Triisooctylamine (TIOA) was used to extract Cr(VI) from several acid solutions of HCl, HBr, HNO3 and HClO4. The results were divided into two groups according to the acid solutions used: HCl and HBr, or HNO3 and HClO4. Chromium(VI) is extracted according to an ion-exchange reaction which has been found for amine-Cr(VI) extraction systems. However, the ion-exchange mechanism holds only at low acidities. At high acidities, Cr(VI) is coextracted with HCl or HBr by TIOA. It was found that TIOA extracts Cr(Vl) as Cr2O72– from HClO4 solution and as HCrO4– from HCl, HBr, and NHO3 solutions. An extraction model which could explain the distribution equilibria was proposed, and the extraction constants were then determined.
Air-mixing in an expanding and contracting vessel influences the deposition of Brownian particles. In the present work, by using a previously developed technique to measure mixing and deposition of aerosol particles in a balloon (Otani et al., 1990), extensive data were collected to cover a wide range of balloon expansion and contraction conditions. The results were examined in terms of dimensionless parameters derived from the convective diffusion equation to predict the behavior of submicron particles in actual human alveoli. It was found that (1) convective mixing in the balloon is characterized by the Reynolds number at the balloon throat; (2) for Re > 5, the mixing is determined by fluid convection; (3) for Re < 5, Brownian diffusion enhances the mixing; and (4) when dimensionless duration of balloon expansion and contraction is longer than 5 × 10–4, the mixing does not influence particle deposition and the deposition efficiency can be predicted by diffusional deposition of particles in a spherical vessel with constant volume.
Single-component gas-phase adsorption isotherms for a commercial activated carbon fiber, KF-1500, are determined for acetone, diethyl ether and methanol at 273.15, 298.15 and 323.15 K and for water at 298.15 K by a gravimetric method. No hysteresis was found for the three solvent gases, but a reproducible hysteresis loop was observed for water. The isosteric heats of adsorption for the three organic solvent gases decrease with increase in adsorption in a low-surface coverage region while they increase in a high-loading region. Dubinin’s characteristic curve method and the multi-site occupancy model were applied to correlate the adsorption data. The latter model represents the experimental isotherms of the organic solvent gases satisfactorily in the region of surface coverage greater than 0.1.
Production of hydrogen peroxide by electroreduction of oxygen in alkaline solutions was experimentally carried out using a packed-bed electrode of graphite particles or graphite felts. Oxygen sparging into the bed was found to increase overall yield of hydrogen peroxide for a one-pass operation because of the increase in concentration of dissolved oxygen. The current efficiency decreased with increasing concentration of hydrogen peroxide. Effects of packing materials, sparging rate of oxygen and operational conditions on production rate were analysed by measuring potential distribution within the bed. The production rate for the electrolysis using graphite felts was found to be larger than that using graphite particles. The felt cathode had a large electrode surface area and a large volumetric ratio of gas to electrolyte, which might enhance the mass transfer rate of dissolving oxygen. The consumption of electric power for producing hydrogen peroxide was found to range from 2 to 6 kWh·kg–1-H2O2 in the packed-bed electrode with graphite felts.
To estimate the distribution of turbulent eddy viscosity in the flow related to complicated geometries, a simple model was developed based on the mixing-length hypothesis. This model was applied to two cases of turbulent separated flows around a surface-mounted obstacle of circular cross section. The first case is film flow on the outer wall of a vertical tube, and the second is flow between parallel plates. Calculations were performed by using the Galerkin finite-element method. In view of heat transfer augmentation, the structure of the separating and reattaching flow around a surface-mounted obstacle was investigated in detail, together with the corresponding temperature field. It was shown that the reattachment distance depends on the Reynolds number; the isotherm contours are strongly distorted and pushed down toward the wall due to the back flow in the recirculating zone; the maximum Nusselt number occurs on the upstream side of the reattachment point; and the augmentation effect due to the obstacle is remarkable only in a limited turbulent range of relatively low Reynolds number.
The effects of pH and the addition of such salts as NaCl on the structure of the gel-cake in dead-end ultrafiltration of BSA solutions are examined by using a batchwise filter which has an abrupt reduction in its filtration area. Both the average porosity εav and the average specific filtration resistance αav of the gel-cake may be determined accurately on the basis of measurements of the variation of filtrate volume with time. The most compact gel-cake, which provides a large flow resistance, forms on the membrane around the isoelectric pH. By the addition of NaCl at pH extremes the gel-cake structure becomes compact, thereby decreasing the filtration rate. Such internal structures as the solute concentration and the compressive pressure acting on the solutes within the gel-cake, which may serve as a basis for evaluating the sweeping performance of the gel-cake in crossflow ultrafiltration, are evaluated on the basis of the overall filtration characteristics obtained under various filtration pressure conditions.
Experiments in bubbling fluidized beds show that gases which adsorb on solids experience up to ten times more rapid bubble-emulsion interchange than do nonadsorbing gases. We propose that the very small fraction of solids dispersed in and passing through the bubbles are the vehicle for this behavior. A theory is developed which shows that the enhancement in the overall bubble-emulsion interchange coefficient is strongly dependent on the value of the gas–solid adsorption equilibrium constant. The enhancement predictions from theory are then compared with values reported in the literature. This mass transfer phenomenon is then extended to gas–solid heat transfer in fluidized beds. The results indicate that the particles dispersed in bubbles usually play the dominant role in the heat interchange between bubble gas and emulsion gas.
For an aerated agitated vessel (AAV) treating various foaming liquids, the foam-breaking characteristics of a rotating-disk mechanical foam-breaker (MFRD) fitted to the AAV were evaluated. Foaming behavior of the AAV and foam-breaking behavior of the MFRD under varying operating conditions were related to the changes in liquid holdup in the foam. The agitation power in required the AAV with MFRD was found to be smaller than that in the AAV using antifoam agents (AFs), due to increased gas holdup. Comparison of the power input between two AAVs when foaming was controlled by the MFRD and AFs also demonstrated the superiority of mechanically controlled foaming.
In an internally circulating fluidized bed, the bubble properties (bubble chord length, volume fraction, rising velocity and frequency) in the fluidized bed and the solids circulation rate from a fluidized bed to a moving bed were determined, along with the effects of gas velocity on solids circulation rate and bubble properties in the fluidized bed. The solids circulation rate increases with increase in gas velocity. The measured bubble chord length distribution in the fluidized bed is well represented by the logarithmic normal distribution. The measured bubble chord length, volume fraction and rising velocity increase with bed height and gas velocity to the fluidized bed. However, bubble chord length, volume fraction and local bubble frequency decrease with increasing gas velocity to the moving bed. The bubble chord length, volume fraction and rising velocity are lower in the bed wall region and exhibit a peak value at a dimensionless lateral distance of 0.5–0.8, and the peak moves to the center region of the bed with increasing bed height. The rate of increase in bubble rising velocity increases with bubble size up to about 0.07 m, at which the flow regime may lie between bubbly and slugging flows. The bubble properties in the internally circulating fluidized bed have been correlated with the solids circulation on the basis of theoretical equations in the literature.
The deposition of aerosol particles of 0.01–2 μm in diameter is investigated in laminar and turbulent flow fields, and deposition by air flow, particle diffusion and gravitational sedimentation is evaluated. In experiments, wall loss rates of monodisperse particles are measured in four stirred tanks (similar in form but different in size) with stirring Reynolds number Re ranging from 10 to 104. The result has shown that deposition velocities vd obtained for particles smaller than 0.1 μm, which range over several orders of magnitude, are almost proportional to Re0.56 when Re is larger than about 1500. At Re < 200, a dimensionless mass transfer rate (Sherwood number) Sh, which expresses a dimensionless deposition rate, is found to be equal to 1.1Re1/2Sc1/3, where Sc is the Schmidt number. The deposition velocities for 200 < Re < 1500 show that the increasing rate of vd with Re depends on Re, indicating the existence of a transitional regime concerning deposition. The influence of gravity is observed when the particle diameter exceeds 0.2 μm, and the deposition rate can be predicted well by summing the contributions of diffusive deposition and gravitational sedimentation of a particle.
The Marangoni convection during steam absorption into aqueous LiBr solution with a small amount of surfactant was observed by Schlieren photography, and the surface tension of the absorbing solution was measured to investigate the mechanism of the convection. Also, a numerical simulation of the Marangoni convection was carried out. It is found experimentally that the surface tension of LiBr solution with surfactant decreases with increasing LiBr concentration and that the Marangoni convection in this system is not essentially induced by the presence of surfactant islands on the surface of the solution as proposed by the previous workers, but is in accordance with the general criterion of Marangoni instability. The calculated results could simulate qualitatively the initiation and growth of convection in this system.
A new type of viscometer for measuring human blood viscosity is made by using a vacuum glass tube and a capillary needle. This new viscometer needs a volume of only about 5 × 10–6–8 × 10–6 m3 of blood. Simple handling, quick measurement and accurate results in a wide shear rate range are its superior features. Additionally, the whole measuring set with the viscometer is easy to use and to carry, and it is not expensive. By using various standard-viscosity Newtonian liquids, it is confirmed experimentally that this viscometer is sufficiently accurate in the range of less than about 10–2 Pa·s. The viscosity of fresh human blood collected in a hospital is measured by the new viscometer and it is confirmed that the viscosity value obtained is reasonable by comparing it with that by a traditional measuring method. Further, it is shown that the new viscometer can measure the viscosities of other liquids which have complex properties.
The transfer rates of two metal ions across a cation exchange membrane were measured for the systems Na+–K+, Na+–Mg2+ and Na+–Ca2+. The ion exchange equilibrium was also measured for these systems. The distribution ratio of metal ion between aqueous phase and membrane phase for a system of two metal ions could be estimated with the ion exchange equilibrium constants obtained for the system of a single metal ion. The permeation rates of metal ions across the membrane were calculated by a model based on the nonequilibrium exchange kinetics at the membrane surface, which was in good agreement with the experimental results. The permeation selectivity of ion, which might depend on the ion exchange rate at both surfaces of the membrane, differed largely with the selectivity by the ion exchange equilibrium.
The equipment layout of a process plant is a multi-objective problem in which not only various costs (piping, site and so on) but also preferences about the equipment arrangement influencing its operability, maintenance and the like. It is difficult to obtain the best solution of this problem analytically. In this study, preferences are weighted as penalities so that they can be evaluated in an objective function with costs. To reduce the calculation load, equipment having spatial relations in a local area is put together into “modules” as units in layout work, and these modules are grouped into “sections” as functional groups of equipment. Furthermore, the module arrangement in each section is considered to be two variables (permutation and partition), and an algorithm based on an evolutionary method is developed to search a good plot plan efficiently. The effectiveness of this proposed method is demonstrated by an example problem.
Unsteady simultaneous measurement of the drag coefficients and rates of mass transfer for a chain of uniform water-, hexane- and pentane drops moving in air were made by using an optical method with a microcomputer technique for RePA = 279–1185, Sc = 0.61–1.30, diameter ratios DB/DA = 1.0, dimensionless distance between the centers of drops L/DA = 8.1–67 and dimensionless driving force BM = 0.01–1.35. The observed drag coefficients of a chain of water drops were compared with the theoretical ones proposed in our previous papers. A new correlation is proposed for the drag coefficients of a chain of uniform volatile drops by taking into account the effect of the interaction between drops and of high mass flux. A new correlation of the diffusion fluxes of a chain of uniform volatile drops is proposed by taking into account the effect of the interaction between two drops, of high mass flux and of variable physical properties.
For the polymorphic forms A and B of the organic compound cimetidine, their solubilities in H2O, 2-propanol (IPA) and H2O–IPA mixture and their crystallization behavior, mainly from IPA, were studied. It became apparent that form A is more soluble than form B in any solvent and that the solubilities of both forms increase in the order H2O, IPA, H2O–IPA mixture. In IPA as a solvent, at high supersaturation ratio (SA ≥ 3.6), form A, which is of thermodynamically metastable form, was preferentially crystallized regardless of the presence or absence (SA ≥ 4.5) and the form of seeds (SA ≥ 3.6). Phase transition from form A to B was not observed. The primary nucleation rate of form A was also measured at SA = 4.8–6.3 by the waiting-time method and surface energy σA was estimated to be 8.1 × 10–3 J·m–2. At lower supersaturation ratio (SA ≤ 2) the form corresponding to that of the seed was crystallized, contrary to results reported in earlier papers that no form B is obtained from IPA solution. The growth rate of form A was also measured at SA = 2.2–1.3, by the light transmittance method, and σA was estimated to be 5.2 × 10–3 J·m–2.
Unstable flow phenomena in a mixing vessel with a marine propeller were investigated experimentally using flow visualization and LDV techniques. It was found that the instantaneous bulk flow profiles were usually asymmetric with respect to the impeller shaft and they fluctuated in the large time scale. The circulation flow patterns in the space between two adjacent baffles were classified into quick-return flow (QR), full-circulation flow (FC) and intermediate-circulation flow (IC), which were replaced randomly one after another with time. Their lifetimes were widely distributed from about half a second to several minutes. Connected with the circulation flow patterns, the impeller stream was directed vertically downward or obliquely outward to the vessel wall and held a high level of angular momentum when pattern FC took place, while the impeller stream was directed obliquely inward to the impeller shaft with a low level of angular momentum when pattern QR took pace.
Pressure drops across a perforated-plate distributor, with and without a bed present, were investigated and a disagreement was found. A “real minimum uniform fluidization velocity,” Urmu f, was defined at which the distributor pressure drop was consistent with that in an empty bed. An approach for modifying the orifice equation when the gas velocity was smaller than Urmu f was proposed.