Search

The mobilization pressure was measured for a train of bubbles in a capillary tube when the liquid film forming between the bubble surface and the inner surface of the duct does (Case B) and does not cover the bubble surface (Case A). In Case A, the measured bubble mobilization pressure mostly agreed with the estimated capillary pressure for different numbers of bubbles in the duct. In Case B, the bubble mobilization pressure was lower, with a larger scatter than in Case A. Increasing the number of bubbles caused the mobilization pressure to converge to the Case B results. Factors that affect the mobilization pressure were considered.

A simple and pump-free chemiluminescence immunoassay based on a compact disc (CD)-type microfluidic platform for the determination of alkylphenol polyethoxylates (APnEOs) is described. The method is based on a competitive immunoreaction of the anti-APnEOs antibody immobilized on the magnetic microbeads between APnEOs and horseradish peroxidase (HRP)-labeled APnEOs in the sample solution. The luminol solution containing H2O2 and enhancer is caused to flow from one reservoir in the platform to another by appropriate adjustment of the speed of the rotation of the disc. The detection limit was similar to that reported in our previous paper, i.e., 10 ppb according to IC80.

When two immiscible liquids make contact in a microchannel, the flow pattern is affected by the affinity between channel walls and liquids. In this study, microchannels (200 µm in width and 200 µm in depth) having a T-shaped bifurcation point were fabricated on PMMA plates. The inner walls of the microchannels were modified in a zone-selective manner to be either hydrophilic or hydrophobic, based on verification accomplished via a laser interference fringe technique. The microchannel was placed horizontally, and water and octane were introduced into the upper-side channel (hydrophilic) and into the lower-side channel (hydrophobic), respectively. The experimental results showed that water and octane formed a stable layered flow, and the two liquids were virtually completely separated at the T-shaped section, even when static pressure was intentionally applied to the outlets. CFD simulation, using FLUENT 6.3 software, was performed to explain the role of zone-selective modification of microchannel walls.

単位操作をマイクロチャンネルで実現する一助として，アルキメデスらせん形のマイクロチャンネル中に，固体粒子（直径22および40 μmのガラス粒子）を懸濁した水を流し，粒子を分離して排出する操作の可能性を調べた．チャンネルの幅および深さはそれぞれ700 μmおよび500 μm，最内半径は2.5 mm，最外半径は5.0 mmとした．マイクロチャンネルは水平または垂直に保った．らせん内側端から懸濁液を送入し，2.5巻きの地点で2方向に分岐して取り出した．チャンネル内で粒子は2次流れ（Dean渦）によって運ばれ，かつ重力で沈降するために，内外出口における濃度が異なった．採取した懸濁液試料は乾燥後に秤量し，粒子の分離度を求めた．水平設置の場合：粒子の沈降速度がDean渦の速度に比べて大きくなると，粒子はチャンネル内側の底部に集まった．粒子の沈降速度が，底部壁近傍を流れて内側壁で上昇する流れの速度よりも小さくなると，底部に集まった粒子は上向きDean渦に巻き上げられ，チャンネル内に分散し，出口の分離度が低下した．垂直設置の場合：粒子が集まる壁面は液流の方向によって変化することが分かった．本実験では，頂部を起点として時計回りに90°回転した位置に分岐点を設置したので，液流速が低い条件下において粒子は内側出口に濃縮された．以上の実験結果はCFDシミュレーションと合致した．また，液と粒子の速度分布，ならびに粒子濃度分布のシミュレーション結果は，液固分離用のマイクロらせんの設計と操作条件の選定に有意義な示唆を与えた．

地熱発電所で実際に用いられる遠心力を利用した2重円筒型ミストセパレータについて, 蒸気の流動と液滴の捕集性能との関係をシミュレーションによって検討した. 装置下部に接線方向に供給された蒸気は, 装置内を回転しながら上昇し, 塔頂近くに開口部がある内筒に入り込み, 下向きに排出されてタービンに至る. 蒸気の中の液滴は回転に基づく遠心力によって外壁に向かって移動し, 壁面に衝突して捕集される. 内筒の直径は1m, 外筒の直径は3m, 気液の供給管の直径は0.8mである. 供給管の高さ位置から排出管の入口までの距離(分離高H)は3.4, 13.4mとした. 気体は0.79MPaの飽和蒸気であり, 22m/sの速度で接線方向に送入されるとした. H=3.4mのとき, 捕集可能な液滴直径は20-30μmである. Hを高くすると蒸気の回転運動が安定化し, 蒸気は上昇するに従って回転半径を小さくしながら内筒に入る. この内側に向かう流れで遠心力の効果が弱められ, H=13.4mの場合でも捕集可能な液滴径は50-60μmにとどまる. 2重円筒型の場合, 流れの安定性と捕集効率の関係から, H=7-8mとするのが妥当と思われる.

The purpose of the present study was to investigate fluidic behavior in a microchannel with hair-pin curves, using a three-dimensional (3-D) computational fluid dynamics simulation, and to observe the 3-D flow pattern, to validate the simulation. The microchannel used was fabricated on a PMMA plate using a flat-end mill. The channel width and depth were 210 and 205 μm, respectively, and the radius of each hair-pin curve was 500 μm. Two liquids; purified water and an aqueous solution of 50 μmol/L fluorescein, were introduced into the microchannel through different inlets and were merged, forming a side-by-side parallel flow in the straight channel. When the average velocity was 25 mm/s, the liquid was thrust outward by centrifugal force and, as a result, the vertical line that crossed the central axis was distorted after passing the first hair-pin curve. At the second hair-pin curve, the centrifugal force was exerted in the opposite direction, and the distorted line returned nearly to an initial vertical line. When the average velocity was 125 mm/s, however, the vertical line, which was distorted at the first hair-pin curve, did not recover to the initial vertical line after the second curve. The interface between the two liquids was permanently waved. The simulation was in good agreement with the experimental data. The results suggest that the diffusion rate through the interface of two liquids in microchannels with hair-pin curves can increase, compared to that in straight microchannels. © 2004 American Institute of Chemical Engineers AIChE J, 50: 1530–1535, 2004

AlPO 4 -type membranes were prepared on porous α-alumina tubes by a hydrothermal synthesis method in the presence of HF, using tetraethylammonium hydroxide (TEAOH) as the organic template. The gas permeation properties of the membranes were determined at permeation temperatures in the range of 308–423 K, using H 2 , CO 2 , O 2 , N 2 and iso -C 4 H 10 . For the membrane prepared by a two-step hydrothermal synthesis (membrane M3), the ideal separation factors for single-component H 2 /N 2 , H 2 /CO 2 , and H 2 / iso -C 4 H 10 systems at 310 K reached 13.3, 23.9 and 40.1, respectively. With increasing permeation temperature, permeances of the membrane increased, and separation factors of H 2 to the other gases decreased. The separation factors for the binary H 2 /N 2 , H 2 /CO 2 , H 2 / iso -C 4 H 10 systems were lower than the corresponding ideal separation factors. The presence of humidity in the feed led to a decrease in permeance, and the separation factors of membrane M3 for the equimolar binary H 2 /N 2 , H 2 /CO 2 , H 2 / iso -C 4 H 10 systems at 310 K were 12.0, 31.7 and 28.2, respectively.

The permeation and separation properties through an FAU-type zeolite membrane for mixtures of benzene and saturated C 4 –C 7 hydrocarbons were investigated at a permeation temperature of 373 K. The data related to permeation were cited from previous papers, and the new data on adsorption were obtained for such binary mixtures with and without benzene. The high benzene/cyclohexane separation factor can be attributed to the high adsorption selectivity in favor of benzene and the blockage against cyclohexane. The proposed model is as follows: For the feed of single-component benzene or cyclohexane, the adsorption of benzene is higher than that of cyclohexane, while the diffusivity of benzene is lower than that of cyclohexane. As a result, permeance to cyclohexane is higher than that to benzene. For the case of the feed consisting of an equimolar mixture of benzene and cyclohexane, the adsorption of benzene is smaller than that for the case of a single-component feed. Meanwhile, cyclohexane molecules are expelled from the adsorption sites by benzene, and the benzene/cyclohexane selectivity for adsorption reaches a value of approximately 19. The benzene molecules which are moving toward the permeate side frequently block the windows, and the diffusivity of cyclohexane through the windows is greatly retarded. The small separation factors for mixtures of saturated C 6 hydrocarbons can be explained based on a single-file diffusion mechanism.

An evaporator unit for use in microreactor systems was constructed. The unit consisted of a micropump driven by piezoelectricity and a microevaporator heated electrically by means of a Pt wire. The micropump was comprised of two tightly bonded plates; a silicon (100) plate having three diaphragms and a glass plate having valve and pump chambers. These flow conduits were patterned by photolithography and formed by wet-etching. The diaphragms, as well as the valve and pump chambers, were 7 mm in diameter, and were arranged triangularly. The discharge rates of the pumps were found to be dependent on the voltage and frequency applied to the piezo discs, the actuation pattern, the backpressure at the outlet, and the liquid viscosity. The pump assembled with an epoxy adhesive showed high discharge rates, while a similar one, assembled by anodic bonding, showed high discharge pressures. The flow rate of the former pump reached 60 mLlmin for water. The microevaporator channel was wet-etched on a 10 mm x 40 mm silicon wafer, and a channel for a Pt wire heater was formed on the reverse side. Both sides of the silicon plate were covered with glass plates. Benzene, when pumped to the microevaporator, was evaporated to give a vapor flow rate of 6.8 cm 3 /min, which is sufficiently large as a feed rate for a gas-phase catalytic microreactor.

NaY-type zeolite membranes were synthesized on the outer surface of an α-alumina support tube by a hydrothermal process. The membranes were ion-exchanged with 0.01–0.50 mol/l aqueous solutions of KCl, RbCl or CsCl at 80 °C for 4 h. Gas permeation properties of the ion-exchanged membranes were determined at 35 °C for single component systems of CO 2 and N 2 and for an equimolar mixture, thereof. The CO 2 /N 2 separation factor for permeation increased with increasing ion-exchange degree, which was dependent on the cation species. When the membrane was treated in a 0.50 mol/l solution of KCl, RbCl and CsCl, the ion-exchange degree reached 96, 78 and 62%, respectively. The CO 2 /N 2 separation factors of the Y-type zeolite membranes ion-exchanged with K, Rb and Cs were 55, 47 and 46%, respectively, at a permeation temperature of 35 °C. The NaY-type zeolite membranes, which were impregnated with Cs ions, showed the highest CO 2 /N 2 separation factor of 149 at 35 °C. These results suggest that the CO 2 /N 2 separation factor was largely determined by the affinity between the membranes and the permeants.

Y-type zeolite membranes were prepared on porous support tubes by hydrothermal synthesis, and then ion exchanged using Pt, Ru, Rh, Co, Ni, Cu or Ag (referred to as PtY, RuY, RhY, CoY, NiY, CuY and AgY membranes, respectively). The permeances to H2 and CO through the NaY membrane at 373–523 K were (8.3–9.6)×10−7 and (4.6–5.3)×10−7 mol m−2 Pa−1 s−1, respectively. The permeances of the other membranes did not differ significantly from those of the NaY membrane. The metal-loaded Y-type zeolite membranes were then used for the selective oxidation of CO in an H2-rich mixture (H2:CO2:CO=89:10:1 on a molar basis). The permeation fluxes of CO for the CoY, NiY and AgY membranes were not affected by the addition of O2. For the case of the PtY, RuY, CuY and RhY membranes, however, the permeation fluxes of CO decreased with increasing O2 concentration, when a small amount of O2 was added to the feed. This suggests that these membranes had activity for the CO oxidation. The PtY membrane was the most highly active under the conditions used.

This article reports on the development of a neutron absorption tracer (NAT) technique for evaluating the hydrodynamic properties in direct coal liquefaction reactors. The plant was constructed at Kashima, Japan, based on the concept of the NEDOL process, and possessed three reactors which were connected in series. The reactors were 1.0 m in diameter and 11.8 m in height, and could treat 150 ton coal/day. In order to determine the residence time distribution (RTD) data of the liquid phase in the reactors, 252 Cf was used as the neutron source, and a fine gadolinium oxide powder was used as the tracer. The tracer slurry was injected into the inlet pipe of the first or third reactor, and the response was detected at the outlet pipe of the first and third reactors. Preliminary tests were performed to obtain the design principles for the neutron howitzer and the instrument system. Low- and middle-energy neutrons were more effective than thermal neutrons for the detection of the RTD of the liquid phase. The tracer particles were estimated to behave in the same manner as the liquid under conditions of coal liquefaction. The tracer injection tests were then performed 64 times during the pilot plant (PP) operation. The accuracy of the NAT technique was confirmed by comparing the gas holdups calculated from the RTD data by the NAT with those directly determined using a shutdown technique of the recycle gas. The RTD data suggested that the flow in the reactors was classified in the homogeneous bubble flow. These results show the applicability and effectiveness of the NAT technique.

An NaY-type zeolite membrane was formed on the outer surface of a 1.7-mm I.D. porous α-Al2O3 support tube using a hydrothermal synthesis. NaY-type zeolite powder was also prepared by the same procedure. The permeation and separation characteristics of single-component benzene (B) and cyclohexane (C), as well as an approximately equimolar mixture thereof, were investigated using the membrane at permeation temperatures of 358–413K. For the single-component systems, C permeated through the membrane faster than B, and the permeances to B and C increased with increase in permeation temperature in the ranges (0.6–3.0)×10−7 and (2.3–6.4)×10−7 mol m−2 sec−1 Pa−1, respectively. When the mixture was fed, however, B permeated faster than C, and the separation factor was dependent on the partial pressures on both the feed and permeate sides. The B/C separation factor reached 107 at a permeation temperature of 373K. The amounts of B and C, which were adsorbed to the NaY-type zeolite powder, were also determined at thi...

A zeolite membrane was formed on a porous α-alumina support tube by hydrothermal reaction using a solution which was appropriate for the synthesis of ETS-4 zeolite. The morphology and x-ray diffraction pattern of the top surface of the support tube were consistent with those of the ETS-4 zeolite. The permeances to single-component H2, CO2, O2, N2, and CH4 were in the range (0.8–2.5)×10− 8 mol m− 2 sec− 1 Pa− 1, while those to n-C4H10 and iso-C4H10 were less than 10− 11 mol m− 2 sec− 1 Pa− 1. This suggests that the membrane functioned as a molecular sieve with pores smaller than 0.4 nm. At a permeation temperature of 310K, the ideal N2/O2 separation factor was 2.8, and the N2/O2 separation factor was 2.3 and 3.2 for a mixture of and 4, respectively. Thus, the membrane is nitrogen-selective and can be used to enrich the oxygen level in air, which is fed on the pressurized side, without much loss of compression energy.

FAU-type zeolite membranes with different Si/Al ratios were hydrothermally synthesized on the outer surface of a porous α-Al2O3 support tube. The permeances of the membranes to CO2, CH4 and N2 were then measured at 308 K for single-component and equimolar binary systems. The separation properties were dependent on both the Si/Al ratio and the ion-exchange treatment. For single-component systems, a lower Si/Al ratio resulted in the incorporation of a larger number of Na+ ions. For a CO2-CH4 mixture, both CO2 permeances and CO2/CH4 selectivities were approximately half the values obtained for a binary CO2-N2 mixture. The highest selectivities, obtained using the NaX(1) zeolite membrane, were 28 for CO2/CH4 and 78 for CO2/N2. The RbY, RbX(1) and RbX(2) zeolite membranes showed larger CO2 permeances, compared with those of the original Na-type membranes. Ion-exchange with K+ ions was the most effective for the NaY zeolite membrane in that both the CO2 permeance and the CO2/CH4 selectivity were increased.

In order to oxidize CO, which is diluted in H2, two types of catalytic membranes have been developed. In this study, the characteristics of the membranes are briefly summarized for a typical reaction temperature of 523 K , and the permeation and oxidation mechanism of those membrane reactors are analyzed. (i) A defect-free Y-type zeolite membrane was synthesized on the inner surface of a porous α-alumina support tube by a hydrothermal process. The membrane was ion-exchanged and calcined in air and, finally, reduced in a flow of H2, to give a Pt-loaded Y-type zeolite membrane (PtY membrane). A mixture of H2, CO (10 000 ppm ) and O2 (0– 12 000 ppm ) was permeated through the PtY membrane at 523 K . The total pressure on the feed and permeate sides was maintained at atmospheric pressure, and the permeate side was swept with an inert gas. The PtY membrane rejected CO at a H2/CO separation factor of approximately 10. When the O2 feed rate exceeded the CO feed rate of the stoichiometry, the CO concentration on the permeate side, 1000 ppm , was decreased to less than 8 ppm . Since H2 and CO are able to enter the Y-type zeolitic pores without size discrimination, the H2/CO separation factor can be attributed to the slow diffusivity of CO in the membrane. Thus, CO has a longer residence time in the pores, and the oxidation proceeds effectively. (ii) A γ-Al2O3 layer was formed on the outer surface of a porous α-Al2O3 support tube using a boehmite sol. The layer was then impregnated with Rh using an aqueous solution of RhCl3, and calcined in air at 573 K . A microporous silica layer was then formed on the surface of the Rh-impregnated γ-Al2O3 layer using a silica sol. The SiO2 membrane showed a H2/CO separation factor of 100. Thus, the CO concentration of 50 000 ppm on the feed side was decreased, with no feed of O2, to 500 ppm at 523 K on the surface of the Rh/γ-Al2O3 layer. When O2 was added to the feed, CO was oxidized in the Rh/γ-Al2O3 layer, and a corresponding decrease in CO concentration on the permeate side was observed.

In order to prepare metal-loaded catalyst membranes for CO oxidation, γ-Al 2 O 3 layers were formed on porous α-Al 2 O 3 support tubes (1.7 mm i.d., 2.1 mm o.d.) using a sol–gel technique. The layers were then impregnated with Pt, Ru, Ni, Co and Rh, using aqueous solutions of H 2 PtCl 6 , RuCl 3 , Ni(NO 3 ) 2 , Co(NO 3 ) 2 and RhCl 3 , respectively, and calcined in air at 573 K. A gaseous mixture of H 2 , CO and O 2 was permeated through these membranes, and their activities were determined with respect to CO oxidation at 423–523 K. The CO oxidation activity was the highest for the Rh/γ-Al 2 O 3 membrane. Based on the report that the CO oxidation activity was much reduced for higher concentrations of CO than the threshold value, a membrane was developed by forming a microporous silica layer on the Rh/γ-Al 2 O 3 membrane. By the selective permeation with no feed of O 2 , the CO concentration of 50,000 ppm on the feed side could be decreased to 610–2700 ppm. When O 2 was added to the feed, the CO concentration was further decreased. Thus, the composite membrane, which was developed in the present study, can play dual roles: H 2 -selective permeation with a high H 2 /CO separation factor, and catalytic oxidation of CO which leaked through the microporous silica layer.

When CO is removed from H2 by selective oxidation using a catalytically active membrane, the CO is oxidized during its permeation through the catalytic layer to the permeate side. Since the oxidation rate of CO is proportional to the inverse of its concentration, the rejection of CO using an H2-selective microporous layer which is overlaid on the zeolite layer is also effective. In this study, Y-type zeolite membranes with different thicknesses were prepared on the outer surface of porous α-Al2O3 tubes by hydrothermal synthesis. The membranes were modified to different extents by ion-exchange with Pt ions. Gas permeation properties of the membranes were then determined using single-component H2 and CO, as well as mixtures of H2, CO, O2 and Ar. Thus, the effects of the membrane thickness, the amount of Pt loading and the size of micropores on CO oxidation could be evaluated. For the Pt/NaY membrane prepared by hydrothermal synthesis for 12 h and ion-exchanged using a 2.0-mmol L–1 solution of [Pt(NH3)4]Cl2, the H2 permeance and the H2/CO separation factor were 1.1 × 10–6 mol m–2 Pa–1 s–1 and 1.9 at 473 K. The thickness of the zeolite layer had no detectable effect on CO oxidation. When the membrane was ion-exchanged with a 10.0-mmol L–1 solution of [Pt(NH3)4]Cl2, the CO/H2 mole fraction ratio on the permeate side was minimized to 0.007 at 493 K for a feed of H2:CO:O2:Ar = 4:2:1:93 on a molar basis. The membranes were further modified with SiO2 using dip-coating and chemical vapor deposition techniques. The latter technique led to a membrane, the outer surface of which was coated with an H2-selective microporous SiO2 layer. This resulted in a decrease in CO concentration in the zeolite layer. Thus the CO oxidation proceeded effectively in the catalytically active zeolite layer, and the CO/H2 mole fraction ratio on the permeate side was 0.01 at 523 K for a feed of H2:CO:O2:Ar = 4:2:1:93 on a molar basis.

The vapor permeation properties of an FAU-type zeolite membrane, which was formed on the outer surface of a porous α-Al2O3 support tube, were determined in the temperature range of 358–413 K. The measurements were carried out using single-component benzene (Bz) and n-alkanes (C4-C7), as well as mixtures of benzene and n-alkanes. The hydrocarbons were diluted with nitrogen, and the partial pressures of hydrocarbons on the feed side were determined to be 7–8 kPa for the single-component systems and 4–5 kPa for each component of the binary systems. The permeate side was swept with a flow of nitrogen. The permeances were dependent on the permeation temperature and, to a greater extent, on the presence of benzene. For the single-component systems, the permeances were dependent on the diffusivities of the permeants, and the diffusivity of benzene through the membrane was the smallest of all the hydrocarbons tested. For the binary systems, however, benzene permeated at faster rates than any of the n-alkanes, and the permeances to n-alkanes were not directly dependent on their chain lengths. This benzene-selective permeation was realized using the FAU-type zeolite membrane with pore openings larger than the molecular size of benzene. The separation factors were affected by the partial pressures of permeants on the permeate side. When the partial pressures on the permeate side were in the range of 100–140 Pa for benzene and 1–4 Pa for n-alkanes, the separation factors at 373 K were 57 for Bz/n-C4, 70 for Bz/n-C5, 63 for Bz/n-C6, and 27 for Bz/n-C7. The adsorption selectivities for the binary systems on the feed-side surface of the membrane were 24 for Bz/n-C4 and 15 for Bz/n-C7, and were much higher than those for the single-component systems. This suggests that the adsorption of the n-alkanes on the FAU-type zeolites was strongly retarded by the selective adsorption of benzene.

In order to evaluate the activities of granular catalysts, a microreactor having six channels was fabricated using silicon and glass plates by means of photolithography and wet etching. Using anodic bonding, the plates were configured so as to be gas tight. The dimensions of the channels were 1 mm in width, 0.3 mm in depth and 6 mm in length. Catalysts for the oxidation of carbon monoxide in hydrogen, as examples, were prepared by impregnating Pt, Rh, Ru, Co, Ni, Cu and Pd on γ-alumina support particles which were sieved to 150–210 μm, and 9-mg aliquots of the catalysts were packed in the channels. The experiments were carried out in the temperature range of 120–200°C, and the compositions in the products were determined by quadrupole mass spectrometry. The findings show that the microchannel reactor could be used for the efficient screening of granular catalysts.

Zirconia membranes were prepared on porous α-alumina tubes using a sol-gel technique, and platinum was then impregnated in the pores of the membranes by dipping in 0.1 mol L–1 H2PtCl6 solution, followed by calcination in air. This procedure was repeated four times maximum. The membranes were then exposed in hydrogen prior to permeation tests. Permeation rates through the membranes were determined for single-component gases (H2, CO2 and i-C4H10) at temperatures of 308–573 K. The permeation rates of the unmodified zirconia membrane obeyed the Knudsen diffusion mechanism. However, hydrogen permeation changed from the Knudsen diffusion to surface diffusion after the third Pt modification, and to the activated diffusion after the fourth Pt modification. The ideal H2/CO2 and H2/i-C4H10 selectivities for the membrane, which was Pt-modified four times with calcination at 573 K, were approximately 30 and 1000, respectively, at a permeation temperature of 523 K. When a mixture of H2 (80 mol%) and CO2 (20 mol%) was fed, the membrane showed an H2/CO2 selectivity higher than 400 and an H2 permeance of 4.21 × 10–8 mol m–2s–1Pa–1 at 523 K.

Zirconia is more chemically stable than silica and γ-alumina, which are commonly used in the preparation of inorganic membranes. However, the hydrolysis of zirconia alkoxides, typical zirconia precursors, proceeds rapidly, leading to difficulties in preparing clear sols. Thus, an attempt was made to reduce the hydrolysis rate of zirconium propoxide by a ligand-exchange reaction with 1,5-diaminopentane (DAP). The availability of water was also reduced by the dropwise addition of glacial acetic acid, which produced an ester by reaction with the solvent, 1-propanol. The addition of DAP was effective in preparing zirconia with a narrow pore-size distribution and a high mesopore volume. DAP also decreased the phase transformation temperature of zirconia and had almost no effect on the micropore structures. The gas permeation properties of the prepared zirconia membranes were fundamentally controlled by the Knudsen diffusion mechanism.

Zeolite membranes, which were composed of ETS-4 with Na cations, were prepared on porous α-alumina tubes by hydrothermal synthesis. The membranes, which were formed under optimized conditions, sharply rejected molecules with sizes larger than 0.4 nm. For mixtures of N 2 –CO 2 , N 2 –O 2 , N 2 –Ar and N 2 –CH 4 systems, N 2 permeated faster than the coexisting gas. The N 2 /O 2 separation factor for an equimolar mixture was in the range of 2.3–3.5, and the N 2 permeance was in the range of (0.55–2.8)×10 −8 mol m −2 s −1 Pa −1 at permeation temperatures of 283–333 K. Moisture had some effect on the permeation properties for N 2 –O 2 mixtures. The separation factor for the N 2 /CH 4 system was larger than that of the N 2 /O 2 system. When the membrane was ion exchanged with either Li + or Sr 2+ , the separation factors for N 2 /O 2 and N 2 /CH 4 systems increased, while the permeances decreased.

Upper Freeport coal was extracted using a mixture of carbon disulfide and N-methyl-2-pyrrolidinone (CS2/NMP) under ultrasonic irradiation at room temperature. The CS2/NMP-soluble fraction (CS2/NMP-...

Microporous zirconia membranes, in which the pore size distribution was controlled by optimizing the calcination temperature, were prepared by a sol-gel technique. The membrane with an average pore size 2 SO 4 ) solutions, followed by heat-treatment. The gas permeation properties of the membranes were evaluated for single-component and binary-component systems of H 2 , N 2 , CO 2 , iso-C 4 H 10 and NH 3 . The zirconia membrane, which was modified by treatment with a 0.50 mol l −1 H 2 SO 4 solution, showed a high hydrogen permeance of (2.8–3.0)×10 −6 mol Pa −1 m −2 s −1 , and the H 2 selectivities over CO 2 and iso-C 4 H 10 were 6 and 9, respectively, for equimolar binary systems. The experimental data showed that the H 2 selectivity over the other gases increased, and the permeance decreased, when increasing concentrations of the H 2 SO 4 solution were used. The permeance to NH 3 for the H 2 SO 4 -modified membranes suggests that acidic sites were introduced as a result of the modification.

NaA-type zeolite membranes were prepared on a 1.7-mm i.d. porous α-alumina tube by repeating a hydrothermal reaction. The membranes, which were prepared by four repetitions, were ion-exchanged with Ca2+ and K+, and the permeation for single-component and mixed gases were investigated at 35°C. The NaA-type membrane showed a hydrogen permeance of (0.8–2.6) × 10−8 mol · m−2 s−1 Pa−1 and a hydrogen/n-butane ideal separation factor of 12–15. The permeability properties of the membrane for binary mixtures were dependent on the size of the permeating molecules and their adsorptivity to the zeolite. The KA-type zeolite membrane showed a higher ideal separation factor for a hydrogen/nitrogen system than did the NaA- and CaA-type zeolite membranes. The pores of the CaA-type zeolite membrane were larger than those in the NaA- and KA-type zeolite membranes, and showed higher permeances to n-butane and i-butane. All the zeolite membranes contained non-zeolitic pores, which decreased the selectivity of permeation.

A defect-free Y-type zeolite membrane was synthesized on the inner surface of a porous α-alumina support tube (i.d.=1.7 mm, o.d.=2.1 mm) by a hydrothermal process. The membrane was ion-exchanged with an aqueous 2.5 mmol l−1 solution of [Pt(NH3)4]Cl2 to give a Pt-loaded Y-type zeolite membrane (PtY membrane), was obtained by calcination in air and by reduction in a flow of H2. The PtY membrane was then applied to a membrane reactor system for selective oxidation of carbon monoxide, which was included in H2 at a concentration of 104 ppm. The total pressure on the feed and permeate sides was maintained at atmospheric pressure, and the permeate side was swept with an inert gas. The H2 flux through the PtY membrane was independent on the O2 concentration in the feed. When the feed rate of O2 was 1.2 times that of CO, the CO concentration on the permeate side was decreased to less than 8 ppm at temperatures of 200–250°C. When steam was added to the feed at a concentration of 2%, the H2 and CO fluxes through the membrane remained fundamentally unchanged.

Since the micropore size of Y-type zeolite is in the range of 0.74– 1.2 nm , small molecules are able to pass one another inside the micropores. Thus ion-exchanged Y-type zeolite membranes are expected to show catalytic activity as well as being useful for gas separation with high permeation rates. In this study, NaY-type zeolite membranes were synthesized by a hydrothermal reaction on the inner surface of a porous α -alumina support tube. The Y-type zeolite membranes were stable to heat at temperatures up to 400°C, and the permeances to single-component hydrocarbons showed a maximum at temperatures that were dependent on the heat of adsorption. The permeances to C 3 H 6 , n -C 4 H 10 and i -C 4 H 10 decreased with time at elevated temperatures. This can be explained by the deposition of carbonaceous matter, which was removed by a heat treatment in an atmosphere of O 2 at 300–400°C. The permeation mechanism for the NaY-type zeolite membrane was investigated for single-component CO 2 . The Maxwell–Stefan diffusivities were compared with the Fickian diffusivities obtained using a sorption–diffusion model, in which linear concentration profiles across the membrane were assumed. The differences between the Fickian diffusivities and the Maxwell–Stefan diffusivities were negligible at elevated temperatures.

The purpose of this article is to demonstrate the applicability of microreactors for use in catalytic reactions at elevated temperatures. Microchannels were fabricated on both sides of a silicon wafer by wet chemical etching after pattern transfer using a negative photoresist. The walls of the reactor channel were coated with a platinum layer, for use as a sample catalyst, by sputtering. A heating element was installed in the channel on the opposite surface of the reactor channel. The reactor channel was sealed gas-tight with a glass plate by using an anodic bonding technique. A small-scale palladium membrane was also prepared on the surface of a 50-Μm thick copper film. In the membrane preparation, a negative photoresist was spin-coated and solidified to serve as a protective film. A palladium layer was then electrodeposited on the other uncovered surface. After the protective film was removed, the resist was again spin-coated on the copper surface, and a pattern of microslits was transferred by photolithography. After development, the microslits were electrolitically etched away, resulting in the formation of a palladium membrane as an assemblage of thin layers formed in the microslits. The integration of the microreactor and the membrane is currently under way.

NaY-type zeolite membranes, which were synthesized on the outer surface of a porous α-alumina tube by a hydrothermal process, were ion-exchanged with 0.1 mol l −1 solutions of either KCl, RbCl or CsCl. The degrees of ion-exchange were 62, 40 and 32% for K, Rb and Cs, respectively. Permeation properties of the ion-exchanged zeolite membranes were investigated at 35°C for single-component and binary mixture systems of CO 2 and N 2 . The CO 2 /N 2 selectivity for the equimolar mixture was in the order: Rb=K>Cs>Na. The sorptivities of CO 2 and N 2 onto the zeolites were also determined from sorption isotherms for the single-component and binary-mixture systems at 35°C. The CO 2 /N 2 sorption selectivity was higher for the binary mixture systems than that for the single-component systems, and this was the major reason for the increase in permeation selectivity for the mixed feed.

The development of nitrogen-selective membranes is essential to realize a membrane process for the production of oxygen-enriched air. Zeolite membranes have potential for this purpose, while polymeric and carbon membranes are largely oxygen-selective. In this study, FAU-type zeolite membranes containing Na cations were prepared by hydrothermal synthesis on a porous α-alumina support tube. Permeation measurements for unary and binary systems for N2 and O2 were performed at 310 K. The membranes showed high permeation rates and high N2/O2 ideal separation factors for unary N2 and O2 gases. However, the separation factor was decreased when an N2-O2 mixture was fed. In order to further examine the permeation mechanism of the membranes, permeation rates were also determined for single-components of H2, CO2, CH4, n-C4H10, i-C4H10 and SF6, as well as for N2-H2, N2-CO2, and N2-n-C4H10 mixtures. The permeation properties can be classified based on competitive adsorptivity and diffusivity in the zeolitic pores. Ion-exchange and impregnation with Li cations were not greatly effective in increasing the N2/O2 separation factor of the membrane.

Microporous silica membranes for hydrogen separation were prepared on a γ-alumina coated α-alumina tube by sol-gel method. The reactants of sol-gel chemistry were tetraethoxysilane (TEOS) and methacryloxypropyl-trimethoxysilane (MOTMS). The silane coupling agent, MOTMS, was added as a template in order to control the pore structure to the silicon alkoxide, TEOS. In particular, the microporous membranes were prepared by changing the molar ratio of MOTMS with respect to other substances, and their pore characteristics were analyzed. Then, the effects of thermal treatment on the micropore structure of the resulting silica membranes were investigated. The pore size of the silica membrane prepared after calcination at 400–700 ‡C was in the range of 0.6–0.7 nm. In addition, permeation rates through the membranes were measured in the range of 100–300 dgC using H2, CO2, N2, CH4, C2H6, C3H6 and SF6. The membrane calcined at 600 ‡C showed a H2 permeance of 2×10-7-7×10-7 molm-2s-1Pa-1 at permeation temperature 300 ‡C, and the separation factors for equimolar gas mixtures were 11 and 36 for a H2/CO2 mixture and 54 and 132 for a H2/CH4 mixture at permeation temperatures of 100 ‡C and 300 ‡C, respectively.

Direct coal liquefaction test plants have been constructed and successfully operated under subsidies from the New Energy and Industrial Technology Development Organization. To analyze the data, which had been collected on these plants, a simulator was constructed for the material and enthalpy balances in the preheaters and reactors. This simulator was evaluated in a process supporting unit (PSU) with a capacity of 1 ton of dry coal/day and a pilot plant (PP) with a capacity of 150 tons of dry coal/day, from the standpoint of vapor−liquid equilibria, reaction rate constants, and hydrodynamics in the reactors. Reaction rate coefficients were determined in the PSU by iteration until the calculated yields at the outlet of the third reactor coincided with the observed data. The kinetic data were then applied to the PP, which was operated under conditions similar to those of the PSU, and the liquefaction product yields at the outlet of the third reactor were calculated. The calculated values were in general agr...

The direct coal liquefaction pilot plant at Kashima, Japan, based on the NEDOL Process, was equipped with three bubble column reactors, each of which was 1 m in diameter and 11.8 m in length, connected in series. In order to investigate the hydrodynamic behavior of liquid in the reactors, a gadolinium tracer was injected to the inlet of the first or third reactor, and the concentration of the tracer was measured at the outlet of the reactor by a neutron absorption tracer technique. Residence time distribution curves were thus obtained under liquefaction conditions. A pseudo-three-dimensional two-phase fluid dynamic model was developed based on a mixed Eulerian–Lagrangian representation with the Kato–Launder k–e model for turbulent stresses in the reactors. A dynamic simulation was then carried out using a finite-volume solution algorithm. The physical properties of gas and liquid in the reactor were estimated, taking into consideration the evaporation of oil to the gas phase. The model adequately described the measured residence time distribution curves.

Hydrodynamic properties in the coal liquefaction reactors at the Kashima pilot plant, which was constructed based on the NEDOL process, were investigated using the neutron absorption tracer technique. The reactor system is composed of three vessels, each with 1.0 m ID and 11.8 m in height. The gas velocity in the reactors under coal liquefaction conditions was estimated using a reaction simulator that contained reaction rates and vapor–liquid equilibrium. The axial dispersion coefficients in the first and third reactors at superficial gas velocities of 0.06–0.07 m/s were much smaller than those reported for air–water systems under ambient conditions. This suggests that the pilot-plant reactors operated fundamentally in the homogeneous bubble flow regime.

An iridium (100) layer was epitaxially coated on a MgO (100) plate by sputtering at 1123 K, and was then utilized in the formation of diamond by microwave plasma-assisted chemical vapor deposition (MPCVD) using methane as the carbon source. The electric contact between the substrate and holder was confirmed by coating the entire MgO surface with iridium. The iridium substrate was then treated by bias-enhanced nucleation under optimized conditions. It was found that diamond particles formed by MPCVD were essentially oriented to the iridium substrate. The diamond particles were then grown to the 〈100〉 and further to the 〈111〉, and a smooth diamond film was obtained. The full width at half maximum of the (400) rocking curve of the diamond film was 0.16°, which was close to that of a diamond single crystal.

Prime novelty: The smoothness of the synthesized boron-doped diamond was improved by the pre-treatment of a hydrogen plasma. Moreover, the Hall mobility also increased with this pre-treatment. Surface morphology and electrical properties, such as electrical conductivity, hole concentration and Hall mobility, were investigated for boron-doped diamond films, which were synthesized by microwave-assisted chemical vapor deposition (MPCVD) on a (100) diamond substrate. Trimethylboron (TMB) was used as a dopant source and methane (CH4) was used as a carbon source. The morphology of the synthesized diamond surface depended on the MPCVD conditions such as TMB and CH4 concentrations in the gas phase, and lower concentrations of TMB and CH4 lead to a smoother surface. When the substrate was treated in a hydrogen plasma, the electrical properties of the boron-doped diamond films, as well as the smoothness of the surface, were improved. After optimizing the synthesis conditions, Hall mobility reached to 2020 cm2 V−1 s−1 at 243 K for a diamond film with a hole concentration of 5×1012 cm−3.

A direct coal liquefaction plant was constructed in Kashima, Japan, based on the concept of “NEDOL Process”, and successfully processed 150 tons of Tanitorarum coal per day during 1997−1998. The plant was equipped with three reactors: 1 m in diameter and 11.8 m in length, connected in series. During the operation, solid particles were accumulated mainly in the first reactor. Slurry samples were directly removed from the reactors, and size distributions of solid particles were determined. Two types of particles were found: particles with cores and particles without cores. The size of the former particles was in the range of 10−200 μm, while that of the latter particles was 1−80 μm. The growth rate of the particles was estimated to be 0.10 nm s-1 under the reaction conditions. The solid accumulation in the first reactor was classified into a dense region in the lower part of the reactor and a lean region extending above the dense region. The former region was expressed as a three-phase fluidized bed model...

A highly hydrophilic A-type zeolite membrane was prepared by a hydrothermal synthesis on the outer surface of a porous α-alumina tube. Increasing and decreasing the permeation temperature between 308 and 473 K over 40 cycles in a period of 4 months did not damage the membrane. Permeation tests with a variety of permeates showed that the membrane possessed two types of pores: zeolitic pores of 0.4−0.43 nm diameter and nonzeolitic pores. Molecules larger than C2H6 were not able to permeate through the zeolitic pores. The separation factors of the membrane were dependent on molecular size, affinity to the pore walls, and hydrophobicity of the permeating molecules. The combination of permeants also affected the permeation rates. H2O molecules could be concentrated in the zeolitic and nonzeolitic pores, thus reducing the permeation of the hydrophobic permeants. The H2O/H2 separation factor was larger than 160 at permeation temperatures of 303−473 K.

A coal liquefaction pilot plant of the NEDOL process, supported by the New Energy and Industrial Technology Development Organization (NEDO), was in successful operation for a total of 269 days at Kashima, Japan. The liquefaction section consisted of three reactors, 1 m in diameter, connected in series, and processed 150 tons of coal per day. Product yields were modeled assuming 13 reaction paths, and a steady-state simulator was developed. The performance of the entire process, including distillation and solvent hydrogenation sections, was further incorporated into the simulation. The design procedure for the liquefaction reactors, which would possess a diameter of 4 m and process 2500 tons of coal per day, was discussed in sight of the simulator.

Hydrogen permeation through a palladium disk membrane (5 mm in diameter and 50 μm in thickness) was investigated in the presence of steam, methane, propane and propylene. Co-existing methane and propane had a negligible effect on hydrogen permeation. When propylene was fed with hydrogen, however, the permeance was greatly decreased, and the extent of deactivation increased with time. Hydrogen permeation was independent of the carbon concentration in the membrane, and a carbonaceous matter, which was formed by the decomposition of propylene was the major origin of the deactivation. The addition of steam decreased the hydrogen permeation rate for the hydrogen–methane and hydrogen–propane mixtures but attenuated the decrease in the hydrogen permeation rate for the hydrogen–propylene mixture. The carbonaceous matter which was chemisorbed to the membrane could be gradually removed by treatment with hydrogen at 600°C, and the hydrogen permeation rate was recovered to 80% of the initial value after a week of treatment time. Regeneration using oxygen at 600°C was effective in removing the carbonaceous matter, and the permeation rate was fully recovered within 15 min.

Skeletal isomerization of 4,6-dimethyldibenzothiophene (4,6-DMDBT) was performed at 270 °C for 1 h at a hydrogen pressure of 2.5 MPa using a Ni-supported Y-type zeolite catalyst. Alkyldibenzothiophenes (alkyl-DBTs), which were produced by methylmigration and transalkylation, were further desulfurized at 255−300 °C for 2−120 min at a hydrogen pressure of 2.5 MPa using a CoMo/Al2O3 catalyst. The produced alkyl-DBTs, which carried 1−3 methyl groups, could be classified into three groups on the basis of activation energy for desulfurization. Group I included 2,8-DMDBT and C1-DBT, which carried no methyl groups at either the 4- or 6-position. Group III included unreacted 4,6-DMDBT, and group II contained other alkyl-DBTs with 1−3 methyl groups. No alkyl-DBTs with 4 methyl groups were found in the products of 4,6-DMDBT. The steric hindrance of the methyl groups on 4,6-DMDBT was relieved by methylmigration and transalkylation, and, as a result, the desulfurization activation energy, 37.2 kcal/mol, for 4,6-DMDBT ...

Illinois No. 6 coal was depolymerized by treatment with trifluoromethanesulfonic acid in methylcyclopentane at 120 °C. The yield of tetrahydrofuran (THF) extract was increased to 89 wt %. This incr...

Hydrogenated diamond powder was treated with an anhydrous chloroform solution of 2,2′-azobisisobutyronitrile as the radical initiator and sulfuryl chloride as the chlorination reagent at 323 K under an Ar atmosphere. After the chlorination process, the solution was exhaustively evaporated, and n-butylamine was added without exposure to air. The diamond was then butylaminated at 303 K for 1–24 h. The treated diamond powder was washed with tetrahydrofuran and water to remove by-products, and characterized by infrared spectroscopy and flash pyrolysis. It could be confirmed that the hydrogenated diamond surface was modified by the radical reactions in the liquid phase. However, the formation of covalent bonds between butylamine groups and carbons on the chlorinated diamond surface did not constitute a major reaction under the conditions of the present study.

The direct coal liquefaction plant at Kashima, Japan, was equipped with three reactors, each of which was 1 m in diameter and 11.8 m in length, connected in series. This plant was designed on the basis of the concept of the NEDOL Process, and processed 150 tons of coal per day. The steady-state behavior of the reactors was simulated using an axial dispersion model which took into consideration the liquefaction reactions and the evaporation of oil to the gas phase. The model was validated from axial temperature profiles as well as coal conversion and hydrogen consumption data, obtained from the actual operation of the reactors. The heat of reaction, estimated from the heat balance between the inlet and outlet streams in each reactor, was determined to be 2.1 MJ per 1 m 3 (STP) of hydrogen consumed. The axial dispersion coefficient of the slurry phase was estimated to be 0.03-0.04 m 2 /s at a superficial gas velocity of 0.06 m/s. The axial dispersion coefficient was smaller and the gas holdup was larger than values reported in previous studies.

A coal liquefaction pilot plant of the NEDOL process, supported by New Energy and Industrial Technology Development Organization (NEDO), was in successful operation for a total of 269 days at Kashima, Japan. The liquefaction section involved three bubble-column liquefaction reactors, 1m in diameter, connected in series, and processed 150 tons of coal per day. The hydrodynamics and thermal behavior of the reactors were investigated. The relations of yields and actual slurry residence time measured by the neutron absorption tracer technique were shown in the case of Tanitoharum coal.The procedure for designing the 4m diameter reactors to process 2, 500 tons of coal per day, was developed based on the yields estimated by the reaction simulator validated by the data obtained on the pilot plant. The ratio of heavy oil fraction in recycled solvent was a determinant factor to maintain the stable operation. The design was also supported by the model study estimating the thermal behaviors, discussing the effect of mixing on the thermal efficiency.

In this study, we fabricated oriented A-type zeolite membranes by repeating hydrothermal reactions, and determined permeances to single-component hydrogen and nitrogen. The crystal orientation of the membrane was characterized by SEM and XRD.