PROMANCOA Modular Technology for the Valorization of Mango ( Mangifera indica L.) and Cocoa ( Theobroma cacao L.) Agricultural Biowastes

: PROMANCOA modular technology (PMT) aims at the development of modular agricultural biowaste valorization of mango ( Mangifera indica L.) and cocoa ( Theobroma cacao L.) cultivars within the concept of circular economy in agriculture management. The modular design includes four modules: (1) green raw material (GRM) selection and collection, (2) GRM processing, (3) GRM extraction, in order to obtain bioactive green extracts (BGE) and bioactive green ingredients (BGI), and (4) quality control, which lead to formula components for food, feed, nutraceutical and/or cosme-ceutical products. PMT was applied to mango stem bark and tree branches, and cocoa pod husk and bean shells, from cultivars of mango and cocoa in provinces of the Dominican Republic (DR). PMT might be applied to other agricultural biowastes, where a potential of value-added BGE/BGI may be present. Alongside the market potential of these bioactive ingredients, the reduction of carbon dioxide and methane emissions of agricultural biowastes would be a signiﬁcant contribution in order to reduce the greenhouse effect of these residuals.


Introduction
The goal of circular economy (CE) is the recovery and valorization of waste materials in order to re-use it as value-added products and reduce the environmental impact of wastes by using renewable sources when possible [1]. This technological development is creating new business models and employment opportunities with a positive impact on social development [2]. Several technologies have been developed within the CE concept in order to re-use plastics [3], textiles [4], fruit-industrial by-products [5], and agricultural products [6], among others. The present manuscript describes the development of a modular technology for the utilization of agricultural biowastes from mango (Mangifera indica L.) and cocoa (Theobroma cacao L.) in the Dominican Republic (DR), leading to value-added products for cosmetic, food, and pharma industries named PROMANCOA (products from mango and cocoa).
The DR has large plantations of mango and cocoa, which produce a considerable amount of waste, such as tree branches (mango), during the annual prune period, and pod husks and shells (cocoa), which are produced in two harvest periods (spring and autumn). Mango branch trees (MBT) and cocoa pod husks (CPH) are usually left in the field as organic matter to the soil, while cocoa bean shells (CBS) might be incinerated or used as a food additive. Previous research has demonstrated that mango stem bark (MSB) can be collected, without affecting the tree phytosanitary condition, having attractive antioxidant and anti-inflammatory properties, which may lead to several innovative food, cosmetic, and pharmaceutical formulations [7]. The antioxidant effect of cocoa is one of the main advantages for the consumers, and it is the major claim of chocolate producers [8].

Cocoa
Two farms were selected for cocoa GRM collection in Villa Altagracia and San Francisco de Macoris, DR ( Table 2). The collected CPH were around four tons (three tons from green-yellowish and one ton from red-orange pods) and 810 kg of CBS (630 kg from green-yellowish pods and 180 kg from red-orange pods). SOP about GRM collection from CPH was filled in the UNEV Quality Assessment Department [24]. CBS were collected at the cocoa factory directly, without an SOP, but considered the variety origin. Although the difference between green-yellowish and red-orange pod husks has not been relevant for the cocoa industry, we assumed that such color difference might influence the antioxidant activity of subsequent cocoa extracts. The GRM water content from MSB was reduced from 42.7% to 11.8% with solar drying, and from 11.8% to 5.8% with rotary hot air drying according to results shown in Table 3. The MSB from Baní had a significantly higher water content than the one from Puerto Plata, which corroborated the importance of geographical origin for the GRM yield. However, GRM from both origins had similar water content after the drying operation units, which allowed for the subsequent standardization of the following bioactive green extract (BGE) extraction. MBT, which was allowed to solar dry in the field for a number of weeks, was directly submitted to hot air drying due to its low water content (around 10%). MBT yields were considerably higher compared Processes 2021, 9, 1312 4 of 18 to MSB, as shown in Table 4. A significant lower yield of MSB was obtained in the Baní farm as compared to the Puerto Plata farm; • Cocoa: The CPH water content was reduced from an average of 84.5% to 14.6% after solar drying, and from 14.6% to 5.5% after rotary hot air drying, as shown in Table 5. No significant differences were observed between both geographical origins and between different colored pods. The GRM from all origins and varieties had a similar water content for subsequent extraction. The average yield of dried CPH was around 6%.

GRM Milling and Pelletizing
The physical characterization of GRM pellets for both mango and cocoa are shown in Table 6. It was not necessary to use binders to achieve the consolidation of the pellets, because of the presence of lignin as a natural binder in the GRM, which contributed to the consolidation of the pellets under our experimental conditions. Figure 1 shows photos of the pellets obtained from each GRM.
consolidation of the pellets under our experimental conditions. F the pellets obtained from each GRM.

GRM Extraction
• Mango: BGE yields from MSB were higher than MBT for bo Tommy Atkins). The best yield was obtained from the Hade farm, with 12%. The lowest yield was obtained for MBT f from the Azua farm, however, this GRM had the highest po and, therefore, the highest mangiferin (MF) content. MF con fraction ranged from 41.6% (MBT Tommy Atkins variety) to riety). The higher MF contents were found for the Haden v MBT. All the BGE had similar water content. Results of th quality control are shown in Table 7. Microbiological quality successfully for all lots (results not shown);      Table 7. Microbiological quality control was performed successfully for all lots (results not shown); • Cocoa: BGE yields from CBS are shown in Table 8. Since no significant differences were observed in the PPC values for both geographical origin and pod color, the samples were homogenized in order to run the pilot plant batches. Due to the low yields of CPH (below 6%), BGE extraction from this GRM was not performed. Pellets from the CPH will be used for feed formulations and the results will be shown elsewhere. CBS-PPC was around 28 g/100 g and approximately 50% of its content (13.9 g) was a mixture of five different types of OPA.  The results of the HPLC-DAD analysis of the MSB-BGE and MBT-BGE for the Haden variety are shown in Table 9. The components were identified per retention times as referred to pure standards. Gallic acid (0.7-1.3%), methyl gallate (1.6-1.8%), propyl gallate (1.7%), catechin (5.0-7.4%), epicatechin (4.4-6.3%), MF (76.2-77.5%), iso-MF (5.2-6.9%), and quercetin (0.8-1.9%) could be identified in BGE from both mango GRM (Haden variety). The MF content ranged from 10.5 g/100 g d.w. (MBT) to 12.0 g/100 g d.w. (MSB). In terms of MF content, the Tommy Atkins variety had lower concentrations than the Haden variety (results not shown). The MF extracted from both MSB and MBT were identical qualitatively, in terms of proportion of MF:isoMF. We could obtain pure MF (90.3% purity, HPLC), with 80% yield from BGE.

Cocoa
The results of the HPLC-DAD analysis of the CBS-BGE and OPA-rich extract are shown in Table 10. Catechin, 9.2 ± 2.2 g/100 g, and epicatechin, 18.4 ± 2.6 g/100 g were the main single BGE polyphenols. The proportions of oligomers in the OPA-rich extract were estimated by internal normalization using catechin as the internal standard as follows: 12% dimer, 19% trimer, 22% tetramer, 25% pentamer, and 19% hexamer, with retention times between 31 and 40 min in our experimental conditions. These oligomers could be identified per comparison of similar retention times with reported results [21].

PROMANCOA Technology Development for Mango Agricultural Biowastes
The GRM collection and storage following an SOP is crucial for the quality specifications of the end-product after its processing. Variables, such as land geographic site, soil type, part to be collected (i.e., fruit, peel, kernel, SB or BT), and procedure for collection, must be considered for the adequate elaboration of SOP. PROMANCOA modular technology (PMT) considered two GRM for standardization, SB and BT. The mango GRM from the Haden variety was better than the Tommy Atkins variety in terms of PPC and MF; therefore, future commercial exploitation of antioxidant-rich extracts from mango GRM should consider this variety difference. Mango variety may affect these parameters as was demonstrated i.e., for its mineral content [25]. Another factor to be considered for industrial production is the GRM potential to be collected for extraction. MBT could be collected in higher amounts than MSB, which has additional difficulties for its collection, including risks of tree damage.
The low density of mango GRM was a problem for handling and storage, and therefore procedures for producing mango GRM pellets was needed. We could obtain a GRM volume reduction of 18% for both MSB and MBT with a low water content (5%), which facilitated the subsequent extraction process. These wood-derived pellets were similar physically to pellets from bamboo and pine stem bark [26].
Phenolic antioxidants of mango GRM were studied being that MF is the main component of the polyphenol fraction (7 g/100 g d.w.) [27]. The presence of iso-MF in purified MF was reported in several Chinese mango varieties [28]. MF purity could not be increased above 90% by further recrystallizations, which may be done by preparative chromatography with macroporous resins as reported elsewhere [29]. Structures of both isomers are highly similar (see Figure 2), and therefore its separation is time-consuming.

PROMANCOA Technology Development for Mango Agricultural Biowastes
The GRM collection and storage following an SOP is crucial for the quality specifications of the end-product after its processing. Variables, such as land geographic site, soil type, part to be collected (i.e., fruit, peel, kernel, SB or BT), and procedure for collection, must be considered for the adequate elaboration of SOP. PROMANCOA modular technology (PMT) considered two GRM for standardization, SB and BT. The mango GRM from the Haden variety was better than the Tommy Atkins variety in terms of PPC and MF; therefore, future commercial exploitation of antioxidant-rich extracts from mango GRM should consider this variety difference. Mango variety may affect these parameters as was demonstrated i.e., for its mineral content [25]. Another factor to be considered for industrial production is the GRM potential to be collected for extraction. MBT could be collected in higher amounts than MSB, which has additional difficulties for its collection, including risks of tree damage.
The low density of mango GRM was a problem for handling and storage, and therefore procedures for producing mango GRM pellets was needed. We could obtain a GRM volume reduction of 18% for both MSB and MBT with a low water content (5%), which facilitated the subsequent extraction process. These wood-derived pellets were similar physically to pellets from bamboo and pine stem bark [26].
Phenolic antioxidants of mango GRM were studied being that MF is the main component of the polyphenol fraction (7 g/100 g d.w.) [27]. The presence of iso-MF in purified MF was reported in several Chinese mango varieties [28]. MF purity could not be increased above 90% by further recrystallizations, which may be done by preparative chromatography with macroporous resins as reported elsewhere [29]. Structures of both isomers are highly similar (see Figure 2), and therefore its separation is time-consuming. MF has received considerable attention not only for its antioxidant properties, but as a potential drug candidate for degenerative diseases [30], alongside its benefits as a nutraceutical supplement [31,32]. We demonstrated recently that isolated MF may not be the main responsible party for the antioxidant activity from mango extracts, but, rather, several MF derivatives, such as glucomangiferin and MF-galloyl and benzoyl derivatives [33]. The contribution of bioactive volatile compounds in mango extracts, such as terpenoids (eudesmane type), single phenols, and some steroids, should also be considered when analyzing the possible contribution of mango extract components to its antioxidant and other biological effects [34].
Classical solid-liquid extraction techniques similar to PMT have been reported for the valorization of mango fruit peel and seed kernel [35,36], but novel technologies, such as pressurized liquid extraction, supercritical fluid extraction, and sub-critical solvent extraction were also used for obtaining BGE from mango by-products [37]. The main advantage for PMT is the accessibility to equipment and processes that were established MF has received considerable attention not only for its antioxidant properties, but as a potential drug candidate for degenerative diseases [30], alongside its benefits as a nutraceutical supplement [31,32]. We demonstrated recently that isolated MF may not be the main responsible party for the antioxidant activity from mango extracts, but, rather, several MF derivatives, such as glucomangiferin and MF-galloyl and benzoyl derivatives [33]. The contribution of bioactive volatile compounds in mango extracts, such as terpenoids (eudesmane type), single phenols, and some steroids, should also be considered when analyzing the possible contribution of mango extract components to its antioxidant and other biological effects [34].
Classical solid-liquid extraction techniques similar to PMT have been reported for the valorization of mango fruit peel and seed kernel [35,36], but novel technologies, such as pressurized liquid extraction, supercritical fluid extraction, and sub-critical solvent extraction were also used for obtaining BGE from mango by-products [37]. The main advantage for PMT is the accessibility to equipment and processes that were established in the classical extraction industry. However, novel environmentally friendly technologies show several advantages in terms of selectivity and recovery efficiency [38].

PROMANCOA Technology Development for Cocoa Agricultural Biowastes
PMT considered one GRM for standardization (CPH), since CBS was collected directly from the cocoa factory using the standard industrial technology for cocoa production. The time between CPH collection and drying is critical because of its possible microbial contamination due to its high water content (above 85%). In our case, CPH was collected carefully for research purposes and processed within 24 h after collection, but considering its possible utilization for industrial processing it would be difficult to standardize. Therefore, CPH technological development was studied only up to pelletization for animal feeding, and animal trials have to be designed in the future to prove its utility [39]. The very low CPH density (around 200 kg/m 3 ) was a problem for handling and storage, and we could increase that value up to 652 kg/m 3 through pelletization. The main use reported for CPH pellets has been as biofuel [40,41], although problems in boilers may be present due to its high mineral content [42].
CPH has not been studied extensively as a source of antioxidant extracts [43]. OPA content in the exocarp of this GRM was reported to be between 8 mg/g and 170 mg/g [44]. The PPC in the pericarp was reported between 12.4 mg/g and 13.6 mg/g, using supercritical CO 2 extraction [45], which is considerably lower than those values in the exocarp, and the one reported in the present work for CBS-BGE (18.1 ± 0.2 mg/g). The main polyphenolic components, which were identified in dried CPH, were similar to those of CBS (quercetin, catechin, epicatechin, and gallic acid) [46]. The most extensive uses of CPH are as an organic fertilizer, for animal feeding, and as a source of renewable energy [47].
The single polyphenols and oligomers found in CBS-OPA-rich extract were quercetin, catechin, and epicatechin. The chemical structures of polyphenols and OPA are shown in Figure 3. Oligomers may be chains of catechin and epicatechin or its combinations, and were assigned by comparison of similar retention times with standards.
Processes 2021, 9, x FOR PEER REVIEW 9 of 18 in the classical extraction industry. However, novel environmentally friendly technologies show several advantages in terms of selectivity and recovery efficiency [38].

PROMANCOA Technology Development for Cocoa Agricultural Biowastes
PMT considered one GRM for standardization (CPH), since CBS was collected directly from the cocoa factory using the standard industrial technology for cocoa production. The time between CPH collection and drying is critical because of its possible microbial contamination due to its high water content (above 85%). In our case, CPH was collected carefully for research purposes and processed within 24 h after collection, but considering its possible utilization for industrial processing it would be difficult to standardize. Therefore, CPH technological development was studied only up to pelletization for animal feeding, and animal trials have to be designed in the future to prove its utility [39]. The very low CPH density (around 200 kg/m 3 ) was a problem for handling and storage, and we could increase that value up to 652 kg/m 3 through pelletization. The main use reported for CPH pellets has been as biofuel [40,41], although problems in boilers may be present due to its high mineral content [42].
CPH has not been studied extensively as a source of antioxidant extracts [43]. OPA content in the exocarp of this GRM was reported to be between 8 mg/g and 170 mg/g [44]. The PPC in the pericarp was reported between 12.4 mg/g and 13.6 mg/g, using supercritical CO2 extraction [45], which is considerably lower than those values in the exocarp, and the one reported in the present work for CBS-BGE (18.1 ± 0.2 mg/g). The main polyphenolic components, which were identified in dried CPH, were similar to those of CBS (quercetin, catechin, epicatechin, and gallic acid) [46]. The most extensive uses of CPH are as an organic fertilizer, for animal feeding, and as a source of renewable energy [47].
The single polyphenols and oligomers found in CBS-OPA-rich extract were quercetin, catechin, and epicatechin. The chemical structures of polyphenols and OPA are shown in Figure 3. Oligomers may be chains of catechin and epicatechin or its combinations, and were assigned by comparison of similar retention times with standards.  CBS extracts showed higher levels of procyanidin oligomers, such as tetramers, pentamers, and hexamers, than grape seed extract [48]. However, concerning monomeric flavan-3-ols, such as catechin or epicatechin, grape seed extract has the highest proportion. In terms of dimer proanthocyanidins (B-type and gallate derivatives), no differences were observed [21]. Other authors [49] reported that the CBS-PPC (above 200 mg/g) is much higher than in cocoa powder obtained from cocoa beans (max. 60 mg/g). The most abundant reported flavonoids in CBS were catechin and epicatechin, between 6 mg/g and 17 mg/g, respectively, and these values depend on the extraction conditions (i.e., solvents and temperature) and geographical origin [50]. Our CBS extract had significantly lower CBS extracts showed higher levels of procyanidin oligomers, such as tetramers, pentamers, and hexamers, than grape seed extract [48]. However, concerning monomeric flavan-3-ols, such as catechin or epicatechin, grape seed extract has the highest proportion. In terms of dimer proanthocyanidins (B-type and gallate derivatives), no differences were observed [21]. Other authors [49] reported that the CBS-PPC (above 200 mg/g) is much higher than in cocoa powder obtained from cocoa beans (max. 60 mg/g). The most abundant reported flavonoids in CBS were catechin and epicatechin, between 6 mg/g and 17 mg/g, respectively, and these values depend on the extraction conditions (i.e., solvents and temperature) and geographical origin [50]. Our CBS extract had significantly lower values of catechin (2.1 ± 0.5 g/100 g) and epicathechin (11.1 ± 0.4 g/100 g) as compared with those previous reports, probably due to its major condensation into oligomers, which were comparatively higher in our cocoa GRM (13.9 ± 0.2 g/100 g) as compared with CBS collected from cocoa of different geographic origins [51].
Several researchers have tried to improve CBS-PPC by using different techniques such as supercritical solvent extraction [52], pulsed electric fields [53], or ultrasound techniques [54], which are not common industrial procedures and require high investment. The high PPC and the high-molecular-weight OPA content found in Dominican CBS deserves further research considering not only the extraction technique, but the type of cocoa fermentation and drying solar temperature. Nevertheless, it seems that PMT offers an attractive alternative in terms of producing OPA-rich extracts for the food and pharmaceutical industries.

PROMANCOA Technology for Other Agricultural Biowastes
The discovery of new and sustainable approaches for growing foods and efficiently utilizing biomass resources is a global priority, with the human population on Earth predicted to reach 9 billion in 2050. To achieve this, more sustainable farming practices need to be developed for food production. There is currently a significant change occurring in consumer purchasing behavior worldwide. Foods that are perceived as healthy or marketed as health-promoting are attracting increasing demand [55]. Agricultural commodities do produce a huge amount of waste, termed as agricultural waste or biomass. Innovation in managing such a vast amount of agricultural waste or biomass is a continuous challenge and recent trends favor the utilization of this biomass for value-added purposes to fulfill needs in areas such as renewable energy, fiber composites and textiles, food alternatives, and livestock feed [56]. Therefore, the development of a more integrated approach to resource management, based on sustainable strategies along the whole supply chain (to valorize residues, by-products, and wastes), is essential.
The public perception regarding utilization of agricultural biowastes as raw material for producing food and food supplements depends mainly on two factors: (i) whether genetic material is involved or not, and (ii) the particular cultural attitude to technological development. Compared with other food hazards, genetically engineered food was perceived (81%) as a moderately severe risk, or very unknown risk [57]. When genetic manipulations are not involved in food production, new technologies are perceived (>80%) as beneficial for human consumption and the environment [58]. Most of the studies that were performed on public perception of new food technologies, genetically modified or not, were conducted in developed nations, but this perception changes dramatically when analyzing the perception in non-developed countries. Generally, there is a positive perception towards new food technologies in developing nations, where more urgent needs in terms of food availability and nutritional content are present. Additionally, perceived levels of risk may be smaller due to trust in government, positive perceptions of science, and positive media influences [59].
Antioxidants are known to be present in many plant-derived products, and are accepted worldwide as food supplements, which are legally defined as foodstuffs with the purpose to supplement the normal diet with a nutritional or physiological effect, alone or in combination. Novel food ingredients concerns the placing on the market of foods and food ingredients, including food plant supplements, which have not been used for human consumption to a significant degree, must not present a danger for the consumer, mislead the consumer, and differ from foods or food ingredients that they are intended to replace to such an extent that their normal consumption would be nutritionally disadvantageous for the consumer [60]. Mango and cocoa agricultural production fulfill the requisites previously mentioned as acceptable sources for producing novel ingredients for food, cosmetics, and pharmaceutical industries, and therefore our results are a step forward in introducing new BGE into these markets. • Stem was marked with a chalk at 25 cm below the lowest tree branch and 25 cm above the highest root. Using a compass, the collection was performed first in the stem side looking to the north, and in the south afterwards; • With the use of a manual circular saw (diameter 5 cm) bark was marked, with not more than 2 cm depth, making a rectangle of around 10 cm width and up to 50 cm height, depending on the tree size; •

Materials and Methods
The bark was separated without damaging the stem, with specially designed tools, in clockwise direction, first from the north side, and after that from the south side; • MSB pieces were cleaned manually from dust and residues and collected in sealed polystyrene bags, approximately 50 kg per bag, and identified with a card indicating location, variety, date of collection, and the operator's name. The bags were stored in the dark (at room temperature), and transported to the processing plant within 7 days of collection.

Mango Branch Trees (MBT)
MBT was collected between September 2018 and December 2019 as follows: • Branch trees were cut and left in the field for one week for solar drying; • Solar dried MBT was milled in the field with a crusher and collected in polystyrene sealed bags, approximately 50 kg per bag, and identified with a card indicating location, variety, date of collection, and operator name. The bags were stored in the dark (at room temperature) and transported to the processing plant within 7 days of collection.

Cocoa Pod Husk (CPH)
CPH was collected during 2018 and 2019 (spring season) as follows: • Cocoa plants in the field were selected randomly according to the presence of greenyellowish or red-orange pods. Green-yellowish and red-orange pods were collected separately; • The outer mass from the pods were collected after a careful longitudinal cut, taking care not to cut the cocoa seeds; • Crude CPH was collected in plastic tanks (around 200 kg/tank), and seeds were collected in seed containers for subsequent fermentation. CPH water content was around 85%; • The plastic tanks were transported to the pilot plant within 24 to 48 h after collection for subsequent drying.
Samples were processed within 24 h after collection.

Cocoa Bean Shell (CBS)
Collected cocoa seeds in the field, as described above, were submitted to fermentation, hot-air drying, and subsequent milling in order to obtain dried cocoa seeds for chocolate production in an industrial plant (Rizek Cacao, San Francisco de Macorís, Dominican Republic). A residue from this process is the CBS, which was directly submitted for BGI extraction without any other treatment. CBS was collected in sealed polystyrene bags (approx. 30 kg each), and stored in the dark (at room temperature). CBS samples were processed within 30 days of collection. The dryer drum was covered with a 60-mesh cloth in order to avoid particle entrance into the dryer motor. Each lot from solar drying was divided in two lots of 50 kg each and dried with hot air for 3 h. MBT had low water content (between 10% and 15%), and therefore solar drying was not necessary; • Cocoa: Approximately 1.5 kg of CPH was put in each tray of the solar drying unit as described above. Each lot (75 kg) was dried within 96 and 168 h depending on the initial water content, and divided in two dryer sections, A and B, in a similar array as describe for mango. CPH from the solar drying unit was submitted for drying at a rotary hot-air dryer (Girbau, Barcelona, Spain, Model E660) at 60 • C, with a rotary speed of 80 rpm. The dryer drum was covered with a 120-mesh cloth in order to avoid particle entrance into the dryer motor. Each lot from solar drying was divided into two lots of approx. 37 kg each. and dried with hot air for 4 h.

GRM Milling
• Mango: Both MSB and MBT were milled after drying in a hammer mill (Buskirk, Indiana, USA, Model HM1000) equipped with a 20 HP (3 p) motor and screen holes of 10 mm (first step) and 3 mm (second step). Starting GRM sizes were between 5 and 100 mm; • Cocoa: CBS was milled in the same device in a single step with 3 mm screen holes. Starting CBS sizes were between 2 and 20 mm.

GRM Pelletization
Milled GRM from both sources (mango and cocoa) were pelletized for subsequent storage until extraction in a pelletizer (Buskirk, Indiana, USA, Model PM1230) connected to the hammer mill through a conveyor. The pellets were obtained by pressing the milled GRM against a SS metal plate at high pressure. The average pellets size was 8 mm (diameter) × 15 mm (length). The collected pellets were placed in PE bags, which were sealed under vacuum (32 kg each) and stored at a controlled temperature (25 • C) until extraction within the next 30 days (see Figure 4). The physical quality of the GRM pellets was evaluated according to the Pellet Fuel Institute, USA [62] as follows: • where V p is volume of an individual pellet (cm 3 ); d is diameter of an individual pellet (mm); L is length of an individual pellet (mm); ρ p is density of an individual pellet (kg/m 3 ); and m p is mass of an individual pellet (g). where Vp is volume of an individual pellet (cm 3 ); d is diameter of an individual pellet (mm); L is length of an individual pellet (mm); ρp is density of an individual pellet (kg/m 3 ); and mp is mass of an individual pellet (g).

Stirred-Tank Extraction
Thirty-two kilograms of dried and pelletized MBS or MBT were put into a 250 L SS jacketed-stirred tank extractor, previously filled with 220 L of deionized water (mango) or water:ethanol (1:1) for CBS. Extraction temperature was fixed at 80 ± 5 °C, and agitation speed at 70 ± 5 rpm. Extraction was performed by 2 h, once the mixture reached the appropriate temperature. After this, heating was stopped and the mixture was cooled until room temperature (approx. 30 °C). Five batches for each GRM were performed and the results were expressed as a mean value ± SD.

Extract Filtration
Extract was filtered through a polystyrene multifilament cloth using a filter press (Chayo Ltd., Hangzhou, China, Model XAJ5/400-Uk.) with five gasketed plates, 40 × 40 cm, at 10 L/min flow. Cloth mesh diameters were 60 and 120 for MSB/MBT and CBS, respectively. The filtered extract was collected in 200 L PP tanks, with an adequate amount of sodium benzoate (preservative), and stored in a cool room (10 ± 2 °C) until spray drying.

Spray Drying
The extract was left to stand at room temperature before spray drying and brought into the spray dryer feeding tank. The input flow was set at 70 mL/min; inlet air temperature, 170 °C; outlet air temperature, 90 °C; and disc revolutions at 21,700 rpm, for obtaining a drying yield about 4 L/h (Pilotech, Shanghai, China, Model YC-018). The BGE was put into PE bags, weighed, and sealed under vacuum for storage.

Stirred-Tank Extraction
Thirty-two kilograms of dried and pelletized MBS or MBT were put into a 250 L SS jacketed-stirred tank extractor, previously filled with 220 L of deionized water (mango) or water:ethanol (1:1) for CBS. Extraction temperature was fixed at 80 ± 5 • C, and agitation speed at 70 ± 5 rpm. Extraction was performed by 2 h, once the mixture reached the appropriate temperature. After this, heating was stopped and the mixture was cooled until room temperature (approx. 30 • C). Five batches for each GRM were performed and the results were expressed as a mean value ± SD.

Extract Filtration
Extract was filtered through a polystyrene multifilament cloth using a filter press (Chayo Ltd., Hangzhou, China, Model XAJ5/400-Uk.) with five gasketed plates, 40 × 40 cm, at 10 L/min flow. Cloth mesh diameters were 60 and 120 for MSB/MBT and CBS, respectively. The filtered extract was collected in 200 L PP tanks, with an adequate amount of sodium benzoate (preservative), and stored in a cool room (10 ± 2 • C) until spray drying.

Spray Drying
The extract was left to stand at room temperature before spray drying and brought into the spray dryer feeding tank. The input flow was set at 70 mL/min; inlet air temperature, 170 • C; outlet air temperature, 90 • C; and disc revolutions at 21,700 rpm, for obtaining a drying yield about 4 L/h (Pilotech, Shanghai, China, Model YC-018). The BGE was put into PE bags, weighed, and sealed under vacuum for storage. The BGE was analyzed by standard methods of microbiological quality control for total coliforms, mesophilic content, mold, fungi, Clostridium sp., and Salmonella sp. [63], and for its physical-chemical quality specifications. Water content was determined with a moisture analyzer (Radwag, Puszczykowo, Poland, Model PMR 50). MF and OPA content for MSB-BGE, MBT-BGE, and CBS-BGE were determined by HPLC-DAD as described hereinafter.

Bioactive Green Ingredient (BGI)
• Polyphenol Content (PPC): The polyphenol content (PPC) of mango-and cocoaextracts was determined by a modified Folin-Ciocalteu method using a catechin-equivalent standard [27]. The extracts (1.15 mg) were dissolved in methanol (2 mL) and the solution was diluted ten-fold with distilled water. Folin-Ciocalteu reagent (0.5 mL) was added to the diluted solution, followed by 0.5 mL of sodium carbonate, 100 g/L solution. The absorbance was measured at 700 nm (Thermo Scientific, MA, USA, Genesys 10 spectrophotometer) with a blank sample (water plus reagents) in the reference cell (1 cm-depth quartz). Quantification was performed by plotting the absorbance value in a calibration curve of (+) catechin used as standard phenol; • Mangiferin: Pure MF was obtained from MSB-BGE and MBT-BGE by recrystallization in a mixture of acetone:water (5:1) through a modified procedure [64]. The purity of the re-crystallized MF was checked by HPLC (Young Lin, Korea) with a quaternary pump (Model YL-9110), autosampler (YL-9150), and a DAD (diode-array detector, YL-9160). The column (RP-18, 5 µm, 250 × 4 mm i.d., Merck, Darmstadt, Germany) was placed in a column oven (YL-9131) at 30 • C. The solvents were degassed (YL-9101) and the injection volume was 20 µL. The mobile phase used was acetic acid (0.1%) in water (A) and acetic acid (0.1%) in methanol (B). The ratio of A:B increased from 9:1 to 1:9 in 35 min at a flow rate of 1 mL/min. The data acquisition and peak integration analysis was performed using Clarity software (Data Apex, Czech Republic); • OPA-rich Extract: CBS-BGE was defatted with n-hexane in a Soxhlet apparatus, and subsequently extracted with a mixture of ethanol:water (7:3), pH = 6.5 (with acetic acid) in a ratio of 1:5 at room temperature with agitation (30 rpm) for 30 min. The mixture was filtered under vacuum, the filtrate was spray dried and the solid was recrystallized in a mixture of acetone:water (5:1). OPA content (catechin-and epicatechin types) was determined by HPLC-DAD as described above for MF, with the following changes in experimental conditions: in the mobile phase, the formic acid (0.1%) was in water (A) and acetonitrile (B); the ratio of A:B was increased from 1:9 to 9:1 in 35 min; the flow rate was 0.5 mL/min; and the column temperature, 40 • C.

Chemicals and Standards
PA and HPLC grade solvents (n-hexane, methanol, acetone, and acetonitrile) and reagents (acetic acid, formic acid, sodium carbonate, and sodium benzoate) were purchased from J.T. Baker (USA). Folin-Ciocalteu reagent was purchased from Merck (Darmstadt, Germany). AquaMax (Ultra 360 and Ultra 370, Young Lin, Korea) was used for producing ultrapure water. The following standards were purchased from Sigma-Aldrich Co. See Flowchart S1 in Supplementary Material.

Conclusions
PROMANCOA modular technology (PMT) was developed at a pilot plant scale with four modules: (1) green raw material collection and storage; (2) green raw material processing (drying, milling, and pelletizing); (3) green raw material extraction (extraction, filtration, and spray drying); and (4) quality control (water content, polyphenol content, mangiferin content (mango), and oligomeric proanthocyanidin content (cocoa)) for the valorization of agricultural by-products of mango and cocoa crops. Mango bioactive green extracts from branch trees and stem bark had a polyphenol content of 13.5 ± 0.8 mg/100 g and 16.2 ± 1.2 mg/100 g d.w., respectively, with more than 75% of the main bioactive ingredient (mangiferin), and cocoa bioactive green extract from cocoa bean shell had a polyphenol content of 28.1 ± 0.2 mg/100 g d.w. with around 50% of oligomeric proanthocyanidins. Our results show the feasibility of producing antioxidant-rich extracts from mango and cocoa agricultural by-products as acceptable sources for producing novel ingredients for food, cosmetics, and pharmaceutical industries, and, therefore, are a step forward for introducing new bioactive green ingredients into these markets. The main advantage for PMT is the accessibility to equipment and processes that have been established in the classical extraction industry.