forward osmosis_principles, applications, and recent developments

Journal of Membrane Science 281 (2006) 7087

Review

Forward osmosis: Principles, applications, and recent developmentsTzahi Y. Cath a,, Amy E. Childress b, Menachem Elimelech c

a

Abstract

Osmosiresearchersyntheticin engineeseparationresearch idrug relealimitation 2006 E

Keywords:

Content

1. Int2. Os

2.12.2

3. Co3.13.2

3.34. Me

4.14.2

Abbreviatinational pototal Kjeld

CorrespE-mail

0376-7388doi:10.101. Membranes for forward osmosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

. Membrane modules and devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764.2.1. Plate-and-frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774.2.2. Spiral-wound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774.2.3. Tubular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774.2.4. Hydration bags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

ons: CP, concentration polarization; DOC, direct osmotic concentration; DS, draw solution; FO, forward osmosis; MD, membrane distillation; NPDES,llutant discharge elimination system; OD, osmotic distillation; PRO, pressure-retarded osmosis; RO, reverse osmosis; TDS, total dissolved solids; TKN,ahl nitrogen (sum of organic nitrogen and ammonia); TMDL, total maximum daily loadonding author. Tel.: +1 303 273 3402; fax: +1 303 273 3413.address: [email protected] (T.Y. Cath).

/$ see front matter 2006 Elsevier B.V. All rights reserved.6/j.memsci.2006.05.048s

roduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71motic processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71. Classification of osmotic processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71. Draw solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72ncentration polarization in osmotic processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73. External concentration polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73. Internal concentration polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

3.2.1. Modeling concentrative internal concentration polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743.2.2. Modeling dilutive internal concentration polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

. Influence of internal concentration polarization on water flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75mbranes and modules for forward osmosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75Colorado School of Mines, Division of Environmental Science and Engineering, Golden, CO 80401-1887, United Statesb University of Nevada, Department of Civil and Environmental Engineering, Reno, NV 89557-0152, United States

c Yale University, Department of Chemical Engineering, Environmental Engineering Program, New Haven, CT 06520-8286, United StatesReceived 30 December 2005; received in revised form 30 May 2006; accepted 31 May 2006

Available online 6 June 2006

s is a physical phenomenon that has been extensively studied by scientists in various disciplines of science and engineering. Earlys studied the mechanism of osmosis through natural materials, and from the 1960s, special attention has been given to osmosis throughmaterials. Following the progress in membrane science in the last few decades, especially for reverse osmosis applications, the interestsred applications of osmosis has been spurred. Osmosis, or as it is currently referred to as forward osmosis, has new applications inprocesses for wastewater treatment, food processing, and seawater/brackish water desalination. Other unique areas of forward osmosis

nclude pressure-retarded osmosis for generation of electricity from saline and fresh water and implantable osmotic pumps for controlledse. This paper provides the state-of-the-art of the physical principles and applications of forward osmosis as well as their strengths ands.lsevier B.V. All rights reserved.

Osmosis; Forward osmosis; Direct osmosis; Desalination; Reverse osmosis; Pressure-retarded osmosis

T.Y. Cath et al. / Journal of Membrane Science 281 (2006) 7087 71

5. Modern applications of forward osmosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785.1. Wastewater treatment and water purification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

5.1.1. Concentration of dilute industrial wastewater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78.

.

. .

s5.2. . .5.3. . .5.4. .5.5. . .

6. Conc .Ackn . .Refer . .

1. Introdu

Osmosiby human brealized thapreservatioother potendie or becoventionallyacross a se

ence in osmpermeablemolecules oof the applPresent-dayfrom waterand novel mfocus on othe field ofapplicationunderstand

In the fia more famdescriptionof osmosisthe osmotipurified waforce for mosmotic prNumerouswastewater

Fewerosmosisment/engtheless, Fbench-scfull-scale(at benching wastdemonstrand forRecent d

f-rer

eppse

o

c

et

re

nblFre

tatva

pae

o

sotrros

d.5.1.2. Concentration of landfill leachate . . . . . . . . . . . . . . . . . . . . . . . . . .5.1.3. Direct potable reuse for advanced life support systems . . . . . . .5.1.4. Concentration of digested sludge liquids . . . . . . . . . . . . . . . . . . .5.1.5. Forward osmosis for source water purificationhydration bagSeawater desalination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Food processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The pharmaceutical industryosmotic pumps . . . . . . . . . . . . . . . . . . . . . .Osmotic powerpressure-retarded osmosis . . . . . . . . . . . . . . . . . . . . . . . .

luding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .owledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ction

s is a physical phenomenon that has been exploitedeings since the early days of mankind. Early culturest salt could be used to desiccate foods for long-termn. In saline environments, most bacteria, fungi, andtially pathogenic organisms become dehydrated andme temporarily inactivated because of osmosis. Con-, osmosis is defined as the net movement of waterlectively permeable membrane driven by a differ-otic pressure across the membrane. A selectively

membrane allows passage of water, but rejects soluter ions. Osmotic pressure is the driving force for many

ications that will be reviewed and discussed herein.applications of the osmosis phenomenon extend

treatment and food processing to power generationethods for controlled drug release. This review will

smosis through polymeric membranes, primarily inwater treatment and desalination. References to others of osmosis will be made when necessary to furthertheoretical and applied aspects of osmosis.eld of water treatment, reverse osmosis is generallyiliar process than osmosis. Because of this, a briefof the RO process is given prior to further discussion

. RO uses hydraulic pressure to oppose, and exceed,c pressure of an aqueous feed solution to produceter [1]. In RO, the applied pressure is the drivingass transport through the membrane; in osmosis, theessure itself is the driving force for mass transport.publications on the use of RO for water treatment and

the use oPressurebeen testfor powedifferencwater toosmoticcan be u

The mlow or nrange ofing propBecauseto flowequipmeof a processing,withoutdetrimencan assisoral bioa[25].

Thisrelated mductioncompariof massfor the pFO procediscussereclamation appear in the literature.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

FO in controlled drug release in the body [2427].etarded osmosis (PRO), a closely related process, hasd and evaluated since the 1960s as a potential processgeneration [2839]. PRO uses the osmotic pressurebetween seawater, or concentrated brine, and freshressurize the saline stream, thereby converting theressure of seawater into a hydrostatic pressure thatd to produce electricity.ain advantages of using FO are that it operates athydraulic pressures, it has high rejection of a wide

ontaminants, and it may have a lower membrane foul-nsity than pressure-driven membrane processes [4].he only pressure involved in the FO process is duesistance in the membrane module (a few bars), thet used is very simple and membrane support is lessem. Furthermore, for food and pharmaceutical pro-O has the benefit of concentrating the feed streamquiring high pressures or temperatures that may bel to the feed solution. For medical applications, FO

in the slow and accurate release of drugs that have lowilability due to their limited solubility or permeability

per presents an extensive review of FO and closely-mbrane processes. The review begins with an intro-

f the basic principles of the FO process, includingn to other closely related processes. Special aspectsansport in the process as well as the membranes usedcess are described. Strengths and limitations of the

s in a broad spectrum of applications are reviewed andAnd last, the future of FO technology is considered.publications on the use of osmosis or forward(FO), or direct osmosis (DO) for water treat-ineering applications appear in the literature. Never-O has been used to treat industrial wastewaters (at

ale) [24], to concentrate landfill leachate (at pilot- and) [57], and to treat liquid foods in the food industry-scale) [816]. FO is also being evaluated for reclaim-ewater for potable reuse in life support systems (atation-scale) [1720], for desalinating seawater [21,22],purifying water in emergency relief situations [23].evelopments in materials science have also allowed

2. Osmotic processes

2.1. Classication of osmotic processes

Osmosis is the transport of water across a selectively perme-able membrane from a region of higher water chemical potentialto a region of lower water chemical potential. It is driven by adifference in solute concentrations across the membrane thatallows passage of water, but rejects most solute molecules orions. Osmotic pressure () is the pressure which, if applied tothe more concentrated solution, would prevent transport of water

72 T.Y. Cath et al. / Journal of Membrane Science 281 (2006) 7087

across the membrane. FO uses the osmotic pressure differential() across the membrane, rather than hydraulic pressure dif-ferential (as in RO), as the driving force for transport of waterthrough the membrane. The FO process results in concentrationof a feed stream and dilution of a highly concentrated stream(referred to as the draw solution).

PRO can be viewed as an intermediate process between FOand RO, where hydraulic pressure is applied in the oppositedirection of the osmotic pressure gradient (similar to RO). How-ever, the net water flux is still in the direction of the concentrateddraw solution (similar to FO).

The general equation describing water transport in FO, RO,and PRO is

Jw = A( P) (1)where Jw is the water flux, A the water permeability constantof the membrane, the reflection coefficient, and P is theapplied prePRO, >FO, PRO,driving forearly 1980RO zone, aFig. 2.

2.2. Draw

The conbrane is thDifferent teincluding dsolution, osity, the termWhen seleca higher ospressures osolutions wSystems Inas a functimodeling bpropertiestemperatur

irection and magnitude of water flux as a function of applied pressureRO, and RO. FO takes place when the hydraulic pressure difference ise PRO zone is where the applied pressure difference is between zero andreversal point, and the RO zone is where the applied pressure differencer than the osmotic pressure difference. Figure adapted from [30].

smotic pressure as a function of solution concentration at 25 C for var-ential draw solutions. Data were calculated using OLI Stream Analyzer.

Fig. 1. Solven uses to the more saline side of the membrane. For PRO, water diffusesto the more sa the less saline side due to hydraulic pressure (P > ).ssure. For FO, P is zero; for RO, P > ; and forP. The flux directions of the permeating water in

and RO are illustrated in Fig. 1. Flux directions andces for the three processes were characterized in thes by Lee et al. [30]. The FO point, PRO zone, andlong with the flux reversal point, are illustrated in

solutions

centrated solution on the permeate side of the mem-e source of the driving force in the FO process.rms are used in the literature to name this solutionraw solution, osmotic agent, osmotic media, drivingmotic engine, sample solution, or just brine. For clar-draw solution is being used exclusively in this paper.ting a draw solution, the main criterion is that it hasmotic pressure than the feed solution. The osmoticf several solutions being considered for use as drawere calculated using OLI Stream Analyzer 2.0 (OLIc., Morris Plains, NJ) [40] and are presented in Fig. 3on of molarity. This software uses thermodynamicased on published experimental data to predict the

of solutions over a wide range of concentrations andes.

Fig. 2. Din FO, Pzero. Ththe fluxis greate

Fig. 3. Oious pot2.0 [40]

t flows in FO, PRO, and RO. For FO, P is approximately zero and water diffline liquid that is under positive pressure ( > P). For RO, water diffuses to


Another important criterion in some applications of FO isthe selection of a suitable process for reconcentrating the drawsolution afan NaCl soatively simwithout rissolution thrcific applicion solutioin PRO, seaseawater [2have been utigations of

Variousas solutes ftion applicasolution asGlew [42]of water analiphatic althe first tojunction wiKravath antion of seawof glucose[45] usedtious drinka two-stagedependent sgested solu(SO2) as drapplicationwas demongases in sptions of theproduced Fof 250 atmwater fromin brine dilogical apptested as aferritin canmagnetic fi

3. Concen

The wadescribed bpressure diIn such proactive layeference, wh[30,39,48,5attributed tena. Specifiphenomenaosmotic-dr

3.1. External concentration polarization

ressca

. Reon r

re the toraneranembr

em

ay bressolutis bermeatrati

ve os

osm

easiby

fluxredua inre-dr

toind

red teenmem

than

tern

en aetely, theulk

ranesena

of aFO

orienraneshedpropive iive eand ttionf theyer fhe d. Than

iffereter it has been diluted in the FO process. Very oftenlution is used because it has high solubility and is rel-ple to reconcentrate to high concentration with ROk of scaling. Diffusion of the solute from the drawough the membrane must also be considered. In spe-ations where high rejection is desired, multivalentns may be preferable. In some applications such aswater may be used as the draw solution. In the past,8], Dead Sea water [33], and Salt Lake water [34]sed or considered for draw solutions in various inves-FO and PRO.

other chemicals have also been suggested and testedor draw solutions, particularly in seawater desalina-tions. Batchelder [41] suggested using sulfur dioxidethe draw solution in FO desalination of seawater.

expanded on this idea and suggested using mixturesd another gas (e.g., sulfur dioxide) or liquid (e.g.,

cohols) as the draw solutions for FO. Glew was alsopropose the recycling of the draw solution in con-th FO. Frank [43] used an aluminum sulfate solution,d Davis [21] used a glucose solution in FO desalina-ater, Kessler and Moody [44] used a mixed solutionand fructose for seawater desalination, and Stachea concentrated fructose solution to create a nutri-during FO of seawater. McGinnis [46] suggestedFO process that takes advantage of the temperatureolubilities of the solutes. Specifically, McGinnis sug-tions of potassium nitrate (KNO3) and sulfur dioxideaw solutions for seawater desalination. In a later novel

of FO by McGinnis and coworkers [22,47,48], itstrated that combining ammonia and carbon dioxideecific ratios created highly concentrated draw solu-rmally removable ammonium salts. This approachO draw solutions with osmotic pressures in excess

, allowing unprecedented high recoveries of potableconcentrated saline feeds and substantial reductionsscharges from desalination. In a new nanotechno-roach, naturally non-toxic magnetoferritin is beingpotential solute for draw solutions [49]. Magneto-be rapidly separated from aqueous streams using a

eld.

tration polarization in osmotic processes

ter flux in osmotic-driven membrane processes isy Eq. (1). In this equation, represents the osmoticfference across the active layer of the membrane.cesses, the osmotic pressure difference across the

r is much lower than the bulk osmotic pressure dif-ich results in much lower water flux than expected0,51]. The lower-than-expected water flux is ofteno several membrane-associated transport phenom-cally, two types of concentration polarization (CP) external CP and internal CP can take place in

iven membrane processes as discussed below.

In pate flowsurfacenomen

pressuCP dumembmembthe meof the mThis mCP in pdraw sbranethe peconcen

effectiCP onby incrface orwaterCP bynomen

pressuDue

foulingcompaIt has bdrivenlower-

3.2. In

WhcomplFig. 4aof the bmembphenomsistingused inbranemembestablisolutecentratcentratlayer,applicalayer oport lalayer, tdiluted

It csure dure-driven membrane processes, convective perme-uses a buildup of solute at the membrane active layerferred to as concentration polarization (CP), this phe-educes permeate water flux due to increased osmoticat must be overcome with hydraulic pressure [5254].water permeation is not limited to pressure-drivenprocesses and also occurs during osmotic-drivenprocesses, on both the feed and permeate sides of

ane. When the feed solution flows on the active layerbrane (like in RO), solutes build up at the active layer.e called concentrative external CP and is similar toure-driven membrane processes. Simultaneously, theion in contact with the permeate side of the mem-ing diluted at the permeatemembrane interface byting water. This is called dilutive external CP. Bothve and dilutive external CP phenomena reduce themotic driving force. The adverse effect of externalotic-driven membrane processes can be minimizedng flow velocity and turbulence at the membrane sur-manipulating the water flux [55]. However, becausein FO is already low, the ability to diminish externalcing flux is limited. For modeling external CP phe-FO, equations similar to those developed for CP ofiven membranes can be used [53,54].the low hydraulic pressure used in FO, membraneuced by external CP has milder effects on water fluxo the effects in pressure-driven membrane processes.shown that external CP plays a minor role in osmotic-

brane processes and is not the main cause for the-expected water flux in such processes [48].

al concentration polarization

n osmotic pressure gradient is established across arejecting dense symmetric membrane, as depicted indriving force is the difference in osmotic pressures

solutions in the absence of external CP. However, FOare asymmetric, adding more complexity to the CP. When a composite or asymmetric membrane con-dense separating layer and a porous support layer is, two phenomena can occur depending on the mem-tation. If the porous support layer of an asymmetricfaces the feed solution, as in PRO, a polarized layer isalong the inside of the dense active layer as water andagate the porous layer (Fig. 4b). Referred to as con-nternal CP [48], this phenomenon is similar to con-xternal CP, except that it takes place within the porousherefore, cannot be minimized by cross-flow. In FOs for desalination and water treatment, the activemembrane faces the feed solution and the porous sup-aces the draw solution. As water permeates the activeraw solution within the porous substructure becomesis is referred to as dilutive internal CP (Fig. 4c) [48].be clearly seen in Fig. 5 that the osmotic pres-nce between the bulk feed and bulk draw solution


Fig. 4. Illustr osmosymmetric de cing(c) An asymm le illurepresented b this

(bulk) isthe membrtive osmotto internalexchangersration (feedbrane but ithe membr[56].

3.2.1. Modpolarizatio

Unlike Ris mostly amembrane

lso srane.sitein Find dtionsces (cen

ptinet alter flations of driving force profiles, expressed as water chemical potential, w, fornse membrane. (b) An asymmetric membrane with the porous support layer faetric membrane with the dense active layer facing the feed solution; the profi

y w. External CP effects on the driving force are assumed to be negligible in

higher than the osmotic pressure difference acrossane (m) due to external CP and that the effec-ic pressure driving force (eff) is even lower dueCP. Furthermore, similar to the operation of heat, operation of FO in a counter-current flow configu-

and draw solution flowing tangential to the mem-n opposite directions) provides constant alongane module and makes the process more efficient

eling concentrative internal concentration

layer amembcompotratedfeed acentrainterfathe con

AdoLoebthe wanO, for which water transport through the membraneffected by the hydraulic resistance created by thestructure, in FO, internal CP in the porous support

orientation

Jw = 1K

ln

Fig. 5. (a) Concentrative internal CP and (b) dilutive internal CP across a csis through several membrane types and orientations [48]. (a) Athe feed solution; the profile illustrates concentrative internal CP.strates dilutive internal CP. The actual (effective) driving force isdiagram.

ubstantially affects mass transfer of water across theThe concentration profile across an asymmetric or

FO membrane for concentrative internal CP is illus-g. 5a. C1 and C5 are the concentrations of the bulkraw solution, respectively; C2 and C4 are the con-of the feed-membrane and draw solution-membraneresulting from external CP), respectively; and C3 istration at the active layersupport layer interface.g the models that were developed by Lee et al. [30],. [50] introduced a simplified equation to describeux during FO without consideration for membrane:

HiLow

(2)

omposite or asymmetric membrane in FO.


whereK is the resistance to solute diffusion within the membraneporous support layer, and Hi and Low are the osmotic pressuresof the bulktively, negl

K = tDs

where t, ,porosity, resolute. Howvalid only fthis equatiofor concent

K =(

1Jw

)

where B isof the memexperiment

B = (1 R

where R isseverity ofmore sever

3.2.2. Modpolarizatio

The conFO membrLoeb et al.dilutive int

K =(

1Jw

)

Recent studequation assuccessfullan FO-desi

3.3. Inuewater ux

Mehta asupport laytion concethe membrfluids on ththe permea(Fig. 6).

Experimfiber) Permout that A (not constanpressure (iThe declindehydratio

O tesal CPto th

declin

dryitic drt. Rl CP

mbr

emb

eralal ca,39,nfig

mem

raned fisaters196evelrane separation from a laboratory to an industrial pro-as the development of the LoebSourirajan process forg defect-free, high-flux, anisotropic RO membranes [61].nd co-workers [50,59] investigated the use of asymmetrictic polyamide membranes for FO and PRO.he 1970s, all studies involving osmosis (mostly PRO stud-ed RO membranes, either flat sheet or tubular, and ines, researchers observed much lower flux than expected.et al. [2] and Anderson [3] used several commerciallyle RO membranes and an in-house cellulose acetate mem-o treat dilute wastewater by FO using a simulated seawa-w solution. Kravath and Davis [21] used flat sheet ROranes from Eastman and cellulose acetate hollow fiberranes from Dow to desalinate seawater by FO using glu-s the draw solution. Goosens and Van-Haute [60] usedse acetate membranes reinforced with mineral fillers tote whether membrane performance under RO conditionspredicted through FO testing.draw solution (C5) and feed solution (C1), respec-ecting external polarization effects. K is defined as

(3)

and are the membrane thickness, tortuosity, andspectively, and Ds is the diffusion coefficient of theever, it has been recently demonstrated that Eq. (2) is

or very low water fluxes [57]. Further development ofn [50] has led to a more general governing equationrative internal CP:

ln(

B + AHi JwB + ALow

)(4)

the solute permeability coefficient of the active layerbrane, which can be determined from an RO-type[50] using)A(P )

R(5)

the salt rejection. Eq. (4) can be used to quantify theinternal CP; larger values of K are associated withe internal CP.

eling dilutive internal concentrationncentration profile across an asymmetric or compositeane for dilutive internal CP is illustrated in Fig. 5b.[50] presented a simplified governing equation for

ernal CP in asymmetric membranes:

ln(

B + AHiB + Jw + ALow

)(6)

ies [48,57] have shown that this simplified governingwell as the equation for concentrative internal CP,

y predict the results obtained during experiments withgned membrane.

nce of internal concentration polarization on

nd Loeb [39,58] studied the effect of the porouser on internal CP and the effect of high draw solu-ntrations on the overall permeability coefficient ofane. Results show that upon swapping the workinge two sides of the membrane, flux (as indicated bybility coefficient) sharply declines due to internal CP

enting with DuPont B-9 (flat sheet) and B-10 (hollowasep RO membranes, Mehta and Loeb [58] pointedthe membrane permeability constant from Eq. (1)) ist in FO and PRO; it declines with increasing osmotic.e., increasing concentration) of the draw solution.e of A was explained by partial drying or osmoticn of the membrane at high osmotic pressures. Such

Fig. 6. Fof internexposedsharply

partialosmo

transpointerna

4. Me

4.1. M

Genmateri[2,3,21lary coEarlymembtle, angoldbe

Bybeen dmembcess w

makinLoeb aaroma

In ties) usall casVottaavailabbrane tter dramembmembcose a

celluloevaluacan bets with NaCl and MgSO4 solutions as draw solutions. The effectcan be seen after 2 h when the support side of the membrane is

e draw solution instead of DI water. The water permeation ratees after dilutive internal CP starts. Figure adapted from [58].

ng can be accompanied by pore contraction, known aseswelling, and hence increased resistance to wateresults from recent studies [48,57] confirmed thatis actually the cause of the substantial flux decline.

anes and modules for forward osmosis

ranes for forward osmosis

ly, any dense, non-porous, selectively permeablen be used as a membrane for FO. Such membranes50,5860] have been tested (in flat sheet and capil-urations) in the past for various applications of FO.brane researchers experimented with every type ofmaterial available, including bladders of pigs, cat-

h; collodion (nitrocellulose); rubber; porcelain; and skin [3].

0, the elements of modern membrane science hadoped and the important discovery that transformed


Fig. 7. A crosis embedded wthickness is le

Also inwith DuPomasep ROFlux throug(approxima

Duringoped by OTechnologia wide va[5,6,8,16,1applicationand recreatthe membrbrane is thobe seen thaand it is evquite differtypically coa thick porthick suppovides mech(see SectioHTI is supmajor contmembrane

In summFO wouldrejection; aport layer fhydrophilicand high mused for PRmembranesMembraneminimal inport high hin current ations for FO

4.2. Membrane modules and devices

ferent module configurations can be used to hold orembranes for FO. Laboratory-scale modules have been

ed for use with either flat sheet or tubular/capillary mem-. Larger-scale applications [5,6,18] have been designedilt with flat sheet membranes in plate-and-frame configu-. Each configuration has advantages and limitations thate taken into consideration when planning the research orping the application. Prior to discussing the advantagesitations of plate-and-frame, spiral-wound, tubular, and

nfigurations, differences between continuous flow andperation must first be considered.

continuous flow FO applications, the draw solution isedly reconcentrated/refreshed and reused. In this mode,d solution is recirculated on the feed side of the membranee reconcentrated/refreshed draw solution is recirculatedpermeate side (Fig. 8). For this reason, modules that use

eet membranes are more complicated to build and oper-the FO process compared to pressure-driven processes.

ample, the spiral-wound module, one of the most com-acking configurations in the membrane industry, cannotd in its current design for FO because a liquid stream can-forcepplicolutwer

ted aat thpplicwere

of datchrecon

ice uations-sectional SEM image of HTIs FO membrane. A polyester meshithin the polymer material for mechanical support. The membraness than 50m. Figure taken from [22].

the 1970s, Mehta and Loeb [39,58] experimentednt B-9 (flat sheet) and B-10 (hollow fiber) Per-

membranes made of an aromatic polyamide polymer.h the B-10 Permasep membrane was relatively lowtely 8 l/m2 h at a driving force of 80 bar).the 1990s, a special membrane for FO was devel-smotek Inc. (Albany, Oregon) (currently Hydrationes Inc. (HTI)). This membrane has been tested inriety of applications by different research groups8,19,48]. It is also used successfully in commercials of water purification for military, emergency relief,ional purposes [62]. A cross-sectional SEM image ofane is shown in Fig. 7 [22]. This proprietary mem-ught to be made of cellulose triacetate (CTA). It cant the thickness of the membrane is less than 50mident that the structure of the CTA FO membrane isent from standard RO membranes. RO membranesnsist of a very thin active layer (less than 1m) andous support layer. The CTA FO membrane lacks art layer. Instead, the embedded polyester mesh pro-anical support. Results from various investigations

Difpack mdesignbranesand burationsmust bdeveloand limbag cobatch o

Inrepeatthe feeand thon theflat shate forFor exmon pbe usenot bePRO adraw sfor posupporand thscale atanks)supply

In bis notthe devApplicn 5) indicate that the CTA FO membrane made byerior to RO membranes operated in FO mode. Theributing factors are likely the relative thinness of theand the lack of a fabric support layer.ary, the desired characteristics of membranes for

be high density of the active layer for high solutethin membrane with minimum porosity of the sup-

or low internal CP, and therefore, higher water flux;ity for enhanced flux and reduced membrane fouling;

echanical strength to sustain hydraulic pressure whenO. The development of improved semi-permeablefor FO is critical for advancing the field of FO.

s that can achieve high flux and salt rejection, haveternal CP, and have high mechanical strength to sup-ydraulic pressures will lead to improved performancepplications as well as development of new applica-.

Fig. 8. Flowsof the membrof the membrturbine to prod to flow on the support side (inside the envelope). Inations, the pressure of the receiving stream (i.e., the

ion) is elevated to achieve the high pressure neededgeneration. This requires that the membrane is wellnd able to withstand pressure on the permeate side,e flow channels are not blocked. In advanced large-ations of PRO, additional accessories (e.g., pressuresuggested by Loeb [33,34] to handle the continuous

raw solution at elevated pressures.FO applications, the draw solution is diluted once andcentrated for further use. In this mode of operation,sed for FO is most often disposable and is not reused.s using this mode of operation include hydration

in FO and PRO processes. Feed water flows on the active sideane and draw solution flows counter-currently on the support sideane. In PRO, the draw solution is pressurized and is released in aduce electricity.


bags for water purification and osmotic pumps for drug delivery[26,63].

Even coable semi-pbranes. Fosheet memfiguration o

4.2.1. PlatThe sim

plate-and-fstructed indevices thascale systedesign anders is well emembranetations of pare lack ofsity. Lack ohydraulic psides of thtrol). Lowhigher capibrane replaconfiguratiing, difficurange of osures) [18]

4.2.2. SpirComme

(e.g., spirastream (thevelocity tanvery slowlyand its comerties of thein its currebe operatedforced to fl

Mehta [wound elemcan be usedflows throuthe same wfor RO. Hois blocked hto the otheof the meminside the efirst half ofthe envelopperforatedcan also beof the envein spiral-w

low patterns in a spiral-wound module modified for FO. The feed solu-s through the central tube into the inner side of the membrane envelope

draw solution flows in the space between the rolled envelopes. Figurefrom [36].

of the main limitations in operating a spiral-woundrane element for RO is the inability to induce flow in theate channels (i.e., the inner side of the envelopes) of theraneue a

embtopposi

f thec baembn. As to(Figuld

Tubuuse

uous

Effect of osmotic backwashing on the performance of a spiral woundbrane after desalination. The feed solution consisted of 0.5% NaCl andturated CaCO3. Figure adapted from [64].nsidering their limitations, the most readily avail-ermeable polymeric membranes are flat sheet mem-

r continuous flow operation of an FO process, flatbranes can be used in either a plate-and-frame con-r in a unique spiral-wound configuration.

e-and-frameplest device for packing flat sheet membranes is arame module. Plate-and-frame modules can be con-different sizes and shapes ranging from lab-scale

t hold single, small-size membrane coupons to full-ms that hold more than 1700 membranes. While theconstruction of large plate-and-frame heat exchang-stablished, the construction of large plate-and-framemodules is more complicated. Two of the main limi-late-and-frame elements for membrane applicationsadequate membrane support and low packing den-f adequate membrane support limits operation to lowressure and/or operation at similar pressures on bothe membrane (requiring relatively high process con-packing density leads to a larger system footprint,tal costs, and higher operating costs (labor for mem-cement). Other limitations of the plate-and-frame

on include problems with internal and external seal-lty in monitoring membrane integrity, and a limitedperating conditions (e.g., flow velocities and pres-.

al-woundrcially marketed spiral-wound membrane elementsl-wound RO elements) are operated with only onefeed stream) flowing under direct control of its flowgential to the membrane. The permeate stream flowsin the channel formed by the two glued membranes

position and flow velocity are controlled by the prop-membrane and the operating conditions. Therefore,

nt design, spiral-wound membrane elements cannotin FO mode because the draw solution cannot be

ow inside the envelope formed by the membranes.36] designed and successfully tested a unique spiral-

ent for FO. Both outside-in and inside-out operation. In Fig. 9 (inside-out operation), the draw solution

gh the spacers and between the rolled membranes, inay that a feed stream flows in a spiral-wound elementwever, unlike RO elements, the central collecting tubealfway through so that the feed solution cannot flow

r side. Instead, an additional glue line at the centerbrane envelope provides a path for the feed to flownvelope. In this configuration, the feed flows into thethe perforated central pipe, is then forced to flow intoe, and then flows out through the second half of the

central pipe. It is worth noting that this configurationefficiently used for PRO. The draw solution outside

lope can be pressurized similar to the way it is doneound membrane elements for RO.

Fig. 9. Ftion flowand theadapted

OnemembpermemembA uniqRO mUpon sby osmside oosmotiRO msolutiofor 20resultsflux co

4.2.3.The

contin

Fig. 10.RO memsuper-saelement for the purpose of cleaning or backwashing.pplication of FO for backwashing of spiral-wound

ranes was investigated by Sagiv and Semiat [64].ing the RO process after desalination, water diffused

s from the permeate channels back into the feedmembrane, thus cleaning the membrane through

ckwashing. In their investigation, a spiral-woundrane was purposely scaled with saturated CaCO3t predetermined intervals, the system was stoppedallow osmotic backwashing to occur. Preliminary. 10) indicate that almost complete recovery of waterbe achieved.

larof tubular membranes (tubes or hollow fibers) for

ly operated FO processes is more practical for three


main reasons. First, tubular membranes are self-supported. Thismeans they can support high hydraulic pressure without defor-mation anda holding vlar modulemodules amembraneit is the aumost suitablayer as ininternal CP

The maibranes is thhollow fibebe achievelimited. Uning can redmembranesachieved an

That saibranes (ROsemi-permmarket as fldue to theRO membrexcessive ffiber RO mspecially dhave to takhydraulic cbranes for

4.2.4. HydThe hyd

membrane.an FO memsucrose) anthe FO baghydration b

5. Modern

Forwardtions. Coming in the wthe pharmaing sectionwastewatertion, foodgeneration.

5.1. Waste

Severaltreatment han early stu[2,3], an in

study on direct potable reuse of wastewater in advanced life sup-port systems for space applications [5,18,19], and an investiga-

con

mest in mte prmate

Conof t

for w3].dustmetay thembrs conrnalrang

thanr eqin Rgateaw sulat

potene ofon oed a

Naon. Dgateolyv

ersin. Won, tcreas

to tcalededed f

Condfillts ahightings, did to

te iste isaste

cs, hfor Tse TDo co

efficthey can be easily packed in bundles directly insideessel. Second, it is much simpler to fabricate tubu-

s and packing density is relatively high. Third, thesellow liquids to flow freely on both sides of the

a flow pattern necessary for FO [65]. Furthermore,thors opinion that hollow fiber membranes may bele for FO because there is no need for a thick supportflat sheet RO membranes. This will result in reducedand enhanced performance.

n difference between hollow fiber and tubular mem-e flow regime that can be achieved on the bore side. Inrs (internal diameter


Table 1Raw leachate and final effluent water quality characteristics [6]Contaminant m lea

AluminumArsenicBariumCadmiumCalciumChromiumCopperIronLeadMagnesiumManganeseNickelPhosphorusPotassiumSeleniumSiliconSilverSodiumStrontiumTitaniumVanadiumZinc

Contaminant achat

AlkalinityBOD-5Chloride-wateCODFluoride-freeN-AmmoniaTKNpHTDSTSSSpecific cond

ND = not dete

evaporationhorizontalprocesses.mechanicavery effecticonstructedof landfillOregon [6,receives mo20,00040,to meet the(NPDES) tbe treated tcation.

An FO pCTA membWater recoinant rejeccessing of robserved dcomplete fl(metals or salts) Untreated leachate (g/l) Recovered water fro1320 ND39 ND305 14 0.03691600 150146 ND18 ND8670 2412 ND73700 331480 ND81 ND5740 ND56000 543ND ND35500 ND7 ND1620000 39902370 ND468 ND132 ND531 ND

Untreated leachate (mg/l) Recovered water from le

5000 4

472 2r 1580 23

3190 ND1 ND

1110 1.6780 ND

8 5.62380 48

100 NDuctance 9940S/cm 25S/cm

cted.

(e.g., vapor compression, vertical tube falling film,tube spray film, forced circulation) and membraneA comprehensive evaluation of vapor recompressionl evaporation, RO, and FO revealed that FO can beve in treating landfill leachate [6]. In 1998, Osmotek

a pilot-scale FO system to test the concentrationleachate at the Coffin Butte Landfill in Corvallis,7]. Because this landfill is located in an area thatre than 1400 mm per year of rainfall, approximately000 m3 of leachate is generated annually. In order

National Pollutant Discharge Elimination Systemotal maximum daily load (TMDL), the leachate musto a TDS level lower than 100 mg/l prior to land appli-

ilot system was tested for 3 months using Osmoteksrane (Section 4.1) and NaCl as the draw solution.

veries of 9496% were achieved with high contam-tion. Flux decline was not apparent during the pro-aw leachate; however, a flux decline of 3050% wasuring processing of concentrated leachate. Almostux restoration was achieved after cleaning.

The succonstructiosystem, theis extractedstaged) ROapplicationapproximafied before

Betweentreated overecovery of35S/cm [feed and efTable 1. Mfinal efflueNPDES TM

5.1.3. Diresystems

Long-teself-sufficiechate (g/l) Rejection (%) NPDES TMDL (g/l)100100

99.6799.10 1.899.84

100 16100 18

99.72100 5.399.96

100100100

99.0310010010099.75

100100100100 120

e (mg/l) Rejection (%) NPDES TMDL (mg/l)

99.9299.58 4598.54 378

10010099.86 2.8

10030.0097.98

10099.75

cess of the pilot-scale system led to the design andn of a full-scale system (Fig. 11). In the full-scaleraw leachate is collected and pretreated before waterin six stages of FO cells. A three-pass (permeate-system produces a stream of purified water for landand a reconcentrated stream of draw solution at

tely 75 g/l NaCl. The concentrated leachate is solidi-disposal [6].

June 1998 and March 1999, the treatment plantr 18,500 m3 of leachate, achieving an average water91.9% and an average RO permeate conductivity of

6]. The concentrations of major contaminants in thefluent of the full-scale FO system are summarized inost contaminants had greater than 99% rejection andnt concentrations were substantially lower than the

DL levels.

ct potable reuse for advanced life support

rm human missions in space require a continuous andnt supply of fresh water for consumption, hygiene,


Fig. 11. A flofrom [6].

and maintemissions owater treatmater generaThe threeand reusedter, urine, awastewaterhigh perceally, the synance, min[66].

A pilot-centration (reuse in spare being e

Table 2Water flux through one FO membrane and four RO membranes in FO mode [18]Membrane used Water flux (l/m2 h)CTAOsmotek 17.4OsmonicsCE 1.90OsmonicsCD 1.97HydranauticsLFC1 0.54HydranauticsLFC3 0.66

100 g/l NaCl solution was used as the draw solution and DI water as the feed.

Administration (NASA) for water reuse in space. The NASADOC test unit originally consisted of a permeate-staged ROcascade and two pretreatment subsystems. The first subsystem(DOC#1) utilized an FO process only and the second (DOC#2)utilized a unique combination of FO and osmotic distillation(OD). The OD process targeted the rejection of small com-pounds, like urea, that easily diffuse through semi-permeablemembranes [17,19]. A schematic drawing of the original DOCtest unit is shown in Fig. 12.

he fiuctioccom

ase (iverzatioain gce,r a mer p

rren

desigstemO wiwit

o of

Fig. 12. Flowdrinking watew diagram of the full-scale FO leachate treatment process. Adapted

nance. Long-range/long-duration missions, like lunarr a human Mars exploration mission, depend on a

ent system that recovers potable water from wastew-ted on board a spacecraft or in the planetary habitat.main sources of wastewater that can be reclaimedin long-term space missions are hygiene wastewa-nd humidity condensate. The system to treat theses must be reliable, durable, capable of recovering antage of the wastewater, and lightweight. Addition-stem should operate autonomously with low mainte-imal power consumption, and minimal consumables

In tconstrwere a

ond phthe UnoptimiThe mformandata foand othIn a cubeingthis sy(i.e., Fsystem

Tw

scale FO system, referred to as the direct osmotic con-DOC) system, was developed for direct potable waterace [5,17]. DOC is one of several technologies thatvaluated by the U.S. National Aeronautics and Space

relate to meprocesses.brane usedRO membr

diagram of the original NASA DOC test unit. Three waste streams are pretreatedr is produced by the RO subsystem [18].rst phase of the DOC study (19941999), design,n, and testing of the basic functions of the DOC unitplished by NASA and Osmotek [17,20]. In the sec-

20022004), the NASA DOC unit was transferred tosity of Nevada, Reno for thorough investigation andn of the operating conditions of the system [18,19].oals of this phase were to study the long-term per-

advantages, and limitations of DOC, and to validateass metrics comparison between the DOC system

otential advanced wastewater reclamation processes.t third phase (20042007), a new prototype system isned and fabricated for delivery to NASA. Althoughwill have similarities to the original DOC system

ll be used as pretreatment for RO), it will be a uniqueh a separate process for urine treatment [67].the important findings in Phase 2 of this researchmbrane performance and energy consumption in FOResults in Table 2 indicate that the FO CTA mem-in this study outperformed commercially availableanes. This is likely due to the lower internal CP inin DOC#1 and DOC#2. The draw solution is reconcentrated and


Fig. 13. Specific energy consumption as a function of draw solution flow rate,salt loading, a

the CTA mspecific memeasured dThe paramof the procand the amvaried. Thevariable opalmost alwproduced. Finvestigatio

5.1.4. ConSludges

often treatethe recalcitAfter digecentrifugestream (i.e.

of nutrients (e.g., ammonia, ortho-phosphate, organic nitrogen),heavy metals, TOC, TDS, color, and TSS. In some agriculturalcommunities the centrate is used as a soil fertilizer [68]; how-ever, the common practice for centrate treatment is to return itback to the headworks of the wastewater treatment facility forre-treatment. When these species are returned to the head of thetreatment facility, they increase the facility loading and operat-ing cost, and it is suspected that some of these constituents arerecalcitrant and end up in the effluent as nitrogen and phospho-rous species.

In a current study [4], FO is being investigated as a potentialprocess for concentration of centrate at the Truckee MeadowsWater Reclamation Facility (TMWRF) in Reno, Nevada. Themain objectives of this study are to determine the advantagesand limitations of using FO to concentrate centrate; to designa hybrid system consisting of FO as pretreatment for RO forconcentration of centrate; to develop an economic model anddetermine the cost effectiveness of the FO process; and to eval-uate the potential agricultural use of the concentrated stream as

izer.r to

esh (cen

l drtrati

ch tras w

12 hThuew, 71 thrd ann foa) wind

te thranetude

Fig. 14. Watedecline betwend RO pump capacity for the DOC system [18].

embrane stemming from the unique structure of thismbrane (see Section 4.1). Power consumption wasuring operation under variable operating conditions.

eters that were found to be important in the operationess (e.g., draw solution flow rate, RO pump capacity,ount of NaCl loaded in the draw solution loop) were

results summarized in Fig. 13 indicate that undererating conditions, specific power consumption is

ays less than 30 kWh for every 1 m3 of purified waterurther optimization of the process is currently undern.

centration of digested sludge liquidsproduced in wastewater treatment facilities are mostd in anaerobic digesters for further degradation of

rant organic solids and for stabilization of the sludge.stion, the sludge is commonly dewatered using a

a fertilPrio

150 mtreatedan NaCconcen

For eaNaCl;matelyNaCl.with n(trial #stoppesolutioFig. 14a curve

illustramembmagniwhich produces concentrated biosolids and a liquid, centrate). The centrate contains high concentrations

membranetrial, flux d

r flux as a function of draw solution (DS) concentration during FO concentration of (en trials is due to organic and suspended solids fouling. Flux was markedly restoredFO, the centrate is pretreated by filtering it through a


decline between each trial (for trials #1 through #4) is the resultof membrane fouling over time. After the short cleaning cycle(between trials #4 and #5), the flux was completely restored to itsinitial level. A similar performance was observed when raw (notpretreated) centrate was used as the feed solution (Fig. 14b).However, the extent of flux decline between trials was muchmore apparent and was attributed to greater fouling caused byhigh concentrations of suspended solids (Fig. 14). Some fluxrestoration after cleaning was achieved; however, complete fluxrecovery was not obtained. Phosphorous rejection exceeded 99%and ammonia and total Kjeldahl nitrogen (TKN) rejections wereapproximately 87% and 92%, respectively. Color and odor com-pounds were almost completely rejected. It is expected that acombination of FO and RO (for draw solution reconcentration)will result in even higher rejections of pollutants from the finaleffluent.

5.1.5. Forward osmosis for source waterpurication

The usein Sectionmilitary, reable drinkibags are onslower thanrequire nomuddy watthat in mosmeating waand most io

In the hor beveragsemi-permbag in an athe osmotisolid drawdraw solutnutrients anan ultimatFor small,34 h to co

Fig.

bags can be placed directly in the source water or they canbe suspended in another sealed plastic bag that holds thesource watuser. In recemergencyis debate atreatment psweet drink

5.2. Seawa

Severalsystems fohowever, mVery few pbe found. KAtlantic Ocand hollow

n. Ked anbjec

encyr seaecen

en uanddio

O. Tniumtionsonatesis ofrce (trati

withe am

highe frory.chemcess

andmodto amt wa

eral sdistilrodun avch-s

e FOreat

ed wce oux, m

fore pehydration bagsof FO for water purification was briefly introduced

4. The concept of hydration bags was developed forcreational, and emergency relief situations when reli-ng water is scarce or not available [63,69]. Hydratione of the few commercial applications of FO. Althoughother water purification devices, FO hydration bags

power and only foul minimally, even when used wither. The high selectivity of an FO membrane ensurest situations and for most sources of water, the per-ter is free of microorganisms, most macromolecules,ns.

ydration bags, an edible draw solution (e.g., a sugare powder) is packed in a sealed bag made of aeable FO membrane [63]. Upon immersion of thequeous solution, water diffuses into the bag due to

c pressure difference and slowly dilutes the initiallysolution. At the end of the process the diluted

ion can be consumed as a sweet drink containingd minerals. In this regard, hydration bags represent

e treatment process; not a pretreatment process.personal devices (Fig. 15), the process can takempletely hydrate a 12 oz beverage. The extraction

15. Illustration of water purification hydration bag [69].

solutiomodelTheir oemergcess fo

In rthat whbrane)carbonwith Fammo

proporbicarbAnalying foconcen

waterwith thratherto brinrecove

A sFO prowaterUponposesproducby sevbranepure psolutio

Bendioxidtions gachieving forhigh fldrivingthat therproviding better mobility and autonomy to theent years, the military procured hydration bags forand relief efforts around the world [23]. Yet, there

mong experts whether hydration bags provide waterer se because the product is not pure water but athat can only be used for specific applications.

ter desalination

patents have been awarded for different methods andr water desalination by FO [4143,4547,7073];ost of them have not matured or proven feasible.

eer-reviewed publications on FO desalination couldravath and Davis [21] investigated desalination of

ean seawater by FO using cellulose acetate flat sheetfiber membranes and glucose solution as a draw

essler and Moody [44] and Moody and Kessler [74]d tested similar applications of FO for desalination.tive was to develop a batch desalination process forwater supply on lifeboats, not as a continuous pro-water desalination.t bench-scale studies [22,47,48], it was demonstratedsing a suitable FO membrane (e.g., the FO CTA mem-a strong draw solution (highly soluble ammonia andxide gases), seawater can be efficiently desalinatedhe draw solution was formed by mixing togethercarbonate and ammonium hydroxide in specific[48]. The salt species formed include ammonium

, ammonium carbonate, and ammonium carbamate.the process has shown that an osmotic pressure driv-

) as high as 238 bar for a feed water with a salton of 0.05 M NaCl, and as high as 127 bar for a feeda salt concentration of 2 M NaCl, can be achievedmonia/carbon dioxide draw solution [48]. This is a

driving force considering that 2 M NaCl is equivalentm seawater desalination at approximately 70% water

atic drawing of the novel ammoniacarbon dioxideis illustrated in Fig. 16. Water is extracted from sea-

dilutes the ammoniacarbon dioxide draw solution.erate heating (near 60 C), the draw solution decom-

monia and carbon dioxide. Separation of the freshter from the diluted draw solution can be achievedeparation methods (e.g., column distillation or mem-lation (MD)). The degasified solution left behind isct water and the distillate is a reconcentrated drawailable for reuse in the FO desalination process.cale FO data demonstrates that the ammoniacarbonprocess is a viable desalination process. Salt rejec-

er than 95% and fluxes as high as 25 l/m2 h wereith the FO CTA membrane with a calculated driv-

f more than 200 bar [22]. Although this is a relativelyuch greater flux is actually expected for such a high

ce. Further analysis of the results [48] has indicatedrformance ratio (defined as experimental water flux


Fig. 16. Sche[22].

divided bywas at moslower-than

All of thdesalinatioFOlack oan easily seseawater deis desired, Fa high osm

5.3. Food

Althougfruits andFO treatmehas been shas severaages and liand low prtaste, aromrejection, apressure-dr

Severaldevices an[76,77]. Shin chronolo

Two detof FO in th[14] reviewMD forvided a mo(including

Popper efirst generajuices. Themembranesusing saturfruit juicesthe membr

Based on the development of a new plate-and-frame mem-brane module for FO [77] by Osmotek Inc., Beaudry and Lampi

lishtrati

solutwardcentd waffee,the hma

508olstapber

eva

n. Dncel diftratecos

s theptin

] furtd intrate-ma

ranesalciuhylenhickllingr mewat

ition, espocessodeleterison ppolyam thantagmatic drawing of the novel ammoniacarbon dioxide FO process

theoretical water flux) of the FO CTA membrane usedt 20%, and on average between 5% and 10%. This

-expected flux is attributed to internal CP.e previous, yet limited, work on FO as an alternativen process has exposed the two major limitations off high-performance membranes and the necessity forparable draw solution. Moreover, when consideringsalination, and especially when high water recoveryO can be utilized only if the draw solution can induce

otic pressure.

processing

h osmotic treatment of food products (e.g., preservedmeats) is very common in the food industry [75],nt for concentration of beverages and liquid foodstudied at laboratory-scale only [8,1114,16]. FO

l advantages as a process for concentrating bever-quid foods, including operation at low temperaturesessures that promote high retention of sensory (e.g.,a, color) and nutritional (e.g., vitamin) value, highnd potentially low membrane fouling compared toiven membrane processes.patents have been awarded for inventions of FOd methods for the concentration of liquid foods

[8] pubconcen

sugarwere a

to conshoweand codue tothe sumuse of

Wrred rasvacuum

solutiodifferesitionaconcen

capitaladdres

Ado[12,13be useconcen

custommembride, cpolyetbrane tcontroThinnehigherIn addtrationFO prand mcharacrates)matic

Froof advort summaries of the main studies are provided belowgical order.ailed summaries were published on the potential usee processing of liquid foods. Petrotos and Lazaridesed low pressure membrane processes FO, OD, andconcentration of liquid foods. Jiao et al. [11] pro-re comprehensive summary on membrane processesFO (DOC)) for concentration of fruit juices.t al. [15] were the first to use modern techniques andtion polymeric RO membranes to concentrate fruity used both tubular and flat sheet cellulose acetateand achieved an average water flux of 2.5 kg/m2 h

ated NaCl as the draw solution. Highly concentratedwere produced but salt was found to diffuse throughane into the grape concentrate.

processes flow operaticentrationsindustries,recovery prto transform

5.4. The ph

The oraadministratachieving drelease, targedy in theillnesses. Aed a short summary on the use of the DOC process foron of fruit juices. They recommended that 72Brixion be used as the draw solution. Herron et al. [77]ed a patent on a membrane module and a method

rate fruit juices and wines. Laboratory test resultster fluxes as high as 4 and 6 l/m2 h for orange juicerespectively. These relatively high fluxes are partiallyigh turbulence achieved in their FO module [77]. Inry of the invention, the inventors recommended the5 wt.% sugar solution as the draw solution.

d et al. [16] compared the sensory characteristics ofry juice concentrated by FO with that concentrated byporation. High fructose corn syrup was used as drawescriptive sensory analysis revealed no significant

between the products of the two processes. Compo-ferences between the two types of processed juice

were minor. No comparison of other aspects (e.g.,t or energy requirements) was provided to furtherbenefits or limitations of the FO process.g the methods of Herron et al. [77], Petrotos et al.her explored the optimal conditions by which FO canthe concentration of tomato juiceone of the mostd vegetable juices. Petrotos and co-workers usedde tubular thin-film composite aromatic polyamideand different draw solutions including sodium chlo-m chloride, calcium nitrate, glucose, sucrose, ande glycol 400 (PEG400). Results showed that mem-

ness and draw solution viscosity are the main factorswater flux in the FO concentration of tomato juice.mbranes and lower viscosity draw solutions yielded

er fluxes, and therefore, faster juice concentration., it was shown that pretreatment in the form of fil-ecially ultrafiltration, enhances the efficiency of the. In further studies, Dova et al. [9,10] investigatedd the impact of process parameters (e.g., membrane

tics, feed and draw solution concentrations, and flowrocess performance using thin-film composite aro-mide RO membranes.

ese studies, the FO process appears to have a numberes over evaporation and pressure-driven membraneor concentration of liquid foods. Low energy use,ng temperatures and pressures, and high product con-are the main advantages. However, similar to otherthe lack of optimized membranes and an effectiveocess for the draw solution are the main limitationsing FO into a full-scale process in the food industry.

armaceutical industryosmotic pumps

l route, which is still the most acceptable mode ofion of prescription drugs, may have limitations inesired outcomes. In special circumstances, extendedeted delivery, or especially accurate dosage of a rem-body is requiredparticularly in cases of chronicdditional benefits of the controlled release of drugs


include decreased dosing frequency, more regular drug concen-tration in the bloodstream, enhanced bioavailability, improvedpatient compliance, and reduced side effects [25]. For example,delivering therapies to specific cells can be an obstacle in treat-ing neurodegenerative diseases, such as Parkinsons disease andHuntingtons disease. Targeting treatment to the brain is diffi-cult because the bloodbrain barrier blocks many drugs [78]. Inother cases, a drug can have low oral bioavailability (e.g., it canhave low solubility or limited permeability) in the gastrointesti-nal tract [25].

Controlled- or modified-release of drugs is possible throughthe use of osmotic pumps. Osmosis offers several advantagesas a driving force for constant pumping of drugs, includingaccurate mass transfer. For example, when considering targetedtreatment, osmotic pumps can deliver medicine directly to thecerebrospin

The devcommerciastudies in thhave beenment of druOROS Purelease drupumps havtion, metersamples [7

The prinsystem arein a titaniumand 40 mmfrom enzymight deacbrane covemembranesimpermeabtion) occupThe draw spharmaceuton separatdrug reservsion and eitbe stable atime, usualsmall orific

Fig. 17. Crosmembrane cosolution) occtomeric pistonreservoir. Upoand the drug i

der. Exit ports can range from simple, straight channels to morecomplicated design configurations [27].

When taqueous sothe membrbuilds up,the piston,consequentfice. The raspect of athe drug wthe principmethodicalrate.

The volnt oKa

is plcan

DS

DSresse

P

Pdelivefor tnt wide o

botmbr

P

can

ic puic prion chumtartiing aed thic prluesllingn beo, w

n,

smo

ewant saal fluid, where it is ultimately taken up into neurons.elopment of osmotic pumps was pioneered with thelization of the ALZET osmotic pump for animale mid-1970s [26]. More recently, osmotic principles

applied to human therapy, resulting in the develop-g delivery systems such as the DUROS system [27],sh-PullTM, L-OROSTM, and EnSoTrol [25] that cangs continuously for up to 1 year. Recently, osmotice been studied as regulated systems for the acquisi-ing, buffering, delivery, and assay of fluid biological9].cipal components of a typical osmotic drug-deliveryillustrated in Fig. 17. The osmotic pump is contained

alloy cylindrical reservoir that is 4 mm in diameterin length. This reservoir protects the drug molecules

mes, body moisture, and cellular components thattivate the drug prior to release. A polyurethane mem-rs one end of the reservoir. Like other semi-permeable, it is permeable to water but almost completelyle to ions. The osmotic engine (i.e., the draw solu-ies a portion of the cylinder behind the membrane.olution is most often NaCl and a small amount oftical excipients in a tablet form. An elastomeric pis-es the draw solution from the drug formulation in theoir. The drug may either be a solution or a suspen-her aqueous or non-aqueous in nature. The drug mustt body temperature (37 C) for extended periods ofly from 3 months to 1 year. The drug exit port is ae located at the opposite end of the titanium cylin-

s-section of the implanted DUROS System [27]. A polyurethanevers one end of the reservoir. The osmotic engine (i.e., the drawupies a portion of the cylinder behind the membrane. An elas-

separates the draw solution from the drug formulation in the drugn diffusion of water into the osmotic engine, the piston is pusheds released through the drug outlet orifice.

partmeKedempumpe,

=wherebe exp

P =wheredrug dessaryconstahand sonly ifthe me

(DSand

Pd Eq. (9)osmotosmotreflectin the[26]. Sassum

saturatosmotThe vacontropartitioEven ssolutio

5.5. O

Rendifferehe osmotic pump is brought into contact with anlution or wet environment, water diffuses through

ane into the draw solution compartment. As pressureit expands the draw solution compartment, pushesincreases the pressure in the drug compartment, andly induces the release of the drug through the ori-ate of water diffusion is the most important designn osmotic pump because it dictates the rate at whichill be released. Theeuwes and Yum [26] summarizedles of design and operation of osmotic pumps andly described the design criteria for constant delivery

umetric flux of water into the draw solution com-f osmotic pumps can be described by the simplifiedtchalsky equation (Eq. (1) in Section 2). When theaced in an environment with an osmotic pressure of

be expressed as

e (7)is the osmotic pressure of the draw solution. P cand as

d Pe (8)is the internal pressure difference associated with thery through the orifice and Pe is the pressure nec-he piston to compress the drug reservoir. To achieveater flux, and hence constant delivery rate, the right-f Eq. (1) must be constant. This condition can be meth terms, and P, are constant (assuming thatane permeability coefficient is constant)e) = constant (9)

e = constant (10)be satisfied if (1) the amount of draw solution in the

mp can ensure saturation in the water solution, (2) theessure of the environment is negligible, and (3) theoefficient is constant. Implantable osmotic pumpsan body can easily meet these three requirements

ng from solid or very concentrated draw solution, andsmall drug reservoir, the draw solution can be keptroughout the delivery time period and at much higher

essure than the osmotic pressure in the human body.of P in Eq. (10) can be reduced to almost zero bythe orifice size and by selecting a highly deformabletween the draw solution and drug compartments [26].hen a low molecular weight salt is used as a draw will generally always be much greater than P.

tic powerpressure-retarded osmosis

ble energy can be extracted wherever two streams oflinity or different chemical potential meet. Consid-


Fig. 1

ering that tapproximawater is recan be usedSection 2 abe used to

A schemtrated in Fitaining meFO membrside of thethe pressurdiluted andone is deprond passesthe incominity are the[33,80] sugdiffer in ththe permeation of thepressurizatble of with[33,80].

Power gthat make iit is renewacompared tdensity (i.eEstimates ssalinity gra

For the leration has[3135,39,generatedpared PROfrom saliniPRO for pUnion andof NorwayForschungsTechnologymembranesequivalent

The deveration wil

for water treatment and desalination. Both processes depend onthe same semi-permeable membranes that at this time do not

ercially exist.

ncluding remarks

increased attention to FO from various disciplines ariseshe fact that FO can be employed in many fields of sci-nd engineering including water and wastewater treat-seawater/brackish water desalination, food processing,elivery, and electric power production. Despite the lackust membranes and membrane modules for FO, basich on FO and the development of new applications of FOadily growing. Currently, the most important measure ton in order to advance the field of FO is the developmentmembranes in both flat-sheet and hollow fiber configura-he membranes need to provide high water permeability,

jection of solutes, substantially reduced internal CP, highal stability, and high mechanical strength. For the future,

ensity packing methods for well-performing flat-sheet FOranes (such as the modified spiral-wound module shown9) must be further developed. Development of draw solu-at can induce high osmotic pressure is another important

wards further progress of the FO process. Most desirablew so

entrathate wi

wled

authch,andAdm

enc8. Simplified process layout for a PRO power plant [28].

he salinity of seawater yields osmotic pressures oftely 2.7 MPa and that the osmotic pressure of riverlatively insignificant, a large portion of the 2.7 MPa

for power generation. PRO (as briefly described innd illustrated in Figs. 1 and 2) is one method that canrealize this energy.atic drawing of power generation by PRO is illus-

g. 18. Fresh water is pumped into a PRO module con-mbranes, in principle, similar to the semi-permeableanes described earlier. The fresh water flows on onemembrane and diffuses through the membrane intoized side of the membrane filled with seawater. Thepressurized seawater is then split into two streams;

essurized in a turbine to generate power and the sec-through a pressure exchanger to assist in pressurizingg seawater. The two key components of a PRO facil-pressure exchanger and the membrane. Loeb et al.gested two possible steady-state flow systems that

eir means of pressurizing the draw solution enteringte-receiver side of the membrane: pump pressuriza-draw solution (as depicted in Fig. 18) and permeateion of the draw solution using at least two tanks capa-standing the pressure of the permeate-receiver side

enerated by PRO from seawater has certain featurest very attractive. It is a large and unexploited resource;ble; its use has minimal environmental impact; ando other potential sources of energy from the ocean, its., power capacity per physical size) is high [38,81].uggest that global power production potential fromdients is on the order of 2000 TWh per year [28].ast several decades, the PRO process for power gen-been continuously studied by Loeb and co-workers

50,58,59,82]. Additionally, Jellinek and Masuda [29]power from PRO pilot testing. Wick [38] com-

and reverse-electrodialysis for power generationty gradient sources. Currently, an intensive study of

comm

6. Co

Thefrom tence a

ment,drug dof robresearc

are stebe takeof newtions. Thigh rechemichigh-dmembin Fig.tions thstep toare drareconc

water,reactiv

Ackno

TheResear0801,Space

Nom

ABCDJwK

Ptower generation is being funded by the Europeanbeing conducted by a joint effort of Statkraft SF

, ICTPOL of Portugal, SINTEF of Norway, GKSS-zentrum of Germany, and Helsinki University ofof Finland with the main objective being to developfor PRO power that have a production capacity

to at least 4 W/m2 [28].elopment and future success of PRO for power gen-l, in turn, have great influence on the success of FO

Greek le

lutions that require low energy for regeneration ortion, that are easily separable from the product fresh

have low or no toxicity, and that are chemically non-th polymeric membranes.

gments

ors acknowledge the support of the Office of NavalAward Nos. N00014-03-1-1004 and N00014-05-1-the support of the U.S. National Aeronautics andinistration (NASA), Award No. NNA04-CL12A.

lature

water permeability coefficient (m3/m2 s Pa)solute permeability coefficient (m/s)concentration (mol/l)diffusivity (m2/s)water flux (m3/m2 s)resistance to salt transport in the porous support(s/m)hydraulic pressure (Pa)membrane thickness (m)

ttersdifference operatormembrane porositychemical potentialosmotic pressure (Pa)reflection coefficientpore tortuosity


References

[1] S. Sourirajan, Reverse Osmosis, Academic Press, Inc., New York, NY,1970.

[2] F. Votta,by direc

[3] D.K. AnOsmosis

[4] R.W. Hosis for cthe AWWAZ, 200

[5] E.G. Beain: ProcSystems

[6] R.J. Yorosmosisof the Ssium (S1999.

[7] Osmotekhttp://ww

[8] E.G. Beconcentr

[9] M.I. Dovtration operforma

[10] M.I. DocentratioJ. Food

[11] B. Jiao,for the c303324

[12] K.B. Peosmoticfiguratioprocess

[13] K.B. Petion of teffect ofMembr.

[14] K.B. PeJ. Food

[15] K. Poppbeverage

[16] R.E. WrE.G. Beajuice co(1993) 6

[17] E.G. Betration o1999.

[18] T.Y. Catbrane coosmoticSci. 257

[19] T.Y. Catfor wasosmoticwastewa

[20] M. Flynsit VehicMoffett

[21] R.E. KrDesalina

[22] J.R. Mccarbon dtion 174

[23] D. Cohen, Chemical Processing magazine, Mixing moves osmosistechnology forward, October 2004. http://www.chemicalprocessing.com/articles/2004/346.html.

ire Lrelabs. BhaarmaV

s/11Theeu

oticn. Bi. Writems2003). Aabrce, R.G.

osmo

. LeepressLoeb,Loebardedmbr.Loeb,: chalLoeb,eat SaardedLoebng rivion 1. Me

mbraSepp

arded138. Wi. Me

emi-p. 4 (1I Stre

p://ww. Ba

71,79. Gl

ent 3. Fra. Kesosis,

Stachter or79,03. Mc(200. Mc

T/US. McCmoni

feed. 278noMaw.na

Loeb,ric one, J.Sepperm.S.M. Barnett, D.K. Anderson, Concentration of industrial wastet osmosis: completion report, Providence, RI, 1974.derson, Concentration of Dilute Industrial Wastes by Direct

, University of Rhode Island, Providence, 1977.lloway, T.Y. Cath, K.E. Dennett, A.E. Childress, Forward osmo-oncentration of anaerobic digester centrate, in: Proceedings ofA Membrane Technology Conference and Exposition, Phoenix,

5.udry, J.R. Herron, Direct osmosis for concentrating wastewater,

eedings of the 27th International Conference on Environmental, Lake Tahoe, NV, July 1417, 1997.k, R.S. Thiel, E.G. Beaudry, Full-scale experience of directconcentration applied to leachate management, in: Proceedingseventh International Waste Management and Landfill Sympo-ardinia 99), S. Margherita di Pula, Cagliari, Sardinia, Italy,

Inc., Landfill leachate treatment, 2003, Electronic Source:w.rimnetics.com/osmotek.htm.

audry, K.A. Lampi, Membrane technology for direct osmosisation of fruit juices, Food Technol. 44 (1990) 121.a, K.B. Petrotos, H.N. Lazarides, On the direct osmotic concen-f liquid foods. Part I. Impact of process parameters on processnce, J. Food Eng., in press.

va, K.B. Petrotos, H.N. Lazarides, On the direct osmotic con-n of liquid foods. Part II. Development of a generalized model,Eng., in press.A. Cassano, E. Drioli, Recent advances on membrane processesoncentration of fruit juices: a review, J. Food Eng. 63 (2004).trotos, P.C. Quantick, H. Petropakis, A study of the directconcentration of tomato juice in tubular membrane-module con-n. I. The effect of certain basic process parameters on theperformance, J. Membr. Sci. 150 (1998) 99110.trotos, P.C. Quantick, H. Petropakis, Direct osmotic concentra-omato juice in tubular membrane-module configuration. II. The

using clarified tomato juice on the process performance, J.Sci. 160 (1999) 171177.trotos, H.N. Lazarides, Osmotic concentration of liquid foods,Eng. 49 (2001) 201206.er, W.M. Camirand, F. Nury, W.L. Stanley, Dialyzer concentratess, Food Eng. 38 (1966) 102104.olstad, M.R. McDaniel, R.W. Durst, N. Micheals, K.A. Lampi,udry, Composition and sensory characterization of red raspberry

ncentrated by direct-osmosis or evaporation, J. Food Sci. 5833637.audry, J.R. Herron, S.W. Peterson, Direct osmosis concen-f waste water: Final report, Osmotek Inc., Corvallis, OR,

h, S. Gormly, E.G. Beaudry, V.D. Adams, A.E. Childress, Mem-ntactor processes for wastewater reclamation in space. I. Directconcentration as pretreatment for reverse osmosis, J. Membr.(2005) 8598.

h, V.D. Adams, A.E. Childress, Membrane contactor processestewater reclamation in space. II. Combined direct osmosis,distillation, and membrane distillation for treatment of metabolicter, J. Membr. Sci. 257 (2005) 111119.n, J. Fisher, B. Borchers, An Evaluation of Potential Mars Tran-le Water Treatment Systems, NASA Ames Research Center,

Field, CA, 1998.avath, J.A. Davis, Desalination of seawater by direct osmosis,tion 16 (1975) 151155.Cutcheon, R.L. McGinnis, M. Elimelech, A novel ammoniaioxide forward (direct) osmosis desalination process, Desalina-(2005) 111.

[24] Shshi

[25] P.PPhnew

[26] F.osm

An[27] J.C

sys3 (

[28] R.Jsou

[29] H.Han

[30] K.Lby

[31] S.[32] S.

retMe

[33] S.sis

[34] S.Grret

[35] S.usinat

[36] G.Dme

[37] A.ret115

[38] G.L[39] G.D

a s

Sci[40] OL

htt[41] G.W

3,1[42] D.N

Pat[43] B.S[44] J.O

osm

[45] K.wa

4,8[46] R.L

B1[47] R.L

PC[48] J.R

am

andSci

[49] Naww

[50] S.fabbra

[51] A.Thaboratories Inc., EnSoTrol, Electronic Source: http://www..com/techguide/ensotrol.html.tt, Osmotic drug delivery systems for poorly soluble drugs,entures Ltd., Oxford, UK, 2004. http://www.shirelabs.com/30 04.pdf.wes, S.I. Yum, Principles of the design and operation of genericpumps for the delivery of semisolid or liquid drug formulations,omed. Eng. 4 (1976) 343353.ght, R.M. Johnson, S.I. Yum, DUROS osmotic pharmaceuticalfor parenteral & site-directed therapy, Drug Delivery Technol..erg, Osmotic powera new and powerful renewable energyeFocus 4 (2003) 4850.

Jellinek, H. Masuda, Osmo-power: theory and performance of-power pilot plant, Ocean Eng. 8 (1981) 103128., R.W. Baker, H.K. Lonsdale, Membranes for power generationure-retarded osmosis, J. Membr. Sci. 8 (1981) 141171.Osmotic power plants, Science 189 (1974) 350.

, Production of energy from concentrated brines by pressure-osmosis. I. Preliminary technical and economic correlations, J.Sci. 1 (1976) 49.Energy production at the Dead Sea by pressure-retarded osmo-

lenge or chimera? Desalination 120 (1998) 247262.One hundred and thirty benign and renewable megawatts fromlt Lake? The possibilities of hydroelectric power by pressure-osmosis, Desalination 141 (2001) 8591.

, Large-scale power production by pressure-retarded osmosiser water and sea water passing through spiral modules, Desali-43 (2002) 115122.hta, Further results on the performance of present-day osmotic

nes in various osmotic regions, J. Membr. Sci. 10 (1982) 319.ala, M.J. Lampinen, Thermodynamic optimizing of pressure-osmosis power generation systems, J. Membr. Sci. 161 (1999).ck, Energy from salinity gradients, Energy 3 (1978) 95.hta, S. Loeb, Internal polarization in the porous substructure ofermeable membrane under pressure-retarded osmosis, J. Membr.978) 261.am Analyzer 2.0, OLI Systems Inc., Morris Plains, NJ, 2005.w.olisystems.com/.tchelder, Process for the demineralization of water, US Patent9 (1965).ew, Process for liquid recovery and solution concentration, US,216,930 (1965).nk, Desalination of Sea Water, US Patent 3,670,897 (1972).sler, C.D. Moody, Drinking water from sea water by forwardDesalination 18 (1976) 297306.

e, Apparatus for transforming sea water, brackish water, pollutedthe like into a nutritious drink by means of osmosis, US Patent0 (1989).Ginnis, Osmotic desalination process (1), US Patent 6,391,2052).Ginnis, Osmotic desalination process (2), US Patent Pending02/02740 (2002).utcheon, R.L. McGinnis, M. Elimelech, Desalination by a novel

acarbon dioxide forward osmosis process: influence of drawsolution concentrations on process performance, J. Membr.

(2006) 114123.gnetics, Water purification, 2005. Electronic Source: http://nomagnetics.com/water purification.asp.L. Titelman, E. Korngold, J. Freiman, Effect of porous support

n osmosis through a LoebSourirajan type asymmetric mem-Membr. Sci. 129 (1997) 243249.

ala, M.J. Lampinen, On the non-linearity of osmotic flow, Exp.Fluid Sci. 28 (2004) 283296.


[52] L. Song, M. Elimelech, Theory of concentration polarization in cross-flow filtration, J. Chem. Soc., Faraday Trans. 91 (1995) 33893398.

[53] S.S. Sablani, M.F.A. Goosen, R. Al-Belushi, M. Wilf, Concentrationpolarization in ultrafiltration and reverse osmosis: a critical review,Desalination 141 (2001) 269289.

[54] M. Elimelech, S. Bhattacharjee, A novel approach for modeling con-centration polarization in crossflow membrane filtration based on theequivalence of osmotic pressure model and filtration theory, J. Membr.Sci. 145 (1998) 223241.

[55] M. Mulder, Basic Principles of Membrane Technology, 2nd ed., KluwerAcademic Publishers, Dordrecht, The Netherlands, 1997.

[56] S. Loeb, M.R. Bloch, Countercurrent flow osmotic processes for theproduction of solutions having a high osmotic pressure, Desalination 13(1973) 207215.

[57] G.T. Gray, J.R. McCutcheon, M. Elimelech, Internal concentration polar-ization in forward osmosis: role of membrane orientation, Desalination197 (2006) 18.

[58] G.D. Mehta, S. Loeb, Performance of Permasep B-9 and B-10 mem-branes in various osmotic regions and at high osmotic pressures, J.Membr. Sci. 4 (1979) 335349.

[59] S. Loeb, Pressure-retarded osmosis revisited: the prospects for osmoticpower at the dead sea, in: Proceedings of the Euromembrane 95 Con-ference, University of Bath, UK, 1995.

[60] I. Goosens, A. Van-Haute, The use of direct osmosis tests as complemen-tary experiments to determine the water and salt permeabilities of rein-forced cellulose acetate membranes, Desalination 26 (1978) 299308.

[61] R.W. Baker, Membrane Technology and Applications, 2nd ed., JohnWiley & Sons, Ltd., New York, NY, 2004.

[62] C. Biberdorf, Filter in a pouch, The Warrior (2004) 3.[63] Hydratio

ElectronCTGY&

[64] A. Sagivnation 1

[65] J. MallevProcesse

[66] P.O. Wito enviroMarshallnasa.gov

[67] T.Y. CaProgress

ater recovery process for advanced life support systems, in: Proceedingsof the 35th International Conference on Environmental Systems (ICES),Rome, Italy, 2005.

[68] BiogasWorks, Nutrient utilization and anaerobic digestion, 2000. http://www.biogasworks.com/Index/Nutrient%20Utilization%20&%20AD.htm.

[69] HTI, Military/X Pack, 2005. Electronic Source: http://www.hydrationtech.com.

[70] W.T. Hough, Process for extracting solvent from a solution, US Patent3,532,621 (1970).

[71] W.T. Hough, Solvent extraction into a comestible solute, US Patent3,702,820 (1972).

[72] J. Yaeli, Method and apparatus for processing liquid solutions of suspen-sions particularly useful in the desalination of saline water, US Patent5,098,575 (1992).

[73] K. Lampi, E.G. Beaudry, J. Herron, Forward osmosis pressurized deviceand process for generating potable water, US Patent 6,849,184 (2005).

[74] C.D. Moody, J.O. Kessler, Forward osmosis extractors, Desalination 18(1976) 283295.

[75] M.D. Rosa, F. Giroux, Osmotic treatment (OT) and problems related tothe solution management, J. Food Eng. 49 (2001) 223236.

[76] R.D. Scott, Dehydration and packaging of foodstuff, US Patent3,858,499 (1975).

[77] J.R. Herron, E.G. Beaudry, C.E. Jochums, L.E. Medina, Osmotic con-centration apparatus and method for direct osmosis concentration of fruitjuices, US Patent 5,281,430 (1994).

[78] E. Singer, New technologies deliver in treating neurological diseasesRNA interference, osmotic pumps may solve drug delivery problems,Nat. Med. 10 (2004).

[79] Y.C. Su, L. Lin, A water-powered micro drug delivery system, J. Micro-ctromLoeberal pverte

ergyvancevelophnoloSess

licy RLoebted brjectedn Technologies Inc., Hydration bagstechnology overview,ic Source: http://www.hydrationtech.com/merchant.mv?Screen=Store Code=MHTI&Category Code=TECH-FO., R. Semiat, Backwash of RO spiral wound membranes, Desali-79 (2005) 19.ialle, P.E. Odendaal, M.R. Wiesner, Water Treatment Membranes, McGraw-Hill, USA, 1996.eland, Design for human presence in space: an introductionnmental control and life support systems, NASA George C.Space Flight Center, Alabama, 1994. http://flightprojects.msfc.

/book/rp1324.pdf.th, V.D. Adams, A.E. Childress, S.J. Gormly, M.T. Flynn,

in the development of direct osmotic concentration wastew-

ele[80] S.

sev

con

[81] EnAdDeTecondPo

[82] S.traproech. Syst. 13 (2004) 7582., T. Honda, M. Reali, Comparative mechanical efficiency oflant configurations using a pressure-retarded osmosis energy

r, J. Membr. Sci. 51 (1990) 323335.from the ocean, A report prepared for the Subcommittee ond Energy Technologies and Energy Conservation Research,ment, and Demonstration of the Committee on Science andgy, U.S. House of Representatives, Ninety-fifth Congress, Sec-ion, Prepared by the Congressional Research Service, Scienceesearch Division, 2002.

, F.V. Hassen, D. Shahaf, Production of energy from concen-ines by pressure-retarded osmosis. II. Experimental results and

energy costs, J. Membr. Sci. 1 (1976) 249269.

forward osmosis_principles, applications, and recent developments

Documents