concentration of peptides. In addition, the sequential strategy produced the smallest peptide (dipeptide). Thus, it was proved that ultrasound-treatment affected the enzymatic hydrolysis. Further investigation should be carried out on the identification of amino acid sequences, purification strategies, mainly membrane-based technology, and application of these peptides.

Biological Properties of Chicken Feathers Hydrolyzate
CFs have low digestibility (9.6-15.6%) [5,26]. When compared to the enzymatic hydrolysis, the sequential hydrolysis composed of ultrasound and enzymatic steps showed higher digestibility, ≈80 and 100%, respectively (Table 8). These results are higher than chicken feather hydrolyzates that are commercially available ≈56% and closer to feather protein hydrolyzate (86%) produced by Kshetri et al. [26,37], considering in vitro digestibility. It is worth noting that further experiments can be complementary to these in vitro protein digestibility trends, such as an approach with digestive enzymes. In addition, very often, there is a discrepancy between in vitro and in vivo results.

Hydrolyzate Digestibility (%)
Enzymatic hydrolysis 80.23 a ± 0.20 Sequential hydrolysis (ultrasound and enzymatic steps) 100.00 b ± 0.09 n = 3 repetitions per treatment; it means followed by different superscript lowercase letters indicate a significant difference between treatments at a 5% level by Tukey's test.

Conclusions
An environmentally friendly promising methodology for the production of chicken feather peptides was achieved. Thus, this approach is an interesting alternative to the valorization of an abundant by-product. When compared to enzymatic hydrolysis, the strategy using a sequential hydrolysis with ultrasound and enzymatic steps reached a higher degree of hydrolysis, higher digestibility, and higher antioxidant activity. The sequential hydrolysis also produced unique dipeptides. Further investigation should be carried out on the identification of amino acids sequences, purification strategies, mainly membrane-based technology, and the biological application of the chicken feather peptides. It is worth noting that a similar strategy could be applied to other by-products (residue) composed of high protein content. The chicken feather peptides could be applied, mainly, into food, cosmetic, and pharmaceutical formulations.