Journal of Food Engineering 88 (2008) 411–418

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Influence of process conditions on the physicochemical properties of acßai

(Euterpe oleraceae Mart.) powder produced by spray drying
Renata V. Tonon a, Catherine Brabet b, Míriam D. Hubinger a,*
a
Faculty of Food Engineering, State University of Campinas, P.O. Box 6121, 13083-862 Campinas, SP, Brazil
b
Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD) – Department PERSYST – UMR QualiSud, Montpellier, France

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this work was to study the influence of spray drying conditions on the physicochemical
Received 7 December 2007 properties of acßai powder. The process was carried out on a mini spray dryer and maltodextrin 10DE was
Received in revised form 27 February 2008 used as carrier agent. Seventeen tests were made, according to a central composite design. Independent
Accepted 28 February 2008
variables were: inlet air temperature (138–202 °C), feed flow rate (5–25 g/min) and maltodextrin concen-
Available online 8 March 2008
tration (10–30%). Moisture content, hygroscopicity, process yield and anthocyanin retention were ana-
lysed as responses. Powder moisture content and process yield were positively affected by inlet air
Keywords:
temperature and negatively affected by feed flow rate, which are directly related to heat and mass trans-
Spray drying
Acßai
fer. Process yield was also negatively influenced by maltodextrin concentration, due to the increase on
Physicochemical properties mixture viscosity. Powders hygroscopicity decreased with increasing maltodextrin concentration,
Morphology decreasing temperature and increasing feed flow rate. Powders with lower moisture content were more
Anthocyanins hygroscopic, which is related to the greater water concentration gradient between the product and the
surrounding air. Anthocyanin retention was only affected by temperature, due to its high sensitivity.
In respect to morphology, the particles produced at higher temperature were larger and a great number
of them showed smooth surface.
Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction Spray drying is a process widely used to produce fruit juices

powders (Abadio et al., 2004; Cano-Chauca et al., 2005; Quek
Acßai (Euterpe oleraceae Mart.) is a typical fruit from Amazonia, et al., 2007). It results in powders with good quality, low water
with great occurrence and economical importance in the Brazilian activity and easier transport and storage. The physicochemical
state of Pará. Beyond being a highly energetic fruit, acßai has been properties of powders produced by spray drying depend on some
recognized for its functional properties for use in food and nutra- process variables, such as the characteristics of the liquid feed (vis-
ceutical products, due to its high antioxidant activity, which is re- cosity, particles size, flow rate) and of the drying air (temperature,
lated to its high anthocyanin and phenolic content (Coisson et al., pressure), as well as the type of atomizer. Therefore, it is important
2005; Schauss et al., 2006). Del Pozo-Insfran et al. (2004) reported to optimize the drying process, in order to obtain products with
that acßai was found to have higher antioxidant capacity than other better sensory and nutritional characteristics and better process
anthocyanin-rich fruits, such as highbush blueberries, blackberries, yield.
cranberries and others, and verified that the predominant anthocy- Fruit juice powders obtained by spray drying may have some
anin present in acßai pulp was cyanidin-3-glucoside (1040 mg/L problems in their properties, such as stickiness, hygroscopicity
pulp). Wang et al. (1997) evaluated the antioxidant activity of 14 and solubility, due to the presence of low molecular weight sugars
anthocyanins and observed that cyanidin-3-glucoside was the and acids, which have low glass transition temperature (Bhandari
one with the highest antioxidant power, about 3.5 times superior et al., 1993). Thus they can stick on the dryer chamber wall during
to Trolox (equivalent to vitamin E). drying, leading to low product yield and operational problems.
However, acßai is a highly perishable fruit with short shelf life. Part of these problems can be solved by the addition of some
Moreover, anthocyanins are pigments very instable to processing carrier agents, like polymers and gums, to the product before being
and storage. Thus, the food industry is constantly looking for novel, atomized. Besides reducing powder hygroscopicity, such agents,
inexpensive and more stable source of these pigments. normally used for microencapsulation, can protect sensitive food
components against unfavorable ambient conditions, mask or pre-
serve flavors and aromas, reduce the volatility and reactivity and
* Corresponding author. Tel.: +55 19 35214036; fax: +55 19 35214027. provide additional attractiveness for the merchandising of food
E-mail address: [email protected] (M.D. Hubinger). products (Ré, 1998).

0260-8774/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jfoodeng.2008.02.029
412 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418

Nomenclature

b regression coefficients of Eq. (1) Subscripts

D[4,3] mean diameter over the volume distribution (lm) 1 inlet air temperature
x coded independent variable of Eq. (1) 2 feed flow rate
y dependent variable of Eq. (1) 3 maltodextrin concentration

Some authors have used carrier agents in order to protect sen- pump and the feed flow rate was controlled by the pump rotation
sitive components like vitamin C in fruits such as camu–camu (Dib speed. Drying air flow rate was 73 m3/h and compressor air pres-
Taxi et al., 2003) and to increase product stability in acerola pow- sure was 0.06 MPa. Inlet air temperature varied from 138 °C to
der (Righetto and Netto, 2005). In the case of acßai juice, spray dry- 202 °C and feed flow rate varied from 5 to 25 g/min, according to
ing with carrier agents represents an interesting process, which an experimental design, described at the item 2.5.
can promote the protection of anthocyanins against adverse condi- Process yield was calculated as the relationship between total
tions like heat, light and oxygen, besides resulting in less hygro- solids content in the resulting powder and total solids content in
scopic powders. the feed mixture.
The objective of this work was to study the influence of inlet air
temperature, feed flow rate and maltodextrin concentration on 2.4. Analytical methods
process yield, powder moisture content, powder hygroscopicity
and anthocyanin retention, during the microencapsulation of acßai Feed mixtures were analysed for moisture and anthocyanin con-
pulp by spray drying. The effect of inlet air temperature and mal- tent, as well as for viscosity. Spray-dried powders were analysed for
todextrin concentration on powders morphology and particle size moisture content, hygroscopicity and anthocyanin content. More-
distribution was also evaluated. over, the powders produced with different maltodextrin concentra-
tions and at different inlet air temperatures were analysed in respect
2. Materials and methods to their morphology, by means of scanning electron microscopy and
particles size distribution.
2.1. Materials
2.4.1. Mixture viscosity
Frozen acßai pulp was purchased from Palamaz Ind. e Com. Ltda. Feed mixtures viscosity was analysed through the determina-
(Belém, Brazil). The pulp was stored in a freezing chamber at – tion of steady-shear flow curves (shear stress  shear rate), using
18 °C and thawed according to the quantity required for each test. a controlled stress Carri-Med CSL2 500 rheometer (TA Instruments,
Table 1 shows its physicochemical composition. Crawley, England) with acrylic concentric cylinders geometry.
Maltodextrin MOR-REXÒ 1910 (Corn Products, Mogi-Guacßu, Three flow ramps were obtained in a range of shear stress corre-
Brazil), with 9.0 6 DE 6 12.0, was used as carrier agent. sponding to shear rates from 0 to 300 s1. The first one with
increasing shear rate, the second one with decreasing shear rate
2.2. Sample preparation and the third one also with increasing shear rate. This assured that
the viscosity time-dependency was eliminated, once the second
Before entering the spray dryer, acßai pulp was filtered through a and the third curves matched. Trials were performed in triplicate,
qualitative filter paper, in order to reduce fat content, thus reduc- using a new sample for each repetition. Rheograms were analysed
ing the risk of lipid oxidation (lipids content was reduced from according to empirical models and apparent viscosity was calcu-
6.83 ± 0.03% to 0.21 ± 0.01%) and increasing anthocyanin content lated as the relationship between shear stress (r) and shear rate (_c).
per g of dried mass (from 234 ± 4 to 766 ± 5 mg/g of dried mass).
Maltodextrin was added to the filtered pulp under magnetic agita- 2.4.2. Moisture content
tion, until complete dissolution. Powders and mixtures moisture contents were determined
gravimetrically by drying in a vacuum oven at 70 °C until constant
2.3. Spray drying weight (A.O.A.C., 1990).

Spray drying process was performed in a laboratory scale spray 2.4.3. Hygroscopicity
dryer LabPlant SD-05 (Huddersfield, England), with a 1.5 mm Hygroscopicity was determined according to the method pro-
diameter nozzle and main spray chamber of 500 mm  215 mm. posed by Cai and Corke (2000), with some modifications. Samples
The mixture was fed into the main chamber through a peristaltic of each powder (approximately 1 g) were placed at 25 °C in a con-
tainer with NaCl saturated solution (75.29%RH). After one week,
samples were weighed and hygroscopicity was expressed as g of
adsorbed moisture per 100 g dry solids (g/100 g).
Table 1
Composition of acßai (Euterpe oleraceae Mart.)
2.4.4. Anthocyanin content
Analyzed item Mean value Analysis method Anthocyanins were extracted with 95% ethanol/1.5 N HCl
Moisture (wet basis) 85.96 ± 0.11 A.O.A.C. (1990) (85:15, v:v), according to the procedure described by Francis
Proteins 1.43 ± 0.04 A.O.A.C. (1990) (1982). Absorbance was measured in a UV–vis spectrophotometer
Lipids 6.83 ± 0.03 Bligh and Dyer (1959)
at a kmax of 535 nm, and the total anthocyanin content was calcu-
Fibers 4.52 ± 0.22 A.O.A.C. (1990)
Total sugars 0.48 ± 0.05 A.O.A.C. (1990) lated by using the absorption coefficient of 982, corresponding to
Ash 0.44 ± 0.01 A.O.A.C. (1990) cyanidin-3-glucoside in ethanol/1.5 N HCl (Zanatta et al., 2005).
Acidity (% citric acid) 0.34 ± 0.02 A.O.A.C. (1990) For determination of anthocyanin retention, total anthocyanin con-
Anthocyanins (mg/100 g) 32.81 ± 0.57 Francis (1982) tent was calculated as mg/100 g of dry matter, in the mixture be-
pH 5.18 ± 0.02 pH meter
fore entering the spray dryer as well as in the resulting powder.
 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418 413

Table 2 tween 0.909 and 0.961) for all the maltodextrin concentrations.
Experimental design for the spray drying tests Thus, the mixture ‘‘filtered acßai pulp + maltodextrin” could be
Tests Inlet air temperature Feed flow rate Maltodextrin characterized as a fluid with shear-thinning behavior, which is typ-
(°C) (g/min) concentration (%) ical for most of the food materials, specially fruit pulps (Dak et al.,
1 150 (1) 9 (1) 14 (1) 2006; Nindo et al., 2007).
2 190 (+1) 9 (1) 14 (1) The apparent viscosity was calculated as the relationship
3 150 (1) 21 (+1) 14 (1) between shear stress and shear rate. In order to evaluate this
4 190 (+1) 21 (+1) 14 (1)
5 150 (1) 9 (1) 26 (+1)
property as a function of maltodextrin concentration, experimental
6 190 (+1) 9 (1) 26 (+1) data for a shear rate of 100 s1 were considered. The mixtures
7 150 (1) 21 (+1) 26 (+1) apparent viscosity varied from 3.88  103 to 23.70  103 Pa s
8 190 (+1) 21 (+1) 26 (+1) and showed an exponential increase with maltodextrin concentra-
9 138 (1.68) 15 (0) 20 (0)
tion, as shown in Fig. 1.
10 202 (+1.68) 15 (0) 20 (0)
11 170 (0) 5 (1.68) 20 (0)
12 170 (0) 25 (+1.68) 20 (0) 3.2. Response surface analysis
13 170 (0) 15 (0) 10 (1.68)
14 170 (0) 15 (0) 30 (+1.68) The values of process yield, powders moisture content, hygro-
15 170 (0) 15 (0) 20 (0)
scopicity and anthocyanin retention, as well as the outlet air tem-
16 170 (0) 15 (0) 20 (0)
17 170 (0) 15 (0) 20 (0) perature, are shown in Table 3.
Table 4 shows the regression coefficients for the coded second-
order polynomial equation, the F and p values and the determina-
tion coefficients (R2). Some non-significant terms were eliminated
2.4.5. Particles morphology and the resulting equations were tested for adequacy and fitness
Particles morphology was evaluated by scanning electron by the analysis of variance (ANOVA). The fitted models were suit-
microscopy (SEM). Powders were attached to a double-sided adhe- able, showing significant regression, low residual values, no lack of
sive tape mounted on SEM stubs, coated with 3–5 mA gold/palla- fit and satisfactory determination coefficients.
dium under vacuum and examined with a LEO440i scanning
electron microscope (LEICA Electron Microscopy Ltd., Cambridge,
England). SEM was operated at 5 kV with magnifications of
2000 and 7000. 25
Apparent viscosity (s-1)

2.4.6. Particles size distribution

20
The particles size distribution was measured using a laser light
diffraction instrument, Mastersizer S (Malvern Instruments, Mal-
vern, UK). A small powder sample was suspended in isopropanol 15
under magnetic agitation and the particle size distribution was
monitored during each measurement until successive readings be- 10
came constant. The particle size was expressed as D[4,3], the mean
0,0909x
diameter over the volume distribution. y = 1,5391e
5 2
R = 0,9991
2.5. Experimental design
0
0 5 10 15 20 25 30 35
A rotatable central composite design was used to design the
tests for the spray drying of acßai pulp, considering three factors % Maltodextrin
(independent variables): inlet air temperature (138–202 °C), feed Fig. 1. Apparent viscosity at 100 s1 as a function of maltodextrin concentration.
flow rate (5–25 g/min) and maltodextrin concentration (10–30%).
Five levels of each variable were chosen for the trials, including
the central point and two axial points (Khuri and Cornell, 1996),
Table 3
giving a total of 17 combinations (Table 2). The following polyno- Process yield, powders moisture content, hygroscopicity, anthocyanin retention and
mial equation was fitted to the data: outlet air temperature, for the 17 trials of the experimental design

y ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b11 x21 þ b22 x22 þ b33 x23 þ b12 x1 x2 Tests Process Moisture Hygroscopicity Anthocyanin Outlet air
yield (%) content (%) (g/100 g) retention (%) temperature (°C)
þ b13 x1 x3 þ b23 x2 x3 ð1Þ
1 50.33 1.23 ± 0.07 14.73 ± 0.20 84.86 ± 1.74 89 ± 2
The analysis of variance (ANOVA), test for the lack of fit, deter- 2 52.47 0.69 ± 0.08 15.42 ± 0.20 81.40 ± 2.04 111 ± 3
mination of the regression coefficients and the generation of three- 3 42.53 2.89 ± 0.10 15.17 ± 0.06 86.01 ± 2.86 82 ± 2
4 49.32 0.86 ± 0.04 15.10 ± 0.29 79.09 ± 2.78 106 ± 2
dimensional graphs were carried out using the Statistica 5.5 soft- 5 45.74 1.92 ± 0.27 13.83 ± 0.09 84.08 ± 2.31 90 ± 2
ware (StatSoft, Tulsa, USA). 6 48.17 0.64 ± 0.02 14.77 ± 0.10 79.38 ± 1.05 114 ± 2
7 34.39 2.34 ± 0.06 12.54 ± 0.20 83.99 ± 1.23 86 ± 2
8 38.38 1.05 ± 0.07 14.30 ± 0.20 79.29 ± 1.66 106 ± 2
3. Results and discussion 9 46.01 2.56 ± 0.17 12.98 ± 0.11 84.62 ± 1.58 83 ± 2
10 55.66 0.66 ± 0.18 15.13 ± 0.15 77.21 ± 1.10 112 ± 3
11 50.27 0.95 ± 0.24 14.31 ± 0.04 80.05 ± 1.15 111 ± 2
3.1. Mixtures viscosity
12 39.02 1.88 ± 0.02 13.62 ± 0.20 80.82 ± 1.37 94 ± 2
13 50.83 1.78 ± 0.06 15.79 ± 0.29 83.13 ± 1.28 96 ± 2
As stated before, feed mixtures viscosity was determined 14 41.40 1.68 ± 0.40 12.48 ± 0.10 84.06 ± 3.15 99±3
through steady-shear flow curves. The most appropriate mathe- 15 46.04 1.29 ± 0.05 13.71 ± 0.05 84.65 ± 1.44 97 ± 2
16 48.85 1.45 ± 0.10 13.68 ± 0.14 82.42 ± 2.33 100 ± 2
matical model for describing the flow characteristics was the
17 47.21 1.30 ± 0.11 13.34 ± 0.30 79.76 ± 1.43 97 ± 2
Power Law, with flow behavior index values (n) inferior to 1 (be-
414 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418

Table 4 of solids available to be in contact with the chamber wall and to

Coded second-order regression coefficients for process yield, moisture content, paste in it. Thus, the lower the process yield.
hygroscopicity and anthocyanin retention

Coefficients Process Moisture Hygroscopicity Anthocyanin 3.2.2. Powders moisture content

yield (%) content (%) (g/100 g) retention (%) Powders moisture content varied from 0.64% to 2.89%, similarly
b0 48.49 1.48 13.55 82.05 to the range of the moisture contents for spray-dried raisin juice,
b1 2.31 0.61 0.53 2.36 obtained by Papadakis et al. (2006). Moisture content was signifi-
b2 3.74 0.31 0.18 NS
b3 3.21 NS 0.79 NS
cantly influenced by inlet air temperature and feed flow rate.
b11 NS NS 0.25 NS According to Fig. 3, temperature was the variable that showed
b22 1.64 NS 0.22 NS the greatest influence on powders moisture content. At higher inlet
b33 1.12 NS 0.28 NS air temperatures, there is a greater temperature gradient between
b12 NS 0.19 NS NS
the atomized feed and the drying air, resulting in a greater driving
b13 NS NS 0.30 NS
b23 1.27 NS NS NS force for water evaporation and thus producing powders with low-
R2 0.926 0.907 0.909 0.703 er moisture content. Quek et al. (2007), Rattes and Oliveira (2007)
F 20.84 42.16 12.83 35.53 and Grabowski et al. (2006) also observed a reduction of powders
p-value <0.0001 <0.0001 <0.0005 <0.0001 moisture content with increasing air temperatures, studying the
NS: Non-significant (p > 0.05). spray drying of watermelon juice, sodium diclofenac and sweet po-
tato puree, respectively.
The feed flow rate negatively affected powders moisture con-
3.2.1. Process yield tent. Higher flow rates imply in a shorter contact time between
Fig. 2 shows the influence of inlet air temperature, feed flow the feed and the drying air, making the heat transfer less efficient
rate and maltodextrin concentration on the spray drying process and resulting in lower water evaporation. Hong and Choi (2007)
yield. This response was significantly influenced by all the inde- investigated the physicochemical properties of protein-bound
pendent variables. polysaccharide from Agaricus blazei Murill prepared by ultrafiltra-
Increasing temperatures led to higher process yield, which can tion and spray drying process. The authors verified that the pow-
be attributed to the greater efficiency of heat and mass transfer ders moisture content increased with increasing pump rate and
processes occurring when higher inlet air temperatures are used. with decreasing inlet air temperature, and the effect of tempera-
This is in agreement with the results published by Cai and Corke ture was greater than the effect of pump rate.
(2000), working with spray drying of Amaranthus betacyanin
pigments. 3.2.3. Powders hygroscopicity
The feed flow rate showed a negative effect on process yield, Maltodextrin concentration was the variable that most affected
that is, the increase on this parameter resulted in lower process powders hygroscopicity. The lowest hygroscopicity values were
yields, which is also probably related to the slower heat and mass obtained when the highest maltodextrin concentrations were used
transfer occurring when the process was carried out with higher (Fig. 4). This is due to the fact that maltodextrin is a material with
feed flow rates. In addition, when higher feed rates were used, a low hygroscopicity, and confirms its efficiency as a carrier agent.
dripping inside the main chamber was observed, that is, part of Rodríguez-Hernández et al. (2005), working with spray drying of
the mixture passed straight to the chamber and was not atomized, cactus pear juice, and Cai and Corke (2000), working with spray
resulting in a lower process yield. Toneli et al. (2006), working with drying of betacyanin pigments, also verified a reduction of hygro-
spray drying of inulin, also verified an increase on mass production scopicity with increasing maltodextrin concentrations.
rate with increasing air temperatures and decreasing pump speeds. Inlet air temperature and feed flow rate also influenced parti-
Maltodextrin concentration also showed a negative effect on cles hygroscopicity. The lowest hygroscopicity values were ob-
process yield, probably due to the mixture viscosity, which expo- tained with increasing flow rates and decreasing temperatures,
nentially increased with this variable. The increase on feed viscos- which were the variables that affected powders moisture content,
ity can cause more solids to paste in the main chamber wall, thus in an opposite way. This indicates that the lower the particles
reducing the process yield (Cai and Corke, 2000). In addition, the moisture content, the higher their hygroscopicity, i.e., the greater
higher the solids content in the mixture, the higher the amount their capacity to adsorb ambient moisture, which is related to

Fig. 2. Response surface for process yield, for (a) 20% maltodextrin, (b) feed flow rate of 15 g/min and (c) inlet air temperature of 170 °C.
 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418 415

Fig. 3. Response surface for powder moisture content, for (a) 20% maltodextrin, (b) feed flow rate of 15 g/min and (c) inlet air temperature of 170 °C.

Fig. 4. Response surface for powder hygroscopicity, for (a) 20% maltodextrin, (b) feed flow rate of 15 g/min and (c) inlet air temperature of 170 °C.

the greater water concentration gradient between the product and tion (20%). As inlet air temperature was the only variable affecting
the surrounding air. These results are in agreement with those anthocyanin retention (Table 3), the response surface relating inlet
published by Goula et al. (2004), in their work about spray drying air temperature and maltodextrin concentration is identical to the
of tomato pulp. Varying the compressed air flow rate, the drying air one showed at Fig. 5, while the surface relating feed flow rate and
flow rate and the inlet air temperature, the authors verified that maltodextrin concentration is a flat surface without inclination,
powders hygroscopicity inversely increased with powders mois- once none of these variables showed significant effect on this
ture content. response.
Increasing inlet air temperatures led to an increase on anthocy-
3.2.4. Anthocyanin retention anins loss, which is due to the high sensitivity of these pigments to
Fig. 5 shows the surface response obtained for anthocyanin high temperatures. Moreover, according to Quek et al. (2007),
retention, considering the middle point of maltodextrin concentra- powders produced at lower temperatures have a tendency to
agglomerate, because of their higher moisture content. This
agglomeration reduces the powder exposition to oxygen, protect-
ing the pigments against oxidation. Working with spray drying of
watermelon juice, these authors also verified a decrease of lyco-
pene and b-carotene content in the final powder, when increasing
the drying air temperature.
Ersus and Yurdagel (2007) studied the microencapsulation by
spray drying of anthocyanins extracted from black carrot, using
three drying temperatures (160, 180 and 200 °C) and three types
of maltodextrin (10, 20 and 30DE). The authors verified that for
the maltodextrins with higher DE, the powders produced at
160 °C showed higher pigments retention than the ones produced
at higher temperatures. On the other hand, for maltodextrin 10DE,
the drying temperature did not show significant effect on the
retention of anthocyanins.
Cai and Corke (2000) also verified a greater pigment loss with
Fig. 5. Response surface for anthocyanin retention, for 20% maltodextrin. increasing air temperatures in the spray drying of Amaranthus
416 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418

betacyanin pigments and concluded that temperatures higher than obtained at different temperatures may also arise from the physi-
180 °C are not suitable for spray drying of betacyanins. The pow- cal characteristics of the crust, which may be pliable and collapsed
ders produced at lower temperatures showed better storage stabil- (when low and intermediate temperatures are used) or rigid and
ity (better retention of betacyanins) after 16 weeks, at 25 °C and porous (when high temperatures are used).
32%RH. Nijdam and Langrish (2006) also verified the formation of more
rigid particles when using higher temperatures in the spray drying
3.3. Powders morphology of milk. According to the authors, a vacuole forms within the par-
ticle soon after a skin develops on the surface, and it inflates once
Fig. 6 shows the SEM microphotographs of powders produced the particle temperature exceeds the local ambient boiling point
with feed flow rate of 15 g/min and 20% of maltodextrin, at three and the vapour pressure within the vacuole rises above the local
different temperatures. The particles produced with different mal- ambient pressure. When the drying temperature is sufficiently
todextrin concentrations (at 170 °C and feed flow rate of 15 g/min) high, moisture is evaporated very quickly and the skin becomes
were similar to those at Fig. 6c and d and are not shown here. dry and hard, so that the hollow particle cannot deflate when va-
The particles showed spherical shape and various sizes, which is pour condenses within the vacuole as the particle moves into cool-
typical of materials produced by spray drying. When the inlet air er regions of the dryer. However, when the drying temperature is
temperature was low, most of the particles showed a shriveled sur- lower, the skin remains moist and supple for longer, so that the
face, while increasing drying temperatures resulted in a larger hollow particle can deflate and shrivel as it cools.
number of particles with smooth surface. This is related to the dif-
ferences in the drying rate, which is higher for higher tempera- 3.4. Particle size distribution
tures, causing faster water evaporation and leading to the
formation of a smooth and hard crust. According to Allamilla-Belt- Figs. 7 and 8 show the particle size distribution for the powders
rán et al. (2005), the morphological differences between powders produced at different temperatures (with feed flow rate of 15 g/min

Fig. 6. Micrographs of particles at different temperatures and in different magnifications, with feed flow rate of 15 g/min and 20% maltodextrin: (a) 138 °C, 2000; (b) 138 °C,
7000; (c) 170 °C, 2000; (d) 170 °C, 7000; (e) 202 °C, 2000; (f) 202 °C, 7000.
 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418 417

Table 6
10
138ºC Mean diameter D[4,3] of powders produced at different temperatures, with feed flow
rate of 15 g/min and 20% maltodextrin
170ºC
8 202ºC Maltodextrin concentration (%) Mean diameter (lm)
10 13.27
Volume (%)

20 16.12
6 30 21.35

trin) and with different maltodextrin concentrations (at 170 °C and

2 15 g/min), respectively.
The increase on inlet air temperature resulted in larger particles
(Table 5), which is related to the higher swelling caused by higher
0
temperatures. According to Reiniccius (2001), drying at conditions
0.01 0.1 1 10 100 1000
that result in faster drying rates produce larger particles than drying
Particle size (μm)
under conditions that result in slower drying, which is due to the
Fig. 7. Particle size distribution of powders produced at different temperatures, fact that very fast drying sets up a structure early and does not allow
with feed flow rate of 15 g/min and 20% maltodextrin. the particles to shrink during drying. When inlet air temperature is
low, the particle remains more shrunk and thus, with smaller diam-
eter. Nijdam and Langrish (2006) obtained similar results, working
10 with the production of milk powder at 120 °C and 200 °C.
10%MD Higher maltodextrin concentrations also led to the production
20%MD of larger particles (Table 6), which may be related to the feed vis-
8 30%MD cosity, which exponentially increased with maltodextrin concen-
tration. According to Masters (1991), the mean liquid droplet size
Volume (%)

6 varies directly with the liquid viscosity at constant atomizer speed.

The higher the liquid viscosity, the larger the droplets formed dur-
ing atomization and thus, the larger the particles obtained by spray
4 drying. This is in agreement with the results published by Jinapong
et al. (2008), for instant soymilk powders produced by ultrafiltra-
tion and spray-dried in a rotary atomizer. Keogh et al. (2003) ob-
2
served a linear increase of the particles size with feed viscosity,
working with spray drying of ultrafiltered whole milk concen-
0 trated, in a two-fluided nozzle atomizer. In both works, the authors
0.01 0.1 1 10 100 1000 attributed the increase on particles size to the increase on feed
viscosity.
Particle size (μm)

Fig. 8. Particle size distribution of powders produced at different maltodextrin 4. Conclusions

concentrations, with feed flow rate of 15 g/min, at 170 °C.
Inlet air temperature showed significant effect on all the re-
sponses studied. Increasing temperature led to higher process yield
and 20% maltodextrin) and with different maltodextrin concentra- and powder hygroscopicity, and to lower moisture content and
tions (at 170 °C and 15 g/min), respectively. These conditions are anthocyanin retention. Feed flow rate negatively influenced pro-
relative to the axial points of the experimental design, in which cess yield and hygroscopicity, and positively influenced moisture
two variables are fixed and another varies in three levels. content. Maltodextrin concentration had negative effect on powder
According to Figs. 7 and 8, particles showed a bimodal distribu- hygroscopicity, confirming its efficiency as a carrier agent. The in-
tion, that is, there are two distinct peaks, each one representing a crease on this variable also caused a reduction on process yield,
predominant size. This is particularly interesting in the case of probably due to the increase on feed viscosity. In respect to powder
powders, once the ‘‘population” of smaller particles can penetrate morphology, increasing temperatures resulted in a greater number
into the spaces between the larger ones, thus occupying less space. of particles with smooth surface and with larger sizes, due to the
The presence of larger particles may be attributed to a beginning of higher drying rates. The increase on maltodextrin concentration
the agglomeration process, where the formation of irreversible link also led to the production of larger particles, which is related to
bridges leads to the production of particles with greater size. the increase on feed viscosity.
Tables 5 and 6 show the values of D[4,3] (mean diameter over
the volume distribution) for the powders produced at different Acknowledgements
temperatures (with feed flow rate of 15 g/min and 20% maltodex-
The authors thank CNPq and Capes for the financial support.

Table 5 References
Mean diameter D[4,3] of powders produced at different temperatures, with feed flow
rate of 15 g/min and 20% maltodextrin Abadio, F.D.B., Domingues, A.M., Borges, S.V., Oliveira, V.M., 2004. Physical
properties of powdered pineapple (Ananás comosus) juice – effect of malt
Temperature (°C) Mean diameter (lm)
dextrin concentration and atomization speed. Journal of Food Engineering 64
138 13.38 (3), 285–287.
170 16.12 Allamilla-Beltrán, L., Chanona-Pérez, J.J., Jiménez-Aparicio, A.R., Gutiérrez-López,
202 20.11 G.F., 2005. Description of morphological changes of particles along spray drying.
Journal of Food Engineering 67 (1–2), 179–184.
418 R.V. Tonon et al. / Journal of Food Engineering 88 (2008) 411–418

A.O.A.C., 1990. Official Methods of Analysis, 14th ed. Association of Official Khuri, A.J., Cornell, F.A., 1996. Response Surfaces: Design and Analyses, second ed.
Analytical Chemists, Washington, D.C. Marcel Dekker, NY.
Bhandari, B.R., Senoussi, A., Dumoulin, E.D., Lebert, A., 1993. Spray drying of Masters, K., 1991. Spray Drying Handbook, fifth ed. Longman Scientific and
concentrated fruit juices. Drying Technology 11 (5), 1081–1092. Technical, London.
Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and Nijdam, J.J., Langrish, T.A.J., 2006. The effect of surface composition on the
purification. Canadian Journal of Biochemistry and Physiology 37 (8), 911–917. functional properties of milk powders. Journal of Food Engineering 77 (4),
Cai, Y.Z., Corke, H., 2000. Production and properties of spray-dried Amaranthus 919–925.
betacyanin pigments. Journal of Food Science 65 (6), 1248–1252. Nindo, C.I., Tang, J., Powers, J.R., Takhar, P.S., 2007. Rheological properties of
Cano-Chauca, M., Stringheta, P.C., Ramos, A.M., Cal-Vidal, J., 2005. Effect of the blueberry puree for processing applications. LWT – Food Science and
carriers on the microstructure of mango powder obtained by spray drying and Technology 40 (2), 292–299.
its functional characterization. Innovative Food Science and Emerging Papadakis, S.E., Gardeli, C., Tzia, C., 2006. Spray drying of raisin juice concentrate.
Technologies 5 (4), 420–428. Drying Technology 24 (2), 173–180.
Coisson, J.D., Travaglia, F., Piana, G., Capasso, M., Arlorio, M., 2005. Euterpe oleracea Quek, S.Y., Chok, N.K., Swedlund, P., 2007. The physicochemical properties of spray-
juice as a functional pigment for yogurt. Food Research International 38 (8–9), dried watermelon powder. Chemical Engineering and Processing 46 (5), 386–
893–897. 392.
Dak, M., Verma, R.C., Jaaffrey, S.N.A., 2006. Effect of temperature and concentration Rattes, A.L.R., Oliveira, W.P., 2007. Spray drying conditions and encapsulating
on rheological properties of ‘‘Kesar” mango juice. Journal of Food Engineering composition effects on formation and properties of sodium diclofenac
80 (4), 1011–1015. microparticles. Powder Technology 171 (1), 7–14.
Del Pozo-Insfran, D., Brenes, C.H., Talcott, S.T., 2004. Phytochemical composition Reiniccius, G.A., 2001. Multiple-core encapsulation – the spray drying of food
and pigment stability of acßai (Euterpe oleraceae Mart.). Journal of Agricultural ingredients. In: Vilstrup, P. (Ed.), Microencapsulation of Food Ingredients.
and Food Chemistry 52 (6), 1539–1545. Leatherhead Publishing, Surrey, pp. 151–185.
Dib Taxi, C.M., Menezes, H.C., Santos, A.B., Grosso, C.R., 2003. Study of the Ré, M.I., 1998. Microencapsulation by spray drying. Drying Technology 16 (6),
microencapsulation of camu–camu (Myrciaria dubia) juice. Journal of 1195–1236.
Microencapsulation 20 (4), 443–448. Righetto, A.M., Netto, F.M., 2005. Effect of encapsulating materials on water
Ersus, S., Yurdagel, U., 2007. Microencapsulation of anthocyanin pigments of black sorption, glass transition and stability of juice from immature acerola.
carrot (Daucuscarota L.) by spray dryer. Journal of Food Engineering 80 (3), 805– International Journal of Food Properties 8 (2), 337–346.
812. Rodríguez-Hernández, G.R., González-García, R., Grajales-Lagunes, A., Ruiz-Cabrera,
Francis, F.J., 1982. Analysis of anthocyanins. In: Markakis, P. (Ed.), Anthocyanins as M.A., 2005. Spray-drying of cactus pear juice (Opuntia streptacantha): effect on
Food Colors. Academic Press, NY, pp. 182–205. the physicochemical properties of powder and reconstituted product. Drying
Goula, A.M., Adamopoulos, K.G., Kazakis, N.A., 2004. Influence of spray drying Technology 23 (4), 955–973.
conditions on tomato powder properties. Drying Technology 22 (5), 1129–1151. Schauss, A.G., Wu, X., Prior, R.L., Ou, B., Huang, D., Owens, J., Agarwal, A., Jensen, G.S.,
Grabowski, J.A., Truong, V.D., Daubert, C.R., 2006. Spray-drying of amylase Hart, A.N., Shanbrom, E., 2006. Antioxidant capacity ond other bioactivities of
hydrolyzed sweet potato puree and physicochemical properties of powder. the freeze-dried amazoniam palm berry, Euterpe oleracea Mart. (Acai). Journal of
Journal of Food Science 71 (5), E209–E217. Agricultural and Food Chemistry 54 (22), 8604–8610.
Hong, J.H., Choi, Y.H., 2007. Physico-chemical properties of protein-bound Toneli, J., Park, K.J., Murr, F., Negreiros, A., 2006. Spray drying optimization to obtain
polysaccharide from Agaricus blazei Murill prepared by ultrafiltration and spray inulin powder. In: Proceedings of the 15th International Drying Symposium
drying process. International Journal of Food Science and Technology 42 (1), 1–8. (IDS 2006). Budapest, Hungary.
Jinapong, N., Suphantharika, M., Jamnong, P., 2008. Production of instant soymilk Wang, H., Cao, G., Prior, R.L., 1997. Oxygen radical absorbing capacity of
powders by ultrafiltration, spray drying and fluidized bed agglomeration. anthocyanins. Journal of Agricultural and Food Chemistry 45 (2), 304–309.
Journal of Food Engineering 84 (2), 194–205. Zanatta, C.F., Cuevas, E., Bobbio, F.O., Winterhalter, P., Mercadante, A.Z., 2005.
Keogh, M.K., Murray, C.A., O’Kennedy, B.T., 2003. Effects of ultrafiltration of whole Determination of anthocyanins from camu–camu (Myrciaria dubia) by HPLC-
milk on some properties of spray-dried milk powders. International Dairy PDA, HPLC-MS and NMR. Journal of Agricultural and Food Chemistry 53 (24),
Journal 13 (12), 995–1002. 9531–9535.