Condensable volatile decomposition products and free fatty acid in the deteriorated oils were examined to know what kind of reaction may play the main part of deterioration during deep fat frying in the absence or the presence of air. Steam distillated materials from soybean oil under the continuous water-spraying and heating system with and without air blowing, were collected as shown in Fig.-1. Collections were continued during 014, 1428 and 2842 hr respectively. The quantities of acidic, carbonyl and alcoholic fractions of the volatile products at various intervals were determined, and the methyl ester of acidic fractions was gas chromatographed using succinate polyester column and each peak was identified by the tentative method (Fig-3, 6, Table-4, 5). The free fatty acid was isolated from the final deteriorated oils by means of ion-exchange resin and fractionated into monomer, dimer and secondary products by silica gel chromatography mentioned by Frankel et al., and the methyl ester of monomer acid was gas chromatographed (Table-2, 3, Fig.-2). Acid value, viscosity increasing ratio, carbonyl value and iodine value of the deteriorated oils were deteamined every 14 hr heating (Table-1). In the absence of air, free fatty acid in the final deteriorated oil was recovered with the yeald of 99.0% with ion-exchange resin and 93.5% of this acid fraction was monomeric. The composition of monomer acid was practically similar to the constituent acid of original soybean oil. Also, in the volatile products, acidic compound was main product and its composition was almostly similar to that of soybean oil. These results indicate that the free aftty acid increased in the deteriorated oil in the absence of air, was produced by the hydrolysis of triglyceride. In the presence of air, free fatty acid was removed 90.5% by ion-exchange resin and 82.3% of this acid fraction was monomeric. Acidic materials which cannot be removed by ion-exchange resin, was considered as that combined with glyceride. The composition of monomer acid was similar to that of soybean oil. These results indicate that, even in the case of presence of air, the free fatty acid increased in the deteriorated oil was produced mainly by hydrolysis. On the other hand, volatile products were the mixture of acidic, carbonyl and alcoholic compounds, and acidic compounds contained short chain fatty acids, dibasic and keto acids.
Experiments were carried out with hardened coconut oil to find the correlation between crystal growth and hydrolysis of esterifird lower fatty acid at lower temperature. 1) A.V. showed practically no change, both in hardened and non-hardened oils, when stored in anhydrous state, and there was no difference in the values due to storage temperature. When the oil was saturated with vapor, the values increased in the order of those stored 15 and 5°C, and the tendency was especially marked in hardened coconut oil stored at 15°C. 2) Methyl ester of hardened coconut oil and those dissolved in ethanol, n-hexane and benzen did not show rise in A.V. values even when stored at a low temperature. 3) The fraction obtained by molecular distillation showed less change, even when stored at 5°C, than the original hardened coconut oil. The fact that A.V. was higher in Fraction 2 than in Fraction 1, containing larger amount of lower acids, indicated that when molecules of a same size were collected, more impacted intermolecular space should be presumably produced. This was borne by the fact that electron microscopic observation showed a smooth surface although there were fine single crystals.
On studying phenomena relating to emulsification, it becomes necessary to study on intermolecular forces between emulsifier and oil or water. On this view point, it was found that there exists, at a given oil/water system, required HLB value for interfacial viscosity as well the required HLB value for emulsification, and also there exists linear correlation between required HLB value for interfacial viscosity and solubility parameter of oil. The empirical equations to determine required HLB value for interfacial viscosity at a given oil/water interface were deduced experimentally as follows; Y=6.146δ0-46.0 or Y=127.3logδ0-112.2 where Y=required HLB value for interfacial viscosity, and δ0=solubility parameter of oil. Above empirical equations are both applicable in the range of Y=515 (corresponding to δ0=8.310.0) independent of the kind of emulsifiers (at least, in the case of nonionic emulsifiers). Required HLB value for interfacial viscosity (now this value may be calculated only from δ0 value for a given oil), as well required HLB value for emulsification, may be useful for determination of HLB value of a given surfactant experimentally. In the course of this study, the author found that heat of vaporization (accordingly δ0 value) may be calculated by the following empirical relation; ΔHv (T') =ΔHv (T0) √T0/T' where T', T0 are absolute temperature and ΔHv (T') , ΔHv (T0) are heat of vaporization at T' and T0 respectively.
The hydrothermal reaction of silica-lime-water-surface active agent was studied by statistical method. The elements of this system consist of reaction time and temperature, molding pressure and temperature, surface area of silica, kind of surface active agent, ratio of lime/silica and NaOH/lime. The elements of the hydrothermal reaction were analysed by orthogonal array. The main effects, interaction and partition of contribution (%) were corrected and estimated for each element and level. The results were as follows. 1) The molded sample at high temperature (100°C) and high pressure (100kg/cm2) cracked often when seperated from metal mold, and decreased bending strength. 2) Surface area of silica did not affect in the hydrothermal reaction when powder of silica as start material (in hydration system) possessed area over 3500cm2/g. 3) Generally, bending strength and unreacted lime of hydration reaction products were decreased when small amount of surface active agent was initially added to silica-lime-water-system, in spite of resulting surface gloss and smoothness of the products. 4) Especially, the anionic and nonionic surface active agent was absorbed to silanol group of silica and interfered hydration reaction of initial composite. Also, addition of sodium hydroxide to system as catalyst accelerated the reaction, while decreased the bending strength. 5) Estimations of the population mean of bending strength of test pieces were decided as 39.49±1.91 (kg/cm2). 6) The best result was obtained when the initial mixture of C/S 0.51.0 reacted at 200250°C for 2025hrs. The reaction was almost decided through statistical treatment of above mentioned elements. although the existence of several other unknown elements possessing slight effect to the hydration reaction was presumed.
New Polyesters prepared by the esterification of succinic or adipic acid with α-buthyl-α'-hydroxyethyl glyceryl diether were found to be useful liquid phase for gas liquid chromatography for separation of linolenic, arachidic and eicosenoic acids. Rape seed oil methyl esters were successfully analyzed with the polyester columns.
Authors attempted the preparation of a lot of phosphatidyl inoshitol (PI) for the physico-chemical and biochemical researches according to the combined method of solvent fractionation, DEAE cellulose and silicic acid column chromatography. The analytical data of pure PI obtained by this method showed by Table-1. The fatty acid composition of PI showed by Table-2. Both preparations gave the patterns of IR spectrum which accorded to their chemical structure. They showed single spot on the thin-layer and paper chromatgram.