The Chemical Engineering Journal, 5 (1973) 77-84 0 Elsevier Sequoia S.A., Lausanne.

Printed in the Netherlands

77

Absorption of Chlorine into Aqueous Sodium Hydroxide Solutions

H. HIKITA,
Department

S. ASAI, Y: HIMUKASHI
of Chemical Engineering,

and T. TAKATSUKA
Sakai, Osaka (Japan)

University of Osaka Prefecture,

(Received 27 October 1972; in final form 26 December 1972)

Abstract The rates of absorp tion of pure chlorine in to aqueous sodium hydroxide solutions of vanous concentrations were measured at 30C using a liquid-jet column. The experimental results and also the previous data of Spalding and of Takahashi et al. were analysed on the basis of the penetration theory for gas absorption accompanied by a two-step instantaneous chemical reaction, which was presented by the authors in a previous paper. The data were in good agreement with the theoretical predictions.
INTRODUCTION There have been very few studies of the absorption

cock type absorber. Their experimental results on the absorption rate also showed good agreement with penetration theory predictions for absorption accompanied by an instantaneous irreversible reaction between the dissolved Cl* and OH- ions. In these studies, it was assumed that the forward part of reaction (b) was the only reaction governing the absorption rate of Clz. However, this assumption seems unreasonable because, in a strong hydroxide solution, HOC1 (hypochlorous acid) formed by reaction (b) can react again with OH- ions according to the very rapid reaction HOC1 + OH- = OCl- t H20 (c)

of chlorine into aqueous hydroxide solutions, although this chemical absorption process is of considerable industrial importance and also of theoretical interest. Spalding studied the rate of absorption of Cl2 into water and aqueous solutions of HzS04 and NaOH in a liquid-jet column, and found that the absorption rate of Cla was affected by reactions (a) and/or (b): Cl? + Hz0 = HOC1 + H+ + ClClz t OH- = HOC1 f Cl(a) (b)

The equilibrium constant of this reaction is very large, as will be mentioned later, and it is to be expected that this reaction also affects the absorption rate of c12. The present work was undertaken to clarify the mechanism of chemical absorption of Cl;! into aqueous hydroxide solutions. Absorption experiments with the C12-NaOH system were performed using a liquid-jet column. The results were interpreted on the basis of the penetration theory for absorption with a two-step instantaneous reaction, which was proposed in a previous papers.
MECHANISM OF CHEMICAL ABSORPTION

depending upon the pH value of the solution. He showed that in solutions with pH values higher than 12.6 (which corresponded to an OH- concentration of 0.04 g mol/Q) the forward part of reaction (b), i.e. Cl* + OH--+ HOC1 + Cl-, was rate-controlling and the effect of this reaction on the absorption rate could be predicted by the penetration theory for absorption accompanied by an instantaneous irreversible reaction2*3 of the type A + B + E. Takahashi et aL4 measured the absorption rates of Cl2 into aqueous NaOH solutions of concentrations from 0.05 to 0.2 g mol/Q over a wide range of exposure times, using a liquid-jet column and a stop-

When Cl;! is absorbed into an aqueous hydroxide solution, the following over-all reaction between the dissolved Cl2 and OH- ions may take place in the liquid phase: Cl2 + 20H-= Cl- t OCl- t H,O (d)

Reaction (d) takes place in two steps: Cl2 +OH-=HOCltClHOC1 + OH- = OCl- t H20 (b)
cc>

78 The values of the equilibrium constants of reactions (b) and (c), K 1 and K2 respectively, are
K =

H. HIKITA et al.

WC'1

[cl-l

3.

,010

[Cl21 W-1

[ocl-] K2 = ,HOC1l ,OH-l = 2.2 x IO6

Q/gmol

tration theory, and found that the mechanism of chemical absorption in this case depends on the equilibrium-constant ratio P (= KI/K2). In the limiting case where P = 0, it was shown that there exists only one reaction plane, where the over-all reaction of steps (e) and (f), i.e. A + 2B * E, proceeds irreversibly, and thus the average rate of absorption of the solute gas can be calculated from the well-known equations

at 30C and infinite dilution. These values were estimated from the equilibrium constants for the Cl* hydrolysis reaction6 and the HOC1 dissociation reaction7. The first-step reaction (b) is very rapid, with a rate constant around lOlo Q/g mol seca9. Since the second-step reaction (c) is a proton-transfer reaction, its rate constant will be very much greater than that of reaction (b). Therefore, both reactions (b) and (c) may be considered to be instantaneous. The dissolved Cl* will also react with water according to the hydrolysis reaction Cl, + Ha0 = HOC1 + H + + ClThe value of the equilibrium K3. is K
3

(1)
/3= 1/erf(o) erfc
DA -_

(2)
1

DB

1I
u2
(3)

(a) of this reaction,

constant

[HoC1luH+l [-I =4.5

[Cl21

10e4 (g mol/Q)*

at 30C and infinite dilution6. The rate constant of the forward reaction is 2 1.4 set- at 30C lo. These values are very low compared with those of reaction (b). Hence in a strong hydroxide solution reaction (a) will make a negligible contribution to the total reaction rate of Cl,. Therefore, the absorption of Cl2 into aqueous hydroxide solutions can be regarded as a process of absorption accompanied by an instantaneous two-step reaction, i.e. reaction (b) followed by reaction (c). In a previous papers, Hikita etal. considered the problem of gas absorption with an instantaneous twostep reaction, ie. A+B=C followed by C+B=E (0 (e)

which were derived by Danckwerts and Sherwood and Pigford for an instantaneous irreversible single reaction of the type A + 2B + products. This absorption mechanism will be called here a one reaction-plane model. In another limiting case where P = 00, it was shown that two reaction planes are formed within the liquid, and the reaction A + E + 2C (the sum of the first-step reaction (e) and the secondstep reaction (f) in the reverse direction) and the reaction (f) take place irreversibly at the first reaction plane (which is located closer to the gas-liquid interface) and the second reaction plane, respectively. Further, the reaction factor in this case is given by fl= l/erf(ut) (4)

where the constant ur can be determined by solving the following pair of simultaneous equations:

erfc(&ur]exp[(?2erfc(&ui)exp[(~t($?)oa2]

l]ur2]

~ l)0,2t

wher-e the equilibrium constants of steps (e) and (f), K, and K2 respectively, are both very large. They nresented a theoretical analvsis based on the nene-

=k?erf(ur)

(6)

ABSORPTION

OF CHLORINE

79 will take place at the first and second reaction planes, respectively, Further, the reaction factor for this case is given by eqns. (4) (5) and (6).
EXPERIMENTAL

This absorption mechanism will be called here a two reaction-plane model. The reaction scheme for the Cla-NaOH system studied in this work is of the form A+B=C+C CtB=E (8) (f)

which is somewhat different from that treated previously5. However, one of the intermediate products, C, is non-reactive and hence its presence does not affect the chemical absorption mechanism, so that the theoretical analysis given in the previous paper may also be used for the present system. As described above, the values of the equilibrium constants of reactions (b) and (c) are 3.1 x 10 and 2.2 x lo6 S?/gmol, respectively, and hence the equilibrium-constant ratio P is 1.4 x lo4 g mol/Q. Therefore the two reaction-plane model is applicable to the present system. According to this model, the concentration profile of each species in the liquid near the interface will be similar to that shown in Fig. 1, and the irreversible reactions Cl* + OCl- + Ha0 -+ 2HOCl + Cland 2HOCl t 20H- + 2OCl- + 2HaO (i)
-

(h)

Apparatus and procedure The experimental apparatus is shown schematically in Fig. 2. The liquid-jet column was the same as that used by Hikita et al. in previous work11,12. The laminar liquid jet was formed by a bell-shaped glass nozzle with an exit diameter of 1.27 mm. The gas phase was pure Cl2 saturated with water vapour, and the absorbing liquids were 0.125, 0.25, 0.50 and 1.O g mol/R NaOH aqueous solutions. The liquid flow rate was kept constant at about 4 cm3/sec, and the exposure time of the liquid to the gas was varied from 0.0012 to 0.025 set by changing the jet length. Cl2 absorption rates were determined volumetrically by means of a soap-film meter. All the experiments were conducted at atmospheric pressure and at a temperature of 30 + O.lC. Prior to the chemical absorption experiments, preliminary runs on the physical absorption of pure CO2 into water were carried out at 30C using the same apparatus, in order to check the hydrodynamics of the liquid jet.

kregion

l+region

Aregion

d-:
interface

I I
first reaction - plane

reaction-plane

X2
distance

Fig. 1. Concentration profiles aqueous NaOH solution.

for absorption

of Cl2 into

Fig. 2. Schematic diagram of experimental apparatus: 1 liquid storage tank; 2 liquid feed tank; 3 exit liquid tank; 4 pump; 5 constant head device; 6 overflow tube; 7 liquidjet column; 8 soap-film meter; 9 water vapour saturator; 10 gas preheater; 11 Cl2 or CO2 cylinder; 12 air bath; CG gas cock; CL liquid cock; TG thermometer for gas; TL thermometer for liquid.

H. HIKITA

ef al.

Liquid-jet

Stop-cock

t x10
Fig. 3. Absorption

, set
at 30C.
lo+ lo-

rate of CO2 into water

100 18

10

lo2

10

Results

set
NaOH solutions

Experimental results on the physical absorption of CO2 into water are shown in Fig. 3, where the average absorption rates/V** are plotted against the exposure time t on logarithmic coordinates. The straight line in this figure represents the theoretical line for physical absorption calculated from the wellknown Higbie equation
*=2Ji

Fig. 5. Absorption rate of Cl2 into aqueous at 25C (data of Takahashi el aL4).

NA

J
-

DA

Trt

taking the physical solubility Ai and the liquid phase of CO2 in water at 30C as 2.84 x 10d5 diffusivityDA g mol/cm3 r3 and 2.25 x 10e5 cm2/sec 14, respect20

ively. It can be seen that the measured absorption rates agree very well with the theoretical line, lending strong support to the applicability of the penetration theory to the analysis of results of chemical absorption experiments. Figure 4 shows the experimental data for the Cl,-NaOH system as a log-log plot of the average absorption rate NA versus the exposure time t. Data points for each solution with a given concentration fall on a straight line with a slope of -4. A similar dependency of NA on t can also be seen from the experimental data of Takahashi et a1.4 obtained at 25C using a liquid-jet column and a stop-cock type absorber. Their data are shown as a log-log plot OfNA versus tin Fig. 5, which shows that the relation that NA varies with t-l /2 holds over a wide range of exposure times. Thus it may be concluded that the absorption of Cl2 into aqueous NaOH solutions is accompanied by an instantaneous chemical reaction between the dissolved Cl2 and OHions.

0 x z

DISCUSSION 2

OF RESULTS

Prediction of physical properties

1.0 0.8 0.6 3
I

Illll
6 610 set

L tx103,

20

Fig. 4. Absorption at 30C.

rate of Cl2 into aqueous

NaOH solutions

In order to compare the experimental results with the theoretical eqns. (1) (4), (5) and (6) it is necessary to know the values of the physical properties of the present system, such asAi, DA, Dg, DC, DE etc. According to the two reaction-plane model, an aqueous solution containing HOC1 and NaCl (= Na+ + Cl-) exists in the region between the gasliquid interface and the first reaction plane (region 1 in Fig. l), concentrations of these species, Cr and C;

ABSORPTION

OF CHLORINE

81
TABLE 1 for pure Clz-aqueous NaOH solution

respectively,

being given by the following equations:

Physical properties system at 30C

BO X Cl

erf (+A/&ca)

erf (

D..@~or)
(8)

(g mol/Q) 0 (water) 0.125 0.25 0.50 1 .oo z-2.32

(g mol/Q) 0 0.147 0.293 0.580 1.172 $=1.04

Ci k mol/Q) 0 0.141 0.275 0.538 1.079 2=1.40

Ai (g mol/Q) 0.0525 0.0510 0.0496 0.0469 0.0420 %=0.786

DAXlO (cm2/sec) 1.68 1.64 1.60 1.52 1.35

erfc(~D~lW4

+( %?)&]

eer;;$z;;

(9)

where Ai and DA are, respectively, the physical solubility and the liquid phase diffusivity of Cl2 in this solution. (Derivation of these equations is given in Appendix 1.) Since the effect of the presence of HOC1 on the solubilify of Cl2 is not known, the solubility was assumed to be the same as that in aqueous NaCl solution of concentration given by eqn. (9). The physical solubility of Cl2 in aqueous NaCl solution was estimated from the following equation (see Appendix 2): log (Ai/&)

= - k,C;

(10)

where Ai, is the physical solubility of Cl* in water and k, is the salting-out parameter. The value of Ai, was obtained from the Henrys law constant for the Clz-water system, and was taken as 0.0525 g mol/P at 30C. The value of k, at 30C was taken to be equal to 0.0903 Q/g mol (see Appendix 2). The liquid phase diffusivity of Cl2 in a solution containing HOC1 and NaCl was predicted from the value of 1.48 x 10ms cm2/sec 1,14, obtained at 25C for molecular Cl2 in water, by correcting for temperature and also for the viscosity of the solution according to the Stokes-Einstein relation: D.&T = constant (11)

DB~DA, DC/DA, Dc/DA and DE/DA, respectively, were assumed to be equal to those at infinite dilution or in water. The diffusivity of HOC1 in water was estimated from the measured value of 1.54 x 1O- cm2 /set l4 at 25C by correcting for temperature. The effective diffusivities of OH-, Cl- and OCl- at infinite dilution were estimated by the method of Vinograd and McBain, taking the ionic conductances of these ionic species at 30C as 216, 84 and 40 cm2/S2 g mol, respectively. (The value of the ionic conductance of OCl- ions is not known and was taken as 40, which was estimated from the variation of the ionic conductance of Cl,- ions (n = 2, 3 and 4) with the formula weight.) The calculated values of the effective diffusivities for species C (Cl- ions) and species E (OCl- ions) were somewhat different in regions 2 and 3 (see Fig. l), respectively; the arithmetic mean of the values estimated in both regions was employed for each species. As can be seen from eqns. (9) and (5), the concentration of NaCl in region 1, C;, is a function of Ai, the interfacial concentration of dissolved Cl,. Therefore the values of At and C; were obtained by solving eqns. (9) and (10) together with eqns. (5) and (6) by trial and error. Estimated values of the physical properties for the present system are listed in Table 1. Comparison of experimental results with theory The present experimental data and the data of Spalding and Takahashi et al.4 are shown in Fig. 6 as a log-log plot of the reaction factor /3 versus the concentration ratio Be/At. Values of /3were calculated from the measured absorption rates NA using the values of Ai and DA estimated by the abovementioned methods. The solid line in this figure represents the two reaction-plane model (eqns. (4), (5) and (6)), while the chain line (-. - - -) shows

where 1-1 is the viscosity of the solution and T the absolute temperature. The viscosity of the solution was assumed to be the same as that of aqueous NaCl solution of concentration equal to the sum of the concentrations of HOC1 and NaCl, i.e. C1 + C;, ignoring the unknown effect of the presence of HOC1 on the viscosity. The ratios of the liquid phase diffusivities of OH-, HOCl, Cl- and OCl- to the diffusivity of Cl,, i.e.

82 ii - I 60 _ Symbol
40 0 . A

H. HIKITA

et al.

I
Temp. 3OC 2 5 C 25 C

II

/ /-

Investigator Present work Spalding Takahashi et a

Single

reaction,

10 0 6

diffusivity of OH- ions is different from the salt diffusivity of NaOH due to the presence of Cl- ions. The effective diffusivity of OH- ions estimated by the method of Vinograd and McBain l5 is approximately 90% greater than the salt diffusivity of NaOH. Thus, the good agreement observed by Spalding and Takahashi et al. between the experimental data and their theoretical predictions might be a coincidence due to underestimated values of the diffusivity of OH- ions.

NOMENCLATURE
two reaction-plane model

2
. one reaction-plpne model

A I 2 I L I II 6 810 I 20 I I LO
I

1.01

0.6 1.0

Ai

Ba/Ai
Fig. 6. Reaction NaOH solutions. factor for absorption of Cl2 into aqueous

Aiw

B Bo c, c Cl 3c; G 2,

the one reaction-plane model (eqns. (2) and (3)). Further, the broken line represents the model proposed by Spalding and Takahashi et al. 4 on the assumption that only the first-step reaction (b) takes place instantaneously and irreversibly at a reaction plane, and has been calculated from eqn. (2) and the equation

(12) which is identical with eqn. (3) except that 2 in the right-hand side of eqn. (3) is replaced by unity. As seen in Fig. 6, measured values of b are in good agreement with the predictions from the two reactionplane model, and do not support the absorption mechanism based on the one reaction-plane model or the model proposed by Spalding and Takahashi et al. As mentioned earlier, Spalding and Takahashi et aL4 showed that values of the reaction factor for the Cl,-NaOH system were in good agreement with the theoretical predictions based on the model they proposed. In calculating theoretical values of 0, they used the salt diffusivity of NaOH as Dg, and not the effective diffusivity of OH- ions. The use of the salt diffusivity for Dg, however, is considered unreasonable in the present problem, because the

DB DC, D,?
DE

E Eo I i+, i-, zg

K1
K2 K3

concentration of solute gas A in solution, g mol/Q interfacial concentration or physical solubility of solute gas A in solution, g mol/P physical solubility of solute gas A in water, g mol/Q concentration of reactant B in solution, g mol/Q concentration of reactant B in bulk of solution, g mol/Q concentrations of intermediate products C and C, respectively, in solution, g mol/Q concentrations of intermediate products C and C, respectively, in region 1, g mol/Q concentration of intermediate product C in bulk of solution, g mol/Q concentration of ions, g ion/Q liquid phase diffusivity of solute gas A, cm2 /set liquid phase diffusivi ty of reactant B, cm2/sec liquid phase diffusivities of intermediate products C and C, respectively, cm2/sec liquid phase diffusivity of final product E, cm*/sec concentration of final product E in solution, g mol/Q concentration of final product E in bulk of solution, g mol/Q ionic strength of solution, g ion/Q contributions of positive ion, negative ion and solute gas, respectively, to salting-out parameter, Q/g ion equilibrium constant of reaction (b) equilibrium constant of reaction (c), Q/g mol equilibrium constant of reaction (a), (g moW2)

ABSORPTION ks
NA

OF CHLORINE

83 When gas absorption is accompanied by a two-step instantaneous reaction, A + B + C + C followed by C + B =+E, where the equilibrium constants of the first and second steps, K1 and K2, respectively, and also their ratio P (=KI/K2)are all very large, two reaction planes will be formed in the liquid and the concentration profile for each species will be similar to that in Fig. 1. Then the basic differential equations describing the diffusion of all species, based on the penetration theory, are

NA*

P T
t
X X1,X2

zi

salting-out parameter, Q/g mol or Q/g ion average rate of absorption with chemical reaction, g mol/cm* set average rate of absorption without chemical reaction, g mol/cm* set equilibrium-constant ratio, K1/K2 absolute temperature of solution, K exposure time of liquid to gas, set distance from interface into liquid, cm distances from interface to first and second reaction planes, respectively, cm valency of ions

a*A
DA,,z=,,

aA
O<x
_?

<xl

(1A)

Greek symbols reaction factor, NA/NA* P viscosity of solution, g/cm set P (3 parameter defined by eqn. (3) or eqn. (12) parameters defined by eqns. (5) and (6) (51>02

D B

ax2

at
ac

x2 <x <m

(2A)

DC&T=%
REFERENCES 1 2 3 4 5 6 I 8 9 10 11 12 13 14 15 16 C. W. Spalding, A.I.Ch.E.J., 8 (1962) 685. P. V. Danckwerts, Trans. Faraday Sot., 46 (1950) 300. T. K. Sherwood and R. L. Pigford, Absorption and Extraction, 2nd edn, McGraw-Hill, New York, 1952. H. Takahashi, M. Hatanaka and R. Konaka, Can. J. Chem Engrs, 45 (1967) 145. H. Hikita, S. Asai and T. Takatsuka, Chem. Eng. J., 4 (1972) 31. A. E. Connick and Y. T. Chia, J. Am. Chem. SOC., 81 (1959) 1280. J. C. Morris, J. Phys. Chem., 70 (1966) 3798. M. Eigen and K. Kustin, J. Am. Chem. Sot., 84 (1962) 1355. A. Lifshitz and B. Perlmutter-Hayman, J. Phys. Chem., 65 (1961) 753: 66 (1962) 701. P. L. T. Brian, j. E. Vivian and C. Piazza, Chem. Eng. Sci., 21 (1966) 551. H. Hikita and S. Asai, Kagaku Kogaku, 28 (1964) 1017. H. Hikita, S. Asai, Y. Himukashi and T. Ishiwara, Kagaku Kogaku, 32 (1968) 278. C. Bohr, Ann. Phys., 68 (1894) 500. D. W. Peaceman, SC. D. Thesis, M. I. T., Mass., 1951. J. R. Vinograd and J. W. McBain, J. Am. Chem. Sot., 163 (1941) 2008. D. W. van Krevelen and P. J. Hoftijzer, Chim. Ind. XXIme Conrr. Intern. Chim. Ind., Sot. Chim. Ind., Paris, 1948, p. 168. E. Oliveri-Mandala, Gazz. Chim. Ifal., 50 (1920) II 89. N. Yokota, Kagaku Kogaku, 22 (1958) 476. 1

a*c

O<x

<x2

(3A)

DC' ax2 5 a*E

DES=%

a*c' ac'
aE

O<x<m

(4A)

x1

<x<m
conditions:

(5A)

with the following initial and boundary

x = 0, t > 0;
A=A~,

-_=-=O

ac ac' ax ax

x=x1,t>O;

A=E=O

x=x*,

t>O;

B=C=O Dgg=-DCE=DF
x=-,

17 18

[(3*

- (Z)jA)

APPENDIX

t>O;

The concentration profiles of the intermediate products of a two-step instantaneous reaction, A + B + C + C followed by C + B + E, were calculated as follows.

B=Bo,
Solution

C'=C&

E=Eo
gives the concentration

(9A)

to these equations

84
II I I I I I1 II

H. HIKITA
11

et al.

TABLE

20
c 10 ~~_ 0 nt E r!n
0 x

Symbol
0

Temp.
~~~

Investigator

Values of k, and & for Clz-aqueous

0

NaCl solution system

s (g

. d .

4OC 50C

Temperature (0 20 30 40 50

ks (Q/g ion) 0.1005 0.0903 0.0854 0.0793

ion/Q)

-0.0145 PO.0247 --0.0296 -0.0357

-z
_
0 1 2 3
I

log (Ai/Ai,)

= - k,I

(1 W

1
solutions.

g-ion/l

Fig. 7. Physical solubility of Cl2 into aqueous

NaCl

where Ai and Ai, are the physical solubilities in the solution and in water, respectively, and I is the ionic strength of the solution defined by
I =:

profiles in the liquid as functions of the exposure time. The expressions for the concentration profiles of all species except species C are the same as those obtained in the previous paper5 for the case of a twostep reaction, A + B * C and C + B + E. Therefore only the expressions for the species C are shown here: o<x<x,:

CCiZi2

(124

ci being the concentration of ions of valency zi. The coefficient k, is the salting-out parameter and is expressed as the sum of the contributions due to the positive and negative ions present and the dissolved gas:
k, = i, + i_ + ig

(13A)

APPENDIX

Fhe physical solubility of Cl2 in aqueous NaCl solutions was estimated as follows. The physical solubility of a gas in aqueous electrolyte solutions can be correlated by the following expression16 :

The values of i for various ions and gases were reported by van Krevelen and Hoftijzer16. However, since the value of ig in the present case (contribution due to Cl;?) is not known, it was estimated by analysing the data of Oliveri-Mandala17 and Yokota18 for the total solubility of Cl2 in aqueous NaCl solutions. The physical solubility of Cl*, i.e. A i, was calculated from the total solubility data, correcting for the effect of the hydrolysis reaction (a) by use of the values for the hydrolysis equilibrium constant reported by Connick and Chia6. The results obtained are shown in Fig. 7 as a semi-logarithmic plot ofAi against the ionic strength I of the solution. This figure indicates that the physical solubility of Cl2 in aqueous NaCl solutions is well correlated by eqn. (11A). The values of k, were calculated from the slopes of the straight lines in Fig. 7. From these values of ks and the contributions of Na+ and Clions16, the values of i, for Cl2 at several temperatures were obtained. The estimated values of i, and k, are given in Table 2.