Membrane Handbook

Membranes are primarily used for separation, and membrane processes are generally separation processes. Over the last 30 years, such processes have been widely adopted by different industries. Large-scale commercial uses of membrane separations have displaced conventional separation processes. More are expected in the future. Membrane separation processes are often more capital and energy efficient when compared with conventional separation processes. Membrane devices and systems are almost always compact and modular. In addition, membrane processes can sometimes achieve totally novel results. Membrane processes treated in this handbook are, therefore, important from current use and future development points of view.

The concept of separating gases with polymeric membranes is more than 100 years old, but the widespread use of gas separation membranes has occurred only within the last 10 to 15 years. Commercialization depended on the development of membranes with sufficient productivity and separating ability to make them economically attractive in industrial applications.

Gas diffusion through nonporous membranes is a concentration gradient driven process, which is generally well described by Fick’s first law (Crank 1975) 3-1$$J = - D\nabla c,$$ where D is the diffusion coefficient and c refers to the local gas or penetrant concentration. For unidirectional diffusion through a flat membrane, Eq. (3-1) can be written for species i as (Crank 1975) 3-2$${J_i} = - {D_i}({c_1})\frac{{d{c_i}}}{{dx}},$$ where D _i (c _i ) indicates that the diffusion coefficient can be dependent on the local composition of penetrant i.

In Chapter 2 we stated that membranes can be considered to have four structural levels and that three additional organizational levels exist between the membrane itself and the final membrane process. The design at all of these levels must be aimed at the ultimate goal of providing a cost-effective, robust separation technique.

About forty years ago Weller and Steiner (1950) considered membrane processes for the separation of hydrogen from hydrogenation tail gas, enrichment of refinery gas, air separation, and helium recovery from natural gas. However, commercial membranes capable of performing these separations economically have only become available within the last 15 years.

The economic analysis of membrane processes for gas separation is intrinsically no different from that for other separations. However, Spillman (1989), in his recent overview of the economics of gas separation with membranes, notes that cost comparisons involving membranes can be particularly difficult for the following reasons: 1.Membranes do not generally perform under the same operating conditions or with the same product split as alternative processes.2.The costs for membrane-based processes have been changing due to improved membrane performance and increased competition among membrane suppliers.

Over the past five years membrane pervaporation has gained acceptance by the chemical industry as an effective process tool for separation and recovery of liquid mixtures. It is currently best identified with dehydration of liquid hydrocarbons to yield high-purity organics, most notably ethanol, isopropyl alcohol, and ethylene glycol. Due to its favorable economics, efficacy, and simplicity, it can be easily integrated into distillation and rectification processes and, depending on the specific process, even replace them. Presently, considerable data are available on industrial-scale processes utilizing pervaporation to evaluate its performance. Chapters 7 through 10 review the historical perspectives of pervaporation, its underlying fundamental principles, design considerations, current commercial applications and processes, and general economics with respect to alternative process technologies.

Pervaporation through a nonporous membrane can be described by the widely accepted solution-diffusion mechanism (Binning and James 1958a, 1958b; Binning et al. 1961). The underlying assumptions of pervaporative transport are: 1Sorption of the liquid mixture on the feed (upstream) side of the membrane2Diffusion through the membrane3Desorption on the permeate (downstream) side of the membrane in the vapor phase.

In its simplest form, pervaporation is a separation process in which a liquid mixture is contacted with one side of a membrane and one of the components permeates more quickly than the other(s) to the permeate side of the membrane. At the permeate side, the component is evaporated and collected downstream. Generally, a mechanical vacuum is applied to drive this process. This so-called vacuum-driven pervaporation is illustrated in Figure 9–1, with a vacuum pump shown to generate the necessary partial pressure driving force.

Over the past few years, the number and variety of industrial pervaporation plants have dramatically increased. At least 20 to 50 plants of a minimum of 5000 L/day product capacity are in operation, with many more in development and pilot phases. In Europe and Asia, the primary driving forces have been (1) reduced energy costs, (2) low overall system capital costs, and (3) superior separations possible, with no limitations imposed by thermodynamic azeotropes relative to azeotropic distillation. In North America, the driving forces have been somewhat different: (1) pollution-free, closed-loop operation, minimum wastewater, and no entrainers, and (2) small, compact units with low capital costs for retrofitting existing plants to increase existing bottlenecked capacity versus distillation and adsorption with molecular 132 sieves. Although the level of energy consumption is considerably less in pervaporation than other competing processes, it is a much less important factor in the United States than is pollution abatement in the selection of pervaporation or integrated pervaporation.

Although nineteenth century scientists attempting to understand the phenomenon of osmosis (e.g.Graham 1861) are generally credited with the discovery of semipermeable membranes, modern exploitation for industrial applications of dialysis must be credited to twentieth century workers. In the first part of this century, two applications of dialysis were widely described. One was the recovery of NaOH from cellulose steeping liquors. These contained both the desired alkali and contaminating hemicelluloses. The steeping liquors, containing 20% NaOH, were dialyzed against water and the NaOH—which permeated the parchment “membranes”—was recovered as a dilute (4 to 5%) solution. The contaminating hemicelluloses remained behind and were used as fuel or partially recycled into the cellulose feed streams. A second early application of dialysis was the recovery of acid from copper leaching solutions. Dilute sulfuric acid was recovered by dialysis because the acid diffused much more rapidly than did the metal salts.

To model effectively membrane transport phenomena that pertain to dialysis, two general types of membrane models can be considered: homogeneous and porous. Homogeneous membranes are thought of as structureless continua, which, in the case of dialysis, consist of polymer-liquid gels. Porous membranes are viewed as an impervious polymer phase, interpenetrated by liquid-filled pores.

When a dialysis membrane separates two aqueous phases, the problem of material choice is relatively simple. Clearly, the membrane pores must be filled with water during use, so wettability is a primary consideration. As a consequence of this requirement, the predominant materials for dialysis membranes are relatively hydrophilic polymers. At one end of the spectrum are cellulose and poly(ethylene-covinyl alcohol) (Eval), and on the other end poly(methylmethacrylate), the latter prewetted by the manufacturer. In between fall cellulose acetate with a degree of substitution (DS) of 2.5, poly(acrylonitrile-co-methallylsulfonic acid), and other acrylonitrile copolymers.

The traditional industrial applications of dialysis, such as recovery of sodium hydroxide in rayon processing or separation of nickel sulfate from sulfuric acid in electrolytic copper refining, have been reviewed in detail by Tuwiner (1962). The dialyzers used in these processes were described in the preceding chapter. This chapter is devoted to applications of dialysis that are currently of greater interest.

The cost of a process that incorporates a dialysis procedure depends not only on the dialytic device, but also on the ancillary equipment to complete the separation process. Depending on the type of separation considered, the fractional weights of these two cost components may vary over extreme ranges. Unfortunately, dialytic separation processes are not as amenable to cost prediction as are the more classical separations such as distillation. Although the acquisition and replacement costs of the dialytic separator can be predicted with some accuracy, this may represent only a small fraction of the total process costs.

The history of electrodialysis began with the work of Ostwald (1890). He investigated the properties of semipermeable membranes and was the first to discover that a membrane is impermeable for any electrolyte if it is impermeable either for its cation or its anion. To illustrate this, he postulated the existence of the so-called “membrane potential” at the boundary between the membrane and the solution as a consequence of the difference of concentration. Later, Donnan (1911) confirmed this postulate for the boundary of an ion-exchange membrane and its surrounding solution. Simultaneously, he developed a mathematical equation describing the concentration equilibrium, which resulted in the so-called “Donnan exclusion potential.”

In this chapter, the various aspects of energy requirements in electrodialysis are considered first, and then the nature of equilibrium distribution of ions between the solution and the ion exchange membrane is considered.

The properties and preparation procedures of ion-exchange membranes are closely related to those of ion-exchange resins. As with resins, many types of membranes are possible with different polymer matrixes and different functional groups to confer ion-exchange properties on the product. Although a number of inorganic ion-exchange materials are available, most of 230 them based on zeolites and bentonites (Helfferich 1962), these materials are rather unimportant in ion-exchange membranes and will not be discussed further.

The performance of electrodialysis as a unit operation is determined by several process and equipment design parameters, such as feed flow velocities, cell and spacer construction, stack design, etc. These parameters affect directly the costs of the process as well as indirectly by means of the limiting current density and the current utilization.

Electrodialysis was first developed for the desalination of saline solutions, in particular, brackish water. The production of potable water is still currently the most important industrial application of electrodialysis. But other applications, such as the treatment of industrial effluents, the production of boiler feed water, the demineralization of whey, the deacidification of fruit juices, etc., are gaining increasing importance with large-scale industrial installations (Ahlgren 1972; Korngold, Kock, and Strathmann 1978). Another application of electrodialysis, which is limited regionally to Japan and Kuwait, has gained considerable com-mercial importance: the production of table salt from seawater. Diffusion dialysis and the use of bipolar membranes have significantly expanded the application of electrodialysis in recent years (Strathmann 1985). Some of the more important industrial applications of electrodialysis and related processes are listed in Table 20–1.

Reverse osmosis (RO) technology has grown extensively in recent years; many new types of membranes are now available, leading to an increase in various applications. Improvements have been made in membrane materials making them more pH, temperature, and chlorine resistant than the traditional cellulose acetate membranes. The ability of membranes to separate simultaneously, or selectively, organic and inorganic solutes from aqueous systems without phase change offers substantial energy savings and flexibility in the design of separation processes. The industrial development of noncellulosic, thin-film composite (TFC) membranes has provided better flux performance and enhanced separations of organics under lower operating pressures than those obtained with cellulosic membranes.

Various reviews (Pusch 1986;Soltanieh and Gill 1981;Mazid 1984;Dickson 1988;Dresner and Johnson 1980;Jonsson 1980;Rautenbach and Albrecht 1989; and Sourirajan and Matsuura 1985) on reverse osmosis (RO) transport mechanisms and models have been reported in the literature. Several models have been developed to describe solute and solvent fluxes through RO membranes. The transport models can be divided into three types: nonporous or homogeneous membrane models (solution-diffusion, extended solution-diffusion, and solution-diffusion-imperfection models), pore-based models (preferential sorption—capillary flow, finely porous, and surface force—pore flow models), and irreversible thermodynamics phenomenological models (such as KedemKatchalsky and Spiegler-Kedem models). Most models for reverse osmosis membranes assume diffusion or pore flow through the membrane while charged membrane theories include electrostatic effects. A model that includes prediction of reverse osmosis rejection to multi-ionic salt solutions was reported by Brusilovsky and Hasson (1989). For nanofiltration membranes, which are often negatively charged, Donnan exclusion models and the extended Nernst-Planck model can be used to determine solute fluxes.

Membrane performance prediction in reverse osmosis (RO) processes is important both in designing membrane systems and module arrangements and in choosing appropriate operating conditions to obtain a desired separation with maximum water flux (Sirkar, Dang, and Rao 1982; Belfort 1984; Rautenbach and Albrecht 1989;Bhattacharyya et al. 1990;DOE 1990). Some of the limiting factors in RO operations are concentration polarization, particulate fouling, bacterial fouling, and organics adsorption (Gilron and Hasson 1987;Siler 1987;Potts, Ahlert, and Wang 1981;Ridgway, Rigby, and Argo 1985;Williams, Deshmukh, and Bhattacharyya 1990).

The development of new-generation membranes that can tolerate wide pH ranges, higher temperatures, and harsh chemical environments and that have improved water flux and solute rejection characteristics has resulted in many applications for the reverse osmosis (RO) process. In addition to the traditional seawater and brackish water desalination processes, RO membranes have found applications in wastewater treatment, production of ultrapure water, water softening, food processing, and many other applications.Table 24-1 shows some applications of RO along with selected references. Membrane processes for these applications have several advantages over many of the traditional separation techniques such as distillation, extraction, ion exchange, and adsorption. No energy-intensive phase changes or potentially expensive solvents or adsorbents are needed for membrane separations, and simultaneous separation and concentration of both inorganic and organic compounds is possible with the RO process. Also, the RO process is inherently simple to design and operate compared to many traditional separation processes. Reverse osmosis can also be combined with ultrafiltration, pervaporation, distillation, and other separation techniques to produce hybrid processes that result in highly efficient and selective separations. ApplicationSelectedReferencesSeawater and brackish water desalinationSpiegler and Laird 1980; Ko and Guy1988Wastewater treatmentGeneralSlater. Ahlert, and Uchrin 1983a; GhabrisAbdel-Jawad, and Aly 1989Industrial and municipal wastewaterBhattacharyya et al. 1987; Bhattacharyyaand Williams 1990; Nusbaum and Argo1984Electroplating wastewaterSato et al. 1977; Spatz 1979; Imasu 1985Pulp and paper wastewaterOlsen 1980; Paulson and Spatz 1983Chakravorty and Srivastava 1987Textile wastewaterBrandon et al. 1981; Porter and Goodman1984; Slater, Ahlert, and Uchrin 1987Nanofiltration applications in wastewatertreatmentTextilePulp and paperElectroplatingDye manufacturingFood processingHazardous wastewaterSimpson, Kerr, and Buckley 1987Bindoff et al. 1987Cadotte et al. 1988Perry and Linder 1989Ikeda et al. 1988; Cadotte et al. 1988Bhattacharyya, Adams, and Williams1989; Bhattacharyya and Williams 1990Surface water and groundwater treatmentReverse osmosisWilson and Duran 1982; Eisenberg andMiddlebrooks 1986; Baier et al. 1987NanofiltrationConlon 1985; Simpson, Kerr, and Buckley1987; Cadotte et al. 1988; Taylor et al.1989Table 24-1Selected Applications of Reverse Osmosis.

The key factor in determining the cost of water treatment by reverse osmosis (RO) is the inherent variation in capital costs and operating costs for RO systems. Capital costs for RO systems put out for bid can easily vary by up to a factor of 2 (Birkett 1988; Glueckstern and Arad 1984). Operating costs also vary over a wide range, even for the same type of application (Birkett 1988; Applegate 1984). Several reasons cause these cost variations, including (1) widely different feed stream compositions; (2) widely different capabilities in RO technology, including membranes, membrane modules, equipment, and systems design; (3) the competitive commercial environment of the RO industry and associated technological breakthroughs; (4) system size; and (5) desired product purity.

Ultrafiltration (UF) is primarily a sizeexclusion-based pressure-driven membrane separation process. UF membranes typically have pore sizes in the range from 10 to 1000 Å and are capable of retaining species in the molecular weight range of 300 to 500,000 dal-tons. Typical rejected species include sugars, biomolecules, polymers, and colloidal particles. Most UF membranes are described by their nominal molecular weight cutoff (MWCO), which is usually defined as the smallest molecular weight species for which the membrane has more than 90% rejection.

It is sometimes useful to think of ultrafiltration (UF) as simply flow through pores in which separation is a sieving process based on relative molecular size. However, starting from this simple concept, more detailed models have evolved.

Most ultrafiltration (UF) membranes are polymeric in nature, although recently inorganic membranes have also become available. The formation of an asymmetric membrane structure, i.e., an upper skin that is permselective and a more porous substructure for mechanical support, is an important element in the success of UF membranes. While many polymers, in particular, have been examined for use as membrane materials (Lloyd and Meluch 1985), only a few are widely used. Table 28-1 lists various polymeric and inorganic materials for UF membrane manufacture. Various polymer membranes have been reviewed by Pusch and Walch (1982), Kesting (1971), and Cabasso (1980a). An excellent review of inorganic membranes has been given by Hsieh (1988). An exhaustive list of membrane types made by various manufacturers has been compiled by Cheryan (1986). Table 28-2 lists product information from various UF membrane manufacturers. A list of manufacturers of UF membranes is given in Table 28-3.

Several membrane configurations are available both commercially (see Figure29-1) and at the laboratory level. Polymeric membranes can be cast or extruded, as either flat sheets or in cylindrical geometry. Flat-sheet membranes are used in spiral-wound or plate-and-frame (or “plate-frame”) modules while other modules are of the tubular or hollow-fiber type. Inorganic membranes are used commercially only in tubular or monolith form; flat disks are available for laboratory use.

Many ultrafiltration (UF) applications are already being practiced in industry and more are being studied.Porter (1977) listed 34 UF applications in 1977; these included applications in the food industry, pharmaceuticals and biotechnology, water purification, and waste treatment in the chemical and paper industries. Various applications are discussed in detail by Cheryan (1986) and by Aptel and Clifton (1986).

When pressure-driven flow through a membrane or other filter medium is used to separate micron-sized particles from fluids, the process is called microfiltration. Although the exact size range is a matter of debate, microfiltration is generally defined to be the filtering of a suspension containing colloidal or fine particles with linear dimensions in the approximate range of 0.02 to 10 µm. This size range encompasses a wide variety of natural and industrial particles, as shown in Figure 31-1. These particles are generally larger than the solutes that are separated by reverse osmosis and ultrafiltration. Consequently, the osmotic pressure for micro-filtration is negligible, and the transmembrane pressure drop, which drives the microfiltration process, is relatively small (1 to 50 psi, typically). Also, the membrane pore size and permeate flux are typically larger for microfiltration than for ultrafiltration and reverse osmosis.

In this chapter, theories for deadend microfiltration are presented. The first part of the chapter focuses on the sieving or surface filtration mechanism, which is dominant when the particles are physically too large to pass through the pores of the filter medium. The primary goal is to predict the flux decline due to the buildup of the rejected particles on the membrane surface. We start with Darcy’s law for the relationship between flux and pressure drop across a cake layer and membrane in series. This relationship is used to describe transient cake buildup and flux decline for batch operation of deadend microfilters. The analysis is then extended to continuous operation of rotary drum vacuum filters.

In this chapter, recent theories describing crossflow microfiltration behavior are presented. First, the use of macroscopic balances to describe the overall behavior of various microfiltration module configurations is briefly reviewed. A major portion of this chapter is then devoted to recent models that predict the steady-state and transient permeate flux for crossflow microfiltration. These models are summarized in a brief section that describes the predicted dependence of the permeate flux on the material properties of the suspensions and the operating conditions of the filter. The focus is on the use of microporous membranes, which accomplish the desired separation using the sieving mechanism of surface filtration. The ssumption is made that the membrane completely rejects the particles reaching its surface. The chapter concludes with a review of cross-flow filtration experiments and their comparison with theory.

The previous three chapters covered a general description of microfiltration and the theories of deadend and crossflow microfiltration. This chapter focuses on the practical aspects of microfiltration with special emphasis on deadend microfiltration using commercial membranes, important applications, design criteria, and cost estimates.

Compared to deadend filtration, in crossflow filtration (also called tangential flow or inertial filtration) pressure drives only part of the feed through the medium; the remaining feed flows tangentially along the surface of the medium, continuously sweeping particles from the medium’s surface back into the feed. Generally, crossflow filters are operated as surface filters and have pores that are smaller than the particles to be removed.

The emulsion liquid membrane process is unique and different from the membrane processes discussed above. The membrane is a liquid phase involving an emulsion configuration. Emulsion liquid membranes, also called surfactant liquid membranes or liquid surfactant membranes, are essentially double emulsions, i.e., water/oil/water (W/O/W) systems or oil/ water/oil (O/W/O) systems. For the W/O/W systems, the oil phase separating the two aqueous phases is the liquid membrane. For the O/W/O systems, the liquid membrane is the water phase that is between the two oil phases.

This chapter describes the theory behind batch extraction with emulsion liquid membranes (ELMs). The extension of the theory to continuous ELM processes is given in Chapter 38 on design considerations. The theory of batch extraction may be classified into two categories: (1) diffusion-type mass transfer models for Type 1 facilitation and (2) carrier facilitated transport models for Type 2 facilitation (Lorbach and Marr 1987; Ho 1990). The definitions of Type 1 and Type 2 facilitations (Matulevicius and Li 1975; Li 1978 Li 1981) are given in Chapter 36.

As discussed in Chapter 36, an emulsion liquid membrane (ELM) process includes four steps: (1) emulsification, (2) dispersion and extraction, (3) settling, and (4) demulsification (breaking of the emulsion). These four steps are shown schematically in Figure 36–3. This chapter presents the design considerations for these steps in the ELM process.

In the viscose rayon industry, zinc ions are used in the spinbaths to improve the spinning process and the properties of the fibers. In the subsequent rinsing steps, zinc is lost with the wastewater. In addition to the economic disadvantage, the loss of zinc causes pollution problems. Therefore, most viscose companies in the industry have installed plants for zinc removal. The mere removal of zinc from the wastewater, e.g., by precipitation, merely represents a shift from a wastewater problem to a landfill problem. Therefore, some new processes have been tested and installed recently that enable the recovery of zinc. For the reuse of zinc in the viscose industry, a sufficient selectivity of zinc to calcium is needed to avoid any precipitation of gypsum in evaporation steps.

Contrary to most other membrane processes, emulsion liquid membrane (ELM) technology is not based on modules but is tailor-made for each specific problem. Therefore, it does not make sense to give prices, for example, in dollars per square meter of membrane area. Even figures in dollars per cubic meter of treated wastewater cannot be given in a general way because it depends on a number of parameters. To supply an estimate of the costs of an ELM plant, a complete calculation of a plant for the viscose industry according to Figure 39–2 is presented in this chapter (prepared in cooperation with VOEST-Alpine Industrieanlagenbau GmbH). Some effects of flow rate and metal concentration on capital and operating costs are given in the next section.

Solvent extraction is a common industrially used equilibrium-based separation process. In such a process, a solute (or solutes) in a solution, aqueous or organic, is extracted into an immiscible solvent, organic or aqueous, by dispersing one of the immiscible phases as drops in the other phase. This creates a large interfacial area and increases the extraction rate considerably. After the extraction is over, the phases are separated and the dispersed phase coalesced. There are two general categories of equipment for solvent extraction. A mixer-settler arrangement provides a single equilibrium stage; a number of them connected together provide multistage extraction. Continuous countercurrent contacting equipment whether in the form of columns or centrifugal devices can generate the equivalent of many stages in one device (Treybal 1963).

In gas separations by permeation through a nonporous polymeric membrane, gas molecules undergo dissolution in the membrane at the feed gas/membrane interface. The dissolved species diffuse through the membrane and are desorbed at the other membrane surface, the permeate gas/membrane interface. For permanent gases at temperatures greater than the critical temperature of the gases, the dissolution/desorption behavior of a gas has been found to obey Hen-ry’s law if the membrane is made of a rubbery material; the gas species dissolve in the membrane or desorb from it as if the membrane were a liquid. It is then obvious that a thin liquid layer ought to be able to function as a selective membrane provided it can withstand the pressure difference between the feed gas and the permeate gas and preserve itself. Not only a pure liquid but any solution could then be used as a permselective membrane

The term membrane reactor first began to appear in the chemical processing literature around 1980. Over the past decade, it has attained a proper niche in the lexicon of membrane technology and the topic is now a regular feature at membrane conferences and symposia as well as being the subject of a growing technical literature. Although there is no commonly accepted definition of a membrane reactor, the term is usually applied to membrane processes/ devices whose function is to perform net chemical conversion under conditions in which the unique contacting features of membrane devices are exploited. In particular, the term membrane reactor is reserved for those processes wherein the membrane functions as more than simply a reactive membrane i.e., a membrane matrix used for catalyst immobilization. These special features of membrane reactors have been demonstrated with multilayer devices (Matson 1979; Matson and Quinn 1986) and, more recently, with multiphase membrane contactors (Matson 1989a 1989b; Matson and Lopez 1989; Lopez et al. 1990). These important developments appear to be among the first ones in this emerging new area of reaction engineering and, therefore, a review such as the present one can serve as but a snapshot of the field in 1990, presenting underlying concepts and illustrating typical applications.

The commonly accepted mechanism for the transport of a penetrant in nonporous polymer membranes is solution-diffusion (Crank and Park 1968). The penetrant species dissolves in the membrane and diffuses across the membrane due to an imposed concentration gradient. Facilitated transport membranes also involve a reversible complexation reaction in addition to penetrant dissolution and diffusion. The addition of the complexation reaction makes facilitated transport analogous to a chemical absorption process on the feed (high partial pressure) side of the membrane and a stripping process on the product, or permeate, side of the membrane. Facilitated transport membranes, which are similar to emulsion liquid membranes and hollow-fiber contained liquid membranes described in previous chapters, have several general characteristics: 1They are highly selective.2A maximum flux or minimum permeability is reached at high concentration driving forces.3Very high permeabilities can be obtained at very low concentration driving forces.4They are often unstable in the conventional immobilized liquid membrane configuration.

Solvent extraction can be said to have started in the 1940s. With the growing needs of separation engineering (especially in nuclear fuel cycles) and the continuing synthesis of new extractants, solvent extraction technology developed rapidly and has played an increasingly important role in hydrometallurgy.

In conventional gas absorption, gas is dispersed in the absorbent liquid as bubbles in a vessel acting as a single stage or in a multistage column with the gas and the liquid in countercurrent flow. The dispersion of the gas increases the gas/liquid contact area and increases the mass transfer rate of the species to be absorbed in the liquid. After contacting is over, the gas bubbles coalesce as they leave the liquid phase except when foaming tendencies appear. Con-ventional gas absorption is also carried out with liquid dispersed as drops or thin films in the gas flowing countercurrently as in spray towers, packed towers, venturi scrubbers, etc. (Treybal 1980). Current gas stripping processes operate in an identical manner and in the devices used for gas absorption except that species are transferred from the liquid to the gas phase.

Controlled-release technologies are designed to deliver a wide variety of active ingredients (drugs, pesticides, fragrances, etc.) at a specified rate, for a specified period of time, and at a desired location. This might mean steady, small quantities of an active ingredient released over a long period of time, or short bursts of specified quantities released at designated intervals. Regardless of the rate or manner of release, the release kinetics from controlled-release formulations are characteristically dependent on the formulation rather than on the properties of the active ingredient (e.g., solubility or volatility) in the environment of use. In many cases, the release of the active ingredient is controlled by its rate of diffusion through a membrane. Controlled-release formulations offer (1) greater efficacy because optimal concentrations of active ingredient can be maintained in the environment of use, (2) improved safety from reliable release kinetics and control over the amount of active ingredient available at any one time, and (3) greater convenience because fewer applications or treatments are needed.