Adsorption

1586_book.fm Page 1 Friday, May 7, 2004 3:56 PM

61 Adsorption

Shivaji Sircar Air Products and Chemicals, Inc.

61.1 61.2 61.3 61.4 61.5 61.6 61.7

Adsorbent Materials Adsorption Equilibria Heat of Adsorption Thermodynamic Selectivity of Adsorption Adsorption Kinetics Adsorption Column Dynamics Adsorptive Separation Processes and Design

Adsorption is a surface phenomenon. When a pure fluid (gas or liquid) is contacted with a solid surface (adsorbent), fluid–solid intermolecular forces of attraction cause some of the fluid molecules (adsorbates) to be concentrated at the surface. This creates a denser region of fluid molecules, which extends several molecular diameters near the surface (adsorbed phase). For a multicomponent fluid mixture, certain components of the mixture are preferentially concentrated (selectively adsorbed) at the surface due to differences in the fluid–solid forces of attraction between the components. This creation of an adsorbed phase having a composition different from that of the bulk fluid phase forms the basis of separation by adsorption technology. Adsorption is a thermodynamically spontaneous process. Energy is released (exothermic) during the process. The reverse process by which the adsorbed molecules are removed from the surface to the bulk fluid phase is called desorption. Energy must be supplied to the adsorbed phase (endothermic) for the desorption process. Both adsorption and desorption form vital steps in a practical separation process in which the adsorbent is repeatedly used. This concept of regenerative ad(de)sorption is key to the practical use of this technology. It has found numerous commercial applications in chemical, petrochemical, biochemical, and environmental industries for separation and purification of fluid mixtures, as listed in Table 61.1.

61.1 Adsorbent Materials A key element in the development of adsorption technology has been the availability of a large spectrum of micro- and mesoporous adsorbents. They have large specific surface areas (500 to 1500 m2/g), varying pore structures, and surface properties (polar and nonpolar) that are responsible for selective adsorption of specific components of a fluid mixture. These include activated carbons, zeolites, aluminas, silica gels, polymeric adsorbents, and ion-exchange resins. Adsorbents may be energetically homogeneous, containing adsorption sites of identical adsorption energy, or heterogeneous, containing a distribution of sites of varying adsorption energies.

© 2005 by CRC Press LLC


61-2

The Engineering Handbook, Second Edition

TABLE 61.1 Key Commercial Applications of Adsorption Technology Gas Separation

Liquid Separation

Gas drying Trace impurity removal Air separation Carbon dioxide-methane separation Carbon monoxidehydrogen separation Hydrogen and carbon dioxide recovery from SMR off gas Production of ammoniasynthesis gas Normal-iso paraffin separation Ozone enrichment Solvent vapor recovery

Environmental Separation

Liquid drying Trace impurity removal Olefin-paraffin separation Xylene, cresol, cymene isomer separation Fructose and glucose separation Breaking azeotropes

Municipal and industrial waste treatment Ground and surface water treatment VOC removal

Bioseparation Recovery of antibiotics Purification and recovery of enzymes Purification of proteins Removal of microorganisms Recovery of vitamins

61.2 Adsorption Equilibria Adsorption equilibria determine the thermodynamic limits of specific amounts of adsorption (moles/g) of a pure fluid or the components of a mixture under a given set of conditions [pressure (P), temperature (T), and mole fraction (y)] of the bulk fluid phase. A convenient way to represent adsorption equilibria is in terms of adsorption isotherms in which the specific amount adsorbed of a pure gas ( nio ) or that of component i(ni) from a multicomponent gas mixture is expressed as a function of P (pure gas) or as functions of P and yi (fluid mixtures) at constant T. The values nio and ni decrease with increasing T for a given P and yi. Adsorption isotherms can have many different shapes, but most microporous adsorbents exhibit the shape (type I) shown in Figure 61.1 (pure gas) and Figure 61.2 (binary gas). The simplest thermodynamic system for adsorption from liquids consists of a binary liquid mixture. The adsorption isotherm in this case is measured in terms of Gibbs surface excess (moles/g) of component i(nie ) : nie = ni - yi

Ân ; Ân

e i

i

i

= 0,

nie = 0 for pure liquid

i = 1, 2;

(61.1)

i

1.0

1.0

ni mmoles/g

ni mmoles/g

30°C N2

0.5 70°C

O2

0.5

EXPERIMENT THEORY 30°C 70°C

EXPERIMENT THEORY

0

1 p, atm

2

0

1.0 p, atm

2.0

FIGURE 61.1 Langmuirian pure gas adsorption isotherms on Na-mordenite. (Source: Kumar, R. and Sircar, S. 1986. Chem. Eng. Sci. 41:2215–2223.)



61-3

Adsorption

N2 (1) O2 (2) EXPERIMENT THEORY 1.0

0°C

ni, mmoles/g

P = 1.0 Bar

30°C 0.5 0°C 30°C

85°C 85°C

0

0.2

0.4

0.6

0.8

1.0

y1

FIGURE 61.2 Langmuirian binary gas adsorption isotherms on Na-mordenite. (Source: Kumar, R. and Sircar, S. 1986. Chem. Eng. Sci. 41:2215–2223.) 2.0 Experimental 0°C. 30°C. 60°C.

1.5

nei, mmoles/g

Theory

1.0

0.5

0

0

0.2

0.4 0.6 Mole fraction benzene

0.8

1.0

FIGURE 61.3 Benzene–cyclohexane liquid phase surface excess isotherms on silica gel. (Source: Sircar, S., Novosad, J., and Myers, A. L. 1972. Ind. Eng. Chem. Fundam. 11:249–254.)

The liquid phase surface excess is a function of yi and T. Figure 61.3 shows an example of bulk binary liquid phase isotherm. For adsorption of a very selective trace component (yi << 1), nie is approximately equal to ni. The isotherm in that case (Figure 61.4) is type I in shape. The simplest model to describe pure and multicomponent gas adsorption isotherms on a homogeneous adsorbent is the Langmuir equation,



61-4


3.0 pyridine from n -heptane

n1e, mmoles/g

Silica gel 2.0

1.0 Alumina 30°C

0

0.1 c1, moles/I

0.2

FIGURE 61.4 Surface excess isotherms for a trace component from liquids. (Source: Rahman, M. A. and Ghosh, A. K. 1980. J. Colloid Interface Sci. 77:50–52.)

Pure gas: nio =

mbi P ; 1 + bi P

Mixed gas: ni =

mbi Pyi 1 + Â i bi Pyi

(61.2)

where bi is the gas–solid interaction parameter for component i. The term bi is a function of temperature [Equation (61.7)] only. The term m is the saturation-specific adsorption capacity (moles/g) for the components. The Toth equation is often satisfactory for describing gas adsorption isotherms on heterogeneous adsorbents: Pure gas: nio =

mbi P ; [1 + (bi P)k ]1/k

Mixed gas: ni =

mbi Pyi [1 + (Â i bi Pyi )k ]1/k

(61.3)

The term k (£ 1) is the heterogeneity parameter. The lower the value of k is, the larger is the degree of adsorbent heterogeneity. The adsorbent is homogeneous when k is unity (Langmuir model). Statistical mechanics of adsorption dictates that the gas adsorption isotherms become linear functions of pressure (pure gas) and component partial pressures (mixed gas) when the total gas pressure approaches zero (P Æ 0): Pure gas: nio = K i P;

Mixed gas: ni = K i Pyi

(61.4)

Ki is called the Henry’s law constant for pure gas i. It is a function of temperature only [Equation (61.7)]. The simplest model isotherm for adsorption of an ideal binary liquid mixture of equal adsorbate sizes on a homogeneous adsorbent is given by n1e =

my1 y 2 (S - 1) = -n2e Sy1 + y 2

where S is the selectivity of adsorption of component 1 over component 2.


(61.5)


61-5

Adsorption

61.3 Heat of Adsorption The thermodynamic variable of practical use that describes the differential heat evolved (consumed) during the ad(de)sorption process due to a differential change in the adsorbate loading ( nio or ni) of a pure gas ( qio ) or the components of a gas mixture (qi) is called the isosteric heat of adsorption (kcal/ mole). It is given by È ∂ ln P ˘ Pure gas: qio = RT 2 Í ˙ ; Î ∂ T ˚ nio

È ∂ ln Pyi ˘ Mixed gas: qi = RT 2 Í ˙ , Î ∂ T ˚ ni

i = 1, 2,K

(61.6)

It follows from Equation (61.2), Equation (61.3), Equation (61.4), and Equation (61.6) that K i = K i* exp[qi* / RT ];

bi = bi* exp[qi* / RT ]

(61.7)

where qi* is the isosteric heat of adsorption of pure gas i at the limit P Æ 0 (Henry’s law region), and K i* and bi* are constants. The terms qio and qi are constants and equal to qi* for any adsorbate loading for a homogeneous adsorbent. They are functions of loadings for a heterogeneous adsorbent. Thus, for the Toth model, Ê RT 2 ˆ Ê d ln k ˆ 0 qi0 = qi* + Á ˜ Fi (qi ); ˜Á Ë k ¯Ë d T ¯

F(qio ) =

Ê RT 2 ˆ Ê d ln k ˆ qi = qi* + Á ˜ F(q) ˜Á Ë k ¯Ë d T ¯

[1 - (qio )k ]ln[1 - (qio )k ] + (qio )k ln(qio )k [1 - (qio )k ]

(61.8)

(61.9)

where (qio ) is fractional coverage (= nio /m) by pure gas i. The term q (= S qi) is the total fractional coverage (= S ni/m) by all adsorbates for the mixture. F(q) has the same mathematical form as Equation (61.9) except that qio is replaced by q. Equation (61.8) shows that qio or qi decreases with increasing qio or qi for a heterogeneous adsorbent. The higher energy sites of the adsorbent are predominantly filled at lower adsorbate loadings, and the lower energy sites are progressively filled at higher coverages. Figure 61.5 shows examples of isosteric heats of adsorption of pure gases on heterogeneous adsorbents. Figure 61.6 shows the variation of isosteric heats of the components of a binary gas mixture with coverage (or gas composition) at a constant total gas pressure according to the Toth model. An integral heat of adsorption called heat of immersion (kcal/mole) can be measured by contacting a clean adsorbent with a pure liquid (DH io ) or a liquid mixture (DH). The typical variation of (DH) as a function of bulk liquid phase concentration is shown in Figure 61.7. For an ideal binary liquid system [Equation (61.5)], DH is given by

DH =

Syi DH 1o + y 2 DH 20 Sy1 + y 2

(61.10)

61.4 Thermodynamic Selectivity of Adsorption Most practical adsorptive separations are based on thermodynamic selectivity. The selectivity Sij (= niyj/ njyi) of adsorption of component i over component j from a mixture determines the maximum achievable separation between the components under equilibrium conditions. Component i is selectively adsorbed



61-6


ne,µ MOLES/m2 2

6

10

14

18 qo or (qo−L) KCAL/MOLE

I II III IV V VI VII

1.1 V IV II

0.9

I

METHANE ON ACTIVATED CARBON 4.76 ETHANE ON ACTIVATED CARBON 7.68 BENENE ON SILICA GEL 5.61 CARBON DIOXIDE ON 13X ZEOLILE 11.02 ETHYLENE ON CARBON BLACK 1.51 METHANOL ON CARBON BLACK 0.77 BENZENE ON CARBON BLACK 1.99 VI

(q − L)/(qo− L) OR q/qo

III

I

0.7 III

II

VII

IV

0.5

VI

V

0.3

0.1

0

1

3

5

7

9

ne, MMOLES/GM

FIGURE 61.5 Isosteric heats of adsorption of pure gases on heterogeneous adsorbents. (Source: Sircar, S. and Gupta, R. 1981. Am. Inst. Chem. Eng. J. 27:806–812.)

over component j when Sij > 1. Sij can approach infinity if component j is excluded from entering the pores of the adsorbent (molecular sieving). The selectivity of adsorption (Sij* ) in the Henry’s law region (P Æ 0) is given by È qi* - q *j ˘ K i Ê K i* ˆ S = =Á ˙ ˜ exp Í K j Ë K *j ¯ ÍÎ RT ˙˚ * ij

(61.11)

Sij values can be strong functions of adsorbate loadings of the components when the adsorbates have different molecular sizes and when the adsorbent is heterogeneous, as shown in Figure 61.8. For adsorption of an ideal binary liquid system [Equation (61.5)], S = exp[f o2 - f1o ]/m RT

(61.12)

where fio is the surface potential for adsorption of pure liquid i. S for liquid mixtures can be strong functions of bulk phase concentrations when the adsorbates have different sizes or when the adsorbent is heterogeneous. © 2005 by CRC Press LLC


61-7

Adsorption

10 P = 99.2 (KPa) T = 293.1 K

9

BINARY

Heat of immersion (cal/g)

8 qi, Kcal/mole

26

PURE GAS AT SAME θi

C2H4(1)

7 6

CH4(2)

5 4 3

Decolorizing Charcoal

24 22 20 18

Coconut Shell Charcoal

16 benzene + cyclohexane mixtures at 30°C

14

0

0.2

0.4 y1

0.6

0.8

1.0

FIGURE 61.6 Isosteric heats of adsorption of binary gas on heterogeneous adsorbents. (Source: Sircar, S. 1991a. Langmuir 7:3065–3069.)

12

0

0.2

0.4 0.6 0.8 Mole fraction of benzene

1.0

FIGURE 61.7 Heats of immersion of binary liquid mixtures. (Source: Wright, E. H. M. 1967. Trans. Faraday Soc. 63:3026–3038.)

20 23.5% ethene and 76.5% methane

212.7 K

Selectivity

16

12

260.2 K

8

4

0

301.4 K

0

2

4

6

8

Total amount adsorbed (mmol g−1)

FIGURE 61.8 Binary selectivity of adsorption on activated carbon. (Source: Sircar, S. and Myers, A. L. 1985. Adsorp. Sci. Technol. 2:69–87.)

61.5 Adsorption Kinetics The actual physisorption process is very fast (milliseconds to reach equilibrium). However, a finite amount of time may be required for an adsorbate molecule to travel from the bulk fluid phase to the adsorption site in a microporous adsorbent. This rate process is generally referred to as adsorption kinetics. Adsorbate mass transfer resistance may be caused by: 1. Fluid film outside the adsorbent particle (for mixture adsorption) 2. Anisotropic skin at the particle surface 3. Internal macro- and microporous diffusional resistances (pore and surface diffusion) © 2005 by CRC Press LLC


61-8


1.0 water from cyclohexane

1.0 T = 20°C

0.8

0.8 Kinetic Uptakes on a CMS 0.6

0.6

y / yo

FRACTION OF EQUILIBRIUM

O2

N2

0.4

0.4

0.2

0.2 0

21°C

0

30 60 TIME, MINUTES

90

FIGURE 61.9 Pure gas adsorption kinetics on carbon molecular sieve. (Source: Sircar, S. 1994. In Separation Technology: The Next Ten Years, ed. J. Garside, Institute of Chemical Engineers, Rugby, Warwickshire, U.K. p. 49.)

0

2

4 6 Time, hours

8

10

12

FIGURE 61.10 Adsorption kinetics of a trace liquid adsorbate on 3A zeolite. (Source: Sircar, S. and Myers, A. L. 1986. Sep. Sci. Technol. 21:535–562.)

The transport of an adsorbate can be severely affected by: 1. The presence of other adsorbates in the pores 2. Local temperature changes caused by heat of ad(de)sorption Gas phase adsorption is generally faster than liquid phase adsorption. The overall mass transfer coefficient for adsorption can be increased by reducing the length of diffusional path (adsorbent particle size). Figure 61.9 shows examples of uptakes of pure gases by a carbon molecular sieve. Figure 61.10 shows an example of uptake of a trace adsorbate from a bulk liquid mixture by a zeolite. The terms y and yo are, respectively, bulk phase mole fractions at time t and at the start of the test. The simplest model for describing gas phase adsorption kinetics is known as the linear driving force (LDF) model: Pure gas: Mixed gas:

d nio (t ) = kio[nio *(t ) - nio (t )]; dt d ni (t ) = kii[ni*(t ) - ni (t )] + dt

(61.13)

Â k [n (t ) - n (t )] ij

* j

j

j

The terms nio (t ) and ni are, respectively, the transient adsorbate loading of pure component i, and that for component i from a mixture at time t. The terms ni*(t ) and nio (t) are, respectively, the corresponding equilibrium adsorbate loadings of component i at the instantaneous bulk phase conditions. The term kio (sec-1) is the overall mass transfer coefficient for pure component i. The terms kii and kjj are the overall straight and cross (between component i and j) mass transfer coefficients for component i in the mixture. The temperature and adsorbate loading dependence of the mass transfer coefficients is determined by the governing transport mechanism. The LDF model can also describe the kinetics of adsorption of liquid mixtures by rewriting Equation (61.13) in terms of transient surface excess nie (t ) and equilibrium surface excess nie* of component i. An experimental gas adsorption kinetics process is generally nonisothermal, whereas liquid phase adsorption kinetics can be measured isothermally due to high heat capacity of the liquid. Separation of the components of a fluid mixture can also be achieved by a kinetic selectivity when certain components of the mixture are adsorbed at a much faster rate than the others, even though there © 2005 by CRC Press LLC


61-9

Adsorption

V

IV

III

II

I

y* T*

Temperature

Mole fraction

yo

Flow direction 0

To

Distance in column

FIGURE 61.11 Type I column dynamics. (Source: Sircar, S. and Myers, A. L. 1985. Adsorp. Sci. Technol. 2:69–87.)

is no thermodynamic selectivity of adsorption between the components. Figure 61.9 is an example of such a case.

61.6 Adsorption Column Dynamics Practical separation and purification of fluid mixtures are carried out in packed columns of adsorbent materials. The dynamics of the ad(de)sorption process in columns is determined by the adsorption equilibria, heats, and kinetics and by the modes of operation of the process. The simplest case of the adsorption dynamics is to flow a binary gas mixture consisting of a single adsorbate (mole fraction yo) and an inert carrier gas through a packed column that has previously been saturated with the pure inert gas at the pressure (Po) and temperature (To) of the feed gas. Two types of behavior may be observed: Type I. Two pairs of mass and heat transfer zones are formed in the column, as shown in Figure 61.11. The column ahead (section I) of the front zones (section II) remains saturated with the carrier gas at initial conditions. The column (section III) between the front and rear zones (section IV) is equilibrated with a gas mixture of mole fraction y* (< yo) at temperature T* (>To). The column (section V) behind the rear zones is equilibrated with feed gas mixture at feed conditions. Type II. A pure heat transfer zone (section II) is formed, followed by a pair of mass and heat transfer zones (section IV), as shown in Figure 61.12. The adsorbate is absent in sections I to III in this case. The column behind (section V) the rear zones remains equilibrated with feed gas at feed conditions.

V

IV

III

II

I

T*

Temperature

Mole fraction

yo

Flow direction 0

Distance in column

To

FIGURE 61.12 Type II column dynamics. (Source: Sircar, S. and Myers, A. L. 1985. Adsorp. Sci. Technol. 2:69–87.)



61-10


Component 2

1.0

y/yo

Component 1

Flow direction

0

Distance in column

FIGURE 61.13 Rollover effect for binary adsorbate system. (Source: Sircar, S. and Myers, A. L. 1985. Adsorp. Sci. Technol. 2:69–87.)

Approximate criteria for formation of these two types of systems are as follows: Type I: n o / y o < C s /C g ;

Type II: n o / y o > C s /C g

(61.14)

The term no is adsorbate loading at feed conditions. Cs and Cg are heat capacities of the adsorbent and the feed gas. The zones move through the column as more feed gas is introduced. Adsorbates with type I adsorption equilibria generally yield constant pattern (unchanging shape and size) front zones (type I dynamics) and rear zones (type II dynamics) in a long column. The zones can be proportionate pattern (expanding in size with time) when the adsorption isotherms are linear. The zones eventually leave the column, and the measured adsorbate concentration–time and temperature–time profiles (breakthrough curves) are mirror images of these profiles within the column. Multiple transfer zones and equilibrium sections are formed in systems with multicomponent adsorbates. They are often characterized by rollover effects in which the more strongly adsorbed species (component 1) displaces the weaker species (component 2) as the zones propagate. This effect is illustrated in Figure 61.13 for isothermal adsorption of a binary adsorbate from inert carrier gas. Well-defined mass and heat transfer zones and equilibrium sections can also be formed in the column during the desorption process. Figure 61.14 is an example for isobaric desorption of a single adsorbate from a column saturated with the adsorbate at mole fraction yo and temperature To that is being purged with an inert gas at To. Again, two pairs of mass and heat transfer zones (sections II and IV) and three equilibrium sections (I, III, V) are formed. Nonisothermal column dynamics is generally a rule. Near-isothermal dynamics (mass transfer zones only) can be obtained for ad(de)sorption from (a) liquid mixtures and (b) trace gaseous adsorbates. Figure 61.15 shows two examples of isothermal breakthrough curves for adsorption of trace single adsorbate from (a) an inert gas and (b) an inert liquid. For isothermal–isobaric adsorption of a trace Langmuirian adsorbate i from an inert carrier gas in a column, which is initially free of the adsorbate, and where a constant pattern mass transfer zone is formed, the LDF mechanism of mass transfer yields

(t 2 - t 1 ) @

È f (1 - f1 ) ˘ f1 b1m ln Í 2 ˙ + ln o o f2 (1 + b1 )ki ni Î f1(1 - f 2 ) ˚

(61.15)

The terms t1 and t2 correspond to time difference in the breakthrough curve corresponding to two arbitrary composition levels f1 and f2 [f = yi(t)/ yio ]. © 2005 by CRC Press LLC


61-11

Adsorption

V

IV

III

II

I

Mole fraction

Flow direction

To

Temperature

yo

T*

y* 0

Distance in column

FIGURE 61.14 Column dynamics for desorption by purge. (Source: Sircar, S. and Myers, A. L. 1985. Adsorp. Sci. Technol. 2:69–87. 1.0 1.0 0.8

0.8

ethane from helium

dilute p -chlorophenol from water

0.6

φ

φ

0.6

0.4

0.4

ts = 82.8 min

0.2 0 −20

−10 0 (t − t s ) (a)

10

20

0.2 0

0

5

10

15 20 Time, hours

25

30

35

(b)

FIGURE 61.15 Isothermal breakthrough profiles for a single-trace adsorbate. [Sources: (a) Garg, D. R. and Ruthven, D. M. 1974. Chem. Eng. Sci. 29:571–581. (b) Mathews, A. P. 1984. In Proceedings of First International Conference on Fundamentals of Adsorption, ed. A. L. Myers, Engineering Foundation, New York. p. 345.]

61.7 Adsorptive Separation Processes and Design Three generic adsorptive process schemes have been developed to serve most of the applications described by Table 61.1. These are (a) thermal swing adsorption (TSA), (b) pressure swing adsorption (PSA), and (c) concentration swing adsorption (CSA). TSA is by far the most frequently used industrial application of this technology. It is used for removal of trace impurities as well as for drying gases and liquids. The adsorption is carried out at near ambient temperature and the desorption is achieved by directly heating the adsorbent using a part of the cleaned fluid or steam. The adsorbent is then cooled and reused. PSA is primarily used for bulk gas separations and for gas drying. These processes consist of a series of sequential cyclic steps. The adsorption step is carried out at higher partial pressures of the adsorbates and the desorption is achieved by lowering their partial pressures in the column by (a) decreasing the total pressure of the column and (b) flowing an adsorbate free gas through the column. Many complementary process steps are also used in modern PSA processes in order to: 1. Increase the recovery and purity of desired products from a multicomponent feed gas mixture © 2005 by CRC Press LLC


61-12


A O2 ENRICHED GAS

D ADSORBER

DESORBED GAS

DISTILLATION

RV AC

A/D

D

FEED B B/D DISTILLATION

FEED AIR COMPRESSOR

FIGURE 61.16 Schematic flowsheet of a PSA process. (Source: Sircar, S. 1989. In Adsorption Science and Technology, ed. A. I. Rodrigues et al. NATO ASI Series E158. Kluwer Academic, Dordrecht, The Netherlands. p. 285.)

A+B FEED

FIGURE 61.17 Schematic flowsheet of an SMB process. (Source: Keller, G. E., Anderson, R. A., and Yon, C. M. 1987. In Handbook of Separation Process Technology, ed. R. W. Rousseau, John Wiley & Sons, New York. p. 644.)

2. Produce more than one pure product 3. Decrease adsorbent inventory 4. Reduce energy of separation Figure 61.16 shows a schematic flowsheet for a PSA process. The CSA processes are generally designed for separation of bulk liquid mixtures. Certain components of a liquid mixture are adsorbed at ambient conditions, and the desorption is effected by flowing a less strongly adsorbed liquid (eluent) over the adsorbent. Simple distillation may be necessary to separate the eluent from the components of the feed mixture. Simulated moving bed (SMB) adsorbers have been designed for this purpose, as shown schematically in Figure 61.17. The feed and eluent injection points as well as the liquid withdrawal points are changed periodically in a fixed column to simulate continuous countercurrent operation. The design of adsorptive processes requires simultaneous solutions of differential mass, heat, and momentum balance equations describing the operations of the cyclic process steps in the adsorbent column using the appropriate initial and boundary conditions for each step. The final column conditions at the end of a step become the initial conditions for the next step. Numerical integration of the equations is often necessary in order to reach a cyclic steady state solution. Multicomponent adsorption equilibria, heats, and kinetics from the key input variables for the solution. Bench- and pilot-scale process performance data are generally needed to confirm design calculations.

Defining Terms Adsorbent — A material used for carrying out the adsorption process. Adsorption — The surface phenomenon by which the molecules of a bulk fluid phase are attracted by a solid surface in contact with the fluid. Adsorption equilibria — The thermodynamic property describing the extent of adsorption of a fluid species by a solid surface. Adsorption kinetics — The measure of travel time of an adsorbate molecule from bulk fluid phase to the adsorption site. Column dynamics — Defines behavior of mass and heat transfer zone movements within an adsorption column during the ad(de)sorption process. Concentration swing adsorption — An adsorptive process in which desorption is effected by changing the fluid phase concentration of the adsorbate. Desorption — The process of removing the adsorbed molecules from the solid surface to the bulk fluid phase. © 2005 by CRC Press LLC


Adsorption

61-13

Heat of adsorption — The measure of thermal energy released during the exothermic adsorption process. Pressure swing adsorption — An adsorptive process in which desorption is effected by lowering the partial pressure of the adsorbate. Selectivity of adsorption — The measure of extent of separation of a component of a fluid mixture by adsorption process. Thermal swing adsorption — An adsorptive process in which desorption is effected by heating the adsorbent.

References Basmadjian, D. 1997. The Little Adsorption Book, CRC Press, Boca Raton, FL. Broughton, D. B. and Gembicki, S. A. 1984. Adsorptive separations by simulated moving bed technology. In Fundamentals of Adsorption, ed. A. L. Myers and G. Belfort, Engineering Foundation, New York. p. 115. Crittenden, B. and Thomas, W. J. 1998. Adsorption Technology and Design, Butterworth-Heinemann, Oxford. Do, D. D. 1998. Adsorption Analysis: Equilibria and Kinetics, Imperial College Press, London. Garg, D. R. and Ruthven, D. M. 1974. The performance of molecular sieve adsorption columns: system with micropore diffusion control. Chem. Eng. Sci. 29:571–581. Keller, G. E., Anderson, R. A., and Yon, C. M. 1987. Adsorption. In Handbook of Separation Process Technology, ed. R. W. Rousseau, John Wiley & Sons, New York. p. 644. Kumar, R. and Sircar, S. 1986. Skin resistance for adsorbate mass transfer into extruded adsorbent pellets. Chem. Eng. Sci. 41:2215–2223. Mathews, A. P. 1984. Dynamics of adsorption in a fixed bed of polydisperse particles. In Proceedings of First International Conference on Fundamentals of Adsorption, ed. A. L. Myers, Engineering Foundation, New York. p. 345. Rahman, M. A. and Ghosh, A. K. 1980. Determination of specific surface area of ferric oxide, alumina and silica gel powders. J. Colloid Interface Sci. 77:50–52. Rouquerol, F., Rouquerol, J., and Sing, K.S.W. 1999. Adsorption by Powders and Porous Solids, Academic Press, London. Ruthven, D. M. 1984. Principles of Adsorption and Adsorption Processes, John Wiley & Sons, New York. Ruthven, D. M., Farouq, S., and Knaebel, K. S. 1994. Pressure Swing Adsorption, VCH, New York. Sircar, S. 1989. Pressure swing adsorption technology. In Adsorption Science and Technology, ed. A. I. Rodrigues et al. NATO ASI Series E158. Kluwer Academic, Dordrecht, The Netherlands. p. 285. Sircar, S. 1991a. Isosteric heats of multicomponent gas adsorption on heterogeneous absorbents. Langmuir 7:3065–3069. Sircar, S. 1991b. Pressure swing adsorption — research needs by industry. In Proceedings of Third International Conference on Fundamentals of Adsorption, ed. A. Mersmann, Engineering Foundation, New York. p. 815. Sircar, S. 1993. Novel applications of adsorption technology. In Proceedings of Fourth International Conference on Fundamentals of Adsorption, ed. M. Suzuki, Kodansha, Tokyo. p. 3. Sircar, S. 1994. Adsorption technology: A versatile separation tool. In Separation Technology: The Next Ten Years, ed. J. Garside, Institute of Chemical Engineers, Rugby, Warwickshire, U.K. p. 49. Sircar, S. and Gupta, R. 1981. A semi empirical adsorption equation for single component gas solid equilibrium. Am. Inst. Chem. Eng. J. 27:806–812. Sircar, S. and Myers, A. L. 1985. Gas adsorption operations: equilibrium, kinetics, column dynamics and design. Adsorp. Sci. Technol. 2:69–87. Sircar, S. and Myers, A. L. 1986. Liquid adsorption operations: equilibrium, kinetics, column dynamics, and applications. Sep. Sci. Technol. 21:535–562. Sircar, S., Novosad, J., and Myers, A. L. 1972. Adsorption from liquid mixtures on solids: thermodynamics of excess properties and their temperature coefficients. Ind. Eng. Chem. Fundam. 11:249–254. © 2005 by CRC Press LLC


61-14


Wright, E. H. M. 1967. Thermodynamic correlation for adsorption from non-ideal solutions at the solidsolution interface. Trans. Faraday Soc. 63:3026–3038. Yang, R. T. 1987. Gas Separation by Adsorption Processes, Butterworths, London. Young, D. M. and Crowell, A. D. 1962. Physical Adsorption of Gases, Butterworths, London.

Further Information Barrer, R. M. 1978. Zeolites and Clay Minerals as Sorbents and Molecular Sieves, Academic Press, New York. Breck, D. W. 1974. Zeolite Molecular Sieves, Wiley-Interscience, New York. Capelle, A. and deVooys, F. (Eds.) 1983. Activated Carbon — A Fascinating Material, Norit N. V., Amersfoort, The Netherlands. Gregg, S. J. and Sing, K. S. W. 1982. Adsorption Surface Area and Porosity, Academic Press, London. Karger, J. and Ruthven, D. M. 1992. Diffusion in Zeolites, Wiley-Interscience, New York. Rousseau, R. W. (Ed.) 1981. Handbook of Separation Technology, John Wiley & Sons, New York. Suzuki, M. 1990. Adsorption Engineering, Kodansha, Tokyo. Wankat, P. C. 1986. Large Scale Adsorption and Chromatography, CRC Press, Boca Raton, FL.


Adsorption

Recommend Documents