United States                     EPA-600/R-01-054
            Environmental Protection                 August 2001
            Agency                         °
&EPA     Research and
            Development
             Chlorine Absorption
             In S(IV) Solutions
            Prepared for
            Office of Research and Development
            Prepared by

            National Risk Management
            Research Laboratory
            Research Triangle Park, NC 27711

-------
                                 Foreword
      The U.S. Environmental Protection  Agency is  charged by  Congress  with
protecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and  implement actions leading to
a compatible balance  between human activities and the ability of natural systems to
support and nurture life. To meet this mandate, EPA's research program is providing
data and technical support for solving environmental problems today and building a
science knowledge base  necessary to manage  our ecological resources wisely,
understand how pollutants affect our health, and prevent or reduce environmental risks
in the future.

      The National Risk Management Research Laboratory (NRMRL) is the Agency's
center for investigation of technological and  management approaches for preventing
and reducing risks from pollution that threaten human health and the environment.  The
focus of the Laboratory's research program is on methods and their cost-effectiveness
for prevention and control  of pollution to air, land,  water, and subsurface resources,
protection of water quality in public water systems; remediation of contaminated sites,
sediments and  ground  water; prevention and control  of  indoor air  pollution;  and
restoration of ecosystems.  NRMRL collaborates with both public and private sector
partners to foster technologies that reduce the cost of compliance and to anticipate
emerging problems. NRMRL's research provides solutions to environmental problems
by: developing and promoting technologies that protect and  improve the environment;
advancing scientific and engineering  information  to support regulatory and policy
decisions; and  providing the technical support  and information transfer to ensure
implementation  of environmental regulations and strategies  at the national, state, and
community levels.

      This publication has  been  produced as part of the Laboratory's  strategic
long-term  research plan.  It  is published and made available  by  EPA's Office of
Research and Development to assist the user community and to link researchers with
their clients.
                                 E. Timothy Oppelt, Director
                                 National Risk Management Research Laboratory

                           EPA REVIEW NOTICE

     This report has been peer and administratively reviewed by the U.S. Environmental
     Protection Agency, and approved for  publication.  Mention of trade names or
     commercial products does not constitute endorsement or recommendation for use.

     This document is available to the public through the National Technical Information
     Service, Springfield, Virginia 22161.

-------
                                                      EPA-600/R-01-054
                                                      August 2001
Chlorine Absorption in S(IV)  Solutions
                                 By
                     Sharmistha Roy and Gary T. Rochelle
                     Department of Chemical Engineering
                       The University of Texas at Austin
                            Austin, TX 78712
                U.S. EPA Cooperative Agreement CR 827608-01-1
                          U.S. EPA Project Officer
                            Theodore G. Brna
                 National Risk Management Research Laboratory
                  Air Pollution Prevention and Control Division
                      Research Triangle Park, NC 27711
                              Prepared for:
                  United States Environmental Protection Agency
                      Office of Research and Development
                          Washington DC 20460

-------
                                   Abstract

The rate of chlorine (Cb) absorption into aqueous sulfite/bisulfite [S(IV)] solutions was
measured at ambient temperature using a highly characterized stirred cell reactor. The
reactor media were 0 to 10 mM S(IV) with pH ranging from 3.5 to 8.5.  Experiments
were performed using 20 to 300 ppm Cbin nitrogen (TS^).  Chlorine absorption was
modeled using the theory of mass transfer with chemical reaction. Chlorine reacts
quickly with S(IV) to form chloride and sulfate. Chlorine absorption is enhanced by
increasing pH and S(IV) concentration. The rate constant for the reaction of chlorine with
S(IV) was too rapid to be precisely measured using the existing stirred cell reactor, due to
mass transfer limitations. However, the most probable value of the rate constant was
determined to be 2 x 109 L/mol-s.

These results are relevant in the simultaneous removal of chlorine, sulfur dioxide (862),
and elemental mercury (Hg) from flue gas. The developed model shows that good
removal of both chlorine and mercury should be possible with the injection of 1 to 10
ppm chlorine to an existing limestone slurry scrubber.  These results may also be
applicable to scrubber design for removal of chlorine in the pulp and paper and other
industries.

-------

-------
                                             Tables

Table 7-1. Parameters for gas film mass transfer correlations	16
Table 7-2. Diffusivity for species used to correct k°L correlations	18
Table 8-1. Initial chlorine absorption results, electrochemical sensor analyzer	19
Table 8-2. Chlorine absorption in pH 4-4.5 S(IV) with incorrect analyzer calibration of electrochemical
     sensor analyzer	20
Table 8-3. Chlorine absorption in buffered S(IV) using multi-point calibration	21
Table 8-4. Electrochemical chlorine analyzer calibration after fixing electrolyte level	22
Table 8-5. Chlorine absorption in S(IV) solutions using improved calibration of the electrochemical
     analyzer, 1.29L/min gas, 50 mM buffer, kg = 0.6 mol/s-atm-m2	23
Table 8-6. Chlorine absorption in pH 4.3 - 4.5 S(IV), measured with electrochemical analyzer, 1.2 L/min
     gas, 50 mM succinate buffer, and kg = 0.60 - 0.66 mol/s-atm-m2	24
Table 8-7. Chlorine absorption measured with electrochemical analyzer in pH 4.5 S(IV), Cl2,in= 21 ppm,
     1.19 L/min gas, 50 mM succinate buffer	24
Table 8-8. Chlorine and oxygen absorption measured  with IMS analyzer in pH 4.5 S(IV), 50 mM succinate
     buffer, and 1.15 L/min gas	25
Table 8-9. S(IV) depletion resulting from oxygen absorption in pH 4.5 S(IV) at ambient temperature, 50
     mM succinate buffer, and 1.15 L/min gas	26
Table 8-10. Chlorine absorption in pH 4.5 S(IV), 50 mM succinate buffer, 1.15 L/min gas	27
Table 8-11. Chlorine absorption as  a function of buffer concentration at ambient temperature in pH 4.5 with
     1.15 L/min gas	28
Table 8-12. Chlorine absorption in pH 4.5 S(IV), 5 mM buffer, 1.18 L/min gas	29
Table 8-13. Chlorine absorption in S(IV) solutions with 5 mM buffer at various agitation rates	29
Table 9-1. Values of parameters in global model	31
Table 9-2. Flux variance with liquid agitation rate	36
Table 9-3. Parameters used to predict mercury removal	44
Table 9-4. Mercury penetration in limestone slurry scrubber at 25°C	45
Table 9-5. Mercury penetration in limestone slurry scrubber at 55°C	45
Table A-l. Gas film mass transfer coefficient (kg), IMS analyzer	49
Table A-2. Gas film mass transfer coefficient (kg), electrochemical sensor  analyzer	49
Table A-3. Gas film mass transfer coefficient (kg), electrochemical analyzer after modifications to reduce
     scatter	50
Table B-l. Data used to determine k°cl  correlations	51
Table B-2. Physical liquid film mass transfer coefficient for chlorine	56
                                                 IV

-------
                                             Figures

Figure 1-1. Chlorine injection for Hg removal in limestone slurry scrubbing	1
Figure 6-1. Stirred cell reactor apparatus	10
Figure 7-1. Data and correlations for gas film mass transfer coefficient	16
Figure 7-2. Data and correlations for physical liquid film mass transfer coefficient	18
Figure 8-1. Electrochemical chlorine analyzer multi-point calibration	21
Figure 8-2. Improved calibration of electrochemical sensor analyzer	23
Figure 9-1. Chlorine absorption in buffered S(IV), k2,s(iv)= 2 x 109 L/mol-s	32
Figure 9-2. Chlorine penetration in buffered S(IV), k2jS(IV) = 2 x 109 L/mol-s	34
Figure 9-3. Chlorine penetration in buffered S(IV), k2jS(IV) = °°	35
Figure 9-4. Chlorine penetration in buffered S(IV), k2iS(IV) = 2.5 x 108 L/mol-s	35
Figure 9-5. Data (Table 9-2 Series B4) limited by gas film mass transfer coefficient	37
Figure 9-6. Chlorine absorption in succinate buffer with chlorine inlet of 264 ppm	38
Figure 9-7. Chlorine absorption in succinate buffer with chlorine inlet of 21 ppm	39
Figure 9-8. S(IV) oxidation by 275 ppm chlorine and 14.5% oxygen	40
Figure 9-9. S(IV) oxidation by 21 ppm chlorine and 20.5% oxygen	40
Figure 9-10. Electrochemical analyzer data overlaid onto IMS data	42
Figure 9-11. Chlorine absorption in 0 -2 mM S(IV) in 50 mM buffer using electrochemical analyzer	43
Figure 9-12. Effect of chloride seen from data obtained using electrochemical analyzer	43
Figure 9-13. Predicted mercury penetration	46
Figure B-l. Extracting k°LjC12at 729 rpm	53
Figure B-2. Extracting k°LjC12 at 305 rpm 	53
Figure B-3. Extracting k°LjC12 at 504 rpm 	53
Figure B-4. Extracting k°LjC12 at 734 rpm	54
Figure B-5. Extracting k°LjC12 at 699 rpm	54
Figure B-6. Extracting k°LjC12at228 rpm	54
Figure B-7. Extracting k°LjC12 at 600 rpm	55
Figure B-8. Extracting k°LjC12 at 306 rpm	55
Figure B-9. Extracting k°LjC12at 514 rpm	55
Figure B-10. Extracting k°LjCi2 at 718 rpm	56

-------
                                 Nomenclature

A     gas/liquid contact area (m2)
A"     generic anion
Ccl2   concentration of chlorine in liquid (mol/L = M = kmol/m3)
[Cl~]   concentration of chloride in liquid (M)
DC1   diffusion coefficient for chlorine in water (m2/s)
DHS   diffusion coefficient for mercury in water (m /s)
DS(iv)  diffusion coefficient for S(IV) in water (m2/s)
E     enhancement factor (dimensionless)
FC    mass flow controller
O     reactant stoichiometric coefficient (dimensionless)
G     gas flow rate to reactor (m3/s)
HA   generic acid
Hcl2   Henry's law constant for chlorine (atm-m3/kmol)
HHg   Henry's law constant for mercury (atm-m3/kmol)
IMS   ion mobility spectrometry
K     equilibrium constant
kg     individual gas film mass transfer coefficient (kmol/s-atm-m )
kL;ci2  individual physical liquid film mass transfer coefficient for chlorine (m/s)
KOG   overall gas phase mass transfer coefficient (kmol/s-atm-m2)
ki,H2o  first order rate constant for chlorine hydrolysis reaction (s"1)
       second order rate constant for chlorine/buffer reaction (L/mol-s)
k2,Hg   second order rate constant for mercury/chlorine reaction (L/mol-s)
k2,s(iv) second order rate constant for chlorine/S(IV) reaction (L/mol-s)
k2,oH   second order rate constant for chlorine/hydroxide reaction (L/mol-s)
Ncl2   flux of chlorine (kmol/m2-s)
Ng     number of gas phase mass transfer units, defined as kgA/G (dimensionless)
ng     gas phase agitation rate (rpm)
nL     liquid phase agitation rate (rpm)
Pcl2   partial pressure of chlorine (atm)

P     partial pressure of chlorine in equilibrium with chlorine in bulk liquid (atm)
pKa   negative logarithm of acid dissociation constant
R     gas constant (8.205 x 10"5 m3-atm/mol-°C)
t       time (s)
T     temperature (°C)
                                        VI

-------
V     reactor volume (m3)
yng    mole fraction of mercury in the gas phase (dimensionless)

Subscripts

b      in bulk
i      at gas/liquid interface
in     inlet
init    initial
out    outlet
T     total
                                         Vll

-------
1.  Introduction
Mercury (Hg) pollution is an important problem because of its behavior in the
environment (bioaccumulation) and the potential for deleterious health effects. Roughly
85% of anthropogenic mercury emissions are from combustion sources (Keating et al.,
1997).  The flue gas from these sources contains sulfur dioxide (SO2) and hydrogen
chloride (HC1) at much higher concentrations than the mercury compounds. Aqueous
scrubbing is currently used to remove SO2 and HC1 from these flue gases. It should be
possible to remove Hg by conventional aqueous scrubbing technologies with the addition
of reagents to produce chlorine, which will oxidize the Hg to a more soluble form
through reaction in the mass transfer boundary layer. Some researchers, such as Zhao
(1997) and Livengood and Mendelsohn (1997), have had success in removing Hg via
reactions with chlorine compounds. Mercury reacts with chlorine to form mercuric
chloride, HgCli, which is very soluble (Ernst et al., 1997) and can thus be easily removed
through aqueous scrubbing. Figure 1-1 depicts the process in a limestone slurry scrubber.
                 Exit Gas
      HoSCX/NaOCl
      Flue Gas
 1000ppmSO2
  1 ppb Hg
                  C12|
                 (lOppm)
                                 /Y7\
                                     7\
                                 /\7\
Hold Tank
                              Air
                    pH5-6

                    0.1-10 mM S(IV)
                    pH3-5

                    2-10 mM S(IV)
                                               .> CaSO4(s)
Figure 1-1. Chlorine injection for Hg removal in limestone slurry scrubbing

In the proposed technology, either hypochlorite solution will be sprayed into the scrubber
to generate chlorine in-situ or chlorine gas (< 10 ppm) will be directly injected into the
gaseous feed as shown in Figure 1-1. The chlorine cannot be introduced with the bulk
solution. If an oxidant were put in the bulk solution, it would be completely depleted by
reaction with dissolved S(IV).  Hypochlorite will release Cl2 upon acidification by
absorption of SCh and HC1 or by addition of sulfuric acid. The C12 should react with
elemental Hg in the solution at the gas/liquid interface and should greatly enhance the
rate of absorption of Hg.  The mercury will be oxidized and absorbed into the scrubber
solution. The chlorine will also react at the gas/liquid interface with any elemental Hg
formed by sulfite reduction of HgCl2 in the bulk solution.

-------
The success of this approach requires that the mercury react with the chlorine before it
gets reduced by the dissolved SO2 present as S(IV).  S(IV) represents sulfur in the +4
oxidation state (sulfite and bisulfite).  Therefore, the kinetics of the reaction between
chlorine and S(IV) need to be quantified to ensure that the chlorine will be available to
react with the  Hg. Measuring the reaction rate of chlorine with S(IV) is the topic of this
report.

-------
2. Conclusions

The rate constant for the Cl2/S(IV) reaction was too rapid to be precisely measured using
the stirred cell reactor, due to mass transfer limitations.  However, the most probable
value for the rate constant was determined to be 2 x 109 L/mol-s.  At low S(IV), the
chlorine absorption was limited by the buffer-enhanced hydrolysis reaction.  At
moderate S(IV), it was limited by diffusion of S(IV) from the bulk solution to the
interface.  At high S(IV), the absorption was limited by diffusion of chlorine from the
bulk gas to the interface.

Chlorine injection to enhance mercury removal may be  a feasible process.  In a typical
limestone slurry scrubber, chlorine absorption will be gas film controlled because of the
rapid Ci2/S(IV) reaction rate.  Thus, 99 to 99.99% chlorine removal will be achieved in
typical scrubbers. Also, there will be enough chlorine at the interface to react with
mercury. The model shows that only 1 ppm chlorine is needed to get 99% mercury
removal.

The succinate buffer enhances chlorine absorption.  However, lowering the succinate
buffer concentration did not aid in extracting kinetics  because there  is not much of a
range between the chlorine flux due to absorption in water and the maximum flux
resulting from complete gas film control.  Therefore, extracting kinetics for the S(IV)
reaction will always be difficult in the existing apparatus. On one end, absorption is
limited by the chlorine hydrolysis reaction, and on the other end, it is limited by gas film
control in the stirred cell contactor.

Chloride does not affect chlorine absorption in S(IV) since the chlorine/S(IV) reaction is
irreversible. Oxygen does not affect chlorine absorption in S(IV) either, nor does it seem
to catalyze S(IV) depletion at the ranges investigated.

-------
3. Recommendations

In order to accurately predict mercury and chlorine removal in a scrubber, a better model
with precise kinetics is needed. The Ci2/S(IV) reaction rate needs to be precisely
measured in a gas/liquid contactor with higher mass transfer coefficients. Furthermore,
this reaction rate should be measured at 55°C to simulate a typical limestone slurry
scrubber.

Simultaneous absorption of mercury and chlorine must be measured and modeled to
obtain a precise value for k2,Hg. These experiments should also be done at 55°C.
Simultaneous absorption of Hg, Ci2, and SOi should also be studied. Furthermore, in
order to completely simulate flue gas, CCh, NOX, and Ch should be added to the inlet gas.

Results have shown that chloride does not affect chlorine absorption. However,
experiments were not done in sodium chloride solutions higher than 0.02 M. Limestone
slurry may have 1 M Cl".  Thus, absorption into  1 M chloride must be quantified.

-------
4. Literature Review

4.1 Mercury removal with chlorine

Previous researchers have performed screening experiments on mercury removal through
reaction with chlorine oxidants.  Zhao (1999) investigated Hg absorption in
hypochlorite/chloride solutions and showed that low pH, high temperature, and high Cl"
concentration favored Hg absorption.  In aqueous hypochlorite  solution, the distribution
of OC1", HOC1, and Cl" depends on solution pH and [Cl"]. Since lower pH results in
higher Hg removal, it is probable that free Cl2 is the active species that reacts with Hg.
The activity of free Cl2 can be obtained from the  following two equilibria:

                                                OC1"                         (4-1)
                                                                             (4-2)

Thus, at low pH (high H+) and high Cl", the formation of C12 is favored.  The chlorine
reacts with the Hg to form HgCli in an apparent overall second order reaction and greatly
enhances the  rate of Hg absorption. The rate constant was obtained from modeling the
Hg absorption using surface renewal theory for mass transfer with fast chemical reaction
in the boundary layer. The rate constant measured by Zhao in hypochlorite solutions at
pH 9 to 11 was 1.7 x 10*5 L/mol-s at 25°C and 1.4 x 101? L/mol-s at 55°C. Furthermore,
preliminary experiments with simultaneous absorption of chlorine and Hg have
demonstrated that 1 to 10 ppm of chlorine can be effective in removing 0.1 ppm Hg
(Zhao, 1999).

Fedorovskaya et al. (1979) said that Hg removal with an acidic  chlorine-containing
solution can be represented by two mechanisms:  (1) chlorine from the solution is swept
into the gas phase where it oxidizes the Hg and (2) Hg diffuses  from the gas into the
solution and reacts with the chlorine. Their experiments showed that when the
chlorine/mercury ratio is less than 20:1, mercurous chloride is the primary product
formed. At ratios greater than 20:1, the oxidation of Hg with chlorine yields mercuric
chloride. This reaction takes place rapidly (in 1-2 seconds). Fedorovskaya et al. (1979)
also showed that oxidation of Hg can occur in alkaline medium in the presence of
hypochlorite/chloride. They found that the oxidation of Hg is still fast under these
conditions, but the reaction is twice as fast in  acid because of higher oxidizing potentials
in acid.

Mercury is also known to react with Cli in the gas phase with a reaction rate constant of 2
x 105 L/mol-s at 20°C (Hall, 1992). Hall's experiments showed that the reaction rate was
relatively independent of temperature from 20°C to 700°C. Thus, the apparent activation
energy is probably not greater than 10 kJ/mol. Mercury removal via gas phase reaction
with chlorine can be quantified using this rate constant and a typical commercial gas
phase residence time of 2 seconds. If the gas  inlet were 1 ppb Hg and 10 ppm chlorine,
0.84 ppb Hg would exit the scrubber.  Therefore, gas phase reaction with chlorine is not
enough to remove Hg.

-------
4.2 Chlorine absorption in aqueous solutions

There are several reactions which contribute to chlorine absorption in aqueous solutions.
At a pH of 3 to 10.5 (with no S(IV) present), chlorine hydrolysis to form hypochlorous
acid and hydrochloric acid is the dominant reaction controlling chlorine absorption
(Spalding, 1962). This reaction is relatively slow, ki>H2o = 15.4 s"1 at 25°C (Brian et al.,
1966). Under typical limestone slurry scrubber conditions, if 10 ppm C12 were injected
into the gaseous feed, 8.5 ppm C12 would exit if chlorine absorption in water were the
only reaction enhancing chlorine absorption.  Therefore, the chlorine hydrolysis reaction
alone will not cause the chlorine at the gas-liquid interface to be depleted.

Chlorine hydrolysis can be enhanced by the presence of buffer anions (Wang and
Maregerum, 1994; Lifshitz and Perlmutter-Hayman,  1962). The following overall
reaction occurs:

                         C12 + H2O + A" ^ HOC1 + Cl" + HA                   (4-3)

The kinetics of this reaction have been studied for the following anions (A"): acetate,
chloroacetate, formate, and phosphate (Lifshitz and Perlmutter-Hayman, 1962).

At a pH greater than 10.5, the chlorine/hydroxide reaction shown below is the dominant
reaction:

                          C12 + 2OH" -^ Cl" + OC1" + H2O                      (4-4)

(Spalding, 1962). This reaction is relatively fast, k2;on = 1.57 x  109 L/mol-s (Ashour et
al., 1996); thus, at high pH, there will be no chlorine  at the gas/liquid interface.

Chlorine can also react with sulfite to form chloride and sulfate (Askew and Morisani,
1989; Gordon et al., 1990).

                      C12 + H2O + SO3"2 -» 2C1"+  SO4"2 + 2H+                  (4-5)

This reaction has not been studied much; thus, the kinetics have not been quantified.

4.3 Reactions ofS(IV) with chlorine oxidants

Even though the C12/S(IV) reaction kinetics have not been studied, researchers have
investigated S(IV)  reactions with various chlorine oxidants. Hypochlorous acid (HOC1)
and hypochlorite (OQ~) can react with sulfite (SOs"2) (Fogelman et al., 1989).
Hypochlorous acid reacts as follows:

                           HOC1 + SO3"2 -* OH"+ C1SO3"                      (4-6)
                         C1SO3" + H2O -^ Cl"+ SO4"2 + 2H+                    (4-7)

The first reaction has a rate constant of 7.6 x 108 L/mol-s at 25°C and ionic strength of
0.5, but the rate limiting step is the second reaction, which has a rate constant of 270 s"1.

-------
Hypochlorite reacts with sulfite as shown below:

                             ocr + so3"2 -»cr+ so4"2                        (4-8)

The rate of oxidation of sulfite with HOC1 is more than four orders of magnitude faster
than the rate with OC1". A shift in mechanism is proposed to account for the huge
increase in reactivity (Fogelman et al., 1989).

Suzuki and Gordon (1978) investigated the reaction of chlorine dioxide (C1O2) with
S(IV) in basic solutions, where S(IV) represents sulfur in the +4 oxidation state
(primarily sulfite and bisulfite). The overall stoichiometry is:

                 nC!O2 + mS(IV) -+ pC!O2" + qC!O3" + rCl"+ mSO4"2            (4-9)

The coefficients n, m, p, q, and r depend  on both the pH and the specific buffer solution
used.

Reactions of S(IV) with chlorine oxidants are important in water and wastewater
treatment.  Here, sulfite is used to deplete residual chlorine (such as chloramines and
chloropeptides) which remains after water disinfection using chlorine.  A
monochloropeptide reacts with sulfite as  shown (Jensen and Helz, 1998).

          CINH-peptide + SO3"2 + 2H2O -* NH2-peptide + Cl"+ SO4"2 + H3O+    (4-10)

Jensen and Helz (1998) say that usually bisulfite (HSO3~) is a much poorer reducing agent
than sulfite (SO3~2).
than with bisulfite.
than sulfite (SO3~2). So, reaction rates of chlorine oxidants are much faster with sulfite
Sulfite is also used to remove chlorite (which can result from using chlorine dioxide as a
disinfectant) from treated water. The reaction is (Gordon et al., 1990):

                           C1O2" + 2SO3"2 -+ Cl"+ 2SO4"2                      (4-11)

-------
5. Mass Transfer with Chemical Reaction

Chlorine absorption into S(IV) solutions occurs by mass transfer with simultaneous
chemical reaction. Chlorine must first diffuse from the bulk gas to the gas liquid
interface with the flux (Na ) given by:

                              Na2=kg(Pa2ib-Pa2.)                          (5-1)

Then, chlorine absorption into the liquid occurs by mass transfer with fast chemical
reaction in the boundary layer with the same flux:

                            XT   _ EkL,Cl2(PCl2,i -PCl2,b)
                            iNci2 -         -                                  P-A>
                                           Hci2

According to surface renewal theory (Danckwerts, 1970), the enhancement factor (E) can
be expressed as (Critchfield, 1988; Shen, 1997; Zhao, 1997):


        E =  1 + -f- (ki,H2o + k 2,buf [buffer]1 + k 2,OH [OH" ], + k2 S(IV) [S(IV)]1 )    (5-3)
which incorporates the reactions which contribute to chlorine absorption. If the
chlorine/S(IV) reaction is the dominant reaction and equilibrium effects are negligible,
then the flux expression simplifies to:

                                pcl     pcl    _
                 Nci2 = Ek°>cl2 — -^ = — ^^Dci2k2;S(IV)[S(IV)]1             (5-4)
                              2 Hci2    Hci2 v

The enhancement factor expression is derived assuming that the chlorine/S(IV) reaction
is first order in chlorine and first order in S(IV). If this model is correct, the extracted
rate constant, k2,s(iv), can be  used to extrapolate chlorine removal at low chlorine
concentrations. The corresponding rate expression is:

                          reaction rate = k2>S(iv) [C12][S(IV)]                     (5-5)

The concentrations, physical properties (diffusivity, D, and Henry's law constant, H) and
rate constants for the water  and hydroxide reactions are known. At 25°C, the Henry's
law constant for chlorine, Hcl2 , was taken to be 16.7 atm-m3/kmol (Brian et al.,  1966),
and the diffusion coefficient for chlorine through water, DC1 , was taken to be 1.48 x 10"9
m2/s (Spalding, 1962). The chlorine flux was determined experimentally from the gas
phase material balance. Thus, the only unknown is the rate constant for the
chlorine/S(IV) reaction. This rate constant, k2,s(iv> can be calculated by substituting the
enhancement factor into the flux equation (5-2).

-------
The interfacial liquid S(IV) concentration is obtained by assuming that C\2 reacts with
S(IV) at the gas/liquid interface.
                    NCi2 = ONs(iv) = Ok°L, S(iv)([S(IV)]b - [S(IV)]i)                (5-6)

                                  ]1 = [S(IV)]b -- ^—                        (5-7)
O represents the stoichiometric relationship between the reactants. For example, in the
reaction C12 + 2OH" -^ Cl" + OC1" + H2O, O = l/2 since 1 mol of C12 reacts with 2 mol of
OH.

The rate constant can only be extracted if mass transfer does not limit the chlorine
absorption.  When the S(IV) concentration is high relative to the chlorine concentration,
the chlorine flux is limited by the resistance in the gas phase, and the flux from Equation
5-1 simplifies to:
                                   Nci2=kgpci2,b                               (5-8)

Under these conditions, there is essentially no chlorine at the interface since all the
chlorine reacts with S(IV) as soon as the chlorine reaches the interface. Thus, the
chlorine absorption only depends on how fast the chlorine diffuses from the bulk gas to
the gas/liquid interface, not on the kinetics.

When the chlorine concentration is high relative to the S(IV) concentration, the flux is
limited by S(IV) depletion at the interface. This means that there is essentially no S(IV)
at the interface since whatever S(IV) diffuses to the interface is readily depleted through
reaction with chlorine. Under these conditions, the flux in Equation 5-6 simplifies to:

                          NCi2 = ONS(iv) = Ok°L, S(iv)[S(IV)]b                      (5-9)

showing that the flux of chlorine is linear with the bulk S(IV) concentration.

The fraction gas film resistance is a parameter used in analyzing some of the data.  The
fraction gas film resistance is directly related to reaction kinetics:

                    K           1
                   — — = - - — — — = fraction gas film resistance              (5-10)
As the enhancement factor (E) increases, which corresponds to fast reaction rates, the
total resistance to mass transfer (1/Koo) becomes limited by gas film resistance (l/kg).
Under these conditions, the fraction gas film resistance (5-10) approaches unity and
becomes independent of reaction kinetics.  Thus,  data which approach gas film resistance
cannot be used for extracting kinetics.  For these data, the gas film mass transfer
coefficient (as opposed to kinetics) is being measured.

-------
6. Experimental Apparatus and Methods

All chlorine absorption experiments were performed at ambient temperature (22 to 25°C)
in the well-characterized stirred cell contactor with Teflon surfaces shown in Figure 6-1
(Zhao and Rochelle, 1999; Zhao, 1997). Teflon tubing, fittings, and valves were used for
all the connections. Mass flow controllers are labeled as "FC."
  Cl2/N2
                                                         Rotameter
                                                                 Hood
Figure 6-1. Stirred cell reactor apparatus

6.1 Description of stirred cell reactor apparatus

The stirred cell contactor allowed gas/liquid contact, for which mass transfer properties
were known or measured, at a known interfacial area (A) of 8.1 x 10~3 m2. The
cylindrical reactor had a 0.01 m inner diameter and 0.016 m height. The reactor vessel
consisted of a thick glass cylinder with Teflon-coated 316 stainless steel plates sealed to
the top and bottom by thick gasket clamps. Four equally-spaced, Teflon-coated, 316
stainless steel baffles were welded to the bottom plate.  The length of the baffles was long
enough to extend to the main body of the gas phase. The bottom plate contained ports for
liquid inlet and outlet. The top plate contained ports for the gas inlet and outlet, solution
injection, and pH probe. The total volume of the reactor was 1.295 x 10"3 m3.

The stirred cell contactor was equipped with Teflon-coated independently controlled
agitators for gas and liquid phase mixing.  Each agitator was driven by a Fisher StedFast
Stirrer (Model SL 1200).  The gas inlet was at the near center of the top  plate, directly
above the gas agitator blade, to ensure that the inlet gas was properly mixed. Gas  and
liquid agitation speeds were measured using a tachometer.  The mass transfer coefficients
(kg, !CL) were a function of the agitation rates.
                                        10

-------
6.2 Gas source and flow path

Gas feed was prepared by quantitatively mixing 0.1% (1000 ppm) Cli (in NI) with
nitrogen. The flow rates of all gas streams were controlled by Brooks mass flow
controllers. The synthesized gas stream, typically at a flow rate of 1.2 L/min, was
continuously fed to the reactor.  After exiting the reactor, the gas stream was diluted with
house air and continuously analyzed for chlorine. An empty 125-mL Erlenmeyer flask
was connected after the reactor outlet to capture any water vapor or liquid. Since this
flask stayed empty throughout an experiment, no liquid exited the reactor through the gas
outlet. When the chlorine concentration to the reactor was less than 30  ppm,
approximately 3  L/min of dilution air was used. When the chlorine concentration was
greater, 36 L/min of dilution air was used.  The chlorine analyzer output was connected
to a strip chart recorder.  The flux of chlorine (rate of chlorine absorption) was calculated
from the gas phase material balance. An analyzer with an electrochemical sensor
(NOVA Model 540P) was initially used. Later experiments were done  using ion mobility
spectrometry (Molecular Analytics AirSentry 10-C12).

6.3 Analyzer calibration

The chlorine analyzer was calibrated at the beginning and end of each experimental series
to check for analyzer drift.  There was essentially no drift for the IMS analyzer.  During
calibration, the gas flow rate was identical to that in an experiment. Other than bypassing
the reactor, the gas flow path during calibration was the same as during an experimental
run. To calibrate the analyzer zero, nitrogen (without chlorine) was supplied and diluted
with house air. To calibrate the span, the gas flow rates were  adjusted to give different
chlorine concentrations spanning the range of interest. For the later experiments (those
analyzed by IMS), both dry gas and wet gas calibrations were done to ensure that
moisture did not affect the analyzer reading. Wet gas calibrations were done by having
the nitrogen (not the chlorine) go through the stirred cell reactor filled with aqueous
solution. For chlorine concentrations less than  30 ppm, there  was a slight difference in
wet and dry gas calibrations, probably resulting from the  increased humidity of the gas.
Therefore, wet calibrations were always used whenever experiments were done in which
the chlorine concentration would be less than 30 ppm.

6.4 Electrochemical analyzer

The analyzer contains an electrochemical sensor with a platinum measuring electrode and
silver reference electrode. The electrolyte used is a 3% lithium chloride solution. The
electrolyte continuously weeps over the active surface of the sensor. When the chlorine
contacts the electrochemical sensor, it  reacts to form silver chloride (AgCl) which
releases two electrons. The current produced is proportional to the chlorine
concentration. Nitrogen dioxide (NOi) and sulfur dioxide (SOi)  interfere with the
electrochemical sensor analysis.
                                        11

-------
6.5 IMS analyzer

The analyzer is based on ion mobility spectrometry (IMS), similar in principle to time-of-
flight mass spectrometry. The sample is passed over a semi-permeable membrane
through which the chlorine diffuses. Purified dry instrument air (supplied externally)
sweeps the chlorine from the interior of the membrane and into an ionization region
supplied with a (3" source  (Ni63).  The ionized molecules then drift through a cell under
the influence of an electric field.  An electronic shutter grid allows periodic introduction
of the ions into a drift tube where they separate based on charge, mass, and shape.
Smaller ions move faster  than larger ions through the drift tube and arrive at the detector.
The current created at the detector is amplified, measured as a  function of time, and a
spectrum is generated.  A microprocessor evaluates the spectrum for the chlorine and
determines the concentration based on peak height.

The IMS analyzer is linear throughout the entire range and can detect chlorine at very
low concentrations. The IMS analyzer also has much better repeatability than the
electrochemical sensor analyzer.  The following do not interfere with the chlorine
analysis: COi, Hg, SOi, NO, hydrocarbons, and chlorinated hydrocarbons. However,
NC>2 and HC1 do interfere with the IMS analysis due to peak overlap at high
concentrations.

6.6 Absolute chlorine analysis through wet chemical methods

The absolute gas phase chlorine concentration was measured using a wet chemical
method. During these analyses, the gas flowed from the cylinder through the same tubing
it would normally flow through in an experiment.  The absolute chlorine concentration
was measured (instead  of relying on that of the gas supplier) since the analysis would
yield the actual chlorine concentration that the reactor sees by  accounting for chlorine
loss between the cylinder and reactor (such as adsorption in tubing).  At the beginning
and end of each experiment, the tubing was flushed with nitrogen.  Since no chlorine was
detected by the analyzer under these conditions, chlorine desorption from tubing did not
occur.

For many of the data analyses, absolute values for the chlorine concentration do not
matter as much since relative chlorine concentrations are the important parameter.
Usually in the data interpretation, the flux is normalized by the chlorine concentration,
which is why only relative values are important. Also, in order to calculate the fraction
gas film resistance (which is used to analyze some of the data), only relative
concentrations are needed. The absolute value of the chlorine  concentration does matter
if only chlorine flux is looked at alone without normalizing it.

Initially, the absolute chlorine concentration was analyzed by sparging chlorine into a
potassium iodide solution buffered at pH 5. Chlorine reacts with iodide to form iodine,
which is then titrated with sodium thiosulfate to give the chlorine concentration.  This
method was the basis used for the initial analysis.  Later on, the absolute chlorine was
analyzed by sparging chlorine into sodium sulfite solution at pH 12.5 and then analyzing
the chloride concentration using ion chromatography.  For each mole of chlorine
                                        12

-------
absorbed, two moles of chloride are formed. This method resulted in a chlorine
concentration 1.4 times what was seen in the early data (using the iodide method).  The
sulfite method gave more reproducible results and recovered more of the chlorine than
the previously used potassium iodide method.  For both of these methods, the zero  was
calibrated by supplying only nitrogen (without chlorine) through the tubing. Also,
different chlorine concentrations were used in the experiment in order to verify the wet
chemical method. Using the sulfite analysis, the cylinder chlorine concentration (after
passing from the cylinder through the tubing) was 840 ppm, while Air Products stated
that the cylinder concentration was 1000 ppm.  Air Products also stated that the cylinder
chlorine concentration may decrease over time, but they were not sure what the time span
would be. A basis of 840 ppm was assumed for most of the data. For the  earlier data, a
basis of 600 ppm was used since the iodide method was employed.  Multiplying the old
data by 1.4 (840/600) will convert the magnitudes of the earlier data to that of the new
data.

6.7 Reactor solution and analysis

The reactor contained the aqueous S(IV) solution, ranging from 0 to 10 mM, used in
absorbing chlorine.  The reactor fluid volume in a typical experiment was  1.06 x 10"3 m3.
Distilled water was first added to the reactor.  For experiments at pH ~ 4, the reactor
solution was buffered by injecting a stock solution of equimolar succinic acid/sodium
succinate. The buffer concentration in the reactor ranged from 5 to 50 mM total
succinate. The S(IV) solution was obtained by injecting a stock solution containing
equimolar sodium sulfite and sodium bisulfite.  For experiments at pH > 7, stock
solutions of only sodium sulfite were used.

Liquid samples  (2 to 25 mL, depending  on S(IV) concentration) were periodically taken
from the bulk of the reactor and  analyzed for S(IV) concentration by iodometric titration
(Kolthoff and Belcher, 1957), and some samples were analyzed for chloride using ion
chromatography. Withdrawing samples did not affect the reaction since each withdrawal
was followed by subsequent injection of fluid.

The pH of the bulk reactor solution was continuously monitored and recorded using a
strip chart recorder. The pH probe was  calibrated by placing it into a standard pH 4 and
pH 7 buffer solution. After calibration,  the pH probe was inserted into the reactor fluid
for continuous pH monitoring of the bulk reactor fluid. The pH was measured to verify
that the experiments were being  conducted at the desired pH. The concentration of each
S(IV) species (bisulfite and sulfite) can be calculated by knowing the pH and total  S(IV)
concentration. In the buffered S(IV) experiments, essentially all of the S(IV) was present
as bisulfite since the pH was much lower than the pKa of the sulfite/bisulfite reaction.

6.8 Iodometric titration for S(IV)

After the S(IV)  sample was withdrawn from the reactor, it was directly injected into
excess iodine solution to avoid air oxidation to sulfate. The S(IV) reduced the iodine to
iodide. The excess iodine was titrated with sodium thiosulfate.  When the  yellow color of
the iodide started to fade (as the  iodine was reduced to iodide by the thiosulfate), a  couple
                                        13

-------
drops of starch indicator were added to enhance the endpoint detection. The endpoint
was reached when the blue solution turned clear.

The S(IV) concentration was determined from the difference between the amount of
thiosulfate used to titrate the excess iodine and the amount needed if no S(IV) were added
to the iodine. The difference indicates how much of the iodine reacted with S(IV).

The S(IV) analysis procedure was modified as experiments were done. The S(IV)
concentrations were more precise in the data taken after that in Table 8-5 because of
procedural modifications such as checking blanks and verifying standard solutions daily
and withdrawing larger samples from the reactor. However, at very low S(IV), the
difference to determine the amount of iodine which reacted with S(IV) was very low.  At
times, it was close to the errors in measurements (buret readings). Thus, the expected
precision at S(IV) concentrations below 0.09 mM is ± 0.04 mM. For example, when the
S(IV) concentration is reported as 0.08 mM, the actual concentration could range from
0.04to0.12mM.
                                        14

-------
7. Characterizing Stirred Cell Contactor

The overall resistance to mass transfer is equal to the sum of the resistances in the gas
and liquid phases:
                                 KOO    k,   Ek°L
                                                                               (7-D
If a reaction is very fast (high E) and/or if the gas is extremely soluble in the liquid (high
H), the last term vanishes, and the overall gas mass transfer coefficient is equivalent to
the individual gas film mass transfer coefficient. Therefore, kg can be easily determined
by absorbing a gas into a liquid which has a rapid reaction rate.  On the other hand, if the
gas has a low solubility (low H) and does not react quickly in the liquid, the resistance in
the liquid film dominates the overall mass transfer resistance. Thus, to determine the
liquid film mass transfer coefficient, gas should be desorbed from a liquid in which the
gas is not very soluble.  These criteria led to the experiments which were conducted to
characterize the stirred cell contactor.  Mass transfer coefficients are independent of
chlorine concentration.

7.1 Gas film mass transfer coefficient

The gas phase mass transfer coefficient was obtained by measuring chlorine absorption in
0.28 M sodium hydroxide.  Since the chlorine/hydroxide reaction is very fast, there  is
negligible resistance in the liquid phase; thus, K0o is equivalent to kg. Under complete
gas film control, there is essentially no chlorine at the interface.

                                          N
                                          rci2>0

Figure 7-1 depicts the data and correlations for obtaining kg, and Table 7-1 lists the
correlation parameters. Detailed experimental data are in Appendix A.  Data were taken
under three analyzer conditions.  Series A- 1 displays the correlation obtained using the
IMS analyzer. This correlation was used for all data obtained using the IMS analyzer.
Series A-2 displays the correlation obtained from using the improved calibration (Figure
8-2) of the electrochemical analyzer. This correlation was used to analyze the chlorine
absorption data in Table 8-5.  Series A- 3 shows the correlation obtained from using the
electrochemical analyzer after experimental modifications were made to reduce scatter.
This correlation was used for the data in Tables 8-6 and 8-7. Tables A-l through A-3
tabulate the data used to obtain the above correlations.

The kg correlations were compared with each other and with other correlations developed
by previous researchers who used a similar apparatus.  Zhao, who used the exact same
apparatus, developed a kg correlation for mercury by absorbing mercury into aqueous
permanganate.  Her correlation was: kg(mol/s-atm-m2) = 0.0344(ng)038 (Zhao and
Rochelle, 1996).  Dutchuk, who used a similar apparatus (but not identical), developed a
kg correlation for sulfur dioxide by absorbing SCh into sodium hydroxide.  His correlation

                                         15

-------
was: kg(mol/s-atm-m2) = 0.0552(ng)a385 (Dutchuk, 1999). Figure 7-1 and Table 7-1
display all the correlations.
     0.8
                Zhao-Hg
                Dut-SO2
             - - A-2
                A-3
            - - - A-1 (IMS)
                A-1 data
            •   A-2 data
            A   A-3 data
     0.2
       250       350        450       550
                                  ng(rpm)
650
750
850
Figure 7-1. Data and correlations for gas film mass transfer coefficient

Table 7-1. Parameters for gas film mass transfer correlations
kq=a<
Series a b
A-1
A-2
A-3
Zhao-Hg
Dutchuk-SO2
0.0252
0.0218
0.0089
0.0344
0.0552
0.5142
0.5018
0.655
0.38
0.385
Typical gas phase agitation rates (ng) for the chlorine absorption in S(IV) experiments
ranged from 650 to 750 rpm.

7.2 Liquid film mass transfer coefficient

The liquid phase mass transfer coefficient was obtained by measuring chlorine desorption
from hypochlorous acid (HOC1) solution in 0.1 M HC1 at ambient temperature.  In a
typical experiment, sodium hypochlorite (NaOCl) was injected into 1.06 L of 0.1 M HC1.
At a flow rate of 1.2 L/min, nitrogen gas flowed over the solution, desorbing chlorine.
The chlorine formation resulted from the following reaction:
                           C12 + H2O <-» HOC1 + Cl" + H+
At low pH and high chloride, the formation of chlorine is favored.
                             (7-3)
                                        16

-------
The liquid film mass transfer coefficient can be determined by correlating the chlorine
concentration as a function of time. Liquid phase mass balance gives:
                            dCr
                             dt
                                                                               (7-4)
With excess gas, the chlorine at the interface is negligible compared to the chlorine in the
bulk liquid. Thus, the mass balance in Equation 7-4 can be simplified and rearranged to
yield:
                                dCr
                                C
                                  Cl2,b
                                             V
                                                 -dt
                                                                               (7-5)
Integrating the above differential equation results in:

                                             ( k°
                           Cci2,b=Cci2,imtexp --
                                             v

Since the gas phase flux equals the liquid phase flux:

                     G(P      -P    )
                     ^Tc^out  Fci2;in;    0
                                                  V
                                                                               (7-6)
                             RT
                                                       _r
                                                         *^
                                                          C12,i
                                                                               (7-7)
Since there is no chlorine in the entering gas, Prio.   = 0, and since the interfacial chlorine
                                     o o     <-i2;ln
concentration is much less than the bulk chlorine concentration:
                                       RT,
                                                                               (7-8)
Combining the gas balance with the liquid balance and taking natural logarithms results
in:
                    InP      = In
                       U2,out
                                  RTCCl2.imtkL.Cl2A

                                          G
                                                          V
(7-9)
Therefore, from a plot of lnPr-     vs time, the liquid film mass transfer coefficient can
                 r         U2,out              n
be extracted from the slope. Figure 7-2 shows the data and correlations for the liquid
film mass transfer coefficient. All data were taken with the IMS analyzer.  Detailed
experimental data are in Appendix B.

The kL;Ci2 correlation was compared with Zhao's correlation. Zhao performed mercury
desorption experiments in the stirred cell contactor and found the correlation for mercury
                                         17

-------
to be k^Hg (m/s) = 2.42 x 10"7(nL)a73 (Zhao, 1997). The correlation for mercury was
converted to a correlation for chlorine by correcting for diffusivities:
                                k°    -k°
                                KL,C12 ~ KL,Hg
                            (7-10)
Applying this correction resulted in: k^ci  (m/s) = 2.7 x 10"7(nL)°'73. Figure 7-2 shows
how the correlation from this work compares with Zhao's corrected correlation.  The
solid line represents Zhao's correlation.
 1.E-04

 1.E-05
                       k°L,ci2 = 2.7 x10~7nL    (Zhao)
                                  L*|2 = 6.591 x10~7nL
,-7_ 0.5621
       100                     nL(rpm)                     1000
Figure 7-2. Data and correlations for physical liquid film mass transfer coefficient
In order to calculate the S(IV) at the interface, the liquid film transfer coefficient for
S(IV) was needed. Thus, the above correlation, k^ (m/s) = 6.591 x 10"7(nL)a56, wa

corrected for S(IV) by correcting for the diffusivities, which resulted ink^y) (m/s) =
6.248 x 10"7(nL)°'56. Table 7-2 lists the values for the diffusivities of mercury, chlorine,
and S(IV) through water.

Table 7-2. Diffusivity for species used to correct k°L correlations
DHg(m2/s)a
Dcl2 (m2/s)b
Dsav) (m2/s)c
1.2 x 10"y
1.48 x 10"y
1.33 x 10"y
 obtained from Zhao (1997)
 1 obtained from Spalding (1962)
 obtained from Chang (1979)
                                         18

-------
8. Tabulated Results

8.1 Preliminary results with the electrochemical analyzer

Tables 8-1 and 8-2 display initial chlorine absorption data. Gaseous chlorine was
analyzed using the electrochemical sensor analyzer for these data except for data in Table
8-1 Series A, which were analyzed using the wet chemical iodide method (since the
continuous chlorine analyzer had not been purchased yet). A linear calibration was used
for the electrochemical analyzer. For the absolute values of the chlorine concentration,
the iodide wet chemical method was used. As noted earlier, the magnitude of these data
are 1.4 times less than later data calculated using the sulfite method.
Table 8-1. Initial chlorine absorption results, electrochemical sensor analyzer
Series
A



B


C


D





pH
8
8.2
8.6
7.6
water
6-7
6-7
water
4.5
4.5
4-4.5
4-4.5
4-4.5
4-4.5
4-4.5
4-4.5
[S(IV)](mM)
5.5
8.1
10
1.9
0
2.51
2.48
0
0
2.9
0
3.44
1.3
1.3
0.68
2.12
nq(rpm)
432
482
498
484
533
533
533
555
555
555
521
521
554
554
554
554
r\(rpm)
663
693
660
680
609
609
609
620
620
620
640
640
640
640
640
640
CI2,in(ppm)
200
200
200
200
204
204
204
204
204
204
204
204
204
204
204
204
CI2,out(ppm)
45
40
40
65
170
63
50.7
161
175
93.3
172
79.6
113
136
142
61.9
NCi2(kmol/m^-s)
1 .67E-08
1.72E-08
1.72E-08
1.45E-08
3.65E-09
1.51E-08
1 .65E-08
4.58E-09
3.12E-09
1.19E-08
3.49E-09
1 .34E-08
9.78E-09
7.30E-09
6.66E-09
1 .53E-08

All the data in Table 8-2 were obtained with a succinic acid/succinate buffer. These data
were also taken with the electrochemical sensor analyzer using a linear calibration.
However, it was later discovered that the calibration was not entirely linear, and a linear
calibration overpredicted the concentration.
                                         19

-------
Table 8-2. Chlorine absorption in pH 4-4.5 S(IV) with incorrect analyzer calibration
of electrochemical sensor analyzer
Series
A


B






C







[S(IV)](mM)
4.2
3.58
2.19
0
4.32
3.32
1.16
1.4
1.19
1.16
0
3.84
3
2.68
2.25
0.67
0.60
5.27
nq (rpm)
619
619
619
592
604
604
604
605
603
603
535
535
590
573
573
579
579
579
r\ (rpm)
740
740
740
578
567
567
567
574
575
575
560
560
569
568
568
573
573
573
CI2,in(ppm)
200
200
200
200
200
200
200
200
155
155
200
200
200
200
200
200
168
168
CI2,out(ppm)
97
105
94
183
106
156
161
113
83
93
166
96
103
111
128
157
148
79
NCi2(kmol/m^-s)
1.11E-08
1.02E-08
1.14E-08
1 .83E-09
1 .01 E-08
4.73E-09
4.19E-09
9.35E-09
7.75 E-09
6.60E-09
3.71 E-09
1.12E-08
1 .04 E-08
9.56E-09
7.73E-09
4.62 E-09
2.18E-09
9.54E-09
8.2 Results obtained using multi-point calibration of the electrochemical analyzer

Since the analyzer output was not linear over the entire range, a multi-point calibration
was performed. Also due to problems with analyzer drift, calibration was checked
several times within a series of experiments.  Figure 8-1 displays a typical chlorine
analyzer multi-point calibration.
                                        20

-------
    200  -i
    180
                      10            20            30
                              Recorder Output
40
50
Figure 8-1. Electrochemical chlorine analyzer multi-point calibration

The "old" line represents the linear calibration previously used for much of the data.
Using the linear calibration greatly overpredicts the chlorine concentration. The two-line
(multi-point) calibration demonstrates the nonlinearity and reproducibility problem of the
analyzer.  Using this two-line calibration, errors in chlorine concentration can be as much
as ± 20 ppm. Table 8-3 displays data using the two-line calibration shown in Figure 8-1.
Table 8-3. Chlorine absorption in buffered S(IV) using multi-point calibration
buffer(mM)
10
10
10
50
10
50
50
10
50
50
10
50
10
10
50
[S(IV)](mM)
0
0
0
0.31
0.36
0.46
0.50
0.50
0.77
0.88
0.92
0.94
1.12
1.85
2.57
CI2,in(ppm)
179
170
170
184
179
183
186
179
183
183
151
184
179
151
186
CI2,out(ppm)
127
129
129
118
90
156
153
104
90
80
65
113
67
21
53
NCi2(kmol/m^-s)
5.75E-09
4.53E-09
4.53E-09
7.24E-09
9.82E-09
2.96E-09
3.65E-09
8.31 E-09
1 .03E-08
1.14E-08
9.51 E-09
7.82E-09
1 .24E-08
1 .44E-08
1 .48E-08
KOG/kg(%)
11.4
8.9
8.9
15.3
27.5
4.8
5.9
20.2
28.7
35.8
36.6
17.3
46.3
172.5
70.4
[Cf] (mM)
0.19
0.21
0.30
2.0
0.87
2.49
5.4
0.71
2.24
1.74
1.47
1.0
0.33
0.41
2.5

                                        21

-------
8.3 Results obtained using improved calibration of electrochemical analyzer
After the above data were taken, the nonlinearity and reproducibility problem shown in
Figure 8-1 was solved.  The problem occurred because the electrochemical cell was not
completely full of electrolyte. Table 8-4 and Figure 8-2 display the new improved
calibration after this problem was remedied.

Table 8-4. Electrochemical chlorine analyzer calibration after fixing electrolyte level
Recorder output
88.7
74.3
55.5
74.7
88.8
74.6
54.9
41.4
33.4
28.3
24.8
17.7
25.0
28.6
17.5
28.2
33.5
43.8
ppm C\2
187
143
87.5
143
187
143
87.5
53.5
31.3
20.4
13.7
4.79
13.7
20.4
4.79
20.4
31.3
53.5
                                        22

-------
CI2 Concentration (ppm)
200 -,
1 RO
1 RO
I OU
1 /in
1 9n
I iL\)
1 nn
on
fin
An
*f U
on
n
•

•


•

• •
•
^
•
•
                      20
40           60
Recorder Output
80
100
Figure 8-2. Improved calibration of electrochemical chlorine analyzer

As seen from Table 8-4 and Figure 8-2, fixing the electrolyte problem greatly improved
reproducibility. Using this improved calibration for the electrochemical sensor analyzer,
chlorine concentrations greater than 100 ppm usually had a reproducibility of ± 8 ppm
while concentrations less than 100 ppm had a reproducibility of ± 2 ppm. Table 8-5
displays data which were taken using this type of calibration.

Table 8-5. Chlorine absorption in S(IV) solutions using improved calibration of the
electrochemical analyzer, 1.29 L/min gas, 50 mM buffer, kg = 0.6 mol/s-atm-m2
Series
A





B
C






[S(IV)](mM)
0*
1.76*
1.3*
1.31
1.14
0.94
0
0
1.19
0.81
0.68
2.04
1.05
0.90
CI2,in(PPm)
197
197
197
197
197
197
197
197
197
197
197
197
197
197
Cl2,out(ppm)
136
44
74
80
141
145
120
129
101
132
135
59
93
116
NCi2(kmol/m'i-s)
6.07E-09
1 .54E-08
1 .23E-08
1.17E-08
5.60E-09
5.20E-09
7.72E-09
6.82E-09
9.60E-09
6.55E-09
6.20E-09
1 .39E-08
1 .04E-08
8.08E-09
KOG/kg(%)
8.2
64.7
30.1
26.7
7.2
6.5
11.8
9.4
16.9
8.9
8.2
41.8
19.9
12.3
[Cr](mM)
0.18
1.15
1.92
1.75
3.18
2.87
0.09
0.09
0.87
1.22
1.51
1.97
2.68
3.04
* Experiments in 10 mM succinate buffer instead of 50 mM succinate buffer
                                       23

-------
The experimental scatter in the above data was resolved by modifying the experimental
analysis. The precision in the S(IV) concentration measurements was greatly enhanced
by increasing sample size and verifying standard solutions daily. Also, scatter was
reduced by lowering the pressure in the reactor, which minimized leaks.  The absolute
chlorine concentration was now obtained using the sulfite method since it gave better
results than the previously used iodide method.  So, the magnitudes are all 1.4 times
higher than what was previously reported. Tables 8-6 and 8-7 show data obtained after
these modifications were made. The data in Table 8-6 are for chlorine inlet
concentrations greater  than 100 ppm, while the data in Table 8-7 are for chlorine inlet
concentrations of 21 ppm.

Table 8-6. Chlorine absorption in pH 4.3 - 4.5 S(IV), measured with electrochemical
analyzer, 1.2 L/min gas, 50 mM succinate buffer, and kg = 0.60 - 0.66 mol/s-atm-m2
Series
A






B





C




[S(IV)](mM)
0
1.19
0.88
0.43
1.21
0.87
0.78
0
1.45
0.95
0.84
0.62
0.58
0
1.37
1.25
0.82
0.69
CI2,in(ppm)
276
276
276
276
276
276
276
276
276
276
129
129
276
276
276
276
276
129
CI2,out(ppm)
185
118
177
222
132
162
184
174
70
141
43
50.4
181
179
51.5
80.6
94.1
29.1
NCi2(kmol/m^-s)
9.13E-09
1.59E-08
9.94E-09
5.37E-09
1 .45E-08
1.15E-08
9.18E-09
1 .02E-08
2.06E-08
1 .35E-08
8.86E-09
8.07E-09
9.55E-09
9.69E-09
2.25E-08
1 .96E-08
1 .82E-08
1 .03E-08
KOG/kg(%)
8.2
22.5
9.4
4.0
18.3
11.8
8.2
9.5
47.7
15.4
33.5
25.9
8.5
8.3
37.0
29.5
53.7
8.8
[CH (mM)
0.36
0.5
0.76
1.1
1.65
6.56
19.7
0.12
1.13
1.32
1.3
2.88
1.49
0.17
0.86
1.61
1.9
2.31
Table 8-7. Chlorine absorption measured with electrochemical analyzer in pH 4.5
S(IV), C\2,m = 21 ppm, 1.19 L/min gas, 50 mM succinate buffer
Series
A

B


C
D

[S(IV)]buik(mM)
1.16
1.14
0.23
0.20
0.13
0.20
0.31
0.27
CI2jn(ppm)
21
21
21
21
21
21
21
21
CI2,out(ppm)
2.9
2.9
9.5
12.2
14.0
9.9
6.4
7.9
Nc^kmol/m^-s)
1.85E-09
1.85E-09
1.17E-09
8.98E-10
7.17E-10
1.13E-09
1.49E-09
1 .34E-09
                                       24

-------
8.4 Results obtained using the IMS chlorine analyzer

The above data were the last set obtained using the electrochemical analyzer. The next set
of experiments were done using the IMS analyzer with the following conditions in mind:
1) experiments at low chlorine and S(IV) concentrations to extract kinetics, 2)
experiments with oxygen in the inlet gas, 3) experiments at high S(IV) concentrations to
quantify S(IV) oxidation rates and see if chlorine catalyzes S(IV) oxidation. Table 8-8
displays the first set of data using the IMS analyzer, including data with oxygen. Table 8-
9 lists data for oxygen absorption in S(IV).  Table 8-10 lists the final set of data in 50 mM
succinate buffer.
Table 8-8. Chlorine and oxygen absorption measured with IMS analyzer in pH 4.5
S(IV), 50 mM succinate buffer, and 1.15 L/min gas
Series
A









B





C



D



E







F
[S(IV)]bUik(mM)
0*
0*
0
1.15
0.78
0.634
4
4
4
4
0*
0*
0
3.85
2.44
1.86
0
5.04
3
1.74
0*
0*
0
5
0
4.43
3.83
3.22
2.02
1.37
0.664
0.35
0
Cl2,in(ppm)
276
276
276
276
129
21
276
129
129
21
276
276
276
276
276
276
276
276
276
276
21
21
21
21.2
264
264
264
264
264
264
264
264
21
Cl2,out(ppm)
235
242
150
94.6
21.6
2.5
33.8
10.8
13.2
0.33
225
236
151
34.5
39.3
39.7
149
37.5
38.5
43.7
15.6
17
8.3
1.3
134
32.7
28
31.7
39
41.9
116
159
8.1
Nc^kmol/m^-s)
4.09 E-09
3.37E-09
1 .26E-08
1.82E-08
1.10E-08
1 .93E-09
2.43E-08
1 .21 E-08
1.19E-08
2.14E-09
5.10E-09
3.99E-09
1.25 E-08
2. 42 E-08
2.37E-08
2.37E-08
1 .27E-08
2.39E-08
2.38E-08
2. 33 E-08
5.46E-10
4.05E-10
1 .28E-09
2.01 E-09
1 .26E-08
2.24E-08
2.29E-08
2.25E-08
2.18E-08
2.15E-08
1 .44 E-08
1.02 E-08
1 .31 E-09
02 (%)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
14.5
14.5
14.5
14.5
14.5
14.5
14.5
14.5
0

                                                  (Continued)
                                        25

-------
Table 8-8. Continued
Series
G









H









[S(IV)]buik(mM)
0
0.26
0.176
1.08
1.05
1.02
3.74
3.44
3.43
3.59
0
0.519
0.336
0.18
1.56
1.18
0.642
4.74
4.23
3.88
CI2,in(ppm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
CI2,out(ppm)
8
3.8
4.7
2.2
2.04
2.15
0.9
1.2
1
1
7.9
2.99
3.6
4.3
1.5
1.9
2.5
0.8
0.94
1
NCi2(kmol/m^-s)
1 .33E-09
1 .76E-09
1 .66E-09
1.92E-09
1 .93E-09
1.92E-09
2.05E-09
2.02E-09
2.04E-09
2.04E-09
1 .34E-09
1 .84E-09
1 .78E-09
1 .70E-09
1.99E-09
1.95E-09
1.89E-09
2.06E-09
2.04E-09
2.04E-09
02 (%)
0
0
0
0
0
0
0
0
0
0
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
* not buffered; absorption in pure water
Table 8-9. S(IV) depletion resulting from oxygen absorption in pH 4.5 S(IV) at
ambient temperature, 50 mM succinate buffer, and 1.15 L/min gas
Series
A





B

C









At(min)
0
33
37
29
57
103
0
53
0
53
42
65
60
36
37
45
57
29
[S(IV)]bUik(mM)
6.06
4.23
4.07
3.89
3.86
3.13
2.72
2.36
4.88
3.99
3.64
3.21
2.66
2.33
2.18
1.69
1.27
0.99
02(%)
14.5
14.5
14.5
14.5
14.5
14.5
14.5
14.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
                                     26

-------
Table 8-10. Chlorine absorption in pH 4.5 S(IV), 50 mM succinate buffer, 1.15
L/min gas
Series
A










B









C









D







[S(IV)]bUik(mM)
0
0
0.38
0.169
0.59
0.334
0.208
1.56
1.22
4.95
10
0
0
0.182
0.075
0.276
0.178
1.38
1.32
5.23
10
0
0
0.27
0
0.61
0.437
1.25
0.74
4.7
10
0
0.183
0.098
0.26
0.128
1.43
1.39
5.45
Cl2,in(ppm)
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
Cl2,out(ppm)
134.1
142.3
154.0
163.5
131.2
171.7
178.2
43.3
49.8
29.9
23.0
4.35
4.89
3.67
7.75
3.84
4.62
2
2
0.7
0.3
132.5
142.2
153.5
163.5
125.0
160.0
46.5
94.7
32.0
24.9
7.25
4.24
6.68
3.83
5.69
2.2
2.2
0.6
Nc^kmol/m^-s)
1 .26E-08
1.18E-08
1 .06E-08
9.73E-09
1.29E-08
8.93E-09
8.30E-09
2.14E-08
2.07E-08
2.27E-08
2.33E-08
1 .70E-09
1.65E-09
1 .77E-09
1 .36E-09
1.75E-09
1 .67E-09
1 .94E-09
1 .94E-09
2.07E-09
2.11E-09
1 .27E-08
1.18E-08
1 .07E-08
9.73E-09
1.35E-08
1 .01 E-08
2.11E-08
1 .64 E-08
2.25E-08
2. 32 E-08
1 .41 E-09
1 .71 E-09
1 .46 E-09
1.75 E-09
1 .56E-09
1.92 E-09
1.92 E-09
2.08E-09
                                      27

-------
8.5 Chlorine absorption as a function ofsuccinate buffer concentration

Experiments were done to study the effect of the succinate buffer on chlorine absorption.
The reactor solution was buffered by injecting a stock solution of equimolar succinic
acid/sodium succinate. The buffer concentration was varied from 0 to 150 mM total
succinate. Table 8-11 lists data for chlorine absorption at various buffer concentrations in
which no S(IV) is present.

Table 8-11. Chlorine absorption as a function of buffer concentration at ambient
temperature in pH 4.5 with 1.15 L/min gas
Series
A









B








C







[buffer] (mM)
0
0
10.3
48
48
66
102
102
137
154
0
0
10.3
48
66
102
102
137
154
0
0
10.3
48
66
102
137
154
CI2,in (ppm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
263.9
Cl2,out(ppm)
13.83
14.85
11.82
8.61
9.00
8.04
7.33
7.15
6.71
6.76
211.8
225.0
187.4
148.6
143.0
132.0
136.4
129.2
126.7
175.4
203.1
182.2
148.1
139.6
126.8
118.3
116.5
NCi2 (kmol/m^-s)
7.40E-10
6.37E-10
9.44E-10
1 .27E-09
1 .23E-09
1 .33E-09
1 .40E-09
1 .42E-09
1 .46E-09
1 .46E-09
5.05E-09
3.77E-09
7.41 E-09
1.12E-08
1.17E-08
1 .28E-08
1 .24E-08
1 .31 E-08
1 .33E-08
8.57E-09
5.89E-09
7.92E-09
1.12E-08
1 .20E-08
1 .33E-08
1 .41 E-08
1 .43E-08
8.6 Chlorine absorption in 5 mM succinate buffer

Since chlorine reacts with the buffer, future experiments were done in 5 mM buffer
instead of 50 mM.  These data are tabulated in Table 8-12. Lowering the buffer
concentration lowered the chlorine flux in the buffer, resulting in a wider range of fluxes
between the limitations of the chlorine/buffer and of gas film control.
                                        28

-------
Table 8-12. Chlorine absorption in pH 4.5 S(IV), 5 mM buffer, 1.18 L/min gas
Series
E





F






[S(IV)]bUik(mM)
0
0.213
0.155
0.103
0.089
1.48
0
0.117
0.076
0.064
0.06
0.15
0.12
CI2,in(PPm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
Cl2,out(ppm)
12.9
4.33
5.51
8.88
13.6
2.23
13.0
7.49
11.6
13.51
15.1
6.09
9.26
Nci^kmol/m^-s)
8.39E-10
1 .70E-09
1 .58E-09
1 .24E-09
7.63E-10
1 .91 E-09
8.21 E-10
1 .38E-09
9.66E-10
7.73E-10
6.17E-10
1.52 E-09
1 .20E-09
Table 8-13 tabulates the data in which agitation rates were varied.  S(IV) concentrations
were not measured for each point. The series of experiments began with chlorine
absorption in 5 mM buffer.  S(IV) was then injected and agitation rates were varied. At
the end of the series, a sample was taken from the reactor and analyzed for S(IV). After
the S(IV) was depleted, more was injected, and the agitation rates were varied.

Table 8-13.  Chlorine absorption in S(IV) solutions with 5 mM buffer at various
agitation rates

A
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A2
A2
A2
A2
A2
A2
A2
A2
A2
[S(IV)](mM)
0
Inject S(IV)









0.11
Inject S(IV)








nq (rpm)
720
720
720
720
720
720
720
720
887
1007
1007
752
752
752
745
382
382
678
926
926
926
nL (rpm)
780
780
780
472
472
472
898
1029
1036
1039
1039
757
757
757
1053
1063
330
319
319
695
1051
CI2,in(ppm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
CI2,out(ppm)
16.8
4.69
5.91
9.87
10.5
10.8
6.82
6.67
7.16
7.43
7.74
12.3
5.30
5.91
5.23
6.06
13.9
14.5
14.8
10.8
7.74
NC|2(kmol/m'£-s)
4.43E-10
1 .67E-09
1 .54E-09
1.14E-09
1 .08E-09
1 .05E-09
1 .45E-09
1 .47E-09
1 .42E-09
1 .39E-09
1 .36E-09
8.97E-10
1 .60E-09
1 .54E-09
1 .61 E-09
1 .53E-09
7.30E-10
6.72E-10
6.44E-10
1 .05E-09
1 .36E-09
                                                     (Continued)
                                        29

-------
Table 8-13. Continued

A2
A2
B
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B2
B3
B3
B3
B3
B3
B3
B3
B3
B3
B4
B4
B4
B4
B4
B4
B4
B4
[S(IV)](mM)

0.078
0
Inject S(IV)










0.070





0.028
Inject S(IV)







0.13
Inject S(IV)






0.64
nq (rpm)
1051
1051
730
730
730
1151
344
344
332
332
330
330
974
974
749
748
749
1140
1153
734
734
734
1054
1054
1069
478
467
1103
720
715
715
711
369
366
1100
1113
1117
719
nL (rpm)
757
757
708
708
708
713
713
713
402
402
1044
1044
1058
1058
762
408
1057
1083
722
717
717
717
715
715
341
336
990
1004
1009
731
731
408
403
938
946
383
713
717
Cl2,in(ppm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
Cl2,out(ppm)
12.0
13.1
14.8
7.10
7.71
8.94
9.86
10.5
14.8
15.1
10.2
10.5
11.1
11.4
14.6
15.8
14.4
14.8
16.1
16.3
16.6
4.35
4.10
4.38
8.48
9.24
5.14
4.20
4.75
6.28
2.91
3.06
4.10
3.98
2.42
2.54
2.48
3.06
Nc^kmol/m^-s)
9.25E-10
8.17E-10
6.44E-10
1 .42E-09
1 .36E-09
1 .24E-09
1.14E-09
1 .08E-09
6.44E-10
6.16E-10
1.11E-09
1 .08E-09
1 .02E-09
9.88E-10
6.60E-10
5.42E-10
6.78E-10
6.47E-10
5.08E-10
4.92E-10
4.61 E-10
1 .70E-09
1 .72E-09
1 .70E-09
1 .28E-09
1 .20E-09
1 .62E-09
1 .72E-09
1 .66E-09
1 .50E-09
1 .85E-09
1 .83E-09
1.72E-09
1 .74E-09
1.89E-09
1 .88E-09
1.89E-09
1 .83E-09
                                     30

-------
9. Discussion of Results

9.1 Rate of reaction for chlorine with S(IV)

Figure 9-1 shows the data with the IMS analyzer (Tables 8-8, 8-10, and 8-12).  Data at
the two chlorine inlet concentrations of 265 ppm and 21 ppm are shown.  The points at
0.01 mM S(IV) are actually in succinate buffer with no S(IV). The inverted triangles in
Figure 9-1 represent points in which oxygen was added.  At an inlet concentration of 265
ppm, the oxygen level was 14.5%, and at 21 ppm, the oxygen level was 20.5%. All the
data are in 50 mM buffer except for the points represented by the squares with diagonal
lines which are in 5 mM buffer.

The  curves are calculated using the model for mass transfer with fast chemical reaction in
the boundary layer.  The model was similar to Equation 5-4, but included the buffer
effect also:
                  NCI- =
                          Pci
                             2i
                          Ha,
[S(IV)] i
                                                 [buffer] )
(9-1)
The partial pressure of chlorine at the interface can be calculated from Equation 5-1, and
the interfacial S(IV) concentration can be obtained from Equation 5-7.  Substituting these
into Equation 9-1 yields:
NCI,  =
                      N
                        C12
        Hci,
        Cl2,b
Dei 2
[i- frcroi Nci2 1
K2,8(1V) l^\L * )lb 0
I ^k L,S(IV)
V V /
\
f k2 buf [buffer]
                                                                       (9-2)
The partial pressure of chlorine in the bulk (which is equivalent to the chlorine exiting the
reactor) can be written in terms of the inlet chlorine concentration through a gas phase
material balance.  Thus, Equation 9-3 displays the model used to calculate the curves, and
Table 9-1 lists the parameters that were supplied to the model.
Hci0
                             N
             rC!2,in
                               C12
                       G
                                     Da,
                                                     N
                                    2,S(IV)
                                                       ci2
                          -k2;buf [buffer]
                                                                           (9-3)
Table 9-1. Values of parameters in global model
Dcl2 (m2/s)
G (m3/s)
0
Hcl2 (atm-m3/kmol)
kg (kmol/s-atm-m2)
k°L,s(iv) (m/s)
k2,sov) (L/mol-s)
1.5 x 10"y
0.0708 (1.1 8 L/min)
1
16.7
0.00075
2.45 x 10"s
2xlOy
                                        31

-------
The gas flow rate and mass transfer coefficients used in the model were representative of
the experimental data.  The stoichiometric coefficient, , was chosen to be one because
the overall stoichiometry of the reaction shows that one mole of chlorine reacts for every
mole of S(IV). The buffer rate constant was obtained from the analysis of the chlorine
absorption in succinate buffer experiments discussed in Section 9.3.  The rate constant of
the chlorine/S(IV) reaction (k2,s(iv)) was chosen to best fit the data.
     10"
  o
  E
  O
     10"
                yz
                                                                 265 ppm
                          265 ppm
                          21 ppm
                          oxygen
                          5 mM buffer
        0.01
01   [S(IV)](mM)
1
10
Figure 9-1. Chlorine absorption in buffered S(IV), k2,s(iv) = 2 x 10 L/mol-s
Figure 9-1 shows that at high S(IV), the chlorine flux does not depend on the S(IV)
concentration since the limit of gas film resistance is approached. At lower chlorine
concentrations, gas film control is achieved at lower S(IV) since it takes less S(IV) to
react with the chlorine. Because of complete gas film control, the chlorine flux should
not increase after 3 mM S(IV) with the 265 ppm chlorine inlet or after 0.6 mM S(IV)
with 21 ppm inlet chlorine. However, the data at high S(IV) concentrations in Tables 8-8
and 8-10 show that additional chlorine removal is occurring at high  S(IV). The
additional chlorine removal is due to the increased levels of gaseous SO2, which forms at
high S(IV) concentrations, in the reactor:
                           SO2(g) + H2O ^ H+ + HSO3"
                                             (9-4)
At high S(IV) and low pH, the above reaction favors SO2 production.  The SO2 reacts
with the chlorine in the moist areas on the reactor surface. Therefore, Figure 9-1 plots all
of the IMS analyzer data except for the values at S(IV) concentrations greater than 5 mM
for the 265 ppm data and greater than 1 mM for the 21 ppm data. At pH 4.5, 5 mM S(IV)
results in generation of about 70 ppm SO2. Therefore, the existing stirred cell contactor
cannot be used for experiments at high S(IV)/low pH since SO2 would form inside the
reactor.
                                       32

-------
Figure 9-1 shows that at low S(IV), the flux is limited (in some cases inhibited) by the
buffer-enhanced chlorine hydrolysis reaction. In this region, the flux depends only on the
buffer reaction rate. However, the data show that when very little S(IV) was injected, the
chlorine flux was less than what it was initially in buffer alone. Thus, S(IV) inhibited
chlorine absorption at very low S(IV) concentrations. At a chlorine concentration of 265
ppm and S(IV) concentration of 0.5 mM, the flux was equivalent to the flux in buffer
alone.  When the S(IV) concentration was lower, the flux was lower than what it was
initially without S(IV).  As the S(IV) increased, the chlorine flux increased until the gas
film limit was reached.  At the low inlet chlorine of 21 ppm, when the S(IV)
concentration was 0.06  mM, the chlorine flux was the same as the flux in buffer alone.
At lower concentrations, the flux is lower, and thus, the reaction seems to be inhibited by
a little S(IV) but enhanced by greater amounts of S(IV).

Since Fogelman et al. (1989) have shown that HOC1 reacts with sulfite, one possible
mechanism for chlorine reaction with S(IV) is that the chlorine first hydrolyzes in water
to form HOC1, and then the HOC1 (not C12 directly) reacts with S(IV). These overall
reactions are shown below:

                           C12 + H2O -+ HOC1 + H+ + Cl"                        (9-5)
                          HOC1 + SO3"2 -^ SO4"2 + H+ + Cl"                      (9-6)

If this were the case, the rate of chlorine absorption in S(IV) would be equivalent to the
rate of chlorine hydrolysis to form HOC1 since chlorine hydrolysis is the rate limiting
step. Then, the HOC1 would react with S(IV). However, since the addition of S(IV)
results in a greater chlorine removal rate than the chlorine hydrolysis rate, it must not
depend on HOC1 formation.  Thus, chlorine itself reacts with S(IV) directly, and it is not
necessary for HOC1 to form before chlorine reaction with S(IV) occurs.

In the intermediate region of Figure 9-1, the flux is limited by S(IV) diffusion to the
interface [depicted by flux increasing linearly with S(IV)] and/or kinetics (depicted by
curvature). Looking at  the 265 ppm data, for S(IV) between 0.5 and 0.8 mM, the flux
increases linearly with S(IV), which is consistent with the model of S(IV)  depletion. At
the lower inlet concentration of 21 ppm, the data do not fall on the model curve at S(IV)
concentrations below 0.1 mM. These deviations result from the experimental uncertainty
in the S(IV) concentration measurements at low S(IV).  For example, if the iodometric
analysis yielded an S(IV) concentration of 0.07 mM, the actual value could be 0.03 mM
due to the analysis procedure not being as accurate at low S(IV) concentrations. Also, at
the lower concentrations, there could be less S(IV) at the interface than perceived due to
oxidation by residual oxygen in the inlet gas.  Up to 5 ppm oxygen may be present in the
"pure" nitrogen.

Figure 9-1 shows that there is only a very small range (depicted by curvature) where the
chlorine flux should be  limited by the kinetics of the Cl2/S(IV) reaction. Looking at the
265 ppm curve, the range of kinetics-limited data would be from  0.8 to 1.2 mM S(IV).
For the 21 ppm inlet, the range is from 0.07 to 0.1 mM S(IV). Thus, it is hard to extract  a
rate constant from the data since most of the data falls in a region where the chlorine
absorption is not limited by kinetics.

                                        33

-------
Even though it was difficult to obtain a precise value for the rate constant because of the
mass transfer limitations of the reactor, an approximate value can be determined. Model
curves were calculated for various rate constants to see what value fitted the data best.
Instead of plotting flux as a function of bulk S(IV) as was done in Figure 9-1, plotting
chlorine penetration (Cl2,0ut/Cl2,in) as a function of [S(IV)]b/Pci2,in allows the errors in the
data to be magnified.  This allows better observation of which value for k2,s(iv) fits the
data the best.  Figure 9-2 plots the same data, without separately labeling the points with
oxygen, and uses the same model as in Figure 9-1. The points on the y-axis (at a value of
0.5 M/atm) are actually in 0 mM S(IV).  These values are plotted on the y-axis since a
value of zero cannot be shown on a  log-log plot.
   1
 o
 o
 •(-»
 ro
 &_
 *j
 o>
 c
 0)
 Q_
   C
 O
 0.1
      - 21  ppm
      - 5 mM^buffer
         265 ppm
        21 ppm;50 mM buffer
                                                    265 ppm

                                                    21 ppm

                                                    5 mM buffer
                                               10

Figure 9-2. Chlorine penetration in buffered S(IV), k2,s(iv) = 2 x 109 L/mol-s
                                                                              100
Figure 9-2 shows the same trends as Figure 9-1.  The chlorine penetration is the greatest
when the chlorine absorption is controlled by the buffer rate and the least when it is
controlled by gas film resistance.  Figures 9-2, 9-3 and 9-4 plot the same data but use
different values for k2,s(iv). For all of these figures, the points on the y-axis represent
points with no S(IV).  The model curve in Figure 9-3 is calculated using an infinite rate
constant while the curve in Figure 9-4 uses a lower rate constant of 2.5 x 108 L/mol-s.
Figure 9-3 shows that the Cl2/S(IV) reaction rate is not instantaneous since much of the
data lie above the model curve. This instantaneous reaction model predicts a greater rate
of chlorine reaction than what the data show.  On the other hand, the rate constant used in
Figure 9-4 is too low. The penetration is overpredicted in most of the data, signifying
greater chlorine removal than what is predicted from the model.
                                         34

-------
  .1
       21 ppm
       5 m M buffer
       21 ppm ; 50 m M buffer
                                                         265 ppm

                                                         21 ppm
                                                         5 mM buffer
                                              •      •
                   [S(IV)] /P     (M/atm)
              1           b  CI2,in           10
100
Figure 9-3. Chlorine penetration in buffered S(IV), k2,s(iv) = °°
 o
 o
 +-»
 ro
 i_
 *-
 o>
 c
 0)
 a.
  C
 o
 0.1
      - 21 ppm
      - 5 mM^buffer
                                                   265 ppm

                                                   21  ppm

                                                   5 m M buffer
        21  ppm;50 mM buffer
                            IP    (M/atm)
Figure 9-4. Chlorine penetration in buffered S(IV), k2,S(iv) = 2.5 x 108 L/mol-s

Therefore, the most probable value of the rate constant is 2 x 109 L/mol-s, although it
could be an order of magnitude smaller or larger.  However, many of the low S(IV)
points, especially for the 21 ppm data, do not fall on the curve, but that could be due to
the inability to accurately measure low S(IV) concentrations.  In order to get a more
precise rate constant, an apparatus with higher mass transfer coefficients is needed so that
                                        35

-------
the absorption falls in a region controlled by reaction kinetics instead of being controlled
by mass transfer.

9.2 Chlorine absorption as a function of agitation rates

In order to further investigate  if it was possible to obtain a precise rate constant in the
stirred cell reactor, experiments were done in which the mass transfer coefficients were
varied by varying the agitation rates (ng and nL).  These data are listed in Table 8-13. For
fast reactions that are controlled by kinetics, absorption should not be affected by changes
in mass transfer coefficients.  Therefore, if varying the agitation rates changes the flux of
chlorine (signifying that the chlorine absorption depends on the mass transfer
coefficient), kinetics cannot be extracted since mass transfer is being measured instead of
kinetics. To  analyze the data, the variation of flux with nL was investigated.

Table 9-2. Flux variance with liquid agitation rate

A1
A1
A1
A1
A1
A1
A1
A2
A2
A2
A2
A2
A2
A2
A2
B1
B1
B1
B1
B1
B1
B1
B1
B1
B1
B3
B3
B3
B3
B3
B3
nq (rpm)
720
720
720
720
720
1007
752
752
745
382
382
926
926
926
1051
344
332
332
330
974
749
748
749
1140
1153
1054
1069
478
467
720
715
r\ (rpm)
780
472
472
898
1029
1039
757
757
1053
1063
330
319
695
1051
757
713
402
402
1044
1058
762
408
1057
1083
722
715
341
336
990
1009
731
CI2,in(ppm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
21.2
CI2,out(ppm)
5.91
9.87
10.8
6.82
6.67
7.74
12.3
5.91
5.23
6.06
13.9
14.8
10.8
7.74
12.0
10.5
14.8
15.1
10.2
11.4
14.6
15.8
14.4
14.8
16.1
4.38
8.48
9.24
5.14
4.75
6.28
Nci2(kmol/m^s)
1 .54E-09
1.14E-09
1 .05E-09
1 .45E-09
1 .47E-09
1 .36E-09
8.97E-10
1 .54E-09
1 .61 E-09
1 .53E-09
7.30E-10
6.44E-10
1 .05E-09
1 .36E-09
9.25E-10
1 .08E-09
6.44E-10
6.16E-10
1.11 E-09
9.88E-10
6.60E-10
5.42E-10
6.78E-10
6.47E-10
5.08E-10
1 .70E-09
1 .28E-09
1 .20E-09
1 .62E-09
1 .66E-09
1 .50E-09
d(ln NCi2/ln r\)
0.600

0.504
0.078

1.31

0.131

0.630

0.626
0.624
1.17

0.903

0.618

1.23
0.314
0.236

0.598

0.379

0.274

0.304

                                                         (Continued)
                                          36

-------
Table 9-2. Continued

B4
B4
B4
B4
B4
B4
B4
nq (rpm)
715
711
369
366
1100
1113
1117
nL (rpm)
731
408
403
938
946
383
713
Cl2,in(ppm)
21.2
21.2
21.2
21.2
21.2
21.2
21.2
Cl2,out(ppm) Nc^kmol/nr^s)
2.91
3.06
4.10
3.98
2.42
2.54
2.48
1.85E-09
1 .83E-09
1.72E-09
1 .74E-09
1.89E-09
1 .88E-09
1.89E-09
d(ln NC|2/ln r\)
0.014

0.008

0.007
0.005

From the first set of points in Table 9-2, when nL was lowered from 780 rpm to 472 rpm,
                               0.6
the flux changed by a factor of nL '  . Based on the experiments to obtain the liquid film
mass transfer coefficient of the reactor, k°L,ci2 should be proportional to n^'56. Since the
flux depends on nL, the rate of S(IV) diffusion to the interface (which is controlled by the
liquid film mass transfer coefficient) is limiting the rate of chlorine absorption. For the
points in Series Al, A2, and Bl, the flux changes as nL changes.  For the points in Series
B3, the overall dependence is less than the dependence in the above series, but there is
still a dependence on nL.  For Series B4, the flux does not depend on the liquid agitation
rate. Therefore, the flux is not being limited by S(IV) depletion at the interface.

Figure 9-5 was plotted to check if Series B4 was limited by the diffusion of chlorine to
the interface, which is controlled by the gas film mass transfer coefficient, kg.  If the
chlorine absorption were limited by kg, the data in B4 would fall  on a straight line
corresponding to kg.
  0.1
      100            ng(rpm)     1000                       10000

Figure 9-5. Data (Table 9-2 Series B4) limited by gas film mass transfer coefficient

Figure 9-5 shows that the points in Series B4 are close to the kg line.  Since there is a
dependence on ng, kinetics cannot be extracted since these points are dependent on kg.
Thus, there may be a very small region in between the S(IV) concentrations represented
                                        37

-------
in Series B3 and B4 for kinetics to limit the chlorine absorption. But in most of the data,
chlorine absorption is not being controlled by kinetics.

9.3 Effect ofsuccinate buffer on chlorine absorption

Table 8-11 tabulates the data obtained from varying buffer concentration, and these data
are shown in Figures 9-6 and 9-7.  Figures 9-6 and 9-7 show that the flux of chlorine
increases as the total succinate buffer concentration increases.  The maximum flux
resulting from complete gas film control (corresponding to a fraction gas film resistance
of one) is plotted as a line.  These results show that the succinate buffer does enhance
chlorine absorption. In the beginning of each series when the chlorine was run through
water, there was an initial "dip" in the flux before it stabilized at the higher value. This
"dip" is reflected in the graphs by the lower flux in water (0 mM buffer). The multiple
data at a given [succinateji in Figure 9-6 represent data from two experiments as shown
in Table 8-11 Series B and C. The multiple data in Figure 9-7 represent the slight drift in
flux that occurred with time. The rate constant, k2,buf, was extracted from the data and
found to have a slight dependence on the chlorine concentration. At the chlorine inlet
concentration of 265 ppm, k2,buf = 29,000 L/mol-s; at inlet of 21 ppm, k2,buf = 149,000
L/mol-s. These values were used in the model calculations.
     2.5E-08
  
-------

    2.0E-09

    1.5E-09H
   =
   o1.0E-09 •]
   E
  .^
   S5.0E-10
    O.OE+00
Maximum TIUX ror compieie gas Turn coniroi
             0            50            100          150          200
                                [succinate]T (mM)

Figure 9-7. Chlorine absorption in succinate buffer with chlorine inlet of 21 ppm
There is a very narrow range for S(IV) data to determine kinetics since the chlorine flux
in the 50 mM buffer is not that far from the maximum flux which is controlled by the
mass transfer in the gas phase. Therefore, later experiments were done in 5 mM buffer.

9.4 S(IV) oxidation by chlorine and oxygen

Chlorine and oxygen both oxidize S(IV) to sulfate, S(VI).  Several experiments were
performed which established that Cl2 is not a catalyst for the oxidation of S(IV) by
oxygen. Figures 9-8 and 9-9 show how the oxidation of S(IV) depends on oxygen and
chlorine. The different symbols represent each series of experiments.  The dashed lines
represent experiments in which chlorine and oxygen are simultaneously absorbed. The
values associated with each line represent the S(IV) oxidation rate for that series.  The
chlorine absorption data are  from Table 8-8, and the oxygen absorption data are from
Table 8-9. Most of the data in Table 8-8 were at high S(IV) concentrations since it is
easier to observe S(IV) oxidation when the chlorine flux is constant due to complete gas
film control.
                                       39

-------
    7

    6

    5
    0
    S(IV) oxidation from physical absorption of O2 = 0.039 mol/m -hr
                                S(IV) ox = 0.037 mol/m -hr
       0
       100
200
300        400
   time (min)
500
600
Figure 9-8. S(IV) oxidation by 275 ppm chlorine and 14.5% oxygen
          S(IV) oxidation from physical absorption of O2 = 0.055 mol/m2-hr
                                                          NoO,
         A
         X
         Xj
1-CI2
2-CI2
02
1-CI2 and O2
2-CI2 and O2
3-CI2 and O2
                                   ox = 0.068 mol/m2-hr   /
                                                      /
               100      200       300      _400
                                    time (min)
                                          500
                                 600
                               700
Figure 9-9. S(IV) oxidation by 21 ppm chlorine and 20.5% oxygen

Figure 9-8 shows that the effects of chlorine and oxygen on S(IV) oxidation may be
additive when the inlet concentration is 275 ppm. For the points with no chlorine, the
first point at 6 mM was not used in the regression since the point seemed to deviate
greatly from the rest.  When only oxygen is absorbed into S(IV), the S(IV) oxidation
corresponds to the oxidation rate resulting from the physical absorption of 14.5 %
                                      40

-------
oxygen, which is 0.039 mol/m2-hr. The S(IV) oxidation is calculated from the depletion
rate of S(IV) observed over time.  When only chlorine is absorbed, the oxidation of S(IV)
is 0.095 mol/m2-hr. There were two separate data series for these, and both data sets fell
on the same line, resulting in the same S(IV) oxidation rate of 0.095 mol/m2-hr.  When
chlorine and oxygen are absorbed (corresponding to S(IV) oxidation rate of 0.13 mol/m2-
hr), about three-fourths of the S(IV) oxidation is due to the reaction of chlorine with
S(IV). In this case, the chlorine does not seem to be catalyzing S(IV) oxidation  since the
S(IV) oxidation seems to be additive.

Figure 9-9 shows data in which the chlorine inlet concentration is significantly lower at
21 ppm.  In this case, the oxidation of S(IV) due to reaction with chlorine (Series 1-Cli
and 2-C12) is practically negligible. Physical absorption of 20.5% oxygen corresponds to
S(IV) oxidation of 0.055 mol/m2-hr.  Thus, the oxidation of S(IV) without chlorine
(Series Ch) is in the range of oxidation due to physical absorption of Ch. It should be
noted that in the data in Series C>2, the last four points contained 0.025 mM Fe+2. Fe+2
was added in order to enhance oxygen absorption since ferrous ion is a well-known
catalyst.  However, Figure 9-9 shows that adding ferrous ion did not enhance oxygen
absorption (since the S(IV) oxidation rate in Series Ch did not change after Fe+2  addition).

At very low S(IV)  concentrations  (< 0.5 mM), the S(IV) oxidation is much less (0.017
mol/m2-hr) than that expected from physical absorption of oxygen. This may occur
because at these low concentrations, there is barely any S(IV) at the interface. When
S(IV) is between 0.5 and 1.5 mM  (Series 2-C12 and Oi), it seems that the S(IV) oxidation
is equivalent to that which would result from the physical absorption of oxygen. Thus,
chlorine does not seem to enhance S(IV) oxidation in this case. At high S(IV)
concentrations (around 4 mM), the slope of the line for Series 3-C12 and C>2 is steeper, and
the S(IV) oxidation is much greater at 0.099 mol/m2-hr.  Chlorine may be catalyzing
S(IV) oxidation, but it is hard to tell from looking at the figure. More data would need to
be taken in this range.

9.5 Discussion of electrochemical analyzer data

The discussion so far has only dealt with data from the IMS analyzer.  Figure 9-10
overlays the electrochemical data from Tables 8-6 and 8-7 with the IMS data in  Figure 9-
1.  Tables 8-6 and 8-7 only include data after the calibration of the electrochemical
analyzer was improved.
                                        41

-------
                                                           265 ppm
                                                           electrochemical data
                                                           21 ppm
                                                           oxygen
                                                           5 mM buffer
O
   10
         0.01
                              0.1
                                     [S(IV)HmM)
10
Figure 9-10. Electrochemical analyzer data overlaid onto IMS data

The trends are similar, but the data from the electrochemical analyzer do not fit the
model. The data obtained using the electrochemical analyzer are at a slightly lower k£
value of 2.3 x  10"5 m/s, while the model curves were calculated using the k£ value of
2.45 x 10"5 m/s (since this was the value for the data taken with the IMS analyzer). Even
if the lower  k£ value was used in the above model calculations, the electrochemical
analyzer data would not fit the model. In order for the data to fit the model, the liquid
film mass transfer coefficient used in the model would need to be approximately 1.23 x
10"5 m/s.

Figure 9-11  shows the above electrochemical data in a separate figure to better see trends.
As expected, the percent gas film resistance increases (signifying enhanced reaction) as
S(IV) increases until 100% gas film control is reached. Figure 9-11 also shows that at the
low chlorine concentration, the absorption is more likely to approach gas film control at
lower S(IV) concentrations.  This is expected since at lower chlorine concentrations, it
takes less  S(IV) to react completely with the chlorine.  Figure 9-11 also shows that the
point at 0.43 mM S(IV) seems to have a lower reaction rate than  the points with no S(IV).
This is equivalent to the S(IV) inhibition seen with the IMS data.
                                        42

-------
      100
                     275 ppm
                     130ppm
                                                            \
       0)
       0
            0
                               0.5  [S(IV)]b(mM)  1
       1.5
Figure 9-11. Chlorine absorption in 0 - 2 mM S(IV) in 50 mM buffer using
electrochemical analyzer
Figure 9-12 shows that the flux of chlorine is linear with bulk S(IV), signifying a region
controlled by S(IV) diffusion to the interface.
     2.5E-08

M*   2.0E-08

1   1.5E-08
 *  1.0E-08
 ^  5.0E-09

    O.OE+00
                       + 275 ppm
                       • 130 ppm
                0
                               0.5[S(IV)]b(mM)i
1.5
Figure 9-12. Effect of chloride seen from data obtained using electrochemical
analyzer

Figure 9-12 also shows that chloride (up to 20 mM) has no effect on the rate of chlorine
absorption in S(IV). The two marked points have increased levels of chloride (stock
solution of NaCl was added to reactor). The other points have chloride concentrations
resulting from only chlorine absorbing to form chloride (no external addition of chloride).
The chloride concentrations for these points range from 0.1 to 2 mM.  The chloride does
                                     43

-------
not seem to have any effect on the chlorine absorption since the two points with elevated
chlorine seem to follow the trend of the other points. Thus, at chloride concentrations
less than 0.02 M, there is no effect on the chlorine reaction with S(IV). This makes sense
since chloride should not affect the Ci2/S(IV) reaction if the reaction is irreversible.

Even the earlier electrochemical analyzer data (prior to data in Table 8-6) show that
S(IV) enhances chlorine absorption. However, these data cannot be rigorously analyzed
because of the analyzer problems mentioned earlier. Table 8-1 does show that pure
sulfite (which is at a higher pH than the buffered S(IV) solutions) may enhance
absorption more than the buffered S(IV) solutions. For the pH 7-8.5 data, the reaction
rate is so fast that the system is essentially gas film controlled.

9.6 Mercury removal in a typical limestone slurry scrubber

The expected mercury removal in a limestone slurry scrubber can be predicted using the
extracted rate constant for the Cl2/S(IV) reaction, a preliminary rate constant for Hg/Cl2
(Zhao and Rochelle, 1999), and typical mass transfer characteristics for a scrubber. Table
9-3 tabulates the parameters used in the model. The value for k2,s(iv) at 55°C was
estimated from the value at 25°C.  The model must be supplied with a given chlorine
inlet and a constant S(IV) concentration. The model accounts for the two simultaneous
reactions occurring at the gas/liquid interface: the depletion of chlorine through reaction
with S(IV) (k2,s(iv)) and the reaction of elemental mercury with chlorine (k2,Hg).

Table 9-3. Parameters used to predict mercury removal

k2,sov) (L/mol-s)
k2,Hg (L/mol-s)
kg (kmol/s-atm-m2)
Dcl2 (m2/s)
DHg (m2/s)
HHS (atm-m3/kmol)
25°C
2xlOy
1.7 x 101S
0.001
1.48 x 10"y
1.19 x 10"y
8.91
55°C
2x 1011
1.4 x 1017
0.001
2.15xlO"y
2.21 x 10"y
35.64
In a limestone slurry scrubber, chlorine absorption will be gas film controlled.  Thus,
Equation 5-8 can be used to calculate the flux of chlorine. The bulk chlorine in the
scrubber depends on the number of gas phase mass transfer units (Ng), which is defined
as kgA/G. Equation 9-7 shows this dependence. The total number of gas phase mass
transfer units in a typical scrubber is 6.9.
                              Pci2;b =pci2;mexp(-Ng)
(9-7)
The chlorine flux calculated from Equation 5-8 must equal the flux from Equation 5-4,
thus allowing the concentration of chlorine at the interface to be determined when the
S(IV) concentration is provided.  The interfacial chlorine concentration is very important
in predicting mercury absorption.
                                        44

-------
To predict mercury absorption due to reaction with chlorine, an expression similar to
Equation 5-4 is used. The enhancement of mercury removal is:
                                                                            (9-8)
The rate at which mercury is absorbed is the product of the driving force (yng) and the
overall gas phase mass transfer coefficient (K0o) given in Equation 5-10. Thus, the rate
of mercury absorption in a scrubber is given by Equation 9-9:
•GdyHg=yHg—;-%— dA
Substituting Ng for kgA/G and integrating:
                                                                            (9-9)
                         YHg.in  3%
                                                  -dN
                                               (9-10)
Equation 9-10 was used to quantify mercury removal. Tables 9-4 and 9-5 and Figure 9-
13 show the mercury penetration (Hgout/Hgin) as a result of chlorine injection to the
scrubber. The model curves were calculated at constant S(IV) concentrations of 1 and  10
mM.

Table 9-4. Mercury penetration in limestone slurry scrubber at 25°C
CI2,in(ppm)
0.1
1
10
1 mM S(IV)
0.0249
0.0057
0.0021
10mMS(IV)
0.0554
0.0114
0.0032
Table 9-5. Mercury penetration in limestone slurry scrubber at 55°C
Cl2,in(ppm)
0.1
1
10
1 mM S(IV)
0.0293
0.0065
0.0023
10mMS(IV)
0.0649
0.0133
0.0036
                                       45

-------
    0.1
  o>
  c
 = 0.01
  a>
  a>
  Q.
  0)
        0.1                     C'2,;n(PPm)   1
Figure 9-13. Predicted mercury penetration
10
Mercury removal increases (penetration decreases) as the chlorine injected increases.
Mercury removal decreases as S(IV) increases due to greater depletion of chlorine at the
interface.  At the higher temperature, there is less chlorine at the interface due to the
higher reaction rate of chlorine with S(IV).  However, since the reaction rate of mercury
and chlorine also increases with temperature, significant mercury removal still occurs.
Based on this model, only 1 ppm chlorine is needed to obtain 99 % mercury removal.
Since chlorine absorption is gas film controlled, 99.9% chlorine removal will be achieved
due to reaction with S(IV).
                                        46

-------
 References

Ashour, S.S., E.B. Rinker, and O.C. Sandall, "Absorption of Chlorine into Aqueous
       Bicarbonate Solutions and into Aqueous Hydroxide Solutions," AIChE J., 42(3),
       671-682(1996).

Askew, W. C., and S. J. Morisani, "Determining Chlorine Concentrations in Air and
       Water Samples for Scrubbing  Studies Using Ion Chromatography," J.
       Chromatogr. Sci., 27, 42-46 (1989).

Brian, P.L.T., J.E. Vivian, and C. Piazza, "The Effect of Temperature on the Rate of
       Absorption of Chlorine into Water," Chem. Eng. Sci., 21, 551-558 (1966).

Chang, C.S., "Mass Transfer with Equilibrium Chemical Reaction, Sulfur Dioxide
       Absorption in Aqueous Solutions," Ph.D. dissertation, The University of Texas at
       Austin, December 1979.

Critchfield, J.E., "CCh Absorption/Desorption in Methyldiethanolamine Solutions
       Promoted with Monoethanolamine and Diethanolamine: Mass Transfer and
       Reaction Kinetics," Ph.D. Dissertation, The University of Texas at Austin, 1988.

Danckwerts, P.V., Gas-Liquid Reactions, McGraw-Hill Book Co., New York, NY
       (1970).

Dutchuk, M.J., "Nitrogen Dioxide Absorption in Aqueous Dithionite," M.S. Thesis, The
       University of Texas at Austin, May 1999.

Ernst, W.R., B. Indu,  and M.F. Hoq, "Influence of Mercuric Nitrate on Species and
       Reactions Related to Chlorine Dioxide Formation,"  Ind. Eng. Chem. Res., 36, 11-
       16 (1997).

Fedorovskaya, L.F., V.A. Skripnik, L.I. Kravetskii, and I.M. Umanskaya, "Mechanism
       and Kinetics of Mercury Oxidation by Chlorine-Containing Solutions,"
       Translated from Zhurnal Prikladnoi Khimii, 52(6), 1233-1237 (1979).

Fogelman, K.D., D.M. Walker, and D.W. Margerum, "Non-metal Redox Kinetics:
       Hypochlorite and Hypochlorous Acid Reactions with Sulfite," Inorg. Chem., 28,
       986-993 (1989).

Gordon, G., B. Slootmaekers, S. Tachiyashiki, and D. Wood, "Minimizing Chlorite Ion
       and Chlorate Ion in Water Treated with Chlorine Dioxide," Amer. Water Works
       Assoc. L, 82,  160-165 (1990).

Hall, B.B., "An Experimental Study of Mercury Reactions in Combustion Flue Gases,"
       Ph.D. dissertation, Goteborgs Universitet (Sweden), 1992.

Keating, M.H., K.R. Mahaffey, R. Schoeny, G.E. Rice, and O.R. Bullock, "Executive
       Summary," Mercury Study Report to Congress (1),  EPA/452/R-97-003 (NTIS
                                       47

-------
       PB 98-124738), U.S. Environmental Protection Agency, Office of Air Quality
       Planning and Standards, Research Triangle Park, NC,  December 1997.

Kolthoff, I.M., and R. Belcher, Volumetric Analysis, Volume III, 199-374, Interscience
       Publishers, Inc., New York, NY (1957).

Jensen, J.S., and G. R. Helz, "Rates of Reduction of N-Chlorinated Peptides by Sulfite:
       Relevance to Incomplete Dechlorination of Wastewaters," Environ. Sci. Technol.,
       32,516-522(1998).

Lifshitz, A., and B. Perlmutter-Hayman, "The Kinetics of the Hydrolysis of Chlorine. III.
       The Reaction in the Presence of Various Bases, and a Discussion of the
       Mechanism," J. Phys. Chem., 66, 701-705 (1962).

Livengood, C.D., and M.H. Mendelsohn, "Improved Mercury Control in Wet Scrubbing
       Through Modified Speciation," presented at the EPRI-DOE-EPA Combined
       Utility Air Pollutant Control Symposium, Washington, D.C., August 25-29,1997.

Shen, C.H., "Nitrogen Dioxide Absorption in Aqueous Sodium Sulfite," Ph.D.
       dissertation, The University of Texas at Austin, May 1997.

Spalding, C.W., "Reaction Kinetics in the  Absorption of Chlorine into Aqueous Media,"
       AIChEL, 8(5), 685(1962).

Suzuki, K., and G. Gordon, "Stoichiometry and  Kinetics of the Reaction between
       Chlorine Dioxide and Sulfur(IV) in Basic Solutions," Inorg. Chem., 17(11), 3115-
       3118(1978).

Wang, T. X., and D. W. Maregerum, "Kinetics of Reversible Chlorine Hydrolysis:
       Temperature Dependence and General-Acid/Base-Assisted Mechanisms," Inorg.
       Chem., 33, 1050-1055 (1994).

Zhao, L.L., "Mercury Absorption in Aqueous Solutions," Ph.D. dissertation, The
       University of Texas at Austin, May 1997.

Zhao, L.L., and G.T. Rochelle, "Mercury Absorption in Aqueous Hypochlorite," Chem.
       Eng Sci., 54, 655-662 (1999).

Zhao, L.L., and G.T. Rochelle, "Mercury Absorption in Aqueous Permanganate," AIChE
      L, 42, (1996).
                                       48

-------
Appendix A Gas film mass transfer coefficient data
Table A-l shows the kg values calculated using the IMS analyzer data.  All of the data are
for chlorine absorption in 0.28 M sodium hydroxide (NaOH).
Table A-l. Gas film mass transfer coefficient (kg), IMS analyze
nq (rpm)
750
623
500
762
CI2,in (ppm)
21.2
21.2
21.2
21.2
CI2,out (ppm)
2.47
2.69
3.00
2.50
Nci2(kmol/m^-s)
1.89E-09
1 .87E-09
1 .84E-09
1.89E-09
kq(mol/s-atm-m^)
0.765
0.694
0.613
0.755
Table A-2 shows the kg values calculated using the electrochemical analyzer data.  The
data use the improved calibration shown in Figure 8-2, but the data were taken before the
experimental modifications were made to reduce scatter.

Table A-2. Gas film mass transfer coefficient (kg), electrochemical sensor  analyzer
nq (rpm)
428
357
292
546
691
514
403
303
714
CI2,in (ppm)
197
197
197
197
197
197
197
197
197
CI2,out (ppm)
35.0
37.8
40.4
32.2
28.5
34.3
37.6
41.5
28.5
Nci2(kmol/m'i-s)
1 .63E-08
1 .60E-08
1 .57E-08
1 .65E-08
1 .69E-08
1 .63E-08
1 .60E-08
1 .56E-08
1 .69E-08
kq(mol/s-atm-m'£)
0.46
0.42
0.39
0.51
0.59
0.48
0.43
0.38
0.59
                                       49

-------
Table A-3 shows the kg values calculated using the electrochemical analyzer data after
the experimental modifications were made to reduce scatter.
Table A-3. Gas film mass transfer coefficient (kg), electrochemical analyzer after
modifications to reduce scatter
nq (rpm)
629
474
352
707
484
710
710
710
490
362
711
616
519
412
330
597
342
720
428
288
704
CI2,in (ppm)
195
195
195
195
195
195
149
93
93
93
93
197
197
197
197
197
197
197
197
197
197
CI2,out (ppm)
26.2
30.6
34.5
24.9
31.6
25.3
19.7
12.8
16.5
18.2
12.6
27.8
31.7
36.7
41.3
30.9
39.8
26.6
36.1
41.9
27.0
NC|2(kmol/m'i-s)
1 .69E-08
1 .65E-08
1 .61 E-08
1.71E-08
1 .64E-08
1 .70E-08
1 .31 E-08
8.25E-09
7.87E-09
7.69E-09
8.27E-09
1 .70E-08
1 .66E-08
1 .61 E-08
1 .56E-08
1 .67E-08
1 .58E-08
1.71 E-08
1 .61 E-08
1 .56E-08
1.71 E-08
kq(mol/s-atm-m'£)
0.65
0.54
0.47
0.69
0.52
0.67
0.66
0.65
0.48
0.42
0.65
0.61
0.52
0.44
0.38
0.54
0.40
0.64
0.45
0.37
0.63
                                       50

-------
Appendix B Liquid film mass transfer coefficient data and correlations

Table B-l lists all the data used to determine^c\2 •  Chlorine desorption was measured
from a sodium hypochlorite solution in 0.1 M HC1. These data were all obtained using
the IMS analyzer.

Table B-l. Data used to determine kLjCi2 correlations
nL(rpm)
729













305




504





734





699






t(min)
0
6
12
18
20.6
24
30
36
42
48
54
60
64.8
75
0
6
12
18
25.1
0
6
12
14.2
18
24
0
4.2
12
18
24
32.4
0
6
9.6
12
18
24
30
CI2,out (ppm)
284
264
248
232
222
213
203
186
168
158
148
135
129
112
56.7
53.8
51.2
50.2
47.3
65.4
61.9
58.6
57.0
56.3
52.1
66.4
63.5
57.0
53.4
50.5
45.7
289
275
264
257
238
218
203
lnPcl2(atm)
-8.17
-8.24
-8.30
-8.37
-8.41
-8.46
-8.50
-8.59
-8.69
-8.76
-8.82
-8.91
-8.96
-9.10
-9.78
-9.83
-9.88
-9.90
-9.96
-9.63
-9.69
-9.74
-9.77
-9.78
-9.86
-9.62
-9.66
-9.77
-9.84
-9.89
-9.99
-8.15
-8.20
-8.24
-8.27
-8.34
-8.43
-8.50
 (Continued)
                                       51

-------
Table B-l. Continued
nL(rpm)



228





600






306




514





718





t(min)
36
42
48
0
6
12
18
24
27.6
0
3.6
7.2
9.6
15.6
21.6
25.2
0
6
12
18
24
0
2.4
8.4
14.4
20.4
26.4
0
2.4
8.4
14.4
20.4
26.4
CI2,out(ppm)
186
171
156
62.3
58.7
58.0
54.8
54.1
51.5
104.0
98.4
95.8
91.9
88.7
84.1
78.6
45.0
43.4
40.4
39.1
36.5
54.4
52.8
49.5
47.3
44.3
41.1
52.5
50.5
46.6
43.4
39.8
37.5
lnPcl2(atm)
-8.59
-8.67
-8.76
-9.68
-9.74
-9.75
-9.81
-9.82
-9.87
-9.17
-9.23
-9.25
-9.29
-9.33
-9.38
-9.45
-10.01
-10.05
-10.12
-10.15
-10.22
-9.82
-9.85
-9.91
-9.96
-10.02
-10.10
-9.86
-9.89
-9.97
-10.05
-10.13
-10.19
For each HL, a plot of In Pcl2 against time was generated, and the liquid film mass transfer
coefficient was determined from the slope of the line. Table B-2 lists the k^a values
obtained from the slopes of these lines.
                                        52

-------
     -9.2

         0          20         40

Figure B-l. Extracting k°L,Ci2 at 729 rpm

   -9.75
         0             10            20

Figure B-2. Extracting k°L,ci2 at 305 rpm

     -9.6 -,

    -9.65 4-
   w-9.75 -|
   o
  £ -9.8

    -9.85 -

     -9.9
                            nL=504 rpm
        0             10 t™11        20

Figure B-3. Extracting k°L,ci2 at 504 rpm
30
  30
                                      53

-------
      -9.6
    -9.65

  I "975 "
  ^  -9.8
  CL -9.85
  ~   -9.9 H
    -9.95
      -10
                             nL=734 rpm
          0         10        20  t(min)so
Figure B-4. Extracting k°L,ci2 at 734 rpm
            0             20   t(min)   40
Figure B-5. Extracting k°Ljci2 at 699 rpm
                                                  40
-8 n
a -\
-o. I
<
a o
^->
C Q Q
4.1 -O.O
(0
^ a A
CM -Q.'f
o
Q. 0 C
C'b'b
(3 C
-O.D
o 7
-o./
Q Q

^^^
^^v^
^X* nL= 699 rpm

V.
^s^
^
                                                      60
-9.65 n
4
Q "7^
-y./o
Q £
-y.o

>
*>»v^
^^"^....^^ nL= 228 rpm
^\^
* 	
"%
0 5 10t(min)15 20 25 3
Figure B-6. Extracting k°L,ci2 at 228 rpm
                                     54

-------
-9.15 n
<
Q O
Q OR
4-1 93
CM
n° Q ^
Q_ -y.oo
c
Q /I

9c
.O n
(
»
^^^^
s^ nL=600rpm
^^^
^x»
^\

) 5 10 15 20 25 3
t(min)
Figure B-7. Extracting k°L,Ci2 at 600 rpm
      -9.95 -,
           0      5      10t(min)15
Figure B-8. Extracting k°L,Ci2 at 306 rpm
       -9.8 n
      -9.85 -
  ^   -9.9
             20
           0
10
                           t(min)
Figure B-9. Extracting k°L,ci2 at 514 rpm
20
      25
30
  30
                                      55

-------
            0       5      10     15      20      25      30
                              t(min)

Figure B-10. Extracting k°L,Ci2 at 718 rpm

Table B-2. Physical liquid film mass transfer coefficient for chlorine
r\(rpm)
729
305
504
734
699
228
600
306
514
718
k°>cl2 (m/s)
2.73E-05
1 .51 E-05
2.01 E-05
2.51 E-05
2.86E-05
1 .37 E-05
2. 23 E-05
1 .90E-05
2. 25 E-05
2.80E-05
                                      56

-------
1. REPORT NO. 2.
EPA-600/R-01-054
4. TITLE AND SUBTITLE
Chlorine Absorption in S(IV) Solutions
7. AUTHORS
Sharmistha Roy and Gary T. Rochelle
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The University of Texas at Austin
Department of Chemical Engineering
Austin, TX 78712
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
August 2001
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1 1 . CONTRACT/GRANT NO.
EPA Cooperative Agreement
CR 827608-01-1
13. TYPE OF REPORT AND PERIOD COVERED
Final Report; 6/99-12/00
14. SPONSORING AGENCY CODE
EPA/ 600/1 3
                                 TECHNICAL REPORT DATA
                       (Please read instructions on the reverse before completing)
15. SUPPLEMENTARY NOTES

 2683.
APPCD project officer is Theodore G. Brna, Mail Drop 04, 919/541-
16.ABSTRACT jhe report gives results of measurements of the rate of Chlorine (CI2) ab-
sorption into aqueous sulfite/bisulfite -- S(IV) -- solutions at ambient temperature
using a highly characterized stirred-cell reactor.  The reactor media were 0 to 10 mM
S(IV) with pHs of 3.5-8.5. Experiments were performed using 20-300 ppm CI2 in nitrogen
(N2) or air.  CI2 absorption was modeled using the theory of mass transfer with  chemi-
cal reaction.  CI2 reacts quickly with S(IV) to form chlorine and sulfate.  CI2 absorp-
tion  is enhanced by increasing pH and S(IV) concentration.  The rate constant for the
reaction of CI2 with S(IV) was too rapid to be precisely measured using the existing
stirred-cell reactor, due to mass transfer limitations.  However, The most probable
value of the rate constant was determined to be 2 x 10 to the 9th power L/mol-s.
These results are relevant to the  simultaneous removal of CI2, sulfur dioxide, and
elemental mercury (Hg) from flue gas. The developed model shows that good removal of
both CI2 and Hg should be possible with the injection of 1-10 ppm CI2 to an existing
limestone slurry scrubber. These results may also be applicable to scrubber design for
removal  of CI2 in the pulp and  paper and other industries.
17.
                                KEYWORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Pollution Slurries
Flue Gases
Chlorine
Sulfur Dioxide
Mercury (Metal)
Scrubbers
Limestone
18. DISTRIBUTION STATEMENT
Release to Public
b. IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This Page)
Unclassified
c. COSATI Field/Group
13B 11G
21B
07B
17A, 131
08G
21. NO. OF PAGES
64
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE

-------