DRYING KINETICS OF BANANAS BY NATURAL CONVECTION : INFLUENCE OF TEMPERATURE , SHAPE , BLANCHING AND CULTIVAR

The influence of variables as temperature, solid shape, temperature, branching and cultivar on drying of bananas were studied. Bananas from cv. Prata and D’água, on disk and cylindrical shape, blanched or not, were dehydrated in a tray dryer in natural convection at 40 and 70 C. Drying behavior was analyzed by using a mathematical model. The exponential model showed good agreement to experimental data (r 0.93 0.99 and standard error: 0.010.05). Temperature presented positive influence on drying rate in all the tested conditions. With respect to the sample shape, the disk shape carried out to significantly higher drying rates only for D’água cultivar without blanching. Blanching was significantly influent, with positive influence, on drying rate of banana cv. Prata at 40o C. The influence of the cultivar did not present a defined tendency.

The goal of this work was the study of the influence of variables as temperature, fruit shape, blanching and cultivar in drying kinetics of bananas cv.Prata and D'água, in a natural convection tray dryer.A mathematical model was applied to fit the drying kinetics.

Raw material
The bananas were purchased at a local market, in uniform ripening grade, appropriated for drying (yellow with some black points, according to Travaglini et al. (1993)), with average moisture content of 3.5 g/g (d.b.) for cv.Prata and 4.4 g/g (d.b.) for cv.D'água.

Material preparation
The fruits were washed with tap water and manually peeled.The bananas were cut in disk shape (0.5 cm of thickness with the fruit average diameter of about 3.5 ± 0.3 cm) and cylindrical shape (10 cm of length with the fruit average diameter of about 3.5 ± 0.3 cm).

Blanching
The experiments with blanching were performed by immersing the samples in boiling water for 1 minute and after in a cold bath (0º C) for 1 minute.

Drying
Natural convection drying was performed in a tray dryer at 40 and at 70º C in time intervals (1,2,3,4,5,6,7,8 and 24 hours) to determine the drying kinetics and moisture content.The experiment was done in triplicate.The moisture content analysis was obtained according to Association of Official Analytical Chemists -AOAC (2002).

Mathematical model
Kinetics drying curves were adjusted with and exponential model that relates the ratio moisture content / initial moisture content with drying time.Such model was also used by Alcina et al. (1997) and Cano-Chauca et al. (2004).
This model considers that: where Y st is the solid moisture content at a given time, Y s0 , solid moisture content before drying and Y s * is an non dimensional solid moisture (ratio).
The initial condition is: The term Y s is then given by: (3) where k 0 corresponds to the initial condition and should be around the unity.
The results were adjusted by equation ( 3) and evaluated with respect to the correlation coeficient (r 2 ) and standard error (SE), according to Pimentel-Gomes (1999).

Influence of the temperature
The influence of the temperature can be evaluated by the inspection of Figures 1 to 6.In all these figures, one can see that drying rate increases in a proportional way to the drying temperature.This can be explained with respect to the diffusion coefficient of water that increases with temperature and is decisive in natural convection (Wang & Chen, 1999).The influence of the temperature is more relevant in the first period of drying, called the period of constant rate.In this period, gradients of water are higher and also the diffusivity.Such behavior is largely published in drying works (Kouris & Vagenas, 1991;Tsami & Katsioti, 2000;Queiroz & Nebra, 2001;Krokida et al., 2003;Sacilik, 2007;Silva et al., 2008;Simioni et al., 2008).This can be corroborate by the parameter of experimental adjustment, k 1 (Table 1), which corresponds to a rate between ln(Y st /Y 0 ) and t and is presented in equation 3.

Influence of the shape
Table 1 shows that the difference with respect to the shape is more evident at 70º C. Except for banana 'Prata' without blanching, the disk shape carried out to a higher k 1 .However, if the average standard error of the experimental values of Y is considered as 10%, the disk shape presented a significant higher drying rate for banana cv.D'água at 70º C (Figure 2).

Influence of blanching
Blanching presented a significant and positive influence only for cultivar Prata at 40º C in both shapes (Figures 3 to 6 and Table 1).This trend was also observed in similar works with other fruits (Dandamrongrak et al., 2002;Arévalo-Pinedo & Murr, 2005).Blanching increases drying rate due to the physical changes of the tissue and destruction of cellular membrane promoted by the heating (Mazza, 1983).On the other hand, at 70º C, blanching was not influent.This can be related to the gelatinization of banana starch, which occurs at 65º C (Demirel & Turham, 2003).Demirel & Figure 1 -Drying kinetics for banana cv.Prata at disk and cylinder shape, without blanching.

Influence of the cultivar
Table 1 shows that cultivar did not present a clear tendency.The inference of this table shows that with blanching, cv.Prata presented higher k 1 values for both shapes and both temperatures.Otherwise, without blanching, disk shape carried out to higher k 1 for cv.D'água and cylinder shape presented higher k 1 for cv.Prata, for both studied temperatures.

Mathematical model
Table 1 presents the parameters obtained with equation 3 and the adjustment data r 2 and SE.It was verified that the parameter k 0 was always nearby the unity, what is in accordance with the physical meaning of initial Figure 3 -Drying kinetics for banana cv.Prata, disk shape (CB-with blanching; SB-without blanching).
Figure 4 -Drying kinetics for banana cv.Prata, cylinder shape (CB-with blanching; SB-without blanching).moisture content.It can be observed (Table 1 and Figures 1 to 6) that the mathematical model used here gave a satisfactory agreement (r 2 0.937 and SE 0.048).The non dimensional moisture content, Y is in the range 0.0 < Y < 1.0 and the highest estimative standard error of (SE) is 0.048.With this in mind, it can be said that SE is much appropriated.Moreover, a correlation coefficient (r 2 ) higher our equal to 0.937 is a pertinent parameter of adjustment.
The curves obtained with the model were explored to estimate the critical moisture content.The values for each condition are presented in Table 1.It can be seen that critical moisture decrease with increasing.temperature.Temperature raises kinetic energy of water vapor and, as a consequence, water vapor diffusivity increasing.Such situation causes drying rate increase and critical moisture reduction.

CONCLUSIONS
In experiments of drying kinetics of bananas cultivar D'água and Prata in heat natural convection condition, at 40 and 70º C, the influence of temperature, shape, blanching and cultivar were as follows: Temperature presented a positive influence on the drying rate for both cultivar and both shape.
The Disk shape presented significant higher drying rate with respect to the cylindrical shape only for banana cv.D'água at 70º C.
Blanching presented a significant influence only for cv.Prata at 40º C, in both shapes.
There was not a well defined trend on the curves of drying kinetics with respect to the cultivar.
The exponential mathematical model resulted in good adjustments and helped in the phenomena analysis.

Figure 2 -
Figure 2 -Drying kinetics for banana cv.D'água at disk and cylinder shape, without blanching.

Table 1 -
Parameters of Equation 3, adjustment data and critical moisture for drying kinetics of banana cv.Prata and D'água in disk and cylinder shape, with and without blanching.