c 2011 Ismail Tosun

Chapter 10 Solubility of Gases in Liquids

In some separation processes, we are mainly interested in the transfer of a component from the gas phase to the liquid phase. For example, stringent environmental protection laws in many countries require the concentrations of pollutants in the industrial emission gases (CO2 , H2 S, SO2 , NOx ) to be kept below specied limits. For this purpose gas absorbers, scrubbers, and wetted wall columns are used to bring the gas stream into contact with liquid solvent so as to transfer the undesired component from the gas to the liquid. A typical gas absorber is shown in Figure 10.1. Within the column, the contact between the gas and liquid streams is achieved either on trays, as in the case of distillation, or in a packed bed. The purpose of using a packed bed is to enhance the mass transfer between the phases by increasing the interfacial area. While the gas stream to be cleaned enters from the bottom of the column, the liquid solvent is fed from the top of the column and is distributed over the packing by a distribution plate (or a spray nozzle). Absorption in the liquid phase takes place either by physical absorption or by a chemical reaction. The euent gas leaves the column from the top. The liquid stream leaving the absorber is rst sent to a stripper to recover the dissolved solute and then recycled back to the absorption column.
Gas out

Liquid in

Gas in

Liquid out

Figure 10.1 A packed gas absorber. In recent years, hollow ber membrane contactors have found wide application in the biological treatment of gas streams containing volatile organic compounds (VOCs), and also in the removal of acid gas1 from waste gas streams. These devices, in a way, resemble shell-andtube heat exchangers. The membrane wall not only keeps liquid and gas phases separated from each other but also provides an interfacial area for the two phases. The gas stream may ow either in the tube or shell side, in cocurrent or countercurrent ow with the liquid stream. For more detail about membrane contactors, see Gabelman and Hwang (1999) and Sirkar (1997).
1

An acid gas is a gas containing signicant amounts of acidic gases, such as CO2 and H2 S.

357

10.1 HENRYS LAW The solubility of a gas in a liquid is determined by the equations of phase equilibrium. Consider the gaseous and liquid phases of a binary system that are in equilibrium. The condition of equilibrium states that bL (T, P, xi ) bV (T, P, yi ) = f i = 1, 2 (10.1-1) f i i bV (T, P, yi ) = (T, P, xi ) xi f L (T, P ) yi P i i i (10.1-2)

Expressing the left- and right-hand sides of Eq. (10.1-1) in terms of fugacity coecient and activity coecient, respectively, yields

The diculty in using Eq. (10.1-2) comes from the fact that if the temperature of the mixture, T , is greater than the critical temperature of component i, then the pure component i exists only as a gas and it is impossible to calculate fiL (T, P ). For example, consider a rigid cylinder containing water (1) and methane (2) in equilibrium at 298 K and 1 bar. Since the critical temperature of methane is 190.6 K, methane does not exist in liquid form at the temperature L (298 K, 1 bar) cannot be determined. of the mixture and f2 To circumvent this problem, Henrys law is dened as bL (T, P, xi ) = Hi (T, P ) xi f i when xi 0 (10.1-3)

where Hi is called the Henrys law constant and has the units of pressure. Henrys law "constant" is not constant but rather dependent on temperature and pressure2 . It should be kept in mind that Eq. (10.1-3) is valid not only when xi 0, but also when the species entering the liquid phase do not react or dissociate. The mole fraction of species i in the liquid phase, xi , is generally termed as the solubility of species i. Substitution of Eq. (10.1-3) into Eq. (10.1-1) gives bV (T, P, yi ) = Hi (T, P ) xi yi P i (10.1-4)

indicating that the amount of gas dissolved in a solvent is directly proportional to the partial pressure of the gas above the solvent. An extensive list for the Henrys law constants is presented by Sander (1999). Table 10.1 gives the Henrys law constants for various gases in water. Table 10.1 Henrys law constants for common gases in water at 298 K. Gas N2 CO O2 CH4 CO2 H2 S SO2 NH3
2

At low pressures, the gas phase may be considered an ideal gas mixture and Eq. (10.1-4) becomes yi P = Hi xi P is low and xi 1 (10.1-5)

85.1 103 58.2 103 42.5 103 39.5 103 1.63 103 0.553 103 0.461 102 0.988

Hi ( atm)

While Hi is a strong function of temperature, its dependence on pressure is generally weak. Dependence of Hi on temperature and pressure will be discussed in detail in Section 10.2.

358

Example 10.1 Calculate the number of moles of oxygen dissolved in 1 L of water under atmospheric pressure and 298 K. Solution Air contains approximately 21 mol % O2 . Assuming that a liter of solution contains practically 1000 g of water, Eq. (10.1-5) is written as nO 2 3 (0.21)(1) = (42.5 10 ) (1) nO 2 + (1000/18) where nO 2 is the number of moles of O2 per liter of solution. The denominator of Eq. (1) can be approximated as 1000 1000 ' nO 2 + 18 18 so that the number of moles of O2 dissolved, nO 2 , is given by nO 2 = (0.21)(1000/18) = 2.75 104 mol/L 42.5 103

Comment: Water quality is frequently measured by dissolved oxygen amount.

Note that the Henrys law constant is dened by ! bL f i Hi = lim xi 0 xi

(10.1-6)

bL versus xi curve at xi = 0 gives the Henrys law Therefore, the slope of the tangent to the f i constant as shown in Figure 10.2.
L f 1 df L H1 = 1 dx1 x1 =0

indicating that Hi is simply a proportionality constant between the fugacity of component i and its mole fraction in the liquid phase in the limit of innite dilution. Since both the numerator and the denominator of Eq. (10.1-6) go to zero when xi goes to zero, the application of Hopitals rule gives ! bL df i (10.1-7) Hi = dxi
xi =0

L f1 = H1x1

L f1

x1

bL as a function of composition. Figure 10.2 For a binary system, variation of f 1 359

In the literature, Henrys law is expressed in several dierent forms. Besides Eq. (10.1-5), one of the most frequently used forms is
ci = Hi P yi

(10.1-8)

has the units of ( mol/L)/ atm or M/ atm, where M stands for The Henrys law constant Hi 3 molarity . The conversion between dierent units of the Henrys law constant is given as (M/ atm) = 101.325 H ( mol/ m3 )/ Pa Hi i (10.1-9) 55.3 (M/ atm) = Hi Hi ( atm)

Sander (1999) also provided a calculator for conversions between dierent units of Henrys law constant. 10.1.1 Relation Between Henrys Law and the Lewis-Randall Rule For a binary system of components 1 and 2 at constant temperature and pressure, the use of the Gibbs-Duhem equation, Eq. (6.4-9), gives dG1 dG2 x1 + x2 =0 dx1 dx1 or dG1 dG2 x2 =0 x1 dx1 dx2 The partial molar Gibbs energies are expressed by Eq. (7.3-6), i.e.,
L Gi L L L L

(10.1-10)

(10.1-11)

Substitution of Eq. (10.1-12) into Eq. (10.1-11) and rearrangement lead to ! ! bL bL d f df 1 2 dx1 dx2 ! = ! bL bL f f 1 2 x1 x2 Taking the limit of Eq. (10.1-13) as x2 0 (or x1 1) gives ! ! bL bL df d f 1 2 dx1 dx2 lim ! = lim ! x1 1 x2 0 bL bL f f
1 2

bL = i (T ) + RT ln f i

bL dGi RT df i = L dxi b dxi f i

(10.1-12)

(10.1-13)

(10.1-14)

x1

x2

Let us assume that component 2 obeys Henrys law, i.e., x1 x2 . Noting that L! L! L! b b b f df f 1 2 2 L lim lim lim = f1 = H2 = H2 x1 1 x2 0 x2 0 x1 dx2 x2
3 The ocial SI unit for Hi is ( mol/ m3 )/ Pa.

(10.1-15)

360

Eq. (10.1-14) reduces to

bL df 1 dx1

L = f1

(10.1-16)

x1 =1

bL = f L x1 . Hence, whenever Henrys which is nothing more than the Lewis-Randall rule, i.e., f 1 1 law applies to component 2 (x2 is very close to zero), the Lewis-Randall rule is applicable to component 1 (x1 is close to unity). Therefore, the Lewis-Randall rule and Henrys law are not independent, but related to each other by the Gibbs-Duhem equation. In a similar way, it can be shown that when Henrys law is applicable to component 1 (x1 is very close to zero), the Lewis-Randall rule applies to component 2 (x2 is close to unity). Figure 10.3 shows the regions of applicability of the Lewis-Randall rule and Henrys law for components 1 and 2.
H1
Region over which the Lewis-Randall rule applies to Component 1 and Henrys law applies to Component 2 Region over which the Lewis-Randall rule applies to Component 2 and Henrys law applies to Component 1

H2

L f 1 f
L 1

L f 2

f 2L

x2

Figure 10.3 Representative variation of fugacities in a binary mixture of 1 and 2 as a function of composition. Consider an equilibrium between the vapor and liquid phases of a binary system of components 1 and 2. Let us assume that x1 x2 so that Henrys law is applicable to component 2. This automatically implies that the Lewis-Randall rule applies to component 1. Therefore, the condition of equilibrium is expressed as " L # e (P P vap ) V V 1 1 vap vap b = f L x1 = P V (T, P ) exp (10.1-17) y1 P x1 1 1 1 1 1 RT Addition of Eqs. (10.1-17) and (10.1-18) gives h i vap V vap e L (P P vap )/RT 1 (T, P1 ) exp V P1 1 1 H2 x1 + x2 P = V V b b 1 2 b = H2 x2 y2 P 2
V

(10.1-18)

(10.1-19)

At low pressures, the Lewis-Randall rule leads to Raoults law. In this case, Eq. (10.1-19) simplies to vap x1 + H2 x2 (10.1-20) P = P1 361

Example 10.2 The solubility of carbon dioxide (2) in water (1) follows Henrys law and the Henrys law constant at 323 K was reported as 2939 bar (Dalmolin et al., 2006). The vapor is represented by the virial equation of state with the following virial coecients B11 = 834.2 cm3 / mol B22 = 104.2 cm3 / mol B12 = 153.4 cm3 / mol

Estimate the solubility of carbon dioxide in water at 323 K and 3.65 bar. The vapor pressure of water at 323 K is 0.123 bar. Solution Since Henrys law applies to carbon dioxide, then the Lewis-Randall rule is applicable to water and Eqs. (10.1-17) and (10.1-18) can be used for this system. Fugacity coecients of water and carbon dioxide are calculated from Eq. (7.5-6) as i V P h 2 b B11 + y2 (2 B12 B11 B22 ) 1 = exp RT n h io 3.65 2 834.2 + y2 2( 153.4) + 834.2 + 104.2 = exp (83.14)(323) 2 = exp 0.1134 + 0.0858 y2 (1) bV = exp 2 i P h 2 B22 + y1 (2 B12 B11 B22 ) RT h io n 3.65 2 = exp 2( 153.4) + 834.2 + 104.2 104.2 + y1 (83.14)(323) 2 = exp 0.0142 + 0.0858 y1

(2)

Fugacity coecient of pure water under saturation condition is calculated from Eq. (5.3-5) as vap ! B P 11 1 vap V 1 (T, P1 ) = exp RT ( 834.2)(0.123) = 0.996 (3) = exp (83.14)(323) e L = 18 cm3 / mol, Eqs. (10.1-17) and (10.1-18) are expressed as Taking V 1 (18)(3.65 0.123) 2 y1 (3.65) exp 0.1134 + 0.0858 y2 = (0.123)(0.996) exp x1 (83.14)(323) 2 = (2939) x2 y2 (3.65) exp 0.0142 + 0.0858 y1

(4) (5)

The solution of Eqs. (4) and (5) requires the following iterative procedure :

1. Assume y1 (this will automatically x the value of y2 ), 2. Solve Eq. (4) for x1 , 3. Solve Eq. (5) for x2 , 4. Check the summation of mole fractions in the liquid phase. The assumed y1 value is correct when x1 + x2 = 1. The results are given in the following table :

362

y1 0.0300 0.0320 0.0370 0.0347

x1 0.8630 0.9203 1.0632 0.9975

x2 103 1.1878 1.1853 1.1792 1.1820

x1 + x2 0.8642 0.9215 1.0644 0.9987

Thus, the solubility of carbon dioxide in water at 323 K and 3.65 bar is x2 = 1.182 103 . The experimental value reported by Dalmolin et al. (2006) is x2 = 1.28 103 .

Example 10.3 The following data are provided at 353 K for a binary system of components 1 and 2 vap vap = 0.8 bar P2 = 1.5 bar P1 H1 = 1.53 bar Solution Since x1 is almost equal to unity and pressure is low, it is possible to apply Raoults law to component 1 and Henrys law to component 2, i.e.,
vap x1 y1 P = P1

H2 = 7.38 bar

Calculate the pressure and the composition of the vapor when x1 = 0.99.

(1) (2)

y2 P = H2 x2 Addition of Eqs. (1) and (2) gives the total pressure as

vap P = P1 x1 + H2 x2 = (0.8)(0.99) + (7.38)(0.01) = 0.866 bar

The composition of the vapor phase is calculated from either Eq. (1) or Eq. (2), i.e., y1 = y2 =
vap P1 x1

(0.8)(0.99) = 0.915 0.866

H2 x2 (7.38)(0.01) = = 0.085 P 0.866

10.1.2 Relationship Between Henrys Law Constant and Innite Dilution Activity Coecient From Eq. (8.3-5) bL f i xi = fiL (T, P ) i (T, P, xi ) (10.1-21)

Substitution of Eq. (10.1-21) into Eq. (10.1-6) yields h i Hi (T, P ) = lim fiL (T, P ) i (T, P, xi )
xi 0

(10.1-22)

363

From Eq. (5.4-8) lim h fiL (T, P ) i = lim (

vap ) exp Pivap V i (T, Pi

xi 0

xi 0

=
L

vap ) exp Pivap V i (T, Pi

"

"

Vi

(P Pivap ) RT

e L (P P vap ) V i i RT #

#) (10.1-23)

where V i is the partial molar volume of component i in the liquid phase at innite dilution. On the other hand, i h lim i (T, P, xi ) = (10.1-24) i (T, P ) Substitution of Eqs. (10.1-23) and (10.1-24) into Eq. (10.1-22) leads to " L # vap V ( P P ) i i vap vap Hi = V ) exp i (T, Pi i Pi RT
xi 0

(10.1-25)

At high pressures, the Poynting correction factor cannot be neglected and the estimation of Hi L L requires the value of V i to be known. In the absence of an experimental value, V i can be e L , at its normal boiling point. approximated by the molar volume of pure i, V i Example 10.4 The following data are provided by Hildebrand and Scott (1950) for the partial molar volumes of various gases in solvents at 298 K. V Solvent Acetone Benzene Carbon tetrachloride Water N2 55 53 53 40 O2 48 46 45 31
L

( cm3 / mol) CO2 33 SO2 68 48 54 CH4 55 52 52 37

CO 53 52 53 36

The normal boiling points (NBP) of these components are given as follows: Species NBP ( K) N2 77.4 O2 90.2 CO 81.6 CO2 194.7 SO2 263.1 CH4 111.7

Calculate the molar volume of pure liquids at their normal boiling points and check the validity L e L. of approximating V i by V i Solution From Appendix A Component Nitrogen Oxygen Carbon monoxide Carbon dioxide Sulfur dioxide Methane Tc ( K) 126.2 154.6 134.5 304.2 430.3 190.6 364 Pc ( bar) 34.0 50.4 35.0 73.8 78.8 46.1 0.039 0.025 0.066 0.239 0.256 0.011

The molar volumes of pure liquids are calculated by using the Rackett equation, Eq. (5.4-9), and are presented below : Species e L ( cm3 / mol) V N2 34.2 O2 27.9 CO 34.8 CO2 34.7 SO2 44.6 CH4 38.0

e L values underestimate the values of V L Note that V i . i

Example 10.5 A cylinder of propane (2) at 60 bar and 303 K contains some liquid water (1). If propane is sparingly soluble in water under these conditions, estimate the Henrys law constant. The activity coecient of propane follows the NRTL equation with 12 = 1.156, 21 = 1.214, and = 0.3. Propane vapor is represented by the Peng-Robinson equation of state and the vapor pressure of propane at 303 K is 10.84 bar. Solution From Appendix A Tc = 369.9 K Assuming
L V2

Pc = 42.5 bar "

= 0.153 #

H2 =

e L , Eq. (10.1-25) takes the form 'V 2

vap V 2 (303 K, 10.84 bar) exp 2 P2

(8.314 105 )(303)

e L (60 10.84) V 2

(1)

The use of Eq. (5.3-12) gives the fugacity coecient of pure propane gas as V 2 (303 K, 10.84 bar) = 0.834

From Eq. (8.4-32) h i n h io = exp + exp( ) = exp 1 . 156 + 1 . 214 exp (0 . 3)(1 . 214) = 7.385 12 21 21 2

The molar volume of liquid propane can be calculated from the modied Rackett equation, Eq. (5.4-9), 2/7 e2L = RTc Z 1+(1Tr ) (2) V RA Pc where the Rackett parameter, ZRA , is given by Eq. (5.4-10), i.e., ZRA = 0.29056 0.08775 The reduced temperature is Tr = and the Rackett parameter is calculated as ZRA = 0.29056 (0.08775)(0.153) = 0.2771 303 = 0.819 369.9 (3)

365

Therefore, Eq. (2) gives the molar volume as

5 e2L = (8.314 10 )(369.9) (0.2771)1+(10.819)2/7 = 9.12 105 m3 / mol V 42.5

Substitution of the numerical values into Eq. (1) yields " # 9.12 105 (60 10.84) H2 = (7.385)(10.84)(0.834) exp = 79.8 bar (8.314 105 )(303)

At low and moderate pressures, both the fugacity coecient of pure i and the Poynting correction factor are almost unity, and Eq. (10.1-25) simplies to
vap Hi = i Pi

(10.1-26)

Example 10.6 The Henrys law constant for benzene (2) in water (1) is experimentally determined at 296 K as 271 bar. Estimate the innite dilution activity coecient for benzene in a binary liquid mixture of benzene and water. Solution From Appendix C 2788.51 T 52.36 = 0.115 bar. Therefore, from Eq. (10.1-26)
vap = 9.2806 ln P2

vap At 296 K, P2

2 =

vap P2

H2

271 = 2357 0.115

Comment: Such a high value for the innite dilution activity coecient indicates that benzene is practically insoluble in water. In other words, benzene and water molecules repel each other.

10.1.3 Henrys Law and Activity Coecients Consider a binary mixture4 of solvent (1) and solute (2). As long as the solute concentration is dilute, i.e., x2 x1 , Henrys law is applicable to the solute and the liquid phase fugacity is expressed as bL = H2 x2 (10.1-27) f 2 bL = H2 x2 f 2 2 (10.1-28)

For concentrated mixtures, however, it is necessary to incorporate a correction factor (or activity coecient) into Eq. (10.1-27) in the form

where the term 2 is called the asymmetric activity coecient. Note that 2 must satisfy the following constraint x2 0 (10.1-29) 2 1

The terms "solvent" and "solute" are used for the components present in the mixture in larger and in smaller amounts, respectively.

366

The activity coecient based on the pure component fugacity at the same temperature and pressure as the mixture was dened by Eq. (8.3-5) as 2 = bL f 2 = Actual fugacity of component 2 Fugacity of component 2 based on the Lewis-Randall rule 2 1 Rearrangement of Eq. (10.1-28) in the form 2 = bL f 2 = Actual fugacity of component 2 Fugacity of component 2 based on Henrys rule (10.1-32) x2 1 (10.1-30)

Lx f2 2

with the constraint that (10.1-31)

H2 x2

indicates that the standard state fugacity in the denition of the asymmetric activity coecient is simply the Henrys law constant. The standard states are shown in Figure 10.4 for two dierent cases.

L f2

2 > 1

* 2 <1

H2

L f2

2 < 1

* 2 >1

L f2 = H2 x2

L f2
L L f2 = f2 x2

L f2

H2

L L f2 = f2 x2

L f2 = H2 x2

x2 (a)

x2 (b)

Figure 10.4 Standard states used in the denition of activity coecients. The relationship between 2 and 2 can be found as follows. Comparison of Eqs. (10.1-30) and (10.1-32) gives L (10.1-33) H2 2 = f2 2 Taking the limit as x2 0 yields
L 2 H2 = f2

(10.1-34)

The use of Eq. (10.1-34) in Eq. (10.1-33) gives 2 = 2 2

(10.1-35)

Example 10.7 Consider a binary mixture of solvent (1) and solute (2). If the system is represented by the two-sux Margules equation, determine the asymmetric activity coecient, 2. 367

Solution From Eq. (8.4-3) ln 2 = A x2 1 Taking the limit as x2 0 gives The use of Eq. (10.1-35) gives
ln 2 = ln 2 ln 2

(1) (2)

ln 2 =A

2 = A x2 1 A = A (x1 1)

(3)

10.2 FACTORS AFFECTING GAS SOLUBILITY Consider a binary mixture of solvent (1) and solute (2). The condition of equilibrium between the gas and liquid phases is expressed as bV (T, P, y2 ) = f bL (T, P, x2 ) f 2 2 (10.2-1)

or

At a given temperature, pressure, and gas composition, the left-hand side of Eq. (10.2-2) is xed. Therefore, there is an inverse relationship between solubility of solute, x2 , and Henrys law constant, i.e., the higher the Henrys law constant, the lower the solubility. 10.2.1 Eect of Pressure At low pressures, the dependence of Henrys law constant on pressure may be neglected. At high pressures, however, a relationship indicating the variation of Henrys law constant with pressure is necessary. For this purpose, note that Eq. (7.7-1) gives the change in fugacity with pressure as ! L bL ln f V2 2 (10.2-3) = P RT
T,x

bV (T, P, y2 ) = H2 (T, P ) x2 (T, P, x2 ) f 2 2

(10.2-2)

Taking the limit of Eq. (10.2-3) as x2 0 leads to

x2 0

lim

bL ln f 2 P

= lim

T,x

x2 0

V2 RT

L!

(10.2-4)

Since

V2 V2 Eq. (10.2-4) becomes

bL H2 x2 f 2
L L

ln H2 P

As x2 0
L

(10.2-5)

T,x

V2 = RT

(10.2-6)

368

where V 2 is the partial molar volume of solute in the liquid phase at innite dilution. Intevap , to gration of Eq. (10.2-6) from the vapor pressure of solvent at the given temperature, P1 any pressure P gives ! Z L P V 2 vap dP (10.2-7) ) exp H2 = H2 (P1 vap RT P1 Substitution of Eq. (10.2-7) in Eq. (10.2-2) results in bV f 2 =
vap H2 (P1 ) exp

vap P1

V2 dP RT

x2 2

(10.2-8)

Krichevsky-Kasarnovsky Equation To simplify Eq. (10.2-8), Krichevsky and Kasarnovsky (1935) made the following assumptions: i) x2 1 so that 2 = 1, ii) V 2
L

is independent of pressure.

The resulting equation becomes bV f 2 !

vap )+ ln H2 (P1

ln

x2

V2 vap (P P1 ) RT

(10.2-9)

which is known as the Krichevsky-Kasarnovsky equation. It predicts solubilities of sparingly soluble gases, i.e., H2 or N2 in water, fairly well. vap ' 0, the gas phase is almost When the vapor pressure of solvent is negligible, i.e., P1 V V b pure solute and f2 can be substituted by f2 , fugacity of pure solute in the gas phase. Under these circumstances, Eq. (10.2-9) reduces to ln
V f2

x2

vap )+ = ln H2 (P1

V2 P RT

(10.2-10)

V /x ) is plotted as a function of P a straight line is obtained with a suggesting that when ln(f2 2 L vap slope of V 2 /RT and an intercept of ln H2 (P1 ).

Example 10.8 Yuan et al. (2007) reported the solubilities of carbon dioxide (2) in 2-hydroxyl ethylammonium formate (1) at 313 K as a function of pressure. Using the following data, estimate the Henrys law constant and partial molar volume of CO2 at innite dilution. Use the Soave-Redlich-Kwong equation of state to calculate fugacities. x2 0.0304 0.0831 0.1001 P ( bar) 5.1 16.0 19.8 x2 0.1099 0.1796 0.2101 P ( bar) 22.2 43.3 58.6 x2 0.2428 0.2468 0.2437 P ( bar) 74.6 86.9 90.9

369

Solution
V /x ) versus P , it is rst necessary to calculate the fugacity of pure carbon In order to plot ln(f2 2 dioxide using Eq. (5.3-11) at dierent pressures. The results are given in the following table :

x2 0.0304 0.0831 0.1001

P ( bar) 5.1 16.0 19.8

V ( bar) f2

x2 0.1099 0.1796 0.2101

P ( bar) 22.2 43.3 58.6

V ( bar) f2

x2 0.2428 0.2468 0.2437

P ( bar) 74.6 86.9 90.9

V ( bar) f2

5.0 14.9 18.2

20.2 35.7 44.7

52.2 56.2 57.2

V /x ) versus P is shown below. The plot of ln(f2 2

5.458

5.4

5.3 y z 5.2

5.1

5 5 0 5.1 20 40 P 60 80 90.9 100

The correlation coecient for the best straight line passing through the data points is 0.987. The slope and the intercept of the straight line are 3.668 103 and 5.121, respectively. Thus, H2 = exp(5.121) = 167.5 bar V2
L

= (83.14)(313)(3.668 103 ) = 94.45 cm3 / mol

Example 10.9 Increased concentrations of carbon dioxide in the atmosphere lead to global warming. The purpose of "carbon dioxide sequestration" is to store CO2 in the earth rather than releasing it into the atmosphere. One of the proposed methods is to discharge liqueed CO2 through submerged pipelines below the mixed layer of the ocean, typically at depths of 5003000 m. Since the solubility of CO2 in water forms a basis for the prediction of the solubility of CO2 in seawater, Teng et al. (1997) studied the solubility of CO2 (2) in water (1) at various temperatures and pressures. The reported solubility values at 283 K are as follows : x2 = 0.0279 x2 = 0.0305 at 64.4 bar at 147.7 bar

Estimate the solubility of carbon dioxide at 245.8 bar and 283 K. Solution
vap ) and V 2 can be calculated by using the From the given data points, the parameters H2 (P1 Krichevsky-Kasarnovsky equation, Eq. (10.2-10). Using the Soave-Redlich-Kwong equation of state, the fugacities of pure carbon dioxide are calculated from Eq. (5.3-11) as L

370

V = 35 bar f2 V = 42.3 bar f2

at 64.4 bar at 147.7 bar

The use of Eq. (10.2-10) gives ln 35 0.0279 42.3 0.0305 =

vap ln H2 (P1 )+

V 2 (64.4) (83.14)(283) V 2 (147.7) (83.14)(283)

L

(1)

ln

vap ln H2 (P1 )+

(2)

Simultaneous solution of Eqs. (1) and (2) gives

vap ln H2 (P1 ) = 7.057

and

V2

= 28.341 cm3 / mol

Therefore, Eq. (10.2-10) takes the form V! f2 28.341 P ln = 7.057 + x2 RT At P = 245.8 bar the fugacity of CO2 is 52 bar. Hence, 52 (28.341)(245.8) = ln = 7.057 + x2 (83.14)(283)

(3)

x2 = 0.0333

Comment: Teng et al. (1997) measured the solubility of CO2 at 245.8 bar as 0.0326.

Krichevsky-Ilinskaya Equation The assumptions used by Krichevsky and Ilinskaya (1945) in the simplication of Eq. (10.2-7) are as follows: i) The system is represented by the two-sux Margules equation, 2 For the two-sux Margules equation, Example 10.7 indicates that 2 = exp A (x1 1) . Therefore, Eq. (10.2-7) becomes ln bV f 2 ! =
vap )+ ln H2 (P1 L

ii) V 2

is independent of pressure.

x2

V2 vap (P P1 ) + A (x2 1 1) RT

(10.2-11)

which is known as the Krichevsky-Ilinskaya equation. The use of this equation requires the L vap ), V 2 , and A. evaluation of three parameters, H2 (P1 10.2.2 Eect of Temperature Intuitively, lower solubilities are expected as the temperature increases. From our everyday experience we notice that the amount of carbon dioxide escaping from the soda pop bottle when the bottle is opened is greater if the bottle is warm. Although solubility decreases with

371

increasing temperature for most gases, the reverse may be true especially at very high temperatures. In general, at low temperatures the Henrys law constant increases with temperature, reaches a maximum, and then decreases at higher temperatures, i.e., T Tc . Since solubility is inversely proportional to Henrys law constant, solubility decreases with increasing temperature at low temperatures, reaches a minimum, and then increases with increasing temperature at high temperatures. As stated by Smith and Harvey (2007), even a variation as small as 10 K may change the Henrys law constant by a factor of two, leading to serious errors in the design of process equipment. To quantify the variation of Henrys law constant with temperature, note that Eq. (7.7-5) gives the change in fugacity with temperature as bL ln f 2 T ! =
L e IG H2 H 2

P,xj

RT 2

(10.2-12)

Taking the limit of Eq. (10.2-12) as x2 0 leads to ! L bL e IG H2 H ln f 2 2 lim = lim x2 0 x2 0 T RT 2

P,xj

(10.2-13)

Since

H2 H2 Eq. (10.2-4) becomes

L

bL H2 x2 f 2
L L

As x2 0

(10.2-14)

ln H2 T

P,x

H2

RT 2

e IG H 2

H 2

RT 2

(10.2-15)

where H 2 is the partial molar enthalpy of solute in the liquid phase at innite dilution, and H 2 is the enthalpy change associated with the transfer of one mole of a solute from an ideal gas state to an innitely dilute solution. The term H 2 can be regarded as a heat of solution 5 . For small temperature dierences, H 2 may be considered constant. Under these circumstances, rearrangement of Eq. (10.2-3) under isobaric conditions gives H 2 d ln H2 = d(1/T ) R Integration of Eq. (10.2-16) gives
H 2 1 H2 (T2 ) 1 = ln H2 (T1 ) R T2 T1

(10.2-16)

(10.2-17)

H 2 can be either positive or negative, and must be measured experimentally. If it is negative, i.e., exothermic process, the solubility decreases (or H increases) with increasing temperature.
The heat eect associated with the dissolving of a solute (gas or solid) in a liquid is called a heat of solution. Its value is closely related to the interactions between solute and solvent. Heat of solution may be either endothermic or exothermic depending on whether more energy is used to break bonds, or more energy is released as a result of new bond formation.
5

372

Besides Henrys law constants, Sander (1999) also published the values of H 2 /R for various substances. In the absence of experimental information, the use of
e vap H 2 ' H 2

(10.2-18)

e vap is the molar enthalpy of vaporization of pure solute, may provide satisfactory where H 2 results. Example 10.10 Indicate the conditions under which the approximation given by Eq. (10.218) is valid. Solution At low and moderate pressures Eq. (10.1-26) gives
vap H2 = 2 P2

(1)

Thus,

vap H2 (T2 ) 2 (T2 ) P2 (T2 ) = H2 (T1 ) 2 (T1 ) P vap (T1 ) 2

(2)

Assuming the ratio of the innite dilution activity coecients to be unity, Eq. (2) can be rearranged as " vap # P2 (T2 ) H2 (T2 ) = ln (3) ln vap H2 (T1 ) P2 (T1 ) Using the Clausius-Clapeyron equation, i.e., Eq. (5.6-13), the dependence of vapor pressures on temperature is expressed as " vap # e vap H P2 (T2 ) 1 1 2 ln (4) = vap R T2 T1 P2 (T1 ) Substitution of Eq. (4) into Eq. (3) gives e vap H H2 (T2 ) 1 1 2 = ln H2 (T1 ) R T2 T1 e vap . which is identical with Eq. (10.2-17) if H 2 = H 2 10.2.3 Eect of Electrolyte Addition Addition of a small amount of electrolyte to a solvent stabilizes the charged groups on the solute. In this way, intermolecular forces between solute and solvent becomes stronger, leading to increasing solubility of a gas in a given solvent. This phenomenon is known as salting-in. At higher concentrations of electrolyte, however, solubility of a particular gas in a solvent decreases, which is known as salting-out. The reason for the salting-out phenomenon is the weakening of the intermolecular forces between solute and solvent as a result of the greater anity of an electrolyte for the solvent.

(5)

373

10.2.4 Temperature and Pressure in a Gas Absorber For a binary system of components 1 and 2, the vapor-liquid equilibrium for solute (2) gives H2 x2 = m2 x2 (10.2-19) y2 = P The x-y diagram for a solute is given in Figure 10.5 in which m2 (or y2 /x2 ) is the slope of the equilibrium line. During gas absorption operation, for a xed gas composition, the purpose is to make x2 as high as possible. This is possible by keeping m2 as small as possible. Small values of m2 can be achieved either by decreasing H2 (low temperature) or by increasing pressure. As a result, gas absorption is preferred at low temperatures and high pressures.
y2 y2 = m2 x2

x2 = 0

x2 < 0.1

Figure 10.5 The x-y diagram for a solute.

10.3 APPLICATIONS OF HENRYS LAW 10.3.1 Carbonated Beverages All carbonated beverages, such as Coca Cola, Sprite, 7-Up, champagne, and sparkling wine, are bottled under pressure to increase the carbon dioxide dissolved in solution. When the bottle or can is opened, the pressure in the head space decreases. Since the liquid can no longer keep a large amount of carbon dioxide dissolved in it (in other words, the liquid is supersaturated with carbon dioxide), the solution eervesces and the carbon dioxide leaves the solution in the form of bubbles. These bubbles not only provide zz, but also change the avor of the drink. The CO2 in the soda forms carbonic acid, which counteracts the sweet taste of the drink. Example 10.11 On the average, the 330 mL soda cans have a pressure of roughly 1.2 atm when canned at 4 C, and 2.5 atm when stored at 20 C. In an unopened can, CO2 gas dissolved in the liquid is in equilibrium with gas in the space between the can and the liquid. Once the can is opened, the pressure of CO2 above the liquid becomes essentially zero (the partial pressure of CO 2 in the atmosphere is about 3 104 atm) and the liquid enters a metastable state, i.e., it contains CO2 in excess of what Henrys law states. To reach a new equilibrium state corresponding to the partial pressure of CO2 in the atmosphere, almost all the CO2 dissolved in the liquid must escape6 . In other words, the soda goes "at". Lack of CO2 makes "at" soda taste much sweeter than when it is carbonated. Calculate the concentration of CO2 in soda (in mol/L) before and after the can is opened. Solution Let us assume that the Henrys law constant for CO2 , 1.63 103 atm, at 25 C is also applicable at 20 C. Choosing 1 L of solution ( 1000 g of water) as a basis, the concentrations are
6

It takes a nite amount of time for the system to reach this new equilibrium state.

374

Closed can 2.5 = 1.63 10 Open can 3 10

4 3

nCO 2 nCO 2 + (1000/18)

nCO 2 = 0.085 mol/L

= 1.63 10

nCO2 nCO2 + (1000/18)

nCO2 = 1.02 105 mol/L

Comment: Boiled water also tastes "at" because all of the oxygen gas has been removed by heating. This is the reason why people do not prefer to prepare tea with water boiled for extended periods. Similarly, addition of boiled water to an aquarium will cause the sh to suocate.

10.3.2 The Bends The hydrostatic pressure in the sea increases almost linearly with depth according to the relation (10.3-1) P = Patm + gh where h is the distance measured from sea level. The density of saltwater, , is about 1025 kg/ m3 . Therefore, the pressure at a depth of 40 m is7 P = 1 + (1025)(9.8)(40)(9.8692 106 ) = 4.97 atm | {z }
atm/( kg. m1 . s2 )

The muscles we use to expand and contract our lungs during breathing are not strong enough to overcome such a pressure8 . One way of overcoming this problem is to protect the divers body from ambient pressure by placing him/her in a deep sea submersible in which pressure is maintained at 1 atm. Another way of solving this problem is to provide a pressurized breathing gas mixture to the diver. If the breathing gas supply is delivered at the same pressure as the surrounding ambient pressure, the divers lungs do not have to work against the water pressure. Therefore, as scuba divers dive deeper, the pressure of air they breathe must match the pressure of the water around them. This increase in air pressure, however, increases the solubility of nitrogen in the blood as shown in Table 10.2. In the calculations, density of seawater is taken as 1025 kg/ m3 , blood properties are considered to be equal to those of water and the variation of the Henrys law constant with temperature is considered negligible. As a result of breathing compressed air and being subjected to high pressures caused by water depth, the amount of nitrogen dissolved in blood and other tissues increases. If the diver swims quickly to the surface, the nitrogen forms bubbles in the blood as it becomes less soluble due to a decrease in pressure. These bubbles, by blocking the capillaries in the body and stopping the blood ow, produce the symptoms of decompression sickness (numbness, diculty in breathing, itching, pain in the joints or abdomen), which is also known as the bends. The bends is very painful and is sometimes fatal. To avoid the eects of the bends, the diver must rise slowly and/or make intermittent stops on the way up, called decompression stops, so that the gas can come out of solution slowly.

7 8

As a rule of thumb, keep in mind that 10 m of water height creates a pressure of 1 atm. This is the reason why we cannot simply use a snorkel of the necessary length for diving.

375

Table 10.2 Increase in nitrogen solubility in blood as a function of depth. Depth ( m) 0 2 4 6 8 10 12 14 16 18 20 P ( atm) 1 1.198 1.397 1.595 1.793 1.991 2.190 2.388 2.586 2.784 2.983 PN 2 ( atm) 0.780 0.935 1.089 1.244 1.399 1.553 1.708 1.863 2.017 2.172 2.327 mol N2 /L blood 5.092 104 6.102 104 7.111 104 8.121 104 9.130 104 1.014 103 1.115 103 1.216 103 1.317 103 1.418 103 1.519 103

10.3.3 Natural Disaster at Lake Nyos9 On August 26, 1986, ocials found 1746 people and several thousand livestock lying dead mostly within 3 km of Lake Nyos, a remote lake in the highlands of northwest Cameroon. Lake Nyos rests in a volcanic crater called a maar. Maars are inverted volcanoes that form as a result of a series of powerful explosions that occur when magma (T 1400 K) encounters cooler groundwater. The resulting steam explosions excavate a crater. Initial reports from local authorities suggested that the tragedy was the result of a volcanic eruption. Investigators arriving at the lake several days after the incident were puzzled by the lack of evidence for hot, caustic gases that normally accompany volcanic eruptions; plants on the lake perimeter were uncharred and apparently unharmed. Later, scientists discovered that water at the bottom of the lake (maximum depth = 208 m) contained high concentrations of dissolved CO2 gas. The amount of CO2 dissolved in water was estimated to be 300 106 m3 . Apparently, CO2 released from the volcano into the lake was dissolved in the lower layers of the lake by the high pressure of the water above. Application of the Henrys law indicates that 1 L of water at 40 m holds 2 to 3 L of CO2 , at 190 m up to 11 L of CO2 . Scientists speculated that a landslide into the lake stirred up the waters. As the CO2 -rich waters moved to the surface of the lake, where hydrostatic pressure was less, CO2 solubility decreased resulting in the release of the gas, which erupted explosively from the lake surface. At Lake Nyos, the volume of the gas vented to the atmosphere was sucient to cause a 1 m drop in lake level (the surface area of the lake is 1.49 km2 ). The research team estimated that the volume of CO2 released was approximately equal to 1 km3 , or about 2 106 tons. The exit velocity of the gas was estimated as 80 m/ s. Since CO2 (molecular weight = 44) is heavier than air (molecular weight = 29), the released gas did not simply rise into the atmosphere but formed a "river" of CO2 about 50 m thick that swept through the village nearby Lake Nyos, suocating human and animal life. To prevent the occurrence of a similar disaster, scientists have been degassing Lake Nyos since April 200110 . For this purpose, a 205 m long polyethylene pipe (ID = 315 mm, OD = 258 mm) was placed vertically into the lake and some water was pumped up through the pipe to start the process. As the water moves up in the pipe, pressure decreases and CO2 bubbles start to form within the liquid phase. The dierence in densities of liquid and gas
9 10

For more information see: http://www.ched-ccce.org/confchem/1998/donnelly See http://mhalb.pagesperso-orange.fr/nyos/2006/index2006.htm

376

bubbles is large enough to keep the process going, i.e., once the movement of the bottom layers is initiated, no pumping is necessary. The resulting fountain shoots 6-10 m in the air. 10.3.4 Knuckle Cracking Have you ever wondered where the noise comes from when you crack your knuckles? Joints are where two bones meet and they are covered by a capsule lled with synovial uid, which is responsible for the lubrication of the joint. Gases, such as O2 , N2 , and CO2 , remain dissolved in this uid as long as the pressure in the joint is fairly constant. When the joint is stretched, i.e., the volume of the joint is increased, the pressure in the capsule decreases, leading to the formation of gas bubbles. The formation of this bubble creates a popping sound, known as "cracking." It takes about 25-30 minutes for the gas to be dissolved in the synovial uid again. This is why a joint cannot be "recracked" immediately. REFERENCES Dalmolin, I., E. Skovroinski, A. Biasi, M.L. Corazza, C. Dariva and J.V. Oliveira, 2006, Fluid Phase Equilibria, 245, 193-200. Gabelman, A. and S.T. Hwang, 1999, J. Memb. Sci., 159, 61-106. Hildebrand, J.H. and R.L. Scott, 1950, Solubility of Nonelectrolytes, 3rd Ed., Reinhold, New York. Krichevsky, I.R. and J.S. Kasarnovsky, 1935, J. Am. Chem. Soc., 57, 2168-2172. Krichevsky, I.R. and A.A. Ilinskaya, 1945, Acta Physicochim. USSR, 20, 327-348. Miyano, Y., 2004, J. Chem. Thermodynamics, 36, 865-869. Miyano, Y., I. Fujihara and K. Sato, 2006, Fluid Phase Equilibria, 240, 56-62. OSullivan, T.D. and N.O. Smith, 1970, Journal of Physical Chemistry, 74 (7), 1460-1466. Rigby, M. and J.M. Prausnitz, 1968, J. Phys. Chem., 78 (1), 330-334. Sander, R., 1999, Compilation of Henrys law constants for inorganic and organic species of potential importance in environmental chemistry" (Version 3), http://www.henrys-law.org. Sciamanna, S.F. and S. Lynn, 1988, Ind. Eng. Chem. Res., 27 (3), 492-499. Sirkar, K.K., 1997, Chem. Eng. Comm., 157, 145-184. Smith, F.L. and A.H. Harvey, 2007, Chemical Engineering Progress, 103 (9), 33-39. Yaws, C.L., J.R. Hopper, X. Wang, A.K. Rathinsamy and R.W. Pike, 1999, Chemical Engineering, 106 (June), 102-105. Yuan, X., S. Zhang, J. Liu and X. Lu, 2007, Fluid Phase Equilibria, 257, 195-200.

377

PROBLEMS Problems related to Section 10.1 10.1 A gaseous species (component 2) dissolves in liquid (component 1) to form an ideal mixture of composition x2 = 0.015. If the normal boiling point of this mixture is 337 K, estimate the Henrys law constant of component 2. The vapor pressures of components 1 and 2 at 337 K are 0.98 bar and 0.35 bar, respectively. (Answer: 3.18 bar) 10.2 Henrys law constants of sulfur dioxide in dierent organic solvents at 298 K are given in the following table. Which solvent would you choose to absorb SO2 ?

Solvent Tetrahydrofuran 1, 2-dichloroethane Pyridine Tetramethylsulfone

H ( kPa) 39.2 428.0 8.8 101.0

10.3 A binary system of species 1 and 2 consists of vapor and liquid phases in equilibrium at 350 K and 0.9 bar. The liquid phase is concentrated in species 2, i.e., x2 x1 . Estimate the compositions of the liquid and vapor phases using the following data
vap P1 = 0.324 bar H1 = 130 bar vap P2 = 0.199 bar H2 = 70 bar

(Answer: x1 = 0.0054 y1 = 0.78) 10.4 A storage tank contains a mixture of n-hexane (1) and methane (2) at 310.9 K and 10.8 bar. The mole fractions of methane in the liquid and vapor phases are reported as x2 = 0.0496 y2 = 0.9613

Estimate the Henrys law constant of methane in n-hexane. The vapor phase is represented by the Peng-Robinson equation of state with k12 = 0.04. (Answer: 205 bar) 10.5 When the Poynting correction factor is almost unity, Eq. (10.1-25) reduces to
vap V Hi = i (T, Pivap ) i Pi

(1)

Under atmospheric pressure, Miyano (2004) reported the innite dilution activity coecients of n-butane (2) in isobutanol (1) at dierent temperatures as T ( K) 2 250 4.03 270 4.00 290 3.97

378

Estimate the Henrys law constant for n-butane in isobutanol using Eq. (1). Use the virial equation of state to determine fugacity coecients of pure n-butane. (Answer: 1.545 bar, 3.524 bar, 6.993 bar) 10.6 From Eq. (10.1-6) Hi = lim
xi 0

bL f i xi

a) For a binary system of species 1 and 2, if the fugacity coecient is calculated by the Redlich-Kwong equation of state, show that s ! B1 L A B B2 A 22 11 1 L (Z 1) ln(Z2 B2 ) 2 (1 k12 ) ln 1 + L (2) ln H1 = ln P + B2 2 B2 A22 B2 Z
2

L b = lim P i
xi 0

(1)

b) For a binary system of n-butane (1) and methanol (2) at 374 K and 1.013 bar, Miyano et al. (2006) reported that 1 = 9.3. First use Eq. (3) of Problem 8.27 to estimate the binary interaction parameter, k12 . Then use Eq. (2) to estimate the Henrys law constant of n-butane in methanol. (Answer: 0.151, 117 bar) 10.7 A system of water (1) and methane (2) exists as two phases at 298 K and 23.23 atm. Rigby and Prausnitz (1968) reported the water mole fraction in the vapor phase as 1.483 103 . The vapor phase is represented by the virial equation of state with the following parameters: B11 = 1165 cm3 / mol B22 = 43.4 cm3 / mol

a) Estimate the liquid phase composition. b) Determine the second virial cross coecient, B12 . The vapor pressure of water at 298 K is 0.03126 atm. (Answer: a) x1 = 5.64 104 b) 64.9 cm3 / mol) 10.8 A binary liquid mixture of 28 mol % benzene (1) and 72% n-heptane (2) is at a temperature of 318.15 K. Estimate the composition of the vapor phase in equilibrium with this liquid mixture if the system is represented by the three-sux Margules equation. DATA: At 318.15 K Henrys law constant for benzene in n-heptane = 0.522 bar Henrys law constant for n-heptane in benzene = 0.285 bar (Answer: y1 = 0.495) Problems related to Section 10.2 10.9 The release of excessive waste heat into rivers and lakes leads to an increase in water temperature, known as "thermal pollution". Discuss the probable harmful eects of thermal pollution on the environment. 10.10 Hyperbaric oxygen therapy consists of breathing 100% oxygen under increased pressure. It is especially applied to patients suering from carbon monoxide poisoning. Why? 10.11 The following data are provided by OSullivan and Smith (1970) for the solubility of nitrogen and methane in water (1) at 324.5 K: 379

P ( bar) 101.3 202.6 303.9 405.2 506.5 607.8

x2 104 Nitrogen 7.99 14.54 20.17 24.90 29.20 33.50

L

Methane 14.27 22.79 28.70 33.40 37.30 40.90

vap Use Eq. (10.2-9) and estimate H2 (P1 ) and V 2 for nitrogen and methane. Use the PengRobinson equation of state to calculate pure component fugacities.

(Answer: N2 H2 = 1.14 105 bar, V 2 L V 2 = 39.4 cm3 / mol)

= 34.8 cm3 / mol; CH4

H2 = 5.555 104 bar,

10.12 Henrys law coecients for propane in heavy hydrocarbons are reported by Sciamanna and Lynn (1988) as T ( C) 25 35 45 816 985 1175 H ( kPa)

Dodecane

Hexadecane 775 926 1096

Estimate the heat of solution of propane in dodecane and hexadecane. (Answer: 14, 367 J/ mol, 13, 652 J/ mol)

380