Spray-drying technology is a suitable process for large-scale production of various advanced materials. This study describes the effect of the concentration (0.5, 1.0, and 1.5%) of novel carrier agents, agave fructans (FOS) and maltodextrin (MD), in a chlorophyll extract spray-drying process on colour, antioxidant activity, and stability. A pilot plant was employed with a spray-dryer feed flow rate of 13% and an atomisation level of 28000 rpm. The air inlet and outlet temperatures were 180 and 80°C, respectively. A MD concentration of 0.5% was enough to obtain a chemically and physiochemically stable encapsulated chlorophyll powder. High inlet air temperatures reduced the chlorophyll content from 18 to 6 mg/g; high concentrations of the carrier agent also decreased the chlorophyll content. The results showed that, under conditions of low inlet temperature and concentration, both wall materials were efficient in microencapsulating chlorophyll for potential use in the food industry.

1. Introduction

There is growing interest in the food industry for dyes, flavours, and food products with natural characteristics [1]. However, colourants from natural sources, such as chlorophyll, are unsuitable for use as dyes due to their rapid degradation by enzymatic reactions or other factors, such as acidic conditions and the presence of oxygen, light, and heat, resulting in the formation of chlorophyll derivatives (pheophytin, pheophorbides, pyropheophytin, and pyropheophorbides). Stable chlorophyll molecules can be generated by replacing the magnesium ion in the porphyrin ring with a divalent cation, such as zinc or copper, thus changing the native form of the molecular structure of chlorophyll to a more stable form [2]. The derivatives produced exhibit the same green colour as native chlorophyll but are more stable under heat and acidic conditions and more efficient antioxidants [3]. Spray drying is a useful powder-forming technique for converting liquids to solids, which is easy to use and improves shelf life and product stability [4]. The initial step in drying dyes involves the selection of a material, known as a carrier or encapsulant. Arabic gum, hydrolysed starches, and modified starches are the three most important classes of agents used in the microencapsulation process [5]. Agave fructans (FOS) consist of fructose units with β-type (2-1) bonds with glucose units at the end of the chain, and fructans (FOS) are a branched variant of inulin linked by β-(2 ⟶ 1) and β-(2 ⟶ 6) [6, 7]. Both compounds, like maltodextrin (MD), are very interesting encapsulants. FOS are dietary fibres that show prebiotic effects that can also improve the bioavailability of calcium [8]. In addition, FOS are officially recognised as natural food ingredients. However, FOS are carbohydrates that undergo many changes, such as hydrolysis, Maillard reaction, and caramelisation [8]. Thus, full or partial encapsulants such as Arabic gum substitutes have been sought, and MD and modified starches have been identified for the first time as suitable substitutes because they are abundant and inexpensive materials [9]; in addition, inulin and FOS are beneficial to human health. For this reason, the objective of this work was to find a novel formulation of agave FOS and MD as carrier agents to produce spray-dried chlorophyll extracts.

2. Materials and Methods

2.1. Extraction of Chlorophyll and Generation of Zn2+ Chlorophyllin Derivatives

Spinach leaves (Spinacia oleracea) were obtained from a local market and dried using convective drying on trays until the moisture content was 12.5% (dry basis) (094 TABML Dryer Polinox, Mexico DF). The dried leaves were ground in a hammer mill (hammer mill JERSA M.D.M.-01, Edo de Mexico, Mexico) until an approximate particle size of 1 mm was obtained. Then, solid-liquid extraction was carried out for 24 hours by maceration with ethanol (80% v/v). Subsequently, the pigments were stabilised by the regreening effect, and the temperature was increased to 115°C for 30 minutes; then, zinc acetate was added at a concentration of 300 ppm relative to the amount of vegetal material used for the extraction. Finally, the extract was concentrated in a rotavapor (BÜCHI Rotavapor R-220, Switzerland).

2.2. Spray Drying

The extracts were dried using a spray dryer (GEA NIRO A/S production model minor, Denmark) under two inlet air temperature conditions (180 and 190°C). The outlet air temperature was constant at 80°C, the disc rotation speed was 28140 rpm, and the air flow rate was 100%. A total of 2.5 kg of the extract diluted in deionised water (1 : 1) was fed to the dryer. The agave fructans and maltodextrin (DE 10) concentration as carrier agents were added in base to relation of 0.5, 1, and 1.5% w/v. The powder collected from the cyclone was stored in sealed plastic bottles and protected against sunlight until analysis.

2.3. Moisture Content and Water Activity Determination

The residual moisture was determined gravimetrically by convective drying of 2 g of powder in a thermobalance (model PRECISA 310, PRECISA SA 300, USA) at 105°C to a constant weight. The results are expressed as a percentage based on the weight of the powder (100 × g water/g powder). The value of aw was determined using 1 g of powder in a portable digital water activity meter (Pawkit Water Activity Meter Inc. Services, Decagon, Pullman, WA, USA) at 26°C [10].

2.4. Bulk Density

Volumetric density was measured by adding 2 g of powder to 15 ml graduated centrifuge tubes (Falcon). The tubes were gently dropped on a surface after lifting them to a height of 10 cm. This procedure was repeated 100 times. The bulk density was calculated by dividing the mass of the powders by the volume occupied by these powders in the tubes (g/cm3) [11].

2.5. Analysis of Solubility

Solubility analysis was carried out according to the method by Al-Kahtani and Hassan [11]. Specifically, 100 ml of deionised water was transferred to a blender cup. One gram of powder was added carefully into the blender operating at full speed for 5 minutes. The solution was transferred to Falcon tubes and centrifuged at 5000 ×g for 5 minutes. A 25 ml aliquot of the supernatant was transferred into reweighed Petri dishes and then oven-dried at 105°C for 5 hours. The solubility is reported as percentage (%).

2.6. Colour Values

The colour values lightness (), greenness (), yellowness (), chroma (C), and hue (°h) were determined. A total of 0.25 g of powder sample was diluted in 25 ml of distilled water by mixing for 3 minutes. Ten-millilitre volumes were transferred to a container with a white background to be measured using a colorimeter (X-rite SP62, Grandville, MI, USA) following the method of Farías Cervantes et al. [6].

2.7. Determination of Zn2+ Chlorophyllin Derivative Content

The Zn2+ chlorophyll content of both powders and extracts was quantified according to an AOAC method [12] with certain modifications. Fifty millilitres of the sample was prefiltered using #1 Whatman filter paper and transferred to a volumetric flask with 85% acetone. Then, a 50 ml aliquot was transferred to a separatory funnel containing 50 ml of petroleum ether. Washings were performed by adding water to the separator until all fat-soluble pigments were in the ether layer. Then, the water layer was drained and discarded. The ether solution containing the pigments was washed 5 times within the funnels until all acetone was removed. Then, the ether solution containing the pigments was transferred into a volumetric flask of 100 ml followed by stirring for a short time. Five grams of anhydrous Na2SO4 was added to a 60 ml reaction flask, and the ether solution containing the pigments was added, followed by spectrophotometric measurements. Once the solution was optically clear, it was measured spectrophotometrically; the solution may need to be diluted enough to give absorbance values of 0.2–0.8 for both chlorophyll a and b. Calculations of the total Zn2+-chlorophyll concentration can be carried out according to the formulas shown in the following equations using the absorbance values obtained at 642 and 660 nm:

2.8. Antioxidant Activity (DPPH)

The 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical method was used to measure the antioxidant activity of the microencapsulated extracts [13]. A 0.1 µm solution of DPPH was prepared, and 2900 µl was transferred to covered glass tubes. Then, 100 µl of the sample was added to each tube, followed immediately by vortexing and then placing the tubes in darkness for 30 minutes. Once the reaction time was reached, the absorbance at 517 nm was measured. A calibration curve using the Trolox standard reagent was prepared in a range of micromolar concentrations, and the concentration results are expressed as µM Trolox equivalents (TE)/g powder.

2.9. Sorption Studies (Sorption Isotherms)

Dynamic dew point isotherms (DDIs) were generated using an isotherm generator (AquaSorp Decagon Devices Inc., Pullman, WA, USA). Samples (1 to 2 g) were weighed into a stainless-steel container to completely cover the bottom of the container. The samples were run with a flow rate of 300 ml/min at 30°C for water activity values of aw from 0.20 to 0.90, using the initial aw of each sample as the initial value (AquaSorp firmware version ASIG 1.19, Decagon Devices, Pullman, WA, USA). Sections of sorption and desorption were obtained in a single run by treatment with repetition. The change in weight for each aw value was converted to the moisture content (% dry basis) (SorpTrac version 1.03.3, Decagon Devices, Pullman, WA, USA).

2.10. Differential Scanning Calorimetry (DSC)

Glass-transition temperatures (Tg) were generated using DSC (2000CE TA Instruments, USA). Thermograms were generated during heating of the sample in a temperature range from 5°C to 200°C at a heating rate of 5°C/min. Indium was used to calibrate the temperature and enthalpy reading. All DSC measurements were performed using hermetic aluminium trays of known mass; for example, an empty sample tray and a reference tray had equal masses within the range of ±0.10 mg, whereas the mass of the sample was generally 2.50 ± 0.50 mg. The DSC equipment was purged with dry nitrogen.

2.11. Scanning Electron Microscopy (SEM)

The morphology of the microcapsules was determined by SEM (JEOL scanning microscope, model JSM-6390LV). The sample was placed on a slide with double-sided ionised carbon tape, and there was no need to cover the samples with gold. The images were digitally captured at a resolution of 10 µm and an accelerating voltage of 20 kV and were amplified 2000 times.

2.12. Statistical Analysis

The means and standard deviations were calculated and analysed by one-way analysis of variance (ANOVA) with value = 0.05. All treatments were replicated three times. All analyses were carried out using Minitab software (Minitab 15; Minitab Inc., Minneapolis, MN, USA).

3. Results and Discussion

3.1. Physicochemical Properties

The physicochemical properties are shown in Table 1. For the reconstituted powder, the colour values on the CIELab scale do not show a significant difference, indicating that the temperature and carrier agent type are not determining factors for obtaining a change in the colour of the reconstituted powders. However, the opposite was observed with the increase of concentration of carrier agents, as increasing the concentration of any carrier agent increases the brightness of 25.12 to 32.96, greenness () of 3.88 to 7.99, yellowness () of 9.0 to 19.94, and chroma values of 9.72 to 21.62, no change for hue angle values in high concentrations of carrier agents, where no significant differences were observed. An increase in the brightness value indicates a lighter colour, most likely related to a loss in chlorophyll, or a dilution for the increase of the carrier agent concentration. Porrarud and Pranee [4] and Al-Kahtani and Hassan [11] mentioned that due to the residence time of particles in the dryer, brightness is affected, which can be described by the CIELab scale; in other words, colour is affected during the drying process. In the quantification of residual moisture, a relatively similar behaviour was observed for both carrier agents, when the concentration increases and moisture content decreases from 14.0 to 9.62%, which may be attributable to an increase in feed solids and a reduction in total moisture for evaporation. The temperature showed a change in the moisture with MD10 and FOS, and this can be because the heat transfer between particles is greater which will give a greater driving force for moisture evaporation. Goula and Adamopoulos [14] observed a similar behaviour in the drying process of tomato pulp when maltodextrin was used as a wall material, and they explained that the higher the temperature difference between the drying medium and solid particles was, the higher the heat transfer rate within the particles was. In the aw measurements, the values obtained with MD were visibly higher than those with FOS, so the inlet air temperature showed the same effect as in the moisture content study, i.e., an increase in the inlet air temperature reduces the water activity. Similar behaviour has been reported by authors such as Benstain et al. [15], Grabowski et al. [16], Chattopadhyay et al. [17], Goula and Adamopoulos [14], and Tonon et al. [18]. In the bulk density study, an increase of 10°C in the inlet air temperature considerably increases the bulk density; likewise, an increase in the concentration of the wall material decreases the bulk density. This decrease may be due to the increase in feed viscosity with an increase in the wall material in the feed material, which results in the formation of larger powder particles, thus decreasing the bulk density. This fact is in contrast to the report made by Goula and Adamopoulos [14], who explained that an increase in the air-drying temperature decreases the bulk density. Additionally, an increase in the MD concentration decreases the bulk density by reducing the thermoplastic properties of particles, making them fall into a tacky state; this state of a material is associated with a high bulk density.

3.2. Zn2+ Chlorophyllin Derivative Content

The content of Zn2+ chlorophyllin derivatives can be seen in Figure 1; the concentration of Zn2+ chlorophyllin ranged from 18.14 to 5.19 mg/gfw. The highest concentration of 18.14 mg/gfw was obtained with the use of maltodextrin as a carrier agent at a concentration of 0.5% and with an inlet air temperature of 180°C. When comparing the effect of MD with FOS as carriers, it was observed that when using the concentration of 1% at the two inlet air temperatures, they did not show significant differences. It was observed that, at higher inlet temperatures, lower concentrations of pigment are produced. Zn2+ chlorophyll is an easily oxidized composite when exposed to high temperature, and one of the reasons for the low efficiency found may be related to inlet air temperature used in the spray dryer (180 and 190°C). Shu et al. (2006) state that high inlet air temperature can break the balance between the evaporation and the particle formation rate, reducing encapsulation efficiency. Additionally, a higher concentration of carrier agent leads to a decrease in the concentration of the same pigment. The effect of FOS concentration was similar reported by Farías Cervantes et al. [6], where an increased FOS concentration reduced the Zn2+ chlorophyllin. This behaviour could be due to the increased particle size and low spray-dried rate during spray drying. These results are different from those reported by Al-Kahtani and Hassan [11] who found that higher inlet temperatures increase anthocyanin retention, but higher concentrations of the carrier agent lead to a decrease in pigment retention. Porrarud and Pranee [4] found a lower concentration of Zn2+ chlorophyllin derivatives for microencapsulation using MD than other carrier agents, such as arabic gum and modified starches, and explained that MD is derived from acidic or enzymatic modification of starch into monomeric or chained polymers that subsequently form short, thin layers that protect the microencapsulated compounds. However, the emulsification properties and lack of oil-water linkages to MD interfere with surface activity, resulting in a loss in microencapsulated content [19].

3.3. Antioxidant Activity (DPPH)

The powders that used a high MD concentration as a carrier agent have high percentages of inhibition with 65.98% (Figure 2). The values ranged between 14.86 and 65.98%. However, at low MD concentrations, lower inhibition percentages were obtained. Moreover, the percentage of inhibition in those powders in which FOS was used as the carrier agent indicated more stability. The antioxidant capacity is related to the content of Zn2+ chlorophyllin derivatives. Thus, a higher content of these Zn2+ chlorophyllin derivatives increases the antioxidant capacity [20]. The highest content of Zn2+ chlorophyllin derivatives indicates a greater breakdown of radical chains because these derivatives are electron donors. Hoshina et al. [21] found that the porphyrinic structure of chlorophyll molecules was able to inhibit the formation of lipid hydroperoxides and ferric thiocyanate from nitrilotriacetate. Consequently, the formation of a metalochlorophyll complex by changing chlorophyll to Zn2+ chlorophyllin derivatives not only stabilise the colour due to the stability of the porphyrin group but also result in a higher antioxidant capacity than that of native chlorophyll [22]. Because FOS is able to bind to both the hydrophilic and hydrophobic groups of the chlorophyll molecules, its antioxidant values were more stable than those of MD.

3.4. Studies of Sorption (Sorption Isotherms)

The powders exhibited similar sorption behaviour with a typical sigmoid shape (Figures 3 and 4), indicating relatively rapid internal diffusion of water vapour. Additionally, this similarity indicates that different concentrations of the carrier agent and different drying temperatures do not affect the internal diffusion of water vapour in the powders. Interestingly, there is a slight curve in the central part of each isotherm, suggesting that the physical changes were dependent on time and perhaps related to the initial conditions of the material (e.g., moisture content and aw) and/or storage history of the sample. Figures 3 and 4 show isotherms where MD and FOS were used as a carrier agent at different concentrations and spray-drying temperatures. The sorption isotherms show a typical increase in moisture content in response to increasing water activity. At a constant temperature, the behaviour of all curves is type II, according to the classification of Brunauer [23, 24]. The beginning of the curves represents the monolayer where water is tightly bound to the product. The middle region of the curve represents the multilayer where the water is weakly bound, and the end of the curve represents the region where water is free.

3.5. Differential Scanning Calorimetry (DSC)

In Figure 5, the glass-transition temperatures of the FOS samples are identified near 40°C. The ΔCp values of FOS varied from 0.25 J/gK at an aw value of 0.31 to 0.03 J/gK at an aw value of 0.64. Beirão-da-costa et al. [25] explained that the pattern may differ when different types of crystals are melted, and the differences are likely to be due to their shape, size, and/or molecular weight. Generally, the glass-transition temperature of amorphous materials is gradually changed by the heating flow due to changes in the heating capacity and the transition phase temperature. With the use of MD as a carrier agent (Figure 6), for spray-drying conditions of 190°C and 1% concentration, Tg values were found to be 60.03 and 51.75°C, respectively. However, when the concentration of MD was increased to 1.5%, the Tg value was only 59.89°C. The above values are consistent with those obtained by Fabra et al. [26] who emphasised that this reduction in the Tg values is expected because the water activity is expected to decrease due to the plasticising effects of water. MD shows higher glass-transition temperatures depending on the average molecular weight of the MD used.

3.6. Morphology

Particles in which FOS were used as a carrier agent presented spherical shapes with smooth walls and agglomerated structure, and those in which MD was used showed spherical shapes with dents (Figure 7). Porrarud and Pranee [4] suggested that spray-dried, spherically shaped particles have a greater surface/volume ratio, which is an appropriate characteristic of a spray-dried product. Additionally, particles with spherical shapes are capable of retaining different agents, such as flavourings. Treatment of samples under spray-drying conditions with 1.5% FOS as the carrier agent and 190°C inlet temperature showed slight signs of “cracking”. This is an important factor to ensure low gas permeability and better protection of the Zn2+ chlorophyllin derivatives. There are noticeable differences in the surface characteristics of each type of particle. The central mechanism for these differences is related to the expansion of the particles during later stages of the drying process. Thermal expansion of air or steam can reduce particle shrinkage, depending on the drying rate or the viscoelastic properties of the matrix carrier agent [27].

4. Conclusions

The efficiency of MD and FOS in maintaining physicochemical properties under conditions of low inlet air temperature during drying in combination with low wall material concentrations (180°C, 0.5%) was comparable. FOS presented a more stable inhibition percentage than MD. Sorption studies showed weak adsorbate-adsorbent interactions and an expected decrease in moisture content at equilibrium with high process temperatures and high carrier agent concentrations, which indicate that microencapsulation of both MD and FOS can extend the product shelf life when stored under normal temperature and humidity conditions. Calorimetric studies showed that both wall materials have low glass-transition temperatures, which is an important variable in determining processing and storage criteria. Although Tg for FOS was below the glass-transition temperature, which was an influencing factor at these temperatures, the material became viscous very rapidly during the drying process, indicating an increase in the probability of agglomeration related to the initial characteristics of FOS used exclusively for this study. However, when FOS are used as a carrier agent, the material shows a smooth and spherical form.

Data Availability

The data (DataSecado.xlxs) used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.


The authors express their acknowledgments to Mexican Consejo Nacional de Ciencia y Tecnología (CONACyT) for scholarships to the master’s student Cesar Femat-Castañeda and for the financial support through project PEI2017. Additionally, the authors wish to acknowledge the partner companies BIOBERSA® through Mr. Victor Berber and AGAVEN® through Mr. Pablo Orozco for their gentle economic support in the acquisition of the raw material and development of this project, as well as the Washington State University through Dr. Gustavo Barbosa for supporting the student Cesar Femat-Castañeda in the research stay.