Catalytic Control of NOX Emissions
from Mobile Sources
FINAL REPORT
by
M. I. Seegall
J. C. Napier
W. A. Compton
Prepared for
Environmental Protection Agency
Office of Air Programs
-------
~13 ,(d¥~//
Catalytic Control of NOx Emissions
from Mobile Sources
FINAL REPORT
by
M. I. Seegall
J. C. Napier
W. A. Compton
Prepared for
Environmental Protection Agency
Office of Air Programs
Division of Emission Control Technology
LkJ
HI"SOLAR
DIVISION OF INTERNATIONAL HARVESTER COMPANY
2200 PACIFIC HIGHWAY' SAN DIEGO, CALIFORNIA 92112
RDH ] 700
-------
FOREWORD
This Final Report covers the work performed for the Environmental
Protection Agency during the work period July 1, 1970 to July 31, 1971.
The contract with the Research Laboratories of Solar Division of International
Harvester Company, San Diego, California, was initiated by EPA under contract
number EHS 70-114 and monitored at the start by Mr. Steve Quick, and since
November 1970, by Dr. Joseph H. Somers.
The program was under the direction of Mr. W. A. Compton, Assistant
Director-Research, who served as Program Director. Mr. M. I. Seegall, Senior
Research Engineer, was the principal investigator at Solar. Mr. J. C. Napier,
Research Engineer, conducted the catalytic gas reaction tests and collected the
material for the ammonia testing. The Reaction Kinetics program was conducted by
Dynamic Science, a Division of Marshall Industries. Mr. Edward Fort, Staff Engineer,
,
Engineering Research, International Harvester (Hinsdale, Illinois) acted as a consultant
throughout the program, providing valuable data on Otto cycle exhaust composition and
general guidance on emission monitoring. Mr. T. E. Duffy, set up and built the
apparatus for the measurement of the catalyst surface area and Mr. J. F. Nachman,
Chief of Applied Science, wrote most of Appendix E. Special gratitude is due to
Mr. J. V. Long, Director-Research, who supported the project and encouraged the
research team throughout the entire program.
This report is identified by Solar as RDR 1700.
-------
ABSTRACT
This is a program studying the catalytic approach toward the control of NOx
emissions from motor vehicle exhaust. An investigation of rare earth oxide catalysts
and a comparison with selected transition metal and noble metal catalysts for the pro-
motion of the reaction
2 NO + 2 CO ~ N + 2 CO
2 2
was made.
The initial criteria for the catalyst were:
a.
to have high effectivity for NO reduction;
b.
not to be permanently damaged by lead oxide and lead bromide
compounds in the exhaust;
c.
to be physically stable for long periods in exhaust gases - to last an
equivalent of about 50,000 miles of engine operation;
d.
to be economical.
Point b, on the compatibility of the catalyst with lead, became less important with
the probability that unleaded fuel for catalytic systems will be widely available in
1975.
The study involved the design of experimental equipment for the gas reaction
process and for the investigation of physical and chemical properties of the rare earth
oxide catalysts. Also included were a theoretical nitric oxide equilibrium analysis, a
reaction kinetics study, and a study in concentration of NO, CO, and C02 as a function
of several reaction parameters of the catalytical gas reaction. The data were analyzed
by the non-dispersive infrared (NDIR) analysi s method for the concentration of these
three gases, and by titration for the measurement of ammonia generated.
The following catalysts were investigated in the course of the program:
. Catalyst RE(23)75-P
. Catalyst RE(M3)75
-------
. Monel 400
. Solar Monel 400 Mod
. Solar Monel 400 Core
. Copper Dichromate, CuCrZ07
. Palladium Catd.lyst, 0.5% Pd on AIZ03 pellets
. Platinum Catalyst, 0.5% pt on AIZ03 pellets.
The first two catalysts listed represent rare earth oxide formulations.
Although the effectiveness of the rare earth oxide catalysts was less than
desired, the goals listed under points b, c and d, were met. The other catalysts
evaluated by comparison have been found to possess a higher effectivity, but they
fall far short of goals band c.
The Monel 400 catalysts were found to be substantially more efficient than all
others for NO reduction and minimum ammonia formation. The noble metal and copper
dichromate catalysts were in the same NO reduction efficiency range, however,
ammonia formation for pt and Pd catalysts were well beyond acceptable limits.
-------
Section
1
2
3
4
CONTENTS
INTRODUCTION
1.1
1.2
Background
Goals of Solar Program
THEORETICAL ANALYSIS
2.1
2.2
2.3
2.4
Definition of Concepts
Equilibrium Composition Analysis
Study of CO-NO Reaction Kinetics
2.3.1
Catalytic Reactor Program -- Thermochemical
Calculations
Significance of Theoretical Study of Experimental Work
3.1
EXPERIMENTAL PROCEDURE
3.2
3.3
3.1.1
Rare Earth Oxide Catalyst Study
Catalyst Composition Selection and Screening
Experimental Study With Other Catalysts
The Catalytic Gas Reaction Experiment
3.3.1
3.3.2
3.3.3
3.3.4
3.3.5
Gas Flow Apparatus
The Synthetic Exhaust Gas Mixture
Description of Analytical Apparatus for Emissions
Monitoring
Procedure for Conducting Experiment
Reduction of Data
4.1
EXPERIMENTAL RESULTS
Catalysts and Catalytical Reaction Parameters
Investigated
4.1.1
4.1. 2
4.1. 3
Catalysts Considered in the Program
Parameters Considered in the Program
Parametric Investigation of the NO-Reducing
Efficiency of the Catalysts in This Program
v
Page
1
1
3
5
5
5
16
16
19
21
21
21
29
29
29
35
40
43
46
47
47
47
47
-------
Section
5
6
CONTENTS (Contd)
4.2
NO Reduction Efficiency of the Catalysts Tested
Experimental Data for Rare Earth Oxide Catalysts
Experimental Data for Transition Metal and Noble
Metal Catalysts
The Ammonia Generating Factor for all Catalysts Tested
4.2.1
4.2.2
4.3
5.1
EVALUATION OF DATA AND DISCUSSION
5.1.1
Evaluation of Experimental Data
5.1. 2
5.2
Graphical Presentation of Catalytical Efficiency
Parameters
Comparison of Ammonia-Generating Factor E
for all Catalysts Tested
Parametric Design Charts for Catalyst Formula
RE(23) 75-P
SUMMARY AND CONCLUSIONS
REFERENCES
APPENDIX A
APPENDIX B
APPENDIX C
APPE NDIX D
APPENDIX E
APPENDIX F
vi
Page
50
50
55
57
61
61
61
70
72
77
81
85
95
105
113
127
-------
.Figure
1
2
3
4
5
6
7
8
9
10
11
12
13
14
ILLUSTRA TIONS
Equilibrium Concentration of NO (Y) Vs. Temperature (T)
Effect of CO/NO Concentration Ratios on NO Equilibrium
Concentration Normalized for Y at Ratio of 2
Effect of Pressure on NO Equilibrium Concentration,
Normalized for Y at 1 Atmosphere
Dependence of NO Equilibrium Concentration of Either N2
or C02
Influence of Initial Reaction Product Concentration on NO
Equilibrium Concentration Normalized to NO Initial Product
Concentration
Rare Earth Oxides as Raw Material for Solar Catalyst Formulas
Exhibit of Solar Sponsored Catalyst Development Program,
Covering Five Formulations, Poisoning Studies, and Raw
Materials Utilized
View of Newly Pelletized REO Catalysts, Segregated by Formula
and Preparation Parameters, Prior to Sintering
Catalyst Samples of Figure 8 After Sintering
Schematic of Catalytic Gas Reaction Apparatus for Experi-
mentation With Synthetic E chaust Gas Mixtures
Detailed Sketch of Catalyst Bed
Gas Temperature Profile in Catalyst Bed
General View of Catalytic Gas Reaction Apparatus in Final
Version
Researcher Rolls Safety Screen in Front of Gas Accumulator
Underneath Gas Manifold Section of Catalytic Gas Reaction
Apparatus
vii
Page
11
12
13
14
15
23
25
26
28
30
33
33
34
-------
Figure
15
16
17
18A
18B
19
20
21
22
23
24
25
26
27
28
29
ILLUSTRA TIONS (Contd)
Main Part of Gas Reaction Apparatus With Flow-Monitoring
Section, Catalyst Furnace, Heat Exchanger, and Dynamic and
Static Gas Sampling Stations Located in Front of and Behind the
Catalyst Furnace
NO in Exhaust Gas
Exhaust Gas Composition Typical of S. I. Engine Using Fuel
With H/C: 1. 9
The Non-Dispersive Infrared Radiation Detector and Readout
Recorder of the Beckman Gas Analyzer
Total View of Beckman Gas Analyzer Including the Flame
Ionization Detector for the Measurement of HC Shown at Left
Principle of Operation of Infrared Analyzers
The Gas Chromatograph for Static Measurements of Gas
Exhaust Components
Three-Dimensional Schematic Representation of Experimental
Conditions for the Testing of Catalysts Noted
Three-Dimensional Schematic Representation of Experimental
Conditions for the Testing of Rare Earth Catalyst Noted
Three-Dimensional Schematic Representation of Experimental
Conditions of Transition Metal and Noble Metal Catalysts
NO Reducing Efficiency T'f Versus Flowrate cP for RE(23)75-P
NO Reducing Efficiency.,., of Catalyst RE(23)75-P Versus Tem-
perature T, Averaged Over Four Gas Compositions Tested
NO Reducing Efficiency.,., of Catalyst RE(23)75-P Versus Tem-
perature T, for Various NO Concentration and CO/NO Ratios R
Mapping of Catalyst Efficiency TJ Versus Space Velocity II Within
a Given Temperature and Flowrate Region
NO-Reducing Efficiency.,., Versus Flowrate cP
NO-Reducing Efficiency T'f Versus Temperature T, for Various
NO Concentrations and CO/NO Ratios R
viii
~
36
38
39
41
41
42
43
49
50
51
62
64
64
65
67
-------
ILLUSTRATIONS (Contd)
Figure Page
30 NO Reducing Efficiency TJ For the Empty Bed Versus Tempera-
ture T, for Various NO Concentrations and CO/NO Ratios R, in
the Catalyst Bed in Absence of Catalyst 68
31 Comparison of the NO Reducing Efficiency.,., as a Function of T
for the V cirious Catalysts Tested 69
32 Version of Parametric Design Chart for RE(23)75-P in
Nomograph Form 73
33 Revised Version of Parametric Design Chart for RE(23) 75-P 74
B-1 Schematic Diagram of the Apparatus Used for the Determin-
ation of Adsorption Isotherms 100
B-2 Adsorption Isotherm for BET Surface Area Measurement of
CuCr207 Catalyst 101
B-3 View of BET Surface Area Measurement Apparatus 103
C-1 Solar Catalyst Formulas as they Appeared Before Spalling Test 108
C-2 Components of the Humidity Chamber 109
C-3 Catalyst Formulas Inserted into Humidity Chamber 109
C-4 Solar Catalyst Formulas as they Appeared After Spalling Test 111
D-1 Solar Catalyst RE(23)75-P and Standard Catalyst CuCr207
Before Lead Poisoning Test 116
D-2 The Retort and its Components for the Lead Poisoning Test
of the Catalyst 116
D-3 The Boat With the 2PbO- PbBr2 is Inserted into the Retort 117
D-4 The Catalyst Boat is Placed on Top of the Boat Containing
the Lead Oxide-Lead Bromide, and the Retort is then Covered
With the Ceramic Block at Left 118
D-5 Operator Places Retort into Furnace to Melt the Lead 119
Compounds
D-6 Retort and Its Contents After the Lead Poisoning Test 120
D-7A Close-Up of Lead Poisoned Solar Catalyst Formula RE(23)75-P 121
D-7B Close- Up of Lead Poisoned Standard Catalyst CuCr207 121
-------
Figure
D-8
D-9
D-10
D-ll
E-1
E-2
ILL USTRA TIONS (Contd)
Both Catalysts are Immersed in Dilute Nitric Acid (Dilution
1:1) to Wash Off Poisoning Lead Deposits
Symbolic Representation of Heating Step For Washing of
Poisoned Catalysts
Results of Restoration of Poisoned Catalysts
Close-Up of Restored Solar Catalyst Formula RE(23)75-P at
Conclusion of Test
Cost of Catalytic Agents of Varying Purity
Economics of Catalysts
x
Page
122
123
124
125
131
-------
Table
I
IT
ill
IV
v
VI
VTI
VIII
IX
X
XI
XII
XITI
XIV
TABLES
Concentration Quantities for Computer Equilibrium Analysis
Values of the CO-NO Reaction Equilibrium Constant at
Constant Pressures Vs. Temperature
Coding of Rare Earth Oxide Catalyst Formulas
Certified Contents of Synthetic Exhaust Gas Components
Preheater Temperatures Used in Restoring Mixture Values
of CO and NO
Actually Used Synthetic Exhaust Gas Mixtures for Catalyst
NO Reduction Efficiency Experimentation
Definition of Gas Mixture Denoted "A"
Flowrate Range
The NO-Reducing Efficiency, and Percent Loss of CO, for
Catalyst RE(23)75-P at a Flowrate of 9000 cclmin
NO Reducing Efficiency, and Percent Loss of CO, as a
Function of Flowrate for Catalyst RE(23)75-P, at T = 1200° F,
Gas Mixture A
The NO Reducing Efficiency, and Percent Loss of CO, for
Catalyst RE(M3)75 at a flow rate of 9000 cclmin
NO Reducing Efficiency, and Percent Loss of CO, as a Function
of Flowrate, for Catalyst RE(M3)75, at T = 1200° F, Gas Mixture
A
Percent Loss of NO and CO, Versus Gas Mixture and Temper-
ature, for Empty Catalyst Bed, at a Gas Flowrate of 9000 eel
min
Percent Loss of NO and CO, as a Function of Flowrate for
Empty Catalyst Bed at T = 1200° F, Gas Mixture A
xi
Page
8
9
23
36
37
40
40
48
52
53
54
55
56
-------
TABLES (Contc1)
Table Page
XV NO Reducing Efficiency of Selected Transition Metal and
Noble Metal Catalysts, VersuS Temperature, in Synthetic
Exhaust Gas Apparatus 58
XVI NO Reducing Efficiency of Monel 400 Catalyst in Different
Structural Configuration 59
XVII Ammonia Formation for All Catalysts Studied in this Program
at T = 14000 F (Except Where Noted Otherwise), and Gas Flow-
rate of 9. 0 l/min 60
XVIII Comparison of Ammonia Generation Factor E (in % of NO Re
Reduced) for the Various Catalysts Tested and With Preheater
Plus Empty Bed Only 71
XIX State of Art in NO Reducing Catalysts 79
A-I Iso-Octane/Air Equilibrium Exhaust Composition 88
A-II Iso-Octane/Air Equilibrium Exhaust Composition 89
A-III Iso-Octane/Air Equilibrium Exhaust Composition 90
A-IV Iso-Octane/Air Equilibrium Exhaust Composition 91
A-V Iso-Octane/Air Equilibrium Exhaust Composition 92
A-VI Iso-Octane/Air Equilibrium Exhaust Composition 93
B-1 Results of Beta Surface Area Measurements of Catalysts 104
C-I Neights of Catalysts Before, During and After Spalling Test 110
D-I Catalyst Weights During Poisoning and Regeneration Experi-
ments, in Grams 121
F-I Catalyst Effectiveness 141
F-II Catalyst Effectiveness 142
F - III NH3 Production Over Solar Catalysts 143
-------
1
INTRODUCTION
1.1
BACKGROUND
The deterioration of the quality of the air we breathe, and the increase in air
pollution, is caused to a significant extent by the emissions of the internal combustion
engine, which among other things, powers our automobiles. These mobile sources
emit about
.40 percent of the oxides of nitrogen
.60 percent of the unburned hydrocarbons, and
.64 percent of the carbon monoxide,
released daily into the atmosphere within the USA (Ref. 1).
The oxides of nitrogen, in the form of nitric oxide and nitrogen dioxide,
absorb ultraviolet radiation from the sun and generate ozone and more nitric oxide,
which in turn later oxidizes to nitrogen dioxide again. The ozone oxidizes the unburned
hydrocarbons, and thus yields aldehydes and carboxylic acids. Further oxidation of
the aldehydes which then combine with N02 generates the very noxious peroxyacyl
nitrates (PAN) (Ref. 2), and thus all the principal elements of photochemical smog
have come into being. Meantime, the oxides of nitrogen, regenerated early in the
cycle, produce more ozone, the chemical reactions form a closed loop, and the smog
multiplies and spreads.
In order to control this problem, the obvious solution is to eliminate the
precursor materials from the engine exhaust, namely NOx and HC. Since the mobile
source population is increasing at a faster rate than the general population, a marked
reduction (EPA standards are for 90%) in the emission values of carbon monoxide,
oxides of nitrogen, and hydrocarbons will be needed in the next few years in order to
prevent further increases in atmospheric contamination. This means that new methods
will be required, both for the actual removal of pollutants from the exhaust gases and
for preventing the formation of pollutants in the internal combustion engine. It is
expected that a combination of these two approaches will be required in order for the
existing vehicle power plants to meet the new 1975 Federal standards for exhaust
emissions.
-------
The work"discussed in this report concerns itself solely with the control of
the emission of the oxides of nitrogen, which is created as NO in the internal com-
bustion engine, and oxidized to N02 in the air. The mixture of NO plus N02 is
commonly referred to as NOx'
Several possible approaches exist for the control of NOx in the emission
gases from mobile sources. The two main ones are the prevention of the formation of
NO in the engine, and the decomposition of the NO gas formed, either by dissociation,
by recycling, or by catalytic reduction. The Solar program concentrates on the
reduction of NO by CO under the action of :rare earth based catalysts.
The catalytic reaction under consideration is
2NO + 2CO catalys.t N + 2CO + 173 kcal
2 2
This reaction proceeds essentially in two steps, as follows:
2NO + 2CO catalyst CO + 2N
. 2
2N + M . N + M
catalyst 2 catalyst
with the second reaction being rate controlling. The literature (Ref. 3) describes
various iron, copper, and noble metal based oxide catalysts, which take three steps
to reduce NO by CO, as follows:
2NO + CO cataly~t
N20 + C02
catalyst
N20 + CO . 2N + C02
2N + M . N + M
catalyst 2 catalyst
:h.e rate controlling step is a function of catalyst temperature, at lower temperatures
It IS step 2, at higher ones, step 3. Solar's studies of rare earth based catalysts
have shown that the three step reduction process of most commercial catalysts does
not occur, the two step process is the one controlling the gas reaction.
. In addition, since the rare earth based catalyst material has refractory
proper~es,. unlike the copper, nickel and iron based ones, it can be reclaimed from
lead pOlsomng by chemical means.
-------
For these reasons, the use of rare earth based catalysts in the control of
NOx emissions from mobile sources has promised to show certain advantages over
the conventional commercial catalysts. It is the purpose of this program to demon-
strate these advantages, so that this catalyst, either by itself or in combined formu-
lations with other catalysts, can be considered by EPA to reduce NO emissions from
motor vehicles.
The use of a catalyst to remove NO from mobile sources is particularly
promising since it requires no modification to the engine and might function to control
NO emissions while extensive engine modifications and redesigning (or conversion to
a different type of power plant) are carried out. In addition, if very low emission
levels are required, a system incorporating both low emissions from a properly
modified engine and reaction of the exhaust after it leaves the cylinder has practical
inte re st.
1.2
GOALS OF SOLAR PROGRAM
The objectives of this program are limited to the selection of a rare earth
oxide catalyst or catalysts from a group of the lighter rare earth oxides, to be used
either singly or in certain mixtures, and the study of the nitric oxide reducing
efficiency of this catalyst as a function of various parameters. The program goals do
not include life tests, road tests and studies of catalyst warm-up times, as has been
done in other catalytic investigations described in the literature (Ref. 4, 5, 6, 7).
The study of the efficiency of the rare earth oxide catalyst or catalysts in the
reduction of the NO by CO and in part by the hydrogen in the exhaust, includes the
initial NO concentration for which two initial levels have been decided upon, namely,
concentrations of 1000 and 500 parts per million. The second parameter is the ratio
of the CO to the NO concentration. This ratio is related to the air to fuel ratio in an
actual exhaust gas mixture. The air to fuel ratio selected for all original testing is
close to 14 which gives us a CO/NO ratio of 10. This parameter is varied downward
to 5 in order to obtain a true measure of the NO reduction efficiency of the catalyst.
As shall be shown in the theoretical section, the equilibrium concentration of NO is a
function of the CO/NO ratio and decreases with an increase in the CO/NO ratio.
Therefore, a large excess of carbon monoxide aids in the reduction of the nitric oxide
to nitrogen and carbon dioxide. The third parameter of importance for the parametric
study of the efficiency of the catalyst is the temperature range. The temperature
range selected lies between 600 and 1600° F. Finally, the flow rates selected lie
between 460 and 18,000 cc per minute, which corresponds for our catalyst bed con-
figuration to automobile flow rates of between 5 to 200 cubic feet per minute.
One of the secondary parameters in the catalytic reduction reaction is the
catalyst surface, whereby we are using the maximum surface obtainable under the
conditions and methods of the preparation of the catalyst. The goal is to be in the
-------
neighborhood of lO'square meters per gram. Another parameter is the effect of
competing reactions on our main catalytical reaction. This competing reaction con-
sists of the oxidation of carbon monoxide by the oxygen present in the exhaust gas
under the fuel-rich conditions. Thus the "effective" CO/NO ratio is decreased by the
oxygen. Another parameter includes the effect of the water vapor on the overall gas
reaction kinetics. Another competing reaction is the generation of ammonia by the
free hydrogen in the exhaust gas. For our purposes we have introduced pure hydrogen,
mixed with nitrogen into our synthetic gas exhaust mixture, in order to see whether
the catalyst will generate an appreciable amount of ammonia from the reduced nitric
oxide. Experience has shown in the literature cited that the ammonia will be oxidized
back to nitric oxide in the oxidizing stage of the double catalyst bed and therefore, a
generation of ammonia will reduce the overall NO removing efficiency of the catalyst.
These are the parameters studied in the present program. In the following
section we shall introduce the theoretical analysis describing the CO/NO equilibrium
conditions and their dependence upon the various parameters defined above and the
CO/NO reaction kinetics. This section is followed by the experimental procedure
section, which summarizes the studies made on the rare earth oxide catalyst as such
and the catalytic gas reaction experiment, including the analytical apparatus and the
procedures used to obtain the data. Section 4 discusses the experimental results,
first as the function of parameters investigated, second as a reduction efficiency for
each of the catalysts selected, and finally the ammonia generation factor for each of
the catalysts. Then follows the discussion which will show the relationship of this
catalyst to some of the others on the basis of our data. Section 6 concludes this
report. The economics of the rare earth oxide catalysts, as well as results of
secondary studies regarding their physical and chemical properties, are discussed
in the Appendices.
-------
2
THEORETICAL ANALYSIS
In this chapter we shall derive the equilibrium concentration of nitric oxide
as a function of various reaction parameters, and calculate the reaction kinetics of
our equation 2NO + 2CO ~ N2 + C02 using a simulated catalyst.
2.1
DEFINITION OF CONCEPTS
The main concept to be defined is the one of catalytic efficiency. The
efficiency can be denoted as the ability of the catalyst to obtain the equilibrium con-
centration of the reaction rapidly. It is thus a measure of the reduction below the
initial NO concentration and toward the equilibrium NO concentration. The catalyst
is 100 percent efficient if within a short time (a few seconds) it will reduce the NO
concentration down to equilibrium. The equilibrium concentration in turn is defined
as the chemically lowest possible concentration of the reagent left after the reaction.
It itself is a function of various parameters, as shall be seen in the next section.
Another concept entering into the calculation of the efficiency is the one of the
reaction kinetics. This can be defined as a measure of the time rate of progress of
the reaction toward equilibrium. The catalytic reaction rate therefore is the difference
in the progress of the reaction in the presence and absence of the catalyst. If in the
absence of the catalyst the reaction does not proceed at all, as in the case with the
NO-CO reaction, then the catalyst determines by itself the reaction kinetics. This
concept will be analyzed for a simulated catalyst in Section 2. 3.
The conclusion of the chapter will summarize the results and their significance
for the experimental determination of reaction efficiency of the rare earth oxide
catalysts considered in this program.
2.2
EQUILIBRIUM COMPOSITION ANALYSIS
The equilibrium constant Kp for the reaction
1
CO + NO ---.- - N2 + CO
2 2
(1)
-------
may be derived in'terms of the equilibrium constants with respect to the elements
in their standard states. The three equilibrium expressions pertinent to Equation (1)
are the following:
.!N 1 ~NO (P-6)
+ -0
2 2 2 2
C(g) + 1:..0 ~ CO (P- 9)
2 2 -.-
C(g) + 02 ~ C02 (P-10)
where the three elemental reactions are labelled according to Penner's Table of
Equilibrium constants (Ref. 8) Kp, #, which are constants based on equilibrium
pressures, the temperature dependence of which are tabulated in the above reference.
In particular, the following equilibrium constants are tabulated as a function
of temperature, which corresponds to the above three equilibrium expressions:
(graphite, C(g), which is not a gas, and does not have a partial pressure, thus does
not appear in Kp, 4#:
1/2
Kp, 6 = p NOI (P N PO) ,
2 2
(2)
K = P Ip 1/2
P, 9 CO 0 '
2
(3)
Kp,10 = PCO Ipo
2 2
(4)
Combining Equations (2), (3) and (4) yields an expression for Kp, i. e. ,
K = K IK K = (p )1/2 I
P P,lO P,9 P,6 N PCO PCO PNO'
2 2
(5)
FO: this analysis it is preferable to work with mole fractions (or percent
b~ v~lume) Instead of partial pressures since the experimental data is in parts per
mIllIon. The mole fraction of species i is defined as
x = Pip
i iT'
(6)
-------
where Pi is the partial pressure of species i and PT is the total pressure, i. e. ,
PT = ~.p.,
1 1
( 7)
Mole fraction values can be easily converted to parts per million by the relation
6
PPM = X x 10 .
i
(8)
Introducing Equation (6) into Equation (5) leads to the equation
1/2 = X 1/2 X /X X
P T Kp N CO CO NO'
2 2
(9)
'.Vhen equilibrium is reached according to Equation (1), the following mole
fractions of each component will be present:
XNO = XNO - ~,
XCO = Xeo - ~,
XN
2
= X ° + 1/2 ~,
N2
(10)
X
CO
2
= XO + ~
CO '
2
where XN'o, XCO' XN ' and XC02 represent, respectively, the initial mole fractions
of NO, CO, N2, and Cb2' The above equations indicate that for every ~ mole fraction
C02 formed, 1/2 ~ mole fraction of N2 is formed, and ~ mole fraction of each, NO
and CO, disappears. F rom Equation (9) it now follows that
PT1/2Kp = (X n + 1/2~) 1/2 (X ° + ~) ~X 0 - ~)(XCOO -~) (11)
N2 C02 l' NO
It is evident that, for given values of the total pressure and the equilibrium constant
Kp, the value of ~ can be calculated readily for given initial mole fractions of the
four reactants.
-------
A computer program has been written for several values of Kp, P, and initial
reagent and reactant compositions, over the temperature range between 600 and 1600° F,
and a pressure range between 1. 0 and 2.0 atmospheres. Table I summarizes the
quantities of Equation (10) as they are redefined for the computer program, to simplify
the expression in Equation (11).
T ABLE I
CONCENTRATION QUANTITIES FOR COMPUTER
EQUILIBRIUM ANALYSIS
~Quantity Symbol in Symbol for
Equation (10) Computer Analysis
NO Equilibrium Mole Reaction XNO Y
Initial CO Mole Fraction XCoO PI
Initial NO Mole Fraction °
XNO P2
Initial N2 Mole Fraction X ° 0'1
N2
Ini tial C02 Mole Fraction °
XCO 0'2
2
Mole Fraction of NO Reduced ~ X
Thus, from Equation (11) we obtain, in the new terminology,
and
[X2 + {3 {3 - X({3 + (3 )] P 1/2K = 1/2
1 2 1 2 T P (0!1 + 1/2X) (0!2 + X)
-Y = X-{3
2'
(12)
The. expression X = {32 is very close to the solution, i. e., the equilibrium concen-
tration of NO approaches zero, therefore
(X - (3 ) (X - (3) =
1 2
1/2
(O! 1 + 1/2 (32) (0!2 + (32)
P 1/2
T Kp
= D
(13)
If {3 =
1
{32 '
2
(X - (3 )
2
=D; -Y = X- {3 = -.fD
2
-------
If {31;c {32'
-D
-y = X - {32 =
f32 - f31
The temperature dependence of Equation (13) is given by the equilibrium
constant Kp' which is listed in Table II in 2000 F increments, starting from 6000 F and
ending at 16000 F. The constant Kp has dimensions of inverse pressure to the 1/2
power and is therefore called K for pressure. Equation (13) has be'en solved for
various parameters, as followsf
TABLE II
VALUES OF THE CO-NO REACTION EQUILIBRIUM CONSTANT AT
CONST ANT PRESSURES VS. TEMPERATURE
Temperature K
(0 F) P
600 1. 688 x 1033
800 4.28 x 1022
1000 9.88 x 1018
1200 1. 70 x 1016
1400 1. 54 x 1014
1600 1. 155 x 1012
The temperature changes from 600 to 16000 F in accordance with the values
of Table II for K. Pressure values selected are 1 atmosphere,!. 5 atmospheres, 2
atmospheres. ~e original CO concentration are defined as 0.001 mole fraction,
0.003, 0.006, and 0.010. The original NO concentrations considered is 1,000 parts
per million and 500 parts per million. Original concentrations of C02' the reaction
product, is none and 0.002 mole fraction; the original nitrogen concentrations, another
product of the reaction, is in 0, 0.002, and 0.2 mole fractions. The equilibrium
concentration has been calculated by means of a computer for all combinations of
these points at each temperature, so that the total of 648 points was obtained. The
data were then analyzed according to the various parameters introduced. During this
analysis, no regard was taken of the fact that the equilibrium concentration computed
was negligibly small, which was due to the value of the equilibrium constant K .
From the tabulation set up by the computer, data were collected for equilibriu~
-------
concentration Y versus temperature, with all the other parameters held consta~t.
In further analysis it was shown that the value of the NO equilibrium concentratIon Y
was independent for each parameter, and the effect of all the parameter~ w~s merely
to multiply their total factors to obtain the actual equilibrium concentratIon III mole
fraction for a given set of parameters. Therefore, it was possible to plot the data
separately for each parameter varied and this has been done in the following figures.
Figure 1 shows the temperature dependence of the equilibrium NO concentra-
tion between 600 and 1600° F. As can be observed from Figure 1, the curve is of a
parabolic type; increasing from utmost small numbers in the low temperature range,
rapidly at first and slower later, to still very small numbers at the higher tempera-
ture range, reaching, for a given set of parameters, the value of 10-12 mole fraction
for 1600° F. The normalizing set of parameters used is a CO-NO ratio of 2, a gas
pressure of 1 atmosphere, and the absence of any initial N2 and C02 concentrations.
After eliminating the effect of the temperature on the NO equilibrium concen-
tration, the factor Y, which is the equilibrium NO concentration, has been normalized
to the parameters given above, and the effect of the remaining parameters on it has
been plotted out in the following figures.
Figure 2 shows the effect of the CO- NO ratio on the NO equilibrium concen-
tration. Since the CO-NO ratio effect is independent of the temperature, pressure, or
any other equilibrium parameter effect, it has been normalized and the change of Y N'
the normalized NO equilibrium concentration versus the CO-NO ratio has been plotted.
As can be seen, the CO-NO ratio reduces the NO equilibrium concentration rapidly at
first and slower later. It looks similar to an exponential curve. We have had also
data for a CO-NO ratio of one, however, in terms of normalized YN' the amount is
orders of magnitude greater for the CO-NO ratio of one than it is for values of two
and higher. In addition, a slight temperature dependence of the CO-NO ratio value of
one has been found. However, this temperature dependence disappears for higher
values of the ratio, so that a division of the parametric influences could safely be
made.
The effect of pressure on the NO equilibrium concentration is shown in
Figure 3. As can be seen, with increased pressure the equilibrium concentration is
reduced and it oecreases to about 70 percent of the value of the pressure data point at
1 atmosphere by the time it reaches 2 atmospheres. This effect, however, is not as
pronounced as was the CO-NO ratio effect.
After considering the pressure, temperature, and ratio parameters, the
effect of reactant concentrations on the CO-NO reaction and the equilibrium concen-
tration of NO was evaluated. Figure 4 shows the data for NO concentrations between
10-3 mole fraction and 0.7 mole fraction of N2' Similarly, the data for C02 go from
-------
900 1100 1300
TEMPERATURE, . F
EQUILIBRIUM CONCENTRATION OF NO (Y) VS. TEMPERATURE (T)
-12
-13
-16
-17
-18
-19
z
o -20
~
U
-------
1.0
6
~
N
~
~ 0.5
I:!::
o
~
"..
0.1
FIGURE 2.
0.9
0.8
0.7
0.6
0.4
0.3
0.2
2
3
6
CO/NO RATIO
8
9
10
4
7
5
EFFECT OF CO/NO CONCENTRATION RATIOS ON NO EQUILIBRIUM
CONCENTRATION NORMALIZED FOR Y AT RATIO OF 2
a mole fraction of 10-3 to mole fraction of 0.2 with an extension to 0.7 also. In
normal conditions, the N2 concentration would be 0.7 mole, or 70 percent of nitrogen
in the exhaust gas; the initial C02 concentration would be about 14 percent or 0.14
mole. Figure 4 shows the increase in the NO equilibrium concentration normalized
to zero C02 and zero N2' with increases in N2' but C02 remaining constant at zero,
which is the top curve; and increase in C02 concentration with N2 remaining a constant
at zero in the bottom curve. As can be seen, the N2 concentration has a bigger effect
-------
1.0
0.9
0-
w
N
.....
~
<1::
::a 0.8
p::
o
~
~
0.7
0.6
FIGURE 3.
1.0
2.0
1.5
P (ATM)
EFFECT OF PRESSURE ON NO EQUILIBRIUM CONCENTRATION,
NORMALIZ ED FOR Y AT 1 ATMOSPHERE
on the NO equilibrium concentrations than the C02 concentration. From the obser-
vation that the effect of the combined N2 and C02 the concentrations on Y N equals the
product of both parameters taken separately, several graphs have been selected for
Figure 5, where the NO equilibrium concentration is shown as a function of initial N2
and initial C02 concentration. The N2 concentration ranges continuously from 10-3
to 0.7 mole fraction. Selected values for the C02 concentration from zero to 0.1
mole fraction have been chosen.
This analysis shows that, in order to obtain the equilibrium NO concentration
at a given set of parameters, the value from Figure 1 for the temperature is multiplied
by the YN value for the ratio effect in Figure 2, the pressure effect in Figure 3, the
initial nitrogen concentration effect and the C02 concentration effects in Figure 4, and
thus for any given set, the NO equilibrium concentration Y is obtained. We shall give
some examplies of this below.
-------
1000
800
4
EFFECT OF INITIAL N2 CONCENTRATIOO
FOR INITIAL C02 CONCENTRATION - 0
I
-t-
I
600
400
200
c::1
~ 100
N
~ 80
-<
~ 60
p::
o
z
Z 40
o
-
E-<
-<
p::
~ 20
~
C)
~
C)
o 10
Z
8
6
EFFECT OF INITIAL C02 CONCENTRATION
FOR INITIAL N2 CONCENTRATION - 0
10-3
2
4
6 8 10-2 2 4 6 8 10-
ALL CONCENTRATIONS IN MOLE FRACTION
2
4
6
8 10C
FIGURE 4.
DEPENDENCE OF NO EQUILIBRIUM CONCENTRATION OF EITHER NZ
OR C02
As our discussion has shown, the equilibrium concentration of NO for various
values of T, P, R, N2' and C02 initially is given by:
NO [equil.] = Y(T)YN(P)YN(R)YN(N2)YN(C02)
These values are given in Figures 1 through 4. We shall demonstrate this on the
following examples:
-------
0.7
0.5
0.3
10-1
~ 0.07
!:
U
< 0.05
0:
'"
'"
-'I
j 0.03
'"
z
-'I
<
E
~
10-l
0'007~
V'OOl
0.00:1
]0-3
1
FIGURE 5.
I
co . 0.0
.
CO~ IN MOLE FRACTIONS
60 80 100
400 600 800 1000
10.000
.woo
4000 6000
lOO
NO CONCENTRATION, NORMALIZED (NO.1 H CO. . O. N 1.0)
INFLUENCE OF INITIAL REACTION PRODUCT CONCENTRATION ON
NO EQUILIBRIUM CONCENTRATION NORMALIZED TO NO INITIAL
PRODUCT CONCENTRATION
1.
At T = 1000° F, R = 3, P = 1 atm., C02 = 0 moles, N2 = 0 moles.
F rom the figures we find mole fraction of
Y(T) = 10-19
Y N(R) = 0.29
YN(P) = 1
Y N(N 2) = 1
Y N(C02) = 1
-------
Therefore, the mole fraction of
-19 -20
NO = 10 x 0.29 x 1 x 1 x 1 = 2.9 x 10
[equi1.]
2.
In the above example, we have an initial concentration of C02 = 0.01,
and N2 = 0.70 mole fraction. Then Y N(N2) = 700, Y N(C02) = 4.6, and
the mole fraction of
-19
NO = 10
[equil.]
x
-17
0.29 x 1 x 700 x 4.6 = 9.3 x 10 .
3.
At T = 1600° F, P = 1. 25 atm., R = 2, C02 = 0.02, N2 - 0.70. Here
we find mole fraction of
-12
Y(T) = 10
YN(P) = 0.89
Y (R) = 1
N
Y (N) = 700
N 2
Y (CO) = 6.4
N 2
and the mole fraction of
-12 -9
NO = 10 x O. 89 x 1 x 700 x 6. 4 = 3. 987 x 10
[equil.]
or about 4 parts per billion.
It may be observed that for the conditions of Example 3, the NO equilibrium
concentration no longer is significantly small, and for a CO/NO ratio of 1 will grow
into the ppm range.
2.3 STUDY OF CO-NO REACTION KINETICS
2.3.1
Catalytic Reactor Program -- Thermochemical Calculations
The work accomplished has centered on estimating the effect of the preheater
located upstream of the catalytic reactor on the composition of a simulated exhaust
gas mixed at room temperature. The variables are initial concentration, preheater
-------
tempe ratu re, and residence time at tempe rature. The procedure then was to use the
Dynamic Science Generalized Kinetics Program to predict the change in equilibrium
concentration of a gas mixture when heated to automotive exhaust temperatures.
The initial effort was to gather data representative of the range of automotive
exhaust emissions. This provided a realistic basis for the simulation of automobile
exhaust in terms of the trace contaminant species (CO, NOx, He). Most of the data
presented in Volume 6 and 12 of the SAE Technical Progress Series on exhaust
emissions were compiled. Several distinct operating modes were recognized; these
being: idle, cruise at 30, 50 and 60 mph, acceleration and deceleration from low
velocity (30 mph) and high velocity (50 mph).
Generally CO decreases as automobile speed increases. During acceleration
and deceleration, CO reaches levels neaT those at idle (nearly 10% of entire exhaust
composition). NO generally increases with velocity but is very low (:::::50 ppm) on
deceleration and at idle. Unburned hydrocarbon measurements show that the trend is
generally toward decreasing HC with velocity. Increases are noted during acceleration.
HC measurements during deceleration from high velocity are an order of magnitude
higher than all the others (Ref. 9). These results then represent a rather large scale
summary of existing measured emission levels. To assess preheater effects, it is
necessary to define the rest of the chemical system, i. e., major species concentrations,
A series of chemical equilibrium calculations was made for two reasons:
1.
To determine the major species concentrations in the exhaust which are
at or very near equilibrium values.
2.
Determine the degree of departure of the measured emission data from
equilibrium concentrations; for if the measured values are in equilibrium
at the exhaust temperature, no conversion can occur in the reactor.
A scattering of computer runs, designed to look at the range of exhaust
conditions, were made with the Generalized Kinetics Program. The results, in
tabular form, are given in Appendix A. Since a wide scattering of conditions were
run, the presentation of these restults in graphical format was difficult, and the
tabular form (Table A-I) preferred. In summary, the results show, however, that:
1.
Nonequilibrium CO and NO are "frozen" in the exhaust gases at tempera-
tures and times representative of the preheater in the absence of oxygen.
2.
At temperatures higher than approximately 1600° F, trace amounts of
oxygen can oxidize from 10-50 percent of the initial CO to C02'
-------
3.
Tl1e addition of oxygen has no effect on the NO and only affects CO.
4.
At A/F ratio of 14 ammonia generation is limited to the lower tempera-
tures and does not exceed 20 ppm. At richer A/F ratios, NH3 concen-
tration may reach up to 30 ppm.
2.3.2
Effect of Gas Phase Reactions on Simulated Catalytic Conversion of NO
A crude attempt was made to assess whether gas phase reactions would effect
the catalytic conversion of NO. For this evaluation, the catalytic process was first
modeled by assuming the catalyst affected only the reaction:
NO + CO Ji. C02 + N
(14)
The rate constant k was determined by integrating, in closed form, the rate expression
for [NO].
d [NO] = - [NO] [CO] k
dt
which gives
1
k = t[CO]
[NO]. 't. I
1m la
log [NO]
final
(15)
It was then assumed that the catalytic reactor would reduce [NO] to approximately 10
percent of its initial value in 0.1 sec, i. e., [NO)f' l/[NO]. .t' I ~ 0.1 for t = 100
Ina 1m la
msec. Using these values, an effective rate constant k was computed. This k was
then the input into the kinetics program for the reaction NO + CO - C02 + N, along
with the other gas phase reactions and the program was then run for 100 milliseconds.
The calculated concentration of NO after this time interval for an initial NO concen-
tration of 7.12 x 1015 molecule/cc("'1000 ppm) was 1. 58 x 1014 molecule/cc as com-
pared to an NO concentration of 8.47 x 1014 molecule/cc computed us ing Equation
(16). This result was somewhat surprising, since it indicates that the gas phase
reactions accelerate the catalytic NO conversion process.
In order to evaluate whether this conclusion was valid or only a result of the
constraints on the chemistry due to the input catalytic reaction, a second set of
calculations were made. For these calculations, the reaction
2NO + 2CO k 1 2CO + N
- 2 2
(16)
-------
was used to represent the catalytic reaction, thereby eliminating the production of
free N atoms (which could reduce the NO concentration by reactions such as N + NO=
N2 + 0). The rate constant for this reaction k1 was similarly evaluated by a closed
form integration for [NO]final/[NO]initial ~ 0.1, at t = 100 msec. This constant and
reaction (16) were input into the final kinetics program using the same initial conditions
used previously. This calculation yielded an NO concentration of 2. 12 x 1014 molecule/
cc compared to 3.78 x 1014 molecule/cc for the closed form calculation. Again the
kinetic NO concentration is lower, indicating a gas phase enhancement of the NO
conversion.
2.4
SIGNIFICANCE OF THEORETICAL STUDY OF EXPERIMENTAL WORK
The data of the theoretical analysis in the previous two sections make it
possible to practically define the reaction efficiency of the catalyst studied experi-
mentally. Regarding the equilibrium concentration of the nitric oxide, it is zero on
our scale, which is in the parts per million range. The catalytic efficiency therefore
can be defined as the ratio of the difference of the initial minus the final concentration
of the nitric oxide with the initial concentration of nitric oxide, multiplied by 100
percent.
As far as the reaction kinetics is concerned, there is none without the catalyst
for the reaction under study. Therefore, the catalyst is the sole determining factor of
the reaction kinetics, and can be evaluated accordingly. Finally, regarding the com-
peting reactions, the analysis showed that some of the secondary reactions are support-
ing our basic relation and some are hindering the catalytic action. This particular
result is important, since no single reaction component experiments were being run,
but the experiment was run in the total synthetic exhaust gas mixture containing all
the components of such a mixture. The effect of the ammonium generation as a
fraction of the nitric oxide reduced, which is another competing reaction, was not
considered in the reaction kinetic program as such.
We shall now proceed with the discussion of the experimental phase of the
program.
-------
3
EXPERIMENTAL PROCEDURE
The experimental part of the program was divided essentially into two phases:
the rare earth oxide catalyst study and the study of the gas reaction. Therefore, this
chapter will discuss first the screening and selection of the rare earth catalyst,
briefly summarize its properties, and then pass on to the description of the experi-
mental procedure and apparatus. The actual data obtained from the catalytical gas
reaction study, as well as the ammonia generation factor, will be discussed in the
next chapter, headed Experimental Results.
3.1
RARE EARTH OXIDE CATALYST STUDY
3.1.1
Catalyst Composition Selection and Screening
In selecting a catalyst for the reduction of NO by CO, the following criteria
originally were used by Solar:
The catalyst should be stable and refractory
The catalyst should be able to be regenerated after poisoning.
Based on these criteria, the best class of catalyst material suited under these condi-
tions appeared to be the class of the rare earth oxide catalysts. In the following
sections, we shall describe our choice of the rare earth oxide catalysts, the way we
coded the catalyst formulas, the narrowing down of the selection of the REO catalyst,
and the economics of the REO catalyst compared with other standard catalytical
materials.
(a) Choice of Rare Earth Oxides
In the group of rare earth oxides, the light rare earth oxides are of special
interest because they contain two elemental oxides which are highly unstable and
therefore very well suited for catalytic action. The instability is a phase change
between various of the oxidation stages, i. e., as a function of temperature and oxygen
pressure, the element increases its valence for oxygen or decreases it. The two
elements of interest are cerium and praseodymium. Especially in the case of
-------
praseodymium, various phase changes can be noted, whereby the degree of oxidation
of the praseodymium varies from pr02 to Pr203' The change from one to the other
does not occur in one step but in several steps (Ref. 10). The change from pr02 to
PrO occurs in nine steps or nine phases, which are designated in order from the
origt~~l Pr02 phase, which is the alpha-phase, along .the Greek alphabet as beta,
gamma, delta, epsilon, phi, sigma, pi, and theta. PI and th~ta are. two. forms of
Pr01 5' Due to these various phase changes, the praseodymIUm oXIde IS very often
denot~d PrOx, where the X can have any value between 1. 5 and 2. Most common is
the value of X = 1. 833, which is also written as Pr6011' Other values of X are 1. 8,
1. 78, 1. 71, 1. 70, 1. 67, and 1. 5. In a similar fashion, but not quite as complicated,
the cerium oxide switches its oxygen content between Ce02 and Ce203' The Ce203
oxide is white in color, while the Ce02 is a light yellow. Similarly, the Pr6011 is a
black powder and the Pr203 is a light green one. The neighboring oxides, including
those of lanthanum and neodymium, and the combination of praseodymium and
neodymium known as didymium, have also been studied in this connection.
The raw materials from which our catalysts are formulated, are shown in
Figure 6. The upper row shows the pure rare earth oxides of 99 percent purity,
namely, from left to right, the white La203' the light yellow Ce02' the black Pr6011'
and the light blue Nd203' The bottom row shows commercially available mixed oxides
which we are also using due to their economy, namely the yellowish brown commercial
didymium oxide on the left, and the reddish brown mixed rare earth oxide on the right.
(b) Coding of the Catalyst Formulas
The catalyst formulas based on the light rare earth oxides, especially the
four mentioned, have been coded with a designation of RE(a number). RE stands for
a pure rare earth oxide formula, and the value of the number in parentheses following
the RE is shown in Table III. In the case where the number in parentheses ends in 0,
a pure rare earth oxide of the particular element, as shown in Table III, is being con-
sidered. If the number in parentheses is two digits, neither of which is 0, the formula
contains the two rare earth oxides designated. RE(43) denotes the didymium oxide,
whereby the majority oxide present is the neodymium. Formula RE(43)M denotes a
commercial mixture of the didymium oxide which, beside neodymium and praseo-
dymium, contains especially lanthanum and a number of other rare earth oxides, and
only a trace of cerium oxide. The majority oxides in this case is lanthanum and
neodymium. In Solar Formula No.8, the cerium oxide is the majority. The formula
RE(M) denotes a commercially available mixed rare earth oxide, wherein the majority
component is cerium oxide, followed by those of lanthanum and neodymium. The Solar
Formula No. 10 consists of a mixture of these mixed oxides with pure praseodymium
oxide. Therefore, the number in parentheses is M3. The mi.<:ed oxides are the
majority component in this particular formula. In the study reported here, we have
-------
"
FIGURE 6.
RARE EARTH OXIDES AS RAW MATERIAL FOR SOLAR CATALYST
FORMULAS
TABLE III
CODING OF RARE EARTH OXIDE CATALYST FORMULAS
Component
Formula Symbol Name
RE(10) La203 Lanthanum Oxide
RE(20) Ce02 Cerium Oxide
RE (30) PrOx Praseodymium Oxide
RE(40) Nd203 Neodymium Oxide
RE(43) Nd203 + PrOx (D203) Didymium Oxide, Pure
RE(43)M D203 Didymium Oxide, Commercial
RE(23)75 Ce02+ PrOx Solar Formula No.8
RE (M) REO Mixed Rare-Earth Oxides
RE(M3)75 REO + PrOx So]~r Formula No. 10
-------
been concentrating on the pure and commercial didymium and on Solar Formula No.
Sand 10. The total catalyst producing effort, and its result, is shown in Figure 7
which encompasses all RE catalyst formulations prepared for this program.
(c) Elimination of Various Formulations and Selection of Catalysts to be Studied
Elimination of Pure Elemental Oxides. Formulas RE(20) and RE(30) each showed
good catalytic testing and catalytic properties in early testing. However, their com-
bination, RE(23), was superior in performance to that of each of their components and
reduced the lower temperature threshold of catalytic activity by 100° F. Therefore,
the pure elemental oxides, RE(20) and RE(30), were not considered for the program
in their unmixed state.
Formula RE(10) by itself had not been tested; when added, however, to
Formula RE(20), which resulted in the formulation called RE(21) 75, it was found that
the overall catalytic efficiency of RE(21) was lower than that of the pure oxide formula
RE(20). Thus, RE(10) served merely to dilute the efficiency of the pure Ce02' and on
this ground the elemental oxide RE(10) was eliminated.
Elimination of Simple Oxide Mixtures. Didymium oxide, RE(43), comes in the form of
a mixture of pure oxides, the latter of which contains a quantity of neodymium oxide.
In Table lIT, our formula RE(43) denotes the pure didymium oxide, and RE(43)M, the
commercial mixed oxides containin formulation. Both formulas were eliminated on the
basis of the physical unstability due to hydration causing the pellets to disintegrate.
Formula RE(M), the commercially available mixture of RE oxides consists principally
of cerium oxide and lanthanum oxide, with smaller admixtures of the more abundant
rare earth oxides and a low praseodymium oxide content. This essentially diluted
the cerium oxide, and the conclusions reached for formula RE(21) are valid for
RE(M) also, causing its elimination. An enrichment of RE(M) with RE(30) however,
resulted in the effective and economical formulation of RE(M3), which remained in
the program.
Finally, the low purity RE(23) 75 catalyst as designated in Table III, Solar
formula No.8, was eliminated due to lack of sufficient spalling resistance, which
shall be shown in Appendix C, and the presence of a low melting impurity in the
formula, which interferred with its processing and its effective surface area.
RE Catalysts Evaluated for Their NO Reducing Efficiency. The catalyst formulas
selected for further study are thus the following:
First, RE(23)75-P -- a high purity oxide mixture of cerium dioxide and
praeseodymium oxide and second, RE(M3)75, a mixed oxide variation of the above
mentioned formulation, which has better economic features.
-------
. '(.{,} .'c.'<
,,~~~
--
11 -
I
"t;r:"'tf~~
1"'4l'~" ,
~~;:;"
I""''' "~-"",",,,'
, t7~~~'.,
,~~,,,
'"" ';~'\'<\~I~;~t
I:\:)
Q1
,'"' -
!"",''''''''~,
\!"h,,~~ ;.""'r-.
, ,
, '.~
,,::;-~;;;.."'--.
f{t..\~W:"I' "K.'~ "'.'!!I>::t1r,>..
11'\# 'r~......-.'
REd;j:f3.X ,~ 1'r<-~""",
~F;«VSI-.41'q,'-i<'
-------
3.1. 2
Catalyst Porm
The selected catalyst formulations have been sintered into pellet form, a
right circular cylinder, of radius of 0.1 inch and height of 0.2 inch.
Variations in the preparation of parameters of the catalyst pellets have been
made with the object of increasing the surface area. The preparation parameters have
been changed in various steps in order to obtain a surface area comparable with that
of the copper dichromate. The parameters were changed for both rare earth catalyst
formulas, the pure RE(23)75-P and the mixed RE(M3)75. Hence, four variations in
the preparation parameters have been decided upon and executed for each of the two
formulas being used. The result is a grouping of eight various catalysts. These
have been pelletized and prepared for sintering, as shown in Figure 8.
The first row on top shows the grayish formula RE(23)75-P, and the second
row the dark brown formula RE(M3)75. Each two boxes contain one-half pound of
catalyst and these same pairs of boxes contain catalysts made under the same prepara-
tion conditions. The preparation parameters vary from one group of two boxes to the
next.
The catalyst pellets are shown as they appeared after completion of sintering
in Figure 9. Formula RE(23)75-P has turned black, while RE(M3)75 is now reddish
brown.
The results of the variation of the catalyst preparation parameters on the
surface area indicated only a limited success within the parameter range selected.
The two RE catalyst formulations showed a greater difference in surface area among
themselves, than the variation in preparation parameters could achieve for either of
them.
3.1. 3
Rare Earth Oxide Catalyst Properties
(a) Surface Area
The catalyst surface area was measured by the BET method for both rare
earth oxide formulations and the copper dichromate. Details are given i~ Appendix B.
The results are that for catalyst RE(M3)75 the surface areas were found to be in the
range from 0.1 to 0.2 m2/gram, for catalyst RE(23)75-P in the range between 0.4
and 0.8 m2/gram, and for CuCr207 at 11 m2/gram. The range for the rare earth
catalyst was obtained in part by the variation of the preparation parameters, men-
tioned in the prior subsections.
-------
r""""'~?'"';-'
.=-=-~. . ;; " ...> ,....,,*.,,~ '
.. -"
,I
!
JI
,u'/-~_. .' _.,,-_..._-~.,.-
~
FIGURE 8.
VIEW OF NEWLY PELLETIZED REO CATALYSTS, SEGREGATED BY
FORMULA AND PREPARATION PARAMETERS, PRIOR TO SINTERING
-------
.1
FIGURE 9.
,~~".~\,':'7V.. ,~.~.-..-:;;,;; ,
.:,~ .
:i'
l
,
CATALYST SAMPLES OF FIGURE 8 AFTER SINTERING
-------
(b) Mechanical Properties - Spalling Resistance
Details of the measurement on spalling resistance are presented in Appendix
C. Of the two rare earth oxide formulations of interest, RE(23)75-P showed the best
resistance to cracking due to the influence of atmospheric water vapor.
(c) Chemical Properties - Lead Poisoning and Regeneration Study
A study of the mechanism of lead poisoning and preliminary catalyst regen-
eration was made on catalyst formula RE(23)75-P, and compared with the transition
metal oxide combination CuCr207' The experimental procedures and results are given
in Appendix D. The result indicates that rare earth oxide catalysts can be regenerated
which is one of the reasons REO catalysts were considered for NOx control in this
program. The loss to catalyst RE(23)75-P was only 4 percent after one regeneration
run.
(d) Economics of Rare Earth Oxide Catalysts
A study of the availability, cost, life cycle aspects, and comparative
economics of the rare earth oxide catalysts is presented in Appendix E.
3.2
EXPERIMENTAL STUDY NITH OTHER CATALYSTS
The study of catalytic efficiency for the reduction of NO by CO, with the
minimum quantity of NH3 generation as a byproduct, was also extended to Monel 400,
discus sed by L. S. Bernstein et aI, in Reference 4, and coppe r dichromate. The
inclusion of these two catalysts allows a comparative analysis of the efficiency of the
REO catalysts with the other ones, evaluated under identical conditions.
Data obtained with Monel 400 and CuCr207 will be presented in Chapter 4,
and discussed along with the results of RE(23)75-P and RE(M3)75, in Chapter 5.
3.3
THE CAT AL YTIC GAS REACTION EXPERIMENT
This section describes the experimental apparatus, the synthetic exhaust gas
mixture used for the experiment, the analytical apparatus used for the monitoring of
the emissions for our catalytic gas reaction, the procedure for conducting the experi-
ment, and the method of reducing the data.
3.3.1
Gas Flow Apparatus
The experimental apparatus used for the catalytical gas reaction 2NO + 2CO =
N2 + 2C02 is shown in Figure 10. According to this figure, we can distinguish several
main components of the apparatus:
-------
PREHEATER FURNACE
,---------------,
I PREHEATER iH 316 SS :
L__-------------- _I
~
'"
'" '" '" Z '"
Z z :z: 2S z
is 2S '" '" is 2S
o z '"
o 0 C) ", :I: '"
Z C) 0 "' :I:
C)
FIGURE 10.
SCHEMATIC OF CATALYTIC GAS REACTION APPARATUS FOR
EXPERIMENTA TION TNITH SYNTHETIC EXHAUST GAS MIXTURES
1.
On the bottom left of the schematic we have the manifold which is
connecting the various gas tanks. The gas tanks involved are those
containing NO, CO, 02' C3HS and H2 in nitrogen, a C02 gas bottle
and a nitrogen gas bottle.
2.
The manifold leads the gases into the gas mixing container which has
a volume of about 10 cubic feet at elevated pressure. This gas mixing
container is an accumulator, that is a constant pressure vessel, con-
taining a diaphragm which divides accumulator's interior into two parts.
The gas is admitted into the accumulator by the method of partial
pressures, read on the pressure gauge Po' Once the full pressure,
made up of the sum of partial pressure of all the gas components, is
attained, one can, with the use of the diaphragm, keep the pressure in
the accumulator constant by admitting nitrogen into the bottom part of
the container (not shown in the sketch of the apparatus). In this manner,
-------
the diaphragm will exert a constant pressure to the gas remaining in the
accumulator; thus the flowrate will not decrease as a function of pressure
in the mixing chamber. However, the maintenance of constant pressure
is not necessary for the successful conduction of the experiment.
3.
The flowrate entering the catalyst bed is determined by three flowmeters,
which follow the accumulator. These three flowmeters, shown in the
schematic as only one flowmeter, are placed in parallel. They consist
of a large flowmeter which has a capability of flowrates up to 100,000
cubic centimeters per minute, a medium flowmeter with a capability of
flowrates up to 13,600 cc/minute and a small flowmeter with a capability
of up to 1,400 cc/minute. The three flowmeters cover adequately the
flow ranges considered in this experiment.
4.
The gas enters a preheated furnace, which raises the temperature of the
gas to about 6300 F. The preheater is made from 316 stainless steel.
Its length is approximately 40 feet. The purpose of this independent
preheater furnace is to decompose any nitric dioxide formed in the
accumulator into its components, nitric oxide and oxygen, so that a
true measure of the catalytical activity on the nitric oxide can be obtained
from our apparatus.
5.
Leaving the preheater, the gas is passed into a water containing flask.
This water is heated to a temperature of 1750 F. According to literature,
at this temperature the water-vapor will be 10 percent of the total gas
leaving the flask through the exit tube, regardless of the actual flow rate
(Ref. 11 and 12). There is a bypass valve, and additional inlet and outlet
valves on the flask, allowing the synthetic exhaust gas to remain dry.
In most instances of the experimentation however, the exhaust gas was
properly moistened by this warm flask feature.
6.
Following this step the reaction gases, the pressure of which is meas-
ured by the pressure gauge PI' enter the catalyst furnace, which con-
tains a small preheater No.2 and the catalyst bed. There is a valve
leading out of the gas line between the preheater furnace and the catalyst
furnace, and which leads through a water vapor trap to our continuous
gas sampling station. The purpose of this station is to obtain a composition
of the gas regarding its NO, CO and C02 concentration, before the
reaction. The gas which, through the preheater No.2, passes into the
catalyst bed, undergoes a catalytic reaction in the bed and leaves the
furnace through a heat exchanger in order to cool it back down. The gas
pressure beyond the catalyst bed is measured by the pressure gauge P 2'
The gas then enters a water vapor trap and can be led either to the con-
-------
timious gas sampling station for ob taining the data on NO, CO, and C02
after the reaction, or to a titration experiment to measure ammonia.
The remainder of the gas is led through a vacuum pump into the exhaust.
(b) Catalyst Bed
The various components and features of the catalyst bed itself, which is
placed inside the catalyst furnace, is shown in the next illustration.
In Figure 11 the catalyst bed, consisting of a small tubing with an 1. D. of one
inch, and length of 8 inches, which is placed in a larger tubing with a slide fit, is
sketched in detail. A perforated ending is placed at the bottom of the smaller tubing
and the catalyst put into this container. The gas coming from the preheater flows)
first into the larger tubing; and from there it continues into the smaller tubing, which
is the only way toward the exit from the catalyst bed area. A flange is placed on the
top of the larger tubing, which has been raised above the level of the furnace, so that
the catalyst can be exchanged from time to time without the need of taking the entire
apparatus apart.
The geometry of the catalyst bed and the method of monitoring its temperature
profile are shown in Figure 12. The gas reaction chamber is 19 inches long, extruding
on both sides of the 15 inch long furnace. The gas flow in the figure is from left to
right, in the apparatus from top to bottom. The furnace itself has three resistance
heaters, of dimensions as shown, with a 3 inch diameter. The location of the catalyst
bed in the reaction chamber, with respect to the furnace heating elements, is given.
It is 7.25 inches high with an 1. D. of 1 inch and filled to a depth of 7 inches with the
catalyst pellets. Starting at the bottom of the catalyst bed, at one inch intervals and
with a clockwise rotation of 90 degrees each time, seven thermocouples are located
such that the last thermocouple, No.7, is placed one inch beneath the top surface of
the catalyst.
Measuring the temperature profile of the catalyst bed, the variations from
nominal are observed as plotted in Figure 12. A temperature close to the one
indicated by thermocouple #3 has been selected as the nominal temperature. The
indicated temperature gradient across the catalyst bed is the one as measured with
the flowrate of 1000 cc/min.
(c) The Assembled Apparatus
A general view of the assembled apparatus is given in Figure 13. On the left
is the gas manifold and the accumulator, placed underneath the table. The center
background gives the flow section, preheater furnace and catalyst furnace, with the
furnace controls on the right. The right foreground pictures a mass spectrograph
-------
;-.
--1 +10
<
z;
~ 0
1.
::E
~ -10
:..
;£:
.2 -20
E-<
<
;;
[iI-30
Q
::.~
CATALYST
GAS INLET
PRE HEAT COIL
t.
FIGURE 11.
TOP OF REAC'TWN CHAMBER
FU ACE
TOP
FIGURE 12.
~t\::;
INSULA TION
THERMOCOUPLE
SCREEN
FURNACE
DETAILED SKETCH OF CATALYST BED
GAS FLOW
I
TOP
.117 .'6 .'5 .*4 .'3
CATALYST BED IN R.C.
TOP HEATER
CENTER HEATER
BOTTOM OF REACTION
CHAMBER
.'2 ~l
BOTTOM
BOTTOM HEATER
FURNACE
BOTTOM
GAS TEMPERA TURE PROFILE IN CA TAL YST BED
-------
,
~--_/
<,
FIGURE 13.
GENERAL VIEW OF CATALYTIC GAS REACTION
APP ARA TUS IN FINAL VERSION
analyzer, which is on standby. A more detailed view of the manifold section of the
catalytical gas reaction apparatus is shown in Figure 14. The gas bottles belonging
to the equipment are clustered around the table, containing the gas accumulator,
where the synthetic exhaust gas mixture is prepared. The gas bottles against the rear
wall in this figure go with the Beckman NDIR Analyzer in the next room, which they
serve to calibrate.
Figure 15, finally, gives a close-up of the gas flow measuring and catalyst
bed sections of the apparatus. Also shown is the dynamic and static gas sampling
stations. The white sampling lines lead from the apparatus to the Beckman NDIR
Analyzer. The line on top leads to the station from which the gas composition before
the catalytical reaction can be monitored; along with this dynamic outlet there is a
manifold with three gas sampling bottles for gas chromatographic analysis. To the
right of it is the station for gas analysis afte r the reaction. A water trap, standing
on the floor, is hooked up into the sample line to the NDIR analyzer. A gas sample
bottle for GC testing and/or ammonia analysis is connected to this same gas sampling
station. Also visible in this figure is the heat exchanger and the temperature readout.
-------
~j
~
._~.. ..,.... _.",,~y.,.~.,-- ~..,_.._-_.- ~~"-
.~..~
~
.Jr: ~ . .'7~~¥
~,;.. &.;;"'t-" ,,~.(.!. :.~~~-. 'It:
.' -'-r/:~~\~...~.if-~~-~.'.f..i~I:.i"'
", ' ,-."'.'.! ]1 .~
:,~ ", ~
. ~J f
,. ; ~.... ~
: ..' ..
'..-' ,..~;-.
or, .
FIGURE 14.
RESEARCHER ROLLS SAFETY
SCREEN IN FRONT OF GAS
ACCUMULA TOR UNDERNEATH
GAS MANIFOLD SECTION OF
CATALYTIC GAS REACTION
APPARATUS
(d) Preliminary Experimentation for Qualification of Apparatus for Testing
After completion of the modifications in the experimental apparatus, pre-
liminary test runs in the absence of catalyst were made to determine optimum opera-
tion parameters, especially regarding the optimum preheater temperature, the
flow rate scans and temperature scans. The synthetic exhaust gas was mixed according
to a formula, tabulated later on in this section. Table IV, below, gives the actual,
certified gas mixtures obtained from Matheson and the deviations from the requested
composi tions.
The first task was to establish the best preheater temperature for the cata-
lytical reaction study. The results, as functions of gas composition and flowrate, are
given in the aforementioned Table V, below. At optimum preheater temperatures
and constant catalyst furnace temperatures over the entire range, stable conditions
prevailed in the absence of the catalyst.
3.3.2
The Synthetic Exhaust Gas Mixture
The synthetic exhaust gas mixture which is being used in this experiment,
has been defined to approach the relative composition of automobile exhaust at an
-------
:If..~__r"~._)~ .'
--i
f; -~
7-[
E~ ,
5.;
~.
,:; _.
FIGURE 15.
MAIN PART OF GAS REACTION APPARATUS WITH FLOW-MONITORING
SECTION, CATALYST FURNACE, HEAT EXCHANGER, AND DYNAMIC
AND STATIC GAS SAMPLING STATIONS LOCATED IN FRONT OF AND
BEHIND THE CATALYST FURNACE
TABLE IV
CERTIFIED CONTENTS OF SYNTHETIC EXHAUST GAS COMPONENTS
Gas Deviation
Component Requested Analysis (%)
CO in N2 8.8% 8.84% +0.5
H2 in N 2 2900 ppm 3130 ppm +7.9
NO in N2 6500 ppm 6650 ppm +2.3
02 in N 2 1. 75% 1. 742 % -0.5
C3H8 in N2 1800 ppm 1755 ppm -2.5
-------
TABLE V
PREHEA TER TEMPERATURES USED IN RESTORING
MIXTURE VALUES OF CO AND NO
Flowrate Preheater
Gas Composition cc/min. Temperature
NO: 10S0 or 540 ppm 1S000 SOO° F :J: 20° F
CO: 10500 or 5250 ppm 9000 700° F :J: 20° F
02: 2900 ppm, H2: 530 ppm 4000 630° F :J: 20° F
C3HS: 300 ppm 1000 500°F:J:20°F
C02: 16%, H20: 10%
N2: 72.5%
1000 ppm NO (Nominal) 9000 5000 F:J: 200F
1. 0% CO (Nominal)
2000 ppm H2 (Nominal)
I Balance N2
11000 ppm NO (Nominal) 9000 5000 F :J: 200 F
2.4% CO (Nominal)
1. 0% H2 (Nominal)
Balance N2
air/fuel ratio near to 14. The components of this exhaust gas mixture have been
obtained from actual experiment3;t"data obtained by the IH Truck Division (Communi-
cation from Mr. Ed Fort), which are summarized in Figure 16, the exhaust gas
composition of a typical Otto cycle engine, as a function of air/fuel ratio, and in
Figure 17, which shows the NO concentration in exhaust gases as a function of speed,
power, and A/F ratio.
The concentration of H2 in the exhaust gas as a function of the A/F ratio was
obtained from Reference 13, which, though dating from 1936, is, according to Mr.
Fort, still considered as the best reference on the subject. Mr. Fort adds that
recent unpublished work confirms D'Alleva's and Lozell's data on the H2 concentration
levels.
Based on the data of Figure s 16 and 17, the gas components, and the
synthetic exhaust gas mixture are shown in Tables VI and VII.
-------
1() [
(i ,
\
\
\
\ H2
\
\
\
\
\
\
\
'"
'"
""
"
16
18
20
14
12
I
~ lUi
~ i
~ I
cf2 I-i -
-1
2
() _L__-
8
J--
10
FIGURE 16.
I
14
AIR/FUEL RATIO
NO IN EXHAUST GAS. TYPICAL OF LOW-QUENCH S. I. ENGINES.
FUEL Hlc 1. 9, RELATIVELY INDEPENDENT OF PPM PROVIDED
IGNITION TIMING IS NEAR THAT FOR BEST TORQUE
-------
I I
12 14
AIR/FUEL RATIO
EXHAUST GAS COMPOSITION TYPICAL OF S. I. ENGINE USING FUEL
WITH H/C: 1. 9. SAMPLE DRIED TO SATURATED AT 32° F
(ESSENTIALLY DRY SAMPLE)
4000 -
--
S
0-
.5
o
z
300+
I
I
I
2000
I
1000 ~
I
I
I
I
L
8
FIGURE 17.
-. ----------- ----- -- - ---.--
3 " of Hg
Manif. Vacuum
(Near Full Throttle)
10" of Hg
Manif. Vacuum
(-65% Load)
10" of Hg
Manif. Vacuum
(Near Road Load
at 50 mph)
I
10
I
16
18
-------
TABLE VI
ACTUALLY USED SYNTHETIC EXHAUST GAS MIXTURES FOR CA TAL YST
NO REDUCTION EFFICIENCY EXPERIMENTATION
Gas Mixture Nominal NO CO/NO Nominal 02
Code ppm Ratio R ppm
A 1080 10 2900
B 1080 5 2900
C 540 10 725
D 540 5 725
TABLE VII
DEFINITION OF GAS MIXTURE DENOTED "A"
Component ppm Percent Range in Accuracy
NO 1,080 0.1 :f:10%
CO 10,500 1.0 :f:10%
C02 160,000 16.0 :f:8.5%
02 2,900 0.29 :f:10%
C3H8 300 trac e :f:10%
H2 530 trac e :f:10%
H20 100,000 10.0
N2 724,690 72.5 :f:10%
3.3.3 Description of Analytical Apparatus for Emissions Monitoring
Dynamic Measuring Apparatus
Analytical equipment available at Solar for the evaluation of the catalytical
efficiency in the gas reaction
2CO + 2NO ~ N + CO
2 2
in the synthetic exhaust gas mixture is shown in Figures 18 and 19. This equipment
consists essentially of the Beckman Gas Analyzer with its Non-Dispersive Infrared
-------
FIGURE 18A.
FIGURE 18B.
~ I" w i I
D
/"
THE NON-DISPERSIVE INFRARED RADIATION DETECTOR AND
READOUT RECORDER OF THE BECKMAN GAS ANALYZER
C.""""".
,
\
;
~
~4~~\/ .
:;. t~
TOTAL VIEW OF BECKMAN GAS ANALYZER INCLUDING THE
FLAME IONIZATION DETECTOR FOR THE MEASUREMENT OF
HC SHOWN AT LEFT
-------
EXHAUST GAS SIDE
1. R. SOURCE
WJ
(REFERENCE SIDE~
I. R. SOURCE
WJ
IN AIR
-
-
2.5 IN
SAMPLE
CELL
.
.
DETECTOR
.
.
FIGURE 19.
PRINCIPLE OF OPERATION OF INFRARED ANALYZERS
Radiation detector, shown in Figure 18A, for the measurement of CO, C02' and NO,
and its Flame Ionization Detector, shown on the left in Figure 18B for measuring
the HC with the NDIR being pictured at right.
The principle of NDIR analysis is shown in Figure 19 taken from Reference
14. In the Beckman Instrument, the sample flows through the C02' CO and NO detect-
ing chambers in series.
The flame ionization detector utilizes a hydrogen-air flame, which generates
a minute current when a combustible material, such as hydrocarbons, in the carrier
(nitrogen) enters the region of the flame. Sensitivities are extremely high, on a
picogram full scale under ideal conditions for high response organic materials. The
readout for the Beckman Flame Ionization Detector is located on the NDIR instrument
panel, so that CH, NO, CO and C02 concentrations all can be read out toge;her on
one chart.
Static Measuring Apparatus
This apparatus used for this type of measurement is a Gas-Chromatograph
with a sensitivity for CO and NO into the parts per billion range, shown in Figure 20.
The Gas-Chromatograph has been used to spot-check and calibrate readings obtained
from the two other pieces of analytical equipment available.
-------
"
,I
I
~
FIGURE 20. THE GAS CHROMATOGRAPH FOR STATIC MEASUREMENTS
OF GAS EXHAUST COMPONENTS
Principles of Gas Chromatography
Gas chromatography is the physical separation of two or more molecular
compounds based on their differential distribution between two phases, one of which
is stationary, and the other gaseous. The compounds, after separation, may then be
detected by various methods (see Ref. 15 and 16). The one used at Solar is the Helium
Ionization Detection method.
This method permits analysis of inorganics and permanent gases in the
fractional parts per million range. The helium ionization detector is nonradioactive,
utilizing photoionization of the helium gas, and analyses with high sensitivity, one to
two ppm full scale for most components.
Further details on gas chromatography are found in References 15 and 16.
3.3.4
Procedure for Conducting Experiment
The procedure which has been used for each sequential experimental step
(determined by a given set of parameters for that step is as follows:
-------
The accumulator is filled with the appropriate synthetic exhaust gas mixture
to a pressure of 368 psig, the preheater and catalyst bed temperatures are set, a flow-
rate determined, and the gas, moistured by the vapor from the water in the heated
flash, is allowed to pass through the catalyst bed and to the sampling stations. Two
sample lines to the Beckman NDIR Analyzer are used, one to transport the gas before
the reaction, the other one after the reaction. The gas is allowed to push its own way
through the instrument cells, by means of a slightly higher pressure (.....3 psig), in
order to keep a good control on the flowrate. Data are taken on the NDIR instrument
directly, or else gas is passed through an absorber for the titration method of NH~
dete rmination.
It was found that testing of the catalysts required various periods of running
time for stabilization of catalyst effectivity. Stabilization times ranged from 10 min-
utes to almost an hour, depending on flowrate, temperature, and gas composition.
Accordingly, all data points were accumulated at steady- state conditions and only
when efficiency values were devoid of any drift in the NDIR Analyzer.
Evacuation of the system caused drastic variation from equilibrium efficiency
states and, therefore, one atmosphere of pressure was maintained between runs. This
method decreased the required stabilization time as long as pressure and the range
of the gas compositions were maintained between runs. (System evacuation also had a
transient effect on preheater gas conversion effectiveness. )
Rare earth catalyst test procedures included its exposure to a reducing
atmosphere of 8.8 percent CO in N2 for a period of 10 minutes at 1200° F. Such
procedures have been used by other investigators (Ref. 5 and 7). After this the
catalyst is not exposed to air. (This procedure was also observed to deviate the
catalyst from its equilibrium efficiency.) It seems such a reducing condition is
favorable to NO reduction and thus should give more optimal results than in motor
vehicle s.
Data points are accumulated by running synthetic gas through the entire
system and sampling concentrations with the Beckman NDIR Analyzer until values of
CO, NO and C02 are observed to be completely stable. At this point the gas sample
pickup is switched instantaneously to catalyst bed inlet and allowed to remain there
for the minimum time required for sampling equipment reading to stabilize. This
procedure is normally repeated three times for each accumulator charge of 368 psig.
The NDIR equipment records the concentrations of NO, CO and C02 con-
currently, using three pens with different colors of ink on the same strip chart. The
analyzer is recalibrated before each set of experimental data by the use of the
appropriate zero and span gases.
-------
Chemical Procedure for Determination of Ammonia Formed in Synthetic Exhaust Gas
The general procedure for collection and determination of ammonia is as
follows:
The sample is collected by allowing exhaust gas to bubble through 80. 0 ml
of an absorbing solution of boric acid, containing 20.0 grams of boric acid per liter,
at a rate of 9.0:1: 0.5 1/minute for a specified length of time, generally 10 minutes.
The collection apparatus is similar to a standard gas washing bottle with a fritted gas
dispersion tube. The diameter of the bottle is such that 80.0 ml yields a solution
height of 3.5 inches above the fritted dispersion tube. The solution is maintained at
low temperature in a dry ice water bath in order to ensure maximum absorbing
efficiency during each experimental run.
After collection, the sample is titrated with 0.040 N hydrochloric acid to a
Brom Cresol Green endpoint. One milliliter of 0.040 N hydrochloric acid is equivalent
to 0.68 mg of ammonia. A blank titration is also performed, not only to determine
the blank titration value, but also as an endpoint color standard. Parts pe r million
ammonia (/11/1) is calculated from the following equation:
ppm NH
3
I 22,400 ml/mole 1
= N x m x x
HCl HCl 170 g/mole Litres of gas bubbled
The total percent error involved in this analytical determination is as follows:
[(.1N)2 (.1 1 )2 ( .1L)211/2
Total % Error = N + :::1: + - L 100
where
.1N = :1:0.002
.1ml
= :1:0. 02 mls
.1L = :1:5.0 liters (for ten minute runs)
L = flow rate x time
L = (9.0:1: 0.5 f/min) x time
The maximum error attributable to the analytical technique used for the
ammonia determination is calculated to be less than one percent. The largest factor
contributing to the total error is the flow rate deviation. If the flow rate changes by
0.5 i/min as a minimum deviation, this would result in a total flow error of 5.6
percent. Therefore, the error due to the analytical technique must be considered
minimal when compared to the error introduced by the flowrate conditions.
-------
3.3.5
Reducti'on 9f Data
The data measured are all in the form of gas constituent concentrations, by
the NDIR method for NO, CO, C02 and by the titration method of Reference 17 for
NH3 concentrations. All the data obtained by various gas analysis methods, except
for the titration, are dependent upon the calibration of the respective instruments
with calibration gases, thus the accuracy of the data is a function of the calibration.
Since the NDIR method allows a dynamic monitoring of the gas component
concentrations, the data can be obtained directly as a function of temperature and
flowrate. The effects of initial NO concentration and NO/CO ratio has to be obtained
only after a change in the corresponding parameters has been made.
As we have defined the NO reduction efficiency 1] of the catalyst to be
Initial NO Concentration-Final CO Concentration
1] = Initial NO Concentration x 100%
we obtain the efficiency of the catalyst directly as a function of the temperature and
flowrate, and as functions of CO/NO ratio and NO concentration after changing the
appropriate parameters in the experimental sequence.
The results have been tabulated, plotted in graphs, and assembled in
parametric form in the parametric design chart.
-------
4
EXPERIMENTAL RESULTS
4.1
CATALYSTS AND CATALYTICAL REACTION PARAMETERS INVESTIGATED
The experimental program consisted of a parametric investigation of a number
of catalysts, with emphasis on the rare earth oxide formulations. Additional catalysts
were included after the shortcomings of the original rare earth formulations became
apparent.
4.1.1
Catalysts Considered in the Program
The catalysts investigated in the course of this program were the following:
. Catalyst RE(23)75-P
. Catalyst RE(M3)75
. Monel 400
. Solar Monel 400 Mod.
. Solar Monel 400 Core
. Copper Dichromate, CuCr207
. Palladium Catalyst, 0.5% Pd on Al203 pellets
. Platinum Catalyst, 0.5% pt on Al203 pellets
Also considered was the effect of the empty catalyst bed, consisting of stainless steel,
with a catalytic action ,of its own.
4.1. 2
Parameters Considered in the Program
The basic parameters are:
. Initial NO concentration
. CO/NO concentration ratio
-------
. Catalyst temperature
. Flowrate of gas through the catalyst bed
The first two parameters can be combined into an entity defined as "gas mixture".
With essentially two NO concentrations, 500 ppm and 1000 ppm nominal, and two CO/
NO concentration ratios, four gas mixtures are needed, denoted A, B, C and D.
Table VI in the previous chapter defines these gas mixtures in detail. Gas mixture A
moreover, has been arrived at as the synthetic exhaust gas mixture for an A/F ratio
near 14, and it is defined in Table VIT, also in the prior chapter.
Catalyst temperatures have been selected in the range from 6000 F to 16000 F
in 2000 F increments. The experimental flowrates, coded 1-5, are defined in
Table VITI, below, both in terms of car flowrate in CFM, and experimental flowrates
in cc/min.
TABLE VITI
FLOWRATE RANGE (REF. 18)
Flowrate Car Flowrate Experimental
Code Car Mode CFM Flowrate cc/min
1 Decelerating 5 460
2 Idling 10 1000
3 Partial Load, Cruising 45 4000
4 Full Load, Cruising 100 9000
5 Accelerating 200 18000
-
Though the concept of the space velocity combines both, catalyst bed tem-
perature (assumed to be equal to gas temperature in the bed) and gas flowrate into
one parameter, experimental data obtained have shown the NO-reducing efficiency of
the catalyst to be more temperature dependent than flow rate dependent or space-
velocity dependent.
4.1. 3
Parametric Investigation of the NO-Reducing Efficiency of the Catalysts in
This Program
. Having reduced the parameters to three independent quantities, namely gas
ml.xture, temperature, and flowrate, we may plot these on a three dimensional graph,
USIng an "M" ax. f th . ~
IS or e gas mIxtures; a "T" axis for the temperature, and a "'f/"
-------
axis for the flowrates. The result is a three-dimensional grid, such as shown in
Figures 21, 22, and 23. These figures give, at a glance, the parametric range of
investigation of the NO- reducing efficiency rJ for each catalyst studied in the program.
Figure 21 shows the parameters along the M, T, and if> axes, varied for
catalyst RE(23)75-P. The same parameters have also been used for the empty bed
readings.
Figure 22 gives the parameters for the catalyst RE(M3) 75. Here the lower
temperature range has been omitted due to the relatively poor performance of this
catalyst at the intermediate and higher temperatures.
Figure 23 gives the range of parameters used for all the remaining catalysts.
As can be seen, the only parameter changed here was the temperature, thus all data
points are on the "T" axis.
Each block in Figures 21 to 23 represents a data point. It is labelled with the
values of the parameters at that particular location of the three dimensional parametric
grid. The first number on the front-face of each block refers to the flowrate code, as
RE(23)75-P
EMPTY BED
T
FIGURE 21.
THREE-DIMENSIONAL SCHEMA TIC REPRESENTA TION OF EXPERI-
MENTAL CONDITIONS FOR THE TESTING OF CATALYSTS NOTED
-------
RE(M3) 75
T
FIGURE 22.
THREE-DIMENSIONAL SCHEMATIC REPRESENTATION OF EXPERI-
MENTAL CONDITIONS FOR THE TESTING OF RARE EARTH CATALYST
NOTED
defined in Table VIII. This is followed with the letter denoting the gas mixture code,
as defined in Table VI. The number following the code letter, finally, gives the
catalyst bed temperature in degrees F.
4.2
NO REDUCTION EFFICIENCY OF THE CATALYSTS TESTED
In this section, we shall first present, in tabular form, the experimental
data for the rare earth oxide catalysts, i. e., RE(23) 75-P, RE(M3)75, and the empty
bed catalyst. Next follows the graphical presentation of these same data, and finally
the NO reduction efficiency data of the other catalysts tested.
4.2.1
Experimental Data for Rare Earth Oxide Catalysts
A. Catalyst Formula RE(23)75-P
The data for the temperature, as well as gas mixture scan, is given in
Table IX. These data correspond to the plane of data points given by the T-M axes
in Figure 21, above. The table gives the quantities of the CO, NO, and C02 components
-------
T
C/J
MONEL 400
MONEL 400 - MODIFIED
CuCr207
pt on A1203
Pd on A1203
T
FIGURE 23. THREE-DIMENSIONAL SCHEMATIC REPRESENTATION OF EXPERI-
MENTAL CONDITIONS OF TRANSITION METAL AND NOBLE METAL
CATALYSTS
of the synthetic exhaust gases, before and after the catalytic gas reactions, as
measured by the Beckman NDIR apparatus. The loss percentages for the CO, and
for the NO, the latter being equal to the NO-reducing efficiency 1], are also shown
in this table.
The flowrate scan data are given in Table X, below. These data correspond
to the cp axis in Figure 21. Again, the quantities of CO, NO, C02' and the loss of CO
and NO, the latter being again the NO-reducing efficiency 1], are given in the table as
a function of flowrate shown both in code number and actual dimensions of liters/
minute.
B. Catalyst Formula RE(M3)75
The data for the temperature and gas mixture scans, at constant flowrate,
are tabulated in Table XI. These data, corresponding to the data points shown on
the T-M plane in Figure 22, above, are arranged in the same way as those in Table
IX, for the catalyst formula RE(23)75-P. The table shows the catalytic efficiency, 1],
as a function of temperature, CO/NO concentration, and initial NO concentration.
-------
TABLE IX
THE NO-REDUCING EFFICIENCY, AND PERCENT LOSS OF CO, FOR
CATALYST RE(23)75-P AT A FLOWRATE OF 9000 cc/min
-1 --~:2~O)--1
Before 1- ~fte~_j
19.0 19.2 I
19.0 19.2
--- -----1-- - ----:O-~)----- -- -1 - -- - NO (ppm)
Composition I Temperature -+B~f~r~1~~~r- - T~~~~s_~-_i_HB~f;re -- After r %1.oss
-- ---- -----------~- -- --- r --; ,
A 1600. F : 1. 20 : 0.72 : 3 '1060
, 1.17 ! 0.72 ! 9.2 1060
!
-----
1400. F
I
!
~
I
1200. F
1000. F
1400. F
1200° F
1000. F
600° F
C 1400.F
1200° F
1000° F
600. F
D 1400. F
1200° F
1000° F
L--
500
510
52.4
0.95
1.01
I 0.57
I 0.59
1120
1060
587
565
47.6
46.7
15.8 I 16.0
15.8 16.0
! 40
41.6
I 1.02
:
! 1.01
I 0.986
i 1.05
0.635
37.7
1057
580
16.0 ~ 16.0
45.6
0.80 20.8 1200 660 45.0
I 0.80 18.9 ; 1120 740 33.9
0.695 33.8 ! 1090 830 23.8
--~-- -~----l-- n- -- - --~---------~- __n____-
i
15.2 i 15.2
15.2 15.2
-~~'~j-~
I
I
i
~ 0.443
I
i
I 0.50
0.227
48.8
43.8
15.33
15.4
I 1120
I
I 1180
I
630
0.250
50.0
812
31.2
15.4 I 15.6
I 0.447 0.287 35.8 I
I 1203 lOgO 10.2 I 15.2 I 15.2
0.49 0.24 51.0 1140 1080 I
I 5.3 15.2 I 15.2
I : I
I I :
I I
0.48 0.48 0 1040 : 1000 3.8 I 15.6 I 15.6
I I
I : I
I -------,-- -- -- -- -- ---- - -- - -- . _J._- - -- - -- -- -- - -+ -- - -- -i-- ----~
0.587 0.365 ! I
I : 37.8 645 I 300 53.5 14.831 14.8 I
1 0.60 ' i
0.443 26.2 617 317 48.6 15.07 15.2 I
I I I I
0.57 I 0.515 I I
! 9.6 735 485 I 34.0 15.6 15.6 I
I I
I
0.52 I 0.52 0 440 440 I 0 15.0 15.0 I
-1--- _m___+__---- ---- - ---- -- - - -- i
- - - -- -- --- ------j
I
0.277 ] 0.16 42.2 688 358 48.0 15.0 14.8 I
I
0.345 0.225 34.8 643 353 45.1 14.8 14.8 :
I
0.25 0.19 24.0 645 560 13.2 15.0 15.0 !
i
---1--
-------
TABLE X
NO REDUCING EFFICIENCY, AND PERCENT LOSS OF CO, AS A FUNCTION
OF FLOWRATE FOR CATALYST RE(23)75-P, AT T = 12000 F, GAS
MIXTURE A
-
Flowrate
Code # I /min
--- --
5 18
4 9
3 4
2 1
----
--.-- --- ----
-------
CO(%) NO (ppm) C °2(%)
------ ----
After %Loss Before After %Loss Before After
--------
1. 09 0.82 24.8 1155 700 39.4 15.3 15.5
1. 02 0.635 37.7 1067 580 45.6 16.0 I 16.0
1.167 0.60 48.6 1115 427 61. 7 15.5 15.8
0.91 0.37 59.3 955 80 91. 6 16.25 16.4
----- "---
Before
---~ ---- ---
The flow rate scan data, at constant gas mixture and temperature, corres-
ponding to the cP axis in Figure 22, is given in Table XII, completing the parametric
study of the NO reducing efficiency of catalyst RE(M3)75.
C. Data for the Empty Catalyst Bed
In order to determine what contribution, if any, the empty catalyst bed made
to the NO-reducing efficiency of the two rare earth oxide catalysts, a temperature-
gas mixture scan, corresponding to most of the data points in Figure 21 on the T-M
plane, and a flowrate scan, corresponding to the cP axis of Figure 21, were made.
These data are presented in T d.bles XIII and XIV, constructed along the same lines as
the corresponding tables for catalysts RE(23) 75-P and RE(M3)75. If the experimental
parameters of some data points indicated that continued measurements would not
yield measurable data, as it occurred at the lower catalyst bed temperatures,
these data points were omitted.
D. American Oil Company Study
A supporting experiment for the measurement of the NO reducing efficiency
of catalysts RE(23)75-P and RE(M3)75 was made by the American Oil Company
Research Department. These data, obtained by the use of the pulse-flame apparatus,
bear out the measurements obtained by Solar as to the relative efficiencies of the two
rare earth oxide catalysts studied. These data, as well as data obtained by American
Oil on the generation of ammonia by these two catalysts, are reproduced in
Appendix F.
-------
T ABLE XI
THE NO REDUCING EFFICIENCY, AND PERCENT LOSS OF CO, FOR
CATALYST RE(M3)75 AT A FLOWRATE OF 9000 cc/min
co (%) NO (ppm) C02 (%)
Average Average
Temperature Conversion Conversion
Gas Composition rF) Before After (%) Before After (%) Before After
A 1400 1. 03 0.83 18.2 1060 1000 5.6 16.0 16.0
1. 00 0.83 1080 1020 16.4 16.0
I
1400 1. 06 0.92 1090 1060 15.6 15.8
1. 04 0.93 12.3 1100 1065 3.2 15.8 15.8
1. 07 0.93 1110 1070 15.8 15.8
1200 1. 10 0.98 1060 1000 16.0 16.0
9.2 6.5
1. 08 1. 00 1090 1010 16.4 16.3
1000 1. 09 1. 03 1050 1020 16.2 16.6
1. 06 1. 00 5.6 1050 1020 2.8 16.3 16.3
1. 06 1. 00 1060 1030 16.6 16.6
B 1400 0.49 0.40 1060 1020 16.0 15.8
I 0.49 0.40 18.4 1060 1020 3.8 15.8 15.8
0.49 0.40 1080 1040 15.8 15.8
1200 0.53 0.46 1060 1020 15.2 15.0
0.53 0.46 12.5 1070 1020 4.4 15.2 15.2
0.54 0.48 1080 1030 15.2 15.2
1000 0.48 0.47 1055 1040 16.4 16.4
0.48 0.47 2.1 1. 18
1060 1050 16.4 16.4
C 1400 0.64 0.53 640 480 14.4 14.4
0.64 0.55 16.0 650 480 25.6 14.4 14.4
0.66 0.55 660 490 14.6 14.6
1200 0.53 0.50 580 540 13.9 13.8
0.53 0.49 6.6 6.9
580 540 14.0 13.9
1000 0.54 0.53 610 580 14.6 14.4
0.54 0.53 1. 85 4.5
610 585 14.6 14.6
D 1400 0.30 0.25 620 600
14.0 13.6
0.30 0.25 14.8 620 600 4.3 13.6 13.6
0.28 0.25 640 600 13.6 13.6
120U 0.31 0.28 560 520
15.8 15.8
0.31 0.28 9.6 565 520 7.4 15.8 15.8
0.32 0.29 570 530 16.0 15.8
1000 0.34 0.33 610 590
2.9 15.8 15.8
0.34 0.33 2.9
615 600 15.8 15.8
-------
TABLE XII
NO REDUCING EFFICIENCY, AND PERCENT LOSS OF CO, AS A FUNCTION
OF FLOWRATE, FOR CATALYST RE(M3)75, AT T = 1200° F,
GAS MIXTURE A
co (%) NO (ppm)
Flowrate Average Average C02(%)
Conversion Conversion
Code # i!min Before After % Before After % Before After
-
5 18 1. 05 0.94 8.6 1100 1060 4.7 15.9 15.8
1. 05 0.94 1140 1090 15.6 15.6
1. 05 1. 00 1170 1110 16.2 16.0
4 9 1.10 0.98 9.2 1060 1000 6.5 16.0 16.0
1. 08 1. 00 1090 1010 16.4 16.3
3 4 1.15 0.96 18.4 1210 1140 6.3 15.0 15.0
1.17 0.92 1230 1140 15.0 15.0
1.15 0.95 1220 1150 15.0 15.0
2 1 1. 00 0.42 55.4 1180 140 88.0 15.0 15.2
1. 02 0.48 1200 145 14.9 15.1
4.2.2
Experimental Data for Transition Metal and Noble Metal Catalysts
The program was extended to cover additional catalysts after it became
apparent that the rare earth oxide catalysts do not meet expectations regarding both
NO reducing efficiency and ammonia generation factor, to be discussed below. Three
transition metal catalysts were investigated, namely Monel 400, Solar Monel 400-Mod. ,
which contains a minor additive in order to modify the oxidation resistance of Monel
400, and Copper Dichromate. The first two were non-oxidized metal sheets, bent
into the form of burl saddles, as described in Reference 4. The last consisted of
pressed oxide pellets, of 0.18 inch in height and diameter.
The two noble metal catalysts contained 0.5 percent of Palladium or Platinum,
respectively, mixed into aluminum trioxide pellets of 0.15 inch height and 0.125 inch
diameter.
The data obtained in the temperature scan, at constant gas mixture and flow-
rate, as shown in Figure 23, is tabulated in Table XV, for all five catalysts. The
data are averaged again in the same fashion as those for the temperature scans of the
rare earth oxide catalysts and the empty catalyst bed. The evaluation and discussion
of these data will follow in Chapter 5.
-------
TABLE XIn
PERCENT LOSS OF NO AND CO, VERSUS GAS MIXTURE AND
TEMPERATURE, FOR EMPTY CATALYST BED, AT A GAS
FLOWRA TE OF 9000 cc/min
..- -T~~~;t";:[&[O~~~A:'(%) - %;; -r-~;-(;;m'-- --=r=. C021t1
CompoSition. - =--i--- -1 - - .~ 1- -- B~1~1--;3-3~~1.~Oe9r8---~3~.lSS---- B1~6!.o6re After
A -;~OO°F ----~:~~ --~-:~~--~-~:~ 1118 I ~083 3.1 16.1 ~::~3
1200° F 1.15 1.13 1. 7 1120 I 1100 1 8 16 0 16.0
1000'F 'l--'---L-!'.'!" , .
I I T i
1400° F 0.487\ 0.433 I 11.1 i 1203 1193 I 0.8 15.8
1200°F 0.49 i 0.466 4.9 I 1080 1080 0 16.0
! I * * I
~J . ,:::: ~.-i..:. 6~:1- ::+~:+;~3- - :6~~ 2:~-- --:5.87 :5.87
I 0.65 10.63 I 3.1 660 573 13.2 15.7 15.63
I 1200°F 0.57: 0.55 I 3.5 650 640 i 1.5 ! 15.0 I 15.6
-.1 1:::::-.1__:._L- L:- :_~l: .1_~_J.~
1400" F 0.32 . 'I '-:"--'1 713 4.0 ! 14.53 1,4.53
::::::J 0:25'J 0:2~3J :,0- -., 6:J_~~. :.0.1:5.33 j ',5.33
..
B
15.8
16.0
1000° F
*
*
*
*
*
D
. Not Measured Due to Lack of Effect.
-------
TABLE XIV
PERCENT LOSS OF NO AND CO, AS A FUNCTION OF FLOWRATE, FOR
EMPTY CATALYST BED AT T = 1200° F, GAS MIXTURE A
I C:::ra~m: ---
. 5 18 0.925
------
I
C 02(0/0) I
Before Afte~
CO (0/0)
Befo;;-lAfter
NO (ppm)
%Loss
Before
After
%Loss
0.885
4.3
1065
1045
1.9
17.13
17.03
4
9
1. 15 1.13 1. 7
I 0.98 I 0.857)12.6
I .
~.047_1_0_~66_- -~~:_~-
1120 I 1100 1.8 16.0
1083 I 857 11. 6 1,5. 93
10~~~8!_d2~~~15.8
I
116.0
116.0
I
i 16.0
I
3
4
2
1
----
An additional catalyst, denoted Solar Monel 400-Core:" was also tested. It
was placed into a tube 4 inches long, with a diameter of slightly over one inch. Its
volume was thus 61. 5 cc, its weight was 110 grams. Four experimental runs at
constant temperature were made with this new catalyst bed. The results are shown
in Table XVI, which also lists the space velocity involved for each run.
The data indicate a slight space velocity dependence of the efficiency rj, but
in general the percentage of NO reduced is the same as that obtained with the burl
saddles of Monel.
4.3
THE AMMONIA GENERATING FACTOR FOR ALL CATALYSTS TESTED
The study of the catalytic reduction of NO by CO in exhaust gases showed
both in the literature (Ref. 4 and 5) and at Solar, that ammonia is formed as a
byproduct. Since, with a two-stage catalyst bed, all ammonia is reconverted into
NO at the oxidizing stage, ammonia generation will effectively limit the NO reduction
efficiency of the catalyst. Experiments have therefore been run, measuring the
quantity of NH3 generated for each of the catalysts tested.
The method of analysis has been discussed already in Chapter 3. The data
obtained, along with the experimental conditions, are given in Table XVIT. The
evaluation of the catalysts, along with the calibration of their ammonia-producing
factors, will follow in Chapter 5.
* Solar Monel 400-Core is a design consisting of extended surfaces of Monel inter-
leaved between stainless steel sheet collected in a right cylinder configuration.
-------
TABLE XV
NO REDUCING EFFICIENCY OF SELECTED TRANSITION METAL AND NOBLE
METAL CATALYSTS, VERSUS TEMPERATURE, IN SYNTHETIC
EXHAUST GAS APPARATUS
(Gas Composition A and Flowrate of 9000 cc/min Apply to All Tests,
Otherwise Noted)
I--~--r--
co (%) !
I Catalyst
~------
I
! Type 400
~ Monel
Type 400
Monel-
Modtfled
Copper;
Dichromate I
CuCr207
P alladl urn
(0.5% In
A1203)
Platinum
(0.5% In
A1203)
I Temperature
1400
1200
1000
800
1400
1400
1200
1200
1000
1000
800
1400
1200
1000
800
1400
1000
1400
1200
1000
800
800
I No Water
OF
I Befo re
--I
After %LOSB i Before
- - - -t-
1.23 0.14 88.6 1070
1. 22 0.19 84.4 1100
1. 22 0.07 94.3 1100
1.21 0.10 91. 7 910
-- ---
1.28 0.10 1170
1.28 0.10 92.2 1170
1.30 0.10 1190
1. 23 0.11 I 91.1 1230
1.18 0.10 i 91.5 1170
1.175 0.07 I 94.0 1210
1. 27 i 0.18 I 1170
1. 27 0.18 85.8
1190
I
1.15 0.11 1210
1.17 0.22 1210
NO (ppm)
1.13 0.13 88.5 i 1130 I 220
1. 26 0.15 88.1 i 1160 i 220
1.16 O.-~; ---~~2-80
1.16 0.68 41 I 1150 i 280
I I
1. 20 0.88 t8 10S0! 460
1. 22 0.88 1090 I 450
0.65 46 --1160 T3~
0.63 ! 48 1170: 320
0.65 I 47 1190 i 320
0.52 I 56 1200! 3>0
0.52 ! 56 1200 I 3;5
0.53 I 55 1220 I 325
I'
I 1.16
1.13
1.11
1. 20
1. 20
1. 22
1.18
1.17
1.19
1.17
1.16
1.15
0.33 I 72
0.34 ! 70
0.33 i 70
I
0.375 i 68
0.36 ,69
0.35 : 70
1190
1190
: 1200
, 1000
I 1030
I 1030
I
I
1. 25
1. 25
1. 25
O. 73 i 42
0.75 : 40
0.73 I 42
----+-- - ----..-.0.-_--
58
1260
1270
1280
After
55 I 91.9
65 91.1
65 91.1
70 i 92.6
f-
80
80 93.2
80
85 I 93.1
I
85
92.7
80
93.4
80
80
93.2
80
100
93.4
91. 7
79.8
81. 0
80.5
81. 0
75.6
75.6
57.4
58.7
72.4
72.6
73.1
73.3
72.9
73.4
240
250
250
76.0
75.7
75.7
325 72.7
330 72.2
330 I 72.5
2tO I 80.9
250 I 80.3
260 i 79.7
-'--
Except Where
C02(%)
Before
After
--1----
17.2 18.0
17.2 I 18.0
17.2 I 18.0
17.0 ~ 18.0
17.4 18.2
17.4 18.2
17.6 18.3
17.4 18.2
17.2 18.0
17.0 18.0
17.4 18.0
17.2 18.0
17.0 I 18.0
17.8 18.0
17.8 18.6
17.6 18.4
16.9 17.7
17.6 18.4
I
16.9 i 17.2
16.8 I 17.2
I
16.0 i 16.5
16.5 ! 16.5
16.8 : 17.0
17.0 ' 17.2
17.0 I 17.2
17.0 17.4
17.1 17.4
17.2 17.4
15.9 16.8
13.9 16.8
16.1 16.9
16.8 17.2
16.8 17.2
16.8 17.2
17.0 17.2
17.0 17.2
-------
TABLE XVI
NO REDUCING EFFICIENCY OF MONEL 400 CATALYST IN DIFFERENT
STRUCTURAL CONFIGURATION
(Solar Monel 400-Core)
Log Book Ref. 148/9; Gas Mixture A, NO = 1160, R
Vcat= 61.5 cc, T=1400°F
10,
cJ 1.1 CO (%) NO (ppm) 1] 1] average C02 (%)
cc/min hr-1 Before Afte r Before After (%) (%) Before After
4000 12874 1. 00 0.375 1160 55 95.26 18.1 18.7
4000 12874 1. 00 0.375 1160 55 95.26 95.26 18.2 18.7
9000 28967 1.17 0.50 1180 90 92.37 18.2 18.5
9000 28967 1.20 0.48 1200 75 93.75 93.06 18.2 18.7
Note:
The runs with the gas composition other than gas mixture A, using the
catalyst Monel 400, were made to duplicate and verify the data reported
in Reference 4. 'Vith the use of these special gas mixtures the results
reported in the literature cited were verified, which at the same time,
therefore, confirms the validity of the data obtained with Solar's experi-
mental apparatus utilizing gas mixture A (1000 ppm NO, CO /NO = 10).
-------
TABLE XVII
AMMONIA FORMATION FOR ALL CATALYSTS STUDIED IN TillS PROGRAM
AT T = 1400° F (EXCEPT WHERE NOTED OTHERWISE), AND
GAS FLOWRATE OF 9.0 l/min
I~
I
Catalyst
: RE(23)75-P
RE(M3)75
~~-
None
Bed Empty
Monel 400
: Monel 400
;
1
I
L---
Monel 100
, Solar Monel
: 400 Modified
i
1--
CuCr207
Pd on A 1203
r-
I
Pt on A P 203
Pt on Af203
Pt on A P203
~---- -
._.. ._-- I --- .- -. r--T ----I
-_..£O~~;~l~n ., _.~:~ - j_~;:i~~~t-~;:;~Ia
A II, 10 mln i 19(j
) 2 j i 146
3 I 157 ,
4 ! 197 i
---- ;;.~.__.I__.- -.2~~~1
2 49.2
3 43.0
4 ~ 30.0
5 i ,11. 0
-------T -----;-----
IT 1 I" 7.4
2 5.3
3 I 3.7
4 3.7
- --- ..--. . - ... . ---t-- _.
III I! 16.3
2 i 16.4
3 I" ~ 9.3
~ NO: 870 ppm; CO: 1.15% - -I~ -:---~-7':'~-:1:t-;;:~1
I H2: 2000 ppm; N2: Bal. 2 i 10 min I 21. 3
3 ! 8 min I 21. 3
4.~_~_m!n_j 42.6
VI' 10 min I 83
, 6 mln 72
6 mln /128
: 10 mln 103
I 10 min 156
A
A
A
- - ------
NO: 820 ppm; CO: 2.4%;
H2: 1. 0%; N2: Bal.
3
4
5
A
10 min -r. :::: I
11. 5 I
8.2 I
-p:rJ
! 462
I 491
408
XI 1
2
3
A
, Xli 1
I 2
3
. ;
A
'XIV 1
2
A
'vrn
A
800' F
IX
1
2
3
663
354
385
A, No Water
800' F
X
1
2
3
776
423
404
60
Average
For Set
of Runs
(ppm)
174
31. 6
~
5.0 I
-----r
12.6
26.2
108
10.8
----,
I
!
I
I
I
I
,
--~-I
I
9.1
218
454
~~~
-------
5
EV AL UA nON OF DATA AND DISCUSSION
This chapter has been divided into two sections, the first one of which will
give a graphical presentation of all the experimental data obtained in measuring the
parameters for the catalytical NO reduction efficiency, and then evaluate the NH3
generation data; the second section will summarize the data into a parametric design
chart and discuss its limitations.
5.1
EVALUATION OF EXPERIMENTAL DATA
5.1.1
Graphical Presentation of Catalytical Efficiency Parameters
In presenting the data in graph form, we shall proceed in the same order as
in the previous chapter, namely, first show the data for catalysts RE(23)75-P,
RE(M3)75, and the empty bed, and then those for the other catalysts in the program.
A. Data for Catalyst RE(23)75-P and Relative Empty Bed Effect
The experimental data for the high purity rare earth oxide catalyst, RE(23)
75-P, are presented in the next four figures. The first one, Figure 24, gives the
flow rate dependence of the catalytical efficiency. Since the data obtained with the
"empty bed" showed a large contribution of this factor for low flowrates, the "empty
bed" efficiency versus flowrate has also been entered into Figure 24. Thus, the only
problem was to find a proper method of evaluating the actual catalyst efficiency, in
the face of the significant empty bed contribution at the two lower flowrates. (This
effect was the reason why no measurements at the flow rate of 460 cc/min, code #1,
have been made. )
Let us denote the efficiency of the bed-catalyst system TJc' and of the empty
bed TJo' A direct subtraction, in order to obtain the catalyst efficiency TJ, of the form
TJ= TJ-TJ
c 0
(17)
would yield an anomalous result for the TJ vs. cP graph, namely the catalyst efficiency
would vanish as the gas residence time increases. Since a result like that does not
-------
10(1
17%
:;0
90
EMPTY BED EFFECT
~o
iO
T .0 1200' F
1000 PPM NO
R ~ 10
liO
COMBINED EFFECT
----
-to
CATALYST EFFECT
30
20
10
o
o
14,000
16.000
18,000 20.000
10.000
~0. CC /MIN
12. 000
(j000
8000
2000
4000
NO REDUCING EFFICIENCY 1] VERSUS FLOWRATE
-------
the efficiency is 63 percent, or abrnt equal to that at the 4000 cc/min flowrate, which
appears to be the most reasonable result. Thus, this figure shows that the flowrate
dependence of the NO reducing efficiency of RE(23)75-P is only very slightly tempera-
ture dependent.
Figure 25 gives the temperature dependence of." for RE(23)75-P at CfJ = 9000
cc/min. This figure gives the mean of ." as measured for all four gas compositions
tested, so that the general trend is visible. With increase in temperature, ." first
grows slowly, then faster between 1000 and 1200° F, only to slow down again above
12000 F.
The actually measured values of." as a function of temperature for both
initial NO concentrations and both CO/NO ratios is shown in Figure 26. Here it is
interesting to note that while the ratio effect behaved as generally expected, even
though not so pronounced, the catalyst appeared to be slightly more efficient for the
smaller initial NO concentration than for the larger one. According to theory there
should have been no effect due to initial concentration changes. The average values
of TJ, shown in Figure 25, are designed by the triangular points and the full curve in
Figure 26.
Figure 27 represents a map of the catalyst efficiency in NO reduction versus
the space velocity v, within a given region of temperatures and flowrates. The space
velocity is defined as
v=
60CfJP 0 T1
V P T
1 0
[hr-1]
(19)
where 60CfJis the flowrate in cc/hr (60 min/hr and (jJ in cc/min), V is the volume of the
catalyst in cc, Po and To are the standard temperature and pressure, in absolute
units, and PI and T1 are the gas pressure and temperature, respectively, also in
absolute units. For a given'catalyst of volume V and entering gas pressure PI' the
space velocity can be expressed as
v= ACPT
(20)
where
60 Po
A =
V PI To
(21)
Thus, v can be plotted out on the CP, T scale as a family of parabolas, determined by
A. This was done in Figure 27, for the range of T between 1460 and 1860° R, or 1000
and 1400° F, and flowrates between 0 and 23,000 cc/min. The abscissa itself is the
-------
FIGURE 25.
FIGURE 26.
50
40
30
7]%
20
10
o
o 600
1000 1200
TEMPERA TITRE (0 F)
NO REDUCING EFFICIENCY 17 OF CATALYST RE(23)75-P VERSUS
TEMPERATURE T, AVERAGED OVER FOUR GAS COMPOSITIONS
TESTED
800
1400
1600
f)O
------------
........--
,.. -- - -----=- - - -.:....=.= - -
;' -
R 10"/ ----:. ~
,,/ ~, -
/ R = 10 "
/' /'" ,,/
// /' //
/' / /
,,/ /
// /
// R 5'/
/ /
I
40
---
_AVERAGE VALUE OF /"
EXPERIMENTAL /'
POINTS /'
/
/'
/
:JO
/-EXPERIMENTAL POINTS
/
R."/
/
/
/
/
/
/
/
.
ry%
I
/
I
I
I
I
I
I
I
I
I
I
I
I
.
--
1000 PPM NO
~o
CO = 9000 CC/MIN.
-----
,,00 PPM NO
10
MEAN VALUE~
o
o
""
1000
I
1300
I
1400
1100
1200
TEMPERATURE (0 F)
NO REDUCING EFFICIENCY 17 OF CATALYST RE(23)75-P VERSUS
TEMPERATURE T, FOR VARIOUS NO CONCENTRATION AND CO/NO
RA TIOS R
-------
25,000
20,000
RE(23)75-P
R = 10
V
cat
90 CC
DIRECTION OF MAXIMUM
SPACE VELOCITY
GRADIENT
/
I -1
29,090 hr
T] = 39. 4%
PI
4.3 PSIG
15,000
LINE OF CONSTANT
SPACE VELOCITY
Z I
:2
'- I
w
u I
S 16,296 hr-l
14,545 hr-l ."O~ 47.17,
T] = 45. 6%
I 12,704 hr-l
I
I 600hr-l
I
5,000 I
16464 hr-l ." = 60.6%
I
I 4000 hr-l
I 1616 hr-l
I ." = 54.70/0
0 I 0 hr -1
1000 1100 1200 1300 1400
TEMPERATURE (0 F)
FIGURE 27.
MAPPING OF CATALYST EFFICIENCY." VERSUS SPACE VELOCITY
V WITHIN A GIVEN TEMPERATURE AND FLOWRATE REGION.
INITIAL NO CONCENTRATION IS 1000 PPM NOMINAL
-------
parabolic curve 'for V = 0 hr-1, and selected segments of the family of parabolas for
various values between 1616 hr-1 and 33,200 hr-1 are given. The dashed curve denotes
the direction for steepest gradient of increase in v. It shows the magnitude of the
relative contributions of the temperature and flow rate on spece velocity, by its degree
of inclination. This is also seen from the experimental points: V changes from 12,704
hr-1 to 16,296 hr-1 at
-------
100
90 .
80
70
l 60
J::-
I
>-
(.) 50
z
L.LJ
(.)
l.L 40
l.L
L.LJ
30
20
10
0
0
30
C
s:o-
>- 20
(.)
Z
L.LJ
S2
l.L
l.L
L.LJ
10
40
- CATALYST RE(M3) 75 +BED
.---- EMPTY BED
~
,
,
'"
,~
"
'-
6000
14,000 16,000 18,000
FIGURE 28.
8000 10,000 12,000
FLOW RATE -l"t:c/MIN)
NO-REDUCING EFFICIENCY 1] VERSUS FLOWRA TE
-------
EM PTY BED
- 1000 PPM NO
--- 500 PPM NO
<(i= 9000 ccjmin
40
f-
30-
~ f-
10"
>-
u 20 f-
Z
W
U
LL.
LL.
W
10 -
o
1000
" R=10
"
".--
",,"
"
",,"
,,"
"
"
,," ---+-- R=5
fill" ------ y- R=lO
'-.t,..------ - R 5
,,~l T
1200 1300 1400
TEMPERATURE CF)
NO REDUCING EFFICIENCY." FOR THE EMPTY BED VERSUS TEM-
PERATURE T, FOR VARIOUS NO CONCENTRATIONS AND CO/NO
RATIOS R, IN THE CATALYST BED IN ABSENCE OF CATALYST
I
1100
I
1500
FIGURE 30.
Figure 30, were also plotted. This figure shows, too, essentially no temperature
dependence of.", with the exception of R = 10, NO initial = 1000 ppm. The anomalous
behavior of that curve in Figure 29 thus finds its explanation.
Comparing Figures 29 and 30, one sees that." for RE(M3)75 is only by a factor
of two greater than that of the empty stainless steel bed. This catalyst, therefore,
showed no behavior different from that of the empty bed, and has to be adjudged as not
suited for catalytic reactions of the type studied in this program.
C. Comparison of." Versus T for Transition Metal Catalysts and Noble Metal Catalysts
With the Rare Earth Oxide Catalysts
The data of Table XV, regarding the NO reducing efficiency of the five non-
rare earth oxide catalysts, have been plotted out in Figure 31, with the data for the
two REO catalysts and the empty bed, all taken under the same experimental conditions,
being added. In this way, Figure 31 becomes a comparison criterion for all eight
catalyst configurations considered. One data point for Solar catalyst Monel 400-Core,
taken at 1400° F under the same gas mixture and flow conditions, has been added, for
purposes of comparison. However', due to its smaller volume, this catalyst is being
-------
100
o
700
I
800
MONEL 400
1-------.-- .-1.-.-- -------=- .-.--
MONEL 400 - MODIFIED
CuCr 0
..-------- . 2 7 .
- --------------------e--
. PT CATALYST ..--.
~. .~ ....:::.;.-
...-. ...-
i~L"Si ...-.
1'0 ~.--.
...--
----. .....-.
---- . ......- .
--
~.
/. 1000 PPM NO, CO/NO = 10
/ ~=9000 oo/m;"
.
RE(M3)75
.~.-._..-._._._._._.-.
- - ~ - -. - - - 't- - -' :~:_E~~!.Y_C!.~~l:~T_B!~__---'f=':'
900 1000 HOO 1200 1300 1400
TEMPERATURE (OF)
..~
MONEL 400 -CORE
90r.-
80~
70r-
.-
~ 60-
s::-
I
~ 50-
z
UJ
!:2 40-
u.
u.
UJ
-
30 -
20-
10-
1500
FIGURE 31.
COMPARISON OF THE NO REDUCING EFFICIENCY 7J AS A FUNCTION
OF T FOR THE VARIOUS CATALYSTS TESTED
measured at a larger space velocity than all the others. The conditions of flow rate,
CO/NO ratio, NO concentration, and temperature range, are all noted in that figure.
Monel catalyst had the highest effectivity at NO reduction plus NH3 genera-
tion. Although it is well known Monel 400 cannot meet the goal -- long term catalyst
stability, Solar Modified Monel had essentially the same effectivity however, and is
expected to show substantially increased stability. The data point for Solar Monel
400-Core falls between those of the two other Monel catalysts.
Next comes CuCr207' in the 80 percent efficiency range. Its efficiency
decreases with rising catalyst temperatures, indicating its maximum efficiency tem-
perature to be below 8000 F. Next follows the platinum catalyst over the temperature
range up to 13000 F. It shows an increase in 7J between 800 and 10000 F, but falls off
thereafter. The palladium catalyst, starting out in the 50 percent range, appears to
have a strong temperature dependence, and it overtakes the faltering platinum catalyst
above 13000 F. However, only two data points have been obtained with the Pd catalyst,
thus its curve is only an indication of its behavior.
Further down, rising from 10 to about 47 percent efficiency, we find the
catalyst RE(23)75-P. It shows the strongest temperature dependence of all in the
-------
lower temperat~re region, flattening out somewhat, but still rising, in the higher
regions of temperature. Last, just slightly above the empty bed levels, is RE(M3)75.
The conclusion to be drawn from Figure 31 is that while the pure REO have
a vastly better NO-reducing efficiency than the mixed REO, which contain also other
impurities, even the pure rare earth oxide catalysts trail far behind those consisting
of the transition metals and precious metals, at least in their present formulation.
Monel appears to be the best NO reducing catalyst of all.
Figure 31 shows, essentially, that the REO catalysts are not suited for the
reduction of NO by CO in an exhaust gas environment, containing 02, H20, hydrogen
and other components. However, of the other catalysts tested, both the Pd and pt
catalysts were also disappointing. The Pd catalyst appears to be regarded as the
leader in some circles. However, it is possible that its performance improves with
an increase of the precious metal content. That approach, however, would destroy
the economic feature of these catalysts, as compared with the others in this program.
In addition, Table XV shows that the pt catalyst, which is essentially an oxidation
catalyst, did not do too well with the oxidation of CO, either. The pt catalyst does
not exceed a 70 percent CO oxidation efficiency, and the Pd catalyst is at most 41
percent effective here. This shows that for the pt catalyst oxygen is needed to remove
CO from the exhaust, while the Mo nel catalysts, and also CuC r207, are highly effect-
ive on the CO front as well.
5.1. 2
Comparison of Ammonia-Generating Factor E for All Catalysts Tested
In completing the evaluation of the five catalysts studied in this program, their
ammonia generating factors would also have to be compared. This ammonia-generating
factor, denoted here E, is defined as the fraction of the ammonia generated in ppm,
divided by the nitric oxide reduced. It is based on the assumption that all NH3 gener-
ated derives from the original NO concentration, and none from the free nitrogen
molecules in the gas mixture. To be expressed in percent E, is given by the
expression
E =
c
XN01]
x 100%
where c is the concentration of NH3 generated, XNO is the initial NO concentration,
and 1], as before, the NO reducing efficiency; however, it is now expressed as a
number between 0 and 1, and not in percent:
o $ 1] $ 1
-------
This factor E has been tabulated in Table XVITI where the catalysts are arranged
according to their values of Tf, in descending order. The mean of the concentrations
of NH3' plus the percent error due to the maximum deviations in c are noted in the
table, resulting in a mean value of E, as well as in the tabulation of the total range
in E for each catalyst, due to the error in c. The data for c have been obtained from
Table XVII in Chapter 4. The experimental conditions are denoted beneath the title
of Table XVTII.
TABLE XVIII
COMPARISON OF AMMONIA GENERATION FACTOR E (IN % OF NO REDUCED)
FOR THE VARIOUS CATALYSTS TESTED AND YVITH PREHEATER
PL US EMPTY BED ONLY
NO = 1000 ppm Nominal Initially, CO/NO = 10
T = 1400° F, cP = 9 l/min, in Basic Exhaust Gas Mixture A
Overall NO Ave. NH3 Maximum Deviation % Error in NH3 Mean Value Total Range in E I
Reducing Eff. Generated in NH3 Generated Ganerated of E (based on max. dev. I
Catalyst TJ ppm ppm (from max. dev.) % in NH3) %
I
Monel 400 0.95 12.6 ,j, 4.2 l ,j, 33 1. 21 0.81 - 1.61
Monel 400-Mod. 0.93 10.8 ,j, 3.8 ,j, 35 1. 06 0.69-1.41
CuCr207 0.80 9.1 t 2.4 ,j, 26 1.03 0.76 - 1. 30
Pd on A 1 203 0.76 218 ,j, 2 ,j, 0.92 26.23 25.97 - 26.49
PtonAl203 0.73 454 ,j, 46 ,j, 10.13 56.75 51.00-62.50
I
RE(23) 75-P 0.47 174 ,j, 28 ,j, 16 33.65 28.27 - 39.03
RE(M3) 75 0.06 31. 6 z 20.6 ,j, 65 47.88 16.76 - 79.00
Empty Bed 0.03 5.0 ,j, 2.4 ,j, 48 15.15 7.88 - 22.42
As a result of the data of Table xvrn, the two Monel catalysts and CuCr207
appear all to be in the same leading class, generating the least percentage of ammonia.
Solar's Monel 400-Mod. seems here to lead the ordinary Monel formulation, pointing
to the conclusion that modifications of Monel's 0
-------
The next higher class in ammonia formation is represented by the two rare
earth oxide catalysts. Formula RE(23)75-P generates about twice as much ammonia
than does the "empty catalyst bed". Formula RE(M3)75 appears to have a slightly
higher value of E than its pure oxides sister formulation, however, due to the large
scatter in the data on its NH3 generation, the range in E is so large tha.t nothing.
definite can be said about it on the basis of these results. If one takes mto consIdera-
tion, however, the data obtained by the American Oil Company, as listed in Table III
of Appendix F, one sees that the true value of E for RE(M3) 75 falls into the low part
of the range as shONn in Table XVITI, and this catalyst generates less ammonia than
RE(23)75-P and even the palladium catalyst; as a matter of fact, RE(M3)75 falls into
the same class as the "empty bed"-
The catalyst with the highest ammonia formation factor is the platinum one,
which turns over one half of the NO reduced into NH3' This suggests that the pt
catalyst is not suited as a catalytic agent for the reduction of NO. The catalyst was
included in this study for comparative purposes.
A new theory of the generation of ammonia in automobile exhaust gases,
indicating the complexity of the catalytical reactions in these gas mixtures, has been
proposed by Richard L. Klimisch and Gerald J. Barnes of GM (Ref. 20). They believe
that the carbon monoxide is not necessarily the prime reducing agent for the nitric
oxide in automobile exhaust gases. Instead the carbon monoxide reacts with water to
yield carbon dioxide and hydrogen; the hydrogen then reduces the nitric oxide, producing
ammonia in the process. As exhaust temperature increases, less ammonia is genera-
ted because it becomes unstable and decomposes into nitrogen and hydrogen. From
this theory there follows that a catalyst generates less ammonia if it can bring
preferentially CO and NO togetter, instead of H2 and NO.
Both of the above theories, namely that NH3 is a byproduct of the NO-CO
reaction, or NH3 is an intermediate step, assume essentially that all NH3 generated
derives from the original NO concentration. However, there is another school of
thought which states that at sufficiently high temperatures in exhaust gases ammonia
is generated from nitrogen itself by reactions with hydroxide radicals and hydrogen
atoms, yielding both ammonia and nitric oxide (Ref. 21). Under such conditions our
ammonia generating factor E would lose all significance. However, in the exhaust
gases the temperature is too small for free hydroxide radicals to exist, and, following
Klimisch's reasoning, any ammonia generated at the combustion temperatures will
automatically decompose into its elements, before getting into the exhaust stream.
5.2
PARAMETRIC DESIGN CHARTS FOR CATALYST FORMULA RE(23)75-P
The experimental data for catalyst RE(23)75-P have been summarized in a
nomograph type chart and a standard version parametric design chart has been con-
structed. Both these figures shall be discussed in detail below.
-------
No charts have been made for RE(M3)75, since its efficiency is insignificant.
The other catalysts considered in this program are well known in the literature.
Solar's catalyst Monel 400-Mod. behaves essentially in the same manner as the
standard Monel 400.
Figure 32 presents a summary of all experimental data for the catalyst
RE(23)75-P in the form of a nomograph, with an extension to the CO/NO ratio of 15.
The variables have been combined into two groups: the CO/NO ratio and the
initial NO concentration being the first variable, and the catalyst temperature and
exhaust gas flow rate being the second variable. By connecting a point starting with
the selected CO/NO ratio and the NO concentration, to a point showing the correct
catalyst bed temperature and gas flow rate , and extending this line to the efficiency
scale, the efficiency of the catalyst RE (23) 75- P is obtained. The data points for the
CO/NO ratio of 5 and 10 for all the temperature-flowrate combinations listed in the
figure have actually been measured. The higher CO/NO ratio of 15 is the only exten-
sion of the experimental data in this figure.
EFFICIENCY - ry (70)
75
NO
RATIO CONCENTRATION
CO/NO (PPM)
5 1000
5 500
FLOW
RATE
<.?
TEMPERATURE U/MIN)
(0 F)
1600 1
70 70.0
1600 4
1400 1
1200 1
1400 4
1200 4
66.2
65 65.3
63.0
62.1
60 59.8
10
10
1000
500
55
1600 9 50 51.1
1400 9 47.8
12001000 9 46.0
16001000 118 45 45.5
4 44.6
1400 18 43.2
1200 18 41.4
40 40.0
1000 9 35
1000 18 33.1
30
29.0
25
15
15
1000
500
FIGURE 32.
VERSION OF PARAMETRIC DESIGN CHART FOR RE(23)75-P
IN NOMOGRAPH FORM
-------
A conv'entional Parametric Design Chart is given in Figure 33. The curves
for the initial NO concentration had, however, to be reversed from the originally
planned design, in order to better approximate the effect of the parameter of the
catalyst. This Parametric Design Chart is only an approximation of the actual function
of the several parameters of the NO reducing efficiency of the catalyst RE (23) 75- P,
because it had been found during experimentation that the parameters were not indep-
endent of each other, as had been predicted by the theoretical analysis of the CO-NO
reaction as presented in Chapter 2. The reason for this might be due to the fact that
the catalytical reaction in the presence of oxygen, hydrogen, water vapor and hydro-
carbons proceeds according to the theory of Klimisch, via the production of ammonia
and other intermediaries, rather than the straight CO + NO equals N2 + C02 reaction.
However, for the data points, around which this Parametric Design Chart was con-
structed, the resulting curves are correct. These data points are indicated by the
dashed line in Figure 33. In addition, for each independent quadrant of this figure the
parameter curves are also correct. It is only due to the fact that the various param-
eters can clearly be seen. The efficiency increases both with flowrate and temperature
though not uniformly. For instance, for an efficiency of 80 percent at the flowrate
5
/ L/j
/ ~ ,,~v
~ ~/ ~'
/ I ~V.~
~ ~ r/ (,')C) ,<::-
I ~~' ~
" v ."
~ ~ '/ / I I ~ /o,""'~" v v\~'-
'/ <:>." " v
}~ // /. / '~'-' -.K- v # ~""
A
~A' / '/ ~ VI ~
~ <:>. 0"«.
); ,," 0"«. I 0'<. V ~ a
~'\: ,," . ,," 0<';
i/ v: V~"-+- 1~ ~ ~ i
1'/ '/ '0" ~
1/ f'l V /1/< ~ 17/ ~ i
! 0 20 40 60 80 100
EFFICIENCY - ~ (oj.)
~ ~
. .-
I ~
~.
I <.;:~.'\ ,\,,,
iii-C; Q;-'I-
I c;" ~'\
~c iii-C;'<;
'- .- f+. ~C;"
-. --:~ ~c
-------
of 9000 cc/min we see that for a 1000 ppm NO concentration and the 9000 F catalyst
bed temperature, the CO/NO ratio has to be near 10. As the temperature increases
to, for instance, 12000 F, the CO/NO ratio can be much less than 5.
Theory does not predict any effect of the initial NO concentration on the final
catalyst efficiency. The difference obtained in our experimental data may be due,
again, to the secondary reactions, such as described by Klimisch, for instance. The
chart does follow the experimental results, however, and indicates a slightly higher
efficiency for the 500 ppm initial concentration.
-------
6
SUMMARY AND CONCLUSIONS
In this program, two rare earth oxide catalysts,
. Catalyst RE(23)75-P
. Catalyst RE(M3)75;
four transition metal catalysts, namely
. Monel 400
.Solar Monel 400-Mod
.Solar Monel 400-Core
.Copper Dichromate;
and two noble metal catalysts, i. e. ,
. Palladium impregnated on alumina pellets
. Platinum impregnated on alumina pellets,
were investigated as to their respective nitric oxide reducing efficiency and their
ammonia generating factors. These catalytic reactions were taking place in a synthetic
automobile gas exhaust mixture, prepared to specifications approximating an air/fuel
ratio of 14.
These results would indicate that in order to obtain an efficient rare earth
oxide catalyst, the economy feature would be lost, due to the fact that the impurities,
contained in the mixed oxide form, tend to inhibit the catalyst reaction. This seems
to be valid on both counts -- the direct NO reduction and the indirect ammonia for-
mation. Hence, the RE(M3)75 formulation showed a superior characteristic with
respect to the ammonia formation than did the high purity formulation. However,
both rare earth oxide catalysts, as presently formulated, show an insufficient catalytic
nitric oxide reducing efficiency, and an unacceptable high rate of ammonia formation.
Thus, this program, which started out to demonstrate the capabilities of rare earth
oxide catalyst, which had appeared rather promising on the basis of earlier experi-
-------
mentation in the absence of secondary reactions, proved to be insufficient for the
purpose intended; namely, the control of NO emissions from mobile sources.
However, it might be pointed out here that additional experimentation with
the catalyst RE(23)75-P showed a NO reducing efficiency of over 90 percent in the
absence of H20 in the synthetic exhaust gases. All the other competing gas compon-
ents were present for those tests, except the water vapor.
On the other hand, Monel 400 was confirmed to be the best NO reducing catalyst
on the basis of our data. These data show that Monel outranked even the precious metal
catalysts such as Platinum and Palladium. These results confirm the data in the
literature regarding Monel 400, but the noble metal catalysts did not turn out too
well. In addition, the ammonia generating factors of the noble metal catalysts are
also unacceptable.
A comparison of various catalysts, reported in the literature (Ref. 4, 5, 6,
7, and 22) has been made, to which have been added the data on catalyst RE(23)75 -P,
obtained in the program. This data is tabulated in Table XIX. All important physical
and chemical features of the catalysts have been placed into the table, as far as they
have been reported. The data verified that the space velocity used for the rare earth
oxide catalyst corresponds to that which has been applied on most of the other catalysts
studied in the literature. Again, it is apparent from Table XIX that Monel catalyst is
the leader, both in terms of efficiency and space velocity. In terms of ammonia
formation, Monel again is the leader with the least quantity generated while the rare
earth oxide catalyst occupies a middle position. Thus, the data agree generally with
those in the literature.
The secondary parameter of the surface area has been shown to be rather
insignificant with the Monel catalyst, a fact which also has been verified by our own
testing. It is probable that DOPfs noble metal catalyst makes up on surface area
what it lacks on catalytic activity, and thus remains competitive. However, the data
published on it (Ref. 22) are meager.
In conclusion, the following observations can be stated:
The rare earth oxide catalysts by themselves do not appear to be suited for
the function of controlling the NOx emissions of exhaust gases.
Furthermore, it appears that more knowledge of the intermediate and side
reactions in the exhaust system for the CO/NO reaction is needed for the design of
an optimum catalyst.
-------
-;J
c.o
T ABLE XIX
STATE OF ART IN NO REDUCING CATALYSTS
Producer Esso American Oil and Ford UOP Solar) I Chemetron Girdler I Harshaw Chemical I Lmde(UC) I
Catalyst Monel Copper Based Iron Oxide Mixed Oxide(Pt) "O('~"'''-"t . , Noo:~r-::::;.:':~"" Cobalt, Vanadium. I
Stainless Steel on Alumina Alumina
Surface Area Less than 0.01 m2/g 390 to 57 m2/g 400 m 2 /g 0.2 till ]J)m2/g _n- --- -- - - Not Measured
-1 -1 -1 -1 -1 -1 -1 -1
Space Velocity 100,000 hr. 15000-50000 hr. 15000-50000 hr. 15,OOOhr.-1 15,000 hr. 15,000 hr. 15,000 hr. 15,000 hr. 15,000 hr.
DlrsbUity 31,000 mi. 10,000 mi. Insufficient Probably some Poor, so far I I !
Activation None Useful High temp. Yes Yes Yes Yes, 3 hrs. 550°C Yes Yes
active + CO 3% Co in N2
Deactivation None Yes, after Yes I
2+ hours I
Temperature 1300° F up 660-800-1OO00F 1600-19000F Around 12oo0F 10oo0F 10000F 10000F 1OO0°F
1700F pretreated
Ammooia Formation 10% of NO 50% Or attrition resistance
CuCr 204
12 to 25 Ibs. per unit
$1. -2. 5011b.
-------
The program resulted in the design and putting in ope ration of an experimental
approach capable of accurately measuring catalyst performance under synthetic exhaust
gas conditions for an air-fuel ratio near 14.
It appears from this program that the best future approach in the development
of an appropriate catalyst for the CO/NO reaction lies in increasing the life of the
catalyst material -- Monel 400 (Solar Monel 400-Mod series).
-------
1.
2.
3.
4.
REFERENCES
Anon, "Air Quality Criteria for: Carbon Monoxide, Nitrogen Oxide From
Mobile Sources, Hydrocarbons", U. S. Dept. of Health, Education and
Welfare, Public Health Service, National Air Pollution Control Administration
,
Washington, D. C. (March 1970).
Lewis, James, "peroxyacyl Nitrates", Task Force 7, Section 4, Project
Clean Air, University of California Task Force Assessments (9/1/70).
Shelf, M., "Catalytic Reduction of Nitric Oxide", First National Symposium
on Heterogeneous Catalysts for Control of Air Pollution, Franklin Institute,
Philadelphia, Pa., published by Scientific Research Staff, Ford Motor
Company (November 21-22, 1969).
Bernstein, L. S., et al., "Application of Catalysts to Automotive NOx
Emissions Control", Esso Research and Engineering Co., Standard Oil
Co. (N) SAE Paper #710014 (January 1971).
5.
Meguerian, G. H. (American Oil Company), Lang, C. R.
"NOx Reduction Catalysts for Vehicle Emission Control",
#71 0291 (January 1971).
(Ford Motor Co.),
SAE Pape r
6.
Yolles, R. S., Wise, H., and Berriman, L. P., "Study of Catalytic Control
of Exhaust Emissions for Otto Cycle Engines", SRI Project PSU-8028,
Standford Research Institute (April 1970) .
7.
Balgard, W. D., "Novel Catalysts and Catalyst Supports for Control of NOx
in Automobile Exhaust", presented at the Second International Clean Air
Congress, Washington, D. C. (Dec. 6-11, 1970).
8.
Penner, S. S., Thermodynamics,
Wesley Publishing Co., Inc. (1968).
Vol. 7, Reading, Mass.: Addison-
9.
(a) Progress in Technology, Vol. 6, "Vehicle Emissions - A Selection of
SAE Technical Papers on the Definition, Measurement, and Control of
Emissions from Motor Vehicles to the Atmosphere", covers period 1955-
1963. Includes bibliography with abstracts of SAE material.
-------
9
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
(b) Progress in Technology, Vol. 12, "Vehicle Emissions - Part 2". Thirty-
six technical presentations dating from 1963 to 1969 form the second volume
of a comprehensive, evaluated technical work in the emissions field. Major
sections include Fundamentals, Control, Measurement, Diesel,
and Evaporation Loss.
Bevan, D. J. M., Hyde, B. G., and Eyring, LeRoy, "Phase Relationships
in the PrOx System", Third Rare Earth Conference, Session III, pg. 72
(1963) .
Kuo, J. C. (Mobile R&D Corp.), Lassen, H. G. (Ford Motor Co.), and
Morgan, C. R. (Mobile R&D Corp.), "Mathematical Modeling of Catalytic
Converter System", presented at the Automotive Engineering Congress,
Detroit, Mich., SAE paper #710289, Appendix D (Jan. 11-15, 1971).
Jagel, K. I., and Dwyer, F. G., "HC/CO Oxidation Catalysts for Vehicle
Exhaust Emission Control", presented at the Automotive Engineering
Congress, Detroit, Michigan, SAE paper #710290 (January 11-15, 1971).
D'Alleva, B. A. and Lozell, W. G., SAE Journal, Vol. 38, No.3 (March 1936)
'NoIf, P. C., "Carbon Monoxide Measurement and Monitoring in Urban Air",
Environmental Science and Technology, pg. 214 (1971).
General Course in Gas Chromatography, Hewlett Packard Publication.
Jones, Theron, "Beckman Gas Chron.atography Applications Manual",
Beckman Bulletin 756-A.
'Vestwood, L. C., "A Chemical Method for the Determination of Ammonia in
Automotive Exhaust", Technical Memorandum SRM 70-25, EPA-Technical
Services and Administration Dept. (9/22/70).
Obert, Edward F. , Internal Combustion Engines - Analysis and Practice,
2nd ed., 4th printing, Chapter 11 - Carburation, pp. 327 ff, especially
Figures 11-3 and 11-13, Scranton, Penna.: International Textbook Company,
(June 1952).
Eyring, LeRoy, Ed., Progress in Science and Technology of the Rare-Earths,
New York: The McMillan Co. (1964).
Environmental Science and Technology, Vol. 5, No.7, pg. 579 (July 1971).
-------
21.
22.
23.
24.
25.
26.
Bowman, C. T., "Investigation of Nitric Oxide Formation Kinetics in
Combustion Processes: The Hydrogen-Oxygen-Nitrogen Reaction", United
Aircraft Research Laboratories paper, submitted for publication to Combustion
Science and Technology.
Zmuda, J. P., "New: A Catalytic Converter that Really Cleans up Auto
Exhaust", Popular Science Monthly, pg. 47 (Dec. 1970).
Gschneidner, Karl A., Jr., "Rare Earths, The Fraternal Fifteen", U. S.
Atomic Energy Publication.
Anon, "Present and Potential Sources of Rare-Earth Minerals", Industrial
Minerals, London, England (Nov. 1968).
Private communication with J. G. Cannon, Manager, Rare-Earth Develop-
ment, Molybdenum Corp. of America, New York, New York.
"Problems and Issues of a National Materials Policy, " compiled by Science
Policy Research Division, Legislative Reference Service, Library of
Congress, for Committee on Public Works., U. S. Senate, requested by
Hon. J. Caleb Boggs (December 1970).
-------
-------
APPENDIX A
CALCULATED ISO-OCT ANE/ Am EQUILIBRIUM EXHAUST COMPOSITION
USING THE "DYNAMIC SCIENCE" GENERALIZED KINETICS PROGRAM
Tables A-I through A-VI show the calculated equilibrium exhaust composition
at OfF ratios from 10-15 when octane and air are burned. For each O/F, compositions
are presented at temperatures from 2300-10000 F. These are read columnwise and the
major species concentrations presented directly below. It is interesting to note that
no NO exists at these temperatures at equilibrium. NO is a product that is formed at
high temperature and "freezes" during temperature decreases. It also can be shown
that CO concentrations measured in engine exhaust can be an order of magnitude
different from that at equilibrium.
A scattering of runs designed to look at the range of exhaust conditions were
made with the Generalized Kinetics Program. Measured emission values coupled with
equilibrium values of the major species (C02, H20, N2) were used as exhaust com-
position. Major species were selected from Tables A-I through A-VI at conditions
representative of the particular engine operating mode being examined. Runs were
made with and without the addition of trace (2000 ppm) 02.
The results of these data are summarized in Section 2.2.
-------
TABLE A-I
ISO-OCTANE/AIR EQUILIBRIUM EXHAUST COMPOSITION
o /F = 10
CASE NO.1
Chemical Formula
Fuel C 8.00000 H 18.00000
Oxidant 0 2.00000
Oxidant N 2.00000
O/F = 1. 0000 + 01, Percent Fuel = 0.0909 + 00, Equivalence Ratio = 1. 4371 + 00
THERMODYNAMIC PROPERTIES
P, psi 14.70 14.70 14.70 14.70 14.70 14.70
T, oR 2300 2000 2000 1500 1250 1000
H, BTU/lb -595.8 -700.9 -771. 2 -894.2 -1064.0 -1170.2
S, BTU/(lb)(R) 2.2408 2.1918 2.1548 2.0794 1.9559 1. 8620
M, MOL wt. 26.266 26.267 26.268 26.503 27.817 28.426
(DLV /DLP)T -1. 00000 -1. 00001 -1. 00010 -1. 01184 -1. 01079 -1. 00185
(DLV /DLP)P 1.0000 1. 0001 1. 0014 1. 2017 1.1920 1. 0361
CP, BTU/(lb)(R) .3509 .3505 .3530 .6095 .5585 .3331
Gamma (S) 1. 2749 1. 2753 1. 2736 1. 1988 1.2062 1. 2878
Son Vel. ft/sec 2355.9 2.197.3 2083.0 1836.6 1641. 6 1500.8
MOLE FRACTIONS
CH4 . 000000 .000001 .000021 . 004469 .029484 .041087
CO .085861 .075740 .066736 . 043207 . 006440 . 000171
C02 .081367 .091489 .100480 .121055 . 1411 78 .139721
H2 .073081 .083194 .092113 .099209 . 043838 .007419
H20 .115048 .104929 .095967 . 081595 .096318 .113932
NH3 . 000002 .000006 .000013 . 000055 . 000076 . 000050
N2 .644640 .644642 .644669 .650410 .682665 .697620
-------
T ABLE A-II
ISO-OCT ANE/ AIR EQUILIBRIUM EXHAUST COMPOSITIONS
OfF = 11
CASE NO.1
Chemical Formula
Fuel C 8.00000 H 18.00000
Oxidant 0 2.00000
Oxidant N 2.00000
OfF = 1.1000 + 01, Percent Fuel = 8.3333 + 00, Equivalence Ratio = 1. 3064 + 00
THERMODYNAMIC PROPERTIES
P, psi 14.70 14.70 14.70 14.70 14.70 14.70
T, oR 2300 2000 1800 1500 1250 1000
H, BTU/lb -652.3 -755.3 -823.8 -931. 5 -1070.0 -1172.3
S, BTU/(lb)(R) 2.1985 2.1506 2. 1145 2.0488 1. 9476 1. 8571
M, MOL wt. 26. 919 26.919 26.919 26.995 27.903 28.461
(DLV /DLP)T -1. 00000 -1. 00000 -1.00003 -1. 00462 -1. 00955 -1. 00175
(DLV/DLP)P 1. 0000 1. 0000 1.0004 1.0773 1. 1679 1.0341
CP, BTU/(lb) (R) .3442 .3424 . 3425 .4355 .5193 .3278
Gamma (S) 1. 2731 1. 2749 1. 2750 1. 2369 1. 2158 1.2920
Son Vel. ft/sec 2325.5 2170.1 2058.8 1.848.6 1645.6 1502.3
MOLE FRACTIONS
CH4 .000000 .000000 .000005 .001384 .018247 .028619
CO .063708 .055002 .047309 .030662 .005395 .000149
C02 .093393 .102099 .109790 .125496 .139203 .137333
H2 .051448 .060149 .067817 .079222 .040880 .007059
H20 .125289 .116584 .108902 .095184 .105722 .122496
NH3 .000001 .000004 .000008 .000040 .000069 .000047
N2 .666161 .666162 .666169 .668011 .690483 .704298
-------
TABLE A-lIT
ISO-OCT ANE/ AIR EQUILIBRIUM EXHAUST COMPOSITIONS
OfF = 12
CASE NO.1
Chemical Formula
Fuel
Oxidant
Oxidant
C 8.00000
o 2.00000
N 2.00000
H 18.00000
OfF = 1. 2000 + 01, Percent Fuel = 7.6923 + 00, Equivalence Ratio = 1. 1976 + 00
THERMODYNAMIC PROPERTIES
P, psia 14.70 14.70 14.70 14.70 14.70 14.70
T, oR 2300 2000 1800 1500 1250 1000
H, BTU/lb -700.6 -801. 4 -868.0 -968.0 -1076.9 -1174.2
S, BTU/(Lb)(R) 2.1614 2.1145 2.0794 2.0186 1. 9387 1. 8525
M, MOL wt. 27.497 27.497 27.497 27.512 28.010 28.494
(DLV /DLP)T -1. 00000 -1. 00000 -1. 00001 -1. 00103 -1. 00768 -1. 00161
(DLV /DLP)P 1. 0000 1.0000 1.0001 1. 0168 1.1337 1.0312
CP, BTU/(lb)(R) .3374 .3340 . 3322 .3469 .4687 .3216
Gamma (S) 1. 2725 1. 2761 1. 2781 1.2726 1. 2298 1. 2970
Son Vel. ft/sec 2300.5 2148.2 2039.6 1857.3 1651. 8 1504.3
MOLE FRACTIONS
CH4 . 000000 .000000 .000001 . 000266 . 009309 .018111
CO . 043238 .036561 .030685 .919241 .004327 .000127
C02 .104892 .111570 .117445 .128705 .137260 .135264
H2 .033131 .039807 .044675 . 056071 .036146 .006508
H20 . 133514 .126837 .120964 .110100 . 114908 .129898
NH3 .000001 .000002 .000005 .000024 .000058 .000042
N2 .685224 .685224 .685226 .685593 .697992 . 710051
-------
T ABLE A-IV
ISO-OCT ANE/ Am EQUILIBRIUM EXHAUST COMPOSITIONS
OfF = 13
CASE NO.1
Chemical Formula
Fuel C 8.00000 H 18.00000
Oxidant 0 2.00000
Oxidant N 2.00000
OfF = 1. 3000 + 01, Percent Fuel = 7.1429 + 00, Equivalence Ratio = 1.1055 + 00
THERMODYNAMIC PROPERTIES
P, psia 14.70 14.70 14.70 14.70 14.70 14.70
T, oR 2300 2000 1800 1500 1250 1000
H, BTU/lb 1742.4 -840.9 -905.6 -1001. 0 -1086.3 -1176.2
S, BTU/(lb)(R) 2.1281 2.0822 2.0482 1.9902 1. 9277 1. 8479
M, MOL wt. 28.013 28.013 28.013 28.014 28.168 28.528
(DLV /DLP)T -1. 00000 -1.00000 -1. 00000 -1. 00010 -1. 00446 -1. 00137
(DLV /DLP)P 1.0000 1.0000 1.0000 1. 0015 1. 0768 1. 0267
CP, BTU/(lb)(R) . 3307 .3254 .3216 .3149 . 3908 .3135
Gamma (S) 1. 2731 1.2788 1. 2830 1. 2917 1.2576 1.3035
Son Vel. ft/ sec 2279.7 2130.6 2024.5 1854.3 1665.8 1507.2
MOLE FRACTIONS
CH4 .000000 .000000 .000000 .000022 .002752 .009185
CO .024186 .020045 .016411 . 009730 .003023 .000104
C02 . 115943 .120084 .123717 .130384 . 135130 .133419
H2 .017587 .021728 .025359 .031943 . 027913 .005648
H20 .140057 .135916 .132283 .125651 .125041 .136477
NH3 . 000000 .000001 .000002 .000010 .000040 .000034
N2 .702227 .702227 .702227 . 702259 .706100 .715133
-------
TABLE A-V
ISO-OCTANE/Am EQUILIBRIUM EXHAUST COMPOSITIONS
OfF = 14
CASE NO.1
Chemical Formula
Fuel
Oxidant
Oxidant
C 8.00000
o 2.00000
N 2.00000
H 18.00000
OfF = 1. 4000 + 01, Percent Fuel = 6.6667 + 00, Equivalence Ratio := 1. 0265 + 00
THERMODYNAMIC PROPERTIES
P, psia 14.70 14.70 14.70 14.70 14.70 14.70
T, oR 2300 2000 1800 1500 1250 1000
H, BTU/lb -779.0 -875.1 -937.9 -1029.6 -1103.4 ..,1178.6
S, BTU/(lb)(R) 2.0971 2.0523 2.0192 1. 9635 1. 9098 1.8426
M, MOL wt 28.475 28.475 38.475 38.475 28.477 38.573
(DLV /DLP)T -1.00000 -1.00000 -1. 00000 -1. 00000 -1.00013 -1. 00085
(DLV /DLP)P 1. 0000 1.0000 1.0000 1.0000 1. 0021 1. 0164
CP, BTU/(lb)(R) .3241 .3166 .3109 .3001 .2912 .2977
Gamma (S) 1. 2744 1. 2828 1. 2894 1.3030 1. 3167 1. 3167
Son VeL ft/sec 2262.3 2116.5 2013.1 1847.3 1695.2 1513.6
MOLE FRACTIONS
CH4 . 000000 .000000 .000000 .000000 .000031 . 001714
CO .006344 .005157 .004118 .002291 .000915 . 000066
C02 .126603 .127790 .128829 .130656 .132010 .131626
H2 .004379 .005566 .006604 .008430 .009672 .003810
H20 .145186 .144000 .142961 .141134 . 139829 .142814
NH3 . 000000 .000000 .000000 .000001 .000008 .000019
N2 .717487 .717488 .717488 .717488 .717534 . 719951
-------
TABLE A-VI
ISO-OCTANE/All EQUILIBRIUM EXHAUST COMPOSITION
OfF = 15
CASE NO.1
Chemical Formula
Fuel C 8.00000 H 18.00000
Oxidant 0 2.00000
Oxidant N 2.00000
OfF = 1. 5000 + 01, Percent Fuel = 6.2500 + 00, Equivalence Ratio = 9.5806 - 01
THERMODYNAMIC PROPERTIES
P, psia 14.70 14.70 14.70 14.70 14.70 14.70
T, oR 2300 2000 1800 1500 1250 1000
H, BTU /lb -742.0 -836.9 -898.7 -988.8 -1061. 0 -1130.7
S, BTU/(lb)(R) 2.0853 2.0411 2.0086 1.9539 1. 9012 1.8391
M, MOL wt 28.645 28.645 28.645 28.645 28.645 28.645
(DLV /DLP)T -1.00000 -1. 00000 -1. 00000 -1.00000 -1.00000 -1.00000
(DLV /DLP)P 1. 0000 1.0000 1.0000 1.0000 1.0000 1.0000
CP, BTU/(lb)(R) .3206 .3120 .3057 .2943 .2838 .2733
Gamma (S) 1. 2761 1. 2859 1.2936 1.3084 1.3235 1.3401
Son Vel, ft/sec 2257.1 2112.7 2010.3 1845.6 1694.5 1525.1
MOLE FRACTIONS
C02 .125381 . 125381 .125381 .125381 .125381 .125381
H20 .141052 .141054 .141054 .141054 .141054 .141054
NO . 000072 . 000020 . 000007 . 000001 . 000000 . 000000
N2 . 724952 .724979 .724986 .724989 .724989 .724989
OH . 000004 . 000000 . 000000 . 000000 .000000 . 000000
02 . 008538 . 008566 .008572 . 008575 . 008576 . 008576
-------
-------
APPENDIX B
CATALYST SURFACE AREA MEASUREMENTS BY THE BET METHOD
CATALYST SURFACE AREA DETERMINATION USING THE BET EQUATION
The surface area of porous and non-porous solids may be estimated from
measurements of the adsorption of gases. The starting point is the experimental
determination of the adsorption isotherm, and then calculating from it. using the
BET (Brunauer, Emmett and Teller) equation, the monolayer capacity of the
adsorbent solid. The monolayer capacity is defined as the amount of adsorbed gas
required to form a completely filled, single layer of molecules on the surface of the
solid. The surface area, per unit mass of solid, can be calculated from the mono-
layer capacity if the area occupied by a single molecule of adsorbed gas in the
completed monolayer is known.
The adsorption isotherm is a plot of the amount adsorbed, per gram of
solid, as a function of the equilibrium pressure, for a given gas adsorbed on a
given solid at a constant temperature. A general expression for the adsorption
isotherm is of the form
X = f(P/Po)T, gas, solid
(1)
where X is the amount of gas adsorbed per gram of adsorbent. P is the equilibrium
pressure, Po is the saturation vapor pressure of the gas, and T is the temperature.
For this ana}ysis X will be expr'essed as a volume reduced to (N. T. P.) standard
temperature and pressure (0° C, 760 Torr), i. e., the volume the adsorbed gas
would occupy at normal temperature and pressure.
The BET Equation and Surface Area
Braunauer, Emmett. and Teller developed a mathematical expression for
the adsorption isotherm from a simple kinetic model. This expression, the "BET
equation", has proved remarkably successful in the calculation of surface areas
and net heats of adsorption from adsorption isotherms.
-------
The form of the BET equation most often used in interpreting experimental
data is
P
X(Po - P)
= ~ + C -1
XmC XmC
(P: )
(2)
where X is the amount adsorbed (cm3 at N. P. T.) per gram of adsorbent at pressure
P, Xm is the amount adsorbed when the monolayer is completed, Po is the saturation
vapor pressure of the adsorbed gas, and C is a constant related to the latent heat of
condensation and the net heat of adsorption. According to Equation (2), when the
experimentally determined adsorption isotherm data (X vs. P) is plotted in the form
P P
X(Po - P) vs. Po '
a straight line should result with slope
S = (C - 1)/XmC
(3a)
and intercept
I
= 1/XmC .
(3b)
Solution of these two simultaneous equations immediately yields Xm and C, viz.,
Xm = 1/(S + I)
(4a)
C
= (S + I) II .
(4b)
The surface area As(m2/gm) is related to the monolayer capacity Xm by
the equation
A
s
=
XmNAmx 10-20
22414
where N is Avogadro's number and Am is the molecular cross sectional area of the
adsorbed gas in square Angstroms. The cross sectional area Am is usually calcu-
lated from the dens ity of the gas in the ordinary liquid form, i. e., it is assumed
that the arrangement of the molecules on the surface is the same as it is within the
bulk liquid. For nitrogen at -195°C, Am = 16.2 A2.
-------
Measurement of Adsorption Isotherms
A number of different experimental methods have been successfully used for
the determination of adsorption isotherms. The most common is the volumetric
method which consists essentially in admitting to the adsorbent successive charges
of gas from some sort of volumetric measuring device; when equilibrium is reached
the pressure of the gas in the dead space is measured, and the quantity of gas re-
maining unadsorbed is then calculated with the aid of the gas laws. Knowing the dead
space volume, and by subtracting the amount remaining unadsorbed from the total
amount which has been admitted, the quantity which has been adsorbed can be
calculated.
Figure B-1 shows a schematic of the volumetric adsorption apparatus used
for determining the adsorption isotherm of N2 at -1950 C on a number of perspective
catalyst materials. The catalyst sample is contained in a small glass bulb which is
immersed in a liquid nitrogen bath. Gas pressures are measured by the manometer.
The reference volume is used for calibrating the various parts of the dead space and
must be of accurately known volume. The volume of the line J!:. should be either small
compared to the ref~rence volume or measured and included in the reference volume.
The total dead space volume of the system is defined as the sum of:
l.
2.
The volume of lines..Q, and~
The space above and surrounding the catalyst particles.
3.
The pore volume of the particles.
This means that the amount adsorbed is equal to the amount of gas admitted
(by valve 1 from the reference volume) minus the amount of gas present in the dead
space, or, mathematically,
X=Va-Vd'
where X is the volume of gas adsorbed, Va is the amount admitted, and V d is the
amount in the dead space. The dead space volume is determined using a charge of
helium in the reference volume since it will act to fill the dead spaces but not be
adsorbed.
-------
TO VACUUM PUMP
...
LIQUID N2
BATH
MANOMETER
a
GAS INLET
...
N2' He
FIGURE B-l. SCHEMATIC DIAGRAM OF THE APPARATUS USED FOR THE
DETERMINATION OF ADSORPTION ISOTHERMS
The BET Surface Measurement of the Catalyst
A BET surface measurement apparatus has been constructed and the
surface area of a charge of the catalysts has been measured. The procedure used
is as follows.
After the determination of the dead space, which has been discussed in the
above subsection, and which for the copper dichromate catalyst came out to be
28.9 cm3, the nitrogen adsorption isotherm measurements were made. For this,
various amounts of volumes of nitrogen gas were allowed into the system and the
gas pressure before and after adsorption onto the catalyst was measured. Several
points have been obtained, whereby the highest nitrogen gas pressure has to be
kept below 1/2 atm. The volume measured for each absorption consists of the
total volume admitted minus the volume in the dead space and the volume left in
the reference vessel. This volume is plotted versus a function Pf/Ps' where Pf
is the gas pressure remaining after adsorption and Psis the saturation vapor
pressure of nitrogen at the catalyst temperature of 74° K, that is the boiling point
of liquid nitrogen which is equal'to 1 atm. Using the BET equation, the data are
then replotted in the form of Pf divided by V(Ps - Pf) versus the ratio Pf/Ps and
the best straight line through the data points is drawn. For copper dichromate,
this is shown in Figure B-2. The slope S and the intercept I is measured and the
-------
O.ROO
0.770
0.740
0.710
0.680
0.650
0.620
0.590
0.560
0.530
0.:;00
M
0
~
>< 0.470
.....
~ 0.440
.... I
~ <0
& 0.410
> @
0.380
0.3:;0
@
0.320
0.290
0.260
0.2:30
0.200
0.170
10
20
30 40
50
60
70
80
90 100 110 120 130 140 150 x 10-:3
0.020-
PIPs
FIGURE B-2. ADSORPTION ISOTHERM FOR BET SURFACE AREA MEASUREMENT
OF CuCr207 CATALYST
-------
volume absorbed" by the monolayer of nitrogen is given by the equation
* 1
VM = S + I
(cm3 at STP)
The volume V per unit mass is obtained by dividing V M * by the mass M, which is
the weight of t~e catalyst charge. The catalyst surface then is given by the expression
AS =
-20
VMNAMx 10
22414
where
N = 6.0226 x 1023 molecules/mole
AM = 16, 2 A 2 /molecule for nitrogen
Thus, the expression for AS' using the constants defined, becomes
AS = 4.3529 x VM in m2/gram
For our charge of CuCr207' the value of the surface area is
AS = 6. 5 m 2 / gr am. **
This is, according to Reference 6, a very small surface area. A good
catalyst should have an area of the order of 70 m2/gtam. Therefore, if it takes
12 pounds of CuCr207 at 6.5 m2/gram to achieve a given catalytic conversion
efficiency for the reduction of NO, the same job can be completed by 1. 2 pounds
of a catalyst with a 65 m2/gram surface area. Our experimentation has shown that
such surface areas cannot be obtained with our present pelletized catalysts, but
would require supported ones.
The equipment for the BET apparatus is shown in Figure B-3. The
reference volume of 500 ml is clearly visible, while the volume with the catalyst
bed has been placed into the bath of liquid nitrogen in the vessel covered by black
tape. The mercury manometer, with which the pressure difference is measured,
is also seen. The system is evaluated by the vacuum pump system underneath the
** Later measurements, with better degassed CuCr207 pellets, yielded surface
areas of 10.3 and 11.0 m2/gram (see Table B-1).
-------
FIGURE B-3. VIEW OF BET SURFACE AREA MEASUREMENT APPARATUS
assembly, and the helium and nitrogen gas tanks used for measuring the dead space
and the absorbed volumes, respectively, is seen on the right hand side of the figure.
The results of the surface area measurements by the BET method are shown in
Table B-I.
In the BET data obtained, the material factor as well as preparation para-
meters were shown to play an important role. The high purity oxide pellets had a
larger surface area per unit mass than the mixed oxide pellets; the latter had initially
a much greater hardness as well. However) after a storage of a few months, many
of the RE(M3)75 pellets showed signs of lamination cracking, while the RE(23)75-P
pellets proved to be much more stable.
Preparation parameters showing some influence on the pellet surface area
were temperature and pressure, with preparation temperature to be the one with
greater effectiveness.
-------
TABLE B-1
RESULTS OF BETA SURFACE AREA MEASUREMENTS OF CATALYSTS
Catalyst Weight After Total Surface Spe cifi c
Sample Degassing Surface Area
(Pellet Form) 2 m2jg
g m
CuCr ° * 19,9887 60.0 10.3
2 7
CuCr207 19,9887 64.1 11. 0
RE(M3)75 22,3959 0.8 0.109
RE(M3) 75** 20,6360 1.4 0.24
RE(23)75-P 20,9296 5.0 0.81
* Incompletely outgassed
** Catalyst Pellets with moderate lamination cracking.
-------
-------
APPENDIX C
RARE EARTH OXIDE CATALYST MECHANICAL
PROPERTIES - SPALLING RESISTANCE
The spalling resistance was measured for four catalyst formulations, namely
the copper dichromate catalyst, formula RE(23)75 of lower purity, formula RE(23)75-P
of highest purity, and formula RE(M3)75.
Figure C-1 shows the four catalyst formulas selected for the spalling test
as they appear originally. These four catalyst formulas were then placed for five
days in a closed chamber, as shown in Figure C -2. Inside the chamber a sieve-
like container was put, and the four formulas were placed inside this container. The
container, divided into four chambers, one for each formula, was then put into the
humidity chamber, which at its bottom had about one inch high of pure, distilled
water. The arrangement is shown in Figure C -3. After five days, during which the
average temperature was 740 F and the relative humidity in the chamber 100 percent,
which results in a water vapor content of 0.03 moles, the catalyst formulas were
removed from the chamber and each catalyst, after being weighed, was placed
separately into the catalyst bed of the flow apparatus and heated to about 2000 F in
thl'J first hour and to about 8600 F in the second hour, under a constant flow of high
purity helium of 7600 cc per minute. After completion of the testing, the various
catalysts were removed from the catalyst bed, weighed and photographed.
Table C-I shows the weight of the catalyst before, during and after the
spalling test, whereby the weight during the test is the one obtained after the
catalyst was removed from the humidity chamber and before it was exposed to heat
and flow in the catalyst bed. As can be seen all catalysts gained slightly in weight
due to the adsorption of moisture on the catalyst surface, whereby the copper
dichromate catalyst gained most moisture, and after the spalling test was completed
all catalysts lost weight to even below their original value, indicating that some
moisture was adsorbed to them even before the beginning of the test.
-------
1
. ,
" .
'" t~""'.
.... ",'I.
-.. ......
SOLAR CATALYST FORMULA
RE(M3) 75 .
;
.t
SOLAR CATALYST FORMULA
RE(23) 75-P
- - .... - ~ 1'"-'- ~ -- m
~
o
00
"-s:::: . "
SOLAR CATALYST FOHMULA
HE(23) 75
ST ANDAIlD CAT AL YST
CuCr.)o~
~ I
-------
-- - - --- -
- - - - - -- -
'" ,
--
1>
,
~
FIGURE C-2. COMPONENTS OF THE HUMIDITY CHAMBER
/ . .~
- """!'I!"""t f\
r 'p ~
fIlr~IJDrTY C]L\~1 ~~..~{ fl) . 1
~p..-\LLrNGTLSJ ,
'"
~
~
\:"
FIGURE C-3. CATALYST FORMULAS INSERTED INTO HUMIDITY CHAMBER
-------
TABLE C-I
WEIGHTS OF CATALYSTS BEFORE. DURING AND AFTER SPALLING TEST
Weight Weight Weight
Before Test During Test After Test
Formula g g g
RE(23)75 9.819 9.828 9.806
RE(23)75-P' 9.112 9.280 9.025
RE(M3) 75 10.033 10.199 10.016
CuCr207 10.050 10.232 9.925
Figure C-4 gives a view of all four catalysts after the spalling test. For
the copper dichromate no change in appearance was noted. The results for the other
three catalyst formulations are as follows:
. Solar Catalyst Formula RE(23)75-P: No change in appearance noted.
. Solar Catalyst Formula RE(M3)75: Only minor changes in the surface
noted. Small longitudinal cracks, which were present in the original
pellets before the spalling tests, became slightly enlarged. These
cracks were a function of the preparation of the catalysts and have now
been eliminated by improving the preparation procedures. therefore,
it can be said that the spalling test had no effect on the formula
RE(M3)75 either.
. Solar Catalyst Formula RE(23)75: Small cracks which were present
. I
tn the pellets before the test became noticeably enlarged. This
formula showed the least spalling resistance and can be expected
to fall apart on the repeated humidity-and-heat treatment.
-------
, ."
\\
/ .11/
(.oil:
. r
~
SOLAR CATALYST FORMULA
RE(M3) 75
SOLAR CATALYST FORMULA
RE(23) 75-P
......
......
......
I .
I
I
I
~
-- --- ----
i{
~,
.
.,
SOLAH CATALYST FORl\IULA
HE(23) 75
STANDARD CATALYST
CuCr.)O
- 7
-------
-------
APPENDIX D
RARE EARTH OXIDE CATALYST CHEMICAL PROPERTIES -
LEAD POISONING AND REGENERATIVE STUDY OF
CATALYST FORMULAS RE(23)75-P and CuCr207
The object of this test was to show that the rare earth oxide catalysL once
it is covered by lead deposits, can be regenerated by chemical means, while the
copper dichromate catalyst and others in its class would be lost. The nature of this
study was preliminary, since no catalytic efficiency studies after the restoration
of the pelleL were made. This experiment was not completed in that respect due
to the change of emphasis in the program, since it is now expected that unleaded
fuels will be widely available in 1975, if required by catalytic systems, which will
prevent the lead poisoning of catalytic systems.
For the lead-poisoning and regeneration study described in this appendix, we
selected
10 grams of catalyst RE(23)75-P, which was 25 pellets, and
10 grams of copper dichromate - 38 pellets,
both of which were dessicated in order to remove the moisture adsorbed onto it.
The following figures graphically describe the experimental procedure followed in
this catalyst poisoning and regeneration experiment.
\
(a) Poisoning of the Catalysts
The view of these catalysts as they appear before the test, after the bakeout,
is seen in Figure D-l. The purpose of these photographs is to give a comparison of
the catalyst appearance before and during the lead poisoning test. Figure D-2 shows
the retort in which the lead poisoning is taking place and its components. A stain-
less steel tube with a double bottom is seen in the background and in front of it a
ceramic beaker, containing 10 grams of lead bromide and 12 grams of lead oxide,
to yield the mixture 2PbO -PbBr2. A sieve -like boat for the two catalyst formulas
is placed on top of the ceramic vessel containing the lead oxide and lead bromide,
and the thermocouple is attached to the former to show the inside temperature of
the retort,' The Solar catalyst formula has been placed into the left cell of the boat
-------
, --i.~ 'i
SOLAR CATALYST FORl\lULA
RE(23) 75-P
ST ANDARD CAT AL YST
CuCr 207
:..,
FIGURE D-l. SOLAR CATALYST RE(23)75-P AND STANDARD CATALYST CuCr207
BEFORE LEAD POISONING TEST
Cu<
. -.. ......... .......
~,
~ -:<
\
i\
II.
I,
Z''I>O - '>h!J,;t
-~' -
-:-'j .
' .
";<:IJ.AJlC.\1'.\I.YST t.O[i.\H,L.-\ I' . ", "
It! U;J) 7,.-1' ,
~" ,- ~- """ "
.-. ~ ~m.
~s i\tU)AtH) 1..:""1 ,1" I
C:u{:'"ll'1
FIGURE D-2. THE RETORT AND ITS COMPONENTS FOR THE LEAD POISONING
TEST OF THE CATALYST
-------
and the standard catalyst into the right cell Figure D 3 shows the t 1.
. . . . - opera or, p acmg
the ceramiC b.oat contammg the lead oxide-lead bromide into the retorL and Figure
D-4, the placmg of the catalyst boat on top of the lead salts. The retort is then
placed into the furnace, as pictured in D-5, which is the same furnace being used
for our gas flow apparatus.
"
~
\
I
I
I
I
I
I 1
'J ~
~.
FIGURE D-3. THE BOAT WITH THE 2PbO-PbBr2 IS INSERTED INTO THE RETORT
The retort was heated for 2 hours and the catalyst was then removed,
weighed and returned into the heated fumes for 3 more hours. After that the poison-
ing phase of the test was completed.
Figure D-6 shows the dissembled retort after the test. All components were
found to be coated with lead. The lead salts in the ceramic boat were completely
boiled out. Table D-I will summarize the weights of the two catalysts as originally
selected, as they appeared after the bake-ouL as they appeared after 2 hours of ex-
posure to lead vapor, then after 5 hours exposure, and finally after the washing.
-------
C,,' '
~ ~-'"
,~". -,
~\. .'
';).~. "
t,.'.""',,, ~","'. ,l, ''',',F,',', -'j
. ,,~, l'pd'
,~-1 ,~ <~j'.
I
. I
t
'~
FIGURE D-4. THE CATALYST BOAT IS PLACED ON TOP OF THE BOAT CONTAINING
THE LEAD OXIDE -LEAD BROMIDE. AND THE RETORT IS THEN
COVERED WITH THE CERAMIC BLOCK AT LEFT
-------
II
I":.
,. ~-
FIGURE D-5. OPERATOR PLACES RETORT INTO FURNACE TO MELT THE
LEAD COMPOUNDS
The appearance of the catalysts, after being coated by the lead vapors, is
shown in Figures D-7A and 7B, which are close-ups of the poisoned catalyst formulas.
Of special interest here was the observation that the copper dichromate did not gain
any weight but instead lost some. This is ascribed to be due to possible reactions of
the copper oxides and chrome oxides with the lead vapors. A check of the dimens ions
of the copper dichromate pellets showed that there was no change from those prior to
the test in spite of the loss of weight.
(b) Restoration of the Catalysts
After the poisoning test was completed, both catalysts were placed in test
tubes to which diluted nitric acid was added. This is seen in Figure D-8. Note the
immediate discloration of the washing liquid containing the standard copper dichro-
mate catalyst. Figure D-9 shows that the catalysts in their nitric acid baths are
heated, though not over the flame as indicated in the picture, but in boiling water
baths. During this procedure the liquid containing the copper dichromate turned
green, while the liquid containing the rare earth oxide catalyst remained clear.
-------
~
\
\
(/
~ l'
" .
,
?
,,'
.1
f-'
I:\:)
o
\~
~."'!
"\
,
~-'-..
# ~
.~
k.'~~~"'-'i1'11iii!ii"_rifI
"
'"
".r..;.!.",~--...-=-,.
~T .\XD\!tD C.\.T.\LYST
C\l.Cr"o-
SOL.\H C.\T\LYST f"fJB\ItJ'L\
FIGURE D-6. RETORT AND ITS CONTENTS AFTER THE LEAD POISONING TEST. NOTE
-------
TABLE D-I
CATALYST WEIGHTS DURING POISONING AND
REGENERATION EXPERIMENTS, IN GRAMS
RE(23)75-P CuCr207
Date Condition net weight net weight
11-20 Original weight 10.0049 10.1144
11-23 Weight after dessication, start of expo 10.0016 10.0996
11-23 Weight after 2 hours exposure to Pb 9.9846 9.6651
11-24 Weight after 5 hours exposure to Pb 9.9942 9.5963
11-25 Final weight, after restoration 9.5904 None
Remarks 4% loss 100% loss
-- ..........-
-------
;..
*' ~ti! ,.; ~---<'"
~. .~ ...;;.~/ t ,:~;
..~...~.~n /.~,
~,;:\ .. ~..' '''';W'
~"'t'" '~ " \ .'" . "..<~
"""~< NITRIC ACID ~-.:;f
~~,~ ,- 7'
-~ ~ "tllG.t~ ".C$. ~ ceor 11 ~'
.~i;" ...oJ.
-'"'-'.. '~lAtoRAT08IYOtt~T~ ~1'<
~ "+ JotO'fOR~1J&f ~,~
"..
~. -.. -::~"~... -'"
..' -~...~~'
;~ ~
~
.
~,
r...
,.
~
~,
'"'
~.
~(
'.'
~
fJ!
I ,~
J;'~'~ ~
8'. " .' ::",:,mmHoo.mc,
, RE(23) -;j-P
T..."'."",_"'.,\). ~-
~.........;
~'
FIGURE D-8.
BOTH CAT ALYSTS ARE IMMERSED IN DILUTE NITRIC ACID
(DILUTION 1: 1) TO WASH OFF POISONING LEAD DEPOSITS.
NOTE DISCOLORATION IN LIQUID OF CuCr207 BATH
After 75 minutes boiling time the copper dichromate-containing test tube showed
advanced dissolution of the pellets. The liquid of the test tube containing catalyst
formula RE(23)75-P had turned slightly yellow. The yellow liquid from the REO
catalyst was then recovered and fresh dilute nitric acid substituted. The copper
dichromate containing test tube was left unchanged and the boiling continued. The
liquid recovered from the REO catalyst was tested for contents of lead by the
following method:
To the washing liquid dilute sulfuric acid is added. The precipitation of
white crystals indicate the presence of lead sulfate. In this case the washing of the
pellets at 2500 F in diluted nitric acid was continued, until no precipitate in the new
washing liquid was found. The result of this procedure is shown in Figure D-10.
Here we have the REO catalyst formula, which has been washed in two nitric acid
baths and then rinsed in distilled water, dried and placed on a Petrie dish. The
lead sulfate precipitate from the first washing water is shown in the test tube appear-
ing behind the rare earth oxide dish. On the other hand, the copper dichromate
completely dissolved in its nitric acid bath, and its test tube is shown along with the
empty Petrie dish for the standard catalyst. This result then shows that standard
commercial catalysts and those in their class cannot be recovered once they have
been poisoned by lead oxides and sulfur oxides, while the REO based catalysts can
be regenerated.
-------
I.
II
I
--~-- - -
J
if<
WI
If(-
r
.o.'"i",»'
-
\..
..l
i
:-:-
f
i:..
FIGURE D-9. SYMBOLIC REPRESENTATION OF HEATING STEP
FOR WASHING OF POISONED C:\TALYSTS
-------
.
.
!' <-,
- -"..,
.r ~ . ' -;;..,¥
'~'" "
,~-,;- .. ~:; ,
L
. .
t..... oiiililloli!
SOLAIl C.\T/\LYST foRMUL.\
~~ ]lE!~3} 1!>-P
roPI'f.H ~n'I'UT'"
.\ND
«:11110"": Nrr'LATL
ST,uWARD <':AT,\U'S'f
CUCf:!°7
e.- ~.... II.
-~",'
I
f~.~j ffn. ;I"r,
-
~--
",---
~/
-
FIGURE D-10.
RESULTS OF RESTORATION OF POISONED CATALYSTS. LEAD
DEPOSIT FROM RE(~3)75-P WASHED RIGHT OFF, FORMING A
WHITE DEPOSIT WITH ADDITION OF DILUTE SULPHURIC ACID.
CuCr207 PELLETS DISSOLVED IN WASHING LIQUID
Figure D-ll is a close-up view of the REO catalyst after the lead poisoning
and restoration test. All 25 pellets are restored. As seen from Table D-I there
was a four percent loss in weight in this catalyst after the described procedure.
The conclusion of the experiments shows that copper dichromate cannot be
regenerated by this procedure from lead poisoning and thus is lost. The high purity
REO catalyst, RE(~3)75-P, can be regenerated from lead poisoning and can be
reused.
-------
r-SO:LAR-CA T ALYST: F-OR-MULA -,
RE(23) 75-P ,
}~',,~,~" ',' "",'" ',., ",~: ~,::".,',', "'C", :'" ...... ~ ,",.... ,'," ""-', ','>','~, -...,' "'-.,,, '."~"',,",::,':'.~'~, .,~.,',.'~: ~""'-"~'.'" ".,' ,,'~i. '.-,,"::~',:'~:.'::~,:,',:.'.,
:)-~
~
- -. -
I
~
I
" I
j
!J
-
FIGURE D-11. CLOSE-UP OF RESTORED SOLAR CATALYST
FORMULA RE(23)75-P AT CONCLUSION OF TEST
-------
-------
APPENDIX E
THE ECONOMICS OF THE REO CATALYSTS
Availability of Rare Earth Oxides
In any applications contemplating the use of rare earths because of their
unique properties, the first thing one should do is dispel the notion that these elements
are rare, when as a matter of fact, this group of elements, including scandium and
yttrium, comprise about 25 percent by weight of the metallic elements in the periodic
table. (Ref. 23). Cerium is more abundant than tin, and yttrium and neodymium are
more abundant than c.obalt. Even the scarcest of the rare earths are more abundant
than the more commonly known metals such as silver, gold, and platinum.
Of the various sources of rare earths, as eU4enite, xenotine, monozite, and
bastnasite, the bastnasite source is the most significant. The United States is for-
tunate in this respect as it possesses the world's largest bastnasite deposit at
Mountain Pass, California, which is located in the Sierra Nevada Mountains near
Las Vegas, Nevada. This particular orebody supplied 60 percent of the 1969 world
shipments of the "light" rare earth oxides, which include oxides of lanthanum,
cerium, neodymium, praseodymium, samarium, and europium. Present production
of rare earth oxides from this particular source is 25,000 tons annually, with the
capability of increasing this to a 50,000 ton annual capacity. It is reported that proven
reserves for the Mountain Pass deposit are five billion pounds of REO (rare earth
oxides), with indicated reserves of several times that amount (Ref. 24). Thus it
appears that an ample supply of "light" rare earth oxides is assured for some time
to come.
Analysis of Requirements and Cost of REO for Automotive Catalysts
Two forms of rare earth oxide catalysts for use in automotive exhaust
I
systems are contemplated:
. Pressed and sintered pellets of varying porosity
. Relatively thin coatings of the oxides upon some sort of
support medium exhibiting a large active surface area.
-------
Because of convenience, preliminary studies at Solar have made use of the
pellet form. But of the two, for reasons of economy and effectiveness, the supported
catalyst would appear to have the greatest potential in the long run.
The following cost analysis is based upon the latest and best price infor-
mation available to date, and upon the assumption that, automobile sales will average
about 10 million annually. Historically, the price of rare earths has decreased
approximately 10 percent per year over the past several years. The figures obtained
for this study have taken this factor into account and were projected into the near
future, using the best facts and judgment available to the industry. (Ref. 25).
Another fact which must be kept in mind in any cost analysis is the saleability
of the byproducts when considering the use of a single oxide, or say a mixture of just
two oxides. For example, the distribution of the rare earths in bastnasite is given
below. If we were only interested in praseodymium oxide, it would mean the production
of 25 times that amount of cerium. lanthanum, and neodymium oxides as byproducts.
Unless these were saleable, the entire '(ost would have to be borne by the praseodymium
oxide. This, of course, is another incentive for the use of the mixed oxides, if at all
possible. Preliminary investigations at Solar indicated that the proper blend of oxides
may indeed provide a satisfactory automotive catalyst for NO reduction.
Rare Earth
Percent in Bastnasite
Cerium
Lanthanum
50
34
Neodymium
Praseodymium
11
4
Samarium
0.5
0.2
Gadolinium
Europium
Others
0.1
0.2
Pelletized REO Catalysts
Assuming a production of 107 autQmobiles per year, a requirement of 15
pounds of oxide catalyst per unit, and use of the maximum amount of anyone of the
oxide constituents. the question of supply and availability would appear as follows:
-------
REO (Mixed RE Oxides) - Proven reserves of bastnasite deposits in
the U.S. indicate 5 x 109lb of REO or, on the basis of 1.5 x 10S
lb/yr, a supply for at least 33 years.
Cerium Oxide (Ce02) - Bastnasite contains 50 percent, which trans-
lates to about 2.5 x 109 lb total. This amounts to at least a 20 year
supply.
Lanthanum - Neodymium - Praseodymium Oxides ("Didymium")
Since this mixture comprises 47 percent of bastnasite, the supply
situation will be roughly the same as the cerium oxide above.
. Praseodymium Oxide (PrOx) - At four percent of bastnasite, the
requirements of 4 x 107 lb/yr would mean a supply of about five
years.
Figure E -1 summarizes the results of a cost analysis based upon the presently
proposed Solar formulations. Depending upon the purity of the component oxides or
mixtures of the RE oxides, the cost of the catalytic agent would vary from a high of
$4. OS/lb for RE(23)75 of 99 percent purity to a low of $0.45/lb for RE(M) of a 90 per-
cent purity level. It would be well to remember at this point, that the cost of the higher
priced formulations could no doubt be lowered substantially by optimizing and refining
the compositions of the present catalytic agents.
Q
Z
;:J
:>
i:l.
......
~
-r.
... :3. 00
...
o
Q
2S
f-o
~
U ~.oo
5.00
.
. 99 PERCENT
095 PERCENT
t:>. !IO PERCENT
1. 00
HE(23)75
HE(M3)75 HE(~O)
SULAR FOHMULA
RE(M)
FIGURE E-l. COST OF CATALYTIC AGENTS OF VARYING PURITY
-------
.
REO Coatings on Support Medium
Estimated requirements for a supported catalyst call for approximately one
pound of rare earth oxides per automobile, or 107 lb annually. Although the required
supply is roughly an order of magnitude smaller than what is needed for the pelletized
form of catalyst, the cost of the raw materials is not expected to be any different.
Using the same approach as before, the supply picture would appear as given below.
. REO (Mixed RE Oxides) - Bastnasite reserves of 5 x 109 lb of REO
and a need for 107 lb annually would translate to a supply of the
mixed oxides that would last at least 500 years.
Cerium Oxide (Ce02) - At 50 percent of 5 x 109 lb, a supply would
be assured for a minimum of 200 years.
Lanthal).um - Neodymium - Praseodymium Oxides ("Didymium") -
As in the case of cerium oxide, enough of the material should be
available for approximately 200 years use.
. Praseodymium Oxide (PrOx) - At the lower level of use, we could
count on at least 50 years supply.
In summary, it is apparent that the supply of rare earth oxides is more
than adequate to serve the needs of an automotive catalyst based on 107 new cars
per year for some time 'to come, without any consideration at the present time for
reactivating used catalytic material.
Life Cycle Aspects of REO Catalysts
The foregoing discussions were based on the assumption that there would be
no need to reuse the catalytic material, because the supply appears adequate at this
time. In reality, however, the national concern for environmental pollution, including
the generation of solid wastes, would dictate that serious consideration be given to
the recycling of the catalytic agents.
It is generally known that lead and lead compounds tend to poison catalysts
and thus render them useless. Furthermore, lead and its compounds are extremely
corrosive to most metals and their oxides, particularly at elevated temperatures.
Studies at Solar with the copper dichromate and REO pellets in a molten
2PbO-PbBr2 mixture indicate a drastic deterioration of the copper dichromate
catalyst, whereas the REO pellets appeared to be fully resistant to this highly
corrosive environment.
-------
Another aspect which is also worthy of consideration is that of conservation
of relatively scarce and critical materials. (Ref. 26). Copper dichromate would
certainly fit into this category, as copper occurs to the extent of only 64 ppm average
in the earth's crust and the chromium supply is located in areas which are inaccessible
J
to the U. S. for political reasons. Although it is true that the relative average abun-
dance of an element such as cerium (46 ppm) is less than that of copper, the reactivation
aspect mentioned above could minimize this apparent shortcoming. In addition, there
are so many other applications requiring the use of copper and chromium th:;1t the demand
could conceivably exceed the supply of these materials.
The Comparative Economics of the Rare Earth Oxide Catalysts With Common Noble
Metal Catalysts
In our study on the cost per pound of catalysts, we are comparing the various
catalyst materials considered earlier with platinum metal catalysts, both in pure and
supported and in pellet forms, and with the commercial copper dichromate catalysts.
Figure E -2 gives a schematic view of the costs of the various catalysts starting from
the most expensive one, osmium powder, down to the least expensive one. The cost
ranges between $3,300 for the most expensive pure noble powder per pound to $1. 20
per pound for copper dichromate or pure supported ceri~m oxide catalyst. Figure
E-2 divides the catalysts into several groups. Group 1 is the pure noble metal powders,
which are of course the most expensive, ranging in cost from $3,300 to $540 per pound. *
Next follows the noble metal powders supported on carbon. These range from $230 per
pound to $68 per pound. * Then follow the supported noble metal catalyst pellets which
range between $135 and $36 per pound. Next down in cost is the category of the high
purity rare earth oxide catalyst pellets, whereby the high purity is 96 or 99 percent
rare earth oxide. Here the high goes from $30 per pound for praseodymium oxide down
to a low of $6 per pOund for cerium oxide. (Ref. 25). In between these two are cost-
wise the didymium oxides and the pure Solaform 8 and the pure Solaform 10 formulas.
The commercial mixed oxides, unenriched with anything, are comparable in price with
the commercially available copper dichromate: The rare earth oxides as such, cost
$2.40 per pound and the didymium oxide $1. 85 (Ref. 25); the copper dichromate in
comparison is $1. 20. If we support our catalyst formulas on alumina pellets, which
come to about 60 cents per pound, using the approximation that for a support ed catalyst,
only 10 percent of the quantity of the unsup~rted catalyst is needed, the cost for the
pure praseodymium catalyst supported is $4.60, while the pure cerium oxide supported
catalyst would be equal to the copper dichromate cost of $1. 20. The range of all these,
catalysts in cost per pound is shown on Figure E -2~ which also shows the acceptable
upper limit for the catalyst material cost of $8. 30 and the mean catalyst cost per pound
* Phone conversation with Mr. T. Romeo of Engelhard Co., 6001 BandinL Los Angeles,
California, week of September 20, 1970.
-------
10.000
~ 000
6.000
,>.000
4,000
3,000
2,000
1,500
1,000
800
i 600
! 500
.. 400
~ 300
€ 200
I-
i-' ~ 150
~ ...:I
*", ..c
I-
~ 100
~ 80
~ 60
::> 50
~ 40
<=
'" 30
'"
I-
~ 20
U
15
10
8
!:j
~
'"
::II
1. e
:II!
~
1-1
<=
<= '"
'" ~
~ 0
It I ~
:II! --+- e
e 8
8 ==
1'5 <=
.
0;
'" 0 ~
7J t:. 4! c:!
. j .. - '" ~ "5~
~ c 4!..!: '" ~~
-8.8 ... ft "'~
"3 ~ ~ ~. :.'" i'J 0 -
o ~ U .. fS
~ '" CJ "'" - '" ;;
o . "iJ=::: 7J '"
. 38 ~~ £j' 'Pet - s 11
:0 ~'" l~ iCJ i:..
:t~ . t II Jl
t~ ~~ Ii ~~ E.., 15
~/ - J;! 0 .. 0 "
~" "'CJ "'0 QQ QifS "11:
.,
-*---~--*-_. -*'~-~ ~--_. x *-- x
<=
'"
~
o
'"
:E
~
I-
:S
p,
f:j ~
~~-~
p, '"
:E :E
e e
~ j
.... ...:I
=> ..:
a: p,
<=
'"
~
o
p,
~
o
"
<=
'"
~
p,
:E
::>
15
....
:S
p,
~
f:i
~
p,
:E
e
j
...:I
..:
p,
~
z
~
~
u
Z
o
p;
~
a:
'"
~
'"
::Ii
::>
1:1
:S
...:I
..:
'"
't;
a:
'"
~
o
'"
::Ii
::>
15
....
:S
p,
~
FIGURE E-2.
""
'"
8
...
...:I
::>
'"
0:
~
:Ii
-3--
<
z
o
t:i
'"
8
...
...:I
::>
'"
"'
~
o
::Ii
e
::Ii
>-
8 '"
"" t-
~ll IT
'" '"
"" a:
'" '"
a: a:
=> ::>
p, p,
"" If'
'" '"
'" '"
<=
~
o
'"
z
~
<=
..:
u
~
"
p,
~
g
::Ii
::>
...:I
<
Z
o
"
p,
~
a
< ..:
~ ~
::> ::>
...:I ...:I
..: <
~t' ~
p, '"
~ ~
...
0'"
"
t.
'"
a:
::>
p,
"
'"
't;
a
~
a
oR
~
a
a
o
CATALYST FORMULAS
ECONOMICS OF CATALYSTS
ACCEPTABLE UPPER LIMIT FOR CATALYST
_~n.~c\.._~-----
'"
t-
M
::Ii
r;r
<=
'"
a:
::>
p,
~
ACCEPTABLE CATALYST MATERIAL C06T ,
(REF. 1) / ~ ~ ..
/ - ::! '" !:!. M II:
-~ ~ ~ ::J_~-~-
u ... '" a:
p,
'1,';
'"
o
'"
-------
of $2. 00 (obtained from Ref. 6). If a catalyst is more expensive than this upper
limit, it would be economical to support it on alumina pellets or on carbon.
Catalysts, which by themselves cost less, could be used in their pure state. It
has to be considered here, however, that with the decrease in the purity of the rare
earth oxides, the price of the rare earth oxides will decrease. The prices for the
rare earth oxides are based on 100 pound quotations. For quantities in the amounts
of tons, the prices might decrease by 50-70 percent. Therefore, it appears rea-
sonable that Solar Formulas 8 and 10 both could be economically used in large
quantities and lower purity in the unsupported state as well as the supported state.
-------
-------
APPENDIX F
AMERICAN OIL COMPANY EVALUATION OF SOLAR NOx CATALYST
BY PULSE -FLAME METHOD
d~
AMERICAN
OIL
COMPANY
RESEARCH AND DEVELOPMENT DEPARTMENT 2500 NEW YORK AVENUE
WHITING. INDIANA 46394
H. R. TALIAFERRO
Coordinator-A WOn1Mt \'('
Emissions Research
July 1, 1971
AIR MAIL--SPECIAL DELIVERY
Mr. W. A. Compton
Assistant Director--Research
Solar
2200 Pacific Highway
San Diego, California 92112
Dear Sir:
Re:
Evaluation of Solar NOx Catalysts
Your Solar catalysts were evaluated using the pulse-flame apparatus
developed by Dr. Meguerian and described in some detail in the at-
tached paper.* Isooctane was used as the fuel. The results of our
evaluation are listed in the attached tables.
We wish to make the following observations based on these data and
visual examination of the catalysts:
(1)
Both catalysts are mechanically very weak.
us, considerable pellet break-up occurred.
graph shows the condition of the catalysts
During shipment to
The attached photo-
as received.
* Reference 5
-------
(2)
At 6500e and soooe, significant NOx reduction was obtained at
a eo concentration of about 1% (Tables I and II). However,
higher eo levels were required to obtain a NOx reduction level
of 90% or more. Because this level was attained at a lower co
concentration with catalyst No.2 than with No. l;*No. 2 is the
better catalyst. At 550oe, both catalysts were not active ex-
cept at very high eo concentrations.
(3 )
A substantial fraction of the NOx was reduced to NH3 over both
catalysts. Although less NH3 formed over catalyst No.1 than
over No.2, NH3 is unacceptably high in both cases. At soooe,
we have observed 10% or less conversion to NH3 with some catalysts.
We are returning both the used and unused portions of your catalysts
under separate cover.
If you have any questions with regard to our data, please call me.
Very truly yours,
/-/ /-) - /~L ( ~ .j
IA.. t ,I. L (. ~
H. R. TALIAFERRO
Attachments
** Catalyst No.1 is Catalyst RE(M3)75, here denoted also M3
Catalyst No.2 is Catalyst RE(23)75-P, here denoted also 75-P
-------
TABLE F-I
CATALYST EFFECTIVENESS
Solar No.1 M-3
Space Velocity = 33,000 hr.-l
NOx' ppm % NOx
% 02 % CO Inlet Outlet Reduction
5500C
1.04 0.70 560 550 2
0.99 0.82 560 515 8
0.64 1.50 560 540 4
0.57 1.98 550 480 13
0.54 2.86 520 190 64
6500C
0.74 1.11 540 535 1
0.65 1.23 585 345 41
0.58 1.55 570 85 85
0.57 2.01 560 125 78
0.55 2.78 535 85 84
8000e
0.88 0.65 630 560 11
0.97 0.78 560 545 3
0.85 0.91 580 500 14
0.68 1.12 630 260 59
0.76 1.18 555 245 56
0.64 1.32 520 185 64
0.61 1.38 675 150 78
0.70 1.40 550 185 66
0.60 1.83 540 5 99
~: Space velocity is volume of gas per
volume of catalyst per hour.
FWR/ ac
7/1/71
-------
TABLE F-II
CATALYST EFFECTIVENESS
Solar No.2 75-P
Space Velocity = 26,000 hr.~l
NOx, ppm % NOx
% 02 % CO Inlet Outlet Reduction
-
550°C
0.98 0.59 670 575 14
0.77 0.89 595 560 6
0.73 1.01 605 560 7
0.73 1. 28 550 475 14
0.64 1.36 590 455 23
0.57 1.80 520 375 28
0.67 2.32 650 170 74
0.54 2.60 610 225 63
0.58 3.67 565 190 66
650°C
1.04 0.57 610 600 1
0.75 1.11 530 240 55
0.74 1.13 580 330 43
0.61 1.61 610 90 85
0.58 2.24 620 45 93
0.57 2.30 560 15 97
800°C
0.90 0.59 -
615 580 6
0.83 0.89 550 270 51
0.76 1.02 550 220 60
0.69 1.27 600 125 79
0.66 1.43 560 60 89
0.63 1.50 510 15 97
0.61 1.69 600 15 98
FWR/ ac
7/1/71
-------
TABLE F-III
NH3 PRODUCTION OVER SOLAR CATALYSTS
Space Velocity = 27,000 hr.-1
Inlet NOx = 1,350 ppm
Temp. , % NOx Conversion
Catalyst °C % 02 % CO Overall To NH3
No. 1 - M3 550 0.33 2.12 8 4
650 0.32 2.39 52 37
800 0.36 1.81 62 17
800 0.33 1.63 80 16
No.2 - 75P 550 0.42 1.72 7 5
650 0.32 1.72 44 29
800 0.42 1.85 83 63
800 0.37 1.72 86 64
800 0.41 1.72 82 58
FWR/ ac
7/1/71
------- |