Radian Corporation
8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512/454-9535
FINAL REPORT
VOLUME I
GAP Contract No. EHSD 71-5
A THEORETICAL STUDY OF NOX
ABSORPTION USING AQUEOUS ALKALINE
AND DRY SORBENTS
CHEMICAL RESEARCH • SYSTEMS ANALYSIS • COMPUTER SCIENCE • CHEMICAL ENGINEERING
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512/454-9535
RAD-7l-200-007-0l
FINAL REPORT
VOLUME I
OAP Contract No. EHSD 71-5
A THEORETicAL STUDY OF NOx
ABSORPTION USING AQUEOUS ALKALINE
AND DRY SORBENT S
Presented. to:
OFFICE OF AIR PROGRAMS
ENVIRONMENTAL PROTECTION AGENCY
411 West Chapel Hill Street
Durham, North Carolina 27701
31 December 1971
Prepared by:
Philip S. Lowell
Principal Scientist
Terry B. Parsons
Engineer/Scientist
CHEMICAL RESEARCH. SYSTEMS ANALYSIS. COMPUTER SCIENCE. CHEMICAL ENGINEERING
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. 80X 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
ABSTRACT
This study was carried out to develop a theoretical
description of aqueous scrubbing processes for NOx emission
control. The theoretical description is necessary to provide
a basis for process development. The important chemical species
present and the reactions they undergo in the sorption process
were defined. An equilibrium model was developed so that the
concentrations of the significant nitrogen-oxygen species present
in the gas phase could be calculated under given conditions. A
sorption mechanism and rate limiting step were proposed and an
experimental program for testing them and providing engineering
data was defined. Potential metal oxide sorbents were evaluated
for effectiveness and efficiency on the basis of the thermo-
dynamics of the sorption reactions.
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
ACKNOWLEDGEMENTS
The authors wish to acknowledge the assistance of
EPA personnel under whose guidance this program was carried out.
In particular we wish to thank Dr. Joshua Bowen and Mr. Thomas
A. Kitt1eman (Project Officer, 1 August 1970 through 15 August
1971). We also appreciate the cooperative spirit of Mr. Luis
Garcia who directed the EPA experimental program and was Project
Officer, 15 August through 30 December 1971.
We also wish to thank the following members of the
Radian staff for their contributions: Klaus Schwitzgebe1 and
Nancy Phillips in chemistry, Murray Wells and James Phillips in
chemical engineering, and Thomas Strange in computer assistance.
We are indebted to Dr. Kurt H. Stern of the
Electrochemistry Branch, Naval Research Laboratory, Washington,
D. C. for sending us a copy of his monograph "High Temperature
Properties and Decomposition of Inorganic Salts. Part 3.
Nitrates and Nitrites."
ii
-------
'-
Radian Corporation
VOLUME I
1.0
2.0
2.1
2.2
2.3
3.0
3.1
3.2
3.2.1
3.2.2
3.2.2.1
3.2.2.2
3.3
4.0
4.1
4.2
4.3
4.4
5.0
5.1
5.2
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
TABLE OF CONTENTS
INTRODUCTION. . .
Page
................1
LITERATURE SURVEY
. . . . . .
. . . .
. . . . . .
Information Sought. . . . . . . . . . . . . . . .
Sources of Information. . . . . . . . . . . . . .
Information Found. . . . . . . . . . . . . . . .
PROBLEM DEFINITION. .
. . . .
. . . .
. . . . . .
Scope of the Problem. .............
Theoretical Framework. . . . . . . . . . . . . .
Species and Reactions. . . . . . . . . . . . . .
Rate Controlling Step. . . . . . . . . . . . . .
Mechanisms Suggested in the Literature. . . . . .
Proposed Method of Determining Rate Controlling
Step for Flue Gas Treatment. . . . . . . . . . .
Summary of Problem Definition. . . . . . . . . .
THE COMPOSITION-CONCENTRATION MODEL.
. . . . . .
Reactions Considered in the Gas Phase
Equilibrium Model. . . . . . . . . . . . . . . .
Problem Formulation and Method of Solution. . . .
Comparison of Calculated and Measured Results. .
Application of the Gas Phase Equilibrium Model
to Predict Rate Controlling Step and Absorp-
tion Mechanism. . . . . . . . . . . . . . . . . . 38
EXPERIMENTAL PROGRAM. . . . . . . . . . . . . . . 43
Apparatus and Data Collec ted. . . . . . . . . . . 44
Practical Experience Gained. . . . . . . . . . . 44
iii
5
5
6
7
8
8
11
12
18
18
28
29
31
33
33
36
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
TABLE OF CONTENTS
6.0
6.1
6.2
6.3
7.0
7.1
7.2
7.3
7.3.1
7.3.2
7.3.3
7.4
7.4.1
7.4.2
7.4.3
7.4.4
7.4.5
8.0
8.1
8.2
9.0
10.0
ENGINEERING ANALYSIS OF SORPTION DATA. . .
. . . .
Problem Formulation. . . . . . . . . . . . . . . .
Data Used iri Correlation. . . . . . . . . . . . .
Results. . . . . . . . . . . . . . . . . . . . . .
SCREENING OF CANDIDATE SORBENT S ON THE BASIS OF
THE THERMODYNAMICS OF THE SORPTION RFACTIONS . . .
Description of the Processes for which Candidate
Sorbents were Screened. . . . . . . . . . . . . .
Thermodynamic Basis for Screening. . . . . . . . .
Data Collection and Calculations. . . . . . . . .
Calculation of N20a and N206 Pressures Over the
Sar be n t . . . . . . . . . . . . . . . . . . . . . . .
Calculation of N20a and Na06 Pressures in the
Flue Gas. . . . . . . . . . . . . . . . . . . . .
Results of Calculations. . . . . . . . . . . . . .
Results of Thermodynamic Screening. . . . . . . .
General Considerations. . . . . . . . . . . . . .
Screening for Dry and Aqueous Processes Based on
Nitrite Formation. . . . . . . . . . . . . . . . .
Screening for Dry and Aqueous Processes Based on
Nitrate Formation. . . . . . . . . .
. . . . . . .
Screening Metal Carbonate Sorbents.
.......
Summary and Conclusions.
. . . . . . . . .
. . . .
THERMODYNAMIC PROPERTIES. . . . . . . . . . . . .
Standard State Thermodynamic Properties and Heat
Capacity Data. . . . . . . . . . . . . . . . . . .
Measured Equilibrium Constants. . . . . . . . . .
SUMMARY OF RESULTS
. . . . . . . . . . . . . . . . 82.
BIBLIOGRAPHY. . . . .
. . . . . . . . . . . . . . 84
iv
Page
49
49
51
53
55
55
59
62
62
63
63
66
66
69
72
72
75
77
77
79
-------
Radian Corporation
TABLE 3-1
TABLE 3-2
TABLE 3-3
TABLE 3-4
TABLE 4-1
TABLE 4-2
TABLE 4-3
TABLE 5-1
TABLE 6-1
TABLE 6-2.
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
LIST OF TABLES
Composition of Gas Mixture from which
Nitrogen Oxides Must be Removed. . .
. . . . .
Solutions in Equilibrium with a Gas
Containing 0.10 atm COa. . . . . . .
. . . . .
Gas Phase Species in the NOx-HaO System.
. . .
Aqueous Phase Species in the NOx-HaO System. .
Comparison of Calculated and Measured
Partial Pressures of NOa . . . . . . .
. . . .
Comparison of Calculated and Measured Total
Number of Moles. . . . . . . . . . . . . . . .
Equilibrium Compositions in the Gas Phase
System NOx-HaO at 60°C, 1 atm Total Pressure,
and 8 mo1e% HaO. . . . . . . . . . . . . . . .
Summary of Data Collected at OAP Laboratory,
May, 1971 (mo1es/min x lOa). . . . . . . . . .
Recalculated Values of CNO and CNO in the
a
Inlet and Effluent Gas in NOx Sorption
Experiments. . . . . . .
. . . . . . .'
. . . .
Results of Mass Transfer Coefficient
Calculations. . . . . . . . . . . .
. . . . .
v
Page
9
10
12
14
37
38
39
46
52
53
-------
!--
I
I
I
Radian Corporation
LIST OF TABLES
TABLE 7-1
TABLE 7-2
TABLE 7-3
TABLE 7-4
TABLE 8-1
FIGURE 3-1
FIGURE 3-2
FIGURE 3-3
FIGURE 5-1
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
Page
Metal Oxides Screened for Applicability in
NOx Removal Processes. . . . . . . . . . .
. . . 56
Metal Oxides Unsuitable for Use in Processes
Based on Nitrite Formation. . . . . . . . .
. . 71
Metal Oxides Unsuitable for Use in Processes
Based on Nitrate Formation. . . . . . . . .
. . 73
Potential Sorbents After Screening
. . . .
. . . 76
Selected Values for Equilibrium Constants.
. . . 8]
LIST OF ILLUSTRATIONS
Species and Reactions in the Gas and
Aqueous Phase System NOx-H20 . . . .
. . .
. .. 17
Predicted Plate Efficiencies Based on the
Theoretical Rate Equations of Andrew and
Hanson. .
. . . 23
. . .
. . . .
. . . . . . . . .
Comparison of Andrew and Hanson's Measured
and Predicted Total Plate Efficiencies. .
. . . 24
Flow Diagram of Equipment Used in OAP
In-House Experiments. . . . . . . . .
. . . . . 45
vi
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. 80)\ 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
LIST OF ILLUSTRATIONS
Page
FIGURE 7-1.
FIGURE 7-2
FIGURE 7-3
FIGURE 7-4
FIGURE 7-5
Flow Diagram for Aqueous Process. . .
. . . . . 57
Comparison of N;Os
and Vapor Pressure
Sorbents . . . . .
Partial Pressure in Flue Gas
of N20s over Metal Oxide
. . .
. . 64
. . . .
. . . . . .
Comparison of NaOs Partial Pressure in Flue
Gas and Vapor Pressure or NaOs over Metal
Oxide Sorbents . . . . . . . . . . . . . . .
. . 65
Comparison of N203 Partial Pressure in Flue
Gas and Vapor Pressure of NiOs over Metal
Carbonate Sorbents . . . . . . . . . . . .
. . . 67
Comparison of N20s Partial Pressure in Flue
Gas and Vapor Pressure of N20s over Metal
Carbonate Sorbents . . . . . . . . . . . .
. . . 68
vii
-------
Radian Corporation 8500 SHOAL CREEK ILVD. . P. O. lOX 9948 . AUSTIN, TEXAS 78766 .
TELEPHONE 512 - 454-9535
1.0
INTRODUCTION
Nitrogen oxide emissions from stationary sources have
been identified as the potential source of over half the NOx
emissions projected for the next thirty years. A recently
published survey (BA-003) estimated that over three and one-
half milli.on tons of NOx calculated as NOz were emitted by
electric utilities alone in one year. The Environmental
Protection Agency (EPA) has funded investigations into methods of
reducing NOx emissions in flue gases such as those produced by
electric utilities. Only a few methods, namely combustion
modification and flue gas treatment including aqueous scrubbing
and selective reduction, were found to offer potential for NOx
control. Further, none of the suggested flue gas treatment pro-
cesses have been applied to NOx control for electric utility flue
gases. The methods are still in the early stages of process
development.
This report presents the results of an effort to
theoretically describe a process for reducing NOx emissions in
flu~ gases such as those produced by electric utilities. The
process under investigation was flue gas treatment by aqueous
scrubbing. The goal of this work was to develop a theoretical
description of the aqueous scrubbing process. A theoretical
description was a necessary prerequisite for successful process
development because the basic chemical and chemical engineering
aspects of the process had not been previously described. Four
steps needed to make a theoretical description to aid in process
development will now be given.
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512.454-9535
1.
Literature Survey and Problem Definition
These tasks are described in Sections 2.0 and 3.0
of the report. They involved becoming familiar with
what had already been published concerning aqueous
scrubbing for flue gas treatment. In addition, they
involved defining the scope of the problem by specify-
ing what important chemical species and reactions
take place in the process and what the rate controlling
steps might be. The problem definition revealed that
flue gas composition and the concentration of the many
different nitrogen species present determined some
very basic aspects of the process. Specifically they
gave a strong indication of what the rate controlling
step and the sorption mechanism might be. This fact
pointed the way for the second step in developing a
theoretical description.
2.
Development of the Composition-Concentration
Model
The second step in providing a theoretical
description of the aqueous scrubbing process was
to develop a model by which the concentrations of
the various nitrogen-oxygen compounds present in
aqueous scrubbing systems could be calculated.
This model is described in Section 4.0 of the
report.
-2-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 45'-15;5
3.
Cooperation with EPA In-House Experimental
Program
A third phase of the program involved experimentally
testing the concepts proposed in the theoretical
description. An in-house experimental program to
study NOx sorption in packed towers was in progress
during the period of performance of this study.
While the hypotheses developed from the theoretical
description were not completely tested during the
rough screening studies carried out by EPA, some
valuable information as well as considerable practical
experience were gained. These studies are discussed
in Sections 5.0 and 6.0 of the report.
4.
Screening of Candidate Metal Oxide (Hydroxide)
Sorbents
One of the main goals of the program was to use the
theoretical description to predict which metal
oxides (hydroxides) would be the most effective and
efficient sorbents in an aqueous or a dry scrubbing
process. The screening of candidate sorbents was
done on the basis of the thermodynamics of their
reactions with nitrogen oxides. This task is discussed
in Section 7.0 of the report. The results are limited
to processes in which a solid nitrate is formed in the
sorption step and decomposed in the regeneration step.
Throughout the course of this study there were numerous
cases in which it was necessary to know the thermodynamic proper-
ties of the nitrogen oxides, oxyacids, the metal oxide (hydroxide)
sorbents and the sorption products. In addition, values for
equilibrium constants for decomposition, dissociation, ionization,
-3-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454.9535
vaporization and sorption reactions were needed. Finally,
activity coefficients for aqueous phase species were necessary.
These thermodynamic data were obtained from the literature using
careful evaluation and in some cases, recalculation. Some of
the properties for solids were estimated. The acquisition and
use \of thermodynamic properties are discussed in Section 8.0
of the report. Section 9.0 gives a summary of the results of
the program. The references cited in Volume I are given in
Section 10.0. Volume II contains the actual details of most
of the work. As work packages were completed during the study,
they were documented in detail in technical notes. The technical
notes which are referred to quite frequently in Volume I are
contained in Volume II. Volume II also contains a listing of'
the computer program written to perform the composition-concentra-
tion model calculations.
-4-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 5'12.454-9535
2.0
LITERATURE SURVEY
The first task in the investigation of aqueous alkaline
scrubbing was to collect existing published data and become
familiar with what had already been reported concerning the
problem. This section describes the type of information that
was sought, the sources that were used to collect information
and the general types of information that were found.
2.1
Information Sought
The goal of the program was to describe the
thermodync~ics of regenerative aqueous alkaline processes
for removing NO and N02 from gas mixtures. This was to be
the thermodyn~ic basis for screening and selecting potential
metal hydroxide sorbents. A further goal of the program was
to determine the sorption mechanism and develop a description
of the chemical engineering unit operations involved. This
was to be done in conjunction with EPA inhouse experiments.
The accomplishment of these goals required the kinds
of information described below.
Reactions between nitrogen oxides and aqueous
solutions: equilibrium constants, kinetic
data, mechanisms.
Physical, chemical and thermodynamic properties
of gaseous, aqueous and solid nitrogen oxides
and oxyacids, metal nitrates, nitrites, hydroxides,
and carbonates: standard state thermodynamic
properties, vapor pressures, solubilities, activ-
ity coefficients, thermal stabilities.
-5-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 787S7 . TELEPHONE 512 - 454-9535
Descriptions of existing processes for
aqueous sorption of NO and/or NOa from
gaseous mixtures.
2.2
Sources of Information
Three sources for acquisition of pertinent literature
were used. The source from which most of the references of
interest were obtained was Chemical Abstracts. Cumulative
subject indices covering references published from 1947 to
1966, as well as semiannual subject indices for the period
1967 to June, 1969, were searched. The biweekly issues from
July, 1969, to October, 1970, were searched using the key word
index at the end of each issue. Slightly less than 1000 abstracts
were selected.
A second useful information source was publications
from EPA's Air Pollution Technical Information Center. Air
'Pollution Abstracts and Nitrogen Oxides: An Annotated Bibli-
, ography were both consulted. In addition, Tom Kittleman of EPA
provided the computer listing of abstracts from an APTIC search
on nitrogen oxides and absorption.
\
A third source of information was the final report
"Systems Study of Nitrogen Oxide Control Methods for Stationary
Sources" prepared for EPA under contract PH 22-68-55. Section
5.5.1 dealt with aqueous absorption of NOx and the bibliography
for that section gave several pertinent references. In addition,
the supplementary bibliography of 750 references was searched
and some additional titles were selected as potentially interest-
ing. Many of these were duplications of references already
found using APTIC publications.
-6-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
2.3
Information Found
The abstracts and titles collected during the
literature search were reread and filed in categories with
descriptive titles. These files were consulted at appropriate
later times when new phases of work were begun. Much of
the literature is reviewed in technical notes. Some of the
references collected were not found to be useful at any time
during the period of contract performance, but they are
potentially useful for future work. Some of the references
served to point out what isn't known rather than what is
known. Finally, the numerous abstracts collected concerning
descriptions of and operating data for aqueous scrubbing
processes were particularly useless for this work.
The details of what types of information were found
are discussed in Technical Note 200-007-14 in Volume II of
this report. The note contains a supplemental bibliography
which lists most of the references found during the literature
search in categories with descriptive titles. It should be
kept in mind that the categories are only broadly descriptive,
and some references could undoubtedly be filed in more than
one category. The filing of references in each category was
done by reading the abstract, since the titles alone are not
always indicative of the content.
-7-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
3.0
PROBLEM DEFINITION
The goal of this program
study of aqueous alkaline sorption
oxide (hydroxide) sorbents and use
help answer the following:
was to conduct a theoretical
processes involving metal
the theoretical framework to
What is the sorption mechanism (to be done
in conjunction with EPA experiments)?
What metal oxide (hydroxide) sorbents
are capable of or most effective for
removing NOx from flue gases?
3.1
?cope of the Problem
The most important goal was to develop a theoretical
description of sorption of NOx in the scrubber. This descrip-
tion, combined with experimental data, would indicate the
technical feasibility of aqueous sorption processes. In view
of the EPA experimental results, it appeared that the most
important question was whether or not the nitrogen oxides could
be removed in the scrubbing step. However, it is realistic
to consider also what will be done with the scrubbing liquor
after it contains the nitrogen oxides which have been removed
from the flue gas. Regenerative processes were to be considered.
After all, converting an air pollution problem into a water
pollution problem is not a practical solution. Therefore, the
goal was to derive a theoretical description not only for the
scrubber but for an entire process.
-8-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
A flue gas composition was assumed for
purposes in the thermodynamic screening studies.
tion is given in Table 3-1.
computational
The composi-
TABLE 3-1
COMPOSITION OF GAS MIXTURE FROM WHICH
NITROGEN OXIDES MUST BE REMOVED
Component
N~
CO~
H~O
O~
NOx
Approximate Concentration
(mo1e%)
76.
13
7
3
0.05
(500 ppm)
Gas mixtures having the composition shown in Table 3-1
are typical of those emitted by natural gas burning power plants
or oil or coal burning plants whose off gases have been treated
for S02 removal (LO-007). Vapor phase and aqueous phase inter-
action of nitrogen oxides and sulfur oxides is qui'te complicated.
The number of possible reactions which can take place imposes an
almost insurmountable problem for the development of a theoretical
description. Therefore, it is logical and useful to develop
a theoretical description for the already complicated system
described in Table 3-1 before the description of an even more
complicated system is attempted.
The composition of the gas mixture described in
Table 3-1 has important implications for the study of aqueous
alkaline scrubbing. The nitrogen oxides make up only 0.05 mole%
(500 ppm) of a typical flue gas, while 13 mole% CO2 is present.
An alkaline sorbent will remove CO2 as well as nitrogen oxides
-9-
-------
Radian Corporation 8SOO SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766' TELEPHONE 512. 454-953S
especially since the CO2 is present in much greater concentration.
The Radian aqueous phase equilibrium model (LO-007) was used to
calculate the results of sorption of a gas containing 10% CO2 into
scrubber solutions composed of .05 molar and 1.0 molar NaOH. The
calculated values of pH and ionic strength at 40, 50, and 60°C are
shown in Table 3-2.
TABLE 3-2
SOLUTIONS IN EQUILIBRIUM WITH A GAS
GAS CONTAINING 0.10 atm CO~
T (0 C)
40
50
60
pH at Equilibrium
Initial Conc. NaOH = .05 M Initial Conc. NaOH = 1.0 M
7.53
7.61
7.71
(I = .05)
8.51
8.56
8.62
(I = .84)
The calculated results shown in Table 3-2 have been
confirmed experimentally. Streight (81-012) reported absorbing
a synthetic waste gas containing .05% N02 and 2.4% CO2 using
"dilute caustic" of pH 12.8 in a 12-inch i. d. column with five
feet of packing. The CO2 converted the sorbent to sodium car-
bonate and the pH decreased to a little above seven. The impli-
cation is that it is impossible to carry out alkaline (OH-)
scrubbing on flue gas containing 13% CO2, The CO2 will be re-
moved by an alkaline solution until equilibrium is reached and
the pH is lowered to a value in the range of seven to nine.
Nitrogen oxides will then continue to be absorbed, liberating
CO2 .
-10-
()
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512 - 454-95)5
3.2
Theoretical Framework
A theoretical description of the processes considered
had to be developed as a basis for predicting which sorbents
would be effective in removing NO and NOa. In order to be useful
for engineering predictions, the theoretical framework had to
include the following.
What chemical species are present and
in what reactions are they involved during
sorption and regeneration?
What is the rate limiting step in the
sorption process and what species are
involved in the rate limiting step?
What are the free energy changes involved in
the processes and which sorbent has the most
efficient free energy change. An efficient
free energy change is one that is large enough
to supply sufficient driving force for the
process but not so large that a large free
energy reversal will be required in the re-
generation step.
The theoretical framework was then developed in the
following way. The possible species present and the reactions
they can undergo were listed and studied to discern what takes
place in the system. Next, all the diffusion, reaction, and
mass transfer steps involved in going from NOx(g) to NO;(~) and
NO;(L) were studied and the rate controlling step was proposed.
Then the problem was formulated in terms of the species signifi-
cant for the mass balance and in terms of the species involved
in the rate controlling step. As will be seen from the following
-11-
-------
Radian Corporation 8500 SHOAL CREEK ILVD. . P. O. lOX 9948 . AUSTIN, TEXAS '8766 . TELEPHONE 512 - 454-9535
discussion the problem of describing the path from NOx(g) to
NO;(,e) and N03(,e) is very complicated. First the possible species
and reactions are described. Then the rate controlling step is
discussed.
3.2.1
Species and Reactions
The nitrogen oxides to be removed from the flue gas
are NO, nitric oxide, and NOg, nitrogen dioxide. The sum of
their concentrations is usually referred to as NOx. Even though
the 500 ppm NOx typically present in flue gases is 90% nitric
oxide, NOx concentrations are usually reported as equivalent NOa.
There are several reasons for the preference for NOa. First,
many of the analytical methods used involve oxidation of nitric
oxide to NO:;j and measurement of total NOa concentration. Nitrogen
dioxide is easily measured spectrophotometric ally; it is a brown
colored gas. Also, NOg is more reactive than NO and can be
absorbed more easily by more solvents. Most research has
centered on studying the absorption of NOa rather than NO.
The concentration of NO. as equivalent NOa is a useless
description of the nitrogen oxides in the vapor phase if one is
concerned with the development of a theoretical framework. The
species listed in Table 3-3 all exist in some concentration in
the gas phase. The subscript (g) has been eliminated, so that
NO, for example, is understood to be NO(g)'
TABLE 3-3
GAS PHASE SPECIES IN THE NOx-H~O SYSTEM
NO
NO~
..
Ng03
Ng 04,
Og
Na
NaO
NaOs
HaO
HNOa
HN03
-12-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512. 450m
The species listed in Table 3-3 are present in widely
varying concentrations. They interact with each other in many
possible reactions. Which of these species and reactions should
be used to describe the gas phase for the sorption process is
a question of basic importance. Some species and reactions c~n
be eliminated from consideration in the theoretical description.
For instance, some species are present in relative concentrations
that cause them to be unimportant for mass balance. These
species can be disregarded in the problem formulation, but only
if they are not involved in a process or step that is rate limit-
ing. On the other hand, some species that are present only in
small concentrations and are not involved in rate limiting steps
are still included in the problem formulation for convenience.
An example is Ng05, which is a convenient species to consider in
nitrate thermal decomposition calculations. Stern (ST-026)
reported that the decompositions of NO and NOa to form the ele-
ments need not be considered since the reactions are very slow
at temperatures below lOOooK. Reactions involving N~O are also
slow. Based on the reasons just discussed, the reactions in
Equations 3-1 through 3-6 were used to describe interactions in
the gas phase.
NO + NO:; ~ N203 3-1
2NOg ~ Na0,. 3-2
Na03+ HaO ~ 2 HNOa 3-3
Na 04, + Ha ° ~ HNOa+ HN03 3-4
NOa+ Na 04 t NO + Na 05 3-5
NO + ~Oa ~ NOa 3-6
-13-
-------
Radian Corporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766 . TELEPHONE 512.454-9535
Reactions 3-1, 3-2 and 3-3 have been reported to be
rapid enough to be considered at equilibrium for the purposes of
this study (CH-032, p. 55; WA-014; WA-015, pp. 11-12; WA-016).
lnere is some disagreement concerning whether or not reaction 3-4
is a homogeneous reaction and whether or not it can be considered
to be at equilibrium. Reaction 3-6 is a slower reaction; it is
the rate limiting step for nitric acid production (CH-032). It
cannot be considered to be at equilibrium for purposes of des-
cribing mass transfer in a scrubber where the residence time
~ould be on the order of seconds. It could be considered to
have reached equilibrium in a static nitrate thermal decomposition
experiment where the product gases are not removed and the duration
of the decomposition is on the order of several minutes to an hour.
Reaction 3-5 is included for convenience in the screening process.
Thl~ number of possible species in existence in the
aqueous phase is even greater than the number in the gas phase.
In addition, the set of possible reactions is much more compli-
c-~ed and the equilibria are not completely described in the
literature. Waldorf (WA-015) discussed the problem at length in
his dissertation, "Reactions and Equilibria in the Nitrogen Oxides-
Water System", which included a comprehensive literature review.
lne species listed in Table 3-4 can exist in the aqueous phase.
lnese were identified from a review of the chemistry of the system
NO-NO:a-H:aO. The subscript (t) has been eliminated for convenience
so that NO, for example, is understood to be NO(t)'
TABLE 3-4
AQUEOUS PHASE SPECIES IN THE NO~-H~O SYSTEM
NO
NO:a
N:a03
N:a04
HN 0:2
H20
HN0;3
H+
OH-
NO;
NO;
NO+
NO~
HONOIrt
HONONO+
HN03'H:aO
N02NO+
-14-
-------
l
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN. TEXAS 7B757 . TELEPHONE 512.454.'1535
It has been proposed by numerous authors including Moll
(MO-008) and Gray and Yoffe (GR-004) that the dimers Na03 and
N~04 either ionize or form ion pairs according to the following
scheme.
Na°4,
~
NO+NO-
a a
i!
NO+NO;
(3-7)
Na03
~
NO+NO-
a
(3-8)
The species HONOH+ (BU-009), HONONO+ (WA-006) and
NOaNO+ (MI-004) have had tentative spectroscopic identification.
Some of the species listed in Table 3-4 exist only under extreme
conditions of pH which probably will not be encountered in the
processes studied. These can therefore be eliminated from our
formulation. They may become of importance later if processes
other than aqueous alkaline scrubbing are investigated. Addi-
tional species can be eliminated from the aqueous phase formulation
with no difficulty if their relative concentrations cause them
to be unimportant in the mass balance and they are not involved
in a rate limiting step.
The aqueous phase reactions which involve ionic
species are considered to be quite rapid. If an aqueous phase
reaction is the rate controlling step for sorption of NO and
NOa, it will be necessary to know the mechanism of the reaction,
i.e., the individual, serial, usually bimolecular steps which
occur and which result in the overall stoichiometry. The
mechanism according to which nitrogen oxides react with water
to form nitrous and nitric acids is not known. Several possi-
bilities have been discussed in the literature. The overall
stoichiometries are given in (3-9) and (3-10).
-15-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454.9535
NO + NOiOl+ HaO
~
2HNOa
~
2H+ + 2NO-
a
(3-9)
2NOa+ HiOIO
~
HNOa+ HNOs
~
2H+ + NO; + NO;
(3-10)
Some proposed mechanisms are given in (3-11) through (3-13).
Na04
i!
NO~O;
HaO
'-
HONOH+ + NO;
....
2W + NO; + NO;
(3-11)
or
NO+ + NO;
Na03
i!
NO~O;
H:;J0
~
HONOH+ + NO;
....
2H+ + 2NO-
a
(3-12)
or
NO+ + NO;
NOa+ HaO ~ x (3-13a)
HaO NO; + NO; + 2W
x + N0:;J ~ (3-13b)
x + NO HaO 2NO- + 2H+ (3-13c)
--+ :;J
The mechanism for the nitrogen oxides-water
in the aqueous phase will not have to be known unless
is rate controlling for the sorption process.
reaction
the reaction
A summary of the possible species and reactions in the
gas and aqueous phase systems containing nitrogen oxides and
water is shown in Figure 3-1. The reactions or mass transfer
steps which have been studied experimentally and for which
equilibrium constants have been reported in the literature are
indicated in the figure by those numbers which are circled.
-16-
-------
~
I N ° (L'
....... I
'IS, +
I NO
, : d ~2 C L)
'9 ~
I .......
.......: N203(L).......
I +
I
H20(g) I H20(1,)
01 ~ + !31 ~ @ '-
2HN02(g), I 2HN02(1,)"
,IS
I '2NO
, I 2 ( 1,)
-A.~ ~
....... I N204(J,)'
: +
: H20 ( L)
2~' ~41 l
,: H N 03 ( J,)
@, +
,I H N 0 2 (.e) "
I
'-
Gas
N ° (g)
+
N02(9)
01 ~
N203(9)
+
2N02(g)
01l
N204(9)
+
H20(g)
31 ~
H N 03(9)
+
HN02(g)
L iq u i d
I ;0:) N O~+N 0;+2 H+
2 J6&
J6 c
H20 r.: J NO - +
J6a>~02H2~ H20>2N02+2H
I
7
, NO+NO-+H °
812~ 2
HONOH++NO-
2
'Sb 1 ~
2 H++ 2 N 0-
2
6
....... N O+N 0- + H °
3 2
'4a 1 ~
H 0 N 0 H+ + N 0-
3
0"
H+ + N 0-
3
'4 b
,
@"
H+ + N 0-
2
o Reactions for which equilibrium or dissociation constants have been measured
FIGURE 3-1
Specics and Rcactions in the Gas amI Aqucous Phase System
NOx-H20.
-17-
-------
Radian Corporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766. TELEPHONE 512 - 454-9515
3.2.2
Rate Controlling Step
Figure 3-1 shows the various paths by which the gaseous
nitrogen species can be transformed into aqueous nitrogen contain-
ing ions or neutral aqueous species. The process involves chemical
reaction and molecular diffusion in the bulk gas up to the gas-
liquid interface, gas film diffusion, liquid film diffusion, and
then diffusion and chemical reaction in the a.queous phase. All
of these steps are seen as resistances to mass transfer and one
step will probably be the rate controlling step. The problem
discussed in this section is how to describe the rate of sorption.
One of the paths shown in Figure 3-1 will be the fastest path.
The slowest step in this fastest path will be the rate controlling
step. Rate equations must be written in order to size equipment
for a sorption process. The rate equation must describe the rate
of removal of gaseous nitrogen oxides in the scrubber in terms of
the change in concentration of one of the species in Figure 3-1.
It is best to use either the concentration of the species involved
in the rate controlling step or the concentration of the species
with which it is in equilibrium.
3.2.2.1
Mechanism Suggested in the Literature
Numerous studies have been conducted to investigate
the rate of nitrogen oxides sorption by aqueous solutions in
y-~ying types of equipment. Other studies have been carried out
to determine simply the rate of reaction between nitrogen oxides
and water. The experimental results of the various investigators
and the mechanisms and rate determining steps they proposed were
studied in detail and are reviewed in Technical Note 200-007-02
entitled "Review of the Literature on Experimental Studies of the
-18-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512.454'1535
Aqueous Absorption of Nitrogen Oxides". The note is included
in Volume II of this report. The results were obtained for
widely varying experimental conditions using acidic neutral and
alkaline sorbents for gas mixtures containing NO alone, NO~
alone, and mixtures of the two. With the exception of one or
two studies, the lowest concentration of nitrogen oxides studied
was 0.5 mole%, which is a whole order of magnitude greater than
the concentration range for flue gases. Few of the investigators
made material balances or even measured anything except NOa(g)'
Many of the results appear to be contradictory. Since the
experimental conditions were so diverse, it was difficult in the
beginning to make generalizations about the results or to apply
them to the problem at hand. To facilitate comparisons, each of
the results was summarized consistently in a table with separate
columns describing gas composition, absorbing medium, apparatus
and operating conditions, quantities measured, and results and
discussion.
The data concerning sorption of pure nitric oxide
or pure nitrogen dioxide or those gases diluted only with nitrogen
will not be discussed in detail here. They are summarized in
the technical note. Nitric oxide sorption with no NOa present
has been investigated to a limited extent by workers in the field
of coke oven gas purification. Nitric o~ide impurities of
even a few ppm must be removed from coke oven gas to prevent
their reaction with dienes to form explosive resins. Nitrogen
dioxide sorption with no NO present has been studied widely
because of nitric acid manufacture where re1at.ively high concen-
trations of NOa are absorbed by dilute nitric acid solutions.
Much early effort was devoted to showing that the rate of N02
sorption is proportional to the concentration of Na04 and that
the rate of the Na04 reaction with water might be rate controlling
for NOa sorption. Not until fairly recently have investigators
-19-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
considered how the presence of nitric oxide might affect the
rate of NO~ sorption. These investigations were still carried
out in the interes~ of nitric acid manufacture, and virtually
no one has investigated nitrogen oxides sorption in the context
of flue gas purification.
Investigations concerning the rate of sorption and
the rate limiting step for sorption from gas mixtures containing
both NO and NO~ were studied in detail. None of the investiga-
tors actually answered the questions of what is the rate limiting
step and what species should be used in the rate equations for
the particular case of interest, i.e., flue gas cleaning. It is
tempting, nevertheless, to select the most reasonable mechanism
and the resulting rate equations, and to apply them to the problem.
While it would save a lot of hard thinking, the results would not
be useful. The following paragraphs give a general description
of what was found in the literature. Then the research that has
a bearing on the problem of flue gas NOx removal is discussed in
more detail. The contributions are subtle since only points of
view and ideas from the literature were used rather than detailed
mechanisms or actual rate equations.
Twenty two of the publications reviewed concerning
the rate of nitrogen oxides sorption dealt with absorption
from gases initially containing both nitric oxide and nitrogen
dioxide. 'The initial total concentrations varied from about
0.5 to 100%. Absorbents studied were water and aqueous solutions
of HN03, NaOH, CaCla, Nag C03, Ca(NOsk, Ca(OH) a' CaC03, Na:; C03
+ NaOH + NaHC03, and NagC03+ NaNOg+ NaN03. A variety of equip-
ment was used including wetted wall columns, packed columns,
venturi scrubbers, spray columns, laminar jet apparatus, sieve
plate absorbers, and a horizontal mechanical absorber with a
rapidly revolving axial shaft.
-20-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. 80X 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
There was some agreement (VA-009, VA-006, MI-OOS) that
the absorption rate from NO-NOa mixtures in one particular con-
centration range is a maximum at a 1:1 mole ratio of NOa to NO or
at SO% oxidation. These results are well expressed graphically
by Hofmeister and Kohlhaas (HO-009). Some agreed that the
absorption rate is proportional to the concentration of Na03 or
to the product of NO and NOa concentrations (KR-007, ZH-001,
EL-004, KO-026). Perelman (PE-010) and Zhavoronkov (ZH-002) and
several other workers have stated that the rate of Na03 absorp-
tion is greater than the rate of Na04 absorption. Atroshchenko
(AT-OOS) found that the relative rates depend on the experimental
apparatus. Ganz (GA-009, GA-010) found the rates to be about
equal in his apparatus. Koval (KO-024, KO-026) reported the rate
of reaction of Na03 with water to exceed the rate of reaction of
Na04 with water. Kramers, et a1. (KR-006) and Hofmeister and
Kohlhaas (HO-009) measured the rates of the individual reactions
using a laminar jet apparatus and the penetration theory.
There was wide agreement (GA-014, PE-010, KR-007,
MU-004, GA-008, KR-008) that the concentration of nitrate and
nitrite ions in the sorbent has an effect on the rate of sorp-
tion. All reported that an increase in nitrate and nitrite
concentration reduces the, rate of sorption. Perelman (PE-010)
reported the rate to increase for sorbents in the order CaC03,
NaaC03, Ca(OH)a. Atroshchenko (AT-OOS) and Caudle and Denbigh
(CA-014) reported the sorption rate was greater in water than
in HN03 or CaCla solutions, respectively. There was agreement
among several workers (KR-008, VA-006, PO-01S) that, above a
lower limiting value, the concentration of NaaC03 in the sorbent
had no effect on the sorption rate.
-21-
-------
Radian Corporation
8500 5HOAL CREEK BLVD. . P. O. BOX 9948 . AU5T1N, TEXAS 78757 . TELEPHONE 512 - 454.9535
The work of several authors merits more detailed
discussion. Of particular interest for the problem at hand
is the work of Andrew and Hanson (AN-OOl) who stated that the
mechanism and thus the rate of absorption is dependent on the
relative and total concentrations of NO and NO:;!. Most important,
they showed that to explain the rate of NOx sorption over a wide
range of gas concentrations, it is necessary to use more than one
mechanism. They described four possible absorption mechanisms
and developed theoretical rate equations for a laboratory sieve
plate for each mechanism. The rate equations were expressed in
terms of the plate efficiency, ~, defined as:
Chemical NO~ Absorbed
Chemical NOa Entering
Chemical N02 was defined by Andrew and Hanson as N02 + 2N:a 04 +
N203 + ~HN()2' A discussion of the mass balance equations for
NO and N02 is given on pages 34-35. They compared plate effi-
ciencies predicted from the theoretical rate equations with
actual measured plate efficiencies. Figure 3-2 shows the plate
efficiencies predicted for each mechanism as a function of chem-
ical NOG concentration for a fixed NO:N02 ratio of 1.0. Some
physical constants had to be estimated to construct the figure.
lne mechanisms numbered one through four in Figure 3-2 are
described in the legend below.
-22-
-------
Radian Corporation
FIGURE 3-2 -
Tl2
8500 SHOAL CREEK 8LVD. . P. O. 80X 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 . 454.'i5~5
100%
100%
T . 10 SEC.
G. 10 CM/SEe.
~
°
25°C
~
Q I~o
'"
G
ii:
I:J
10'-
'"
~
, 73 Ot 0 I , 10' 'Yc
(NO~. CHEMICAL NOJ CONCENTRATION (GM.MOl/o.-r.xI06)
9
PREDICTED PLATE EFFICIENCIES BASED ON THE
THEORETICAL RATE EQUATIONS OF ANDREW AND HANSON
[Taken from: Chern. Eng. Sci., 14, 105 -13 (1961)~
Tll
Percent of total plate efficiency due to
Mechanism 1: diffusion across gas and liquid
films as NOg and Na04; subsequent hydrolysis
to HNO~ and HN03; decomposition of HN02 to Na03)
Na03(g) or HNOa(g) given off. N02-N~04 diffusion
rate limiting.
Percent of total plate efficiency due to
Mechanism 2: diffusion as HN02(g)-Na03(g)
equilibrium mixture; HN02 decomposition in
aqueous phase; nitric oxide given off. Aqueous
HN02 decomposition and NO(g) diffusion rate
limiting.
-23-
-------
Radian Corporation
by Andrew
different
dioxide.
while the
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
Tls
Percent of total plate efficiency due to
Mechanism 3: gas phase formation of HNOg
and HNOs, HN03(g) dissolution in water
vapor forming mist, HNOs(g) diffusion into
aqueous phase. Rate limited by gas phase
HNOs formation.
Tl,t
Percent of total plate efficiency due to
Mechanism 4: liquid film limited diffusion
of NOa, dimerization in solution and hydrolysis
of Ng04. HNOg(g) given off.
Figure 3-3 shows the total plate efficiencies measured
and Hanson as a function of chemical NOg for two
relative concentrations of nitric oxide and nitrogen
The experimental data are indicated by O's and Xis,
predicted efficiencies are indicated by the solid lines.
100 Ofo
100'0
25<1c
10'0
IOfo
001
1°10
~ I ~
[N02J: O£MICAL N02CONCENTRATION (GMMOl/Ct.fx 106)
COMPARISON OF ANDREW AND HANSON'S MEASURED
AND PREDICTED TOTAL PLATE EFFICIENCIES
[Taken from: Chern. Eng. Sci., 14, 105-13 (1961)J
FIGURE 3-3 -
-24-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . p, O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.4505)5
Using Figure 3-2 one could determine the mechanism
which makes the greatest contribution to plate efficiency
(absorption rate) for different chemical NOa concentrations at
a fixed ratio of NO/NO::!. Since Andrew and Hanson were primarily
interested in nitric acid manufacture, they only showed data
for a fixed NO/N02 ratio and not for a variety of compositions.
Their main concern was how a change in chemical NOa concentra-
tion was reflected in a change in rate determining step and
mechanism. In addition, they used mass transfer coefficients
determined for a laboratory sieve plate absorber and equilibrium
constants which might or might not agree with the ones used by
Radian. As a result, the rate equations and mechanisms are not
directly applicable to the problem of flue gas sorption in a
packed tower. A valuable result of their work, however, is the
idea that the rate determining step and the sorption mechanism
are peculiar to the gas composition and concentration. However
reasonable and theoretically sound a mechanism appears, if it
describes gas mixtures whose compositions and concentrations
differ from those in flue gas, the mechanism may not be appJicab~
to flue gas cleaning.
The work of Koval (KO-026) concerns the effect of
adding nitric oxide on the absorption of NOa+ Na04 by water.
Again the investigations were prompted by a desire to understand
the mechanism of nitric acid manufacture. The work
seems to have been performed with great care and the proposed
mechanism did explain some results found by previous workers.
Although Koval did not measure the rates predicted by the rate
equations resulting from his mechanism,he did make material
balances and demonstrate that the stoichiometry resulting from
his mechanism agreed with the actual stoichiometry.
-25-
-------
Radian Cor'poration
8500 SHOAL CREEK BLVD, . p, 0, BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
Koval pointed out indirectly one of the main problems
in evaluating NO. sorption systems and in NOx sorption process
design. The problem is that nitrogen exists in so many different
oxidation states and ionic or molecular forms that it is very
complicated to keep track ,of them all. Analytical chemistry
techniques and calculation of material balances become a problem.
Koval realized these problems because he was one of the few
investigators to attempt to make a material balance. Even when
sorption of pure N02-N204-nitrogen mixtures is studied, the con-
centration of HNO;-N~03-NO must be taken into account. Some
investigators ignored their affect or treated their concentratfuns
as negligible. In order to account for the different forms of
nitrogen, analytical techniques had to be developed. Koval
observed that many investigators mistakenly titrated acidic
liquid absorption products directly with sodium hydroxide. He
showed that decomposition of HN02 between the time of sampling
and analysis could result in values for total acid from four to
twenty percent too low. In addition the titration of nitrous
acid with sodium hydroxide does not have a sharp end point since
HN02 is always decomposing and the pH is not constant with time.
Waldorf (WA-015) studied instrumental analytical techniques
which provide one means of solving the analytical problems. He
also pointed out the lack of reliable activity data and the fact
that most investigators did not analyze the aqueous phase for
anything other than nitrate ("nitric acid").
/
The mechanism and rate determining step for NOx
sorption also depend on the apparatus and the amount of gas-
liquid contact area that can be provided. Koval (KO-026)
demonstrated this by conducting experiments in a variety of
equipment types with other factors held constant. Ganz and
coworkers (GA-010, GA-013, GA-014) describe a high speed
horizontal mechanical absorber with a revolving axial shaft
which is said to provide 'an increase in sorption capability
-26-
-------
I .
Radian Corporation
8500 SHOAL CREEK 8LVD. ' P. O. BOX 9948 ' AUSTIN, TEXAS 78757 ' TELEPHONE 512 - 454-9535
over conventional gas-liquid contactors by increasing gas and
liquid turbulence. They reported precipitation of basic nitrate
salts beyond some critical concentrations of CaO, Ca(NO~)~ and
Ca(N03)2' Gorfunke1 (GO-007) also reported the possibility of
formation of CaO'Ca(N03)2 '2H20. The mechanism proposed by Ganz
for the mechanical absorber includes the suggestion that "liquid
phase" oxidation according to Equation 3-14 takes place.
NO; (L) +N02 (g)
~
NO; (.~)+ NO(g)
(3 -14 )
The reaction was suggested to account for the greater than
theoretical amounts of nitrate and less than theoretical amounts
of nitrite produced in the mechanical absorber.
In summary, none of the mechanisms and rate controlling
steps proposed in the literature for NOx aqueous sorption were
directly applicable to sorption of NO and N02 from power plant
flue gases. Most of the research results are inapplicable
because they deal with gas compositions and concentrations out-
side the range of those in flue gas. In particu1a~ a great
amount of work has been published concerning the problem of N02
sorption in nitric acid production. The concentration of nitric
oxide in those systems does not exceed that of N02. However,
several important ideas concerning the mechanism and rate determin-
ing step for aqueous sorption were gained from the literature
published on NOx sorption, First, the mechanism and rate deter-
mining step are dependent on the gas composition (ratio of
NO/N02), the total nitrogen oxides concentration, and the type
of gas-liquid contactor employed. The mechanism and rate equa-
tions proposed to date are not applicable to flue gas cleaning
processes. Second, since the nitrogen oxides-water system is
-27-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 7B757 . TELEPHONE 512 - 454-9535
complicated in both the aqueous and gaseous phases by the many
stable oxidation states of nitrogen, material balances and
analytical techniques are very important in establishing a
mechanism.
3.2.2.2
Proposed Method for Determining Rate Controlling
Step for Flue Gas Treatment Processes
Since no applicable work has been published concerning
the rate controlling step, it was decided one should be proposed
for flue gas treatment on the basis of gas composition and con-
centration and mass transfer characteristics for packed towers.
Then the proposed mechanism had to be tested by conducting
absorption experiments and measuring enough gas and liquid phase
concentrations to be able to calculate material balances. One
important tool was needed to describe mass transfer from a gas
phase containing numerous nitrogen species at equilibrium to a
liquid phase containing numerous nitrogen species at equilibrium.
This was the ability to calculate the concentrations of each of
the many species present at equilibrium under given conditions of
temperature, total pressure, and total chemical NO and NO~ con-
centrations. This "equilibrium model" was necessary since not
every species could be measured.
In summary, it was found that a mechanism for NOx
sorption from flue gases had to be proposed on the basis of
chemical composition and total concentration. An "equilibrium
model" was therefore needed to (1) give a detailed description
of the equilibrium composition at a given temperature, pressure
and total concentration, and (2) allow a description of the
changing composition as nitrogen species were transferred from
-28-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
the gas to the aqueous phase. The proposed mechanism then had
to be tested in experiments in which sufficient data were
collected to calculate material balances with the help of the
equilibrium model.
3.3
Summary of Problem Definition
The problem of removing NO and N02 from flue gases by
aqueous sorption has not heretofore been completely defined.
The scope of the problem as it was investigated by Radian was
narrowed by the following factors:
50 ppm N02 and 450 ppm NO had to be removed
from flue gas also containing about 13% CO2,
While the primary goal was to obtain a
theoretical description of NO and N02 sorption
in a wet scrubber, the theoretical description
was developed for a complete process, i.e., not
just the scrubbing step. Two processes, a
"dry" and an aqueous process, were investigated.
The theoretical description had to be used to
evaluate the performance of potential dry metal
oxide sorbents in the process of interest.
To properly define the problem whose scope was
described above, the following questions had to be answered:
What nitrogen-containing oxide species are present
in the flue gas; what aqueous species are formed
by sorption; and what reactions take p~ace in the
system?
-29-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
What is the mechanism for the sorption
process and which of the steps controls
the rate of sorption from gas mixtures
having the composition and concentration
of flue gases?
The answer to the first question was found from the
chemical literature. The second question had not been previously
investigated. Research had been conducted on related problems
and some of the ideas could be applied to aqueous sorption from
flue gases. It was found that to fully define the problem two
things were needed:
an equilibrium model which would allow the
concentration of each of the many species
present at equilibrium to be calculated,
experimental studies to verify a proposed
sorption mechanism. The equilibrium model
would also be required to perform some
aspects of the mass balance and mass transfer
calculations.
-30-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHGfJi: 512. <5<-15;;
4.0
THE COMPOSITION-CONCENTRATION MODEL
From the process of defining the problem for develop-
ment of a theoretical description of aqueous scrubbing of NOx
from flue gases, the complexity of the nitrogen oxides-water
system became apparent. It was then obvious that a basic
description of the system required that the concentrations of
each of the gaseous and aqueous species present be known. It
was recognized that some species might be present in concentra-
tions too small to be of significance for a mass balance.
At the same time, if the species present in relatively small
concentrations were involved in the rate controlling step for
sorption) then it would be necessary to know their concentration
and how it changes as sorption occurs. A knowledge of the rela-
tive concentrations of the gas phase species was also needed in
order to predict which compound of nitrogen and oxygen would be
involved in the rate-limiting step for mass transfer.
While the gas phase NOx-HaO system is quite compli-
cated, the aqueous phase system presents an even greater
problem since ionization creates a greater number of possible
reactants and reaction pathways. In addition, concentrations
of the ionic species suggested to be involved in aqueous phase
rate limiting steps are more difficult to measure. Waldorf
(WA-01S) has studied the application of ultraviolet spectroscopy
to the identification of many of the ionic and molecular aqueous
phase species, but routine analytical methods have not yet been
developed. Evaluation of the relative rates of different
reaction pathways to determine which might be rate limiting
presents a formidable task. Finally, the aqueous solutions of
interest for this problem are of relatively high ionic strength
and are far from ideal solutions. Calculations must therefore
be based on activities rather than measured molalities. Radian
-31-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512 - 454-9535
has previously developed an aqueous equilibrium model which
describes ionic interactions in scrubbing liquors for S02
removal processeso The model calculates activity coefficients
from ionic strength and takes into account the equilibria of
ion-ion interactions; therefore, it is applicable to the non-
ideal solutions involved in NOx sorptiono The physical and
thermochemical data for some molecular and ionic nitrogen-
oxygen species were added to the existing aqueous model. [The
data are described in Technical Note 200-007-11 (Volume II) and
in Section 8.0 of this report.] However, since the kinetics of
liquid phase reactions for NOx sorption have not yet been fully
defined, it is not known whether the aqueous equilibrium model
will be sufficient for describing aq~eous phase interactions
in NOx sorptionprocesseso
The same problems do not exist for the gas mixtureo
The -species present oth~r than NO, N02, H20 and their reaction
products act as inerts so that the system is effectively quite
dilute, At the temperatures and low NO + N02 pressures which
prevail the gases almost obey the ideal gas law (PV = RT).
Fugacity coefficients are almost unity. Half times of some of
the gas phase reactions were reported in the literature to be on
the order of centiseconds (WA-016) and microseconds (CA-069).
Other reactions were described as instantaneous (WA-016). The
reactions are considered to have reached equilibrium within
scrubber gas residence times. A description of the relative
concentrations of the gas phase species at equilibrium was
selected as a useful and logical starting point. Accordingly,
a gas phase equilibrium model was developedo It is described
in this section and in detail in Technical Note 200-007-03a,
included in Volume II of the reporto A listing of the computer
subroutines written to calculate the gas phase equilibrium con-
centrations is also included in Volume II in Technical Note
200-007-16.
-32-
-------
Radian Corporation, 8500 SHOAL CREEK 8LVD, . p, 0, BOX 9948 . AUSTIN, TEXAS 78757 '. TELEPHONE 512.454-9535
4.1
Reactions Considered in the Gas Phase Equilibrium Model
In Section 3.2.1, Table 3-3, the gas phase species
in the NOX-HgO system were listed, and some equations were given
to describe their interactions. The selection of the reactio~s
used to describe equilibrium in the gas phase is documented in
Technical Note 200-007-03a (Volume II of this report). The
selected reactions are given in equations (4-1) through (4-5).
NO + NOa ~ Na03 (4-1)
2NOa ~ Na04 (4-2)
Na03 + HaO ~ 2 HNOa (4-3)
Na 04 + Ha 0 ? HNOa + HN03 (4-4)
NO + NaOs ? NOa + Na 04 (4-5)
There is little question that the reactions in (4-1), (4-2) and
(4-3) are equilibrium reactions (CH-032, p.55; WA-014; WA-015,
pp. 11-12; WA-016). There is great controversy concerning
whether nitric acid is formed in a homogeneous gas phase reac-
tion such as (4-4). The facts presented to date are discussed
in detail in the technical note. Reaction (4-5) was included
so that equilibrium partial pressures of NaOs could be cal-
culated. The reaction for oxidation of NO to NOa was not
included in the equilibrium formulation since it would not
reach equilibrium within gas residence times in the scrubber.
The half time for the oxidation reaction is about 21 minutes
at scrubber conditions of 43°C, 8 mole % oxygen and .05 mole
% NO. Under the same conditions, about 1% would be oxidized
after 5 seconds.
4.2
Problem Formulation and Method of Solution
The purpose of the gas phase equilibrium model was to
calculate the partial pressure or mole fraction of each of the
-33-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, T~S 78766. TELEPHONE 512.454-9535
gaseous species present at equilibrium under given conditions
of total pressure and temperature 0 A computer program was
written to perform the calculations.
There are eight components for which the mole fraction
was to be calculated 0 These eight unknowns are calculated from
a system of eight nonlinear equations. The eight nonlinear equa-
tions consist of five equations for the equilibrium constants of
the reactions in (4-1) through (4-5) plus three mass balance equa-
tions. The mass halance equations are defined in (4-6), (4-7) and
(4-8), where n is the number of moles of each component, NT is the
total number of moles, and Y is the mole fraction of each component.
Chemical NO
CNO
= nNO + nN 0
2 S
+ ~nHNO
a
1:n
- 2 HNO
3
- n
NaOs
= NT (YNO + YNaOs + ~YHNOa - ~YHN03 - YNaOs)
(4-6)
Chemical NOa
CNO:;! = nNOa + 2nNa04 + nN:
-------
Radian Corrporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766 .
TELEPHONE 512 - 454-9535
Chemical NO, NOz, and ~O are inputs to the program.
They are quantities which can be measured. The mass balance
equations describe the distribution of the total number of moles
among the possible molecular species. Chemical NO is the term
used to describe the number of moles of nitrogen in the +2 oxi-
dation state as shown in Equation 4-6. Chemical NOz describes
the number of moles of nitrogen in the +4 oxidation state,
Equation 4-7. The mass balance equations were derived by con-
sidering the stoichiometry of Equations 4-1 through 4-5. For
example, since one mole of NO is used in the formation of a mole
of NZ03, the term nNZ03 is included in the mass balance equation
for +2 nitrogen. One mole of NZ03 contains one mole of NO and
forms two moles of HN02 (Equation 4-3). However, some HNOz is
also formed from +4 nitrogen according to Equation 4-4. The
total number of moles of HN02 formed is shown as follows:
nHN02 (TOTAL)
=
nHNOz (4-3)
+
nHN02 (4-4)
(4-9)
We would like to include the term ~nHNOz (4-3) in the +2 nitrogen
balance but it is impossible to distinguish between the HN02
formed according to Equation 4-3 and that formed as in Equation
4-4. Only nHN02 (TOTAL) is known. However, the amount of HN02
formed according to Equation 4-4 is equal to the amount of HN03
formed:
nHN02 (4-4)
=
nHN03 (4-4)
=
nHN03 (TOTAL)
(4 -10)
Since this is the only path by which HN03 is formed in the
Radian formulation, the amount of HN02 formed according to
Equation 4-3 can be indicated by substituting nHN03 (4-4) for
nHN02 (4-4) :
nHNOz (TOTAL)
=
nHN02 (4-3) +
nHN03 (4-4)
(4-11)
-35-
-------
Radian Corporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766. TELEPHONE 512 - 454.9535
and
nHN02 (4-3)
=
nHN02 (TOTAL) -
nHNOs (4-4)
=
nHN02 - nHNOs
(4 -12)
Therefore the term ~nHN02 (4-3) can be replaced by ~nHN02 - ~nHN03
.in the +2 nitrogen balance.
The five equilibrium constant equations are the non-
linear ones. They are expressed in terms of total pressure,
mole fractions of products and reactants and fugacity coefficients,
The total pressure is a program input. The equilibrium constants
themselves are calculated as a function of temperature from
standard state thermodynamic properties and heat capacities 0
The equilibrium constant temperature dependence data are stored
by the program and the temperature is a program input. As men-
tioned previously, the fugacity coefficients are very nearly
unity since the gas mixture is effectively an ideal gas.
The system of eight nonlinear equations is solved in
the logarithmic domain using an iterative procedure originally
developed for calculating aqueous solution equilibria (LO-007).
4,3
Comparison of Calculated and Measured Results
A concentrated effort was made to compare measured
and calculated values of the mole fractions or partial pressures
present at equilibrium in the NOx' -Ha 0 vapor phase system. The
effort is described in detail in Technical Note 200-007-03a.
An experimental program to measure gas phase equilibrium concen-
trations was beyond the scope of Contract EHSD 71-50 Therefore,
-36-
-------
Radian Corporation 8500 SHOAL CREEK ILVD. . P. O. lOX 9948 . AUSTIN, TEXAS 78766 .
TELEPHONE 512 - 454-9535
it was neces sary to re ly on published measurement s of equilibrilJm
concentrations. The only measured equilibrium concentrations
reported in the literature were for the component NOa. A
thorough search revealed that analytical methods do not exist
for some of the species and the few existing methods for the
other species have not been applied to equilibrium mixtures.
The comparison of some measured and calculated N02 partial
pressures is shown in Table 4-1 and a comparison of measured
and calculated total number of moles, N., is shown in Table 4-2.
TABLE 4-1
COMPARISON OF CALCULATED AND
MEASURED PARTIAL PRESSURES OF NOa
PNOa (atm)
Temperature
0C Calculated Measured Reference
29095 .0227 .0204 AS-004
25 0 01+ .073 0067 VO-007
,053 ,050
.032 .031
.018 .017
80.7 .0263 .0233 AS-004
-37-
-------
Radian Corporation
8S00 SHOAL CREEK 8LVD. . P. O. 80X 9948 . AUSTIN, TEXAS 787S7 . TELEPHONE S12. 454-9535
TABLE 4- 2
COMP ARISON OF CALCULATED AND
MEASURED TOTAL NUMBER OF MOLES
Temperature
°c
NT Calculated
by Equilibrium
Program
NT Calculated
from Measured Total
Pressure (BE-023)
25.04
25.04
25.04
45.12
45.04
.0358
.0345
.0395
.0144
.0452
.0353
.0341
.0391
.0152
. 0463
4.4
Application of the Gas Phase Equilibrium Model to
Predict the Rate Controlling Step
Table 4 - 3 shows the results obtained when the gas
phase equilibrium model was used to calculate the mole fractions
of NO, NOa, NaOs, Na04, HNOa and HN03 present at flue gas
scrubber conditions of 60°C, 1 atmosphere total pressure, and
8 mole % H20. The calculations were made to demonstrate the
effect of total concentration of input chemical NO + NOa and
the effect of relative concentrations of input chemical NO and
NO:! on the equilibrium concentrations. Additional calculations
were performed to determine the effect of temperature and total
pressure on the equilibrium distribution, and they were found
to have a negligible effect on the relative distributions.
The calculated equilibrium concentrations can be used
to explain which compounds should be involved in the rate limit-
ing step for nitrogen oxides absorption at different gas composi-
tions and concentrations. Investigators studying absorption of
-38-
-------
TABLJ 4-3
EQUILIBRIUM COMPOSITIONS IN THE GAS PHASE SYSTEM
NOx-H20 AT 60°C, 1 ATM TOTAL PRESSURE, AND 8 MOLE % H20
Total Input Relative Eauilibrium Concentrations (mole -Fraction)
Concentrations Concentrations .
NO+N02 (mole%) Input NO and NO? NO NO? N?03 N204 HN02 HN03
.05 Equimolar 2.43xlO-4 2.37xlO-4 5.33xlO-g 3. 99xlO-s 1. 70xlO-s 2. 7lxlO-s
10% NO;o, 90% NO 4.45xlO-4 4.48xlO-s 1.84xlO-g 1.42xlO-g 9. 99xlo-e 1.64xlO-7
90% NO;o, 10% NO 5.l8xlO-s 4.23xlO-4 2.03xlO-g 1. 27xlO-7 1.05xlO-s 1.40xlO-s
.5 Equimolar 2.43xlO-3 2.37xlO-3 5. 3lxlO-7 3.97x10-e 1 . 7 Ox 10- 4 2.7x10-s
10% NO;o, 90% NO 4.45xlO-3 4.47x10-4 1. 84x10-7 1.42x10-7 9.98xlO-s 1.64x10-s
90% NO;o, 10% NO 5.l7x10-4 4.2lxlO-3 2.01x10-7 1. 26x10-s 1.04x10-4 1.39x10-4
1.0 Equimolar 4.85xlO-3 4.72xlO - 3 2.l2xlo-e 1. 58x10-s 3. 38xlO-4 5.36xlO-s
10% NO;o, 90% NO 8.9x10-s 8.94x10-4 7.35xlO-7 5.65x10-7 1.99xlO-4 3.27xlO-s
I 90% N02, 10% NO 1.03xlO-3 B.38x10-3 8 . Ox 10 - 7 4.98x10""s 2. 08x10-4 2. 75x10-4
W
\.0 -
I 2.0 Equimolar 9 . 71x 10- s 9. 37xlO s 8.4lxlo-e 6. 22xlO-s 6.73xlO-4 1. 06x10-4
10% N02, 90% NO 1. 78x10-;o 1. 78xlO-s 2. 93xlo-e 2.25xlo-e 3.98xlO-4 6.5lxlo-e
90% NO;o, 10% NO 2.06xlO s 1. 66xlO-;o 3.l6xlO-s 1.95xlO-4 4.l2xlO-4 5.42xlO-4
5.0 Equimo1ar 2.42xlO-;o 2. 30xlO -2 5.l6x10-S 3.75x10-4 1.66x10-3 2.57xlO-4
10% NO;o, 90% NO 4 .45x10 -2 4.43xlO-S 1.82x10-s 1.39x10-s 9.91x10-4 1.6lxlO-s
90% N02, 10% NO 5.l2xlO-3 4.02xlO-2 1.9lxlO-s 1.15x10-s 1 . 0 Ix 10- 3 1.29xlO-s
10.0 Equimolar 4.84xlO-~ 4.46xI0-2 2 . Ox 10 - 4 1.41x10-3 3.26x10-s 4 . 8 9x 10 - 4
10% N02, 90% NO 8. 90x10-2 8. 79x10-s 7.23xlO-s 5.47x10-s 1.97x10-s 3.17x10-s
90% N02, 10% NO 1. 02x10 -2 7 .69x10-2 7 .26x10-s 4.21x10-s 1.96x10-3 2.42x10-3
50.0 Equimo1ar 2. 39x10:1 1.85x10 - 1 4.26x10-S 2.83x10-2 1.44x10-2 1. 82x10-s
10% N02, 90% NO 4.44x10 1 4.10x10-2 1.70x10-3 1. 20x10- 3 9.30x10-3 1.40x10-4
90% N02, 10% NO 4.85x10-2 2.97x10-1 1.44x10-3 6.76x10-2 8.37x10-3 8.34x10-;J
-------
Radian Corporation
8500 5HOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512-454-9515
concentrated gas mixtures containing mostly N02 found the
removal rate proportional to the concentration of NaO!;," The
equilibrium concentrations for the case of 50 mole % chemical
NO + NOa input concentration and composition 90% N02- 10% NO
are (in mole fractions):
NO;a
.297
Na04
.0676
NO
.0485
HNO:a
.00837
HN03
.00834
N;3 03
.00144
The properties which affect the rate at which a molecule can
be absorbed are its diffusivity, solubility, chemical reactivity
in the aqueous phase, and relative concentration and diffusivity
in the gas phase. The diffusivities of all these species are
of the same order of magnitude, so that property probably doesn't
significantly affect the absorption rate. Nitric oxide, NO, is
the least soluble of the species. In fact, its solubility is
so small that even if equilibrium conditions existed for NO
dissolution, the amount of nitric oxide removal observed would
not be accounted for by mass transfer of NO. The other species:
NOa, Na03, N:a04, HNOg and HNO~ then remain as candidates for
participant in the mass transfer mechanism and rate limiting step.
The aqueous phase chemical reactivities and relative gas'phase
concentrations remain as properties which influence the rate of
sorption. The relative concentrations are shown above in descend-
ing order. Clearly NOg and N304 are present in concentrations
orders of magnitude greater than HN02, HN03 and N203' The fact
that the absorption rate in gas mixtures of this type was found
to be proportional to the concentration of N:a04 rather than NOa
must then be due to the fact that N:a04 is much more chemically
reactive than N02 in aqueous solutions. The greater reactivity
could be accounted for by greater ease of ionization for N:a04
than for N02. Indeed, the proposed mechanism for reaction of
NOa-Na0'* mixtures with water (MO-008, p. 20; WA-015, p. 18; and
KO-026, p. 32) involves the sequence of reactions in equation (4-9).
-40-
-------
Radian Corporation asoo SHOAL CREEK 8LVD. . P. O. BOX '~\8 . AUSTIN, TEX~S 1am. TELEPHONE SI2 - 4SHm
N;} 04
NO+ +
\OH-
.
HNO~
+r
H + NO:,
~
NO,~
\H+
HNO:~
I ,
i i
H+ + NO;
(4-9)
Now that it has been shown how equilibrium concen-
trations can be used to determine mass transfer mechanism and
rate limiting step for a known case, let us consider the unknown
case of sorption from flue gas mixtures. From Table 4-3, the
calculated equilibrium concentrations for a gas mixture contain-
ing .05 mole % NO + NO:? composed of 90% NO and 10% NO:? are as
follows (mole fractions x 104) :
NO
NO::
HNO~
4.45
.448
.0999
HNOs
N20s
N:.;! 04
.00164
.0000184
.0000142
Consideration of solubilities and relative gas phase concentrations
shows clearly that NO, N20~, and N~O~ could not be the species
for describing mass transfer or rate limiting step. Again NO is
too insoluble to account for the amount of NO sorbed. Because
the vapor phase diffusivities of HNO~, HN03, N~02 and N~O.~ are
almost equal, the vapor film coefficients for all of these species
are almost equal. The relative amounts of mass transfer for the
various species will depend on the relative value of the driving
forces (concentrations). The concentrations of HNO~ and HNO~ are
orders of magnitude greater than N203 or N~O~. For the liquid
film the diffusivities of the unionized species are approximately
equal. The solubilities of HN02 and HN03 would be expected to
be larger than N203 or N204 because of the rapid ionization rates
of HNO~ and HN03 as compared to the rates of the chemical reac-
tions for N~03 and Na04' Therefore every step in the mass transfer
sequence is faster for HN02 and HN03 than for N403 or N20~. Ob-
viously the mechanisms proposed for absorption from gas mixtures
-41-
-------
Radian Corporation
8S00 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 787S7 . TELEPHONE SI2 - 454.9535
of the concentration and composition typical of nitric acid
plant tail gases do not apply to this case.
The species remaining to be considered are then NOa,
HNOa, and HN03. Although NOa is present in much greater gas
phase concentrations than HNOa or HN03, there is reason to
believe that it does not react rapidly in the aqueous phase
until it has dimerized to form Na04(,e) after which subsequent
ionization, hydrolysis, and ionization steps (equation 4-9)
must take place before nitrite and nitrate are formed. However,
HNOa and HN03 do not have to react further once they are dis-
soived in the aqueous phase. Only one ionization step has to
take place before nitrate and nitrite are formed. Furthermore,
if N02 diffusion and reaction were the only mechanism, nitrite
and nitrate would be formed in equimolar quantities in the
aqueous phase, which is not the case. The conclusion is that
HNOa and HN03 must be the species through which significant mass
transfer takes place during sorption from flue gas mixtures
since they offer the path of least resistance. The steps in-
volved in HNOa and HN03 mass transfer are bulk gas diffusion,
gas film diffusion, liquid film diffusion and ionization in the
bulk liquid. The slowest of these steps is probably gas film
diffusion, so that the proposed mechanism and rate controlling
step is gas film limited diffusion of HNOa and HN03. It is
also possible that the aqueous phase reactions in Equation 4-10
NOg(,e) ~ Na04(£) ~ NO+ + NO;
H:
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512. 454.'1$;5
5.0
EXPERIMENTAL PROGRAM
It is clear from the problem definition that insuffi-
cient physical and thermochemical data exist to enable one to
verify the predicted mass transfer mechanism and rate limiting
step for NOx aqueous sorption from flue gas mixtures. The usual
practice is to verify the predictions experimentally. An experi-
mental program was not included in the scope of work of Contract
EHSD 71-5. However, an inhouse experimental program at the OAP
Laboratory Research Branch Cincinnati laboratories was being
carried out during the period of contract performance. The
sor'ption of equimolar NO-NO:;! mixtures in a variety of aqueous
sorbents was under investigation. In order that the inhouse
program might receive the full benefit of Radian theoretical
studies, a contract modification was made according to which
Radian would cooperate closely with the existing experimental
program in an effort to verify the mass transfer mechanism.
Much valuable experience was gained as a result of the close
cooperation that was possible. Most of the experimental results
were obtained before the time that Radian became closely involved
with the program. These results were reported by Garcia
(GA-046) at the Second International Lime/Limestone Wet
Scrubbing Symposium. Radian participated directly in data
collection and analysis for only three experiments. Much
additional experience was obtained through discussions about
the experimental apparatus and the data obtained in experiments
in which Radian was not directly involved. In section 5 the
apparatus is described, but only the raw data obtained specif-
ically for Radian are presented. In addition, the practical
experience gained through cooperation with the experimental
program is discussed. In section, 6.0, an engineering
analysis of the data collected during Radian participation is
described.
-43-
-------
:';'
.. Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN. TEXAS 7B757 . TELEPHONE 512.454.9535
5.1
Apparatus and Data Collected
A flow diagram of the experimental apparatus used in
OAP inhouse studies is shown in Figure 5-1. The apparatus was
modified slightly during the experiments in which Radian partici-
pated, since the object of these experiments differed from the
general goals of the inhouse program. Instead of using a
simulated flue gas produced by combustion of methane in air
and addition of 802 and N02, the inlet gas was a mixture of
NO and N02 in nitrogen. The entering gas stream was not pre-
heated. The scrubber was a 2~-inch ID glass column packed with
18 inches of 3/8-inch Berl saddles. The measured gas flow rates
were corrected for pressure drop and density. Using the measured
gas and liquid flows, IR and UV analyses for chemical NO and
chemical N02, respectively, in the entering and effluent gas,
and UV analysis for nitrite and nitrate in scrubbing liquid
and condensate, the data shown in Table 5-1 were calculated.
A detailed description of the material balance calculations is
given in Technical Note 200-007-09, which is included in
Volume II of this report.
5.2
Practical Experience Gained
Experience gained by OAP and Radian investigators
during the study of aqueous sorption of nitrogen oxides revealed
the 6ifficulties involved in calculating material balances and
mass transfer coefficients in a system as complicated as the
NOx-H20 system.
-44-
-------
FLOW.
. I METER
I
FIGURE 5-1
FLOW DIAGRAM OF EQUIPMENT
USED IN EPA IN-HOUSE EXPERIMENTS
f RECORDER
INFRARED
ANALYZER
S02
/ RECORDER
INFRARED
ANALYZER
NO
f RECORDER
ULTRAVIOLET
ANALYZER
N02
CONDENSERS
FLUE GAS
PREHEATER
FURNACE
N02
TANK
CATCH
TANK
-45-
NO
TANK
CH4
AIR
502
TANK
FEED
TANK
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512. .454.9535
TABLE 5-1
SUMMARY OF DATA COLLECTED AT EPA LABORATORY. MAY 1971
(Mo1es/min x 106)
In Out
Scrubber
Run No. Entering Gas Exit Gas Liquid Condensate
1 CNO* 102.76 97.54 21 . 14 **
CNO 101.03 52.84 60.08
:2
2 CNO 135.60 134.05 21. 72 -3.2
CNO 130.80 68.79 62.95 9.0
2
3 CNO 172.80 185.12 23.61
CNO 162.00 85.95 70.56
2
*CNO and CNO refer to chemical NO and chemical NO:2 as
2
defined in the mass balance equations in the problem
definition.
**Ana1ysis conducted only for Run 2.
-46-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512.454.9535
Since the!re are many possible compounds of nitrogen and oxygen
and several oxidation states that are exhibited by the nitrogen,
it is important either to exclude or to account for the possibi:-
ity of oxidation. This factor was even more important for the
method of data analysis employed in this study. The material
balance E~quations were based on conservation of the +4 and +2
oxidation states of nitrogen. Therefore, experiments to verify
the mass transfer mechanism had to be performed in an atmosphere
of nitrogen rather than air or oxygen.
Second, it is necessary but quite difficult in practice
to obtain starting materials of known purity. It was found that
one commercial NO-Na cylinder contained about 25% NOa. An over-
sight of such an impurity would lead to obvious difficulties in
data analysis.
Difficulties in chemical analysis, as first pointed
out in the problem definition from literature reports, were
confirmed during the experimental studies. Some of these
problems were solved satisfactorily. It was found that at
elevated temperatures, less than 1% of the nitrogen-oxygen
species present exist in molecular forms other than NO and NOg.
Therefore, if analytical measurements are conducted at elevated
temperatures, the molecular NO and NOa concentrations are
essentially equal to the chemical NO and chemical NOa required
as inputs to the equilibrium model. The relative concentrations
of other gas phase molecules can then be calculated by the
equilibrium model at the lower temperatures of interest.
The fact that water vapor interferes in the gas phase
analyses was not solved satisfactorily. The method used to
eliminate the interference was the application of two water-
cooled condensers. Unfortunately, some of nitt'ogen oxides are
-47-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. 80X 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
removed in the condensate along with the water vapor. The
problem was circumvented by chemical analysis of the condensate
for nitrite and nitrate.
-48-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 7B757 . TELEPHONE 5i2. 450535
6.0
ENGINEERING ANALYSIS OF SORPTION DATA
Insufficient data were taken and not a wide enough
range of vapor and liquid rates was investigated to prove or
disprove the proposed mass transfer mechanism or to provide a
correlation that can be applied with confidence. The limited
amount of data was analyzed on the basis of vapor film limited
mass transfer of HNO,. and HNO.o. and mass transfer coefficients
tG .. .
for HN02 and HN03 were calculated which c~n be applied for
vapor rates of from 0.7 to 1.2 normal meter3/hr. No attempt was
made to account for the influence of NO~ mass transfer.
6.1
Problem Formulation
The data ana1ys is proced ure differs slight 1y from
standard methods of calculating mass transfer coefficients
since the species being removed from the gas phase, HNO~ and
HNOs, are not conserved, i.e., sorbed HNO~ and HNOs are quickly
replaced due to the reactions between N~03' N204 and H20.
The problem formulation is given in detail in Techni-
cal Note 200-007-12, included in Volume II of this report. It
is summarized in the following paragraphs. An expression for
the number of moles of HN02 or HNOs transferred in a differen-
tial height of packing, dz, of crossectional area Ac is given
in equation (6.1).
-dnx = kxaArP(Yx-y~)dz
(6-1)
The subscript x refers to HN02 (x = 1) or HNO~ (x = 2). P is
the total pressure, a is the sUl:face area of packing per unit
volume, k is the mass transfer coefficient, and y is the mole
fraction in the bulk gas. Assuming the equilibrium partial
-49-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
pressure over the liquid, y*, is zero results in equation (6-2).
-dnx = kxaAcPyx dz
(6-2)
The quantities that can be measured in sorption
experiments are chemical NO CNO and chemical NO~ CN02' These
quantities were defined by the mass balance equations in equations
(4-6) and (4-7). The change in the numbe~ of moles of chemical
NO and N02 corresponding to the change in number of moles of
HNO; and HN03 are therefore shown in equations (6-3) and (6-4)
from the mass balance equations.
dCNO = ~(dnHNO dnHNO)
a 3
dCNO = ~(dnHNO + 3dnHNO )
a a :3
(6-3)
(6-4)
Substituting (6-2) into (6-3) and (6-4) gives (6-5a) and (6-5b)
where ~l is kHNOaaPAc and ~2 is kHNO:3aPAc'
-dCNO = ~(~lYHNO 0'2 YHNO:3) dz
2
-dCNO = ~(~lYHNO + 3~QYHNO )dz
2 a :3
(6-5a)
(6-5b)
The amount of chemical NOa and NO removed in a column
of height H can then be calculated by numerically integrating
(6-5a) and (6-5b). Values for al and ~a are then calculated
such that the difference between the amount removed calculated
using (6-5a) and (6-5b) and the measured amount removed is
minimized. A computer program was written to perform the
calculations. It utilizes the gas phase equilibrium model
subroutine for calculating mole fractions of HNOa and HNOs for
-50-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.,535
given chemical NO and NO:!, pressure, and temperature, A first
guess for 01 and as is obtained by calculating mass transfer
coefficients for HNO;G and HN03. The first trial mass transfer
coefficients were calculated using a correlation for the vapor
film coefficient for the ammonia-water system (BR-002, p, 530)
and correcting for the difference in diffusivities.
6.2
Data Used in Correlation
The data shown in Table 5-1 were used in the correla-
tion for calculating mass transfer coefficients, The measure-
ments of gas phase concentrations were not as accurate as the
results of the liquid phase chemical analyses. Therefore, the
data were recalculated to place more emphasis on the most
A II "'1 '
accurate measurements. n average 1n et concentrat10n,
In was recalculated, All of the inlet CNO and CNO entered
avg. 2
in the gas. The CNO and CNO left in both the gas and the
2
liquid streams. The material balance can then be represented
by equation (6-6),
In = Out + Outl' 'd
gas gas 1qu1
(6-6)
Although In , Out, and Outl' 'd were all measured, the
gas gas 1qU1
liquid phase measurements are the ones in which one should
have the most confidence. Therefore, the inlet concentration
was recalculated as shown in equation (6-7).
2(In ) = In + Out + Outl' 'd
avg gas gas 1qu1
or
(6-7)
In = k(In + Out + Outl' 'd)
avg 2 gas gas 1qu1
-51-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512 - 454.9535
The recalculated outlet concentration is shown in equation (6-8).
Out = In
avg,gas avg
Out 1 . . d
~qu~
(6-8)
The values for In and Out used in the correlation
avg avg,gas
were calculated from the data in Table 5-1. They are shown
in Table 6-1.
TABLE 6-1
RECALCULATED VALUES OF CNO AND CNO
:2
IN THE INLET AND EFFLUENT GAS IN NOx
SORPTION EXPERIMENTS
Gas In Gas Out
(mo1es/min x 106) (mo1es/min x 106)
Run 1 CNO 110.72 89.58
CNO 106.97 46.89
::I
Run 2 CNO 144.10 122.38
CNO 135.75 72.80
~
Run 3 CNO 190.75 167.14
CNO 159.25 88.69
::<
-52-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
6.3
Results
The vapor film coefficients for HNOa and HN03 calcu-
lated as descr ibed in Section 6.1 with the data given in
Section 6.2 are shown in Table 6-2.
TABLE 6-2
RESULTS OF MASS TRANSFER COEFFICIENT CALCULATIONS
Vapor Flow Rate Liquid Rate KgaHNOa KgaHN03
Run r-normal M3 ] [~T~J [ fm mole J I fm mole]
No. - hour (hr) cm3) (atmL __(hr) cm3) (atm)
1 0.728 710 0.277 0.972
2 0.948 710 0.275 0.962
3 1.18 710 0.302 1.22
The vapor and liquid rate dependence of the coefficients can be
described by equations (6-9) and (6-10).
kgaHNOa = 0.795 (:L)0.8 (L \0.39
7\;)
.Ac
kgaHN03 = 2.96 (V \0.8 (L \0.39
-) p;;)
\Ac I
(6-9)
(6-10)
-53-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
These equations and this mechanism must be considered
as being tentative 0 The ratio of kgaHNO:;! /kgaHNOs is 00.27 and
not 1.04 as predicted by film theory. This could be explained
by inaccuracy of equilibrium constants used to calculate mole
fractions of HNO:2 or HN030
-54-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHCNE 512.454.95)5
7.0
SCREENING OF CANDIDATE SORBENTS ON THE BASIS OF
THE THERMODYNAMICS OF THE SORPTION REACTIONS
The ultimate goal of this study was to
theoretic,a1 description of aqueous scrubbing and
theoretical description for two purposes:
develop a
to use the
to satisfy some of the engineering
requirements of new process development
to predict the most effective and
efficient sorbents for use in the pro-
cess.
This section describes how candidate sorbents were screened for
effectiveness
Screening was
the reactions
in two different processes for NOx removal.
conducted on the basis of the thermodynamics of
between nitrogen oxides and the candidate sor-
bent.
7.1
Description of the Processes for which Candidate
Sorbents were Screened
The metal oxide sorbents considered for applicability
in NOx removal processes are listed in Table 7-1. Two processes
were considered: a dry process and an aqueous process. Both
start with dry metal oxides and result in the formation of
solid metal nitrites and/or nitrates as end products.
The processes are regenerative. Regeneration could
be accomplished either by temperature, pressure, or chemical
changes. One example of a method for regenerating the starting
-55-
-------
Radian Corporation
8000 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78707 . TELEPHONE 512.454.9535
TABLE 7-1
METAL OXIDES SCREENED FOR APPLICABILITY
IN NOx REMOVAL PROCESSES
Ag20 Fe:;j°3
Aln 03 K20
BaO LisO
BeO MgO
Bi203 MnO
CaO Mn:;j03
CdO Na20
CeO.. NiO
~.
CoO PbO
CugO Sn02
CuO SrO
FeO ZnO
Zr02
-156-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.95;5
material is thermal decomposition. Many metal nitrites or ni-
trates can be ther~ally decomposed at temperatures of severa)
hundred °c to yield the starting material metal oxide and gaseous
nitrogen oxides. The published data on thermal decomposition of
nitrites and nitrates are summarized in Technical Note 200-007-06
(see Volume II).
The dry process under consideration involves sorption
of gaseous nitrogen-oxygen compounds on dry metal oxides to
form metal nitrates and/or nitrites. Regeneration of the sor~
bent is by thermal decomposition of the sorption products. The
gaseous nitrogen compounds are to be removed from a flue gas
containing both water vapor and CO2 so that formation of
carbonates and hydroxides must also be considered.
The aqueous process for which screening was carried
ou t is described in Figure 7-1.
P;
Flue
PI
I
(1) (5)
Scrub- Disso1v
ber er
i
Gas
~
(2)
Crystal-
lizer
(4)
Hydra- --
tor
(3)
NO+t\O; ~ Decomposer
H;O
FLOW DIAGRAM FOR AQUEOUS PROCESS
FIGURE 7-1
-57-
-------
Radian Corporation
steps:
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN. TEXAS 78757 . TELEPHONE 512 - 454-9535
The process shown in Figure 7-1 involves the following
1.
hydration of the metal oxide to form
the hydroxide
2.
dissolution of the hydroxide
3.
sorption of gaseous nitrogen oxide
species from flue gas containing CO2 (g) ,
and H20(g) into the aqueous solution
4.
crystallization of nitrates and/or
nitrites
5.
thermal decomposition of the nitrates
and/or nitrites to regenerate the dry
metal oxide.
One aspect of the screening procedure should be
emphasized. The material or chemicals that were selected as
potential sorbents are dependent on the process. Should a
different process be used, e.g., heating of the aqueous sorbent
to drive off NOx as the regeneration step, a different list of
potential sorbents might result.
There were several assumptions on which the screening
process was based. It was assumed that except for the sorption
step, all the other process steps were reversible and could be
carried out at equilibrium. However, there are some steps
which are known to be irreversible. Therefore, it was necessary
to account for the free energy loss resulting from hydration of
the sorbent or reaction with cO~ to form carbonates.
-58-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHO~JE 512.454,,5)5
7.2
Thermodynamic B'as is for Screening
Screening of candidate metal oxides can be carried
out by considering the free energy changes for the reactions
involved in the processes of interest. The overall change in
free energy provides the driving force for a process. The
object of screening is to select the metal oxide for which the
reactions have the most efficient free energy change. It must
be large enough to supply driving force for the process, but
not too large. If the free energy change in one step of the
process is too large, then it is likely that an expensive
energy input will be required at some other stage in the pro-
cess. An example of this is the regeneration step where the
free energy change has to be reversed.
The free energy change that we wish to evaluate is
the difference between the free energies of two states. The
first state of interest is NO+N02 in the effluent flue gas in
which the concentration of NO+NOI;} has been reduced to some accept-
able level, say 50 ppm. The second state of interest is NO+NO~
gas with the concentration of NO+N03 equal to the equilibrium
partial pressure of NO+N02 over the scrubber liquid or the
dry sorption medium. The free energy of either state is given
in equat :lon (7 -1) .
t-.G = -RT£.nP
NOx
(7-1)
It is convenient to express PNOx in terms of one component of
the mixture of gaseous nitrogen compounds. This is acceptable
since all of the components are at equilibrium. We choose
either NaOs for nitrite based processes or N~05 for nitrate
based processes. The free energy difference between the two
states of interest is then given in equation (7-2) for processes
based on nitrate formation.
-59-
-------
Radian Corporation 8500 SHOAL ClEEK BLVD. . P. O. lOX 9948 . AUSTIN. TEXAS 78766' TELEPHONE 512 - 454-9535
6GState 1 - 6GState2 = [-RTJ,nPN;Os (State 1)] - [-RTJ,nPNj306 (State 2) ]
[PN20s(State 1) ]
= -RTJ,n
P
NaOs(State 2)
(7-2)
In equation (7-2)~ PNQOs(State 1) is the partial pressure of
N;Os desired in the exit flue gas, i.e., PN~05 when PNO+NOQ =
50 ppm. PN20s(State 2) is the equilibrium pressure of NaOs
over the dry or aqueous sorbent.
According to equation (7-2) the task of thermodynamic
screening is as follows: for each metal oxide calculate the
partial pressure of N20s or NaOs over its aqueous solution or
over the dry sorbent, PN~Os in State 2, and compare that
pressure with the pressure of NaOs or NaOs desired in the exit
flue gas, PN ° in State 1. Obviously the pressure of NaOs or
a s
N20s must be lower over the sorbent than that desired in the
exit flue gas for sorption to occur. In addition, it must be
low enough that the term -RTJ,n\PState l/PState 2) supplies a
free energy difference sufficient to provide driving force for
the process.
Now let us consider how to calculate the partial
pressures of N~03 or NaOs at equilibrium over the metal oxide
sorbent. Since both processes start with a dry metal oxide
and end with a solid metal nitrite or nitrate, equations (7-3)
and (7-4) are valid ones for describing both overall processes.
Me(NOs)a(s)
~
NaOs(g) + MeO(s)
(7-3)
Me(NOa)a(s)
~
N203(g) + MeO(s)
(7-4)
-60-
-------
Radian Corporation
8500 SHOAL CREEK BLVD.' P. O. BOX 9948 . .AUSTIN. TEXAS 78757 . TELEPHONE 512 - 454-9535
It was assumed that all the steps in the process approach
equilibrium and there are no steps other than the sorption
step which have excessively large free energy changes. If all
other steps could be carried out reversibly then Equations (7-3)
and (7-4) can be used for sorption free energy calculations.
However, we know there are some steps which cannot be conveniently
. carried out in a reversible manner. We must therefore account
for free energy loss that will occur, for example, due to carbo-
nate fo~nation in either process. This can be done by con-
sidering the reverse of reactions (7-5) and (7-b).
Me(NOs)2(S) + CO2 (g)
+!
NaOs(g) + MeCOs(s)
(7-5)
Me(NOa)2(S) + COa(g)
+!
NaOS(g) + MeCOs(s)
(7-6)
The equilibrium constant for reaction (7-3) is shown
in Equation (7-7) where a is the activity.
a 0 a
= Na 5(g) MeO(s)
aMe(NOs)2(S)
K (T)
eq
(7-7)
If it is assumed that the activities of the solid species are
one (which is a good assumption since there are usually no
solid solutions formed) then the equilibrium constant is numeri-
cally equal to the activity of the gaseous product. The total
pressures in which we are interested are low and the tempera-
tures are high. We can therefore assume the fugacity coefficient
for the gaseous species are one. Under these conditions the
equilibrium constant is numerically equal to the partial pres-
sure of the gaseous product in atmospheres.
-61-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454-9535
The equilibrium constant for a reaction at the
temperature of interest can be calculated from standard state
thermodynamic properties and heat capacities as a function of
temperature. Since the equilibrium constant is numercially
equal to PN 0 or PN 0 in atmospheres, calculation of K at
2 5 a 3 eq
the sorption temperature yields the data for PNa 05 or PN" 03(State 2)
needed for screening.
For reactions (7-5) and (7-6) the same procedure
was used. The equilibrium constant is shown in Equation (7-8)
where activities of the solid species have been assumed equal
to one.
Keq(T)
a
= Na05(g)
a
CO2 (g)
(7-8)
The partial pressure of COa in the flue gas was taken to be
0.147 atm, i.e., 14.7%. The standard state for CO2 used in
these calculations was 0.147 atm so that the equilibrium con-
stant is again numercially equal to PN 0 or PN 0 in atmospheres.
2 3 2 5
7.3
Data Collection and Calculations
7.3.1
Calculation of N70~ and NaOs Pressures
Over the Sorbent
The basic data needed for calculating PN 0 or PN 0
2 3 2 5
over the sorbent for both the dry process and the process
illustrated in Figure 7-1 are standard state thermodynamic
properties and heat capacities as a function of temperature.
The data were needed for products and reactants in nitrate and
nitrite decomposition reactions (7-3), (7-4), (7..5), and (7-6).
Section 8.0 describes the collection and tabulation of the
-62-
-------
<-
Radian Cc)rporation
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 7B757 . TELEPHONE 512.454.9535
thermodynamic properties. No thermodyn<:tmic data had been
published. for some of the products and reactants of interest,
so the properties were estimated. The estimation methods and
the results are also described in Section 8.0.
Using the thermodynamic properties and heat capacities
of reacta.nts and products, the equilibrium constants for reac-
tions (7-3), (7-4), (7-5), and (7-6) were calculated as a
function of temperature using a previously developed computer
program, AIRPOL (PA-016). AIRPOL generates graphs of 10gl~K as
a function of temperature for reactions involving products and
reactants for which thermodynamic data have been stored. The
computer program normalizes the equilibrium equation to one mole
of the first product formed. It is for this reason that the
reactions have been written as decompositions with the nitrogen
oxides as the first product. The calculated equilibria are
numerically equal to the nitrogen oxide partial pressure.
7.3.2
Calculation of Na03 and N20s Pressures in the
Flue Gas
The method for calculating PN ° and PN ° at
equilibrium in the flue gas when 50 ppm2NO+NOa ar~ ~resent has
been previously described (see T.N. 200-007-03a). Radian's
gas phase equilibrium model was used to perform the calculations.
Two compositions were used for input to the program: 40 ppm
NO + 10 ppm NOa and 25 ppm NO + 25 ppm NOa. The calculated
amounts of Na03 present were approximately the same for the two
compositions, but the amount of NaOs present differed for
different compositions.
7.3.3
Results of Calculations
The results of the
(7-3) and (7-4) are given in
calculations of 10gloK for reactions
Figures 7-2 and 7-3 respectively.
-63-
-------
~~ ---------=::--~ ~
.--.: -- -::- --- -------=~- - -
-- ~- -:;::::;..--
-- -- - - - ~ - --=-===-
--- ...- ~ -- ----
~ --- ~ ~-----
o C~';~/" //'~ ~ ~~~-----
~J /~~ ~ -~---=:- ---
~cO~~~o ~~/ ~ ".---;:. - -- -
0;-/ ~v -' ~ ~ --- --
c~ / O':y ~'/ , ---
-6 _:/ ~~/f ~~ / /" --- --------
~ O--~;h.OO h # / ~
~ +",.' ~ "/" /
~ -10 ~~~",Op / / .-/ ~
o~ ~/f' /' // -!.0IOPNO
~~ ~(~~ /' In fluo !a ~ -- ~
~ {'/. / ~ -
£'\ -
0' -15 '<:;~. i / /
~ I/o / /
- /,~ / / /
8 "\'/ / / ./
.. /'~ il I '/ / /' /' /'
/ ~~'/ /
/ '/ /
/ I /
0' / /
j' ~ if 0
'! "/ ¥
1.1, I I I I ,'t I I , I I , I II I I I, I 1'" UlJ 1'1 , , I I I' 1" I , U " I 1.1 I , I I I I I I, I I I 111'-
100 200 300 400 500 GOO 700 800
FIGURE 7 - 2
Comparison of N203 Partial Pressure io Flue Gas and
Vapor Pressure of N203 over Metal Oxide Sorbents
10
5
--
~
Q
~
-25
-30
o
TEMPERATURE - DEGREES CENTIGRADE
-64-
-------
,-
Comparison of N FIGURE 7 - 3
Vapor Pr :106 Partia'l P
essure of N rcssure in F
. ,0, over Metal 0 lue Cas aod
I ' xide Sorb
cnts
I I
5
o
000
-65-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 994B . AUSTIN, TEXA5 78757 . TELEPHONE 512 - 454-9535
The data were calculated by the computer program AIRPOL (PA-
Olo), and they show 10gl~Keq for reactions (7-3) and (7-4) per
. mole of N203 or N20s produced as a function of temperature. The
common logarithm of the partial pressure of N20s or N20s in the
flue gas at 50 ppm NOx is also plotted on the graphs as a
function of temperature. Since the partial pressure of N203 did
not vary greatly for the two compositions considered, only one
line was plotted for N~03' However, a broken line and a solid
line were plotted for N~Os' The broken line indicates 10gloPN 0=
- Ii! ~
when the flue gas composition is 25 ppm NO and 25 ppm N02. The
solid line indicates PN ° when the flue gas composition is 40
2 S
ppm NO and 10 ppm N02.
The results of the calculations of 10gloK for
reactions (7-5) and (7-6) are shown in Figures 7-4 and 7-5.
The logarithms of the partial pressures of N203 and N20s in
flue gas at 50 ppm NOx are also plotted on the figures.
7.4
Results of Thermodynamic Screening
7.4.1
General Considerations
The object of the screening process was to select
metal oxide sorbents over which the equilibrium vapor pressures
of NaOs or N206 were lower than the desired effluent flue gas
pressures of N203 or NaOs. The equilibrium vapor preSSla.es over
the sorbent must be low enough compared to flue gas pressures
so that the resulting free energy difference is great enough to
supply driving force for the process. On the other hand, the
vapor pressures cannot be so much smaller that too great a free
energy difference results. The reason for this is that the
-66-
-------
FIGURE 7 - 4
Comparison of N203 Partial Pressure in Flue Gas and
Vapor Pressure of Na03 over Metal Carbonate Sorbents
10
5
Cs.C03 - - - - - - - - -
- --
- .-
- -
~
~
~
~
~
~
o
-6
-10
CO)
o
""
-
o
-15
'"'
C>
o
....:I
-25
-30 -
o
----------
---------
--
--...-.......
srC03 ---- -
-----
-------
. . --------- r-v--
. ----- .. ~.~ -
CO~ .' ....-::=::-.'_~-_'--
'" . ~.. ~--
~h.' . ~ . -""" --=--- -
. .-b~ --- ~,..:::-:::=---
. . ~';>.G ~ ~--
/'" . ~0 -::::'-
/'" ~ yj\.\?~~-
,~~~~J. Y C6~
.~-F::' /' 't-?
. /'
'"
:::? ~ /'
//"
"
- ,/
o
log,o PfJ 0 In fluo go s
2 3
1111111'111111111111 "IIIIIIIIIIII!IIIIIIIIIII'II~' ,11111111111'111111'-
100 200 300 400 500 COO 100 000
TEMPERATURE-DEGREES CENTIGRADE
-67-
-------
FIGURE 7 - 5
CompDrison of "20& Partial Pressure in Flue Gas and
Vapor Pressure of ";0& over Metal Carbonate Sorbcnts
5
~
~
-6
-10
,....
o
....
-
8
~
---------
-20
10010 PN ° In flue go s
2 5
o
100
I 1'1 [I t f I I I I r I I I I I I I,'
100
200
300
400
500
GOO
TEMPERATURE-DEGREES CENTIGRADE
-68-
000
-------
.Radian Corporation
8500 SHOAL CREEK BLVD. . P. o. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-1535
free energy change
step. An expensive
a large free energy
will have to be reversed in the regeneration
energy input will be required to reverse
change.
The vapor pressures over metal oxides sorbents can
be compared to flue gas pressures using Figures 7-2 and 7-3.
For the aqueous process screening the pressures were compared at
a scrubber temperature of 50°C. For the dry process screening
the pressures were compared in the temperature range of 100 to
750°C. To describe processes resulting in nitrite formation
the pressures of NQ03 over the sorbent were compared with the
NaOs pressure in the flue gas. To describe processes resulting
in nitrate formation, the pressures of Na05 over the sorbent
were compared with the Na05 pressure in the flue gas.
7.4.2
Screening for Dry and Aqueous Processes Based
on Nitrite Formation
First let us consider the results of screening for
processes which result in nitrite formation. Screening for
the aqueous process can be accomplished by comparing loglcPN °
. :a 3
in the flue gas at 50°C with the vapor pressure of N~03 over
the metal oxide sorbent at 50°C. If the vapor pressure over
the sorbent is greater than, equal to, or only slightly less
than the flue. gas pressure at 50°C, then the sorbent would not
~ be suitable for an aqueous removal process involving nitrite
formation. If the vapor pressure over the sorbent is at least
two orders of magnitude less than the flue gas pressure,. then
sufficient driving force to run the process may result. The
metal oxide would be suitable for further consideration as a
sorbent. Based on the reasoning described above and the data
in Figure 7-2, the oxides BeO, CdO, CoO, CuaO, CuO, FeO, MgO,
-69-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512 - 454-9535
MnO, NiO, and ZnO were judged unsuitable for use in an aqueous
sorption process involving nitrite formation. The nitrites
of these metal oxides are too unstable to sorb NO+NOa at 50°C.
In fact, no evidence for the existence of nitrites of BeO, Cu~O,
FeO, MnO, or ZnO has been reported in the literature. It was
not possible to calculate Na03 pressures over the nitrites of
Ala03' Bia03' CeCa, Fea03' MnaOa, SnOe or ZrOe because of our
inability to estimate some of the thermodynamic properties
needed. However, again we found no evidence in the literature
for the existence of these nitrites.
Screening for the dry process resulting in nitrite
formation can also be accomplished using the data in Figure 7-2.
For the dry process comparison of N~Oa vapor pressures over the
sorbent with flue gas pressures was made at temperatures above
100°C. The same reasoning was used for the dry process screen-
ing as was used for the aqueous process. It was found that all
of the metal oxides that were unsuitable for the aqueous process
at 50°C were also unsuitable for the dry process at 100°C for
the same reason: vapor pressure of Na03 over the sorbent was
greater than or equal to PN;Os in the flue gas. In addition,
the above was also true for PbO and AgaO at 100°C, although.
these oxides were suitable for the aqueous nitrite process at
50°C. The same results for AleOa, Bia03' CeCa, FeeOa, Mna03'
SnaG, and ZrOa hold for the dry process as well as the aqueous.
In summary, it appears that the oxides Ag20, BaO, CaO,
KaO, LieO, NaeO, PbO, and SrO remain to be considered as sorbents
for an aqueous process at 50°C involving nitrite formation. For
a dry process at 100°C involving nitrite formation, the same
candidate oxides except for PbO and AgeO remain. These results
are summarized in Table 7-2.
-70-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 78757 . TELEPHONf 512.454-9535
TABLE 7-2
METAL OXIDES. UNSUITABLE FOR USE IN PROCESSES
BASED ON NITRITE FORMATION
Dry Process Aqueous Process
AgaO Be01
Be01 CdO
CdO CoO
CoO Cu 01
a
Cu:;! 01 CuO
CuO Fe01
Fe01 MgO
NgO Mn01
MnOl NiO
NiO 2n01
Al 01, 2
PbO a 3
21101 B. 01' a
1.a 3
Al a O~ ' 2 CeO;':;!
B. 01,2 F 01' 2
1:;1 3 e:;1 3
CeO~ ,2 Mn 01,2
a 3
F 01,::! SnO~ ' 2
ea 3
M 01,2 2rO~,2
n:;1 3
SnO~,2
2rO; ,2
1.
2.
No evidence for existence of the nitrite was found in the literature.
Thermodynamic properties for the nitrite could not be estimated.
Therefore, no partial pressure calculations could be made to predict
nitrite stability.
-71-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
7.4.3
Screening for Dry and Aqueous Processes Based
on Nitrate Formation
Screening the sorbents for use in processes based on
nitrate formation was accomplished by comparing vapor pressures
of NaOs over the sorbent with flue gas NaOs pressure at 50°C
for the aqueous process and 100°C for the dry process. Using
the data in Figure 7-3 it was found that the vapor pressure of
NaOs over the sorbent at 50°C was greater than or approximately
equal to that in the flue gas for the oxides Al;O~, BeO, Bia03'
CoO, CuaO, CuO, FeO, Fea03, Mna03' NiO, SnOa, and ZnO. These
oxides were therefore eliminated from consideration as sorbents
in aqueous nitrate forming processes. At 100°C those oxides
mentioned above plus th~ oxides CdO, CeOa, MgO, and ZrOa were
found to give vapor pressures of NaOs greater than or approxi-
mately equal to the NaOs pressure in the flue gas. These
oxides were therefore no longer considered as candidate sorbents
for dry nitrate-forming processes.
In summary it was found that the oxides AgaO, BaO,
CaO, CsaO, KaO, LiaO, NaaO, PbO, and SrO would be suitable for
use in both the dry and the aqueous process based on nitrate
formation. In addition, the oxides CdO, CeO;, MgO, MnO, and
ZrOa remain as possible sorbent candidates for the aqueous
nitrate-forming process only. These results are summarized
in Table 7-3.
7.4.4
Screening Metal Carbonate Sorbents
will probably be
free energy tha t
Some of the steps in the processes under consideration
irreversible and will involve dissipation of
would otherwise be available for driving force
-72-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 78757 . TELEPHONE 512.454-15;5
TABLE 7-3
METAL OXIDES UNSUITABLE FOR USE IN PROCESSES
BASED ON NITRATE FORMATION
.Dry Process
A1g 03
Aqueous Process
BeO
A1g 03
BeO
Big 03
Bia 03
CdO
CoO
CeOg
Cu 01
:2
CoO
CuO
Cu 01
2
FeO
CuO
Fea 03
1
Mn203
FeO
Feg03
NiO
MgO
SnOg
MnO
ZnO
1
Mn:a03
NiO
SnO~
ZnO
ZrOg
1.
No evidence was found in the literature for existence of
the nitrate.
-73-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
for the process. One step that is anticipated to so reduce
the available free energy is that of carbonate formation by
sorption of CO2 present in the flue gas. The reduction in
available free energy due to carbonate formation can be taken
into account by considering reactions (7-5) and (7-6). V3por
pressures of NgOS and N205 over metal carbonate sorbents in
the presence of 14.7% CO2 were obtained by calculating log K
for reactions (7-5) and (7-6). Those pressures calculated in
the presence of CO2 were again compared to flue gas pressures
to complete the screening process. Calculations were done
only for those metal carbonates whose oxides were not eliminated
on the basis of nitrite or nitrate instability.
The data in Figure 7-4 were used for screening
carbonate sorbents for dry and aqueous processes involving
nitrite formation. Again, the vapor pressure of N~03 over the
sorbent must be at least two orders of magnitude less than the
pressure of NaOs in the flue gas. For the aqueous nitrite
forming process at 50°C, potassium is the only suitable sorbent.
Lithium is a borderline case. For the dry nitrite forming
process at IOO°C, no metal carbonate sorbent is suitable.
To screen metal carbonate sorbents for nitrate forming
processes, the data shown in Figure 7-5 were used. It was
found that for aqueous nitrate forming processes, all of the
sorbents considered were suitable except for CdO, MgO, MnO,
and ZrOa' It was found that for dry nitrate forming processes,
none of the sorbents were suitable except for BaO, KaO, NaaO,
and SrO (borderline).
-74-
-------
L
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. 80X 9948 . AUSTIN. TEXAS 78757 . TELEPHONE512. 4S4.mS
7.4.5
Summary and Conclusions
Metal oxide sorbents were screened for applicability
in dry or aqueous, regenerative NOx sorption processes. The
screening was done on the basis of the thermodynamics of the
reactions between the metal oxides and nitrogen oxides. The
free energy difference between two states of interest was
investigated. One state was the equilibrium vapor pressure of
the nitrogen oxide species over the dry or aqueous sorbent.
The other state of interest was the equilibrium concentration
of the nitrogen oxide species in the flue gas at 50 ppm NO+N02.
A sorbent was sought such that the free energy
difference between the two states of interest was great enough
to provide driving force for the removal process. For some
sorbents, the vapor pressure of the gaseous nitrogen compound
over the sorbent was greater than the equilibrium partial
pressure desired in the flue gas. Those sorbents which were
judged unsuitable on the basis of nitrite instability were BeO,
CdO, CoO, Cu20, Cuo, FeO, MgO, MnO, NiO, 2nO, A1203, Bi203,
Ce02, Fe~03' Mn203, Sn02' and 21"02. Those which were judged
unsuitable on the basis of nitrate instability were A1203, BeO,
Bi203, CoO, Cu20, CuO, FeO, Fe203, Mn203, NiO, Sn02 and 2nO.
Further calculations were made taking into account
that the presence of CO2 in the flue gas would reduce the free
energy difference available for driving force. The result of
those calculations was that potassium is the only oxide suitable
for use in a nitrite forming process. For nitrate forming
processes Ag20, BaO, CaO, K20, Li20, Na~O, PbO, and SrO are
potential sorbents. The results are summarized in Table 7-4.
-75-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 7B757 . TELEPHONE 512.454-9535
TABLE 7-4
POTENTIAL SORBENTS AFTER SCREENING
Nitrite Forming Processes
Nitrate Forming Processes
Aqueous Dry Aqueous Dry
Process Process Process Process
~O None Ag:aO BaO
(Li:a 0) BaO ~O
CaO Na:aO
K:aO SrO
Li:aO
NaaO
PbO
SrO
It should be re-emphasized that the screening
described here is based upon the formation of a solid nitrate
or nitrite. Processes in which solids are not formed, e.g.,
heating of the liquid to drive off sorbed NOx, could be based
upon some of the metal cations screened out in this study.
-76-
-------
Radian Corporation
8S00 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 787S7 . TELEPHONE SI2 - 45nS~5
8.0
THERMODYNAMIC PROPERTIES
One of the most important requirements for creating
a theoretical description of aqueous NOx scrubbing are thermo-
dynamic properties of nitrogen oxides, oxyacids and their.
sorption products. The thermodynamic data of interest include
standard state properties of formation and heat capacities of
compounds and temperature dependence of equilibrium constants
for dissociation, dissolution, decomposition, and vaporization
reactions. Data such as these were used in the gas phase and
aqueous phase equilibrium models and in thermodynamic screening.
They are also needed in engineering evaluation of experi-
mental sorption data.
The data must be collected mainly from the literature.
To measure even one equilibrium constant or heat of formation
would require several months of work. After data were located in
the literature .they had to be considered carefully and usually
converted to a form consistent with other data being used.
Often" differing values for the same cons tan tare repor ted.
Then, a dec.ision must be made concerning which value to use.
If a value for a particular constant of interest is not re-
ported in the literature, it must be calculated or estimated.
The following sections describe the collection of thermodynamic
properties of interest, the evaluation of data, and how the
data were used in the theoretical description.
8.1
Standard State Thermodynamic Properties and
Heat Capacity Data
entropy,
products
If the standard state heat of formation, absolute
and heat capacity as a function of temperature of
and reactants are known, the change in free energy
-77-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
and thus the equilibrium constant as a function of temperature
for a reaction can be calculated directly. Therefore, values
of these thermodynamic properties were compiled for the gaseous
nitrogen oxides and oxyacids, and solid metal nitrates, nitrites,
and hydroxides. Data for single and double metal oxides, metal
carbonates, sulfites, sulfates, and sulfides and elements had
been collected, evaluated, and compiled as described previously
(PA-016). The sources of data were compilations of thermody-
namic properties (ST-006, ST-017, ST-018, ST-019, RO-007, WA-OOl,
WA-Ol8, LA-008, KE-009 - KE-014, CO-048 and KU-003) as well as
the open literature which was surveyed using Chemical Abstracts.
Published data were availabie for all the gaseous compounds of
interest. However, some of the data for solid metal nitrates
and nitrites could not be found in the literature. Those data
were estimated using estimation methods developed previously
(PA-01b). The standard state properties and heat capacities were
stored in a computer data base. A storage and retrieval program
was used including options by which the free energy change and
equilibrium constant for a reaction can be calculated from the
stored thermodynamic properties. The details of the standard
state thermodynamic property compilation are given in T.N. 200-
007-04a which is included in Volume II of this report. The
note includes a copy of the entire data base including the
bibliography as well as a description of the selection of a
value when differing values were reported.
The standard state properties and heat capacities
stored in the data base were used as follows:
to calculate equilibrium constants as a function
of temperature for reactions considered in the
gas phase equilibrium model (Section 4.0),
-78-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454-,5)5
to calculate equilibrium constants as a
function of temperature for chemical
reactions involved in the thermodynamic
screening process (Section 7.0).
The computer program used for data storage and
retrieval also calculates and plots equilibrium constants
versus temperature. The calculations and the plots
generated for this work are contained in Technical Note 200-
007-15 in Volume II of this report.
8.2
Measured Equilibrium Constants
Equilibrium constants for some of the reactions of
interest for the theoretical description had been measured and
reported in the literature. The reactions of interest for
which constants were reported and evaluated are shown in
equations 8-1 through 8-5.
HN03(g) i! HN03 (.g,) (8-1)
HNO:. (t) ~ H+ + NO; (8-2)
HNO~(g) 4"! HNO:2(t) (8-3)
HNOs.? (t) 41 H+ + NO; (8-4)
NO(g) ~ NO(t) (8-5)
I
I
i
I
For the reactions involving liquid phase species,
both equilibrium constants and activity coefficients were needed.
The literature was searched from 1960 using Chemical Abstracts
-79-
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512.454.9535
to obtain equilibrium constant and activity coefficient data.
A compilation (SI-001) covered the open literature up to 1960.
The original articles were studied; some had to be translated.
In some cases the data had to be recalculated because activity
coefficients were not used to calculate the equilibrium constant
or the units were not on a consistent basis with other units
used by Radian. A detailed description of the references and
data considered for each constant is given in Technical Note
200-007-11, entitled "Selected Values for Equilibrium Constants
Used in the Aqueous Equilibrium Formulation". The note is .
included in Volume II of this report. The selected and recal-
culated constants are shown in Table 8-1.
-80-
-------
TABLE 8-1
Selected Values for Equilibrium Constants
Reaction
Form of
Constant
Selected Values
Reference
+ -
HNO 3 (p) ~ H + NO 3
aH+aNO-
K = 3
a
HNO 3 ( .e,)
K250C = 26.9
log K = 6.557
HE-001
320.88 - .01359T
T
HNO 3 (g) ~ HNO 3(.e,)
a
K = HN03 ( 1,)
P
HN03(g)
4. - 1
K250C = 5.80 x 10 atm
DA - 012
I
00
t--'
I
HN02(t) ~ H+ + NO;
~+aNO-
K = ~
a
HNOa( .e,)
_4
K250C = 7.24 x 10
log K = 34.558 -
- 60.571
3
5.8554 x 10
T
_3
X 10 T
LU-005
W-007
HN02(g) ~ HNOa(.e,)
a
K = HNO a (.e,)
. P
HNO 2 (g)
K25°C
=
:a
2.155 x 10
_1
atm
AB- 006
NO(g) ~ NO(.e,)
K = aNO ( .e,)
PNO (g)
_3
K200C = 2.10 x 10
_1
atm
WI-029
-------
Radian Corporation
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHON(512', 454.9535
9.0
SUMMARY OF RESULTS
This study was undertaken to develop a theoretical
description of alkaline scrubbing processes for removing NO
and NO a from electric utility plant flue gases. The first task
was to define the problem using what had already been published
in the literature. The problem was defined for the SO~-free
system. It was found that the nitrogen oxides - water system
is very complicated due to the many stable oxidation states of
nitrogen and the variety of gaseous and aqueous molecules and
ions that exist in equilibrium. It was further found that the
application of aqueous scrubbing to treatment of gases having
the composition of power plant flue gases had not been studied,
previously.
The molar composition of a gas mixture containing
NO, NOa, and HaO affects the concentration of each individual
nitrogen-oxygen species quite markedly due to the large number
of possible equilibrium reactions. Therefore a composition-
concentration model was developed which calculates the gas
phase concentrations of each of the species present at equilibrium
under specified conditions of temperature, total pressure, and
chemical composition (total moles of NO, NO~, and H20). Calcu-
lations done using this model were used to explain why Na04 is
the molecule involved in the mass transfer mechanism and in the
rate limiting step for processes such as nitric acid manufacture
where N02 is present in larger quantities than NO. The calcu-
lations showed further that the same mechanism and rate limiting
step proposed for a particular sorption process such as nitric
acid manufacture or sorption from nitric acid plant tail gases
cannot be applied to aqueous processes for treating flue gas.
-82-
-------
Radian Corporation
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-,535
A new mechanism was proposed for this previously
undefined problem. From calculations done using the equilibrium
concentration model it was proposed that vapor film limited
diffusion of HN02 and HN03 is the mechanism for NOx sorption
from flue gases. Liquid film limited sorption of NOR may also
be important in the mechanism.
A limited amount of sorption data were analyzed
according to this mechanism. The data were collected in
connection with an EPA in-house experimental program for
studying alkaline sorption of NOx' Data analysis using the
proposed mechanism gave results which agreed with experimental
findings.
Thermodynamic properties of gaseous nitrogen-oxygen
compounds, aqueous sorption products and solid metal nitrates
and nitrites were compiled and unpublished values for the
solids were estimated. The thermodynamic data were used to
screen metal oxide sorbents for wet or dry NOx removal proces~es
based on formation of a solid metal nitrate or nitrite and
regeneration of the metal oxide sorbent. The results of the
thermodynamic screening apply to any regenerative process in-
volving formation of solid metal nitrates or nitrites.
-83-
-------
Radian Corporation
10.0
AN-001
t.:
AS-004
AT-005
BA-003
BE-023
BR-002
BU - 009
CA-014
BSOO SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 7B7S7 . .:TELEPHONE 512 - 454.9535
BIBLIOGRAPHY
Andrew, S. P. S. and D. Hanson, "The Dynamics of
Ni trous Gas Absorption", Chern. Eng. Sci. 14, 105-13
(1961).
Ashmore, P. G. and B.J. Tyler, "The Formation and
Thermodynamic Properties of Nitrous Acid Vapour",
J. Chern. Soc., London 1961, 1017-21 (1961).
Atroshchenko, V. I., "Absorption of Nitrogen Oxides
by Alkaline Solutions and the Influence of the State
of the Gas and Liquid Phases'., Ukrain. Khem. Zhur. 12
442-59 (1937).
Bartok, W., et a1., Systems Study of Nitrogen Oxide
Control Methods for Stationary Sources, prepared for
NAPCA under Contract No. PH 22-68-55, Esso Research
and Engineering Co., 1969.
Beattie, I. R. and S. W. Bell, "Dinitrogen Trioxide.
Part I. Stability in the Gaseous Phase", J. Chern.
Soc., London 1957, 1681-6 (1957).
Brown, G. G., et a1., Unit Operations, John Wiley and
Sons, New York, 1950.
Bunton, C. A. and G. Stedman, "Absorption Spectra
of Nitrous Acid in Aqueous Perchloric Acid", J. Chern.
Soc., London 1958, 2440-4 (1958).
Caudle, P. G. and K. G. Denbigh, "Kinetics of the
Absorption of Nitrogen Peroxide into Water and Aqueous
Solutionsll, Trans. Far. Soc. 49, 39-52 (1953).
-84-
-------
Radian Corporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS ?8766' TELEPHONE 512 - 454-9535
CA-069
CH-032
CO-048
EL-004
GA-008
GA-009
GA-010
Carrington, Tucker and Norman Davidson, "Photoelectric
Observation of the Rate of Dissociation of Nitrogen
Tetroxide by a Shock Wave", J. Chern. Phys. 12, 1313
(1951) .
Chi 1 ton, T. H., "The Chemis try 0 f Ni trogen Oxides'l,
Strong Water, Chapter 2, ppo 25-59, MIT Press,
Cambridge, Mass., (1968).
Coughlin, Jamp-s P., Contributions to the Data on
----
Theoretical Metallurgy. Part XII: Heats and Free
Energies of Formation of Inorganic Oxides, U.S.
Bureau of Mines Bulletin 542, (1954).
E1enkov, D. and Iv. Krustev, "Kinetics and Mechanism
of the Absorption of Nitrogen Oxides in Alkaline
Solutions", Izv. Otd. Khim. Nauki, Bu1g. Akad. Nauk
1(4), 771-80 (1969).
Ganz, S. N., et a1., "Absorption of Ni trogen Oxides
wi th Milk 0 f Lime in Mechanical Absorbers", Izv.
Vysshikh Uchebn. Zavedeni~, Khim. i Khim. Tekhnol.
2(1), 155-9 (1962).
Ganz, S. N., IIKinetics of Nitric Acid Formation in
Rapidly Revolving Mechanical Absorbers", Zhur. Prikl.
Khim. 28, 1037-48 (1955).
Ganz, S. N. and S. B.
of Nitrogen Oxides by
Absorbers with Higher
28, 133-41 (1955).
Kravchinskaya, "R..i te of Absorption
Ca(OH)~ Solutions in Mechanical
R.P.M.", J. ~. Chern. (USSR)
-85-
-------
Radian" Corporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN. TEXAS 18166. TELEPHONE 512 . 454-9535
GA-Ol3
GA - 014
GA-046
GO-007
GR-004
HO-009
KE-009
Ganz, S., N., "Ef£8c t of Hydrodynamic Condi tion on the
, '
Rate of Absorption ,of Nitrogen Oxides by Solutions of
Calcium Hydroxide in a Mechanical Absorber 011 a Semi-
plant Scale. 1.", J. AP.El.. Chern. (USSR) 30, 1383-91
(1957) .
Ganz, S. N. and M. A. Lokshin, "Influence of Fundamental
Physicochemical Factors on the Absorption Rate of
Nitrogen Oxides in Ca(OH)~ Solution in High-Speed
Hechanical Absorbers", Zh. Prikl. Khim. 30(10), 1525-
35 (1957); J. ~. Chern. USSR 30, 1592-1600 (1957).
Garcia', L. H., "Absorption Studies of Equimolar
Concentrations of NO and NOQ in Alkaline Solutions",
Presented at the Second International Lime/Limestone
Wet Scrubbing Symposium, New Orleans, Louisiana,
'November 8-12, 1971.
Gorfunkel, V. E. and Ya. I,~ Killman, "Preparation of
Solutions of Calcium Nitrite-Nitrate by Absorption of
Nitrogen Oxides from Nitrose Gas Tailing with Milk of
Lime. 11.", Re~erat. Zhur., Khim. 1958, Abstr. No.
1875, Chern. Abstracts 53: l2599g.
Gray, P. and A. D. Yoffee, "The Reactivity and Structure
of Nitrogen Dioxide", Chern. Rev. 55, 1069-1154 (1955).
Hofmeister, H. K. and R. Koh1haas,
NO-NO~ Mixtures by a Laminar Water
Physik. Chern. 69(3), 232-8 (1965).
ItAbsorption of
Jet", Ber. Bunsenges.
Kelley, K. K., Contributions to the Data ~ Theoretical
Metallurgy. Part XIII: High-Temperature Heat Content,
Heat-Capacity, and Entropy Data for the Elements and
Inorganic Compounds~ U.S. Bureau of Mines Bulletin
584, (1960).
-86-
-------
Radian Corporation 8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS ;8766. TELEPHONE 512.454.9535
KE-010
KE-011
. KE-012
KE-013
KE-014
KO-024
KO-026
Kelley, K. K., Contributions to the Data .2!! Theoretical
Metallurgy. Part X: High-Temperature Heat-Content,
Heat Capacity, and Entropy Data for Inorganic Compounds,
U.S. Bureau of Mines Bulletin 476, (1949).
Kelley, K. K., "Contributions to the Data on Theoretical
----
Metallurgy. Part VIII: The Thermodynamic Prqperties
of Metal Carbides and Nitrides, U.S. Bureau of Mines
Bulletin 407, (1937).
Kelley, K. K. and E. G. King, Contributions to the Data
~ Theoretical Metallurgy. Part XIV: Entropies of the
Elements and Inorganic Compounds, U. S. Bureau of Mines
Bulletin 592, (1961).
Kelley, K. K., Contributions to the Data on Theoretical
Metallurgy. Part V: Heats of Fusion of Inorganic
Substances, U.S. Bureau of Mines Bulletin 393, (1936).
Kelley, K. K., Contributions to the ~ on Theoretical
Metallurgy. Part III: The Free Energies of Vaporiza-
tion and Vapor Pressures of Inorganic Sut'stances, U.S.
Bureau of Mines Bulletin 383, (1935).
Koval, "E. J. and M. S. Peters, "How Does Ni tric Oxide
Affect Reactions of Aqueous Nitrogen Dioxide?", Ind. Eng.
Chern. 21, 1011-14 (1960).
Koval, E. J., Influence of Nitric Oxide ~ Aqueous
Nitrogen Dioxide Reactions, Dissertation, Univ. of
Illinois (Urbana), (1958).
-87-
-------
Radian Corporation 8SOO SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766 .
KR-006
KR-007
KR-008
KU-003
LA-008
LO-007
MI-004
MI-005
. .
TELEPH~NE 512.4$:4-9535
Kramers, H., M. P. P.
o~ Nitrogen'Tet~oxide
J5, 115- 23( 1961) .
Blind, and E. Snoeck, "Absorption
by Water Jets", Chern.' Eng. Sci.
Krus tev, Iv., "Absorption of Nitrogen Oxides by
Solutions Containing Sodium Carbonate, Sodium Nitrite.
and Sodium Nitrate", Izv. Otd. Khim. Nauki, Bulg.
Akad. Nauk. 1(1), 109-23 (1968).
Krustev, 'Iv.., '~bsorption of Nitrogen Oxides by
Sodium Carbonate Solutions under Industrial Conditions",
Khim. Ind. (Sofia) 1968(4),147-51 (1968).
Kubaschewski, 0., et al., Metallurgical Thermochemistry,
4th Edition, Pergamon Press, 1967.
Landolt-BBrnstein, "Kalorische ZustandsgrHssen", 1(4),
Springer-Verlag, Berlin, 1961.
Lowell, P. s., et al., A Theoretical Description of the
Limestone In;ection Wet Scrubbing Process, Vol. 1,
Final report fof NAPCA Contract No. CPA-22-69-l38,
Austin, Texas, Radian Corp., 1970, PB 193-029.
Mill en, D. J.,
"The Ion N!:IO~:
Nature 165, 565
J. D. S. Goulden, and Co K. Ingold,
Binding by One-Electron Transfer",
(1950).
Mirev, D., et al., "Absorption of Nitrogen Oxides by
Vibrating Layers of NaOH Solutions", Compt. Rend.
Acad. Bulgare Sci. 14, 259-62 (1961)0
-88-
-------
Radian Corporation
MO-008
MU-004
PA-016
PE-010
PO-015
RO-007
SI-OOl
8SOO SHOAL CREEK BLVD. . P. O. BOX 994B . AUSTIN, TEXAS 78757 . TELEPHONE 512 - 454-9535
Moll, Albert James, The Rate of Hydrolysis of
Nitrogen Tetroxide, Dissertation, Univ. of Washington,
1966.
Mucskai, L., "Absorption Kinetics of Nitrogen Oxides.
II.", ~. Kern. Folyoirat 72(8), 354-62 (1966).
Parsons, Terry, Gary D. Schroeder, and David DeBerry,
Applicability of Metal Oxides to the Development of
New Processes for Removing SO~ from Flue Gases, 2
vols., Final Report for NAPCA Contract No. PH 86-68-68,
Austin, Texas, Tracor, Inc., 1969, PB 185 562.
P ere l' man, S. and L. Kan torovich, "The Al ka 1 ine
Absorption of Nitrogen Oxides", J. Chern. Ind. (USSR)
11(6),3-9 (1940).
Pozin, M. E., et al., liThe Absorption of Nitrogen
Oxides in Soda Solutions in the Penn Apparatus for
the Production of Sodium Nitrate", Trudy Leningrad.
Teknol. Inst. im. Lensoveta 36, 120-32 (1956).
Rossini, F. D., et al., Selected Values of Chemical
Thermodynamic Properties, National Bureau of
Standards Circular 500, 1952.
Sill~n, L. G., Stability Constants of Metal-Ion
Complexes, Section!: Inorganic Ligands, Special
Publication No. 17, London: The Chemical Society,
Burlington House, 1964.
-89-
-------
Radian Corporation
ST-006
ST-012
ST-017
ST-018
ST-019
ST-026
VA-006
VA-009
8500 SHOAL CREEK 8LVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78757 . TELEPHONE 512- 454.9535
Stull, D. R., et al., JANAF Thermochemical Tables,
Midland, Michigan, 'Dow Chemical Co., PB 168 370, 1965.
Streight, H. R. L., IIReduction of Oxides of Nitrogen
in Vent Gases", Can. J. Chern. Eng. 36, 3-11 (1958).
Stull, D. R., et a1., JANAF Thermochemical Tables -
First Addendum, Midland, Michigan, Dow Chemical Co.,
PB 168 370-1, August, 1966.
Stull, D. R., et a1., JANAF Thermochemical Tables -
Second Addendum, Midland, Michigan, Dow Chemical Co.,
PB 168 370-2, August, 1967.
'Stull, D. R., et a1., JANAF Thermochemical Tables -
Third Addendum, Midland, Michigan, Dow Chemical Co.,
PB 168 370-3, August, 1968.
Stern, KurtH., High Temperature Properties and
Decomposition of Inorganic Salts, Part 1, Nitrates
and Nitrites, Preprint, Available from NSRDS-NBS
around May, 1971.
Varlamov, M. L. and O. M. Drobysheva, "Absorption of
Nitrogen Oxides in Low Concentrations by Sodium
Carbonate Solutions in an Apparatus of the Venturi-
v
Tube Type", Izvest. Vysshikh. Ucheb. Zavedenii, Khim.
i Khim. Tekhno1. 1(1), 146-50 (1960).
Var1amov, M. L. and O. M. Drobysheva, 'Mass Transfer
and Chemical Reaction in a Venturi Scrubber", Zhur.
Prik1ad. Khim~ 33, 2020-9 (1960); J. ~. Chern.
(USSR) 33, 1997-2004 (1960).
-90-
-------
Radian Corporation
VO-007
WA-OOl
WA-006
WA-014
WA-015
WA-016
WA-018
8500 SHOAL CREEK BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 7B7S7 . TELEPHONE 512 - 454.9535
Vosper, Alan J., "Dissociation of Dinitrogen Tetroxide
in the Gas Phase", J. Chern. Soc. A. 1970, 625-27 (1970).
Wagman, D. D., et al., "Selected Values of Chemical
Thermodynamic Properties. Tables for Elements 35
through 53 in the Standard Order of Arrangement",
NBS Techn~cal Note 270-4, Washington, D. C., GPO,
May, 1969.
Waldorf, D. M. and A. L. Babb, "Absorption Spectrum
at 357 mu in the Nitrogen Oxides-Water System",
J. Chern. Physics 40(6), 1476-7 (1964).
Wayne, L. G. and D. M. Yos t, "Kinetics
Gas Phase Reaction between NO, NOg and
Physics 19(1),41-7 (1951).
of the Rapid
H~O", J. Chern.
Waldorf, D. M., Reac tions and
Nitrogen Oxides-Water System,
of Washington, 1962.
Equilibria in the
Dissertation, Univ.
Wayne, L. G. and D. M. Yost, "Rate of the Rapid Gas
Phase Reaction between NO, NOg, and HgO", J. Chern.
Physics 18,767-8 (1950).
Wagman, D. D. et al., "Selected Values of Chemical
Thermodynamic Properties. Tables for the First
Thirty-Four Elements in the Standard Order of
Arrangement", NBS Technical Note 270-3, Washington,
D. C., GPO, January, 1968.
-.91-
-------
Radian Corporation 8SOO SHOAL CREEk BLVD. . P. O. BOX 9948 . AUSTIN, TEXAS 78766 '. TELEPHONE 512.454-9535
ZH-OOl
Zhavoronkov, N. M., et al., "The Study of Absorption
of Nitroge~ Oxides by Alakaline Solutions in Columns
with Filler Bodies", Khim. Prom. 1954, 419-23 (1954).
ZH-002
Zhavoronkov, N. M. and Yu. M. Mar tynov , "The Kine tic s
of the Absorption of Nitrogen Oxides by Water and
Aqueous Solutions of Nitric Acid", Khim. Prom. 1959,
150-5 (1959).
-92-
------- |