Vous êtes sur la page 1sur 2

Optimization of Blanching for Crispness of Banana Chips

Using Response Surface Methodology



Whole green bananas were blanched in water at 50, 60, 70, 80, 90, and Materials
1007C for 2, 15, and 30 min using a factorial design then peeled, sliced Musa cavendish bananas from local supermarkets were used. The ba-
and fried in oil to make chips. Crispness was measured using a bending- nanas were bought and processed at color index No. 1, when the fruit
snapping test in the TA.XT2 Texture Analyzer. Significant interactions is green, but mature (Von Loesecke, 1950).
were found between blanch time and temperature and crispness of chips.
A second experiment was then performed using a central composite de-
sign and blanch temperatures from 41.7 to 98.37C. Response surface Methods
analysis predicted that crispiest chips should be produced at blanching
conditions of 697C and 22 min. Banana chips preparation. The study involved two separate frying
experiments: the first (the control) had no blanch treatment; the second
Key Words: banana chips, blanching, optimization, crispness, response involved blanching whole green fruit according to designed experiments.

FoodScience95 0855 Mp
surface After blanching, fruit were peeled manually and sliced in a Hobart Slicer
to a thickness of 1.6 mm. The slices were fried in soybean oil at 1907C
for 2 min or until crisp, then cooled and drained. The chips were stored
INTRODUCTION for 2 wk at '257C in a desiccator containing MgCl2z6H2O to equilibrate
them to a water activity (aw) of 0.34. This aw level was selected because
BANANA CHIPS are deep-fried snack foods that are very popular it is well below the level of aw 0.50 above which crispness decreases
in many countries. Because of the growth in snack foods the rapidly (Sauvageot and Blond 1991).
Instrumental tests. A force in compression test was set up in the
processing of bananas into chips has the potential to become an TA.XT2 Texture Analyzer (Texture Technologies, Scarsdale, New
important industry in countries where bananas are produced. York/Stable Microsystems, Hoslemere, Surrey, U.K.) using a constant
Blanching has been reported in the preparation of banana chips speed of 0.1 mm/sec. This slow speed was selected so that the sequence
(Jain et al., 1962; Adeva et al., 1968; CARIRI, 1993). It inac-

of events leading to fracture could be observed. A banana chip was
tivates enzymes that cause browning, and facilitates peeling of placed on a 3 cm diameter ring, and a flat headed probe of diameter 2.5
green fruit where the peel adheres firmly to the pulp. However, mm was pressed down to pierce its surface to a distance of 10 mm from

Friday Jan 26 08:02 PM

most commercial chips are made with no blanch before frying. the starting position. The slope before the first major fracturability peak
Jackson and Bourne (1994) reported on the effects of blanch- was taken as the crispness of the banana chip (Fig. 1) and it was meas-
ing treatments on sensory textural characteristics perceived in ured automatically by the computer program of the TA.XT2
Sampling. Samples (19) from a factorial design, and others (15) from
banana chips using the sensory texture profile method of anal- a central composite design (CCD) were used. Five replicates of each
ysis. However, few studies have used instrumental methods to sample from the designed experiment were tested for crispness in the
determine effects of blanching on texture of banana chips. TA.XT2.
A main desired textural characteristic for dry foods is crisp- Statistical analysis. The initial 3 3 6 factorial design using blanch
ness. It is a highly valued and universally liked textural char- times of 2, 15, and 30 min and blanch temperatures of 50, 60, 70, 80,
acteristic that signifies freshness and high quality. In a word 90, and 1007C was analyzed using the Genstat program to examine main
association test determining consumer awareness of food attrib- and interaction treatment effects. Based on these results a second series
utes, particularly texture, (Szczesniak and Kleyn, 1963), the of experiments was performed following a CCD design using 13 com-
word ‘‘crisp’’ was mentioned more often than any other texture binations of variables including five replicates of the center region to
generate a second order response surface using the JMP program. In this
descriptor. According to Szczesniak (1988), ‘‘crispness appears set, blanch temperatures ranged from 41.77C to 987C and blanch times
to be the most versatile single texture parameter. It is particu- from 3.4 to 34.6 minutes. Lack of fit and R2 were computed to determine
larly good as an appetizer, and as a stimulant to active eating; adequacy of the model, and optimum blanching conditions for increasing
and is very prominent in texture combinations that mark excel-
lent cooking and is nearly synonymous with freshness and
A review of published reports on analysis of crispness indi-
cates that an exact method for its determination has not been
reported. In describing objective methods for crispy foods,
Bourne (1975) suggested that since they fracture on first bite
giving sharp force peaks and steep force-distance curves, a sat-
isfactory method of measuring this textural property may be the
steepness of the force-distance curve in a bending-snapping test.
Our objective was to examine the effects of blanch temper-
ature and blanch time on crispness of banana chips. Response
surface methodology was then used to determine optimum con-
ditions for producing crispy chips.
Authors Jackson and Bourne are with the Dept. of Food Science,
and Author Barnard with the Statistics Dept., Cornell Univ., New
York State Agricultural Experiment Station Geneva, NY 14456.
Author Jackson’s present address: Dept. of Food Science and Hu- Fig. 1—Force deformation curve of banana chips (blanch temp 5
man Nutrition, Michigan State Univ., East Lansing, MI 48824. 70&C, blanch time 5 19 min).

Volume 61, No. 1, 1996—JOURNAL OF FOOD SCIENCE—165


Table 1—Mean crispness and standard deviations of banana chips from

the factorial experimental design
Blanch time Blanch temp Crispness Standard deviation
(min) (&C) (N/m) (N/m)
0 (control) 0 4.19 1.39
2 50 4.18 0.12
15 50 6.75 2.53
30 50 4.15 0.38
2 60 4.46 1.80
15 60 4.69 2.41
30 60 7.11 0.95
2 70 8.15 2.28
15 70 11.43 2.07
30 70 8.59 0.80
2 80 7.26 0.66
15 80 6.90 0.42
30 80 3.85 0.98
2 90 8.15 4.57
15 90 6.29 1.68
30 90 4.84 0.19
2 100 4.17 0.86
15 100 5.48 1.39
30 100 5.39 1.27

Table 2—Analysis of variance for banana chips from the factorial experi-
mental design with added control
Source of variation d.f. s.s m.s. v.r.
Fig. 2—Optimum blanching conditions for crispy banana chips,
relationships of blanching time and temperatures.
control vs treatments 1 62.02 62.02 41.63
temperature 5 455.18 91.04 61.11
time 2 134.02 67.01 44.98
too short or long times and low or high temperature combina-
temperature 3 time 10 679.48 67.95 45.61 tions had a negative effect on crispness, while medium range
residual 32 509.47 1.49 treatments had positive effects. The crispiest chip (crispness
Total 360 1840.17 value 10.6) was predicted to occur at blanch temperature 697C
and blanch time 22 min. The 3-D plot demonstrated the com-
bination of factors that would produce crispier chips, and how
crispness was determined. A 3-D response surface graph which provided a specific crispness response would be affected by a given set
better visualization of interactions was generated by the Splus program. of factor levels.


A TYPICAL FORCE DEFORMATION CURVE for a crisp banana chip BLANCH TIME and blanch temperature have significant effects
is shown (Fig. 1). In the first 2 sec the curve shows the force on crispness of banana chips, and a choice of specific levels of
necessary to bring the probe into full contact with the chip, and these factors can increase crispness. Blanching whole green ba-
the chip in full contact with the supporting ring. The next 4 sec nanas for 22 min at 697C before peeling and frying was pre-
shows a long, steep, almost linear increase in force, the slope dicted to produce crispiest chips, almost three times higher than
of which was defined as crispness. A minor fracture is seen after for chips made from unblanched bananas. However, before this
4 sec, but the slope before and after was practically constant. blanching treatment could be recommended a cost analysis
At about 6 secs there was a sharp decrease in force when the would need to be conducted. Further research involving an he-
chip suddenly fractured, followed by multiple peaks which de- donic study would also be required to evaluate consumer pref-
note further fracture events. The vertical line at the end of the erences for crispier banana chips.
test indicated complete fracture and snapping of the chip.
This curve is typical of crispy foods, and indicates the phe- REFERENCES
nomenon known as ‘‘brittle fracture.’’ The steepness of the Adeva, L.V., Gopez, M.D., and Payumo, E.M. 1968. Studies on the prepa-
slope indicates the resistance of the chip to bending. A crisp ration and storage qualities of banana chips. Philippine J. Sci. 97(1): 27–
chip has a steep slope and is more resistant to bending than a 35.
Bourne, M.C. 1975. Texture properties and evaluations of fabricated foods.
less crisp chip. The mean crispness values and standard devia- Ch. 11 in Fabricated Foods, G.E. Inglett (Ed.), p. 127–158. AVI Publishing
tions for the factorial experiment were compared (Table 1). The Co., Westport, CT.
CARIRI (Caribbean Agriculture Research Institute). 1993. Processing of
highest crispness value was observed for 15 min blanch at 707C. plantain chips. Technical Report, Trinidad.
An analysis of variance determined significant main and inter- Jackson, J.C. and Bourne, M.C. 1994. Sensory texture profile analysis of
action effects at p , 0.001 (Table 2). banana chips. Presented at the 30th Annual Meeting of the Caribbean
Food Crops Society (CFCS), St. Thomas, U.S. Virgin Islands, July 31–Aug.
Based on these data, a CCD was then carried out. In this 5, 1994.
experiment the highest crispness value was observed for the Jain, N.L., Nair, K.G., Siddappa, G.S., and Lal, G. 1962. Studies to improve
the keeping quality of fried salted banana chips. Food Science 11: 335–
707C-19 min sample. The coefficient of determination (R2) was 338.
0.77, and there was no significant lack of fit to the general quad- Sauvageot, F. and Blond, G. 1991. Effects of water activity on crispness of
breakfast cereals. J. Texture Studies 22: 423–442.
ratic model (F3,4 5 0.96). Szczesniak, A.S. and Kleyn, D. 1963. Consumer awareness of texture and
A 3-dimensional representation of the relationship between other food attributes. Food Technol. 17(1): 74–77.
banana chip crispness with blanch temperature and blanch time Szczesniak, A.S. 1988. The meaning of textural characteristics—crispness.
J. Texture Studies 19: 51–59.
was developed (Fig. 2). Blanch time and temperature were plot- Von Loesecke, H.W. 1950. Bananas, 2nd ed. Interscience Publishers, Inc.,
ted on the two horizontal axes, and crispness on the vertical New York.
axis. In moving along the horizontal axes, it can be seen that

166—JOURNAL OF FOOD SCIENCE—Volume 61, No. 1, 1996