Separation Processes

Lecture Notes 13 - Adsorption 1
Engr. Caesar P. Llapitan
I. INTRODUCTION
Adsorption, ion exchange, and chromatography are sorption operations, in which certain
components in a mixture are selectively transferred to insoluble materials suspended or packed in a
column. The components that were selectively transferred are known as solutes. The sorbed solutes
are known as sorbates, and the materials used for sorbing the solutes are known as sorbents.
Adsorption is a process whereby a substance (adsorbate, or sorbate) is accumulated on the surface

of a solid (adsorbent, or sorbent).
The adsorbate can be in a gas or liquid phase. The driving force for adsorption is unsaturated forces
at the solid surface which can form bonds with the adsorbate. These forces are typically electrostatic
or van der Waals interactions (reversible). Stronger interactions involve direct electron transfer
between the sorbate and the sorbent (irreversible). The strength of this interaction dictates the relative
ease or difficulty in removing (desorbing) the adsorbate for adsorbent regeneration and adsorbate
recovery. The selective nature of the adsorbent is primarily due to the relative access and strength of
the surface interaction for one component in a feed mixture. The solid is the mass-separating agent
and the separating mechanism is the partitioning between the fluid and solid phases. An energy-
separating agent, typically a pressure or temperature change, is used to reverse the process and
regenerate the sorbent.
Figure 1 Sorption operations with solid-particle sorbents

Retrieved from Ref [4]
Adsorption processes are used economically in a wide variety of separations in the chemical process
industries. Activated carbon is the most common adsorbent, with annual worldwide sales estimated
at $380 million [1]. One common adsorption process is dehydration for the drying of gas streams.
Adsorption offers several advantages as a separation process. It can be used in situations where
distillation is difficult or impossible due to components with similar boiling points, vapor–liquid
azeotropes, or species with low relative volatilities. A high loading of solute is possible in adsorption
processes and it works well for dilute systems. In addition, the energy-separating agent needs are
usually low. The disadvantages of adsorption processes are due to the use of an adsorbent mass-
separating agent. Each adsorbent bed must be removed from the process before regeneration, so
that more than one column in series or parallel is usually required. The regeneration process can
involve losses in the sorbent amount and loading over time. Column operation affects solute
breakthrough that limits actual bed loading.
Definition of Adsorption Terms

 Adsorption: the process where a substance is accumulated on an interface between phases.
 Adsorbent: the phase that collects the substance to be removed at its surface.
 Adsorbate: the solute that is to be adsorbed (removed from the gas or liquid stream).
 Isotherm: A relation between the equilibrium amount of a substance adsorbed per weight of
sorbent and its concentration in the liquid or gas stream at constant temperature.
Examples of industrial applications of sorption operations are presented below:

A. Adsorption
1. Gas Purifications
Removal of organics from vent streams
Removal of SO2 from vent streams
Removal of sulfur compounds from gas streams
Removal of water vapor from air and other gas streams
Removal of solvents and odors from air
Removal of NOX from N2
Removal of CO2 from natural gas
Gas bulk separation:
N2/O2
H2O/Ethanol
Acetone/vent streams
C2H4/vent streams
Normal paraffins/iso-paraffins, aromatics
2. Liquid purifications
Removal of H2O from organic solutions
Removal of organics from H2O
Removal of sulfur compounds from organic solutions
Decolorization of solutions
Liquid bulk separation:
Normal paraffins / iso-paraffins
Normal paraffins / olefins
p-xylene / other C8 aromatics
p- or m-cymene / other cymene isomers
Fructose / dextrose, polysaccharides
B. Ion Exchange
Water softening
Water demineralization
Water dealkalization
Decolonization of sugar solutions
Recovery of uranium from acid leach solutions
Recovery of antibiotics from fermentation broths
Recovery of vitamins from fermentation broths
C. Chromatography
Separation of sugars
Separation of perfume ingredients
Separation of C4 - C10 normal and iso-paraffins
Sorbents
To be suitable for commercial applications, a sorbent should have the following characteristics:
1. High selectivity to enable sharp separation
2. High capacity to minimize the amount of sorbent needed
3. Favorable kinetic and transport properties for rapid sorption
4. Chemical and thermal stability, including extremely low solubility in the contacting fluid, to
preserve the amount of sorbent and its properties
5. Hardness and mechanical strength to prevent crushing and erosion
6. Free-flowing tendency for ease of filling or emptying vessels
7. High resistance to fouling for long life

8. No tendency to promote undesirable chemical reactions
9. Capacity of being easily regenerable
10. Relatively low cost
Adsorbents
Major types of adsorbents in use are: activated alumina, silica gel, activated carbon, molecular sieve
carbon, molecular sieve zeolites and polymeric adsorbents.
Most adsorbents are manufactured (such as activated carbons), but a few, such as some zeolites,
occur naturally. Each material has its own characteristics such as porosity, pore structure and nature
of its adsorbing surfaces.
Pore sizes in adsorbents may be distributed throughout the solid. Pore sizes are classified generally into
3 ranges: macropores have "diameters" in excess of 50-nm, mesopores (also known as transitional
pores) have "diameters" in the range 2 - 50-nm, and micropores have "diameters" which are smaller
than 2-nm.
Many adsorbent materials, such as carbons, silica gels and aluminas, are amorphous and contain
complex networks of inter-connected micropores, mesopores and macropores. In contrast, pores in
zeolitic adsorbents have precise dimensions.
Table 1 Adsorbent types. Reproduced with permission from Ref [1]. All rights reserved
Adsorbent Characteristics Commercial uses Strengths Weaknesses

Hydrophobic Removal of Cheapest Difficult to
surface, favors Organic hydrophobic regenerate if
organics over air pollutants from adsorbent, fouling occurs,
Activated carbon
or water aqueous and workhorse of the sometimes can
gaseous effluents pollution-control catch fire during
business regeneration
Separates on the Production of N2 The only practical Has not found
Carbon basis of different from air adsorbent any uses except
molecular intra-particle favoring O2 for air separation
Sieves diffusivities adsorption
over N2
High-capacity, Primarily drying Higher capacity Not as effective
hydrophilic of gas streams, than zeolite as zms in
adsorbent sometimes used molecular sieves removing traces
Silica gel
for hydrocarbon (zms) of H2O from
removal from gases
gases
High-capacity Primarily drying Higher capacity Not as effective
hydrophilic of gas streams than zms as zms in
Activated
adsorbent removing traces
alumina
of H2O from
gases
Hydrophilic Dehydration, air Can make Lower total
surface; polar, separation, separations capacity than
Zeolite molecular regular geometry-based based both on many other
sieves (zms) chemicals separations, polarity adsorbents
many and geometry
others
Hydrophobic Removal of Can be burned Costlier than
surface; organics from off more easily activated carbon
Silicalite
adsorption gas streams than activated
characteristics carbon
similar to those of
activated carbon
Usually styrene Removal of Not as subject to Much more costly
Polymer
organics from fouling as than activated
adsorbents
gas streams activated carbon carbon
Surfaces that Removal of low Excellent in Only economical
Irreversible react selectively levels of H2S, SO2, removing trace for removal of
adsorbents with components etc., from gases contaminates <100 kg/day of
of gas streams Adsorbents
Activated sludge Removal of No regeneration Percent removals
on a porous organics from needed often lower than
Biosorbents
support gas streams for other
adsorbents
Typical applications of commercial adsorbents:

Adsorbent Applications
Drying of gases, refrigerants, organic solvents, transformer oils
Silica Gel Desiccant in packings and double glazing
Dew point control of natural gas
Drying of gases, organic solvents, transformer oils
Activated Alumina Removal of HCl from hydrogen
Removal of fluorine in alkylation process
Nitrogen from air
Hydrogen from syngas
Ethene from methane and hydrogen
Vinyl chloride monomer (VCM) from air
Removal of odors from gases
Carbons
Recovery of solvent vapors
Removal of SOX and NOX
Purification of helium
Clean-up of nuclear off-gases
Water purification
Oxygen from air
Drying of gasses
Removing water from azeotropes
Sweetening sour gases and liquids
Purification of hydrogen
Separation of ammonia and hydrogen
Recovery of carbon dioxide
Separation of oxygen and argon
Zeolites
Removal of acetylene, propane and butane from air
Separation of xylenes and ethyl benzene
Separation of normal from branched paraffins
Separation of olefins and aromatics from paraffins
Recovery of carbon monoxide from methane and hydrogen
Drying of refrigerants and organic liquids
Pollution control, including removal of Hg, NOX and SOX
Recovery of fructose from corn syrup
Water purification
Recovery and purification of steroids, amino acids
Polymers & Resins Separation of fatty acids from water and toluene
Separation of aromatics from aliphatics
Recovery of proteins and enzymes
Removal of colors from syrups

Removal of organics from Hydrogen peroxide
Treatment of edible oils
Removal of organic pigments
Clay
Refining of mineral oils
Removal of polychlorinated biphenyls (PCBs)
Criteria for Use

Distillation is usually the first choice for bulk separation of components in a liquid phase. Another option
often used is air stripping. In both cases, the ease of separation is based on the relative volatility of the
various components in the feed mixture. In comparison, adsorption is a good choice when:
1. The relative volatility between the key components to be separated is in the range of 1.2 to
1.5 or less. In this case, distillation is not an obvious choice and the separation factor for
adsorption could be very large based on criteria other than relative volatility.
2. The bulk of the feed is a relatively low value, more volatile component, and the product of
interest is in relatively low concentration. For this case, large reflux ratios (and large energy
consumption) would be required if distillation is considered. This situation is often found in
environmental applications where one has diluted concentrations of a pollutant in an air or
water stream.
3. The two groups of components to be separated have overlapping boiling ranges. Again,
compared to distillation, several distillation columns are required even if the relative volatilities
are large. For adsorption, this separation can be done effectively if the two groups contain
chemically or geometrically dissimilar molecules so as to obtain a high separation factor.
4. A low temperature and a high pressure are required for liquefaction (for gases).
5. Components in the feed stream can be damaged or altered by heating. These alterations
can include chemical degradation, increases in chemical reactivity and polymerization
reactions. This chemical change can cause fouling in the equipment which reduces the
effective lifetime as well as the production of undesirable components.
6. Chemical corrosions, precipitation, and/or undesirable chemical reactions (explosive
conditions are an example) are problems using operating conditions for distillation.
7. Factors favorable for separation by adsorption exist. Adsorption selectivity between the key
components should be greater than 2, the sorbent can be easily regenerated, does not easily
foul for the given feed mixtures, and does not act as a catalyst or other reaction medium to
promote the production of undesirable side products. The feed throughput and product purity
are also important considerations. The costs for adsorption separation are generally lower than
distillation for small to medium throughputs (less than a few tons per day), and when high-purity
products are not required (function of selectivity of the sorbent with respect to the feed
components). The major cost for pressure-swing adsorption (discussed in the next section) is
the compressor costs. If a feed gas mixture is available at an elevated pressure, the separation
costs are substantially reduced.
II. PRINCIPLES OF ADSORPTION
Adsorption (also known as adsorptive separation) can be simply defined as the concentration of a
solute, which may be molecules in a gas stream or a dissolved or suspended substance in a liquid
stream, on the surface of a solid. The major applications had been in the separation of solutes from
liquid streams and removal of impurities from gas streams.
In an adsorption process (see Figure 2), molecules or atoms or ions in a gas or liquid diffuse to the
surface of a solid, where they bond with the solid surface or are held there by weak inter-molecular
forces (see Figure 3). The adsorbed solute is called the adsorbate, and the solid material is the
adsorbent.
Figure 2 Flow of feed stream through a solid adsorbent
Figure 3 Bulk fluid phase
To achieve a very large surface area for adsorption per unit volume, highly porous solid adsorbents
with small diameter inter-connected pores are used. The adsorbents are less than 5-mm in diameter,
with pore sizes of the order of 0.01- m in diameter. The porous structures can account for up to 50% of
the volume of the material.
During adsorption, the solid adsorbent becomes saturated or nearly saturated with the adsorbate. To
recover the adsorbate and allow the adsorbent to be reused, it is regenerated by desorbing the
adsorbed substances (i.e. the adsorbates).
As shown above, we see that a feed stream containing a contaminated component is passed
through a solid adsorbent, and the contaminant component is retained in the adsorbent. Now, if a
second liquid feed (known as the desorbent) containing a different component that also has an
affinity for adsorbing to the solid adsorbent, there will be competition between the contaminant and
this component for the limited number of adsorption sites available. This will result in the removal or
"desorption" of the contaminant component from the solid, as shown in the Figure 4 below:
Figure 4 Flow of desorbent
If sufficient quantity of desorbent is fed to the solid adsorbent, the contaminant component can be
completely removed. Because the contaminant is removed and the solid adsorbent, the adsorbent
is said to have been "regenerated" and so can be used again to adsorb and concentrate more
contaminant from fresh liquid streams.
Removal of adsorbates can also be achieved by changing the pressure or the temperature. This will
be covered in more details in later sections.
Adsorption processes can be divided into 2 groups: bulk separation, which involves the separation of
up to half of the components from a process stream, and purification, a process in which a small
amount of impurity is removed from the gas stream.
Physical Adsorption v.s. Chemisorption

The phenomenon of adsorption is essentially an attraction of adsorbate molecules to an adsorbent
surface. The preferential concentration of molecules in the proximity of a surface arises because the
surface forces of an adsorbent solid are unsaturated. Both repulsive and attractive forces become
balanced when adsorption occurs.
Adsorption is nearly always an exothermic process.
We can distinguish between 2 types of adsorption process depending on which of these 2 force types
plays the bigger role in the process. Adsorption processes can be classified as either physical
adsorption (van der Waals adsorption) or chemisorption (activated adsorption) depending on the
type of forces between the adsorbate and the adsorbent.
In physical adsorption, the individuality of the adsorbate and the adsorbent are preserved. In
chemisorption, there is a transfer or sharing of electron, or breakage of the adsorbate into atoms or
radicals which are bound separately.
Physical adsorption from a gas occurs when the inter-molecular attractive forces between molecules
of the solid adsorbent and the gas are greater than those between molecules of the gas itself. In
effect, the resulting adsorption is like condensation, which is exothermic and thus is accompanied by
the release of heat.
Physical adsorption occurs quickly and may be mono-molecular (unimolecular) layer or monolayer,
or 2, 3 or more layers thick (multi-molecular). As physical adsorption takes place, it begins as a
monolayer. It can then become multi-layer, and then, if the pores are close to the size of the
molecules, more adsorption occurs until the pores are filled with adsorbate. Accordingly, the
maximum capacity of a porous adsorbent can be more related to the pore volume than to the
surface area.
In contrast, chemisorption involves the formation of chemical bonds between the adsorbate and
adsorbent is a monolayer, often with a release of heat much larger than the heat of condensation.
Chemisorption from a gas generally takes place only at temperatures greater than 200 0C, and may
be slow and irreversible. Most commercial adsorbents rely on physical adsorption; while catalysis relies
on chemisorption.
Although the theoretical difference between physical and chemical adsorption is clear, the
distinction is not as simple as it may seem. The following parameters can be used to evaluate an
adsorbate-adsorbent system to establish the type of adsorption:
 The heat of physical adsorption is in the same order of magnitude as the heat of condensation,
while the heat of chemisorption is of the same order as of the corresponding.
 Physical adsorption will occur under suitable temperature-pressure conditions in any gas solid
system, while chemisorption takes place only if the gas is capable of forming a chemical bond
with the surface.
 A physically adsorbed molecule can be removed unchanged at a reduced pressure at the
same temperature where the adsorption took place. The removal of the chemisorbed layer is
far more difficult.
 Physical adsorption can involve the formation of multi-molecular layers, while chemisorption is
always completed by the formation of a monolayer. In some cases, physical adsorption may
take place on the top of a chemisorbed monolayer.
 Physical adsorption is instantaneous (although the diffusion into porous adsorbent may be time
consuming), while chemisorption, though may be instantaneous, and generally requires
activation energy.
Heat Effects
When a molecule adsorbs from the gas phase, the heat it liberates is perhaps 2 to 3 times larger than
the heat of vaporization, while a molecule adsorbing from the liquid phase liberates an amount of
heat nearly as great. There are only 2 ways in which the heat of adsorption can be dissipated: by
heating up the adsorbent (and the vessel containing it) or by heating up the fluid exiting from the
vessel.
When the concentration of adsorbate in the feed is low, say only a few wt % or less, then a large
fraction of the heat of adsorption passes out with the effluent stream, and the adsorbent remains
close to the temperature of the incoming fluid.
When the concentration of the adsorbate is high, say 10 % wt or more, a large fraction of heat will be
trapped in the bed and causes large temperature increase with time.
Factors Influencing Adsorption

(1) Surface Area
The extent of adsorption is proportional to specific area. Specific area can be defined as that
portion of the total area that is available for adsorption.
(2) Nature of the Adsorbate

a. Solubility: Adsorption of a solute is inversely proportional to its solubility in the solvent. The greater
the solubility, the stronger the solute-solvent bond and the smaller the extent of adsorption
(Lundelius's rule).
In general, the solubility of any organic compound in water decreases with increasing chain
length, because the compound becomes more hydrocarbon-like as the number of carbon atoms
becomes greater (Traube's rule).
Adsorption of aliphatic series of organic acids increases in the order formic-acetic-propionic-

butyric, whereas the order is reversed for adsorption from toluene.
b. Molecular sizes: If the rate is controlled by intraparticle transport, the reaction generally will
proceed more rapidly the smaller the adsorbate molecule, within a given class of compound.
c. Charge (Ionic species): As long as the compounds are structurally simple. Adsorption is at
minimum for the charged species and at a maximum for the neutral species. As compounds
become more complex, the effect of ionization becomes of decreasing important.
Decreasing adsorption with increasing ionization has been observed for many organic acids.
Adsorption of propionic acid on carbon decreases markedly with increasing pH to a minimum in
the range pH 3.5 to 5.5 (succinic acid to a min. in the range pH 4.0 and 7.0, caproic acid between
pH 4.8 and 7.5).
Studies on amphoteric compound indicate an adsorption maximum at the isoelectric point, or

that pH at which both the acidic end and the basic end of the compound are ionized and the
compound bears a net charge of zero.
d. pH
Because hydrogen and hydroxide ions are adsorbed quite strongly, the adsorption of other ions is
influence by the pH of the solution.
In general, adsorption of typical organic pollutant from water is increased with decreasing pH.
e. Temperature
Adsorption reaction are normally exothermic; thus, the extent of adsorption generally increases
with decreasing temperature.
f. Adsorption of Mixed Solutes

The degree of mutual inhibition of competing adsorbates should be related to the relative sizes of
the molecules being adsorbed, to the relative adsorptive affinities, and to the relative
concentrations of solutes.
g. Nature of the Adsorbent

The physicochemical nature of the adsorbent can have profound effects on both rate and
capacity for adsorption. Commercial carbons can be prepared from a variety of raw materials,
including wood, lignite, coal, petroleum residues, and nut shells. The raw material is generally
activated in an atmosphere of CO2, CO, O2, H2O, vapor, air, or other selected gases, at a
temperature between 300 and 100oC often followed by quenching in air or water. The most
characteristic physical property of activated carbon is its extremely large surface area, which is
composed mainly or surfaces bordering inner pore spaces.
Sorbent Regeneration
In most instances, it is necessary to regenerate the sorbent after each cycle of use. This step is
important to reduce the replacement costs of the sorbent, extend the useful life of the sorbent and
minimize the quantity of solid material that must be discarded. For gases, there are two methods
generally used.
a) Pressure Swing Adsorption (PSA)

Regeneration in a PSA process is achieved by reducing the partial pressure of the adsorbate. There
are 2 ways in which this can be achieved: (1) a reduction in the system total pressure, and (2)
introduction of an inert gas while maintaining the total system pressure. In the majority of pressure
swing separations, a combination of the 2 methods is employed. Use of a purge fluid alone is unusual.
The Figure below shows the effect of partial pressure on equilibrium loading for a Type I isotherm at a
temperature of T1. Reducing the partial pressure from p 1 to p2 causes the equilibrium loading to be
reduced from q1 to q2.
Changes in pressure can be effected very much more quickly than changes in temperature, thus
cycle time of pressure swing adsorption (PSA) processes are typically in the order of minutes or even
seconds.
PSA processes are often operated at low adsorbent loadings because selectivity between gaseous
components is often greatest in the Henry's Law region. It is desirable to operate PSA processes close
to ambient temperature to take advantage of the fact that for a given partial pressure the loading is
increased as the temperature is decreased.
Typical PSA processes consist of 2-Bed system, although other systems (e.g. 1-Bed system or complex,
multiple-beds system) had also been developed.
Uses of PSA processes

PSA processes are a popular process for performing bulk separations of gases. Separations by PSA
and VSA are controlled by adsorption equilibrium or adsorption kinetics. Both types of control are
important commercially. For the separation of air with zeolites, adsorption equilibrium is the controlling
factor. Nitrogen is more strongly adsorbed than oxygen. For air with about 21% oxygen and 79%
nitrogen, a product of nearly 96% oxygen purity can be obtained. When carbon molecular sieves are
used, oxygen and nitrogen have almost the same adsorption isotherms, but the effective diffusivity of
oxygen is much larger than nitrogen. Hence more oxygen is adsorbed than nitrogen, and a product
of very high purity nitrogen (99%) can be obtained.
b) Temperature Swing Adsorption (TSA)

Regeneration of adsorbent in a TSA process is achieved by an increase in temperature. The Figure
below showed schematically the effect of temperature on the adsorption equilibrium (Type I isotherm)
of a single adsorbate.
For any given partial pressure of the adsorbate in the gas phase (or concentration in the liquid phase),
an increase in temperature leads to a decrease in the quantity adsorbed. If the partial pressure
remains constant at p1, increasing the temperature from T1 to T2 will decrease the equilibrium loading
from q1 to q2.
A relatively modest increase in temperature can effect a relatively large decrease in loading. It is
therefore generally possible to desorb any components provided that the temperature is high
enough. However, it is important to ensure that the regeneration temperature does not cause
degradation of the adsorbents.
A change in temperature alone is not used in commercial processes because there is no mechanism
for removing the adsorbate from the adsorption unit once desorption from the adsorbents has
occurred. Passage of a hot purge gas or steam, through the bed to sweep out the desorbed
components is almost always used in conjunction with the increase in temperature.
A very important characteristic of TSA processes is that they are used virtually exclusively for treating
feeds with low concentrations of adsorbates.
Two additional methods which can shift equilibrium conditions to favor desorption are purge stripping
and displacement, as described by Ruthven [5].
c) Displacement Purge Adsorption (DPA)

Adsorbates can be removed from the adsorbent surface by replacing them with a more preferentially
adsorbed species. This displacement fluid, which can be a gas, a vapor or a liquid, should adsorb
about as strongly as the components which are to be desorbed. If the displacement fluid is adsorbed
too strongly then there may be subsequent difficulties in removing it from the adsorbent.
The mechanism for desorption of the original adsorbate involves 2 aspects:
1) Partial pressure (or concentration) of original adsorbate in the gas phase surrounding the
adsorbent is reduced.
2) There is competitive adsorption for the displacement fluid. The displacement fluid is present on
the adsorbent and thus will contaminate the product.
One advantage of the displacement fluid method of regeneration is that the net heat generated or
consumed in the adsorbent will be close to zero because the heat of adsorption of the displacement
fluid is likely to be close to that of the original adsorbate. Thus, the temperature of the adsorbent should
remain more or less constant throughout the cycle.
With neither pressure nor temperature changes from adsorption to desorption, regeneration by
displacement purge depend solely on the ability of the displacement fluid to cleanse the bed in
readiness for the next adsorption step. A typical Displacement Purge Adsorption (DPA) process is
shown in the Figure below.
A is the more strongly adsorbed component in the binary feed mixture of (A and B) while D is the
displacement purge gas. The feed mixture of (A and B) is passed through Bed 1 acting as the adsorber,
which is preloaded with D from the previous cycle (when Bed 1 was the regenerator).
A is adsorbed and the product of a mixture of (B and D) emerges from the top of the column. (B and
D) are easily separated by distillation so that B is collected in a relatively pure state.
The displacement gas D then enters Bed 2 acting as regenerator and from which emerges a mixture
of (A and D). (A and D) can be separated without difficulty in another distillation column.
In effect the original mixture of (A and B), which would have been difficult to separate by PSA or TSA,
is separated by the "intervention" of another strongly adsorbed component D. The ease of separation
of A from D, and B from D, in the additional distillation stages, is crucial in determining the economies
of displacement purge cycle operation.
Table 2 Factors governing choice of regeneration method. Reproduced with permission from Ref [1].
All rights reserved.
Method Advantages Disadvantages

Thermal swing  Good for strongly adsorbed  Thermal aging of adsorbent
species: small change in T gives  Heat loss means inefficiency in
large change in q (heat of energy usage
sorption)
 Desorbate may be recovered at  Unsuitable for rapid cycling so

high concentration adsorbent cannot be used with
 Gases and liquids maximum efficiency
 In liquid systems, high latent heat of
interstitial liquid must be added.
Pressure swing  Good where weakly adsorbed  Very low P may be required
species is required in high purity  Mechanical energy more
 Rapid cycling-efficient use of expensive than heat
Adsorbent  Desorbate recovered at low Purity
Purge stripping Operation at constant temperature Large purge volume required
and total pressure
Displacement  Good for strongly held species Product separation and recovery
 Avoids risk of cracking reactions needed (choice of desorbent is
during regeneration crucial)
 Avoids thermal aging of Adsorbent
III. ADSORPTION ISOTHERMS
A. Gas-Phase Adsorption Isotherms

In adsorption, dynamic phase equilibrium is established for the distribution of the solute between the
fluid (gas, vapor or liquid) and the solid surface.
The equilibrium is usually expressed in terms of partial pressure (gas, vapor) or concentration (liquid) of
the adsorbate in the fluid and the solute loading on the adsorbent, expressed as mass, mole or volume
of adsorbate per unit mass, mole or volume of the adsorbent.
Unlike vapor-liquid and liquid-liquid equilibria, where theory is often applied to estimate phase
distribution, no acceptable theory has been developed to estimate fluid-solid adsorption equilibria.
Thus, it is necessary to obtain experimental equilibrium data for a particular solute, or mixtures of solutes
and/or solvent, and a sample of the actual solid adsorbent material of interest.
If the data are taken over a range of fluid concentrations at a constant temperature, a plot of solute
loading on the adsorbent versus concentration or partial pressure in the fluid can be made. Such a
plot is called the adsorption isotherm. We focused primarily on gas-phase adsorption. Liquid-phase
adsorption exhibits their own isotherms as well.
Gas-Phase Adsorption
For pure gases, experimental physical adsorption isotherms have shapes that are classified into 5 types
as discussed below. Each of these types is observed in practice but by far the most common are types
I, II and IV.
Figure 5 Brunauer’s five types of adsorption isotherms. (P=P0 ¼ total pressure/vapor pressure.)
An inherent property of Type I isotherms is that adsorption is limited to the completion of a single
monolayer of adsorbate at the adsorbent surface. Type I isotherms are observed for the adsorption of
gases on microporous solids whose pore sizes are not much larger than the molecular diameter of the
adsorbate. Complete filling of these narrow pores corresponds to the completion of a molecular
monolayer. An example is the adsorption of oxygen on carbon black at -183 oC. Adsorption in all other
types do not reach a limit corresponding to the completion of a monolayer.
Type II isotherms do not exhibit a saturation limit as did Type I. This type of isotherm indicates an
indefinite multi-layer formation after completion of the monolayer and is found in adsorbents with a
wide distribution of pore sizes. Near to the first point of inflexion (point A) a monolayer is completed,
following which adsorption occurs in successive layers. An example is the adsorption of water vapor
on carbon black at 30 0C.
Type III isotherm is obtained when the amount of gas adsorbed increases without limit as its relative
saturation approaches unity. This type of isotherm is obtained when bromine is adsorbed on silica gel
at 20 0C.
Type IV isotherm is a variation of Type II, but with a finite multi-layer formation corresponding to
complete filling of the capillaries. The adsorption terminates near to a relative pressure of unity. This
type of isotherm is obtained by the adsorption of water vapor on activated carbon at 30 0C.
Type V isotherm is similar variation of Type III obtained when water vapor is adsorbed on activated
carbon at 100 0C.
Uses of Adsorption Isotherms

Adsorption isotherms are most commonly used to select the adsorbent or even the adsorption process
as a unit operation for the adsorptive separation of gases.
If the adsorption isotherm shape is Type I, II or IV, adsorption can be used to separate the adsorbate
from the carrier gas. If it is Type III or V, adsorption will probably not be economical for the separation.
Although isotherms are indicative of the efficiency of an adsorbent for a particular adsorbate removal,
they do not supply data to permit the calculation of contact time or the amount of adsorbent required
to reduce the solute concentration below prescribed limits.
Adsorption Equilibria
A variety of different isotherm equations have been proposed, some of which have a theoretical
foundation and some being of a more empirical nature. Many of these equations are valid over small
relative pressure ranges but do not fit experimental data when tested over the full range of relative
pressures.
Adsorption is a reversible process. As some molecules attach themselves to the surface, others are
kicked off back into the fluid. At equilibrium, the rate of movement of molecules onto the surface
(adsorption) and the rate movement off the surface (desorption) are equal. Both rates are
dependent on concentrations. The concentration in the fluid is usually denoted by Ce, the mass
concentration in mg/L or other such units. The concentration on the surface is denoted by x/m where
x is the mass of adsorbate adsorbed onto mass m of solid.
1. The Langmuir Isotherm

This isotherm described adsorbate-adsorbent systems in which the extent of adsorbate coverage is
limited to one molecular layer at or before a relative pressure of unity is reached (i.e. Type I). Although
the isotherm is more usually appropriate for the description of chemisorption, the equation is
nevertheless obeyed at moderately low coverage by a number of systems and can be readily
extended to describe the behavior of binary adsorbate systems.
The isotherm was formulated on the basis of a dynamic equilibrium between the adsorbed phase and
the gaseous or vapor phase. The implicit assumptions are:
 No lateral interaction between adsorbed molecules, thus the heat of adsorption is constant
and independent of coverage
 Each adsorbate molecule occupies only one site
 The adsorbed molecule remained at the site of adsorption until it is desorbed (i.e. the
adsorption is localized)
If  is the fraction of adsorbent surface covered by adsorbate then (1 - ) is the fraction of surface
remaining uncovered and therefore available as adsorption sites. The rate of adsorption is
proportional to (1 - ) while the rate of desorption is proportional to . At equilibrium, these 2 rates are
equal. Thus,
k a p1  θ   k d θ
where ka and kd are the rate constants for adsorption and desorption respectively, p is the partial
pressure of the gas.
q bp
θ 
qm 1  bp
ka
where: b
kd
and qm is the maximum loading corresponding to complete coverage of the surface by solutes.
On linearization, the equation becomes:
p 1 p
 
q bqm qm
The quantities b and qm are treated as empirical constants. Thus, the plot of p/q vs. p is linear and the
slope (gradient) gives (1/qm) and the y-intercept gives (1/bqm).
At low pressures where bp tends to zero, the equation reduces to the linear form:
q
θ  bp
qm
At high pressures, bp tends to infinity, and q = qm, i.e.  = 1.
Examples of gas adsorption onto porous solids which obey the Langmuir equation are CH 4-sodalite,
Kr-carbon, C3H8-5A zeolite.
2. The Freundlich Isotherm

The heat of adsorption is many instances decreases in magnitude with increasing extent of adsorption.
This decline in heat of adsorption is logarithmic, implying that adsorption sites are distributed
exponentially with respect to adsorption energy. This isotherm does not indicate an adsorption limit
when coverage is sufficient to fill a monolayer ( = 1). The equation that describes such isotherm is the
Freundlich Isotherm, given as:
1/n
x/m  q  kp
where: n > 1.
k is a temperature-dependent constant.
With n = 1, the equation reduces to the linear form: q = k p
On linearization, the equation becomes:
1
log q  log k  log p
n
The plot of log (q) vs. log (p) gives a straight line with slope of (1/n) and y-intercept of log (k).
3. The Brunauer-Emmett-Teller (BET) Isotherm

In its most useful form, the BET equation is written as:
p 1 C 1 p
  
qp s  p qm C qm C p s
where ps is the saturated vapor pressure.
At fixed temperature, a plot of the LHS against p/ps would yield a slope of (C – 1)/ qmC and a y-
intercept of 1/qmC thus enabling both qm and C to be determined.
By appropriate choice of C and qm the equation can be made to fit any of the isotherm types II to V
inclusive.
For small values of p/ps, the BET equation reduces to the Langmuir equation.
Other isotherms based on various other models, e.g. Grant and Manes, Vacancy Solution Model, etc.
is also available.
Gas Mixtures and Extended Isotherms

Commercial application of physical adsorption involves mixtures rather than pure gases. Experimental
data had shown that in adsorption of 2 or more components, one component can increase,
decrease, or have no influence on the adsorption of the other component, depending on the
interactions of the adsorbed molecules.
A simple theoretical treatment is to the extension of the Langmuir equation, by neglecting such
interactions and assumes that the only effect is the reduction of vacant surface area available for
adsorption.
Other Adsorption Equilibria

Other diagrams for displaying adsorption equilibrium data includes: adsorption isobars (e.g. amount
of gas adsorbed per gram of adsorbent vs. temperature at various pressures), and adsorption isosteres
(pressure vs. temperature graphs at various constant amount adsorbed).
B. Liquid-Phase Adsorption Isotherms
Sample Problems:
1. An adsorption study was conducted by adding varying amounts of activated carbon to a series
of seven flasks containing 500mL of feed water used in soft drink preparation having an initial TOC
of 2.0mg/L. The flasks were agitated for 14h, and the residual, steady-state TOC concentrations
were determined. Plot the Langmuir and Freundlich isotherms for the data presented below and
determine the values of the appropriate constants. Only hand in one of the isotherms.
Flask No. Carbon Dosage (mg) Final TOC (mg/L)
1 0 2.0
2 4.4 1.7
3 9.7 1.4
4 14 1.2
5 28 1.0
6 56 0.9
7 140 0.8
(Tip: Find the non-adsorbable fraction of the TOC. As this does not participate in any way to the
adsorption, its value should be disregarded, i.e. subtracted from the TOC to leave an adsorbable
dissolved organic carbon, ADOC, which may then fit the adsorption models better.)
Solution
Langmuir isotherm
1. Calculate the values of x and x/m from the data, e.g. for flask 1:
250mL
x  1.27mg/L  0.97mg/L  0.15mg
1000mL/L
Assume the non-absorbable TOC is 0.73 mg/L (from a linear extrapolation)
x 0.15mg
  0.034mg/mg
m 4.4mg
Mass of
Initial Final 1
Carbon Adsorbate x/m
Flask No. ATOC ATOC 1/C
Dosage Adsorbed (mg/mg) x/m
(mg/L) (mg/L) (C)
(mg) (mg) (x)
1 0 1.27 1.27 0.00 0.000

2 4.4 1.27 0.97 0.15 0.034 1.031 29.333
3 9.7 1.27 0.67 0.30 0.031 1.493 32.333
4 14 1.27 0.47 0.40 0.029 2.128 35.000
5 28 1.27 0.27 0.50 0.018 3.704 56.000
6 56 1.27 0.17 0.55 0.010 5.882 101.818
7 140 1.27 0.07 0.60 0.004 14.286 233.333
1
2. Plot values of 1/C vs. as shown in Figure 6.
x/m
3. Determine the values of the constants a and b.
Read value of intercept 1/b = 6.02, b = 0.17
Slope 1/ab = 15.84, a = 1/15.84b = 1/(15.84)(0.17) = 0.38
4. The Langmuir equation is
1 1 1 15.84
   6.02 
x m b abC C
250.000
y = 15.837x + 6.0192
R² = 0.9949
200.000
1/(x/m)
150.000
100.000
50.000
0.000
0.000 5.000 1/C 10.000 15.000
Figure 6 Plot of Langmuir isotherm
Note: The non-adsorbable value should be a bit lower than 0.8. If you choose 0.8, the graph will
move horizontally between the two first (left) two points. Also, physically speaking, it should be
a bit lower than 0.8 as we have not applied such a huge dosage yet, and larger dosages would
result in a TOC < 0.8...
2. The Freundlich isotherm can be accepted to represent equilibrium concentrations of phenol in the
g/L concentration range and x/m represents g phenol/mg C.
a) If the limiting concentration of phenol is set at 0.2 g/L and the source contains 30 g/L,
calculate the required dosage of powdered activated carbon.
b) If instead of a single dosage, the carbon was dosed twice, first to achieve a concentration of
3g/L and then in a second tank to the final requirement, how much carbon will be required?
c) How could we minimize carbon requirements over two dosages?
d) Qualitatively consider the effect of even more tanks and possibilities of reusing the carbon
from tank to tank. The isotherm may be extrapolated.
Figure 7 Freundlich isotherm for Problem 2

Solution
a) According to the graph, when c = 0.2  g/L, x/m = 0.019, assuming 1L phenol solution:
(30 μg/L  0.2μ.2μg 1L 29.8 μg
 0.019 μg phenol/mg C  m   1568 mg C
m (mg) 0.019 μg phenol/mg C
b) According to the graph, when c = 3  g/L, x/m = 0.089, assuming 1L phenol solution:
(30 μg/L  3 μg/L)  1L 27 μg
 0.019 μg phenol/mg C  m   303.4 mg C
According to the graph, when c = 0.2 g/L, x/m = 0.019, assuming 1L phenol solution:
(3 μg/L  0.2 μg/L)  1L 2.8 μg
 0.019 μg phenol/mg C  m   147.4 mg C
So, the total carbon required is: 303.4 mg C + 147.4 mg C = 450.8 mg C
c) In order to minimize the carbon usage over two dosages, a concentration of the effluent of
phenol from the first activated carbon dosage must be determined. If this concentration is too
close to the initial phenol concentration, the dosage used in the second step will be decreased
while that used in the first step will be increased, and vice versa. More elegantly, write and
equation of C as a function of PAC dose, differentiate and solve when = 0. PAC dosages of
about 160 mg/L each will represent the lowest dosage for a two-step process.
d) The total dosage needed will decrease with an increase of tanks. If the carbon from the last step
is recycled towards the first tank, countercurrent to the water flow, the adsorption capacity of
carbon can be maximized and minimize the carbon requirement even further. You can see that
the PAC from the last step above would have much adsorption capacity left when reused in the
first tank.
3. A laboratory column 10cm in diameter and 3m deep, is found to produce good results in COD
removal when operated at a flow of 50L/h. Calculate the following:
a) The application rate in m/h.
b) The residence time, t, in the column.
c) The volumetric flow rate, Vb, in bed volumes per hour, for this residence time.
d) You need to design a GAC system consisting of two columns in series for treating the same
water at a rate of 0.5 ML/d. Practical considerations limit the diameter of the columns to 3m.
Calculate the required height of the columns allowing for sufficient backwash space.
Solution
2
2  10 cm  10 2 m/cm 
a) The section area A = πr  3.14     0.00785 m2
 2 
Volume of GAC = 3m x 0.00785 = 0.02355 m3
3 3
Q 50 L/h  10 m /L
The application rate or approach velocity v   2
 6.37 m/h
A 0.00785 m
h 3m
b) The residence time t =   0.471h
v 6.37 m/h
c) Assume the fractional void volumes for granular carbon columns is 0.5 (normally range from
0.40 to 0.55)
Volumetric flow rate = 1/residence time = 1/0.471 = 2.12 bed volumes/h
d) Q = 0.5 ML/d = 0.5 ML/d  1000 m3/ML  1 d/24 h = 20.83 m3/h
The volume of the GAC in the columns has to be V = Qt = 20.83 m3/h  0.471 = 9.82 m3
The required sectional area of the column is calculated from the approach velocity:
A = 20.83/6.37 = 3.27 m2, i.e 2m diameter
The height of GAC in each column has to be h = V/A (2) = 1.5 m
Assume backwash space 50% (typical value is 20 ~ 50%)
The height for each column then has to be 2.25m.
Note: The loading rate on pilot scale was low. We might want to work at 10-20 m/h.
4. Ref [6] Water, following conventional treatment and chlorination, contains 100 μg/L chloroform.
This exceeds permissible levels and you need to check the implications of using GAC to keep this
below 50 μg/L. Estimate the bed life and carbon usage rate. Comment on the suitability of GAC
for this purpose.
According to Table 14-2: K = 0.0925 (mg/g) (L/μg)1/n, i.e. 1/n = 0.67
Using Freundlich: (qc)0 = 0.0925 (mg/g) (L/μg)0.669  (100 μg/L)0.67 = 2.0 mg/g
Eq 14-20: Y = [(qe)0 / (C0 - C1)]  ρ = 2.0 / (0.1 - 0.05)  500 kg/m3 = 20,000 m3/m3 = bed capacity
A typical bed capacity of 15 minutes would then have a life of 20,000/ (24×4) = 208 days.
Find carbon utilization rate from 14-21: CUR = (C0 – C1)/qe0 = 0.05/2 = 0.025 kg GAC/ m3 water
GAC suitability Practical limitation will be competition. The adsorbability of chloroform is poor
compared with most other substances, so GAC life may be much more limiting. Furthermore,
chloroform is poorly biodegradable. There is a need to keep chloroform low. Adapting pretreatment
and disinfection practices to minimize formation of chloroform is better, but GAC can play a role, if
not a very effective one, to keep this DBP to an acceptable level.
Exercise 1:
1. Ref [3] An adsorption study is set up in laboratory by adding a known amount of activated
carbon to six flasks which contain 200 mL of an industrial waste. An additional flask containing
200 mL of waste but no carbon is run as a blank. Plot the Langmuir isotherm and determine the
values of the constants.
Flask No Mass of C Volume in Final COD

(mg) Flask (mL) (mg C /L)
1 804 200 4.7
2 668 200 7.0
3 512 200 9.31
4 393 200 16.6
5 313 200 32.5
6 238 200 62.8
7 0 200 250
2. Ref [5] The following laboratory data were collected in a batch adsorption study. Plot the data
according to Freundlich Isotherm and determine the values for the constants n and K F. A
volume of 500 mL is placed in each flask, and the waste has an initial COD of 100 mg/L.
Flask No Mass of Volume in Final COD

C (mg) Flask (mL) (mg C /L)
1 965 500 3.5

2 740 500 5.2
3 548 500 8.0
4 398 500 12.5
5 265 500 20.5
6 168 500 33
7 0 500 100
3. Ref [5] Effluent water from a biological process is contacted with activated carbon and
allowed to come to equilibrium. The equilibrium data in terms of total organic carbon (TOC)
are given below. Determine the best values of the constants for the Langmuir and Freundlich
isotherms.
Ce , TOC in solution, qe , TOC on Carbon,

mg/L mg/mg C
1.8 0.011
4.2 0.029
7.4 0.046
11.7 0.062
15.9 0.085
20.3 0.097
4. Ref [3] Iron – coated clays (“red clays”) can adsorb phosphorus. A municipality is considering
the use of these materials for the removal of P from the town’s wastewater prior to discharge
into a lake. The following adsorption data are obtained in the laboratory using the clay and
secondary effluent from the town’s treatment facility.
a) Determine the maximum P adsorption capacity of this clay (mg P/g clay).
b) The P concentration in the secondary effluent is 8 mg P/L and the flow is 3 MGD. Make a
preliminary estimate of the clay required (tons/month) to remove 99% of P in a batch
reactor. State any simplifying assumptions that you make. (MGD: million galloon/day)
qe Ce
P adsorbed, Equilibrium P in solution,
mg P / g clay mg P / L
0.2 0.055
0.3 0.088
0.5 0.182
1.0 0.50
1.7 2.78
IV. ADSORPTION OPERATIONS
Equilibrium Relations for Adsorbents

The concentration between the concentration of a solute in the fluid phase and its concentration on
the solid resembles somewhat the equilibrium solubility of a gas in a liquid.
Linear equation:
q  kc
 This linear isotherm is not common, but in dilute region, it can be used to approximate
data of many systems
Freundlich isotherm equation

n
q  kc
log q  n  log c  log k
 Empirical, often approximate data for many physical adsorption systems and is
particularly useful for liquids.
Langmuir isotherm equation

q c
q 0 Theoretical
kc
1 k 1 1
 
q q0 c q0
A. Stagewise Operations
Liquid solutions, where the solute to be removed is adsorbed relatively strongly compared with the
remainder of the solution, are treated in batch, semi-continuous or continuous operations in a manner
analogous to the mixer-settler operations of liquid extraction (contact filtration). Continuous
countercurrent cascades can be simulated or actually realized by use of such techniques as fluidized
beds.
Gases are treated for solute removal or for fractionation, usually with fluidized-bed technique.
1. Single-stage operation
L = adsorbent Y = mass solute/mass solvent
V = solutions X = mass solute/mass adsorbent
Solute removed from the liquid equals that picked-up by the solid.
L0 (mass of adsorbent)
X0 (mass solute/mass adsorbent)
V0 V1
Y0 (mass solute/mass solvent) Y1
L1
X1
Figure 8 Material balance of single stage operation

Note:
o The amount of adsorbent used is ordinarily very small with respect to the amount of solution
treated, and since the solute to be removed is adsorbed mush more strongly than the other
constituents present, the adsorption of the latter can be ignored. Thus
V0  V1  V
L0  L1  L
VY0  Y1   LX1  X 0 
L Y0  Y1 mass of adsorbent
 
V X1  X 0 mass of solution
Equilibrium Curve
Y0 (X0, Y0) y* v.s. x
Operating Line
m = - L/V
Y1 (X1, Y1)
X0 X1
Figure 9 Operating and equilibrium lines
For small adsorbable solute concentration involved, the Freundlich equation is applied:
* n
Y  mX
At final equilibrium condition: n

1/n
Y 
X1   1 
m
Thus,
L Y0  Y1

V  Y1  1/n
 
m
where X0  0 for pure fresh adsorbent used
It is generally stated that the values of n in the range of 2 to 10 represent good, 1 to 2 moderately
difficult, and less than 1, poor adsorption characteristics (although the value of m is important also). In
the last case, impractically large adsorbent dosages may be required for appreciable fractional
removal of solute.
Exercise 2: Retrieved from Ref [2]

1. Fifty parts of copperas and lime per million parts of dychrome effluent removes 50 % of the
color. How much will be required to reduce the color to 5 % if n = 8.5?
2. Water contains organic colors which is to be extracted with alum and lime. Five parts of alum
and lime per million parts of water will reduce the color to 25 % of the original color and 10
parts will reduce the color to 3.5 %. Estimate how much alum and lime as parts per million are
required to reduce the color to 0.5 % of the original color.
3. Two identical aqueous solutions containing valuable solute are colored by small amounts of
an impurity to different extent. The dark solution has a color intensity of 50 units and the light
solution has 20 color units, measured on the same arbitrary scale. It is desired to treat an equal
amount of light and dark solutions to a color unit of 0.05. The equilibrium adsorption isotherm
on activated carbon is given by Freundlich equation Y = mX n, where n = 1 and m = 0.005 with
color units color units adsorbed
Y and X
lb solution lb carbon
The color scale is such that colors are additive; i.e., equal weights of solution color 20 and 50
will give a solution color (20 + 50)/2 = 35. In a single process, will it be more economical of
carbon to first blend the original solutions and to treat the blend, or to treat each separately
to color 0.5 and to blend the finished products?
2. Multi-Stage Cross-Current Operations

L1 L2
(X0)1 (X0)2
V Y1 Y2
Y0 1 2
X1 = (Y1/m)1/n X2 = (Y2/m)1/n
Figure 10 Material balance of a 2-staged operation

The removal of a given amount of solute can be accomplished with greater economy of adsorbent
if the solution is treated with separate small batches of adsorbent rather than in a single batch, with
filtration between each stage. This method of operation, sometimes called split-feed treatment, is
usually done in batch fashion, although continuous operation is also possible. Economy is particularly
important when activated carbon, a fairly expensive adsorbent, is used. The savings are greater the
larger the number of batches used but are at the expense of greater filtration and other handling
cost. It is therefore seldom economical to use more than 2 stages. In rare instances, the adsorption
may be irreversible, so that separate adsorbent dosages can be applied without intermediate
filtration at considerable savings in operating costs. This is by far the exception rather than the rule,
and when applied to ordinary reversible adsorption, it will provide the same end result as if all
adsorbent had been used in a single stage.
Stage 1 Balance:
L1 Y0  Y1 Y0  Y1 Y0  Y1
   (1)
X1  X 0 1
1/n 1/n
V  Y1   Y1 
   X 0 1  
m m
Stage 2 Balance:
L2 Y1  Y2 Y1  Y2 Y1  Y2
   (2)
X 2  X 0 2
1/n 1/n
V  Y2   Y2 
   X 0 2  
m m
Thus,
L total L1 L2 1/n  Y0  Y1 Y1  Y2 
  m  1/n  1/n 
V V V  Y1 Y2 
L total
m
1/n
Y Y
0 1
1/n
Y
1 1/n
 Y1Y2
1/n
 Y2
1 1/n

V
For minimum total adsorbent:
dL total  1 (1/n) 1  1 1/n 1/n  1/n

  Y0 Y1  1  Y1  Y2  0m 0
dY1  n  n 
1
 Y1  n
1 Y0 1
   1 (3)
 Y2  n Y1 n
Since m, n, Y0 and Y2 are constants, the equation above can be solved for intermediate concentration
Y1, and the adsorbed quantities can be calculated using equations (1) and (2).
Graphical representation:
Y0
m = -L1/V
Y1
m = -L2/V
Y2
X
X0 X2 X1
3. Multi-Stage Countercurrent Operation

 Better economy of adsorbent is obtained
X1 X2 X3 XN XN+1
L
V 1 2 N
Y0
Y1 Y2 Yn-1 YN
Note: Y0, n, Xn + 1 are known
Overall solute balance:
VY0  Yn   LX1  Xn1 

L Y0  Yn

V X1  Xn1
Consider a two-stage:
X1 X2 X3
L
V 1 2
Y0
Y1 Y2
Overall solute balance:
L Y  Y2 Y0  Y1
   0  1
 V overall X1  X 3  Y1  n
   X3
m
L 1  Y Y 
0 1
   m n  1/n 
 overall
V Y
 1  X 3
Stage 2 balance:
L Y  Y2 Y1  Y2
   1  1
 V  2 X1  X 3  Y2  n
   X3
m
1  Y Y 
1 2
 m n  1/n 
 Y2  X 3 
But X3  0 for fresh adsorbent
L L
    m
 V overall  V 2
Y0  Y2 Y1  Y2
1/n
 1/n
Y1 Y2
Exercise 3: Retrieved from Ref [1]

1. An aqueous solution containing a valuable solute is colored by small amounts of an impurity.
Before crystallization, the impurity is to be removed by adsorption on a decolorizing carbon
which adsorbs only insignificant amounts of the principal solute. A series of laboratory tests was
made by stirring various amounts of the adsorbent into batches of the original solution until
equilibrium was established, yielding the following data at constant temperature.
kg carbon/kg solution 0 0.001 0.004 0.008 0.02 0.04

Equilibrium color 9.6 8.6 6.3 4.3 1.7 0.7
The color intensity was measured on an arbitrary scale, proportional to the concentration of
the colored substance. It is desired to reduce the color to 10 % of its original value, 9.6.
Determine the quantity of fresh carbon required per 1000 kg of solution for a single-stage
operation, for a two-stage crosscurrent process using the minimum total amount of carbon,
and for a two-stage countercurrent operation.
2. In bleaching or removal of coloring substance of material such as coconut oil or sugar,

decolorizing agents such as activated carbon is used. Bleaching can be carried out in a single
pass or multi-pass or other methods. If 80 % of the color is removed in a single decolorization
with 4 % activated carbon, how much activated carbon will be needed to remove the same
amount of color in a 2-stage countercurrent decolorization if n = 2?
B. Continuous Contact
In these operations, the fluid and adsorbent are in contact throughout the entire apparatus, without
periodic separation of the phases. The operation can be carried out in strictly continuous, steady-state
fashion, characterized by movement of the solid as well as the fluid. Alternatively, owing to the rigidity
of the solid adsorbent particles, it is also possible to operate advantageously in semi-continuous
fashion characterized by a moving fluid but stationary solid. This results in unsteady-state conditions,
where compositions in the system change with time.
Steady-State Conditions
Steady-state conditions require continuous movement of both fluid and adsorbent through the
equipment at constant rate, with no change in composition at any point in the system with passage
of time. If parallel flow of solid and fluid is used, the net result is at best a condition of equilibrium
between the effluent streams, or the equivalent of one stage. It is the purpose of these applications
to develop separations equivalent to many stages, however, and hence only countercurrent
operations need be considered. There are applications especially in treating gases and liquids, for
purposes of collecting solute and fractionation, through ordinary adsorption and ion exchange.
One Component Adsorbed

For purposes of computation the operation can conveniently be considered analogous to gas
absorption, with solid absorbent replacing the liquid solvent. Refer to Figure 11, where GS and SS are
the solute-free fluid and mass velocities, respectively, mass/(cross-sectional area of column)(time).
Solute concentrations are expressed as mass solute/mass solute-free substance. The development is
shown for treating a gas but can be applied equally well to treating a liquid by replacing GS by LS. A
solute balance about the entire tower is
GS(Y1 – Y2) = SS(X1 – X2) (1)
and about the upper part
GS(Y – Y2) = SS(X – X2) (2)
These establish the operating line on X, Y coordinates, a straight line of slope

SS/GS joining the terminal conditions (X1, Y1) and (X2, Y2) (Fig. 11). The solute concentrations X and Y at
any level in the tower fall upon this line. An equilibrium-distribution curve appropriate to the system
and to the prevailing temperature and pressure can also be plotted on the figure as shown. This will
fall below the operating line for adsorption and above for desorption. In the same fashion as for
absorbers, the minimum solid/fluid ratio is given by the operating line of maximum slope which
anywhere touches the equilibrium curve.
SS GS
X2 Y2
SS GS Y1
X Y
Operating Line
Y Slope = SS/GS
dZ
Equilibrium curve
SS GS Y2 Y* vx. X
X+dX Y+dY
X2 X1
X
SS GS
X1 Y1
Figure 11 Continuous countercurrent adsorption of one component.
The resistance to mass transfer of solute from the fluid from the fluid to the adsorbed state on the solid
will include that residing in the gas surrounding the solid particles, that corresponding to the diffusion
of solute through the gas within the pores of the solid, and possibly an additional resistance at the time
of adsorption. During physical adsorption, the last of these will probably be negligible. If the remaining
resistances can be characterized by an overall gas mass-transfer coefficient KYap based on ap, the
outside surface of the solid particles, the rate of solute transfer over the differential height of adsorber
dZ (Fig. 1) can be written in the usual manner as
SSdX = GSdY = KYap(Y – Y*)dZ (3)
where Y* is the equilibrium composition in the gas corresponding to the adsorbate composition of X.
The driving force (Y – Y*) is then represented by the vertical distance between operating line and
equilibrium curve (Fig. 2). Rearranging Eq. (3) and integrating define the number of transfer units NtOG,
Y dY K Y ap Z Z
NtOG  Y21  0 dZ  (4)
* GS HtOG
YY
where
GS
HtOG 
K Y ap
The integral of Eq. (4) is ordinarily evaluated graphically and the active height Z determined through
knowledge of the height of a transfer unit HtOG, characteristic of the system.
Two Components Adsorbed

When several components of a gas are appreciably adsorbed, fractionation is required for their
separation.
For purposes of computation, it is easiest to recall the similarity between the adsorption operation and
continuous countercurrent extraction with reflux. Solid adsorbent as the added insoluble phase is
analogous to extraction solvent, the adsorbate is analogous to the solvent-free extract, and the fluid
stream is similar to the raffinate.
Figure 12 Continuous fractionation

Refer to Figure 12, a schematic representation of the adsorber. Feed enters at the rate of F
mass/(area)(time), containing components A and C. The adsorbate-free adsorbent enters the top of
the tower at the rate of B mass/(area)(time) and flows countercurrent to the gas entering at the
bottom at the rate of R1 mass/(area)(time). Compositions in the gas stream are expressed as x weight
fraction C, the more strongly adsorbed substance. E represents the weight of the adsorbent-free
adsorbate A + C upon the solid, mass/(area)(time), and N the ratio, mass adsorbent/mass adsorbate.
At any point in the adsorber, B  NE. The adsorbate composition is expressed as y weight fraction
component C, on an adsorbent-free basis. Solid leaves the adsorber at the bottom and is stripped of
its adsorbate in the desorption section, and the desorbed fluid is split into two streams, the reflux R1
and the C-rich product PE. At the top the fluid leaving is the A-rich product, R2 mass/(area)(time).
Calculations are made on a phase diagram of the type shown in Figure 13, with N plotted as ordinate,
x and y as abscissa.
Figure 13 Adsorption of acetylene-ethylene on silica gel at 250C, 1 std. atm

Let ∆𝐸 represent the net adsorbent-free flow downward and out of the adsorber. Then E = PE, and the
coordinates of point E on the phase diagram are (NE = B/PE, yE = x1). At the bottom of the adsorber,
B = N1E1 and
E1 = E' = PE + R1 = E + R1 (6)
An A-C balance below section e of the enriching section is
Ee = Pe + Re = E + Re (7)
and, for substance C,
Eeye = PEx1 + Rexe = Ex1 + Rexe (8)
while for adsorbent it becomes
NeEe = NEE = B1 (9)
E therefore represents the difference Ee – Re for any level in the enriching section. Equations (7) and
(8) provide a measure of the internal reflux ratio,
Re NΔE  Ne Ne
  1 (10)
Ee NΔE NΔE
and at the bottom of the tower the external reflux ratio is

R1 R1 NΔE  N1 NΔE
  1 (11)
PE E1  R1 N1 N1
At the top of the adsorber let R represent the difference between flows of adsorbate and unadsorbed
gas, and since E2 = 0, R= -R2. The coordinates of R on the phase diagram are then (NR = -B/R2, xR =
x2). Material balances above section s are then, for A and C,
RS = ES + R2 = ES – R (12)
for C,
RSxS = BSyS + R2x2 = ESyS - Rx2 (13)
and for adsorbent,
B = NSES = RNR (14)
Equations (12) and (14) provide the internal reflux ratio.
RS NS  NΔR NS
  1 (15)
ES  NΔR NΔR
Overall balances about the entire plant are
F = R2 + PE (16)
FxF = R2x2 + PEx1 (17)
The definitions of R and E, together with Eq. (17), provide
 R + F = E (18)
The graphical interpretation of these relations on the phase diagram is the same as that for the
corresponding situations in extraction.
If a stage-type device, e.g., fluidized beds, is used, the number of theoretical stages is readily obtained
by the usual step construction on the lower part of a diagram such as Fig. 4. For continuous contact,
z p dp x dx 1  r  1 x1
NtOG   p21  x21  ln (19)
HtOG p p *
xx *
1  r  1 x 2
where r = MA/MC and HtOG is defined by Eq. (5). This can be applied separately to the enriching and
adsorption sections. It assumes equimolar counter-diffusion of A and C, which is not strictly the case,
but refinement at this time is unwarranted in view of the sparsity of the data. It is also subject to the
same limitations considered for one component adsorbed (solute collection).
Sample Problem. Ref [3] Determine the number of transfer units and the adsorbent circulation rate
required to separate a gas containing 60% ethylene C2H4 and 40% propane C3H8 by volume into
products containing 5 and 95% C2H4 by volume, isothermally at 25℃ and 2.25 std atm, using activated
carbon as the adsorbent and a reflux ratio of twice the minimum. Note: These adsorbers will
customarily operate at reflux ratios closer to the minimum in order to reduce the adsorbent circulation
rate. The present value is used in order to make the graphical solution clear.
Solution:
Equilibrium data for this mixture at 25℃ and 2.25 std atm have been estimated from the data of Lewis
et al., Ind. Eng. Chem., 42, 1319, 1326 (1950), and are plotted in Fig. 4. C3H8 is the more strongly
adsorbed component, and compositions in the gas and adsorbate are expressed as weight fraction
C3H8. Tie lines have been omitted in the upper part of the plot, and the equilibrium compositions are
shown instead in the lower part.
Molecular weights are 28.0 for C2H4 and 44.1 for C3H8. The feed-gas composition is then
xF = 0.4(44.1)/[0.4(44.1) + 0.6(28.0)] = 0.512 wt fraction C3H8.
Similarly, x1 = 0.967, and x2 = 0.0763 wt fraction C3H8.
Basis: 100 kg feed gas. Eqs. (16) and (17):
100 = R2 + PE
100(0.512) = R2(0.0763) + PE(0.967)
Solving simultaneously gives R2 = 51.1 kg, PE = 48.9 kg.
Point F at xF and point E1 at x1 are located on the diagram as shown. From the diagram, N1 (at point
E1) = 4.57 kg carbon/kg adsorbate. The minimum reflux ratio is found as it is for extraction. In this case,
the tie line through point F locates (E)m and (E)m = 5.80.
 R1  5.80
    1  0.269 kg reflux/kg product
 PE m 4.57
(R1)m = 0.269 PE = 0.269(48.9) = 13.15 kg
(E1)m = (R1)m + PE = 13.15 + 48 9 = 62.1 kg

Figure 14 Solution to sample problem

Bm = N1(E1)m = 4.57(62.1) = 284 kg carbon/100 kg feed
At twice the minimum reflux ratio, R1/PE = 2(0.269) = 0.538, Eq. 11:
0.538 = NE/4.57 – 1
NE = 7.03 kg carbon/kg adsorbate
and point ∆𝐸 is located on the diagram.

R1 = 0.538PE = 0.538(48.9) = 26.3 kg
E1 = R1 + PE = 26.3 + 48.9 = 75.2 kg
B = N1E1 = 4.57(75.2) = 344 kg carbon/100 kg feed
Point R can be located by extending line EF to intersect the line x = x2. Alternatively, NR = -B/R2 = -
344/51.1 = -6.74 kg carbon/kg adsorbate. Random lines such as line RK are drawn from R, and the
intersections with the equilibrium curves are projected downward in the manner shown to provide the
adsorption-section operating curve. Similarly, random lines such as line EJ are drawn from E the
horizontal distance between operating and equilibrium curves on the lower diagram is the driving
force x – x* of Eq. (19). The following were determined from the diagram:
x x* 1/(x – x*) x x* 1/(x – x*)

x1 = 0.967 0.825 7.05 xF = 0.512 0.39 8.20
0.90 0.710 5.26 0.40 0.193 4.83
0.80 0.60 5.00 0.30 0.090 4.76
0.70 0.50 5.00 0.20 0.041 6.29
0.60 0.43 5.89 x2 = 0.0763 0.003 13.65
The third column was plotted as ordinate against the first as abscissa, the area under the curve
between x1 and xF was 2.65, and that between xF and x2 was 2.67. Further, r = 28.0/44.1 = 0.635.
Applying Eq. (19) to the enriching section gives,
1 0.635  10.967 
NtOG  2.65  ln  2.52
1 0.635  10.512 
and to the adsorption section,
1 0.635  10.512 
NtOG  2.65  ln  2.53
1 0.635  10.0763 
The total NtOG = 2.52 + 2.53 = 5.1.

References
[1] (1988). American Institute of Chemical Engineers (AIChE).
[2] Dozdi. (2010). Sample Problem #3. Retrieved from Scribd:
https://www.scribd.com/presentation/29901923/Sample-Problem-3
[3] Faust, S. D., & Aly, O. M. (1988). Chemistry of Water Treatment (2nd ed.). Boca Raton, Florida , USA:
Lewis Publishers.
[4] Seader, J. D., Henley, E. J., & Roper, D. K. (2016). Separation Process Principles with Applications
Using Process Simuators (4th ed.). (L. Ratts, Ed.) Hoboken, New Jersey, USA: John Wiley & Sons, Inc.
[5] Sundstrom, D. W., & Klei, H. E. (1979). Wastewater treatment. Prentice-Hall.
[6] Wankat, P. C. (2012). Separation Process Engineering (Includes Mass Transfer Analysis) (3rd ed.).
Westford, Massachusetts, USA: Pearson Education, Inc.

Separation Processes

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Separation Processes

Hochgeladen von

Copyright:

Verfügbare Formate

Lecture Notes 13 - Adsorption 1

Engr. Caesar P. Llapitan

Adsorption is a process whereby a substance (adsorbate, or sorbate) is accumulated on the surface

Figure 1 Sorption operations with solid-particle sorbents

Definition of Adsorption Terms

Examples of industrial applications of sorption operations are presented below:

7. High resistance to fouling for long life

Adsorbent Characteristics Commercial uses Strengths Weaknesses

Typical applications of commercial adsorbents:

Removal of colors from syrups

Criteria for Use

II. PRINCIPLES OF ADSORPTION

Figure 2 Flow of feed stream through a solid adsorbent

Figure 3 Bulk fluid phase

Figure 4 Flow of desorbent

Physical Adsorption v.s. Chemisorption

Adsorption is nearly always an exothermic process.

Factors Influencing Adsorption

(2) Nature of the Adsorbate

Adsorption of aliphatic series of organic acids increases in the order formic-acetic-propionic-

Studies on amphoteric compound indicate an adsorption maximum at the isoelectric point, or

f. Adsorption of Mixed Solutes

g. Nature of the Adsorbent

a) Pressure Swing Adsorption (PSA)

Uses of PSA processes

b) Temperature Swing Adsorption (TSA)

c) Displacement Purge Adsorption (DPA)

Method Advantages Disadvantages

 Desorbate may be recovered at  Unsuitable for rapid cycling so

III. ADSORPTION ISOTHERMS

A. Gas-Phase Adsorption Isotherms

Uses of Adsorption Isotherms

1. The Langmuir Isotherm

At high pressures, bp tends to infinity, and q = qm, i.e.  = 1.

2. The Freundlich Isotherm

With n = 1, the equation reduces to the linear form: q = k p

On linearization, the equation becomes:

3. The Brunauer-Emmett-Teller (BET) Isotherm

where ps is the saturated vapor pressure.

Gas Mixtures and Extended Isotherms

Other Adsorption Equilibria

B. Liquid-Phase Adsorption Isotherms

Flask No. Carbon Dosage (mg) Final TOC (mg/L)

1 0 1.27 1.27 0.00 0.000

Figure 6 Plot of Langmuir isotherm

Figure 7 Freundlich isotherm for Problem 2

So, the total carbon required is: 303.4 mg C + 147.4 mg C = 450.8 mg C

Volumetric flow rate = 1/residence time = 1/0.471 = 2.12 bed volumes/h

d) Q = 0.5 ML/d = 0.5 ML/d  1000 m3/ML  1 d/24 h = 20.83 m3/h

The height of GAC in each column has to be h = V/A (2) = 1.5 m

Assume backwash space 50% (typical value is 20 ~ 50%)

The height for each column then has to be 2.25m.

According to Table 14-2: K = 0.0925 (mg/g) (L/μg)1/n, i.e. 1/n = 0.67

Flask No Mass of C Volume in Final COD

Flask No Mass of Volume in Final COD

1 965 500 3.5

Ce , TOC in solution, qe , TOC on Carbon,

IV. ADSORPTION OPERATIONS

Equilibrium Relations for Adsorbents

Freundlich isotherm equation

Langmuir isotherm equation

Figure 8 Material balance of single stage operation

x x* 1/(x – x) x x 1/(x – x*)