ULTRAFILTRATION

Definition ……..
The average diameter
of pores of membrane =
10 – 1000 Angstrom

Ultrafiltration (UF) is a variety of membrane filtration in which forces like

pressure or concentration gradients lead to a separation through a
semipermeable membrane. Suspended solids and solutes of high
molecular weight are retained in the so-called retentate, while water and
low molecular weight solutes pass through the membrane in the
permeate. This separation process is used in industry and research for
purifying and concentrating macromolecular (103 - 106 Da) solutions,
especially protein solutions. Ultrafiltration is not fundamentally different
from microfiltration. Both of these separate based on size exclusion or
particle capture. It is fundamentally different from membrane gas
separation, which separate based on different amounts of absorption and
different rates of diffusion.
History ……..
 The first synthetic ultrafiltration membranes were prepared by
Bechhold from collodion (nitro cellulose); he also coined the term
“ultrafilter”
 By the mid-1920s, collodion ultrafiltration and microfiltration
membranes were commercially available for laboratory use; no
industrial applications existed until 1960s
 The crucial breakthrough was the development of the anisotropic
cellulose acetate membrane by Loeb Sourirajan in 1963; their goal
was to produce high-flux reverse osmosis membranes, but Michaels
at Amicon realized the general applicability of the technique.
History ……..

 Michaels and his coworkers produced ultrafiltration membranes

from cellulose acetate and many other polymers including
polyacrylonitrile copolymers, aromatic polyamides, polysulfone,
and poly(vinylidene fluoride)
 In 1969 Abcor (now a division of Koch Industries) installed the first
commercial ultrafiltration equipped with tubular membrane
module to recover electrocoat paint from automobile paint shop
rinse water
 In 1970 the first cheese whey ultrafiltration system was installed
 The early systems used tubular or plate and frame module
 Hollow fiber (capillary) modules produced by Romicon in 1973, and
spiral wound modules (by Abcor) became a commercial item by
1979-1980
History ……..

 The principal problem inhibiting wider application of ultrafiltration

is membrane fouling. The problem is controlled, but not eliminated,
by module and system design and by regular membrane cleaning.
Development of membranes with surface properties designed to
minimize fouling has also helped.
Process Diagram ……..
Process Diagram ……..
Plant ……..
Plant ……..
Characterization of Ultrafiltration Membranes ……..
Ultrafiltration membranes are usually anisotropic structures made as
Loeb-Sourirajan technique. They have a finely porous surface layer
or skin, supported on much more open microporous substrate. The
finely porous surface layer performs the separation; whereas, the
microporous substrate provides mechanical strength.

Ultrafiltration and microfiltration are related process; the distinction

between the two lies in the pore size of membrane. Microfiltration
membranes have larger pores and are used to separate particles in
the 0.1 – 10 μm range; whereas, ultrafiltration ones have 10 –
1000 Angstrom of pore diameter.
Characterization of Ultrafiltration Membranes ……..
The cut-off of ultrafiltration membranes is usually characterized by
solute molecular weight, but several other factors affect
permeation through these membranes. When membrane retention
measurements are performed with linear water soluble molecules
such as polydextran, poly(ethylene glycol) or poly(vinyl pirrolidone),
the measured rejection is much more lower than the rejection
measured for proteins of the same molecular weight.

It’s believed that linear water soluble polymer molecules are able to
snake through the pores; however, proteins exist in the solution as
tightly wound globular coils held together by hydrogen bonds.
Characterization of Ultrafiltration Membranes ……..
The membrane shows significant rejection to globular protein
molecules as small as pepsin (MW 35000) and cytochrome c (MW
13000), but is completely permeable to a flexible linear polydextran
(MW > 100,000)

Globular Proteins Linear Polymer

Solute Pepsin Cytochrome c Polydextran

MW 35000 13000 100000

Rejection (%) 90 70 0
Characterization of Ultrafiltration Membranes ……..
The pH of the feed solution also affects the permeation through the
ultrafiltration membranes, particularly with polyelectrolytes. For
example, poly(acrylic acid) is well rejected by ultrafiltration
membranes at pH 5 and above, but is completely permeable
through the same membrane at pH 3 and below.

The reason relates to the change in configuration of the polyacid. At

pH 5 or above, poly(acrylic acid) is ionized; the negatively charged
carboxyl groups along the polymer backbone repel each other; the
polymer coil is then very extended and relatively inflexible. At pH 3
or below, the carboxyl groups along the poly(acrylic acid) polymer
backbone are all protonated; the resulting neutral molecule is much
more flexible and can pass through the pores.
Concentration Polarization & Membrane Fouling ……..

A key factor determining the performance of ultrafiltration membranes

is concentration polarization, which causes membrane fouling due
to the deposition of retained colloidal and macromolecul material
on the membrane surface. The pure water flux of ultrafiltration
membranes is often very high (> 1 cm3/cm2.min = 350 gal/ft2.day)
Concentration Polarization & Membrane Fouling ……..

The pure water flux of ultrafiltration membranes is often very high (> 1
cm3/cm2.min = 350 gal/ft2.day). However, to separate
macromolecul or colloidal solutions, the flux drops to typically
0.1 cm3/cm2.min. The drop is caused by the formation of a gel layer
of retained solute on the membrane surface due to concentration
polarization.
Concentration Polarization & Membrane Fouling ……..

Solvent molecules permeate the membrane, but the larger solutes

accumulate at the membrane surface. Because of their size, the
rate at which the rejected solute molecules can from the
membrane surface back to the bulk solution is relatively low; their
concentration at the membrane surface = 20-50 times higher than
that in the feed solution.
Concentration Polarization
& Membrane Fouling ……..

Because of the effect of the secondary layer on selectivity,

ultrafiltration membranes are not commonly used to fractionate
macromolecules mixtures. Most commercial ultrafiltration
applications involve processes in which the membrane completely
rejects all the dissolved macromolecular and colloidal material in
the feed solution, while completely passing water and dissolved
microsolutes.

Efficient fractination by ultrafiltration is only possible if the species

differ in molecular weight by a factor of 10 or more.
Membrane Fouling
Cleaning ……..

The easiest way to clean the membrane is by circulating an appropriate

cleaning solution through the membrane modules for 1 or 2 hours.
The most common ultrafiltration fouling layers – organic polymer
colloids and gelatinous materials – are best treated with alkaline
solutions followed by hot detergent solutions.
Membrane Fouling
Cleaning ……..

Enzymatic detergents are effective when the fouling layer is a

proteinaceous gel. Ca, Mg, and silica scales, which usually give a
problem in RO, can permeate through the ultrafiltration membrane.
Many feed waters contain small amount of soluble ferrous iron
salts, hydrated iron oxide scaling is a problem. In the ultrafiltation
system, these salts are oxidized to ferric iron by entrained air. Ferric
iron is insoluble in water so insoluble iron hydroxide gel is formed.
Such deposits are usually removed with a citric or hydrochloric acid
wash
Membrane Fouling Cleaning ……..

The period of the cleaning cycle can vary from daily for food
applications, such as ultrafilttration of whey, to once a month or
more in ultrafiltration used as polishing unit in ultra pure water
system. A typical cleaning cycle is as follows:
1. Flush the system several times with hot water at the highest
possible circulation rate
2. Treat the system with an appropriate acid or alkali wash,
depending on the nature of the layer
3. Treat the system with a hot detergent solution
4. Flush the system thoroughly with water to remove all traces of
detergent; measure the pure flux water through the membrane
modules under standard test conditions. If the restoration of flux is
less than expected, repeat steps 1-3
Membrane Fouling Cleaning ……..

Mechanical cleaning may also be used, particularly if chemical

cleaning doesn’t restore the membrane flux. Early electrocoat paint
systems used 1-in-diameter tubular membrane modulew. These
tubes could be effectively cleaned by forcing sponge balls with a
slightly larger diameter than the tube through the tube; the balls
gently scaped the membrane surface, removing deposit material.
This way is effective but time-consuming process.

Backflushing is another way of cleaning heavily fouled membranes. It’s

widely used to clean capillary ang ceramic membrane modules that
can withstand a flow of solution from permeate to feed without
damaging the membrane. Spiral wound modules are not good to
use this method. To avoid membrane damage, the backflushing
pressures are 5 – 15 psi.
Membrane Fouling
Cleaning ……..

The densified gel layer could be removed by periodic cleaning of the

membrane. If regular cleaning is repeated many times, the
membrane flux eventually does not return to the original value due
to: some particles remain on the membrane surface, or internal
fouling which is caused by penetration of solid material into the
membrane
Membrane Fouling
Cleaning ……..

Ultrafiltration module lifetimes are rarely more than 2-3 years. For
cheese, whey, and electrocoat paint applications, the modules may
be replaced annually. In contrats, RO membranes are normally not
cleaned more than once or twice per year, and can last 4-5 years.
Mass Transfer Theory ……..
Formation of gel layer:
At any point within the boundary layer, the convective flux of solute to
the membrane surface is given by the volume flux (Jv) of the
solution through the membrane multiplied by the concentration of
solute ci. At steady state, this convective flow is balanced by the
diffusive flux of retaibed solute in the opposite direction, as follows:

dci
J v .ci  Di .
dx

Integrating the equation gives:

c gel  J v . 
 exp  
cib  Di 
Mass Transfer Theory ……..

The formation of a gel layer produces a limiting or plateau flux that

cannot be exceeded at any particular operating condition. Once a
gel layer is formed, increasing the applied pressure does not
increase the flux, but merely increases the gel thickness.

The maximum flux (plateau or limiting flux) can be determined from:

.ln cib  ln c gel 

D
J max  

Mass Transfer Theory ……..
The selectivity can be measured as the salt rejection coefficient R:

 cp 
R  1  .100%
 cb 

p = permeate side
b = bulk solution (feed)