Journal of Food Engineering 94 (2009) 215–221
Contents lists available at ScienceDirect
Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng
Water sorption and glass transition temperature of spray dried açai (Euterpe oleracea Mart.) juice Renata V. Tonon a,*, Alessandra F. Baroni b, Catherine Brabet c, Olivier Gibert c, Dominique Pallet c, Míriam D. Hubinger a a
Faculty of Food Engineering, State University of Campinas, Rua Monteiro Lobato, 80, P.O. Box 6121, 13083-862 Campinas, SP, Brazil Faculty of Engineering, Mauá Institute of Technology, Praça Mauá 1, 09580-900, São Caetano do Sul, SP, Brazil c Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD), Department PERSYST, UMR QualiSud, Montpellier, France b
a r t i c l e
i n f o
Article history: Received 21 January 2009 Received in revised form 6 March 2009 Accepted 12 March 2009 Available online 20 March 2009 Keywords: Açai Sorption isotherms Water activity Glass transition Stability
a b s t r a c t Sorption isotherms and glass transition temperature (Tg) of powdered açai juice were evaluated in this work. Powders were produced by spray drying using different materials as carrier agents: maltodextrin 10DE, maltodextrin 20DE, gum Arabic and tapioca starch. The sorption isotherms were determined by the gravimetric method, while the Tg of powders conditioned at various water activities were determined by differential scanning calorimetry. As results, experimental data of water adsorption were well ﬁtted to both BET and GAB models. Powders produced with maltodextrin 20DE and gum Arabic showed the highest water adsorption, followed by those produced with maltodextrin 10DE and with tapioca starch, respectively. With respect to the glass transition temperature, Gordon–Taylor model was able to predict the strong plasticizing effect of water on this property. Both aw and Tg were used to determine the critical conditions for food storage, at which powders are not susceptible to deteriorative changes such as collapse, stickiness and caking. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction For many years, water activity has been considered more important than total amount of water, concerning to quality and stability of foodstuff. Water sorption isotherms are important thermodynamic tools for predicting the interactions between water and food components. They describe the relationship between water activity and the equilibrium moisture content of a food product and provide useful information for food processing operations such as drying, packaging and storage (Lomauro et al., 1985). Several authors have coupled the concepts related to water activity with those of glass transition temperature (Tg) in order to evaluate food stability, thus providing an integrated approach to the role of water in food (Slade and Levine, 1991; Sablani et al., 2004; Shrestha et al., 2007a; Symaladevi et al., 2009). The glass transition temperature (Tg) is deﬁned as the temperature at which an amorphous system changes from the glassy to the rubbery state. Theoretically, in the glassy state, the high viscosity of the matrix (about 1012 Pa s) does not allow the occurrence of diffusion-controlled reactions. However, some authors have demonstrated that some diffusion-controlled reactions, such as non-enzimatic browning, may occur, even at the glassy state (Schebor et al., 1999; Miao and Roos, 2004). As the temperature increases above Tg, various * Corresponding author. Tel.: +55 19 37884036; fax: +55 19 37884027. E-mail address: [email protected]
(R.V. Tonon). 0260-8774/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jfoodeng.2009.03.009
changes such as increase of free volume and speciﬁc heat, as well as decrease of viscosity, are noticed. These factors control various time-dependent structural transformations, such as stickiness, collapse and crystallization during food processing and storage. Açai is a Brazilian fruit of Euterpe oleracea Mart. palm, highly energetic and with high anthocyanin and phenolic content, being recognized for its great antioxidant capacity (Schauss et al., 2006; Coisson et al., 2005). However, it is a highly perishable fruit with short shelf life, thus needing additional processes in order to increase its shelf life and improve its stability. Spray drying is a process that results in products with good quality, low water activity and easier transport and storage, and has been widely used to produce fruit juice powders (Quek et al., 2007; Cano-Chauca et al., 2005; Abadio et al., 2004; Dib Taxi et al., 2003). However, such powders may have some problems in their properties, such as stickiness and high hygroscopicity, due to the presence of low molecular weight sugars and acids, which have low glass transition temperatures (Bhandari et al., 1993). Thus, they can stick on the dryer chamber wall during drying, leading to low product yield and operational problems. An alternative widely used to dry such products has been the addition of high molecular weight additives to the product before being atomized, in order to increase its glass transition temperature (Bhandari and Howes, 1999; Truong et al., 2005a; Shrestha et al., 2007b). The more common carrier agents used for fruit juices are maltodextrins and gum Arabic (Gabas et al., 2007; Righetto and Netto, 2005; Cano-Chauca et al., 2005).
R.V. Tonon et al. / Journal of Food Engineering 94 (2009) 215–221
Nomenclature aw CBET CGAB k KGAB n N Tg VE
water activity constant of Eqs. (1) and (3) constant of Eq. (2) constant of Eq. (5) constant of Eq. (3) number of adsorbed layers population of experimental data glass transition temperature (K) experimental value
Maltodextrins are products of starch hydrolysis, consisting of aunits linked mainly by (1?4) glycosidic bonds and described by their dextrose equivalency (DE), which determines their reducing capacity and is inversely related to their average molecular weight (BeMiller and Whistler, 1996). Gum Arabic is a natural plant exudates of Acacia trees, which consists of a complex heteropolysaccharide with highly ramiﬁed structure, with a main chain formed of D-galactopyranose units joined by b-D-glycosidic bonds 1?3). Side chains with different chemical structures are linked to the main chain by b-(1?6) bonds (BeMiller and Whistler, 1996). Tapioca starch is a ﬁne ﬂour reﬁned from cassava root and obtained by natural fermentation, which has been used by some Brazilian companies as a carrier agent in the production of powdered fruit juices, in order to assure that no genetically modiﬁed material was used in the powder production, showing advantages like absence of ﬂavor and taste, high availability and low cost. The objective of this work was to provide experimental data of water sorption and glass transition temperature of spray dried açai juice produced with different carrier agents, in order to obtain useful information about powders stability. Sorption isotherms were modeled according to BET and GAB models, while Tg was modeled according to the Gordon–Taylor model. Critical storage conditions were determined based on the water adsorption and the glass transition temperature. D-glucose
2. Material and methods 2.1. Material Frozen açai pulp was purchased from Palamaz Ind. e Com. Ltda. (Belém, Brazil). The pulp was stored in a freezing chamber at 18 °C and thawed according to the quantity required for each test. The use of frozen pulp was necessary due to the short shelf life of açai, associated to distance between the harvest and the processing places. Açai is a very perishable fruit and must be pulped up to 24 h after being collected. The pulp, even when stored under refrigeration, has a maximal shelf life of 12 hours. The carrier agents used were: maltodextrin MOR-REXÒ 1910 (Corn Products, Mogi-Guaçu, Brazil) with 10DE, maltodextrin MOR-REXÒ 1920 (Corn Products, Mogi-Guaçu, Brazil) with 20DE, gum Arabic Instantgum BA (Colloïdes Naturels, São Paulo, Brazil) and tapioca starch (National Starch and Chemical Company, São Paulo, Brazil). 2.2. Sample preparation Before entering the spray dryer, açai pulp was ﬁltered through a qualitative ﬁlter paper, in order to reduce fat content, thus lowering the risk of lipid oxidation (lipids content was reduced from 6.53 ± 0.03% to 0.21 ± 0.01%). Moreover, solids in suspension were eliminated, making easier the product passage through the nozzle
VP Xe Xm w
predicted value equilibrium moisture content (g water/g dry matter) monolayer moisture content (g water/g dry matter) weight fractions (g/g total)
Subscripts c critical s solids w water
atomizer (total solids content were reduced from 14% to 3%). Table 1 shows the physicochemical composition of açai pulp before and after ﬁltration. The carrier agents were then added to the ﬁltered pulp, in a concentration of 6% (w/w), under magnetic agitation, until complete dissolution. This concentration was selected in a preliminary study, as the minimum concentration at which it was possible to dry the juice without excessive powder stickiness on the chamber wall. 2.3. Spray drying Spray drying process was performed in a laboratory scale spray dryer LabPlant SD-05 (Huddersﬁeld, England), with a 1.5 mm diameter nozzle and main spray chamber of 500 215 mm. The mixture (at 20 °C) was fed into the main chamber through a peristaltic pump, drying air ﬂow rate was 73 m3/h and compressor air pressure was 0.06 MPa. The feed ﬂow rate used was 15 g/min, inlet and outlet air temperature were 140 ± 2 °C and 78 ± 2 °C. It was not possible to spray dry the juice without adding a carrier agent, due to the high powder stickiness on the chamber wall, resulting from the presence of sugars and acids in the açai juice. 2.4. Sorption isotherms Sorption isotherms were determined by the gravimetric method. Eight saturated salt solutions were prepared (LiCl, CH3COOK, MgCl2, K2CO3, Mg(NO3)2, KI, NaCl and KCl) in order to provide relative humidity values of 11.3%, 22.6%, 32.8%, 43.2%, 52.9%, 68.9%, 75.3% and 84.3%, respectively (Greenspan, 1977). Triplicate samples of 1 g of açai powder were weighed into aluminum vials and equilibrated over such saturated solutions, in dessicators at 25 °C. The required time for equilibration was 3–4 weeks, based on the change in samples weight, which did not exceed 0.1%. The equilibrium moisture content was determined in a vacuum oven, at 70 °C until constant weight (AOAC, 1990). Sorption isotherms are generally described by mathematical models based on empirical and/or theoretical criteria, which can be easily found in the literature. The most commonly used equations are those of Brunauer–Emmett–Teller (BET) and GuggenTable 1 Composition of pure and ﬁltered açai pulp (Euterpe oleraceae Mart.). Analyzed item
Moisture (% wet basis) Proteins (%) Lipids (%) Fibers (%) Total sugars (%) Ash (%) Acidity (% citric acid)
85.96 ± 0.11 1.43 ± 0.04 6.53 ± 0.03 4.52 ± 0.22 0.48 ± 0.05 0.44 ± 0.01 0.34 ± 0.02
96.94 ± 0.01 0.50 ± 0.05 0.21 ± 0.01 n.d.* 1.69 ± 0.12 0.39 ± 0.01 0.32 ± 0.02
AOAC (1990) AOAC (1990) Bligh and Dyer (1959) AOAC (1990) AOAC (1990) AOAC (1990) AOAC (1990)
n.d. = Not detected.
R.V. Tonon et al. / Journal of Food Engineering 94 (2009) 215–221
heim–Anderson–de Boer (GAB) models (Eqs. (1) and (2), respectively), once they have some theoretical background and their parameters provide a physical meaning related to the sorption process, as compared to empirical models
X m C BET aw ; ð1 aw Þ ð1 aw þ C BET aw Þ X m C GAB K GAB aw Xe ¼ : ½ð1 K GAB aw Þð1 K GAB aw þ C GAB K GAB aw Þ Xe ¼
However, the BET model fails for water activities above 0.5 and the Eq. (2) is not able to accurately predict the sorption behavior (Jonquières and Fane, 1998; Goula et al., 2008). Brunauer et al. (1938), in their original publication, derived a modiﬁed model with three parameters, considering a limited number of adsorbed layers and allowing the modeling for water activities up to 0.9 (Eq. (3))
h i X m C BET aw 1 ðn þ 1Þðaw Þn þ nðaw Þnþ1 h i: Xe ¼ ð1 aw Þ 1 þ ðC BET 1Þaw C BET ðaw Þnþ1
Thus, in this work, sorption isotherms data were modeled according to GAB and modiﬁed BET models (Eqs. (2) and (3), respectively), using the Solver algorithm of Microsoft Excel (Microsoft, Redmond, USA). The goodness of ﬁt was evaluated by the determination coefﬁcient (R2) and the mean relative deviation modulus (E):
N 100 X jV E V P j : VE N i¼1
about 5 mg were placed into differential scanning calorimetry (DSC) aluminum pans (20 ll) and equilibrated over saturated salt solutions in desiccators at 25 °C. After equilibrium was reached, samples were hermetically sealed, weighed and taken for DSC analysis. Samples were heated at 10 °C/min from 70 to 120 °C and an empty pan was used as reference. Depending on the sample moisture contents, different initial and ﬁnal temperatures were used. Two runs were performed for each sample, once the second scanning reduces the enthalpy relaxation of the amorphous powder, which appears in the ﬁrst scan, thereby enhancing the accuracy of Tg measurement on the DSC thermogram. Equipment calibration was performed with indium (Tmelting = 156.6 °C) and veriﬁcation with azobenzol (Tmelting = 68.0 °C). Dry helium, 25 ml/ min, was used as the purge gas. All analyses were done in triplicate and data were treated by the software Universal Analysis 2.6 (TA Instruments, New Castle, USA). The plasticizing effect of water on glass transition was described by the Gordon–Taylor model (Gordon and Taylor, 1952), represented by Eq. (5), where Tgw was taken as 135 °C (Johari et al., 1987)
ws T gs þ kww T gw : ws þ kww
The parameters for this model were estimated using the Solver algorithm of MS Excel (Microsoft, Redmond, USA) and the goodness of ﬁt was evaluated by the determination coefﬁcient (R2) and the mean relative deviation modulus (E).
3. Results and discussion 2.5. Glass transition temperature 3.1. Sorption isotherms The glass transition temperature was determined by differential scanning calorimetry in a TA–MDSC-2920 (TA Instruments, New Castle, USA) equipped with a mechanical refrigeration system (RCS – refrigerated cooling accessory). Açai powder samples of
Equilibrium moisture content for açai powders produced with the different carrier agents and stored at the eight water activities are shown in Table 2.
Table 2 Equilibrium moisture content of powders produced with different carrier agents. aw
0.112 0.226 0.328 0.432 0.529 0.689 0.753 0.843
Equilibrium moisture content, Xe (g/g dry matter) Maltodextrin 10DE
0.0142 ± 0.0003 0.0370 ± 0.0019 0.0390 ± 0.0003 0.0482 ± 0.0012 0.0715 ± 0.0005 0.1305 ± 0.0005 0.1604 ± 0.0023 0.2447 ± 0.0057
0.0171 ± 0.0004 0.0332 ± 0.0017 0.0362 ± 0.0020 0.0553 ± 0.0017 0.0775 ± 0.0008 0.1501 ± 0.0010 0.1969 ± 0.0003 0.3221 ± 0.0032
0.0167 ± 0.0003 0.0429 ± 0.0004 0.0418 ± 0.0017 0.0627 ± 0.0011 0.0822 ± 0.0011 0.1505 ± 0.0007 0.1974 ± 0.0019 0.3145 ± 0.0033
0.0068 ± 0.0001 0.0328 ± 0.0020 0.0340 ± 0.0002 0.0470 ± 0.0006 0.0581 ± 0.0019 0.0907 ± 0.0004 0.1180 ± 0.0003 0.1850 ± 0.0041
Table 3 Estimated BET and GAB parameters for açai juice powder produced with different carrier agents. Model
Carrier agents Maltodextrin 10DE
Xm CBET N R2 E (%)
0.045 3.45 21.47 0.997 6.58
0.058 1.67 27.45 0.999 8.60
0.054 2.96 31.41 0.998 6.08
0.031 6.33 28.03 0.995 15.46
Xm CGAB KGAB R2 E (%)
0.050 2.83 0.962 0.996 6.29
0.063 1.51 0.981 0.999 9.05
0.053 3.07 0.996 0.997 6.08
0.032 5.75 0.986 0.995 14.73
R.V. Tonon et al. / Journal of Food Engineering 94 (2009) 215–221
Experimental data of the sorption isotherms for spray dried açai juice produced with different carrier agents were ﬁtted to GAB and BET (three-parameters) models. The estimated parameters are presented in Table 3. According to Table 3, both BET and GAB models showed a good ﬁt to experimental data, with high R2 values and satisfactory mean relative deviation modulus (E). As the R2 values calculated by the BET model were slightly higher (for maltodextrin 10DE and gum Arabic) and the mean relative deviation moduli were slightly lower (for maltodextrin 20DE), this model was chosen to represent the sorption data of powdered açai juice. The sorption isotherms ﬁtted to the BET model are presented in Fig. 1 and they were of type III according to Brunauer’s classiﬁcation (Brunauer et al., 1938). This type of curve was also observed by Wang et al. (2008) for freezedried Chinese gooseberry, Gabas et al. (2007) for vacuum dried pineapple added of maltodextrin and gum Arabic, and Kurozawa et al. (2009) for spray dried chicken meat hydrolysate protein produced with these same carrier agents. Both BET and GAB models are based on the monolayer moisture concept and provide the value of the monolayer moisture content of the material (Xm), considered as the safe moisture for dried foods during preservation, while most other models lack this parameter. The monolayer moisture content (Xm) indicates the amount of water that is strongly adsorbed to speciﬁc sites at the food surface and is considered an important value to assure food stability. The Xm values obtained for spray dried açai juice varied from 3.1% to 5.8% according to the BET model and from 3.2% to 6.3% according to the GAB model, which are in agreement with the values obtained by Righetto and Netto (2005) and Moraga et al. (2006), for spray dried acerola and freeze-dried kiwi, respectively. Pérez-Alonso et al. (2006) determined the sorption isotherms for maltodextrin 10DE and they also obtained lower monolayer moisture contents for maltodextrin 10DE (6.96–7.35%) than for gum Arabic (8.11– 11.0%) at temperatures of 25, 35 and 40 °C. The authors attributed such results to a combination of factors, such as the conformation and topology of molecule and the hydrophilic/hydrophobic sites adsorbed at the interface. According to Fig. 1, the powders produced with tapioca starch showed the lowest water adsorption, followed by that produced with maltodextrin 10DE, while the samples produced with maltodextrin 20DE and gum Arabic were the most hygroscopic ones. Such differences in water adsorption can be explained by the chemical structure of each agent. Maltodextrin 20DE and gum Arabic have a great number of ramiﬁcations with hydrophilic groups and therefore, can easily adsorb moisture from the ambient air. Maltodextrin 10DE is less hydrolyzed, showing less hydrophilic groups and thus adsorbing less water. Tapioca starch is a native
Equilibrium moisture content (g water/g solids)
starch (not hydrolyzed), which explains its lower hygroscopicity. Cai and Corke (2000) and Ersus and Yurdagel (2007), working with microencapsulation of betacyanins and anthocyanins, respectively, using maltodextrins with different DE’s, also veriﬁed an increase of hygroscopicity with increasing DE’s and attributed such increase to the lower molecular weight of the maltodextrins with higher DE, which have shorter chains and, therefore, more hydrophilic groups. Changes on the physical characteristics of the powders stored at 25 °C at different relative humidities could be observed. When stored at relative humidities of 43% or lower, particles remained as a free-ﬂowing powder, for all the carrier agents used. At 53%, particles showed a beginning of agglomeration, and the powder could not ﬂow so easily. When stored at higher relative humidities, physical transformations were more evident. At aw’s above 0.69, the particles produced with maltodextrins and gum Arabic showed the formation of hard and dark blocks, resulting from the compaction, an advanced stage in caking associated with a pronounced loss of system integrity as a result of thickening of interparticle bridges owing to ﬂow, reduction of interparticle spaces and deformation of particle clumps under pressure (Aguilera et al., 1995). The samples produced with maltodextrin 20DE and with gum Arabic, when stored at the highest relative humidity (84%), had the appearance of a highly sticky liquid. According to Aguilera et al. (1995), in this stage of caking the interparticles bridges disappear as a result of sample liquefaction and low molecular weigh fractions are solubilized. The particles produced with tapioca starch, even when stored at higher water activities, were agglomerated but did not show the formation of blocks or liquefaction, which is probably related to its lower water adsorption as compared to the powders produced with the other agents. 3.2. Glass transition temperature A typical DSC thermogram obtained for the sample produced with tapioca starch at aw = 0.328 is shown in Fig. 2. Similar curves were obtained for all the other samples conditioned at the various water activities. In general, the thermograms showed the typical second-order transition that produces a step change in the heat ﬂow due to changes in heat capacity at the temperature of phase transition. The glass transition temperature was taken as the mid point of the glass transition. Table 4 shows the glass transition temperatures obtained for each powder stored at different water activities. The Tg values obtained for spray dried açai juice were in a similar range to those obtained by Silva et al. (2006) for freeze-dried camu–camu pulp added of maltodextrin. The authors observed lower values for the freeze-dried pulp produced without maltodextrin, conﬁrming the efﬁciency of the addition of a carrier agent on
MD10 - experimental MD10 - BET model MD20 - experimental MD20 - BET model GA - experimental GA - BET model TS - experimental TS - BET model
0.6 0.5 0.4 0.3 0.2 0.1 0 0
aw Fig. 1. Sorption isotherms of spray dried açai juice produced with different carrier agents (MD10 = maltodextrin 10DE; MD20 = maltodextrin 20DE; GA = gum Arabic; TS = tapioca starch).
Fig. 2. DSC proﬁle for spray dried açai juice produced with tapioca starch and conditioned at aw = 0.328.
R.V. Tonon et al. / Journal of Food Engineering 94 (2009) 215–221 Table 4 Glass transition temperature of powders produced with different carrier agents.
0.112 0.226 0.328 0.432 0.529 0.689 0.753 0.843
Glass transition temperature (°C) Maltodextrin 10DE
70.20 ± 3.06 61.53 ± 0.22 60.31 ± 2.71 51.79 ± 2.24 40.70 ± 3.31 5.35 ± 0.93 14.15 ± 2.75 54.59 ± 1.22
64.20 ± 0.24 57.20 ± 0.31 53.37 ± 0.23 43.45 ± 2.42 32.95 ± 3.65 3.63 ± 0.82 28.85 ± 1.17 61.93 ± 0.45
74.37 ± 0.34 63.62 ± 0.83 60.87 ± 1.01 45.46 ± 2.72 40.70 ± 0.21 6.11 ± 0.97 14.21 ± 0.93 56.32 ± 1.08
73.95 ± 0.25 62.26 ± 1.69 52.20 ± 2.31 40.59 ± 0.20 31.33 ± 1.63 6.44 ± 0.23 15.58 ± 1.96 56.62 ± 0.63
the improvement of powder characteristics and stability. Righetto and Netto (2005) obtained lower Tg values for spray dried acerola juice with maltodextrin 25DE and gum Arabic, as compared to açai, which can be attributed to the higher sugar and acid content present in acerola, or to the lower ratio of fruit solids: added solids used by them. The authors also observed higher Tg for the sample produced with gum Arabic as compared to that produced with maltodextrin 25DE, which is in agreement with the results of this study. The powder produced with maltodextrin 10DE showed higher glass transition temperatures as compared to that produced with maltodextrin 20DE, which is related to the decrease in the molecular weight, which decreases the Tg (Roos et al., 1996). The powder produced with tapioca starch was expected to have higher Tg than those produced with maltodextrins, since it is a native starch with higher molecular weight. However, the values obtained for such agent were similar or even lower than the obtained for the other powders. This can be attributed to the low solubility of tapioca starch at room temperature. When this agent was added to açai juice, it did not reach complete dissolution and a little quantity was precipitated inside the pipe of the spray dryer, during the process. Thus, the content of tapioca starch in the ﬁnal powder was not the same that for maltodextrins and gum Arabic, which were totally soluble. As the amount of carrier agent was lower, the amount of solids content of juice was higher, which can explain the Tg values obtained, lower than the expected. According to Table 4, the glass transition temperature decreased with increasing moisture content due to the plasticizing effect of water. The same trend was observed by several authors working with many fruits such as tomato, gooseberry, kiwi, strawberry and pineapple (Goula et al., 2008; Wang et al., 2008; Moraga et al., 2004, 2006; Telis and Sobral, 2001). Experimental data of Tg well ﬁtted to the Gordon–Taylor model, showing satisfactory values of R2 and E. The estimated parameters are presented in Table 5 and the ﬁtted curves are shown in Fig. 3. According to Table 5, Tgs values varied from 79 to 97 °C. With respect to the parameter k, the values obtained by Gordon–Taylor model were between 3.56 and 6.87, similar to those obtained for tomato, camu–camu, kiwi, garlic powder and some berries (Goula et al., 2008; Silva et al., 2006; Moraga et al., 2006; Rahman et al., 2005; Khallouﬁ et al., 2000). This parameter controls the degree of curvature of Tg dependence on water content (in a binary sys-
MD10 experimental MD10 - Gordon-Taylor model MD20 experimental MD20 - Gordon-Taylor model GA experimental GA - Gordon-Taylor model TS experimental TS - Gordon-Taylor model
0 -50 -100 -150 0
Ws (g solids/g total) Fig. 3. Glass transition temperature as a function of solids content for spray dried açai juice produced with different carrier agents (MD10 = maltodextrin 10DE; MD20 = maltodextrin 20DE; GA = gum Arabic; TS = tapioca starch).
tem) and can be related to the strength of the interaction between the system components (Gordon and Taylor, 1952). The Tgs values obtained can explain the difﬁculty on drying the pure açai juice (without adding a carrier agent). It has been shown that the addition of carrier agents such as maltodextrins and gum Arabic leads to a considerable increase on Tgs. Silva et al. (2006) veriﬁed an increase in the Tgs of freeze-dried camu–camu from 74.59 (pure pulp) to 125.45 °C, when 30% of maltodextrin was added. Kurozawa et al. (2009) obtained a Tgs of 44.43 °C for the pure chicken meat hydrolysate protein powder, while the addition of 10% of maltodextrin or gum Arabic led to Tgs values of 91.90 and 94.70, respectively. Thus, the spray dried pure açai juice might have a glass transition temperature much lower than the values obtained for the powders produced with additives. According to Truong et al. (2005b), the sticky-point temperature is normally about 10–23 °C higher than the glass transition temperature and, in spray drying, particles which are above this temperature stick to the dryer wall and degrade, and/or clump together, adversely affecting the free-ﬂowing property. In the case of pure açai juice, considering its sugars and acids level, the sticky-temperature is much lower than 78 °C (the outlet air temperature) and that would result in a high degree of stickiness and thus in an insigniﬁcant powder yield.
Table 5 Estimated Gordon–Taylor parameters for açai juice powder produced with different carrier agents. Parameters
Carrier agents Maltodextrin 10DE
Tgs (°C) k R2 E (%)
93.99 4.60 0.978 1.83
79.12 3.75 0.987 1.49
88.29 3.56 0.993 1.32
96.72 6.87 0.974 1.89
R.V. Tonon et al. / Journal of Food Engineering 94 (2009) 215–221
3.3. Product stability based on water activity and glass transition Both water activity and glass transition temperature have been widely used to evaluate storage stability. Roos (1995) reported that the plasticization of biosolids is a result of combined effects of water and temperature. According to the author, the prediction of food stability based only on sorption isotherms data is not enough, since certain physicochemical and structural processes such as stickiness, crispness, collapse, amorphous-to-crystalline transformations and the rates of non-enzymatic browning are not related to a monolayer value and they are better correlated to the glass transition temperature through plasticization by water or temperature. Thus, the use of state diagrams that indicate the material’s physical state, combined with the sorption isotherms, helps in the prediction of food stability, regarding to its physical characteristics. Several authors have coupled the data of sorption isotherms with those of glass transition temperature, in order to obtain the critical conditions for food storage (Roos, 1993; Moraga et al., 2004, 2006; Kurozawa et al., 2009). The critical water content/ water activity is the value at which the glass transition temperature of the product is equal to the room temperature. Above this temperature, the amorphous powders are susceptible to deteriorative changes like collapse, stickiness and caking, resulting in quality loss. Hence, in order to calculate the critical conditions of storage for açai juice powder, sorption isotherms and Tg data were plotted as a function of aw and the critical values of water activity and moisture content were obtained considering a room temperature of 25 °C (Fig. 4). The water content and Tg value were predicted by the BET and Gordon–Taylor models, respectively. Table 6 shows the critical aw and moisture content for the powders produced with the different carrier agents. The critical water activities were similar for all the carrier agents, varying between 0.535 and 0.574. The powder produced
Table 6 Critical values for water activity (awc) and moisture content (Xc) of spray dried açai juice. Carrier agent
Xc (g/g dry matter)
Maltodextrin 10DE Maltodextrin 20DE Gum Arabic Tapioca starch
0.574 0.535 0.571 0.554
0.086 0.083 0.100 0.061
with maltodextrin 10DE can be considered as the most stable, once it showed the highest critical aw, equal to 0.574. This means that when the powder is stored at 25 °C, the maximum relative humidity to which it can be exposed is 57.4% and its moisture content is 8.6%. However, when stored at a relative humidity higher than 57.4% (at 25 °C), or at a higher temperature (at aw = 0.574), the powder will suffer physical transformations such as collapse, stickiness and caking. Moraga et al. (2004, 2006) obtained very lower values of critical water activity and moisture content for freeze-dried kiwi (0.034% and 1.4%, respectively) and strawberry (0.110% and 2.0–4.0%) at 30 °C, which is probably related to the higher sugar and acid content present in these fruits, as compared to açai. Moreover, the authors did not use any additive in the powder production, resulting in lower Tg values. 4. Conclusions Both BET and GAB models well described water adsorption of açai powder produced with the four different carrier agents. The glass transition temperature of powders stored at different water activities was measured and it decreased with the increase in moisture content, conﬁrming the strong plasticizing effect of water on this property. The critical conditions for storage at 25 °C were determined based on the sorption isotherms and the glass transi-
Fig. 4. Variation of glass transition temperature (solid line) and equilibrium moisture content (dashed line) with water activity for spray dried açai juice produced with: (a) maltodextrin 10DE, (b) maltodextrin 20DE, (c) gum Arabic and (d) tapioca starch.
R.V. Tonon et al. / Journal of Food Engineering 94 (2009) 215–221
tion temperature. Critical water activities varied from 0.535 to 0.574 and critical moisture contents varied from 0.061 to 0.100 g/g dry matter, respectively. Above these values, açai powder can become collapsed and sticky. However, for a complete study of powder stability, it is necessary to evaluate the occurrence of chemical reactions (such as anthocyanin degradation, in the case of açai) during storage, in order to verify the possible ability of small molecules such as oxygen and water, to diffuse in materials, even in the glassy state. Acknowledgement The authors thank CNPq, CAPES and FAPESP for the ﬁnancial support. References AOAC, 1990. Ofﬁcial Methods of Analysis, 14th ed. Association of Ofﬁcial Analytical Chemists, Washington, DC. 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 dextrin concentration and atomization speed. Journal of Food Engineering 64 (3), 285–287. Aguilera, M.J., Del Valle, J.M., Karel, M., 1995. Caking phenomena in amorphous food powder. Trends in Food Science and Technology 6 (5), 149–155. BeMiller, J.N., Whistler, R.L., 1996. Carbohydrates. In: Fennema, O.R. (Ed.), Food Chemistry, third ed. Marcel Dekker, New York. Bhandari, B.R., Howes, T., 1999. Implication of glass transition for the drying and stability of dried foods. Journal of Food Engineering 40 (1–2), 71–79. Bhandari, B.R., Senoussi, A., Dumoulin, E.D., Lebert, A., 1993. Spray drying of concentrated fruit juices. Drying Technology 11 (5), 1081–1092. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and puriﬁcation. Canadian Journal of Biochemistry and Physiology 37 (8), 911–917. Brunauer, S., Emmett, P.H., Teller, E., 1938. Adsorption of gases in multimolecular layers. Journal of the American Chemists’ Society 60, 309–319. Cai, Y.Z., Corke, H., 2000. Production and properties of spray-dried Amaranthus betacyanin pigments. Journal of Food Science 65 (6), 1248–1252. Cano-Chauca, M., Stringheta, P.C., Ramos, A.M., Cal-Vidal, J., 2005. Effect of the carriers on the microstructure of mango powder obtained by spray drying and its functional characterization. Innovative Food Science and Emerging Technologies 5 (4), 420–428. Coisson, J.D., Travaglia, F., Piana, G., Capasso, M., Arlorio, M., 2005. Euterpe oleracea juice as a functional pigment for yogurt. Food Research International 38 (8–9), 893–897. Dib Taxi, C.M., Menezes, H.C., Santos, A.B., Grosso, C.R., 2003. Study of the microencapsulation of camu–camu (Myrciaria dubia) juice. Journal of Microencapsulation 20 (4), 443–448. Ersus, S., Yurdagel, U., 2007. Microencapsulation of anthocyanin pigments of black carrot (Daucuscarota L.) by spray dryer. Journal of Food Engineering 80 (3), 805– 812. Gabas, A.L., Telis, V.R.N., Sobral, P.J.A., Telis-Romero, J., 2007. Effect of maltodextrin and Arabic gum in water vapor sorption thermodynamic properties of vacuum dried pineapple pulp powder. Journal of Food Engineering 82 (2), 246–252. Gordon, M., Taylor, J.S., 1952. Ideal copolymers and the second-order transitions of syntetic rubbers. I. Non-crystalline copolymers. Journal of Applied Chemistry 2 (9), 493–500. Goula, A.M., Karapantsios, T.D., Achilias, D.S., Adamopoulos, K.G., 2008. Water sorption isotherms and glass transition temperature of spray dried tomato pulp. Journal of Food Engineering 85 (1), 73–83. Greenspan, L., 1977. Humidity ﬁxed points of binary saturated aqueous solutions. Journal of Research of the National Bureau of Standards – Physics and Chemistry 81 (1), 89–96. Johari, G.P., Hallbrucker, A., Mayer, E., 1987. The glass–liquid transition of hyperquenched water. Nature 330 (10), 552–553. Jonquières, A., Fane, A., 1998. Modiﬁed BET models for modeling water vapor sorption in hydrophilic glassy polymers and systems deviation strongly from ideality. Journal of Applied Polymer Science 67 (8), 1415–1430.
Khallouﬁ, S., El-Maslouhi, Y., Ratti, C., 2000. Mathematical model for prediction of glass transition temperature of fruit powders. Journal of Food Science 65 (5), 842–848. Kurozawa, L.E., Park, K.J., Hubinger, M.D., 2009. Effect of maltodextrin and gum Arabic on water sorption and glass transition temperature of spray dried chicken meat hydrolysate protein. Journal of Food Engineering 91 (2), 287–296. Lomauro, C.J., Bakshi, A.S., Labuza, T.P., 1985. Evaluation of food moisture isotherm equations. 1: fruit, vegetable and meat products. Lebensmittel–Wissenschaft und Technologie 18 (2), 111–117. Miao, S., Roos, Y.H., 2004. Comparison of nonenzymatic browning kinetics in spraydried and freeze-dried carbohydrate-based food model systems. Journal of Food Science 69 (7), E222–E331. Moraga, G., Martínez-Navarrete, N., Chiralt, A., 2004. Water sorption isotherms and glass transition in strawberry. Journal of Food Engineering 62 (4), 315–321. Moraga, G., Martínez-Navarrete, N., Chiralt, A., 2006. Water sorption isotherms and phase transitions in kiwifruit. Journal of Food Engineering 72 (2), 147–156. Pérez-Alonso, C., Beristain, C.I., Lobato-Calleros, C., Rodriguez-Huezo, M.E., VernonCarter, E.J., 2006. Thermodynamic analysis of the sorption isotherms of pure and blended carbohydrate polymers. Journal of Food Engineering 77 (4), 753– 760. Quek, S.Y., Chok, N.K., Swedlund, P., 2007. The physicochemical properties of spraydried watermelon powder. Chemical Engineering and Processing 46 (5), 386– 392. Rahman, M.S., Sablani, S.S., Al-Habsi, N., Al-Maskri, S., Al-Belushi, R., 2005. State diagram of freeze-dried garlic powder by differential scanning calorimetry and cooling curve methods. Journal of Food Science 70 (2), E135–E141. Righetto, A.M., Netto, F.M., 2005. Effect of encapsulating materials on water sorption, glass transition and stability of juice from immature acerola. International Journal of Food Properties 8 (2), 337–346. Roos, Y.H., 1993. Water activity and physical state effects on amorphous food stability. Journal of Food Processes and Preservation 16 (6), 433–447. Ross, Y.H., 1995. Phase Transitions in Foods, ﬁrst ed. Academic Press, San Diego. 360p. Roos, Y.H., Karel, M., Kokini, J.L., 1996. Glass transitions in low moisture and frozen foods: effects on shelf life and quality. Scientiﬁc status summary. Food Technology 50 (11), 95–108. Sablani, S.S., Kasapis, S., Rahman, M.S., Al-Jabri, A., Al-Habsi, N., 2004. Sorption isotherm and the state diagram for evaluating stability criteria of abalone. Food Research International 37 (10), 915–924. Schauss, A.G., Wu, X., Prior, R.L., Ou, B., Huang, D., Owens, J., Agarwal, A., Jensen, G.S., Hart, A.N., Shanbrom, E., 2006. Antioxidant capacity and other bioactivities of the freeze-dried amazoniam palm berry, Euterpe oleracea Mart. (Acai). Journal of Agricultural and Food Chemistry 54 (22), 8604–8610. Schebor, C., Buera, M.P., Karel, M., Chirife, J., 1999. Color formation due to nonenzymatic browning in amorphous, glassy, anhydrous, model systems. Food Chemistry 65 (4), 427–432. Shrestha, A.K., Howes, T., Adhikari, B.P., Bhandari, B.R., 2007a. Water sorption and glass transition properties of spray dried lactose hydrolysed skim milk powder. LWT – Food Science and Technology 40 (9), 1593–1600. Shrestha, A.K., Ua-arak, T., Adhikari, B.P., Howes, T., Bhandari, B.R., 2007b. Glass transition behavior of spray dried orange juice powder measured by differential scanning calorimetry (DSC) and thermal mechanical compression test (TMCT). International Journal of Food Properties 10 (3), 661–673. Silva, M.A., Sobral, P.J.A., Kieckbusch, T.G., 2006. State diagrams of freeze-dried camu–camu (Myrciaria dubia (HBK) Mc Vaugh) pulp with and without maltodextrin addition. Journal of Food Engineering 77 (3), 426–432. Slade, L., Levine, H., 1991. Beyond water activity: recent advances based on an alternative approach to the assessment of food quality and safety. Critical Reviews in Food Science and Nutrition 30 (2–3), 115–360. Symaladevi, R.M., Sablani, S.S., Tang, J., Powers, J., Swanson, B.G., 2009. State diagram and water adsorption isotherm of raspberry (Rubus idaeus). Journal of Food Engineering 91 (3), 460–467. Telis, V.R.N., Sobral, P.J., 2001. Glass transitions and state diagram for freeze-dried pineapple. Lebensmittel–Wissenschaft und Technologie 34 (4), 199–205. Truong, V., Bhandari, B.R., Howes, T., 2005a. Optimization of co-current spray drying process of sugar-rich foods. Part I – moisture and glass transition temperature proﬁle during drying. Journal of Food Engineering 71 (1), 55–65. Truong, V., Bhandari, B.R., Howes, T., 2005b. Optimization of co-current spray drying process of sugar-rich foods. Part II – optimization of spray drying process based on glass transition temperature. Journal of Food Engineering 71 (1), 66–72. Wang, H., Zhang, S., Chen, G., 2008. Glass transition and state diagram for fresh and freeze-dried Chinese gooseberry. Journal of Food Engineering 84 (2), 307–312.