RELATIONSHIPS BETWEEN PHYSICOCHEMICAL PROPERTIES AND HARD - TO - COOK PHENOMENON OF DRY COMMON BEANS By ROSETTE NAKALEMA A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Food Science - Master of Science 2015 ABSTRACT RELATIONSHIPS BETWEEN PHYSICOCHEMICAL PROPERTIES AND HARD - TO - COOK PHENOMENON OF DRY BEANS By ROSETTE NAKALEMA Common beans ( Phaseolus vulgaris ) are one of the most important legumes cultivated worldwide today. De spite the fact that all beans have a long cooking time, some beans cook faster than others. The objective of this study was to determine the physicochemical properties of dry common beans influencing their cooking time. Three genotypes of yellow and three of red mottled beans from both Puerto Rico and Tanzania locations were used in t he study. In order to ascertain the relationships among physiochemical properties and cooking time s, the proper ties such as seed coat weight and thickness, calcium and magnesiu m contents, protein content , starch content, starch gelatinization and pr otein denaturation temperatures , and pectin content were determined . Ascending order of cooking time for yellow bean varieties was ADP - 521, ADP - 111 and ADP - 513 , and for red mottled be an varieties was ADP - 4 43, ADP - 436 and ADP - 434. There was no significant difference in cooking time of soaked and dehulled studied bean varieties . Country of origin had no effect on the trend of cooking time. Among the studied physicochemical properties, s e ed coat thickness , starch content, total pectin content , and hot water s oluble pectin content each was associated with cooking time of the studied dry bea iii ACKNOWLEDGEMENTS I am very much grateful to the Almighty God for his grace, love and care throughout the two years at Michigan State University . My sincere gratitude goes out to MasterCa rd Foundation Scholars program for offering me a great opportunity to study here at MSU. My earnest appreciation goes to my adviser Professor Perry Ng and my research com mittee members , Dr. Karen Cichy and Professor Gale Strasburg , for all their guidance, encouragement and supervision they offered in accomplishing this w ork. I am appreciative of their confidence in me, patience, correction and aid during the course of this project. Great thanks go to Dr. YongFeng Ai , Dr. Jin Yini ng , Yasmin Salat and Jason Wi e singer for sharing with me their technical expertise and guidance . Heartfe lt gratitude goes to my family and friends who have given me pecuniary, devout and moral suppo rt even when I was a thousand miles away . iv TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........................ vii LIST OF FIGURES ................................ ................................ ................................ ..................... viii KEY TO ABBREVIATIONS ................................ ................................ ................................ ......... x CHAPTER 1 ................................ ................................ ................................ ................................ ... 1 INTRODUCTION AND PROBLEM STATEMENT ................................ ................................ .... 1 CHAPTER 2 ................................ ................................ ................................ ................................ ... 4 LITERATURE REVIEW ................................ ................................ ................................ ............... 4 2.1. Physicochemical properties influencing cooking ti me of beans ................................ .......... 4 2.1.1 Introduction ................................ ................................ ................................ .................... 4 2.1.2. Bean seed coat ................................ ................................ ................................ ............... 4 2.1.3 . Bean cotyledon ................................ ................................ ................................ .............. 8 2.1.4. Proteins ................................ ................................ ................................ .......................... 8 2.1.5. Starch ................................ ................................ ................................ ........................... 10 2.2. Propo sed mechanisms of the HTC Phenomenon ................................ ............................... 13 2.3. Factors influencing the HTC phenomenon ................................ ................................ ........ 15 2.3.1.1 Time ................................ ................................ ................................ ........................... 15 2.3.1.2. Relative Humidity and Temperature ................................ ................................ ........ 15 2.3.2.1 Polyphenolic compounds ................................ ................................ ........................... 16 2.3.2.2. Phytic acid ................................ ................................ ................................ ................ 18 2.3.2.3. Dietary fiber ................................ ................................ ................................ .............. 19 2.3.2.4. Pectin ................................ ................................ ................................ ........................ 20 2.3. 2.5. Divalent cations ................................ ................................ ................................ ........ 21 2.3.2.6. Starch content and starch gelatinization temperature ................................ ............... 25 2.4. Methods used to reduce cooking ti me of dry common beans ................................ ............ 26 2.4.1. Soaking ................................ ................................ ................................ ........................ 26 2.4.2. Micronization ................................ ................................ ................................ ............... 27 2.4.3. Dehulling ................................ ................................ ................................ ..................... 27 2.5. Knowledge gaps ................................ ................................ ................................ ................. 28 v 2.6. Hypothesis ................................ ................................ ................................ .......................... 29 2.7. Objectives ................................ ................................ ................................ ........................... 29 CHAPTER 3 ................................ ................................ ................................ ................................ . 30 METHODS AND PROCEDURES ................................ ................................ ............................... 30 3.1. Materials: Dry common bean samples ................................ ................................ ........... 30 3.2. Sample preparation ................................ ................................ ................................ ......... 32 3.3. Determination of cooki ng time ................................ ................................ ....................... 32 3.4. Determination of the percentage of seed coat weight ................................ ..................... 33 3.5. Measurement of the seed coat thickness ................................ ................................ ......... 33 3.6. Determination of calcium and magnesium contents ................................ ....................... 36 3.7. Determination of total protein content ................................ ................................ ............ 36 3.8. Determination of total starch content of the common bean varieties ............................. 37 3.9. Thermal properties of common dry beans ................................ ................................ ...... 38 3.10. Extraction and determination of contents of hot water soluble pectin and hot water insoluble pectin and total pectin of dry common beans. ................................ ....................... 39 3.11. Statistical anal ysis of data ................................ ................................ ............................. 41 CHAPTER 4 ................................ ................................ ................................ ................................ . 42 RESULTS AND DISCUSSION ................................ ................................ ................................ ... 42 4.1. Effe ct of processing on cooking time of common beans ................................ ................ 42 4.2. Percentage seed coat weight ................................ ................................ ........................... 48 4.3. Seed coat thickness ................................ ................................ ................................ ......... 50 4.4. Calcium and magnesium contents ................................ ................................ .................. 56 4.5. Protein content ................................ ................................ ................................ ................ 59 4.6. Starch c ontent ................................ ................................ ................................ ................. 61 4.7. Hot water soluble pectin (HWSP) content ................................ ................................ ..... 64 4.8. Hot water insoluble pectin (HWIP) content ................................ ................................ ... 67 4.9. Total Pectin ................................ ................................ ................................ ..................... 68 4.10. Starch gelatinization temperature ................................ ................................ ................. 69 4.11. Protein denatu ration temperature ................................ ................................ .................. 71 CHAPTER 5 ................................ ................................ ................................ ................................ . 72 CONCLUSIONS AND RECOMMENDATIONS ................................ ................................ ....... 72 vi APPENDIX ................................ ................................ ................................ ................................ ... 73 REFERENCES ................................ ................................ ................................ ............................. 79 vii LIST OF TABLES Table 1. Overall Mineral Content of Phaseolus vul garis (Common beans). Source: Salunkhe and Kadam (1989) ................................ ................................ ................................ ................................ 24 Table 2. Cooking times of raw tempered yellow (Y) and red mottled (R) beans from Puerto Rico (PR) and Tanzania (TZ); beans were raw (not soaked or dehulled), soaked, or dehulled before cooking 1 ................................ ................................ ................................ ................................ ......... 43 Table 3. Seed coat weight ( %) and thickness (µm) of fast, moderate and slow cooking yellow and red mottled beans from Puerto R ico and Tanzania 1 ................................ ................................ 49 Table 4 . Correlation coefficients between cooking time of raw common beans and their physicochemical properties: protein content, gelatinization temperature, denaturation temperature, seed coat weight, seed coat thickness, magnesium content, and calcium contents, starch content, hot water soluble pectin, hot water insoluble pectin, and total pectin for dry common beans from Tanzania and Puerto Rico locations 1 ................................ ........................... 55 Table 5. Calcium and magnesium contents of whole seeds of fast, moderate, and slow cooking yellow and red mottled common dry beans from Puerto Rico and Tanzania locations 1 ............... 57 Table 6. Protein concentration of fast, moderate, and slow cooking yellow and red mottled common dry beans varieties from Puerto Rico and Tanzania locations 1 ................................ ....... 60 Table 7. Starch contents of raw fast, moderate, and slow cooking yellow and red mottled dry common beans from Puerto Rico and Tanzania locations 1 ................................ ........................... 63 Table 8. Hot water soluble pectin (HWSP), hot water insoluble pectin (HWIP), and total pectin (Total) of six fast (F), moderate (M), and slow (S) cooking common bean varieties of yellow (Y) and red mottled (R) beans grown in Puerto Rico and Tanzania locations 1 ................................ ... 66 Table 9. Starch gelatinization and protein denaturation temperatures of dry common bean varieties of yellow and red mottled beans from Puerto Rico and Tanzania locations 1 .................. 70 Table 10. Correlation coefficients among physicochemical properties of dry common beans: cooking time (CK) protein content (PC), protein solubility (PS), Gelatinization temperature (GT), Denaturation temperature (DT), seed coat weigh t (SCW), seed coat thickness (SCT), magnesium content (MC) and calcium content(CA), starch content (SC), HWSP, HWSIP and total pectin (PC) for dry common beans ................................ ................................ ........................ 76 viii LIST OF FIGURES Figure 1. A cross section of a mature broad bean seed with one cotyledon removed. Source: Salunkhe and Kadam (1991) ................................ ................................ ................................ ........... 5 Figure 2. A microstructure of mung bean seed coat and cot yledon: PC= Palisade cells; MC= Mesophyll cells; SP= Spongy cells. Source: Salunkhe and Kadam (1991) ................................ .... 7 Figure 3. Scanning electron microscope (SEM) photomicrographs for chick pea cotyledon; ( A) dry and (B) soaked; S - starch globule, ECS - Extra cellular space. Source: Tiwari and others (2011). ................................ ................................ ................................ ................................ ............. 9 Figure 4. Chemical structure of amylose and amylopectin Source: Tester and others (2004) . .... 12 Figure 5. Scanning Electron Microscope images of the middle lamella between three cells of bean cotyledon stored under different conditions for 6.5 months. (A) Control (5 o C/40% RH) and ( B) HTC (35 o C/75% RH). Source: Mohan and others (2011). ................................ .................... 14 Figure 6. The polyphenol units commonly found in common beans. Source: Petry and others, (2015). ................................ ................................ ................................ ................................ ........... 17 Figure 7. Dry common bean genotypes used in the study. ................................ ........................... 31 Figure 8. Cross sections of raw beans mounted on aluminum stubs in a Turbo Pumped Coater for platinu m coating. ................................ ................................ ................................ ........................... 34 Figure 9. SEM image of a cross section of a raw ADP - 513 - S common bean seed from Puerto Rico with four seed coat thickness measurements indicated (in orange). ................................ .... 35 Figure 10. Cross - sections of seed coats for fast, moderate and slow cooking common beans by Scanning Electron Microscope (SEM). Y - 521 - F = fast cooking, Y - 111 - M = moderate cooking and Y - 513 - S = slow cooking f or yellow beans. R - 443 - F = fast cooking, R - 436 - M = moderate cooking and R - 434 - S = slow cooking for red mottled beans ................................ ....................... 53 Figure 11. The peak gelatinization (first peak) and peak denaturation ( second peak) temperatures of dry common beans from Tanzania. ................................ ................................ .......................... 74 Figure 12. The peak gelatinization (first peak) and peak denaturation (second peak) temperatures of dry common beans from Puert o Rico. ................................ ................................ ...................... 75 Figure 13. Cooking time of raw tempered yellow (Y) and red mottled beans (R) from Puerto Rico (PR) and Tanzania (TZ) that were raw (not soaked or dehulled), soaked and dehulled bef ore ix cooking. Data represent means and standard deviations. Values of each bar topped by the same ................................ .................. 77 Figure 14. Hot water sol uble pectin (HWSP), hot water insoluble pectin (HWIP), and total pectin (Total) of six fast (F), moderate (M), and slow (S) cooking common bean varieties of yellow (Y) and red mottled (R) beans grown in Puerto Rico and Tanzania locations . Data represent me ans and standard deviations of n=2. Values of each bar topped by the same letter are significantly ................................ ................................ ............................ 78 x KEY TO ABBREVIATIONS ADP - Andean Diversity Panel HTC - Hard - to - cook ETC - Easy - to - cook HWSP - Hot water soluble pectin HWIP - Hot water insoluble pectin TP Total Pectin SEM - Scanning Electron Microscope Tg - Gelatinization temperature Td Denaturation temperature 1 CHAPTER 1 INTRODUCTION AND PROBLEM STATEMENT Common beans ( Phaseolus vulgaris L. ) are one of the most important legumes cultivated worldwide today (El - Ta bey Shehata, 1992) and a major source of food to large populations of people in the world. They originated from South and Central America. There are different market classes of beans which are differentiated according to seed size, color and s hape. Examples include yellow beans, red mottled beans, pinto beans, kidney beans, black beans , etc. They are a good source of dietary protein, fiber, starch, vitamins and minerals (Kaur and others, 2013). Unlike fresh co mmon beans, dry beans require a lo ng cooking time , which has been a problem for cen turies. Cooking dry beans is time consuming and , most importantly , energy consuming. Beans being a staple food commonly eaten as a protein source, they are the b iggest component of most di et in Uganda. In m ost African countries , particularly Uganda, firewood is the main source of energy for cooking since the biggest percentage of population ca n not afford gas and electricity. This implies that a lot of trees are cut down per day for this cau se , which has detrimentally degraded the environment. For developed countries that mainly use gas or electricity, cooking beans is costly since more gas or electricity is needed to cook dry beans than other foods like rice or beef . This justifies the impor tance of study ing relationships among cooking time and physiochemical properties of dry beans in hopes of eventually producing varieties with shorter cooking times. Past studies have proven that long cooking time of dry beans is a result of beans hardeni ng during storage . This is known as the h ard - to - cook (HTC) phenomenon. Long storage time, high 2 temperatures and high relative humidity are factors that were reported to enhance th e hardening phenomenon of beans (Aguilera and Stanley, 1985) . There are two possible mechanisms by which storage hardens the beans : (1) pectin - phytate crosslinking in the cell walls and (2) lignification (Aguilera and Stanley, 1985) . The latter is a result of deposition of lignin - like material into the cell walls from the seed coat , thus hardening them (Stanley and others , 1989; Stanley, 1992). Pectin - phytate crosslinking is promoted by high temperature and relative humidity storage condi tions. Under these conditions, phytic acid undergoes hydrolysis , releasing the bound divalen t cations , i.e. , calcium and magnesium . These cations then move to the middle lamella where they participate in crosslinking of the enzymatically demethoxylated pectin substances and phenolic compounds , thus hardening the cell walls (Del Valle and Stanl ey, 1995). This hardening of the bean cell walls increase s rgy and time, several methods have been adopted to reduce the cooking time (i.e. , easy - to - cook, ETC, methods), such as soaking (in water, sodium and po tassium salt solutions), micronization , pressure cooking , and dehulling. The HTC effect develops in all dry beans regardless of the variety. However, some varieties have inherently longer cooking time s than others. Maryange and others (2010) reported a si gnificant difference in cooking time among 30 bean lines that were tested , with times ranging from 29 - 83 minutes using the Matson method ( Maryange and others , 2010) . These differences were maintained even when dry beans were preconditioned (soaking of dry beans in distilled water for 4 hours) before cooking. Cooking time of beans is said to be a genetic trait (Singh, 1999) , but it is also associated with factors such as seed size, seed coat color and thickness, the cotyledon and seed coat chemical compositi on s , and the size of the micropyle and hilum (Mkanda and others, 3 2007). However, it is still not fully understood why some varieties cook faster than oth ers. The aim of this study was to understand if the physicochemical properties influencing the HTC mech anism are associated with the varietal differences in cooking time of different market classes of dry common beans. Information generated from this study could be used by plant breeders to perform targeted genetic engineering in order to deve lop bean varie ties with shorten ed cooking time s . 4 CHAPTER 2 LITERATURE REVIEW 2.1 . P hysicochemical properties influencing cooking time of beans 2.1.1 Introduction A mature dry common bean seed (hereafter referred to as bean) is mainly comprised of two cotyledon s and a seed coat. The cotyledon s comprise 80 - 90% and the seed coat 8 - 20% of the total seed weight (Sathe and Deshpande, 2003). There are different market classes of beans , which are differentiated according to seed size, color and shape. The cooking time of be ans is greatly influenced by both physicochemical properties a nd environmental factors. The physicochemical properties include seed size, seed coat color and thickness, the cotyledon and seed coat chemical composition s , and size of the micropyle and hilum (Mkanda and others, 2007). The environmental factors are temperature and relative humid ity of the storage environment. 2.1.2 . Bean seed coat A seed coat is an outer layer of a bean seed (Figure 1) that protects the inner part of the seed from the external environment (Reyes - Moreno and Paredes - Lopes, 1993). It is rich in minerals such as calcium and iron (Sathe and Dashpande, 2003). In addition, the cell walls of the seed coa t are reported to have high amounts of fiber including cellulose (59.4 to 60.7%), hemicelluloses (17.4 to 25.8%), pec tin substances (11.1 to 15.9%) and lignin (1.4 to 1.9%). 5 Figure 1 . A cross section of a mature broad bean seed with one cotyledon removed. Source: Salunkhe and Kadam (1991) 6 The seed coat has also been reported to contain phenolic compounds , especial ly for the colored common beans ( Phaseolus vulgaris ) , e.g., red, black and bronze beans (Bressani and Elias, 1980). Depending on the variety of the bean, the microstructure of the seed coat ma y have a morphous palisade cells (Figure 2) as observed in the black eyed peas, or organized palisade cells, as observed in adzuki beans (Dedeh and Stanley, 1979). The organized palisade cells could be related to thick seed coats as a result of old age/matu rity (Sathe and Deshpande, 2003). On the other hand, the amo rphous palisade cells could be associated with thin seed coats with loosely packed palisade cells (S efa - Dedeh and Stanley, 1979). Each seed coat has a micropyle (Figure 1) , a small pore through wh ich water enters the seed for germination. This pore, too, plays a role in hydration of the seed. Agbo and others (1987) reported that beans with open micropyles and pores in their seed coats showed rapid water uptake as compared to those with closed micro pyles and no pores in the seed coat. Also, thick seed coats are associated with low water uptake since the packed palisade cells act as a barrier to water penetration. In addition, thick seed coats were reported to be associated with high lipid content of seed coats (Sathe and Deshpande, 2003) which would further hinder hydrat ion of the bean cotyledon. T hi s would imply that beans wi th thick seed coats could have longer cooking times than those with thin seed coats. In contrast, Pirhayati and others (2011) r eported that ETC beans had thicker seed coats (38.00 ) than HTC beans (32.00 ), and therefore cooking time was not necessarily attributed to seed coat thickness. In addition, there were no significant difference s between the microst ructure s of the seed coat s of the ETC and HTC beans in the ir study . On t he contrary, Bhatty (1995) reported structural differences between the microstructure s of seed coats of HTC and ETC lentils. More research is needed to clarify these contradictions . 7 Figure 2 . A microstructure of mung bean se ed coat and cotyledon: PC= Palisade cells ; MC= Mesophyll cells ; SP= Spongy cells. Source: Salunkhe and Kadam (1991) PC SC MC 8 2.1.3 . Bean cotyledon A common bean seed is composed of two cotyledons covered with a seed coat. The cotyle dons contain parenchyma cells ( Figure 3) which are the storag nutrients, e.g., protein bodies and starch granules (Sefa - Dedeh and Stanley, 1979b). Each cell wall contains 15 to 25% protein, 50 to 75% carbohydrates and 0.4 to 0.6% lignin (Sathe and Deshpande , 2003). Reyes and Paredes - Lopez (1993) reported that cotyledon cell walls also contain non - starch polysaccharides including cellulose (25.9 to 30.9%) and, pectin substances (28.5 to 41.2%) in the middle lamella . 2.1.4. Proteins The common bean ( Phaseolu s vulgaris L .) contains 18 - 2 5 % proteins , most of which are salt - soluble globulins (45 to 70%) and water - soluble albumins (10 to 30%) ( Chung and others, 2008). Proteins play a major role in water absorption in the parenchyma cells of the cotyledon. However, the cells do not have direct contact with water because of the seed coat which is a barrier to water movement (Sefah - Dedeh and Stanley, 1979b). Proteins form a matrix around the starch granules , which hinders hydration of the granules. The different pack ing densities of the parenchyma cells also affect hydration (Jones and Boutler, 1983). Bernal - Lugo and ot hers (1997) discovered that an easy - to - cook (ETC) bean ( variety Michigan 800) had a h igher protein content (252.4 mg/g) than a HTC bean (variety Ojo de Cabra, 219.0 mg/g). 9 Extra cellular space Figure 3 . Scanning electron m icroscope (SEM) photomicrographs for chi ck pea cotyledon; (A) dry and (B ) soaked; S - starch globule, E C S - Extra cellular s pace . Source: Tiwari a nd others ( 2011). Parenchyma cells (storage units) inside the cotyledon. Cell wall B A 10 It is not known however, if other genotypes of the same market classes would behave in the same way in terms of cooking time. If protein content does not have an influence on cooking time, perhaps its functional properties such as denat uration temperature could play a role. Reyes - Moreno and other s (1994) reported that ETC beans ( variety Michigan 800) had a lower protein den aturation temperature (88.6 o C) than that of HT C beans (variety Ojo de Cabra, 95.3 o C), and indicated that this affect ed the cooking time of beans. On the contrary, Bernal - Lugo and others (1997) reported that both ETC and HTC beans had similar protein denaturation temperatures (103 o C and 104 o C , respect ively) and therefore protein denaturation temperature did not appear to contribute to cooking time. The other functional property that may influence cooking time of beans is protein solubility. Coelho and others (2011) reported that in their studied beans, a decrease in soluble protein s was associated with increase s in stor age time and cooking time. They concluded that the high storage temperature induced protein denaturation , which caused coagulation of protein thus reducing its solubility. It is not known if common beans stored under similar conditions (e.g., temperature a nd time) , but with different cooking times , will exhibit similar protein solubility properties. 2.1.5 . Starch Legumes contain about 13 - 49% sta rch content on dry matter basis. Starch is the most abundant carbohydrate in the common bean, constituting abou t 22 - 45% by weight (Chibbar and others, 2010). It is a polymer of a few thousands of D - glucose units and it consists mainly of two polysaccharides, amylose and amylopectin (Figure 4 ) . Amylose is a linea r molecule composed of 11 D - - 4 linkages. Amylopectin is a branched molecule of D - glucose units with about 5% - 6 branch linkages. It has a molecular weight of 10 9 g/mole (Jane, 2012). 12 Figure 4 . Chemical structure of amylose and amylopectin Source: Test er and others ( 2004). 13 2.2 . Proposed mechanisms of the HTC Phenomenon (cell wall thickening) and pectin - phytate being the most pos sible mechanism s (Hincks and Stanley, 1987). These mechanisms result in increased difficulty in achieving cell separa tion during cooking (Stanley and Aguilera, 1985). Soft texture of the seed coat simplifies separation of cell walls during cooking of beans , which signif ies the role of the middle lamella and cell wall in this phenomenon ( El Tabey Shehata, 1992). This was confirmed when Garcia and others (1998) reported with scanning electron micrographs of common bean cell walls stored at 5°C/40% RH and 35°C/75% RH which showed thickening of the middle lamella (Figure 5). Pectin - phytate crosslinking is promoted by high temperature and relative humidity storage condi tions. Under these conditions, pectin methylesterase (PME) hydrolyses pectin molecules forming pectic acid and methanol. Also, phytase hydrolyses phytic acid releasing the bound divalent cations , i.e. , calcium and magnesium . These cations then migrate to the middle lamella where they react with pectic acid and phenolic compounds forming insoluble substances thu s hardening the cell walls ( El Tabey Shehata, 1992). 14 Figure 5 . Scanning Electron Mic roscope images of the middle lamella between three cells of bean cotyledon stored under different conditions for 6.5 months. (A) Contr ol (5 o C/40% RH) and (B) HTC (35 o C/75% RH). Source: Mohan and others (20 11). Middle lamella Thickened middle lamella 15 2.3 . Factors influencing the HTC phenomenon 2.3.1.1 Time The cooking time of dry beans depend s not only on the genotype of the common bean (Singh, 1999), but also on storag e conditions, including storage time, temperature and relative humidity (Nasar - Abbas and others, 2008). Duration of storage of beans is one of the main factors influencing cooking time of beans. The longer the storage time, the harder the texture of the dr y beans. However, rate of bean texture hardening depends on the storage co nditions such as relative humidity and temperature. Moscosco and others (1984) reported an increase in hardness with increasing storage time of red kidney beans stored for 9 months. Furthermore, Curtis (1991) studied a relationship between bean texture hardness and soaking time before cooking , where fresh and aged beans were soaked for 0, 20 and 44 hours and their texture hardness es determined. Aged beans were observed to have the gre atest hardness at each of the soaking time s . 2.3.1.2 . Relative Humidity and T emperature Jones and Boulter (1983) reported that high relative humidity (RH) of the storage environment is an initiation stage of the hardening effect of beans. It allows metabo lic breakdown of the cell wall allowing bivalent ions from hydrolyzed phytate to access pectin making the latter insoluble. Hentges and others (1991) studied the effects of storage temperature and humidity on physical and chemical components of beans. They r eported that seeds stored at 29 o C and 65 % RH required a longer cook ing time than seeds stored at 5 o C and 30% RH, or 29 o C and 30% RH , or 5 o C and 65% RH. There is a linear relationship between HTC effect and high storage temperature. Nasar - Abbas and others (2008) reported a li near increase in the HTC effect on Faba beans with high storage 16 temperatures of > 37 o C. In addition, Reyes - Moreno and others (2 001) r eported a long cooking time of chick peas after storing them at temperatures > 25 o C and relative humid ities > 65%. Furthermore , storage of beans under high temperature and relative humidity conditions was reported by Kon and Sanshuck (1981) to have increas ed cooking time by about 5 - fold . Similarly, prolonged storage of dry beans at high temperatures (30 - 35 o C) and high relative humidities (60 - 80%) has been reported to elevate the HTC phenomenon (Taiwo, 1998). It is not clear which one of the three factors (time, temperature , or relative humidity ) is the major contributor to the HTC effect. 2.3.2.1 Polyphenol ic compounds Polyphenolic compounds are a heterogeneous class of compounds derived from secondary plant metabolism. They protect the plant against pathogens and UV radiation ; they are also responsible for pigmentation in beans . In addition, they play an i mportant role in pollination by insects (Petry and others, 2015). Some of the polyphenols present in common beans include phenolic acids, anthocyanidins, and flavonoids ( Figure 6 ) (Petry and others, 2015). Flavanols are commonly referred to as proanthocyan idins or condensed tannins ; they are the major polyphenols found in colored beans and are located in the seed coat (Reddy and others, 1985). Diaz and others (2010) determined polyphenol concentration in bean seed coats and they reported an average concent ration of about 20% of seed coat weight, with an overall average of about 20% in 250 bean varieties . 17 Figure 6 . The polyphenol units commonly found in common beans . Source: Petry and others, ( 2015) . 18 Phenolic compounds are als o responsible for the browning of the seed coat of legumes during storage (Nozzolillo and Bezada, 1984). Browning is ascribed to the polymerization of soluble tannins forming insoluble brown high molecular weight tannins. The presence of condensed tannins in the seed coat of lentils was supporte d by the deep color (brown) of not - soaked seeds after four days of storage as compared to the yellow color of the presoaked seeds after the same storage time. Phenolics in these lentil sampl es were abundant in the n ot - soaked seeds and most originated from the seed coat (Nozzolillo and Bezada, 1984). 2.3.2.2 . Phytic acid Phytic acid is a natural reservoir of phosphorous in the bean seed and it has been implicated in influencing the HTC effect in legumes. Stanley and Aguilera (1985) reported that phytic acid chelates divalent ions (Ca 2+ and Mg 2+ ), hindering them from precipitating pectic acid in the cell wall. More evidence was reported by Chang and others (1997) wh o studied the phytase enzyme (myo - inosito l hexakisphos phat e phosphohydrolase, EC. 3.1.3.8). The activity of this enzyme increases with storage time at high er te mperatures and relative humidities . Phytase hydrolyses phy t ate, releasing the divalent cations that move to the cell w alls where they precipitate pect ic acid and p henolic compounds resulting in insoluble substances forming in the middle lamella , thus hardening the bean. Upon blanching the beans before storage, the hardening effect was reduced , implying that the enzyme had been inactivated (Vindiola and others, 1986). Bhatty and Slinkard (1989) studied the effect of phytic acid and cooking quality of 36 samples of lentils; they reported that the cooking time was predominantly influenced by phytic acid. Seeds with a long cooking time had a low phytic acid content which implied t hat most of it had been hydrolyzed. Increasing 19 phytic acid levels in legumes would be a remedy to HTC effect ; on the contrary, phytic acid binds divalent cations making them unavailable for human nutrition . Moscosco and others (19 84) reported a decrease in the phytic acid content of beans during storage for 9 months under elevated temperatures and high relative humidity conditions which was associated with a subsequent increase in cooking time. After the ninth month , beans were lef t with less phytic acid and fewer of the monovalent ions (Na + and K + ) that are typically responsible for solubilizing the pectin substances via ion exchange and chelation during cooking. It was conc luded that the degraded cell wall allowed redistribution o f divalent ions to the middle lamella during storage where they formed insoluble subst ances with pectin. This was in agreement with the pectin - phytate theory. Furthermore, Chitra (1994) studied the effect of storage on phytic acid and cooking time . He repo rted a notable loss of phytic acid associated with elevated cooking time for chick peas and soy beans stored at 25 o C and 37 o C, respectively. For short cooking time , a high phytic acid content is needed, however, it binds divalent ions making them una vailab le for human nutrition. It was suggested that genotypes with high phytic acid could be identified and bred to improve their cooking time s ( Chitra , 1994). 2.3.2.3 . Dietary fiber Dietary fiber is defined as the component of plant cells that is resistant to hydrolysis by human digestive enzymes. Dietary fiber includes cellulose, hemicellulose, pect in , lignin , and resistant starch. Dietary fiber can be classified according to its solubility, i.e., water - soluble and water - insoluble (Salunkhe and Kadam, 1989). The dietary fiber content of legumes can be affected by variety, environmental conditions and agronomical practices (Wang and others, 2008). T he total dietary fiber (TDF) of pulses ranges from 8 - 27.5% , with soluble dietary fiber (SDF) in the range 20 of 3.3 - 1 3.8% (Gullion and Champ, 2002). Gullion and Champ further reported that DF is located in both the cotyledon s and the seed coat of the common bean. The cell walls of the cotyledons contain pectin substances (approximately 55% w/w) and non - lignified cellulos e (about 9%). On the other hand, the seed coat contains a large amount of cellulose (35 - 57% w/w) and small amounts of hemicellulo se and pectin (Van - Laar and others, 1999). Kutos and others (2003) determined contents of SDF , IDF ( i nsoluble d ietary f iber) , T DF and res istant starch (RS) of raw and processed dry beans. The results indicated that thermal processing reduced IDF content and the overall TDF of the beans. Thermal pr ocessing also increased RS of beans. Shiga and others (2009) investigated effects of cooking on fibers of HTC beans, and reported that the HTC effect did not change the amounts of soluble and insoluble fiber contents in cooked beans, but rather the physico - chemical characteri stics of the bean carbohydrates like solubility of the fiber. Fu rthermore, Shiga and others (2011) studied the effect of storage on the solubility of bean hull fiber . It was observed that aging resulted in a decrease in water - soluble fib e r and an increase in water - insoluble fiber. This was caused by the insolubilizatio n of galacturonans and xyloglucan s . Hul ls make up 7% of all water - insolubl e fiber of the whole bean while cotyl edons contribute 10% of the water - insoluble fiber. 2.3.2.4 . Pectin Pectin is a complex polysaccharide in plant cell walls and consists of thre e components i.e., rhamnogalact uronan I, homogalacturonan with homogalacturonan accounting for about 60% of the total pectin in the cell walls (Mohnen, 2008 ). Pectin molecules can also be tightly bound to other polysaccharides like hemicellulose, wall prot eins and phenolic compounds (Caffall and Mohnen, 2009). It consists of about 100 - 200 D - galacturonic acids linked via (1 - - D - galacturonic acid linkage. Shiga and others ( 2006 ) reported that the pectin in common beans consists of rhamnogalacturonans, gala ctans, xylogalacturonans and ramified arabinans. During 21 storage of beans under higher temperature and relative humid i ty, phytase hydrolyses phytate releasing cal c ium and magnesium cations. These cations migrate to the cell walls where they bind demethylate d pectin molecules , forming insoluble substances thickening the cell wall (Reyes - Moreno and Paredes - Lopez, 1993) . There are changes in pectin during storage of beans at high temperature and humidity , which result in increased bean cooking time. Jones and B oulter (1983) observed a reduction in water - solubility of pectin during storage at 4 o C and RH of 65%. In addition, Salvador (2007) observed reduced pectin water - solubility of cowpeas stored at an elevated temperatures of 42 o C and RH of 67% and this was as sociated with increased cooking time from 89 to more than 270 minutes. Bernal - Lugo and others (1997) reported that the fast cooking common beans ( variety Michigan 800) and the slow cooking beans ( variety Ojo de Cabra), both grown in Mexico, had similar tot al pectin contents (25.0 and 25.7 mg/g , respectively). On the other hand, the fast cooking beans ( variety Michigan 800) had more hot - water - soluble pectin (13.6±1.5 mg/g) than the slow cooking beans ( variety Ojo de Cabra; 6.0±1.0 mg/g). 2.3.2.5 . Divalent c ations Common beans contain 70 - 210 /g calcium and 160 - 230 /g magnesium (Table 2. 1). The cations of these minerals have been reported to be involved in the HTC effect of common beans. Soaking cow peas in Ca 2+ salt solutions prior to cooking enhanced the hardening effect (Shewfelt and McW atters, 1992). Aguilera and Rivera (1992) also reported that there was a positive correlation between calcium content and the HTC effect, however, it was small. Furthermore, Kyriakidis and others (1997) boiled beans in both plain water and divalent cation solutions (CaCO 3 and MgCO 3 ) and they reported that beans boiled in the divalent solutions were much harder than 22 those boiled in plain water. However, they did not report which of the two solutions , i.e., CaCO 3 or MgCO 3 solution , caused more hardness. 2 4 Table 1 . Overall Mineral Content of Phaseolus vulgaris (Common beans). Source: Salunkhe and Kadam (1989) Common beans Ca ( /g) Cu ( /g) Fe ( /g) Mg ( /g) Mn ( /g) P ( /g) K ( /g) Na ( /g) Zn ( /g) Raw 70 - 21 0 0.0 - 1.40 3.34 - 8.0 160 - 230 1.0 - 2.0 380 - 570 1320 - 1780 4.0 - 21.0 1.9 - 6.5 Cooked 70 - 260 0.50 - 1.10 2.88 - 7.93 130 - 220 1.0 - 2.1 360 - 510 1100 - 1710 1.5 - 6.90 1.9 - 4.0 25 2.3.2.6 . Starch content and starch gelatinization temperature There are two physicochemical pro perties of starch that are proposed to influence cooking time , i.e. , starch content and starch gelatinization temperature. Bernal - Lugo and others (1997) reported tha t the fast - cooking common beans of variety Michigan 800 had a lower starch content (240.0 m g/g) than the slow - cooking bea n variety Ojo de Cabra (260.0 mg/g) , indicatin g that starch content influenced cook ing time. Furthermore, the fast - cooking bean variety was reported to have a lower amylose content (38.0%) than the slow - cooking bean variety ( 4 4% ) . This implied that amylose content could influence cooking time. Starch gelatinization is the disruption of molecular orders within the starch granule when hea ted in water and it results in irreversible changes in properties such as granular swelling , native crystalline melting, loss of birefringence, and starch solubilization (Jane, 2012) . Gelatinization is usually followed by a process known as pasting. When starch is heated in water, there is disruption of hydrogen bonds between starch molecules, w hich weakens the granules. This allows the starch granules to imbibe water and swell , which co ntributes to an increased viscos ity of the starch - water solution (Jane, 2012). The initial temperature at which gelatinization begins is known as t he onset gelati nization temperature. Studies have shown that high onset gelatinization temperatures are exhibited by starch es with long - branch chains of amylopectin as these can form much more stable crystalline structures than the short chains (Jane and others, 1999). L ower onset gelatinization temperature has been associated with short amylopectin chains (Ovando - Martinez and others, 2011). The peak gelatinization temperatures of varieties Black 8025 and Pinto Durango beans grown in two different locations and with diffe rent water ing regimes were reported to range from 26 70.14 o C to 75.42 o C ; starch of beans grown in a rainy location had a lower peak gelatinization temperature than that of those grown in the irrigat ion locality (Ovando - Martínez and others, 2011) . When coo king beans, starch granules present in the beans are heated in water and therefo re undergo gelatinization. It has been proposed that if beans have a low peak gelatinization temperature, they could have a shorter cooking time. Reyes - Moreno and others (1994) studied physicochemical properties influencing cooking time of fresh, storage - hardened , and c hemical - hardened variety Mayocoba (fast - cooking) and variety Flor de Mayo (slow - cooking) common b eans. He reported that the fast - cooking beans had a lower peak ge latiniza tion temperature (76.6 o C) than the slow - cooking beans (78.7 o C), thus influencing cooking time of the beans. On the contrary, Bernal - Lugo and others (1997) r eported that both fast and slow - cooking beans had similar peak gelatinization temperatures ( 86 o C and 87 o C, respectively), therefore peak gelatiniz atio n temperature may not relate much to differences in cooking time. In both cases, it is not mentioned if the two varieties studied belonged to the same market class; it would be important to know how bean genotypes of the other market classes grown in the same location would behave. 2.4. Methods used to reduce cooking time of dry common beans 2.4 .1 . Soaking Soaking beans before cooking is a common practice used in many parts of the world with a go al of reducing subsequent cooking time. The amount of water abs orbed depends on the permeability of the seed coat (Valle and others, 1992). Soaking dry beans in monovalent ( N a + ) c ation solutions has also been used to re duce the cooking time of beans. De Le o and othe rs (1992) reported that soaking HTC beans in a salt solution of 2.5% K 2 CO 3 or of 0.5% NaHCO 3 , w/v , significantly 27 reduced their cooking time. Furthermore, Salvador (2007) reported that cowpeas soaked in a sodium chloride solution were observed to have improved pectin water - solubility, which was associated with reduced cooking time. Soaking activates rhamnogalacturonase, galactanase and polygalacturonase cell wall enzymes of beans. These enzymes break down the pectin polysaccharides, rhamnogalactur onan I and polygalacturonan giving cell walls new polysaccharide arrangements that result in higher thermo solubi lity hence shorter cooking time (Martinez - Manrique, 2011). 2.4 .2 . Micronization Micronization of legumes is a process in which moisture - conditi oned seeds are treated with infrared heat (1800 to 3400 nm) to soften their texture by breaking the pectin molecules into smaller and more soluble molecules before cooking. This increases water uptake during cooking thus reducing cooking time of beans (Mwa ngwela and others, 2007). According to Fasina and others (2001) , the infrared rays cause vibration of water molecules at 60 000 to 150000 MHz resulting in internal heating of the bean . This is a high efficiency method as compared to other technologies (Bell ido and others, 2003). Salvador (2007) preconditioned cow peas using sodium ion solution prior to micronization, and the combination was seen to reduce cooking time from more than 270 to 59 minutes. 2. 4 .3 . Dehulling Dehulling has been reported to reduce cooking time of pulses. Kon and others (1973) reported a 70% reduction in coo king time of p into beans resulting from dehulling. This was attributed to the removal of the impermeable seed coat that hinders water uptake of the bean during cooking. However , d ehulling has been reported to result in a significant reduction in minerals, especially 28 calcium, copper, magnesium and man ganese (Wang and others, 2009) , which are important for human health. 2.5. K nowledge gaps Pirhayati and others (2011) reported that t he seed coat thickness of the fast - cooking beans (38.00 ) was greater than that of the HTC beans (32.00 ) ; therefore , cooking time might not be solely attributed to seed coat thickness. There were also no significant differences in the microstructure s of the seed coats of the fast and slow - cooking beans. O n the contrary, according to the other published literature, varieties with thick s eed coats are expected to have longer cooking time s ) observations on fast and slow - cooking l entils showed structural differences in seed coat s of slow - cooking lentils. To resolve this contradiction, this current thesis research was aimed at characterizing the dry common bean seed coat thickness and weight, and their relationship s with cooking time. The influence of calcium and magnesium salt so lutions on the cooking time of beans has been extensively studied. Pirhayati and others (2011) reported an increase in hardness of beans after soaking them in tap water (Ca 2+ and Mg 2+ ). This was attributed to the assumption that the calcium and magnesium ions in the tap water migrated to the middle lamella and participated in the crosslinking of pectate substances and phenolic compounds hardening the cell walls. However, the influence of the inherent calcium and magnesium of the entire bean is not yet know n. Protein solubility of dry common beans was reported to decrease with increase in storage time , and result in an inc rease in the cooking time of dry beans . Coelho and others (2012) concluded that the high storage temperature induced protein denaturation which caused coagulation of protein thus reducing its solubilit y in water. It is not known , however, if protein from other dry common bean 29 varieties will behave in a similar manner. Slow - cooking beans were reported to have less hot water soluble pectin (HW SP) and higher starch content than the fast cooking beans (Bernal - Lugo and others, 1997). However , determinations were done on only two va rieties grown in Mexico. It is not known if other var ieties from other locations would follow the same trend. 2.6. Hyp othesis Dehulling reduce s cooking time of raw dry common beans more than does soaking the whole raw bean . Cooking time of dry common beans is associated with seed coat thickness and weight, starch content, starch gelatinization and protein denaturatio n tem peratures, protein content, hot water soluble pectin content , and inherent calcium and magnesium content s in the entire bean . 2.7. Objectives Overall objective To determine the physicochemical properties of components in dry common beans associated with co oking time of dry common beans . Specific objectives 1. To determine the seed coat weight and thickness and their correlation s with cooking time of dry common beans. 2. To determine the thermal properties of various components of dry common beans and to quantify their correlation s with cooking time s of the studied varieties . 3. To quantify the storage protein, starch, and pectin contents , and to determine the contents of inherent calcium and magnesium of dry common beans and to quantify their correlations with cooki ng time . 30 CHAPTER 3 METHODS AND PROCEDURES 3.1 . Materials: Dry common bean samples D ry common beans ( Phaseolus vulgaris L.) belonging to the Andean Diversity Panel (ADP) were used in this study. Bean varieties in the ADP are large seed beans that are comm only consumed in Southern and Eastern Africa. Two market classes were used in this study, i.e., yel low beans and red mottled beans. These market classes were chosen because the y showed large variations in cooking time (Cichy and others, 2015). Each mar ket class had three genotypes, one slow - cooking (S), one moderate - cooking (M), and one fast - cooking (F) variety. Y ellow bean s included Cebo cela ( ADP - 521 - F) that originated from a market Place in Angola , Uyole 98 ( ADP - 111 - M) that originated from Tanzania and w as developed by the Tanzania National breeding Program in 1999 and Canario ( ADP - 513 - S) that originated from Angola (hereafter referred to as Y - 521 - F, Y - 111 - M, and Y - 513 - S, respectively). Genotypes of red mottled beans all originated from Puerto Rico. JB17 8 ( ADP - 443 - F) was released for its superior agronomic traits and disease resistance , Vazon 7 was released in 2005 ( ADP - 436 - M) and PR0737 - 1 ( ADP - 434 - S) is a virus resistant variety developed jointly by University of Puerto Rico, Haiti National Program and U SDA - ARS in 2013 (Cichy and others, 2015). H ereafter referred to as R - 443 - F, R - 436 - M, and R - 434 - S, respectively ( Figure 7) . All six genotypes were grown in two locations , i.e., Puerto Rico and Tanzania (Arusha) in 2014. These beans were sorted and cleaned t o remove dirt and foreign matter. 31 Figure 7 . Dry common bean genotypes used in the study. Y - 521 - F Y - 111 - M Y - 513 - S R - 443 - F R - 436 - M R - 434 - S 32 3.2. Sample preparation The bean samples were tempered by storing whole beans in a moisture equalizer at 14% moisture content at room temperature for 3 weeks ( h ereafter referred to as raw beans ) prior to any analyses. For analyses requiring ground s ample, raw beans were freeze - dried (Refrigeration for Science, Inc. Island Park, New York) for 48 hours, then ground by a Wiley laboratory mill (Arthur H. Thomas Co., Scientific Apparatus , Philadelphia. PA., USA) to pass through a 0.5 mm screen before analysis. The ground sample s were stored in airtight plastic tubes at room temperature . 3. 3 . Determin ation of cooking time Cooking time was determined using a Mattson cooker (Customized Machining and Hydraulics Co., Winnipeg, Canada) ( Wang and Daun, 2005 ) . For each bean variety , 25 raw beans were cooked as is , another 25 raw beans were soaked for 12h bef ore cooking, and another 25 raw beans were dehulled manually using a razor blade before cooking . The M attson cooker has 25 metallic pins under which beans are to be placed, one under each pin . The device is c onnected to a monitor with the M attson software. Distilled water ( 1800 ml ) was boiled in the pan of the Mat t son cooker. The Mattson device , loaded with 25 raw, soaked, or dehulled beans of one variety , was placed in the boiling water , the start button pressed, and the starting time recorded. During cook ing, the bean s softened and the pin s dropped as they pierced the bean s . The time at which each pin dropped was recorded. When 80% of the pins had dropped (i.e. , 20 beans), the time was recorded as the end cooking time of that particular treated bean variet y . Cooking time was calculated by subtracting time at the beginning of cooking from end cooking time . Cooking t ime determinations were done in two replicates for each bean variety and treatment as described above. 33 3. 4 . Determination of the percentage of s eed coat weight The seed coat weight of the beans was determined gravimetrically. Total weight of five raw whole beans from each variety was determined. The beans were then dehulled manually using a sharp razor blade and the total weight of the five dehul led beans was determined. The total seed coat weight of the five beans was the calculated difference between the total weight of five raw beans and the total weight of the same five beans dehulled. The percentage of seed coat weight was calculated as total weight of the five seed coats divided by total weight of the five raw beans multiplied by 100. This was done in two replicates and results for each variety averaged. 3. 5 . Measurement of the seed coat thickness The thickness of the common bean seed c o at was measured using the Scanning Electron Microscope (SEM) . Raw whole bean s of each variety wer e first freeze dried for 48h ( Refrigeration for Science, Inc. Island P ark, New York) to 3 % moisture content. Cross sections of s ix raw beans of each variety were made using sharp razor blades. The cross sections (2 per bean) were then mounted on a luminum stubs using high vacuum carbon tabs (SPI Supplies, West Chester, PA). They were then coated with platinum for sixty seconds while rotatin g to a thickness of 8nm in a Quorum Technologies/ Electron Microscopy Sciences Q15OT Turbo Pumped Coater (Quorum Technologies, Laughton, East Sussex, and England BN8 6BN ) (Figure 8 ). They were then examined using a JEOL JSM - 7500F (cold field emission electron emitter) scanning elec tron microscope (JOEL Ltd, Tokyo, Japan). Images of bean cross section s (Figure 9 ) were taken and the seed coat thickness es measured using the micrograms. This was done in duplicate ( i.e., for a total of 12 beans examined per variety ), with two micrograms produced per bean (one for each cross section). The averages for each variety were recorded . 34 Figure 8 . C ross sections of raw beans mounted on aluminum stubs in a Turbo Pumped Coater for platinum coating. 35 Figure 9 . SEM image of a cross section of a raw ADP - 513 - S common bean seed f ro m Puerto Rico with four seed coat thickness measurements indicated (in orange). 36 3. 6 . Determination of calcium and magnesium contents Calcium and magnesium c ontents were determined using Inductively Coupled Argon Plasma Emission Spectroscopy (ICAP ES ) following AOAC M ethod 985.01 . C alcium or m agnesium standards or ground raw whole beans (2 g each , prepared according to the description in 3.1 ) were each weighed into microwave digestive tubes and 2.0 mL of concentrated nitric acid was added int o each tube. Samples were chemically digested according to the Open Vessel Microwa ve SW846 - 3051A procedure (AOAC 991 - 10D ) at 175 o C for 15 minutes . After the digestion, sampl es were diluted to a volume of 25 ml with deionized water and analyzed on a Thermo iCAP 6000 Series IC A P ES . This analysis was done in duplicate and the averages for each variety were recorded . 3. 7 . Determination of total protein content Ground freeze dried bean samples were analyzed by A & L Great L akes L aboratories, Inc, Fort W ayne, USA. Total protein was calculated by determining total nitrogen of the bean sample s according to the Dumas M ethod ( AOAC 990.03 ) . For each variety, g round raw bean sample ( about 120 mg) was weighed and a sample press was then used t o compress the sample to remove any atmos pheric air. The sample weight was recorded and checked using the in the Dumas software to verify that the weight was recor ded accurate ly . During the run, 100 mg of aspartic acid was used in the blank position to calibrate throughout the run. The samples were loaded into the rapid nitrogen cube auto - sampler carousel for analysis. The percentage nitrogen 37 content was recorded and converted to protein content by multiplying by 6.25. All determinati ons were conducted in duplicate and the averages for each variety were recorded. 3. 8 . Determination of total starch content of the common bean varieties B ean s of each variety were prepared for ana lysis by placing raw whole beans in a freeze drier ( Refrigeration for Science, Inc. , Island Park, New York) for 48 h , then grin d ing each sample using a Wiley laboratory mill to pass th rough a 0.5 mm screen. Total starch content was determined according to Megazyme Total Starch Kit procedure ( Megazyme Internationa l , Wicklow, Ireland). For e a ch variety, ground bean sample ( about 100 mg) was weighed accurately into a glass tube. It was then wetted with 0.2 ml of 80% v/ v aqueous ethanol to aid dispersion and th e mixture was mixed vigorously on a vortex mixer (Fisher Vortex, Genie 2 TM ) for 30 seconds . Im mediately, 2 ml of dimethyl sulf oxide (DMSO) were a dded and the mixture was shaken vigorously using the above vortex mixer. The tube was then placed in a vigorous ly boiling water bath and was removed after five minutes. Immediately, 3 ml of thermal - amylase ( conten ts of bottle 1 diluted 1:30 in R eagent 1; 100 mM sodium acetate buffer, pH 5.0) was added to the tube . The tube was then incubated in a boiling w ater bath for 6 minutes , and stir r ed vigorously using the above vortex mixer after 2, 4 and 6 minutes. The tube was then placed in a water bath (Julabo SW22, USA) at 50 o C and 0.1 ml of the contents of bottle 2 of the kit ( amyloglucosidase, 300 U of starch) was added . The tube w as vortex ed and incubated at 50 o C for 30 minutes. The entire content s of the test tube w as transf e rre d to a 100 ml volumetric flask using a funnel to assist transfer , and a wash bottle with distilled water to rinse out the tube s cont ents thoroughly. The volume was adjusted to the volume mark with distilled wate r and mixed thoroughly by hand swirling . An a liquot of this solution (10 ml) was centrifuged using the above centrifuge at 3000 x g for 10 minutes and the 38 clear undiluted filtra te was used for the assay. Duplicate aliquots of 0.1 ml of each sample were transferre d to glass test tubes. Glucose oxidase/peroxidase ( GOPOD ) reagent (3.0 ml) was added to each tube. In addition, D - glucose c ontrols consisting of 0.1 ml of D - glucose stand ard solution (1m g /ml) and 3.0 ml of GOPOD reagent were prepared and stirred. Furthermore, a blank solution was prepared by adding 0.1 ml distilled water and 3.0 ml of GOPOD reagent to a glass test tube and vortexed . All the tubes were then incubated at 50 o C for 20 minutes. Absorbance for each sample and D - glucose control was read at 510 nm against the reagent blank using a spectrophotometer (Spectronic Genesys TM 5 , USA ). Starch % was calculated usin g the formula : starch % = A * F/W*FV*0.9 , where A = absorbance again st the reagent blank, F = 100 µg of D - glucose/absorbance of 100 µg of D - glucose , FV = 100ml, W = weight of sample flour in mg . This e xperiment was done in duplicate . 3. 9 . Thermal properties of common dry beans Both starch gelatinization and protein denaturation temperature s were measured using a differential scanning c a lorimeter (DSC) (TA Instruments, DSC Q100, DE, USA) and the method reported by Yin and oth ers ( 2008) . R aw whole freeze dried bean s were ground to pass t hrough a 0.5 mm screen. For each study variety, t he sample (5 mg) was weighed into a standard DSC aluminum pan and d istilled water ( 15 µ l) was added to the sample. A rubber ring was placed on the edge of an aluminum standard DSC lid , the lid positi oned on top of the pan , and the pan then sealed tightly with a standard sealer . The samples were then left in the pan to stand at room temperature for 2h before analysis. A reference was prepared by sealing an empty pan to be a control. The reference pan a nd a sample pan were placed in the DSC machine with the reference 39 pan taking the back position. The DSC was closed and analysis was run between 30 o C and 140 o C at a temperature increase rate of 10 o C per minute . This was done in duplicate . 3. 10 . Extraction and determination of contents of hot water soluble pectin and hot water insoluble pectin and total pectin of dry common beans. Pectin content was determined spectrophotometrically using the m eta hydroxydiphenyl method of Blumenkrantz and Asboe - Hansen (1973 ) and using galacturonic acid as a standard ( Berna - Lugo and others, 1996) . Raw bean s ample s were dehulled manually with a blade and freeze dried for 48 hours. The raw - dehulled , freeze - dried beans were then ground to pass through a 0.5 mm screen , and stored in airtight plastic tubes at room temperature until analysis . Before extraction of pectin, soluble sugars were removed from the ground samples by extracting with 75% ethanol . Ground s am ple ( 3 g) was weighed into a beaker and 3 0 ml of 75% ethanol was added . The mixture was stirred at 25 o C for 2h . It was then centrifuged by the above centrifuge for 15 minutes at 4000 x g at 20 o C to remove soluble sugars . The residue , alcohol insoluble solids (AIS) , was washed with 3 0 ml of absolute ethanol twice . It was then oven dried for 24h at 25 o C. Hot water soluble pectin ( HWSP ) and hot water insoluble pectin ( HW I P ) were extracted in the following manners. The HWSP was extracted by weighing 1g of AIS into a beaker , adding 10 ml of distilled water , and then stirring for 2 hrs at 85 o C using the VWR 575 Digital Hotplate S tirrer made in the USA. The mixture was then centrifuged for 15 minutes at 4000 x g and 20 o C . T he extract was frozen overnight an d then freeze dried for 48h using the above freeze drier. The HWIP was extracte d using 2% sodium hexametaphosp h ate. One g of AIS was weighed into a glass beaker and 10 ml of 2% hexametaphosphate w ere added to the sam ple. The mixture was stirred at 90 o C for 6 h . The mixture was centrifuged with the above centrifuge for 15 minutes at 40 00 x g and 20 o C. The extract 40 in the supernatant was frozen overnight and then freeze dried for 48 hours using the above freeze drier. The samples we re kept in a ir tight plastic tubes at room temperature until further analysis. Pectin content was ex pressed a s galacturonic acid . This is because galacturonic acid is mainly used as a representative of acid mucopolysaccharides (including pectin) in biol ogical substances , since it is the main comp onent of repeating units of most acid muco poly saccharides. The colo rimetric assay using m - hydroxydiphenyl solution for analysis of galacturonic acids was used. While all carbohydrates react in concentrated acid to form pink - colored compounds, galact uronic acids react with m - hydroxydiphenyl solution in a strong basic envir onment to form pink - colored complexes. Reag ents used include d the following: (1) g alacturonic acid stock solution, prepared by dissolving 100 mg dry galacturonic acid powder in 100 ml distilled water to give a solution concentration of 1 mg/ml ( t his was k ept refrigerated ); (2) M/80 s odium tetraborate in sul phuric acid (0.0125M ) ; ( 3 ) 0.5% sodium hydroxide ; ( 4 ) m - h y dro xydiphenyl solution (0.15%), c over ed in a container with aluminium fo il to pr otect it from the light , and kept refrigerated. Solutions for the stan dard curve were made as follows: 2 ml stock galacturonic acid solution were pipetted into a 100 ml volumetric flask and diluted to final volume with distilled water. The concentration of gal acturonic acid in this sample was 20 µg/ml. This was subsequ ently repeated using 4 ml , 6 ml , 8 ml , 10 ml , 12 ml, 14 ml and 16 ml of stock galacturonic acid solutions to make 40 , 60, 80,100 , 120, 140 and 160 µg/ ml concentrations of standard solutions . To determine the pectin contents of each bean va r iety, c lean gla ss t est tubes were placed in an ice bath to cool before being used . T he sulf uric acid/sodium tetraborate solution was kept in an ice bath throughout the experiment . An aliquot (1.0 ml) of the standard solution or sample was pipetted into a test tube and al lowed to cool for 1 minute. T he tetraborate solution (6.0 ml) was then added to the test tube. This was mixed thoroughly using a vortex mixer (Fisher Vortex, Genie 2 TM , USA ) 41 for 1 min ute. The glass test tubes were then heated at 80 o C in a water bath (Jul abo SW22, USA ) for precisely 6 minutes. The tubes were then returned to th e ice bath and allowed to cool to 4 o C. M - hydroxydiphenyl solution (0.1 ml) was pipetted into each tube and m ix ed thoroughly using the above vortex mixer . A blank was prepared by mixi ng 10 ml deionized water plus 6.0 ml tetraborate solution and 0.1 ml of 0.5% sodium hydroxide solution. The glass tubes were then allowed to stand for 15 - 20 minutes at room temperature to allow any bubbles formed to dissipate. A bsorbance was measured at 52 0 nm using a spectrophotometer (Spectronic Genesys TM 5 , USA ) by reading sample against the blank . A calibration curve of a bsorbance (y - axis) against concentration of galacturonic acid (x - axis) was plotted and was used to determine the pectin cont ent s of th e bean varieties studied. The experiment was done in duplicate and the averages were recorded . 3.1 1 . Statistical analysis of data All data obtaine d from above experiments were analyzed using Statistical Package for Social Sciences ( SPSS version 2009, Inter national Business Machines , New York, US A) . Multiple Linear regression and Analysis of Variance (AN OVA) were performed and means were separated by 42 CHAPTER 4 RESULTS AND DISCUSSION 4.1. Effect of processing on cooking time of common beans Cooking times of tempered raw, tempered soaked and tempered dehulled common beans (hereafter referred to as raw, soaked and dehulled, respectively) from Puerto Rico and Tanzania are presented in Table 2. Generally, cooking time of the studied bean varieties was specific to variety for both yellow and red mottled beans. For the raw yellow (Y) beans from Puerto Rico, it was confirmed that the Y - 521 - F had the shortest cooking time ( 80. 93±0.87 minutes ) followed by Y - 111 - M ( 83.33±1.27 minutes ) and Y - 513 - S had the longest cooking time ( 106.37±1.69 minutes ); F, M, and S, refer to fast, moderate and slow cooking time, respectively. The cooking time of variety Y - 521 - F was 2.4 minutes less than that of Y - 111 - M and 25 minutes less than that of variety Y - 5 13 - S. This implies that less energy and money is required to cook variety Y - 521 - F than variety Y - 513 - S. These results show that genotypes within a market class have significantly different cooking times. All the yellow genotypes maintained the same cookin g time trend when grown in Tanzania, i.e., Y - 521 - F, Y - 111 - M, and Y - 513 - S in ascending order. The cooking time of variety Y - 521 - F was about 5 minutes less than that of Y - 111 - M and 9 minutes less than that of variety Y - 513 - S. However, the cooking time of ra w yellow beans grown in Tanzania were significantly greater than their counterparts from Puerto Rico (P<0.05). The cooking times of genotypes Y - 521 - F, Y - 111 - M, and Y - 513 - S grown in Tanzania were 31, 33, and 14 minutes longer than those of their counterpart s grown in Puerto Rico. 43 Table 2 . Cooking time s in minutes of raw tempered yell ow (Y) and red mottled (R) beans grown in Puerto Rico and Tanzania ; beans were raw (not soaked or dehulled), soaked for 12 hours , or dehulled before co oking 1 Bean Sample 2 Puerto Rico Tanzania Raw Soaked Dehulled Raw Soaked Dehulled Yellow Y - 521 - F 80.93±0.87 a 31.37±2.71 d 23.42±1.06 d 111.92±8.00 s 29.42±0.93 p 29.28±1.44 p Y - 111 - M 83.33±1.27 b 42.24±2.89 de 26.96±0.37 de 116.53±2.84 x 40.43±0.48 p 30.42 ±0.03 p Y - 513 - S 106.37±1.69 c 42.55±2.94 de 32.36±0.73 de 120.64±2.84 z 48.51±2.37 h 35.19±0.27 h Red Mottled R - 443 - F 95.14±1.51 f 42.28±4.95 m 33.00±0.59 o 118.04±17.01 f 62.66±1.87 u 38.71±0.86 s R - 436 - M 103.09±2.28 g 46.24±9.32 l 37.07±0.79 p 125.63±12.82 v 43 .71±2.60 sr 40.16±1.18 s R - 434 - S 154.88±14.97 y 105.56±0.63 n 40.87±0.82 q 149.07±16.16 y 84.01±4.93 k 44.65±0.64 r 1 Data represent means and standard deviations of n= 36 . Values within the same column or row followed by the same superscript letter are not signi 2 F: Fast - cooking genotype, M: Moderate - cooking genotype and S: Slow - cooking genotype. 44 This could be attributed to environmental differences like temperature and relative humidity during their growth. The variety Y - 513 - S had the least difference in cooking time between Tanzania and Puerto Rico. As expected, soaking significantly reduced the cooking time of beans (P<0.05) for all the yellow genotypes from Puerto Rico or Tanzania, however, the magnitude of th is decrease varied among genotypes. There were 61.2 %, 49.3 %, and 60 % reductions in cooking time of Y - 521 - F, Y - 111 - M, and Y - 513 - S bean genotypes , respectively, from Puerto Rico when soaked before cooking. When grown in Tanzania, there was a 73.7 %, 65 %, and 60 % reduction in cooking time of Y - 521 - F, Y - 111 - M, and Y - 513 - S bean genotypes, respectively, when soaked before cooking. Similar to soaking, dehulling significantly reduced the cooking time for all the yellow genotypes from Puerto Rico or Tanzania (P<0.05 ). The magnitude of this decrease also varied among genotypes. There were 71 %, 68 %, and 70 % reductions in cooking time of Y - 521 - F, Y - 111 - M, and Y - 513 - S bean genotypes , respectively, from Puerto Rico when dehulled before cooking. When grown in Tanzania, the re were 73 %, 74 %, and 7 1 % reductions in cooking time of Y - 521 - F, Y - 111 - M, and Y - 513 - S bean genotypes, respectively, when dehulled before cooking. Generally, Y - 521 - F exhibited a greater reduction in cooking time when soaked than Y - 513 - S. This could imply th at perhaps it had a better soaking ability than Y - 513 - S. There were no significant differences between the cooking times of soaked yellow beans and their dehulled counterparts from Puerto Rico or Tanzania. For the raw red mottled (R) beans from Puerto Ric o, it was confirmed that the R - 443 - F had the shortest cooking time followed by R - 436 - M and R - 434 - S had the longest cooking time; F, M, and S, refer to fast, moderate and slow cooking time, respectively. The cooking time of variety R - 443 - F was 8 minutes les s than that of R - 436 - M and 60 minutes less than that of variety R - 434 - S. This 45 implies that less energy and money is required to cook variety R - 443 - F than varieties R - 436 - M and R - 434 - S. As yellow beans, red mottled beans maintained the same cooking time tre nd when grown in Tanzania, i.e., R - 443 - F, R - 436 - M, and R - 434 - S in ascending order. The cooking time of variety R - 443 - F was about 8 minutes less than that of R - 436 - M, and 31 minutes less than that of variety R - 434 - S. However, the cooking time of raw red mot tled beans grown in Tanzania were significantly greater than their counterparts from Puerto Rico (P<0.05). The cooking times of genotypes R - 443 - F, R - 436 - M, and R - 434 - S grown in Tanzania were 23, 23 and 6 minutes longer than those of their counterparts grow n in Puerto Rico. This could be attributed to environmental differences like temperature and relative humidity during their growth. Like yellow beans, the slow - cooking genotype, R - 434 - S, had the least difference in cooking time when grown in Tanzania and P uerto Rico. This could imply that environmental differences did not have much effect on its cooking time. Like yellow beans, soaking significantly reduced the cooking time of red mottled beans (P<0.05) for all the genotypes from Puerto Rico or Tanzania, ho wever, the magnitude of this decrease varied among genotypes. There were 56 %, 55 %, and 32 % reductions in cooking time of R - 443 - F, R - 436 - M, and R - 434 - S bean genotypes, respectively , from Puerto Rico when soaked before cooking. When grown in Tanzania, there were 47 %, 65 %, and 44 % reductions in cooking time of R - 443 - F, R - 436 - M, and R - 434 - S bean genotypes, respectively, when soaked before cooking. As yellow beans, the fast - cooking red mottled genotype, R - 443 - F, grown in Puerto Rico had the greatest reduction in cooking time when soaked implying that it could have the best soaking ability among the three genotypes. However, for Tanzania red mottled beans, the moderate - cooking genotype, R - 436 - M, had the greatest reduction in cooking time when soaked before cooking . 46 Similar to soaking, dehulling significantly reduced the cooking time of beans (P<0.05) for all the red mottled genotypes from Puerto Rico or Tanzania. The magnitude of this decrease also varied among genotypes. There were 65 %, 64 %, and 74 % reductions in cooking time of R - 443 - F, R - 436 - M, and R - 434 - S bean genotypes, respectively from Puerto Rico when dehulled before cooking. When grown in Tanzania, there were 67 %, 68 %, and 70 % reductions in cooking time of R - 443 - F, R - 436 - M, and R - 434 - S bean genotypes, respe ctively, when dehulled before cooking. Among the three red mottled genotypes grown in Puerto Rico or Tanzania, the slow - cooking genotype, R - 434 - S, had the greatest reduction in cooking time when dehulled before cooking. There were no significant difference s between the cooking times of dehulled red mottled genotypes and their soaked counterparts for both locations. This implies that dehulling reduced cooking time of beans as soaking. The results in the present study show that genotypes within a market class have significantly different cooking times. It is commonly known that yellow beans cook faster than red mottled beans however this is not true for all genotypes. The raw genotype R - 443 - F had a shorter cooking time ( 95.14±1.51 ) than the yellow genotype Y - 5 13 - S ( 106.37±1.69 minutes). This information is very vital when making a choice among many bean varieties for consumption since fast - cooking beans will require less energy and money to cook them. The findings in the present study are in agreement with past studies that have shown that soaking reduced cooking time of legumes, (Taiwo and others, 1998; Martinez - Manrique and others, 2011). The results are also consistent with the common experience of reduced cooking time in households. Soaking activates the pec tic enzymes in the cell walls of b eans which hydrolyze the main pectin polymer rhamnogalacturonan I reducing its degree of polymerization and increasing 47 polygalactururonan and galactan solubility. This results into a re duced thermoso lubility of pectic poly saccharides thus shorter cooking time (Martinez - Manrique and others, 20011). The results of the present study are also in line with Kon and others (1973) who reported that d ehulling reduces cooking time of pulses. This is attributed to the removal of the impermeable seed coat that hinders water up take of the bean during cooking. It has also been reported that dehulling resul ts in a significant reduction of minerals especially calcium, copper, magnesium and m anganese (Wang and others, 2009) and dietary fib er . Removal of these minerals compromises the nutritional quality of the beans, thus soaking is often preferred. From the above results of the current study , we can conclude that the studied b ean genotypes within a market class have different cooking time s , and that s oaking and dehulling each reduced cooking time , relative to raw bean cooking, in a similar way . Since dehul ling is very cumbersome and compromises the nutritional quality of the beans, soaking is the easiest way to reduce cooking time. Although the cooking times of the studied bean genotypes from different locations may be different, the trend s within each location were consistent , i.e., the variety that cooked fast when grown in Puerto Rico , still cooked fast when grown in Tanzania . 48 4.2 . Perc entage seed coat weight The percentages seed coat we ight of the studied raw beans are presented in Table 3. Generally, beans from Puerto Rico had larger percentage seed coat weights than the ir counterparts from Tanzania. The percentage seed coat weight s o f ye llow beans were not significantly different among fast, moderate and slow cooking genotypes for both locations (P> 0.05). Except for the fast - cook ing red m ottled beans variety R - 433 - F being significantly smaller than the slow - cooking variety R - 434 - S fro m Puerto Rico, the percentage seed coat weight values of all the three genotyp es from both locations were not significantly different (P> 0.05). This implies that percentage seed coat weight did not have an association with cooking time of bean genotypes wi thin a market class and between market classes in the studied varieties . Also, locations did not display significant effect on the percentage seed coat weight value s . 49 Table 3 . Seed coat weight ( %) and thickness (µm) of fast, mode rate and slow cooking yellow and red mottled beans from Puerto Rico and Tanzania 1 Bean Variety 2 Seed Coat W eight (%) 3 Seed Coat Thickness ( µm ) Puerto Rico Tanzania Puerto Rico Tanzania Yellow Y - 521 - F 8.65±1.02 a 7.47±0.57 a 51.29±2.52 d 72.47±5.7 9 e Y - 111 - M 9.18±0.6 ac 7.66±0.51 a 59.46±4.86 de 72.47±5.79 e Y - 513 - S 9.27±0.82 ac 7.89±0.11 a 71.89±2.86 e 76.8±9.66 e Red Mottled R - 443 - F 8.29±0.89 a 7.92±0.31 a 70.86±7.18 e 73.83±4.14 e R - 436 - M 9.55±1.04 b ac 8.71±0.43 a c 84.34±3.73 f g 86.68±7.75 fg R - 43 4 - S 10.23±0.59 c 9.71±0.34 ac 85.17±8.06 fg 89.54±0.79 g f 1 Data represent means and standard deviations of n=2 4 . For seed coat weight and n= 7 2 for seed coat thickness. Values within the same column or row followed by the same superscript letter are not sign 2 F: fast - cooking genotype; M: moderate - cooking genotype; S: slow - cooking genotype. 3 S eed coat weight (%) = Total weight of seed coats of 5 beans divided b y total weight of 5 whole beans x 100. 50 4.3 . Seed coat thickness The seed coat thickness es of the studied raw beans are presented in Table 3 , and Figure 10 shows images of cross sections of raw beans with seed coats. G enerally, red mottled beans had thicker seed coats than yellow beans. Also , beans from Tanzania had thicker seed coats than their counterparts from Pue rto Rico. The seed coat thickness of yellow beans from Puerto Rico was significantly different among genotypes (P<0.05) and increase in thickness was associated with an increase in cooking ti me. The seed coat thickness of the slow - cooking genotype Y - 513 - S from Puerto Rico was 1.4 - fold thicker than genotype Y - 521 - F . On the contrary, there were no significant difference s in the seed coat thickness es of fast , moderate and slow cooking yellow bean genotypes (P>0.05) from Tanzania . For red mottled beans, the fast cooking genotype R - 443 - F had a significantly smaller seed coat thickness than the moderate and slow cooking genotypes for both Puerto Rico and Tanzania locations. Generally, seed coat thick ness es of all the three red mottled bean genotypes f rom Puerto Rico and Tanzania were not significantly different (P>0.05) . This would imply that location had n o effect on seed coat thickness. 53 Figure 10 . Cro ss - sections of seed coats for fast, moderate and slow cooking common beans by Scanning Electron Microscope (SEM). Y - 521 - F = fast cooking, Y - 111 - M = moderate cooking and Y - 513 - S = slow cooking for yellow beans. R - 443 - F = fast cooking, R - 436 - M = moderate coo king and R - 434 - S = slow cooking for red mottled beans Y - 521 - F Y - 111 - M Y - 513 - S R - 436 - M R - 443 - F R - 434 - S 54 Regardless of the market class, increase in seed coat thickness was associated with an increase in bean cooking time. There was a 74.3 % positive correlation (P<0.05) between seed coat thickness and coo k ing time of raw beans (Table 4 ) , which implies that cooking time of beans was largely associated with the seed coat thickness. Furthermore, seed coat thickness made a unique association with cooking time (beta coefficient = .824 ) (P< 0.05) ; however , this v alue was lower than that of hot water soluble pectin (beta coefficient = 3.548) implying that it made less of a contribution when the variance explained by all other variables in the model is controlled for . These results are in agreement with (Sathe and Deshpande, 2003) who rep orted that thick seed coats hinder hydration of the bean cotyledon. Hydration is a requirement for any bean to cook. The results from the current study generally show that red mottled beans have thicker seed coats than yellow beans. T his could explain why yellow beans cook faster than red mottled beans. Selecting for a thinner seed coat would be a good method of reducing cooking time of beans , however, it i s a protective mechanism against pests , and so perhaps selecting for thin but tough seed coats would address the problem. 55 Table 4 . Correlation s and beta coefficients between cooking time (CT) of raw common beans and their physicochemic al properties: seed coat thickness (SCT) , protein content (PC) , starch co ntent (SC) , hot water soluble pectin (HWSP) , hot water insoluble pectin (HWIP) , and total pectin (TP) for six varieties of dry common beans grown in Tanzania and Puerto Rico 1 1 Data represent correlation and beta coeffi ci ents for n = 24 . V alues followed by * are significant The R square for this model is 70.2 % (P<0.05) . CT = 10.995 + 1.27 ( SCT) 0.344 (SC) + 5.590 (HWSP) + 6.096 (HWIP) 8.421 (TP) Physicochemical properties Correlation coefficient (r ) Unstandardiz ed coeffi ci ent s ( B ) Standardiz ed coeffi cie nts (Beta) (Constant) 10.995 Seed coat thickness 0 .743* 1.270 * 0 .824 * Protein content - 0 .26 0 - 1.756 - 0 .251 Starch content 0 .417* - 0 .344 - 0 .150 Hot water soluble pectin - 0 .484* 5.590 * 3.548 * Hot water insoluble pectin - 0 . 346* 6.096 2.462 Total pectin - 0 .668* - 8.421 * - 3.987 * 56 4.4 . Calcium and m agnesium content s The calcium and magnesium contents of the studied raw common beans are presented in Table 5 . The calcium con tent ranged between 0.0 8 9 - 0.1 88 g /100 g of beans , and magnesium co ntent ranged between 0.17 2 - 0.22 g/100 g. There were no significant difference s in the c alcium and magnesium contents of fast, moderate and slow cooking yellow or red mottled beans from either the Puerto Rico or the Tanzania location (P>0.05 ). Kyriakidis an d others (1997) reported that beans cooked in divalent cation solutions (CaCO 3 and MgCO 3 ) had a longer cooking time than those boiled in plain water. This could imply that the total inherent calcium and magnesium contents of dry beans is not associated wit h cooking time of beans, but rather only those found in cooking solutions. Since these minerals were proved to play a role in the HTC phenomenon ( Kyriakidis and others, 1997), perhaps it is about how much of these minerals are available for a chemical reac tion. Fast and slow cooking beans might inherently have similar amounts of these minerals, but perhaps the slow cooking beans release more divalent cations during phytic acid hydrolysis which in turn participate in crosslinking of pectin molecules, as comp ared to the fast cooking beans that might be releasing fewer of these divalent ions. Growing location did not have an effect on calcium and magnesium contents of the beans according to the findings in the current study. These findings are in agreement with Muyonga and others (2008) who reported that calcium and magnesium contents of common beans grown in different locations (Kabale, Lira and Naari) were not significantly different (P>0.05). 57 Table 5 . Calcium and magnesium content s of whole seeds of fast, moderate , and slow cooking yellow and red mottled common dry beans from Puerto Rico and Tanzania locations 1 Bean Sample 2 Calcium (% by weight ) Magnesium (% by weight ) Puerto Rico Tanzania Puerto Rico Tanzania Yellow Y - 52 1 - F 0.188±0.026 b 0.15 0 ±0.02 0 b 0.2 06 ±0.00 8 a 0.2 06 ±0.01 3 a Y - 111 - M 0.1 17±0.004 b 0.11 4 ±0.00 1 b 0.2 09 ±0.00 1 a 0.1 8 9±0.00 1 a Y - 513 - S 0.155 ±0.00 7 b 0.12 4 ±0.00 9 b 0.17 4 ±0.00 4 a 0.17 3 ±0.00 4 a Red Mottled R - 443 - F 0.15 4 ±0.00 6 b 0.089 ±0.00 2 b 0.22 0 ±0.00 4 a 0.20 0 ±0. 00 4 a R - 436 - M 0.135 ±0.00 7 b 0.1 5 6±0.00 8 b 0.20 0 ±0.00 3 a 0.19 0 ±0.00 3 a R - 434 - S 0.1 37 ±0.01 8 b 0.1 4 5±0.00 2 b 0.19 2 ±0.01 1 a 0.17 2 ±0.00 2 a 1 Data represent means and standard deviations of n=2 4 . Values within the same column or row followed by the same superscript 2 F: fast - cooking genotype; M: moderate - cooking genotype; S: slow - cooking genotype. 58 Therefore inherent calcium and magnesium contents in beans are n ot associated with cooking time of t he six common beans studied here . Perhaps the distribution of these minerals in the bean cells might be different for the fast and slow cooking beans. Further research should be conducted to investigate this. 59 4.5 . Protein content The protein content s of the studied raw common beans are presented in Table 6 . They ranged between 20.17 and 25.97 g/100 g of raw beans . T he protein content s of the fast, moderate, and slow cooking yellow beans were not significantly different from each other (P> 0.05) for both Puerto Rico and Tanzania locations. However, genotypes Y - 111 - M and Y - 513 - S from Puerto Rico had 4 g and 2.3 g , respectively , greater protein content than their counterparts from Tanzania. This could be attributed to the fertility of the soil where they we re grown. Except for the genotype R - 443 - F, t he protein content s of all the red mottled genotypes were not significantly different from each other (P> 0.05) for both Puerto Rico and Tanzania locations. The fast - cooking genotype R - 443 - F from Puerto Rico had a bout 3 g /100 g of protein greater than the slow - cooking genotypes. There was also no correlation between protein content and cooking time. This however does not necessarily imply that protein has no association with cooking time of beans. Perhaps the struc ture of the proteins in the fast and slow cooking beans could be different. Further research should be done to study the structure of protein in these beans. 60 Table 6 . Protein content of fast, moderate , and slow cooking yellow a nd red mottled common dry bean varieties from Puerto Rico and Tanzania locations 1 1 Data represent means and standard deviations of n=2 4 . Values within the same column and row followed by the same superscript 2 F: fast - cooking genotype; M: moderate - cooking genotype; S: slow - cooking genotype. Common Bean V arieties 2 Protein C ontent (%) P u e rto Rico Tanzania Yellow Y - 521 - F 22.34±0.04 fb 21.67±0.05 b Y - 111 - M 24.15±0.23 fe 20.17±0.14 b Y - 513 - S 23.57±0.16 f 21.28±0 .36 b Red mottled R - 443 - F 25.97±0.39 e 23.82±0.17 f R - 436 - M 23.03±0.04 f 23.42±0.12 f R - 434 - S 23.29±0.03 f 22.59±0.03 fb 61 4.6 . Starch content The starch content s of the studied raw common be an genotypes ranged between 22. 8 and 39.75 g/100 g ( Table 7 ) . There was a significant difference among the starch contents of fast, moderate and slowing - cooking yellow bean genotypes (P<0.05) from Puerto Rico and Tanzania. As expected, s tarch content increased with increase in cooking t ime for all yellow genotypes in both locations except genotype Y - 521 - F from Puerto Rico that had 7 g and 5 g higher starch content than genotypes Y - 111 - M and Y - 513 - S, respectively. T he slow - cooking genotype Y - 531 - S from Tanzania had 4.3 g greater starch th an the fast - cooking genotype Y - 521 - F. The genotypes Y - 111 - M and Y - 513 - S from Tanzania had 27% and 26%, respectively, greater starch content than their counterparts from Puerto Rico. Like yellow beans, starch content of fast, moderate and slow - cooking red mottled beans was significantly different from each other for both locations (P<0.05) with fast - cooking beans having the lowest starch content and the slow - cooking beans having the highest content. Like yellow beans, the starch content of red mottled beans from Tanzania was significantly higher than their counterparts from Puerto Rico however the differences varied according to genotypes. The genotypes R - 443 - F, R - 436 - M, and R - 434 - S from Tanzania had 8 g, 10 g, and 9 g, respectively , greater starch content t han their counterparts from Puerto Rico. There was a 41.7% positive correlation (P<0.05) between starch content and cooking time of raw beans (Table 4), which implies that cooking time of beans was associated with the ir starch content. These results are i n agreement with Reyes - Moreno and others (1994 ) who reported tha t the fast cooking common bean variety Michigan 800 had a lower starch content (24 .0 g/ 100 g) than that 62 of the slow cooking bean variety Ojo de Cabra ( 26 .0 g/ 100 g ), thus starch content could be associated with cooking time . Regardless of the positive correlation with cooking time, the present study also showed that starch content did not make a significant contribution to cooking time (beta coefficient = - .150 ) (P> 0.05) in this model. This co uld have been due to the small number of samples used in the present study. Further research should be conducted using a larger number of samples , and perhaps the different components of starch and the starch : protein ratio c ould be studied in relation to cooking time. 63 Table 7 . Starch content s of raw fast, moderate , and slow cooking yellow and red mottled dry common beans from Puerto Rico and Tanzania locations 1 1 Data represent means and sta ndard deviations of n=2 4 . Values within the same column or row 2 F: fast - cooking genotype; M: moderate - cooking genotype; S: slow - cooking genotype. Common Bean V arieties 2 Starch C ontent (%) P u e rto Rico Tanzania Yellow Y - 521 - F 34.30±1.23 b c 35.48±1.89 b c Y - 111 - M 26.84±0.99 e 36.59±1.29 c Y - 513 - S 29.31±0.06 f 39.75±4.32 c Red mottled R - 443 - F 25.31±0.95 e 33.44±3.17 b R - 436 - M 22.84±0.06 a 32.94±0.45 b R - 434 - S 27.1 0 ±0.44 e 36.01±2.57 c 64 4.7. Hot water soluble p ectin ( HWSP ) content Total pectin content and its two fractions for the studied raw common beans are presented in Table 8 . All bean genotypes for both market classes from both locations had less HWSP than HWIP. The HWSP contents of fast, modera te and slow cooking y ellow beans from Puerto Rico were significantly different (P< 0.05) , with moderate cooking beans having the highest HWSP . The HWSP content of Y - 521 - F was about 2.4 - fold greater than the slow cooking beans with the lowest HWSP value. HWSP content of fast, moderate, and slow - cooking yellow from Tanzania location were not significantly different from each other (P> 0.05) . HWSP contents of red mottled beans from Puerto Rico were not significantly different from each other (P>0.05) . HWSP content of fast, moder ate, and slow - cooking red mottled be ans from Puerto Rico and Tanzania location were not significantly different from each other except for the fast - cooking red mottled bean genotype, R - 443 - F, that was significantly greater than all of them and had about 2.5 - fold greater HWSP content ( 14.78±0 .62 mg/g) than the slow - cooking red mottled bean genotype (5.94±1.26 mg/g ). There was a 48.4 % (P< 0.05) negative correlation between HWSP and cooking time of raw beans (Table 4), which implies that cooking time of beans decreased with increase in HWSP. Fur thermore, HWSP made the greatest contribution to cooking time (beta coefficient = 3.548) (P<0.05) when the variance explained by all other variables in the model is controlled for . These results were in agreement with Bernal - Lugo and others (1996) and Njo roge and others (2014) who reported that fast cooking bean ( variety Rose coco ) had more hot water soluble pectin 65 ( 8.44 mg/g) than the slow cooking beans ( variety Pinto , 5.51 mg/g ) . Further res earch with a large sample size should to be conducted to verify t hese findings. 66 Table 8 . Hot water soluble pectin (HWSP), hot wa ter insoluble pectin (HWIP), and tot al pectin (Total ) of fast (F) , moderate (M), and slow (S) cooking common bean varieties of yellow (Y) and red mottled (R) beans grown in Puerto Rico and Tanzania locations 1 1 Data represent means and standard deviations of n=2 4 . Values within the same column or row followe 2 F: fast - cooking genotype; M: moderate - cooking genotype; S: slow - cooking genotype. Bean Sample 2 Puerto Rico Tanzania HWSP (mg/g) HWIP (mg/g) Total (mg/g) HWSP (mg/g) HWIP (mg/g) Total(mg/g) Yellow Y - 521 - F 11.69± 0.23 b 14.5 0 ±2.19 e 26.18±0.41 h 6.13±0.45 a 12.78±0.82 de 18.90±0.12 fg Y - 111 - M 16.54±1.20 c 16.86±0.11 e 33.39±1.7i 5.23±0.52 a 14.40±1.26 e 19.63±0.68 g Y - 513 - S 6.91±0.80 a 13.77±1.28 e 20.69±0.27 g 6.63±1.25 a 9.99±0.29 d 16.63±0.98 fg Red Mottled R - 443 - F 8.62±0.51 a 14.43±2.43 e 23.05±4.13 g 14.78±0.62 b 11.76±0.85 de 26.54±0.19 h R - 436 - M 6.45±0.87 a 17.60±1.22 e 24.06±1.39 gh 7.11±0.90 a 13.44±0.87 e 20.51±2.05 g R - 434 - S 7.79±0.74 a 12.13±0.67 de 19.90±2.35 g 5.94±1.26 a 9.98±1.78 d 15.93±1.33 f 67 4. 8 . Hot water insoluble p ectin ( HWIP ) content The HWIP of the studied r aw beans ranged between 9.98±1.78 mg/g and 17.60±1.22 mg/g (Table 8 ). Except for R - 443 - F, each of the bean varieties of both yellow and red mottled beans from Puerto Rico and Tanzania had a greater HW I P content than HWSP content . The HWIP contents of fast, moderat e and slow - cooki ng beans of both market classes within one location were generally not significantly different from each other (P>0.05). This could mean that HWIP is formed at a more uniform rate during bean growth, and these beans were all about t therefore they had similar amounts of HWIP. If some had been stored longer, then perhaps they wou ld have higher HWIP contents. The re was a slight negative correlation 34.6% (P>0.05) between HWIP and cooking time (Table 4 ) . This implies tha t cooking time increased with decrease in HWIP. These results are not in line with the findings of Bernal - Lugo and others (1996) who reported that the slow cooking beans had a higher HWIP pectin content than fast cooking beans. Some pectin molecules are ti ghtly bound to other polysaccharides like hemicellulose, wall proteins and phenolic compounds (Caffall and Mohnen, 2009). Perhaps most of the HWIP in the hard to cook beans in the present study were tightly bound to other polysaccharides like hemicellulose , wall proteins and phenolic compounds making it hard to be extracted. Further research with a larger sample size should be conducted to verify these findings. 68 4. 9 . Total Pectin The total pectin of the studied raw beans ranged between 15.92 mg/g and 33. 39 mg/g ( Table 8 ). The total pectin contents of yellow beans from Puerto Rico were significantly different with Y - 111 - M having the highest pectin and Y - 513 having the lowest pectin content. On the contrary, TP of yellow beans from Tanzania was not signific antly different. Also, TP of yellow beans from Puerto Rico were slightly higher than their counterparts from Tanzania. The red mottled fast, moderate and slow cookin g beans from Puerto Rico were not significantly different (P>0.05) from each other . On the contrary, the red mottled fast, moderate and slow cooking beans from Tanzania were significantly different (P< 0.05) from each other. The genotype R - 443 - F had 5 mg and 10 mg higher TP content than genotypes R - 436 - M and R - 434 - S, respectively. For both marke t classes in both locations, fast cooking beans were associated with a high total pectin content. T here was a 66.8 % (P< 0.05) negative correlation betwee n cooking time and total pectin (Table 4 ) and it made the largest contribution to cooking time (beta coe fficient = 3.987) (P<0.05). Since the largest composition of TP is HWIP, it could mean that most of the pectin polyme rs in the slow cooking beans are tightly bound to the c ell wall polymers like hemicellulose , phenolic compounds and wall proteins making i t h ard to extract enough of it. These results also imply that no matter the amount of pectin present, its soluble fraction is what is associated with the cooking time . Total pectin in this study is the sum of HWSP and HWIP so the strong negative correlatio n of TP with cooking time could be largely influenced by HWSP which has been reported in this study to also have a strong negative correlation with cooking time. 69 4. 10 . Starch gelatinization temperature Starch gelatinization temperatures of the studied r aw beans are presented in Table 9 . Except for genotype R - 443 - F from Puerto Rico, there were no significant differences in Tg (P>0.05) among fast, moderate and slow cooking yellow and red mottled beans from Puerto Rico and Tanzania. The Tg of genotype R - 443 - F from Puerto Rico was at least 10 o C lower than genotypes R - 436 - M and R - 434 - S. Except for R - 443 - F, each of the bean genotypes of both yellow and red mottled beans from Puerto Rico had 5 - 6 o C higher gelatinization temperature (Tg) than its counterpart from Tanzania . This suggests that location could have an influence on the gelatinization temperatures of these dry bean varieties. These results are in agreement with Bernal - Lugo and others (1996) who reported that both fast and slow cooking beans had simila r gelatinization temperatures (86 o C and 87 o C, respectively) and therefore that Tg did not have much contribution to cooking time. Their Tg values were 4 to 6 o C higher than Tg values in the present study probably because of the environmental difference in w hich the beans were grown. Since the beans had similar gelatinization temperatures, it can be concluded that Tg was not associated with their cooking time. 70 Table 9 . Starch gelatinization and protein denaturation temperatures of d ry common bean varieties of yellow and red mottled beans from Puerto Rico and Tanzania locations 1 Bean Variety 2 T g 3 ( o C) Td 4 ( o C) Puerto Rico Tanzania Puerto Rico Tanzania Yellow Y - 521 - F 82.23±0.05 e 76.97 ±0.50 a 94.26±0.24 t 92.07 ±1.23 r Y - 111 - M 80.46 ±0.60 e 75.78 ±1.29 a 92.97±1.33 r 92.65 ±0.79 r Y - 513 - S 81.85±0.41 e 76.58±0.12 a 94.34±0.22 t 91.06 ±1.41 r Red Mottled R - 443 - F 70.29 ±0.07 d 73.8 0 ±0.37 a 98.35 ±0.17 t s 96.63±0.25 t R - 436 - M 81.34 ±0.12 e 75.79 ±0.42 a 95.06±1.30 t 92.43±0.01 r R - 434 - S 81 .38±0.47 e 75.75±0.14 a 94.37±2.07 r t 92.62 ±0.26 r 1 Data represent means and standard deviations of n=2 4 . Values within the same column or ro w 2 F: fast - cook ing genotype; M: moderate - cooking genotype; S: slow - cooking genotype. 3 Tg = starch gelatinization temperature, 4 Td = protein denaturation temperature 71 4. 11 . Protein denaturation temperature Protein denaturation (Td) temperatures of the studied raw beans are presented in Table 9. There were no significant differences in Td among fast, moderate and slow cooking yellow beans within the Puerto Rico location and within the Tanzania location (P>0.05). For red mottled beans, with the exception of R - 443 - F, ther e were also no significant differences in Td among fast, moderate and slow cooking yellow beans within the Puerto Rico location and within the Tanzania location (P>0.05). R - 443 - F had at least 3 o C higher Td than other varieties in Puerto Rico. For both yell ow and red mottled beans, each genotype grown in Puerto Rico had a slightly significant higher Td (2 - 3 o C) than its counterpart grown in Tanzania. Moreno and others (1994) reported that fast cooking beans (variety Michigan 800) had a lower protein denatura tion temperature (88.6±0.8 o C) than that of slow cooking bean (variety Ojo de Cabra, 95.3±0.7 o C), and indicated that this affected the cooking time of beans. On the contrary, Bernal - Lugo and others (1996) reported that both fast and slow cooking beans had s imilar protein denaturation temperatures (103 o C and 104 o C, respectively) and that therefore denaturation temperature does not contribute to cooking time. The results of this thesis, in Table 9, are in agreement with Bernal - Lugo and others (1996) although t he Td values for their bean varieties were at least 11 o C higher than those of the beans in the present study. The results imply that Td did not have an association with cooking time of beans. Based on results in the present study (Table 9), a difference in growing location had a slight effect on protein denaturation temperature of most of the counterpart varieties. Due to a limited number of bean varieties, further tests are needed from more different locations to verify the findings. 72 CHAPTER 5 CONCLUSION S AND RECOMMENDATIONS The findings of this study showed that yellow beans cook ed slightly faster than red mottled beans and location had no effect on the cooking time trend. There was no significant difference in cooking time s of soaked and dehulled beans , therefore soaking is as good as dehulling . S eed coat thickness and starch content were positively correlated with cooking time. H ot water soluble pectin and total pectin content were negatively correlated to cooking time . Starch gelatinization temperatur e, protein denaturatio n temperature, protein content, and ca lcium and magnesium contents each w as not associated with cooking time of beans. Households should always soak dry common beans before cooking to shorten cooking time. Also, s electing for bean va rieties with a thinner seed coat would be a good method of reducing coo king time of beans ; however, the seed coat is a protective mechanism against pests , and so may be selecting for thin but tough seed coats would result into varieties with a shorter cooki ng time. Further investigation on the structure of the studied components of fast and slow cooking common beans will provide more knowledge towards understanding the physicochemical properties influencing cooking time of dry common beans. Also larger numbe r of bean genotypes and market classes should be used . 73 APPENDIX 74 Figure 11 . T he peak gelatinization (first peak) an d peak denaturation (second peak) temperature s of dry common bean s from Tanzania. 75 Figure 12 . T he peak gelatinization (first peak) and peak denaturation (second peak) temperatures of dry common beans from Puerto Rico. 76 Table 10 . Correlation coefficients among physicochemical properties of dry common beans : cooking time (CT ) protein content (PC), seed coat thickness (SCT), starch content (SC), hot water soluble pectin ( HWSP ) , hot water insol uble pectin (HW IP ) and total pectin (TP ) for dry common beans Correlations CT PC SCT SC HWSP HWI SP TP Pearson Correlation CT 1.000 - .226 .743 .417 - .484 - .346 - .668 PC - .226 1.000 - .144 - .605 .423 .172 .470 SCT .743 - .144 1.000 .288 - .372 - .567 - .66 3 SC .417 - .605 .288 1.000 - .573 .065 - .514 HWSP - .484 .423 - .372 - .573 1.000 - .184 .816 HWI P - .346 .172 - .567 .065 - .184 1.000 .410 TP - .668 .470 - .663 - .514 .816 .410 1.000 Sig. (1 - tailed) CT . .144 .000 .021 .008 .049 .000 PC .144 . .252 .001 .020 .210 .010 SCT .000 .252 . .086 .037 .002 .000 SC .021 .001 .086 . .002 .381 .005 HWSP .008 .020 .037 .002 . .195 .000 HWI P .049 .210 .002 .381 .195 . .023 TP .000 .010 .000 .005 .000 .023 . N CT 24 24 24 24 24 24 24 PC 24 24 24 24 24 24 24 SCT 24 24 24 24 24 24 24 SC 24 24 24 24 24 24 24 HWSP 24 24 24 24 24 24 24 HWI P 24 24 24 24 24 24 24 TP 24 24 24 24 24 24 24 77 Figure 13 . Cooking time of raw tempered yellow (Y) and red mottled beans (R) from Puerto Rico (PR) and Tanzania (TZ) that were raw (not soaked or dehulled), soaked and dehulled before cooking. Data represent means and standard deviations. Values of each bar topped by the same other. 0 20 40 60 80 100 120 140 160 180 Raw (PR) Soaked(PR) Dehulled (PR) Raw (TZ) Soaked (TZ) Dehulled(TZ) Cooking time (minutes) Common bean varieties Y-521-F Y-111-M Y-513-S R-443-F R-436-M R-434-S a b c f g y d d e d e m l n d d e d e o p q s x z f v y h p h p h u s r k p p h s s r 78 Figure 14 . Hot water soluble pectin (HWSP), hot wa ter insoluble pectin (HWIP), and tot al pectin (Total ) of six fast (F) , moderate (M), and slow (S) cooking common bean varieties of yellow (Y) and red mottled (R) beans grown in Puerto Rico and Tanzania locations . Data represent means and standard deviations of n=2 . Values of each bar topped by the same letter are significantly 0 5 10 15 20 25 30 35 40 HWSP(PR) HWIP(PR) Total (PR) HWSP(TZ) HWIP(TZ) Total (TZ) Pectin content (mg/g) Y-521-F Y-111-M Y-513-S R-443-F R-436-M R-434-S b c a a a a e e e e e d e h i g g g h g a a a b a a d e d e d e e d f g g f h g f 79 REFERENCES 80 REFERENCES Agbo, G. N., Hosfield, G. L., Uebersax, M. A., & Klomparens, K. (1987). Seed microstructure and its relationship to water uptake in isogenic lines and a cultivar of dry beans (Phaseolus vulgaris L.). Food Microstructure , 6 , 91 - 102. Aguilera, J. M. & Stanley, D. W. (1985). A revie w of textural defects in cooked reconstituted legumes - The influence of storage and processing. J. Food Process. Preserve , 9, 145. Aguilera, J., & Rivera, R. (1992). Hard - to - cook defect in black beans: hardening rates, water imbibition and multiple mechanism hypothesis. Food research international , 25 (2), 101 - 108 . Bellido, G. G., Arntfield, S. D., Scanlon, M. G., & Cenkowski, S. (2003). The effect of micronization operational conditions on the physicochemical properties of navy beans ( Phaseolus vulgaris L.). Journal of food science , 68 (5), 1731 - 1735. Bernal - Lugo, I., Parra, C., Porti lla, M., Peña - Valdivia, C. B., & Moreno, E. (1997). Cotyledon thermal behavior and pectic solubility as related to cooking quality in common beans. Plant foods for human nutrition , 50 (2), 141 - 150. Bhatty, R. S. (1995). Lentils as a dietary cereal complement. Cereal foods world , 40 (5), 387 - 392. Bhatty, R. S., & Slinkard, A. E. (1989). Relationship between phy tic acid and cooking quality in lentil. Canadian Institute of Food Science and Technology Journal , 22 (2), 137 - 142. Bressani, R., & Elías, L. G. (1980). The nutritional role of polyphenols in beans. Cereal Chemistry , 87 (4), 342 352. Caffall, K. H., & Mohnen, D. (2009). The structure, function, and biosynthesis of plant cell wall pectic polysaccharides. Carbohydrate Research, 344, 1879 - 1900. Chang, R., Schwimmer, S., & Burr, H. K. (1977). Phytate: removal from whole dry beans by enzymatic hydrolysis and diffusion. Journal of Food Science , 42 (4), 1098 - 1101. Chibbar, R. N., Ambigaipal an, P., & Hoover, R. (2010). Review: Molecular Diversity in Pulse . Cichy, K. A., J. A. Wiesinger & F. A. Mendoza (2015). "Genetic diversity and genome - wide association analysis of cooking time in dry bean (Phaseolus vulgaris L.)." Theor Appl Genet 128(8): 1555 - 1 567. Chitra, U. (1994 ). Effect of storage and processing on phytic acid levels in legumes and its interference with the utilization of protein and iron . Andhra Pradesh Agricultural University . 81 Chung, H. J., Shin, D. H., & Lim, S. T. (2008). In vitro starch diges tibility and estimated glycemic index of chemically modified corn starches. Food Research International , 41 (6), 579 - 585. Coelho, S. R. M., Prudêncio, S. H., Christ, D., & Sampaio, S. C. (2011). Storage - induced ch anges in protein interactions and protein solubility in common beans. Journal of Food, Agriculture & Environment , 9 (3&4), 44 - 48. Curtis, H. (1991). The Effect of Soaking and Pectin Composition on the Hardness of Dried Legumes (phaseolus Vulga ris) (Doctoral dissertation, Brigh am Young University. Department of Food and Science and Nutrition.). De Leo n, L. F., Elias, L. G., & Bressani, R. (1992). Effect of salt solutions on the cooking time, nutritional and sensory characteristics of common beans ( Phaseolus vulgaris ). Food Research International, 25, 131±136. Del Valle, J. M., & Stanley, D. W. (1995). Reversible and irreversible components of bean hardening. Food research international , 28 (5), 455 - 463. D iaz, A.M.; Caldas, G.V.; Blair, M.W. Concentrations of condensed tannins and anthocyanins in common bean seed coats. Food Res. Int. 2010 , 43 , 595 601. El Tabey Shehata, A. M. (1992). Hard to cook phenomenon in legumes. Food Reviews International , 8 (2), 191 - 221. Fasina, O., Tyler, B., Pickard, M., Zheng, G. H., & Wang, N. (2001). Effect of infrared heating on the properties of legume seeds*. I nternational journal of food science & technology , 36 (1), 79 - 90. Garcia, E., Filisetti, T. M., Udaeta, J. E., & Lajolo, F. M. (1998). Hard - to - cook beans (Phaseolus vulgaris): involvement of phenolic compounds and pectates. Journal of Agricult ural and Food Chemistry , 46 (6), 2110 - 2116. Guillon, F., & Champ, M. J. (2002). Carbohydrate fractions of legumes: uses in human nutrition and potential for health. British Journal of Nutrition , 88 (S3), 293 - 306. Hentges, D. L., Weaver, C. M., & Nielsen, S. S. (1991). Changes of Selected Physical and Chemical Components in the Development of the Hard to Cook Bean Defect. Journal of Food Science , 56 (2), 436 - 442. Hincks, M. J., & Stanley, D. W. (1987). Lignification: evidence for a role in hard to cook beans. Journal of Food Biochemistry , 11 (1), 41 - 58. Jane, J. L. (2000). Carbohydrates: basic concepts. Food chemistry: Principles and applications . California : Science technology systems , 43 . 82 Jane, J. L. (2012). Carbohydrates: basic concepts. Food chemistry: Principles and applications. California : Science technology systems , 43 . Jane, J., Chen, Y., Lee, L., McPherson, E., Wong, K., Radosavlijevic, M., & Kasemsuwan, T. (1999). Effects of a mylopectin branch chain length and amylose content on the morphology, molecular structure, and physicochemical properties of starches from newly released chickpea (Cicer arietinum L.) cultivars grown in Canada. Food Research Jones, P. M. B., & Boulter, D. (1983). The cause of reduced cooking rate in Phaseolus vulgaris following adverse storage conditions. Journal of Food Science , 48 (2), 623 - 626. Journal of Cereal Science, 39, 151 - 165. Kadam S.S., Deshpande S.S., & Jambhale D.N. (1 991). Seed structure and composition. In: Ng, P. K. W. Book Review -- Handbook of World Food Legumes: Nutritional Chemistry, Processing Technology, and Utilization (Vol. I ) . Kaur, A., Kaur, P., Singh, N., Virdi, A. S., Singh, P., & Rana, J. C. (2013) . Grains, starch and protein characteristics of rice bean (Vigna umbellata) grown in Indian Himalaya regions. Food Research International , 54 (1), 102 110. Ken, S., Brown. A.H., Ohanneson. J.G., and Booth, A.N. 1973.Split peeled bean s: Preparation and some properties. J. Food Sci .38: 496. Kon, S., & Sanshuck, D. W. (1981). Phytate content and its effect on cooking quality of beans. Journal of Food Processing and Preservation , 5 (3), 169 - 178. fiber content of dry and processed beans. Food chemistry , 80 (2), 231 - 235. Kyriakidis, N. B., Apostolidis, A., Papazoglou, L. E., & Karathanos, V. T. (1997). Physicochemical Studies of Hard to Cook Beans (Phaseolus vulgaris). Journal of the Science of Food and Agriculture , 74 (2), 186 - 192. L iu, K., Shewfelt, R. L., Phillips, R. D., & Hung, Y. C. (1992). Hard to cook state in cowpeas influence of pretreatment and cooking on electr olyte leakage solids loss and water absorption. International journal of food science & technology , 27 (6), 683 - 690. Martínez Manrique, E., Jacinto Hernández, C., Garza García, R., Campos, A., Moreno, E., & Bernal Lugo, I. (2011). Enzymatic changes in pect ic polysaccharides related to the beneficial effect of soaking on bean cooking time. Journal of the Science of Food and Agriculture , 91 (13), 2394 - 2398. Maryange, M., Msolla, N., Sawargaonkar, S. L., Hudge, B. V., & Thanki, H. P. (2010).Screeni ng of 30 advanced common bean (P. vulgaris, L) lines for short cooking time using two different methods. Electronic Journal of Plant Breeding , 1 (4), 505 - 511. 83 Mkanda, A. V., Minnaar, A., & De Kock, H. L. (2007). Relating consumer preferences t o sensory and physicochemical properties of dry beans (Phaseolus vulgaris). Journal of the Science of Food and Agriculture , 87 (15), 2868 - 2879. Mohan, J. R, Sangeetha, A., Narasimha, V. H and Tiwari, K. B . (2011).Chapter 7:Post - harvest tec hnology of pulses. In Tiwari, K. B., Gowen, A. and Mckenna, B. (2011).Pulse Foods: Processing, Quality and Nutraceutical Applications. Mohnen, D. (2008). Pectin structure and biosynthesis. Current Opinion in Plant Biology, 11, 266 - 277. Mosco so, W., Bourne, M. C., & Hoo d, L. F. (1984). Relationships b etween the Hard to Cook Phenomenon in Red Kidney Beans and Water Absorption, Puncture Force, Pectin, Phytic Acid, and Minerals. Journal of Food Science , 49 (6), 1577 - 1583. Mwangwela , A. M., Waniska, R. D., & Minnaa r, A. (2007). Effect of microniz ation temperature (130 and 170 C) on functional properties of cowpea flour. Food chemistry , 104 (2), 650 - 657. Nasar - Abbas, S. M., Plummer, J. A., Siddique, K. H., White, P., H arris, D., & Dods, K. (2008). Cooking quality of faba bean after storage at high temperature and the role of lignins and other phenolics in bean hardening. LWT - Food Science and Technology , 41 (7), 1260 - 1267. Njoroge, M. D., Kinyanjui, K. P., Makokha, O. A., Christiaens, S., Shpigelman, A., Sila, N. D., & Hendrickx, E. M. (2014). Extraction and characterization of pect i c polysaccharides from easy - and hard - to - cook common beans (Phaseolus vulgaris). Food Research International 64, 314 - 322 Nozzolillo , C., & Bezada, M. D . (1984). Browning of lentil seeds, concomitant loss of viability, and the possible role of soluble tannins in both phenomena. Canadian journal of plant science , 64 (4), 815 - 824. Ovando - Martínez, M., Bello - Pérez, L. A., Whitney, K ., Osorio - Díaz, P., & Simsek, S. (2011). Starch characteristics of bean (Phaseolus vulgaris L.) grown in different localities. Carbohydrate polymers , 85 (1), 54 - 64. Petry, N., Boy, E., Wirth, J. P., & Hurrell, R. F. (2015). Revi ew: The Potentia l of the Common Bean (Phaseolus vulgaris) as a Vehicle for Iron Bio fortification . Nutrients , 7 (2), 1144 - 1173. Pirhayati, M., Soltanizadeh, N., & Kadivar, M. (2011). Chemical and microstructural to cook phenomenon in legumes (pin to bean and small type lentil). International Journal of Food Science & Technology , 46 (9), 1884 - 1890. 84 Reddy, N.R.; Pierson, M.D.; Sathe, S.K.; Salunkhe, D.K. Dry bean tannins: A review of nutritional implications. J. Am. Oil Chem. Soc. 1985 , 62 , 541 549. Reyes - Moreno, C., Cárabez - Trejo, A., Paredes - López, O., & Ordorica - Falomir, C. (1994). Physicochemical and structural properties of two bean varieties which differ in cooking time and the HTC characteristic. LWT - Food Science and Techno logy , 27 (4), 331 - 336. Reyes Moreno, C., Paredes López, O., & Gonzalez, E. (1993). Hard to cook phenomenon in common beans A review. Critical Reviews in Food Science & Nutrition , 33 (3), 227 - 286. Reyes Moreno, C., Rouzaud Sandez, O., Milán Car rillo, J., Garzón Tiznado, J. A., & Camacho Hernández, L. (2001). Hard to cook tendency of chickpea (Cicer arietinum L) varieties. Journal of the Science of Food and Agriculture , 81 (10), 1008 - 1012. Salunkhe, D. K., & Kadam, S. S. (1989). Handb ook of world food legumes: nutritional chemistry, processing technology, and utilization. Volume 1 . CRC Press, Inc. Salvador, B. D. V. (2007). Improvement in the cooking and physico - chemical characteristics of hard - to - cook cowpeas by pre - condition ing and micronization (Doctoral dissertation, University of Pretoria). Sathe, S.K. and Deshpande, S.S. (2003). Beans. Encyclopedia of Food Science and Nutrition, 2 nd Ed. Vol.1.Pp 403 - 412. Sefa - Dedeh, S., & Stanley, D. W. (1979a). The relationship of microstructure of cowpeas to water absorption and dehulling properties. Cereal Chem , 56 (4), 379 - 386. Sefa - Dedeh, S., & Stanley, D. W. (1979b). The relationship of microstructure of cowpeas to water absorption and dehulling properties. Cereal Che m , 56 (4), 379 - 386. Shiga, T. M., Cordenunsi, B. R., & Lajolo, F. M . (2006). Cell wall polysaccharides of common beans ( Phaseolus vulgaris L.) composition and structure. Carbohydrate Polym ers 63 ( 1 )1 - 12 Shiga, T. M., Cordenunsi, B. R., & Lajo lo, F. M. (2009). Effect of cooking on non - starch polysaccharides of hard - to - cook beans. Carbohydrate polymers , 76 (1), 100 - 109. Shiga, T. M., Cordenunsi, B. R., & Lajolo, F. M. (2011). The effect of storage on the solubilization pattern of bean hull non - starch polysaccharides. Carbohydrate Polymers , 83 (2), 362 - 367. Singh, U. (1999). Cooking quality of pulses. J. Food Sci. Technol. Mysore 36, 1 14. Stanley, D. W. & Aguilera, J. M. (1985). A review of textural defects in cooked rec onstituted legumes ~ the influence of structure and composition. J. Food Biochem. 9. 277 - 323. 85 Stanley, D. W. (1992). A possible role for condensed tannins in bean hardening. Food Research International, 25, 187 - 92. Stanley, D. W., Wu, X., & Plhak, L. C. (1989). Seed coat effects in cooked reconstituted bean texture. Journal of texture studies , 20 (4), 419 - 429. Taiwo, K. A., Akanbi, C. T., & Ajibola, O. O. (1998). Regression relationships for the soaking and cooking properties of two cowpea varieties. Journal of food engineering , 37 (3), 331 - 344. Tester, R., Karkalas, J., & Qi, X. (2004). Starch composition, fine structure and architecture. Tiwari, B. K., Gowen, A., & McKenna, B. (Eds.). (2011). Pulse foods: proce ssing, quality and nutraceutical applications . Academic Press. Valle, J., Stanley, D. W., & Bourne, M. C. (1992). Water absorption and swelling in dry bean seeds. Journal of Food Processing and Preservation , 16 (2), 75 - 98. Va n Laar, H., Tamminga, S., Williams, B. A., Verstegen, M. W., & Engels, F. M. (1999). Fermentation characteristics of cell - wall sugars from soya bean meal, and from separated endosperm and hulls of soya beans. Animal feed science and technology , 79 (3), 179 - 193. Vindiola C, O. L. (1986). Accelerated development of the hard - to - cook state in beans (Master's thesis, Kansas State University). Wang, N., Hatcher, D. W., & Gawalko, E. J. (2008). Effe ct of variety and processing of nutrients and certain anti - nutrients in field peas (Pisum sativum). Food Chemistry , 111 (1), 132 - 138. Wang, N., Hatcher, D. W., Toews, R., & Gawalko, E. J. ( 2009). Influence of cooking and dehulling on nutritional composition of several varieties of lentils (Lens culinaris). LW T - Food Science and Technology , 42 (4), 842 - 848. Wang N . , Daun J . K . (2005) . Determination of cooking times of pulses using an automated Mattson cooker apparatus. J ournal of Sci Food Agric 85:1631 1635 .