The non-linear regression procedure required that ini¬tial parameter estimates be chosen close to the true value. These estimates were obtained by linearization of Eq. (10) through logarithmic transformation and ap-' plication of linear regression analysis. The linear least square estimates were then used as the initial esti¬mates in the non-linear regression procedure. The values of coefficient a and b obtained were 2.85 with a stan¬dard error of 0.162 and 47.592 with a standard error of 1.645 respectively for melon seed, and 3.3446 with a standard error of 1.065 and 10.9569 with a standard error of 4.54 for cassava. The standard error of the estimate of L/h(g was 0.0152 for melon seed and 0.39 for cassava.
Eq. (10) becomes
for melon seed:
L/h(a = 1 +2.85exp(-47.592M) (11)
and for cassava:
L/h(i, = 1 + 3.3446 exp(-10.9569M) (12)
Eqs. (11) and (12) were used to compute the values of L/hf, at different moisture contents and these values were plotted against moisture content. The monolayer moisture contents of melon seed and cassava at different temperatures were obtained by applying the BET equation to the desorption equilibrium moisture content data for the products in the water activity range of 0.05 0.45. A plot of aw/(l - aw)M versus aw yielded a straight line with slope and intercept on the K-axis from which Mm was calculated. The spreading pressures of the products at different temperatures were calculated using Eq. (4). The value of spreading pressure was noted to be indeterminate at aw = 0, therefore, the lower limit em¬ployed was aw = 0.05. The computed value of spreading pressure was adjusted by adding the values corre¬sponding to the interval aw = 0 to aw = 0.05 which was calculated by assuming~that a linear relationship (Hen¬ry's law) exists between aw and 0 within the water ac¬tivity range (Fasina et al., 1999). For the interval aw = 0 to aw = 0.05, Eq. (4), therefore becomes
, KT0
The spreading pressure was plotted against water ac¬tivity at various temperatures to establish the spreading pressure isotherms and determine the effect of temper¬ature on the spreading pressure. Ln(aw) was plotted against 1/7" at constant spreading pressure. The net in¬tegral enthalpy was determined from the slope of the straight line obtained and plotted against moisture content. Eq. (6) was used to determine the net integral




entropy of melon seed and cassava. The values of inte¬gral entropy obtained were plotted against moisture content.
3. Results and discussion 3.1. Heat of vaporization
The plot of Ln(P„) versus Ln(Ps) for melon "seed and cassava at various moisture contents are presented in Fig. 1(a) and (b). The effect of moisture content on L/h[g for melon seed, cassava and gari is shown in Fig. 2. Data on the L/h(g of gari, which is a dried form of gelatinized cassava mash, was obtained from Fasina et al. (1999). From Fig. 2, it can be seen that heat of vaporization decreased with increase in moisture content. This con¬firms the fact that at higher moisture levels, the strength of water binding decreases. The heat of sorption of melon seed approached that of pure water at the mois¬ture content of about 13% dry basis, while that of cas¬sava was at moisture content of about 36% dry basis. Fasina and Sokhansanj (1993) and Fasina et al. (1997) reported that the heat of sorption of alfalfa pellets ap¬proached the latent heat of pure water at a moisture content between 16% and 18% dry basis, while Fasina et al. (1999) reported that the heat of sorption o( gari and winged bean seed approached thai of pure water at a moisture content of about 15 . Iziesias and Ounfe (1976b) explained that the moisture content at which the heat of sorption approaches the heat of vaporization of water can be an indication of the point at which water exists in free form in a product.
The heat of sorption for cassava and gari were found to be higher than that of melon seed. This could be at¬tributed to the presence of a substantial amount of oil in melon seed. Iglesias and Chirife (1982) reported that the presence of fat and oil in agricultural products con¬tributes to the lowering of their sorption capacity. The heat of sorption for cassava was higher than that of gari. This indicates that gelatinization of starch may have a lowering effect on the binding energy of water sorbed by the product.
3.2. Net integral enthalpy and entropy
The values of the monolayer moisture content of melon seed, cassava and gari obtained at different tem¬peratures are presented in Table 1. Monolayer moisture content of melon seed and cassava decreased with in¬crease in temperature. Similar results were obtained for the monolayer moisture content of winged bean seed and gari (Fasina et al., 1999). However, the monolayer moisture content of gari at any temperature is higher than that of winged bean seed, cassava and melon seed. From Table 1, it can be seen that the monolayer mois¬ture content of cassava at any temperature is higher than that of melon seed. Rizvi (1986) noted that a survey of literature data on the effect of temperature on different dehydrated food products shows that monolayer mois ture content decreases with increase in temperature. This behavior is generally ascribed to a reduction in the number of active sites due to chemical and physical changes induced by temperature, and the extent of de¬crease depends on the nature of the food.
The spreading pressure isotherms of melon seed and cassava are presented in Fig. 3(a) and (b). The spreading pressure decreased with increase in temperature and increased with increase in water activity more markedly in the high range of water activity (aw > 0.6). The variation of net integral enthalpy with the moisture content for melon seed, cassava and gari is shown in Fig. 4. Net integral enthalpy of melon seed decreased from a value of 847 kJ/kg as the moisture content increased from 3.2% (db). The trend seemed to become asymptotic at a net integral enthalpy value of 161.7 kJ/kg as the moisture content of 10% was approached. Similar trends have been reported for the enthalpy of sugar beet root, its insoluble fraction and sucrose (Iglesias et al., 1976), grain sorghum (Rizvi & Benado, 1983), horseradish root (Mazza, 1980) and yellow globe onion (Mazza & Le Maguer, 1978). For both cassava and gari, the net in¬tegral enthalpy increased with increase in moisture content up to a certain point and then decreased with further increase in moisture content up to a certain point and' then decreases with further increase in moisture content. The net integral enthalpy of both products seemed to reach maximum values at about 10% mois¬ture content. Benado and Rizvi (1985), Fasina et al. (1997) and Fasina et al. (1999) respectively obtained a similar response for the integral enthalpy of rice, alfalfa pellets and winged bean seed. This response has been attributed to the influence of volume change and monolayer covering on the thermodynamic behavior of the materials (Fasina et al., 1999). At low moisture contents, water is absorbed on the most accessible lo¬cations on the exterior surface of the solid. As the moisture content increases, the material swells and therefore, new high-energy sites are opened up for water to get bound to. This causes the net integral enthalpy to increase with moisture at low values of moisture con¬tent. The increase continues until the strongest binding sites are covered. The net integral enthalpy then reaches a maximum value and declines as less favorable loca¬tions are covered and as multiple layers of sorbed water form.

The variation of net integral entropy with moisture content for melon seed, cassava and gari is shown in Fig. 5. Net integral entropy of melon seed increased contin¬ually with moisture content but remained negative in value. Similar trends have been observed in the entropy of water sorbed on soil (Taylor & Kijne, 1963), water insoluble fraction of sugar beet root (Iglesias et al., 1976) and grain sorghum- (Rizvi & Benado, 1983). Iglesias et al. (1976) explained that this behavior might be attributed to the existence of chemical adsorption and or structural modifications of the adsorbent, while Rizvi (1986) attributed it to the fact that the products contain more polar groups, which bind water more strongly. A combination of the above factors with the oil content may be responsible for the trend in integral entropy that was obtained with melon seed. The net integral entropy of both cassava and gari decreased with increase in moisture content to a point and then in¬creased with increase in moisture content. Similar trends have been reported on the entropy of moisture sorption in potato starch gel (Fish, 1958), sugar beet root (Iglesias et al., 1976), full fat peanut flakes (Hill & Rizvi, 1982), rice (Benado & Rizvi, 1985), and winged bean seed (Fasina et al., 1999). The decrease of entropy in the low water activity range has been attributed to lateral interactions in the adsorbed film caused by restrictive effect (loss of rotational freedom) of adsorbed water molecules as the available sites become saturated, and structural alteration of the adsorbing food towards in¬creased crystallinity (Kapsalis, 1987). The increase in entropy after-a point indicates that the newly bound water molecules are held less strongly, they possess more degrees of freedom as a result of a gradual opening and swelling of the polymer (Bettelheim, Block, & Kaufman, 1970). Net integral entropy of cassava and gari reached the minimum value at about 12% moisture, and '.here¬after, increased. At the same moisture content, the in¬tegral entropy of cassava is lower than that oi gari. This confirms the fact that gelatinization increases the mois¬ture absorption, solubilization and swelling of starch containing food materials.




4. Conclusions

This study determined the thermodynamic functions associated with moisture sorption by melon seed and cassava. The heat of vaporization of melon seed was lower than that of cassava indicating that cassava has a higher affinity for moisture than melon seed. Cassava has higher heat of vaporization than gari indicating that gelatinization of starch lowers the binding energy of the moisture sorbed by the product. The monolayer mois¬ture content of melon seed at any temperature is lower than that of cassava. This indicates that melon seed could maintain stability at lower moisture contents than cassava at a given temperature. The spreading pressure of both melon seed and cassava decreased with increase in temperature and increased with increase in water activity more significantly in the high range of water activity. At lower moisture levels, the net integral enthalpy of melon seed was higher than that of cassava, but as the moisture content increased, it became lower. The variation of net integral enthalpy and entropy of melon seed with moisture content was not similar in trend to that of the enthalpy and of cassava.
A combination of chemical adsorption, structural modification of the adsorbent and the presence of oil and more polar groups in melon seed are believed to be responsible for the trends observed in the net integral enthalpy and entropy of the seed.