Unformatted Document Excerpt
University of Texas
Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
Eng. Ind. Chem. Res. 2001, 40, 1277-1300
REVIEWS Progress and New Perspectives on Integrated Membrane Operations for Sustainable Industrial Growth
Enrico Drioli* and Maria Romano
Institute on Membranes and Modeling of Chemical Reactors, CNR, and Department of Chemical Engineering and Materials, University of Calabria, 87030 Arcavacata di Rende (CS), Italy
Membrane science and technology has led to significant innovation in both processes and products over the last few decades, offering interesting opportunities in the design, rationalization, and optimization of innovative productions. The most interesting developments for industrial membrane technologies are related to the possibility of integrating various of these membrane operations in the same industrial cycle, with overall important benefits in product quality, plant compactness, environmental impact, and energetic aspects. Possibilities for membrane engineering might also be of importance in new areas. The case of transportation technologies is of particular interest. The purpose here is to present a summary review of the extent to which membrane processes have been integrated into industrial practice. Some of the most interesting results already achieved in membrane engineering will be presented, and predictions about future developments and the possible impact of new membrane science and technology on sustainable industrial growth will be analyzed.
Introduction Membrane science and technology has led to significant innovation in both processes and products, particularly appropriate for sustainable industrial growth, over the past few decades. The purpose here is to present a summary review of the extent to which membrane processes have been integrated into industrial practice. The preparation of asymmetric cellulose acetate membranes in the early 1960s by Loeb and Sourirajan is generally recognized as a pivotal moment for membrane technology. They discovered an effective method for significantly increasing the permeation flux of polymeric membranes without significant changes in selectivity, which made possible the use of membranes in largescale operations for desalting brackish water and seawater by reverse osmosis and for various other molecular separations in different industrial areas. Today, reverse osmosis is a well-recognized basic unit operation, together with ultrafiltration, cross-flow microfiltration, and nanofiltration, all pressure-driven membrane processes. In 1999, the total capacity of reverse osmosis (RO) desalination plants was more than 10 millions m3/day, which exceeds the amount produced by the thermal method,1 and more than 250 000 m2 of ultrafiltration membranes were installed for the treatment of whey and milk. Composite polymeric membranes developed in the 1970s made the separation of components from gas
* Corresponding author: IRMERC-CNR c/o Department of Chemical Engineering and Materials, via Ponte P. Bucci, 87030 Arcavacata di Rende (CS), Italy. Tel.: (39) 0984492039/492025. Fax: (39) 0984-402103. E-mail: firstname.lastname@example.org.
streams commercially feasible. Billions of cubic meters of pure gases are now produced via selective permeation in polymeric membranes. The combination of molecular separation with a chemical reaction, or membrane reactors, offers important new opportunities for improving the production efficiency in biotechnology and in the chemical industry. In 1997, five large petrochemical companies announced a research project devoted to the development of inorganic membranes to be used in syngas production. At about the same time, an $84 million project, partly supported by the U.S. Department of Energy (DOE), that has Air Products and Chemical Inc. working together on the same objective has been promoted. The availability of new high-temperature-resistant membranes and of new membrane operations as membrane contactors offers an important tool for the design of alternative production systems appropriate for sustainable growth. The basic properties of membrane operations make them ideal for industrial production: they are generally athermal and do not involve phase changes or chemical additives, they are simple in concept and operation, they are modular and easy to scale-up, and they are low in energy consumption with a potential for more rational utilization of raw materials and recovery and reuse of byproducts. Membrane technologies, compared to those commonly used today, respond efficiently to the requirements of so-called "process intensification", because they permit drastic improvements in manufacturing and processing, substantially decreasing the equipment-size/ production-capacity ratio, energy consumption, and/or waste production and resulting in cheaper, sustainable technical solutions.2 The possibilities of redesigning innovative integrated
10.1021/ie0006209 CCC: $20.00 2001 American Chemical Society Published on Web 02/13/2001
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Table 1. Sales of Membranes and Modules in Various Membrane Processes5 1998 sales (millions of U.S. dollars) 900 500 400 230 110 70 >10 30 growth (%/year) 8 10 10 15 5 5 ? 10
membrane process microfiltration ultrafiltration reverse osmosis gas separation electrodialysis electrolysis pervaporation miscellaneous
membrane processes in various industrial sectors characterized by low environmental impacts, low energy consumption, and high quality of final products have been studied and in some cases realized industrially. Interesting examples are in the dairy industry and in the pharmaceutical industry. Research projects are in progress in the leather industry and in the agrofood industry based on the same concept. In this review, some of the most interesting results already achieved in membrane engineering will be presented, and predictions about future developments and the possible impact of new membrane science and technology on sustainable industrial growth will be analyzed. Actual possibilities and future perspectives of medical and biomedical applications of membrane technology are not discussed in this work. This theme is the object of another recent paper.3 The continuous interest and growth of the various new industrial processes related to life sciences, as evidenced also by the strategies and reorganization adopted by large chemical groups worldwide in this area (e.g., Aventis, Novartis, Vivendi Water, etc.) will also require significant contributions from membrane engineering. We will, however, not concentrate our analysis on this subject in this review. Membrane Operations Various membrane operations are available today for a wide spectrum of industrial applications. Most of them can be considered as basic unit operations, particularly the pressure-driven processes such as microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and RO; electrodialysis (ED) is another example of a mature technology.4 Their worldwide sales are reported in Table 1.5 The significant variety of existing membrane operations is based on relatively simple, compact, and largely clarified fundamental mechanisms characterizing transport phenomena in the dense or microporous membrane phases and at the membrane-solution interface. The understanding and prediction of transport phenomena in the membrane phase is today at least qualitatively possible, also theoretically through the newly available tools provided by molecular simulation.6,7 Much progress has been made in this area in recent years in the design of polymeric materials, such as polyimides, etc., and in the calculation of the diffusion coefficients of simple gases in the dense phase. An interesting agreement can be found between the theoretical and experimental values.8,9 It is the integration of advanced knowledge about transport phenomena in dense or microporous thin phases, combined with the understanding of interfacial
phenomena controlling the adsorption and desorption of penetrants and other species at the membrane surfaces, with the correct flow-dynamic analysis of the tangential flow and concentration profile built up in the bulk solutions upstream and in the membranes downstream and with the reology of often concentrated nonNewtonian fluids, that permits the design of correct membrane separation units. Membrane operations show potential in molecular separations, clarifications, fractionations, concentrations, etc. in the liquid phase, in the gas phase, or in suspensions. They cover practically all existing and requested unit operations used in process engineering. All of the operations are modular, easy to scale-up, and simple to design. Other important aspects are the lack of moving parts; ability to work totally unattended; lower cost; operational flexibility; and, when necessary, portability. Coupling of molecular separations with chemical reactions can be realized in a simple unit efficiently, having ideal reaction surfaces where the products can be continuously removed and the reagents continuously supplied at stoichiometric values. These overall properties make membrane operations ideal for the design of innovative processes where they will carry on the various necessary functions integrated eventually with other traditional unit operations, optimizing their positive synergic effects. It is interesting to mention that statistical analysis carried out by Electricite de France on 174 different membrane installations in France using MF, UF, RO, and ED mainly in small- and medium-sized industries found a normal percentage of satisfaction between 70 and 95%, one of the highest positive responses received in this kind of analysis. This result is, in part, surprising because of the high innovative content of the technology and the lack of education still existing on their basic properties. It is, however, consistent with the important contributions that membrane operations can make in terms of cost reduction, quality improvement, pollution control, etc. Several examples of successful applications of membrane technology as alternatives to traditional processes can be mentioned. Ion-Exchange Membranes. The use of ion-exchange membrane cells in chloro-soda production represents, for example, an interesting case study for analyzing the possibilities of membrane operations and one of the first successes in terms of their electrochemical application in minimizing environmental impacts and energy consumption. The technology is based on the discovery and utilization of fluorinated polymeric membranes stable in a specific environment, such as Nafion. Today, membrane systems in which the anodic and cathodic species are directly produced in separate compartments without mixing and final separation problems permit one to overcome the limitations of traditional mercury cells, related to the need for Hg recovery, and of diaphragm cells, in which the separation and concentration of final products still create difficulties. All new chloro-soda installations are now practically based on this design, which represents a typical rationalization of the process, removing all of the pollution problems that characterized chloro-soda productions in the past. In principle, other molecular halogens could be produced from their respective gases. The direct production
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1279
Table 2. Worldwide Desalination Production Capacitya country Saudi Arabia U.S. United Arab Emirates Kuwait Japan Libya Qatar Spain Italy Bahrain Oman total capacity (m3/day) 5 250 000 3 100 00 2 200 00 1 500 00 746 000 685 000 570 000 530 000 520 000 310 000 190 000 % of world production 25.9 15.2 10.7 7.6 3.7 3.4 2.8 2.6 2.6 1.5 0.9 MSF (%) 67.5 1.7 89.8 95.5 4.7 67.7 94.4 10.6 43.2 52.0 84.1 MEE (%) 0.3 1.8 0.4 0.7 2.0 0.9 0.6 0.9 1.9 0.0 2.2 VC (%) 1.2 4.5 3.0 0.0 0.0 1.8 3.3 15.1 15.1 1.5 0.0 RO (%) 31.0 78.0 6.5 3.4 86.4 19.6 0.0 20.4 20.4 41.7 11.7 ED (%) 1.9 11.4 0.2 0.3 6.8 9.8 0.0 19.2 19.2 4.5 0.0
a Phase-change processes: MSF (multistage flash), ME (multi-effect evaporation), VC (vapor condensation). Single-phase processes: RO (reverse osmosis), ED (electrodialysis).
Table 3. Costs Related to Various Sea Water Desalination Processes energy consumption thermal thermal mechanical mechanical electric energy equivalent (kWh/m3) 10-14.5 6-9 7-15 4-8 scale of application small-large small-medium small small-large cost for 1 m3 of freshwater produced (ECU) 0.6-1.9 0.5-2.0 0.6-2.4 0.4-1.4
processa MSF ME VC RO
maturity very partly partly yes
a Phase-change processes: MSF (multistage flash), ME (multi-effect evaporation), VC (vapor condensation). Single-phase processes: RO (reverse osmosis), ED (electrodialysis).
of essentially dry chlorine gas would also reduce oxygen formation, which allows the reaction to be run at much higher current densities, with much less purification and drying required compared to the chlorine produced by other systems. Reverse Osmosis and Nanofiltration. As already mentioned, desalination of seawater and brackish water has been at the origin of the interest for membrane operations, and the research efforts on reverse osmosis membranes have had an impact on all of the progress in membrane science and technology. Evaporation plants have been substituted with RO systems in different part of the world (Table 2).10 The relatively low energy consumption is one of the reasons for this success (Table 3). In seawater desalting, in fact, the global energy consumption of RO, with a recovery factor of 30% and energy recovery, has been 5.32 kWh/m3 corresponding to a primary energy consumption of 59.94 kJ/kg.11 Costs for brackish water desalination are 60-70% lower than those for seawater desalination. RO desalination is not only devoted to the production of drinkable water but today is also strategic in various industrial sectors and particularly in ultrapure water production for the electronic industry. It is interesting to realize that, in Japan, the largest part of the water produced by RO is for the electronics industry, in which the country has worldwide leadership. Reverse osmosis has not generally been used until recently in the purification, separation, or concentration of chemicals, particularly because of osmotic limitations and the low chemical and thermal resistance of the existing membranes. The recent development of nanofiltration and lowpressure reverse osmosis membranes with interesting selectivities and fluxes, as well as higher chemical and thermal resistances, has been rapidly utilized in the realization of innovative processes in various industrial sectors. An interesting case studied in Italy is represented by the preparation of Iopamidol in the pharmaceutical
industry.12 Recently, in X-ray diagnosis, the use as contrast media of new nonionic iodinated compounds as opacifying agents was studied and introduced to the markets as a substitution for the traditional iodinated ionic compounds. However, the preparation and, particularly, the final purification of these products were much more complex and expensive than for those previously used. In particular, the neutral iodinated agents cannot be isolated by precipitation in water because of their high solubilities. The problems to be solved were particularly the removal of ionic species, usually inorganic salts present in the final reaction mixture and the recovery of valuable reagents present in excess and of the water-soluble reaction media. A technique was developed based on the treatment of the raw solutions of the contrast media with a complex series of operations such as removal of the solvent (DMAC or DMF) by evaporation; extraction of the residual reaction medium by a chlorinated solvent; elution of the aqueous phase on a system of cationic and anionic ion-exchange resins; concentration by evaporation; and crystallization of the crude residue to remove the last impurities. Various drawbacks are present in this system. A much better system has recently been realized based on the use of two nanofiltration stages operating on highly concentrated raw solutions containing the contrast media, inorganic salts, organic compounds with a relatively low mass (about 200), and the solvents (Figure 1). The first NF unit operates in diafiltration mode and the retentate, partially concentrated and purified with respect to contrast media, is recycled at the first stage after dilution with a small amount of deionized makeup water; the permeate (water, inorganic salts, solvents, etc.), which still contains small amounts of the iodinated compound, proceeds toward the second NF unit. The permeate from this second step is completely contrast agent free. The degree of purification that can be reached is such that the total amount of residual impurities in the final
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 1. Recovery of Iopamidol by membrane process.12
recycled retentate does not exceed 10% of the initial amount and is generally on the order of 5%. The process is simple, economical, and environmentally acceptable; it permits the elimination of acid and basic reactants necessary for the regeneration of the resins; and it avoids the use of toxic organic solvents, etc. Also, the integrated membrane processes proposed for chromium recovery in the leather industry13 and for the treatment of secondary textile effluents for their direct reuse,14 which will be described and discussed later in this work, show efficient applications of nanofiltration and low-pressure reverse osmosis operations. Microfiltration and Ultrafiltration. Recently, in the food industry, membrane technology made realistic the possibility of cold sterilization. Tetra Pak (BactoCatch System) developed a cross-flow microfiltration system that debacterizes fresh milk, avoiding any thermal treatment and taste alteration. An industrial process using this technology is already in operation at Villefranche (Lyon) producing 2000 L/day of fresh milk registered with the trade name Marguerite (Figure 2). The skimmed milk, obtained by whole milk centrifugation, is sterilized at low temperature by microfiltration. Then, it is mixed with pasteurized cream. After homogenization and cooling, a debacterized whole milk is obtained using a process alternative to the classical UHT (ultrahigh-temperature) treatment. Similar products have also recently been commercialized in Italy by Parmalat S.p.A. The current systems for cleaning oil-water streams via cross-flow microfiltration or ultrafiltration are very reliable and compact. They can decrease the oil content of water from 10-30 mass % to less than 5 ppm. Nitrogen blanketing helps to prevent oxidation of oils during mechanical oilseed pressing, while also reducing explosion risks in extraction and during desolventizing. New possibilities exist, however. Solvent recovery, dehydration of solvents, use of membrane reactors, winterization, and fractionation of fats are interesting cases. More than two million tons of extraction solvents, mainly hexane, is used in the U.S. alone. Its recovery is by distillation and condensation. It is estimated that, also in the most modern units, 0.7 kg of hexane per ton of seed is still released into the environment. The
possibilities of recovering solvents from the oil-micelle mixture and from air exist today with membrane operations that might significantly reduce these losses. A reduction of the solvent content of the oil-micelle mixture from 70 to 40% has been demonstrated, with an energy saving of about 50%. An important aspect of the utilization of membrane operations in this area will be the possibility of using other solvents such as alcohols. Their higher evaporation heats make them unattractive in traditional evaporation units. Better solvent-resistant membranes, eventually inorganic ones, however, will be necessary for large-scale applications in this area. Cross-flow microfiltration can also be used successfully for the removal of long-chain traces of saturated fat that are present in, e.g., sunflower oil.15 Considerable advances in UF and MF technologies in water purification processes for drinking water production have been achieved to such a point that, presently, more than 1 000 000 m3/day of water are treated using these membrane operations.16,17 The employment of integrated membrane systems in the production of drinking water is growing rapidly with excellent results. The reliability of the reverse osmosis membrane is greatly increased when UF or MF operationsswhich emerged in the past decade as an efficient way to remove suspended solids and organic and microbiological contaminantssare used in the pretreatment step. Furthermore, economical considerations have shown that multiple membrane systems are more competitive than conventional processes, resulting in the reduction of capital and operating costs. In addition to the already-mentioned membrane operations, gas separation, pervaporation, and some others membrane processes, which have recently shown significant possibilities for their application in various industrial areas, must be cited; among these, a class of membrane-based unit operations identified as membrane contactors, membrane bioreactors, and catalytic membrane reactors will be discussed. Gas Separation. Membrane processes for gaseous mixture separation are, today, technically well-consolidated and apt to substitute for traditional techniques.18 Separation of air components, natural gas dehumidification, and separation and recovery of CO2 from biogas
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1281
Figure 2. Flow sheet of an industrial system for the debacterization of fresh milk by cross-flow microfiltration (Villefranche, Lyon, France).
and of H2 from industrial gases are some examples in which membrane technology is applied at the industrial level. The gas separation business was evaluated in 199619 at $85 million in the U.S., with growth of about 8% per year. Asymmetric polymeric membranes, used for gas mixture separations, are made either as plane sheets and assembled in spiral-wound modules or as hollow fibers. These modules are made and commercialized by various companies all over the world. Although the kind of module used is declared, the type of polymer is still protected as industrial know-how. In Table 4, some permeability and selectivity data for the various polymers used in the manufacture of the most commercial membranes are reported.20,22 The separation of air components or oxygen enrichment has advanced substantially during the past 10 years. The oxygen-enriched air produced by membranes has been used in various fields, including chemical and related industries, the medical field, food packaging, etc. In industrial furnaces and burners, for example, injection of oxygen-enriched air (25-35% oxygen) leads to higher flame temperatures and reduces the volume of parasite nitrogen to be heated; this means lower energy consumption. Mixtures containing more than 40% v/v of O2 or 95% v/v of N2 from the air can be obtained. Industrial nitrogen is used in the chemical industry to protect fuels and oxygen-sensitive materials. Membranes today dominate the fraction of the nitrogen market for applications less than 50 tons/day and
relatively low purity (0.5-5% O2). Single-stage operation is preferred. Oxygen is the third largest commodity chemical in the U.S. with annual sales in excess of $2 billion. Whereas nitrogen membrane separation has been a great success, oxygen separation using membranes is still underdeveloped. The major reason for this is that most of the industrial oxygen applications require purity higher than 90%, which is easily achieved by adsorption or cryogenic technologies but not by membranes. Today's limited application of membrane-based oxygen generation systems operate either under feed compression or permeate vacuum mode (Figure 3). Both methods of separating oxygen are inferior to the adsorption separation processes using various zeolites. New materials are being developed that could possibly have higher permeabilities than conventional solid electrolytes, in which ionized atoms are transported through the crystalline lattice under a driving force provided by partial pressure differences over the membrane (pressure-driven process) or by electrical potential gradients (electrochemical pumping). Mixed conductors with high electronic and oxygen ion conductivities could be used as a membrane alternative to solid electrolytes for oxygen separation. In such materials, both oxygen ions and electronic defects are transported in an internal circuit in the membrane material. Promising oxygen permeation fluxes have been obtained in many perovskite systems, e.g., La-Sr-CoFe-O,23,24 Sr-Fe-Co-O,25,26 and Y-Be-Co-O.27 In particular, in the ITM-oxygen systems, simultaneous conduction of ions and electrons in the same
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Table 4. Permeability and Selectivity Data of Some Polymers Used in the Manufacture of Commercial Membranes for Gas Separationa permeability coefficient, barrer CO2 poly[1-(trimethylsilyl)-1-propyne] poly(dimethylsiloxane) poly(dimethylsilmethylene) poly(cis-isoprene) poly(butadiene-styrene) natural rubber (at 25 c) ethyl cellulose polystyrene butyl rubber poly(ethyl methacrylate) poly(phenylene oxide) (at 25 c) bisphenol A polycarbonate cellulose acetate bisphenol A polysulfone PMDA-4,4-ODA polymide poly(methyl methacrylate) poly(vinyl chloride) (at 25 C) PEEK-WC (at 25 C) polyphosphazeny (at 25 C)
selectivity (ideal) (-) N2 4970 351 35.9 14.5 10.3 9.43 3.4 0.52 0.324 0.33 3.81 0.38 0.15 0.19 0.1 0.02 0.0118 0.1 CO2/N2 5.60 13.0 14.5 13.2 16.6 16.2 22.1 23.8 16.0 21.2 19.9 17.9 31.7 24.2 27.0 31.0 13.3 27.5 21.2-30.5 O2/N2 1.55 2.22 2.53 2.60 3.19 3.65 5.58 5.76 4.21 5.47 6.32 6.10 7.00 5.00 3.71-5.05
O2 7730 781 91.0 37.5 32.9 12.4 2.9 1.9 1.6 0.82 1.2 0.61 0.14 0.5 0.955-1.72
28 000 4550 520 191 171 153.0 5.0 12.4 5.18 7.01 75.7 6.8 4.75 4.6 2.7 0.62 0.157 2.75 5.76-10.2
At 35 C unless otherwise specified.
Figure 3. Oxygen production systems.
Figure 4. Integrated oxygen and power production.
material obviates the need for an external electrical circuit to provide the driving force for the separation, with a significant reduction in cost. The driving force for the separation process is the partial pressure difference across the membrane. High-pressure air (100300 psia) is required to achieve a significant flux of O2 across the membrane. The oxygen flux is directly proportional to the pressure gradient and inversely proportional to the membrane thickness. The pressure of the oxygen product is typically only a fraction of an atmosphere. These dense inorganic perovskite type membranes, today manufactured in tubular configurations, transport oxygen as lattice ions at elevated temperatures with infinite selectivity ratios in O2 separations. The ionic
conductivity of the material studied is mainly equal to the electronic conductivity. Because this oxygen-ion-conducting membrane must operate at temperatures above 700 C, an effective means of recovering the energy contained in the nonpermeate, oxygen-depleted stream is required. An efficient and cost-effective means to accomplish this is to integrate the membrane system with a gas turbine (Figure 4).28 A technology known as OTM syngas (oxygen transport membrane synthesis gas) utilizes these ionconducting membranes able to separate oxygen from air with a high flux in the same temperature region required for the reforming of natural gas. This technology was presented in 1997 by an alliance of five
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1283
Figure 5. Scheme of a plant for H2 recovery from ammonia synthesis.
international companies including AMOCO, BP Chemicals, PRAXAIR, SASOL, and STATOIL. Philippe Petroleum joined the alliance in 1998. The new process, still under development, integrates the separation of oxygen from air, steam reforming, and natural gas oxidation into one step, eliminating the need for a separate oxygen plant. The new technology offers the possibility of reducing the energy and capital costs of syngas production. Considering that 60% of the cost for manufacturing any product from natural gas is related to synthesis gas production, the interest of this innovation technology is evident. Various plants for the recovery of hydrogen from the purge of the synthesis of ammonia have been realized today.29 The unit modules are in general arranged in a "one-stage-two-unit" form. One of the first plants of this type has been realized by Permea in Louisiana (Figure 5).30 The first unit, consisting of eight hollow-fiber modules [total feed capacity about 3800 m3(stp)/h] is operated with a transmembrane pressure difference of 60 bar, the permeate leaving at a pressure of about 70 bar. At this pressure, the permeate can be fed to the second stage of the synthesis feed compressor. The retentate of the first unit is fed to the second unit where the permeate leaving the modules at 25 bar is mixed with fresh feed (suction side of the first stage of the compressor). The retentate is utilized for heating purposes. Gas pretreatment consist of conventional scrubbing to reduce the ammonia content of the bleed from 2% (molar) to less than 200 ppm in order to avoid membrane swelling and, as a consequence, damage of the membrane. The economical and technical advantages related to this membrane system for the recovery of hydrogen are shown in Table 5. Methanol synthesis is another process based on a gaseous feed; in purge recovery, a water scrubber is also used with a similar purpose, and it pays for itself in terms of the recovered methanol. The methanol/water mixture is simply sent to the existing crude methanol distillation column. Hydrogen recovered from this purge can result in energy savings, and if additional carbon oxide is available, it can be used to obtain increased methanol production. PRISM separators operate on
Table 5. Economic and Technical Advantages for a 1000 ton/day Ammonia Plant30 ammonia recovery (scrubbing) heat saving 4 ton/day 522-836 kJ/ton of NH3 produced 50-55 ton/day 20-50 ton/day
additional ammonia production increase in ammonia production (at constant natural gas consumption) reduction in natural gas production (at constant production rate)
stoichiometric as well as nonstoichiometric H2/(CO)x ratio methanol plants at differential pressures up to 70 bar. Figure 6 shows the flow diagram of such a hydrogen recovery unit installed for demonstration purposes.31,30 Before entering the gas permeators, the feed is scrubbed in order to reduce the methanol content to levels below 100 ppm. From a bleed stream of 4000 mol/h, for example, a recovery of 2000 mol/h of hydrogen has been achieved. Gas mixture dehumidification is a process of great industrial interest, especially for natural gas purification and air dehumidification. An efficient membrane system for air dehumidification called the Cactus Membrane Air Dryer, developed in the late 1980s, has been commercialized by Permea.32 When the Cactus dryer is fed with compressed air, water vapor and a small amount of oxygen pass through the walls of the hollow fiber, while nitrogen, argon, and most of the oxygen continue through the hollow core of the fibers to the end of the separator. A small amount of the slower gases passes through the fiber, and this is used to sweep the water vapor through the separator. Cactus membranes work on the principle of dew point depression. For example, a membrane might be sized for inlet conditions of 100 psig and 100 F inlet dew point to achieve a 0 F pressure dew point. If inlet conditions change, e.g., compressed air with a lower inlet dew point is supplied, the separator will provide dry air at an even lower dew point. The removal of hydrogen sulfide (H2S) and carbon dioxide (CO2) from natural gas is an ideal application for membranes (Figure 7); both H2S and CO2 permeate through membranes at a much higher rate than meth-
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 6. Hydrogen recovery from the bleed of a methanol synthesis.
Figure 7. Removal of H2S and CO2 from natural gas.
ane, enabling a high recovery of the acid gases without significant loss of pressure in the methane pipeline product gases. These membrane processes are going to substitute for the more traditional methods of hydrocarbon stream purification. Through a comparison of the separation cost for the membrane process with that for the diethanolamine (DEA) gas-absorption process, it was found that the membrane process is more economical than the DEA gas-absorption process in the range of CO2 concentrations in the feed between 5 and 40 mol %. When the feed also contains H2S, the cost for reducing the CO2 and H2S concentrations in the feed to pipeline specifications increases with increasing H2S concentration (1000 to 10 000 ppm). If membrane processes are not economically competitive because of the high H2S concentration in the feed, the separation cost could be significantly lowered by using hybrid membrane processes. In such processes, the bulk of CO2 and H2S is separated from sour natural gas with membranes, and the final purification to pipeline quality gas is performed by means of suitable gas-absorption processes.33 Despite the high levels of H2S in the feed, membrane selectivities are maintained.34 The possibility of utilizing membrane technology in solving problems such as the greenhouse effect related to CO2 production has also been suggested. Membranes able to remove CO2 from air, having a high CO2/N2 selectivity, might be used at any large-scale industrial CO2 source as a power station in petrochemi-
Figure 8. Recovery of CO2 from exhaust gas and reuse to produce chemicals by hydrogenation.
cal plants. The CO2 separated might be converted by reacting it with H2 in methanol, starting a C1 chemistry cycle. As schematized in Figure 8, a membrane reactor might be ideally used to carry out hydrogenation reactions for chemical production using CO2 recovered from exhaust gases by membrane separation. The separation and recovery of organic solvents from gas stream is also rapidly growing at the industrial level. Polymeric rubbery membranes that selectively permeate organic compounds (VOC) from air or nitrogen have been used. Such systems typically achieve greater than 99% removal of VOC from the feed gas and reduce the VOC content of the stream to 100 ppm or less. The technology has been applied to the recovery of highvalue organic vapors such as vinyl chloride monomer, methyl chloride, and methyl formate. Membrane systems are competitive with carbon ad-
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1285
Figure 9. Flow diagram of compression/condensation and membrane separation for MVC recovery.
Figure 10. Flow sheet of two-stage recovery system of unreacted monomer and other volatile hydrocarbons from the nitrogen used during polymer particle degassing (MTR).
sorption or condensation for streams containing more than 5000 ppm, particularly if high VOC recovery is required. The typical industrial applications of vapor recovery are off-gas treatment in gasoline tank farms, gasoline station vapor return, and end of pipe solvent recovery in the chemical and pharmaceutical industries. Another interesting example of an industrial application is VOC recovery by the compression-condensation and vapor permeation method, presented schematically in Figure 9. This is a scheme of the process developed in Anwil (Wloclawed, Poland), which has been built by MTR (U.S.) for the recovery of monovinyl chloride (MVC). The recovery of ethylene and propylene from nitrogen in polyolefin plant vent streams has been suggested and realized at the industrial level by DSM in Geleen, The Netherlands. To recover the unreacted monomer (up to 25%) and other volatile hydrocarbons from the nitrogen used during polymer particles degassing, MTR35 developed a two-stage operation in which the mixture of N2 and propylene is first compressed and later directed into a membrane vapor separation unit, as shown in Figure 10. The spiral-wound membrane modules (8 in. diameter, 20 m2 surface area) used are 10-100 times more permeable to organic vapors than to air or nitrogen. In 1989, the first vapor recovery unit (VRU) based on membrane technology was commissioned for off-gas treatment in a gasoline tank farm. At present, various membrane VRU's are in operation or under construction. The capacity of these units ranges from 100 to 2000 m3/h. These are single-membrane stages of hybrid systems of a membrane stage combined with a post-
treatment facility, e.g., a catalytic incinerator, gas engine, or pressure swing adsorption unit. These plants are equipped with a modified plate and frame configuration.36 A case of a vapor recovery unit based on membrane technology is that commissioned in a gasoline tank farm in Munich for the treatment of the off-gasses generated from the storage, handling, and distribution of gasoline. The plant capacity was 300 m3/h. The only external available energy source was the electrical power supply. This was planned in the framework of a pilot project for the reduction of emissions at the BP tank farm Hamburg-Finkenwerder. The VRU has a capacity of 1500 m3/h and a hybrid system of a membrane stage and a gas engine. Two gas engines coupled with a generator are permanently in operation to supply the basic electrical power of the side. The gas engines are designed to switch the fuel feed from natural gas to retentate of the membrane stage over a period of VRU operating time. A commercially successful application is a hybrid system of a membrane stage with pressure swing adsorption (PSA) (Figure 11). The liquid ring compressor operating with gasoline as the service liquid sucks the hydrocarbon (HC) contaminated air from the gasometer. The off-gases are compressed and fed into a scrubber. Gasoline from the tank farm is used as a lean absorbent. The HC concentration of the feed gas leaving the scrubber depends on the operating temperature and pressure. The layout of the membrane stage (membrane area and permeate pressure) is governed by the permissible HC intake concentration of the PSA unit. Two parallel PSA units are installed and operated alter-
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 11. Membrane stage with integrated pressure swing adsorption. Table 6. Practical Applications of Pervaporation application separation of water from organic/aqueous mixtures details separation and/or dehydration of water/organic azeotropes (water/ethanol, water/2-propanol, water/pyridine) dehydration of organic solvents shifting of the reaction equilibrium (e.g., esterification) removal of chlorinated hydrocarbons separation of organics from the fermentation broth separation of aroma compounds wine and beer dealcoholization removal of VOCs from air separation of azeotropes (e.g., ethanol/cyclohexane, methanol/MTBE, ethanol-ETBE) separation of isomers (e.g., xylenes)
removal of volatile compounds from aqueous and gas streams
separation of organic/organic mixtures
Table 7. Comparison of the Dehydration Costs of Ethanol from 99.4 to 99.9 vol % by Different Techniques vapor permeation ($/ton) 40.00 4.00 19.00 63.00 pervaporation ($/ton) 12.80 17.60 4.00 30.60 64.0 entrainer distillation ($/ton) 120.00 8.00 15.0 9.60 molecular sieve adsorption ($/ton) 80.00 5.20 10.00 50.00 145.20
utilities vapor electricity cooling water entrainer replacement of membranes and molecular sieves total costs
nately. One is in the adsorption phase while the other is in the desorption and regeneration phase. A bypass of the clean stream is used as a purge gas for regeneration. To maintain a low vacuum, the vacuum pump at the downstream side of the membrane stage can be a liquid ring pump with mineral oil as the service liquid or a rotary vane vacuum pump. This vacuum pump is also used to support the desorption of the PSA column. The adsorber material is activated carbon, a carbon molecular sieve, or an inorganic molecular sieve. A typical VRU combined with a PSA is installed at Shell in Ludwigshafen.36 Other interesting applications of the technology might be in the separation of light hydrocarbons from refinery waste gas streams, the recovery of natural gas liquids and hydrogen, or the separation of propane, butane, and higher hydrocarbons from methane in the processing of natural gas for dew point control.
Pervaporation. Some applications of pervaporation processes are listed in Table 6. Dehydratation of ethanol by PV was the first industrial-scale application proposed by GFT in the 1980s. Today, more than 40 industrial pervaporation plants built by Sulzer Chemtech Membrantechnik (former GFT) are in operation worldwide. They are used for the dehydration of different solvents and/or solvent mixtures. In many practical applications, it might be more economical to use pervaporation or vapor permeation only to break the azeotrope and to concentrate the retentate further by the above-azeotropic distillation (Table 7). Another successful example of PV is its application in the enhancement of chemical reaction efficiency. Examples of such reactions are esterification or phenolacetone condensation. The first industrial plant for the
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1287
Figure 12. Pervaporation-enhanced MTBE production.
Figure 13. Membrane system for CO2 recovery from fermentation broth.
pervaporation-enhanced ester synthesis was built in 1991 by GFT for BASF. A possible application of the removal of organic solutes could be the treatment of industrial and municipal water supplies contaminated with carcinogenic halogenated organic compounds. Such a process would also be attractive for the extraction of organics. The possibility of recovering volatile organic compounds from gases by pervaporation has been demonstrated and applied recently at the industrial level.37 The elimination of volatile solutes from dilute aqueous solutions might be possible by pervaporation. Separation of organic/organic mixtures represents the least-developed application and the largest potential commercial impact of pervaporation, but considerable developments in membrane materials and processes remains to be done. The first industrial application of PV to organic/organic separation was the separation of methanol from a methyl tert-butyl ether (MTBE) stream in the production of octane enhancer for fuel blends (Figure 12). Flexibility with respect to part-load performance and changing product and feed concentrations is one of the advantages of pervaporation over other separation processes. This is especially useful in the production of fine chemicals and in the pharmaceutical industry, where solvents are used and almost no single waste solvent is generated continuously. Pervaporation-based hybrid processes offer significant potential for new economical and efficient solutions to some classical separation problems.38
Membrane Contactors. In these systems the membrane function is to facilitate diffusive mass transfer between two contacting phases, which can be liquidliquid, gas-liquid, gas-gas, etc.39 The traditional stripping, scrubbing, absorption, and liquid/liquid extraction processes can be carried out in this new configuration. With respect to conventional systems, membrane contactors can guarantee some advantages such as nondispersion of the phases in contact, independently variable flow rates without flooding limitations, lack of phase-density difference limitations, lack of phase separation requirements, higher surface area/volume ratios, and direct scale-up due to a modular design. Traditional liquid-supported membranes in which a carrier is immobilized in the microporous hydrophobic structure of the polymeric membranes are the most traditional and well-developed example of a membrane contactor system. Other applications, however, have been studied and realized today or are under investigations. Interesting examples include the removal of trace of oxygen (at levels of <10 ppb) from water for ultrapure water preparation for the electronics industry,40 the removal of CO2 from fermentation broth (Figure 13), and the supply of CO2 as a gas to liquid phases (carbonation of soft drinks).41 The flow sheet of a water carbonation process is presented in Figure 14. Additional examples include the removal of alcohol from wine and beer, the concentration of juice via osmotic or membrane distillation,41 the nitrogenation of beer,42 the degassing of organic solutions, and water ozonation.43
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 14. Simplified flow sheet for the water carbonation process.
In particular, in the carbonation process, hollow fibers have generally been used for industrial units. During operation, an aqueous liquid flows over the shell side (outside) of the hollow fiber. A strip gas or vacuum, either separately or in combination, is applied on the lumen side of the hollow fiber and flows counter current. Because of its hydrophobic character, the membrane acts as an inert support to allow intimate contact between the gas and liquid phases without dispersion. The interface is immobilized at the pores by applying a higher pressure to the aqueous stream than the gas stream. The result is fast diffusive transfer of dissolved gases from or to the liquid phase. Since 1993, a bubble-free membrane-based carbonation line has processed about 112 gal/min of beverage by membrane contactors having a total interfacial area of 193 m2 (Pepsi bottling plant in West Virginia).40 Permea commercializes beer dispensing systems known as CELLARSTREAM Dispense Systems using PULSAR gas/liquid contactors, which increase or decrease the amount of carbon dioxide and nitrogen in draft beer for optimal presentation.42 Membrane distillation and osmotic distillation can be considered examples of membrane contactors for realizing the concentration of aqueous solutions with nonvolatile solutes as salts and sugars.44-47 In the membrane distillation process, two liquids or solutions at different temperatures are separated by a porous membrane acting as a barrier between the two phases, which must not wet the membrane (this implies that hydrophobic membranes must be used in the case of aqueous solutions). Because of the temperature gradient, a vapor pressure difference exists across the membrane, and it is the driving force inducing vapor molecule transport through the pores from the high-vapor-pressure side to the permeate side. In the case of osmotic distillation, the vapor molecule transport is due to a vapor pressure driving force provided by having a low-vapor-pressure solution on the permeate side of the membrane, e.g., a concentrated salt solution. The formation of emulsions or dispersions characterized by very uniform dimensions of droplets or microbubbles can be realized using the same technology. The membrane emulsification process is applied mainly in the preparation of food emulsions. Moreover, microbubble formation increases the stability of the system by minimizing coalescence phenomena. An interesting study evidenced the relationship existing between membrane pore diameter and droplet size.48 The formulation of various products might be realized
using this new concept, and important phenomena such as oil combustion might be optimized. Membrane Reactors The possibility of combining molecular separation and chemical transformations in a single unit soon attracted the interest of membrane engineers.49 The first studies on such reactors were devoted to the immobilization of biocatalysts on polymeric membranes. Recently, hightemperature reactions have been the objective of important studies. Both areas will be analyzed in the following pages. Membrane Bioreactors. Biocatalytic membrane reactors are interesting with respect to conventional membranes as they combine selective mass transport with chemical reactions. The selective removal of products from the reaction site increases conversion of product-inhibited or thermodynamically unfavorable reactions. Biocatalysts can be used suspended in solution and compartmentalized by a membrane in a reaction vessel or immobilized within the membrane matrix itself.50 Since the advent of what has been called solid-phase biochemistry, the advantages of immobilized biocatalytic preparations over homogeneous-phase enzymatic/cellular reactions have been exploited to develop new and less-expensive processes. Synthetic membranes provide an ideal support for biocatalyst immobilization because of a wide available surface area per unit volume and the possibility for the development of new immobilization procedures. Enzymes are retained in the reaction side, do not pollute the products, and can be continuously reused. Immobilization has also been shown to enhance enzyme stability. Moreover, provided that membranes of suitable molecular weight cutoff are used, chemical reaction and physical separation of biocatalysts (and/or substrates) from the products can take place in the same unit. Substrate partition at the membrane/fluid interface can be used to improve the selectivity of the catalytic reaction toward the derived products with minimal side reactions. Membranes are also attractive for retaining in the reaction volume the expensive cofactors that are often required to carry out some enzymatic reactions. At the 1997 Achema conference in Frankfurt, Germany, statements on the impact that innovative bioreactors, and particularly those based on the hollow-fiber design, have in setting new performance standards were
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1289
clearly presented. For example, hollow-fiber bioreactors in which cells attach into a capillary-type space have been designed to mimic biological processes more closely than any other reactor system. Through the fibers, nutrients such as glucose and oxygenase are fed to the cells, and wastes such as CO2 and H2O are removed. Roche Diagnostic declared the use of such reactors to produce monoclonal antibodies for diagnostic tests. Membrane bioreactor technology can also be applied to produce pure enantiomers, in that a membrane separation process can be combined with an enantiospecific reaction to obtain a so-called "enantiospecific membrane reactor". As for general membrane reactors, the result is a more compact system with higher conversion. This technology can respond to the strongly increasing demand for pharmaceuticals, food additives, feeds, flavors, fragrances, agrochemicals, etc., as optically pure isomers.51 Recently, the results achieved in the production of a chiral intermediate used for the preparation of an important calcium channel blocker, diltiazem, were discussed in the open literature,52 confirming the possibilities of membrane reactors also in the large-scale production of biotechnological products. Phase-transfer catalysis can also be realized in membrane reactor configurations, immobilizing the appropriate catalysts in the microporous structure of the hydrophobic membranes. Biphasic membrane reactors have been extensively studied with lipases entrapped or bonded on the membrane surface, which confirms the possibilities of the approach,53,54 as already discussed. Catalytic Membrane Reactors. The development of catalytic membrane reactors for high-temperature applications became realistic only in the last few years with the development of high-temperature-resistant membranes. In particular, the earlier applications involved mainly dehydrogenation reactions, where the role of the membrane was simply hydrogen removal. The earlier studies carried out, particularly in the Soviet Union on palladium and palladium alloys, confirmed the existence of membranes able to permeate H2 with high selectivity. Both capital and operative cost savings were anticipated, as units for hydrocarbon separation from the streams were avoided and the possibility of operating at lower temperatures because of reactor yield enhancement was realized. The fact that the membranes separate intermediates and products from the reacting zone, avoiding possible catalyst deactivation or secondary reactions, is also of practical interest. The kinetic mechanisms might be modified or controlled by the presence of appropriate membrane systems, which can also act only on a reactive interface with no permselectivity, optimizing phase-transfer catalysis.55 Palladium membrane costs and availability, their mechanical and thermal stability, and poisoning and carbon deposition problems are still obstacle to the large-scale industrial application of these dense metal membranes, also when prepared in a composite configuration.56 Hydrogen can be produced by steam reforming and shift conversion of natural gas or other hydrocarbons. In conventional steam reformers, high conversions of natural gas, on the order of 85-90% or even higher, are obtained at reformer outlet temperatures of around 850-900 C. The energy efficiency of steam reforming
processes is relatively high, but the investments are substantial. Pure hydrogen can be produced at significantly lower temperatures by integrating into the reactor a membrane that selectively removes hydrogen during conversion. Potential savings in membrane reformer and downstream processing costs compared to conventional steam reforming apparatuses must, in many cases, be weighed against additional costs associated with recompression of the hydrogen permeate stream. Ag membranes were initially suggested for their H2 permeability. Howevere, they present the same problems that characterize Pd membranes, also having a much lower permeability. Solid oxide membranes have recently been suggested for large-scale applications in syngas production.57 Studies carried out in the U.S. showed the possibility of preparing membranes with improved mechanical and thermal characteristics, able to operate, for example, at 900 C for over 21 days. Integrated Membrane Processes Traditionally, the various membrane operations (RO, UF, MF, etc.) have been introduced in industrial production lines as an alternative to other existing units. Reverse osmosis instead of distillation and ultrafiltration in place of centrifugation are typical examples. The possibility of redesigning overall industrial production by the integration of various already developed membrane operations is becoming of particular interest, because of the synergic effects that can be reached, the simplicity of the units, and the possibility of advanced levels of automatization and remote control that can be realized. The rationalization of industrial production by use of these technologies permits low environmental impacts, low energy consumption, and higher quality of final products. New products also often become available. These results are related to the introduction of new technologies from the very early stages of the same material transformations and not at the end of the pipe, as was often done in the past. The leather industry might be an interesting case study because of (1) the large environmental problems related to its operation, (2) the low technological content of its traditional operations, and (3) the tendency to concentrate a large number of small-medium industries in specific districts. More than 2000 companies are in operation in Italy, which is recognized as a world leather leader for the quality of the leather produced. The traditional flow sheet of the tanning process in its humid phase consists of about 20 steps operating in a discontinuous cascade system. The possibility of rationalizing the overall process by introducing advanced molecular separation systems such as ultrafiltration, cross-flow filtration, microfiltration, nanofiltration, and reverse osmosis was suggested and has recently also become the objective of an Italian National Research Program coordinated and carried out by a consortium representing most of the companies in the country. In Figure 15 is presented an ideal process based on integrated membrane operations.58,59 The innovative integrated scheme suggested in Figure 15 allows the pollution problems of the leather industry to be faced by solving or minimizing them one by one where they originate, thereby avoiding the need for huge wastewater treatments at the end of the overall produc-
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 15. Flow sheet of some humid phases of the tanning process integrated with membrane operations.
tion line. The fact that membrane operations act by physical mechanisms without modification of the chemical procedure at the origin of the final high quality of the leather should also be mentioned. The possibility of also introducing an enzyme membrane reactor as an alternative to the traditional chemical dehairing process and for the optimization of the degreasing step has also been considered. The recovery of the proteins produced in dehairing in the retentate of an ultrafiltration unit and the recovery and reuse in the same process of the excess of sodium sulfide used and separated in the permeate became realistic when UF tubular membranes able to operate at the high pH (>12) characterizing the dehairing bath were prepared commercially. It is interesting to consider that around 40% of the overall pollution in leather processing originates in the dehairing step, where only 10% of the overall liquid stream is generated. The problem of chromium used in the tanning step has always been crucial for its negative environmental impact. The exhaust chromium coming from the tanning bath can be recovered and concentrated by a two-stage process based on MF or UF as a pretreatment and nanofiltration as a concentration technique.13 The concentrated chromium solutions obtained by NF have an improved quality with respect to those obtained by the conventional recovery process of chemical precipitation because of the optimally low ratio of organic lipolytic component/ chromium characterizing the new product. If necessary,
the recovered chromium solution can be further concentrated by traditional techniques. These recovered solutions were used in sheepskin retanning and tanning operations; the skins showed improved physical and chemical characteristics compared to those obtained with the traditional chromium solutions. The permeate from the nanofiltration unit, considering its high content of chlorides, might be used in the pickling phase, realizing an interesting closed-loop process. In Figure 16, the schematic flow sheet of chromium recovery is shown. The possibility of a membrane-based posttreatment of secondary textile wastewater for the direct reuse of polished effluent within the dyeing process was also recently verified by tests at the pilot scale.14 A first treatment scheme examined requires two filtration steps: membrane microfiltration followed by nanofiltration. The preliminary filtration on ceramic MF modules reduces the fractions of pollutants (suspended solids and colloids) that can induce a rapid fouling in the nanofiltration membrane. An addition of high concentrations of aluminum polychloride (10-70 mg/ L) is necessary to obtain performance satisfactory of the treatment system. Permeate quality confirms the possibility of reusing the secondary effluent for textile industry purposes, but approximate preliminary calculations on this coupled membrane process indicate that, at present, this process cannot be transferred to a full-
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1291
Figure 16. Scheme of the proposed process for reuse of exhausted chromium solutions from tanning operations.
scale plant, because of the high price of the ceramic MF membranes and the need for high dosing of coagulants. A second treatment scheme studied requires a clarification-flocculation step followed by multimedia filtration prior to a low-pressure reverse osmosis operation. The clarification-flocculation/filtration is aimed at removing the colloidal fraction that promotes fouling of RO membranes. A polished effluent of high quality (COD < 0.10 mg/L; conductivity < 40 mS; negligible residual color) that can be reused in textile mills is obtained. Costs for the complete polishing by lowpressure RO are comparable to the costs of conventional secondary wastewater treatment and are quite affordable (on the order of 0.20-0.25 Euro) even for the Italian situation, where price of water is much lower than in most industrialized countries.14 Interesting cases of integrated membrane processes can also be found in the agrofood industry, in water desalination, in biotechnological production, etc. In the dairy industry, single-membrane operations such as UF in the treatment of wheys, cross-flow microfiltration in the stabilization of milk, and RO in the concentration of milk or in lactose concentration have been widely applied in the past year. As already mentioned, an overall quantity of about 250 000 m2 of UF membranes were already installed in 1999 and around 165 000 m2 of RO membranes. Successful application of integrated membrane operations in fruit juice concentration (in an osmotic distillation process) has been developed by the Australian company The Wingara Wine Group (Melbourne, NSW). A hybrid pilot plant in which UF/RO and osmotic
distillation are integrated has been realized. It consists of UF and RO pretreatment stages, an osmotic distillation unit, and a single-stage brine evaporator. This plant concentrates fresh juices up to 65-70Brix and has a capacity of 50 L/h. Being athermal, osmotic distillation allows for concentration of the juices without product deterioration or loss of flavors.60 Hogan et al.61 reported a total process cost of osmotic distillation concentration on the order of $1.00/L of concentrate. From 1 L of fresh juice, it is possible to achieve about 200 mL of 70Brix concentrate. The value of this concentrate is between $2.50 and $7.50/L. From these data, the economical advantages of the integrated membrane process seem evident. The coupling of RO and membrane distillation for obtaining high recovery factors has been also tested in fruit juice concentration.62 The potential for osmotic distillation flux enhancement in grape juice concentration by ultrafiltration pretreatment has recently been investigated.63 Today, the integration concept finds interesting success in the use of membrane operations for brackish water treatment. Large-scale applications after many years of trials with other membranes have recently been successfully realized. For instance, at the end of 1999, the world's largest integrated membrane system was put into operation by PWN Water Supply Company North Holland in The Netherlands for drinking water production from lake water. Ultrafiltration and reverse osmosis are the most essential process elements of this treatment plant, having a capacity of 18 000 000 m3/year.64
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 17. Water reuse using MF/UF pretreatment for RO.
In Bahrain, the Addur SWRO desalination plant is planned for rehabilitation utilizing ultrafiltration membranes instead of the traditional flocculation-clarification process65 in water pretreatment. The traditional seawater or brackish water desalination process can be reconsidered by optimizing the pretreatment by MF and/or UF and by adding NF before the RO step. The introduction of a membrane distillation stage operating on the RO retentate might permit recovery factors on the order of 87-88% (while the RO unit alone worked with around a 40% of recovery factor) at costs that might be acceptable in various situations. Typically, the process systems for wastewater treatment are designed with reverse osmosis operation as the final treatment step and require several process steps, a large land area, and high capital investment and operating costs. Microfiltration and ultrafiltration membranes simplify the conventional water reuse process by treating effluent directly from the secondary clarifier, with a simpler process that is easy to operate, requires less land area, and is less vulnerable to process disruption (Figure 17). There are also some advantages in combining a membrane bioreactor with a reverse osmosis step. This process solution increases the life of the RO membrane and overall facility productivity. A new commercial membrane bioreactor for wastewater treatment already used in this type of integrated membrane system is known as the ZenoGem (ZENON Inc.) process. This unit consists of a biological reactor integrated with immersed membranes that form an absolute barrier to solids and bacteria and retain them in the process tank. The benefit of using biocatalytic membrane reactors combined with other membrane processes, such as microfiltration, ultrafiltration, reverse osmosis, membrane extraction, etc., for the production and processing of bioreactive compounds is also apparent. This integration is particularly important for products obtained by fermentation processes, such as organic acids, antibiotics, etc., and in the processing of food and beverages, such as wine, fruit juices, milk, etc. Energy Requirements One of the significant and recognized benefits of membrane operations is their low direct energy consumption (in general electricity) because of the absence of phase transformations. An important possibility for reducing indirect energy consumption through the recycling and reuse of raw materials or secondary products and minimization of the formation of wastes also characterizes these techniques. For a correct evaluation of the total energy involved, an energetic and
exergetic analysis of the overall integrated production lines, if not of the complete system, is recommended.66 The total energy requirements can be estimated on the basis of the overall supply of electrical energy of pump engines and external equipments and the thermal energy supplied. The energy analysis must be elaborated in order to include all real involved variables, which generally are very numerous and variously aggregated. It is necessary to establish the exact size of the unit operations and of all of the flows of mass and energy of the process. The block diagram of the operating phase of interest, connected to the recovery operation (Figure 18), or the block diagram of all of the productive process, integrated with the new operation, can help to report, in a compact way, all of the pertinent information for the elaboration of the estimation. The layout of the "traditional" process and the layout of the alternative process, both completed with all information relative to the fluxes of matter and energy, can then be compared. Because membrane operations utilize primarily electrical energy, the benefit estimate can be done using the "substitution coefficient" introduced by Electricite de France;67 this coefficient compares the primary energy saved to the electrical energy consumed in cycles that utilize electricity-consuming operations in substitution for conventional thermal operations. The substitution coefficient is defined as the ratio between the primary energy (thermal) saved in the new process with respect to the conventional process and the amount of electrical energy consumed, relative to the conventional process: CS ) C1 - C2/E2 - E1, where CS is the substitution coefficient, C is the consumption of thermal primary energy (MJ or Mcal), E is the consumption of electrical energy (kWh), and 1 and 2 are the relative indexes of the conventional and innovating process, respectively. Taking into account that 1 kWh of electrical energy, available at the utilization site, requires to burn, in a power station, 10.5 MJ of primary energy from a combustible source (oil, gas, coal, etc.), the substitution (or process innovation) results are advantageous when CS is greater than 10.5 MJ/kWh (2.5 Mcal/kWh). Other than energy, recovered and recycled materials are also involved; therefore, it is necessary to evaluate their indirect energy content. For example, one should also consider the energy consumed for the production of a material and, therefore, intrinsically associated with it; the energy used for disposal of a material in dumping; the energy used for an inertization treatment of a material, if required; etc. Considering that a substance,
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1293
Figure 18. General scheme of a system on which an energy balance can be planned. Table 8. Total Substitution Coefficients Evaluated for Some Processes Integrated with Membrane Operations membrane process analyzed recovery and recycling of water in the textile industry recovery and recycling of a monomer in the chemical industry recovery and recycling of the sulfide in the tanning industry saving of thermal energy and fat substance recovery in the dairy industry thermal energy savings in tomato juice concentration CS (MJ/kWh) 35.3 30.8 36.0 21.7 137.9
to be produced, needs a certain amount of energy, recycling a substance means, in addition to an economic saving, also an indirect energy saving corresponding to the amount of primary energy that would be utilized for its production. The results summarized in Table 8 indicate clearly the energetic advantages of some suggested membrane operations.68 New Membranes More and more complicated and special separation problems of liquid and gaseous mixtures in industry and biomedical or medical technology require tailored finished products made from potential membrane materials available on the market. The range of application fields involves widely spread uses in micro-, ultra-, and nanofiltration, dialysis, membrane electrolysis, or reverse osmosis, as well as in fields such as hightemperature gas separation, hydrogen recovery from syngases, and also oxygen-enriched air. Accordingly, the number of investigated and established membrane materials has also simultaneously grown. Some examples are listed in Table 9. Among the variety of new thermoplastics developed up to now are some special polymers that have been shown to be particularly suitable for membrane production. For example, the aromatic poly(etheretherketone) called Vitrex (PEEK produced by ICI) shows a remarkable long-term temperature stability of 250 C, where the modulus of elasticity remains sufficient even at 150 C. Because of its mechanical and chemical stability, Vitrex represents a suitable material for the production of hollow fibers. Another polymer tested for this purpose is called Tedur (polyphenyle sulfide, PPS, manufactured
Table 9. Some New Investigated Membranes and Membrane Materials thermostable polymeric membranes (PEEK, PPS, PEEK-WC, PEEK-WC functionalized, etc.) polymeric membranes resistant to hostile environments (HYFLON AD, etc.) H2 permselective dense membranes (Pd-based, dense SiO2, etc.) O2 permselective dense membranes (Brownmillerite, solid oxides, etc.) porous infiltrated composite membranes for MRs (dense silica/Al2O3 composite membranes, VMgO/ZrO2/RAl2O3, perovskites/RAl2O3, VmgO/RAl2O3, VPO/RAl2O3, etc.) inorganic nanofiltration membranes (ZrO2 on ceramic support) hollow-fiber ceramic membranes
by Bayer). It is characterized by a maximum long-term stability of 190 C. Hollow fibers produced from PEEK and PPS show textile-like properties, so that flexible modules with large membrane areas and small volume filling can be realized for the purposes of gas separation. Regarding separation performance, permeability, and thermal and mechanical stability, PEEK membranes are better than the PPS ones.69 Modified PEEK (PEEK-WC), an amorphous polymer exhibiting mechanical and electrical properties equal to or better than those shown by traditional PEEK, is also soluble in DMA, DMF, chlorohydrocarbons, etc., which makes possible its use for asymmetric membrane formation with the phase inversion procedure.70 Interesting results have been obtained with PEEK-WC dense membranes showing high O2/N2 selectivity and good permselectivity to water in the pervaporation of water/ methanol mixtures, as well as in acetic acid aqueous solutions.
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Table 10. Afforded Methods for the Synthesis of Porous Infiltrated Composite Membranes method chemical vapor deposition/infiltration (CVD/CVI) direct impregnation of a porous support with salt solutions sol-gel/infiltration solvothermal (hydrothermal/infiltration) applications synthesis of dense silica/Al2O3 composite membranes highly selective to H2 (supplied by MPT)82 synthesis of VMgO/ZrO2/RAl2O383 (contactor for ODHP) synthesis of perovskites/RAl2O384 (VOC combustion) synthesis of meso- and microporous inert85 or catalytically active86 membranes synthesis of mixed oxides composite membranes (VMgO/RAl2O3, VPO/RAl2O377) synthesis of supported zeolite membranes (silicalite-1, ZSM-5, A-type, mordenite, zeolite Y, ferrierite, AlPO4-5, zeolite L, SAPO, etc.) with insertion of catalytically active sites87
PEEK-WC modified by the introduction of NO2 groups or by sulfunation are also studied for gas separation applications.71,72 The use of PEEK-WC membranes functionalized with o-octyloxycarbonyl -cyclodextrin derivatives to carry out the hydrolysis reaction of p-nitrophenyl acetate to p-nitrophenol in phosphate buffer enhances the reaction rate with an enzyme-like behavior, improving productivity and stability and decreasing costs.73 Much research on the synthesis of more selective, permeable, and stable membrane materials for gas separation has been done and is still ongoing all over the world. For this purpose, some interesting results have been presented in the recent literature. Novel silicone-coated hollow-fiber membrane modules for the removal of toluene and methanol from N2 have been tested.74 This novel membrane offers lower permeation resistance than other silicone-based membranes because the selective barrier is ultrathin (1 m) and the porosity of the polypropylene substrate is high. The bond between the substrate and the coating layer has been obtained by plasma polymerization. High separation factors have been obtained (toluene/nitrogen )10-55; methanol/nitrogen )15-125; more than 96% of VOCs removed from a feed stream of 60 cm3/min when the permeate side was subjected to a high vacuum).74 New 6FDA-DAF polyimide membranes have been obtained by simultaneous suppression of intersegmental packing and inhibition of intrasegmental motion with a significant increase in both permeability and selectivity.75 These membranes have been tested with mixtures of He-CH4 and CO2-CH4. Permselectivities of helium and carbon dioxide over methane are improved with respect to the other polymers: the permeability of He/ CH4 is 2.6 times higher in these membranes than in polysulfone. Roman76 recently presented new fiber spinning and processing technology and streamlined/automated bundle-forming processes to reduce manufacturing costs and enable greatly increased production volumes. An important innovation has been the development of a proprietary co-extruded sheath/core fiber construction, effectively a thin asymmetric layer coated on a rugged porous support. The mechanical support function is uncoupled from the permeation function, so both functions can be optimized and a large fraction of the fiber wall, the core in this case, can be made of an inexpensive polymer to save on material costs. An additional advantage of the sheath/core construction is that it helps form a thinner skin in the sheath layer by allowing the use of sheath spinning solution with low polymer content and low viscosity (lower than could be used for a self-supporting monolithic membrane). The O2 and N2 flow rates in Air Liquide's N2 membrane have been increased 2-fold since 1990,77 via reduction of the
thickness of the separating layer and, to a lesser extent, changes in the material composition of the separating layer and optimization of the fiber size. The good H2 permselectivity and permeability of the recently developed dense (Pd-based) and almost dense SiO2 membranes were successfully exploited for a number of H2-consuming or -generating reactions. For some applications, the thermochemical instability of Pd membranes and the hydrothermal instability of silica are the main problems to solve. Concerning O2-generating or -consuming reactions, the development of O2 permselective membranes with good fluxes in the range of 400-700 C is still needed. Promising Brownmillerite dense membranes were recently developed by Eltron Research Inc.77 Most of the membrane research for membrane reactors aims at the development of thin films on porous supports for obtaining high fluxes. Because of a strict synthesis protocol, large-scale production of such membranes with consistent quality induces high-cost membranes and limits the range of industrial applications. Porous infiltrated composite membranes (in which the membrane material is deposited in the pores of the support) are attractive candidates, with good thermochemical resistance (barrier effect) and easy reproducibility (see Table 10). Furthermore, in the case of catalytic membranes, a high quantity of catalyst can be deposited in such a membrane configuration, which provides for easy diffusion of the reactants to the catalyst.78,79 In particular, zeolite membranes,80 mainly used for gas and vapor separations, have scarcely been used as O2 distributors in membrane reactors81 or for their catalytic properties. The insertion (postsynthesis or in situ) of catalytically active sites (e.g., Pd, Pt, V, etc.) might extend the possibilities of these membranes for membrane reactor applications.82,83 Recent studies on membranes made with perfluorinated polymers show the possibility of their application in the field of separation processes performed in hostile environments, i.e., high temperatures or aggressive nonaqueous media, such as chemicals and solvents. Perfluoropolymers are polymers designed for high demanding applications in hostile environments. The presence of fluorine in the polymer backbone imparts to the structure an ability to withstand very high temperatures and a very high resistance to chemical attack. Copolymers of tetrafluoroethylene (TFE) and 2,2,4-trifluoro-5-trifluoromethoxy-1,3-dioxole (TTD), commercially known as HYFLON AD, are amorphous perfluoropolymers with glass transition temperatures (Tg) higher than room temperature. They show a thermal decomposition temperature exceeding 400 C. An important peculiarity of these polymer systems is that they are highly soluble in fluorinated solvents, with low solution viscosities. This aspect allows for the
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1295
Table 11. New Modules and Strategies for Concentration Polarization and Fouling Control module/strategy for fouling reduction countercurrent transverse flow hollow-fiber module with baffles spirally wound feed flow channels coiled modules (tubular/hollow fiber) gas sparging fluidized bed negative TMP pulsing (back-pulsing) dynamic filtration rotary disk modules vortex flow filtration vibratory shear-enhanced processor immersed membrane with aeration system effect mass-transfer coefficient increase Dean vortex formation Dean vortex formation secondary flow and local mixing generation near membrane surface turbulence, continuous mechanical erosion of particle deposits at wall of membrane periodic removal of particle cakes from membrane surface generation of high shear rates in fluid near membrane applicationsa RO RO, UF, MF UF, MF, PV, MC UF, MF, MBR MF UF, MF (with ceramic membranes) UF, MF
contaminants not forced into membrane pores under high pressure, aeration minimizes settling of solids and both agitates and scrubs the membrane surface
UF, MF, RO
a UF ) ultrafiltration; MF ) microfiltration; PV ) pervaporation; MC ) membrane contactor; MBR ) membrane bioreactor; TMP ) transmembrane pressure.
preparation of self-supported and composite membranes with desired membrane thicknesses.84,85 During the 1990s Techsep (Orelis) initiated an ambitious R&D project to design inorganic nanofiltration membranes. These membranes were first developed in collaboration with the nuclear industry (Commissariat a l'Energie Atomique and its subsidiary SFEC, France). ` The membrane is a pure inorganic ZrO2 layer obtained by sol-gel technology and deposited on a Kerasep TM ceramic support. New possibilities have been opened by nanofiltration ceramic membranes now available on the market for several years: they can be used in a very broad range of operating conditions (pHs from 0 to 14, severe oxidizing or reducing conditions, thermal resistance from 0 to 350 C, high-pressure resistance, inertness toward radiation, etc.), which means that new membrane applications can be examined. Recently, engineers at the TNO Institute of Applied Physics, Materials Research and Technology Division, in Eindhoven, The Netherlands, believe they have solved several of the common problems with ceramic membranes by developing and commercializing them in the shape of hollow fibers instead of tubes. These patented hollow-fiber ceramic membranes have a high surface-to-volume ratio (more than 1000 m2/m3) and are easy to scale-up. TNO also combines this technology with highly selective top layers. Compared to existing flat and tubular membranes, ceramic hollow-fiber membranes have the advantage that they are compact and up to 10 times less expensive to produce. These membranes are now being used in slurry reactors. There are also applications in wastewater treatment to selectively remove pollutants and in gas separation. One ceramic membrane nearing commercialization is being developed by Pall Corp., East Hills, NY, based on DOE classified technology through a Cooperative Research and Development Agreement (CRADA). Pall also recently commercialized a stainless steel membrane with this technology.86 Recently (October 1998), in the U.S., a research project has been partially funded by the DOE to develop a new fabrication process for ultrafiltration membranes based on thermally induced phase separation (TIPS) to produce membranes with more uniform microscopic pore sizes in an appropriate range (down to the 10-50 nm size range), enabling more efficient separation and
purification of biological materials for hemodialysis and virus filtration. In the meantime, Praxair, Inc. and the University of New Mexico are studying a new gas separation technology consisting of porous membranes containing special sites designed to temporarily bind to particular gas molecules, promoting their transport through the membrane (facilitated transport membranes). In particular, polymeric systems, compatible with current manufacturing methods, and mixed inorganic-polymer coatings, offering better pore-size control, are being studied. New Module Design and Strategies for Concentration Polarization and Fouling Control Although involving less novel scientific principles, module technology is also absolutely crucial to the successful implementation of membrane technology. Seals, assembly methods, flow distribution, and pressure-drop minimization require careful attention, and this trend will intensify as the field matures and competition intensifies. In the past few years, additional innovations in module design and new strategies and techniques have been explored, particularly in the reduction of concentration polarization and fouling problems in pressuredriven membrane operations. Some examples are listed in Table 11. A countercurrent transverse-flow hollow-fiber RO module with baffles has been designed for low feed flow velocities transverse to the hollow fibers to achieve high mass transfer coefficients in an overall countercurrent flow (on the module scale) to the permeate flow through the tube side.87 Spirally wound feed flow channels with membrane walls will allow the formation of Dean vortices, which will mix the bulk with the wall layers and reduce concentration polarization without moving the membrane or the module or having flow reversal in the feed stream via inserts or otherwise.88 Such a concept is useful for UF and MF as well (see Figure 19). Although Dean flows satisfy the exigencies of an efficient membrane process, such as high permeate flux by increased wall shearing, radial mixing, and low concentration polarization, the presently employed modules still have
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
Figure 19. Schematic representation of a spiral tubular membrane module.
the problem of low packing density and increased pressure drop, which renders the process operating costs relatively higher than conventional processes.89-91 In laminar conditions, the limiting flux obtained in ultraand microfiltration coiled modules is higher than that obtained in straight ones. The enhancement can reach a factor of up to 5 depending on module characteristics, hydrodynamic conditions, and suspension properties. The energy analysis shows that, for the same energy dissipation, the limiting flow reached in a helical module is still far greater than that reached in a straight module. There are also potential advantages of the Dean flow in some other membrane operations, e.g., membrane oxygenators or pervaporation. In the laminar flow regime, the mass transfer coefficients obtained in gas-liquid contactors (water oxygenation and VOC removal) are higher than those obtained using modules with straight fibers.92,93 It is known that, in ultrafiltration, concentration polarization affects the permeate flux (the productivity) and membrane rejection characteristics (the separation efficiency). Injection of gas bubbles to generate secondary flow and to promote local mixing near the membrane surface proves effective in overcoming concentration polarization. When applied to protein fractionation, gas sparging can also improve the selectivity significantly. One application of such a process might be in enzyme membrane bioreactors. For vertical membrane systems, the observed enhancement, in terms of percentage increase in permeate flux, depends on the severity of the concentration polarization and membrane surface shear. When concentration polarization is severe, for example, at low cross-flow velocities, at high transmembrane pressures, and at high feed concentrations, the observed flux enhancement is higher (up to 320%). With high shear systems, including hollow-fiber module and spiral-wound membranes, or even for highly turbulent flow, injecting gas at low flow can only achieve a marginal flux enhancement (20-36%).94,95 Two-phase flow in ultrafiltration hollow fibers is very efficient in enhancing mass transfer when it is limited by particle deposition. Air sparging seems to expand the particle cake, as it increases both cake porosity and thickness, thus allowing higher water fluxes. This effect can be explained by the mixing and turbulence created by the slug flow. In some cases, intermittence also affects the cake structure.96 The use of a fluidized bed during the microfiltration of suspensions on ceramic membranes results in a significant increase in permeate flux in comparison with results that can be obtained with an empty-tube system. This phenomenon is especially pronounced during the microfiltration of oil emulsions when the permeate flux in a fluidized-bed system is almost three times higher.
The fluidized solids ensure a significant reduction in concentration polarization as well as a continuous mechanical erosion of the particles deposited at the wall of the membrane. The improved permeate flux that is achieved is due to the combined action of turbulence and particle motion.97 Negative rapid transmembrane pressure (TMP) pulsing has been increasingly adopted to control fouling in conventional and ceramic membranes to restore membrane productivity and increase solvant flux.98 An other strategy explored to enhance filtration performances in UF and MF is so-called "dynamic filtration", which consists of creating high shear rates in the fluid near the membrane by relative motion between a fixed membrane and a moving wall or vice versa. The main advantages of this technique are that these high shear rates can be generated independently of the feed flow and that, because of the small head loss in the system, transmembrane pressure can be kept lower than in the classical cross-flow filtration. It is, therefore, advantageous for filtrating highly charged fluids or for some biotechnological applications99 when it is necessary to use low TMPs for solute recovery. The system is very efficient in terms of permeate flux, for example, with highly concentrated mineral suspensions. A rotary disk module was developed based on the patent of NIMIC and Hitachi Plant Engineering and Construction Co. for the low-power separation of highly concentrated liquids. A conventional membrane module circulates the treated liquid at a high flow rate so as to obtain a rapid flow at membrane surface and a high flux through the surface. On the other hand, rotating the membrane would allow a high flow rate at the membrane surface to be maintained without having to maintain high pressure. Therefore, relatively low power would be required to run the system.100 An alternative widely commercialized technique developed by Membrex Inc. is called the Vortex Flow Filtration (VFF) system wherein the feed is introduced into the annular gap between two cylinders, one of which is rotating. The membrane can be placed on either the inner or the outer cylinder. Taylor vortices are generated between the two curved surfaces, creating high shear at the membrane surface, but the feed pumping is low. The VFF technique is employed for smaller systems. For feed streams having high solids fractions, new UF techniques are used commercially. The VSEP system employs membrane leaf elements as parallel disks separated by gaskets in a disk stack, which is spun in a torsional oscillation like the agitator in a washing machine at a fast rate to produce shear rates as high as 15000 s-1 and, therefore, a much higher flux. The cost per square meter is high compared to that of commercial cross-flow systems. A second high solid
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1297
commercialized UF system called Discover achieves high fluid shear at a flat membrane plate surface by spinning a grooved disk between adjacent membrane plates; fluxes are 5-6 times larger than those in competing units with sludges containing oil and suspended solids. The PallSep VMF system101 is another vibrating membrane system commercialized by Pall Co. used as an efficacious and economical alternative to rotary vacuum filters, centrifuges, and cross-flow systems for the treatment of a wide variety of difficult-to-filter process streams in pharmaceutical and bioprocess applications. A characteristic of these UF techniques is the decoupling of the wall shear from the bulk liquid flow rate (and sometimes pressure). This is radically different from fluid management techniques used in conventional membrane devices for controlling polarization, fouling, gel layer formation, etc.102 Drinking water and industrial water on Hokkaido island (Japan) are produced with conventional water production technology from natural river water that contains organic components such as humic substances, but the conventional membrane separation system (UF) is not suitable for treating river water in the summer when turbidity and color become high. To solve the problem mentioned above, vibratory shear-enhanced processing is used. The system has a unique vibration mechanism that generates shear rates on the surface of the membrane so that it is resistant to fouling caused by the PAC [poly(aluminum chloride] used to facilitate coagulation of such natural organic matters for easy removal. The commercial production facility achieves a fairly high permeate flux compared with that of conventional membrane operation technology and has been able to produce the required water quality. The operation with vibration increases flux by about 1.5 times when a relatively high pressure is applied.103 An inorganic membrane module of the external pressure type was developed by Kubota Corporation. The main feature of the equipment is that some tubular membrane modules without casing are submerged in the anaerobic tank and row water is directly filtered through the membrane. The results obtained were succeeded by the Japanese national project Membrane Aqua-Century 21 by the Ministry of Health and Welfare, which was aimed at the development of membrane technology for the purification of drinking water.104 The equipment consists of a coagulation tank, an aeration tank equipped with modules inside, an air blower, a suction tank, and an air compressor for back-washing. The driving force of filtration is hydraulic pressure and the suction force of a pump to keep water flow constant. The system is now commercially available for water purification. Recently, ZENON Environmental Systems Inc.105 developed a new generation of membranes for water and wastewater treatment known as ZeeWeed. This new membrane is of the immersed hollow-fiber type and is able to operate in high solids environments (10 00015 000 ppm). Unlike conventional membranes that are housed in pressure vessels and require a positive pressure, the immersed membrane operates in an open tank environment under a slight vacuum (-2 to -8 psi). The lower operating pressure, which provides increased membrane life and reduced replacement costs, also permits the reduction of the energy requirements as-
sociated with water production and membrane fouling (the contaminants are not forced into the membrane pores under high pressure). An aeration system permits a consistent flux to be maintained because of the generation of a recirculation pattern in the process tank that minimizes the settling of solids and both agitates and scrubs the membrane fiber surface to prevent plugging and fouling. New Areas of Interest for Membrane Engineering The significant results already reached in the development of membrane operations as discussed in the previous pages suggest other areas in which the overall possibilities of membrane engineering might be of importance. The case of transportation technologies is of particular interest. Transportation technologies, for example, will go through an important revolution in the next few years, mainly because of concerns about the poor air quality in many of the cities of industrialized countries, increasing levels of greenhouses gases, and problems with the oil supply. Various interesting projects are in progress worldwide trying to accelerate the solutions to these problems. The Exploratory Technology Research Program in the U.S., which seeks to identify new batteries and fuelcell systems with higher performance and lower lifecycle costs than those available today, is an important example of these actions. Membrane systems represent a significant aspect of these efforts. For example, proton-exchange-membrane (PEM) fuel cells are, in principle, capable of high power density and of changing their power output more quickly than other fuel cell types, making them candidates for replacing internal combustion engines in transportation applications, particularly in the automotive industry. Methanol, ethanol, hydrogen, natural gas, and gasoline are being evaluated as fuels. The technical barriers that must be overcome include size, weight, and cost reductions; fuel storage, conditioning and delivery; durability; reliability; etc. At Argonne National Laboratory, a new 10-kW partial oxidation methanol reformer with 50% H2, less than 4% CO, thermal efficiency of 88-95%, excellent dynamic response, and rapid start-up (<100 s) has already been demonstrated. Direct methanol fuel cells (DMFC), which eliminate the need for an external reformer, reducing the system weight and cost, are under investigation at Los Alamos National Laboratories (LANL). Technical problems include methanol permeation through the membrane. By NMR spectroscopy, the diffusion coefficient of methanol in Nafion membranes has been measured. It is only a factor of 2-3 smaller than the diffusion coefficient in aqueous solutions, and this is an important factor contributing to the sizable methanol crossover rates observed. Solid polymer electrolyte membranes play a vital role in these fuel cell systems. Unfortunately, costs (U.S. $ 70-150/ft2) appear to be still too high.106 In the U.S., an interesting coordinated national program was launched in 1998 for the promotion of industrial research in this area. The goal is to develop a totally new fuel cell system with improved CO tolerance (raising permissible levels
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
by 50 ppm to 3000 ppm), by using a high-temperature, ion-conductive solid polymer membrane electrode assembly and new bipolar separator plates. Other important projects are in progress on fuel cell technologies that are going to depend mainly on the membrane concept design. Conclusions Not many years have passed from the days when Loeb and Sourirajan, with their preparation of asymmetric membranes, made the reverse osmosis process of industrial interest. The early membranologists have always been optimistic about the possibilities of membrane operations, but the scientific and technical results reached today are even superior to the expectations. The intrinsic multidisciplinary character of membrane science has been and is still today one of the major obstacles to the further exploitation of its possibilities. The new logic of membrane engineering based on a drastic rationalization of the existing process design and not on the more traditional approach of adding one more, eventually innovative, unit at the end of the existing pipe, which has been another obstacle to the growth of membrane units, will also contribute to the exploitation of these technologies. A variety of technical challenges must be overcome to permit the successful industrial application of new membrane solutions. For example, the development of catalytic membranes will depend on material advances and increases in module reliability under extremetemperature cycling. The development of affinity membranes will require research on electron-beam grafting and other approaches to the modification of membrane chemistry. The development of tunable membranes will require extensive research on materials (e.g., conducting polymers) and assembly processes (e.g., chemical vapor deposition). In general, advanced membrane and module materials need to be matched with appropriate, economical manufacturing processes. The limitations still existing today to the large-scale industrial applicability of some membrane operations can be attributed only in part to inadequate intrinsic membrane properties (low permeability and selectivity, low thermal and chemical resistance, etc.) but probably more to inadequate module design, hydrodynamic studies, and, in general, engineering analysis. As already evidenced in this work, significant progress has been made in the study and realization of new organic and inorganic membranes, and many academic and industrial research projects in this area are also in progress. Many efforts on new module configuration designs and on the individuation of more efficient strategies for concentration polarization and fouling control are showing growing possibilities. A continuous research effort on fundamental aspects of transport phenomena in the various membrane operations already existing and in the new ones under investigation is evident. However, these efforts need to be combined with new research works in the process dynamics of these processes and in the study of advanced control systems applied to integrated multimembrane operations. These multidisciplinary studies will offer interesting opportunities for the design, rationalization, and optimization of innovative productions.
In fact, as already evidenced, the most interesting developments for industrial membrane technologies are related to the possibility of integrating various of these membrane operations in the same industrial cycle, with overall important benefits in terms of product quality, plant compactness, environmental impact, and energetic aspects. It is known that existing non-membrane-based equilibrium-driven separation technologies (e.g., absorption, adsorption, distillation, extraction, ion exchange, stripping), which represent the core of the traditional chemical and petrochemical industry, have significant shortcomings: inherent operational difficulties, lack of flexibility and modularity, slower rates, need for hazardous chemicals, high capital costs, higher energy requirements, and need for large equipment volume. These shortcomings are exacerbated by new separation demands (for example, environmental pollution control laws). New membrane-based separation concepts and technologies (e.g., vapor permeation, osmotic distillation, facilitated transport, supported liquid membranes, membrane-based extractors, membrane-based absorption, and stripping in contactors) do not suffer from many such deficiencies and are poised to invade more and more the domain of traditional separation technologies. It is, then, realistic to affirm that new wide perspectives of membrane technologies and integrated membrane solutions for sustainable industrial growth are possible. It is also important to recall that the most recent legislation and standardization are finally starting to identify the role of membrane operations in various areas (production of pure water for pharmaceutical uses, production of wine, of drinkable water or of ultrapure water for electronic industry, etc.). In Japan and in the U.S., the introduction of official standards for characterizing the membranes used in various processes is already in progress. Acknowledgment We have to acknowledge various colleagues from the IRMERC-CNR for their data and information on the various processes of their specific interest and, in particular, Dr. L. Giorno, Dr. G. Clarizia, Ms. A. Gordano, and Mr. A. Cassano. Literature Cited
(1) Furukawa, D. H. Conference on Reverse Osmosis Process Status, 1999. Proceedings of ICOM `99, Toronto, Canada, June 1999. (2) Stankiewicz, A. I.; Moulijn, J. A. Process Intensification: Transforming Chemical Engineering. Chem. Eng. Prog. 2000, 96, 22. (3) De Bartolo, L.; Drioli, E. Membranes in artificial organs. Biomed. Health Res. 1998, 16, 167. (4) Baker, R. W.; Cussler, E. L.; Eykamp, W.; Koros, W. J.; Riley, R. L.; Strathmann, H. Membrane separation systems; Noyes Data Corp.: Park Ridge, NJ, 1991. (5) Strathmann, H. Membrane processes for a sustainable industrial growth. New frontiers for catalytic membrane reactors and other membrane systems. Ravello, Italy, May 1999. (6) Hofmann, D.; Fritz, L.; Ulbrich, J.; Schepers, C.; Bohning, M. Detailed-atomistic molecular modeling of small molecule diffusion and solution processes in polymeric membrane materials. Macromol. Theory Simul. 2000, 9, 293-327. (7) Laciak, D. V.; Robeson, L. M.; Smith, C. D. Group Contribution Modeling for Gas Transport in Polymeric Membranes; American Chemical Society: Washington, D.C., 1999; Chapter 12.
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001 1299
(8) Tocci, E.; Hofmann, D.; Paul, D.; Russo, N.; Drioli, E. Polymer 2001, 42, 521. (9) Hofmann, D.; Fritz, L.; Ulbrich, J.; Paul, D. Molecular modelling of amorphous membrane polymers. Polymer 1997, 38 (25), 6145. (10) Ettouney, H. M.; El-Dessouky, H. T.; Alatiqi, I. Understand thermal desalination. Chem. Eng. Prog. 1999, 95 (9), 43. (11) Satone, H. Comparison between MSF distillation and RO. Technol. Proc. 9th Annu. Conf. NWSIA 1981, Vol. I, Session II. (12) Viscardi, C. F.; Piva, R. European Patent EP 0575360, 1991. (13) Cassano, A.; Drioli, E.; Molinari, R.; Bertolutti, C. Quality improvement of recycled chromium in the tanning operation by membrane processes. Desalination 1996, 108, 193. (14) Rozzi, A.; Antonelli, M.; Arcari, M. Membrane treatment of secondary textile effluents for direct reuse. Water Sci. Technol. 1999, Vol. 40, No. 4-5, 409. (15) Cuperus, F. P.; Nijhuis, H. H. Membrane Technology Applied in the Food Industry. Trends Food Sci. Technol. 1993, 4, 277. (16) Macroric, C.; Freoman, S. Design and operation of membrane filtration plants for water treatment. World Filtration Congress, Brighton, U.K., April 2000. In Proceedings, Vol. 1, pp 525-528. (17) Applying membrane technology to drinking water and wastewater treatment. Membr. Technol. 2000, 122, 4. (18) Koros, W. J.; Chern, R. T. In Handbook of Separation Process Technology; Rousseau, R., Ed.; John Wiley & Sons: New York, 1987. (19) Puri, P. Membrane gas separations: An opportunity for gas industry or just a niche market. Preprints of the International Conference on Membrane Science and Technology (ICMST `98), June 9-13, 1998, Beijing, China. (20) Stern, S. A. Polymers for gas separationssThe next decade. J. Membr. Sci. 1994, 94, 1. (21) Toi, K.; More, G.; Paul, D. R. Gas sorption and transport in poly(phenylene oxide)/polystyrene blends. J. Appl. Polym. Sci. 1982, 27, 2997. (22) Drioli, E.; Zhang, S. M.; Basile, A.; Golemme, G.; Gaeta, S. N.; Zhang, H.-C. Gas Permeability of Polyphosphazene Membranes. Gas Sep. Purif. 1991, 5, 252. (23) Tai, L. W.; Nasrallah, M. M.; Anderson, H. U.; Sparlin, D. M.; Sehlin, S. R. Structure and electrical properties of LA1XSRXCO1-YFEYO3.1. The system LA0.8SR0.2CO1-YFEYO3. Solid State Ionics 1995, 76, 259. (24) Teraoka, Y.; Nobunaga, T.; Yamazoe, N. Chem. Lett. 1988, 503. (25) Qiu, L.; Lee, T. H.; Liu, L.-M.; Yang, Y. L.; Jacobson, A. J. Oxygen permeation studies of SRCO0.8FE0.2O3-DELTA. Solid State Ionics 1995, 76, 321. (26) Ma, B.; Balachandran, U.; Park, J.-H.; Segre, C. U. Determination of chemical diffusion coefficient of SrFeCo0.5Ox by the conductivity relaxation method. Solid State Ionics 1996, 83, 65. (27) Brinkman, H. W.; Kruidof, H.; Burggraaf, A. Mixed conducting yttrium barium cobalt oxide for high oxygen permeation. Solid State Ionics 1994, 68, 173. (28) Stiegel, G. J. Mixed conductiong ceramic membranes: a new paradigm for gas separation and reaction. Proceedings of the Annual Membranes Technologies/Separation Planning Conference, December 1998, Newton, MA. (29) Drioli, E.; Santella, F.; Molinari, R. Industrial membrane operations. Ninetieth International Symposium on "Large Chemical PlantssFrom 1995 to the Next Decennium", Antwerp, Belgium, October 1995. (30) Rautenbach, R.; Albrecht, R. Membrane Processes; John Wiley & Sons: New York, 1989. (31) Porter, M. C. Handbook of Industrial Membrane Technology; Noyes Data Corp.: Park Ridge, NJ, 1990. (32) Permea Inc. Private communication. (33) Bhide, B. D.; Stern, S. A. Membrane processes for the removal of acid gases from natural gas. J. Membr. Sci. 1993, 81, 209. (34) Winston, W. S.; Sirkar, K. K. Membrane Handbook; Van Nostrand Reinhold: New York, 1992. (35) Baker, R. W.; Jacobs, M. Improve monomer recovery from polyolefin resin degassing. Hydrocarbon Process. 1996, 75, 49. (36) Peinemann, K. V.; Ohlrogge, K. Separation of Organic Vapors from Air with Membranes. In Membrane Processes in Separation and Purification; Crespo, J. G., Boddeker, K. W., Eds.; Kluwer Academic Publishers: London, U.K., 1994; p 357. (37) Neel, J. Pervaporation; Lavoisier Tec. & Doc.: London, U.K., 1997. (38) Lipnizki, F.; Fiel, R. W.; Ten, P.-K. Pervaporation-based hybrid process: A review of process design, applications and economics. J. Membr. Sci. 1999, 153, 183. (39) Reed, B. W.; Semmens, M. J., Cussler, E. L. Membrane Contactors. In Membrane Separations Technology. Principles and Applications; Noble, R. D., Stern, S. A., Eds.; Elsevier: New York, 1995; Chapter 10. (40) Sengupta, A.; Peterson, P. A.; Miller, B. D.; Schneider, J.; Fulk, C. W., Jr. Large-scale application of membrane contactors for gas transfer from or to ultrapure water. Sep. Purif. Technol. 1998, 14, 189-200. (41) Sirkar, K. K. Membrane Separations: Newer Concepts and Applications for the Food Industry. In Bioseparation Processes in Foods; Singh, R. K., Rizvi, S. S. H., Eds.; Marcel Dekker: New York; Chapter 10. (42) Permea (Air Products). Private communication. (43) Wikol et al. ICCS 14th International Symposium on Contamination Control, 14th Annual Technology Meeting, Phoenix, AZ, April 26-May 1, 1998. (44) Lawson, K. W.; Lloyd, D. R. Membrane Distillation: A Review. J. Membr. Sci. 1996, 124 (1), 1. (45) Calabro, V.; Drioli, E.; Matera, F. Membrane distillation ` in the textile wastewater treatment. Desalination 1991, 83, 209. (46) Johnson, R. A.; Valks, R. H.; Lefevre, M. S. Osmotic distillationsA low-temperature concentration technique. Aust. J. Biotechnol. 1989, 3, 206. (47) Mengual, J. I.; Ortiz De Zarate, J. M.; Pena, L.; Velazquez, A. Osmotic distillation through porous hydrophobic membranes. J. Membr. Sci. 1993, 82 (1-2), 129. (48) Furuya, A.; Asano, Y.; Katoh, R.; Sotoyama, K.; Tomi, M. Preparation of food emulsions using a membrane emulsification system. ICOM `96, August 1996, Yokohama, Japan. (49) Michaels, A. S. In Separation for Biotechnology 2; Pyle, D. L., Ed.; Elsevier Applied Science: Cambridge, U.K., 1990; p 3. (50) Giorno, L.; Drioli, E. Biocatalytic membrane reactors: Applications and perspectives. Trends Biotechnol. 2000, 18, 339. (51) Giorno, L. Enantiospecific membrane processes. Korean Membr. J. 1999, 1 (1), 38. (52) Lopez, J. L.; Matson, S. L. A multiphase/extractive membrane reactor for production of diltiazem chiral intermediate. J. Membr. Sci. 1997, 125 (1), 189 (Alan Michaels Special Issue). (53) Giorno, L.; Molinari, R.; Natoli, M.; Drioli, E. Hydrolysis and regioselective transesterification catalised by immobilised lipases in membrane bioreactors. J. Membr. Sci. 1997, 125, 177. (54) Giorno, L.; Molinari, R.; Drioli, E.; Bianchi, D.; Cesti, P. Performance of biphasic organic/aqueous hollow fibre reactor using immobilised lipase. J. Chem. Technol. Biotechnol. 1995, 64, 345. (55) Falconer, J. L.; Noble, R. D.; Sperry, D. P. Catalitic Membrane Reactors. In Membrane Separations Technology. Principles and Applications; Noble, R. D., Stern, S. A., Eds.; Elsevier: Amsterdam, 1995; p 669. (56) Dixon, A. G. Innovations in Catalytic Inorganic Membrane Reactors. In Catalysis; Spivey, J. J., Ed.; Royal Society of Chemistry: London, U.K., 1999; Vol. 14, p 40. (57) Frost, L. J.; Foster, E. P. T.; Russek, S. L.; Rowley, D. R. Use of ceramic membranes for oxygen separation and syngas production: Syngas for fuel cells. International Business Communications (IBC) Floating Platform Situation Offshore (FPSO)/ Remote Gas Utilisation Conference, London, U.K., December 1997. (58) Cassano, A.; Drioli, E.; Molinari, R. Recovery and reuse of chemicals in unhairing, degreasing and chromium tanning processes by membranes. Desalination 1997, 113, 251. (59) Cassano, A.; Molinari, R.; Romano, M.; Drioli, E. Treatment of aqueous effluents of the leather industry by membrane processes. A review. J. Membr. Sci. 2000, 181, 111. (60) Barbe, A. M.; Bartley, J. P.; Jacobs, A. L.; Jonhson, R. A. Retention of volatile organic flavour/fragrance components in the concentration of liquid foods by osmotic distillation. J. Membr. Sci. 1998, 145, 67. (61) Hogan, P. A.; Canning, R. P.; Peterson, P. A.; Johnson, R. A.; Michaelis, A. S. A New Option: Osmotic Distillation. Chem. Eng. Prog. 1998, 94, 49. (62) Calabro, V.; Jiao, B. L.; Drioli, E. Theoretical and experi` mental study on membrane distillation in the concentration of orange juice. Ind. Eng. Chem. Res. 1994, 33, 1803.
Ind. Eng. Chem. Res., Vol. 40, No. 5, 2001
(87) Futselaar, H.; Reith, T.; Racz, I. G. The countercurrent transverse flow hollow fiber membrane module for the separation of liquid streams. Engineering of Membrane Processes, GramishPartenkirchent, Bavaria, Germany, May 1992. (88) Brewster, M. E.; Chung, K.-Y.; Belfort, G. Dean vortices with wall flux in a curved channel membrane system. A new approach to membrane module design. J. Membr. Sci. 1993, 81, 127. (89) Manno, P.; Moulin, P.; Rouch, J. C.; Clifton, M.; Aptel, P. Mass transfer improvement in helically wound hollow fibre ultrafiltration modules. Yeast suspensionsSep. Purif. Technol. 1998, 14, 175. (90) Moulin, P.; Rouch, J. C.; Serra, C.; Clifton, M. J.; Aptel, P. Mass-transfer improvement by secondary flowssDean vortices in coiled tubular membranes J. Membr. Sci. 1996, 114, 235. (91) Belfort, G. Coiled Membrane Filtration System. U.S. Patent 5,626,758, May 6, 1997. (92) Guigui, C.; Manno, P.; Moulin, P.; Clifton, M. J.; Rouch, J. C.; Aptel, P.; Laine, J. M. The use of Dean vortices in coiled hollow-fibre ultrafiltration membranes for water and wastewater treatment. Desalination 1998, 118, 73. (93) Schnabel, S.; Moulin, P.; Nguyen, Qt.; Roizard, D.; Aptel, P. Removal of volatile organic components (vocs) from water by pervaporationsSeparation improvement by Dean vortices. J. Membr. Sci. 1998, 142, 129. (94) Ghosh, R.; Cui, Z. F. Fractionation of BSA and lysozyme using ultrafiltrationsEffect of gas sparging. AIChE J. 1998, 44, 61. (95) Ghosh, R.; Cui, Z. F. Fractionation of BSA and lysozyme using ultrafiltration: Effect of pH and membrane pretreatment. J. Membr. Sci. 1998, 139, 17. (96) Laborie, S.; Cabassud, C.; Durand-Bourlier, L.; Laine, J. M. Flux enhancement by a continuous tangential gas flow in ultrafiltration hollow fibres for drinking water production: effects of slug flow on cake structure. Filtr. Sep. 1997, 34, 887. (97) Mikulasek, P.; Hrdy, J. Permeate flux enhancement using a fluidized bed in microfiltration with ceramic membranes. Chem. Biochem. Eng. 1999, 13 (3), 133. (98) Ramirez, J. A.; Davis, R. H. Application of cross-flow microfiltration with rapid backpulsing to wastewater treatment. J. Hazard. Mater. 1998, 63, 179. (99) Frenander, U.; Jonsson, A. S. Cell harvesting by crossflow microfiltration using shear-enhanced module. Biotechnol. Bioeng. 1996, 52, 397. (100) Ohkuma, N.; Shinoda, T.; Aoi, T., Okaniwa, Y.; Magara, Y. Performance of rotary disk modules in a collected human excreta treatment plant. Water Sci. Technol. 1994, Vol. 30, No. 4, 141. (101) Pall Corp. Private communication, 2000. (102) Sirkar, K. K. Membrane separation technologies: Current developments. Chem. Eng. Commun. 1997, Vol. 157, 145. (103) Takata, K.; Yamamoto, K.; Bianc, R.; Watanabe, Y. Removal of humic substances with vibratory shear enhanced processing membrane filtration. Desalination 1998, 117, 273. (104) Kubota Corporation. Private communication, 2000. (105) Mourato, D. Water reuse with the immersed membrane & the membrane bioreactor. Int. Desalination Water Reuse Q. 2000, 9/4, 27. (106) Savogado, O. Energing membranes for electrochemical systems: solid polymer electrolyte membranes for fuel cell systems. J. New Mater. Electrochem. Syst. 1988, 1, 47.
(63) Bailey, A. F. G.; Barbe, A. M.; Hogan, P. A.; Johnson, R. A.; Sheng, J. The effect of ultrafiltration on the subsequent concentration of grape juice by osmotic distillation. J. Membr. Sci. 2000, 164, 195. (64) Schippers, J. C. Kiwa NV Research and Consultancy, EDS Newsletters. February 2000. (65) Ammar Ali, H. Ministry of Electricity & Water, Bahrain. Personal communication, 1998. (66) Romano, M.; Drioli, E. Analisi energetica ed exergetica nei processi a membrana. ICP Riv. Ind. Chim. 2000, 3, 76. (67) Electricite de France. Le coefficient de substitution, le gain net de petrole. SEPAC: Paris, 1981. (68) Molinari, R.; Gagliardi, R.; Drioli, E. Methodology for estimating saving of primary energy with membrane operations in industrial processes. Desalination 1995, 100, 125. (69) Vaorbach, D.; Schulze, Th.; Taeger, E. Thermostable hollow membranes for separation processes. Chem. Fibers Int. 1999, 49, 133. (70) Lufrano, F.; Drioli, E.; Golemme, G.; Di Giorgio, L. Transport parameters of carbon dioxide in poly(etheretherketone) membranes. J. Membr. Sci. 1996, 113, 121. (71) Trotta, F.; Drioli, E.; Gordano, A. Nitro derivates of PEEKWC. J. Appl. Polym. Sci., submitted for publication. (72) Trotta, F.; Drioli, E.; Moraglio, G.; Baima Poma, E. Sulfonation of polyetheretherketone by chlorosulfuric acid. J. Appl. Polymer Sci. 1998, 70 (3), 477. (73) Gordano, A.; Trotta, F.; Drioli, E. -cyclodextrins immobilised in PEEK-WC membranes: Kinetic behaviour and catalysis, unpublished results. (74) Cha, J. S. Removal/recovery of VOCs using a rubbery polymeric membrane. Membr. J. 1996, 6 (3), 173. (75) Kim, T.-H.; Koros, W. J.; Husk, G. R. Advanced Gas Separation Membrane Materials: Rigid Aromatic Polyimides. Sep. Sci. Technol. 1988, 23, 1611. (76) Roman, I. C. How Do You Coax 99+% Nitrogen from Membranes? Membrane Technology/Planning Conference, Newton, MA, 1995. (77) Sammels, A. F.; Schwartz, M. 3rd International Conference on "Catalysis in Membrane Reactors", Copenhagen, Denmark, September 1998. (78) Julbe, A.; Farrusseng, D.; Guizard, C. 9th CIMTEC World Ceramic Congress, Firenze, Italy, June 1998. In Advances in Science and Technology; Vicenzini, P., Ed.; Techna Pub. Srl: Faenza, Italy, 1998. (79) Julbe, A.; Guizard, C.; Larbot, A.; Cot, L.; Giroir-Fendler, A. The sol-gel approach to prepare candidate microporous inorganic membranes for membrane reactors. J. Membr. Sci. 1993, 77 (2-3), 137. (80) Tavolaro, A.; Drioli, E. State of the art on zeolite membranes: Preparations and applications. Adv. Mater. 1999, 11 (12), 975. (81) Pantazidis, A.; Dalmon, J. A.; Mirodatos, C. Oxidative dehydrogenation of propane on catalytic membrane reactors. Catal. Today. 1995, 25, 403. (82) Wu, J. C. S.; Sabol, H.; Smith, G. W.; Flowers, D. L.; Liu, P. K. T. Characterization of hydrogen-permselective microporous ceramic membranes. J. Membr. Sci. 1994, 96, 275. (83) Burggraaf, A. J. Fundamentals of inorganic membrane science and technology. In Membrane Science and Technology Series, 4; Burggraaf, A. J., Cot, L., Eds.; Elsevier: Amsterdam, 1996; Chapter 8. (84) Colaianna, P.; Brinati, G.; Arcella, V. European Patent EP 97106156, 1997. (85) Drioli, E. et al. Gas permeability of polyphosphazene membranes. Gas Sep. Purif. 1991, 5, 252. (86) Caruana, C. M. Oxygen Separation Sparks New Ceramics Membranes. Chem. Eng. Prog. 1999, 95 (10), 11.
Received for review June 27, 2000 Revised manuscript received November 16, 2000 Accepted November 17, 2000 IE0006209