Effect of Solvent Extraction on the Gelatinisation Properties of Flour and Starch of Five Cassava Varieties

| Next Page

S N Moorthy, *a June E Wenham,b and john M V Blanshardc

a Central Tuber Crop Research Institute, Thiruvananthapuram 695017, Kerala, India.

b Natural Resources Institute, Chantham Maritime, Chantham, Kent, ME4 4TB, UK

c Department of food Science, School of Agriculture, University of Nottingham, Sutton Boington, LE12 5RD, UK

(Received 11 January 1995; revised version received 18 January 1996;accepted 7 June 1996)

Abstract: the gelatinization properties of native and solvent-extracted (petroleum ether and ethanol) flour and starch processed from five cassava varieties grown in India were studied using a differential; scanning calorimeter (DSC) peaks but these remained unaffected by solvent extraction, indicating that the DSC profiles are characteristic of these starches. In general, the gelatinisation temperatures of the flours were significantly higher than those of the starches but the enthalpy of gelatinization was less for flour. Varietal differences were also observed in the gelatinisation profiles of starch and flour samples obtained using the Brabender amylograph. Lower viscosity values were obtained for the flour samples but values remained more stable throughout the temperature programme. In most varieties, solvent extraction of the samples caused only sight modification of the gelatinisation patterns. There was no apparent relationship between the gelatinisation properties of the starch derived from the different varieties and the size and amylose content of the starch granules. Results of these experiments indicate that differences in gelatinisation values obtained, between starch and flour samples for the same variety, are not due to the presence of sugars and flour samples for the same variety, are not due to the presence of sugars and fats which were removed by ethanol and petroleum ether extraction, respectively. The significant differences observed in some varieties are therefore due to the presence of other root constituents.

Key words: cassava, starch, flour, gelatinisation, solvent extraction.


Starch is the principal component of cassava roots which serve as an important source of calories for human food and animal feed. Cassava, in the form of flour and starch, is also increasingly being used as raw material for processing into a range of food and modified products.

The cooking and eating qualities of fresh roots and the processing characteristics of cassava are well known to vary considerably between varieties and with environmental growth conditions (Wheatley and Gomez 1985). The role of starch properties in determining the end-product potential of the roots has only recently received attention. No consistent correlation has as yet been established between cassava starch structural characteristics and functional properties. Varietal differences have been reported in flour quality (Abraham et al 1982) and in the DSC patterns and Brabender viscosity profiles of cassava starch (Asaoka et al 1991,1992; Rickard et al 1991).

The objective of this study is to evaluate varietal differences in starch and flour properties and to determine the influence of flour constituents on the gelatinisation properties of the starch.

Materials and Methods

The flour and starch samples were isolated from five varieties of cassava (H-97, H-165, S-856, H-1687 and M-4) which were harvested 9 months after planting. The material was grown in the Central Tuber Crop Research Institute ( Thiruvananthapuram, India). The flour was prepared by pulverizing and sieving (50µm) sun-dried chips. Starch was extracted from chopped cassava roots using a Waring type blender, passing the slurry through a 55 µm sieve, and allowing the starch to sediment overnight. Settled starch was dried in the sun light or in an oven at 400C.

Starch sugar and fibre contents were determined according to standard methods (AOAC 1975). The total and soluble amylose contents were determined according to standard procedure (Sowbhagya and Bhattacharya 1971; Santhy et al 1980).

The size and distribution of starch granules was determined on a Coulter Coulter® Model TA II (Coulter Electronics Ltd, Luton, UK) , after dispersing the starch into Isoton II® (filtered electrolyte solution, Coulter Electronic Ltd) and the equipment calibrated by PDVB Latex® (Coulter Electronics Ltd). The starch concentration in the dispersion was 0.5% and 12 replications were carried out for each sample.

Samples were extracted using petroleum ether (boiling range 40-600C) and absolute ethanol (Fisons, UK) for 14 h in a Soxhlet apparatus. The gelatinisation characteristics of the native and solvent extracted samples were determined calorimetrically using a differential scanning calorimeter (Perkin Elmer DSC-2). A slurry of the starch or flour in water (ratio 1:3) was accurately weighed into the pans, sealed and heated over a range of 27-970C at the rate of 100C min-1. The enthalpy of gelatinisation (?H cal g-1 ) was calculated on the basis of starch present in the slurry.

The pasting properties of the samples were determined using a Brabender amylograph. The change in viscosity of sample suspensions (50 g kg-1 dry weight basis) a 75 rev min-1 was continuously recorded using a 700 cmg-1 measuring cartridge. The temperature was increased from 50 to 950C at a heating rate of 1.50C min-1, held at 950C for 10 min and then cooled to 500C.

Results and Discussions

The chemical compositions of the flour and starches are given in Table 1 and 2. The starch content in the flour varied from 791.2 to 859.7 g kg-1, while it was > 985 g kg-1 for all the starch samples. The crude fibre content ranged from 14.9 to 29.8 g kg-1 in the flour samples, while it was < 1.5 g kg-1 in the starch of all the varieties. The total dietary fibre of the flour samples as determined by enzymatic digestion ranged from 47.6 to 54.9g kg-1 The fat content in starch varied from 1.1 to 2.2 g kg-1 and between 2.6 to 4.5 g kg-1 in the Flour samples. In flour, the fat content was two to three times higher than the levels found in starch. The ethanol soluble substances extracted from the flour samples ranged from 25 to 37 g kg-1 which can be accounted by the presence of sugars (Table 1). Values obtained for starch ranged from 9.0 to 13.0 g kg-1. Amylose and soluble amylose determinations showed only mirror variation in content among the different varieties (Table 2).

There was very little variation in the granule size range (3 – 32µm) among the varieties of cassava evaluated. However, there were some distinct differences in the granule size distribution patterns. The peak granular size for varieties H-165, H-1687 and S-856 was 16.4-20.2 µm, 10.3-12.7 µm for M-4 and 12.7-20.2 µm for H-97 (Fig 1). Varieties H-165, H-97 and S-856 had similar granular size range (2.4 – 31.1 µm) in four Colombian varieties of cassava.

The DSC patterns of native starch samples were found to vary between the five cassava varieties. Most of the starches and flour samples produced smooth DSC curves. The noticeable differences were the presence of a right sided shoulder in the peak of H-97 starch and M-4 starch which exhibited a broad peak (Fig 2). Peak profiles of all samples of the five varieties were unaffected by solvent extraction. The results indicate that distinct DSC peak profiles are due to the structure of the starch itself and are not attributable to the presence of solvent extractable constituents. In maize, DSC curves have been reported to be modified by the presence of fats and surfactants (Eliasson et al 1988). It is possible that the fat content in cassava is too low to affect the DSC curve. Similar results were obtained with the native and solvent extracted flour samples substantiating that the DSC peak profiles for cassava are a characteristic of the starch.

The gelatinisation temperatures, as observed using the DSC, are given in Table 3. Values obtained for the Tint ,Tmax, Tend varied with variety by up to 40C and were in all cases considerably higher than the results reported for four Colombian varieties of cassava starch (Asaoka et al 1992). The heat of gelatinisation (ΔH) was also higher for the different Indian cassava starch samples (2.7-3.4 cal g-1) than was reported for the Colombian varieties (1.7-2.1 cal g-1).

The Tmax values of the flours samples were 2 – 4.20C higher than those obtained for the corresponding starch. Similar differences were also obtained with Tint and Tend. These increases can be attributed to the presence of other components in the flour which delay the onset of gelatinisation probably by restricting the entry of water into the starch granules. However, solvent extraction of sugars and fats did not reduce the gelatinisation temperature of the flour samples to the values obtained for the starch samples, indicating that they do not exert any major influence on gelatinisation properties of cassava starch. The observed higher values for the flour could be due to presence of fibers which have been reported to have a role in delaying gelatinisation (Moorthy et al 1993).

Table 1 : Chemical constituents of starch and flour of five cassava varieties (± SD)

Variety Starch ( g kg-1) Sugar( g kg-1) Fat ( g kg-1) Crude fibre( g kg-1) TDFa( g kg-1)
Sugar 981.6 (±0.9) -b 1.8 (±0.4) 1.5 (±0.3) -
Flour 805.8 (±3.4) 27.2 (±3.3) 2.9 (±0.7) 22.3 (±1.3) 47.6 (±2.5)
Sugar 980.0 (±1.8) - 2.2 (±0.4) 1.3 (±0.3) -
Flour 791.2 (±1.5) 34.9 (±2.1) 2.7 (±0.4) 29.8 (±1.8) 54.9 (±2.8)
Sugar 982.7(±3.6) - 2.1 (±0.4) 1.1 (±0.3) -
Flour 825.9 (±4.5) 30.5 (±0.8) 2.6 (±0.3) 27.0 (±2.8) 49.9 (±1.9)
Sugar 985.5(±1.6) - 2.0 (±0.3) 1.2 (±0.6) -
Flour 812.4 (±4.6) 32.3 (±1.3) 3.6 (±0.4) 25.6 (±1.8) 51.1 (±2.9)
Sugar 981.2 (±0.8) - 1.1 (0.5) 1.2 (±0.2) -
Flour 859.7 (±6.9) 22.0 (±3.5) 4.5 (±0.6) 14.9 (±2.1) 46.7 (±1.9)

a TDF, total dietary fibre.

b not determined.

In a variety of other plant starches, fats are known to delay gelatinisation by forming complexes with amylose and amylopectin side chains (Krog 1973). The lack of effect of solvent extraction on the gelatinisation values obtained for cassava starch samples indicate that low levels of fat are associated with the starch granules. These results also indicate that varietal differences are not due to fats or sugars associated with the starch granules.

The data on enthalpy of gelatinisation(Table 2) showed that there was little varietal variation in the values obtained and that flour samples had lower values compared with the starch samples. It was also observed that solvent extraction had little effect on the values obtained.

Table 2 : Amylose contents of starch

Variety Total amylose(g kg-1) Soluble amylose(g kg-1)
H-97 234 57
H-165 226 51
S-856 241 61
H-1687 251 62
M-4 262 59

The Brabender amylograms of the samples examined are presented in Fig 3(a-e). Significant varietal differences were observed in gelatinisation profiles obtained for the starch samples. Varietal differences obtained in these experiments were considerably greater than those reported for four Colombian varieties of cassava (Asaoka et al 1992). Pasting initiation temperatures (PIT) recorded for the Indian varieties were also considerably higher (65-690C) than those reported by Asaoka et al (1992) (58-630C).

Solvent extraction had only a minimal effect on the gelatinisation characteristics of the starch samples, peak viscosity being slightly enhanced for all varieties. These results also indicate that fats and sugars which can be associated within the starch granules are not responsible for the varietal differences observed.

There was a significant difference in the profiles obtained between the starch and flour with three varieties, H-1687, H-97 and especially H-165. Lower viscosity values were obtained for all the flour samples but in the case of S-856 the differences were minimal. Visocity values for the flour samples also remained more stable throughout the temperature programme.

Fig 1

Variation in granule size distribution in cassava varieties

Table 3 : DSC data on native, defatted and ethanol extracted starch and flour

Variety     Starch     Flour
Tint Tmax Tend ΔH Tint Tmax Tend ΔH
  (0C) (0C) (0C) (cal g-1) (0C) (0C) (0C) (cal g-1)
Pl 67.1 71.4 75.4 3.2 70.0 73.9 79.1 2.2
De 67.3 70.9 75.1 3.4 69.5 73.3 78.1 2.0
EE 67.5 71.3 75.4 3.1 69.4 73.0 77.0 2.0
Pl 65.4 69.2 74.9 3.3 68.7 72.0 77.2 2.1
De 66.0 69.4 73.9 3.3 69.9 73.1 78.1 2.0
EE 65.0 68.9 73.7 3.2 68.9 73.0 77.9 2. 1
Pl 69.4 72.3 77.1 3.4 71.8 75.0 79.9 2.3
De 69.0 71.3 76.9 2.8 72.0 75.4 80.0 2.1
EE 68.0 71.0 76.1 2.8 71.7 75.2 79.8 2.0
Pl 65.6 70.1 74.9 2.7 69.1 74.1 79.2 2.4
De 65.2 70.3 75.1 2.8 68.1 73.2 78.3 2.4
EE 65.1 70.7 74.8 2.7 69.3 73.4 79.4 2.3
Pl 68.2 73.2 78.5 3.0 71.1 75.7 81.2 2.0
De 67.9 72.8 77.3 2.9 70.1 75.1 82.1 2.0
EE 67.8 73.0 77.8 3.1 70.1 74.8 81.5 2.0