Functional Properties of the Starchy Flour Extracted from Cassava on Fermentation with a Mixed Culture Inoculum

S.N Moorthy, Mathew George and G padmaja

Division of Post-Harvest Technology, Central Tuber Crops Research Institute, Sreekariyam, Thiruvananthapuram 695 017, Kerala, India

(Received 2 June 1992; revised version received 5 November 1992; accepted 5 February 1993)

Abstract: The fundamental properties of the starchy flour extracted from six varieties of cassava subjected to fermentation by a mixed culture inoculum comprised of lactobacilli, streptococci, corynebacteria and yeast cells are examined. Apparent reduction in total and soluble amylose contents is observed. Differential scanning calorimetry of the samples indicated that the enthalpy of gelatinisation was reduced, while gelatinisation temperature was enhanced. A marked reduction in Brabender viscosity values of starch from fermented tuber was observed, but the X-ray diffraction pattern was unaffected. All these changes were attributed to the presence of fibrous material and consequent reduction of starch in unit volume rather than any change un the starch granule structure.

Keywords: cassava, starch, flour, fermentation, mixed culture.


Cassava (Manihot esculenta Cants) is an important tropical root crop which finds extensive use as a food and in the food and other industries. The tubers are usually consumed after cooking just like potato. Fermentation also is practiced in some countries like Brazil and Nigeria (Akinrele 1964; Lancaster et al.1982) and the process is found to help in detoxifying the cyanogenic glucosides present in the tubers (Collard and Levi 1959). ‘ Polvilho azedo’ is an important fermented product in Brazil while ‘gari’ is widely consumed in Nigeria. Fermentation of cassava tubers was carried out using a mixed culture inoculum and it was found that the tubers became soft within 24-48 h so that the amount of starch recovered from the fermented tuber is higher (George et al. 1991). Various biochemical changes taking place during the fermentation have been identified, the major observations were that the pH dropped considerably within 24 h of fermentation and then remained steady, and that the fibre content in the recovered starchy flour increased noticeably with the time of fermentation (George et al unpublished ). Camargo eta l. (1988) have recently studied the properties of sour starch formed during the fermentation of starch from cassava and found that only minor changes occurede in the chemical composition and that the granular structure was similar to that resulting from short periods of acid hydrolysis. Martinez and Quiroga (1988) have examined the physiochemical properties of cassava starch during fermentation and observed that some starch granules became smaller due to loss of superficial layers while others remained intact. In this paper all effects of mixed culture fermentation on the starch properties of some varieties of cassava are reported.

Materials and Methods

The six varieties of cassava used for this study were grown at the Central Tuber Crops Research Institute and harvested at the tenth month stage. Fermentation was carried out according to a procedure described earlier (George et al. 1991). Peeled tubers which had been cut into prismoid pieces of 3×2 cm were dispensed in double their weight of water, inoculated with 1 ml mother liquor per 100 g tuber pieces and left to ferment for 72 h. the mother liquor contained 7.5 x 105 lactobacilli, 3.5×105 streptococci, 6.5×105 corynebacteria and 0.7×105 ml-1 yeast cells, counts being determined in nutrient agar. Samples were taken at 0, 24, 48 and 72 h of fermentation. The extraction of flour was carried out by disintegration in water, sieving and settling.

The gelatinisation of the starch by differential scanning caorimetry (DSC) was examined using a Perkin elmer DSC-2 calorimeter. Sealed pans were heatee from 27 to 970 at a rate of 100 min-1 and the data plotted using a data plotter. The X-ray diffractometry (XRD) was done using a Philips XRD using CuKα source and the range used was 2θ=4-600. The swelling volume was determined by Schcoh’s procedure (Schcoh 1964) at 950. The pasting behavior was studied with a Brabender Viscoamylograph at a starch concentration of 60 g DM kg-1 and rotation speed of 75 rpm. The temperature was raised at 1.50 min-1 from 40 to 950 allowed to remain constant for 15 min and then cooled to 300 at the same rate. Pasting temperatures were read from the viscographs.

The amylose contents were determined according to standard procedures (Showbhagya and Bhattacharya 1971; Shanthy et al 1980).

Results and Discussion

The amylose content in the starch from fermented and non-fermented tubers is present in Table 1. The trends indicate that there is an apparent reduction in amylose content on fermentation between 24 and 48 h. however, the amount of starch present in the recovered samples also falls during 24, 48, and 72 h fermentation (George et al. 1991) indicating that the apparent reduction in amylose is not simply due to the fall in actual amylose content. The results with the sample containing antibiotic confirm the observation that total amylose is not influenced by fermentation. Camargo et al (1988) in their studies on sour starches observed only a minor difference in the amylose content between fermented and non-fermented starches. Martinez and Quiroga (1988) observed loss in only superficial layers and hence the amylose fraction remained unaffected. The authors were also interested in observing the effect of fermentation on the soluble amylose fraction since it is assumed that soluble amylose is present in the amorphous regions of the starch granules and fermentation may be affecting these which are more easily accessible to the microorganisms. However, the reduction observed in the soluble amylose content is relatively low and the observed reduction can be attributed to the fall in starch and also total amylose contents. The effect of addition of antibiotics is alos quite evident in the soluble amylose data.

The XRD patterns of the starch extracted from fermented and non-fermented tubers are found to be largely similar and all the samples exhibit an ‘A’ pattern (Fig. 1). The presence of a relatively larger quantity of fibrous material in the starch did not affect the XRD pattern or peak heights to any significant extent and the non-crystalline nature of the fibrous residue explains this result. The XRD observation is in conformity with the results of Camargo et al (1988) who also did not observe any effect of fermentation on the XRD of sour starch.

Table 1 : Amylose and soluble amylose contents (blue values) in fermented and non-fermented tubers.

Variety Fermentation Time (h)
0 24 48 72
Amylose content  
M4 0.282 0.260 0.249 0.250
H.165 0.282 0.250 0.240 0.228
T.300 0.332 0.297 0.284 0.294
CI.468 0.314 0.305 0.296 0.290
H.1687 0.320 0.291 0.272 0.237
H.2304 0.2777 0.257 0.238 0.238
Soluble amylose Content  
M4 0.155 0.135 0.148 0.128
H.165 0.147 0.141 0.143 0.114
T.300 0.158 0.161 0.152 0.154
CI.468 0.164 0.182 0.143 0.135
H.1687 0.172 0.146 0.138 0.147
H.2304 0.166 0.154 0.150  

aAntibiotoc added control

XRD Patterns of Flour from Fermented and non-Fermented Tubers

XRD Patterns of Flour from Fermented and non-Fermented Tubers

Table 2 : DSC data (0C) of flour from fermented and non-fermented tubers

Temperature Fermentation Time (h)
0 24 48 72
H .1687  
Initial 70.12 72.72 72.34 -
Maximal 73.99 77.60 75.52 -
Final 80.69 82.24 81.21 -
Enthalpy J g-1 2.03 1.78 1.35 -
T .300  
Initial 69.12 70.97 69.49 69.60
Maximal 71.87 73.85 72.67 73.44
Final 75.27 79.89 80.44 78.94
Enthalpy J g-1 1.86 1.67 1.46 1.37

DSC data of the non-fermented and fermented samples of two varieties studied were T. 300 and H. 1687. The initial temperature for H-1687 increases from 70.72 to 72.72 and 72.340C respectively on 24 and 48 h fermentation; while the maximal temperature increases from 73.99 to 77.600C after 24 h, and then falls to 75.520C after 48 h. The final temperature values are also slightly enhanced. The initial, maximal, and final temperatures for starch of T .300 are also enhanced but less consistently compared with H.1687 starch. Camargo et al. (1988) have obtained an increase of 1-20C which they consider minor. Pasting temperatures obtained from viscosity measurements also are enhanced. An increase in the gelatinisation temperatures of starch by various ingredients has been well documented, especially by lipids and surfactants and it has also been observed that cassava flour always has higher gelatinisation temperatures compared with cassava starch. Hence, the increase in gelatinisation temperature is due to the presence of fibrous material rather than any change in starch structure quality. The enthalpy of gelatinisation falls from 2.01 to 17.6 J g-1 fresh weight of sample after 24 h and finally to 1.34 J g-1 for starch of H . 1687 and from 1.84 to 1.38 J g-1 for starch of T . 300. Again, the reduction in enthalpy can be attributed to the reduced starch content in the fermented sample. Whether samples re-run after being cooled would give a thermogram differing from the initial one was not investigated since the lipid content in tuber starches in negligible. The reduction in enthalpy in the sour starch observed by Camargo et al (1988) is much lower and the absence of fibre can explain the difference between those and the present results (Fig. 2).

Studies with a viscoamylograph showed that both results for the viscosity of fermented and non-fermented starches indicated a similar trend (Table 3). There is a marked reduction in the peak viscosities and the viscosity at 970C for all the starches on fermentation. Though there is wide variation among the different varieties as far as the fall in viscosity is concerned, the trend is similar. As the fall in viscosity is concerned, the trend is similar. For M-4 starch, the peak viscosity drops from 580 BU for the non-fermented sample to 480, 400 and 400 BU after 24, 48 and 72 h, respectively. The viscosities at 970C are nearly the same as peak viscosities for fermented samples. For H .165 starch, the fall in peak viscosity is much more pronounced, having come down from an initial 910 to 700, 660 and 660 BU, respectively, at 24, 48 and 72 h fermentation. Here also the viscosity at 97 0C exhibits only a minor reduction or fermentation. For H.2304 starch viscosity values fall from 1000 BU to 740, 620 and 420 BU; both peak viscosity and the drop in viscosity are higher for this variety. For variety T.300 also the drop in viscosity is quite large, namely 500 BU. However, as with other varieties the break down in viscosity is reduced after 48 h fermentation. Thus, fermentation appears to be effective in reducing viscosity breakdown. The presence of fibrous material seems to help in stabilizing the viscosity. The reduction in viscosity of the starch content and also to the effect of fermentation on the starch quality. The actual contribution of each of the factors is difficult to assess since the starch could not be separated from the fibre either by centrifugation or filtration. However, based on the DSC and swelling volume data, as well as on the earlier reports on the effect of fermentation, the damage to starch may not be very significant. Though the observed reduction in viscosity due to fermentation is quite significant, it is not high enough to indicate extensive starch damage, since breakdown in viscosity of damaged starch is usually much greater. Camargo et al (1988) have also observed reduction in viscosity is considered to be due to the greater solubility of fermented starch in hot water and hence the smaller fractions (by volume) of swollen granules in the paste. Thus, the reduction in viscosity can be attributed to a very small extent to reduction in starch quality but mostly to the presence of fibrous material. Many substances such as lipids and surfactants are known to suppress the viscosity of cassava starch (Krog 1973 ; Moorthy 1985) and the fibre acts as a barrier to free swelling and hence to achieving high viscosity (Fig. 3).

Table 3 : Viscocity data (BU) of flour fermented and non-fermented tubers

Variety Fermentation Time (h)
0 24 48 72
PV V97 PV V97 PV V97 PV V97
M4 580 500 480 480 400 400 400 400
H.165 910 500 700 640 660 660 660 660
H.2304 1000 640 740 540 620 580 420 380
T.300 980 540 780 540 440 420 460 420
H.1687 860 400 640 480 450 390 500 480
CI.468 810 460 620 480 570 500 630 580

aPV, peak viscosity, V97, viscosity at 970C.

DSC Patterns of Flour from Fermented and non-Fermented Tubers

DSC Patterns of Flour from Fermented and non-Fermented Tubers

The observation that viscosity breakdown is only minimal in fermented starch indicates the possibility of the use of fermented starch in food products. Camargo et al (1988) have used fermented starch in baking tests, but the fermented starch may be more useful in pudding-type products, where the cohesive nature of plain starch is a disadvantage.

As observed from the DSC data, the pasting temperatures observed from viscographic data are enhanced by fermentation. The increase in very pronounced, even up to fermentation. The increase is very pronounced, even up to 150C after 72 h fermentation, for all the starches. For M4 starch, the increase is maximal and is reached after only 24 h. the enhancement in pasting temperature, as explained earlier, can be attributed to restriction of entry of water molecules by the fibrous material present.

Table 4 : Swelling Volumes (ml g-1) of flour from fermented and non-fermented tubers

Variety Fermentation Time (h)
0 24 48 72
M4 Starch 26.25 21.87 20.62 20.0
Fibre - 0.12 0.25 0.19
H.165 Starch 36.87 34.37 31.25 26.87
Fibre - 0.20 0.25 3.12
T.300 Starch 30.62 29.37 16.87 16.87
Fibre - 0.12 3.12 3.12
CI.468 Starch 28.75 16.87 23.75 25.62
Fibre - 1.00 2.00 1.621
H.1687 Starch 26.87 21.87 18.75 19.37
Fibre - 1.12 2.5 3.12
H.1687a Starch - - 29.4 -
Fibre - - - -
H.2304 Starch 29.37 25.62 18.12 10.0
Fibre Nil 1.00 2.50 3.12

aAntibiotic added

The swelling volumes of the starches, presented in Table 4, are invariably reduced by fermentation. For M4 starch, the values fall from 26.25 ml g-1 to 21.62 ml g-1 after 24 h and then to 20.0 ml g-1 after 72 h. for H .165 the values progressively fall to 26.87 from 36.87 and for T.300 from 30.62 to 16.87. A similar reduction is observed also for other varieties. The results in this paper are in contrast to those obtained by Camargo et al. (1988) who observed an increase in swelling, especially at 950C. It is also found that a layer of fibrous residue forms during centrifiguration and the volume of this layer increase progressively with fermentation time, again confirming the observation that fibrous material invariably contaminates the starch during its extraction from fermented tubers. Though the reduction in swelling volume is reflected in the fall in viscosity there is no correlation between the fall in viscosity and that in swelling volume. The swelling volume reduction is attributable to the greater solubility of the fermented starch as well as the presence of fibrous residues which can restrict swelling.


During the fermentation of cassava using the mixed culture inoculums, fibrous material is extracted along with starch. The gelatinisation of starch is delayed by the presence of the fibrous material as indicated by DSC data. The enthalpy of gelatinisation is reduced depending on the amount of fibre present. The viscosity and swelling volume fall while the pasting temperature is enhanced. All the results indicate that the starch properties are not very much affected, but the presence of fibre changes the functional properties of the resulting flour to a large extent.