SOME FUNCTIONAL PROPERTIES OF SELECTED GHANAIAN CEREALS AND LEGUMES A Thesis Presented to The Faculty of Graduate Studies of The University of Guelph by AKOSUA FREMA OSEI-OPARE In partial fulfilment of requirements for the degree of Master of Science April, 1976 Akosua Frema Osei-Opare, 1976 University of Ghana http://ugspace.ug.edu.gh Q 275135 University of Ghana http://ugspace.ug.edu.gh ABSTRACT SOME FUNCTIONAL PROPERTIES OF SELECTED GHANAIAN CEREALS AND LEGUMES Akosua Frema Osei-Opare, M.Sc. Supervisor: University of Guelph, 1976 Dr. Elizabeth A. Gullett This study was designed to examine'the functional properties of whole flours and starches of sorghum, millet, cowpea and bambara nuts from Ghana as well as Canadian wljeart. Functional properties tested on starches were the gelatiniza- tion temperature and range, and swelling and solubility patterns. The basic chemical analysis of flours determined and the functional properties tested were viscosity characteristics, gel strength and retrogradation tendency of flour gels. Gelatinization temperature and range of.starches were deter­ mined using a staining technique, and the swelling and solubility patterns were tested using a modified Leach's method. Viscosity of hot and cooled flour pastes were measured with the Brookfield Visco­ meter..-1 The gel strength of flours were determined with the Precision Penetrometer and retrogradation of flour gels was measured as percen­ tage syneresis over different periods of cold storage. The findings of this study showed that all the Ghanaian starches tested gelatinized at higher temperatures and ranges than wheat starch. The trend observed for the swelling and solubility patterns of cereals were^similar to each other than between the legumes. Legumes exhibited higher Tfegree of swelling and solubility than the starches of cereals. Viscosity of flour pastes indicated that sorghum and bambara nut flours University of Ghana http://ugspace.ug.edu.gh had thicker pastes. In cold storage, all flour gels showed increased strength as well as syneresis. Wheat flour gels, however, did not portray any sign of syneresis. Direct relationships were found between gelatini- zation characteristics and the swelling and solubility patterns of starches. There was no direct relationship between swelling and solubility of starches, and viscosity characteristics of flours from the same source. Gel strength of flours in cold storage and solubles of starch at 95°C showed a direct relationship. The functional properties of the Ghanaian flours and starches indicated a high potential for their use in the home as well as in industry. I University of Ghana http://ugspace.ug.edu.gh THIS THESIS IS DEDICATED TO MY PARENTS FOR BEING UNDERSTANDING AND UNSELFISH University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENTS I wish to express my appreciation of the help, encouragement, and labours of my thesis committee, Drs. E.A. Gullett, T.A. Watts and J. Sabry. Especially, I acknowledge the helpful direction and confidence provided by Dr. E.A. Gullett, the supervisor of my studies. The author wishes to thank the Ghana-Guelph Project (C.I.D.A.) for providing her the scholarship to study in Canada, especially the Director, Dr. J.C. Shute, for his valuable contribution to my studies. To the Shute family, my sincere thanks for their very personal interest and support. My warmest thanks are tended to Drs. J. Tanner, J. Burton, D. Pletsch and Miss G. Frank for their individual help at various stages of this study. I wish to thank my friends, Mrs. Debby Ricci, Mrs. Donna Rowe, Mrs. Voil and Mr. Sefa Dede for their assistance and encouragement. Debbie Cornwell, I wish to express my sincere thanks for her time and patience in typing this thesis. Most of all, I am indebted to my husband, Dr. Dua Opare, for the generous understanding, inspiration and encouragement received from him at all stages of my study programme. Finally, I should like to express my gratitude to my parents for their unselfish attitude which has made it possible for me to pursue higher education up to this level. University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS Page ACKNOWLEDGEMENTS ................................................ i LIST OF TABLES ............................................... iv LIST OF FIGURES .............................................. v LIST OF APPENDIX TABLES ...................................... vii INTRODUCTION ................................................. 1 REVIEW OF LITERATURE .......................................... 3 Gelatinization ........................................... 3 Swelling and Solubility .................................. 6 Viscosity ............................................... 9 Gel Strength and Retrogradation .......................... 13 Relation of Functional Properties to Practical Application . 17 EXPERIMENTAL METHODS .......................................... 20 Milling ................................................. 20 Starch Preparation ....................................... 20 Basic Chemical Analysis .................................. 21 Determination of Gelatinization Temperature and Range of Starches ................................................ 22 Determination of the Swelling and Solubility of Starches ... 23 Calculations ............................................ 24 Viscosity Characteristics of Flours ...................... 24 Gel Strength and Retrogradation of Flours ................ 26 RESULTS AND DISCUSSION ........................................ 28 Basic Chemical Analysis .................................. 28 Gelatinization Temperature and Range of Starches .......... 28 Swelling and Solubility Patterns of Starches .............. 33 ii University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS (cont'd.) Page Viscosity ................................................ 44 Gel Strength and Retrogradation ........................... 52 Interrelationships Between Some Functional Properties ..... 56 PRACTICAL IMPLICATIONS OF STUDY ................................ 61 Bread Making ............................................. 61 Porridge and Paste-Like Products .......................... 69 Thickeners ............................................... 75 Gel-Type Products ........................................ 76 SUMMARY AND CONCLUSION ........................................ 78 LIST OF REFERENCES ............................................ 80 APPENDIX TABLES ............................................... 83 iii University of Ghana http://ugspace.ug.edu.gh Table Page 1 Basic chemical analysis of five flours used in this study .................................... 29 2 Basic chemical analysis of five flours as reported by Watt and Merrill (1963) and Watson (1971) .................................... 30 3 Gelatinization characteristics of five starches .... 32 4 Essential amino acids of five grains .............. 62 5 Protein content and amino acid score of wheat- cowpea flour mixtures ............................ 63 6 Protein content and amino acid score of wheat- bambara nut flour mixtures ....................... 64 7 Protein content and amino acid scores of sorghum-cowpea flour mixtures .................... 70 8 Protein content and amino acid scores of sorghum-bambara nut flour mixtures .-............... 71 9 Protein content and amino acid scores of millet-cowpea flour mixtures ..................... 72 iv LIST OF TABLES 10 Protein content and amino acid scores of millet-bambara nut flour mixtures ..... 73 University of Ghana http://ugspace.ug.edu.gh Figi 1 2 3 4 5 6 7 8 9 10 11 12 v LIST OF FIGURES Swelling power and solubility of wheat starch at different temperatures of heating ......... Relationship between swelling power and solu­ bility of wheat starch ...................... Swelling power and solubility of sorghum starch at different temperatures of heating ......... Relationship between swelling and solubility of sorghum starch ........................... Swelling power and solubility of millet starch at different temperatures of heating ......... Relationship between swelling power and solu­ bility of millet starch ..................... Swelling power and solubility of cowpea starch at different temperatures of heating ......... Relationship between solubility and swelling power of cowpea starch ...................... Swelling power and solubility of bambara nut starch at different temperatures of heating ... Relationship between solubility and swelling power of bambara nut starch .............. Swelling patterns of five starches ........ Solubilization patterns of five starches . .. Effect of concentration on the viscosity of wheat flour at three different temperatures University of Ghana http://ugspace.ug.edu.gh vi LIST OF FIGURES (cont'd.) Figure Page 14 Effect of concentration on the viscosity of sorghum flour at three different temperatures ....... 46 15 Effect of concentration on the viscosity of millet flour at three different temperatures ........ 48 16 Effect of concentration on the viscosity of cowpea flour at three different temperatures ........ 49 17 Effect of concentrations on the viscosity of bambara nut flour at three different tempera­ tures ............................................. 50 18 Gel strength of five flour gels ..................... 53 19 Amounts of syneresis of five flour gels ............. 55 University of Ghana http://ugspace.ug.edu.gh 11 12 13 14 15 16 17 18 19 20 21 22 vii LIST OF APPENDIX TABLES Swelling and solubility of wheat starch . Swelling and solubility of sorghum starch Swelling and solubility of millet starch Swelling and solubility of cowpea starch Swelling and solubility of bambara nut starch Viscosity of wheat pastes Viscosity of sorghum flour pastes Viscosity of millet pastes Viscosity of cowpea pastes Viscosity of bambara nut pastes Effect of temperature on gel strength of five flours .................................... Effect of temperature on degree of syneresis of five flour gels ....................... University of Ghana http://ugspace.ug.edu.gh INTRODUCTION A good deal of attention is currently being paid to the pre­ servation and increased use of locally produced foodstuffs in Ghana. Both home and industrial uses of local produce are of growing concern to government, agricultural researchers and home economists. Cereals and legumes in particular, form a great part of the Ghanaian diet and also contribute a large proportion of the total protein to the diets of both adults and children. A good percentage of the cereals and legumes used are locally produced and prepared. However prepared weaning foods, convenience products (breakfast cereals etc.) and wheat flour are regularly imported into the country. Bread, made from imported wheat, has become one of the staples in the diet as an established breakfast and snack food. There has been interest in the use of flours from alternate sources to replace wheat wholly or par­ tially in products requiring it. Very little research into local grains (cereals and legumes) has been carried out and therefore even though the country has recognized the need, the problem has not been resolved. Another problem facing Ghana is that of food distribution. Many foods produced across the country tend to be confined to the areas of production and hence may be unknown or unavailable in other parts of the country. Oke (1964) noted that sorghum is extensively grown in the northern part of Nigeria but that its use is almost confined to that area. A similar situation exists in the southern part of Nigeria with maize. This situation can lead to surpluses of some grains in some parts of the country while there are shortages of similar produce elsewhere. Further, this has limited the variety 1 University of Ghana http://ugspace.ug.edu.gh 2 possible in the diet. Therefore the purpose of this study was three-fold: 1) To determine for some Ghanaian grains and Canadian wheat the follow­ ing functional properties i.e. i) gelatinization temperature and range of the starches, ii) swelling and solubility patterns of starches, iii) viscosity characteristics of flours and iv) gel strength and retrogradation tendencies of flour gels. 2) To compare the functional properties of cereals (wheat, sorghum and millet) with those of legumes (cowpea and bambara nut). 3) To compare all other flour and starch properties of the Ghanaian cereals and legumes to those of wheat. The cereals and legumes selected for these were wheat (High quality, Hard Red, Spring wheat, Triticum aestivum L., "Manitou"), sorghum (Sorghum vulgare "932111"), millet (Pennisetum typhoides "Staph & Hubb"), cowpea (Vigna unguiculate "Caroni"), and bambara nut (Voandzeia subterrana). University of Ghana http://ugspace.ug.edu.gh REVIEW OF LITERATURE In order to appreciate the functional properties of starch, the structure of a starch granule has to be understood. Meyer (1947) described the structure of a starch granule as crystalline in form consisting of intermixed linear (amylose) and branched starch (amylo- pectin) molecules. These linear and branched molecules are arranged in a linear form and whenever adjacent linear segments of either amylose or amylopectin parallel one another, they are pulled together into crystalline bundles by hydrogen bonding. Meyer (1947) then postulated that when hydrated these crystalline areas acting as bonds between granules either permit or prevent dispersion and solu­ tion of individual starch granules. Leach et_ al_. (1959) suggested additional reasons for the swelling behaviour of starches. They postulated that differences in swelling behaviour of various starches may also be attributed to the differences between bonding forces within the starch granule. Gelatinization Smith (1959) defined gelatinization as the complete rupturing of starch granules by moisture, heat, pressure and in some cases mechanical shear. The temperature range over which gelatinization occurs is called the gelatinization range and the temperature at which it is completed is called the gelatinization temperature (MacMasters, 1964). On heating an aqueous suspension of starch, no apparent changes occur until the medium reaches a certain temperature and at this point 3 University of Ghana http://ugspace.ug.edu.gh A some of the granules begin to swell very rapidly (Paul and Palmer, 1972). Schoch et al. (1967) said that different varieties of starch tend to exhibit different gelatinization characteristics. The gelatinization of starch granules may be affected by other factors. Paul and Palmer (1972) stated that small amounts of non-carbohydrate material such as lipids and proteins are closely associated with starch in the plant cell and that in some cases they may not completely be eliminated during starch preparation, either industrially or in the laboratory purification method. Some chemicals have been found to either enhance or inhibit gelatinization. Leach (1967) reported that esterification of starch lowers the gelatiniza­ tion temperature and that sodium sulfate inhibits gelatinization. Another factor that affects gelatinization of starch granules is pH. Normally within a pH range of five to seven, gelatinization charac­ teristics are not affected. Collison (1968) categorized the various methods of determina­ tion of starch gelatinization temperature and range into two groups: 1) those involving microscopic work including direct observation of swelling, loss of birefringence on heating and staining techniques, and 2) those involving physical measurements on starch pastes. Collison (1968) disclosed that the microscopic methods have the advantage of making possible direct observation of the behaviour of individual granules. Leach (1967) said that among the microscopic methods, the measurement of the loss of birefringence is the most sensitive, accurate and reproducible technique for determination of initial gelatinization of starch. The Kofler gelatinization equipment is commonly used to measure loss of birefringence. Staining technique University of Ghana http://ugspace.ug.edu.gh 5 is much simpler to handle than measurement of birefringence. Among the various stains used, iodine and congo-red dye seem to be the most popular. Jones (1940) stated that congo-red dye stained both gelatinized and damaged granules. Even though congo-red dye has been suggested as a preferred stain for gelatinized granules, Schoch et al. (1967) reported that they have had little success with this method because of the difficulty in determining when gelatinization had begun. In spite of this, some investigators have had success with it (Jones, 1940; MacMasters, 1964). Collison (1968) regarded physical methods such as sedimentation as less precise for completion of gelatinization. Rasper et al. (1974) reported the gelatinization range for wheat starch was 52 to 62°C whereas Osman (1967) reported 59 to 64°C. Both researchers measured gelatinization as the loss of birefringence using the Kofler Hot-Stage microscope. The gelatinization range for sorghum was found to be 68 to 74°C by Leach (1967) while Rasper et al. (1974) found it to be 69 to 74°C. Rasper et al. (1974) found the gelatinization range for millet to be 70 to 75°C. The gelatini­ zation range for five varieties of cowpea was determined by Tolmasquim et âL. (1971). They found that all the varieties gelatinized at different temperatures and ranges. The lowest observed initial gelatinization temperature was 64°C and the highest observed was o70 C. For final gelatinization temperature they recorded the lowest value as 69°C and the highest as 73°C. These findings reflected the effect of both variety and method­ ology on the gelatinization characteristics of starches. University of Ghana http://ugspace.ug.edu.gh 6 Swelling and Solubility Although starch is highly hydroxylated and therefore very hydrophillic, it is insoluble in cold water (Leach, 1954). However, when an aqueous suspension of starch is heated beyond the initial gelatinization temperature, hydrogen bonds continue to be disrupted and water molecules become attached to the liberated hydroxyl groups, thus causing the granules to swell (Leach, 1954). As the starch granules swell during heating, some granules become fully hydrated and these become separated from the micellar network. These then diffuse into the surrounding aqueous medium and form the solubles. Radley (1957) stated that the shorter amylose molecules are prefer­ entially solubilized and leached out from the swollen granules. The major factor controlling the swelling behaviour of starch is the strength and nature of the micellar network within the granule and these in turn are dependent on the degree and kind of association of the micellar network. As an example Leach ejt al. (1959) said that starches which are highly associated (with an extensive and strongly bonded micellar network) are relatively resistant to swelling. At the molecular level, factors such as the size, shape and composition and distribution of the micellar areas in the internal lattice, the ratio of amylose and amylopectin all affect swelling characteristics. Different starch samples vary widely in their swelling charac­ teristics (Collison, 1968). Other important features of swelling are the initial gelatinization temperature and the temperature range over which the granules rupture. Collison (1968) pointed out that these three variables are not directly related. It has been found that even though some starches from differing varieties have similar initial University of Ghana http://ugspace.ug.edu.gh 7 gelatinization temperatures in the region of 60°C, subsequent swelling behaviour varies widely from variety to variety. Cereal starches tend to exhibit a two-stage or multi-stage swelling and solubility pattern and thus show the presence of two sets of internal bonding forces (Leach, 1954). When cereal starches were compared with root starches, Leach (1954) found that the latter swelled to a greater extent and at a more uniform rate than the former because cereal starches have been found to have a more compact structure and a different crystallinity. Collison (1968) suggested that cereal and other starches (non­ root) do not undergo complete molecular dispersion and hence tend to have lower swelling power than root starches. Schoch and Mayland (1968) investigating the properties of some legume starches found that these showed rather high solubilities. Deatherage jet̂ al. (1955) reported that various normal legume starches have a rather high con­ tent (30 to 36%) of linear fraction which may indicate restricted swelling. However, lima bean was found not to have restricted swel­ ling even though its content of linear fraction was comparatively high (Deatherage et_ al., 1955) . Leach (1954) documented the basic requirements necessary for the determination of the swelling and solubility patterns of starch. The first concept is that there should be excess water present so that during gelatinization the granules can swell freely without mechanical disintegration. According to Leach (1954), a weighed amount of starch is suspended in distilled water in a centrifuge bottle and heated for 30 minutes at a specified temperature with con­ stant stirring. After 30 minutes in the bath, the sample is centri­ University of Ghana http://ugspace.ug.edu.gh 8 fuged from which an aqueous supernate is obtained and the weight of sedimented paste is determined. The solubilities are determined directly by drying an aliquot of the supernate. Sandstedt and Abbott (1964) reported some results for the solubility and water absorption patterns of wheat starch. They found that wheat starch solubles were the same over the temperature range 40 to 60°C. Beyond these temperatures there began a noticeable rise in solubility values. Sorghum starch (white milo) was examined by Leach et al. (1959). The authors reported that at 5% concentration the swelling power of white milo ranged from 5 at 75°C and 22 at 95°C. Solubility at this temperature range was reported to be between 3% at 75°C and 21% at 95°C. A look at the swelling power and solubility versus temperature revealed that there were sharp increases in these varia­ bles between 90 to 95°C (i.e. solubility increased from 14 to 21% and swelling power from 14 to 22). As Leach (1954) pointed out, for any given starch species, the swelling and solubility curves were similar and this indicates a direct interrelation of these two func­ tions. Tolmasquim et_ al. (1971) determined the swelling and solubility patterns of five varieties of cowpea starches, using a modification of the procedure described by Leach al_. (1959). Generally, they found that at all temperatures the swelling power values were higher than the solubilities for all of the varieties of starch. Even though swelling power values were different among varieties, all starch varie­ ties exhibited similar solubilities. Across the five varieties, solu­ bility values obtained from 75 to 95°C ranged from about 2.5 to 20% University of Ghana http://ugspace.ug.edu.gh 9 respectively and for swelling power, 4 to 32 within the above tempera­ ture range. Even though microscopic examinations of legume starches, including cowpeas, showed limited swelling, Tolmasquim et_ al_. (1971) found that the swelling power values obtained from these varieties of cowpeas showed significant swelling. Viscosity Paul and Palmer (1972) define viscosity as the measure of internal friction between adjacent layers of a liquid. Most food materials have non-Newtonian viscosities, i.e. the viscosity is not a single measure but a function of the shear rate during testing (Scott-Blair, 1969). Therefore, viscosity of non-Newtonian fluids such as starch pastes, custards, etc. is usually referred to as "apparent viscosity". The apparent viscosity may either fall or rise with increasing shear rate. However, the former situation is more commonly encountered (Scott-Blair, 1969). Smith (1964) reported that when an aqueous suspension of granular starches is heated above the gelatinization temperature, these starch granules become highly hydrated so they swell to several times their original volumes. If paste is continuously heated, espe­ cially while shearing (stirring), a paste is produced which consists of swollen granules, granule fragments and molecularly dispersed starch molecules which are leached from the granules. Therefore, changes in viscosity during cooking are related to extent of granule swelling and disintegration of the starch granules. There are a number of factors which affect the viscosity of fluids and these must be taken into account during viscosity determi­ University of Ghana http://ugspace.ug.edu.gh 10 nations. One important factor is temperature (Paul and Palmer, 1972), since generally viscosity decreases rapidly with increasing tempera­ ture provided that other changes are not occurring at the same time. The concentration of the fluid also affects the viscosity of it. Schoch and Mayland (1968) divided the viscosity patterns of starches into four types: Type A. Refers to high-swelling starches such as potato or tapioca and waxy cereals. These starches tend to swell enormously but are fragile towards shear; hence, after a peak viscosity is reached there is a rapid major thinning during cooking. Type B. These are the moderate swelling starches such as normal cereals. These starches do not have as high an initial vis­ cosity peak and tend to thin less. Type C. Refers to the restricted swelling starches such as chemically cross-bonded products. Cross-linkages reduce swelling and solubility and these starches are most stable to mechanical fragmentation and therefore show a continued increase in viscosity during cooking. Type D. Refers to starches which are highly restricted in swelling. They usually contain about 55 to 70% amylose for example, "high amylose" corn starch. "High-amylose" corn starch has such an internal rigidity that the granules do not swell sufficiently to give a viscous paste when cooked at normal concentrations. There are various viscometers available for measuring the apparent viscosities of non-Newtonian fluids such as starch or flour pastes. Smith (1964) reported that the most commonly used visco­ meters are: 1) the Brabender Visco-Amylograph, 2) Corn Industry University of Ghana http://ugspace.ug.edu.gh 11 Viscometer, 3) Scott Viscometer and 4) the Brookfield Viscometer. The choice of equipment naturally will depend on the type of infor­ mation required, and also the availability of the equipment. For instance, the Brabender Visco-Amylograph allows for a continuous determination of viscosity while cooking and cooling of the starch paste. However, the Brookfield Viscometer is a simpler instrument. It is equipped with several spindles of different sizes and can be operated at a range of speeds. Since starch pastes are non-Newtonian in character, these features on the Brookfield Viscometer allow for determination of viscosity using several spindles and speeds to determine shear rate dependence if necessary (Smith, 1964) . Smith (1964) recommended that for comparisons involving a single observa­ tion the same spindle and speed should be used. The procedure for preparing pastes when using the Brookfield Viscometer was discussed by Smith (1964). In his method, the starch is mixed with water in a beaker and placed in a boiling water or steam bath and paste cooked while stirring constantly in a prescribed manner. Smith (1964) warned that in routine sample evaluation certain aspects of the methodology should be kept constant from sample to sample. These were, the type and vigour of stirring, the rate of rise in temperature and cooking time. Rasper et al. (1974) measured the pasting characteristics of starches and flours at four different concentrations and different temperatures. The concentrations used for the flours were 25, 30, 35 and 40 gm per 480 mis water which approximately correspond to 5.2, 6.3, 7.3 and 8.3 per cent respectively. Using the Brabender Visco-Amylograph, Rasper et̂ . (1974) measured the maximum viscosity on heating, the viscosity at 95°C held for 11 minutes and for one hour. University of Ghana http://ugspace.ug.edu.gh 12 Also they measured viscosity on cooling of paste to 50 C and holding for one hour at this temperature. Tolmasquim et_ al. (1971), working on the properties of starches from different varieties of cowpea starch, also used two temperatures of 95°C and 50°C for the determi­ nation of viscosity characteristics. It is a standard procedure to hold pastes for ten minutes at respective temperatures for viscosity determinations (Knight, 1964; Rasper £t̂ al., 1974; Tolmasquim et_ al., 1971). Pastes could be held for one hour at a specific temperature to determine the stability of the paste at that particular temperature (Rasper et_ al., 1974; Tolmasquim e_t al., 1971) . Rasper et_ al. (1974) observed that wheat flour viscosity increased with decreasing temperature and also upon holding for one hour at 95°C and 50°C. Sorghum flour also showed a similar trend as described for wheat (Rasper et_ al., 1974). Leach (1967) reported that waxy sorghum starch showed a tendency to break down i.e. paste viscosity dropped when held for one hour at 95°C. Millet flour eval­ uated by Rasper e_t al. (1974) was found to show increased viscosity on cooling; however on holding millet pastes for one hour at 95°C and 50°C decreases in viscosities were observed. They found that both sorghum and millet starches had higher viscosities than wheat, with sorghum being the most viscous paste. Rasper et_ al. (1974) found that starches and flour from the same source did not necessarily follow the same viscosity patterns. He concluded that non-starch components such as protein may have a greater effect on baking performance than the functional properties of the starch. Tolmasquim et_ al_. (1971) worked on the functional properties of University of Ghana http://ugspace.ug.edu.gh 13 five different varieties of cowpea starches. Using the Brabender Amylograph, they reported that the paste viscosity was very similar in all varieties. It was also revealed that at 95°C the paste vis­ cosity was practically unchanged after keeping it at that temperature for one hour and stirring continuously. This indicated a great resistance of the granules against mechanical disintegration. There was an increase in viscosity when starch pastes were cooled to 50°C and this reflected a high retrogradation tendency (Tolmasquim et_ a1 ., 1971). However, at 50°C, cowpea starch pastes did not change much in viscosity when held for one hour. Gel Strength and Retrogradation The term "retrogradation" describes the process of dissolved starch reverting to a water insoluble form (Watson, 1964). For gels, Paul and Palmer (1972) define retrogradation as a process whereby a starch gel shrinks and squeezes out water coupled with changes in its properties such as the stiffening of the gel and/or skin forma­ tion on the surface. Whistler (1954) found that X-rays could be used to show that during retrogradation amylose molecules rapidly associate and form a precipitate (in dilute solutions). This association is so rapid under general conditions that the precipitate is a mixture of crystalline and amorphous regions instead of completely well formed crystals and thus Whistler (1954) concluded that retrogradation is therefore the result of an attempt toward crystallization on the part of the mole­ cules. At higher concentrations the retrogradation phenomenon takes a different form. Collison (1968) explained that starch granules University of Ghana http://ugspace.ug.edu.gh 14 swell irreversibly on heating above the gelatinization temperature and that some of the starch molecules become partly or completely detached from the granules. On cooling, these molecules form hydro­ gen bonds with similar molecules from adjacent granules and in this way a continuous three-dimensional network of swollen granules is formed (Collison, 1968). This is synonymous with the process of gel formation. Both amylose and amylopectin molecules become involved in the crystalline micelles and that the crystalline areas both within the swollen granules and especially in the aqueous solution between the granules have a significant contribution to the strength and rigidity of the gel which is formed (Osman, 1967). Retrogradation of starch is primarily due to the amylose component. Foster (1965) explained that the amylopectin part of the starch molecule forms a reasonably stable solution in water but amylose is not truly soluble in water. In relation to this, Watson (1964) said that retrogradation of starch is related to the concentration of amylose and the presence of amylopectin which retards amylose deposition. Some other factors affecting retrogradation include the molecular size of starch, temperature, pH, non-starch components in the medium and the concentration of starch. Whistler and Johnson (1948) reported that the motion of the larger molecules is sluggish because of their length and therefore they associate less readily than molecules of medium length. Collison (1968) reported that temperature has a great effect on retrogradation. He remarked that at very low temperatures the movement of starch molecules is so slow that there is hardly any molecular association. Kalb and Sterling (1961) reported that as a University of Ghana http://ugspace.ug.edu.gh 15 rule, retrogradation of starch paste is inhibited to some extent at temperatures in the vicinity of 60°C and tends to be accelerated with decreasing temperature. Also Collison (1968) reported that at very high temperatures, there is little association of macro-mole- cules due to the disordering effect of Brownian motion. The rate of retrogradation increases with increasing starch concentration because increasing concentration tends to limit swelling and result in structural disorientation. Among all these factors influencing the rate of retrogradation, Paul and Palmer (1972) stressed that the most important are tempera­ ture and the size and shape of the starch molecules. Most of the factors that affect retrogradation also apply in considering the characteristics of starch gels. Osman (1967) said that the age and previous treatment of the starch affects gel characteristics. The preparation of gels such as cooking time, temperature and agitation during cooking are the other major factors affecting the characteris­ tics of starch gels (Kesler and Bechtel, 1954). Alkaline solutions have the ability to dissolve starch and have been found to inhibit retrogradation. Small concentrations of acids such as sulphuric acid or hydrochloric acid have been found to increase the rate of retrogradation (Collison, 1968). Acid treat­ ment decreases chain length and it has been observed that gel strength of the retrograded pastes of acid treated starch increases and then decreases as the concentration of acid increases. The strength of a gel is often confused with its rigidity although these are not identical (Hjermstad, 1964). Rigidity is rela­ ted to the elastic character of gels. For instance, a Ridgelimeter can University of Ghana http://ugspace.ug.edu.gh 16 be used to measure the rigidity of a gel as the percentage of it. The strength is an ultimate property in the sense that the structures must be destroyed in order to measure its strength. The values ob­ tained from a measurement of gel strength are affected by the size of the sample and the shape of the device which transmits the break­ ing stress to the sample (Hjermstad, 1964). In the food industry the strength of gels has been measured by different devices. Some devices measure the resistance of a gel to penetration e.g. the Bloom Gelometer and the Penetrometer. Osman (1967) said that sorghum starch forms a viscous paste readily and that the resulting paste is "short" and sets to a stiff gel. She commented that sorghum paste has a pronounced tendency to retrograde especially when frozen and subsequently thawed. Sorghum starch has similar characteristics to corn starch and therefore is used inter­ changeably in the food industry (Osman, 1967; Paul and Palmer, 1972). Collison (1968) reported that various researchers have shown that normal cereal starches retrograde more quickly than tuber starches. In addition he noted that cereal starches as a group have a tendency to swell readily on heating but that the difference in retrogradation tendency might be due to the cereal starches being less dispersed in solution than other starches. In some food systems, retrogradation could be reversed by heating (Osman, 1967). However, Osman (1967) commented that this has not always been the case because even though retrograded amylopectin could be returned to its original dispersed state by merely heating to 50 to 60 C, retrograded amylose cannot be reversed even by auto- claving. University of Ghana http://ugspace.ug.edu.gh 17 In this regard it was not surprising that Osman and Cummisford (1959) found that in food systems, changes associated with retrograded amylopectin have not been as reversible as in simple starch water systems. This is of special interest because it is more common to find flour or starch mixed with other ingredients in food products. Whistler (1953) measured the rate of retrogradation of various common starches including wheat starch. At 2% aqueous dispersion retrogradation of wheat starch was found to be about 35 and 40% during a five and ten day storage period respectively at 0.2°C. Cluskey et_ al. (1958) found that gluten increased in rigidity rela­ tively slowly and therefore a high content of gluten in flour resul­ ted in slower rates of increase in rigidity of bread during storage. Also they found that their research substantiated the general belief that the firming of bread (an aspect of retrogradation) was mainly due to the starch component. Collison (1968) stated that retrogra­ dation of wheat starch suspension takes place most rapidly at -4°C. Relation of Functional Properties to Practical Application Sandstedt (1961) reported that gelatinization plays an impor­ tant role in the early stages of baking and has an effect on the flexibility of cell walls as well as on the susceptibility of starch granule to amylolytic degradation. Degree of swelling and solubility characterize individual starches. Leach et al. (1959) said that the pattern of swelling gives a paste its particular elastic character which is responsible for most of the viscous qualities of a cooked starch paste. Smith (1964) reported that viscosity characteristics of a University of Ghana http://ugspace.ug.edu.gh 18 product are important because they indicate the utility of the pro­ duct in specific applications and also reflect the properties encountered by the user during preparations of products. Minard (1954) outlined three reasons why viscosity is measured. First, he stated that viscosity is a direct measure of a fluid's property in practical application e.g. dipping or coating. Second, that a change in viscosity can indicate a fundamental change in the material under question. For example, viscosity changes can show the extent of a starch cooking process. The third reason suggested by Minard (1954) was that viscosity could often be a very sensitive method of indirect measurement of another property such as solid content. The viscosity of a fluid is closely connected with the fluid's rheological properties in food (Paul and Palmer, 1972). Retrogradation can be evident in food systems in many forms. The skinning of pastes and setback of starch gels have been associated with retrogradation phenomena (Schoch, 1941). Osman (1967) said that although all the changes which occur during staling of bread cannot wholly be explained by retrogradation of the starch, raost research has shown that this is the most important single factor. Increased firmness, crumbliness and opacity of bread crumb on aging are consis­ tent with changes which would be expected from starch retrogradation (Osman, 1967). In the use of flours and starches as thickening and emulsifying agents in frozen foods, e.g. desserts, salad dressing, those starches or flours which show less degree of retrogradation at freezing temperature are of definite advantage (Paul and Palmer, 1972). Gels range in consistency from raw egg white to stiffer gels such as pectin gels (Paul and Palmer, 1972). The properties of flour University of Ghana http://ugspace.ug.edu.gh 19 or starch gels are important in many foods. Some starch gels are quick setting, others are slow setting. Sorghum starch gel is an example of a quick setting type and has been used in the food industry when this property is required (Paul and Palmer, 1972) . The protein quality of these cereals and legumes are very important especially in places where these foods form a staple in both adult and children diets. Cereals are usually deficient in lysine and this applies tô iheat, millet and sorghum (F.A.O., 1970). Legumes, on the other hand, tend to be deficient in the sulfur- containing amino acids, i.e. methionine and cystine. The amino-acid composition of a protein may be used to deter­ mine its quality. Recently a new scoring pattern called the "amino acid score" has been suggested as a way of predicting protein quality (F.A.O., 1973). The amino-acid score of a protein is calculated using the most limiting amino-acid against a suggested level of that particular amino-acid. The formula used is reported (F.A.O., 1973). Amino Acid Score = of amino acid/gm of test protein Mg of ammo acxd xn reference pattern This score is taken merely as an initial approximation to the probable efficiency of utilization of the test protein. The essential amino acid contents of similar cereals and legumes to be tested in this research have been published by F.A.O. (1970). The amino acid scoring pattern with suggested reference levels of the essential amino acids (mg per gm of protein and mg per gm of nitrogen) has been reported in the F.A.O. report of 1973. University of Ghana http://ugspace.ug.edu.gh EXPERIMENTAL METHODS Samples of raw and dried sorghum (Sorghum vulgare "932111") , millet (Pennisetum typhoides "Staph & Hubb"), cowpea (Vigna unguicu- late "Caroni") and bambara nut (Voandzeia subterrana) , were obtained from the Department of Crop Science of the University of Ghana. High quality, Hard Red Spring wheat (Triticum aestivum L., "Manitou") was obtained from the Department of Crop Science, University of Guelph. The legumes and cereals were either made into flour or starches extracted for the appropriate tests. Milling Two days before milling, the grains were washed and allowed to dry at room temperature on paper towels. About 4 Kgs of grain was then put through the mill (Wiley Mill, Model 3). The first run was done using the coarsest mesh of 2 mm and collected into a large bowl. The second run used a 1 mm mesh and the final run a % mm mesh. The resulting flour was very smooth to the touch resembling commercial flour. Flours were stored in sealed jars at room temperature. Starch Preparation A procedure outlined by Tolmasquim et_ _al_. (1971) was adapted. The grain was steeped for 24 hours in distilled water to which half had 0.3 per cent sodium bisulfite added. After the steeping, the steeping water was drawn off and discarded. A sufficient amount of distilled water was added to the soaked product and blended for three minutes to obtain a pulp. The pulp thus obtained was passed 20 University of Ghana http://ugspace.ug.edu.gh 21 through a 60-mesh screen and the residue reground with additional water in a similar manner. This regrinding was done three times. All the wash liquids collected were put together and allowed to stand at room temperature for 24 hours. The starch was treated with 0.15% sodium hydroxide with a quantity equivalent to one-third the volume of the suspension. When the starch had settled the super­ natant was poured off and the sediment suspended in large amounts of distilled water. This suspension was then passed over a 325-mesh screen which allowed only starch and water and finer particles i.e. fat to penetrate. The collected liquid was allowed to settle and the starch treated with 0.1% sodium hydroxide (one-third by volume of water). Again, the starch was allowed to settle and then washed with distilled water until liquid became neutral (pH 7). The starch suspension was defatted in excess methanol and filtered using a grade 515 E + D filter paper. The sediment thus collected on the filter paper was allowed to dry in air. Basic Chemical Analysis Representative samples were taken from the flours for basic chemical analysis. The analysis was carried out at the Department of Nutrition, University of Guelph. The following analyses were conducted on each flour: moisture, protein, fiber, ash and fat con­ tents using AOAC methods (1960). Moisture content was determined by the vacuum oven method (AOAC, 1960, 13.003); ash content by the muffle furnace method (AOAC, 1960, 13.006); and fat content using the Soxhlet thimble apparatus. The protein content was obtained using the improved Kjeldahl method and the fiber content was deter­ University of Ghana http://ugspace.ug.edu.gh mined using AOAC, 1960, 13.061. Determination of Gelatinization Temperature and Range of Starches The gelatinization temperature and range were determined by a method used by MacMasters (1964). A starch suspension, 0.5% (by weight), was made with distilled water in a clean test tube. The test tube was then placed in a water bath at room temperature. A thermometer was hung in the starch suspension so that it touched neither the walls nor the bottom of the test tube. MacMasters (1964) recommended that the rate of increase in temperature be made at 0.5°C each three minutes. The Fisher Versa-Bath used for this research did not have a control device needed for such a precise rate of heating. The slowest rate of heating achieved was 1°C for each two to three minutes. The medium was stirred with the suspended thermometer to keep the starch suspension uniform throughout. When the desired temperature had been reached a drop of the starch sus­ pension was removed using a clean pipette made of glass tubing and drawn to a tip. The starch suspension was dropped on a slide with two drops of congo-red dye on it. The dye and starch were thoroughly yet gently mixed and a coverslip was then placed on the slide. The excess dye which was left around the coverslip was drawn off with absorbent paper. MacMasters (1964) recommended that gelatinization temperature and range be determined, starting at 10°C below and con­ ducted 10 C beyond reported literature values. It was considered adequate in this research to determine the first replicate at 2°C intervals. The second and third replicates were done at 1°C intervals starting at 5 C below and ending 5°C above values obtained from the 22University of Ghana http://ugspace.ug.edu.gh 23 first replicate to obtain precise gelatinization range. To decide when gelatinization had begun, one or two isolated stained granules were ignored until the point where a few stained granules were observed evenly distributed in the field. The end of gelatinization was taken to be the point where all granules were very swollen and clearly stained in all fields observed under the microscope. The gelatinization temperature was recorded as the temperature where all granules were fully swollen and stained and the gelatinization range was recorded as the temperature of begin­ ning of gelatinization to the end. A compound microscope "Vickers Microplan" was used to view the starch granules at 320X magnification. Determination of the Swelling and Solubility of Starches The procedure followed was based on the method described by Leach et̂ al_. (1959) . During the preliminary studies it was found necessary to modify this method slightly. Two grams (dry matter) of starch were suspended in 100 mis of distilled water in a 250 ml centrifuge bottle. The bottle was placed in a bath held at each temperature of determination. The starch suspension was mechanically stirred at 375 rpm for 30 minutes. After 30 minutes the bottle was cooled under running tap water for 20 minutes and then 20 mis of distilled water used to wash the stirrers into the centrifuge bottle. The bottle was then dried and weighed and centrifuged at 2000 rpm for 15 minutes. The clear supernatant was siphoned off as quantita­ tively as possible. An aliquot of 20 mis was taken from the super­ natant and pipetted into a dried and x̂ eighcd moisture dish. This aliquot was first evaporated to dryness over a steam bath and subse­ University of Ghana http://ugspace.ug.edu.gh 24 quently dried in a vacuum oven at 120 C for four hours. The sediment left in the centrifuge bottle was weighed for the determination of swelling power. The dried supernatant in the moisture dish was cooled in a dessicator and weighed for the determination of solubles. All determinations were carried out at five degree intervals from 50 to 95°C. Three replicates were determined and treatments (tempera­ tures) were randomly assigned. Calculations The formulae used for the determination of swelling power and solubility were obtained from Schoch (1964). Per cent solubles _ weight of soluble starch x weight of total water (on dry basis) weight of sample (dry basis) x volume of aliquot Per cent swelling _ _________ weight of sedimented paste__________ power weight of sample on dry basis x (100 - per cent (corrected) solubles on dry basis) Viscosity Characteristics of Flours Viscosity characteristics were derived from tests conducted on flours from the grains. Three concentrations of 30, 35 and 40 gm flour (dry matter) per 620 mis distilled water were chosen for the determinations thus providing a range from thin soup-like to fairly thick porridge consistencies. Temperatures for measurement of visco­ sity were selected on the basis of their practical application. A maximum temperature of 95°C (representing a normal cooking temperature for starch-based products), an intermediate temperature of 50°C (consi dered an average serving temperature) and 4°C (normal refrigeration 100 100 University of Ghana http://ugspace.ug.edu.gh 25 temperature) were used as the other variables in this test. As a standard procedure two holding times were considered i.e. ten minutes and one hour. A weighed sample of flour was put into a 1000 ml beaker and 620 mis of distilled water slowly added to form a smooth paste. The beaker with the mixture was then immersed in a 100 C water bath for cooking. The flour paste was gently stirred in a prescribed manner (about five revolutions each minute). When the temperature of the paste was 75°C (about 15 minutes), the beaker was covered with aluminium foil to prevent steam from escaping the beakers and also to prevent condensed steam on the cover of the bath from entering the paste. The water bath was covered and from then on the paste was stirred (in the same manner) at 15 minute intervals until paste temperature was 95°C. Flours being tested at 95°C were held at this temperature for ten minutes and for one hour. For the 50°C readings the hot pastes were cooled to this temperature and held for either ten minutes or one hour. The other temperature used for viscosity measurement was 4°C and this tempera­ ture was achieved by putting hot paste into a container with crushed ice and salt mixture. When the paste attained 4°C it was treated in similar fashion as for the other temperatures. After 95°C had been attained during the initial cooking period, stirring was carried out during cooling at five minute intervals in the same manner as already described for all temperatures tested. Determination of viscosity was done using the Brookfield Visco­ meter (Model RVF). Spindle No. 2 was used for pastes except sorghum University of Ghana http://ugspace.ug.edu.gh 26 pastes cooled to 50°C and 4°C. Sorghum pastes at these temperatures were too thick to give readings using spindle No. 2, and therefore spindle No. 3 was used to test the paste at 50°C and No. 4 used at 4°C. The rpm was maintained at 20 for all viscosity determinations. Before determination, the appropriate spindle was placed in water that had the same temperature as that of the paste being measured. When the time was up to measure the paste viscosity, the spindle was quickly taken from the water, dried and fitted into the instrument. The beaker was then placed in position and readings taken after two revolutions of the dial on the viscometer. Viscosity determinations were made by random selection of flour concentrations and times (holding for ten minutes or one hour). However, due to limitations of equipment, temperatures were not completely randomized but on any day two temperatures were handled. Three replicates were carried out for each variable tested. Gel Strength and Retrogradation of Flours A method was designed to use the Precision Penetrometer to measure the gel strength of flours. Retrogradation was measured in the form of syneresis of flour gels. A concentration of 12 per cent flour (dry matter) in water was selected. In order to cut down on variability of paste, each replicate of a flour was prepared at one time i.e. enough hot paste was made to partially fill five custard cups (5 oz or approximately 150 mis). Three of these cups had been previously weighed for retrogradation determination. The paste was prepared according to the following description. A weighed sample of flour (enough for one replicate) University of Ghana http://ugspace.ug.edu.gh 27 was put into a beaker and mixed with half the required amount of cold distilled water. The other half of the water was brought to a boil. When the water had boiled it was slowly poured back into the cold flour paste while stirring continuously. The mixture was poured back into the top section of a double boiler and cooked over boiling water. Stirring was continuously carried out until the paste temperature was 85°C. After which the flour paste was stirred every 30 seconds until the temperature was 95°C. The hot paste was poured out quickly to fill all the prepared five cups and covered with Saran wrap and set in dishes (to collect any water that oozed out of cooled pastes). After three hours of cooling at room tempera­ ture, pastes (three cups) to be used for retrogradation determina­ tions were weighed and put in the refrigerator. One of the cups for gel strength determination after 24 hours refrigeration was also placed in the refrigerator and the other held at room temperature. Retrogradation was measured at intervals of 24, 48 and 72 hours of refrigeration. The liquid from the refrigerated paste was collected by inverting gel on a wire mesh over a funnel. The liquid was collected into a previously weighed cylinder and reweighed after one hour drainage. Retrogradation was determined as per cent syneresis. Per cent syneresis = Wjight of collected liquid weight of paste The gel strength of pastes was determined using the Precision Pene­ trometer after five hours of cooling the hot paste at room tempera­ ture and after 24 hours of refrigeration. Readings were taken after dropping the plunger for 30 seconds. University of Ghana http://ugspace.ug.edu.gh RESULTS AND DISCUSSION Basic Chemical Analysis The results of the basic chemical analysis are tabulated in Table 1. The legumes, cowpea and bambara nut were found to contain relatively high amounts of protein. Millet and sorghum were found to be low protein cereals compared to wheat. For purposes of com­ parison, a selection of similar products analysed by Watson (1971) and by Watt and Merrill (1968) were compiled in Table 2. Major differences between results obtained in this study (Table 1) and those shown in Table 2 included moisture and fiber contents. Watson (1971) commented that some of the products tested had certain frac­ tions removed, however he did not clearly show when this was the case. The big differences in fiber contents among the Ghanaian cereals and legumes studied and those tested by Watson (1971) did not use 100% flours as used in this study. Also varieties used by Watson (1971) may not necessarily be the same as those used for this research. The differences in moisture contents of the cereals and legumes tested and those compiled in Table 2 indicated different degrees of drying of the raw grain. Gelatinization Temperature and Range of Starches The gelatinization temperature was determined using the defin­ ition of MacMasters (1964) as the temperature at which all granules were swollen and stained. The gelatinization range was defined as the initial signs of gelatinization (when granules begun to take up dye) to when gelatinization was completed as viewed under the compound 28 University of Ghana http://ugspace.ug.edu.gh 29 BASIC CHEMICAL ANALYSIS OF FIVE FLOURS USED IN THIS STUDY TABLE 1 Product Per cent moisture Per cent ash Per cent fiber Per cent protein* Per cent fat Wheat 10.81 1.42 3.01 13.4 1.06 Sorghum 14.25 1.30 10.36 7.1 2.41 Millet 12.61 1.32 6.97 7.2 4.67 Cowpea 10.15 3.76 6.50 24.3 1.05 Bambara nut 9.30 3.38 7.12 18.2 6.54 * Wheat: nitrogen conversion factor =5.83 Sorghum Millet -v , , „rCowpea \ nltroSen conversion factor =6.25 Bambara nut University of Ghana http://ugspace.ug.edu.gh 30 BASIC CHEMICAL ANALYSIS OF FIVE FLOURS AS REPORTED BY WATT AND MERRILL (1963) AND WATSON (1971) TABLE 2 Product Per cent moisture Per cent ash Per cent fiber Per cent protein Per cent fat Wheat (whole)a 12.0 1.7 - 13.3 2.0 Sorghum (brown)^ 9.9 1.4 1.4 9.9 3.1 MilletC 18.5 1.6 1.2 6.3 4.0 Cowpea^ 10.8 3.0 5.0 19.5 1.1 Bambara nut8 11.4 3.5 5.3 19.7 5.6 Published by Watt and Merrill (1963) b—ePublished by Watson (1971) University of Ghana http://ugspace.ug.edu.gh 31 microscope (320X). Results obtained for the five starches are tabulated in Table 3. The gelatinization range values obtained for the wheat starch fell into a similar range as reported by Rasper et_ al. (1974) i.e. 53 to 62°C. For sorghum the values obtained for initial gelatini­ zation temperature were lower than those reported in the literature. Leach (1968) reported the gelatinization range of sorghum as 68.5 to 75°C and Rasper £t aL. (1974) as 69.5 to 74°C. Even though varieties do make a difference in starch characteristics, there could be other reasons for the lower value obtained for the initial temperature such as differences in methodology of starch extraction and determination of gelatinization characteristics. Rasper et_ al. (1974) used the Kofler hot stage microscope to determine gelatinization whereas staining technique was used in this study. A lower gelatinization temperature was also noted for millet starch which had an initial gelatinization temperature ten degrees lower than that recorded by Rasper et_al. (1974). This could be a result of different milling methods or varietal differences. The variety of cowpea starch "Caroni" tested here fell into a range recorded for five varieties. The cowpea starch tested ("Caroni") was found to have a gelatinization range similar to the "Early Spring" variety of cowpea tested by Tolmasquim. et_ al. (1971). Bambara nut starch was found to have the highest gelatini­ zation range among all the test starches but as already mentioned no published data have been found concerning this characteristic of bambara nuts. University of Ghana http://ugspace.ug.edu.gh 32 TABLE 3 GELATINIZATION CHARACTERISTICS OF FIVE STARCHES Starch Gelatinization range °C Gelatinization temperature °C Wheat 51-61 61 Sorghum 65-73 73 Millet 60-72 72 Cowpea 65-75 75 Bambara nut 68-81 81 University of Ghana http://ugspace.ug.edu.gh 33 All starches examined had higher gelatinization ranges than wheat starch. The importance of this as discussed by Rasper e_t a_l. (1974) is that when part of wheat flour is replaced by such flours or starches, the gelatinization process is retarded by increasing the gelatinization temperature thus affecting baking performance of the dough. Rasper t̂_ al_. (1974) found that when flours instead of starch from non-wheat sources were mixed with wheat flour for bread making, better performance was obtained from flour: flour mixtures than flour: starch mixtures. The conclusion drawn, then, was that flour contributed more positively to bread making performance than pure starches and therefore the fact that gelatinization temperature of the test starches differed from wheat starch should not rule out the potential of the flours as being more compatible with wheat in baking or cooking. Swelling and Solubility Patterns of Starches The standard deviations of swelling power values ranged from 0.01 to 0.52 and for solubility 0.00 to 0.25 indicating good precision of results obtained. The averaged results of three replicates were used to construct graphs to show the effect of temperature on these variables and also the interrelationship between them (Figures 1-10). Figures 11 and 12 were drawn to compare the trends of the various starches with each other. Wheat starch showed higher swelling power than solubility at all temperatures (Figure 1). Swelling of wheat starch began after 50°C and a marked increase in the solubility value was noted at 60°C University of Ghana http://ugspace.ug.edu.gh SW EL LI NG PO W ER 34 TEMPERATURE °C Figure 1. Swelling power and solubility of wheat starch at different temperatures o f heating. Figure 2. Relationship between swelling power and solubility of wheat starch. % S O LU B IL IT Y University of Ghana http://ugspace.ug.edu.gh 35 It was also observed that there were sharp increases in both swelling and solubilities up to 75°C and another sharp increase observed from 85 to 95°C. These trends were very similar to trends observed by Sandstedt and Abbott (1964). However, generally, results obtained were lower than those reported for the differences in absolute values obtained by Sandstedt and Abbott (1964) and the experimental values obtained in this study. The first major difference is that of methodology. Sandstedt and Abbott (1964) pipetted aliquots for solubility determination at 2°C and therefore obtained a higher con­ centration of solubles. Secondly, a very high starch concentration (50%) was used by Sandstedt and Abbott (1964) whereas the starch concentration used in this research was 1.5%. Comparing swelling power values, Collison (1968) quoted the swelling power of wheat starch as 13.5 at 95°C. This value compares favourably with the value obtained in the experiment i.e. 12.7 at 95°C (Figure 1). The graph of solubility against swelling power (Figure 2) showed a direct relationship between these two variables indicating that both swel­ ling and solubility took place at the same rate. The trends of the swelling and solubility curves of sorghum were very similar to those obtained for wheat starch i.e. two stages of more pronounced increases in these values were observed (Figure 3). Leach (1959) observed a two-stage swelling and solubility of white milo at 65°C and 90°C. The swelling power and per cent solubilities observed in this research were lower than the variety (white milo) tested by Leach (1959). Millet starch, like the other cereal starches examined, exhi­ bited a two-stage swelling pattern (Figure 5). This behaviour is University of Ghana http://ugspace.ug.edu.gh S W E LL IN G P O W E R 36 % S W E L L 1 N G P O W E R Figure 3 . Sw elling p ow er and so lubility o f sorghum starch at different temperatures o f heating. S W E L L IN G P O W E R Figure 4 . Relationship between swelling p ow er and so lub ility o f sorghum starch. ^S O L U B IL IT Y University of Ghana http://ugspace.ug.edu.gh S W E LL IN G P O W E R 37 T E M P E R A T U R E ° C Figure 5. Swelling pow er and solubility of m illet starch at different temperatures. S W E L L IN G P O W E R Figure 6. Relationship between swelling power and solubility of m illet starch. ^S OL UB IL IT Y University of Ghana http://ugspace.ug.edu.gh 38 attributed to two sets of bonding forces which relax at different temperature levels (Leach, 1959). Although absolute values varied with the cereal starches till the starches exhibited similar patterns for solubility and swelling power. This can be related to Collison s (1968) observation that cereal starches do not undergo complete mole­ cular dispersion in hot water. The rather low solubilities of these cereal starches could also be related to their low amylose content, for example, wheat contains 30 per cent amylose. Both sorghum and millet starches swelled more than wheat but wheat had more solubles. The swelling and solubility patterns observed for cowpea and bambara nut starch are graphically represented in Figures 7 to 10 respectively. The curves obtained for cowpea were found to be very similar in shape to those obtained by Tolmasquim et al̂. (1971) for other varieties of cowpea. However, the variety of cowpea used in this research ("Caroni") seemed to have slightly more solubles than any of the five varieties of cowpeas tested by Tolmasquim £t̂ al. (1971). In the study by Tolmasquim et_ al. (1971), none of the five varieties behaved exactly the same at all temperatures. However, values were close to each other. The results obtained for cowpea starch "Caroni" were also close to the values reported for the varieties of cowpeas tested by Tolmasquim £t al. (1971) . In conclu­ sion, then, it could be postulated that differences in swelling and solubilities of these starches are mainly due to varietal differences. Bambara nut starch did not begin swelling or become soluble before the temperature of 75°C. The interesting phenomenon observed with bambara nut starch was that 75 C the per cent solubility was University of Ghana http://ugspace.ug.edu.gh 39 T E M P E R A T U R E ° C Figure 7. Swelling pow er and solubility o f co w pea starch at different temperatures o f heating. S W E L L IN G P O W E R Figure 8. Relationship between solubility and swelling pow er of co w pea starch. % S O L U B IL IT Y University of Ghana http://ugspace.ug.edu.gh % S O L U B IL IT Y S W E L L IN G P O W E R 40 T E M P E R A T U R E C Figure 9 . Sw elling p ow er and so lub ility o f bam bara starch at d ifferent tem peratures o f heating. S W E L L I N G P O W E R Figure 10. R elationship between so lub ility and sw elling p o w er of bam bara starch. % S O L U B IL IT Y University of Ghana http://ugspace.ug.edu.gh 41 TEMPERATURE °C Figure 11. Swelling patterns of five starches. University of Ghana http://ugspace.ug.edu.gh % S O L U B IL IT Y 42 T E M P E R A T U R E ° C Figure 12. Solubilization patterns of five starches University of Ghana http://ugspace.ug.edu.gh A3 higher than swelling power. This was not observed for the other starches tested. A similar situation was observed by Schoch and Mayland (1968) in wrinkled pea starch (legume) which was the first high-amylose starch to be discovered. Figure 10 indicates that solubility of bambara nut proceeded at a higher rate than swelling. It seemed from the high solubility of bambara nut starch that this could be a high-amylose starch. As already mentioned, legume starches tend to have higher amylose content (Deatherage et al., 1955). Bambara nut starch, even though it exhibited higher solubility, did not show restricted swelling as evidenced by the graph in Figure 9. Between the two legume starches examined, cowpea showed higher swelling power with bambara nut showing a higher solubility. The results obtained showed that both starches had similar swelling and solubility trends (i.e. a continuous progressive swelling). Also these legume starches did not show significant swelling or solubility at lower temperatures. This trend was found to differ from the cereal starches tested which exhibited pronounced swelling and solubility at lower temperatures. Also, the two-stage swelling and solubility pattern found in the cereal starches was not evident among the legume starches. In spite of the high initial gelatinization temperature of the legume starches, the results obtained in this research showed that for both swelling power and solubility, the legume starches showed significantly higher values than the cereal starches. Since the third objective of this project was to compare the functional properties of these flours (millet, sorghum, cowpea and University of Ghana http://ugspace.ug.edu.gh 44 bambara nut) to wheat starch, graphs were made of the swelling power and solubility curves of all starches as a group (Figures 11 and 12). It became evident from these graphs that the cereal starches had swelling and solubility patterns more similar to wheat starch than the legumes. By comparison, it could be seen that among the cereals, wheat starch and sorghum were more similar in their swelling and solubility patterns, thus confirming Leach et al.'s (1959) results. Viscosity Results obtained for viscosity measurements are graphically presented in Figures 13 to 17- Standard deviations of replicates calculated ranged from 0.00 to 40.47. The standard deviations were generally higher for more viscous pastes indicating that there was more variability in viscosity with increasing concentration of flour paste. Curves obtained for wheat flour (Figure 13) showed that vis­ cosity increased with increasing flour concentration. Cooling of hot pastes also led to increased viscosity. On holding pastes for one hour at the test temperatures, wheat flour showed slight increa­ ses in viscosity. These trends observed for wheat flour agreed with trends observed by Rasper et_ sd. (1974) with straight run flour (all purpose). Sorghum flour pastes followed similar viscosity patterns (Figure 14) as reported for wheat flour pastes. However, sorghum had very viscous pastes at every concentration and temperature tested (Figure 14). As expected, the viscosity of millet pastes increased with University of Ghana http://ugspace.ug.edu.gh 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 45 A • 9 5 ° C (A fte r holding 10 m in .) B O 9 5 0 C (A fte r holding 1 hr. ) C ■ 5 0 ° C (A fte r holding 10 m in .) D □ 5 0 ° C (A fte r holding 1 hr. ) E ▲ 4 o c (A fte r holding 10 m in .) F A 4 O C (A fte r holding 1 h r. ) 30 3 5 40 C O N C E N T R A T I O N jgrams (d ry solids) in 620 m lsj Figure 13. Effect o f concentration on the viscosity o f w heat flour at three different temperatures. University of Ghana http://ugspace.ug.edu.gh V IS C O S IT Y (C E N T IP O IS E ) 46 Figure 14. Effect o f concentration on viscosity o f sorghum flo u r at three d ifferent temperatures. University of Ghana http://ugspace.ug.edu.gh 47 increasing concentration and with decreasing temperature of hot pastes. However unlike wheat or sorghum, millet flour pastes dropped in viscosity on holding for one hour at the respective temperatures (Figure 15). Rasper et_ al. (1974) also observed decreases in visco­ sities when millet flour pastes were held for one hour at 95 and 50°C. All the cereal flours therefore showed an instability of pastes at 95, 50 and 4°C due to that fact that there were substantial changes in viscosities when held at these temperatures for one hour (Tolmasquim et_ al. , 1971; Rasper £t_ jd. , 1974). As expected, cowpea flour pastes increased in viscosity with increasing concentration and decreasing temperature of holding (Figure 16). However, the change in viscosity between 95 and 50°C was found to be relatively smaller in cowpea pastes than wheat and sorghum pastes. Holding cowpea pastes for one hour at the various temperatures produced very small viscosity increases. This pheno­ menon indicated that the cowpea pastes were very stable at these temperatures especially at 50°C. Tolmasquim et_ al. (1971) observed similar trends with cowpea starch thus showing that other non-starch components had not affected the stability of cowpea pastes. Bambara nut pastes did not show good paste stability on hold­ ing (Figure 17) as indicated by the big increases observed in vis­ cosity patterns. Also bambara nut was found to have a more viscous paste at 95 C (holding for one hour) than cooling to 50°C. A similar phenomenon has been observed with some root starches, cassava and cocoyam (Rasper £t , 1974). This finding for bambara nut could be explained in that the flour paste was not affected by stirring and prolonged heating. Also the starch in the bambara nut flour has University of Ghana http://ugspace.ug.edu.gh V IS C O S IT Y (C E N TI P O IS E ) 48 5 0 0 0 “ 4 5 0 0 - 3 0 0 0 - 2 5 0 0 - 2 0 0 0 - 1 5 0 0 “ 100 0 “ 500 A • 9 5 0 C (A fte r holding 10 m in .) B O 9 5 O C (A fte r holding 1 hr. ) C ■ 50 o c (A fte r holding 10 m in .) D □ 5 0 0 C (A fte r holding 1 hr. ) E ▲ 4 0 c (A fte r holding 10 m in .) F A 4 0 c (A fte r holding 1 hr. ) C O N C E N T R A T I O N [grams (d ry solids) in 620 m is ] Figure 15. Effect o f concentration on the viscosity o f m illet flour at three different temperatures. University of Ghana http://ugspace.ug.edu.gh V IS C O S IT Y (C E N TI P O IS E ) 49 4 5 0 0 - 4 000 3500 3 0 0 0 “ 2500 2000 1 5 0 0 ' 1000 “ 5 0 0 - A • 9 5 0 c (A fte r B O 9 5 0 c (A fte r C ■ 50 O C (A fte r D □ 50 O C (A fte r E A 4 0 c (A fte r F A 4 0 C (A fte r holding 10 m in .) holding 1 h r. ) holding 10 m in .) holding 1 hr. ) holding 10 m in .) holding 1 hr. ) 30 C O N C E N T R A T I O N grams (d ry solids) in 620 mis) Figure 16. Effect of concentration on the viscosity of cowpea flour at three different temperatures. University of Ghana http://ugspace.ug.edu.gh V IS C O S IT Y (C E N TI P O IS E ) 50 5000 4500 4 0 0 0 ' 3 5 0 0 " 3 0 0 0 “ 2500 2000 1 5 0 0 “ 1 00 0 - 500 A • 9 5 0 C (A fte r holding 10 m in .) B O 9 5 ° C (A fte r holding 1 hr. ) C ■ 5 0 O C (A fte r holding 10 m in .) D □ 5 0 O C (A fte r holding 1 h r.) E A 4 0 C (A fte r ho ld ing 10 m in .) F A 4 0 C (A fte r holding 1 h r. ) C O N C E N T R A T I O N [gram s (d ry solids) in 620 mis] Figure 17. Effect o f concentration on viscosity o f bambara flo u r at three different temperatures. University of Ghana http://ugspace.ug.edu.gh 51 a high swelling power (Figure 9) and therefore holding paste for °ne hour at 95°C increased swelling. Among the cereal flours tested, sorghum flour was the most viscous, with wheat being the least viscous. However, on holding pastes for one hour at the test temperature, sorghum flour exhi­ bited more similar viscosity patterns to wheat than to millet. Bambara nut flour gave a more viscous paste than cowpea flour especially for the cooled pastes. Also these two legume flours showed somewhat different viscosity patterns. For example, cowpea hardly increased in viscosity on holding at the various temperatures for one hour, whereas bambara nut pastes showed sig­ nificant increases in viscosities. All flours tested except millet flour showed a type C viscosity pattern as described by Schoch and Mayland (1968) in that continuous cooking and stirring did not cause mechanical fragmenta­ tion as exhibited by the thinning of pastes. Rather, continuous cooking (holding at 95°C for one hour) and continuous stirring (while cooking or cooling for one hour at 95, 50 and 4°C) caused increases in viscosities. Viscosity patterns observed for millet flour could not be explained by any of the categories described by Schoch and Mayland (1968). However, it must be pointed out that these categories were originally designed for pure starches rather than for flours. Non-starch components such as protein present in the flour could have affected the viscosity characteristics tested. Proteins are often present in starch-based products either naturally or as part of the ingredients used in the preparation of the product. University of Ghana http://ugspace.ug.edu.gh 52 Osman (1967) reported that proteins in starch pastes have been found to stabilize food systems and also may retard starch swelling. Flours used in the present research contained varying amounts of proteins. It was observed that cowpea flour having the highest protein content (Table 1) of all the other flours did show stability and also was not as viscous as other flours. However, viscosity patterns of the other cereals and bambara nut flour did not confirm this. Bambara nut flour was not found to be any more stable or less viscous than the other flours which had less protein such as millet or wheat. The more logical comparison in connection with the contribution of non-flour components to this functional property would be between starch and flour pastes of the same source. The conclusion drawn, then, is that the comparative viscosity patterns of the flours tested in the present research did not hold any obvious relationship to their protein contents. Gel Strength and Retrogradation Results obtained for gel strength are represented in Figure 18. Standard deviations of results of three replicates were found to range from 0.05 to 0.64 indicating that the precision of results was good. The graph of gel strength of flour pastes stored for five hours at room temperature (0 hours of refrigeration) and in the refrigerator (Figure 18) indicated clearly that all flour gels tested showed increased firmness upon refrigeration. The strengthening of a gel on storage is also an aspect of retrogradation. The strongest gel at room temperature was the bambara nut gel. University of Ghana http://ugspace.ug.edu.gh P E N E TR A TI O N (M IL L IM E TE R S ) 53 H O U R S O F R E F R IG E R A T IO N Figure 18. Gel strength of five flour gels. *(0 hours refers to five hours standing at room temperature) University of Ghana http://ugspace.ug.edu.gh 54 In order of strongest to weakest for the remaining gels, were wheat, cowpea, sorghum and millet. After 24 hours refrigeration, bambara nut flour gel still remained the strongest gel but cowpea became stronger than the wheat flour gel. It is evident from the latter results that the refrigerated legume gels were stronger than the cereal gels, and among the cereal gels, wheat flour was the strongest at room temperature and after 24 hours refrigeration. The amount of syneresis of flour gels tested is presented graphically in Figure 19. Standard deviations obtained for repli­ cates ranged from 0.04 to 0.62. Within the 72 hours of refrigera­ tion, wheat flour gel did not retrograde to any measurable extent. Whistler (1964) reported that wheat amyloses retrograde rapidly, yet in this experiment, wheat flour gel acted to the contrary. It seems probable that other substances in the flour apart from the starch portion helped to suppress the rate of syneresis parti­ cularly the bran and gluten. Cluskey at al. (1959) reported that gluten increases in rigidity rather slowly and hence could contri­ bute to the apparent lack of retrogradation with the wheat gels tested. Also, there could have been some amount of retrogradation but was too small to show up as syneresis. All other flour gels tested showed some amount of retrogradation with the amount of syner­ esis increasing with prolonged cold storage. Sorghum showed signs of retrogradation on cooling at room temperature even before it was put into the refrigerator. Within the period of testing, sorghum showed the highest rate of retrogradation. Osman (1967) reported that grain sorghum starch formed pastes which set to stiff gels and had a pronounced tendency to retrograde on cooling especially when University of Ghana http://ugspace.ug.edu.gh % S Y N E R E S IS 55 H O U R S O F R E F R IG E R A T IO N Figure 19. A m ount of syneresis of five flour gels. University of Ghana http://ugspace.ug.edu.gh 56 frozen. Millet flour showed the least amount of syneresis at the end of 72 hours of refrigeration except for wheat flour. Cowpea flour and bambara nut flour gels retrograded extensively. At 72 hours refrigeration, it was observed that bambara nut and sorghum gels have almost shown the same amount of syneresis. Even though wheat flour gel did not show any syneresis, it should be noted that refrigeration led to increased rigidity in wheat flour gels and hence retrogradation took place to some extent. It should be pointed out that values obtained for gel strength and retrogradation will be applicable to only similar situations of cooking techniques and measurement. However, trends observed could be of general application. Interrelationships Between Some Functional Properties Tests conducted on starches of the products used for this research were gelatinization temperature and range, and the swelling and solubility patterns. Examination of the initial gelatinization temperatures of all five starches revealed their agreement with their respective swelling and solubility patterns. Cereal starches gelatinized at lower temp­ eratures and also showed increased swelling and solubilities at lower temperatures than the legume starches. Mayland and Schoch (1968) stated that starches which have high swelling and solubility characteristics tend to show a drop in vis­ cosity when the peak of the paste has been reached, probably due to the starch structure being too weak from high absorption of water and from high loss of solubles. In this research it was not possible University of Ghana http://ugspace.ug.edu.gh 57 to determine the maximum viscosity of each flour paste since this would require a device which could determine viscosity continuously during the gelatinization period such as the Brabender Visco-Amylo- graph. In the literature review on viscosity, it was mentioned that an increase in viscosity during gelatinization or cooking of starch-based products was primarily due to the increase in swelling and solubility of the starches (Osman, 1967; Paul and Palmer, 1972). It can be assumed then that there will be a direct relationship between swelling power and viscosity values. In this research the most viscous flour pastes were those made from sorghum even though sorghum starch did not have the highest swelling and solubility values. Another interesting point was that cowpea starch had the highest swelling and high solubility values; however, like sorghum flour, the former also showed an opposite trend in viscosity values in that it was not the most viscous paste. Bambara nut flour pastes, however, showed a closer relation­ ship between swelling power and solubles of starch i.e. paste viscosities were very high on holding for one hour at 95°C and also had very high swelling and solubility values at this temperature. However, an interesting relationship was observed for the viscosity and the solubilities among the cereal flour pastes held at 95°C. It was observed that the lower the solubilities of these cereal starches, the higher their flour viscosities and also the higher the swelling power of. the starches, the higher their respec­ tive flour paste viscosities at 95°C. These findings support the theory that starches or flours from similar sources (i.e. cereals) tend to have more similar behavioural patterns than those from other University of Ghana http://ugspace.ug.edu.gh 58 botanically different sources e.g. root or legume (Whistler, 1954). Among the legume starches tested, it was observed that even though cowpea starch had a higher swelling power at 95°C than bambara nut starch, in most cases, bambara nut had higher viscosity values than cowpea and therefore did not follow Osman's (1967) theory. The above comparisons of flour pastes, viscosities and the swelling and solubility values of starches revealed that other non­ starch components can influence flour characteristics. There had been a suggested relationship between amylose content and degree of swelling of starches (Deatherage et_ al., 1955). They reported that in some cases high amylose content or solubles could cause less swelling. However, Collison (1968) pointed out that amylose content alone could not determine swelling ability but, more important, was the crystallinity of the molecules and how compact the starch molecules were arranged internally. Collison (1968) again commented that at the present time, there was insuffi­ cient evidence to put forward a comprehensive theory of starch swelling at the molecular level. In the literature review on viscosity, it was mentioned that an increase in viscosity of a starch paste on cooling reflected the retrogradation tendency of a particular starch (Tolmasquim et al., 1971). Results obtained for viscosity measurements on flours (Figures 13 to 17) showed that all flour pastes became more viscous on cooling; hence, it was logical to expect a certain amount of retrogradation for all flour pastes. This expectation was confirmed in all cases except for wheat flour Which did not show any measur­ able syneresis within the test time (72 hours refrigeration). However, University of Ghana http://ugspace.ug.edu.gh 59 it must be pointed out that syneresis is only one aspect of retro­ gradation in gels and that the fact that wheat flour gel on refri­ geration showed increased firmness indicated that a certain amount of retrogradation took place on cold storage. For flour gels that showed syneresis on refrigeration for 72 hours, the amount of syner­ esis was highest for sorghum then bambara nut, followed by cowpea and then millet gel. The differences in viscosity curves of flour pastes held at 95°C (ten minutes) and those held at 4°C (one hour) revealed that sorghum thickened more than the other flours. Bambara nut, cowpea and millet had relatively less thickening between these two temperatures. It was evident that at these points of consideration, there was a definite relationship between amount of increase in viscosity on cooling to the extent of retrogradation of the gels. Tolmasquim et al. (1971) also stated that increased viscosity on cooling reflected a high rate of retrogradation. As mentioned in the literature (Collison, 1968), it is the breaking off of some part of the starch molecules (linear fraction) which help form the gel network for gelation. Therefore, the amylose content of a starch molecule and its solubility during gelatinization would affect the gel characteristics. Deatherage e_t ed. (1955) reported that legumes as a rule contain more amylose than cereals. Again, the results obtained for the swelling and solubility patterns of these starches reflected higher values in all cases for the legumes as compared to the cereals. Generally, results of gel strength and viscosity bear no appa­ rent relationship to each other. This was not totally unexpected for aspects such as amylose content of the starch affecting gel strength University of Ghana http://ugspace.ug.edu.gh 60 are not entirely synonymous with those affecting viscosity. However, it was observed during the preliminary experiments that cowpea did n°t gel quickly on being held at room temperature. This observation seemed to be consistent with the fact that cowpea paste was very stable (little increase or decrease in viscosity) at specific temperatures (Figure 16). Therefore, keeping the paste longer at room temperature did not produce a marked increase in gel strength. It was again observed that even though sorghum starch had the most viscous paste its gel was not very cohesive and therefore lacked strength. University of Ghana http://ugspace.ug.edu.gh PRACTICAL IMPLICATIONS OF STUDY Perhaps the major importance of this research is in the practical implications of the results obtained. It is of interest to discuss these in terms of protein value and functional properties. The essential amino acid contents of similar products used are tabulated in Table 4. As evident from Table 1 the legumes contain higher amounts of protein than the cereals tested. Furthermore, Table 4 shows that the legumes are rich sources of lysine with sulfur-containing amino acids i.e. methionine and cystine, being the major deficiencies. The cereal grains on the other hand tend to be low in lysine content, and sorghum has a very low content of sulfur-containing amino acids. Therefore it is evident that legume and cereal mixtures can be natural supplements to each other. Bread Making It can be seen from data presented in Tables 5 and 6 that any proportions of wheat-cowpea and wheat-bambara nut mixtures lead to improved amino acid balance and increased total protein content over that of wheat alone. The use of wheat flour in Ghana is mainly in bread making. Cakes, doughnuts and cookies are also among the common foods avail­ able to the Ghanaian consumer in which wheat forms the flour base. It is therefore important to try and predict how the other flours and starches tested in this research will perform when mixed with wheat. 61 University of Ghana http://ugspace.ug.edu.gh 62 ESSENTIAL AMINO ACIDS OF FIVE GRAINS (MG/100 GM FOOD)* TABLE 4 Essential Cereals Legumes amino acids Wheat Sorghum Millet Cowpea Bambara nut Arginine 602 311 512 1498 1121 Histidine 299 217 237 764 535 Isoleucine 426 397 397 895 776 Leucine 871 1348 927 1647 1385 Lysine 374 204 332 1599 1141 Methionine 196 141 239 273 312 Cystine 332 152 229 255 184 Phenylalanine 589 496 467 1209 991 Tyrosine 391 271 315 610 617 Threonine 382 306 374 842 617 Tryptophane - 123 189 254 192 Valine 577 507 535 1060 937 &Food Policy and Food Science Service, Nutrition Division, FAO , 1970. University of Ghana http://ugspace.ug.edu.gh 63 TABLE 5 PROTEIN CONTENT AND AMINO ACID SCORE OF WHEAT-COWPEA FLOUR MIXTURES* Parts of wheat Parts of cowpea Limiting amino acid Amino acid score Protein gms/100 gms 9 1 Lysine 64 13.3 8 2 Lysine 78 14.4 7 3 Lysine 87 15.6 6 4 Lysine 94 16.7 5 5 Lysine 102 17.8 4 6 Methionine + Cystine 113 19.0 3 7 Methionine + Cystine 113 20.0 2 8 Methionine + Cystine 113 21.2 1 9 Methionine + Cystine 113 22.3 10 0 Lysine 53 12.2 0 10 Methionine + Cystine 64 23.4 Calculated from: Food Policy and Food Service, Nutrition Division, FAO, 1970, p. 38, 42, 50, 52; and FAO, Nutrition Report Series No. 53, 1973, p. 63. University of Ghana http://ugspace.ug.edu.gh 64 PROTEIN CONTENT AND TABLE 6 AMINO ACID SCORE FLOUR MIXTURES* OF WHEAT-BAX BARA NUT Parts of Parts of Limiting Amino Protein wheat bambara nut amino acid acid score gms/100 gins 9 1 Lysine 78 12.8 8 2 Lysine 72 13.3 7 3 Lysine 79 13.9 6 4 Lysine 86 14.4 5 5 Lysine 92 15.0 4 6 Methionine + Cystine 94 15.5 3 7 Methionine + Cystine 90 16.1 2 8 Methionine + Cystine 87 16.6 1 9 Methionine 83 17.15 10 0 Lysine 53 12.2 0 10 Methionine + Cystine 80 17.7 Calculated from: Food Policy and Food Service, Nutrition Division, FAO, 1970, p. 38, 42, 50, 52; and FAO, Nutrition Report Series No. 53, 1973, p. 63. University of Ghana http://ugspace.ug.edu.gh 65 Sandstedt (1961) studied the baking characteristics of several non—wheat starches and said that gelatinization plays an important role in the early stages of baking and also that it has an effect on the flexibility of cell walls as well as on the susceptibility of starch to amylolytic degradation. In this respect, gelatinization becomes an important factor when considering composite flour mixtures because gelatinization temperatures of native starches vary. In this research, it was observed that wheat starch gelatinized at a much lower temperature than the other starches studied. Rasper et_ al. (1974) reported that starches including millet and sorghum gave inferior bread products. One of the main reasons given for these observations was the difference in gelatinization temperature. In relation to this research, it can be predicted that starches from millet, sorghum, cowpea and bambara nut, if used as -part of composite flour with wheat, may not give good baking perfor­ mance, especially in bread making. One reason could be that during the early stages of baking, the remaining starch (unfermented) is gelatinized and undergoes limited swelling (due to lack of enough water moisture). The gelatinized starch becomes available for hydro­ lysis by enzymes but factors such as the length of time which elapses between the beginning of the gelatinization of starch and the point of heat inactivation of the amylases does affect this (Osman, 1967). From the above it can be assumed that when starches of much higher initial gelatinization temperature and gelatinization range are mixed with wheat flour, the reactions such as starch hydrolysis taking place in wheat flour during, the early stages of baking, will be com­ pleted or near completion before the other starches (non-wheat University of Ghana http://ugspace.ug.edu.gh 66 sources) become fully gelatinized. Obviously, the foreign starches will not be able to effectively contribute to the resulting baked product. There is an indication that this phenomenon could lead to collapse of cell walls in the interior part of the bread because all films do not develop the necessary degree of rigidity during baking (Sandstedt, 1961). Starch helps absorb some of the water in the dough, so without the required amount of starch, the gluten gets too much water and may collapse during the rising of the dough (Sandstedt, 1961). Hoseney (1971) reported that when starches from barley were mixed with wheat, they were found to be compatible in bread making and this was attributed to the fact that both have gelatinization temperature ranges close to wheat starch, 57 to 61°C and 51.5 to 57°C respectively. Rasper et_ al_. (1974) also found that cassava starch mixed with wheat for bread making gave better results than other starches e.g. millet or sorghum. Again, they suggested that this could be due to the fact that cassava starch had a gelatinization range (57 to 65°C) closest to wheat starch. However, both Hoseney (1971) and Rasper ejt al_. (1974) warned that gelatinization is not the only factor contributing to baking performance. Flours and starches from millet and sorghum have been reported to reduce maximum consistency of wheat flour dough. This finding by Rasper ejt (1974) could be explained by the fact that, before baking, water absorbed by wheat is mainly due to the gluten and, therefore, a composite flour with either of the tested flours will reduce the total gluten content of the mixture and hence the strength of the dough. University of Ghana http://ugspace.ug.edu.gh 67 In spite of the above, non-wheat flours have not necessarily followed similar behaviour with their respective starches. Rasper et al. (1974) reported that non-wheat flours including millet and sorghum gave better baking performance with wheat than their starches. In fact they stated that non-starch components could have a greater effect on their performan