Coupling Flash Vacuum Expansion and Spray Drying to Produce Stable Polyphenolic Extract from Coffee Exocarp

: In the present work, the effect of the Flash Vacuum-Expansion (FVE) process on the recovery of total phenolic content (TPC) from entire coffee peel was evaluated. The resulting extracts were microencapsulated by spray drying using maltodextrin (MD10) and arabic gum (GA). The FVE increased TPC by 20.97% in the ethanolic extracts. The use of MD10 allowed to obtain smaller particles, less hygroscopic, and with better color and flow properties, as well as a higher concentration of TPC. Both powders were considered microbiologically stable, based on humidity and water activity.


Introduction
Coffee is a priority crop in Mexico.This country is the tenth worldwide coffee producer.However, despite its economic importance, the use of this crop is restricted to the use of the grain.During coffee processing, large amounts of waste are generated, and coffee peel is one of the most abundant ones.Recently, the search for alternatives that allow valorization of this by-product has intensified.Most of them focus on the recovery of compounds from the peel to generate extract enriched in natural bioactive compounds or with a sensorial interest [1].An emerging technology with high potential for the recovery of phenolic compounds is Flash Vacuum Expansion (FVE) [2].FVE is a mechanical process in which plant material is first steam-heated, generally between 60 and 95 • C, before the expansion stage, that is to say the introduction into a chamber vacuum (2 to 5 kPa) [3].This pressure drop establishes the physical condition necessary to allow the instantaneous evaporation of constitutive water.It creates micro-channels inside the tissues and breaks a part of them, depending on processing condition intensities [3].Entire fruits can be processed by FVE in order to obtain puree enriched in phenolic compounds due to the liberation of compounds from peels and seeds [4].Its versatility is interesting as an alternative for the recovery of compounds of interest from entire coffee peel, mainly anthocyanins [5].The extracts can be then used as food ingredients in a powder form through the coupling of FVE and the microencapsulation process.Microencapsulation into a coating material can be achieved by several methods.However, spray drying is the most common technology used in the food industry due to its low processing cost and its generalization in the food industry for multiple applications [6].
The objective of the present work was to study the feasibility of the fabrication of a powder which can serve as a food ingredient.Whole coffee peels were extracted by FVE in a solvent as mild green processing and dried by spray drying.The recovery of phenolic compounds as well as the physicochemical characteristics of the microencapsulates generated using two different coating materials was in depth.

Raw Material
The coffee peel of Arabica variety was pulled from cherry coffee in a state of maturity and disinfected with a 100 ppm solution of sodium hypochlorite.The coffee cherries were obtained from the town of Tenampa, Veracruz, Mexico, located at 1060 mbsl.Samples were collected on 10 randomly selected trees from a plantation of 350 coffee trees.

Flash Vacuum Expansion Processing
The FVE process was carried out in laboratory-scale equipment.For the experiment, around 1 kg of the entire coffee peel was introduced into the heating chamber at atmospheric pressure (101.3 kPa) and subjected to a 10 min thermal exposure time to the vapor flow [3].Then, the sample was instantly transferred into the vacuum chamber (3 kPa).The expansion also caused a rapid cooling of the sample which was collected and stored in hermetic amber bottles.The coffee peel treated by FVE was used for polyphenol extraction; the composition of the extraction solvent was selected based on previous work on anthocyanin extraction of coffee husks [7].

Spray Drying Processing
For the spray drying process, two different encapsulation materials were used.The extract was mixed with a concentrated aqueous solutions of maltodextrin 10 DE and arabic gum as wall material to reach a final concentration of 30 ± 1% of soluble solids, with a ratio of core/wall of 1:1 (v/v).The mixture was then spray-dried in a Büchi B-290 mini spray dryer (Flawil, Switzerland).Dry conditions were the following: air inlet temperature of 160 ± 2 • C, air outlet temperature of 80 ± 3 • C, feeding rate of 0.39 L•h −1 and volumetric flow rate of 667 L•h −1 .Samples were named MD10 and GA if they were encapsulated by maltodextrin or arabic gum, respectively.

Physicochemical Characterization of Powders
MD10 and GA powders were characterized in terms of moisture content according to the gravimetric method described by the AOAC [8]; water activity measured using a water activity meter at 25 • C; hygroscopicity determined according to Cai and Corke [9].Briefly, samples (about 2 g) from the Petri dishes were placed at 25 • C in a hermetic desiccator filled with Na 2 SO 4 saturated solution (81% RH).After 1 week, the powders were weighed and expressed as g of moisture per 100 g dry solids (g/100 g); particle size determined by calculating the arithmetic mean of 300 particles, measured using a Motic BA310E optical microscope (San Antonio, Texas, USA); apparent and compact density calculated from the volume occupied by a known mass of dust in a graduated cylinder according to Santhalakshmy [10]; fluidity and cohesiveness of the powders calculated from the apparent and compacted density (expressed in Carr's index and Hausner's ratio, respectively) [11,12]; total content of polyphenols expressed in mg gallic acid equivalent and determined using the Folin-Ciocalteu method as described by Singleton et al. [13], the absorbance was measured at 765 nm using a spectrophotometer (Evolution 260, Thermo Scientific, Waltham, MA, USA), color properties were estimated using a CR400 colorimeter (Minolta, Osaka, Japan).The chroma and Hue parameters were evaluated according to the equations reported by Maskan [14].

Statistic Analysis
All determinations were carried out in triplicate, and the results are presented as means with standard deviation.Analysis of variance (ANOVA) and Tukey's test of means were used to identify significant differences, and all the statistical analyses were performed using Minitab software v. 17 (State College, PA, USA).

Effect of the FVE on the Total Phenol Content
TPC in ethanolic extracts of coffee peel treated by FVE were 485.19 ± 1.53 and 401.08 ± 0.76 mg GAE•L −1 , respectively.Similar results were reported on unroasted Arabica coffee with 473.51 mg GAE•L −1 [15].The application of the FVE allowed an increase of 21% of the total phenolic content.This increase is usual during FVE application on peels and was extensively reported for red grape processing [5,16].It was attributed to a greater cell disruption due to expansion force caused by sudden water evaporation during the expansion, the formation of microchannels which allows a more effective release of the compounds and the solubilization of pectin during the process.The same behavior has been also reported in the literature for different plant matrices treated in their entirety, with peel and seeds, such as avocados and grapes (six times and 49% more polyphenols after FVE, respectively) [17,18].After entire mango FVE processing, an increase of 38.15% on polyphenol content was reported after 20 min of heat exposition at 100 • C [3].This polyphenol release was related to a diffusion of the polyphenols from the non-comestible parts to the puree due to the rupture of cell vacuoles.Vacuoles are organelles most sensitive to the FVE process because they store a high amount of water and phenolic compounds.This phenomenon could explain the observation on coffee peel diffusion.
After FVE, the enriched extracts were microencapsulated by spray drying.

Physicochemical Characterization of the Powders
Physicochemical characterization results are shown in Table 1.Different letters in the same row represent significant difference (ANOVA, p < 0.05).

Moisture Content and Water Activity
Moisture of powders is important for the determination of flowability, stickiness, and stability during storage due to water content effect on glass transition and crystalliza-tion [19].The values obtained in this work are consistent with the reported literature for spray-dried powders using maltodextrin and arabic gum as wall materials with a mean content of 2.45% [20].This value ensures the microbiological stability of the powder [19].Indeed, moisture is an important index because it indicates the availability of free water in a food system, "water activity" that is responsible for biochemical reaction.Water activities of MD10 and GA were found within the interval also indicated by Da silva et al. [21] as necessary to ensure the microbiological stability of powders (typically water activity < 0.35).

Hygroscopicity
Hygroscopicity is the ability of a material to adsorb moisture from the surrounding environment.It is a property to consider because it offers information about the powder behavior during storage [22], it helps to understand both the physical stability and the applicability of the product.GA was significantly (p < 0.05) more hygroscopic than MD10 (14.69 and 13.95 g of H 2 O/100 g of d.b., respectively).The results obtained for both powders were lower than those reported for other microencapsulated products with maltodextrin and arabic gum, hygroscopicity > 19 g of H 2 O/100 g of d.b. [20].There are different operating conditions during spray drying that affect the hygroscopicity of powders, such as the increase in the inlet temperature.The total moisture content also influences this physical parameter.However, in this study, the inlet temperature was the same for both wall materials, and the moisture content did not show significant differences.The higher capacity of water adsorption observed for GA powders may be associated with the difference of hydrophilic groups in the structure of each wall material.Arabic gum, which has a greater number of hydrophilic groups, presented greater adsorption of ambient humidity, while maltodextrin 10 DE had less hydrophilic groups [9].The same behavior was observed by Da Silva et al. [21].

Particle Size and Bulk and Tapped Densities
Bulk density refers to the density of a material when it is packaged.As reported in Table 1, GA value was significantly higher for this parameter, in agreement with Tonon et al. [23].Bulk density is related to the molecular weight of wall materials; the heavier the wall material, the easier it is to accommodate in the spaces between the particles.Therefore, powder occupies less space, resulting in higher bulk density values.GA has a higher molecular weight than MD10, which can contribute to the explanation of why its bulk density is higher.An increase in this value indicates a lower amount of air in the powders, which can help prevent oxidation and promote stability during storage [10].The tapped density is a property that offers information on the packaging, transport and commercialization.It is frequently used to determine the properties of compressibility and cohesiveness of the powders.The augmentation of the compacted density could be related to the fact that GA presented a particle size larger than MD10, with means values of 12.97 ± 2.4 and 8.61 ± 1.58 µm, respectively.A similar behavior was reported by Ozdikicierler et al. [24] for spray-dried powder of Gypsophila plant extracts using maltodextrin as wall material.

Flowability and Cohesiveness
The Carr index (IC) evaluates the free flow characteristics and the Hausner ratio (RH) evaluates the cohesion of the powders.The results of IC and RH are shown in Table 1.The values obtained were consistent with Bhusari et al. [25], who reported results from 19.34 to 34.16 and from 1.22 to 1.51 for IC and RH, respectively.Both powders, MD10 and GA, were characterized by medium fluidity and high cohesiveness according to Carr and Hausner classification [11,12].A higher IC value indicates poor fluidity; therefore, the use of MD10 as wall material improves the fluidity of the powders.This is related to the cohesion of these powders: the lower the cohesion, the better the fluidity of the powders (Ee et al., 2014).Thus, a free-flowing material has a tendency to have a greater consolidation, which is useful to avoid production stops on an industrial scale [26].

Color Properties and Total Phenol Content
The use of MD10 as wall material allowed a significant increase in parameter L*, which indicates the luminosity or opacity of a sample (100 for very bright and 0 for opaque samples).This behavior was associated with the optical characteristics of the maltodextrin, since this compound has a bright white color.On the other hand, arabic gum has a yellowish color, less bright.This difference in color of the wall material can influence the perception of the powder color, since these compounds represent a high proportion of the powder dry matter.GA showed a significant decrease in the intensity of the chroma value and hue angle.This sample presented a greater yellow color (74 • ), while MD10 exhibited a color shade with a greater tendency towards red colors (56 • ).It can also be observed in the evolution of parameter a*, which represents the red shades.MD10 a* was similar to that reported by Parra-Campos and Ordóñez-Santos [7] for ethanolic extracts of coffee exocarp; they reported values in a range of 5.42-6.51.Although this decrease in red could be related to the loss of anthocyanins, a more detailed study is suggested to understand its conformation.Phenol content was significantly higher for MD10 and in agreement with the literature on coffee-based products [15]; therefore, MD10 and GA powders could be considered as antioxidant food ingredients, with a potential beneficial effect on human health.

Conclusions
The FVE process proved to be an interesting alternative for the extraction of phenolic compounds from entire coffee peel waste using ethanolic extracts.The use of maltodextrin and arabic gum during spray drying allowed obtention of particles with low moisture and water activity which might guarantee their microbiological stability.Powders elaborated with maltodextrin as wall material produce less hygroscopic particles, with significantly better flow and cohesiveness properties as well as color properties.Lastly, the contents of total phenol in both powders were in agreement with the literature.
Therefore, the products which can be considered as potential food ingredients with functional properties added value to coffee production since they were obtained from a waste by a green emerging technology.

Table 1 .
Physicochemical characterization of MD10 and GA.