TRACOR Document
Number 69-579-U
Project Serial
Number 004-009
Task Number 10
APPLICABILITY OF METAL OXIDES TO THE
DEVELOPMENT OF NEW PROCESSES FOR
REMOVING S02 FROM FLUE GASES
FINAL REPORT
VOLUME I
Sections 1-7
Contract PH 86-68-68
Submitted to:
_-••
Process Control Engineering Program
National Air Pollution Control
Administration
5710 Wooster Pike
Cincinnati, Ohio 45227
31 July 1969
6500 Tracer Lane, Austin, Texas 78721, AC 512/926-2800
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
TRACOR Document
Number 69-579-U
Project Serial
Number 004-009
Task Number 10
APPLICABILITY OF METAL OXIDES TO THE
DEVELOPMENT OF NEW PROCESSES FOR
REMOVING S02 FROM FLUE GASES
FINAL REPORT
Contract PH 86-68-68
Submitted to:
Process Control Engineering Program
National Air Pollution Control
Administration
5710 Wooster Pike
Cincinnati, Ohio 45227
31 July 1969
Approved by:
,:/[)~,~
A. D. Thomas, Jr., Director
Chemistry and Materials
ReJ!rch D;~;p .
&d~q~d
Chemistry Section of the
Chemistry and Materials
Research Department
Prepared by::
~~
(Mrs.) Terry Parsons
Engineer/~i~ti~t
~,D~
~~~~D. Schroeder
EnRineer/Scientist
IY~ LJo~ J
David DeBerr; ?f
Engineer/Scientist
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
ACKNOWLEDGEMENTS
Contract PH 86-68-68, under which this study was performed,
was supervised by Mr. Leon Stankus of the Process Control Engi-
neering Program of the National Air Pollution Control Administration.
His suggestions and advice have been of great value.
-------
Ui/iJ;jJ;jJ;j/ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
CONTRIBUTING TRACOR STAFF
During the past 19 months, the Research Laboratory's
staff at TRACOR has had the opportunity of investigating a method
for removing S02 from flue gas by the use of dry metal oxides
sponsored under the Process Control Engineering Program of the
National Air Pollution Control Administration. I would like to
take this opportunity to acknowledge the contributions of those
involved in this program and to compliment them on their technical
achievements.
Dr. P. S. Lowell - Planned the program and directed all phases
from 1 January 1968 until 1 March 1969, with particular
emphasis on thermodynamic calculations and analysis.
Mr. D. L.
Davis - Directed the program from 1 March 1969 to
31 July 1969 with particular emphasis on the experi-
mental phase of the program and compilation of the
final report.
Dr. Klaus
Schwitzgebel - Planned the experimental program
performed thermodynamic and kinetic studies and
analysis.
and
X-ray
Mr. Gary Schroeder - Performed the engineering calculations for
the economic studies.
Mrs. Terry Parsons - Conducted thermodynamic studies, BET and
X-ray diffraction experimental work, and presented
the results of this study at the Contractors' Meeting.
Mr. David
DeBerry - Designed experimental apparatus, performed
TGA measurements and kinetic studies, and analyzed
the kinetic data.
-------
"-
I
~ 6500 TRACOR LANE. AUSTIN, TEXAS 78721
Appreciation is also extended to Dr. William Koehler for
help in experimental set up and data analysis; Mr. Floyd Felfe
for building the BET apparatus; James Johnson, Erwin Kouba, and
John Nolley for assistance in the experimental work, our secre-
taries, Betty Danner, Drusilla Johnson, Ann Slack, and Susan Swan;
and Betsy Cates and Hal Blair for programming the computer.
We are
Dr. Karl Sladek,
Desmond Bond for
also indebted to our consultants on the program,
Dr. Ray Hurd, Dr. Joseph Lagowski, and Mr.
their vital contributions.
aLJL,b.
A. D. Thomas, Jr.
Director of the
Research Laboratory
-------
~ 6500 TRA~OR
L.ANE, AUSTIN, TEXAS 78721
ABSTRACT
This report presents the results of a study to determine
the applicability of metal oxides to the development of new
processes for S02 removal. The oxides of 48 metals were screened
according to the thermodynamics of their reaction with sulfur
oxides to eliminate from consideration as potential sorbents
those oxides that were not capable of reducing the sulfur oxide
concentration in exiting flue gas of power plants to 150 ppm.
Thermodynamic studies resulted in an extensive compilation of
experimentally determined and estimated thermodynamic data. The
result of the thermodynamic screening process was reduction of
the field of potential sorbents to the oxides of sixteen metals,
most of which appear in Groups VI, VII, and VIII of the periodic
table. These potential sorbents are not only thermodynamically
capable of reducing the S02 concentration to the desired limit,
but the thermodynamics of the regeneration step are also favorable.
The remaining potential sorbents were prepared in a
kinetically active form and the rates of their reaction with 802
in a simulated flue gas atmosphere were determined using isothermal
gravimetric methods. There were two results of the kinetic studies.
First, the field of potential sorbents was reduced to the oxides
of copper, chromium, cobalt, iron, nickel, and cerium. Second,
the rate data obtained provided inputs to cost estimation calcu-
lations for the economic feasibility studies.
The economic feasibility studies consisted of equipment
design and sizing and estimation of the capital investment and
gross annual operating cost for a sorber-regenerator system using
a fluidized bed model. The cost estimations were carried out for
the three best potential sorbents. Process flowsheets including
heat and material balances were also prepared. The result of the
economic feasibility studies for copper and iron oxides is a
preliminary estimation for a capital investment of $8 million and
a gross annual operating cost of $3 million.
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Section
1.
2.
TABLE OF CONTENTS
INTRODUCTION
..........
........
THERMODYNAMIC STUDIES
2.1
2.2
. . . .
. . . .
. . .
. . .
Introduction to Thermodynamic Studies
Literature Survey.
. . . . . .
. . . . .
. . . .
. . .
. . .
Compounds of Interest
Information of Interest
Availability of Information
2.3
Compilation of Thermodynamic Properties.
. . . .
2.3.1
2.3.2
2.3.3
2.3.4
Properties Tabulated. . . . . . . .
Calculations Done with Thermodynamic
Consistency of Data Base. . . . . .
Sources of Thermodynamic Data. . .
. . .
Da ta.
. . .
. . .
2.3.4.1
2.3.4.2
2.3.4.3
Compilations. . . . . . . . . .
Open Literature. . . . . . . . .
Estimation Methods. . . . . . .
2.4 Thermal Stability Studies . . . . . . . . . . . .
2.4.1 Thermal Stability of Metal Oxides . . . .
2.4.2 Thermal Stability of Metal Sulfites . . .
2.4.3 Thermal Stability of Metal Sulfates . . .
2.5
2.6
Catalytic Oxidation Properties of Metal Oxides
Determination of Price and Availability of
Metal Oxides. . . . . . . . . . . . . . .
. . .
2.7
Preliminary Screening Based on Thermodynamics
and Sorbent Availability. . . . . . . . . . . .
2.7.1
2.7.2
2.7.3
Criteria for the Screening Process. . . .
Metal Oxide Screening. . . . . . . . . .
Mixed Metal Oxide Screening. . . . . . .
Page
1
4
4
4
5
5
5
6
6
6
7
7
27
27
31
35
35
36
36
36
38
38
-------
.. 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Section
3.
3.1
3.2
3.3
3.4
3.5
4.
.4.1
4.2
4.3
TABLE OF CONTENTS (Cont'd.)
KINETIC STUDIES
. . .
. . . .
. . . .
. . . .
Introduction to Kinetic Studies
Approach
. . .
. . . . .
. . .
. . . .
. . . . .
. . .
. . . .
Compound Preparation and Characterization
3.3.1
3.3.2
3.3.3
3.3.4
3.3.5
. . .
Preparation Methods. . . . . . . . . .
Use of X-ray Diffraction. . . . . . . .
BET Surface Area Determinations. . . .
Differential Thermal Analysis. . . . .
Chemical Analysis. . . . . . . . . . .
Kinetic Measurements
3.4.1
3.4.2
3.4.3
3.4.4
. . . .
. . . .
. . . . .
Apparatus. . . . . . . . . . . . . . .
Experimental Procedure. . . . . . . . .
Data Analysis. . . . . . . . . . . . .
Input to Economic Studies. . . . . . .
Results of the Experimental Program
3.5.1
3.5.2
3.5.3
3.5.4
. . . . . .
Page
49
49
49
50
50
50
52
57
57
57
57
67
68
73
74
Rate Equation and Reaction Orders. .. 77
Temperature Dependence of the Reaction Rate 86
Physical Properties and Reactivity. .. 87
Conclusions. . . . . . . . . . . . .. 87
ECONOMIC FEASIBILITY STUDIES
. . . . . . . . .
Introduction to Economic Feasibility Studies
Sorption Unit Design
4.2.1
4.2.2
4.2.3
.........
. . . .
Sorber Design
Draft Fan and
Sorbent Fines
. . .
. . . . .
. . . . .
Driver Design. . . . . .
Collector (Cyclone) Design
Regeneration Unit Design
4.3.1
. . .
. . . . .
. . .
Regenerator Design
. . . . .
. . . . .
ii
89
89
89
92
95
97
101
101
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Section
4.4
4.5
4.6
5.
6.
7.
7.1
Total Capital Investment Cost
Gross Annual Operating Cost.
4.3.2
4.3.3
Resul ts .
SUMMARY.
TABLE OF CONTENTS (Cont'd.)
Draft Fan and Driver Design. . . . . .
Sorbent Fines Collector (Cyclone) Design
. . . . .
. . .
. . .
. . . . . .
. . . . . . .
. . .
. . . . . . . . .
. . .
. . . .
. . . . . . . .
. . . .
REFERENCES
. .. . .
. . . . . . . . . . .
. . .
ABSTRACTS OF TECHNICAL MEMORANDUMS
. . .
. . .
Technical Memorandums on Thermodynamic Studies
and Preliminary Screening. . . . . . . . . .
7.1.1
7.1.2
7.1.3
7.1.4
Technical Memorandums on Estimation of
Heat of Formation. . . . . . . . . . .
T.M. 004-009-Chl
T.M. 004-009-ChlA
Technical Memorandum on Estimation of
Absolute Entropy. . . . . . . . . . . .
T.M. 004-009-Ch5
Technical Memorandums on Estimation of
Heat Capacity. . . . . . . . . . . . . .
T.M. 004-009-Ch4
T.M. 004-009-Ch9
T.M. 004-009-Ch13
T.M. 004-009-Ch13A
Page
102
102
102
108
108
113
115
118
118
118
118
118
Technical Memorandums on the Effect of Errors
in Estimated Thermodynamic Properties 119
T.M. 004-009-Ch2
T.M. 004-009-Ch4
iii
-------
~ 6500 TRACOR LANE. AUSTIN, TEXAS 78721
Section
7.2
7.1.5
7.1.6
7.1.7
7.1.8
7.1.9
TABLE OF CONTENTS (Conti d.)
Technical Memorandum on Conflicting
Reported Thermodynamic Data. . . . . .
T.M. 004-009-Ch18
Technical Memorandums on Thermal
Stability Studies. . . . . . .
. . . .
T.M. 004-009-Ch7
T.M. 004-009-CH8
T.M. 004-009-Ch3
T.M. 004-009-Ch16
Technical Memorandum on the Price and
Availability of Metal Oxides. . . . . .
T.M. 004-009-Ch6
Technical Memorandums on Preliminary
Screening of Metal Oxides and Mixed
Me tal Oxide s . . . . . . . . . . . . . .
T.M. 004-009-Ch16
T.M. 004-009-Ch22
T.M. 004-009-Ch25
Technical Memorandum on Computer Programs
T.M. 004-009-Ch23
Technical Memorandums on the Kinetic Studies
Experimental Program. . . . . . . . . . . . .
7.2.1
Technical Memorandums on Design and
Operation of Experimental Equipment
T.M. 004-009-Ch10
T.M. 004-009-Ch20
T.M. 004-009-Ch24
iv
Page
119
119
120
120
121
121
121
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Section
7.3
8.
8.1
8.2
7.2.2
TABLE OF CONTENTS (Contld.)
Technical Memorandum on Collection of
Experimental Data and Presentation of
Re suI ts ...............
T.M. 004-009-Ch24
Bimonthly Progress Report No.8
Technical Memorandums on Economic
Feasibility Studies. . . . . . .
7.3.1
7.3.2
. . . . . .
Technical Memorandums on Equipment
Design, Size and Purchase Price. . . .
T.M. 004-009-Ch19
T.M. 004-009-Ch14
T.M. 004-009-Ch15
T.M. 004-009-Ch15A
T .M. 004-009-Ch 21
Technical Memorandums on the Estimation
of Capital Investment and Operating Cost
T.M. 004-009-Ch26
T.M. 004-009-Ch11
UNABRIDGED TECHNICAL MEMORANDUMS AND
ASSOCIATED DATA
Unabridged Technical Memorandums
8.1.1
8.1. 2
8.1.3
Technical Memorandums on Thermodynamic
Studies
Technical Memorandums on Kinetic Studies
Technical Memorandums on Economic
Feasibility Studies
Thermodynamic, Kinetic, and Economic Data
8.2.1
Data Base of Thermodynamic Properties
v
Page
122
122
122
123
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
See tion
8.3
8.2.2
8.2.3
8.2.4
8.2.5
TABLE OF CONTENTS (Cont'd.)
Log K Plots for Sulfate Decomposition
Reactions
Plots of treIntegrated Form of the
Rate Equation
Economic Calculations for the Best
Potential Sorbents
Process Flowsheets Including Heat
and Material Balances
References Used in Writing Technical
Memorandums.
vi
Page
-------
L-
I
Il:i11Jii:iJ;i 6500 TRACOR LANE. AUSTIN. TEXAS 78721
LIST OF ILLUSTRATIONS
Figure 1
Periodic Arrangement of Potential
Sorbents before Screening. . . . . . . . . . 2
Decomposition of Lead Oxides. . . . . . . . 32
Logarithm of the Equilibrium Constant
for the Disproportionation of Zinc
Sulfite. . . . . . . . . . . . . . . . . . . 33
Figure 2
Figure 3
Figure 6
Logarithm of the Equilibrium Constant
for the Decomposition of Zinc Sulfite. . . . 34
Logarithm of the Equilibrium Constant
for the Decomposition Reactions for Tin,
Copper, and Barium Sulfates. . . . . . . . . 39
Periodic Arrangement of Potential
Sorbents after Thermodynamic Screening. . . 47
X-Ray Diffraction Patterns for
Different Preparations of COS04 . . . . . . . 51
Adsorption Isotherm for Zirconium Oxide
at 770 K '. . . . . . . . . . . . . . . . . . . 53
Figure 4
Figure 5
Figure 7
Figure 8
Figure 10
Figure 11
Schematic Diagram of Apparatus for
Determining Specific Surface Area Using
the BET Method .0 . . . . 0 . 0 . 0.' 0 . . 56
TRACOR TGA-3C Schematic Diagram. . . . . . . 65
Schematic Diagram of Gas Mixing
Apparatus. . . . . . . . . . . . . . . . . . 66
Representation of Rate Data for
Copper Oxide at 404°C. . . . . . . . . . . . 72
Rate Data for Copper Oxide. . . . . . . . . 80
Rate Data for Iron Oxide. . . . . . . 81
Rate Data for Nickel Oxide. . . . . . . . . 82
Rate Data for Cobalt Oxide. . . . . . . . . 83
Rate Data for Cerium Oxide. . . . . 0 . . . 84
Rate Data for Chromium Oxide. . . . 0 . 0 . 85
Arrhenius Plot for CuO . . . . . . . . . . . 88
Potential Sorbents After Kinetic Screening. . 90
Economic Study Concept. . . . . . . . . . . 91
Cyclone Separator Proportions. . . . 0 . . . 99
Process Flowsheet for the Copper Oxide
Process. . . . . . . . . . . . . . . . . . 112
Figure 9
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
Figure 17
Figure 18
Figure 19
Figure 20
Figure 21
Figure 22
Figure 23
vii
-------
j;Ii1/Jli:i1;jj 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Table I
Table II
Table III
Table IV
Table V
Table VI
Table VII
Table VIII
Table IX
Table X
Table XI
Table XII
Table XIII
Table XIV
Table XV
Table
Table
Table
XVI
XVII
XVIII
LI ST OF TABLES
Heat of Formation at 25°C. . . . . . . . . .
Deviations of Estimated Values of
Absolute Entropy. . . . . . . . . . . . . . 20
Estimated Heat Capacities for Carbonates
with Known Heat Capacities. . . . . . . . . 28
Estimated Heat Capacities for Sulfates
with Known Heat Capacities. . . . . . . .. 29
Estimated Heat Capacities for Sulfides
with Known Heat Capacities. . . . . . . . .
Properties of Oxides, Sulfates, and Su1fites.
Determination of Specific Surface Area
by the BET Method for Zr02 .........
Results of Chemical Analyses. . . . . . . .
Results of Experimental Program. . . . . . .
Copper Oxide Sorber Physical Dimensions. . .
Copper Oxide Sorber Input Variables for
Fan Co st. . . . . . . . . . . . . . . . . . 98
Copper Oxide Sorber Input Parameters for
Cyclone '. . . . . . . . . . . . . . . . . . . 100
Copper Oxide Regenerator Physical Dimensions. 103
Copper Oxide Regenerator Input Variables
for Fan Cost. . . . . . . . . ... . . . . . 104
Copper Oxide Regenerator Input Parameters
for Cyclone. . . . . . . . . . . . . . . . . 105
Installation Cost Factors. . . . . . . . . . 107
Components of Gross Annual Operating Cost. . 109
Summary of Cost Estimation for the Copper
Oxide and Iron Oxide Processes. . . . . . . 111
9
30
40
58
59
75
96
viii
-------
~ 6500 TRACOR LANE. AUSTIN. TEXAS 78721
1.
INTRODUCTION
Studies by the Department of Health, Education and Welfare
have shown that the most pressing air pollution problem is the
generation of sulfur oxides during combustion of fossil fuels.
Processes for preventing sulfur oxide emission have been divided by
NAPCA into nine areas. One of these areas is the sorption of sulfur
oxides by dry metal oxides. The work described in this report was
conducted under contract PH 86-68-68 to determine the applicability
and economic feasibility of the use of dry metal oxides in a sulfur
oxide removal process. The three objectives for the study are
given below.
The first objective was to determine which of the oxides
of the metals shown in Figure I were thermodynamically capable of
lowering the S02 concentration in flue gases to a specified level
and had a sorption product that could be regenerated to the original
metal oxide sorbent with only a small expenditure of energy.
Thermodynamic screening was conducted by determining the free
energy of the sorption and regeneration reactions. Knowing the
standard heat of formation, the absolute entropy, and the heat
capacity as a function of temperature for the products and reactants
of interest, the free energy for the sorption or regeneration
reaction can be calculated. These thermodynamic properties of
interest which had been determined experimentally and reported in
the literature were computer tabulated. Properties of interest
that had not been determined experimentally were estimated using
carefully selected methods. The thermodynamic studies are discussed
in Section 2.
The second objective was to determine which of the oxides
remaining after thermodynamic screening had favorable kinetic
properties. A kinetically favorable sorbent is one which reacts
fast enough to be economically feasible. This objective was
achieved by conducting an experimental program, since few kinetic
-------
GROUP
PERIOD vb Vlb Vllb
la VIII 0
I H H He
2 N 0 F Ne
3 S CI Ar
4 Se Br Kr
5 Te I X.
r-.J 6 Au Hg Po At Rn
7
lANTHANIDE Pr Nd Pm Sm Eu cd Tb Dy Ho Er Tm Yb lu
SERIES
ACTINIDE Np Pu Am Cm Bk cf Es Fm Md lw
SERIES
PERIODIC ARRANGEGMENT OF
POTENTIAL SORBENTS BEFORE SCREENING
Uili1iIiJii
J>
()'\
I
FIGURE 1
-
()'\
.l="
I
--
.;;-
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
data for the reaction of metal oxides and sulfur oxides had been
reported in the literature. The program consisted of preparing
the compounds in a kinetically active form, determining their
physical properties, and determining their rate of reaction with
S02 in a simulated flue gas atmosphere. An isothermal gravimetric
technique was used for the collection of kinetic data. The kinetic
studies are described in Section 3.
The third objective was to determine the economic feasi-
bility of the use of the selected sorbents in a sorber-regenerator
system using a fluidized bed model for the gas-solid contactor.
The economic feasibility studies consisted of designing and sizing
equipment, preparing flowsheets and heat and material balances, and
estimating total capital investment and gross annual operating cost
for the process. The capital investment was estimated on the basis
of the purchase price of the major pieces of equipment to which a
percent of purchase price was added to account for various instal-
lation costs. Modified Lang percentage factors, as supplied by
HEW, were used. The economic studies are discussed in Section 4.
A summary of results is given in Section 5; references
are given in Section 6; Section 7 contains abstracts of the techni-
cal memorandums; and, Section 8 contains the unabridged memorandums.
3
-------
~ 6500 TRACOR LANE. AUSTIN. TEXAS 78721
2.
THERMODYNAMIC STUDIES
2.1
INTRODUCTION TO THERMODYNAMIC STUDIES
The work done during the thermodynamic studies phase of
the contract consisted of six main tasks: (1) a literature survey,
(2) compilation of thermodynamic properties, (3) determination
of cost and availability of metal oxides, (4) thermal stability
studies, (5) study of catalytic oxidation properties, and (6)
thermodynamic screening. The first five tasks were carried out
so that the screening process would be possible. The goal of the
thermodynamic screening process was to select as potential sorbents
oxides which met the following two criteria. First the sorbent
had to be thermodynamically capable of lowering the sulfur oxide
concentration to the desired limit. Second, the regeneration step
had to have only a small negative free energy change so that the
expenditure of a great amount of energy would not be necessary to
recover the sorbent. Thermal, rather than chemical regeneration
was considered to be of importance. At the same time, the oxides
were screened according to price and availability.
2.2
LITERATURE SURVEY
The literature was surveyed using Chemical Abstracts as
a guide to determine existing knowledge about metal oxides and
mixed metal oxides; flue gas components such as sulfur oxides,
oxygen, nitrogen oxides, and carbon dioxide; and the reaction
products, metal sulfates, sulfites, sulfides, and carbonates.
The information sought was of five types: (1) thermodynamic
properties, (2) physical properties, (3) compound preparation
methods, (4) reaction kinetics, and (5) descriptive information
such as the course of a reaction, the existence of compounds, or
the stability of compounds. Much valuable information was found
on thermodynamic properties, compound preparation methods, physical
properties, and descriptive subjects. Little, however, was avail-
able on the kinetics of the reactions of metal oxides and flue gas
components.
4
-------
I:lit/J;iIili 6500 TRACOR LANE. AUSTIN. TEXAS 78721
2.3
2.3.1
COMPILATION OF THERMODYNAMIC PROPERTIES
Properties Tabulated
'The second of the six tasks in the thermodynamic studies
phase was to establish a data base of thermodynamic properties
for the compounds described in Section 2.2. The following thermo-
dynamic properties were tabulated: (1) heat of formation and
absolute entropy at 25°C, (2) heat capacity from 25°C to 1000°C,
or the data limit, whichever is lower, and (3) the temperature,
type, and heat of phase transitions. A copy of the data base is
given in Section 8.2 of this report.
2.3.2
Calculations Done with Thermodynamic Data
Using the thermodynamic properties stored in the data
calculations were made with Equations (1) through (4) to des-
the extent of reactions over any temperature range for compounds
base,
cribe
6Go = 6Ho - T6So (1)
T T T
I [T. + ST CP(T)dT]
0 +2
HT = H~ge Sa~e Cp(T)dT + 6HT. (2)
i=O 1. T.
1.
0 T Cp(T)
ST = S~9 e + S dT (3)
as e T
0
In K -6GT
= Irr (4)
for which data were stored. In Equation (2) T. is the temperature
1.
of the ith phase transition and 6HT. is the enthalpy of the ith
1.
transition. These calculations were an important source of thermo-
dynamic information for the screening process which is discussed in
Section 2.7.1. The compilation of thermodynamic properties was
also used in the economic feasibility studies to make enthalpy
5
-------
Ui!iJiiiIii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
calculations for determining heat balances for the sorber-regenerator
system. The methods are discussed further in Section 4 and in
detail in T.M. 004-009-Ch26. The thermodynamic properties and the
data sources were stored in a digital computer for accuracy and
ease of retrieval. The program, AIRPOL, stored, retrieved, and
printed the thermodynamic properties and the references from which
they were taken as well as calculated the change in enthalpy,
entropy and the logarithm of the equilibrium c'onstant for reactions
involving the compounds for which data were stored. For a detailed
description of this program see T.M. 004-009-Ch23.
2.3.3
Consistency of Data Base
A data base.should be internally consistent; that is,
the value of a property should be the same for all computational
paths. It should also be consistent with the measurements from
which the data were derived. When published literature values of
measured heats of reaction were found, they were added to values
for the heats of formation of the reactants in the TRACOR data
base to calculate the heat of formation of the product. Therefore,
the heat of formation reported in some references cited may differ
from the heat of formation given in this compilation by an amount
equal to the difference in the heats of formation of the reactants
used. Again, these differences are due to an attempt to establish
an internally consistent compilation.
2.3.4
Sources of Thermodynamic Data
The sources of thermodynamic properties were compilations
of properties, open literature, and estimation methods.
2.3.4.1 Compilations: The compilations consulted were Landolt
Boernstein, Chemisch-Physikalische Tabellen (LA-OOl), National Bureau
of Standards, Selected Values of Chemical Thermodynamic Properties
(NB-003, NB-005), JANAF Thermochemical Tables (JA-OOl), O. Kuba-
schewski, Metallurgical Thermochemistry (KU-OOl), and the U. S.
6
-------
BaiIii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Bureau of Mines, Contributions to the Data on Theoretical Metal-
lurgy, Parts I through X. Occasionally several major compilations
reported widely differing values for some thermodynamic property
of a compound. In such cases, the original literature from which
the data for the compilation were taken was consulted. In many
cases the conflicts were clearly resolved. When conflicting values
were reported, the reference TR-OlO was given. This refers to
T.M. 004-009-Ch18, which is a detailed discussion of the selected
values and the reasons for their selection.
2.3.4.2 Open Literature: Another source of data for the thermo-
dynamic properties compilation was the open literature which was
surveyed through 1967 using Chemical Abstracts as a guide. Informa-
tion obtained from the open literature was carefully assessed using
the values reported in Chemical Abstracts as well as the original
literature in many cases. The Reports of Investigations published
by the U. S. Bureau of Mines were important sources of thermodynamic
data from the open literature. Since thermodynamic properties for
all the compounds of interest had not been determined experimentally,
estimation techniques had to be employed.
2.3.4.3 Estimation Methods: Thermodynamic properties had to be
estimated for about 40% of the compounds tabulated. The estimation
methods were selected by first studying the methods published by
various authors. Next the best method was chosen on the basis of
accuracy, theory, and applicability. The accuracy was determined
by comparing accepted values with those calculated by each method
and computing a root mean square (RMS) error. All of the unknown
thermodynamic quantities were estimated from accepted, experimentally
determined values of thermodynamic data using a Univac 1108 digital
computer to perform the calculations and retrieve the data previously
stored in the data base.
7
-------
lI8iJ;ii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
There are several ways to estimate the standard heat of
formation of inorganic compounds. It was decided on the basis
of accuracy and applicability that the method of Erdos would be
applied to our problem.
Erdos gave the following equation for estimating the heat
of reaction 6H~. between base i and acid j to form, for example,
~J
a salt, ij.
R
-6H. .
~J
=
B.. (K. - A.)nJ.
~J ~ J
(5)
In Equation (5), B.. is the number of ion pair bonds formed, Ki is
the cation combini~~ power, Aj is the anion combining power and
n. is the anion exponent. The heat of formation of the oxide
J
from a.. moles of cation or base and b.. moles of anion or acid
~J . ~J
is given in Equation (6).
f
6H..
~J
=
a.. 6 H~ + b.. 6 H ~ + 6 H~ .
~J ~ ~J J ~J
(6)
In applying the method of Erdos the existing data were
correlated to determine the values for K, A, and n for each cation
or anion species of interest. Then the unknown heats of formation
were estimated for all possible combinations of the cations and
anions.
The accepted values, estimated values, and correlation
errors for the heat of formation method are given in Table I. The
root mean square (RMS) error for the correlation was 3 kcal/mole.
The selection of the Erdos method as the best for our data compi-
lation is discussed in T.M. 004-009-Chl. The mathematical details
of the application of the method and the results obtained are given
8
-------
TABLE I
HEAT OJ:' FORMATION AT 25 OEG C
(KCAL/GMOLE)
COMPOUND . J J PAIRS or HEAT Of fORMATION
IONIC OONOS CALCULATED ACTUAL ERROR
.
( AGI2( ' (03) 1 1 1 .0 -1.2&15't+OZ -1.2066+0Z 3.6599+,00
( BA) C (031 Z 1 1 .0 -2.9000+02 -2.8716+02 Z.6&;60+00
I I (fd ( COJ) 3 1 1.0 -Z.8568+02 002.0811+02 -2."366+00
'.1 CO) (' 'OJI &f 1 1.0 -1.7681+02 -1.7B&fb+02 -1.6&f5&f+UO
, ( CO) C (03/ S 1 1 .0 -1.7'iZIf+02 -1.7331+02 9.28JS-O'1
( CU+2/ ( COJ /. 6 1 1.0 . , -'1. If 2 2 1 + 0 2 -1.'+210+02 h168&f-Ol
C fL+2) ( C03) 7 1 1.0 -1.7838+02 -1.7855+02 -1.7829-01
C K ) 2 ( CO:\) fJ 1 1 .0 -2.71102+02 -2.7167+02 2.1S38+00
.\..- I LI/2( CO)) 9 1 1.0 -2.A687+02 -2.9026+02 -3.398,++00
( IiG) ( CO)/ 10 1 1 .0 -2.S980+02 -2.639U+02 -,+.1796+00
',( I1N+2) ( COJ) 1 1 1 1 .0 -2.096)+02 -2.137"1+02 -'f.1091+00
C NA)21 CO)) 12 1 1.0 -2.7001+0Z -2.6972+0Z 2.8867-01
\.0
. ( rJi) « (03) 1 3 1 1 . 0 -1.6'110+02 -1.6296+02 6.1260+00
I PO) ( C03) 1 'f 1 1 .0 " -1.7099+02 -1.6723+02 3.7567+00
. I 510 ( C03) 15 1 1.0 -2.f:l673+02 -2.9098+02 -....2510+00
( ZN) ( C03) 16 1 1 .0 -1.9221+02 "1.9368+02 -1....677+00
,( AL+))Z( CO)3 1 7 1 3.0 -7.0703+02
( tH!) c CO)) 10 1 1.0 -2.'f689+02
( Bl+3)2( C03)3 19 1 3.0 -".7657+02
(:.,CE+&;) ( ,'" C 0:\ / 2' 20 1 Z.() -q.6601+0Z
I (S)21 CO)) 7. 1 1 1 .0 ..2.7257+02 -2.7"''''7+02 -1.9008....00
I CU+I)21 COJ) 22 1 1.0 -1.'f229+0Z
( FE+)Z( CO))3 23 1 )~O -5.0266+02
.(. MIJ+)Z( C03)) 21f 1 3.0 80S . 'f 5 q " + 0 2
(-: RB)l(', C03) 25 1 . 1 .0 -2.7111'i+02 -2'09'18+02 1.9589+00
( SO+))2( COJ)J ' 26 1 J.O -'1.6609+0Z
( SN+'t) I C03)2 Z7 1 2.0 -,).)00"1+02
( Hi + &;) ( C03)2 za 1 2.0 -5.1Y62+0Z
.
( U+~) ( CO))Z 29- 1 Z.O -....7~20+0Z
: ( ZR+...) ( CO))2 30 1 2.0 -5.0261.0Z
\ ( AG)2( SOlf) 1 2 1 .0 -1.7062+02 -1.7036+02 1j'.5900-01
\ ( HA) ( SO...) Z 2 1.0 -3,"'958+02 -3,"'999+02 -'I.06~7-01
i ( CA) I so...) 3 2 1.0 -3.3&2'1+02 -3.&f019+02 "1,9570.00
-------
TABLE I (continued)
,( CD) ( 50q) q 2. 1 .0 -2.2237+02 -2.2120+02 1.1760+00
( CO) ( 50'1) !) 2 1 .0 -2.20&f8+0z
'.. ( CU+2) ( 50q) 6 2 1 . 0 -1.AO[,If+Ol -1.8q22+02 -3.5806+00
\ (' FE+2) ( 50q) " 2 1.0 -2.23/i7+02 -2.20" 1 +02 3.0621+00
( K ) 2 ( 50q) 8 2 1 .0 -3."070+02 -3.Q23Q+02 -1.61f51+00
" 50Q) -3,"267+02 -3.Q258+02 9.053/i-02
( LI)2( 9 2 1 .0
( 11G) ( SO&f) J 0 2 1 .0 -3.0'.>/.1+02 "':'3.0550+02 1.1123-01
( MN+2) ( SOli) 1 1 2 1.0 -2.51,11+02 -2.5395+02 2.1601+00
i '( NA)2( SO If) 12 2 1.0 -3.3192+02 -3.306"+02 1.2868+00
( N I) ( 50'+) 13 2 1.0 -2'1287+02 -2.1350+02 -6.3629-01
\. t PH) ( 50'+) 1 '1 ,2 1 .0 -2.1705+02 -2.1933+02 -1.lffH3+00
\ ( SRI ( SO'i) 15 2 j .0 -3."302+02 -3.Q&fY7+02 .1.9566+00
( I.N) ( Salt) 11) 2 1 .0 -2.J'f&fl+02 -2.3369+02 7.150'1-01
I ( AL.+J)2( 50'1)3 1 7 2 3.0 -0.2135+02 -8.2,038+02 9.7059-01
( uE) ( 50'1) 10 ' 2 1 .0- -2.A610+02 -2.8565+02 ...'+51Q-Ol
( Bl+J)2( SOIf») I? 2 3.0 -6.0932+02 -6.0810+02 1.2200+00
\ ( CI::+If) ( SO'.) 2 20 2 2.0 -5.[,083+02 -5.6000+02 8.2569-01
( CS)Z( .,SOIf)' 21 2 1 .0 -3.'1119+02 -3.3900+02 2.1918+00
,( CIJ+I)2( SOQ) 22 2 I .0 -1'7981+02 -1.7951+02 3.0120-01
UFE+3)2( SOIf») 23 2 3.0 -6.1652+02 -6.1560+02 9.21f18-01
t-4,.( 11/'4+ .) ) 2 ( 50'1)3 Z'. 2 3.0 -6.6615+02 -6,6690+02 -7,.5377-01
o( RA)2( : 50'1) 25 2 1.0 -3.3919+02 -,).1f019+02 -1.UU78+UO
, ~ ( Sll+3)2( 50'+)3 26 .2 ,) .0' -5.7!;,'IZ+02, -r..7'+20+02 Ie 2166+00
l ( SN+,+) ( 50'+)2 27, 2 2.0 -3.9308+02 -3.9.11f0+02 -3.2331-01
{( TH+,+) ( 50'+)2 28 2 2.0 -6.0607+02 -6.0726+02 8.0502-01
'
( . U+'t) ( 50&+)2 '29 2 2.0 - 5 . 6 3 '. Q + 0 2 -5.6300+02 'f.lflS3-01
,
( ZH+'t) ( SO'f)2 30 2 2.0 -5-?801+02 -S.9701+02 9,9767-01
c- ( AG)2(FElOQ) 1 J 1 .0 -2.0553+02
( !}A) (FE20Q) 2 3 1 .0 -3.5555+02 -J.SO"9+02 5.0S6"+00
"( (A) (FllOIf) 3 J 1 .0 -3.5'/2'f"'O,l
( CO) (FE20't) .. 3 1 ,0 -2.51373+02
\ CO, (FF.:20't) 5 3 1 .0 -Z.5526+02
(~ C U + 2) 'CFE20't' 6 3 1 .0 - 2 . .3 1 'I 9 ... 0 7. -2.37S1+02 -6.02,)..+00
( fE+2) (F L 2 O'i I 7 ,) 1 .0 .. 2 . 13 0 9 3 ... 0 2'
(/ ~)2(FE20Q) "- 8 '3 1 . 0 -3.,3105+02
\ . ( LI )2(F'E20't) 9 3 1 .0 -3.5671f+02 -J.'f7't0+02 9.3391+00
( 11G) (FElO't" 1 0 .3 1 . 0 .-3."tl'+0+02 -J.1f922+02 -7.8227+00
( MN+2) (F[20") 1 1 J 1 . U -2.9036+02 -2.9300+02 -2,6383+00
( "'A) 2 ( F E 2 0 't ) 12 3 1 .0 -3.3280+02
\ \ hll) (FE2()'t) 13 3 1.0 -2. 531.?+OZ "'2.~61f6+02 -2.7692+00
( PB) (,£20..) 1 Ii 3 1.0 -2,5122+02
-------
TABLE I (continued)
( SR) (F f. 2 O'f I 15 .3 1 .0 -3.5606+02
I ( ZN) (FEZO'f1 1 b 3 1 .0 -2.7895+02
( AL+3IZ(FEZO'+),j . 1 7 ,j 3.0 '-9.7513+0Z
( BE) (F£:ZO'f) 1 a .3 1.0 -.1.,3585+0Z
\ ( Bl+3)Z(FE20'f)3 19 .3 3.0 -7,.2792+0Z
( eE+'+1 (FE20'f)2 20 J. 2.0 ;'6..2965+0Z
( (5)Z(1-"1::20,+) 2 1 3 1 . 0 -3.2716+0Z
'.. ( eU+l)2(FE20'f) 22 3 1 . 0 -2.3180+02
( fE+3)l(FE20'f)3 23 3 3.0 -7.7090+02
. ( MN+)2(FE20'f13 2'f 3 3.0 -8.1067+02
( NB)2(Fl20&f) 25 3 1 .0 -3.7.713+0Z \
. ( S8+3):l(FE20'f)3 26 3 3.0 ..7.3707+02
J SN+'f) .(F£20'f'2 27 3 2.0 -5.0728+0l
( TH+'t). (,,'E20'*)2 26 3 2.0 -6.8728+02
( U+&f) (fE20it)2 29 3 2.0 -6.5072+02
. ( ZR+'t) (FEZO't)2 30 3 2.0 -6.610'++02
.,'\.
( \.AG)2( (RO'f). 1 " 1 . 0 -1.7180+0Z ""&.7237+02 . -5.7&95-0!
( GA) ( (RO'f) Z If 1 .0 -3.3637+02
I--' ( (A) I (~o'+) 3 " 1.0 -3.3257+0Z .. 3 ,2 9 .. .. + 0 2. 3.!271+00
t-' ( (D ).~. CNO'f) 't '+ 1 .0 -2 . 2 '1 1 0 + 0 2
( (0) ( eROlf). 5 'I 1 .0 -2.21'+9+0Z
( CU+2) ( CRO't) 6 't 1 .0 -1.A980+0Z
( FE+2) ( (ROlf) 7 '+ 1 .0 '-2.;:S69+02
( K ) 2 ( CRO,+) 8 " 1 . 0 -3.\972+02
(. LI)2( CRO'f) 9 'f 1 .0 -3.3353+02
( 11 G) ( (ROlf) 10 '+ 1 . a ..3.0708+02
\( MN+2) ( CRO,+) 1 1 'f 1 .0 -2.5686+02
( r~ A ) 2 ( cr~o'f) 12 " 1 . 0 "3.1617+02 -3.1860+02 -2. ..30.. +00'
( N I) ( (RO,+) 1 3 't 1 ,,0 -2.16'+0+02
( PB) ( CRO'f) 1 'f 't 1.0 -2e10Z1+02 ..2.1750+02 7.1153-01
I SRI ( eRG 'f) 15 't 1 .0 -3.3336+0Z
\ ( zro ( CROll) 1 b 'I 1 .0 -2.3965+02
. ~ ( AL.+)2( CRa'l) 3 17 'f 3.0 -A.'i9Hl+02
!. ( BE) ( (ROli) 1 0 'i 1 .0 -2.9'f'f'f...02
'. ( 1; i .-3 ) Z ( CROii)3 19 £f 3.0 ..6.1862+02
( CE+'f) ( CRO'f)2 20 'i 2.0 -5.6251+02
( CS) 2 ( CRO'i) 21 'i 1 . U ...3.1800+02
( CU+l)2( CRO'f) 22 Ii 1 .0 -1.8990+02
. ( FE+3)2( crw'i) J 23 'i 3.0 -6.'15';6+02
\..V M N + 3) 2 ( CRO'f)3 2'f '1 3.0 "6.8801+02
( I~B)2( CRO"f) 25 'f 1.0 -3.1702+02
"".
-------
TABLE I (continued)
( SB+3)2( eRO")3 26 It 3.0 -6.1109+02
( SN+~) ( CRO'+)2 27 '+ 2.0 -'+.2563+02
. ( TH+~) ( CROlf)2 28 '+ 2.0 -6.1'+33+02
( U+'+) ( (.RO'+)2 29 If 7..0 -5.730,++02
( ZR+'i) ( CROQ)2 ,30 If 2.0 -S.9697+02
(, AG) 2 ( V106) 1 S 1 .0 -3.9907+01
, ( GA) ( V106) 2 5 1 .0 -S.673'++01
( CA) ( V206) .3 S 1 .0 -S'6098+02 , -S.S870+02 2.282S+00
V20bl -'f.5132+02 \
" ( CD) ( If 5 1 .0
( CO) ( V106) S S 1 .0 ..If.'iA77+U2
(, CU+21 ( V206) b S 1 . 0 -Ii,' 1 "1 5 + 0 2
( f(+2) , V20/d 7 5 1 .0 _0'f .5287+02'
( K ) Z ( VZ06) a 5 1 .0 ';'5.5'+66+02
( 1.1)2( V206) 9 S 1.0 ..5.6295+0Z
(' MG) ( V106) 10 S 1 .0 -5'3'+32+02 "'~.279U+02 6.'t16't+OQ
( 11 N + 1) ( V206) " 1 1 S 1.0 -'+.8'+18+02
. ( . NA)2( Vl06) 12 5 1 .0 -5.lfC'19+0Z ..S,S29't+02 ",Cf,6SJa+oc
( 1'4.1) ( VZ06) 1.3 S 1 .0 -'f.'I359+02 I
( PB) ( V206) 1 ~ 5 1 . 0 -Of \>Ii S S 5 + 0 2 0
....... , .
N ( SR) ( V206) 15 !) 1 .0 -5.6296+02
":'(' ZN) ( V206) 16 S 1 .0 -'+.6675+02
" ( Al.+3)2( V206)3 1 7 5 3.0 -1.537.0+03
\C' BE) LV 206) 18 S 1 .0 ..5.2171+02
( ~1+3)2( V206).3 1 «1 S 3.0 -1.3000+03
( ((+'+) ( V206)2 20 S Z.O -1.(]171+03
( 'C S ) 2 c.~v 1 06 ) 2 1 5 1 .0 -5.5'133+02
( CU+l )7.( VlOj,) 22 5 1 . n -,+.\737+02
( 'n. + .3 ) 2 ( V7.0{,)3 23 5 3.0 "1.327".U3
( I'IH+3)2( V2(6)3 Zlf S. 3.0 -1.,36Q6+03
( 'rw ) ~ c. V 2. 0 6 ) 25 S 1 . 0 -5-52/:'6+0Z
( 5B+3)Z( V20bl3 2/. 5 3.0 "- 1 . 2 9 '+ '+ + 0 3
'. ( SN+,+) ( V206)2 27 S 2.0 -0.8'+00+OZ
A TH+'i) ( VZ06)Z 28 '5 2.0 -1.0686+03
'. ( U+'i) ( V206)2 29 S 2.U -1.0272+03
.-
"'( ZR+~) ( V206)Z 30 S 2.0 -1.0519+03
. ( AG)Z( 503) 1 0 1 .0 "1.1207+02 -1.1~'+0+02 -2.33'+0+00
( 8A) ( 503) 2 6 1 .0 - 2 . B 1 't G.. 0 2 -2.0260+01 -1-1171+00
\( CA) ( 503) 3 0 1 .0 "2.75011+02 "Z.7588+0Z -7.9701-01
( CO) , 503) 't' 6 1 .0 -1'0'+12+02 -1.6390+0Z 2-1793-01
-------
TABLE I (continued)
/
'. ( .CO) ( 503) 5 b 1 . a -1.617!)+02
l--CU+2) ( 503) 6 I:. 1 .0 -1'.2728+02
'. t FE+2) C 503) . 7 b 1 .0 -1.I,S5'-1+02
( K ) 2 ( 503) 8 I:. 1 .0 -2.6159+02 -2.6690+02 6.6863-01
( L [ ) 2 ( 50) 9 6 1 . 0 -Z.772)+0~ -2.79'10+02 -Z.1661f+00
( 11 G) ( 503) 1 0 b 1 .0 -Z.'t718+02 ..2,"'10U+OZ 6.18"'9+00
\..( MN+2) ( 503) 1 1 6 1 .0 -1.9722+0Z -1.9737+02 -1.5163-01
( I~ A ) 2 ( 503) 1 2 n 1.0 -7. .6217+.02 -~.60"O+02 1,7728+00
( IH) ( 503) 13 6 1 .0 -1.5506+02
, ( PB) ( 503) 1 Ii 6 1 .0 -1.5869+02 -1.5700+02 1.68a6+00
\. ( SRI ( 503) 15 6 1.0 -2.772'1+02 \. -2.79"0+02 ..2.1613+00
( ZN) ( 503) 1 b 6 1 .0 - 1 . 7 A '. 8 + 0 2
\.C/ AL"'3)Z( 503)3. P I:. 3.0 ~1:..6191+02
t 8E) ( 503) 18 I:. 1 .0 -2.3221+02
( l)[+3)2( 503),j 19 6 3.0 . - Ii . .3 7,Z 8 + 0 l
( ce:+~) t 503>2 20 6 2.U -I1.1.\311.\+OZ'
( C S ) 2 ( :~~~: S 0 3) 21 I:. 1 .0 -2.6/.71+02
« (U+l)2( 503) Z2 I:. 1 .0 . ,,- 1 . 270 I:. ... 0 2
, C fE+3i2( 503)3 23 I:. 3.0 ... If . 5 7 " 0 ... 0 2
( H r~ + 3 > 2 ( 50)3 2'1 6 3.0 -5.0239+02
" , 503)
~ , I~ B i 2 ( 2S I:. 1.0 -2.",532+02
l.U . ( S.b + 3 ) 2 ( 503)3 26 I:. 3.0 -'t.206'-1+0Z
: ( SN+If) ( 503)Z 27 6 2.0 -Z.?S31:.+0Z
\ TH+ 'i) ( 503)2 28 6 Z.O _'~ .931.\1 +02
'( U+') ( 503)Z 29 6 2.0 -''''5100+02
( Zff+") (. 503)2 30 6 2.0 -"'.7052+02
( ,\ G ) 2 ( T(03) 1 7 1.0 ' -2.3937+02
. ( [)A) ( 1103) 2 7 1.0 -3.9'iOl+0Z
" ( C A) ( T(03) 3 7 1 .0 -3.951./8+02
, 1103)
'. t CD) I Ii 7 1 .0 -2.922)+02
( CO) ( 1103) 5 7 1 .0 -2.8910+02 -2.89'f6+02 "3.6't'fO-O~
- ( CU+2) ( 1103) (, 7 1.0 -2.6'+'i9+02
( FE+2) ( 1103) 1 7 1 .0 -2.9'+17+02
... I. K ) 2 ( 1103) 8 7 1 .0 -3.7172+02
( LIIZ( 1103) 9 7 1 .0 -3 .cPt03+02
- ( MG) ( TI03) 10 7 1 .0 -3.7502+02 -3-.7150+02 3.5172+00
( MN+2) ( TI03) 11 7 1 .0 -3.2430+0Z "
I NA)l( TI03) 12 7 1 .0 -3.7199+02
( N 1) ( TI03) 13 7 1 .0 -2.8602+02 ~2.8775.02 -1,7280+00
( PB) ( 1103) 1 'f 7 1.0 -2.8531.j+OZ
(. .SR) ( 1103) 15 7 1 .0 -3.9350+0Z
,
-------
TABLE I (continued)
( I.N) ( TIO,) 1 b 7 1 .0 -3.1051+02 -3.0927+02 &,2""1+00
( AL+'})2( TI03)3 1 7 7 3.0 -1.07'16+03
( bE) ( T(03) I8 7 ' I . 0 -3.~B~Z+02
( 81+3)2(' TI03)3 19 7 3.0 -8.2608+02
( CE+~I ( iI03)2 20 7 2.0 -b.97~6+02
( CS)2( 'il031 2 I 7 I .0 -3.68~'t+02
( CU+I )2( TI03) 22 7 I .0 -2.6520+02
( FE+3)2( TI03)J 23 7 3.0 -n.7U56+02
,( NN+312( TI0.3)3 2'4 7 3.0 - Y . n I. 2 6 + 0 2
( r< B 1 2 '( TI03; 2S 7 1 .0 -3.6613+02
,( Sb+3)2( TI03)3 26 7 3.0 -8.390'4+02
c( SN+~) ( ll03)2 27 7 2.0 -S.771~+02 ,
( TH+~) ( TI0312 28 7 2.0 -7.5268+02
. ( U+~) ( TI03)2 29 7 2.0 -7.ltIO'i+02
\ ( ZR+If) ( TI03)2 30 7 2.0 -7.2998+02
"
- ( AG 1 2 ( WO'4) 1 8 1 .0 -2.3U92+02 -2,3070+02 2,2073-01
( (1 A) ( \'/0'+) 2 a 1.0 -'i.0176+0Z -...0613+02 ....,3662+00
( I' (A) ( '" WO'ii ".' ,";."" ,..~ 3 8 ol" ';""'"'0 .. 1 .0 -.3.9.391f+02
I-'
+' '-' ( CD) ( VlOIf) 't 8, 1 .0 -2.8305+02
., ( CO) ( WOlf) S 8 1 .0 -2.8061+02
( CU+2)' ( WOlf) 6 B I .0 -2.1.J776+02
( FE+.?) ( '110"') 7 6 1 . n -2.(\"'52+02
" r~ K IZ ( : '~' IV O't) 8 8 1 . a -3.9()81f+02
:. (, Ll ) 2 ( " wo...) 9 8 1 .0 -3.91.56+02
',( NG) ( ii'O'" ) 1 0 8 I .0 -3.66U9+02 -3~73't.3+02 -7.,)&f37+00
( MN+21 ( VlO'4) 1 1 8 1 .0 -;3.1606+0Z, -~.12S9+02 3.&f6S1+00
( NA)Z( WOlf) J 2 8 ,1 .0' - 3. A 3 26 + 0 2 '. , -3.8260+02 6.6'tl0-0l
( 141) ( WO't) 13 8 1 .0 -' -2.750"'+02 - 2'~ 7 1 62 + 0 2 3."17"+00
',( PB) ( WOtt) 1 't 8 1 .0 ' -2.77'49+02
"( SRI ( WO"') 1 S 8 1 .0 -3.?667+0z -3.9270+02 3.965...00
( ZN) ( Wo...) 1 0 8 1 .0 -2.?79S+0Z
( AL+3)2( WO~)3 1 7 8 3.0 -1.0237+03
( BE) ( WO"') 1 £; 8 1 .0 -3.S2~S+02
( f.d+312( VI 0 't ) 3 1 9 8 3.0 -7.9';59+02
( ~l+"') ( ~'/O 't ) 2 20 U 2.0 -6.8082+02
( CS)l( WOlf) 21 8 1 .0 - 3 . 9 1 0 I~ + 0 2 .
. '( :,( U+ 1 ) 2 ( VlO"'1 22 R 1 ,0 -2.'1785+02
( Ft:+3)2( WO"')3 23 8 3.0 -8.1932+02
('MN1'3)Z( W0'4)3 2'+ 8 3.0 -R.bl2Li+02
( RB)2( I/o 0 'f ) 25 8 1 .0 -3.091{,+02
( SB+3)2( WO&f)3 26 8 3.0 -7.a500+02
!'
-------
TABLE I (continued)
( SN+'i) ' \'iOli)Z 27 8 2.0 -5.'f2'10+02
, TH+LI) ( WO'!J2 28 a 2.0 -7.3103+0Z
( U+LI) ( WO'f)2 29 a 2.0 -6.8978+02
( ZR+'f) " W 0 'f' ) 2 30 8 2.0 -7.1602+02
( I,GIl( MOOLl) 1 9 1 .0 -Z.ISSI+02 -Z.1660+0Z ..,.0907+00
( HA) , MOO'i) 2 9 1.0 -3.770Ll+02
( C A) ,. MOO'!') 3 9 1 . U -3.7SLll+02
( CD) ( MOO'!) '+ 9 1 .0 -2.6785+02
;, CO) ( t1DOli) 5 9 1 . 0 -2.6521+02
( CU+2) ( 1'10 0 ,+) b 9 1 . 0 -2.3321+02
, f"E+2) ( MOO.LI) ., 9 1.0 -2.,,9'+5+02
( 102 ( MOO'!) a 9 1.0 -3.5690+02
( l.I)l' 1100'!) 9 9 1 .0 -3.75£>2+02
( hG) , MOO'!) 10 9 1 . 0 -3.5002+0l
'. ( MN+2) ( ~1 0 0 If ) 1 1 9 1 .0 -:\.00S8+02
( ti A) 2 ( '100'!) 1 Z 9 1 . 0 -3.5S83+02 -3.5'105+02 1..7776+00
( N I) ( '-100'1) 1 3 9 1 . U -2.6UZ6+02
I-' ( Pu) ' MOOLl) 1 '! 9 1 .0 -2..6188+02 -Z.6565+02 ..3.7668+00
\J1 , 51-/) ( MOO'i) 15 9 1 .0 -3.7520+02
( IN' ( ,., 0 0 '! ) 10 9 1 .0 -2.8339+02
( AL+3)2( 1100 If ) .) 1 7 9 3.0 -9 ..79Q2+02
( BEl ( MOO'!) 18 Y 1 . 0 -3.3.,96+02
'.' UI...)2( r'iO 0 If ) .3 19 9 3.0 -7.'1996+02
, Cl+'il ( NOOLl)2 20 9 2.0 -6.'1992+02
( (S)Z( NOO'f) 21 9 1 . a -3.S'i07+0Z
'.( 'CU+1 )2' MOOq) 22 9 'I .0 -Z.3318+02
" FE+o31l( tWOLf).3 23 9 3.0 -7.n,so+oz
\, I'IN... 3) 2 ( NOO'!)3 2'+ 9 3.0 -B.IH86+02
( RB)2( MOOli) 25 9 I .0 -.3.!)357+02
( 55+3)l.( MOO'!») 26 9 3.0 -7.'1001+02
( SN+'!) ( MOO'i)2 Z'7 9 2.0 -5.0939+02
.( .TH+LI) ( t100Lf) 2 2H 9 2.0 -7.01A9+02
( U+tf) ( 1100/.f)2 29 9 2.0 -6 d,057+0Z
". ( Zf<+'I) ( MOO/.f)2 30 9 2.0 -6.8'HS+uZ
...
( AG)2(ALZO/.f) 1 10 1 .0 -/.f. 136'i+02
( -bA) ,ALlO'!).. :2 1 0 1 .0 -S.6667+02
\.,.. ( . - c: A) (AL20/.f) .3 10 1 .0 -5.6910+0Z
( CD) (AL20/.f) 'f 10 1 .0 -'f.665,++02
~ ( CO) (AI-20Af) 5 10 1 .0 -'+.6337+02 -/.f.66S.)+02 -3.1633+00
~B,
-------
TABLE I (continued)
'" ( CU+2) I/\L20li) b 10 1 .0 ,-li.3?03+02
\ . I FE+2) IAL20liJ 7 1 0 1 .0 -'ie6852+02 --'.7""0+02 -5.68Z9+00
( K)2IAL20lf) 8 10 1 .0, -5,"'292+02
I Ll)ZIALZOli) 9 ) 0 1 .0 - S . 67,29 + 0 2 -5.6858+02 -1.2882"'00
I 11 G) IAL20..) 10 10 1 .0 -~hlf932+0Z -~.5100+02 -1.6829+00
I 'MN+2) IAL20li) 1 ) 10 1 .0 -'f.?8S6+02 -5.0109+02 -Z.S331f+OQ
I NA)2IAL20'f) 12 I 0 1 .0 -5.'t'i22+02 -5.'+13'1+02 Z.8800+00
( 1-1 J) (ALZOlf) 13 10 1 .0, -&f.60lf'1+02 -'t.6332+02 -2.8810.00
( PB) (AL20lf) 1 If 10 1 .0 "'1.5957+02
( SR) (AL20&f) 15 10 1 .0 -5.6669+02 ..-5.598 J +02 6.8826+00
r LN) tAL20't) , 1 b 10 1 .0 -'1.A500+02
( hl~+3) 2 (AL20lf) 3 1 7 10 3.0 -1.591'2+03
( HE) IAL20lf) 18 10 1 .0 -5.'1299+0Z
'( lj 1 ... 3 ) 2 I A (,'2 0 If ) 3 19, 10 3.0 ':'1.3lf93+03
( Cr:+'f) IAL20'1)2 20 10 2.0 -1.0'160+03
( CS)Z(AL20'l) 21 10 1 .0 -5.3916+02
( CU+ I) 2 (,\LlO't) 22 _10 1'.0, -'1..3Q71+02
\ ( FE+3)'l(AL20't)3 23 1°0 3.0 -1'39'1)+03
( I1N+3) 2 (ALlO't) 3 2'1 1 I) 3.0 -1.'1300+03
( HB)2(AL20'i) 25 10 1 .0 -5.3909+02
...... (. SB+3)2(AL20&fJ3 26 10 3.0 -1.3625+03
0'\
,( $11.1+'1) (AL20&f)2 27 10 2.0 -9.2628+02
( TH+'1) (ALZO&f)2 28 10 2.0 -1.1015+03
( U+'i) (flL20'1)2 29 10 2.0 -1.0630+03
( ZH+'t) (AL20&f)2 30 10 2.0 -1.0783+03
( /,(:d ~ (CR20Lt) 1 j 1.0 -3.0tl2u+02
( iJA) (Ck.=:O,+) 1.0 "
2 ;) -~.1632+02
( CA) (CR20-+) 3 ;) 1.0. " -4.9165+02 -4.9240+02-- -7.4862-01
( Cu) «.t{GO~) 4 j 1.0 " -3.6166+02
( CO) . (Ck204) 5 ;) 1.0 -3.5232+02 -3.5064+02 1.6814+00
( Clj+2) (C:~~04) 6 :1 1.0 i' -3.217£.+02
( Ft:+2) (Ci.0189+02
( l~iG) (CH<20-t) 10 3 1.0 -4.4739+02
( j"iN+~) «;1<204) 11 .3 ' 1.0 -3.9672+02
"
( NA) 2 (Cr~204) 12 3 ' 1.0 -5.0814+02 \\
. ( i....U (CR~04) 13 .3 1.0 -3.4874+02 "
( PU) (Ci<~04) 14 j 1.0 -3.572j+02
. .. . .~.. 00
- ~. ~. ~
-------
TABLE I (continued)
( Si\) (Ct\~OLf) 15 3 1.0 -S.1463+02
l Lj--J) (CH~04) 16 3 1.0 : -.3.7283+U2
. ( I\L +3) ~ (i..k~04) 3 17 3 3.0 '; ' -1.2380+03
( UE) «(.;H204) 18 .3 1.0 -4.244~+02
. ( ul+j)G(c.;k~04)j 19 3 3.0 -1.021.U+03
( (.;E.+4) (CH~04)2 j ,
20 2.0' -8.397<:1+02
( CS) C;: (Cf{~04) 21 .3 1.0 I I ,-~.5192.+02
..
( CU+l);;:'(CH2.04) 22 3 l.O .~ -j.l<}0/.t+02
( Ft:+3 Ld CH~04-).3 23 3 3.0 . -1.0331.+03
( "'N+j)~(Ck204)3 24 3 3.0 -1.0802+03
( RiJ Ld (,.1\~04) 25 .3 1.0 I -5.4497+02
( So+J)~(CR~04)j 26 3 j.O -9.9712+02
( ~N+4) (Cf
-------
-- 6500 TRACOR LANE, AUSTIN, TEXAS 78721
in T.M. 004-009-ChlA. A description of FORM, the computer program
used to correlate the data and estimate the unknown heats of
formation, is given in T.M. 004-009-Ch23.
Four methods for estimating absolute entropy were investi-
gated. All of the methods were based on the concept of adding
cation and anion entropy contributions and applying a correction.
Latimer (LA-002) gave Equation (7) for the entropy, 8, of a solid,
where 80 is a constant independent of the substance and m is the
atomic weight. Equation (8) extends his method from elements to
compounds with I atoms.
s
=
~ R In m + ~ R In T - R In P + 80
(7)
S~9 8
=
I
iRL
i=l
ln m. + S '
~ 0
(8)
From the entropy of KCl he calculated 80' for all elements
to be -0.94 e.u.
The author subsequently amended his estimation method in
light of the fact that when ionic attractions are involved in
solid salts, the entropy of a large ion is increased and that of
a small ion is decreased as compared to values obtained using
Equation (8). Assuming a correspondence between size and atomic
weight he corrected his entropy values by revising a graph of
atomic weight vs entropy. He had also notic:ed that the entropy
of a solid depends on the charge on the cation. Latimer obtained
constant anion entropy values by subtracting entropies of elements
forming the cation from known compound entropies.
Entropies were calculated for the TRACOR data base by
determining anion entropy contributions according to the method
of Latimer using Latimer's cation entropies (LA-D02) and the more
complete experimentally obtained entropy data from the TRACOR.
18
-------
~ 6500 TRACOR LANE. AUSTIN. TEXAS 78721
data base. The accepted values and correlation
entropy estimation method are given in Table II.
for the correlation was 2 cal/mole oK.
errors of the
The RMS error
The selection and application of the Latimer method for
estimation of absolute entropy was discussed in detail in T.M.
004-009-Ch5. A description of ENTROP, the computer program used
for estimating the average anion entropy contributions, is given
in T.M. 004-009-Ch23.
There are two basic methods used in heat capacity esti-
mation. The first, Dulong and Petit's rule, states that the heat
capacity of an element is generally 6.2 - 6.5 cal/oK per atom at
the temperature of the first transition. This rule may be applied
to the estimation of heat capacities if the temperature or entropy
of the first transition is known. Since these data are unknown
for many of the compounds of interest, our estimation method was
based on the second basic method, the rule of Neumann and Kopp.
The rule of Neumann and Kopp states that the heat capacity of a
solid is equal to the sum of the heat capacities of the constituent
elements. While the rule is normally applied to compounds at room
temperature, some extrapolation to other temperatures may be made.
It was found that the most accurate way of applying Kopp's
rule for the compounds of interest was, for the example of a metal
sulfate, to add the heat capacity of the metal oxide and the S03
"lattice" contribution. The method of subtracting metal oxide heat
capacities from the heat capacity of another salt of the same metal
and obtaining a constant contribution for the remaining lattice was
. .
found to be widely applicable. It was used for estimating the
heat capacities of sulfates, sulfites, carbonates, sulfides, and
mixed oxides. The selection and application of the method are
discussed in detail in T.M. 004-009-Ch4, Ch9, Ch13, and Ch13A.
HTCAPY and HTCAPS, the computer programs used for the calculations,
are discussed in detail in T.M.004-009-Ch23.
19
-------
TRACOR lANE. AUSTIN. TEXAS 78721
TABLE II
DEVIATIONS OF ESTIMATED VALUES OF ABSOLUTE ENTROPY
COMPOUND
ACCEPTED ENTROPY VALUE
(e.u./gmole)
DEVIATION
(e.u./gmole)
CU20
AGiO'
LI20
NA20
K20
2.2521+01
2.9074+01
9.0519+00
1.6986+01
2.3484+01
-1.782+00
7.708-01
-6.51'1- 01
-7.176-01
2.381+00
R~.,S DEVIATION
1.44.1+0(\
SNO
PBO
BEO
MGO
CAO
SRO
BAO
COO
FEO
COO
POO
NIO
1'10
MNO
CUO
ZNO
1.3555+01
1.5600+01
3.3685+00
6.4025+00
9.4843+00
1.2996+01
1.6795+01
1.3092+01
1.4191+01
, 1.2655+01
1.3378+01
9.0782+00
8.3042+00
1.4265+01
1.0187+01
1.0392+01
-1.16?'-Ol
-4.713-01
-1.503+00
-1.769+00
-3.871-01
4.2'17-01
2.523+00
-3.797-01
3.219+00
1.283+00
1.070-01
-1.993+00
-2.067+00
3.393+00
-1. H\S+OO
-1.079+00
RMS DEVIATION
1.711+00
20
-------
/J1rJ:if...~'!:!!!i 6500 TRACOR LANE, AUSTIN, TEXAS 78721
TABLE I I
(continued)
DEVIATIONS OF ESTI~~TED VALUES OF ABSOLUTE ENTROPY
COMPOUND
ACCEPTED ENTROPY VALUE
(e.u./gmo1e)
B203
AL203
GA203
5C203
Y203
FE203
MN20j
T1203
LA203
CR203
81203
AS203
Rtv\S DEVIATION
U02
TH02
5102
GE02
SN02
PB02
RU02
ZR02
V02
M002
W02
TI02
HF02
MN02
IR02
Rf.1S DEVIATION
NA2S
AG2S
RMS DEVIATION
1.2665+01
1.2170+01
2.0220+01
1.8395+01
2.3682+01
2.0880+01
2.6398+01
1.8816+01
3.0555+01
1.9375+01
3.6122+01
2.5586+01
1.8622+01
1.5586+01
1.0000+01
1.3204+01
1.2504+01
1.8262+01
1.2500+01
1.2021+01
1.2100+01
1.1056+01
. 1.6006+01
1.2000+01
1.4174+01
1.2686+01
1.7201+01
2.2480+01
:5.4000+01
21
DEVIATION
(e.u./gmo1e)
2.201+00
-4.694+00
-3. OL~3+00
-1.868+00
-1.182+00
-7.838-01
4.935+00
-1.648+00
2.092+00
-1.8139+00
4.058+00
1.822+00
2.835+00
1.561+00
-1.376+00
8.385-01
8.425-01
-1.657+00
1.700+00
-1.061+00
";1.140+00
9.385-01
-2.305+00
-5.516-02
1.138+00
-1.688+00
1.324+00
9.393-01
1.336+00
-4.598-01
4.598-01
.-
I 4.598-01
-------
6500 TRACOR LANE, AUSTIN, TEXAS 78721
TABLE II (continued)
DEVIATIONS OF ESTIMATED VALUES OF ABSOLUTE ENTROPY
COMPOUND
BES
MGS
CAS
SRS
BAS
VS
t.'NS
FES
CUS
ZNS
CDS
SNS
pas
ACCEPTED ENTROPY VALUE
(e.u./gmole)
7.4059+00
1.0607+01
1.3450+01
1.6986+01
2.1501+01
1.4501+01
1.9722+01-
1.5200+01
1.5U87+01
1.3785+01
1.6962+01
1.8395+01
.2.1788+01
DEV IATI ON
(e.u./gmole)
-1.916+00
-2.015+00
-0.722-01
-3.650-02
2.779+00
-6.211-01
4.'~00+00
-2.223-01
6.'156-02
. -2.138+00
-9.60'1-01
2.730-01
1.265+00
RMS DEVIATION
1.823+00
M02S3
AL2S3
BI2S.3
2.8000+01
2.2934+01
3.5300+01
-1 . 't 11 + 0 0
2.123+00
-7.115-01
RMS DEVIATION
1.528+00'
MOS2
RES2
FES2
CU2S
SNS2
RUS2
1.5098+01
1.9996+01
1.2686+01
2.8883+01
2.0880+01
1.7487+01
-2.223+00
-2.579-02
-2.736+00
2.261+00
2.758+00
, -3.425-02
RMS DEVIATION
2.047+00' ,
LI2C03
K2C03
NA2C03
AG2C03
2.1582+01
3 . 60 7'H 0 1
3.2490+01
3.9968+01
-1.4'16+00
1.645+00
1.462+00
-1.6()1+00
RMS DEVIATION
1.557+00
22
-------
f!!!!j:}~({~~{16500 TRACOR
LANE, AUSTIN, TEXAS 78721
TABLE II (continued)
DEVIATIONS OF ESTIMATED VALUES OF ABSOLUTE ENTROPY
COMPOUND
ACCEPTED ENTROPY VALUE
(e.u./gmole)
DEVIATION
(e.u./gmole)
MGC03
SRC03
BAC03
MNC03
FEC03
ZNCOj
PBC03
CDC03
1.5696+01
2.3173+01
2.67lH+Ol
2.04.74+01
2.2194+01
1.96A5+01
3.1272+01
2.~180+01
-3.299+00
-2.210-01
1.686+00.
-1.221+00
3.995-01
-2.609+00
4.378+00
6.857-01
RI~S DEVIATION
2.296+00
LI2S04
NA2S04
K2S04
AG2S04
3.5357+01
3.5692+01
4.1975+01
4.7756+01
4.662+00
-3.003+00
-1.202-0 l'
-1.539+00
RMS DEVIATION
2.878+00
MGS04
BES04 .
CAS04
SRSOt.
BAS04
~INS04
COS04
NIS04
CUS04
ZNSOt.
CDS04
POS04
2.1900+01
1'.8620+01
2.5491+01
2.9074+01
3.1487+01
2.6781+01
2.7067+01
1.8586+01
2.7067+01
2.9767+01
2.9409+01
3.5166+01
-1.701+00
-1.681+00
1.894-01
1.073+00
1.786+00
4.794-01
4.661-01
-7.915+00
2.661-01
2.866+00
5.073-01
3.665+00
RMS DEVIATION
2.616+00
23
-------
~m 6500 TRACOR. LANE, AUSTIN. TEXAS 78721
TABLE II (cOlltinued)
DEVIATIONS OF ESTIMATED VALUES OF ABSOLUTE ENTROPY
COMPOUND
LIAL02
NAAL02
RMS DEVIATION
MGAL204
CAAL204
FEAL204
Rr-1S DEVIATION
CACROL~
AG2Ci(04
RMS DEVIATION
MGCR204
FECR204
COCR204
RMS DEvIATION
lIFE02
NAF£02
CUFt:02
RMS DEVIATION
ACCEPTED ENTROPY VALUE
(e.u./gmole)
DEVIATION
(e.u./gmole)
1.2700+01
1.6900+01
-1.000-01
1.000-01
.1.000-01
1.9250+0 j,
2.7300+01
2.5400+01
-3.233+00
3.117+00
1.167-01
2.594+00
3.2013+01
5.1794+01
-1.7l.f0+00
1 . 7 L~ 0 + 0 0
1.740+00
2.5276+01
3.4879+01
2.7000+01
-1.843+00
l~.961+00
-3.118+00
3.546+00
1.8000+01
2.1100+01
2.1200+01
1.667+00
7.667-01
-2.433+00
1.759+00
24
, ...,..
-------
TRACOR LANE. AUSTIN, TEXAS 78721
TABLE II (continued)
DEVIATIONS OF ESTIMATED VALUES OF ABSOLUTE ENTROPY
COMPOUND
ACCEPTED ENTROPY VALUE
(e.u./gmole)
DEVIATION
(e.u./gmole)
MGFE204
NIFE20/~
ZNFE204
CAFE201~
2.8280+01'
3.0070+01'
3 . 2180 + 0 }',
3.4449+01'
-9.898-01
-2.100+00
-3.900-01
3.480+00
RMS DEVIATION
2.101+00
MGH004
CAf-1004
FEM004
2.8L~00+01
2.9300+01
3.0400+01
i 5.333-01
i -2.667-01
-2.667-01
RMS DEVIATION
3.771-01
LI2TI03
NA2TI03
2.1900+01
2.9100+01
4.000-0r
-4.000-01
4.000-01
Rt-IS DEVIATION
MGTI03
CATI03
SRTI03
BATI03
FETI03
COTI03
1.7800+01
2.2400+01
2.5992+01
2.5777+01
2.5276+01
2.3800+01
-2.708+00
. 1.925-01
1.085+00
-8.302-01
1.9613+00
2.925-01
RMS DEVIATION
1.483+00
25
-------
6500 TRACOR LANE, AUSTIN, TEXAS 78721
TABLE II (continued)
DEVIATIONS OF ESTIMATED VALUES OF ABSOLUTE ENTROPY
COMPOUND
ACCEPTED ENTROPY VALUE
(e.u./gmole)
DEVIATION
(e.u./gmole)
ZN2TI04
SR2TI04
3. 2nOO+Ol.' .
3~8000+01'
-1.500+00
1.500+00
RMS OE\lI,\TION
.1.500+00
-- -.-.-
NA2T1205
~1GTI205
4.1500+01
3.0400+01
1.850+00
-1.850+00
RMS DEVIATION
1.850+00
CAV206
MGV206
4.2800+01
3.8270+01
1.415+00
-1.415+00
Rr-1S DEV IA T ION
1.415+00 .
C Mv 0 4
SRvJ04
BA~~04
FE\'i04
ZNvJ04
. 2.7000+01
2.a300+01'
3.2000+01
3.1500+01'
3.3800+01.
-1.560+00
-2.960+00
-9.600-01
1.840+00
3.640+00
RMS DEVIATION
, 2.398+00
26
-------
~ 6500 TRACOR I.ANE, AUSTIN, TEXAS 78721
The accuracy of the correlation was estimated by calcu-
lating the deviation between observed and calculated values for
those compounds for which heat capacities had been determined
experimentally. The accuracy was expressed as the fractional
error, which is the RMS error divided by the average known heat
capacity for the group of compounds of interest. This error along
with the estimated heat capacity is shown in Tables III, IV, and
V for the carbonates, sulfates, and sulfides, respectively.
The uncertainty in calculated log K and temperature
values caused by errors in estimated thermodynamic properties was
studied. It was found that the effect of errors in heat capacity
and entropy was negligible compared to the effect of errors in
heat of formation and that the effect was smaller at higher tempera-
tures. The results of such investigations are described in detail
in T.M.s 004-009-Ch2 and Ch4. The correlation errors for estimation
methods developed at TRACOR were within the limits of experimental
accuracy. The uncertainties in temperature and log K values result-
ing from the known correlation error in predicted ~H values were
taken into account when the log K and temperature values were used
in the thermodynamic screening process.
2.4
THERMAL STABILITY STUDIES
.The third task of importance for thermodynamic screening
was to examine the thermal stability of metal oxides, su1fites
and sulfates using information obtained from the literature survey
and the compilation of thermodynamic properties.
2.4.1
Thermal Stability of Metal Oxides
For a metal having more than one oxidation state, it was
necessary to determine which of its oxides was the most stable
under flue gas conditions. The free energy of reaction for the
oxide going from one oxidation state to another over the tempera-
ture range 20 to 800°C was calculated using the data base of thermo-
dynamic properties from Equations (1) through (4). The free energy
27
-------
TABLE III
~STIMATEa HEAT CAPACIT!~S FOR. CARBONATES WITH KNOWN HEAT CAPACITIES
CAL.I G~OLfl DEG.KfLVIN
C? = C? (OXIOE> + DELTA CP
OE:LTA cp =
1.325&+01 +
5.8661)-03*T
FETAL CARBONATE
A
R
o
1 .~G~CC3 2.03rJ5E8+01 1.2SCLtC3-D2 5.769311+C5
2 2,!\C03 2.57-333.3+01 6.905243-03 7.752731+05
3 C A ':'J ~ 2.491211+01 6.945S51-03 - 7.43025(;+85
4 FE CO:>: 2 .°47 11 4 4 + iJ 1 7.805626-03 6.529613+05
N r.GC03 ~ . 3 2 3 Q 2 6 + 0'1 7.605225-1)3 7.24g702+05
(X) -
5 i"~ N :: 0 3 2.415'333+G1 7.8G5?C1-03 '5.64'3053+05
7 :\,t.~C03 2.374B35+01 1.126278-:02 5.769911+05
c SRCO 3 2.533931+01 6.986474-03 7.575995+05
'.J
5.7699+05/T..2
TEMPERATURE RANGE
DSGREES CENTIGRADE
FRACTIONAL RMS
ERROR
2.500+01 TO 7.CGO+02 6.C86-C2
2.500+01 TO 7.000+02 3.927-;:;2
2.270+02 TO 6.270+02 3.5 G5 - C 2
2.500+01 TO 1.200+03 4.737-02
2.500+01 TO 1.727+03 1 . 04 5 - {} 2
2.500+01 TO 1.200+03 2.37a-C2
2.500+01 TO 7.000+02 1.182-01
2.500+01 TO 7.000+02 6.719-(;2
-------
TABLE I V
[STI~ATEO HE~T CAPACITIES FeR M~T~L SULFATES wITH KNOWN HEAT CAPACITIES
C~L.I GMaL~1 DE~.XELvrN
C? = CP(OXIDE) + DELT~ CP
.,'
DELTA CP =
9.0132+GO +
1.6828-02*T - -4.6680+04/T**2
MS:TAl SULFATE
A
a
D
TtMPERATURE RANGE
DEGREES CENTIGRADE
FRACTIONAL RMS
ERROR
1 !.G2S04 2.225737+01 2.336632-02 -4.668011+04 2.500+01 TO 7.00C+02 a. 1 ']4 - 0 2
2 pL2(~04)3 5.344459+01 5.4824~S-C2 6.522557+05 2.500+01 TO 7.000+02 1 .6 26 -02
3 2AS04 2.174657+01 1.786755-02 1.516869+05 2.500+01 TO 7.000+02 8.643-02
4 BE: S C 4 1.745331+01 2.032752-02 2.701013+05 2.50C+Gl TO 7.000+02 6.482- 82
1\.)5 C A504 2.03S~80+01 1.79C816-02 1.193554+05 2.270+02 TO 6.210+02 2.111-02
\0
G CO 504 1.d.SS93Ei+D1 1.890675-02 -4.06.9011+04 2.500+01 TO 7.000+02 2.016-02
7 CUS01~ 1..328013+01 2.162544-02 -4.568011+04 2.5DO+01 10 7.000+02 8.156-02
8 ~GS04 1.918795+01 1.856752-02 1.011990+05 2.500+01 TO 1.121+03 2.573-02
o t-lNS04 2.;;11127+01 1.875820-02 4.123509+04 2.500+01 TO 1.200+03 3 . 12 c- Q 2
J
IG NA2SC4 2.470E54+01 2.2225~2-a2 -4.6f>8Cll+G4 2.500+01 10 7.0CO+C2 6.424-()2
11 P3S04 1.9G0841+D1 2.082752-02 -4.66-9011+04 2.500+01 TO 6.000+02 4.356-02
12 SRS04 2.134760+01 1.794877-02 1.33~283+05 2.SDO+C1 TO 7.000+02 7.098-02
13 ZNS04 2.071091+01 1.804672-02 1.711961+05 2.500+01 TO 1.200+03 4 .3 3 2- 0 2
-------
TABLE V
ESTIMATED HEAT CAPACITIES FOR METAL SULFIDES
CAL.I GMOLEI nES.KELVIN
CP = CPCOX!OEJ + OELTA CP
DELTA CP =
WITH KNOWN HEAT CAPACITIES
1.7972+00 + -1.0570-03.T - -1.6525+04/T..2
r'1~ T '\L SUL FI DES
13
o
A
Mj25 1.305138+01. 5.930'357-03 -1.652523+04
2 £:1253 3.011778+01 4.8?490]-03 -4.957569+G4
.., CO <:; 1.144399+01 1.0214G3-03 -1.652523+04
.~
4 FE S 1.345314+01 9.425:;57-04 5.944491+04
w 5 t1G S 1.1'37196+01 6.821647-04 1.313539+05
o
6 ~'NS 1.2'30123+01 8.828407-04 7.138997+04
1 NA:?S 1.743055+01 4.339124-03 -1.052523+04
'3 PBS 1.239242+01 2.942159-03 -1.652523+04
o 5B 25! 2.447974+01 1 .3 S H' 2 7 - 02 -4.957569+04
.-
10 SNS 1.134365+01 2.440469-03 -1.652523+04
11 SNS2 2.124074+01 2.345014-04 4.827346+05
12 1152 2.[172594+01 -1.1345f5-U3 3.1669'31+05
13 US2 2.273525+01 - 4.9 <} 312 G- iJ 4 3.625679+05
14 ZNS 1.350092+01 1.613621-04 2.013510+05
TENPERATURE RANGE FRACTIONAL RH~
DEGREES CENTIGRADE ERROR
2.500+01 TO 1.000+02 1.296-01
2.500+01 10 5.000+02 8 . 4 0 If -0 3
2.500+01 TO 7.000 +02 9.865-02
2.500+01 TO 1.200 +0 3 1.060-01
2.500+01 TO 1.121+03 1 . 04 5 - 0 1
2.500+01 TO 1.200+03 4.364-02
2.500+01 TO 7.000+02 2.380-03
2.500+01 TO 6 .000 +0 2 9.568-02
2.500+01 10 9.000+02 9.250-03
2.500+01 10 7.000 +02 5.009-02
2.500+01 10 1.200+03 4.671-02
2.500+01 TO 1.200+03 2.217-02
2.500+01 TO 1 .200 +0 3 3.686-02
2.500+01 TO 1.200+03 3.062-02
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
changes were used to determine which oxidation state was the most
stable under flue gas conditions over the temperature range of
interest. The calculations and results are given in detail in
T.M. 004-009-Ch7. An example of the results for PbO and PbO is
:2
given in Figure 2. Figure 2 shows that PbO is the most stable
:2
oxide under flue gas conditions up to 268°C and that PbO is the
most stable at higher temperatures.
2.4.2
Thermal Stability of Metal Sulfites
Thermal stability of metal sulfites was studied using
both descriptive and quantitative information. From the literature
survey it was found that on heating, metal sulfites either decompose,
disproportionate to form sulfate and sulfide, or both dispropor-
tionate and decompose. It was found that su1fites are generally
unstable at high temperatures and that the occurrence of dispro-
portionation or decomposition depends on the thermodynamic
equilibrium and the reaction kinetics. In particular if the sulfite
decomposition temperature is high, disporportionation occurs to
a large extent. However, for sulfites with lower decomposition
temperatures, disproportionation and decomposition are competing
reactions. These ideas are discussed in detail in T.M. 004-009-Ch3.
The tendency towards disproportionation or decomposition was deter-
mined by calculating the equilibrium constant for the two reactions
over the temperature range of interest using the program AIRPOL.
These data also provided a measure of the sulfur oxide partial
pressure over the reaction product. The results of these calcu-
lations, as well as the results of experimental work described
in the literature, are given in detail in T.M. 004-009-Chl6. An
example of the calculations is given for zinc in Figures 3 and 4.
Figure 3 shows that zinc has a thermodynamic tendency to dispro-
portionate in the temperature range of interest since the free
energy (-RT ln K) is negative. Figure 4 shows that, under flue
gas conditions, the sulfite would decompose around 200°C.
31
-------
,
211
=m
~
~
b-
II> '"
o I
.... c..,
II> I
£ ~
~ 10
'"
o
'"
~
m
a..
LL
o
LIJ
...J
o
L
a:
LIJ
a..
I- .
:z
«
l-
V)
Z
o
u
L
w ::J
N
a:
m
...J
::J
o
LIJ
........
o
.......
CI
o
...J
-10
-15
-20
-25
o
-5
PB02 -7 PB + 02
o
300 400 500
TEMPERATURE - DEGREES CENTIGRADE
100
200
P BO ~ P B + t 02
600
700
800
FIGURE 2 - DECOMPOSITION OF LEAD OXIDES
STANDARD STATE OF 02 IS 0.028 ATM
-------
J-
Z
cc
!-
en
-,.
..,,!.-
o
u
2-:
::J::r
~ a
(r- en
(11 Z
v.r-~ N
c:w .
~ l.!-
-, 0
G .
!.lJ lL'
--I
o
o ::::
.-"
"--I a:
OLLJ
C) CL
-'
"""") ~. ,- .- ==' ,..... !"""
,(o~ o.!L: 00
LOGARITHM OF THE EQUILIBRIUM CONSTANT FOR
THE DISPROPORTIONATION OF ZINC. SULFITE.
t! ZNS03
3 ZNSOt! + ZNS
..
lSr
....1
10
,.1
5
o~ ~
I .... . .' . . . . . l
- ~ ..... j
Sc- ... .. . J
~ . .., . . j ,
-lO ~, . . . ,. . , , I. ".. " . ,I. . . ,.. , ' ,I. ,. r, .' , , 1"" ,. . ...1 'I' ,.' . ,.1,. ,. . , , , .1 , , ,. ,. . , ,~
o 100 200300 ijOQ 500 600 700 600
TEMPERRTURE,
DEGREES CENTIGRADE
FIGURE
3
!r,~qC@
-------
J-
Z
e:
J-
en
~_.,
"
C)
U
.'-
-""'
..-J
>--t
0:: 0J
will 0
.po..-. (l)
-..1
~ Lt-
=J C;
G
L! J L'..J
~
..-. 0
a :z
.-f
--.; rr
o ;,1
c CL
-'
I 25 5~P G3
10 -
[
~
sl:
I
l-
I
.t-
L
L
o t-
~.
t
_5 I
i-
t
r
-10 [
I
~
t
r-
~ . .'
-15 :" , , , I , I , , II II . I , , I I I I , I , , I I I . I, , , , , I , , , I I , , , I , , I , I, , I I , , ,', , I , , , , I , , , I I, , , , , , , I '-..I
o 100 200 300 QOD SOD 600 700 . 800
LOGARITHM OF THE EQUILIBRIUM CONSTANT FOR
THE DECOMPOSITION OF ZINC SULFITE
ZNS03
.. '502 + ZNO
--,
j
j
--'
J
j
-I
~
j
TEMPERATURE - OEG9EES CENTIGRADE
FIGURE 4
Tf'LPC;Or; ,I
----.J
-------
iIiI1iiiIi 6500 TRACOR LANE, AUSTIN, TEXAS 78721
2.4.3
Thermal Stability of Metal Sulfates
The thermal behavior of metal sulfates was also studied
using both calculated equilibrium constants and reports of experi-
mental work described in the literature. From the literature search
it was found that the sulfates can be roughly classified in four
groups: (1) sulfates which show simple decomposition accompanied
by the formation of the corresponding oxides; (2) sulfates which
. decompose to the oxides with the intermediate formation of oxy-
sulfates; (3) sulfates which decompose with oxidation; and (4)
sulfates which decompose with reduction giving the metal as the
product. Naturally, this grouping is crude, since the decomposi-
tion path is influenced by the experimental conditions, especially
the atmosphere. These findings are discussed in detail in T.M.
004-009-Ch8 and are summarized in Table VI on page 40.
Using the compilation of thermodynamic properties, equili-
brium constants for the sulfate decomposition reactions were
calculated. The sulfate decomposition reactions were written with
S03 as the gaseous product rather than S02 and 02' since the former
method was the most accurate way to represent the total sulfur
oxide concentration in the flue gas for the temperature range of
interest. The values of log K for the decomposition reactions
were plotted from 25 to 800°C by Ca1comp plotter with instructions
from the program AIRPOL. These plots were valuable in the thermo-
dynamic screening process since they gave a measure of the sulfur
oxide partial pressure over the sorption product and a prediction
of the sulfate decomposition temperature. The plots are included
in Section 8.2 of this report. Use of the equilibrium plots for
thermodynamic screening is described in. detail in Section 2.7.1.
2.5
CATALYTIC OXIDATION PROPERTIES OF METAL OXIDES
In order for the reaction product of sulfur oxide sorption
by metal oxides to be a sulfate, either the metal sulfite must be
formed and then disproportionate, or oxidation must take place.
Since some metal oxides are known to be catalytically effective
35
-------
l1J1IJ:iJiJ;Ji 6500 TRACOR LANE, AUSTIN. TEXAS 78721
as S02 oxidation catalysts, the properties of these oxides and
the oxidation reaction itself were studied in some detail. An
extensive literature search revealed several ideas. An oxide
that is effective as an S02 oxidizing catalyst is unsuitable as
a sorbent in the temperature range in which it is catalytically
effective, since according to the reaction mechanism proposed
for catalytic oxidation, its sulfate must be unstable at that
temperature. On the other hand, the oxides which are poor cata-
lysts because of the formation of a stable sulfate are potential
sorbents. Thus the main difference between catalytic effectiveness
and sorption of sulfur oxide with the formation of sulfate is the
partial pressure of S03 over the sulfate intermediate. The results
of the literature search including experimentally determined
catalytic activity for many oxides are discussed in detail in
T.M. 004-009-Ch16 in parts III and IV.
2.6
DETERMINATION OF PRICE AND AVAILABILITY OF
METAL OXIDES
The fifth task of importance for the screening process
was to determine the price and availability of the potential metal
oxide sorbents. In determining the availability of a sorbent,
such factors as world and domestic production, consumption, and
stockpiling were considered. All of the data were reported in
short tons per year. The results of the survey are summarized in
Table VI.
2.7
PRELIMINARY SCREENING BASED ON THERMODYNAMICS AND
SORBENT AVAILABILITY
2.7.1
Criteria for the Screening Process
The goal of the preliminary screening process was to use
all the thermodynamic and descriptive information obtained up to
that point to reduce the field of potential sorbents. Oxides not
thermodynamically capable of reducing the sulfur oxide concentra-
tion in the flue gas to the desired level were eliminated from
36
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
the list of potential sorbents. The thermodynamics of the regener-
ation step were also considered in screening the oxides. A third
consideration, though not a decisive factor, was the price or
availability of the oxide. Fourth, physical properties, such as
toxicity, which were a deterrent to the use of the oxide in an
industrial process were also considered in screening the potential
sorben ts .
. The data reported in the literature and calculated from
thermodynamic properties on thermal behavior of the oxides, sulfites
and sulfates, plus catalytic activity of the oxides, were included
in the screening process. These data were used to predict the
sorption products and the course of the decomposition reaction.
Next, the decomposition reactions for the sorption products were
written and the value of the equilibrium constant was calculated
for the temperature range of interest.
Since the sulfur oxide removal process must reduce the
concentration of sulfur oxides in the flue gas emitted from the
largest power plants to 150 ppm, the partial pressure of SOx in
the exiting flue gas must be .00015 atmospheres when the total
pressure of the flue gas is 1 atmosphere. All the other species
in the sorption reaction are at unit activity (solids) so that
the equilibrium constant is numerically equal to the partial
pressure of SO in atmospheres when the reaction is written as
x
a decomposition. Therefore, a compound will be able to reduce
the sulfur oxide concentration to 150 ppm only in the temperature
range where K is less than .00015 or log K is less than -3.82. The
sorption temperature must also be high enough to prevent plume
droop and the resulting localized pollution. Most of the oxides
studied fulfilled this thermodynamic requirement.
The regeneration step imposed another thermodynamic
requirement. The matter of thermal regeneration is important
because chemical regeneration requires reasonably complex equipment
as well as the consumption of an energetic material. A requirement
37
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
of the regeneration process is that it must produce sulfur oxides
in a concentration sufficient for use in a sulfur recovery process.
Such a concentration of sulfur oxide must be obtainable at tempera-
ture limits imposed by materials of construction and availability
of heat energy, or from 100 to 750°C. All of these requirements
demand a value of log K less than -3.8 at the sorption temperature
and greater than -2 at the regeneration temperature. Another way
of saying this is that sarbents were sought that have only a slight
negative free energy change in the sorption step, since a slight
negative free energy change would be easiest to overcome in the.
regeneration step, whether the regeneration would be thermal or
chemical. Figure 5 gives an example of how the thermodynamic re-
quirements of the screening process were applied. The horizontal
lines labeled "regeneration" and "sorption" indicate the log K
values calculated as described above. As can be seen in Figure 5
the regeneration of barium sulfate would be difficult in the required
temperature range. However, copper and tin oxides have favorable
thermodynamic properties for both sorption and regeneration.
2.7.2
Metal Oxide Screening
Most of the data upon which the preliminary screening
process was based are compiled in Table VI. The result of the pre-
liminary screening process was a reduction of the field of potential
sorbents to the oxides of the 16 metals shown in periodic arrange-
ment in Figure 6. For the oxides of tungsten, tantalum, antimony,
and niobium, not enough data were available to allow a decision
to be made about their value as potential sorbents.
2.7.3
Mixed Metal Oxide Screening
It was found during the preliminary screening process
that some oxides such as the alkaline earth oxides or the alkali
metal oxides underwent a large negative free energy change during
the sorption step. As was stated in Section 2.7.1, sorbents were
sought that had only a slight negative free energy change for
38
-------
I
I
v.>
\0
~
0\
I
-
0\
.J:-
I
-
0\
LOGARITHM OF THE EQUILIBRIUM CONSTANT FOR
THE DECOMPOSITION REACTIONS FOR TIN,
COPPER, AND BARIUM SULFATES
lOG (lD)
EOUlllBRIUM CONSTANT
PER MOLE OF S03
5
-5
-to
-/5
-20
o
Sn(S04J,~2S03 + SnO,
REGENERA TlON
SORPTION
CUS04~SOJ + CuD
. SO) + BaD
/
BaS04
o
/00
200
300 400 500
TEMPERATURE, 't
600
FIGURE 5
700
800
....
-------
OXIDE
1) Li20
2) Na20
3) ~O
~
o
4) Rb20
5) Cs20
6) BeO
7) MgO
*
POSSIBLE SORPTION
RANGE WITH
SULFITE FORMATION
25 to 800° C
25 to 800° C
25 to 800° C
Unknown
Unknown
Sorption not
possible at
temp. greater
than 25°C.
Possible at
temp. less
than 150° C
*
*
DISPROPORTIONATION
RANGE OF
SULFITE
Thermodynamically
favored between
25-8000 C. Rate be-
comes appreciable
at 6500C.
Thermodynamically
favored between
25-8000 C. Rate be-
comes appreciable
at 6000C.
Thermodynamically
favored between
25-8000C. Rate be-
comes appreciable at
temperatures above
7 500 C. .
Unknown
Unknown
Shows no tendency
to disproportion-
ate thermodynam-
ically.
Thermodynamically
favored between
25-800°C. The amount
of sulfite converted
to sulfite is 5-18%
between 300 and 6000 C .
TABLE VI
PROPERTIES OF OXIDES. SULFATES, AND SULFITES
*
DECOMPOSITION
TEMPERATURE
OF SULFITE
> 800° C
> 800° C
> 8000 C
Unknown
Unknown
150° C
3500 C
DECOMPOSITION
TEMPERATURE
OF SULFATE
> 8600 C (melt-
ing point)
> 9000 C
Weight loss
when held for
2 hrs. at 12000 C
is 1.05%.
CATALYTIC
BEHAVIOR
OF OXIDE
Inactive
Inactive
> 9000C. Weight Inactive
loss when held for
2 hrs. at 12000 C
is 3.6%.
> 9000 C. Weight
loss when held
for 2 hrs. at
12000 is 6.3%.
> 1036° C. Weight
loss when held
for 2 hrs. at
l2000C is 13.9%.
690 to 760° C
Static method-
" 950° C
Dynamic method-
.. 871° C
Inactive
Inactive
Inactive
Inactive
Thermodynamic calculations were made only in the temperature range from 25 to 8000C.
PRICE
DOLLARS/SHORT TON
Metal:
Carbonate:
Sulfite:
Metal:
Carbonate:
Sulfate:
Salts:
15,000
850
2,400
Metal: 1.5million
Carbonate: 280,000
Metal: 400,000
Carbonate: 64,000
Metal: 130,000
Beryl(ll% BeO):
Buyer- Seller
Basis
Metal:
Carbonate:
REMARKS
Not a potential sorbent.
1) Sulfite disproportion-
ates. 2) Sulfate decom-
position temperature is
too high.
400
30
20
Not a potential sorbent.
1) Sulfite disproportion-
ates. 2) Sulfate decomposi-
tion temperature is too
high
20
Not a potential sorbent.
1) Sulfite disproportion-
ates. 2) Decomposition
temperature of sulfate is
too high.
Not a potential sorbent.
1) Too expensive. 2) The
sulfite will probably dis-
proportionate. 3) The decom-
position temp. of the sulfate
is too high.
Not a potential sorbent.
1) Too expensive. 2) Sul-
fite probably disproportion-
ates. 3) The decomposition
temperature is too high
for the sulfate.
Not a potential sorbent.
1) Poisonous 2) Sulfite
decomposition temp. too low.
3) Oxide is inactive, sorp-
tion with sulfate formation
would be slow.
70S" Not a potential sorbent.
285 1) Percent disproportion-
at ion is small but important.
2) Sulfate decomposition
temperature is too high.
-------
OXIDE
8) CaO
9) SrO
10) BaO
*
POSSIBLE SORPTION
RANGE WITH
SULFITE FORMATION
< 59{)o C
25 to 80{)O C
25 to 800° C
TABLE VI
*
DISPROPORTIONATION
RANGE OF
SULFITE
( Con t ' d . )
*
DECOMPOSITION DECOMPOSITION
TEMPERATURE TEMPERATURE
OF SULFITE OF SULFATE
> 1227-1527°C
1374°C
> 1347 (m.pt.)
CATALYTIC
BEHAVIOR
OF OXIDE
. PRICE
DOLLARS/SHORT TON
REMARKS
Not a potential sorbent. 1)
Sulfite decomposition temp. too
high. 2) Disproportionation of
sulfite. 3) Sulfate decomposi-
tion temp. too high.
Not a potential sorbent. 1) Sul-
fate decomposition temp. too high.
2) Sulfite disproportionates.
3) Sulfite decomposition temp. is
too high.
288
20
112
Not a potential sorbent. 1) Sul-
fite disproportionates. 2) Sul-
fate decomposition temperature
too high. 3) Sulfite decomposition
temperature too high.
Thermodynamic tend- >
ency to disproportion-
ate from 25 to 8000 C .
Rate becomes appre-
ciable at 6000 C.
8000 C
11) SC203 Unknown Found by experiment Decomposition 7800 C Probably Oxide: 3,500,000 Not a potential sorbent. 1) Too
to disproportionate should be inactive to 5,200,000 expensive.
from 300 to 5500 C . appreciable at
Secondary reactions the same temp.
~ with sulfur forma- as dispropor-
..... tion at 3000 C. tionation.
12) Y203 Unknown Found by experiment Decomposition 11240 C. Before Probably Metal: 500,000 Not a potential sorbent. 1) Too
to disproportionate should occur decomposition inactive to 1 ,000 ,000 expensive. 2) Sulfite dispropor-
from 500 to 8000 C . and compete a basic sulfate tionates. 3) Sulfate decomposi-
Secondary reactions. with dispropor- is formed. tion temperature too high.
with sulfur forma- tionation.
tion at 70ooC.
13) La203
14) Ce203
Ce02
Unknown
Unknown
Unknown
Thermodynamic tend-
ency to dispropor-
tionate from 25 to
800°C. Rate becomes
appreciable at 4500 C.
> 800° C
> 1096° C Bas ic
sulfate is form-
ed before decom-
position.
Inactive
CaO : pos-
sibiy high
oxidation
power(see Fig.65)
4,000
Thermodynamic tend-
ency to dispropor-
tionate from 25 to
80{)O C. Rate becomes
appreciable at 500°C.
> 8000 C
Metal;
Inactive
Sulfate: 60
Carbonate: 400-700
Found experimentally Decomposition
to disproportionate and dispropor-
appreciably from 400 tionation are
to 700°C. Secondary are competing
reactions with sul- reactions.
fur formation at 65OOC.
Decomposition and disproportiona-
tion are expected on the basis of
similarity to lanthanum and samar-
1\.1111.
Unknown
Unknown
Inactive
BaO : pos-
sibiy high
oxidation
power(see Fig.65)
Oxide:
Sulfate:
Carbonate:
Probably
inactive
> 6500 C Decom- Probably
pos~tion product active
is Ce02'
> 8600 C
* Thermodynamic calculations were made only in the temperature
range from 25 to 800" C .
Metal:
150,000
1.8-13.7% Oxide:
conversion
of S02 from
600 to 700" C.
Some sulfate.
found in cat-
alyst mass.
8,000
Not a potential sorbent. 1) Prob-
ably too expensive. 2) Sulfite
disproportionates. 3) Sulfate
decomposition temp. too high.
Potential sorbent 1) Sulfate
decomposition is not too high.
2) Shows catalytic activity.
Potential sorbent 1) Sulfate
decomposition is not too high.
2) Shows catalytic activity.
-------
TABL E VI
(Cont'd.)
POSSIBLE SORPTION DISPROPORTIONATION DECO!'IPOS ITION DECOMPOSITION CATALYTIC
RANGE WITH RANGE OF TEMPERATURE TEMPERATURE BEHAVIOR PRICE
OXIDE SULFITE FORMATION SULFITE OF SULFITE OF SULFATE OF OXIDE DOLLARS/SHORT TON REMARKS
15) no" Unknown Unknown Unknown 430 to 600" C Poor since Oxid"e: 120 Potential sorbent: 1) Sulfate decc
Basic sulfate reaction position temp. seems favorable for
is formed be- Ti"03-TiO,, sorption. 2) Tiogreacts too slowly.
fore decompo- goes too In catalytic experiments only smalJ
sition. slowly. amounts of TiOSO. were found.
16) ZrO" Possible at tem- Unknown 600" C\ 576 to 632° C 1 Probably Metal: 10,500 Potential sorbent 1) Sulfate
peratures below Basic sulfate inactive to 36,000 decomposition temp. is favorable.
330° C. formed before Oxide: 600 2) Oxidation power can be expected
decomposition. to 3,000 to be very low.
17) HfO" Unknown Unknown Unknown 550 to 650" C Probably Metal: 150,000 Same as for ZrO"
inactive to 276,000
18) V,,03 Unknown Unknown Unknown > 400" C Good cat- V"Os "2,400 Potential sorbent: 1) Kinetics of
Decomposition alyst.Con- to 2,600 reaction below 400"C discouraging.
VO" Unknown Unknown Unknown product is version 2) Thermodynamically favorable.
mixture of takes place
V"Os Unknown Unknown Unknown V"O. and V"Os from 400-
700" C. Part
of V"O~ is
~ reduce.
N
19) Nb"Os Unknown Unknown Unknown Unknown Unknown Oxide contain- Insufficient information for
ing ore: 2,000 decision.
to 8,000
Metal: 20,000
to 60,000
20) Ta,,05 Unknown Unknown Unknown Unknown Unknown Pentoxide: 15,000 Insufficient information for
to 16,000 a decision.
21) Cr,,03 Unknown Unknown Unknown Decomposition SOiPhYSi- Ore(50% Cr,,03): Potential sorbent: 1) sulfate
to basic sul- ca ly ad- 18 decomposition temperature probably
fate 460-64O"C. sorbed on to 36 favorable. 2) the reactivity with
Sulfate decom- Cr"01:' no 5o" up to 400" C is very poor.
position temp. reac ion
unknown. to 40OOC.
Efficiency
of Cr ° -
snO"mlX'ture
is very high
from 350 to
4000 C .
There seems to be a discrepancy between theoretical and experimental results.
the sulfite should decompose at much lower temperatures.
If sulfate decomposes at 576 to 632°C,
-------
TABLE VI (Con t ' d . )
POSSIBLE SORPTION DISPROPORTION- DECOMPOSITION DECOMPOSITION CATALYTIC
RANGE WI'l1l ATION RANGE TEMPERATURE TEMPERATURE BEHAVIOR PRICE
OXIDE SULFITE FORMATION OF SULFITE OF SULFITE OF SULFATE OF OXIDE DOLLARS/SHORT TON REMARKS
22) Mo03 Unknown Unknown Unknown Unknown Mo03catalyst Me.tal: 3,200 Not a potential sorbent. Sublimes
converts 45% Oxide: 2,800 rapidly at 600" C.
S0f}from 500-
70°C. Cata-
lyst mass
undergoes un-
described
change.
23) W03 Unknown Unknown Unknown Unknown Poor: Conver- Metal: 35-45 No decision possible.
sion reaches
a maximum of
62% at 6750C.
Catalyst is
partly reduc-
ed.
24) MnO < 2400 C Thermodynamic 4400 C MnSO is oxi- Poor catalytic Me ta 1 : 590 Not a possible sorbent. 1) Shows
tendency to dize1i in O2 activity. Cat- excellent sorption behavior
disproportion- containing alyst mass is (especially Mn02) with MnSO. form-
ate from 25 to atmosphere. transformed to at ion. 2) Decomposition tempera-
8000 C Decomposi- MnSO. ture of sulfate is too high. 3)
+=-- tion between Active form of sorbent difficult
VJ 1090-11000 C. to regenerate.
Mn2 03 < 200c --------- 1300c 1090-1100° C Same as MnO. Same as MnO. Same as MnO.
Mn02 Sorbs with forma- --------- 1090-11000 C Same as MnO. Same as MnO. Same as MnO.
tion of sulfate. for MnSO 4 .
25) Re02 Unknown Unknown Unknown Unknwon Unknown Metal: 1,160,000 Not a possible sorbent. 1) Too
expensive.
26) FeO < 1900C Thermodynamic 3800 C Oxidation Good (used Ore: 8.5 Potential sorbent: 1) Thermody-
tendency to takes place as conuner- Oxide Pigments: namically favorable. 2) Kinetics
disproportion- at 520-53OOC. cial catalyst) 60-440 of the reaction below 400°C
ate from 25- discouraging.
800° C .
Fe2 03 Sulfite is un- ------- 600c Literature Same as FeO. Same as FeO. Same as FeO.
stable at temp- disagreement,
eratures above probably 6500 C.
6ooc.
-------
OXIDE
27) CoO
28) NiO
29) RhxOy
30) Ir02
31) PdO
~
~
32) CuO
CU20
33) Ag20
34) ZnO
POSSIBLE SORPTION
RANGE WITH
SULFITE FORMATION
< 2500 C
< 1800 C
Unknown
Unknown
Unknown
Sulfite is un-
stable at tempera-
tures higher than
1100 C .
Sulfite is un-
stable at temp-
eratures higher
than 1000 C.
< 3100 C, Ag20
is stable to
250° C.
< 110° C
DISPROPORTION -
ATION RANGE
OF SULFITE
Thermodynamical-
ly favored f::om
25 to 800° C .
thermodynamical-
ly favored be-
tween 25 and
800° C .
Unknown
Unknown
Unknown
thermodynamical-
ly favored from
25 to 800°C.
thermodynamical-
ly favored from
25 to 8000 C.
thermodynamical-
ly favored be-
tween 25 and
8000 C .
thermodynamical-
ly favored be-
tween 25 and
8000 C .
DECOMPOSITION
TEMPERATURE
OF SULFITE
470° C
410° C
Unknown
Unknown
Unknown
1100C
lO00C
Predicted to
be > 570°C.
The reac.t ion,
2As2 S03 = 2Ag
+ S02+ A~ SO..'
.occurs at 1000 C.
2900 C
TABLE VI
DECOMPOSITION
TEMPERATURE
OF SULFATE
800-1000° C
End product
is C030..
About 8000 C
Unknown
Unknown
Unknown
840-935° C
CU:j SO.. is
oXLdized in
air. Decom-
poses at
840-935° C.
> 8400 C
Decomposition
product is
metal.
> 7400 C
Decomposition
30es via ox i-
sulfate as
intermediate.
( Con t ' d . )
CATALYTIC
BEHAVIOR
OF OXIDE
No datIl.
Unknown
Unknown
Unknown
Unknown
Poor,(sIJ1-
fate form-
ation).
Poor,(sul-
fate form-
ation).
Poor, (sul-
fate form-
ation).
Inactive
PRICE
DOLLARS/SHORT TON
Metal:3,000-3,300
Oxide: No data.
Metal:
Oxide:
Metal:
Metal:
to
Unknown
1,900
1,800
1,200,000
4,958,000
5,542,000
Metal: 680 to 760
Same as CuO.
Metal:
Metal:
Oxide
38,000
to 52,000
240-280
300-355
REMARKS
Potential sorbent: 1) COxOy can
be expected to have a good oxida-
tion power (See Figure 65). 2) De-
composition temperature of sulfate
is a little too high.
Potential sorbent: 1) The higher
oxides, Ni ° , can be expected to
x y
have a high enough oxidation power
for SO~ oxidation. 2) Sulfate
decomposition temperature is reason-
able.
Not a potential sorbent: 1) U.S.
production 2 tons/yr. 2)Too expensive.
Not a potential sorbent, too expen-
sive.
No decision possible, but probably
too expensive.
Potential sorbent: 1) Oxidation
power may be high enough. 2) Decom-
position temperature of sulfate is
not too high.
Same as CuO.
Not a potential sorbent, sulfate
decomposes to metal.
Potential sorbent: 1) Disproportion-
ation is 3% at 3500C. 2) Decomposi-
tion temperature of sulfate is low
enough.
-------
OXIDE
35) CdO
36) Bx 6000 C
-------
-------
TABLE VI
DECOMPOSITION
TEMPERATURE
OF SULFATE
> 1065° C
A basic sul-
fate is form-
ed before de-
composition.
650-9500 C
Unknown
Unknown
About 600°C.
> 8000 C
----t"'--
Unknown
( Con t 'd . )
CATALYTIC
BEHAVIOR
OF OX IDE
Inac tive
Inactive
Unknown
Unknown
Poor
Poor: It was
reported that
nearly all
oxides(start-
ing with PbO.)
form sulfates.
-------
Unknown
PRICE
DOLLARS/SHORT TON
Metal:
5,160
Element:
7,330
Sulfite: 44 to 76
Oxide 15 to 32
Metal: 860,000
to 1,270,000
U.S. production:
1.12 short tons/yr.
Oxide:
137,000
3,060
Metal:
Metal: 280
----------
Oxide: 950 to 1,200
REMARKS
Not a potential sorbent. 1) Since
the decomposition temperature of the
sulfite is higher than that for Zn
disproportionation will be more ex-
tensive. 2) The resulting sulfate
cannot be thermally regenerated.
Not a potential sorbent.
Potential sorbent: 1) Disagreement
between estimated and reported de-
composition temperature of sulfite.
2) No disproportionation predicted.
3) Sulfate decomposition tempera-
ture low enough.
Not a potential sorbent. 1) Price
is too high. 2) Production too low.
Not a potential sorbent.
Not feasible, U.S. prod. 1.5 tons/yr.
Potential sorbent: The poor oxida-
tion power will most probably yield
a very low reaction rate.
Not a potential sorbent. 1) The lead
oxides PbxOy' especially PbO., show
good sorption behavior. 2) Resulting
sulfate cannot be thermally regenerat-
ed. 3) Poisonous.
Not a potential sorbent, too poison-
ous.
No decision possible. According to
Fig. 65, the system Sb.Os/SbQ013 has
a high oxidation power.
-------
TABLE VI (Con t ' d . )
POSSIBLE SORPTION DISPROPORTION- DECOMPOSITION DECOMPOSITION CATALYTIC
RANGE WITH ATION RANGE TEMPERATURE TEMPERATURE BEHAVIOR PRICE
OXIDE SULFITE FORMATION OF SULFITE OF SULFITE OF SULFATE OF OXIDE DOLLARS/SHORT TON REMARKS
45) Bi203 < 1500 C Thermodynamical- 3l0°C "" 8600 C Unknown Element (refined): Potential sorbent.
ly favored in Basic sulfate 6,860
the entire temp. is formed be-
range. fore decompo-
sition.
46) U02 < 1300 C No tendency to 3200 C " 5000 C Poor, due to U30e: 12,000 Potential sorbent, will sorb S02
disproportion- Forms basic sulfate form- probably with U02S04 formation.
ate. sulfate before ation.
decomposition.
47) Th02 < 1900 C No tendency to 3900 C > 8000 C Unknown Metal Pellets: Potential sorbent.
disproportion- 30,000
ate. Oxide: 12,000
to 20,000
.p-
0'1
-------
PERIOD la
f H
2
3
4
.5
6
~
'oJ 7
lANTHANIDE
SERIES
ACTINIDE
SERIES
PERIODIC ARRANGEMENT OF:
GROUP
VIII
vb Vlb vllb 0
H He
C N 0 F Ne
P S CI Ar
''''''''
Se Br Kr
Te I Xe
Po At in
ii!I!II/IIIIIII! il//I/!I!llill/I!I/lililill 0 s
Pt
Au
TI
Nd Pm Sm Eu
cd
Tb
Dy
Ho
Er
Tm yb
lu
Np Pu Am Cm Bk
cf
Es
Fm Md
lw
POTENTIAL SORBENTS BEFORE SCREENING
j/ / // POTENTIAL SORBENTS AFTER THERMODYNAMIC SCREENING
IIiI1JiJlJl
:x:-
0'
,
~
0'
-l:""
,
FIGURE 6
~
"
-------
UiIiiiIiJ:Ji 6500 TRACOR LANE, AUSTIN, TEXAS 78721
the sorption step, since excess free energy would have to be
supplied during the regeneration step. Therefore, oxides such
as the alkaline earth oxides had to be eliminated as potential
sorbents. However, it was further found that these oxides, as
well as others, when in chemical combination with a second oxide
formed a compound with more favorable thermodynamic properties
for regeneration. Since little was known from the literature
about the reactions which actually occurred, the following simple
reaction was used for the purpose of thermodynamic calculations.
MeS04 + Me'O
.... MeMe '0 + 503
(9)
The thermodynamics of sulfur oxide sorption using such mixed
metal oxides were studied by calculating the log of the equilibrium
constant for the reaction shown in Equation (9) and determining
the magnitude of the free energy change. The equilibrium constants
were calculated using the data base of thermodynamic properties
by the program AIRPOL. The majority of the heats of formation for
the mixed metal oxides were estimated by the modified Erdos method
described in T.M. 004-009-Ch1A. It was found that in many cases
the free energy change was more favorable for the mixed metal oxide
than for one or both of the single oxides. These studies were
described in detail in T.M. 004-004-Ch22.
Preliminary screening for metal oxide pairs consisted
of determining the partial pressure of sulfur oxide over the
sorption product from the log of the equilibrium constant for the
decomposition reaction (see paragraph 2.7.1). It was found that
numerous mixed metal oxides had the desired log K values for the
decomposition reaction. These compounds are listed in T.M.
004-009-Ch22. Further investigation concerning the chemistry
of the so-called "mixed metal oxides" revealed that an experimental
program involving these compounds would be beyond the scope of
the contract. The problems involved in such a study are discussed
in T.M. 004-004-Ch25.
48
-------
Ui/iJ;jj;g;j/ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
3.
KINETIC STUDIES
3.1
INTRODUCTION TO KINETIC STUDIES
This section describes the methods used to screen the
sixteen metal oxides shown in Figure 6 on the basis of their rate
of reaction with S02 in a flue gas atmosphere. Kinetic screening
was essential because even the most thermodynamically capable
sorbent m~y be impractical for S02 removal if it reacts too slowly
to be effective or economically feasible. Kinetic measurements
eliminated such sorbents and provided an input to the economic
studies for determining reactor design and operating conditions.
3.2
APPROACH
The purpose of the kinetic studies was to determine which
thermodynamically favorable metal oxides were most reactive with
802. The criteria for reactivity is the rate of reaction. The rates
of noncatalytic gas-solid reactions can often depend on physical
effects, such as diffusion and heat transfer, which may be a result
of the conditions of the experiment or the treatment of the sample.
An attempt was made to try to minimize these physical effects and
thereby evaluate the inherent. chemical reactivity of each candidate
metal oxide. Physical effects were partially minimized by using
samples with small particle size, dilution of the sample, gas flow
through the diluted sample instead of around it, and isothermal
conditions.
The reactivity of a metal oxide is also determined in
part by physical properties that are a result of the compound pre-
paration method. Each sample was characterized by BET surface area
determination, x-ray diffraction, and chemical analysis. These
measurements were useful in the comparison of reaction rate re-
sults. All of these techniques are described in detail below.
49
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
3.3
3.3.1
COMPOUND PREPARATION AND CHARACTERIZATION
Preparation Methods
The metal oxides were prepared by thermal decomposition
(calcination) of suitable compounds. Compounds which could be
calcined at relatively low temperatures were chosen to minimize
sintering of the product oxide. The metal hydroxides or hydrous
oxides were usually employed. Procedures similar to those discussed
below for the hydroxide were used for all intermediates.
The metal hydroxides were prepared by precipitation.
The precipitate was filtered and washed several times with large
quantities of water. The moist product was dried at low tempera-
ture ( 60oC) for 1-3 days in a vacuum oven. It was then ground
if necessary and calcined either in a muffle furnace or vacuum
oven. The metal oxide product was ground, sieved, and stored
in a des sica tor. The -170 +325 mesh fraction was retained for
BET surface area and reaction rate determinations. The actual
conditions employed and the intermediates used to prepare each
metal oxide are summarized in Table IX in Section 3.5.
3.3.2
Use of X-ray Diffraction
The oxides were identified from their x-ray powder
diffraction patterns. The patterns were determined by using Cr
Ka radiation with either a Siemens type F Diffractometer (Eg 4/22e)
or a Siemens short cylindrical camera (Cat. No. 176311) for
Debye-Scherrer patterns. The diffractometer was used in most
cases since the peak width gave a qualitative indication of the
crystallinity of the sample. The lattice spacings were calculated
from the recorded diffraction angles using the Bragg equation.
Observed values were compared with those listed in the ASTM
Inorganic Index to the Powder Diffraction File 1967.
Figure 7 shows the diffraction patterns obtained for
prepared from the nitrate (A) and the hydroxide (B). The
peaks of B indicate that it is less crystalline or has a
C0304
broad
50
-------
I
I .
I
I
"I- . ..-.'- -,_. . .'" uu u "-"'-'--'1----1- -I-"-'-':::'-:~=+:::"'.:._~! --1-"=-1' :~:l:::'-~:I-~-::.r::: --: j
'- - --! -- .--:- --..- --.- .-- ------- .. ----- _u_-- '---'-'j- - ~ ~ - - - - -- --~
u - i--~ ?~t~~:c:c~ :-: :;=~=~ Cr ~ Radiation -, -J_~~-
. _.I:':~-~::~<~,-"--=----=---::'IJ..:.::__.-:-~:~~ A. Full Scale = 4 x 10' Impulses/min.:
_n__~_~. - --.. -. ~-~. -. . . -:.~ ----~ B. Full Scale = 2 x 10' Impulses/min :::'
~I:c i':::::: --::~ :;~:- :ok_=- =-~~ I ::~ ~~--I~=-j::=:C: 1::1::~~~::'~T=':~ -~~ ~~:~-I' - -~.::--:-I~~~:- :
, -, roo -- --- ----- - 1------ -......_----- - - r.--u_-... --~ -..- -- ---- -- Ju ..- ------1--' --... --. ,
.--Tln"i;-.;ij=-: ;;;,;. .1;; ~a ~~. ~~~;i~=,~~-; ;;:~';--:~~:-~i--~II;-:,;- i-~.; .
i- - -k~- F-~-'~~~=~=~' .... .'2;:}----'--=~}~~1~~; i:~.SI;~;~!{~.~Pf~~- I-I. --
..... _u _.._-.-.- ---1--- ----,j. ,,- '.'--" -.------ --- --.--- -._--._- --.--_.. -Tfffi"""u --- -.-- - ---.- -.- !. 1 --.
, : :. . ~ .:~ -~_. ~:. -~h-l ,- .---::=- -::'.-~:- '~~::--~-:j :_::-~~:: -= C' - -:. - ". : .': _I i
, - \_. --- .. - -'-- -----.1.. n" -, -'--- --,-- -----. '---- ,------ -. -- "----1'--- - _I - - - '_00.. -. "i -..
; ::. - --: 1:- :~~: =~~~~ - "::.1 " - .=:~:- -::. - -J~:~=' ::=-,=-~j -~:.:-:- ----- ~--=:-I -. '. , ~ : - ! _:_: I .: I -
; -;--==1-----'------ __nO' ---1--'----- ---------l----------_._-,- -,----,1 -., .
(: I.... ....I--:~hf~';;;i /H_;~~-:;~E: ~~i! ~-<~i=-ti . I;r;- I: .:1 .. .
,..I-~tHJ-1 ---1-- ---+-- _w--11 {-I. 'I .
: ~I-'=#=f,. ~.2g':--:I~=~ ~== :~=-l-::~~~:_-,~-\-.-+t:t:)i--,
i ' ~I'----I:--I,~J~---:--I..--:--- 1-:....:.....-----[-..:...--,:-..:.-..-1-:""-1---:";--":--1 - --J-. . ~ I:-:--! .-.; i
i . t - --- .-..-., -.- ------- _1- - --"1 j i.f" 1 , '
I : - 1- ---,- -'--'-' -- - - u.--------- ---~. --On' ----- t':'(: - " -. I ~ i
, I , ',I . -. .'. ~ !.I. ,Ii I - ~,.'~., \. ..".\,..~..'.f
.,' . !I ~ ";L' ,l.1..Ju .,~';:J; )1 -.-- -I t~; t.I\~1' U! 4 ! r "'... & JJ ~ _I !".... .. .~Ik. t ~~ I - -::: ,;--.<:..~f., '-' " " " ~. . ;""
. ,~ ,"''' l.\" ~~...,. fI, 'f '.1fC .;4,'Y(~/'~\I ". Ih"',':~ !'."",~,'~_\t }lloll!." I'""t. \":.;1"; \,.-.;,.."
:;':', ~1 .'. ~..0~~"" ,--- ~-'.,' . t ~I "'-. . ....-;~ . i: i
: I, -' i --- i-- - ---l-=-~.:i , --~~,i:'::::- :,'- :::_--~::'::~_":""_':': --:=~:':I'.-=---I., h:: _::::: :.:::::::- .:-~- . ::..- 1- - I .- .! I
-: T-:=:-~~-~~ :~: 1=;-:~:~f~~~~~I..~i~-f:~=::------~ =-=--.:.. =~_._-= :..::.-=~-L=~- ':::::.=--'=::1< :::-' -.:'-:::: L.::-~-:-:.I. . - i -. . ~'
n .:. n .tce -- :, :::: :. ,.:.:- HtI~.~- I _:: =-=-l=-:.: :0::-::-1:.:: :::: ::::_: k..'-'-. _.: ..-1- . H i -. . , i
, !' -- - i'- :'_:'-:_:r~-:--J'=~-' ',~I1. h-'~llr'~":~ ~=-~:OO(:_~::h:~_--:-. .'~-~:- I,', :_--- -.'-~- - :_~j-~
- - L '! -::':":-I=-:=-.j:.-, I-:="::':i.t_:J= - E ___:::::::L...:--.-~ :~=--==I.:::::--_: L-: -'=-_1-:-_-: I :-- ! . -: :
- : :._-~---- -- i-: ::-.If- ) -.-)-==i~-\\ -F....!::-.::: ==...1-:-::.-::::- -=~ - :---\- -- ----f- - -- .\-..--! __nO -: ..- ,. - .
: B.1 ._-,--- -1+--.-- ;--- \)----,-.,,-.--__1--.----- --- .. ! ----1---- _1-- - 1- - _I - ---'-I' I' I
: I-I - .- -----: :..'. - - , - !:.: --' ~.--:~-:-:'. --~-: --_:- ,'~- ~ - - '. ~
. -~!.~. - I - :- I ~~ -~--- ------.-1 - !~. , l-:--.---.I_~.._- -----1--------1------ -\-- ,~. -- I . I "
---!------.----..f,.,... '-'f --~ -j7JTI-"'-II-: I ---j--;
, _.; ------1- . j /''.; - --- -----. I -.---- ,{, -, 'O" - -. ---- -, -- -", ,.. Li t-, \~' r,~ --', ' I '
., - . I - , . '.: --- '\-- . - '...., ,...'. ! - - 'I --- - - -- ,- . fJ i-~1.,,t 'I' I' .
J \ '))If '. - - --- 1- -- - -- -- --- - T''i~~ . - -.' ,,~':-' -,..'~ ,-,' I: ' ,
-~-I'-" I'~ .A,"f/-I,u----I-------I=-~_._\...~\...\, \------------ "-, ' ~J.~lll.'''' ---- - .tl'\,} ,.J..II~ --!,
-. - - , \ Pf i'f - -, - - . -. ' ,- ,- - - 1 f I .,...t I'~'''' I' , I ,.f ; ," , !, " ~ ,
~.~', - ~Yj./~,1 -.1--,---',-.---:----.-. -- :",.1"~'~'~~"."~.t~" ~,,~'..r~~\~ I'. I... 1 ~ .""OC;y';. .
;,.' :.~.!Jlf\~~ - i---- --1- - - ..----1...---- -.--- ,. .rr04IiV'V """~l' 1".,( ----- -00_- ',' -,. if ,I i j
.. . ,., ','- , I' ' .----- --, .-- ""'- ' 1 Jo; .-- - '_.__._~ ' '1
-l~~'/':~F~~~c:-~j~~;C~]-==-;:- 2~:~~ :::~.lc~~C~r -. ---r':.Cl~- :::i-~:; I.~-T J~- 2J
-- -1'-- i ----,- -----I~~.---! -----/".-.. -: ..... ,. -----.1------ ---------I---r-'----- r - .-- ~ -- ----1-.-.. I -'- I ." :
.u ,. ,-=- Lt~:~~~t~=:~J~~t.]}:~J==~,-: ::-t~CS:i;~ C i:.;_~ L -__1_, i
Figure 7:
X-Ray Diffraction Patterns for
Different Preparations of C0304
A. Surface Area 12 m2 /g
B. Surface Area 169 m2 / g
51
-------
iIiIJ:iIiJJi 6500 TRACOR LANE, AUSTIN, TEXAS 78721
smaller primary particle size than A. This character is also
reflected in the high surface area of B relative to A.
3.3.3
BET Surface Area Determinations
Knowledge of the surface area of a sample is an aid in
characterizing its reactivity. Samples of the same metal oxide
with different surface areas differ in rate of S02 sorption. The
specific surface of metal oxide sorbents was .determined using the
BET method, which involves the measurement of the amount of gas
(adsorbate) adsorbed on the surface of a sample (adsorbent) at
a particular pressure and constant temperature. The method,
apparatus, and data analysis are discussed in detail in T.M.
004-009-Ch24.
The BET method provides for calculation of the mono-
layer capacity, or the amount of adsorbate adsorbed in a layer
one molecule thick, from the adsorption isotherm using Equation (10).
An adsorption isotherm obtained at TRACOR for ziroconium oxide is
shown in Figure 8.
P.
1.
Vi(Po.-Pi)
1.
=
1 C-l
VmC + VmC
P.
1.
p-
o.
1.
(10)
Vm
The following definitions are necessary:
Monolayer capacity in cm3 (N.T.P.) of adsorbate
per gram adsorbent
V.
1.
Volume of adsorbate in cm3 (N.T.P.) adsorbed per
gram sorbent at equilibrium pressure p. (rom Hg)
1.
The saturation vapor pressure (rom Hg) of adsorbate
at the temperature of adsorption
Relative pressure
P
°i
Pi/Po.
. 1.
52
-------
IIiIJ:iJ;;g;j 6500 TRACOR LANE, AUSTIN, TEXAS 78721
Figure 8:
r-o
00
..........
........
p....
.
E-I
.
Z
--
~
tJ
L.....J
"d
(l)
...0
1-1
o
en
"d
<
.u
!:
;j
o
~
1.4
;. 'L" "," 'j'l: ""Ir:-H T ,.. "!' 'r'
'. :. . :; ;.:: : i: 1 ;.~ ~ 1; i ~ I r:tr t:i: . Ii : i I I
I~ ',: :,;: ::);.': ",. It J" Ifi'
, ; ! :: ..;. ~:;: i-i~ i- j t r ~!! t r:. : " ~ ~
t\... 'I' .,h :~"I~. l~~ ~ ~I ! ;t:t
,L" '''' .", 'r!, 1!' Ii', I" ",
: . ~. I" \ ,', . ~ i tt t ... . .
i1:: ,::: ~m fJ't 1 If #J -OJ :;1:
:;;:,',:: ;1~! .1. :1111' m. 1 ;,!~ I 1:, :lh :~: :it~ tj~l "
~ d T I' ~ , h t ~ f.t.~ ~L,: ..~.
'''I ;;': I:~:: Ii -0 t 1 ~1U J ~ m ~Ii:;i::J ,.., ~H r' 1 ",
iH ill ' .~ t ,tIt t 'tnH11fHHf "b. ,;
1'" 1 t",1 ,. ~ ' .. ft" .. ., -::- .. r ~._.
" l]t, 1 ' HI, jl ", ! tt mu . 4r: 'J '... Ii r "
,~)i+f 'I:: ;1;1 ':1 ;1~ 'Ij 1: mmt~ I:~~:" "" :::: ,~,:,::; -,
1" :-utJiL . 1i ht~ n" :t '!: ' :rm:m: .t; t:::... ~:t~ i ".l~~: : ..
::" !:it 11[1 t" hd 1'\' !I:t:, ,i!" ~ +I 2:: rIrJ- '::;:; 'j' : ", I
. .t- .,'1 I. . 't" t. -tt it 11" .+ .j...... ,..
,," I"i ; :11: tJ i'j!! Iii r ';.; :1:: ,:+ ':..1: , 'tfi: ,. ~
':+; I:!; ,:;, ,: I ." ::I~ r:~ ~ -~ ~;~: ....:;4 j..o-. :::.t,' ':::C' ..
11'1: '- I:: ii:: ,.. Ii::' lif f t ~i!' rH! if; ~'! '~I' t::+ W: " ,'. ,,- :,;:T~::::i' : :::': " ,i
II.' ." ,,;4.-) j. n ~I. Ht :11; Ih .~:t '" t&.+- .... -..'1~.. "..' . . . .-:
jji: ;llj :iU ,,' ,',::::, 'ii' J: .. :"Jit'! + :;J-: ;::,;:: , ...' :::,1", ,": ,!
,1:; ::q :+1+ ~~! ::i~':'8 ~~~i :, ~:: tt t ;r~: 11f:,"'.. :::J' :: ~... .' ...: :1
,: V ,::i!)::;,~j: ;;11 rP ::: hj! iT ,Wrrtt ti:: :H: ':! 'I " ,,: .
, , , '''''. " "'....,, I" "'" '"..' .. .-"--r-----'"
::::i:'; ,:.. ii:' UP ';:1 ,.. ::i: ::fi 'ijtlli~ B-::i- ,I:''': ,
thl;l~'i,'i;:.' .1 :.:~'. ;;t; "}1 ;::; :1 \ t~;~ ;~1~~~:i;~; : t:;: ;:~~ ':~'F- !"-r""
':.I!: '1/: "I;:' ., ;:;: :!It 1~~J;J. Ir ~ .... r:~:.t!~lH;t~ tt:....,: ::.-:-t-;...: ~..- ~--
I: ti. ,.., ," in: II ; 1 T' :'t; Fj'HH:i if i ..~ 'Of H.t= I: I
~._. ~ 1:1~ h ~;.;. ..-+.. :JI jl I I.:: ~tti i;1:t".tt:t it :1: :::tt H~t : ..:..L:.-:- -.. ..... . .
::k: : II : tJ!i ii:! OO"i: !;: ilf :1tl,! JIll il::bi" r:\'~;,: tJ:!: ;1 - " :
,:i:.. ", ..;t :11: oIH1: ~t; I '" 1:11 ,I"., ,1 .'- T-... TIf:1. '
[Li:!;~t~I~:i,> ~!:!I!!I! I, ,W~ in :I'i 1;:: iifi :!:~ !~! ,UT;:: ::r::21, : ," ,-::1.;-r-;'-:-
',::I~i' ::JiJL.;;' !;ii :i1: t:;: OJ:; Wt :~r: k: jf~: ;.::~::~ " ':;..; : .. ,\
~~, i,-, Ur,_jl.'.'..'." -,j:"~ll~~[':-".':.~".' ;:I?i,..,:::,..:,:, """:I"'~' ,""."":1 C,.~,:,!f,~.,..~,., :-,..,.l,..~,<, '.' ..,~~.-, .:-,
"I-}"" '''"'j ..., -'.. l-jL, l
'~-I.cj'~:~'t'.-!--.~-I.-t-IL- --p.' ~'1'81-,:I."" .
1"I:_t'~ji:lh: .j~ .. '-f!'~iJ:~r
i ' I , 'I ,!: I' ' i. ' 'i,
, I ",' I I 1 ' , ,
:': ;; i , 'I' !'t j:," : ,. "1:- : ., ."!. ,
'~_~---L-.:L~..., ,i:. i . ,."." ",. ,
~I f.J, ::r '.' ';!I ::::-T"
.,;~~ iil; J;ii :;H~!;i -':'~I'-~
. ...l f ,... ....! .' t I ~,I. . -
d; a~; :j:t :::~ .~.; i:;; .,:~-jl
1-t+ : i; ; r..} ~~;1 t-t:.t ::::
'''' 'H. ;J: I' U'J I P'
- :H HH ;fj:H~~:Ii: ::
.._-..t...
,
'-,
, --I
...-
-'-
I
1.3
1.2
1.1
1.0
0.9
0.8
0.1
0.2
0.3
0.4
0.5
Relative Pressure p!p
o
Adsorption Isotherm for Zirconium Oxide at 770K
53
-------
Ui!1J:iIiI;li 6500 TRACOR LANE. AUSTIN. TEXAS 78721
From the derivation of Equation (10) (GR-002), C is
defined by Equation (11).
C
=
e(E1 -L)/RT
V 2 al
(11)
--
VI aa
The factor, VI' is the frequency of oscillation of the adsorbate
molecule normal to the sorbent surface and E1 is the energy necessary
for the molecule to condense or be adsorbed on the first layer.
-E /RT
Therefore, VI e 1 is the number of molecules per second
evaporating from a given site in the first layer. Similarly, the
factor, V2 e-L/RT, is the number of molecules per second evaporating
from or adsorbing on the second layer. L is the latent heat of
condensation of the adsorbate.
The values of V and C are determined according to
m
Equation (10) by plotting experimentally obtained values of
Pi vs Pi and finding the slope and intercept of the re-
Vi(Po.-Pi) po.
1. 1.
suIting straight line. .The "normal" range of relative pressures
for which the plot is actually linear and the BET equation is
valid varies from about .05 to .35, although sometimes measurements
in the 0.001-.05 range are necessary. The specific surface area S
in square meters per gram is calculated from V us.ing Equation (12)
where V is the monolayer capacity in cm3 (N.T~P.) per gram, N is
m
Avogadro's number, and A is the cross-sectional area of the
m
adsorbate molecule, 16 A2.
S
=
Vm N Am x 10-20
22414
(12)
The apparatus built and used at TRACOR for BET surface
area measurements consists of a basic high vacuum system with
manometer, calibrated volumetric bulbs, and thermometers. A
54
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
mercury manometer is used to measure the gas pressure in the
system. The temperature of the sample immersed in a liquid
nitrogen bath is monitored with an argon gas thermometer. A
schematic diagram of the apparatus is shown in Figure 9.
The procedure followed to determine the necessary tempera-
ture, pressure, and volume data was to introduce a known amount of
gaseous adsorbate to the sample enclosed in a known volume, allow
the sample to adsorb the gas until equilibrium pressure was reached,
and measure the pressure. Since the temperature and volume of the
system were known, it was possible to calculate the amount of
adsorbate present after equilibrium had been reached using ideal
gas laws. Then the amount adsorbed at the measured pressure could
be determined from the difference between the known, original
amount of adsorbate and the amount remaining unadsorbed at that
particular equilibrium pressure. This procedure was repeated for
several different values of pressure such that the reduced or
relative pressure (equilibrium pressure divided by saturation
vapor pressure) varied in the range of .001 to .35.
The metal oxides were dried in a vacuum oven, and the
samples to be determined were weighed by difference using a
Mettler Type B6 analytical balance. The samples were weighed
into sample tubes constructed so that bumping of the sample
during outgassing would not cause it to be blown into other
parts of the system. The samples were outgassed overnight
until the system pressure, as measured with an ion gage (NRC
-5
Model 74l),was about 10 torr.
A computer program, SSAREA, was written at TRACOR to
calculate V., p., and p from input pressure, temperature, and
~ ~ .oi
volume data and, using a least squares technique, to find the
best straight line with ordinate y. = Pi and abscissa
~ V. (p -p . )
~ o. ~
~
55
-------
V'1
<7'
:x
»
<7'
I
N
I
N
\0
a
DIFFUSION
AND ROUGH
PUMPS
SAMPLE SAMPLE
PORT ~ORT
~ ARGON GAS
THERMOMETER
PRO BE
VOLUMETRIC BULBS
HIGH PURITY GAS
HELIUM
NITROGEN
ARGON
ION GUAGE
TUBE
A
S
LIQUID
NITROGEN
TRAP
2
B
MANOMETER FOR
ARGON THERMOMETER
REFER-
ENCE
MARK
MANOMETER
INDICATING
SYSTEM
PRESSURE
MERCURY
RESERVOIR
K NITROGEN
VACUUM
FIGURE 9 -
NITROGEN
VACUUM
DETERMINING SPECIFIC
NITROGEN
VACUUM
SCHEMATIC DIAGRAM OF APPARATUS FOR
SURFACE AREA USING THE BET METHOD
THERMOCOUPLE
GAUGE TUBE
D
-------
~ 6500 TRACOR LANE, AUSTIN. TEXAS 78721
Pi
x :::::-p
i 0io
adsorption
An example of the output for ziroconium oxide, whose
isotherm was given in Figure 8, is shown in Table VII.
3.3.4
Differential Thermal Analysis
Differential thermal analysis (DTA) was used to determine
the calcination temperature of intermediate compounds if the
temperature was not already known. It was also used in some in-
stances to check for completeness of calcination of an intermediate
compound.
A TRACOR LB-202 Recorder-Controller and JP-202 Furnace
Platform were used to make the measurements. The results obtained
were in the form of difference in temperature between sample and
an inert reference (alumina) as a function of system temperature.
The decomposition reactions of interest appeared as endothermic
peaks.
3.3.5
Chemical Analysis
The metal oxide samples were analyzed by atomic absorption
and emission spectroscopy to determine the impurities present. The
analyses were done by Southern Spectrographic Laboratory in Dallas,
Texas.
The results are summarized in Table VIII, which has been
reprinted by permission of Southern Spectrographic Laboratory. The
only compound having significant impurities was aluminum oxide
with about 1.2% zinc. The impurity was probably only a physical
mixture of 2nO with the A1203 caused by incomplete cleaning of
the sieve.
3.4
3.4.1
KINETIC MEASUREMENTS
Apparatus
I
A TRACOR TGA-3C thermal balance was used to make the
isothermal kinetic measurements. In this instrument the sample
holder is suspended from a spring. A photocell detector activates
a servo motor and linear actuator to maintain a null condition
57
-------
TABLE VII
DETE~MINATION OF SPECIFIC SURFACE AREA BY THE 5 E T METHOD
170-325
.ZR02634-63-1
MESH, OUT GASSED AT 6SC
AM?L: OES~R!PTION
gET -EQUATION:
?/V(PG-P) = «C-l)/VM*C}P/PO + I/VM*C
SA,""1PLE: R::OUCEO VOLUME AI)SC~ Rt:D PER GRAM P/V(PO-P) NO~~~ALIZED
TEr'i?~R'tTUF/£ PRESSUi~E: P/PO SAM?L:: AT .STP (GIf"!L ) WE I G :-; T I t~ G
(O~ G. K) (MM HGJ CML/GJ FUNCTION
OBSERVED CALCULATED OBSERVED CALCULA TED
77.'; .:::145 I.GO::> 1.040 .04815 .0460U 1.0~
77.5 .053 1.O~7 1.Ctl .05941.;. .05 H; 7 I.DC
77..4 ' .08D 1..Q73 1.0'32 .08093 .07947 I.Ge
77.4 . leI 1.117 1.123 .lG002 .10028 1.GU
77.5 .127 1.160 1.159 .12496 .12512 l.ee
\J1 77.4 .14 '3 1.207 1.192 .14532 .14723 1 . DC
00
77.4. zr.'i 1.300 1.270 .19174 .19c.38 1.00
. u....
77.4 .224 1.330 1.310 .21545 .21976 1.CO
77.4 .265 1.353 1.386 .25925 .2E061 1.CO
77.4 .279 1.4G6 1.412 .27516 .27393 1.0;::
77.4 .292 ' 1.411 1.440 .29288 .28705 1.00
VOLUME ~BSOR6ED PER GR ;1.M SH1PLE IN MOLECULAR J-:ONOLAYER (VM) = 1.02 tK/G
RELATIVE LIFETIMES OF MOLECULES IN FIRST AND HIGHER LAYERS (CJ =
f.77.24
CC~RELATION ERROR FOR P/V(PO-P) =
.00268
C-1/VM*C =
.9763
l/Vt",*C =
.00144
RMS ERROR FOR VOLUME AQSOR2ED =
.02
ERROR IN C-I/VM*C =
.G093
ERROR IN l/VM*C =
.00115
ERROR IN VM =
.01
SPECIFIC SU~FACE ~REA =
4.35 . V~
5P~CIFIC SU~FACE AREA =
4.45
S~UARE METE~S/GRAM
ERR~R IN
SPECIFIC
SURFACE AREA =
.04
-------
PHONE FEDERAL 1.3243
TABLE VIII:
RESULTS OF CHEMICAL ANALYSIS
Southern Spectrographic laboratory
P. O. BOX 6014
2824 N. WESTMORELAND
DALLAS. TEXAS 75222
"ay 8, 1969
Tracor
6500 Tracor Lane
Austin, Texa. 78721
IILI69-M-68
. '.0. 11149
Sample: Nine Oxide.
ZnO Cr20] $n02 CO]04 .UO '8203 A1203 Sb205 CuO
Cu 37 18 74 146 32 458 88 119 X
NI 10* 10* 525 1 ,170 X 126 10* 280 10*
Sn 10* 127 X 10* 10* 90 90 10* 10*
Co 327 10* 10* X '127 72 10* 10* 10*
AI 10* 10* 835 20 75 10* X 10* 10*
Zn X 17.6 47 57 39 118 12,200 It4 50
Ag 43 ---
Sb 10 10 765 X
Fe 43 48 40 15 X 76 272 270
Mn --- 11 --- 1.595
Mg --- --- --- 727 14
K --- --- 2.350
Cr 10* X 10* 10* 10* 10* 10* 10* 10*
Note: All Values ppm
~.( - 1 ess than SOUTH IU8 IRetaOlIAPHe C LABORATORY
~ ':$- /?/.~ J
~'"r
'Y..~ I. I .
/ /. I.A'.I'
V'V 1" .
Don McAlpin
59
-------
TABLE VIII (Cont'd.)
SOUTHERN SPECTROGRAPHIC LABORATORY
P. O. lOX 6014 . 2814 N. WESTMORELAND. DALLAS. TEXAS 75222
FEd.ral I-J14J
'''' 1 of ..
Tr.cor
6500 Tr.cor L...
Au.tlngT.x.. 78721
Re: '.0.1 II'"
Report of Semi-quantitative Analysis No. "-196
, Date: 518/6'
No. . No. No.
ZnO CuO Cr203
Over 10% In Cu Cr
--- --- ---
1-10'ft. F
--- Sn
0.1-1.0'fi. F
- ..
Co ---
0.01-0.1 'ft'
Ib, 'b, Cu II, 'b, Cu, ,. Sb, 'b, Cu, 'e
0.001-0.01 ~7,
S I, "I, In, AI, AI, "I. In. AI, At, In, SI. "I, AI, A9. .1, C.
less than 0.001 ';~ .1, '.. c., Cr c.. Cr
Semi-quantitative analysis is a preliminary, low-cost supplement to more refined methods. The basic definition
of semi-quantitative is a "factor of three", denoting that the reported concentration may be one-third to three
. times the amount present. SOUTHERN SPECTROGRAPHIC LABORATORY reports in "ranges of ten". This
indicates an accuracy exceeding the basic definition, reflects the greater accuracy attainable at low concentra-
tions, and relates the increased deviation with increased concentration.
. ,
SOUTHERN SPECTROGRAPHIC/LABORATORY; .'
. . . /<"/.:~~.- /"/~~..{/' (J"~~'--:: . ". '.' .
Chief SpepttotMpti'et'".~--. . . ,,' " '.
This report, or any part thereof, does not imply an endorsement, and may not be reproduced or used for advertising,
unless expressly pennitted by Southern Spectrographic Laboratory.
60
-------
TABLE VIII (Cont'd.)
SOUTHERN SPECTROGRAPHIC LABORATORY
P. O. lOX 6014 . 2824 N. WESTMORElAND. DALLAS. TEXAS 75222
FEderal 1.3243
,... 2 of It
Tr.cor
Re: '.0. 11149
Report of Semi-quantitative Analysis No.
69-196
Date: 518/69
I
I
No. No. No.
Sn02 Co304 HIO
Over 10% In Co HI
--- --- ---
1-10 ',1,.
--- .81 It
0.1-1.0'~,
"
0.01-0.1'/t, In, AI, tn Cu Sb
0.001-0.01 ')\ II, 'b, Zn, '., Cu, AI 'e, II, 'b, AI, AI, F., 51, Pb, AI, Cu, Zn
Zn, ""
Ib, "., C., Cr "I, In, c., C, In, AI, Co, C., Cr
less than 0.001 ')~
Semi-quantitative analysis is a preliminary, low-cost supplement to more refined methods. The basic definition
of semi-quantitative is a "factor of three", denoting that the reported concentration may be one-third to three
times the amount present. SOUTHERN SPECTROGRAPHIC LABORATORY reports in "ranges of ten". This
indicates an accuracy exceeding the basic definition, reflects the greater accuracy attainable at low concentra-
tions, and relates the increased deviation with increased concentration,.
SOUTHERN SPECTROGRAPHIC ,LABORATORY . -<:'.';,
,....':: \.'.":":,
" /"i (
Chief SpectrOgr~~~~~~~, '::
.'."" , .
This report, or any part thereof, does not imply an endorsement, and may not be reproduced or used for advertisinr, '
unless expressly permitted by Southern Spectrographic Laboratory. "
~. "
61
-------
TABLE VIII (Cont'd.)
SOUTHERN SPECTROGRAPHIC LABORATORY
P. O. BOX 6014 . 2824 N. WESTMORELAND. DALLAS, TEXAS 7S222
FEderal 1-3243
11101 'a98 3 of ..
Tr8cor
Re p . 0. '1149
Report of Semi-quantitative Analysis No. 69-196
Date: 5/8/69
No. No. No.
Sb20S Fe20, 1120'
Over 10 'ii, Sb Fe AI
--- -..- Zn
1-10 ',7" tn
-
--- "g
0.1-1.0'fi, F
51, Cu, NI, Fe ---
0.01-0.1 ',7"
0.001-0.01 ',1, AI, Mg, In, Pb, AI 'b" ln, NI, Sn 51, Pb, Cu, Fe, Sn
In, AI, Co, Ca, Cr 51, AI, Ag, Co, Ca, 'Cr Ag, Ca, Cr
less than 0.001 'k
Semi-quantitative analysis is a preliminary, low-cost supplement to more refined methods. The basic definition
of semi-quantitative is a "factor of three", denoting that the reported concentration may be one-third to three
times the amount present. SOUTHERN SPECTROGRAPHIC LABORATORY reports in "ranges of ten". This
indicates an accuracy exceeding the basic definition, reflects the greater accuracy attainable at low concentra-
tions, and relates the increased deviation with increased concentration.
SOUTHERN SPECTROGRAPHIC LABORATORY
...., ( .'
. /'" ...tI~.. ~ -...'
, L . '~,".' ,,:' .~....." ...~\:." ,- ~'
Chief Spectrograpner'''''''v,,-' '_.
This report, or any part thereof, does not imply an endorsement, and may not be reproduced or used fOr advertising,
unless expressly permitted by Southern Spectrographic Laboratory. . .',
6 2
-------
TABLE VIII (Cont'd.)
SOUTHERN SPECTROGRAPHIC LABORATORY
P. O. lOX 6014 . 2824 N. WESTMOItELAND . DALlAS, TEXAS 75222
FEderal I.J24J
'898 It of 4
Tr8Cor
Re:'.O. 11149
Report of Semi-quantitative Analysis No. 69-196
Date: 518/69
No. No. No.
eeo! zre
Over 10% c. Zr
--- ---
1-10';',
--- Zn
0.1-1.0 ',h. F
.'
---
0.01-0.1 ';'.
0.001-0.01 ',1, Ir, lb. Cu, '.. An II, 'b, In, Cu, ,.
"9, AI. Cr .~ AI, AI. Co. Ca. Cr
':
less than 0.001 '/l
Semi-quantitative analysis is a preliminary, low-cost supplement to more refined methods. The basic definition
of semi-quantitative is a "factor of three", denoting that the reported concentration may be one-third to three
times the amount present. SOUTHERN SPECTROGRAPHIC LABORATORY reports in "ranges of ten". This
indicates an accuracy exceeding the basic definition, reflects the greater accuracy attainable at low concentra-
tions, and relates the increased deviation with increased concentra~!on.
SOUTHERN SPECTROGRAPHIC LABOR A TORY
: .../...'.~1.
" , ,'""" ( /<'
Chief S~ctro~~pl}er.(:/(",' ,,"" . . .
. . i. ;....;"...r,"'~ "L..-"\.~."'-"'. ..
This report, or any part thereof, does not imply an endorsement, and may not be reproduced or used for advertising,
unless expressly pennitted by Southern Spectrographic Laboratory. .
.. .
r '
63
-------
ilaiIii 6500 TRACOR LANE, AUSTIN. TEXAS 78721
whenever a weight change occurs. The linear actuator movement
causes the displacement of the core of a linear variable dif-
ferential transformer (LVDT). This displacement results in a
signal which is displayed as the weight change. The instrument
is shown schematically in Figure 10.
The most important feature of the TGA-3C for this type
of rate measurement is that the gas is forced past the solid. The
gas enters the hollow connecting rod between the spring and sample
holder and passes down through the sample exiting from a porous
frit or 400 mesh stainless steel screen at the bottom of the
sample holder. Thus the gas composition actually surrounding the
solid particles is known. The measured reaction rate is not
limited by the rate of diffusion of the gas through the bulk solid.
The temperature was measured with a Pt-Pt/lO% Rh thermo-
couple placed about 1 rom from the sample holder. The cylindrical
lnconel sample holder is 1.3 cm in diameter and 2.8 cm long.
A TRACOR LB-202 Recorder-Controller was used with the
TGA-3C to control and record the temperature and to amplify and
record the weight change signal The weight gain vs system temper-
ature was recorded on a Moseley X-Y recorder. The OOC reference
point was set electronically. A Hew1itt-Packard model 17504 A
strip chart recorder provided the isothermal weight gain vs time
record.
The gas mixing and control apparatus used to supply the
simulated flue gas to the sample has been described in T.M. 004-
009-Ch20. This apparatus is capable of producing a wide range of
gas compositions. Flows-and pressures are reproducible and can
be obtained without affecting the flow through the thermal balance
system. Gas analysis equipment is included to provide an indepen-
dent means of determining compositions and can be used without
disturbing the flow to the experimental equipment. A schematic
diagram of the gas mixing apparatus is given in Figure 11.
64
-------
TRACOR TGA-3C S£HEMATIC DIAGRAM
RECORDER
LIGHT--+O
SERVO MOTOR
LINEAR ACTUATOR
cr-PHOTOCELL
SLIT
GAS IN
. SAMPLE HOLDER
FIGURE 10
65
A6-164-36
-------
,---------------,
I I
I SC-I SC-2 I
I @ @ J
I I-~=~---I
I I CD
I @ I I - ~m1 I
I I X CDtTC\ ~ I
~I ~- \V'l I
I \.Y I I ~ - ----.I I
I G) I i x G2_m_{E) I
g: I :G) I L-- _REACTO~_~
I ~ I
I \.Y I
I CD CD CV (1) I
I " ", 0, t~ :
L GAS MIXING so
------- -_L--.J
FIGURE 11 - SCHEMATIC DIAGRAM OF GAS MIXING APPARATUS
I~
it
>-
'"
I TO DRAIN
REGULATED AIR I
I
I
SCRUBBING I
I
I
I
I
I
----_--.-J
j.,ItIII'I'/01+'6.86-/64-5 /
2/17/69 JM/I.OW£LL
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
3.4.2
Experimental Procedure
The particle size of all of the samples studied was
-170 +325 mesh (44-88 microns). Usually a 10-50 mg sample was
mixed well with 1.0-1.1 g of inert alumina in the sample holder.
The weight of sample and alumina were determined by difference
using a Mettler Type B6 analytical balance. The alumina used as
an inert had been fired and did not react with 8°2. Dilution of
the sample kept the particles separated and thus provided good
gas-solid contact and thermal homogeneity.
The average total flow to the sample was 250 cm3/min.
The sample size was kept small enough that the reaction rate was
not limited by the rate of 802 supply. The experimental condi-
tions used corresponded to a differential reactor.
The temperature range of interest for each compound was
determined by making a linear temperature programmed run using a
simulated flue gas atmosphere. The composition of the atmosphere
is shown below.
Component
Mole Per Cent
8°2
CO2
°2
H20
N2
0.3
14.3
3.4
2.0
balance
The sample was heated at 10°C!min until a measurable
weight gain occurred. This experiment indicated at what tempera-
ture the reaction of the metal oxide with 802 "started", or became
fast enough to give a measureable weight gain, and at what tempera-
ture the product began to decompose.
Two or more isothermal runs
within the range as determined above.
calibrated before each run by hanging
were then made at temperatures
The thermal balance was mass
known weights on the sample
67
-------
il:iIJ:ii:iJ;i 6500 TRACOR LANE, AUSTIN, TEXAS 78721
'holder and adjusting the amplifier. A fresh sample was used for
each run. The sample was heated in a nitrogen atmosphere at 20oC/min
until the chosen temperature was reached. The temperature was then
held constant by the Recorder-Controller. After thermal equili-
brium was established, H:aO, CO:a' and 0:a were admitted to the gas
stream. When a straight baseline (zero weight change) was obtained
on the strip chart recorder, SO:a was added to the gas mixture.
The atmosphere thus obtained was the same as that given above.
This procedure ensured that the weight gain measured would be due
to sulfite or sulfate formation rather than carbonate or hydroxide
formation or oxidation of the metal oxide.
Zero time for the reaction was taken as the point at
which the weight gain began. The run was continued until there
was little or no further weight gain. The temperature range of
the reaction was recorded on the X-Y recorder as a horizontal of
the weight change ~ temperature graph. In most cases there was
no more than + loC variation in temperature. The SO:a concentration
of the stream was determined during each run by titration of a
known amount of iodine (Reich's test).
The procedure used for the determination of the rate
dependence on the partial pressure of SO:a was the same as above
except that the SO:a concentration was set at 0.12 mole % rather
than 0.3 mole %.
3.4.3
Data Analysis
The data obtained from these kinetic measurements were
in the form of weight gain, w, as a function of time. The extent
of reaction can be better represented as the fractional conversion
a = w/wo' Ideally Wo is the theoretical stoichiometric weight
gain. In practice it was found that often the stoichiometry of
the reaction was unknown or that the reaction rate became very
slow before the stoichiometric weight gain was reached. In these
cases the highest ratio of maximum weight gain to initial sorbent
68
-------
.
1-- --
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
weight, me' observed for a given compound was used to calculate
We. The calculations for the other runs of the same compound were
then based on the ratio wo/me which is given in Table IX in Section
3.5.
The rate of product formation can be defined as ~~. In
isothermal conditions the reaction rate should depend only upon the
fraction reacted and the partial pressure of S02. This is expressed
in Equation (13) where k is the rate constant and g(P ) is the
sOa
~
at
=
k f(O') g(P . )
S02
(13)
dependence on S02 partial pressure. Equation (13) can be written
as Equation (14), since P was constant in the experiments.
S02
dO'
crt
=
k' f(O')
(14)
The analytical form of f(O') depends upon the mechanism
of the reaction. For our purposes the general form
f(O')
= (l-O')n
(15)
has been most often used. In Equation (15), n is the order of
reaction with respect to the solid reactant. The values 0, 1/3,
1/2, 2/3, 1, 3/2, and 2 have been used for n. The form derived
by assuming that the rate is controlled by diffusion of the gaseous
reactant through the solid product to the unreacted core has also
been used. This form, given in Equation (16), is referred to as
DCE for Diffusion Controlled Equation.
f(O')
=
(1-0')1/3
1 - (1-0')1/3
(16)
69
-------
i1J/jJ;jjjJ;jj 6500 TRACOR LANE,
AUSTIN. TEXAS 78721
The basis for these various forms of f(a) will be dis-
cussed in Section 3.5.
The rate equation was applied to the data in its inte-
grated form. The integrated form of f(a) is given in Equation (17).
l(a)
=
SO( d
o f (~)
=
k' t
(17)
The integrated forms of f(a) are listed below for each
value of n.
n 1(0()
0 0(
1/3 3/2[1 - (1-0() 2/3J
1/2 2[1 - (1-0() 1/ 2J
2/3 3[1 - (1-0()1/3]
1 -In (I-a)
3/2 2 [1 1J
(1-a)1/2 -
2 a
I=C;
DCE 1/2[1 2/3J
- 3(1-0() + (1-0()
The 1(0() which corresponds to the proper or "best" f(O()
should yield a straight line with slope k' when plotted versus.
time. The Univac 1108 computer was used to calculate the values
of 1(0() for each value of n from the weight gain data. About
70
-------
1---""--
iIJ1iJ:iillJi 6500 TRACOR LANE, AUSTIN, TEXAS 78721
50-100 weight gain-time points were taken from the strip chart
recording for each isothermal run.
The computer first plotted the values of each form of
I(~) versus t. The value of n which gave the best straight line
was determined by inspection. The slope k' of the least squares
regression line for the "best" n was computer calculated and
the data and line were again plotted by the Calcomp plotter. An
example for copper oxide with n = 1/3 is given in Figure 12. The
plots of I(a) versus t for each of the compounds investigated
are given in Section 8.2.3.
It was found that the S02 partial pressure dependence
of the reaction rate could be satisfactorily represented by Equa-
tions (18) and (19).
g(Pso )
2
=
pm
sOa
(18)
or
o~
ot
=
k f(~) pm
sOa
(19)
The exponent m is the reaction order with respect to
SOa' This quantity was determined by calculating k' for two runs
at the same temperature but different SOa partial pressures and
noting that
k. '
1
=
k PI?
1
sOa
(20)
where m =
log (k1 '/k';l ')
log (Pl. / P 2 ).
sOa sOa
71
-------
~ ]
~ [f-U-a )3
'-I
N
.50
.40
.30
.20
.10
00
4 8 12 16
TIME, MINUTES
20
REPRESENTATION OF RATE DATA FOR COPPER OXIDE AT 404°C
SLOPE = RATE CONSTANT (/()
J>
(J'\
I
-
(J'\
+-
I
-. ~
CD
FIGURE 12
~.
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
(20).
Once m was known,
In these experiments
k
p
sOa
of SOa since the sample holder
was easily calculated from Equation
(in atm) was numerically equal to
the measured concentration
atmospheric pressure.
was at
3.4.4
Input to Economic Studies
The data obtained from kinetic studies must be used to
describe the reaction rate in the fluidized bed model used for
economic calculations. The rate quantity of interest in the fluid-
ized bed calculations is the rate at which the SO concentration
a
can be lowered to a specified level as the flue gas passes through
the bed at a specified flow rate. This rate will be a function of
the steady-state degree of conversion or loading of the well-mixed
solids in the bed. Thus the fluid bed design was based on the
following equation for the rate of SOa removal given by Kunii and
Levenspie1 (KU-007).
1
- Vs
d N
SOa
dt
K
r
C
S02
(21)
In Equation (21), V is the volume of sorbent, N is
s . S02
the number of moles of S02 removed, Cs is the concentration of
02
S02 in the flue gas, and Kr is the rate constant which is a function
of the extent of sorbent conversion. Dependence on S02 concentra-
tion was found to be linear from the experimental studies.
The rate constant K in Equation (21) can be related to
r
the experimentally determined rate constant k by noting that the
amount of S02 removed from the flue gas equals the amount sorbed
by the solid where M is the molecular weight of S03 and w is the
d N
S02
dt
=
1 dw
M crt
=
V
s
K C
r sOa
(22)
73
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
weight gain of S03 (for sulfate formation).
P /RT for C ,Equation (21) becomes
so:;:! so:;:!
If we substitute
dw
crt
=
M V
s
K
r
P
~.
RT
(23)
The rate equation used in the experimental studies is
given in Equation (24).
dQl
dt
=
1 dw
Wo crt
=
k f(QI) P
so:;:!
(24)
Combining Equations (23) and (24), noting that V = me/p ,
s s
and converting k from min-l to seC-l with the factor 60 yields the
required expression for K .
r
K
r
=
k f(OI) Ps RT we
60 M mo
(25)
In Equation (25) mo' wo' and f(QI) have been defined in
Section 3.4.3, Ps is the density of the sorbent (g/cm3), R is the
gas constant (cm3 atm mole-1 °K-1), and T is the temperature (oK).
The value for the extent of conversion (QI) used in f(OI) for the
K calculation was determined by inspection of the kinetic data.
r
A large value of QI is desirable for obtaining low sorbent recircu-
lation rates and for efficiency. However, the reaction rate often
decreases drastically at higher values of QI. Thus an QI was chosen
which was just below the region in which the reaction rate became
slow.
In summary, the values input
were the rate constant K , temperature
r
conversion (sorbent loading).
to the Economic Studies
of reaction, and extent of
3.5
RESUL TS OF EXPERIMENTAL PROGRAM
The quantitative results of the experimental program are
given in Table IX. A discussion of these results and the quali-
tative findings is given in the following paragraphs.
74
-------
TABLE IX:
RESUL TS OF THE EXPER IMENTAL PROGRAM
BET KJNET IC DATA FOR SO, SORPT ION SO, SORPT ION
X-RAY SURFACE PROOUCTS,
PREPARA T ION OIFFRACT ION AREA ~ ,0 FROM X-RAY
COMPOUNO METHOO PATTERN "'9~ T('C) k m PSO, b kO Ed A' 0 I FFRACT ION
COPPER THE OXIOE WAS PRE- THE PATTERN AGREED 52,04 111 l. 01 325 0,00209 0.313 0,668 26.94, 27XIO' C. SO, ,C.O
OX I DE PAREO BY PRECIPITA- WITH THAT OF CuO, 352 0.00642 0.299 2.15 ~~~g'a;. ~~~ 0
634-47-1 TION OF THE HYDROXIOE THE PEAKS WERnROAD, 355 0.004~ 0.316 1. 50
FROM THE SULFATE, INDICATING SHALL 376 0,011 0.308 3.70 FORMED ON
THE HYDROXIDE WAS CRYSTAL SIZE OR 404 D,02850 0.306 9,31 CDDLI NG
HEAT EO IN THE VACUUM AMORPHDUS CHARACTER. 482 O. 11100 0,314 35.4 SAMPLE AFTER
OVEN AT 55'C FOR 44.5 404 0,01120 0,122 9,18 RUN),
HOURS.
CO BA L T THE OXIDE WAS PRE- THE PATTERN AGREED 12,05 0.380 453 0.0200 0.318 6.29 19,4 4.36Xl0' P.Co SO,
OXIDE PARED BY CALC INATION WElL WITH THAT FOR 544 0.0916 0,326 28, I Co,O,
634-43-1 OF THE N 11 RATE FOR Co,O.. THE PEAKS 453 0.00864 '0.129 6.70
3 HOURS AT 350'C WERE SHARP AND WELL 0.931 453 0.00671 0.318 2,11 19.2 1.24X10'
AND 3 HDURS AT 100'C DEF INED. 544 0.0302 0.326 9,26
IN VACUO.
COBALT CoO'OH WAS PRECIPITA- THE PEAKS WERE 169,1 0.931 359 0.00566 0,321 1. 76 13.9 1.6Xl0' t/-CoSO,
OXIDE TED FROM Co(NO,), BROAD AND NOT WELL 430 0.0175 0.325 5.38 Co,O,
634-71-2 SOLUTION AND CAL- DEFINED. THE
CINED FOR 2.5 HR. PATTERN AGREED WELL
AT 16D'C IN VACUO. WITH THAT FOR
Co, 0,.
TIN OXIDE THE HYDROXIDE WAS NO DIFFRACT ION PAT- 4,04
634- 55-1 PRECIPITATED FROM TERN WAS OBTAINED
Sn C I. AND HEATED IN US ING THE
THE VACUUM OVEN FOR DIFFRACTOMETER.
31 HOURS AT 60'C,
THE OXIDE WAS
CALCINED AT 150'C
FOR 3 HOURS.
ZINC ZINC HYDROX I DE WAS DIFFRACTION PATTERN 16.46
OXIDE PREC IP ITATED FROM AGREED WELL WITH
634-51-1 THE SULFATE AND THAT FOR ZnO.
HEATED IN THE PEAKS WERE SHARP
VACUUM OVEN FOR 23 AND WELL DEFINED,
HOURS AT 60'C. THE
OXIDE WAS CALCINED
FOR 1.5 HR. AT 145'C
AND 1 HOUR AT 100'C.
NICKEL THE DXIDE WAS PRE- DIFFRACTION PATTERN 7.02
OXIDE PARED BY CALCINING AGREED WEU WITH
634-41-1 THE NITRATE FOR 1 THAT FOil NIO. PEAKS
HOUR AT 300'C AND WERE SHARP AND WELL
45 MINUTES AT DEF INED.
380' C,
NICKEL THE OXIDE WAS PRE- THE DIFFRACT ION PA T- 63.44 1/2 1.07 380 0.D00764 1 0,322 0.232 1'3. 5 7,81 XI O' N I SO.
OXIDE PARED BY PREC I P 1- AGREED WEU WITH 500 0.00391 0.333 1. 17 NiO
634-65-2 TATING THE HYDRO- THAT OF NIO. THE 501 0,00157 O. '24 1. 27
XIDE FROM THE NI- PEAKS WERE BROAD.
TRATE. THE HYDRO-
X I DE WAS HEATED
FOR 18 HOURS AT
70'C IN VACUO, THEN
HEATED IN AIR AT
330' C FOR 2 HOURS,
IRON THE OXIDE WAS PRE- THE PATTERN HAD VERY 225.56 1/3 0.833 342 0.00572 0.309 1,85 9.16 3. 34Xl O' PATTERN OB-
OXIDE PARED BY PRECIPITA- BRDAD, UNEASILY DIS- 400 0.0108 0.306 3.53 TA INED WITH
634-45-1 TION OF THE HYDRO- T I NGU I SHED PEAKS. 4DO 0.00368 0.120 3.07 DIFFRACTOMETER
XIDE FROM THE CHLOR- PRO BA BL Y Fo,O.. I NCONCLUS I VE,
I DE. THE HYDROXIDE PEAKS BROAD
WAS HEATED AT 55'C AND NOT WELL
I N VACUO FOR 39 DEF I NED.
HOURS, AT 120'C IN REGENERAT ION
VACUO FOR 1.5 HOURS PROOUCT CONTAINED
AND AT 170'C IN Fe, 0..
VACUO FOR 2 HOURS.
ALUMINUM THE OXIDE WAS PRE- THE PATTERN OBTAINED 322.54
OXIDE PARED BY PRECIPITA- WITH THE DIFFRACTO-
634-53-1 TION OF THE HYDRO- METER HAD NO PEAKS.
XIDE FROM THE CHLOR-
IDE. PRODUCT WAS
DR I ED I N VACUO FOR
~~EHgmEA~A~O~h-
CINED FOR 3 HOURS AT
280'C.
ZIRCON IUM THE OXIDE WAS PRE- THE PATTERN OBTA INED 4'.45
OXIDE PARED BY PRECIPITAT- WITH THE DIFFRACTO-
634-63-1 ING THE HYDROXIDE MET ER HAD NO' PEAKS.
FROM THE SULFATE.
THE HYDROXIDE WAS
DRIED IN THE VACUUM
OVEN AT 60' C FOR 24
HOURS AND THE OXIDE
WAS CALC INED AT
400' C FOR 2 HOURS.
ANTIMONY THE HYDROXIDE WAS THE PATTERN AGREED 18.28
OX IDE PRECIPATED FROM THE WELL WITH THAT OF
634-67-' PENTACHlORIDE, DRIEO Sb,o" THE PEAKS
AT 70'C FOR IS HOURS WERE BROAD AND NOT
AND CALC INED AT WELL DEF INED.
275'C FOR 2 HOURS,
UN ITS: a. min-I b.otm . 10' c. min-I otm-m d. keal / mol, e. min-' alm-m
75
-------
COMPOUNO
CERIUM
OXIOE
634-37-1
CHROMIUM
OX I OE
634-59-2
TITANIUM
OXIOE
634-69-1
VANAOIUM
OX I OE
634-73-1
TUNGSTEN
OXIOE
634-61-1
PREPARAT ION
METHOD
THE HYDROXIDE WAS
PRECIPITATED FROM
AMMDNIUM HEXANI-
TRATE CERATE
r(NH,), C. (NO,)J
A'ND DRIED IN VACUO
AT 60'c.
THE OXIDE WAS PRE-
PARED BY CALC INA-
T ION DF (NH,), C,O,
AT 180'c FOR 15
MINUTES.
HYDRDl YS I S OF Ti CI,
WITH (NH,), SO,
SOlUT ION.
CAlCINAT ION AT
I} 350'C 2 HR.
2) 500'C 2.5 HR.
3} 600'c 3 HR.
NH, VO, SOlUT ION +
D Il. H, SO, THEN
1) DRYING AT 90'C
IN VACUD FOR
-15 HR.
2) HEAT ING AT 450'C
FOR 2 HR.
CAlCINATIDN OF
H,WO, AT
I) 200' C - 2 HR.
2) 3DO'C - 3 HR.
TABLE IX: EXPERIMENTAL RESULTS (CDNTINUED)
X-RAY
DIFFRACT ION
PA TT ERN
THE DIFFRACT ION PAT-
TERN WAS QUITE BRDAD
AND THE P£AKS ALMOST
UNDEFINED. IT
AGREED WITH THAT DF
C.O, .
THE PATTERN AGREED
WEll WITH THAT OF
C"D,. THE PRODUCT
WAS ALMOST
AMORPHDUS AS
INDICATED BY THE
BROAD PEAKS.
UNITS: o. min-I b.otmIIO. c.min-lotm- d.kcol/motl ..mln"'iotm'"
c'."".'"
BET
SURFACE
AREA
mlQ-1
192.66
66.52
k"
KINET IC DATA FOR SO SORPT ION
m
PIO.'
k'
.
E
0,
m. T('C}
0.251 1940.0426
1940.0122
196 D.0166
238 0.0345
2620.0504
"
0.315 13.5
0.127 9.60
0.190 8.74
0.318 10.8
0.320 15.8
7.892.56XID'
DCE 0.106 335 0.00257
390 0.00481
336 0.00105
D.319 0.8069.43 1.89X10'
0.312 1.54
0.122 0.861
A TEMPERATURE PROGRAMMED TGA WAS HADE IN FLUE
GAS ATMOSPHERE AFTER EACH CALC INAT IDN BUT NO
REACT ION OCCURRED.
SEE COMMENT FOR TITANIUM OXIDE
SEE COMMENT FOR TITANIUM OXIDE
76
A'
SO, SDRPT ION
PRODUCTS
FRDM X-RAY
DIFFRACT ION
PEAKS BROAD.
NDT WEll
DEF INED.
C.O, PRESENT.
POSSIBLY
C. (SO, ), . 4H, 0
EV I DENCE
INCONClUS IVE
DIFFRACT ION
PATTERN OB-
TAINED WITH
D I FFRACTOMET ER
INCONClUS IVE.
THE PRODUCT
APPEARS TO BE
AMORPHOUS.
-------
ll:i1iJ:iIiIii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
With the exception of CeOa' the metal oxides studied
showed low reactivity with SOa below a temperature of about 325°C.
In this case "low reactivity" means that the time necessary to
achieve a significant conversion would be on the order of hours
rather than minutes. CeOa was unique because sorption was rapid
even at 190°C. The behavior of CeOa was anomalous in other ways
which will be discussed later.
Either the reaction rates of the metal oxides other than
CuO, Fea03' CeOa' C030" Cra03' and NiO were very low or the oxide
did not react at all. The unreactive oxides studied were Ala03'
SnOa' ZrOa' ZnO, TiOa' SbaOs' WxOy' and VaOs. These oxides could
possibly be prepared in more active forms than the forms resulting
from our preparations. For those unreactive oxides which had
low specific surface areas, kinetics experiments were made with
large amounts of sample to give about the same surface area as for
the reactive compounds. Even under these conditions, the reaction
ra,tes were very low.' In most cases the reaction rates were too
low to make meaningful kinetic calculations. For these reasons
attention was concentrated on the reactive oxides and little data
are given for many of the others.
3.5.1
Rate EQuation and Reaction Orders
A relationship was sought between the reaction rate and
the extent of sorbent conversion and SOa partial pressure. Equa-
tion (26) was chosen for its flexibility, general applicability,
and simple representation of the reaction rate (see Section 3.4.3).
oa
at
=
k (l-a)n pm
S02
(26)
An alternate approach would have been to attempt to fit
the data for each of the metal oxides to each of the many equations
arising from idealized solid-gas reaction models. Even this method
often does not indicate which is the true reaction mechanism since
some of the equations obtained from these models g1ve similar
77
-------
i:IiIJ:JJ:iIi 6500 TRACOR LANE, AUSTIN, TEXAS 78721
results. The actual reaction mechanism may be intermediate between
several reaction models. In view of the difficulties involved
Wen (WE-002) has pointed out that there is no reason why simple
rate equations which fit the data should not be used as long as
no extrapolation beyond the reaction conditions investigated is
allowed.
Equation (26) corresponds to several models depending on
the value of the reaction order. When surface phenomenon controls
the rate of a solid-gas reaction, the order can be shown to vary
from zero to two depending on whether the gas reactant is strongly
or weakly adsorbed. For n = 2/3, m = 1, Equation (26) represents
the unreacted-core-shrinking spherical particle with constant
radius model with the rate controlled by the chemical reaction.
For n = 1/3, m = 1, the equation represents the unreacted-core-
shrinking spherical particle with constant radius model with the
rate controlled by fluid film resistance (WE-002). For n = 1, m = 1,
Equation (26) represents the continuous reaction for which diffusion
of the gaseous reactant into a particle is rapid enough compared
with chemical reaction rate that the solid reactant is consumed
uniformly throughout the particle. The values for n of 0, 1/2, 1,
and 2 have been considered by different authors for gas-solid
decomposition reactions (PR-002). In addition to the n + m order
rate equation, calculations were also made for the form of f(a)
which corresponds to the unreacted-core-shrinking spherical particle
with constant radius model with the rate controlled by gaseous
diffusion through the porous product layer as shown in Equation (27).
(1-a)1/3
f(a) = 1 ~ (1-a)1/3
(27)
In most cases the data were well represented by Equation
(26). The reaction rates were in general linear (m = 1) with
respect to 802 partial pressure under flue gas conditions (0.1 -
0.3% 802). The data for Ce02 indicated a higher value of m (~1.4),
78
-------
UiIJiiiIii 6500 TRACOR LANE. AUSTIN. TEXAS 78721
but the evidence was not conclusive. Therefore to maintain con-
sistency with the other metal oxides, the values for Ce02 were
calculated using m = 1. The value of n or f(a) which gave the
best fit was found to depend on the metal oxide. This is to be
expected since the reaction mechanism may be dependent on physical
properties and chemical reactivities which vary from one metal
oxide to another. In some cases the "best" apparent reaction
order varied with the extent of conversion of the sorbent. This
is also to be expected since the rate controlling process may
change during the course of reaction. For example the initial
rate of reaction may be limited by the rate of chemical reaction.
But after a layer of inert reaction product is formed, the rate
may be controlled by diffusion of the S02 through this barrier.
A value of n was chosen which gave the best fit up to a conversion
which would be sufficient for use in an S02 removal process. Since
the true reactivity is best shown in the initial stages of the
sorption process, using the value of n and the rate constants
obtained for the early stages of the experiments allows a good
comparison of the chemical reactivity of the oxides studied.
CuO, Fe:a03' and NiO had "low" reaction orders of 1/3,
1/3, and 1/2 respectively. Once again this does .not indicate a
particular model since the data could have been represented almost
as well for these compounds by n = 1/3 to 2/3. It does suggest
that the reaction rate was controlled either by chemical reaction
or fluid film diffusion or both. For C030. the rate was linear
(n = 1) with respect to extent of conversion. The reaction rate
for ceric oxide, with n = 2, was more strongly dependent on extent
of sorbent conversion than the other rates. The data for Cr203
fit the diffusion controlled Equation (27) after a very brief rapid
initial sorption step. The reaction became very slow after a
small extent of conversion. Examples of the rate plots obtained
are given in Figures 13-18.
79
-------
o
\f1
.
II
\J') 0
. :t'
~ .
+
r---
to
. 0
~ (11
'.'
* .
..-.
00 a:
0 :I:
a...
_,
a:
J 0
C'\J
a .
.
~
"-oJ
*
Lf)
.
~ 0
J ..,..
.
('II
('II
.
II c
:z 0
.
O~OO
18 RPR 69
RATE DATA FOR COPPER OXIDE
CUD 634-47-1
ISO TGA 4 TEMP = 352 OEG C
10.00
20.00
30.00
40.00
50.00
60.00
70.00
TIME (MINUTES)
Figure 13
80.00
jrRft.coRj
-------
RATE DATA FOR IRON OXIDE
FE203 634-45-1
ISO-TGR 4 TEMP = 342 DEG C
a
:;::)
.
C"\I
Lf) a
(Q .
. .
.....-4 -
, .
"+
['
(D a
..
}if N
.
* -
".-..
00 cr:
...... :r:
(L
--1
a: c
I
a CD
.
.
.....-4
'-J
*
Lf)
.
.....-4 C
1 =-
0'
('r')
...
11 c
Z 0 J I I I I I I I I
-
0.00 40.00 60.00 . 120.00 160.00 200.00 2~0.OO 280.00 320.00
u
..
JUN 69
TIME (MINUTES)
Figure 14
/fRRctJA/
-------
i
a
.
N
+
L/)
.
*
*
......,.
(X) a:
N :r:
. 0... .
-I.
a:
I
a
.
.......
"-J
~
a
.
N
I
o
L/)
.
1/
~
RATE DATA FOR NICKEL OXIDE
150
NIO 634-65-2
TGA 1 TEMP = 380 DEG C
o
('IJ
.
~.
tD
,....
.
N
. ,...
. ~
8
.
::!'
o
.
0.00
I
40.00
I
120.00
I
160.00
I
320.00
I
50.00
TIME (MINUTES)
18 APR 69
Figure 15
I .
200.00
I
2110.00
I
. 280.00
/TRRCOR!
-------
a:
:r:
.£1..
00 -J
~ a:
I
o
.....
-
Z
-J
I
C)
.....
II
Z
.
.
o
c
.
-'
. .
~.
C
lD
.
C
to
..
c
;:;J"
.
c
C"II
.
c
o
.
0.00
2Q MAY 69
.
..
..
..
. .
..
20.00
Yo.oo
RATE DATA FOR COBALT OXIDE
C030Ll 63Ll-71-2
ISO TGA 1 TEMP = 359 DEG C
o.
60.00
Bu. 00 .
. 120.00
100.00 .
TIME (MINUTES)
° Figure 16
1110.00
160.00
/TAAcoR/
-------
. -
I
RATE DATA FOR CERIUM OXIDE
CE02 63l!-37-1
ISO TGA 3 TEMP = 191 DEG C
CJ
C
.
":3'
--
o
tV
.
'"
-
,......
a:
J: CJ
a... ='
-.J .
a: N
J
00 0
.po ..
-
.......,
'- CJ
II to
.
J: ....
a...
-.J
a:
0 c
.. CD
N -.
II
:z
I
8.00
J
16.00
I
211.00
I
32.00
I
LJO.OO
I
LJ8.00
I
56.00
I
611.00
TIME (MINUTES)
.18 APR 69
Figure 17
jrRqCoAj
-------
a:
:c
a..-
.-1
a:
J
,-.
r"
to
-
~
~
--
a:
00 :c
U1 a..-
-I
a: .
J
a
.-
I
a
-
~
.
U1
-
.-
u
-
C1
RATE DATA FOR CHROMIUM OXIDE
CR203 634-59-2
ISO-TGA-2 TEMP = 335 DEG C.
C
::1"
"-,
N.
tr)
.
".
~
N
to
-
to
a
0.00
40.00
100.00
120.'00
60.00
80.QO
20.00
TIME (MINUTES)
1 L! RPR. 69
Figure 18
l~O.OO
160.00
j=rRRcoRf
-------
izi1IJ;iii1iI6500 TRACOR LANE, AUSTIN, TEXAS 78721
The rate plots are given in Section 8.2.3. The plots
used to obtain the rate constants are given along with the plots
for each f(O') at a single temperature. For CosO, prepared by
nitrate decomposition, plots are given both for wo/mo = 0.38,
which represents saturation conversion, and for wo/mo = 0.931,
which represents the theoretical conversion for CoSO, formation.
Some of the plots seem to oscillate regularly about the least
squares regression line. . No attempt was made to study this
behavior, but it may be of interest to future investigators. As
mentioned above, several of the plots show initial rapid SO uptake
2
immediately followed by a slower continuous weight gain. This is
probably due either to initial chemisorption or reaction of a thin
outer layer of solid.
3.5.2
Temperature Dependence of the Reaction Rate
The temperature dependence of the reaction rate was
another quantity of interest for design purposes. The form of this
relationship depends upon whether the reaction rate is limited by
the chemical reaction or by mass transfer (diffusion). The
Arrhenius Equation (28) can be used when there appears to be
strong temperature dependence.
k
=
Ae-E/RT
(28)
In this equation k is the rate constant, A is the frequency
or pre-exponential factor, and E is the activation energy. Due to
our incomplete knowledge of the reaction mechanisms, the use of
this equation must be interpreted for the present simply as an
empirical means of describing the temperature dependence. If the
rate controlling step is pore diffusion or bulk diffusion of the
gas into the solid, the rate should vary as Tl/2 and T3/2 depending
on which diffusion mechanism predominates.
The relative importance of chemical reaction and mass
transfer temperature dependence effects may change with temperature
86
-------
~ 6500 TRACOR LANE. AUSTIN. TEXAS 78721
and the effects may not always be separable. Therefore, the situ-
ation may become very complex. The sorption temperatures used for
economic calculations were the same as those at which one of the
kinetic runs had been made so that the measured rate constant could
also be used. The temperature dependence was strongest for CuO,
C030~, and NiO. The activation energies are all small (less than
27 kcal/mole), however. An Arrhenius plot for CuO is given in
Figure 19. The deviation at high temperature may be due to a transi-
tion from chemical reaction to diffusion limitation of the reaction
rate. Several runs for Ce02 at about 190°C gave higher rates than
those obtained for 240°C. The reaction may be occuring via different
paths at different temperatures. The temperature dependence tabu-
lated is for the 240-260°C temperature range only. The temperature
dependence for the six metal oxides is summarized in Table IX.
3 5.3
Physical Properties and Reactivity
The effect of compound preparation and resulting physical
properties on reactivity of a metal oxide with S02 was studied
briefly. The surface area and crystallinity of a sample provide
a good guide to relative reactivity of different preparations of
the same metal oxide. A COS04 sample prepared by calcination of
the nitrate had a surface area of 12 m2/g and was fairly crystal-
line, while a sample prepared from CoO.OH had a surface area of
169 ma/g and was nearly amorphous (see Figure 7, Section 3.3.2).
The high surface area, nearly amorphous preparation had a reaction
rate nearly three times that of the nitrate preparation. Such
comparisons. between different compounds must be done with care,
however. High surface area and amorphous character do not guarantee
reactivity. The AlaOs sample had a surface area of 322 m~/g and
was nearly amorphous according to the X-ray diffraction pattern,
but it was nevertheless a poor S02 sorbent.
3.5.4
. Conclusions
One of the aims of the experimental program was to select
the "best" metal oxides on the basis of their chemical reactivity.
87
-------
100
90
80
70
60
50
40
30
20
10
T 9
~ 8
~
T 7
~ 6
~
~
~ 5
4
3
2
1.0
0.9
,0.8
0.7
0.6
0.5
1.35
1.40
1.45
~
1.50
IOII/T(OK)
1.55
ARRHENIUS PLOT FOR CuO
Figure 19
88
o
1.60
1.65
-------
j;J;jfjJ;iiiIii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
The six reactive compounds are given in Figure 20. Of these, Cr203 and
NiO can be eliminated for the time being due to their slow reaction
rate once a thin layer of reaction product has formed. Thus,
these two have only a small practical S02 capacity. Co30. has a
fairly high reaction rate and S02 capacity. It was found, however,
that the reaction product was CoSO. which is not well suited for
thermal regeneration since its S03 partial pressure is fairly low
even at 800°C. The S02 capacity of CeOa was low, but its low
sorption temperature makes it attractive in that respect. The
reaction products could. not be identified, however, and more research
is needed before it can be recommended. CuO and Fe203 are both
potential sorbents. CuO has a high reaction rate and S02 capacity.
The reaction stoichiometry is well defined. Fe203 is somewhat
slower and our preparation had a lower S02 capacity than CuO. More
work is needed also to define the stoichiometry of the reaction.
4.
ECONOMIC FEASIBILITY STUDIES
4.1
INTRODUCTION TO ECONOMIC FEASIBILITY STUDIES
The economic feasibility studies involved equipment design,
equipment sizing and cost estimation. Process f10wsheets including
heat and material balances were also prepared. The estimation
scheme for determining the total capital investment and gross
annual operating cost is shown in Figure 21. The dry metal oxide
sulfur oxide removal process is composed of three steps: 1)
sorption, 2) regeneration, and 3) sulfur recovery. TRACOR esti-
mated the cost of the sorption and regeneration steps using a
fluidized bed contactor model with computer calculated parameters.
Cost information for the sulfur recovery step is to be supplied by
HEW.
4.2
SORPTION UNIT DESIGN
The major pieces of equipment used in the sorption step
are the sorber (see T.M. 004-009-Ch19), draft fan (see T.M. 004-
009-Ch12), and sorbent fines collector (see T.M. 004-009-Ch1SA).
89
-------
PERIOD la
f H
2
3
4
5
6
\.0 7
0
GROUP
VIII
vb vlb Vllb 0
H He
N 0 F Ne
P S CI Ar
Se Br Kr
Te I Xe
Po At Rn
III, ... . ....
111'111111111111
.-.y: : :-
pt Au Hg
TI
lANTHANIDE pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm yb lu
SERIES
ACTINIDE
SERIES
lw
Pu Am Cm Bk
cf
Fm Md
Np
Es
PERIODIC ARRANGEMENT OF:
POTENTIAL SORBENTS BEFORE SCREENING
~ // POTENTIAL SORBENTS AFTER THERMODYNAMIC SCREENING
~Y$$ POTENTIAL SORBENTS AFTER KINETIC SCREENING
»
(j'\
I
-
(j'\
.p-
I
:\.0
i
,
l1J!i1iiiIii.
FIGURE 20
-------
FLUE GAS
SORBER
\.0
REGENERATOR
KINETIC
i FLUID
BED DESIGN
DESIGN
REGENERATOR
EQUIPMENT
SORBEi
SIZE
FAN SIZE
CYCLONE
SIZE
PURCHASE
PRICE
OPERATING
COST
PURCHASE
PRICE
OPERATING
COST
PURCHASE TOTAL
PRICE PLANT
SCALE UP COST
SULFUR FROM
RECOVERY EXISTING
UNIT UNIT
OPERATING SULFUR NET OPERATING
COST CREDIT COST (PROFIT)
]>
0"
,
'7'
.::-
N
ECONOM IC STUDY CONCEPT
FIGURE 21
. .
UI!iIlilIiJ.
-------
li;J/jJ;iIiIJi 6500 TRACOR LANE, AUSTIN, TEXAS 78721
4.2.1
Sorber Design
The most essential piece of equipment used in the sorption
step is the sorber, which is composed of a cylindrical shell, ellip-
tical heads, and a gas distributor plate. The type of sorber
chosen to be used in making economic studies of dry metal oxide
sulfur recovery processes was the fluidized bed.
The fluidized bed model used for the sorber design is
based on the bubbling bed model proposed by Kunii and Levenspiel
(KU-007, KU-008). This simple three-region model for the gas flow
through fluidized beds views uniformly sized bubbles surrounded
by clouds and followed by wakes rising through an emulsion of down-
ward moving solids. Interchange of gas occurs continuously between
the bubble, cloud-wake, and emulsion regions. The bubble diameter
is calculated using equations given by Kato and Wen (KA-004).
The sorber vessel dimensions to be determined were the
sorber diameter, height, and shell thickness. The minimum fluidi-
zation velocity, Umf' is calculated from equations given by Leva
(LE-004). Values of the ratio of superficial gas velocity, Uo'
to minimum fluidization velocity, Umf' along with the' flue gas flow
rates specified by HEW for typical power plant sizes, are used
as input parameters to a computer program which sets the required
sorber diameter.
The second sorber dimension to be determined is the total
vessel height, which is obtained by adding the expanded bed height
to the freeboard height. The freeboard height is the height that
must be added to the expanded bed height to minimize entrainment
losses. The minimum total height was set. at one half the vessel
diameter to minimize gas distribution problems.
Given the extent of conversion, reaction rate constant
(Kr)' particle diameter (dp) and some physical properties of both
the gas and solid, the expanded bed height (Lf) can be calculated
using the following equation.
92
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
where
CAbi
CAbo
Kbc
K
ce
Kr
1
Lf
Ub
Yb
Yc
Ye
C
1 Abo
n-
CAbi
Lf
(' 9dl
j Ub
o
(29)
-K
r
=
Q
1
Yb +K
--E..... +
~c
=
1
1
K
Yc + -!.- + L
Kc eYe
=
concentration of "A" in the inlet gas stream, g-moles/cm3
=
concentration of "A" in the bubble at the top of the
bed, g-moles/cm3
=
interchange coefficient between bubble and cloud,
_1
see
=
-1
interchange coefficient between cloud and emulsion, sec
=
reaction rate constant, cm3 gas/cm3 solid-see
=
height of bed above distributor plate, cm
height of expanded bed, em
=
= bubble velocity, em/see
= fraction of solids in bubble
= fraction of solids in cloud
= fraction of solids in emulsion
Equation (29) was solved by numerical integration using
an iterative method where 1 was increased until the value for Lf
93
-------
j;j;iIiJiIiIii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
was obtained. It was found that a limit on the maximum and minimum
bubble size was needed for calculating Lf for large fluidized beds.
The third sorber dimension to be determined was the
sorber shell thickness. The shell thickness, t, was calculated
using Equation (30) from methods set forth in the American Petroleum
Institute Standard 650 (AM-OOl).
t
.p D
d s
= 2SE + C
=
-4
2.22 x 10 D + 0.159
s
(30)
In Equation (30), Pa is the maximum design pressure, for
which a value of 5 psig was used; Ds is the sorber diameter cal-
culated as described earlier; S is the maximum allowable tensile
stress, for which a value of 15,000 1h/in2 was used; E is the
joint efficiency factor, for which a value of 0.75 was used; and
C is the corrosion allowance, for which a value of 1/16 in. was
used.
The total pressure drop, which is equal to the pressure
drop through the fluidized bed plus the pressure drop across the
gas distributor plate, must be known in order to determine power
requirements and utility costs. Using Equation (31) from Leva
(LE-004) the pressure drop across the fluidized bed, ~P(atm), was
calculated.
~P
=
(9.66 x 10-4) (Lf) (l-e: f) (p s -Pg)
(31)
In Equation (31), Lf is the expanded bed height from
Equation (29), Ps and Pg are the densities of the solids and gas,
respectively, and ef is the bed voidage at expanded height.
In order to prevent channeling effects, Zenz and OthIDer
(ZE-OOl) recommended that the pressure drop across the distributor
94
-------
j;j;JfjJ;iil1Ji 6500 TRACOR LANE, AUSTIN, TEXAS 78721
plate be approximately 40% of the pressure drop through the
fluidized bed. The distributor plate pressure drop was designed
to be within a range of + 10% of one-third of the fluidized bed
pressure drop. The distributor plate was assumed to have holes
one-half inch in diameter to limit bubble size growth.
The terminal velocity, Ut' must be compared
superficial gas velocity, U , to determine the extent
o
ment of sorbent particles from the top of the sorber.
was used to calculate the terminal velocity.
with the
of entrain-
Equation (32)
Ut
=
[d gc ps]~
. CD Pg
(32)
In Equation (32) d is the particle diameter in cm, g
p c
is 980 cm/sec2, P and P are the solid and gas densities in
s . g
g/cm3, and CD is the drag coefficient.
A computer program, FLUBED, was written to calculate the
sorber physical dimensions (height, diameter, and shell thickness)
along with the total pressure drop through the system and the
terminal velocity. An example of the computer output is given in
Table X for the 'case of a 1400 MW power plant handling 2.5 million
SCFM of flue gas using copper oxide as the sorbent.
4.2.2
Draft Fan and Driver Design
It has been shown by Prater and Antonacci (PR-OOl) that the
cost of fans may be estimated from the wheel diameter. The wheel
diameter was calculated from the inlet flow rate, fluid density,
and pressure rise. The design method is described in detail in
T.M. 004-009-Ch12. The type of fan chosen for use in cost esti-
mating was a 1200 RPM, double width, double inlet fan. This type
of fan is typical of large power plant installations.
95
-------
P:'GE
1
F L (1 \~ R 0.. T E (S C P1 )
P~RTICL€ JIAMETf~ (INCHES)
REACTION RATE (1/5fC)
PATIO- SUPERFICIal TO MIN. VEL.
T~MPER~TUPE (CES. F.)
INLET PRESSURE (OSIA)
FLO\.I :::n:=: (.r.CF:1)
~ DlpMET:K GF BEQ (FEET)
~ ~~IGHT OF ~~O (F~ET)
HEIGHT OF Vf-SSFL (FEET)
THICKNE35 OF SHELL (INCHES)
=~J ?RESSURE DROP (IN. WATER)
DISTR. PRiSS. oR.O? (IN. WATER)
TOTAL PR~S5. DROP (I~. WATER)
OUTLET PRESSUR~ (PSIA)
WEIGHT OF SOR5ENT (LBS)
FRACTION OF SOLIDS IN CLOUD
FRACTION OF SOLIOS IN EMULSION
TABLE X:
SORBER PHYSICAL D~ENSIONS
26 JUN 196<;
INPUT FOR FLUIDIZED BED
25000QO.COD
.0380
~ 7 . = L" G G 0
3.000
613.GOG
DENSITY OF SOLID (LBS/FT*-3)
INL~T CONCENT. (HOLE PER:ENT)
OUTLET CONCENT. (HOLE PERCENT)
VOL. WAKE / VOL. aU3SLE .
FRACTION OF SOLIDS IN BUBBLE
100.000
.00300
.00015
.32(;
.000
14.700
. .
OUTPUT FROM FLUIDIZED B~O
5155171.00C
136.:,23
2.675
I~ITIAL GU6BLE DIAMETER (F~ET)
FINt.L BUBBLE DH~IETER (FEET)
INITIAL BUBBLE VEL. (FT/SEC)
FINIAL 8UBB.LE VEL. (FT/SEC)
MIN. MASS FL~X (L3S/HR-FT*.2)
MIN. FLUID VOIDAGE
MIN. FLUID V~L. (FTISEC)
HOLES/IN.**2 IN DISTR. PLATE
PARTICLE TERMINAL VEL. (FT/SEC)
SUPERFICIAL VELOCITY tFT/SEC)
FRACTIONS OF BUBBLES IN 3EO
.205
.328
7.C41
7.523
89.422
.458
.652
1.855
11.835
5.866
.846
58.291
.427
4 .283
1.441
5.724
. 14.493
327088.130
2.791
-2.619
-------
lIiI1iiiJ;i 6500 TRACOR LANE, AUSTIN. TEXAS 78721
The input electrical power required to drive the fan was
calculated in order to choose an appropriate electric motor
(driver). The horse power rating of the motor must be matched
to the maximum horsepower that the fan might require under any
flow conditions at the rated speed. This rating is usually about
35% more than the required fan horsepower. Hence, the motor
horsepower is given in Equation (33) where BHP is the brake
horsepower.
MHP
=
1. 35 (BHP)
(33)
A computer program, FANCST, was written to calculate the
sizes of the draft fan and driver. An example of the computer
output is given in Table XI. The program is described in T.M.
004-009-Ch23.
4.2.3
Sorbent Fines Collector (Cyclone) Design
The most widely used type of dust-collection equipment
is the cyclone separator. A standard cyclone operating with 98%
removal efficiency as described by Perry (PE-OOl) and Bradley
(BR-005) was designed using inputs of flow rate and allowable
pressure drop.
The cyclone diameter was calculated from a given gas
density, flow rate and specified pressure drop (see T.M. 004-009-
Chl5A). The remaining cyclone dimensions were then calculated
from relationships given by Perry (PE-OOl) as shown in Figure 22.
The inlet rectangular duct was assumed to be one-half a cyclone
diameter (Dc/2) long. The outlet duct was assumed to protrude
one-half a cyclone diameter (Dc/2). A computer program, CYCLON,
was written to calculate the cyclone dimensions. An example of
the output is given in Table XII.
97
-------
PAGe.
7
COST INDeX AS O~ ~AY
r~N SPEED (RPM)
OV~?ALL ~FFICIEN~Y
NUMBER OF FAN UNITS
5d'3;'3
FAN CAPACITY (JCFM)
DRIVER BRAKE HORSEPOWER
~ ACTUAL DRIVER ~OqSEPOWER
FRACTIONAL CAPACITY PER FAN
KILOWATTS REQUI~ED BY DRIVER(S}
TABLE XI:
COPPER OXIDE SORBER
INPUT
VARIABLES
For~ ~AN COST
279.1JO
1200.000
.35Q
1 .
TOTAL FAN CAPACITY (ACFM)
AIR uENSITY (LBS/FT**3)
FAN PRESSUrtE ORO? (IN. WATER)
FAN AND 8RIVER ~OTOR OUTPUT PARAMETERS
INDIVIDUAL FAN UNIT PARAM~TERS
51S517G.9G~
10211.332
~HEEL OIAM~TER (INCHES)
FM~ COST (DOLLARS)
MOTOR COST (DOLLARS)
EACH FAN UNIT COST (DOLLARS)
14322.'374
1.00G
7514.752
COST - ALL FAN UNITS (OOLLARS)
-~
25 JuN 196'3
5155171.0GG
.038
10.724
'31.91C
44245.
1~700;;.
1~1?51.
191251 .
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
A
L__-
A
-.1
i
I
. i
Be: 0./4
D.' 0./2
He: 0.12
L.' 2Dc
S. = D.IB
Z.' 20.
J. z arbilroiy,
usually D./4
. ,
':
Seclion A-A
Duel ~ aul
Figure
22: Cyclone Separator Proportions
Perry, J. H., Chemical Engineers Handbook, 4th Ed., McGraw-Hill,
. 1963.
99
-------
? l~ G ~
5
CYCLCN~ REMOVAL EFFICIENCY (Z)
FLUID C~NSITY (LSS/FT**3)
VISC0SITY (L3S/FT-SEC)
EFFECTIVE TURNS
I~PUT FLOW qATE (ACrM)
t-'
o
o
INcur FLC~ RATE (ACFM)
CYCLON~ DIA~~T~~ (F~~T)
?~RT. ~IZ~ (MIC~C~S)- 33.~ EFF.
PAF.T. SIZE (MICRONS)- 50% EFF.
TOTAL COST OF ALL CYCLONE UNITS
TABLE XI I :
COPPER OXIDE SORBER
INPUT PARA~ETERS FOR CYCLONE
sa.ou]
?RfSSURE a~op (INCHES WATER)
SOLID DENSITY (L8S/FT**3)
WALL THICK~ESS (INCHES)
CU~RENT COST INDEX
NUM9ER OF CYCLONES
.-03 S
.000Q2013
3 .~,OCj
5155171.000
OuT?UT ?A~AMtTERS FOR CYCLONE
. -
INDIVIDUAL CYCLONE PARAM~TERS
1282.792 .7 D a
INLET VELOCITY (FT/SEC)
CYCLONE W£IG~T (LSS)
CCST OF EACH CYCLONE (OOLLAKS)
43.202
59.aao
35.370
553312.
25 JUN ~ 9S'
5.000
lOO.CCG
.25C
27g.10L
. ~.
73.965
363394.790
13957B.
-------
ilJt1J:ii;jjjj 6500 TRACOR LANE, AUSTIN, TEXAS 78721
4.3
REGENERATION UNIT DESIGN
The equipment required for the regeneration step
of a regenerator, draft fans and cyclones. The design and
of these pieces of equipment is discussed in the following
consists
sizing
sections.
4.3.1
Regenerator Design
The gas-solids contactor for the regenerator was designed
using the fluidized bed model discussed in Section 4.2.1. The
same parameters had to be calculated for the regenerator fluidized
bed as for the sorber fluidized bed. However, the required flow
rate for the regenerator had not been specified as it had for
the sorber. The required flow rate was calculated from the sorbent
circulation rate. The sorbent circulation rate was calculated
from a material balance around the sorber according to Equation
(34), where Gs is the gas feed rate to the sorber in 1b moles/hr.,
G (Y1 S - Ya s)Mso
s . . 3
-
S(~ - ~)
(34)
Y (1 ,a) 8 is the mole fraction of sulfur oxides (entering, leaving)
the sorber, S is the sorbentcircu1ation rate in Ib/hr, M is the
molecular weight, and X(l, a) is the sorbent loading (entering,
leaving) the sorber in 1b.SOa/1b.sorbent. The sorbent loading
factor, ~, was determined from the extent of conversion data
found in the experimental program as described in Section 3.4.4.
The required flow rate, GR' for the fluidizing medium
calculated from the sorbent circulation rate S using the
equation.
was then
following
GR(Y1R - Y2R)Mso
3
=
S(~ - Xa)
(35)
101
-------
~ 6500 TRACOR LANE. AUSTIN, TEXAS 78721
In Equation (35), the y's are again the mole fraction of
sulfur oxides entering and leaving, and the XIs are again the
sorbent loading factors It was assumed that the sorbent was only
90% regenerated and that the gas leaving the regenerator contained
approximately 6 mo1e% 803' which is a sufficient concentration
for operating a contact sulfuric acid plant. From the required
flow rate of fluidizing medium, GR' the regenerator vessel dimen-
sions were calculated using the fluidized bed model computer
program FLUBED. Since explicit kinetic rate data were not avail-
able for the regeneration step, residence times based on a few
regeneration experiments were used to determine a reasonable value
for K (regeneration) for the fluidized bed model program. An
r
example of the computer output is given in Table XII.
4.3.2
Draft Fan and Driver Design
The same procedure was followed for sizing the draft fans
required for the regenerator as was used on the sorber. An example
of the computer output for the regenerator fan calculations is
given in Table XIV.
4.3.3
Sorbent Fines Collector (Cyclone) Design
The same procedure was followed for sizing the cyclones
required for the regenerator as was used on the sorber. An example
of the computer output for the cyclone calculations is given in
Table XV.
4.4
TOTAL CAPITAL INVESTMENT COST
The basis of computing the capital investment cost is the
purchase price of the major pieces of equipment (see T.M. 004-009-
Ch11). To the purchase price is added amounts of money to account
for the erection cost, piping cost, instrument cost, etc. These
items are added as a percent of purchase price. Several investi-
gators (LA-004, HA-002, AR-001) have shown how percentages of the
purchase price may be applied to cover the cost of erection, instru-
mentation, etc. The modified percentage factors given by Lang
102
-------
TABLE XIII
COPPER XOIDE REGENERATION PHYSICAL DIMENSIONS
FLal~ iHTE (SCFM)
PARTICLE DIAMETER (INCHES)
PEACTICN RATE (l/SEC)
R~TIO- SUP£P.FICIAL TO MIN. VEL.
TEMPERATU?E (DES. F.) ,
INLET PRESSURE (?SIA)
FLOW RarE (ACFM)
I-' DI~~ET[R OF BED (FEET)
8 H!.::IGHT OF ar;:ij (FEET>
HEIGHT OF V~SSEL (FEET)
THICKNES~ OF SHELL (INCHES)
BED PRfSSURE DROP (IN. WATER)
DISTR. PRESS. Q~OP (IN. W~TER)
TOTAL cR~55. DROP (IN. WATER)
OUTLET PRESSURE (PSIA)
WEIGHT OF SORB€NT (L8S)
FRACTION OF SOLIDS IN CLOUD
FRACTION OF SOllOS IN EMULS:ON
INPuT FOR FLUIDIZED 8EO
111.;80.000
.U3.JO
10.0[000
9.000
1377.080
14.700
DENSITY OF SOLID [LBS/FT**3)
INLET CONCENT. (MOLE PERCE~T)
OUTLET CONCENT. (MOLE PERCENT)
VOLr WAKE I VOL. RUqaLE .
FRACTION OF SOLIDS IN SUSBLE
'.
OUTPUT FROM FLUIDIZED BrD
394029.9SQ
43.110
2.343
INITIAL SUaBLE DIAMETER (FEET)
FINAL BUBBLE DIAMETER (FEET)
INITIAL 3uaBU: VEL. (FT/SEC)
FI~IAl gUS3LE VEL. (FT/SEC)
MIN. ~AS5 FLUX (L0S/HR-FT*.2)
MIN. FLUID VOIDAGE
MIN. FLUID VEL. (FT/SEC)
HOL~S/IN.**2 INDIST~. PLATE
PARTICLE TER~INAL VEL. (FT/SEC)
SUPERFICIAL VELOCITY (FT/SEC)
FRACTIONS OF BUBBLES IN BED
21.555
.250
7.035
2.361
9,.397
14.301
53525.159
1.567
-1.326
."-../
108.000
10.00'300
1.;0000C
.320
.DOC
.229
.328
5.328
6.310
40.074
.465
.500
.483
14.207
4.500
.167
-------
SCST INDEX AS OF MAY
F M~ S PEE 0 (\~ P!vl )
OVERALL EFFICIENCY
NUM3~R OF FAN UNITS
Sd959
FAN CAPACITY (ACFM)
~ DRIVER BRAKE HORSEPOWER
~ ACTUAL DRIVER HORSEPOWER
FR'CTIONAL CAPACITY?ER FAN
KILOWATTS REQUIRED 8Y DRIVE~(5)
TABLE XIV
COPPER OXIDE REGENERATOR
INPUT VARIA3LES FOR FAN COST
279.10:]
1280.000
.850
1.
TOTAL FAN CAPACITY CACFH)
AIK DENSITY (LBS/FT**3)
FAN PRESSURE DROP (IN. WATER)
FAN AND
DRIVER MOTOR OUTPUT PARAMETERS
'.
INDIVIDUAL FAN UNIT PARAM~TERS
394029.958
1047.772
1520.960
1.COG
~HcEL DIAMETER (INCHES)
FAN COST (DOLLARS)
MOTOR COST (OOLLARS)
E4CH FAN UNIT COST (DOLLARS)
781.337
COST - ALL FAN UNITS (DOLLARS)
394029.950
.022
14.397
90.840
42938.
16903.
59841.
59841.
-------
CYClO~E RE~ovtL fFFICI~NCY (%)
FLUIO J~NSITY CL8S/rT**3)
VISCOSITY (L8S/FT-SEC)
EFFECTIVE TURNS
INPUT FLOW RATE (ACF~)
INPUT FLO~ RATE (ACFM)
..... eye LON E: 0 I A ~I E: T E R (F E E T )
o PAF,T. SIZE (MICRONS)- 98.~:; EFF.
VI
PART. SIZE (MICRONS)- 50~ EFF.
TOTAL COST OF ALL CYCLONE UNITS
TABLE XV
COPPER OXIDE REGENERATOR
INPUT PARAM~TERS FOR CYCLONE
98.000
. Q.?2
.000028.30
3.500
3~4029.350
PRES~URE DKCP (INCHES WATER)
SOLID DENSITY (LBS/FT**])
~ALL THICKNESS (INCHES)
CURRENT COST INDEX
NUMBER OF CYCLONES
OUTPUT PARAMETERS FOrt CYCLONE
. ,
INDIVIDUAL CYCLONE PARAMETERS
98507.437
11 .55 Q
30.430
INL~T VELOCITY (FT/SEC)
CYCLONE WEIGHT (LaS)
COST Or EACH CYCLONE (DOLLARS}
17.999
32514.
5.COO
100.J08
.25G
279.100
4.
96.784
21228.051
8154.
-------
lliIJ;iillJi 6500 TRACOR LANE. AUSTIN, TEXAS 78721
(LA-004) were used in this study. These factors are given in
XVI. Although the factors have been criticized as being non-
realistic, the Lang factors are still recommended for use in
preliminary design cost estimation.
Table
The purchase costs of the fluidized bed contacter, along
with the associated cyclones and draft fans, were estimated (see
T.M. 004-009-Ch14, Ch15A) for both the sorber and regenerator.
The cost of the sorber, regenerator, and cyclones was conviently
estimated on the basis of weight of steel required knowing the
vessel dimensions. Approximate costs per pound for fabricated
steel vessels and the. purchase price of steel were taken from cost
information supplied by the Graver Tank Company, (f.o.b. Houston,
Texas).
The fluidized bed contactors were composed of the follow-
ing parts:
1.
2.
3.
4.
Cylindrical shell (including support)
Two elliptical heads
Gas Distributor plate
Vessel lining-insulation material
The purchase price of the contactor was estimated to be
the sum of the costs of the individual pieces of equipment. The
cost of the first three parts was determined by the weight of
steel required. The cost of the insulation material was obtained
from Perry (PE-OOl).
The size and purchase price of the draft fans were
mated and they compared quite well with the draft fans being
in the city power plants in Austin, Texas.
Also included in the total capital investment for the
sulfur removal process was an initial sorbent cost of $l/lb.
esti-
used
The cost data used in this study were updated to current
price levels using the Marshall-Stevens cost index factors published
106
-------
~ 6500 TRACOR LANE. AUSTIN. TEXAS 78721
NO.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
TABLE XVI
INSTALLATION COST FACTORS
ITEM
Purchased Equipment
Erection Labor
Foundations and Platforms
Piping
Instruments
Insulation
E1ec trica1
Buildings
Land and Yard Improvements
Utilities
Engineering and Construction
Contractor's Fee
Contingency
Working Capital
107
LANG FACTOR
1.00
.25
.18
.76
.15
.08
.10
.25
.13
.40
.66
.19
.59
10% Fixed Capital
-------
l1;i/iJ;jJ;jJ;j 6500 TRACOR LANE, AUSTIN, TEXAS 78721
in "Chemical Engineering". The use of these cost index factors is
described by Aries and Newton (AR-001).
4.5
GROSS ANNUAL OPERATING COST
The gross annual operating cost for a given process is the
sum of all direct, indirect, and fixed charges (AR-001). These
costs include such items as labor, plant overhead, utilities,
depreciation and insurance. These costs are further defined in
Table XVII.
The utility costs (power requirements) were calculated
from the total pressure drop through the process. The pressure
drop calculations were discussed in Section 4.2.1. Labor costs
were estimated using the guidelines set forth by HEW. The heat
requirement for the process was calculated from an overall process
heat balance.
It was found that the sorber had a small heat credit which
was applied to the requirements of the regenerator, thus lowering
the amount of heat to be purchased to operate the regenerator. A
detailed description of the heat balance calculations is given in
T.M. 004-009-Ch26. Other operating costs were determined using
percentage factors supplied by HEW. Also included in the operating
cost was an attrition loss of 0.05% of the sorbent circulation rate.
4.6
RESULTS
TRACOR estimated the total capital investment and gross
annual operating cost along with an overall heat and material
balance for the following plants using the two most promising single
metal oxide sorbents indicated by thermodynamic and kinetic screening.
Flue Gas Size Concentration
Plant Type Million SCFM MW Inlet (ppm) Outlet (ppm)
Large Power
Plant 2.5 1400 3000 150
Medium Power
Plant 0.5 220 3000 300
Smelter Gas
Plant 0.02 29000 5000
108
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
TABLE XVII
COMPONENTS OF GROSS ANNUAL OPERATING COST
DIRECT COSTS
INDIRECT COSTS
Raw Mat1's and Chemicals
Direct Labor
Supervision
Maintenance
Payroll Burden
Plant Overhead
Packaging & Shipping
Waste Disposal
Plant Supplies
Utilities
109
FIXED COSTS
Depreciation
Taxes
Insurance
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
For the largest power plant, the sorber was designed with
several units linked in parallel for efficient use of available
space, to allow for operating flexibility, and to decrease the required
vessel diameter.
A summary of the results of the economic calculations
for the best sorbent for each plant type are given in Table XVIII.
An example of the simplified process flow sheet for the CuD process
.for the large plant is given in Figure 23. Process f10wsheets
describing all of the above mentioned sulfur oxide removal processes
are given in Section 8.2.5 along with their respective heat and
mass balances. Cost estimation output for the above plants using
copper and iron oxide sorbents is given in Section 8.2.4
The total capital investment for either the copper oxide
or iron oxide process for a 1400 MW power plant processing 2.5
million SCFM is $8 million ($6/KW) and the gross annual operating
cost is $3 million (0.03~/KWH). It is recognized that there are
other costs that will arise when a more detailed study is done.
Some of the possible costs are given in the following list.
1.
The regeneration temperature may require the use
of stainless steel materials.
Costly utility tie-ins due to plant location could
occur.
2.
3.
4.
Special heat exchange equipment may be required.
Sulfur product credit may vary according to market
areas.
5.
Special valves may be required to regulate gas flow.
The required capital investment for the sulfur oxide
removal processes studied by TRACOR, through only a preliminary
estimate, appears to be a reasonable one. The operating cost is
in the same range as the potential sulfur by-product credit,
approximately $3-4 million assuming $35 ton sulfur. In summary,
the dry metal oxide sulfur recovery process appears to be economi-
cally feasible based on the information derived from this study.
110
-------
TABLE XVIII
SUMMARY OF COST ESTIMATES FOR THE COPPER
OXIDE AND IRON OXIDE PROCESSES
GI
UI
o
o
Total Capital Investment Gross Annual Operating Cost. ~
Process :u
>
$ Million $ Million 0
Sorbent 0
:u
Medium Smelter Medium Smelter r
Large Large >
z
Power Power Gas Power. . Power Gas !11
Plant Plant Plant Plant Plant Plant >
c
UI
:!
z
~
I-' m
X
I-' >
I-' 0.647 0.238 UI
CuO 8.25 1.88 0.351 2.63
...,
CIII
...,
N
..
Fe20s 8.01 1.76 0.279 2.53 0.627 0.226
-------
POWER PLANT SIZE: 1400 MW
2.5 MilliON SCFM FLUE GAS
FEED MAKE-UP
14.7 PS/A 14.7 PS/A
6138F 13778F
CD (]ff)
CD)
-61'''' X 30' 53,525 43'",X22'
CD - LIS
CliO
IT) HEAT
ElCHAllfE
CD CD m FLUIDlllllf
FLUE fAS HEAT NEAT MEDIUM
N (All 01
FIOII FtmllAeE ElCNA.E ElCNA.E
SOl'TlOIl UIII T IUEIlEIATlOIl UIlIT FLUE fAS J
CD CD CD CD (IE)
:t>
()'\
I
~.
PIOCESS FLOWSNEET
DATE -JUliE 30, 1969
A6-164-37
AUSTIN TEXAS
()'\
+-
,
FIGURE 23 - PROCE~,S FLOWS:":EEl FOR LiE COPPER OX I DE PROCESS
+-
w
-------
~ 6500 TRACOR LANE,
AUSTIN, TEXAS 78721
5.
SUMMARY
The purpose of TRACOR's study for NAPCA was to determine
which metal oxides were best suited to the removal of sulfur oxides
from flue 'gases by chemical reaction.
The thermodynamic requirements for efficient sulfur oxide
removal and product regeneration were thoroughly investigated.
The requirements for the sulfur removal process were determined
from the specified outlet concentration of SO~ in the flue gas and
from the requirements of sulfur recovery processes. The specifi-
cations were that the logarithm of the equilibrium constant for
the decomposition reaction "must be less than -3.82 at the sorption
temperature and greater than -2.0 at the regeneration temperature.
In order to calculate the equilibrium constants, thermodynamic
properties of interest were compiled using the Univac 1108 computer
for 629 compounds. Properties for about 40% of the compounds had
to be estimated. Additional descriptive data such as thermal
stability properties were also compiled and used in the thermo-
dynamic screening process. Sixteen thermodynamically favorable
sorbents were selected as a result of the screening process. The
potential sorbents were the oxides of titanium, zirconium, hafnium,
vanadium, chromium,iron, cobalt, nickel, copper,' zinc, aluminum,
tin, bismuth, cerium, thorium, and uranium.
The kinetic studies were carried out to determine which
of the 16 potential sorbents reacted fast enough with SO~ to be
economically feasible. The oxides were prepared in a kinetically
active form by calcining a salt which decomposed to the oxide at
a low temperature. The sorbents were identified and characterized
using x-ray diffraction methods, BET surface area determinations,
and chemical analysis. The rate data were collected using an iso-
thermal gravimetric technique whereby weight gain of SO~ was recorded
as a function of time. The experiments were carried out in a
simulated flue gas atmosphere.
113
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
It was found that six of the potential sorbents had a
significant rate of reaction with 502. The other sorbents either
failed to react or took several hours to achieve a measureab1e
weight gain. An economically feasible residence time is on the
order of minutes rather than hours. An economically feasible
reaction rate constant, Kr' was found to be greater than
10 sec-l. The oxides of copper, chromium, iron, nickel, cobalt,
and cerium had reaction rate constants within this range. Copper
and iron oxides were selected as the best potential sorbents, since
the othersorbents were complicated by such factors as undefined
stoichiometry of the sorption reaction, initial formation of a
product layer which slowed the reaction, and low 503 partial
pressure over the sorption product.
Finally, preliminary design and economic studies were
made for a sorber-regenerator system based on a fluidized bed model
for the gas-solid contactor The bubbling bed model of Kunii
and Levenspie1 (KU-007, KU-008) was used, with inputs from the
thermodynamic and kinetic studies, to design and size the gas-solid
contactor. Draft fans and cyclones were also designed and sized.
The capital investment required for the process was estimated based
on the purchase prices of the major pieces of equipment with amounts
of money added to~count for such things as erection costs, piping
costs, and instrument costs. The percentage factors given by
Lang (LA-004) were used to account for these installation costs.
The gross annual operating costs were determined from
percentage factors supplied in HEW guidelines and from power and
heat requirements determined from heat and material balances for
the process. The metal oxides, CuO and Fe203' were found to have
promise as potential sorbents for an economically feasible sulfur
removal process. The preliminary cost estimate of the required
capital investment for the dry metal oxide sulfur removal process
was $8 million, and the annual operating cost was estimated to be
$3 million for a 1400MW power plant.
114
-------
AM-OOI
AR-OOI
BR-OOS
0(-006
ER-Q[Jl
.....
.....
\J1
GR-002
HA-002
J A- 001
KA-OO~
KU-GCl
!
-------
.K U- 00 8
KU-009
LA-OO 1
LA-D02
L A-004
I-' LE-004
I-' .
0\
MA-006
N B- 00 3
N B- 005
PE -00 1
PR-OOI
P R- 002
D. KUNII, Q. LEVENSPIEL, '3U83LING SED MODEL FORTH~ FLOW OF GAS
THROUGH A FLUIDIZED BED.' liE C FUNOAMENTALS. VOL. 7, NO.3
PP. 44G-52 (AUG. 1968).
D. K~NII, Q. LEVENSPIEL, FLUIDIZATION ENGINEERING. JOHN WILEY
AND SONS. INC.. NEW YORK, (1969).
KALORISCHE ZUSTANDGROESSEN IN 'LANDOLT BOERNSTEIN',SPRINGER-VERlAG,
SERLIN,VOL.2,PART4(1961). .
w. LATIMER, J. AM. CHEM. SOC.. VOL. 13, P. 1480 (1951).
H. J. LANG, 'SIMPLIFIED APPROACH TO PRELIMINARY COST ESTIMATION',
CHEMICAL ENGINEERING, JUNE (1948).
MAX LEVA. FLUIDIZATION, MCGRAW-HILL BOOK CO.. NEW YORK, (1959J.
D. S. MACIVER. P. H. EMMETT, J. AM. CHEM. SOC.. VOL. 60. P. 624
(193-3). .
F,ROSSINI,ET AL. SELECTED VALUES OF CHEM. THERM. P~OPERTIES. NATL.
3UREAU OF STANDARDS, CI~CULAR 500. (1952).
SELECTED VALUES OF CHEM. THERM. PROPS.. NATNL. BUR. STOS.. TABLES
FOR FIRST 34 ELEMENTS IN S1D. ORDER OF ARRANGEMENT. TN 270-3 (1968).
J.H.PERRY. CHEMICAL ENGINEERING HANDBOOK,MCGRAW HILL BOOK CO..INC..
4TH EDITION(1953).
N. H. PRATER, D. W. ANTONACCI, 'E5TIMAT~ FORCED-DRAFT FAN COSTS,'
HYDROCARBON PROCESSlhG AND PETROLEUM REFINER. VOL. 40. NO.7.
-PP. 129-33 (1361J.
E. A. PRODAN, M. M. PA\lLYUCHENKO, 'HETEROGENi::OUS CrtEMICAL
PEACTION5', NAUKA I TEKHN!KA, MINSK, U.S.S.R~. P. 20, (1965J.
-------
----~.
- ---- - -~----- -- - -~--
TR-01G
THIS VALUE hAS BEEN SELECTED AS THE BEST OF SEV£RAL DIffERENT
REPORTED VALUES. SEE TECHNICAL MEMORANDUM 004-009-C~18
WE-002
C. Y. WEN. IND. AND fNG. CHEM.. VOL. 60. NO.9, PP. 34-54 (968).
~
~
~
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
7.
ABSTRACTS OF TECHNICAL MEMORANDUMS
7.1
Technical Memorandums on Thermodynamic Studies and
Preliminary Screening
7.1.1
Technical Memorandums on Estimation of Heat of Formation
T.M. 004-009-Ch1 describes
as the best method for estimation
and shows how the estimated heats
values.
the selection of Erdos' method
of heat of formation at 2SoC
compare with some accepted
T.M. 004-009-Ch1A describes in detail the computational
method used for estimating heats of formation and its derivation.
The heats of formation of the su1fites, titanates, a1uminates,
chromates, ferrates, vanadates, wo1framates, and molybdates for
thirty cations estimated from the heats of formation of their
sulfates and carbonates are given along with the errors involved
in the correlation.
7.1.2
Technical ~emorandum on Estimation of Absolute Entropy
T.M. 004-009-ChS is a discussion of the selection of
Latimer's method as the best for estimation of absolute entropies.
A mathematical description of the method, necessary constants for
estimating entropies, a comparison of estimated and accepted
values, and the correlation errors are given.
7.1.3
Technical Memorandums on Estimation of Heat Capacity
T.M. 004-009-Ch4 is concerned with the selection of the
best method for estimating heat capacity as a function of tempera-
ture. A comparison of estimated and accepted values for each
of the methods studied is given. The uncertainty in equilibrium
partial pressures due to errors in heat capacity is also discussed.
118
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
T.M. 004-009-Ch9 presents the computational method used
to estimate mixed metal oxide heat capacities, describes calcula-
tion of the errors involved, and compares estimated and known
mixed oxide heat capacities.
T.M. 004-009-Ch13 describes in detail the computational
method used to estimate heat capacities of metal sulfides.
Estimated and known values are compared. A general mathematical
description and derivation for heat capacity estimations of
carbonates, sulfates, sulfides and mixed oxides, and the errors
involved is given in T.M. 004-009-Ch13A.
7.1.4
Technical Memorandums on the Effect of Errors in
Estimated Thermodynamic Properties
T.M. 004-009-Ch2 is a discussion of the uncertainties
in calculated equilibrium partial pressures caused by errors in .
heat of formation and absolute entropy. The uncertainty in
calculated equilibrium partial pressures due to errors in
estimated heat capacity is discussed in T.M. 004-009-Ch4.
7.1.5
Technical Memorandum on Conflicting Reported
Thermodynamic Data
T.M. 004-009-Ch18 is a discussion of conflicting reported
values for thermodynamic properties and an explanation of the
choices made for accepted values. Conflicting data for 57 com-
pounds were examined. The memorandum is based on 116 references.
7.1.6
Technical Memorandums on Thermal Stability Studies
T.M. 004-009-Ch7 discusse~
thermodynamically stable form for
oxidation states.
the determination of the most
metal oxides having several
119
-------
lli1/J;jj;jjj/ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
T.M. 004-009-Ch8 gives the results of an extensive literature
search concerning the thermal stability of metal sulfates, forma-
tion of sulfates by catalytic oxidation of S02 and reductive
decomposition of sulfates. The memorandum is based on 67 references.
T.M. 004-009-Ch3 is a short discussion of the literature
found in early months of the contract concerning sulfite
decomposition.
T.M. 004-009-Ch16 gives the results of a comprehensive
literature survey and the correlation of all thermodynamic data
concerning the interaction of metal oxides and S02. The subjects
of adsorption with sulfite formation, sulfite decomposition and
disproportionation, adsorption on catalytic effective oxides,
catalytic oxidation of 5°2, sulfate decomposition, and price and
availability of metal oxides are treated.
Technical Memorandum on the Price and Availabilitv
of Metal Oxides
7.1.7
T.M. 004-009-Ch6 presents in tabular form data concerning
the price per ton, amount of production, amount of consumption,
and amount stockpiled for the metal oxides or metals of interest
as potential sorbents. The data were obtained mainly from
publications of the U. S. Bureau of Mines.
7.1.8
Technical Memorandums on Preliminary Screening of
Metal Oxides and Mixed Metal Oxides
T.M. 004-009-Ch16 describes the correlation
data and other information and the selection of 16
metal oxide sorbents on the basis of these data.
of thermal
potential
120
-------
~ 6500 TRACOR LANE, AUSTIN, TEXAS 78721
T.M. 004-009-Ch22 describes the thermodynamic screening
process for the "mixed metal oxides." T.M. 004-009-Ch25 gives
the results of a theoretical study of the chemistry of the mixed
oxides plus a discussion of the problems involved in conducting
an experimental program for the mixed metal oxides.
7.1. 9
Technical Memoramdum on Computer Programs
T.M. 004-009-Ch23 describes the computer programs written
to store, retrieve, correlate, estimate, and print thermodynamic
data. It also describes programs written for data analysis of
experimentally obtained surface area and kinetic data. Finally,
it gives a description of the programs used in the economic
feasibility studies to design and size equipment and estimate
costs.
7.2
Technical Memorandums on the Kinetic Studies
Experimental Program
7.2.1
Technical Memorandums on Design and Operation of
Experimental Equipment
T.M. 004-009-ChlO describes the temperature calibration
of the differential thermal analysis apparatus. The calibration
was done by recording thermograms of high purity metals with
well-defined melting points.
T.M. 004-009-Ch20 is a detailed description of the apparatus
designed and built at TRACOR to simulate the composition of flue
gas. The gas mixing apparatus is used in connection with the
thermal analysis apparatus.
T.M. 004-009-Ch24 describes the apparatus constructed for
determining nitrogen adsorption isotherms used for calculation
of specific surface area by the BET method.
121
-------
iIJ!1J;iilIii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
7.2.2
Technical Memorandums on Collection of Experimental
Data and Presentation of Results
T.M. 004-009-Ch24 describes the experimental procedure and
methods of data analysis used to determine BET surface areas for
metal oxide sorbents.
Bimonthly Progress Report No.8 gives a complete discussion
of the experLmental program up to April 30 including a detailed
description of the kinetic studies. Compound preparation, com-
pound characterization, surface area determinations, kinetic
data collection apparatus and methods, and kinetic data analysis
are discussed. Exper~ental results are tabularized for the metal
oxide sorbents.
7.3
Technical Memorandums on Economic Feasibility Studies
7.3.1
Technical Memorandums on EquiDment Design, Size, and
Purchase Price
T.M. 004-009-Chl9 describes the design of the fluidized
bed sorber based on the bubbling bed model of Kunii and Levenspiel.
The model allows calculation of the bed height from data on
S02 removal efficiency, reaction rate constant, particle diameter,
and gas and solid physical properties.
T.M. 004-009-Ch14 describes the method used to calculate
the sorber purchase price on the basis of the weight of steel
required for the cylindrical shell, elliptical heads, and the
gas distributor plate. Insulation costs were also included.
T.M. 004-009-Ch15 and the revision T.M. 004-009-Ch15A
describe the determination of purchase costs for cyclone dust
collectors. The cost was estimated on the basis of weight of
I
I
122
-------
~
lliIJiiil;ii 6500 TRACOR LANE, AUSTIN, TEXAS 78721
steel required which was determined from the vessel dimensions.
The vessel dimensions were calculated from gas density, flow
rate, and pressure drop.
T.M. 004-009-Ch21 gives the computational procedure for
estimating the purchase price of forced-draft fans such as those
used to move air through power boilers. The technique is based
on estimating the cost of the fan and adding the cost of an
electric motor driver.
7.3.2
Technical Memorandums on the Estimation of Capital
Investment and Operating Cost
T.M. 004-009-Ch26 describes the determination of heat and
material balances around the sorber-regenerator system. These
balances are necessary to determine sorbent circulation rate and
heat requirements so that operating costs can be estimated.
T.M. 004-009-Ch1l describes the calculation of the Total
Capital Investment and the Gross Annual Operating Cost for the
sorber-regenerator system. The procedures used are consistent
with those outlined in the General Guidelines supplied by the
NAPCA. The capital investment is computed from the purchase
prices of the major pieces of equipment. The gross operating
cost consists of direct, indirect, and fixed costs which are
described in detail.
123
-------
~
.( ..
'.' I'
..-
;'
/
f.-:;. ".. -,-,'"":~;;-;-.I
L~:':'t.'-:,:.'::..:~,-.~.):g G500
TRACOR
L,\NE, AUSTIN, TEXAS 78721
P~DITIONS A~~ CORRECTIONS TO THE FINAL
REPORT - .L\PPLICABILITY OF. METAL OXIDES TO THE
DEVELOPMENT OF ~~W PROCESSES FOR
REMOVING S02 FROM FLUE GASES
I. Equation (3) in Volume I) page 5) under paragraph 2.3.2
should be changed to the following:
I
I,
I r oTi
i=O l ~..,
C (T).
P dT +
T
SO
. .a 9 a
ST
+
=
..'" .
,.
I .
T
AST . + S.
~ . T.
. . ~
C (T)
P
T
l
dTJ
(3)
\\Ihere T.: is t~1e temperature of the ith phase transition and ASrr.
. ~~
is the entroPy.oI.the ith phase transition.
II. The following table gives changes and additions to be
mz..de to the TRACOR Data' Base. given in Volume II) paragraph 8..2.1
of the Final Report.
CORRECTIONS
I .
. Heat Df Formation
.at 25QC KCal/g mole
. .76.00
Compound
CsaO
C sa SO,*
-339.00
Rb2SO~ .
-340.20
. .
-------
"..
.'~
. .
.
)
/~7'~I.~ ~;:'~-;-i".-;,-,,-.7
; :,~ ',.~:'.. .l:.'C:'/.<.".J C500
_.--..'...-~-.. .- . - -J
TRACOR
L.ANE. AUSTIN. TEXAS 78721
ADDITIONS
Co::mol1nd
Heat of Formation
at 25UC KCal/g mole
References
Absolute
Entropy
Cal/g mole/
Deg. K
Re'ferences
C So:> Al;::04 -539.18 TR-002 45.8 , TR-003
. -
CS:aCrO", -318.00 ' TR-002 58.4 KE-OOl
CS:a Cr 2q/~ -551.92 TR-002.. 55.0 KE-OOl
Cs2Fe;aO.. -327.16 . TR-002 52.86 TR-003
c 1'1' 0 -354.07 TR-002 .
~2no .:.
. -.- '4'4-
------- |