Kouji Maeda

Plastic pollution is ubiquitous in the environment, and nanoplastics (<1 μm) are of growing concern as they pose more health risks than larger particles. However, because of a lack of appropriate model particles, studies examining the risks of polyolefin nanoplastics are very limited, despite the prevalence of these plastics in the environment. Although nanoprecipitation using organic solvents is a promising method for preparing model nanoplastic particles of polyolefins, there are currently no methods for controlling the particle size. Here, we examined how the concentration and volume of the feedstock polymer solution affect the size of polyethylene particles produced by nanoprecipitation. The mechanisms underlying the particle formation were investigated by using a simple population balance model. Increasing the concentration of the feedstock solution increased the growth rate and decreased the nucleation rate, and increasing the volume of the feedstock solution increased the growth rate, resulting in an increase in the mean particle diameter in both cases. These changes in particle diameter were linearly correlated with the suspension density of the dispersion up to a suspension density of 0.4 mg·mL-1. In addition, at these suspension densities, spherical particles were prepared without generating aggregates. Together, these results show that the diameter of polyethylene particles prepared by nanoprecipitation could be controlled according to the suspension density up to a suspension density of 0.4 mg·mL-1. This study provides a basis for the development of nanoprecipitation-based techniques for the precise, scale-independent production of model nanoplastic particles, which we hope will accelerate the risk assessment of nanoplastics.

Three different salt crystals (K2SO4, KCl, and Na2SO4·10H2O) were produced via batch cooling and continuous crystallization using a bench-scale crystallization device, in order to elucidate the mechanism of inclusion of the mother liquor as a function of crystal size. The inclusion ratio of the mother liquor was higher at small sizes and decreased with crystal growth. All results were represented by the previously proposed model of core-aggregation and shell-growth processes for the three different salt crystals. Crystals with small Vickers hardness aggregated with each other, and the inclusion ratio of the mother liquor of aggregated grown crystals was high for all three salt crystals tested herein.

Continuous crystallization via indirect cooling of potassium aluminum sulfate dodecahydrate (KAl(SO4)(2)center dot 12H(2)O) was carried out using a draft tube-type crystallizer with a circulation flow path. The crystallizer was equipped with a newly designed system to monitor the degree of supersaturation in the circulation channel. The average degree of supercooling and growth rate were measured. The crystal growth rates calculated by the population balance model and the supersaturation monitor were different but close. The average supersaturation calculated from the supercooling degree and the solubility data in the literature was 0.19 g per 100 g H2O. This system is also expected to be applied to industrial crystallization where impurities are present.

To understand the inclusion of mother liquor in crystals at different sizes, the continuous crystallization of potassium sulfate was investigated. A bench-scale crystallizer of the draft-tube type was employed for both batch cooling and continuous crystallizations. A standard solution of potassium sulfate was employed as the saturated solution at 323 K. The batch cooling crystallization was first performed at 283 K at a rate of 5 K/h, after which the continuous crystallization was performed at 283 K at two different residence times. The crystal size distribution (CSD) of potassium sulfate crystals and the inclusion ratio of mother liquor in the crystals at different sizes were measured. The average size of the crystals was approximately 400-500 mu m, and the inclusion ratio was less than 1% for all sizes; however, the small and large crystals contained a considerable amount of mother liquor, and the specific crystals that exhibited the lowest inclusion ratios were investigated. Moreover, an impurity distribution model for the suspension crystals was proposed based on coreaggregation and shell-growth processes.

A new method for determining the liquidus and solidus pressures of mixtures of C18's unsaturated fatty acids at constant temperatures was proposed, and three binary isothermal SLE data under high-pressure were systematically measured in this study. The liquidus and solidus temperatures of oleic acid, linoleic acid and alpha-linolenic acid commonly rise above their normal melting temperatures when pressure is increased. This new high-pressure experimental system can closely track pressure in real time during pressure swinging changes, making it simple to measure liquidus and solidus pressures using this dynamic measurement method due to fast responsiveness of pressure. In this study, the liquidus pressures of pure oleic acid, linoleic acid, alpha-linolenic acid and three binary mixtures consisted of oleic acid, linoleic acid alpha-linolenic acid at constant temperature were determined. The isothermal SLE were first correlated based on a simple thermodynamic model. (C) 2021 Elsevier Ltd.

We found that an AAA-type battery (min. 750 mAh) pressurized with Ar or N₂ at pressures of up to 5 MPa exhibited a significant durability enhancement even under high-current conditions.

The separation of the binary unsaturated fatty acids C18:1 + C18:2 system was examined by high-pressure crystallization at 298 K (25 °C). The binary fatty acid mixture containing 0.8 mol fraction of C18:1 was packed in the glass cell with a free piston, and then pressurized by an aqueous ethanol solution around the glass cell up to 200 MPa. The complete solid mixture was obtained after cooling at 253 K (- 20 °C) as a depressurized binary solid. The solid was divided into four parts, and the composition of C18:1 was analyzed at each part of the solid using a refractometer, which revealed there was a composition distribution throughout the solid with C18:1 crystallized one-directionally from the top to the bottom of the cell. High-pressure crystallization could be effective for the separation of binary fatty acids.

<title>Abstract</title> A nickel-metal hydride (Ni–MH) prototype battery completely immersed in an aqueous electrolyte solution of KOH under high pressure was fabricated to examine the effects of high pressure on the quality of Ni–MH batteries. The small battery cell comprised positive and negative electrode materials, as used in electric vehicles, and an Ag/AgO reference electrode. The electric capacity of the Ni–MH battery was measured at different temperatures and pressures with small currents and charge/discharge voltages of 1.6 – 1.0 V. High pressures were found to clearly and effectively enhance the electric capacity of the Ni–MH battery at larger currents. The considerable effect of high pressure on the Ni–MH battery was elucidated by the change in internal resistance during the charge/discharge cycle life experiment, indicating that the voltage of the positive electrode did not appreciably change at a high pressure compared to that of the negative electrode. Moreover, the use of large currents in rapid charge/discharge cycle tests at high pressures of up to 30 MPa resulted in charge/discharge cycles that were five times faster and a quick recovery of capacity was achieved in the 0.5 – 2.1 V range.

In this study, batch crystallization via indirect cooling and continuous crystallization via direct and indirect cooling of phosphoric acid hemihydrate (H3PO4.1/2H2O) from the mixed acid solution of phosphoric acid and acetic acid are performed. In batch crystallization, needle-shaped crystals are obtained, and the impurity concentration of acetic acid in the crystals is 4.2 wt%. Primary nucleation occurred under a high degree of supersaturation, and the metastable zone width for secondary nuclei is narrow. In contrast, in continuous crystallization via indirect cooling, scaly crystals are obtained, and the acetic acid impurity concentration therein is 2.0 wt%. Additionally, continuous crystallization via direct contact cooling is performed, the crystal shape and impurity concentration under direct cooling are similar to those under indirect cooling, suggesting the possibility to downsize the apparatus. Through utilization of the equipment and approach elaborated herein, the correlation between the operating conditions and crystal quality can be assessed.

© 2019 Elsevier Ltd To enhance the reaction rate and yield during biodiesel synthesis with Novozym 435 (an immobilized lipase), a co-solvent method was applied, after which the relationships between reaction time and reaction yield during the single-phase enzymatic transesterification of triolein and methanol in the presence of co-solvents such as tetrahydrofuran, acetone, and hexane were investigated. The addition of hexane led to inactivation of Novozym 435, whereas in the presence of acetone or tetrahydrofuran, the reaction was accelerated and the yield was higher than with a conventional solvent-free enzymatic reaction. There was a clear difference in the dispersion of Novozym 435 particles between with and without co-solvents. The co-solvents prevented particle aggregation from occurring during the reaction without co-solvent. This is the likely reason for the enhanced reaction yield in the presence of co-solvents. To better understand the difference among the co-solvents employed, we estimated phase compositions during transesterification using the LLE (liquid-liquid equilibrium)-UNIFAC (UNIversal quasichemical Functional group Activity Coefficients) model, and examined the effect of these compositions on the reaction and dispersion of the glycerin by-product phase. Using the model, we also accounted for why hexane led to inactivation by estimating the transfer free energy of methanol, glycerin, and water between the initial bulk reaction solution and the water layer on the enzyme surface.

During the transesterification of triglycerides with methanol catalyzed by immobilized lipase resin (Novozym 435 (N435)) in a glass vial, two types of aggregate of N435 particles were observed: a thin-layer aggregate on the glass surface inside the vial, and bulky aggregates. This prompted us to investigate the conditions under which only one type of aggregate could be formed, by changing the material from which the reaction vial was made and the feed ratio of N435 to the cross section of the reaction glass vial. When a polypropylene vial was used, only the bulky aggregates were formed, and reducing the feed ratio led to only the thin-layer aggregate being formed. We also examined the influence of morphological variation of the aggregates on the reaction yield. The thin-layer aggregate showed high activity, with a reaction yield of 0.932. In contrast, the bulky aggregate grew as the reaction time elapsed, resulting in a reduction in the surface area of the enzyme resin, consequently giving not only a reaction yield of about 0.454 but also low repeatability. These results show that the reactor material and feed ratio strongly affect the reaction yield as well as the aggregation form of enzyme resin. Finally, the durability of the thin-layer aggregate and the effect of accumulation of glycerin were investigated.

As a method for efficiently recovering marine organisms, which is expected as a resource for next generation biofuels, crystallizing concentration which can be considered as application of melt crystallization was studied. An operation of forming an ice phase from the outer wall surface to the central portion in the crystallizer was carried out by a cooling jacket with circulating solution. As the operation conditions, the stirring rotation number, the cooling temperature, and the freezing rate were varied to clarify the influence on the concentration ratio of Chaetoceros gracilis in the solution, the dilution ratio of Chaetoceros gracilis in ice, and the solid-liquid distribution coefficient of Chaetoceros gracilis measured experimentally in this study. Furthermore, it is suggested that the solid-liquid distribution coefficients are correlated by two kinds of solid-liquid distribution coefficients models, and concentration by high-speed freezing is also efficient as a result of comparison of the two models.

The biodiesel fuel ethyl oleate (fatty acid ethyl ester, FAEE) was produced using Novozym® 435 (immobilized lipase) by transesterification of a mixture of ethanol and triolein in a fixed-bed reactor operated in circulating batch mode. The miscibility of the mixtures during transesterification was predicted thermodynamically and the conversion ratio of triolein was studied as a function of reaction time. The yield of ethyl oleate increased when the feed molar composition of ethanol was increased. The glycerin by-product could be removed from the enzyme particles by the circulating feed mixture. The inactivation of Novozym® 435 by glycerin was also examined in the circulating batch mode experiments, and an effective reaction model was proposed to describe the inactivation of Novozym® 435 and ethanol effect during the transesterification reaction by using the modified Michaelis-Menten equation.

The present study evaluates the lipase (Novozym 435)-catalyzed transesterification of triolein and methanol mixtures to generate fatty acid methyl esters (FAMEs) using dimethyl ether (DME) in a batch reactor and a continuous pipe reactor. The addition of DME facilitated the preparation of homogeneous mixtures of triolein and methanol. The effect of the Novozym 435 and DME contents in the feed mixtures on the time-course of FAME production was analyzed. With increasing contents of Novozym 435 and DME in the feed mixtures, the FAME concentration increased more rapidly in the batch reactor than in the continuous reactor. The feed flow rate and DME content were varied, and the FAME concentration at the outlet of the pipe reactor was measured with variation of the residence time and DME content. Increasing the DME content in the feed mixtures and the residence time led to a more rapid increase in the FAME concentration for the batch reactor than that in the continuous pipe reactor.

To account for both the growth of solid layer and solute distribution in melt crystallization, we proposed empirical equations involving heat and mass transfer at the solid-liquid interface between the solid layer and melt. With the aim of separating saturated fatty acid as a solid phase from binary mixture of fatty acids as model biofuel, we performed melt crystallization by employing three different species of binary melts which contained triglyceride as unsaturated fatty acid, and myristic acid, palmitic acid, or stearic acid as saturated fatty acid, respectively. As a result, we confirmed the validity of the proposed equations under the varied crystallization conditions.

Solid-liquid equilibria (SLE) for the binary systems of C12-C18 saturated fatty acids or their triglycerides + hexadecane (systems of dodecanoic acid (lauric acid), tetradecanoic acid (myristic acid), hexadecanoic acid (palmitic acid), octadecarioic acid (stearic acid), 1,2,3-propanetriyl tridodecanoate (trilaurin), 1,2,3-propanetriyl tritetradecanoate (trimyristin), 1)2,3-propanetriyl trihexadecarioate (tripalmitin), or 1,2,3-propanettiyltrioctadecanoate (tristearin), + hexadecane) were measured in the hexadecane: fractions of 0 to and temperatures from (2781 to 353.2) K, using differential scanning calorimetry. These systems were selected as, model systems' for blending oily content extracted from restaurant trap grease with fossil fuel oils. The obtained liquidus curves were discussed in. terms of the effects of carbon chain length, chemical form, and variation in the liquid component. In addition) the experimental liquidus curves were compared with predictions based on ideal solution approximation and on the LLE-UNIFAC and Dortmund-UNIFAC models. The triglyceride,and hexadecane binary systems were close to the ideal solution, while the fatty acid hexadecane binary systems were not. The Dortmund-UNIFAC model successfully predicted all liquidus curves of the nonideal, fatty acid-hexadecane systems.

To study solute distribution at the solid-liquid (S-L) interface during melt crystallization, we examined the applicability of the interfacial solute distribution factor proposed based on a kinetic model involving both mass and heat balances at the interface. The factor derived from the model was compared with the experimental results obtained by employing a binary melt with the different species and concentrations of fatty acids as biodiesel related mixtures. As a result, we were able to reveal the empirical relation between the purity of the crystal and the solidification conditions of the melt. Based on the model, we also numerically calculated the transient changes in the interfacial solute distribution factor as well as the temperature of the S-L interface in the solidification process of the melt. The minimization of the factor was confirmed when the melt was supercooled at the S-L interface after starting solidification. (C) 2015 Elsevier Ltd. All rights reserved.

The present study investigates the charge-discharge processes of a Plante lead-acid battery performed under atmospheric pressure and 10 MPa. The ampere-hour efficiency of the lead electrodes was substantially increased by the secondary formation of the Plante electrode, and the ampere-hour efficiency under 10 MPa was slightly higher than that under atmospheric pressure. The active PbSO4 crystals on both electrodes were observed by SEM at different currents, and the morphology and crystal size distribution (CSD) were obtained. Active PbSO4 crystals nucleated and grew during discharging; whereas, they disappeared during charging. This is similar to the crystal behavior during continuous crystallization in a battery. Therefore, a simple population balance model was used to express the CSD of PbSO4 crystals on both electrodes. The CSD of both electrodes were similar at 10 MPa, but the CSD of the positive electrode was different from the CSD of the negative electrode. The current density on both electrodes was balanced under 10 MPa.

Anti-solvent crystallization from a ternary mixture was examined by an NpT ensemble molecular dynamics simulation. The co-solvent and anti-solvent effects were represented by the Lennard-Jones interaction energy parameter, epsilon(ij). The homogeneous binary solution of the solute and solvent was achieved at a constant temperature and pressure. Anti-solvent crystallization was introduced by changing some co-solvent molecules to anti-solvent molecules, immediately. The configuration of solute molecules was investigated by using the radial distribution function, g(r), and the local composition, X-L. The value of epsilon(ij) affected the configuration of the solute molecules significantly; the decrease in epsilon(ij) provided the localization and crystallization of the solute molecules. The composition of the anti-solvent molecules in the solution also affected the configuration of the solute molecules; the increase in the anti-solvent composition produced the crystal structure of the solute molecules more rapidly. These qualitative results corresponded well to anti-solvent crystallization. The radial distribution function represented the crystal structure for solute molecules, and the local composition of the solute was increased from the bulk composition as the effect of the anti-solvent increased. We proposed that the time variation of the local composition of the solute represents the temporal development of the crystal structure and the time to crystallization well. (c) 2015 Elsevier B.V. All rights reserved.

To make larger capacity of charge-discharge of lead-acid batteries at high pressure after the initial formation of Plante-type electrodes, we studied the effect of high pressure on prolonged life testing. Electrode life was evaluated by determining the time of detachment of active crystals (Pb, PbOx and PbSO4) or the drop in ampere-hour efficiency during the charge-discharge cycle. Prolonged life testing, active crystals (Pb, PbOx and PbSO4) on the electrodes used in charge-discharge cycles at atmospheric pressure were easily detached when the electric current was increased to 2.7 mA/cm(2), but the electrodes used at 100 MPa remained very tough until 8.0 mA/cm(2) was reached. To increase the electric capacity of electrodes (i.e. in the second formation process), long-term charge-discharge was performed at a high electric current (5.3 mA/cm(2)). Faster second formation of electrodes at this high current was impossible at a charge of 3.6 As under atmospheric pressure, but it was possible at 6 As and 100 MPa. (C) 2014 The Electrochemical Society. All rights reserved.

Lead-acid batteries were electrically charged and discharged more quickly under high pressures than under atmospheric pressure due to high-pressure crystallization induced by the former condition. High-pressure crystallization can generate extremely small crystals with large supersaturation and small mass transfer rates. Crystals of PbSO4 on the electrodes were observed by scanning electron microscopy, and the crystals became smaller when the operative pressure increased, even when faster charging was carried out. Additionally, the amount of electrical current transferred to the electrodes was larger for electrodes charged and discharged at high pressure. This high-pressure charge-discharge process is expected to improve the quality of lead-acid batteries.

The electrical charging and discharging of a battery involves the crystallization of electrolytes or metal oxides on both electrodes. Crystallization technology that can control nucleation, growth, and distribution of solute crystals might be effective for improving battery properties. We performed charge/discharge cycling of a lead acid battery under high pressure. The charging efficiency at high pressure was compared with that at atmospheric pressure. Charging efficiency at high pressure was found to be higher than that at atmospheric pressure under a high charging current. Observation of the positive electrode by scanning electron microscopy revealed that high pressure caused the crystals on the electrode to become extremely fine. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved.

The molar equivalent electrical conductivity of aqueous ethanol solutions containing ammonium chloride or ammonium nitrate was measured under high pressure, up to 400 MPa, at 298 K. The concentration range of the ammonium salts in solution was broad because their solubility is relatively high compared with that of other salts. A common result was that the molar equivalent electrical conductivity decreased as a function of solute concentration. Additionally, the molar equivalent electrical conductivity decreased considerably when the ethanol content was increased and decreased slightly as the pressure was increased. The molar equivalent electrical conductivity was treated as a function of the square root of electrolyte molality, in accordance with the Debye-Hückel- Onsager equation, but could not be represented by one function because of dependence on ethanol concentration and on pressure. Therefore, use of the electrolyte-nonrandom two-liquid (NRTL) model is proposed for extending the Debye-Hückel-Onsager equation to these two electrolyte solution systems. © 2013 American Chemical Society.

Transesterification in a homogeneous mixture with the volatile co-solvent dimethyl ether (DME) has been proposed as a new fatty acid methyl ester (FAME) processing technique. Ternary feed mixtures of triolein, methanol and DME form a homogeneous solution, but the product mixtures of FAME, glycerin and DME form an emulsion after transesterification. In order to obtain FAME only, the glycerin phase mush be separated out from their mixtures by a centrifuge or settler. Therefore, we investigated the de-emulsification process at different methanol compositions in the feed mixtures, either with or without DME co-solvent. The rate of de-emulsification of the product mixture increased as the methanol content in the feed mixture increased: moreover, the rate of de-emulsification with DME co-solvent was 3-5 times faster than that without DME. The rate of de-emulsification was evaluated as a function of the molar ratio of dispersed glycerin phases in the product mixture. When the molar ratio of dispersed glycerin phases increased in the product mixture, the coagulation of the dispersed glycerin phases accelerated according to the Smoluchowski model. The de-emulsification was measured by UV spectroscopy as a function of time, and the relative transmissivity measured by UV spectroscopy was fit to the simple Smoluchowski equation, and the rate of de-emulsification was considered with respect to the molar ratio of dispersed glycerin phases, as calculated from the liquid-liquid equilibrium using the universal functional activity coefficient (UNIFAC) model. (C) 2011 Elsevier B.V. All rights reserved.

Used for biofuel production, fatty acids and their corresponding triglycerols are stored in most plants. At room temperature, saturated fatty acids and their triglycerols are solid, whereas unsaturated fatty acids and their triglycerols are liquid. Solid-liquid equilibrium data were measured for various binary systems of fatty acids and their triglycerols by differential scanning calorimetry. The freezing temperatures from liquidus curves for the binary systems containing tripalmitine were higher than those for systems containing palmitic acid. At the eutectic point, the melting temperatures for binary systems containing tripalmitine or palmitic acid were similar. The liquidus curves from the Solid-liquid equilibrium data were compared with the predicted values obtained by using the LLE-LTNIFAC model, and good agreement between the experimental and calculated results was found.

The major marine resources are electrolytes, and they are mainly dissolved as ions. More than 70 wt% of electrolytes are sodium chloride, but other electrolytes are also dissolved in seawater. To develop a separation or recovery of electrolytes in seawater, we discussed the electrolyte solution model applied to several electrolytes in seawater. The electrolyte NRTL (Non-Random Two Liquid model) for activity coefficient can be applied to multicomponent systems. Twelve solubility data of two-component systems (water + one electrolyte) are correlated by the solubility product function with the electrolyte NRTL model, and we determined the parameters of solubility products. Some solubility data of three-component systems (water + two electrolytes) were compared to the predicted solubility by electrolyte NRTL model.

We demonstrate a new apparatus for optical observation under high pressure conditions with 4-way optical glasses. We use the proposed apparatus to study solubility of ammonium salts in aqueous solutions and to measure the nucleation pressure for different concentrations of salts in the solution. Despite the fact that the solubility of electrolytes usually increases as the pressure increases, it is found that the solubility of ammonium salts decreases as the pressure increases. In both cases, high pressure can be used towards a crystallization process. Finally, we observe high pressure induced nucleation of electrolytes in an aqueous solution.

Spray evaporation of aqueous ethanol solution with dilute potassium chloride (KCl) was examined for the generation of extremely small electrolyte crystals in dry state. Ethanol content in the aqueous solution played an important role in decreasing the solubility of KCl. Evaporation of small liquid droplets of the aqueous ethanol solution produced a small amount of KCl crystals. The concentration of ethanol in the solution was the operative variable,and the ultrasonic spray nozzle and flow rate of the solution were fixed in this study. The crystals collected were observed by SEM, and crystal size distributions were obtained for different operative conditions. The average crystal size ranged from 100 to 800 run. When the ethanol content increased in the solution, the crystal size decreased. We consider the crystal size according to the slight solubility of KCl in liquid droplets of aqueous ethanol Solution. (C) 2009 Elsevier B.V. All rights reserved.

The solubilities of electrolyte salts, such as sodium chloride (NaCl) and potassium chloride (KCl), are small in alcohol solution. As an application of such dilute solubility, nucleation phenomena were investigated by the dynamic laser scattering (DLS) method at different supersaturation degrees, prepared by different solubilities of KCl in aqueous ethanol solutions. The supersaturated solution of KCl was very stable, and therefore ultrasonic irradiation was utilized as a trigger of nucleation. The time-change of the sizes of fine KCl crystals was measured, and the nucleus size was obtained by extrapolating the size of the fine crystals to zero-time. The nucleus size of KCl was correlated by the modified Gibbs-Thomson equation as a function of supersaturation. it was found that the nucleus size tended to increase when supersaturation decreased, but three different curves for three different temperatures could be obtained. The modified Gibbs-Thomson equation represented the nucleus size at three different temperatures. The correlated "minimum supersaturation energy" parameters were converged to zero at the saturated temperature. The parameter "minimum supersaturation energy" could be interpreted as "ultrasonic effective energy" that could be defined as the received energy in the solution from ultrasonic irradiation. (c) 2008 Elsevier B.V. All rights reserved.

The solubilities and complex phase equilibria for the system of MnSO4 center dot 4H(2)O+MgSO4 center dot 7H(2)O+H2O+CH3OH were determined at the temperatures 291.2 and 301.2 K over the methanol mole fraction range of 0.00-0.12. The solubility data were used for modelling with the modified extended electrolyte non-random two-liquid equation. The salting-out effect Of MgSO4 and methanol on the solubilities of two manganese salts (MnSO4 center dot H2O and MnSO4 center dot 4H(2)O) are represented in the several thermodynamic figures as a function of temperature. The solventing-out effect was stronger than the salting-out effect, which results in a decrease of the solubilities of manganese, salts even though the solubility of MnSO4 center dot H2O decreased and solubility Of MgSO4 center dot 4H(2)O increased as temperature increased. Crown Copyright (c) 2008 Published by Elsevier B.V. All rights reserved.

BACKGROUND: The use of fatty acid mixtures, natural biochemical compounds, will be extended to various chemical industries for the production of a wide variety of products, and various mixtures of fatty acids are necessary for production. Separation of a binary fatty acid mixture of lauric acid and myristic acid using physical vapour deposition (PVD) on a cold quartz crystal resonator is examined. The extremely small amount of deposits can be measured with the quartz crystal resonator. The vapour phase is prepared by vaporizing a calculated composition of melt according to the vapour-liquid equilibrium (VILE). RESULTS: The composition of lauric acid in the melt and the melt temperature were utilized as operating variables in the PVD. The growth rate of deposit increases when melt temperature and the composition of lauric acid in the melt are increased. The composition of lauric acid in the deposit is significantly lower than that of the melt of 19% lauric acid, but the composition of lauric acid in the deposit is much higher than that of the melts of 50% and 75% lauric acid. CONCLUSION: The distribution coefficient of lauric acid between solid and vapour phases can be correlated as a function of the growth rate of deposit. The possibility of separation of fatty acid mixtures by PVD is suggested experimentally and theoretically. (C) 2008 Society of Chemical Industry

The complex phase equilibria for the solubilities of MnSO4 center dot H2O and MgSO4 center dot 7H(2)O in three different organic solvents (methanol, ethanol, and 2-propanol) were predicted by the ion-specific electrolyte NRTL model. The salting-out effect of Mg2+, solventing-out effects of alcohols, and their combined effects on the solubility of MnSO4 center dot H2O were represented in several thermodynamic figures as a function of temperature. We found that the solventing-out effect was stronger than the salting-out effect for the solubility of MnSO4 center dot H2O even though the solubility of MnSO4 center dot H2O decreased as temperature increased.

Superfast transesterification of triolein (a biodiesel feedstock model) was achieved in the presence of liquefied dimethyl ether using the co-solvent effect and low viscosity. Furthermore, a method for higher ester yield, in which methanol was additionally introduced into the reaction system immediately before the beginning of phase separation due to byproduct glycerin to maintain the reaction system in a homogeneous state, was proposed. This method led to superfast and high-yield transesterification with a yield exceeding 96% at 3 min.

The decomposition temperature and pressure (quadruple point) of chlorodifluoromethane (R22) gas hydrate in aqueous sodium chloride (NaCl) solution was measured at different NaCl concentrations in the Solution as a phase diagram. The operative concentration curve of NaCl was obtained as a function of temperature. The maximum decomposition temperature of R22 hydrate was about 290 K at 0.9 MPa pressure, and it decreased as the NaCl concentration increased in the solution. R22 hydrate caused supercooling, and the supercooling Occurrence temperature was much lower than the decomposition temperature. The ultrasonic charge reduced the supercooling of hydration effectively even though the ultrasonic charge did not change the decomposition temperature at all. The concentration experimental results from the several NaCl solutions having different NaCl concentrations were in good agreement with the theoretical operative concentration curve for NaCl. (C) 2008 Elsevier B.V. All rights reserved.

In biodiesel production, transesterification in a homogeneous mixture is possible if a solution of triolein and methanol could be prepared with the use of a cosolvent. The volatile solvent dimethyl ether (DME) can make the immiscible mixture of triolein and methanol miscible; moreover, it can be easily recovered due to its high volatility after reaction. Therefore, the vapor-liquid-liquid equilibrium for the triolein + fatty acid-methanol + DME system was measured by determining the liquid composition, temperature, and pressure of the system. Results showed that DME and palmitic acid effectively promoted the mutual solubilities of triolein and methanol. In addition, the UNIFAC prediction was compared with experimental data, and it was found that the UNIFAC model could represent the complicated phase equilibrium of biodiesel compounds.

Directional crystallization from a binary mixture was performed by pseudo-NpT ensemble molecular dynamics. The initial crystal phase having a face-centered-cubic (fcc) structure grew toward the whole cell according to the temperature gradient in the universal cell. The growing crystal phase was not planar even though the solute molecules grew in two-dimensional coordinates because the solvent molecules disturbed the crystallization of the solute molecules at the diffusive crystal-solution interface. This represented the essential phenomenon of solute distribution during crystallization. Consequently, the growing crystal phase still contained solvent molecules having a liquid structure. The time change of the solute composition in the early phase of crystal growth showed an increase in solute composition as the time step proceeded. The resulting solute composition in this early phase was considered at different temperature gradients in the universal cell and it increased as the temperature of the initial crystal-solution interface increased. A new distribution coefficient model was proposed as a function of the difference between the local solute composition and bulk solute composition in the solution around the crystal-solution interface. The impurity-solvent distribution coefficient could be represented by the new model for faster growth of the lower temperature&apos;s initial interface. As regards a better distribution coefficient, there was found to be a very dilute solution phase over the crystal phase. The new variable "distribution rate" instead of the ambiguous variable "growth rate" was considered as a function of temperature gradient in the universal cell. (C) 2008 American Institute of Physics.

Ice was formed on it rotating cold cylinder, and the distribution of cations (i.e., K(+), Na(+), Mg(2+) and Ca(2+)) between ice and water as a function of the operating variables of ice crystallization was discussed. Ice growth rate was controlled by the rotation speed and temperature of the cold cylinder. The shape of the ice formed on the rotating cold cylinder depended on (lie flow structure of Taylor vortices in the water solution, The effective distribution coefficients of the cations were extremely, small compared with those in the fatty acid system. The effective distribution coefficients of the cations correlated well as it function of ice growth rate. According to the correlation model, pure ice can form if the ice growth rate is lower than the critical growth rate. The critical temperatures of the four cations showed no difference front each other. and increased with the supercooling temperature.

A fatty acid mixture of lauric acid and myristic acid was deposited on a cold quartz crystal sensor from the binary vapor phase of the mixture by physical, vapor deposition (PVD). The cold quartz crystal sensor could detect a small amount of deposit in real time, and provided the growth rate of the thin solid deposit by the PVD. The composition of lauric acid in vapor phase was fixed at a mole fraction of about 0.5 by controlling the composition of lauric acid in melt phase computed from the vapor-liquid equilibrium (VLE). The growth rate of solid phase in the PVD increased with the elevation of the melt temperature, because the vapor pressure of the melt raised by the temperature elevation was the driving force of the PVD. The composition of lauric acid in the solid phase was considerably lower than that in the vapor, and it was increased as the growth rate in the solid phase rose. The distribution coefficient of lauric acid between solid and vapor phases was much less than that between solid and melt phases. We compared the separation capacity in melt crystallization with that in the PVD using the effective distribution coefficient, and found that the PVD had higher separation capacity. (c) 2006 Elsevier B.V. All rights reserved.

This paper describes the surface topography characteristics of the (100) and (101) faces of potassium dihydrogen phosphate (KDP) crystals by using atomic force microscopy (AFM) for nanoscale observation. Fractal dimensions of the surface pattern were measured, including step height and terrace distance of KDP crystals under different supersaturation levels and dye concentrations, and the relationship between dye concentration and surface structure was examined.

A solid phase of a fatty acid mixture (lauric acid + myristic acid) on a rotating cold cylinder from the binary melt was formed in an annular cylindrical vessel. The growth rate of the solid phase was kept constant by cooling the rotating cold cylinder. The shape of the solid phase formed on the rotating cold cylinder depended on the hydrodynamics of Taylor vortices in the melt phase. The solid phase consisted of a concave part and convex part. We observed the detail structures of the surface and the inside of the solid phase under a microscope, and analyzed the crystallinity of the solid phase by using X-ray diffraction. We measured the weight fraction of myristic acid from the thin slices of the solid phase by gas chromatography. The growth rate of the solid phase decreased as the rotation speed increased or as the temperature of the hot cylinder increased. When the growth rate decreased, the crystalline part of the solid phase obviously increased. The effective distribution coefficient of lauric acid between the solid phase and melt phase increased according to the crystalline of the solid phase. The effective distribution could be well correlated with the growth rate by using our interfacial distribution model.

A single KDP (potassium dihydrogen phosphate) crystal was grown in a supersaturated solution containing a metallic ion (Al3+, Fe3+, or Cr3+). The growth rate, morphology, and distribution of the metallic ions into the KDP crystal were measured as the ionic concentration and supersaturation in the solution changed. It was found that in the KDP crystal, Al3+ and Fe3+ were greatly concentrated, but Cr3+ was diluted. Complete expressions for the effect of metallic ions on all aspects of the growth of KDP crystal were suggested. The growth rates of (100) and (101) faces were well correlated by the empirical equation and resulted in good estimation of morphology. The distribution of metallic ions into KDP crystal was also correlated by the distribution model. (C) 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

A crystallization monitoring system using a quartz crystal oscillator was implemented in the cooling crystallization of dilute lauric acid solutions for the investigation of the nucleation process of the solute. In addition, the microscopic observation of the oscillator surface was conducted to examine the number and size of yielded nuclei, and the observed results and the resonant frequency variation of the oscillator were analyzed to explain the nucleation process. The change of the resonant frequency indicated the beginning moments of emulsification and nucleation of lauric acid near the oscillator surface. The experimental results show that the average crystal size was raised from 139 to 238 nm and the nucleation rate was reduced as the solution concentration increased from 0.05 to 0.25 g/L. When the solution temperature lowers, the crystal grows more rather than new nuclei form in the high concentration solution. The estimated crystal mass from microscopic observation ranges between 18% and 28% of that from the frequency measurement. It was also demonstrated that the proposed monitoring device is useful to investigate the crystallization process in dilute solutions, which often requires large analytical instruments. (c) 2005 Elsevier B.V. All rights reserved.

Crystallization of solute molecules from homogeneous binary Lennard-Jones solutions was simulated by molecular dynamics of an NPT ensemble. Isothermal (equilibrium) and cooling (non-equilibrium) conditions were examined to observe the configuration of solute molecules for the crystallization process. The isothermal simulations could clearly determine the temperature of crystallization or nucleation of solute molecules by the change of density. The cooling simulations could describe the effect of cooling rate on the temperature of crystallization or nucleation. The configurations of solute molecules in the solution during crystallization were discussed from the radial distribution function (RDF) and the local composition of solute (LC) in the solution. The nuclei that have a crystalline structure could be distinguished from clusters that have a liquid structure. It was found that the liquid clusters were not stable, but the crystal nuclei have very rigid structure and were stable. LC was used to image how many solute molecules gathered in the solution with relation to the clusters and nuclei. It was found that LC fluctuated dramatically when the solute molecules crystallized. The cooling simulations showed that the configuration of solute molecules obtained at a slower cooling rate showed the crystalline structure, but the faster cooling rate caused the liquid structure of clusters in the solutions even at low temperature. (c) 2005 Elsevier B.V. All rights reserved.

A hydrophobic compound dissolves very slightly in an aqueous solution. As an application of such dilute solubility, the concept for making nanometer-scale crystals of lauric acid from the aqueous ethanol solution was suggested. There are two different dilute solubilities in the aqueous phase. One thing is defined as the solubility of solid lauric acid; another is defined as the mutual solubility of liquid-liquid equilibrium between aqueous phase and oil phase. Two possible processes for crystallizing lauric acid were shown according to two different solubility curves. One process is direct crystallization of lauric acid; another process is indirect crystallization of lauric acid after emulsification. The crystals formed by two processes were experimentally observed, and then the size of crystals was analyzed by dynamic light scattering (DLS). The crystals formed through both emulsification and crystallization processes were 500 nm in size. The crystals formed through direct crystallization process ranged from 30 to 100 run in size. The size of the latter crystals was a function of ethanol mole fraction in aqueous phase and cooling temperature. The modified Gibbs-Thomson equation was examined to correlate the crystal-nuclei size of the wide range. (c) 2004 Elsevier B.V. All rights reserved.

The crystal size distribution (CSD) of potassium dihydrogen phosphate (KDP) and inclusion ratio of mother liquor have been measured in a mixed suspension mixed product removal (MSMPR) crystallizer. A small amount of sodium dihydrogen phosphate (SDP) was added as an impurity tracer in order to measure the inclusion ratio of mother liquor inside the KDP crystals. The effects of residence time on the CSD and inclusion of mother liquor were investigated. The crystal growth rate and nucleation rate were determined from the population density balance theory by assuming a size-dependent function of the growth rate, because the size-dependent growth rate of KDP crystals was observed under a microscope. MSMPR operation with a shorter residence time produced larger KDP crystals, and showed more size-dependency of the growth rate. The inclusion ratio was depicted as a function of crystal size, growth rate, and agglomeration. The degree of agglomeration was effectively correlated with the inclusion ratio of the KDP crystals that had a large hardness property. (C) 2004 Elsevier B.V. All rights reserved.

Colored potassium dihydrogen phosphate (KDP) crystals from a dye (sky blue) solution were produced in a mixed suspension mixed product removal (MSMPR) crystallizer. The crystal size distribution (CSD) of the KDP crystals was correlated by a size-dependent growth rate model. The growth rate increased with crystal size, but it decreased as the residence time increased. Two different concentrations of dye in the solution were examined for MSMPR crystallization of KDP crystals. The colored crystals from the solution with low dye-concentration generated regular-shaped KDP crystals. The dye distribution coefficient into the regular-shaped KDP crystals was less than unity. However, the colored crystals from the solution with high dye-concentration formed the special-shaped KDP crystals having no (10 0) sector. The dye distribution coefficient into the special-shaped KDP crystals from the dye solution was larger than unity. The dye distribution coefficient as a function of crystal size was presented, and then the relationship between the aspect ratio of the morphology of KDP crystals and dye distribution coefficient was discussed. As a result, a distribution model of dye impurity in KDP crystals for an MSMPR crystallizer could be first proposed in this work. (C) 2004 Elsevier B.V. All rights reserved.

To develop an activity coefficient model for aqueous multielectrolyte systems, ion-specific NRTL parameters were applied to the electrolyte-NRTL model. In total, 132 ion-specific NRTL parameters for 21 ionic species (K+, Na+, Mg2+, H+, Ca2+, Zn2+, Al3+, NH4+, Cd2+, Co2+' Cu2+, Pb2+, Mn2+, Ni2+, Cr3+, Cl-, SO42-, NO3-, OH-, ClO4-, CrO42-) were determined from the binary activity coefficient data of aqueous electrolytes and used to predict the activity coefficients of electrolytes in aqueous multieletrolyte solutions without any additional parameters. Furthermore, solubilities of inorganic salts including heavy metals in aqueous mixed electrolyte systems and the distribution ratio of Cu in a solvent extraction process containing HCl and CuCl2 were predicted using the activity coefficients derived from our ion-specific electrolyte-NRTL model. The results showed that the NRTL parameters obtained in this study accurately predict phase equilibria of systems involving inorganic salts such that this model can be used in the design of recovery processes such as crystallization and solvent extraction.

Single potassium dihydrogen phosphate (KDP) crystals were grown in a supersaturated solution containing an organic dye (sunset yellow FCF, brilliant blue FCF, and sky blue). The growth rate, morphology, and impurity dye distribution of faces, (100) and (101) in a KDP crystal were measured as dye concentration and the supersatutation of KDP were changed. Complete expressions for the effect of dye on all aspects of the growth of KDP crystals were discussed. The growth rates of (100) and (101) faces were well correlated by the empirical equation, and resulted in good estimation of the morphology. The distribution of dye in a KDP crystal was represented by the distribution model containing the minimum growth rate for coloring. The growth rate equation and distribution equation were expressed by functions of the supersaturation and dye concentration, and they could effectively provide the operational conditions with coloring the KDP crystal. (C) 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

Solid-liquid phase transitions were demonstrated by taking into consideration miscibility of the binary Lennard-Jones (LJ) mixtures using NPT molecular dynamics simulation. The effect of the LJ parameters (sigma and epsilon) on solid-liquid phase transition is discussed. As for single component systems, molecular weight and epsilon parameter were not significant, the reduced temperature of solid-liquid phase transition increased when the sigma parameter became large. The binary LJ mixtures of chiral compounds had the same sigma and epsilon parameters, and mixing physical properties were varied by only the epsilon(ij) parameter of the i-j pair. Miscibility of the binary LJ mixture increased as the epsilon(ij) parameter approached the pure epsilon parameter. As for binary mixtures, the fact was found that miscibility influenced the solid-liquid phase transition temperature. The immiscible mixtures having a low epsilon(ij) parameter. showed a novel phase transition diagram by molecular dynamics simulation. (C) 2003 Elsevier Science B.V. All rights reserved.

The crystal size distribution (CSD) and crystal purity of potassium chloride (KCl) have been measured in an MSMPR crystallizer by using two different impellers. A small amount of sodium chloride (NaCl) was added as an impurity tracer in order to measure the inclusion ratio of mother liquor inside KCl crystals. The effects of mixing, residence time, growth rate and crystal size of KCl on the inclusion of mother liquor were investigated. The crystal growth rate and nucleation rate were determined from the population density balance theory. The attrition of crystals changed round shape of the crystal and affected significantly the inclusion of mother liquor in the crystals. The roundness of KCl crystals was proposed as a new factor to take the attrition effects into consideration. An empirical relationship between the inclusion of mother liquor and the operating variables was obtained.

The solubilities of three sulfate salts (Na2SO4, K2SO4, MgSO4) and two chloride salts (NaCl, KCl) with a second salt in water were measured at 313 K. The ternary systems studied were water + NaCl + KCl, water + NaCl + Na2SO4, water + KCl + K2SO4, water + Na2SO4 + K2SO4, water + K2SO4 + MgSO4, and water + Na2SO4 + MgSO4. The solubilities of two salts in the ternary systems were obtained, and the data were used to obtain the eutectic concentrations of the ternary systems. The solubility data were also predicted by the electrolyte-NRTL model, in which parameters were obtained from the binary data. The qualitative effect of the second salt on the solubility of the first salt in water is discussed.

Azo-organic dyes as additives were examined to make colored KDP (potassium dihydrogen phosphate) crystals. The dyes used here were amaranth and sunset-yellow FCF. Influences of supersaturation and dye concentration in the solution were observed on the color and crystal habit of KDP. Amaranth colored the pyramidal section (10 1) of the crystals in the solution at low supersaturation and high dye concentration. The highly supersaturated solutions produced entirely colored crystals. The crystal habit did not significantly change in the amaranth solution. Sunset-yellow FCF had a significant effect on the crystal habit and color of KDP, and the prismatic section of the crystals was decreased. Sunset-yellow FCF colored the whole KDP crystals in the solution at high dye concentration. The habit index of KDP crystal, which is the ratio of the pyramidal size to prismatic size, was discussed as a function of dye concentration and supersaturation. (C) 2002 Elsevier Science B.V. All rights reserved.

The partitioning of n-eicosane (C-20) and n-tetracosane (C-24) between a melt phase (L-1) and an aqueous ethanol phase (L-2) at 333.2 K, and between a solid phase (S) and an aqueous ethanol phase V at 293.2 K was measured. The wax content of aqueous ethanol solutions was greater when the solutions were in equilibrium with a melt wax phase (liquid-liquid equilibrium) than when they were in equilibrium with a solid wax phase (solid-liquid equilibrium). The distribution coefficients of C-20 and C-24 between the aqueous phase and a wax phase were also significantly higher in the case of liquid-liquid equilibrium than in the case of solid-liquid equilibrium. The data could be modeled satisfactorily using simple thermodynamic expressions based on the UNIFAC activity coefficient approach.

Emulsification of lauric acid in an aqueous ethanol solution including lauric acid solute has been observed during cooling before crystallization of lauric acid occurs, The nature of two different solubility curves was explained for the system of lauric acid and aqueous ethanol solution. The mutual solubility of the two liquid phases controls emulsification; the solid solubility of lauric acid controls crystallization. The mutual solubility curve appears at relatively high temperature, and the solid solubility curve at relatively low temperature. Crystallization essentially generates a solid metastable zone under the solid solubility curve, A supersaturated solution can be obtained in the metastable zone. However, no nucleation occurs in the metastable zone. The metastable zone, therefore, still caused emulsification at low temperature before crystallization of lauric acid occurred. The hypothetical mutual solubility curve for the aqueous solution including hydrophobic solutes appeared invariably even at low temperature in the metastable zone under the solid solubility, (C) 2001 Academic Press.

Experiments have been performed on the impurity incorporation in the solid layer grown from a binary melt of myristic acid and palmitic acid using a rectangular cell with well-controlled temperature. The binary melt forms a simple eutectic phase diagram. Direct numerical simulations of melt solidification were carried out in order to validate the new theory of interfacial solute distribution proposed by the present authors [J. Chem. Phys. 109, 7468 (1998)]. The theoretical predictions of purity or impurity concentration profile in the solid layer agreed well with the experimental measurements. The purity profile inside the solid layer was found to be basically controlled by the flow structure of melt that essentially contributed to the transport phenomena of heat and mass near the solid layer surface. (C) 2000 American Institute of Physics. [S0021-9606(00)51303-8].

Solid-liquid equilibria (SLE) data have been measured for the quaternary systems of aqueous solution + binary fatty acid. In addition, binodal data of two liquid phases have also been obtained. Aqueous acetone solution and aqueous ethanol solution were examined as a solvent for the binary lauric acid + myristic acid system. The co-solvent effect of acetone and ethanol could be investigated on the SLE for the binary fatty acid system. However, the anti-solvent effect of water is slightly observed on the SLE for the ternary systems. The binodal region of two liquid phases became larger when the ratio of myristic acid to the total fatty acids increased. The aqueous acetone solution made the binodal region slightly larger than the aqueous ethanol solution. The NRTL equations were examined for the prediction and correlation of the measured SLE and binodal data. The NRTL parameters from the binary systems could describe the SLE phase diagram qualitatively. However, they could not present the binodal curve of two liquid phases. The correlations of SLE and binodal data could be satisfied by the NRTL equation. (C) 1999 Elsevier Science B.V. All rights reserved.

A new separation process of saturated fatty acids (lauric acid-myristic acid) using crystallization from an aqueous ethanol solution has been examined. There were two vessels in this separation process: an extraction vessel and a crystallization vessel. The fatty acids in the aqueous phase were first extracted from their organic phase (melt) in the extraction vessel. The fatty acids in the aqueous phase were continuously introduced to the crystallization vessel, and then the fatty acids were crystallized there. The crystals of the fatty acids were collected continuously above the aqueous phase in the crystallization vessel. In this process, the yield and the purity of the crystals over time were measured, and it was found that the purity of lauric acid increased unsteadily up to 0.98 mole fraction of lauric acid with an increase in the yield of the low yield range. The mole fraction of ethanol in the aqueous phase could be significant to control the relationship between the yield and the purity of the crystals. Three different mole fractions of lauric acid in the organic phase were used to be separated in this process. Moreover, we have considered the effective separations of this process, and the maximum yield and purity of the crystals have been estimated by a simple mass balance.

Face growth rates were measured on (101) and (100) faces of KDP (Potassium Dihydrogen Phosphate) and ADP (Ammonium Dihydrogen Phosphate) crystals at different supersaturations in the presence of aluminium ferric and chromium ions added as habit modifiers. The growth rate of the (100) face decreases dramatically. The relative habit-modifying powers of the trivalent cations (Al3t&lt;Fe3+&lt; Cr3+) are confirmed. When ADP crystal grows adsorption on the ledge is important for the reduction of face growth rate by aluminium and ferric ions. Adsorption at the step is important for chromium ion.

The crystallization of two liquid phases based on the phase diagrams, which plot solid-liquid equilibrium (SLE) and liquid-liquid equilibrium (LLE) is described. The separation of water+ethanol+fatty acids system using crystallization and liquid-liquid extraction is proposed. Both SLE and LLE are predicted using the UNIFAC method. Solids will easily collect on the surface of the aqueous phase and that solid will have high purity.

A model for predicting a distribution coefficient (ki) of solute at the solid-liquid (S-L) interface, when the solid layer is growing, is proposed. The interfacial distribution coefficient is expressed as a function of two gradients of the liquid concentration and equilibrium concentration at the S-L interface. The model is applied to the solidification of a simple eutectic binary liquid of lauric acid and myristic acid in an enclosed rectangular box in which a vertical wall is cooled. The impurity-concentration profile in solid is predicted from the direct numerical computations. © 1998 American Institute of Physics.

Crystallization of fatty acids for a lauric acid + myristic acid system from ethanol aqueous solutions, which extracted the fatty acids from their organic solutions, has been carried out. The mole fractions of lauric acid excluding solvent components in the aqueous solutions after the liquid-liquid extraction were a little higher than these in the organic solutions. As crystallizing of the fatty acids from the aqueous solutions proceeded, the crystals of the fatty acids floated on the surface of the aqueous solution, and the mole fractions of lauric acid in the crystals increased considerably from the mole fractions of lauric acid excluding solvent components in the aqueous phase. The effect of experimental conditions on the mole fraction of lauric acid in the crystals was shown.

To determine the liquid-liquid-solid three-phase equilibrium (LLSE), binodal data of the liquid-liquid equilibrium (LLE) and the solid-liquid equilibrium (SLE) have been measured for the ternary water + acetone or ethanol + lauric acid, myristic acid or palmitic acid systems. The binodal region for two liquid phases increases successively in the following order of the third component: lauric acid &lt; myristic acid &lt; palmitic acid, The binodal regions of the ternary systems with acetone as the second component are larger than those for ethanol. These binodal curves are not highly sensitive to temperature. The isothermal solubility of these fatty acids is considerably reduced when the water content in the aqueous phase increases. The solubility of fatty acids is successively reduced in the following order: lauric acid &gt; myristic acid &gt; palmitic acid. The solubility of fatty acids in aqueous solutions containing acetone is smaller than in those containing ethanol. Predictions of both the binodal curve for LLE and the SLE curve by the UNIFAC method and the NRTL equation do not represent the experimental data. However, data of the binodal curve and solubility are well correlated by the NRTL equation, and also the LLSE relationship could be represented. (C) 1997 Elsevier Science B.V.

First, the phase diagrams for the water+ethanol+fatty acids system were described, and the separation method was considered. Second, the separation of lauric acid+myristic acid mixture in the ethanol aqueous solution including ethanol by liquid-liquid extraction and crystallization has been examined. The extraction of fatty acids from the organic phase to the aqueous solution has been done. Fatty acids in the aqueous solution were crystallized using the aqueous SLE surface. The solvent-free mole fraction of lauric acid in the aqueous solution was a little higher than that in the organic phase. Fine crystals made in the aqueous solution had better purity for lauric acid than the solvent-free mole fraction of lauric acid in the aqueous solution. The effects of the mole fraction of ethanol in the aqueous solution, the mole fraction of lauric acid in the organic solution, extraction temperature and crystallization temperature on the purity of the crystals were reported.

Crystallization of an organic mixture using spray-evaporation of volatile solvents has been examined to obtain solids without inclusion of liquid. A solid mixture of lauric acid + myristic acid was dissolved in a solvent (acetone, isobutane, propane), and the solution was sprayed through a spray nozzle (100 mu m I.D.) and evaporated in a spray column under vacuum. The column was composed of five glass tubes, where powdery solids were collected after spray-evaporation of the solvent. Weight and composition of the solids on each glass tube were measured at various operating conditions. Most of the solids were collected on the glass tube near the nozzle and also on the bottom of the column. When the composition of the solids was analyzed, it was found that the mole fraction of lauric acid decreased from the top to the bottom of the column. The percent of solids collected on the bottom of the. column was correlated well with the Weber Number, which is function of the physical properties of the solution and the operating conditions.

Slurry was first produced batchwise by cooling a ternary mixture of an organic binary eutectic and a liquefied gas, and a wet solid was obtained by filtrating the slurry. To provide a highly purified solid product, continuous vaporization of the wet and dry solids was carried out. As a result of the continuous vaporization, the time course could be divided into two periods. The first period is evaporation of liquid adhering to the wet solid and the second is sublimation of dry solid. The higher the composition of liquefied gas in the feed was, the lower was the vaporization ratio of the adhering liquid removed, and the purity of the dry solid reached more than 99 mole % without loss of yield. The variation of solid purity and yield during evaporation have been formulated by employing material balance as well as phase equilibria. The formula could indicate the shift point from the evaporation to the sublimation. Regardless of the sort or composition of liquefied gas, the purity of the final solid product reached 99.9 mole% after the sublimation. The relation between impurity and yield during sublimation has been probably correlated by an impurity distribution model, and the role of liquefied gas was shown by the model.

Batch crystallization of binary eutectic organic mixtures using liquefied gas as a solvent was carried out. Mean diameter of crystals in a crystallizer and mole ratio of liquid adhering to crystals in wet solid on the filter were determined. Propane or chlorodifluoromethane (HCFC22) was added as a liquefied gas solvent to binary eutectic mixtures : benzene + cyclohexane, P -xylene + m -xylene and P -xylene + o -xylene. Mean diameter of crystals increased with increase in moles of liquefied gas added. It ranges from 100 to 230 mum for benzene, 100 to 120 mum for cyclohexane and 400 to 600 mum for p-xylene. The diameter was represented by an exponentially decreasing curve with concentration ratio. Purity of wet solid increased with increasing moles of liquefied gas added. Ratio of adhering liquid to crystals was found to be well correlated with capillary number, which is calculated from density, surface tension of mother liquor and mean diameter of crystals.

Applying our new crystallization method to separation of xylene isomers, solid-liquid equilibria (SLE) of the binary xylene isomer systems o-xylene+p-xylene and m-xylene+p-xylene were measured by both cooling curve and DSC methods. SLE data by these methods were consistent with each other. To learn the effect of a liquefied gas on the SLE for these binary systems, vapor-liquid-solid equilibria (VLSE) for the ternary systems liquefied gas (propane or chlorodifluoromethane [HCFC22]) + binary xylene isomer systems were also measured by the cooling-curve method. The correlated results of these VLSE data using an activity coefficient equation showed that the liquefied gas component had no capability to shift the eutectic composition of the o-xylene+p-xylene system but could shift the eutectic composition of the m-xylene+p-xylene system. The quantitative effect of addition of the liquefied gas on the eutectic composition was explained by a thermodynamically derived equation.