Drying 2004 Proceedings of the 14th International Drying Symposium (IDS 2004) So Paulo, Brazil, 22-25 August 2004,

, vol. B, pp. 876-883

DRYING OF BANANA PASTE IN ROTARY DRYER WITH INERT BED

Rebeca Y. C. Padilla, Jos R. Limaverde and Jos R. D. Finzer,2 Federal University of Uberlndia Chemical Engineer Program Av. Joo Naves de vila, 2160 Bloco K, CEP 38405-100, Uberlndia MG Brazil rycpadilla@equi.ufu.br, jrlimaverde@ufu.br, jrdfinzer@ufu.br 2 Associated Faculties of Uberaba Food Engineer Course Av. do Tutuna 720, Bairro Tutunas, CEP 38061.500, Uberaba-MG - Brazil
Keywords: banana paste drying, rotary dryer with inert bed, additive, banana flour.
1

ABSTRACT The aim of this is work is to study the drying of banana "nanica" paste (variety, belonging to the subgroup Cavendish) in Rotary Dryer with Inert Bed (RDIB). As banana is a perishable fruit, one of the techniques used to enlarge its lifetime for consumption, reducing packing and transportation costs, improving sensorial attributes and preserving the nutritional value is the banana drying process. Preliminary tests were performed to adjust the operational process parameters. For the final experiments of drying, an experimental planning (central composed planning) was carried out, having as variables: the drying time and the mass of fed paste into the dryer and, as responses, the value of the moisture content and the production of banana flour. As a result of the tests performed following the experimental planning, it was obtained a beige-colored product, composed of powder and banana flakes, and characteristic aroma of banana. The influence of variables in the banana drying process was investigated by means of a multiple regression for the experimental results. The best operating condition made possible a product with 8.0% moisture content (wet base) and a flour production efficiency of 72.9% and finally the final product was characterized. INTRODUCTION The banana, besides constituting an expressive carbohydrate source (highly energetic), it is still rich in potassium, sodium, phosphorus, chlorine, magnesium, sulfur, silicon, calcium; vitamins: A, B1, B2, C and B3 (niacin) that are essential for the operation of the human body (Padovani, 1989). Brazil the second largest world producer of banana produces eight million tons a year, and only exports 2% and losses are about 40% (Knapp, 2001).

876

Banana flour is used in the pap making, soups, as enrichment of milk, in ice creams and other foods. It can also be used in the bread making, where it is mixed in the proportion of 1/3 of banana flour to 2/3 of wheat flour (Padovani, 1989). Among the conventional dryers used in banana drying the following are distinguished: the trays dryer, where the time of drying is significant; the drum dryer, where the product does not present good quality, above all for not presenting uniformity in the drying; the spray dryer, that it makes possible a high quality product; however, the capital investment and operational costs are significant, what makes unfeasible its use for medium and small capacity industries (Hufenuessler & Kachan, 1986; Maskan, 2000; Nury et al., 1973). The rotary dryer with inert bed (RDIB) is an alternative technically viable, because in the processing of other foodstuffs good quality powdered products were obtained (Finzer et al., 1993). In this work the description of the methodology used for banana paste drying (with and without additive) in RDIB, as well as the results of the drying tests are presented. MATERIALS AND METHODS Raw material The raw material used in this work is the banana "nanica" variety of the subgroup Cavendish with 75% of ripening (yellow color with green tips) that is more appropriate for the flour obtaining (Manica,1997). Banana paste preparation In the paste preparation several tests were carried out by adding different water ratios: 5%, 10%, 15%, and 20% in relation to banana mass. The objective was to obtain a paste with possibility of pumping easily. The most appropriate banana paste was that prepared with 10% water. The preparation way of the paste was the following: - selection of bananas with 75% of ripening; - wash in chlorinated water (4 ppm), at temperature of 45C, for preliminary cleaning of bananas (Manica, 1997); - five minute blanching in boiling water to reduce the number of microorganisms and enzymes inactivation, followed by cooling for three minutes (Dandamrongrok et al., 2000); - skin elimination and then cut in disk-like slices of approximately 2 cm in thickness; - immersion in 0.2% sodium bisulfite (NaHSO3) solution during five minutes, to avoid oxidation of the product; - drainage of the solution for five minutes; - triturating in a knife rotary crusher. Addition of: i) water to obtain the content of 10% of the total pulp mass; ii) 1% of additive Dimodan HO-1, a distilled monoglyceride (to improve the texture; to reduce the viscosity and the sugar crystallization), and iii) 0.4% of ascorbic acid to avoid the darkening during the disintegration step. Drying tests To select the inert materials to be used in the RDIB, qualitative tests of coating and drying were carried out in a laboratory oven. Different materials were tested: stainless steel spheres, porcelain spheres and technyl cylinders. The porcelain spheres had been the chosen ones because they had not presented oxidation of product and presented a better heat transfer than the "technyl" cylinders. Besides, they allowed the obtaining of a more homogeneous paste film on surface, during the coating before the drying. The product presented oxidation when stainless steel spheres were used.

877

Rotary dryer with inert bed The drying tests of banana paste were carried out in R&D Laboratory of Chemical Engineering Faculty of Federal University of Uberlndia. Figure 1 is a schematic representation of the unit of drying and its accessories.
O E N C D M K B A I J G

L
K L M N O Balance Peristaltic pump Agitator Cyclone Temperature display

A B C D E

Steel ribbon Gear Rolling Engine Teflon ring

F G H I J

Blower Air discharge Orifice plate and Manometer Electric resistance Voltage transform

Figure 1 - Schematic unit of the Rotary Dryer with Inert Bed and accessories.

In this dryer, banana paste is fed into the drying chamber that contains inert bed of porcelain spheres (2.8 to 3.0 cm in diameter). In the operation the inert materials are covered for a layer of banana paste. The inerts facilitate the drying when they enlarge the contact surface between the material and drying air. The chamber rotation propitiates a great number of collisions and shear stress (inert-inert, and inert-wall), facilitating the mixture and the paste drying. The mechanical shocks and the shear stress provide the grinding of the product during the drying. The dry material is conveyed by the drying air and separates in a cyclone. The angular speed of the drying tunnel was maintained at 18 rpm. Previous studies showed that in this operational condition the dryer presents good operational performance (Yoshida et al.,2001). In the drying tests the operation conditions were: air temperature of 70C (for not exceeding the degradation temperature of the banana's vitamins); the air mass flow of 5.3 kg/min (for making possible a good heat transfer convection coefficient). These operational conditions are to make easy the powder discharge (Limaverde Jr., 2000). After preliminary experiments the drying tests ware carried out according to the factorial planning (central planning). Characterization of the final product The banana flour was characterized by following contents: moisture, ash, total carbohydrate; lipids, and proteins according to the rules of the Adolfo Lutz Institute (1985); and caloric values were calculated by using sum of carbohydrates, lipids and proteins specific calories (Ascar, 1985).

878

RESULTS Drying test of banana paste without supporting. The rotary dryer with inert bed was assayed initially with the following operational conditions: feeding 0.25 kg of banana paste (without additive) by cycle, with initial moisture of 79.5% (wb). The feeding time by cycle was of approximately 120 s. Each operation drying cycle had duration of 35 min, and during 15 min it was operated with the drying air temperature of 70C and the 20 remaining min with drying air temperature of 53C. Five consecutive drying cycles were effected. The airflow was of 8.8x10-2 kg/s. The results are shown in Table 1. It can be observed that, in these operational conditions, an average flour production of 2.92x10-3 kg was obtained, with average moisture content of 9.26% (wb); and the average production efficiency of 5.2%, (ratio of separated powder in the cyclone by the amount of solids in the feeding). The banana flour mass in each cycle was not significant and its color was dark. At the end of the last cycle the dryer was disassembled and observed that most amount of the fed material was contained in the drying tunnel. All the inerts were covered with moist paste and adhered to the central axis and paddles.
Table 1 - Performance of rotary dryer without additive.

Cycle 1 2 3 4 5

Operation Drying air time Temperature Flow rate (min) (C) (kg/s) 15 70,3 8,8x10-2 20 53,3 15 71,1 8,8x10-2 20 53,3 15 70,5 8,8x10-2 20 54,0 15 70,4 8,8x10-2 20 54,9 15 70,4 8,8x10-2 20 52,9

Total feed pastelike (kg) 0,25 0,25 0,25 0,25 0,25

Flour production (kg) x103 4,5 3,0 3,2 2,1 1,8

Flour moisture % (wb) 8,9 10,2 8,9 9,4 8,9

Production efficiency % 8,0 5,3 5,7 3,7 3,2

Drying test of banana paste with additive Technical information regarding the additive (Dimodan HO-1) used in the drying of pastelike materials showed the easiness to detach the dry material from surfaces. This is an indication that the production capacity should increase when the additive is used. Based on this information and on the preliminary tests other drying experiments in the same operational conditions were conducted. The results are shown in Table 2, where it can be observed an average flour production of 28.3x10-3 kg with average content moisture of 7.7% (wb) and average production efficiency of 49.0%. The banana flour production was practically stabilized after the first cycle. The obtained product, in powder and flakes form, presented a beige coloration and a characteristic banana aroma. After last cycle operation the dryer was disassembled and it was observed in the spheres and walls of the drying tunnel that a little amount of paste left adhered. This was an indication of stability of the production, and the dryer capacity was 0.25 kg banana paste by drying cycle. Losses of banana powder occurred on cyclone overflow.

879

Table 2 - Performance of rotary dryer with additive.

Cycle 1 2 3 4 5

Operation time (min) 15 20 15 20 15 20 15 20 15 20

Drying air Temperature Flow rate (C) (kg/s) 70.5 8.8x10-2 53.7 72.1 8.8x10-2 52.0 72.1 8.8x10-2 53.2 72.7 8.8x10-2 52.6 72.2 8.8x10-2 51.4

Total feed pastelike (kg) 0.25 0.25 0.25 0.25 0.25

Flour Flour production moisture (kg) x103 (% ) wb 21.7 31.6 27.2 30.4 30.4 7.7 8.0 7.6 7.7 7.7

Production

efficiency
(%) 37.7 54.7 47.3 52.7 52.7

Figure 2 shows the comparison of the banana flour production with and without additive at the same operational conditions. It can be observed the difference of the flour accumulated production and the additive influence in the paste-like banana drying. This is fundamental for the process technical viability. For the processing time operation of 175 min using additive the production was approximately ten times larger than in the operation without it.

140 120 100 80 60 40 20 0 0

With additive Without additive

50

100
Drying time (min)

150

200

Drying tests of factorial planning.

Figure 2 - Comparison of the production of accumulated flour with and without additive.

The test results of the factorial planning - PCC (central planning) are showed in Table 3. Each test corresponds to two drying stages. In the first stage the temperature of the drying air was maintained at 70C with duration of operation time as showed in the table. In the second stage, in all tests the drying air temperature was reduced to 53C and maintained for 20 minutes.
Table 3 - Results of the drying tests, with the conditions of each experiment

Test 1 2 3 4 5 6 7 8 9 10

Codified variables

X1
-1 -1 1 1 -1.0781 1.0781 0 0 0 0

X2
-1 1 -1 1 0 0 -1.0781 1.0781 0 0

Total pastelike load (kg) 0.25 0.25 0.75 0.75 0.23 0.77 0.50 0.50 0.50 0.50

Drying time 1st. stage (min) 15 45 15 45 30 30 14 46 30 30

Product moisture content (%) 7.7 7.3 11.0 8.0 7.5 10.0 10.6 8.3 9.5 9.3

Banana flour production (kg) 0.028 0.036 0.054 0.122 0.034 0.088 0.010 0.083 0.086 0.085

880

Accumulated prodution (kg).10

Figure 3 presents the comparison of the accumulated banana flour production as a function of accumulated drying time (where it can be seen in Test 4), the largest accumulated production of 0.61kg. Test 7 presents the smallest accumulated production of banana flour (0.050 kg), in which the drying time was insufficient. Multiple regression analyses response Using data presented in Table 3 a multiple regression was carried out for the responses of moisture content and flour production. Moisture content

650 600 550 500 450 400 350 300 250 200 150 100 50 0

Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9 Test10

50

100 150 200 250 Drying time (min)

300

350

Figure 3 - Comparison of the accumulated production of the ten factorial planning (PCC) tests.

The obtained correlation, Equation (1), relates the banana flour moisture as a function of the codified variables. U = 9.40 + 1.08X1 0.94 X2 - 0.75X12 - 0.64X1X2 To illustrate the effects of the process variables on banana flour quality as moisture content (U) the response surface is presented (Figure 4). When the feeding of the paste (X1) increases and the drying time decreases (X2), the moisture content increases. The maximum region at the surface response for the moisture content corresponds to the largest feeding levels of pastelike banana. The condition where the moisture content level is within the expected value, lower than 8%, corresponds to the region where the feedings should be lower than 0.33 kg of pastelike banana. Flour production The Equation (2) presents the correlation found to represent the production of banana flour as a function of the codified variables. P=0.074+0.027X1+ 0.024X2 0.018X22+0.015X1X2 (2) (1)

7.487 7.881 8.276 8.67 9.065 9.459 9.854 10.248 10.643 11.038 above

Figure 4 - Response surface for the moisture content as function of X1 and X2.

The value of the correlation coefficient (0.90) indicates the quality of the obtained adjustment of the experimental data in the response production of banana flour. To illustrate the effects of the process variables in the banana flour quality regarding the flour production (P) the response surface is presented in Figure 5. It was shown that when the pastelike feeding increases

881

(X1), the production increases; decreasing the drying time (X2), the production decreases. The maximum region for the flour production corresponds to the largest levels of feeding of the paste and to drying time. presents the better characteristics as colour, aroma, production and mainly the moisture content (6-8%). The results of the physicochemistry analyses are shown in Table 4. As reference of banana flour composition percentages published by De Martin et al. (1990) are appropriate. Comparing the results of banana flour obtained in the current work with suitable data, it is found that they present similarity, especially in the moisture content that is a decisive parameter of the quality of this product type. Analyzing the data in Table 4, it is found that the banana flour caloric value obtained in this work is within the range of values published in the literature.

Figure 5 - Response surface for the banana flour production as a function of X1 and X2.

0.024 0.033 0.042 0.052 0.061 0.071 0.08 0.089 0.099 0.108 above

Banana flour characterization The obtained product of the Test 4 was used to banana flour characterization because it
Table 4 - Banana flour quality

Analyzed items Moisture content (wb) (%) Ash content (%) Total carbohydrate content (%) Lipids content (% Proteins content (%) Caloric value (kcal/100 g)

Measured value 8.03 2.63 65.66 2.81 3.79 303.10

Data published by De Martin et al. (1990) Gross Michel species Unidentified species 8.50 5.00 2.00 3.06 82.20 79.90 1.05 2.30 2.00 4.98 346.50 283.00

CONCLUSIONS - The largest efficiency of production of dried banana, in the rotary dryer, happened with the addition of Dimodan HO-1 to the paste. The powder production without the presence of additive in the paste was not technically viable. - The adjustment of the empiric equations of moisture content and flour production indicated that about 98% of the variability of data was explained by the equation proposed for the answer humidity content and 90% for the answer production of banana flour. - Analyzing the results obtained, the best operational conditions for accomplishment of the drying of banana's paste in rotary dryer with inert bed can be selected: drying air flow rate 8.8x10-2 kg/s; drying air temperature, operating for 45 minutes at average temperature of 70C and 20 minutes at average temperature of 53C; banana paste feed at the rate of 4.2x10-3 kg/s; total banana paste feed of 0.75 kg by cycle. Under these operating conditions it was obtained an average production of 0.122 kg of banana flour corresponding to powder production efficiency of 72.9%. - The product obtained was in powder form and flakes beige in color and characteristic aroma of banana. - The results obtained: moisture content of 8.0% (wb) it is within the established range for this product type (6 to 8%).

882

ACKNOWLEDGMENT This work received support from: 1) CNPq (National Council of Research); 2) Danisco Brasil Ltda for Dimodan HO 1 supply NOTATION
X1 X2 U P wu db Codified variables (paste feed) Codified variables (drying time) Moisture content Banana flour production Wet base Dry base % kg

LITERATURE
Ascar J.M. (1985), Anlise Percentual, Unisinos, Rio Grande do Sul, Brasil, Vol 1. Dandamrongrak, R., Young, G., Mason, R. (2002), Evaluation of various pre-treatments for the dehydration of banana and selection of suitable drying models, Journal of Food Engineering, Vol. 55, pp.139-146 De Martin,Z.J., Travaglini, D. A.,Okada, M., Quast, D. G., Hashizume, T. (1990), Processamento: produtos, caractersticas e utilizao. In: Instituto de Tecnologia de Alimentos. Banana-cultura, matria prima, processamento e aspectos econmicos: serie frutas tropicais no. 3, 2da ed., ITAL, Campinas, Brasil Finzer, J. R. D.; Burjaili, M. M.; Limaverde, J. R. (1993), Alternativa tecnolgica de secagem de materiais pastosos secador rotatrio com recheio de inertes, Cincia & Engenharia, Vol. 2, pp. 97-119 Hufenuessler, M.; Kachan, G. C. (1986), Estudo da viabilidade de desidratao de pur de banana em secador tipo jorro, Cincia e Tecnologia de Alimentos, Vol. 1, pp. 31 41 Instituto Adolfo Lutz. (1985), Normas analticas, Mtodos fsicos e qumicos para anlise de alimentos, 3rd ed., So Paulo, Brasil, Vol. 1, pp. 21-53 Knapp, L. (2001), Projeto pode substituir farinha por banana, O Estado de So Paulo, 26 de novembro, www.estadao.com.br. Limaverde Jr, J. R. (2000), Secador rotatrio com recheio de inertes aplicado a secagem de matrias pastosas, Uberlndia: Dissertao, Mestrado Universidade Federal de Uberlndia, Programa de Ps-Graduao em Engenharia Qumica, 78 p. Manica, I. (1997), Fruticultura tropical 4: Banana, Cinco Continentes Editora Limitada, Porto Alegre, Brasil. Maskan, M. (2000), Microwave/air and microwave finish drying of banana, Journal of Food Engineering, Vol. 44, pp. 71 18 Nury, F. S.; Brekkle, J. E.; Bolin, H. R. (1973), Fruit. In: Von Arsdel, W. B.; Copley, M. J.; Morgan, A. I. Food dehydration: practices and applications, The AVI Publishing Company, Westport, Vol. 2 Padovani, M. I. 81989), Banana um mercado crescente para este alimento milenar. 2a ed., ICONE, So Paulo, Brasil Yoshida, L. M.; Limaverde J. R.; Finzer, J. R. D.(2001), Secagem de resduo da formao de cido ctrico em secador rotativo com recheio de inertes, XXIX Congresso Brasileiro de Sistemas Particulados, So Joo Del Rei MG. Anais de ENEMP cd- rom.

883