Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

carbohydrate metabolism in the fruit mesocarp,

which is under strict genetic control. There is a large

amount of genetic variability among melon cultigens

for the trait of sucrose accumulation, and most primi-

tive varieties do not accumulate high levels of sugar.

The genetic trait of sweetness in melon is recessive

and was selected for during the later domestication of

the species.

0038 Mature pumpkins and squash generally store sig-

nificant amounts of starch, which accumulates during

the later stages of fruit development. The starch con-

tent is correlated with fruit sensory textural attri-

butes, and the soluble sugar levels in the mature fruit

determine the fruit sweetness. Therefore, the balance

between starch and soluble sugar plays a major role

in determining fruit quality. During storage of the

mature fruit, starch is metabolized, and the soluble

sugar levels often increase, contributing to the flavor

of the cooked fruit.

0039 In addition to the soluble sugar and starch com-

ponents, the organic acid levels of cucurbit fruit can

also contribute to fruit taste, although this subject has

not attracted much attention to date. Cucumber fruit

accumulate high levels (approximately 1% on a fresh

weight basis) of citric acid during the later stages of

fruit development, but acid levels are low in imma-

ture cucumber fruit. Melon fruit have low acid levels,

and their sensory quality is therefore determined

largely by their soluble sugar content. However, the

fruit of many of the nonsweet melon types accumu-

late high levels of organic acids during their develop-

ment, and this genetic trait can be combined with the

genetic trait of high sugar content, leading to melon

varieties with novel tastes.

0040With respect to nutritional value, cucurbit fruits

are generally not ranked high, as they are fleshy

with a high water content. The immature fruits have

lower levels of nutritional components than the

mature fruits. However, some mature cucurbit fruits

contain high levels of vitamins C and A. Of the cucur-

bit fruit, the melons have the highest vitamin C levels

and a 1-cup serving of sweet melon (236g) can meet

the recommended daily requirement for vitamin C.

The orange-fleshed melons as well as the orange-

fleshed pumpkins and squash have high carotenoid

levels and a high vitamin A activity. However, the

green-fleshed honeydew melons have low carotenoid

levels and, hence, a lower vitamin A activity. The

major carotenoids of the orange-flesh fruit are a-

and b-carotene and the xanthophyls, lutein and vio-

laxanthin. The red color of the watermelon fruit is

due to the carotenoid lycopene, which is also respon-

sible for the characteristic red color of tomato fruit.

A daily serving of the orange-fleshed melon meets

the recommended daily requirement for provitamin

A, making melon the only significantly consumed

fruit in the USA that can provide the RDA of both

provitamin A and vitamin C. There exists a large

amount of genetically determined variation for carot-

enoid levels in the cucurbit fruit. In the case of Cucur-

bita pepo, which has been studied most extensively,

over a dozen genes have been characterized, which

affect fruit color ranging from pale cream to intense

orange.

tbl0003 Table 3 Nutrientcompositionofcucurbitsin100-gediblerawportion

CropWater(%)Calories(kcal)Protein(g)Fat(g)Carbohydrate(g)Fiber(g)

Melon

Cantaloupe90350.90.38.40.8

Casaba92260.90.16.20.8

Honeydew90350.50.19.20.6

Wintersquash

Acorn88400.80.110.41.5

Butternut86451.00.111.7

Hubbard88402.00.58.7

Spaghetti92310.60.66.9

Summersquash94201.20.24.41.9

Cucumber 96 13 0.7 0.1 2.8 0.8

Watermelon 92 32 0.6 0.4 7.2 0.5

Bitter gourd (Momordica) 94 17 1.0 0.2 3.7 2.8

Wax gourd (Benincasa) 96 13 0.4 0.2 3.0 2.9

Luffa gourd (Luffa) 94 20 1.2 0.2 4.4

Calabash gourd (Lagenaria) 96 14 0.6 0.02 3.4

Chayote 94 19 0.8 0.1 4.5 1.7

Pumpkin and squash, seeds 7 541 24.5 45.9 17.8 3.9

Watermelon, seeds 5 557 28.3 47.4 15.3

Data obtained from the US Department of Agriculture, Agricultural Research Service (2001) USDA Nutrient Database for Standard Reference, Release 14.

NutrientDataLaboratoryHome(www.nal.usda.gov/fnic/foodcomp).

3824 MELONS, SQUASHES, AND GOURDS

0041 Besides provitamin A, vitamin C and carotenoids,

cucurbit fruits can contain low levels of additional

antioxidants, such as flavonoids, but the total

antioxidant activity of the various cucurbit fruit

remains to be studied. Cucurbit fruit also contain

numerous bioactive compounds, including vitamin

E, biotin, and folic acid, and here too, there is

no doubt a tremendous amount of genetic variabi-

lity in the family waiting to be discovered. Momor-

dica fruit has high levels of both vitamin C and

folic acid.

0042 The cucurbits, especially wild germplasm, fre-

quently contain high levels of the cucurbitacin terpe-

noids. These are highly bitter compounds, which may

also be toxic, and they play roles in herbal and folk

medicine. They are found at high concentrations in

the colocynth gourd (Citrullus colocynthis (L.) Schra-

der), which is likely the poisonous fruit mentioned in

Kings 2:4 of the Bible. Bitterness in the cucurbits is

genetically controlled and was selected against early

in plant domestication. Nevertheless, the problem of

undesirable bitter fruit in melon, squash, and cucum-

ber still occurs.

0043 The fruit of exotic cucurbits may be sources of

important pharmacological and bioactive com-

pounds, and these are attracting increasing research

attention. For example, the mature fruit of Siraitia

grosvenorii (Swingle) Lu & Zhang (common name

luo-han-guo) contains an unusual terpenoid glyco-

side, mogroside, which is over 100 times sweeter

than sucrose. Bitter gourd (Momordica charantia)

and snake melon (Trichosanthes cucumerina) are

just some of the cucurbits that are being tested for

potential medical benefits.

0044 Seeds of many cucurbits, such as pumpkin, squash,

and watermelon, are consumed largely as snack

foods. The seeds generally have high levels of oils

and proteins, and may be considered a highly

nutritious food. The oils contain mainly unsaturated

fatty acids, primarily oleic and linoleic and, here

too, there is much genetic variation in fatty acid

composition within the family. For example, the

seeds of the white-seeded melon have primarily

linolenic acid (18:3), and lines of buffalo gourd

(Cucurbita foetidissima HBK) have been developed,

with seeds containing over 80% linoleic acid (18:2).

One of the interesting advances in cucurbit seed

improvement has been the utilization of the naked

seed, or hull-less, mutant in Cucurbita pepo. In this

mutant, discovered a little over 100 years ago, the seed

coat does not develop, thus allowing for the efficient

extraction and processing of the oils. Oil extracted

from C. pepo seeds is highly prized for use on salads

and in cooking in much of central and eastern

Europe.

Other Uses

0045In addition to their uses in edible consumption, many

cucurbit fruits have also served other utilitarian pur-

poses throughout history. Collectively described as

gourds, the dried lignified fruit rinds of numerous

cucurbit species have been used primarily as contain-

ers of various sizes, and also as fish-net floats, masks,

musical instruments, birdhouses and other uses of

anthropological interest. Species grown for their

dried fruit are primarily from the Lagenaria and

Cucurbita genera. As previously mentioned, Luffa

gourds are grown for the fibrous interior of the

dried fruit, which can serve as a sponge.

Fruit Ripening

0046Melon cultigens can be divided into two major groups

based on their ripening physiology. The reticulatus

and cantalupensis groups are climacteric, meaning

that there is a characteristic sharp and rapid peak of

respiratory activity during ripening. This is accom-

panied by the development of an abscission layer

between the pedicel and the fruit (termed ‘full-slip’).

The nonclimacteric melon fruit of the inodorus

group, such as honeydew and casaba melons, do not

show a sharp respiratory peak, and their development

generally extends over a longer period of time. They

do not develop an abscision layer, and the appropriate

stage of ripeness is determined by changes in skin

color, from white–green to creamy yellow. In all

melons, sugar is accumulated until the fruit naturally

abcisses or is removed from the vine, and therefore

the tastiest melons are those that are not prematurely

harvested. Since no storage starch accumulates in

melon fruit, in contrast to the winter squash and

pumpkins, there is no postharvest increase in sugar

content or sweetness.

0047The netted climacteric melons have a shorter stor-

age life than the nonclimacteric smooth-skinned

melons. However, the traits of rind netting and res-

piration physiology are unlinked, and recently, new

varieties of netted melons with nonclimacteric respir-

ation have been developed, revolutionizing the har-

vesting and shipping technologies of netted melons

due to their long storage life. The climacteric trait in

melons is controlled by only two genes, and the

breeding of the new types of melon is therefore

relatively straightforward.

0048The cucurbit fruit that are consumed while imma-

ture, such as cucumber and summer squash, have very

short storage lives due to their undeveloped skin and

rind. Ideally, storage of immature fruit should be

under a relatively high relative humidity, 85–95%,

to reduce fruit water loss. At the other extreme with

MELONS, SQUASHES, AND GOURDS 3825

respect to storage, pumpkins and winter squash may

be stored for a long time, frequently many months,

and during this period, the fruit undergoes ‘curing,’

part of which is the transition of the starch in the

stored fruit to sugar. For the long-term storage of

winter squash and pumpkins, the optimal relative

humidity is generally 70–75%.

Future Research and Development

0049 The tremendous genetic variability for fruit quality

components in the cucurbits remains to be characterized

and harnessed for the breeding and production of

improved melon, squash, cucumber, pumpkin, water-

melon, and other cucurbit fruit. Within each species

of cucurbit, there exist genotypes, at times primitive

landraces, that are extraordinary in their levels of a

particular quality component or bioactive compound.

This is true for sugar content, acid content, volatile

productions, nutritional content, ripening physiology

and other metabolically controlled processes in the

developing fruit. Future research will survey and

uncover these metabolic variations, and their genetic

control will be determined, allowing for their efficient

utilization. This will set the stage for the production

of new and novel varieties with beneficial character-

isitics, be it improved taste or nutrition. This effort

will be aided by advances in understanding the

molecular-genetic control of these metabolic pro-

cesses. The unfolding of the cucurbit genome, includ-

ing the cataloging of genes involved in plant

metabolism, will help determine the potential for the

genetic improvement of each of the cucurbit crops.

See also: Antioxidants: Natural Antioxidants; Ascorbic

Acid: Properties and Determination; Carotenoids:

Occurrence, Properties, and Determination; Ripening of

Fruit; Starch: Structure, Properties, and Determination;

Storage Stability: Mechanisms of Degradation

Further Reading

Corrigan VK, Irving DE and Potter JF (2000) Sugars and

sweetness in buttercup squash. Food Quality and Prefer-

ence 11: 313–322.

Heiser CB (1979) The Gourd Book. Norman, OK: Univer-

sity of Oklahoma Press.

Kirkbride JH, Jr. (1993) Biosystematic Monograph of the

Genus Cucumis (Cucurbitaceae). Boone, NC: Parkway.

Lester G (1997) Melon (Cucumis melo L.) fruit nutritional

quality and health functionality. HortTechnology 7:

222–227.

Paris HS (2000) History of the cultivar-groups of Cucurbita

pepo. Horticultural Reviews 25: 71–170.

Pitrat M, Hanelt P and Hammer K (2000) Some comments

on infraspecific classification of cultivars of melon. In:

Katzir N and Paris HS (eds) Proceedings of Cucurbita-

ceae 2000: the 7th Eucarpia Meeting on Cucurbit Gen-

etics and Breeding. Acta Horticulturae 510: 29–36.

Robinson RW and Decker-Walters DS (1997) Cucurbits.

Wallingford, UK: CAB International.

Schaffer AA, Pharr DM and Madore M (1996) Cucurbits.

In: Zamski E and Schaffer AA (eds) Photoassimilate

Distribution in Plants and Crops: Source-Sink Relation-

ships, pp. 729–758. New York: Marcel Dekker.

Seymour GB and McGlasson WB (1993) Melons. In: Sey-

mour GB, Taylor JE and Tucker GA (eds) Biochemistry

of Fruit Ripening, pp. 274–290. London: Chapman &

Hall.

Wang H, Cao G and Prior RL (1996) Total antioxidant

capacity of fruits. Journal of Agricultural and Food

Chemistry 44: 701–705.

Whitaker TW and Davis GN (1962) Cucurbits. New York:

Interscience.

Yamaguchi M (1983) Cucurbits. In: World Vegetables,

pp. 312–349. New York: AVI.

3826 MELONS, SQUASHES, AND GOURDS

MEMBRANE TECHNIQUES

Contents

Principles of Reverse Osmosis

Applications of Reverse Osmosis

Principles of Ultrafiltration

Applications of Ultrafiltration

Principles of Reverse Osmosis

H N Lazarides and E Katsanidis, Aristotelian

University of Thessaloniki, Hellas, Thessaloniki,

Greece

Introduction

0001 Vast amounts of liquid food are industrially concen-

trated for preservation purposes or in order to reduce

storage, packaging, handling, and transportation

costs. Vacuum evaporation is the predominant

method used by the food industry to produce liquid

food concentrates, despite serious drawbacks such as

poor product quality due to thermal degradation and

high energy demand for the evaporation process. Re-

verse osmosis was first suggested by Sourirajan in

1959 as a method for water desalination. Since then,

significant developments in both material science and

membrane technology have allowed reverse osmosis

to be used in the food industry, mainly for concen-

trating liquid foods without the disadvantages of

vacuum evaporation.

Theory

0002 A membrane is a permeable or semipermeable mater-

ial (polymer, inorganic, or metal), which restricts the

motion of certain species through itself. The mem-

brane, acting as a barrier, controls the relative rates of

through transport of various species and thus gives

one product depleted in certain components and a

second product concentrated in these components.

0003Membranes are used in many filtration and separ-

ation processes. Some representative processes and

relative particle sizes are shown in Table 1.

0004When a membrane separates two solutions of dif-

ferent solute concentrations, water moves through

the membrane towards the solution with the high

concentration of solutes tending to equalize the

solute concentration on the two sides of the mem-

brane. The driving force for the movement of water

is the difference in osmotic pressures of the two solu-

tions. This phenomenon is called osmosis.

0005The osmotic pressure of a dilute solution can be

calculated by Van’t Hoff’s equation:

¼ cRT=M

where P is the osmotic pressure (Pa), c is the solute

concentration (Kg/m

), R is the gas constant (J/mol.K),

T is the absolute temperature (K), and M is the molecu-

lar weight of the solute. For a wider range of concen-

trations, Gibb’s equation has been found to be more

suitable for the calculation of the osmotic pressure:

¼ðRT ln x



Þ=V

where x

is the mole fraction of pure liquid and V

the molar volume of pure liquid. When external

(hydrostatic) pressure greater than the osmotic pres-

sure is applied, water is expelled from the concentrated

solution and we observe a reverse flow, that is a flow

against the osmotic pressure gradient. This process is

called reverse osmosis. Reverse osmosis involves the

tbl0001 Table 1 Membrane separation processes

Separation process Particle size (microns) Applications

Microfiltration 0.1–10 Clarification, sterile filtration

Ultrafiltration 0.01–0.1 Separation of macromolecular solutions

Nanofiltration 0.001–0.01 Separation of small organic compounds and selected salts from solutions

Reverse osmosis (hyperfiltration) 0.0001–0.001 Separation of microsolutes and salts from solutions

Dialysis 0.0001–0.001 Separation of microsolutes and salts from macromolecular solutions

Electrodialysis 0.0001–0.001 Separation of ions from water and nonionic solutes

Gas permeation 0.00001–0.0001 Separation of gas mixtures

MEMBRANE TECHNIQUES/Principles of Reverse Osmosis 3827

most ‘tight’ membranes, which are capable of separat-

ing the smallest solutes. The pores of the membrane are

so small that, in some materials, they can hardly be

resolved, even by scanning electron microscopy.

0006 The basis for separation is molecular size, but other

factors such as molecular shape and charge can also

play an important role.

0007 Membranes currently used in commercial reverse

osmosis installations are asymmetric, flat sheet mem-

branes of cellulose acetate (CA) or cellulose triacet-

ate, fine hollow fibers of aromatic polyamides (PA)

or cellulose triacetate and thin-film composites

(TFC). In TFC membranes an extremely fine layer

of highly hydrophilic polymer has been placed on a

microporous support usually made from polysulfone.

Figure 1 depicts the structure of a typical CA mem-

brane and shows the chemical affinity (selectivity) of

the membrane for water molecules. The chemically

induced negative sorption (repulsion) of the ions in

solution and the preferential sorption of pure water to

the membrane result in rejection of ions, thus prevent-

ing them from permeating the membrane. It should be

noted that very-low-molecular-weight organics and

uncharged solutes may pass through the membrane.

In general, reverse osmosis membrane efficiency is

expressed as a percentage rejection of NaCl, and the

typical range for such membranes is 96–99%.

Operational Requirements

0008In conventional filtration processes, all the material

to be filtered flows perpendicularly through the filter

where the separation occurs (dead-end flow). Sus-

pended solids captured by the filter tend to build up,

slowing down the filtration rate and requiring

frequent cleaning.

0009In reverse osmosis, cross-flow systems are used

almost exclusively. In a cross-flow system, the feed

Impurity

Membrane skin

OOC

fig0001 Figure 1 Movement of water through a cellulose acetate membrane. Adapted from Lacey (1972).

3828 MEMBRANE TECHNIQUES/Principles of Reverse Osmosis

stream (influent) flows parallel (tangentially) to the

membrane and only a portion of the stream passes

through the membrane as permeate or filtrate. The

remainder of the stream is called retentate or concen-

trate. The main advantage of the cross-flow system is

that the feed stream constantly removes the accumu-

lated solids (cake) or solutes from the membrane

surface, reducing build-up, which allows for less fre-

quent clean-up of the membrane system. The two

types of flow and their representative flux and cake

build-up are depicted in Figure 2.

0010 In reverse osmosis, the external pressure has to

exceed the osmotic pressure of the solution so

that solvent (usually water) may be expelled. For

example, the osmotic pressure of sea water is

approximately 2.3 MPa and a desalination plant

utilizing reverse osmosis may operate at pressures of

up to 7 MPa. This makes it clear that the reverse

osmosis systems must be very robust from a physical

standpoint and the membranes must be able to resist

mechanical compression (compaction), which can

deform their morphology and negatively affect their

performance.

0011 Resistance to mechanical pressure is only one of the

properties that a good membrane must have. Mem-

branes need to function in a variety of environments,

where the operating pressure, temperature, pH, and

chemical compatibility need to be taken into account.

Cellulose acetate membranes are usually used at tem-

peratures up to 40



C, but they may be operated at

temperatures up to 65



0012The typical pH range for CA membranes is 2–8 and

for PA 3–11. Membranes may be used outside the

above ranges, but their useful life will be shorter.

0013A good reverse osmosis membrane needs to with-

stand chemical agents such as chlorine, iodine, other

oxidizing agents, oils, and solvents. CA has good

chlorine resistance (for sanitation purposes) but TFC

PA has poor chlorine resistance.

0014Several reverse osmosis membranes used in com-

mercial applications and their properties are pre-

sented in Table 2.

Membrane Configurations

0015There are several configurations for the containment

and support of the membranes. These configurations

need to address all the requirements discussed previ-

ously and also need to allow for easy access to the

membranes for replacement. There are four major

configurations that have been used in the food

industry.

Spiral Wound

0016This module consists of a flat sheet membrane, a

support system, and a spacer system, all wrapped

around a perforated premeate-collection tube. The

feed stream flows across the membrane through the

spacer and the permeate is collected in the central

permeate-collection tube. Specially designed spacer

systems enhance turbulent flow, which helps reduce

fouling of the membrane. The main advantages of this

Cake thickness

Permeate flow

Time

Dead-end filtration Cross-flow filtration

Time

Feed

Permeate flow

fig0002 Figure 2 Dead-end vs cross-flow filtration. Adapted from Mannapperuma JD (1997) Design and performance evaluation of mem-

brane systems. In: Rotstein E, Singh RP and Valentas KJ (eds) Handbook of Food Engineering Practice, pp. 168–208. Boca Raton, FL:

CRC Press LLC.

MEMBRANE TECHNIQUES/Principles of Reverse Osmosis 3829

module are the high packing density, the high flux,

and the fact that the flat sheet membrane can with-

stand higher pressures than the hollow fiber module.

Capillary (Hollow Fiber)

0017 In this module, the feed stream is pumped into a

vessel that contains a bundle of thin hollow fibers

(about 40 mm internal diameter). The permeate goes

through the thin fibers and is separated from the

concentrate. The advantage of this system is that it

provides a very large membrane area for the total size

of the system (high packing density). The disadvan-

tages are the low flux and the need for pretreatment

of the feed solution (i.e., filtration, centrifugation,

microfitration, even ultrafiltration).

Plate and Frame (Flat Sheet)

0018 This module consists of flat membrane sheets placed

between support frames that create flow channels

(about 0.5 mm) for the feed stream, the concentrate,

and the permeate. The membrane and its supports are

stacked together in large numbers (up to 360 mem-

brane sheets) to create one unit. The main advantages

of this module are that it is possible to remove and

clean the membrane sheets and that the system is

reusable (once the membranes have been cleaned

or replaced). The disadvantages are the high labor

and time investment required to disassemble and

reassemble the module and the low packing density

of the unit.

Tubular

0019 This module is relatively uncommon in reverse

osmosis. The feed and concentrate streams flow

through the inside of the tube (up to 12 mm internal

diameter) and the filtrate passes through the mem-

brane and is collected in the housing around the tube.

The advantages of this system are the ease of cleaning

and the capability to handle high viscosities. The

disadvantages are the relatively small membrane

area for the size of the system and the high feed

rates that are required for the system to operate

efficiently (high energy/pumping cost).

0020The spiral wound and the hollow fiber are the two

modules that are mostly used in the food industry.

Figure 3 illustrates these module designs. In industrial

systems, usually several units are assembled together

in series and/or in parallel configuration to achieve

the desired results. These systems may be single-pass

systems or they may include recirculation loops and

pumps so that part of the feed solution passes through

the membrane multiple times.

Performance Comparisons

0021Reverse osmosis is a very efficient process, allowing

for the simultaneous concentration, fractionation,

and purification of the product and the accomplish-

ment of multiple tasks in a single unit operation. It

does not impart pH or chemical changes in the prod-

uct and, since no significant heating is required,

there is no heat degradation of the product. Thus,

reverse osmosis has a minimal effect on the quality

characteristics and nutritional value of the finished

product, especially when compared to evaporative

concentration where inevitably there is heat degrad-

ation, as well as flavor and nutritional losses.

0022Since reverse osmosis systems do not require

steam, evaporators, and condensers, they require

much less floor space and equipment than evapora-

tive systems. Due to their relative simplicity,

reverse osmosis systems have shorter come-up and

shutdown time. They are also more flexible and easier

to modify or upgrade than conventional evaporative

systems.

0023Reverse osmosis is very energy-efficient, because it

typically operates at ambient temperature (no heating

or cooling required); and, most importantly, there is

no phase change requirement for water removal, as in

evaporative processes. Overall, reverse osmosis

systems require less energy than evaporative systems

per unit of water removed from the product. Reverse

osmosis systems require almost exclusively electric

energy for pumping and recirculating, whereas evap-

orative systems require steam in addition to electric

energy for pumping.

tbl0002 Table 2 Properties of some commercial reverse osmosis membranes

Membrane type Module type pH Chlorine tolerance Oxidation tolerance

Cellulose acetate blend Spiral wound 3–8 Fair Fair

Cellulose triacetate Capillary fiber 4–9 Fair Good

Aromatic polyamide Hollow fiber 4–11 Poor Fair

Cross-linked polyether TFC Spiral wound 1–12 Poor Fair

Aryl-alkyl polyetherurea TFC Spiral wound 3.5–12 Poor Fair

Cross-linked fully aromatic polyamide TFC Spiral wound 1–12 Poor Fair

TFC, thin-film composites.

Adapted from Fell CJD (1995) Reverse osmosis. In: Noble RD and Stern SA (eds) Membrane Separations Technology, pp. 113–144. Amsterdam, The

Netherlands: Elsevier.

3830 MEMBRANE TECHNIQUES/Principles of Reverse Osmosis

0024 Thermal energy (steam) can be generated from oil

at very high efficiency (80–90%), but electrical/mech-

anical energy is generated from oil at much lower

efficiency (about 30%), yielding a heat-to-power

ratio of 3.0. Table 3 compares the energy require-

ments of reverse osmosis to evaporation with either

thermal vapor recompression or mechanical vapor

recompression. The relative cost of electric energy to

oil or natural gas energy varies a lot in different areas

of the world, thus making it hard for generalized

comparisons of energy costs of the two processes.

However, it seems that in countries such as the USA,

France, and Germany, reverse osmosis is much less

costly than any multistage evaporative process, with

either mechanical or thermal vapor recompression.

0025 In large desalination plants, energy can be re-

covered from the rejected brine by running it through

turbines to generate electricity that can be used in the

plant. Such a recovery system can lead to energy

savings of up to 30%.

Operational Issues

0026In membrane systems a reduction of flux over time is

often encountered, while all operating parameters

remain the same. This is typically due to membrane

compaction, concentration polarization, and fouling.

Compaction

0027As discussed previously, the high operating pressures

in reverse osmosis result in increased membrane dens-

ity, decreased thickness, and deformed (compressed)

membrane pores, which reduce the permeate flux

through the membrane.

Concentration Polarization

0028In all cross-flow processes, as soon as the flow begins,

a build-up of a polarized layer or a gel-like layer of

solutes is observed at the membrane surface. This

phenomenon can result in locally increased osmotic

pressure in the vicinity of the membrane surface that

tbl0003 Table 3 Comparison of energy requirements of reverse osmosis and evaporation with either thermal vapor recompression (TVR) or

mechanical vapor recompression (MVR)

Type of plant Direct energyinput Equivalent energy input (kWh kg

1

Steam (kg kg

1

O) Power (kWh kg

1

Evaporation TVR 3 effects 0.20–0.25 0.003 0.17–0.20

Evaporation TVR 5 effects 0.13–0.17 0.003 0.11–0.14

Evaporation TVR 7 effects 0.077–0.13 0.003 0.07–0.10

Evaporation MVR 0.02 0.01–0.03 0.046–0.11

Reverse osmosis 0.005–0.012 0.015–0.048

In this table a heat-to-power ratio of 3 has been used.

Adapted from Hallstro

m B (1988) Preconcentration: new developments. In: Bruin S (ed.) Preconcentration and Drying of Food Materials, pp. 46–47. New

York, NY: Elsevier.

Permeate

out

Tube sheet Hollow fine fibers Feed tube

Feed in

Retentate out

Permeate out

Permeate spacer

Feed out

Feed spacer

Feed in

Membrane

Nub

fig0003 Figure 3 Hollow fine fiber (top) and spiral wound (bottom) module designs. Adapted from Mannapperuma (1997) Design and

performance evaluation of membrane systems. In: Rotstein E, Singh RP and Valentas KJ (eds) Handbook of Food Engineering Prac-

tice, pp. 168–208. Boca Raton, FL: CRC Press LLC.

MEMBRANE TECHNIQUES/Principles of Reverse Osmosis 3831

will negatively affect the flux through the membrane.

Over time, the accumulated solutes can affect the

affinity of the membrane for certain solutes or solv-

ents. Concentration polarization is a reversible phe-

nomenon that can be reduced by periodical changes

in the velocity of the feed solution (tantalizing), by

increasing the velocity, or by decreasing the pressure

and concentration of the feed solution.

Fouling

0029 Fouling refers to the irreversible changes that occur on

the membrane and result in reduced flux over time.

Many factors contribute to membrane fouling. Pore

plugging and solute adsorption in the membrane pores

can result in a decrease of the pore diameter and

reduced flux. Solutes accumulated on the membrane

surface due to concentration polarization may under-

go irreversible changes over time and form a layer that

will cause hydrodynamic resistance to flow. This layer

formation depends on the membrane surface chemis-

try, solute–membrane and solute–solute interactions.

0030 All the above make it clear that pretreatment of the

feed stream is crucial in extending the useful mem-

brane life. The objective of the pretreatment process

would be to remove as many suspended solids and

contaminants as possible before reverse osmosis. The

pretreatment process may be a simple filtration or

centrifugation step or it may be a much more thor-

ough micro- or even ultrafiltration step. Pretreatment

requirements will depend on the type of membrane

used, the membrane configuration, and the compos-

ition and flow rate of the feed solution.

Cleaning

0031 Mechanical cleaning can be applied in large-diameter

tubular modules using oversized sponge balls to clean

the tubes.

0032 Hydraulic cleaning can be achieved by back-

washing (reversing the flow), by back-shock treat-

ment (back-flushing for very short time), or by the

pulsation of flow (increasing and decreasing the

pressure of the feed solution). Back-washing and

back-shock treatment cannot be applied to TFC

membranes because of their tendency to delaminate.

0033 Electric cleaning uses a pulsed electric field resulting

in the movement of charged particles away from the

membrane. Electric cleaning can be applied without

interrupting the process, but requires electricity-

conducting membranes and special module design.

0034 Chemical cleaning is the most important method

for controlling fouling phenomena. Chemical com-

patibility between the cleaning agent and the mem-

brane is the most important factor in selecting the

proper cleaning agent. Acids (nitric, phosphoric,

citric) are usually used for inorganic contaminants

and basic (caustic) detergents or chlorine are typically

used for organic contaminants (i.e., fat, protein).

Other chemicals used may be complexing agents

(ethylenediaminetetraacetic acid; (EDTA), enzymes,

and disinfectants (H

, NaOCl).

Applications

0035Water desalination and/or purification is one of the

major applications of reverse osmosis. Brackish water

or sea water is treated by reverse osmosis to produce

potable water. Also, ultrapure water for medical or

industrial applications can be produced.

0036Reverse osmosis is used in the dairy industry for

milk, whey, or lactose concentration. Concentration

of whey, produced as a byproduct from the cheese-

making process, is the most widely used application

in the food industry. Spiral module configurations

with CA or TFC membranes are most often used in

the dairy industry.

0037Tubular and spiral module systems equipped with

TFC membranes are used for juice concentration.

Reverse osmosis is economically efficient for concen-

trations up to 24% solids, but it could be used to

concentrate juices up to 60% solids, if economically

feasible. Membranes developed for reverse osmosis

have also been used in direct osmosis applications

for the preconcentration of liquid foods (i.e., juices)

with the use of high osmotic pressure brines on one

side of the membrane. In these applications the

driving force is not hydrostatic pressure (pressure

are rather low; < 500 kPa) but the osmotic pressure

difference on the two sides of the membrane.

0038A thorough discussion of reverse osmosis applica-

tions can be found in a following section.

See also: Effluents from Food Processing: On-Site

Processing of Waste; Evaporation: Basic Principles;

Uses in the Food Industry; Filtration of Liquids;

Membrane Techniques: Applications of Reverse

Osmosis; Principles of Ultrafiltration