CONSULTANT REPORT
of the
COMMITTEE ON MOTOR VEHICLE EMISSIONS
COMMISSION ON SOCIOTECHNICAL SYSTEMS
NATIONAL RESEARCH COUNCIL, ON AN
EVALUATION OF CATALYTIC CONVERTERS
FOR CONTROL OF
AUTOMOBILE EXHAUST POLLUTANTS
SEPTEMBER 1974
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Washington, D.C. 20460
CE
-------
CONSULTANT REPORT
to the
Committee on Motor Vehicle Emissions
Commission on Sociotechnical Systems
National Research Council
on
AN EVALUATION OF CATALYTIC CONVERTERS FOR CONTROL OF
AUTOMOBILE EXHAUST POLLUTANTS
PREPARED BY:
Joe W. Hightower, Chairman
John B. Butt
David F. Ollis
Henry Wise
Washington, D.C.
September 1974
-------
NOTICE
This consultant report was prepared by a Panel of Consultants at
the request of the Committee on Motor Vehicle Emissions of the National
Academy of Sciences. Any opinions or conclusions in this consultant re-
port are those of the Panel members and do not necessarily reflect those
of the Committee or of the National Academy of Sciences.
This consultant report has not gone through the Academy review
procedure. It has been reviewed by the Committee on Motor Vehicle Emis-
sions only for its suitability as a partial basis for the report by the
Committee.
The findings of the Committee on Motor Vehicle Emissions, based
in part upon material in this consultant report but not solely dependent
upon it, are found only in the Report by the Committee on Motor Vehicle
Emissions of November 1974.
ii
-------
PREFACE
The National Academy of Sciences, through its Committee on Motor
Vehicle Emissions (CMVE), initiated a study of automobile emissions-
control technologies at the request of the United States Congress and
the Environmental Protection Agency (EPA) in October 1973. To help carry
out its work, the CMVE engaged panels of consultants to collect informa-
tion and to prepare consultant reports on various facets of motor vehicle
emissions control. This Consultant Report on An Evaluation of Catalytic
Converters for Control of Automobile Exhaust Pollutants is one of five
such consultant reports prepared and submitted to the Committee in con-
nection with the Report by the Committee on Motor Vehicle Emissions of
November 1974. The other consultant reports are:
Emissions and Fuel Economy Test Methods and
Procedures, September 1974
Emissions Control of Engine Systems, September 1974
Field Performance of Emissions-Controlled
Automobiles, November 1974
Manufacturability and Costs of Proposed Low-
Emissions Automotive Engine Systems, November 1974
These five consultant reports are NOT reports of the National Academy
of Sciences or its Committee on Motor Vehicle Emissions. They have been
developed for the purpose of providing a partial basis for the report
by the Committee as described more fully in the cover NOTICE.
iii
-------
CONTENTS
I. Conclusions and Recommendations 1
A. Conclusions 1
1. Hydrocarbon and carbon monoxide oxidation
2. Nitrogen oxide reduction
3. Three-way catalyst systems
4. Catalyst poisoning
5. Abnormal driving modes
6. Materials availability and toxicology
B. Recommendations 4
1. Poisoning by lead-motor mix components
2. Environmental considerations
II. Introduction 6
III. Basic Performance Characteristics of Auto Catalysts 8
A. Oxidation Catalysts 8
B. NO Reduction Catalysts 10
x
General considerations
Catalyst details
Laboratory performance studies
C. Three-Way Catalytic Systems 35
D. Discussion 36
IV. Vehicle Performance and Durability of Auto Catalysts 38
A. Deactivation of Oxidation Catalysts 38
B. Deactivation of NO Catalysts 40
X
C. Deactivation of Three-Way Catalysts 42
D. Poisoning of Oxidation Catalysts 43
Sulfur
Phosphorus
Lead (and halide scavengers)
iv
-------
E. Poisoning of NO Catalysts 52
x
F. Long-Term Durability 55
Oxidation catalysts
NO catalysts (dual-bed systems)
X
Three-way catalysts
Overall durability of HC/CO/NO systems
X
V. Materials Supply and Toxicology 69
A. Materials Supply 69
Noble metals
Base metals
Supports
B. Toxicology 78
Debris from catalysts
Effects on fuel components (sulfates, etc.)
Animal tests
VI. Kinetics, Mechanisms, and Modeling 94
A. Introduction 94
B. CO Oxidation . 94
C. HC Oxidation 96
D. NO Reduction 96
x
E. NO Decomposition 100
F. Exhaust Converter Modeling 101
References 105
Appendixes
A. Companies Site-Visited or Interviewed 112
1. Domestic
2. Nondomestic
B. General Questions to Companies from the Members of
the Panel of Consultants on Catalysts 114
v
-------
TABLES
1. Effect of Catalyst on Exhaust Gas Composition H
2. Composition of Monolithic Catalyst Supports ........ 14
3a. Performance of Supported Noble-Metal N0X Catalysts 17
3b. Performance of Supported Noble-Metal N0X Catalysts 17
4a. Performance of Supported Mixed Ru/Pd N0X Catalysts 19
4b. Effect of Redox Cycling on Performance of Stabilized
N0X Catalyst 23
5. Volatilities of RuO^-RuO^ 20
6. Comparison of Ruthenium Loss in a Ru Catalyst and in Ba-
or La-Oxide-stabilized Ru Catalysts 21
7. Performance of a Presynthesized BaRuO^ Monolithic Catalyst . 22
8. Comparison of Activity and Selectivity between Stabilized
and Nonstabilized Ru Catalysts - System N0-C0-H2 22
9. Performance Data for Base-Metal and Promoted Base-Metal
N0X Catalysts 27
10. N0X Catalyst Performance Summary (Qualitative) 37
11. Abnormal Driving Modes - Effect of Misfiring Cylinder ... 42
12. Total Amounts of Primary Poisons Passed Through
Catalytic Converter in 50,000 Miles 45
13. Comparison of Exxon and Chrysler Lead Poisoning Data .... 47
14. Selectivity Characteristics of Three-Way Catalysts 57
15. Summary of NO Catalyst Durability for Ford Vehicles .... 61
16. Supply/Demand Picture for Platinum and Palladium 71
17. Estimated Composition of Noble Metals from Different Sources 73
18. Growth Potential for South African Platinum and Palladium . 74
19. Estimated U.S. Consumption of Selected Base Metals, 1971 . . 76
vi
-------
20. Toxicity of Some Compounds That Might Be Used in
Automobile Emission-Control Catalysts 82
21; Modeling Studies of Sulfate Concentrations ... 90
22a. Parameters for NO Reduction to N- by H. on a Cu-Zn-Cr.0„
Catalyst . . . ... . 98
22b. Parameters for NO Reduction to N. by CO en Silica"
Supported Copper Oxide . 100
23. Summary of Catalytic NO Decomposition Kinetics 102
vii
-------
FIGURES
1. Comparison of Thermal Reactors and Catalytic
Converters for Oxidation of HC and CO 9
2. Relative Activity of Platinum Group Metals for
NO Conversion 16
3. Comparison of Ammonia-Forming Tendencies and NO Reduction of
Supported Pt and Ru N0„ Catalysts 18
iV
4. Selectivity for ^ of a Ni/Pt N0X Catalyst versus
C0/02 Ratios 25
5. Dependence of N£ Selectivity on CO/O2 Ratio 26
6. Dependence of NO Reduction and NH3 Formation on Residence
Time Evaluated for GEM Catalyst at 1200°F and 10 hr"1 ... 28
7. Dependence of NO Reduction Rate on Inlet 02 Concentration . 29
8. Comparison of Ammonia Formation on GEM-67 and GEM-68
as a Function of 0^ Concentration 31
9. Performance of GEM-68 "Getter" System 32
10. Relative Reactivity of Catalysts in Questor's Reverter ... 34
11. N0X Conversion Window for Three-Way Catalyst 35a
12. Effect of Bromide Scavenger on HC Activity
in Lead-Free System 49a
13. Effect of TEL and Chloride on HC Activity 50
14. Oxidation Catalyst Poisoning Test - Effect of Fuel Con-
taining TetraethylLead and Ethylene Dichloride 51
15. Lead Deposits at Entrance of Catalyst Over Lead Content
in Fuel as a Function of Gas Temperature 53
16. Lead and Sulfur Distribution in Lead-Poisoned Catalyst
Pellet 54
17. Effect of Pb and S on Stabilized, Supported Ru Catalysts for
Removal of NO from a Synthetic Exhaust in an Oxidizing
and Reducing Condition 56
18. Durability Test Results of Some N0X Catalysts for a Small
Vehicle (Datsun P610, 2750 lb) 59
viii
-------
19. Durability Data on Ford's N0X Catalysts 62
20. Durability Data on General Motors' N0X Catalysts 63
21. Effect of NOx Control in Advanced Emissions Systems
on CO Concentration 64
22. Durability Test Results of C>2 Sensor 65
23. Evaluation Test Results of Mixed Noble-Base Metal Three-Way
Catalyst on Monolith, Steady State 67
24. Evaluation Test Results of Noble Metal Three-Way Catalyst
on Monolith, Steady State 68
25. Equilibrium Conversion of SC^ to SOg at 1 atm Pressure . . 85
26. Observed Conversion of SO2 to SO^ Compared with Thermo-
dynamic Equilibrium Curve for SO^ to SO^ 87
27. Effect of Fuel Sulfur Content on Sulfate Emissions,
1972 Federal Test Procedure 88
ix
-------
I. CONCLUSIONS AND RECOMMENDATIONS
A. CONCLUSIONS
The members of the Panel of Consultants found data concerning auto-
mobile emission-control catalysts presented by various companies to be in
general agreement. Some of the major conclusions are listed below; each is
discussed in further detail in the body of this report.
1. Hydrocarbon and carbon monoxide oxidation
a. For HC and CO oxidation, noble-metal catalysts (containing
Pt, Pd, and/or Rh supported on oxide pellets or monoliths) have been
demonstrated to meet the ultimate standards (0.4 and 3.4 g/mi, respec-
tively) for 50,000 miles when run on EPA durability fuels.
b. No base-metal catalysts have demonstrated 50,000-mile
durability, although some that have been promoted with as little as
0.01 wt7o noble metal have exceeded 10,000 miles within standards.
c. Noble-metal oxidation catalysts increase the percent sul-
fur emitted as SO^ from a baseline of less than 10% to a range of
15-60%, depending on the temperature and oxygen partial pressure;
addition of an air pump increases S0g/S02 exhaust ratios. The cata-
lysts (especially in the pelleted form) have a capacity to retain
relatively large quantities of sulfur oxides at low temperatures, but
this adsorbed material is then "dumped" during high-temperature opera-
tion.
d. Since oxidation catalysts require lower operating tempera-
tures than do thermal reactors, their use effectively "de-couples"
engine performance and emission control by allowing the engine to be
optimally tuned to give lower exhaust temperatures. When used opti-
mally, oxidation catalysts could result in fuel-economy benefits as
high as 20% over comparable 1974 model cars.
2. Nitrogen oxide reduction
a. N0X control to the interim standards (e.g., 2.0 g/mi in
California) can be achieved by engine adjustments alone without using
catalysts.
1
-------
2
b. Catalysts to meet the ultimate standards (0.4 g/mi for
50,000 miles), whether in dual-bed, triple-bed, or three-way systems,
have not been demonstrated. Furthermore, due to uncertainties about
the future N0X standards, research is not being aggressively pursued
by most manufacturers. Based on data available at this time, pro-
spects for development of a suitable catalyst in the near future that
will meet the 0.4 g/mi standard are not bright, although use of "high-
temperature" base-metal catalysts on wire-mesh supports protected by
an upstream oxygen scrubber show some promise i_f durability can be
demonstrated.
c. Potential dual- (or triple-) bed low-temperature catalytic
systems for N0x control inherently involve fuel economy penalties, due
to the requirement of fuel-rich operation, and necessarily place a
heavier burden on the oxidation catalysts. Ammonia formation remains
a problem for low- and intermediate-temperature NOx catalysts. Among
the low-temperature candidates only Ru catalysts are low-ammonia form-
ers, but to date no satisfactory method has been demonstrated for sta-
bilizing this noble metal against mass loss in an oxidizing atmosphere.
d. No catalyst has been developed that will decompose NO into
its elements at moderate temperatures, even though such is thermodynam-
ically possible. Furthermore, strong inhibition by oxygen makes such
a development highly unlikely.
3. Three-way catalyst systems
It is possible to control all three pollutants simultaneously
in a single converter. However, effective operation requires main-
taining the fuel mixture to within ±0.1 air/fuel ratio units, a pre-
cision that has not been demonstrated with simple carburetion.
Neither the catalyst nor the required oxygen sensor-feedback control
system has demonstrated the necessary durability to meet the stan-
dards. The complexity and component reliability requirements for
this system are quite stringent, and much more research will be neces-
sary in order to make the three-way catalyst a viable approach to
emission control.
-------
3
4. Catalyst poisoning
The preponderance of evidence indicates that both NO* and
oxidation catalysts are unambiguously deactivated by the presence of
compounds containing lead, bromine, and phosphorus at levels above
those of EPA durability fuel and oil phosphorus specifications. Lead
and phosphorus are "permanent" poisons, while bromine is a "temporary"
poison. Synergistic effects between the various poisons may also be
operative. With certain systems there are reports that lead alone
does not poison the oxidation activity, but the members of the Panel of
Consultants have been unable to resolve satisfactorily this apparent
conflict on the basis of available data.
5. Abnormal driving modes
Early prototype catalytic converters reportedly exhibited a
high incidence of catastrophic failure (burnout, cracking, attrition,
etc.) when exposed to abnormal conditions due to ignition failure,
vibration, accidental exposure to leaded gasoline, etc. Improvement
in ignition systems, better catalyst packaging, and methods to avoid
contamination with leaded fuels have greatly reduced these potential
failure modes. Nevertheless, total ignition failure at high speeds
(e.g., above 50 mph) almost certainly will result in temperatures
sufficiently high to deactivate the catalysts permanently.
6. Materials availability and toxicology
a. All noble metals (Pt, Pd, Rh, Ru) must be imported, mainly
from South Africa (and possibly the U.S.S.R.). Their use in catalytic
converters will more than double the requirements for these metals in
the U.S. While the supply is adequate to meet these new demands for
several years, the price will undoubtedly increase and thus affect
the cost of other items (e.g., electrical and telephone components)
requiring their use.
b. The monolithic supports are all made by domestic suppli-
ers, and production capacity has been expanded to meet the demands.
Most of the beaded catalysts are supported on an alumina base. Cur-
rently, preference is given to an alumina manufactured by a French
-------
4
company. No supply problems are anticipated by the domestic catalyst
finishers who use this material.
c. The oxidation catalysts essentially eliminate the emis-
sion of polynuclear aromatics, olefins, and partially oxidized com-
pounds, some of which are physiologically harmful and are active smog
precursors. By requiring use of low-lead gasoline, the use of cata-
lysts also will result in a decrease in the particulate emissions
attributable largely to lead compounds.
d. Loss of noble metals from the exhaust systems is not a
significant problem under normal operating conditions. During 50,000
mile tests, less than 10% of the catalytic material was lost. Furthermore,
the members of the Panel of Consultants have seen no data to suggest emission
of platinum or palladium in other than the metallic (nontoxic) form.
e. Technology for recycle of noble metals from aged auto-
mobile exhaust catalysts containing lead, various other metals, sul-
fur, refractory oxide support materials, etc. has not been proved in
the absence of economic incentives. No tests have been conducted to
explore the possibility of biological solubilization of noble metals
in discarded converters, and thus it is impossible to assess the
potential this may present for entry into the human food cycle through
water contamination. However, the very small amounts of material
(less than 0.1 oz/car) make such a possibility seem extremely remote.
Moreover, more than 807. of all discarded automobiles are now reclaimed,
and this figure is certain to increase as metals become more expen-
sive.
f. While chemical techniques are known that can remove most
lead from le&d-poisoned catalysts in the laboratory and result in
catalyst reactivation, field testing of these techniques on vehicles
has not been done.
B. RECOMMENDATIONS
1. Poisoning by lead motor-mix components
Conflicting data have been received concerning the relative
-------
5
poisoning characteristics of combustion products of ethylene dichlo-
ride, ethylene dibromide, and the alkyl lead compounds in the usual
motor mix used for octane enhancement. Considering the extremely
important implications of this uncertainty, we strongly recommend
that a continuing, independent study be established to resolve this
issue.
2. Environmental considerations
To assess the impact these catalytic converters may have on
the environment, it is recommended that samples of soil and atmosphere
near typical city, suburban, and freeway roads across the country be
analyzed very carefully for Pb, S, P, noble metals, etc. before the
end of 1974. These data will serve as a baseline for comparison with
periodic measurements taken thereafter to monitor the increase (or
decrease) in these components as the population of catalytic converter-
equipped cars increases.
-------
II. INTRODUCTION
The Panel of Consultants on Catalysts, established by the National
Research Council's Committee on Motor Vehicle Emissions, was organized in
January 1974. The charge to this Panel of Consultants was to assess the
status of catalysts for automobile emission control and to estimate the
future developments in catalytic converters, especially for NO removal.
Geographically, the four consultant members came from widely dispersed
areas (Houston, Chicago, Princeton, and San Francisco), and each has had
several years' experience in the field of catalysis.
To gather data from which to draw conclusions, the members of the
Panel of Consultants made site visits, either singly or in groups, to most
domestic automobile producers and potential catalyst suppliers. Information
from nondomestic automobile manufacturers was collected at a meeting in
Washington, DC, at which many such companies were represented. Further-
more, the members of the Panel of Consultants used data from the open
literature, material presented at technical meetings, suggestions from
independent entrepreneurs, and intuitive assessments based on experience
in catalysis when specific data were not available. Appendix A lists the
companies visited; and Appendix B contains a typical questionnaire sent
to each company prior to the visit.
With few exceptions, the members of the Panel of Consultants were
cooperatively received by the various companies; two catalyst companies
strongly discouraged visits, indicating they were unable to supply perti-
nent new information or that the questions asked did not pertain to their
business. With the use of catalysts on light-duty motor vehicles now
much more of a certainty (at least for oxidation), the companies were more
willing to provide data on catalyst compositions, performance, durability,
testing methods, etc. than occurred during a similar exercise two years
ago. Moreover, there was more general agreement on many points than was
apparent previously. This is due to more standardized, more accurate
testing procedures that have evolved from identification and control of
the significant variables that can influence the behavior of the catalysts.
6
-------
7
Still the exact methods of catalyst preparation were seldom revealed,
as these are generally regarded as proprietary secrets. Absence of
such information did not seriously limit the conclusions, although in
some instances it would have been informative to know the distribution
of active components on the support, a factor that could influence the
susceptibility of the material to poisoning.
Not everywhere is there complete agreement. One important dis-
agreement between information from several automobile and oil companies
arose in the effect of various ingredients in the lead-containing motor
mix used in the fuel for octane enhancement. The data obtained appear
contradictory, and the members of the Panel of Consultants have relied
on their best judgment to resolve the conflict. While they are reason-
ably confident about their conclusions, more data are required in order
to resolve this important issue with complete certainty.
The report begins with a general description of basic performance
of the various systems, catalysts, and testing methods. The next chapter
contains durability data and discusses factors that can lead to catalyst
deactivation, e.g., thermal effects, chemical poisons, physical attri-
tion, and pore blockage. The sulfate formation and possible toxicological
effects of debris emitted from the converters are evaluated in the fol-
lowing chapter, and the last chapter summarizes mathematical modeling
studies and what is known about the kinetics and mechanisms of the various
reactions.
-------
III. BASIC PERFORMANCE CHARACTERISTICS OF AUTO CATALYSTS
Consistent with the primary objective of this report to deal
with catalysts for N0X control, this chapter will treat only very gen-
erally the basic performance characteristics of oxidation catalysts
and will focus most attention on the former.
A. Oxidation Catalysts
Carbon monoxide and hydrocarbons are eliminated through burning
(or oxidation) to carbon dioxide and water. This can be accomplished
either homogeneously at high temperatures in a thermal reactor or het-
erogeneously over a catalyst at much lower temperatures. Figure 1
gives some idea of the temperatures that are required to achieve vari-
ous conversion levels in typical representatives of the two systems.
Achieving the temperatures necessary for the thermal reactors requires
"de-tuning" the engine in such a way as to increase the exhaust tem-
perature above the normal 900-1100°F, and this is necessarily accom-
panied by a fuel-economy penalty and presents materials-corrosion
difficulties in the reactor. On the other hand, the lower tempera-
tures required by the catalytic system effectively "de-couple"1 the
exhaust purification and the engine parameters, a factor that should
allow higher performance and more efficient operation of the auto-
mobile. Hence, there is considerable incentive to use catalytic con-
verters for oxidation. It should also be noted in Figure 1 that
whereas CO is the most difficult pollutant to remove in the thermal
system, hydrocarbons pose the more serious problem over oxidation
catalysts.
The only catalysts that are being considered £or immediate use
in automobiles for HC and CO control include the noble metals Pt, Pd,
and Rh, either separately or in combination impregnated on pelleted or
monolithic supports. The primary reasons these materials are favored
over base metals are their resistance to sulfur poisoning, reluctance
to form less active compounds through reaction with the support mate-
rial, better light-off characteristics, and greater thermal stability
to sintering.
*References are listed at the end of the report (page 105).
8
-------
9
100
z
o
V)
DC
UJ
>
Z
o
o
a?
400
800
1200
1600
2000
REACTOR TEMPERATURE (°F|
FIGURE la Comparison of Thermal Reactors and Catalytic
Converters for Oxidation of HC and CO.
30 r
I 20
S
K
m
ui
a
o
u.
a?
10
Lesd Tolerant Syttamt
_L
1973
Federal
1975
Federal
(49 State&l
1975
California
STANDARDS
1977
Federal
1978
Federal
FIGURE lb Predicted Fuel Economy of Various Emissions Control
Systems with the Indicated Compression Ratio.
REF. 1
-------
10
As will be discussed more fully in the next chapter, it is the
lower activity for HC (than for CO) that limits the effectiveness of
these catalysts. Thus, they must be designed primarily for HC removal,
and the CO removal then requires only a fraction of the total catalyst.
When a catalyst becomes poisoned, it is usually the HC activity that
is first impaired. Furthermore, not all classes of HC are equally
oxidized over noble metals, as may be seen in the Ford data in Table 1^
for a pelleted catalyst that has been "aged" on an automobile for
12,000 miles. While fresh catalysts show higher overall activity, the
relative ease of converting the various HC types always decreases in the
order: acetylenes, aromatics, olefins and paraffins. Methane is
the most difficult HC to oxidize, and the oxidation activity for
paraffins increases with molecular weight. This is not thought to be
a particularly severe problem, however, since methane shows almost no
photochemical activity that results in smog formation. There have
been suggestions that methane be removed from the HC standard, which
would make the standard more easily met and would not have an adverse
health effect. Such a relaxation would require modifications in the
normal testing procedure and involve more sophisticated analytical
equipment.
As long as fuel sulfur remains at the 0.03 wt% level, the
Panel does not foresee the possibility of completely removing noble
metals from the catalysts and still meeting the 50,000 mile durability
requirement. It may be possible, however, to reduce the amount of
noble metals by adding some base metals, but there is still consider-
able disagreement on this issue.
Oxidation catalysts also cause oxidation of SO2 to SO^a prob-
lem that will be discussed in Chapter 5.
B. N0X Reduction Catalysts
General considerations - In its application to automobile emis-
sion control, the catalytic conversion of N0x has proceeded along two
approaches: (1) the dual-bed system (and sometimes three-bed systems),
-------
11
TABLE 1
EFFECT OF CATALYST ON EXHAUST GAS COMPOSITION
Cruise 30 - Vehicle 17-0-F - Indolene Clear Fuel
12,000 Miles
HEW
PPM
7»HC
Reactivity
Compounds
Before
After
Before
After
Before
After
Methane
45.4
45.7
8.5
33.4
0
0
Ethane
11.7
10.3
2.2
7.6
0
0
Butane
11.3
5.5
2.1
4.0
11
6
Isopentane
13.6
4.8
2.5
3.5
14
5
Total Paraffins
148.0
84.8
27.6
62.1
91
29
Total Acetylenes
50.7
0.5
9.5
0.4
0
0
Ethylene
97.5
20.0
18.2
14.7
390
80
Propylene
56.0
7.7
10.4
5.6
392
54
1-Butene,Isobutene
22.0
2.9
4.1
2.2
153
21
Total Olefins
223.6
34.6
41.7
25.4
1275
184
Benzene
14.2
3.6
2.6
2.7
0
0
Toluene
59.1
7.9
11.0
5.8
177
24
Total Aromatics
114.0
16.7
21.3
12.2
402
52
GRAND TOTAL
536
136.7
100.0
100.0
1768
265
Avg. Relative Reactivity
3.297
1.937
REF. 2
-------
12
and (2) the three-way system. In the dual-bed system the combustion
process involves net fuel-rich engine operation with passage of the
exhaust gases first over a N0X reduction catalyst that favors the for-
mation of nitrogen rather than ammonia. Subsequently, air is intro-
duced into the stream before it enters a second reactor containing an
oxidation catalyst for conversion of hydrocarbons and carbon monoxide.
Any ammonia formed in the first catalytic reactor will also tend to be
oxidized back to NOx in the oxidation reactor. To bring the catalytic
system to its light-off temperature (warm-up), it has been found use-
ful to operate the reduction catalyst briefly in an oxidizing mode by
air injection on the upstream side of the first reactor. As a result,
the reduction catalyst has to exhibit stability to a net oxidizing
environment for short intervals (about 1 minute) during cold start-up
of the engine.
In the three-way system the control of automobile exhaust is
achieved by operating the engine near the stoichiometric air/fuel
(A/F) ratio and employing a single catalyst for conversion of N0X, HC,
and CO. This catalytic system requires close control of the A/F ratio,
a problem that has not been solved satisfactorily as yet. Consequent-
ly, the following discussion will deal primarily with N0X reduction
catalysts.
It is useful to classify the NOx catalysts in terms of their
respective temperature ranges for optimum performance.^ On this basis
the active components of the system fall into three categories:
(a) Low-temperature catalysts (Type A): This type of cata-
lyst system contains ruthenium with and without added noble metals.
The optimum operating temperatures range from 600 to 1200°F;
(b) Medium-temperature catalysts (Type B): The active compo-
nents in this category are primarily base metals promoted with noble
metals (other than ruthenium). The optimum operating temperatures
range from 1000 to 1600°F; and
-------
13
(c) High-temperature catalysts (Type C): In this group one
finds single and multicomponent base metals on metallic or ceramic sup-
ports operating in a temperature range of 1200 to 1800°F.
In qualitative terms the performance of these three categories
of catalysts may be summarized as follows:
Type A: The high activity of ruthenium-based catalysts with N2
as the major product of N0X conversion at moderate temperatures and
rich carburetion offers considerable potential in terms of system
design and location within the structure of the automobile. However,
the problem of long-term stabilization of Ru-based catalysts to pre-
vent metal loss with operating mileage has not been demonstrated.
Type B: The higher operating temperatures required by this
type of base-metal/noble-metal catalyst combined with the tendency for
ammonia production (as a by-product of N0X reduction) present major
hurdles to the utilization of this type of catalyst system.
Type C: Base metals on metallic supports have shown satisfac-
tory performance at high temperatures O1200°F). However, their sus-
ceptibility to deterioration under oxidizing/reducing (redox) cycling
conditions requires further development and is at present limiting
their utility.
The following sections will consider in more detail the makeup
and basic performance characteristics of N0X catalysts under controlled
test conditions.
Catalyst details - Both monolithic and particulate (spherical
or cylindrical) supports have found application in catalytic emission-
control converters. Typical examples of the chemical composition of
monolithic supports are given in Table 2.2 The crystalline phase is
made up of cordierite, with mullite and alpha-alumina as secondary
phases. A suitable support material needs to withstand operating tem-
peratures of 2200°F and exhibit high resistance to thermal shock and
attrition.
-------
14
TABLE 2
Composition of Monolithic Catalysts Supports
Percent Composition
American Lava Corning
Major Components
Cordierite (2MgO*2A1202*SSK^)
Alumina (Al^O^)
Mullite (3Al203-2Si02)
Minor Components
Iron (Fe)
Titanium (Ti)
Sodium (Na^O)
Potassium (K^O)
49-59%
23-29%
14-18%
0.5%
0.4%
0.8%
86-98%
3-5%
0.5-1.5%
0.5%
0.4%
0.2%
REF. 2
The ceramic support is provided with a"wash coat" which forms
the high surface area substrate onto which the catalytically active
metallic components are dispersed. Typically, the major component of
the wash coat is gamma-alumina passed as a slurry of aluminum oxide
monohydrate through the monolith to deposit a uniform film and sub-
sequently calcined (heated) at high temperature.
By means of conventional wet-impregnation techniques, the noble
metals are deposited on the wash coat. Generally, water-soluble noble-
metal salts are employed at specified concentrations to yield the
desired weight loading. The resulting material is air dried and
reduced under specified conditions to yield highly dispersed metal
crystallites. Typically, on a monolithic support the metal surface
2
area is of the order of 0.5 m /g catalyst.
Ruthenium-containing catalysts need to be stabilized to prevent
volatilization under oxidizing conditions due to the formation of vol-
atile ruthenium oxides. To this end the addition of basic oxides,^'
-------
15
such as BaO, CaO, and La^O^, can result in formation of the thermally
more stable ruthenate compounds. Preparation of stabilized Ru-based
catalysts involves: (a) impregnation of the support material (either
pelleted or monolithic) with a solution of the Ba-, Ca-, or La-salt;
(b) calcination to form the corresponding oxide; (c) exposure to a
solution of RuCl^; (d) drying at 110°C; and (e) reduction in at
450°C. Alternatively, stabilized ruthenium catalysts have been pre-
pared by applying to the ceramic substrate the Ba-, Ca-, or La-
ruthenate compound as a suspension in the washcoat material.
Laboratory performance studies
Low-temperature catalysts - The activity pattern for NO reduc-
tion exhibits marked differences between some of the more active noble
metals examined**»®® as exemplified by the data in Figure 2 and
Tables 3a and 3b. Although the results obtained under different
experimental conditions of space velocity and gas composition point
to some differences in the activity sequence, it becomes quite appar-
ent that ruthenium occupies a unique position not only in terms of
high conversion of NO but also in terms of relatively low production
of ammonia as a by-product with nitrogen. This effect is demonstrated
by the data presented in Figure 3 contrasting the behavior of alumina-
supported Pt and Ru catalysts under similar experimental conditions.*"®
The decrease in ammonia formation exhibited by Ru catalysts is
manifested also in bimetallic catalysts, such as Ru and Pd deposited
on a ceramic support (see Table 4a). ^ It should be noted that the
ammonia fraction in the product stream is significantly higher than
would be expected for a Ru catalyst in the absence of added Pd. Also,
the results indicate that addition of 1 vol% 0£ to the reactant
stream causes a significant increase in NO conversion and a corre-
sponding decrease in ammonia formation. However, the effect of oxygen
appears to be deleterious to the long-term stability of the ruthenium
catalyst, as demonstrated by the activity loss for NO conversion fol-
lowing exposure of the catalyst for 16 hours to 1 vol % 0^ at 1300°F.
-------
16
RECIPROCAL ABSOLUTE TEMPERATURE (K^/r"^1)
FIGURE 2 Relative Activity of Platinum Group Metals for NO
Conversion. Nominal Loading, 1/8 in. Extrudates: 0.2 wt%:
Activity: Rh >¦ lr >¦ Ru >• Pd * Pt (per Atom Basis); Test
Conditions: Synthetic Exhaust Gas with 45 ppm SO2; GHSV -
138,000 hr_1.
REF. 11
-------
17
TABLE 3a
Performance of Supported Noble-Metal NOx Catalysts
907.. Convr. Selec. for
Catalyst Temp., °F NH3 (PFM) N2 (%)
0.37 wt% Ru/A1203 560 225 70
0.30 wt% Pt/Al203 700 650 13
0.30 wt% Pd/Al203 740 570 24
Feed stream: 0.1% NO, 1% CO, 0.3% H2, 10% C02, 10% H20, balance
•
GHSV = 38,000 hr~*; base was preformed A^Og Kaiser KC/SAF,
surface area 260 m^/g.
Surface area of ruthenium about 0.18 m^/g catalyst.
REF. 8
TABLE 3b
Performance of Supported Noble-Metal NO* Catalysts
90% Convr. Selectivity
Catalyst Temp., °F for Ng, (%)
0.5 wt% Ru/Al203 380 92
0.5 wt% Rh/Al203 495 66
0.5 wt% Pd/Al203 590 25
0.5 wt% Pt/Al203 750 22
Feed stream: 0.57» NO, 2% 2% CO, balance He
GHSV: 24,000 hr"1
REF. 6
-------
18
CATALYST TEMPERATURE (°C)
FIGURE 3 Comparison of Ammonia-Forming Tendencies and NO
Reduction of Supported Pt and Ru N0X Catalysts. Feedstream:
0.1% NO, 1.0% CO, 10% C02, and 10% H2O in a N2 Atmosphere;
GHSV - 38,000 hr_1.
REF. 10
-------
19
TABLE 4a
Performance of Supported Mixed Ru/Pd N0X Catalysts
90% Convr. Max. % No
Pretreatment Gas Composition Temp.. Selectivity
1500°F (N )
v 2' a 520 65
1500°F (N2) a + 1% 02 450 90
1300°F (1% 02
for 16 hrs)
1300°F (1% 02
670 80
90
for 16 hrs) a + n ^ 53Q
Noble-metal loading: 0.2 wt% each metal
Gas Composition a: 2000 ppm NO, 200 ppm HC, 2% CO, 3% HO, balance N2
REF. 3
Apparently due to the formation of volatile ruthenium oxides
(RuOj and RuO^ see Table 5),^ loss of the active Ru component occurs
with simultaneous enhancement in ammonia formation. An effort has
been made to stabilize ruthenium (or its oxide) by adding a foreign
basic oxide (e.g., BaO, CaO, or La.O.) to form thermally more stable
£ 7 7a
crystal phases such as the perovskite structure of LaRuO^. » ' The
degree of stabilization can be seen from the data in Table 6.^ At
temperatures in excess of 1000°C, BaRuO^ converts into BaO and Ru
under reducing conditions, but it can be partially reformed under
oxidizing conditions.^ A more oxidation-resistant catalyst, desig-
nated X-22, has now been reported by Kobylinski and co-workers.
Flow-reactor studies employing monolithic and pelleted cata-
lysts (Tables 7 and 8) and synthetic gas mixtures have indicated
that the addition of the foreign oxides impairs neither the N2
selectivity nor the conversion activity of the catalysts. Similar
-------
20
TABLE 5
Volatilities of Ru0_-Ru0.
3 4
Temp.
(°F)
% o2
ppm Ru0x*
900
2
0.005
900
5
0.01
1000
2
0.01
1600
2
2.83
1600
5
5.31
1800
2
19.41
2000
2
79.83
2000
5
130.41
Concentration in static system at
equilibrium vapor pressure, calculated.
Comparison of Stability of Bulk RuOx and
"Stabilized" BaRuO„ Supported
and Unsupported Catalysts
RuOx BaRu03
T (°F) Bulk Supported Bulk Supported
1000
0
0
0
0
1400
14
22
0
1
1600
40
62
2
18
1800
85
100
16
58
(The numbers are weight % loss of Ru metal after
treatment in flowing 4% - 96% He for 8 hours)
REF. 6
-------
21
i
TABLE 6
i Comparison of Ruthenium Loss in a Ruthenium Catalyst and in Barium Oxide- or
Lanthanum Oxide-stabilized Ruthenium Catalysts
Temperature
Exposure
Ruthenium Loss percent
Time
in
Ru
Ru-BaO
°F
°C
Hours
Catalyst
Catalyst
950
510
16
0
0
1350
730
4
13
0
1450
785
4
45
0
1530
830
4
90
25
1630
885
4
—
25
1750
955
4
—
—
1820
995
4
--
75
Ru-La203
Catalyst
0
0
0
5
25
25
Conditions: Space velocity
O2 concentration in nitrogen
Ru content
La or Ba (as metal) content
Support
38,000/h
2.5-3 percent
2000 p.p.m.
3 percent by weight
UOP AI2O3 pellets
REF. 7
-------
22
TABLE 7
Performance of a Presynthesized BaRu03 Monolithic Catalyst
NO-H2 System N0-H2"C0 System NO-C0-H2O System
Temperature
°C
202
255
278
320
360
447
% NO p.p.m. NH3 % NO p.p.m. NH3
Converted Formed Converted Formed
23.9
89.6
100
100
100
100
18.9
75.0
92.4
186
190
63.2
56.8
100
100
100
100
100
Conditions: Space velocity
Inlet NO concentration
Inlet CO concentration
Inlet H2 concentration
Inlet H2O concentration:
Catalyst support
60.5
212
228
233
263
251
20,000/h
100 p.p.m.
1.5 percent
1.43 percent
10 percent
Corning monolith
%N0
Converted
N.M.
N.M.
100
100
100
100
p.p.m. NH3
Formed
N.M.
N.M.
380
310
260
175
N.M. * not measured
REF. 7
TABLE 8
Comparison of Activity and Selectivity between Stabilized and Non-stabilized
Ruthenium Catalysts. System NO-CO-H2 1
Ru Catalyst Ru-BaO Catalyst Ru-La203 Catalyst
Temperature .
oc % NO p.p.m. NH3 % NO p.p.m. NH3 % NO p.p.m. NH3
Converted Formed Converted Formed Converted Formed
250
85
45
35
30
80
55
300
100
85
92.5
80
98
! 150
350
100
118
100
135
100
i 210
400
100
135
100
210
100
j 208
450
100
137
100
250
100
: 190
500
100
100
220
100
1 50
I
Conditions: Space velocity 20,000/h
Inlet NO concentration : 1000 p.p.m.
Inlet CO concentration : 1.5 percent
Inlet H2 concentration : 1.43 percent
Catalyst support : low-shrinkage American Cyanamid
alumina pellets
Ru content : 2000 p.p.m.
la or Ba (as metal) : 3 percent by weight
REF. 7
-------
23
12
conclusions have been reached by others, although at least one com-
pany reported that complete stabilization of the ruthenium resulted
in an inactive catalyst.Since the NOx-reduction catalysts are
designed to operate in an overall-reducing atmosphere on the automo-
bile, deterioration of the mixed-oxide ruthenates remains a serious
problem especially under cyclic redox operation. It is to be expected
that the reduction to metallic Ru in admixture with the basic oxides
will ultimately cause the formation of Ru metal crystallites that will
tend to volatilize on oxidation. Thus, even the stabilized ruthenium
catalysts (Table 4b)^ demonstrate loss in conversion efficiency due
to exposure to oxidizing and reducing conditions. In reference 14,
the following conclusions are reached concerning the ruthenium cata-
lysts :
At this intermediate stage in our work, it looks as
if at least some stabilized ruthenium N0X catalysts
might be viable under normal, average driving con-
ditions. Further improvement is necessary, however,
to permit ruthenium to be used under the more severe
conditions which can be encountered, perhaps even
routinely, in actual automobile operation.
TABLE 4b
Effect of Redox Cycling on Performance of Stabilized N0X Catalyst
907. Convr. Temp.. °F
Catalyst Fresh After Cycling
A 625 700
B 625 >1100
Cycling: Alternating oxidation-reduction atmospheres for 64 hours
at 1200°F
REF. 11
-------
24
Medium-temperature catalysts - Typical catalysts in this cate-
gory are the supported base metals with and without noble-metal addi-
8 15
tives. ' Their performance on exposure to a synthetic gas mixture
is exemplified by the data in Table 9.15 The Cu-Ni system (on a mono-
lith support), the Pt-Ni system (on a pelletized support) are highly
active for N0X conversion, and the reduction of N0X is accompanied by
appreciable NH^ formation.
To elucidate in more detail the performance characteristics of
the Ni-Pt (Ni/Pt = 500/1) catalyst on a monolithic support, laboratory
tests were performed in which the C0/0„ ratio of the reactant gas mix-
11
ture and the temperature were varied over a wide range. The results
of these tests demonstrated that for CO/O^ ^1.5 (net oxidizing) the
activity for NO^ conversion rapidly decreased. In contrast, the nitro-
gen selectivity diminished with increasing CO/O2 exhaust gas ratio,
especially at temperatures below 1100°F (see Figure 4). Thus, for
net NO* conversion efficiency >80%, the operation of this catalyst
required gas mixture compositions with a CO/O2 ratio between 1.5 and
3.0 (A/F - 14.65 to 13.5). In comparing the selectivity of the Ni-
Cu and Ni-Pt catalysts with that of ruthenium as a function of the CO/O^
ratio, one observes marked differences between base- and noble-metal
additions to the Ni catalysts (Figure 5).**
High-temperature catalysts - The development of an all-metal
catalyst system (support and substrate) for N0X reduction offers
several potential benefits—for example: (a) a low-pressure drop
across the open mesh metal/support configuration: (b) a low-mass and
relatively low-heat capacity metallic construction: and (c) a high
specific surface area of active catalyst. Of numerous non-noble
metal elements studied for N0X conversion, the systems based on Ni
and Cu appear to offer some promise for further development. Fore-
most in this area have been the GEM catalysts in which Ni-Cu alloys^
represent the catalytically active component deposited on high-
temperature corrosion-resistant metallic support. In laboratory tests*?
with synthetic reactant gas mixtures, this catalytic system exhibited
-------
25
EXHAUST GAS C0/02 RATIO
FIGURE 4 Selectivity for N2 of a Ni/Pt NO* Catalyst versus
CO/O2 Ratios. Bench Unit Data: Synthetic Exhaust Gas, 45 ppm
S02; GHSV - 138,000 hr"1.
REF. 11
-------
C0/02 RATIO
FIGURE 5 Dependence of N2 Selectivity on CO/O2 Ratio. Test
Conditions: Synthetic Exhaust Gas, 45 ppm SO2; Temperature:
1000°F; 1/8 in. Extrudates.
REF. 11
-------
27
TABLE 9
Performance Data for Base-Metal and Promoted Base-Metal N0X Catalysts
Catalyst % N0X Conversion % N2 Selectivity
Ni-Cu (monolith) 100 72
Ni-Pt (extrudate) 99 59
Gas Mixture: 1.0% CO, 0.337. 0.35% 02> 0.08% NO, 0.10% C^g,
10% H20, 13% C02, balance N2«
Conditions: GHSV - 45,000 hr"1; Exit bed temp. = 1000°F.
Conversion: Fraction NOx converted to products.
Selectivity: Ratio of volume fraction N2 formed to N0X converted.
REF. 15
the performance data shown in Figure 6a. Net NOjj conversion in
excess of 75 vol% has been obtained at residence times of just over
15 milliseconds.
IO
As reported earlier, the N0X conversion is accompanied by
ammonia formation (especially at temperatures below 1200°F). Its
concentration goes through a maximum with increasing residence time
in the catalytic reactor. In the presence of oxygen (Figure 6b), the
formation of ammonia is greatly suppressed. However, it has already
been noted that the presence of oxygen at a level of 1 vol% retards
the conversion of NOjj. This observation is further amplified by the
results derived from experimental studies of the effect of 02 on the
NO reduction rate (Figure 7).^ A sharp decrease in rate is recorded
as the oxygen level approaches 1 vol7a in the gas stream. These data
suggest that at sufficiently high concentrations the oxygen preferen-
tially occupies the surface sites required for dissociative NO sorp-
16
tion on the catalyst surface.
-------
28
1,500
(a) 0.0% O
1,000
500
E
a
_a
CO
I
Z
a
z
<
o
z
U.
O
z
o
I-
<
a:
o
z
o
o
H
LU
—I
I-
Z3
o
0 20 40
RESIDENCE TIME (msec)
1,500 Or
(b) 1.0 %0
1,000-
500 -
RESIDENCE TIME (msec)
FIGURE 6 Dependence of NO Reduction and NHj Formation on
Residence Time Evaluated for GEM Catalyst at 1200°F and l(r
hr"l. Feed Gas: 1,500 ppm NO, 500 ppm HC, 1.5% CO, 0.5% H2,
12% CO2, 10% H2O, O2 as Indicated, and Balance N2.
REF. 17
-------
29
0.01 0.1 1.0
INLET 02 CONCENTRATION (% by Volume)
FIGURE 7 Dependence of NO Reduction Rate on Inlet O2
Concentration. Curve B - Short Exposure to 02, Curve C -
Long Exposure to O2. Feed Gas: 1.5% CO, 12% CO2, 10% H2O,
500 ppm HC, 1,500 ppm NO, and Balance N2. Temperature:
1200°F; Space Velocity: 6xl05 hr"1.
REF. 17
-------
30
In order to decrease the ammonia-forming tendencies, a modified
19
Ni-Cu catalyst was developed containing Ni, Cu, Cr, and Co; this cat-
alyst is known as GEM-68. Its performance characteristics under labo-
ratory conditions are shown in Figure 8. It is apparent that the total
and net NO conversion efficiencies of this material are superior to
those of its predecessors (e.g., GEM-67, etc.). Furthermore, it is a
low-ammonia former even in the absence of gaseous 0^.
Of critical concern to the applicability of the metal-catalyst
system is its susceptibility to deterioration and mass loss in the
presence of excess oxygen. An interesting approach to the solution of
this problem involves the GEM-68 catalyst in combination with an "oxy-
gen getter" placed on the upstream side of the NOx-reduction catalyst.
In one of the configurations, the "getter" contains a noble-metal cat-
alyst that effectively removes oxygen by catalytic .reaction with the
various reducing agents present in the fuel-rich reactant gas mixture,
e.g., H^, CO, and HC. By close control of the space velocity, such a
system yields high NOx-conversion efficiency and low ammonia formation
under the selected laboratory conditions (Figure 9).^
For vehicle application, the laboratory results suggest the fol-
lowing requirements for this all-metal catalyst system: (a) net fuel-
rich exhaust gas mixture with CO/O^ ratios greater than 1.5: (b) ex-
haust bed temperatures near 1200°F (temperature limit 1600-1800°F): and
(c) space velocities of the order of 10"* hr \ In vehicle tests the
light-off temperature of GEM catalysts has been found to be between
850 and 900°F, a temperature range that exceeds by 200-250°F the tem-
perature attained by the catalyst bed during the inital 20 seconds of
the FTP cycle.
20
Similarly, in the "Reverter Emission Control System," the use
of an all-metal catalytic device has been adopted. The "Reverter"
system includes three stages, only one of which involves a catalytic
process (the second). In this second stage, a base-metal catalyst on
a metallic grid reduces NOjj to Nj by reaction with CO, HC, and/or Hj
in the exhaust system.
-------
31
100
5 80
60
40
NO
GEM 68
GEM 67
%o2
FIGURE 8 Comparison of Ammonia Formation on GEM-67 and GEM-68
as a Function of O2 Concentration. Feed Gas: 1,500 ppm NO,
500 ppm HC, 1.5% CO, 12% C02, 0.5% H2, 10% H20,
Space Velocity: 10* hr"^-
Balance N.
2'
REF. 19
-------
32
100
100
%0,
FIGURE 9 Performance of GEM-68 "Getter" System. Curves A and
B: Previous Data; Curve C: GEM-68 System with Upstream PTX-A
"Getter." Temperature of NOx Catalyst: 1200°F; Space
Velocity: 10^ hr"l.
REF. 19
-------
33
A number of high-temperature, base-metal alloys were examined
for their activity as exhaust-control catalysts and their corrosion
resistance under redox conditions. The type of catalytic materials
20
examined and their relative reactivities are shown in Figure 10.
None of these have withstood the rigors of long-duration vehicle tests.
Most recent tests have involved a three-component metal composition
(Cu-Cr-Ni) applied to an expanded-metal substrate of Inconel-601 alloy.
At CO/OjQ.l and operating temperatures of 1500-1800°F, this system
exhibited more than 907o (gross) N0X conversion. The fractional con-
version to NH^ was not given,^ although it was probably quite low at
these temperatures. One strong point favoring this high-temperature
system is insensitivity to lead.
Ammonia-oxidation catalysts ("three-bed" system) - A problem
common to the N0X catalysts discussed in the preceding sections is the
tendency to form ammonia. It appears that this by-product, in the
presence of excess oxygen, is rapidly reoxidized to N0X in a dual-cat-
alyst system (a tandem arrangement in which a N0X reduction reactor is
followed by an HC and CO oxidizing reactor). As a result, the concept
of a triple-bed catalyst has emerged2>13 which incorporates three cata-
lytic reactors in series (one for NOx reduction, one for NH^ oxidation,
and one for HC/CO oxidation). The system provides for the controlled
introduction of air to the NH^ catalyst section for oxidation to nitro-
gen before coming in contact with the oxidation catalyst for HC and CO.
Such a scheme would greatly relax the limitations imposed on a number of
available N0X catalysts that exhibit high activity for gross NOjj but too
much ammonia formation. At the same time, the three-bed system adds some
complexity due to increased catalyst mass, warm-up time, and engineering
design. As of the time of this writing, the selection of a suitable
ammonia-oxidation catalyst has not taken place. The scientific litera-
22 23
ture ' suggests that some base-metal oxides exhibit the needed high
N2 selectivity for ammonia oxidation. At the same time, such catalysts
tend to oxidize hydrocarbons incompletely (oxidative dehydrogenation to
olefins and partial oxidation to aldehydes, ketones, etc.), so that
the third catalytic section would still be required to convert these
intermediate products into CO^ and 1^0. The development of the three-
bed catalytic system represents an interesting approach to the NO*
-------
34
lncone(6T7 ^
Inconel 601
Experimental Alloy
W
jfc
330 Stainless Alloy
w
A
Platinum-Plated Inconel 601
w
m
Cobalt-Plated Inconel 601
W
i i i i i i i i 1 i
0 10 20 30 40 50 60 70 80 90 100
RELATIVE REACTIVITY
FIGURE 10 Relative Reactivity of Catalysts in Questor's
Reverter.
REF. 20
-------
35
conversion problem which, in spite of added complexity, may offer some
promise of meeting the ultimate N0X standards of 0.4 g/mi, at least at
low mileage.
C. Three-Way Catalytic Systems
In the dual-bed and triple-bed catalytic systems, both oxidiz-
ing and reducing conditions are simultaneously required in separate
reactors to effect decomposition of all three pollutants. It is pos-
sible, however, to convert the three pollutants in a single reactor if
just the right set of conditions can be maintained. While NOx removal
requires net reducing conditions and HC and CO require oxidizing con-
ditions, all three pollutants can be converted simultaneously at A/F
ratios at or slightly lean of stoichiometic as illustrated in Figure
11.^ Unfortunately, with most catalysts this "window" for effective
performance is exceedingly narrow, being the order of +0.1 A/F ratio
units. Such control is impossible with present carburetors alone
because of the variation in the mixture with vehicle operating mode.
About the only way to obtain such precision is through the use of an
oxygen sensor coupled with a feedback loop to control the A/F ratio.
Intensive work is in progress in the U.S., and particularly in foreign
countries (e.g., Germany), to develop such a system that will with-
stand the thermal cycling, poisoning effects, and vibration that would
be present in automobile exhaust systems. To minimize instabilities
in the feedback circuit, some manufacturers^ have suggested using
more than one sensor at different points in the exhaust system. Such
an approach offers several advantages since it involves only one cat-
alyst bed and requires operation at a point where neither automobile
performance nor fuel economy will be impaired.
Neither the catalysts nor the sensors for such a system have
been perfected. Work is continuing on the development of catalysts
with wider, windows which will decrease the degree of carburetion con-
trol required. Most catalysts presently being investigated contain
-------
35a
100
80
60
40
20
/
HC ¦
/>
' CO
13 14 15 16
AIR-FUEL RATIO
FIGURE 11 N0X Conversion Window for Three-Way Catalyst.
REF. 24
17
-------
36
noble metals, specifically Pt, Pd, Rh,^ (and possibly Ru). Not only
does the catalytic activity decrease with age, but the "window" has
been observed to shift slightly for some catalysts, as will be dis-
cussed in detail in the next chapter. Also, the sensor response does
not remain exactly constant as it ages, and these two changes could be
sufficient to make the system completely ineffective for one group of
pollutants after it had aged, even though it was perfectly calibrated
when new.
D. Discussion
The laboratory studies offer a valuable guide to the evaluation
of N0X catalysts for application to automobile exhaust control. A
number of parameters need to be considered, some of them exhibiting
supporting, others, opposing effects as shown by the data entered in
Table 10. The choice of ^-selectivity as the important criterion for
catalyst selection will lead to an entirely different system than the
criterion of minimum temperature required for 907» NOjj conversion (activ-
ity). At the same time, it is to be recognized that the feasibility of
a three-bed catalyst system ((1) N0X reduction, (2) NH^ oxidation,
(3) HC/CO oxidation) makes the low-temperature, noble-metal catalysts
contenders for NOx removal. Similarly, the high-temperature, all-
metal catalyst systems become interesting by going to a three-bed cat-
alyst system ((1) 0^ removal, (2) N0X reduction, (3) HC/CO oxidation).
The catalyst technology available at this time offers no other alter-
natives in meeting the standards of 0.4 N0X g/mi. It is somewhat
surprising that the development effort in the use of three-bed cata-
lytic systems has not advanced sufficiently to evaluate in some
detail the potential utility of such systems in vehicle operations.
While the laboratory tests have served as guides to the devel-
opment of new systems, the ultimate test is performance in automo-
biles operated under realistic conditions, which is the topic con-
sidered in the next chapter.
-------
TABLE 10
NOx Catalyst Performance Summary
(Qualitative)
System
NOx Activity Redox
T 90% Convr. Stability
N2
Selectivity
Range Poisoning
CO/O2 Resistance Durability
Supported noble metal
(not Ru)
Supported, stabilized
Ru
Supported base metal
All metal
600-1200°F
1000°F
1000-1600°F
1200-1800°F
B
B
C
A
B
A
3
2-3
10
B
B
A
C
B
B
A » high; B = moderate; C = poor
a - CO/O2 ratio is representative of air/fuel ratio (when allowance is made for hydrocarbons typically
present in the exhaust, a ratio of ^1.5 indicates net oxidizing condition)
b - Poisoning resistance to Pb, P, S
-------
IV. VEHICLE PERFORMANCE AND DURABILITY OF AUTO CATALYSTS
In an earlier report, the overall deterioration of automobile-
exhaust, emission-control catalysts was separated into two components:
deterioration due to deactivation and deterioration due to poisoning.
The same breakdown will be used in this chapter. The first category
would include those factors encountered in operation with contaminant-
free fuel and oil and represents deterioration due primarily to alter-
ation of the physical nature of the catalyst. The second includes
factors encountered upon contamination of the active-catalyst surface
and represents deterioration due primarily to alteration of the
chemical nature of the catalyst.
In this chapter the durability discussion sequence will begin with
oxidation catalysts and then proceed to various HC/CO/NOx systems.
Although a complete division between deactivation and poisioning effects
is not possible, one may generally associate slow, long-term decreases
in activity to deactivation and initial rapid losses in activity to
26 27
poisoning. ' While a reasonably comprehensive picture of both de-
activation and poisoning is now available for oxidation catalysts alone,
there is as yet very little information regarding the long-term dura-
bility of catalytic systems for N0x control.
No attempt will be made to discuss all catalysts here, only those
which, in the case of oxidation, have been developed to the point of
application or, in the case of NO^, are the most promising candidates.
By and large, these are all noble-metal formulations, except for the
high-temperature systems discussed in the preceding chapter.
A. Deactivation of Oxidation Catalysts
The primary mechanism for deactivation is the sintering or ag-
glomeration of the small crystallites of active metallic components by
a thermal mechanism. Associated with sintering of the supported metal
may also be deterioration of the support structure and changes in the
38
-------
39
nature of the support-catalyst interaction. All these processes are
characterized by high activation energies; they occur to only a mini-
mal extent at normal temperatures of operation (900-1200°F). Thus, the
extent of deactivation is primarily a function of the prior thermal
history of the catalyst.
The most severe mode of deactivation is, of course, melting.
For pelleted catalysts that use gamma-alumina, melting temperatures
25
are around 3700°F; this failure mode is not important. However,
changes in pore-size distribution and decreases in the pellet volume
can occur at substantially lower temperatures. Monoliths, melting in
the range of 2400-2500°F, have more of a problem, although some mono-
28
liths are now claimed to be stable to 2800°F. The specific condi-
tions associated with melting are: (1) dieseling after ignition shut-
off; (2) running out of fuel; (3) long downhill deceleration; and
(4) ignition failure. In each case the ultimate effects depend on
the frequency of such incidents. In sintering, as opposed to melting,
occasional temperature excursions in the range of 2000°F can be tolerated
by existing monoliths containing noble metals; fouling of one or even
two plugs can be withstood under normal (50umph,steady-state) condi-
2
tions. Melting of monolithic supports due to abnormal driving modes
is apprently more frequent with small displacement engines (e.g., 4-
29 30
cylinder) than with larger vehicles. ' One manufacturer of small
30
displacement automobiles felt that the incidence of failure due to
overheating from abnormal driving modes could be as high as 25% over
50,000 miles. With manual choke on after a cold start, they have
measured converter temperatures as high as 2400°F on downhill decelera-
tion.
Conditions that result in over-temperature failures of various
ceramic monolithic oxidation catalysts apparently will be almost
eliminated by engine modifications (such as solid-state ignition,
31
higher voltage, and longer spark). Exxon indicated that the catalysts
indeed serve as an unintentioned but highly sensitive detector of
-------
40
ignition malfunction; Engelhard reinforced this belief claiming that
the automobile manufacturers have told them not to be concerned any
longer about over-temperature as a significant problem because of
basic engine improvements.
Catalyst attrition, potentially leading to enhanced particulate
emissions, was not reported as a problem in a single visit to catalyst
or automobile manfacturers. Corning claims to have eliminated wash
coat peeling about two years ago by development of monoliths with pore
28
sizes several times that of alumina particles in the adhering wash coat.
The Rhone-Progil alumina pellet, as well as the W.R. Grace manufactured
extrudate formulation, have led to pellet oxidation catalysts meeting
25
General Motors' constraints for particulates. Remaining attrition
problems now appear associated largely with loading techniques for
"canning" the monolith or pellets and with quality control for the
28 32
large-scale manufacture of these materials, ' although Houdry is
still searching to develop a better pellet support than the imported
Rhone-Progil candidate. Health hazards that may be associated with
attrition are discussed in the next chapter.
These findings underscore a basic mood shift from two years ago.
The successful catalyst competitors appear to have most of their man-
power now on quality control problems and manufacturing debugging for
oxidation-catalyst production. Efforts in new catalyst development
have almost ceased.
B. Deactivation of NO^ Catalysts
At the present state of the art, the various N0x systems are more
susceptible to thermal damage than oxidation catalysts alone. All
automobile and catalyst manufacturers visited voiced this opinion.
The NO^ systems developed by Questor and Gould seem particularly
susceptible to thermal damage, oxidation and loss of particulate oxide
21 19
matter in the former case, melting in the latter. Both of these
manufacturers have developed improved systems to avoid such thermal
damage, which is associated with oxidizing conditions. Questor has a
-------
41
pre- and post-bed air-injection system that is designed to keep inlet
C^/CO ratios in the range of 0.004-0.4 while maintaining the catalyst
(Cu-Cr-Ni alloy, IN-1013, bonded to Inconel 601) operating temperature
from 1500-1800°F, which can be tolerated. Gould has incorporated a
monolithic oxidation catalyst in the system upstream of their NO catalyst
A
(Cu-Ni-Co-Cr on a high Ni-alloy base) to remove from the exhaust
mixture, as discussed in the preceding chapter.
Thermal degradation of Ru-based NO^ catalysts is associated with
oxidation-reduction cycling. This involves a gradual separation and
agglomeration of Ru metal from the stabilizing structure whereupon it
33
is oxidized to volatile ruthenium oxides, as discussed in chapters
III and VI. The net rate of volatilization is probably thermally con-
trolled due to the high activation energy for Ru metal diffusion in
the catalyst matrix. Volatilization of current supported Ru formula-
tions (primarily based on BaRuO^) becomes a problem at temperatures
above 1100-1200°F, although the exact temperature range where trouble
begins depends on the severity and frequency of redox cycling. Normal
operating temperatures for NO systems may vary from 900-1500°F for
x 32
large engines, 1100-1700°F for small engines. Some representative
data on temperature excursions in NO^ portions of a two-bed system
from abnormal driving modes in small displacement autos are given in
29
Table 11.
Noble metal-base metal combinations are of interest to some
manufacturers because of their improved thermal stability. These are
normally mixtures containing Ni and Rh (and in some cases Pt and/or Pd),
and these can be stabilized to have thermal durability comparable to
current oxidation catalysts.
Another form of deactivation appears in the Gould GEM system.
19
This is called "green rot," which is a term describing the flaking
off of the thick active ingredient layer from the base metal. This
embrittlement process arises from attack at the grain boundary of the
support and is associated with oxidation/reduction cycling which the
improved GEM-68 system avoids (see preceding chapter).
-------
42
TABLE 11
Abnormal Driving Modes - Effect of Misfiring Cylinder
Datsun P610 mounted on Chassis Dynamometer
Exhaust Manifold-Mounted NO^ Catalyst System
Rise of NC>x Catalyst Bed Temperature, °F
Vehicle Speed
(km/hr)
Bed Temperature
Normal Opn. 1 Cvl. Misfiring
60
80
100
1060 ^ 1640 (after 80 sees)
1240 > 1740 (after 50 sees)
1330 ^ 1740 (after 33 sees)
REF. 29
C. Deactivation of Three-Way Catalysts
To date there have been few data presented concerning the deacti-
vation of three-way HC/C0/N0x catalysts. However, the composition of
most of these catalysts appears to be a variation on current oxidation
catalysts, involving Pt, Pd, Rh, or combinations thereof. Thus, it
t
would seem reasonable to assume these three-way systems will have
thermal durability comparable to the existing oxidation catalysts. As
important as the thermal stability of the catalysts is the stability
(and thermal response) of the oxygen sensor used with three-way
catalysts. Most sensors that are currently under evaluation or testing
are based on zirconia. The normal configuration employs a doped
zirconia thimble with platinum electrodes mounted on a device which
looks very much like a spark plug. There have been some instances of
sensor failure due to thermal shock, but this seems a minor factor in
the determination of their durability. Poisoning is the problem here,
as will be discussed subsequently.
-------
43
D. Poisoning of Oxidation Catalysts
Catalyst poisoning, in the automobile emission context, is
largely chemical in nature, although physical factors can in some
cases be important. An empirical approach to the catalyst-poisoning
field appears most sensible here. In the context of the charge to
the Panel of Consultants on Catalysts, that of assessing the feasibility
of catalyst usage on vehicles, the order of decreasing utility of catalyst
poisoning studies is apparently: (1) vehicle data; (2) multicylinder
engine data, preferably over a realistic driving cycle rather than at
steady state; (3) single cylinder data; and, finally, (4) bench-
scale synthetic atmosphere tests with potential catalyst poisons.
Under identical engine and exhaust conditions, two catalysts with the
same overall composition may still give different results due to dif-
ferences in properties such as active metal surface area, radial distri-
bution of the active component in the catalyst, and support surface
area and pore size. These useful pieces of information, needed to
understand fully how a sometimes volatile, sometimes particulate
poison will affect the catalyst performance, are not routinely
measured as a function of time-on-stream.
The materials present in exhaust systems capable of poisoning
oxidation catalysts are numerous, but fortunately most are present in
small amounts. The primary poisons are lead and associated alkyl
halide motor-mix components, sulfur, phosphorus, and zinc. Additional
materials such as alkali metals and magnesium are occasionally men-
34
tioned as potential poisons, but the data available indicate them
to be only minor problems. In actual operation with noble-metal oxida-
tion catalysts, however, the list can be shortened considerably. Under
ordinary conditions of oil consumption (1 qt/1000 mi), the amount of
zinc "seen" by the catalyst is not sufficient to be an important factor
in poisoning. Phosphorus introduced from lubricating oil or as a
34 39-
fuel additive presents more of a problem ' but this is not so
serious a problem as posed by the presence of lead in the fuel. To
serve as a point of reference, the total emissions of Pb, S, and F
passed through a converter during 50,000 miles are given in Table 12.
-------
44
The general effects of the major poisons, lead, sulfur, and
26
phosphorus, have been summarized in an earlier report. Aside from
the current controversy concerning the effects of Pb vis-a-vis
scavenger poisoning, the more recent data made available to the
Panel are in general agreement with the information contained in the
previous report. Each poison is considered separately.
Sulfur - At low temperatures sulfur can poison the oxidation
activity of noble metals. However, under normal operation with sulfur
2 25 35
contents of 0.03 wt% in the fuel, sulfur is not an active poison ' '
provided long-term accumulations are prevented. These can be avoided
by attaining catalyst temperatures of about 1400°F from time to time.
Sulfur is a much more severe poison for base-metal catalysts. At the
above-mentioned sulfur level in the fuel, no base-metal catalyst has
consistently met the hydrocarbon standards (0.41 g/mi) for significant
mileage; however, in one case a base-metal catalyst promoted with less
32
than 0.01 wt% noble metal met the standards for over 10,000 miles.
Phosphorus - Phosphorus, even in relatively small quantities,
is a permanent poison for all types of oxidation catalysts. It has
recently been shown that phosphorus (0.327»), added to ashless engine
34
oil, increased catalyst deactivation during a test over 30,000 miles.
This represents exposure at about three times the rate indicated in
36
Table 12 and substantiates earlier data showing the poisoning of
oxidation catalysts by P in the fuel. However, it is the consensus
of automobile and catalyst manufacturers alike that P at current levels
in fuel and oil, given reasonable rates of oil consumption, does not
2 4
deactivate oxidation catalysts. In fact, in one case ' it was noted
that there seems to be some lessening of Pb sensitivity in the presence
of phosphorus. This could be due to the formation of lead phosphate
which has little poisoning activity.
Lead (and halide scavengers) - The severity of poisoning may
depend strongly upon the physical and chemical state of the poison
when it reaches the catalyst, and these factors make it extremely
difficult to establish exactly the origin of poisoning of oxidation
catalysts by lead-containing fuels. For example, lead and sulfur
-------
45
TABLE 12
Total Amounts of Primary Poisons Passed Through
Catalytic Converter in 50,000 Miles
Oil Component (wt%)a g/50,000 miles
P (0.13) 54
S (0.35) 145
Sulfated Ash 1.3 540
Fuel Component''
P (0.01 g/gal) 50
S (0.03 wt%) 4140
Fb (0.05 g/gal) 250
Consumption - 1000 mi/qt
^Consumption - 10 mi/gal
separately may be expected to influence catalyst performance differently
than when mixed together in such a way that they can form lead sulfate.
Furthermore, the temperature, length, and condition (whether fresh or
corroded) of the manifold and exhaust pipe leading to the converter
may determine whether the poison is volatilized or present as small
(or large) particles. Thus, a complete understanding must take into
account all these physical parameters, as well as chemical knowledge
of the poison and the poison-active site interaction.
Lead, a frequently accused poison of catalysts, has been ob-
served in exhaust systems as lead oxide (PbO), lead sulfate (PbSO^),
lead phosphate (Pb^PO^^), lead oxysulfate (PbSO^-PbO), and lead
chlorobromide (Pb(Cl,Br)2).37 The Panel has not seen a definitive
analysis of the poisoning of catalysts by individual, well-defined lead
compounds. Actually, the surface poison moiety, as has been shown re-
peatedly in other surface studies in corrosion, catalysis and epitaxy,
may be a two-dimensional compound that has no three-dimensional analog.
-------
46
The most complete, comparable data available concerning the joint
and separated effects of lead and various halide scavengers are from
38 31
Chrysler and Exxon . In these dynamometer-aging studies, large
amounts of lead and/or halides were passed over the catalysts and the
effects on HC and CO oxidation monitored. Since the conditions of
both catalyst-aging and catalyst-activity measurement are important,
the details of each study are summarized below.
- , 38
Chrysler:
Type I aging - 3 g Pb/gal (except Pb + EtCl- at 2 g Pb/gal).
Schedule: 45 sec at 2400 rpm, 15 sec at idle. Typical
inlet: 1-1.5% CO, 4-5% 02, bed temperature about 1050®F.
Activity measured at T = 1000-1100°F, 1-1%% CO.
Type II aging - same fuels. Schedule: 3300 rpm (48 min
at 0.2% CO, 2% 0-, bed temperature about 1550°F); 12
min at 2% CO, 0-0.1% O-j bed temperature about 1400°F;
Activity measured at T = 1200°F, 2% CO.
Catalyst - Pt/Pd on alumina washcoat over monolith, Chrysler
manufactured.
Exxon:^
Dynamometer aging at rpm_jquivalent to 40 mph cruise,
space velocity 55,000 hr over each catalyst, catalyst
mid-bed temperature about 1200 °F. Fuel 3.5 g Pb/gal
+ 50% halide excess except where indicated. The two
catalysts were Engelhard PTX-IIB (Pt/Pd) and Engelhard
PTX-A (Pt only), both supported on ceramic monoliths.
Aging times were 7-10 hours on modified fuel (15-20
gal ) followed by return to the original "unleaded"
fuel.
These tests are of interest because they relate to typical
average catalyst-bed temperatures during vehicle operation and because
the poisoning derives from actual engine emission products, not from a
synthetic exhaust. The two sets of data are compared in Table 13.
There is general agreement on most points, with one glaring but
critical exception involving the effect of Pb vis-a-vis Pb + EtC^.
With halide scavengers only (no lead present), both sets of data
show similar results, as indicated below.
1. Bromine-containing exhausts will seriously deactivate
Pt/Pd catalysts for both CO and HC oxidation.
2. Chlorine is a much weaker poison of such oxidations.
-------
47
TABLE 13
Comparison of Exxon and Chrysler Lead Poisoning Data
Chrysler Data/Exxon Data
Loss of %C0 Conversion Loss of 7oHC Conversion
Additive Package
Temporary
Permanent
Temporary
Permanent
TEL + EDC + EDB
69/55
7/5
52/47
16/17
TEL + EDC + EDB
--/10
-/10
--/30
— /19
TEL + EDCC
3/11
2/3
0/46
2/33
TEL + EDB
95/70
10/0
76/64
20/13
TELb
(1—5)/2
5/1
0/9
7/9
EDB + EDC
90/67
2/0
54/37
16/0
EDB + EDC
~/5
--/0
— /12
~/0
EDB
55/56
2/0
46/52
8/4
EDC
0/0
0/0
0/13
3/5
a3.5 g Pb/gallon + 1.5 theory scavenger(s) (Exxon)
3.0 g Pb/gallon + 1.0 theory scavenger(s) (Chrysler)
^2.0 g Pb/gallon (Exxon)
c2.0 g Pb/gallon + EtCl2 (Chrysler); Chrysler dynamometer conditions
unspecified.
Poisoning time: 7-10 hours (Exxon), 10 hours (Chrysler)
TEL - tetraethyllead
EDC - ethylenedichloride
EDB - ethylenedibromide
REFS. 31, 38
-------
48
3. Halide catalyst poisoning is largely or completely reversible
on removal of the halide from the exhaust, even within the
time scale of these experiments (a few hours).
4. Comparison of the Pt/Pd results and the Pt results suggests
that Br attacks primarily Pd.
When lead compounds (TEL) are present in the fuel, the results
(including the point of disagreement) can be summarized by the following
statements.
1. TEL alone in the fuel results in lesser deactivation than
TEL + halide scavengers. (What fraction of this unscavenged
lead actually reached the catalyst is not established since
neither group performed lead material balances on their
systems.)
2. Pt/Pd catalysts appear to recover more completely than Pt-
only catalysts, but they are poisoned more easily, especially
for CO oxidation. (This is consistent with other data
showing susceptibility to poisoning of Pd vs. Pt in the
32 39 40
presence of lead + motor mix. ' '
3. With the exception of the Chrysler data for TEL + ethylene
dichloride, all results indicate a greater permanent activity
loss for hydrocarbons in the presence of lead (and scavenger)
than in the presence of scavenger only.
4. It appears that one might infer (perhaps incorrectly) that
the catalyst itself recovers its CO oxidation capacity more
easily than its HC oxidation ability. Since CO oxidation on
a new catalyst is much more rapid than HC conversion, the
recovery of any appreciable fraction of catalyst activity
will bring CO conversion back to a very high level. In
other words, since all of the catalyst is needed for HC
(but not CO) oxidation, HC oxidation recovery is the key to
catalyst poisoning detection.
Based on these data alone, it is not possible to resolve the
Chrysler vs. Exxon differences for hydrocarbon oxidation with tetra-
ethyl lead plus ethylene dichloride in the fuel. The central point of
-------
49
difference is that Chrysler sees little, if any, activity loss, whereas
Exxon reports a substantial activity loss, both temporary and permanent.
41
In an attempt to resolve these differences, General Motors
recently conducted some experiments on both bench-scale apparatus and
automobiles. In general, their results confirm the Exxon data and
contradict the Chrysler results. Their data plotted in Figure 12
clearly confirm that ethylene dibromide by itself is a temporary
poison (in agreement with both Chrysler and Exxon). Similar tests
showed that ethylene dichloride had a smaller temporary effect. How-
ever, Figure 13 indicates extensive poisoning with TEL + EDC, in
sharp contrast with the Chrysler data in Figure 14. Notice that the
poisoning was apparently irreversible (within the time frame of the
experiment). Unfortunately, General Motors did not report tests with
pure TEL in the absence of any scavengers.
41
To explain the contradictory results, General Motors suggested
that interference of lead compounds deposited on the insides of the
exhaust pipes in the absence of scavengers could be removed and trans-
ported to the catalyst when the scavengers were introduced. They
offered experimental evidence to support this claim. However, this
does not help resolve the discrepancies between the Chrysler vs. General
Motors and Exxon data.
While more work is obviously necessary to clarify this issue with
complete certainty, the members of the Panel of Consultants are convinced
that under most conditional, lead, when it reaches oxidation catalysts, acts
as a permanent poison. It may be slowly desorbed in some form during pro-
longed treatment in unleaded streams at high temperature, but it also
diffuses into the catalyst structure. It is doubtful that oxidation
catalysts once poisoned by lead will ever regain their full unpoisoned
activity.
There are numerous reports that catalyst deactivation is
36,37,39-49
directly related to the amount of lead on the catalyst,
although most of these tests were carried out in the presence of
scavengers. But lead deposits are seldom uniform throughout the
-------
49 a
TIME (hours)
FIGURE 12 Effect of Bromide Scavenger on HC Activity in Lead-
Free System. Catalyst: Pt-Pd; Scavenger: 1.4 g C2H*Br2/gal;
Inlet Temperature: 565°C; GHSV: 128,000 hr .
REF. 41
-------
50
TIME (hours)
FIGURE 13 Effect of TEL and Chloride on HC Activity. Catalyst:
Pt-Pd; Fuel Additive: 3 g Pb + 2.8 g C2H^Cl2/gal; Inlet
Temperature: 565°C; GHSV: 128,000 hr"^.
REF. 41
-------
51
30
20-
10
Unleaded Fuel
I
_L
Modified Fuel
I
_L
Unleaded Fuel
j I I
10 15 20 25
HOURS OF TEST
30
35
40
FIGURE 14 Oxidation Catalyst Poisoning Test - Effect of Fuel
Containing Tetraethyl Lead and Ethylene Dichloride. Unleaded
Fuel: MS 3900A; Modified Fuel: MS 3900A at 2.0 g/gal Fb Plus
1.0 T Ethylene Dichloride.
REF. 38
-------
52
catalyst bed, and the deposition is a function of several system param-
eters. With a monolithic catalyst, Daimler-Benz AG (Mercedes) showed
that the deposition of Fb at the entrance of the bed passes through a
maximum as a function of temperature, as shown in Figure 15."^
40
Klimisch, et al. have examined the distribution of lead in a poisoned
pelleted catalyst and also found changes that depend on temperature.
Figure 16a shows that the Pb is concentrated mainly on the external
surface of the pellet under normal operation, but Figure 16b shows
that it becomes more homogeneously distributed throughout the pellet
after treatment in 0^ at 900°C. It is not surprising, then, that the
position of the catalyst bed has a significant effect on the suscepti-
bility to poisoning. As the catalyst is moved further downstream
from the exhaust ports, more lead is deposited. While the degree of
catalyst deactivation varies with fuel lead levels at all locations,
the influence per gram of lead deposited diminishes as the catalyst
position is moved further from the engine. This may be explained by
the lead being deposited as particles which may be larger in size the
further from the engine and thus tend to remain at the external sur-
face of the catalyst. Other chemical factors such as the interaction
with sulfur oxides (perhaps to form lead sulfates) may affect the
38
poisoning behavior, and such formation has been advanced by Chrysler
to account for the decrease in SO^ emissions in the presence of lead.
Other interactions with such exhaust components as oxygen, phosphorus,
halides, etc. could be discussed, but most of these would be in the
realm of speculation and will not be treated.
E. Poisoning of NOx Catalysts
The base-metal systems of Questor and Gould appear to be quite
resistant to chemical poisons, not because of any inherent superiority
of the ingredients in this respect, but because of the very high tem-
peratures of operation involved. However, addition of the upstream
"oxygen getter" catalyst, which is a conventional oxidation catalyst
containing noble metals, and the downstream oxidation catalyst make the
-------
53
TOTAL LEAD SUPPLIED BY FUEL (grams)
FIGURE 15 Lead Deposits at Entrance of Catalyst Over Lead
Content in Fuel as a Function of Gas Temperature.
REF. 50
-------
DEPTH OF PENETRATION INTO CATALYST PELLET (fiml DEPTH OF PENETRATION INTO CATALYST PELLET t/jm)
FIGURE 16 Lead and Sulfur Distribution in Lead-Poisoned
Catalyst Pellet. Right-Hand Graph: After Exposure for 100 hr
to Exhaust from an Engine Using 0.5 g/gal Pb in Fuel. Left-
Hand Graph: After "Regeneration" in Air for 90 hr at 900°C.
REF. 40
-------
55
Gould GEM-68 system subject to poisons. Other types of N0x catalysts--
the Ru formulations, noble metal-base metal combinations, and the
three-way catalysts--again are susceptible to the "big three" poisons
of lead, phosphorus, and sulfur. Generally, the effect of these poisons
2 29
is very similar to that on oxidation catalysts ' if measured in
terms of total activity loss. Some quantitative data on the suscep-
tibility of ruthenium catalysts to the combined effects of Pb and S
under both oxidizing and reducing conditions are shown in Figure 17^.
(Note that there is some variation in space velocities in these figures.)
It is seen that these formulations are much more severely affected by
Pb and S under oxidizing rather than reducing conditions. However,
in order to meet combined HC/CO/NOx standards,the selectivity as well
as the activity of the catalyst must be maintained. Maintenance of
selectivity seems to be a major problem given the current status of
N0x catalyst development. A persistent difficulty seems to be the
selective deactivation of HC oxidation activity while operating under
conditions that decrease NO^ emissions. These troubles are particularly
notable with three-way systems. Some typical data on changes of
2
selectivity in a three-way system are shown in Table 14.
As is evident from the general nature of the comments above,
very little information specific to the effect of individual poisons
on NO systems is currently available. There are, however, more data
2C
available on overall durability of NO^ catalysts systems, which is
discussed in the following section.
F. Long-Term Durability
Oxidation catalysts - Most automobile manufacturers who intend
to sell cars in the U.S. have developed systems with noble-metal oxida-
tion catalysts that will last 50,000 miles when run on EPA durability
fuels, and most have vehicles that are now in the final stages of
certification for the 1975 interim standards. Furthermore, several of
the companies have met the ultimate 3.4 (CO) and 0.41 (HC) g/mi
standards for 50,000 miles. All these systems have been operated on
-------
56
too
80
40
20
VHSV 200,000 hr
\
\
\NH3-
\
13.5 14.0 14.5
NOMINAL A/F RATIO
100
80
60
40
20
Z
o
o
o
Z
<
loor
80
cc 60-
z
o
(J
§ 40
%
o
20
VHSV 100,000 hr
.0
V
13.5
14.0
8-
-I'OO
14.6
NOMINAL A/F RATIO
80
60
-40 O
z
20
16.0
100
13.5
14.0 14.5
NOMINAL A/F RATIO
15.0
13.5
14.0 14.6
NOMINAL A/F RATIO
160
FIGURE 17 Effect of Pb and S on Stabilized, Supported Ru
Catalysts for Removal of NO from a Synthetic Exhaust in an
Oxidizing and Reducing Condition. Test Conditions on
Laboratory Unit: 500°C and VHSV as Noted. Engine Aging
Conditions on Dynamometer and Fuel as Follows: a) Virgin
Catalyst; b) Lead Sterile Fuel for 60 hr, 10% Oxidizing
(15/1 A/F); c) Durability Fuel (0.035 g/gal Pb, 340 ppm S)
for 60 hr; and d) Durability Fuel for 60 hr, 10% Oxidizing
(15/1 A/F) Plus 8 hr Oxidizing (16.4/1 A/F).
REF. 4
-------
TABLE 14
Selectivity Characteristics of Three-Way Catalysts
Engine Dynamometer Activity Test (1200°F)
Supplier
Steady State Selectivity
Ford Code
Fresh
Aged 100 Hours
% Gross !
%
Inlet
% Gross
%
Inlet
NOx !
nh3
% CO
7o HC
C0/02
NOx
nh3
% CO
% HC
C0/02
M152C3-1*
78
2
78
78
1.5
35
0
35
35
1.6
M152C3-1 & 2**
98
0
86
90
1.0
80
0
74
66
1.8
M196-19*
80
2
80
68
1.4
64
3
63
66
1.3
M196-19 & 20**
92
0
84
80
1.8
85
2
82
86 '
1.7
M226L1-1*
92
0
84
82
1.2
16
1
15
15
1.6
M226L1-1 & 2**
90
0
89
92
2.6
56
0
52
60
1.4
M227L1-1*
72
2
70
58
2.2
22
6
16
16
2.2
M227L1-1 & 2**
80
0
70
80
2.0
-
-
-
-
-
DN-1*
84
2
80
74
1.0
53
0
53
53
1.2
DN-1 & 2**
82
0
82
82
1.8
78
0
81
73
1.2
Ui
*Single Element Converter (130,000/Hr Space Velocity)
**Tandem, Two Element Converter (65,000/Hr Space Volocity)
REF. 2
-------
58
the lean side of stoichiometry at the catalyst inlet, this being
accomplished by some with and some without the addition of an air pump.
Future expected improvements in the catalyst will probably involve a
decrease in the amount of noble-metal loading through better dispersion
and/or the addition of amounts of base metals.
NOx catalysts (three-bed systems) - The only serious durability
study now in progress is the Gould GEM-68 system with three catalysts
(an oxygen "getter" - % PTX monolith, a GEM-68 base-metal monolith N0x
catalyst, and a PTX oxidation catalyst) and operated on the rich side
52
of stoichiometric. On a Ford vehicle this system now (7-9-74) has
accumulated 25,100 miles with the last reported emissions as follows:
0.48 (HC), 3.7 (CO), and 0.55 (N0x) in g/mi. Even though the car has
not been "optimized" to give the lowest possible emissions, all emis-
sions curves have been essentially flat, thus indicating no gross
deterioration. Similarly, a Datsun car has been shown to meet the
0.4 NO standards for 25,000 miles. To the Panel of Consultants' knowl-
edge, no U.S. auto manufacturer has tested the GEM-68 getter system.
Three-way catalysts - Development in this area appears to have
diminished continually since two years ago. However, the recent
general concern about both fuel economy and the lack of well-established
N0x catalyst systems has renewed serious three-way development programs.
With a few exceptions, communication channels between the manufacturers
of 0j sensors and of three-way (or other) catalysts do not appear to
be as well developed as might be desired.
Overall durability of HC/C0/N0v systems - With presently developed
NO^ systems (single-bed, dual-bed, triple-bed), the prospect of simul-
taneously meeting the 0.41/3.4/0.4 HC/CO/NO^ standards over a 50,000-
mile cycle are poor. This statement refers to the current status of all
manufacturers visited.
Data typical of performance on small-engine displacement vehicles
29
with dual-bed systems are given in Figure 18. While there is some
difference in the rate of deterioration determined for the various
catalysts evaluated, it would appear that most, if not all, of these
-------
59
FIGURE 18 Durability Test Results of Some NOjj Catalysts for
a Small Vehicle (Datsun P610, 2750 lb). Average Conversion
Efficiency Measured During CVS-CH Test After Maintenance.
Mileage Accumulated on Modified AMA Durability Cycle.
REF. 29
-------
60
systems would exhibit zero NO conversion efficiency well before
attaining 50,000 miles. Corresponding typical durability data for
dual-bed systems on larger displacement vehicles are shown in
2 2 25
Table 15 and in Figures 19 and 20. Again, there is a substantial
deterioration of activity over test mileages much less than 50,000;
2
in Figure 19, the N0x standard of 0.4 g/mi is not met over any por-
tion of the test cycle, although one catalyst (from Nippodenso)
exhibited durability for 24,000 miles as seen in Figure 20.
One extremely important point is that control of the NO cannot
be separated from HC/CO control. For example, an engine tuned to
run fuel rich (as is required for N0x reduction) necessarily increases
the concentrations of HC and CO in the exhaust and thus puts an in-
creased burden on the oxidation catalyst. The effect of NO control
x 25
on the concentration of CO in the exhaust is shown in Figure 21;
similar data are available for HC. This can result in premature
failure of an oxidation catalyst that would be completely adequate
for an engine tuned for HC/CO control alone. Furthermore, should the
oxidation catalyst fail in such a system, the pollution would be
considerably increased over the levels that would occur and be ob-
served as a result of failure in a HC/CO-only catalyst system.
While dual- (or triple-) bed systems may be developed to the
point where they will be effective for 50,000 miles, the prospects are
not bright. Furthermore, the fuel-economy penalty inherent in such
systems due to rich operation, the additional noble-metal requirement
for more than one catalyst bed, the increased demands on the oxidation
catalyst, and the pollution consequences of oxidation-catalyst failure
are substantial hurdles that must be considered for such systems.
For three-way catalyst systems, deterioration of the catalyst
is coupled with deterioration of the sensor, as has been mentioned
previously. Some durability results obtained with a small displace-
ment vehicle (2 liter, 3500 lb inertial weight) using the Bosch
29
sensor-fuel injection system are shown in Figure 22. The major
-------
61
TABLE 15
Summary of NOx Catalyst Durability for Ford Vehicles
A total of eleven monolithic NOx catalysts have been durability testing
in vehicles. In some of the vehicle tests, independent exhaust systems
were installed on each bank of the V8 or V6 engine allowing concurrent
evaluation of two different NOx catalysts. All data obtained from these
tests are summarized below.
Effective
CVS-CH NOx
Total N0X
Feed-gas
Efficiency
Catalyst
Inertia
CVS-CH
Miles
"0" Miles/
Volume
Weight
NOx
Accumu-
Terminal
Vehicle
Catalyst
(Cu In.)
(Pounds)
(g/mi)
lated
Miles
21A91-D
Ford
84
5,000
0.96
12,000
53/31
429-4V
(M-177B)
A/T
Galaxie
11A55-D
Corning
168
5,000
2.40
21,000
78/44
351W-2V
(M-117)
A/T
Galaxie
Ford
72
2,750
2.20*/
12,000
78*//25*/
(M-177)
22C58-D
Noble Metal
2.6L
A/T
Ford
72
2,750
2.20*/
12,000
79*//15*/
Capri
(M-176B)
Noble Metal
31A73-D
Ford
84
5,000
1.27
12,000
73/20
400-2V
(M-190)
A/T
Noble Metal
Galaxie
31A74-D
AMOCO
84
5,000
1.05
12,000
64/32
400-2V
(DA)
A/T
BM/NM
Galaxie
AMOCO
168
5,000
3.08
21,000
77/56
(DB)
11A55-D
Base Metal
351W-2V
A/T
Gulf
168
5,000
2.06
22,000
67/47
Galaxie
(M-166)
110T714
Ford
84
5,000
1.69
25,000
72/40
351W-2V
(M-190)
A/T
Noble Metal
Galaxie
110T718
Union Oil
84
5,500
0.80
5,000
40/28
429-2V
(M-186)
A/T
BM/NM
T-Bird
14A53
GOULD
84
5,000
3.0
3,000
91/26**
351W-2V
Metallic
A/T
Galaxie
**Catalyst Failed Physically
REF. 2
-------
62
MILES MILES
FIGURE 19 Durability Data on Ford's NOjj Catalysts. CVS-CH
Emissions versus EPA Mileage. Concept Package D, Vehicle
11A55-D, 351W-2V Automatic Transmission Galaxie. HC/CO
Catalyst: Aged M-21 (21,000 Miles). Solid Line: CVS-CH,
Dotted Line: CVS Hot. Top Graphs: N0X Feed Gas 3.2 g/mi,
Bottom Graphs: NOjj Feed Gas 2.1 g/mi.
REF. 2
-------
FIGURE 20 Durability Data on General Motors' N0X Catalysts.
Emission Durability Test Results, Dual Catalyst Emission
Control Systems. 1975 Federal Test Procedure.
REF. 25
-------
64
2.5
2.0
1.5
1.0
® ®
I- ®^
® ©
®
S®
® ® ®
®
®® ®®
S
#> ® ®
i- ®®
®
®
®
0
0
0
0
0
0
50
100
® 1975 Systems
0 1978 Systems
0
0
150
CO (g/mi)
FIGURE 21 Effect of N0X Control in Advanced Emissions Systems
on CO Concentration.
REF. 25
200
-------
65
1.0
0.8
0.6
0.4
0.2
19,542-km Run
500 - 660 C
0.5
1.0
1.5
2.0
2.5
500 - 660 °C
/X
0.5 [CO) +0.5 [H2] + 1.45 |HC]
FIGURE 22 Durability Test Results of 0o Sensor. Evaluation
Conditions: Temperature 650°C (at A), 2,000 rpm, 5.3 kg m.
Test Conditions: Modified AMA Temperature Durability Mileage
Accumulation, Temperature Indicated at Points A, B, and C.
REF. 29
-------
66
effect of deterioration on three-way catalysts is a shifting and
narrowing of the operational "window" with increasing operation. This
is illustrated in Figures 23 and 24 for mixed base-noble metal and
29
noble-metal catalysts, respectively. It is apparent in both cases
that the operational "window" of the fresh catalyst with respect to
the air/fuel ratio is considerably altered; further, it is altered
in different ways for the CO, HC, and NO^. In fact, one would con-
clude from the particular data shown here that the "window" is effec-
tively destroyed by deterioration of the catalyst.
Correspondingly typical data for the alteration of the sensor
29
response is shown in Figure 22. Sensor output becomes much less
sensitive to A/F ratio, and decreases in absolute magnitude, with age.
For this reason, most manufacturers consider, with current technology,
the effective life of the sensor to be about 15,000 miles. Even if
the catalyst were still active, the sensor would require replacement
about once a year. However, the device is not expensive and can
probably be replaced with the same ease as a spark plug.
While the preceding statements refer to the "average" state of
development of three-way catalyst systems, the Robert Bosch Company
from Germany has recently indicated significant advances in this
field. They claim to have catalysts and sensors that will last over
25,000 miles. General Motors has now made similar claims. In light
of these developments, the three-way catalyst approach is showing much
more promise than it did two years ago.
-------
67
0.5 CO + 0.5 H2 + 1.45 HC
14 14.5 15
A/F RATIO
FIGURE 23 Evaluation Test Results of Mixed Noble-Base Metal
Three-Way Catalyst on Monolith, Steady State. Catalyst:
GC 301; Size: 4 in. Dx3 in. Lx4 in.; Durability Mileage:
AMA, 25,100 Miles (on Engine Dynamometer); Test Conditions:
Inlet Gas Temperature - 600°C, Inlet CO Concentration -
500 ppm; Space Velocity: 60,000 hr~l.
REF. 29
-------
68
0.5 CO + 0.5 H2+ 1.45 HC
14 14.5 15
A/F RATIO
FIGURE 24 Evaluation Test Results of Noble Metal Three-Way
Catalyst on Monolith, Steady State. Catalyst GD 306; Size:
4 in. Dx3 in. Lx4; Durability Mileage: AMA 27,400 Miles
(on Engine Dynamometer); Test Conditions: Inlet Gas
Temperature 600°C; Inlet NO Concentration 500 ppm; Space
Velocity: 60,000 hr"^.
REF. 29
-------
V. MATERIALS SUPPLY AND TOXICOLOGY
Catalysts that will be used for exhaust-pollution control in over
half of the new automobiles sold in the U.S. beginning in 1975 will in-
clude a small amount of noble metal (specifically platinum, palladium,
rhodium, and/or ruthenium) sparsely distributed on a refractory oxide
support (e.g., alumina, cordierite, etc.). Other systems currently
under development make use of base metals (e.g., nickel, copper,
chromium, and/or cobalt) deposited on a wire mesh made of a high-
temperature, nickel-chromium alloy support (or on a refractory-oxide
support) for NO^ control.
Some of these components are not produced domestically and must
be imported. Furthermore, the quantities of materials required for
this new application will in some cases be comparable to the total
amount presently consumed in all uses. Almost all potential catalytic
components are included in the government's list of strategic and
26
critical materials. This raises the issue of possible materials
shortages, the first topic discussed in this chapter.
Concern has also been expressed about possible health effects
of catalytic components should they be displaced from the converters
and become transported into the environment. Little information ex-
ists regarding the toxicology of noble metals, but what is known is
summarized in this chapter.
Finally, the effects catalysts may have on nonregulated pollutants
(e.g., sulfates, nitrous oxide, and ammonia), and the associated environ-
mental implications are discussed in this chapter as well.
A. Materials Supply
Noble metals - Within the U.S. there are no known ore deposits
containing sufficiently high concentrations of noble metals that can be
mined profitably. Because of this, essentially all noble metals used
here must be imported. A recent review by the Southwest
Research Institute^ adequately summarizes the world supply/demand
69
-------
70
situation for the two principle noble metals (Pt and Pd) considered
for automobile application. The findings from that study are contained
in Tables 16a-e. The scenario in the last table gives a reasonably
realistic estimate of the future demands for noble metals through 1990,
assuming the emission-control catalysts will continue to be used on
U.S. automobiles. Should other countries adopt such uses, these esti-
mates would need to be increased. However, should base metals replace
noble metals as catalytic components, the figures would be too high.
As indicated in Table 16a, the two major producers of noble
metals are South Africa and the U.S.S.R. with Canada a distant third.
However, the relative amounts of noble metals in ores from the countries
are quite different, as indicated in Table 17. While South Africa is
mainly a platinum producer (Pt/Pd = 2.5), the U.S.S.R. produces pri-
marily Pd (Pt/Pd = 0.5); in Canada the two metals are produced in
about equal quantities.
Most platinum-group metals now imported into the U.S. come from
South Africa, and it is not expected that this pattern will shift
significantly in the next five years. For their projected needs in
catalytic converters, the major U.S. automobile companies have separate
contracts directly with the South African producers, although some Pd
for automobiles will initially come from the U.S.S.R. General Motors
has contracted with Impala for about 420,000 oz noble metals/year for
10 years, and Ford has a similar agreement with Engelhard Industries
to provide 380,000 oz/year for three years (with an additional 360,000
oz in 1978) from the Rustenberg mines. Chrysler will purchase their
noble metals from Impala in South Africa, although the Pd and part of
the Pt for their 1975 cars has come from the U.S.S.R.
The catalyst-equipped Chrysler cars will use monoliths relatively
heavily loaded with from 0.08-0.16 oz noble metals/car in a 70/30 Pt/Pd
mixture. Ford also plans to use monoliths with a variety of loadings
ranging from 0.03 to 0.07 oz noble metals/car in mixtures that contain
66/34 Pt/Pd and 93/7 Pt/ (some other noble metal, probably Rh). General
Motors will use pelleted catalysts in two sizes of converters, each
-------
71
TABLE 16
Supply/Demand Picture for Platinum and Palladium
a. WORLD PRODUCTION OF PLATINUM
& PALLADIUM
Average Annual, 1969-1972
(Thousands of Troy Ounces)
1969-1972 Average
' Source
i
Platinum Palladium
Total
Canada
188
187
375
Colombia
26
-
26
Ethiopia
0.3
-
0.3
Finland
0.2
0.3
0.5
Japan
3
5
8
Philippines
0.6
1
1.6
Republic of South Africa
909
321
1230
U.S.S.R.
665
1339
2004
United States
9
13
22
TOTAL
1801
1866
3667
Sources: Minerals Yearbook and Engineering and
Mining Journal
•
WORLD CONSUMPTION OF
PLATINUM & PALLADIUM
1971
(Thousands of Troy Ounces)
Nation
1 Consumption
United States
1376
Japan
758
U.S.S.R.
589
West Germany
451
France
313
Italy
. 74
Canada
68
United Kingdom
49
Netherlands
33
Sweden
17
Switzerland
12
TOTAL i
3740
c. PLATINUM AND PALLADIUM SALES TO
U.S. INDUSTRY
1967-1971 Averages
(Thousands of Troy Ounces per Year)
Industry Category Plati-
Source: Minerals Yearbook.
num
Electrical 97
Chepical j 155
Petroleum Refining 184
Dental and Medical! 23
Glass 49
Jewelry and 31
Decorative ;
Miscellaneous
TOTAL
Palla- Total
dium
29
568
Source: Minerals Yearbook.
REF. 54
-------
72
TABLE 16 (continued)
"' " " " " ' " """ i
d. WORLD RESERVES PLATINUM & PALLADIUM
1970 Estimates
(Thousands of Troy Ounces)
Reserves
Nation Plati- | Palla-
] num [ dium
Republic of South Africa | 142,400 50,200
U.S.S.R. 60,000 120,000
Canada 6,940 6,860
Colombia 5,000
United States , 950 1,960
TOTAL 215,290 179,020
I
Source: Mineral Facts and Problems, 1970.
Total
192,600
180,000
13,800
5,000
2,910
394,310
e. ESTIMATED FUTURE WORLD DEMAND PLATINUM & PALLADIUM
(Thousand of Troy Ounces per year)
Total Demand
United States:
Rest of world;
Grand total:
base platinum
automotive catalysts
total platinum
base palladium
automotive catalysts
total palladium
total base
total automotive
Total
platinum
palladium
Total
platinum
palladium
Total
1971
I
1980
1990
1
i 541
734
1044
I
i
774
866
541
1508
1910
' 760
898
1095
!
332
371
760
1230
1466
; 1301
1632
2139
1106
1237
1301
2738
3376
1283
1826
2703
1163
1655
2450
2446
3481
5153
1824
3334
4613
1923
2885
3916
3747
6219
8529
Source: Mineral Facts & Problems, 1970; A Look at Business in 1990
(a Summary of the White House Conference on the Industrial
World Ahead. February 1972_)^ and SwRI. _
REF. 54
-------
73
TABLE 17
Estimated Composition of Noble Metals from Different Sources
Percentage by Weight
Metal
Canada
U.S.S.R.
South Africa
Platinum
43.4
30
60
2
2
6
64.02
Palladium
Iridium
Rhodium
42.9
2.2
3.0
25.61
0.64
3.20
6.40
0.13
Ruthenium
8.5
Osmium
REFS. 54,55
with 0.05 oz noble metals/car in a 71/29 Pt/Pd mixture. American
Motors will use similar converters on their cars. On the average, it
appears that cars with oxidation catalysts in 1975 will contain between
0.05 and 0.07 oz noble metals/car. Probably about 75% of the domesti-
cally produced cars will be equipped with oxidation catalytic converters,
which means about 410,000-580,000 oz noble metals/year on cars plus an
(estimated) additional 150,000 oz for replacement parts.
As can be seen from Table 16d, the world reserves of noble metals
are sufficient to meet the projected needs for many years. Further-
more, the mining capacity in South Africa is being expanded to cover
54 55
the increased demand, as shown in Table 18. ' It thus appears that
there will be sufficient quantities of noble metals available for the
automobile applications. However, the price for these metals will un-
doubtedly increase, and this will cause higher prices for other products
(e.g., electrical components) that also involve their use. Finally,
the extended dependence on nondomestic sources for these materials is
a source for concern.
As noted in Chapter 3, ruthenium has been suggested as a possible
(although rather unlikely) catalyst for N0x reduction. From Table 17
it is apparent that much less ruthenium is produced than are Pt and Pd.
However, since there are no major uses for ruthenium today, significant
-------
74
quantities (estimated to be in the hundreds of thousands of ounces, but
exactly how much is not disclosed)"^ of the metal have been stockpiled.
Also, should Ru be used in the catalysts, its high effectiveness (pro-
viding it can be stabilized) makes it necessary to use only l/10th
to l/4th as much per car as are required for Pt and Pd.^ Thus, from a
materials supply point of view, Ru is also a possible candidate for
automobile applications, and its use might help reduce the stockpile
for which there is currently little demand.
TABLE 18
Growth Potential for South African Platinum and Palladium
(1000 troy oz)
1972
1973
1974
1975
Platinum
Rustenburg
550
900
1,300
1,500
Impala
345
400
500
650
Western Platinum
93
125
160
200
Atok
12
15
30
200
Total Platinum
1,000
1,440
1,990
2,550
Total Palladium
355
550
700
890
Source: E/MJ, March 1973
More recent figures in Ref. 57 indicate
Pt will increase from 2,100,000 oz in
1974 to over 2,700,000 oz in 1975.
REFS. 54,55
Rhodium is the only other potential noble-metal candidate for
use in catalytic converters. It is produced in quantities even lower
than Ru (Table 17), and it is used as an alloying agent with Pt to
increase its hardness, in Pt/Rh thermocouples, and with Pt as an alloy
for catalytic oxidation of NH^ to produce nitric acid. While the
initial use as an exhaust-oxidation catalyst is small (and it could
probably be replaced with Pd without significant problem), the
material does show considerable promise for use in the three-way
2
catalysts and also for NOx reduction. There are more potential
-------
75
supply problems for Kh than for the other noble metals, and this will
probably limit its use in catalytic converters to that of a minor
component.
In some cases it may be possible to regenerate noble-metal
4 11
catalysts that have been poisoned by lead. ' The techniques sug-
gested involve flowing through the converters a chemical solution that
will specifically dissolve the lead compounds but that will not dis-
solve the catalyst or converter housing. Host of the lead can be
removed and the catalytic activity significantly increased by this
process. However, as currently formulated, the technique would involve
removing the converter from the automobile for regeneration. Only
laboratory tests have been conducted to date; the process has not been
field tested.
In summary, supplies of Pt and Pd, although not available domes-
tically, appear to be adequate to meet the needs of catalytic converters
for the next several years. The same is true of Ru, should it be used
in small quantities. With Rh there are potential supply problems, and
this metal will probably not be extensively used.
Base metals - As explained in the preceding chapter, base-metal
catalysts (whether reduced or oxidized) are very susceptible to
poisoning at low temperatures by sulfur in the fuel. If the fuel
sulfur level could be substantially decreased,there is a good possi-
bility that these materials could be effectively substituted for noble
metals, at least for limited mileages.
Among the primary contenders are Cu, Co, Ni, and Cr. It is
estimated that the catalysts would require no more than two-to-three
pounds of any single one of these metals, which could mean between 20
19
and 35 million pounds/year at most. Gould has estimated that 28.8
and 32.2 million pounds (for two different formulations) of Ni per
year would be required if all new cars in the U.S. were to be equipped
with their N0x catalysts. According to their figures, this would
amount to 5.8-6.5% of the total Ni currently consumed in the U.S.
26
The consumption figures in Table 19 indicate that only for Ni
and Co does there appear to be a potential materials supply problem.
-------
76
TABLE 19
Estimated U.S. Consumption of Selected
Base Metals, 1971
Metal Millions of Pounds/year
Copper3 5906
Nickel 260
Chromium3 2300
Manganese 4000
Cobalt 11
Vanadium 9
Tungsten 800
Rare earths (as R203)^ 12
a Figures include both scrap and primary;
all others represent primary metal use only.
k Harris and Hannay, 1970, ref. 58; all other
figures from Mineral Facts - 1971.
REF.
-------
77
At present, automobiles contain about three pounds of Ni/car, most of
that being utilized in construction of bumpers. Actually, the concern
about base-metal shortages stems not from use in exhaust catalysts per
2
se, but from use in the catalyst containers. Ford has expressed
concern about the availability of 409 stainless steel and Inconel wire
used in the converters.
A few years ago there was some optimism expressed concerning the
potential of perovskite-type materials containing rare earth metals as
possible substitutes for noble metals. At a level of 1 lb/car, such
58
usage would equal the total use now observed. This could again pre-
sent supply problems. Tests to date have not shown acceptable per-
formance of such materials.
In summary, use of base metals as catalysts in automobiles will
probably not increase the consumption of any one metal by more than 10%,
and increased production could probably easily handle the new demands.
However, their use in construction of high temperature-resistant cat-
alyst containers is a more serious problem.
Supports - The active-metal, automobile-exhaust catalysts are
dispersed on oxide supports in the form of pellets (cylindrical or
spherical) or honeycomb-shaped monoliths. The pellets are mainly
gamma alumina that may contain a "stabilizer" such as MgO, CeO-, and
59
Na20. In order to avoid having to certify too many different types
of catalysts, the automobile companies (mainly General Motors) have
made these catalysts as similar as possible, even though they are
produced by different companies. The chemical and physical proper-
ties of alumina depend quite strongly on the method of preparation.
The physical strength, pore-size distribution, density, level of im-
purities, etc. are all important parameters for the automobile appli-
cation, and they can vary markedly from one manufacturer to another.
At present, the most popular source of alumina for automobile catalysts
is Rhone-Progil, a French manufacturer.
The ceramic monoliths, on the other hand, are all made by do-
mestic manufacturers (Corning and American Lava), and they are mainly
-------
78
cordierite (see Table 2) either extruded or preformed in a paper binder
that is subsequently burned away. The channel walls of the monoliths
are then coated with a high surface area gamma-alumina wash coat (about
10-20 wt%) upon which the active metals are dispersed.
The alumina used in this application will represent only a small
fraction (much less than 1%) of the total material used in the U.S.
60
today (about 8 million tons). The stabilizer technology still is
proprietary, but neither the stabilizer nor the alumina should pose a
materials-supply problem. However, the price will probably increase
due to the recent increased cost of imported bauxite from which alumina
is prepared.
While materials are readily available for the monoliths, this is
a relatively new technology. Each of the monolith vendors is in the
process of completing (or has already completed) the necessary plants
to meet the contractual obligations in this area.
Other support materials that have been suggested (either singly
or in mixture) are titania, zirconia, thoria, zinc oxide, magnesia,
silica, silicon nitride, etc., but none of these materials is being
seriously considered at the present.
B. Toxicology
Debris from catalysts - Loss of catalytic material from the auto-
mobile converters can be attributed either to physical removal through
attrition or to chemical processes by forming a compound with suf-
ficiently high vapor pressure that it can be volatilized.
2
Ford data with monolith catalysts indicate that within experi-
mental error there is no measurable loss of noble metal, wash coat, or
ceramic substrated under normal operating conditions. With pelleted
catalysts there is slightly more material loss (presumably due to
attrition), but even here the amounts are exceedingly small. Using
radioactive techniques to increase analytical sensitivity, General
61
Motors was able to show that at speeds less than 30 mph with stops,
the catalyst was lost at a rate of 2-3 mg/mi, but the noble metals
-------
79
were displaced at a rate of only 1-2 ug/mi. Steady-state measurements
at 60 mph were slightly higher, the corresponding numbers being
5 rag/mi and 3 ug/mi. Since the catalyst contained 0.06 wt% Pt, the
ratio of Pt loss to total catalyst loss 0.0006) indicated that both
were being lost at the same rate and that the noble metal was not
being lost preferentially. This amounts to less than 10% weight loss
for 50,000 miles, although the volume loss may run as high as 15% due
to shrinkage. Actually, these attrition tests were conducted on
"fresh" catalysts with the results extrapolated to 50,000 miles. The
true particulate emission may be less than these values, because the
catalyst would be expected to have more attrition when new than after
"aging." On the other hand, an "aged" catalyst may have a larger void
volume and thus experience more movement of the particle than would
6 2
occur with a "fresh" catalyst. Recent tests by Exxon (although with
63
much less sensitive analytical methods) and EPA produced numbers of
the same order of magnitude as those obtained by GM.
Concern has also been expressed about the chemical and physical
properties of these attrited particles. Using a condensation nuclei
64
counter, Balgord detected an increase in particulate matter in the
size range of 0.001 to 0.1 micron from the effluent streams of several
base-metal catalysts "containing Cr, Ni, and Cu under conditions of
temperature and gas composition similar to those of catalytic auto-
motive emissions-control systems." Particles of this size are in the
respirable range and could possibly have physiological activity in
the lungs. No tests were reported with noble-metal catalysts, however.
61
In their tests with radioactive isotopic tracers, General Motors
showed that no measurable fraction of the noble-metal particulate
matter collected from automobile emissions was soluble in water,
ethanol, carbon tetrachloride, or in hydrochloric acid, which clearly
demonstrated that no soluble Pt compounds were in the debris. Since
it is only the soluble platinum salts, and not the metal or its oxides,
that show physiological activity and are responsible for such condi-
tions as platinosis,^'6^'66 this finding is quite significant.
-------
80
Palladium is considered to be less toxic than platinum in this regard.
Another significant point is that of the particulate matter collected
in the General Motors test, 80% was larger than 125 microns (much too
large to be respired). The Pt particles were probably still deposited
on alumina-dust particles as the two left the converter together.
The alumina should also not have any adverse physiological effects.
In a study of workers in an American Cyanamid Company refinery
who handled catalytic material and were exposed to dust containing Pt
3
near the threshold limit (2 ug/m ), no evidence was found for platino-
sis or other disease that could be attributed to Pt after 10 years'
exposure. These levels are significantly above any that could
possibly occur due to particulate matter from automobiles, and it is
thus concluded that airborne Pt particles emitted from automobile
catalytic converters do not constitute a potential health problem.
Another potential problem involves the fate of Pt emitted from
converters and possibly deposited along roadways, or Pt left to decay
in junked automobiles. There is no evidence that Pt metal can be
solubilized by microorganisms, but this subject has apparently not
6 8
been thoroughly investigated. In fact, Wood has predicted that Pt
and Pd can probably be methylated by microorganisms in much the same
69
way as mercury is methylated, but this has not been demonstrated
at the present. If this did occur, there is the possibility that the
Pt could contaminate water supplies and become incorporated into the
human food cycle. While the small amounts of material involved make
the probability of such creating a problem quite remote, studies should
be carried out to determine if there exist strains of highly specific
microorganisms that can accomplish Pt solubilization.
In addition to physical attrition, volatile species can be
formed by reaction of catalyst components with exhaust gases. Ex-
amples are the formation of carbonyls (e.g., Ni(CO),) or oxides (e.g.,
^ 70
RuO^) that have significant vapor pressures and are quite toxic.
Fortunately, most of these species decompose at high temperatures
that would normally be found in the converters, but during warm-up
-------
81
and cool-down the converters will pass through a transient temperature
range where such formation could occur. While this presumably accounts
for the instability of Ru catalysts to oxidizing atmospheres, the
Panel members are not aware of any evidence for carbonyl formation
in catalytic converters.
Table 20 lists several compounds of potential catalytic ma-
terials and their associated toxicities. However, the small quantities
62
of these materials that escape from the converters make the potential
for hazards in this area fairly remote.
It is therefore concluded that debris from catalytic converters
should not represent a potential health hazard.
Effect on fuel components - The effect the oxidation catalysts
have on the various fuel components will be discussed separately.
1. Phenols, polynuclear aromatics, aldehydes. Gasoline
normally contains small quantities of polynuclear aro-
71
matics, some of which have been identified as being
carcinogenic. As larger fractions of aromatics are
used for octane enhancement, more of these compounds
will be present in the fuels. Since not all the
fuel is burned in the engine, some of these com-
pounds will be emitted from the vehicle, and the
measured amounts have been shown to be proportional
to the amount present in the fuel. Also, certain
partial oxidation products (e.g., phenols and
aldehydes)7* have also been identified in the ex-
haust gases; such compounds are eye irritants and
contribute to smog formation. Oxidation catalysts
are extremely effective for removing all these
compounds, their levels being reduced by more
than 98%.2'25,71
2. Particulate matter. Numerous tests have shown
that particulates from the exhausts of unregu-
lated automobiles are due mainly to the presence
-------
82
TABLE 20
Toxicity of Some Compounds That Might Be
Used in Automobile Emission Control Catalysts
Maximum Allow-
able Concentration
Acute Chronic in Air
Compound
Local
Systemic Local
Systemic
(mg/M^)
Lead compounds
0
3 0
3
0.15
Chromium compounds
3
U 3
3
0.1
Nickel compounds
1
1 2
2
0.5
Nickel carbonyl
3
3 1
3
Manganese compounds
U
2 U
3
15.0
Copper compounds
1
2 1
1
Metal carbonyls
3
3 U
3
Alumina
1
0 2
0
Silica
2
0 3
1
Cobalt compounds
1
1 1
1
0.5
Tungsten compounds
U
1 U
1
Magnesium compounds
1
2 2
0
15.0
Ruthenium compounds
(Details unknown, but probably toxic) 0.01
Ruthenium tetroxide
2
U U
U
Platinum compounds
(very low toxicity)
Palladium compounds
(very
low toxicity)
Mercury
0.1
Source: &ax, 1963 (reference
70)
Note: 3 very toxic
2 moderately toxic
1 slightly toxic
0 nontoxic
U unknown toxicity
REFS. 26, 70
-------
83
of lead in the fuel.^'^'^ Removal of lead motor mix
causes a reduction by over a factor of 10 in particulate
2 25
matter. ' However, a small increase (over the
lead-free case) occurs in emission of particulate
matter when an oxidation catalyst is present (using
unleaded gasoline), but this is still far below the
leaded gasoline-no catalyst case. The increase is
due mainly to sulfate formation, as discussed in
the next section. Overall, particulate emission
will be considerably reduced with the use of cata-
lytic converters with unleaded gasoline.
3. Sulfate emission. Sulfur normally present in fuels
(about 0.03 wt7o) is emitted from engines mainly as
S02> which is slowly oxidized (depending on atmospheric
conditions) in the air to SO^. This gas reacts with
water to form sulfuric acid aerosols that can react
with airborne ammonia (or other basic compounds) to
produce particulate sulfates, compounds that may
have adverse physiological effects if breathed.
Oxidation-catalytic converters cause a part of the
S02 to be oxidized to S0^ before it leaves the car,
and concern has been expressed that this may result
in dangerously high sulfate levels near freeways
even though sulfur in gasoline represents less
than 1% of all sulfur emitted into the atmosphere
72
by man-made sources in the U.S.
Catalysts that are active for HC and CO
oxidation also oxidize S0£ to SO^. (It may be of
interest to note that the first patent In catalysis
involved use of Pt for S09 oxidation in the "con-
73
tact process" for sulfuric acid manufacture.)
The percentage of SO2 that can theoretically be
converted depends on the temperature and oxygen
-------
84
4
partial pressure, as shown in Figure 25. Lower
temperatures and higher oxygen concentrations
favor the oxidation. Furthermore, the catalysts
(depending on type, whether pelleted or mono-
lithic) have a capacity to retain (or "store")
72
considerable quantities of sulfur compounds,
particularly at low temperatures, and this
material is subsequently desorbed (or "dumped")
from the catalyst at elevated temperatures.
These factors, coupled with some uncertainties
about the reliability of analytical procedures
for sulfur compounds, make it impossible to pre-
dict exactly how much SO^ will be formed without
considerable knowledge about the system param-
eters and the catalyst's prior exposure to
sulfur.
2 72
Data from Ford and General Motors indi-
cate that there is almost no SO. emitted from
38
the engine, although Chrysler and EPA have
shown that up to 15% of the sulfur is emitted
as SOg. In any case, the oxidation of SC^
is not very high in the absence of a catalyst,
even with air injection into the manifold.
In oxidation catalyst-equipped cars, the
oxidation figures range from only a slight
72 2
enhancement to 84% conversion (based on
the S0j/S02 rat*-° tlie tailpipe and disre-
garding any storage effects). Inmost cases
the conversion achieved is in the range of
10-25% without air injection to 30-60% with
air injection. Usually the observed con-
versions are well below the thermodynamic
equilibrium values, as seen in the Ford data
-------
85
iooL
9oL
«0.
2+0.5.
80
•Son
C9
s
SQL
40L
%b
¦?"» ^
- 3.o
¦ IB
0.S
301
20i
JO
300
40o~
*Te
XZP.
2S
e®sure
s1tin
ib
'**um
-------
86
in Figure 26. The fractional conversion is generally
independent of fuel sulfur content, although the frac-
tional conversion decreases as most catalysts are aged.
There is much less storage of sulfur on monolithic cata-
lysts than on the pelleted forms. The larger amount of
alumina in the pellets makes this result not unexpected
2
if the sulfur is held as Al^(SO^)3 which decomposes
at high temperature to release SO^.
Neglecting any storage effects and assuming a car
gets 15 mpg on fuel with 0.03% by weight S, about 0.07 g
S will pass through the engine per mile. If all this S
were burned to SO2 in the engine, oxidized to SO^ in the
catalytic converter, and transformed into sulfuric acid
(H^SO^) before leaving the tailpipe, then this would
amount to about 0.2 g H^SO^/mi. In view of the partial
conversion that actually occurs in the converter, it
seems reasonable to assume that on the average there
would be about 0.05 g sulfur compounds (calculated as
H^SO^) emitted per mile under steady state conditions
(i.e., after the catalyst has become saturated with
sulfur under the conditions used for the test).
The 1972 Federal Test Procedure (CVS-C, which in-
volves only the "cold start" portion of the cycle) can-
not be used as a reliable measure of the total SO^ (or
H9S0. or sulfate) emissions because of the sulfur storage
25
problem. As may be seen in the GM data in Figure 27,
measured sulfur emissions for 0.03% S fuel fall between
0.001 and 0.020 g/mi for various converter types and
operational modes; these values are considerably lower
than the steady state value.
Several modeling studies have attempted to predict
the concentration of sulfates that might be expected from
various assumptions and meteorological conditions. The
most popular approach has been to assume that sulfates
-------
87
100
90
80
70
60
50
40
30
20
10
550
O Catalyst #1, Fuel S = 0.031 wt%
~ Catalyst #2, Fuel S = 0.059 wt%
Thermodynamic
Equilibrium Line
s
\
\
~
X
~
\
~ \
~
V
600
650
700
CATALYST TEMPERATURE (°C)
FIGURE 26 Observed Conversion of SO2 to SOo Compared with
Thermodynamic Equilibrium Curve for SC>2 to SO3.
KEF. 2
-------
88
0.143
0.09 r-
0.08
0.07 -
0.06 -
0.05 -
0.04 -
0.03 -
0.02-
0.01 -
0.00
0.05 0.10
PERCENT SULFUR IN FUEL
FIGURE 27 Effect of Fuel Sulfur Content on Sulfate Emissions,
1972 Federal Test Procedure.
0.15
REF. 25
-------
89
will be transported through the air near motorways in
much the same way as more stable automobile exhaust
pollutants (e.g., CO or Pb) for which concentration
profiles are available. Mathematical models (called
"surrogate" models) that adequately describe these
known concentrations can then be modified and applied
to predict the behavior of sulfates. A second surro-
gate estimate is based on the observed carboxyhemoglobin
levels (which are related to atmospheric CO levels) in
healthy blood donors from 18 cities in the U.S. By
assuming that all or most of the CO exposure is due
to emissions from automobiles and that a relationship
between CO and sulfate emissions can be established,
it is possible to provide another estimate of sulfate
exposure that might be expected. In all modelsj, both
typical and adverse meteorological conditions were as-
sumed, and sulfate concentrations were than predicted
for roadside and nearby locations. Some of the results
72a
are summarized in Table 21, which has been based on
only two model years (or 25% of the vehicle-miles
driven) with cars having catalytic converters. The GM
data, based on measured SO^ emissions under non-steady
state conditions where sulfur retention is significant,
3
showed 24-hour exposure values from 0.7 to 3.6 ug/m
3
at roadside to 0.4-2.1 ug/m 10 meters from the road.
Actually, these values should all be multiplied by a
factor of 8 to be applicable under steady state condi-
tions, which brings the predicted sulfate concentrations
3
into the range of 3 to 30 ug/m under adverse conditions.
These numbers are similar to predictions of other models
shown in Table 21. Under very adverse atmospheric con-
ditions, the various models predict peak concentrations
3
could go as high as 124 ug/m . Even though such a
-------
90
TABLE 21
Modeling Studies of Sulfate Concentrations
Maximum 24-hour Sulfate Exposures
(Based on 25% of vehicle-miles on catalyst-equipped cars)
Investi-
gator
Estimation
Method
Atm.
Assumed „ _ _ .3
„ . . H„S0, Cone., ug/m
Emissions 2 4 ' °
Condition fug/mile) Roadside 1Q Meters Away
GM Gaussian
90° Wind
Max angle
GM St. Canyon
Expr.
Art. Hwy.
GM Pb tracer
GM CO tracer
Ford Pb Surrogate
EPA CO Dispersion
EPA CAMP Station
EPA Pb Surrogate
EPA COHb
Adverse
Adverse
Adverse
Adverse
Adverse
Adverse
Max. CO
0.007
0.007
0.007
0.007
0.007
0.007
0.05
0.05
0.05
0.05
0.9*
3.6(3°)*
1.6*
0.8*
0.7*
5.0
14.5**
0.7*
2.1(5°)*
0.8*
0.4*
0.4*
0.7*
11.1**
5.3
1.7-9.2
1.0-8.1
*Numbers should be mutiplied by 8 due to low assumed SO^ emission rate
**Could be as high at 124 and 88 for peak adverse conditions
REPS. 72, 72a
-------
91
concentration would only be expected to occur rarely
(if at all), its possibility must be considered in
light of documented adverse health effects these com-
pounds may have on a significant fraction of the pop-
ulation who may be particularly sensitive. The EPA,
10 "K
in its "CHESS" Studies, has concluded that sulfate
(or, more specifically sulfuric acid) concentrations
3
in the range as low as 8-10 ug/m can aggravate symp-
toms of asthmatics and people with cardiopulmonary
diseases. Furthermore, it is not only the mass con-
centrations but also the chemical composition and
physical state of the various compounds that are im-
72c
portant. For example, sulfuric acid is much more
toxic than ammonium sulfate, and particle sizes in
the respirable range (less than lyuw*in diameter) do
more damage than larger particles.
It should be kept in mind that all SC^ emitted
into the atmosphere (less than 1% actually comes
from automobiles) is eventually converted to sulfates
with the rate of conversion depending strongly on
the atmospheric conditions. Thus, the overall sul-
fate loading in the atmosphere will not be altered
by the introduction of catalytic converters. How-
ever, the S0„-H„S0,-sulfate concentrations near
3 2 4
motorways will almost certainly be increased by
use of catalytic converters, and in some cases,
these increases may be substantial. It is there-
fore extremely important that carefully planned
tests be initiated to monitor the sulfate concen-
trations in areas where harmful levels may develop,
such as in downtown street canyons, near arterial
-------
92
freeways, or in shopping centers. If the sulfate
concentrations approach harmful levels, then steps
should be taken to reduce the sulfur content in
fuels and/or modify (or eliminate) converters in
order to avoid street-level formation of SO^.
In practice, the problem may not be quite so
bad as the numbers in Table 21 indicate. As men-
tioned earlier in this section, the SO^ to SO^ con-
version activity decreases considerably as the
catalysts age, and this would lower the predicted
sulfate concentrations. Furthermore, the additional
refining needed to produce lead-free gasoline will
reduce the sulfur content and thus lower the over-
35
all sulfur output. On the other hand, "dumping"
of the accumulated sulfur in congested areas could
even intensify the problem, but this is not very
likely to be the case.
4. Other pollutants. There have been reports that
under a very limited set of conditions (slightly
reducing atmosphere, such as may occur during down-
hill coast), H^S is produced. This was first no-
ticed by Chrysler personnel who smelled the gas.
While the levels are well below those that repre-
sent a health hazard, the odor could be a mild
annoyance. Also, with high phosphorous contents
there is the possibility of producing highly toxic
38
phosphine, but the current phosphorous levels
are much too low to form a basis for concern.
Neither ammonia nor nitrous-oxide (^0) emis-
sions will occur with oxidation catalysts. Besides,
both are present in the atmosphere at levels far
above those that would be emitted from cars, so
neither appears to present a significant addition
to current pollutant levels.
-------
93
Animal tests - To examine directly the health effects of ex-
74
haust catalyst-equipped cars, several experiments were carried
out with Sprague-Dawley rats in special chambers through which diluted
exhaust could be passed. Animals subjected to the catalyst equipped
exhaust could not be distinguished from those breathing purified air,
while animals subjected to noncatalyst exhaust fared much worse. In
the preliminary draft of their report, the experimenters made this
statement:
It can be concluded that the introduction of catalytic
converters into the automobile exhaust system not only
has reduced the levels of certain exhaust constituents
but has effectively decreased or eliminated (harmful)*
biological effects studied.74
However, the conclusions of this report have been challenged
74a
because of the type of animals used in the study. Treon et al.
have shown that rats are not very sensitive to exposure to sulfuric
acid; in fact, animals follow the increasing sensitivity order
rabbits < rats ^mice
-------
VI. KINETICS, MECHANISMS, AND MODELING
A. Introduction
Information on kinetics and mechanism, and reaction and reactor
models developed from that information, are normally the product of
extensive basic research programs. While considerable information is
available concerning the kinetics and mechanisms of oxidation reactions
of HC and CO on the noble metals to be employed as emission-control
catalysts, very little of it is applicable to the conditions peculiar
to these devices. Thus, much of the development work on catalytic
devices for automobile emissions control has been done in terms of
prototype experiments with rather meager knowledge of the chemical
fundamentals.
Fortunately, under most operating conditions, exhaust converters
are transport limited, which means the rates of mass and heat transfer
to and from the catalytic surface, rather than the intrinsic chemical
kinetics, determine the conversion to products. Given this situation,
one would think that reactor modeling would be a fruitful area for
work; in fact, relatively little effort has been expended in this
direction to date. This is due primarily to the difficulty in reactor
analysis of the major catalyst problem, that of long-term durability.
This chapter contains a summary of pertinent information on the
kinetics of CO and HC oxidation and NO^ reduction, together with a
report on the current state of exhaust converter modeling.
B. CO Oxidation
To a first approximation, the kinetics of CO oxidation under
typical automotive exhaust environments can be represented by the
following rate equations
(base metals)
- -kj(o2)/(co)
(platinum)
94
-------
95
In both cases the rate constants and are sensitive to the levels
of water and of hydrocarbons in the reaction mixture, reflecting an
adsorption competition among these species on the catalyst surface.
Carbon monoxide is more strongly adsorbed on transition metal surfaces,
hence the adsorption inhibition noted above for platinum would be
expected to pertain to Pt/Pd formulations as well. At very low CO
levels, the inhibition has been reported to disappear and the rate
to depend on a positive order for CO. In general, rate corre-
lations of the form
^21 = -k(co)x(o2)y
have been found successful for a number of transition metals.
Some typical values are^
Material x 2 Temperature (°C)
Pd wire -2 1 100
Pd foil -1 1 300
Pt wire -1 1
Ag foil 1 1 450
While no experimental Information has been found which would
permit the unequivocal modeling of the adsorption competition
between CO, H^O, HC and 0^, it is not unreasonable to expect that a
Langmuir-Hinshelwood correlation could be developed. This would have
the f°ra K K (CO) (02)
d(C0) = _k C0 2 2
dt (1 + K (Og) + Kcq(CO) + Kw(H20) + (HC))
where the K values are adsorption equilibrium constants. For
strong chemisorption of CO, this form is also compatible with the
power law forms given above.
Conventional interpretation of the mechanism of CO oxidation on
transition metals, based on this kinetic information, suggests reaction
between chemisorbed oxygen and CO, with the active surface mostly
occupied with CO. Recent work, however, indicates that the mechanism
-------
96
at low CO levels is at least partially controlled by a parallel
78
reaction of gaseous CO with adsorbed oxygen atoms. On some base-
metal oxides, especially in those cases where lattice oxygen mobi-
lity is high, the oxidation mechanism involves an oxidation-reduction
cycle of the catalyst, with the active oxygen supplied by the catalyst
and a possible carbonate intermediate. A thorough review of mecha-
79
nistic studies on base-metal oxides is available.
C. HC Oxidation
While CO is more difficult to oxidize than HC in an homogeneous
thermal reactor, it is the latter that is most demanding in a cata-
lytic system. Hydrocarbon oxidation can be conveniently divided into
two parts: oxidation of saturated materials and oxidation of all
other types. Aromatics, olefins, partially oxidized products such as
aldehydes and ketones, are all easily converted to CO2 and water over
most catalysts under mild conditions. However, paraffins behave quite
differently, the ease of oxidation decreasing as the chain length be-
comes less. Except under very harsh conditions, methane is hardly
affected by oxidation catalysts. When catalysts become slightly
poisoned, it is almost always the hydrocarbon conversion that first
begins to decrease.
, Little information is available about the mechanisms of these
extremely complex reactions. Partially oxidized surface species are
undoubtedly involved, but none of these species has been observed in
the exhaust streams after the catalysts. This indicates that once
formed they must be quite rapidly oxidized completely.
D. NO Reduction
x
Unlike the case for CO and HC oxidation, with their long history
in the literature, the catalytic reduction of NO has been studied
most extensively by those with an interest in automobile emission
control. A general review of the behavior of NO in heterogeneous
-------
97
catalytic reactions leading to reduction of the NO has been given by
80
Shelef and Kumnier. In applications to emission control, the re-
duction of NO (considered as an overall process) is very complex due
to the variety of reactions involved, such as
CO + h2o -—* co2 + h2
2C0 + 2NO > N2 + 2C02
5H2 + 2N0 > 2NH3 + 2H20
2NH3 >N2 + 3H2
4NH3 + 6N0 >5N2 + 6H20
2H2 + 2NO *N2 + 2H20
The importance of the water-gas shift (first reaction) in supplying
hydrogen for reduction and the role of NH^ in the overall reduction
are now well recognized and the subject of considerable current re-
search.
The overall kinetics of the reaction sequence outlined above
have not been studied in detail, either on noble- or base-metal cata-
lysts; however, some of the industrial reactions have been investi-
81~83
gated. Kinetic models have been proposed by Peters and co-workers
for reduction of NO with H2> CO and CH^ (reduction to N2), and for
reduction with H2 to NH^, all over various base-metal catalyst formu-
lations containing Cu, Zn, CrjO^, etc. In the former case, the rate
is correlated by
dCNOI _ v KHOKAqPR
dt --1 (1 + vK0 + w2
where R is either co> or CH4* For the reduction to NH^ by H2
PH
llNOi = _k 2 H2
at 2 a +
Some values of the rate and adsorption parameters for these models
are given in Table 22a. Though determined for a particular base-metal
-------
98
TABLE 22a
Parameters for NO Reduction to N2 by H2 on a
Cu-Zn-Cr203 Catalyst (Girdler)
kl = 1.39 exp (-10,300/RT) g moles/min g
Kno = 3.1x10^ exp (-13,300/RT) atm"^
Kh£ = 4.5x10"* exp (-15,460/RT) atm"^
Parameters for NO Reduction to NHg by H2 on a
Cu-Zn-Cr203 Catalyst (Girdler)
T, °C k2xl0^ , g moles/min g ^H2' atrn"^
375 0.77 6.17 16.41
400 2.13 10.32 13.49
425 2.78 11.71 21.6
Note: These parameters do not fit an exponential
temperature relationship, as did the
parameters for reduction to N2.
REP. 81
-------
99
formulation, they are probably good as order-of-magnitude estimates for
related base-metal formulations. The parameters for ammonia formation
vary irregularly with temperature and thus have empirical significance
only.
84
More recently London and Bell have reported the kinetics of NO
reduction with CO on silica-supported copper oxide. They obtained a
kinetic correlation of the form
d (NOT _ _ 2blb2CNO
4t (1 + b2CN0 + b3%0> (1 + b4CC0>
The equation is consistent on a mechanism involving dissociative
adsorption of NO, ^0 acting as intermediate in ^ formation, and CO
maintaining the surface in a reduced state as well as competing for the
sites required for NO dissociation. Typical values of the constants in
the rate equation are given in Table 22b.
Of special interest is the formation of an isocyanate surface
intermediate (NCO) during the reduction of NO by CO. That such a
85 86
surface intermediate is detectable both on noble-metal catalysts '
84
(Pt, FD, Rh, Ir, Ru) and on a base-metal catalyst (Cu) suggests an
interesting interpretation of the mechanism of NO reduction and of ammonia
formation. The existence of an isocyanate surface structure as detected
by infrared spectroscopy under reaction conditions points to dissociative
chemisorption of NO as an essential step in this reaction.
As for the question of NH^ formation, either by reduction of the
nitrogen adatoms with hydrogen species or by hydrolysis of the iso-
85
cyanate adsorbate, the studies by Unland tend to favor the latter
mechanism. On noble-metal surfaces he reports that the tendency for
ammonia formation is related to the (NCO)-forming properties of the
respective metal. In the case of ruthenium, this surface species ex-
hibits a very low surface density as compared with the other metals
studied.
-------
100
Table 22b
Parameters for NO Reduction to ^ by CO
on Silica-Supported Copper Oxide
T C
bl
b2
b3
b4
135
0.011
2403
0
4392
170
0.023
2942
0
2103
200
0.012
2137
0
2098
REF. 84
A most recent study of the reduction of NO by and/or CO
over noble-metal catalysts has indicated a marked difference between
the behavior of Pt and Pd on the one hand and Rh and Ru on the other.
While the reaction of NO with predominates in each of the cases
studied, CO acts as an inhibitor in the case of Pt and Pd. Over Ru,
however, the reaction is accelerated by CO. These observations are
interpretable in terms of the isocyanate surface species acting as an
ammonia precursor and reaction inhibitor, since it competes with NO and
CO for surface sites involving nitrogen adatoms, viz:
N + NO . N„ + 0 (a)
s y 2 s
N + CO » NCO (b)
s s
Reaction (a) leads to formation and CO oxidation by way of the
subsequent reaction
0 + CO »C0o (c)
s I
On the other hand, the NCO surface species are relatively stable
and unreactive under the experimental conditions, except possibly
to an hydrolysis reaction with ammonia as a product.
E. NO Decomposition
The ideal way to eliminate NO from automobile exhausts (or
from stack gases) is simply to decompose it, since NO is thermo-
dynamically unstable except at very high temperatures. All that is
-------
101
required is a catalyst that will somehow activate the N-0 bond. If
this could be realized, NO could be removed under oxidizing condi-
tions rather than requiring a net fuel-rich (or reducing) atmosphere
as seen in the preceding section. Unfortunately, no such catalyst
has been found that will operate at low or moderate temperatures.
Several base metals and noble metals, in their reduced state, will
react quite readily with NO to form however, the oxygen remains
tightly bound to the surface and oxidizes the catalyst. Once the
surface is covered with 0 atoms; the reaction ceases and will not
begin again until a reducing molecule (CO, etc) chemically
removes the interfering 0 atoms. Of course, this then becomes an
example of NO reduction.
87
Amirhazmi, Benson, and Boudart have recently reviewed NO
decomposition. In general, most investigators find the reaction can
be treated by a Langmuir-Hinshelwood equation with strong inhibition
by adsorbed oxygen. Table 23 shows a collection of parameters for
this reaction.
F. Exhaust Converter Modeling
Contacts with both automotive and catalyst manufacturers
revealed, with one exception, no current activity in trying to model
the behavior of exhaust converters, either fixed beds or monoliths.
The single exception is General Motors, where there has been some
effort in modeling deactivation effects in individual catalyst
88
particles. These consider the interrelations between chemical
rates, diffusion rates, and deactivation rates in determination of
the net activity of a given catalyst. They also investigated the
relationship between pore structure and susceptibility to poisoning.
The approach is similar to that reported previously by Masamune and
Smith. ^
The primary interest in modeling of exhaust converters seems
to reside in the academic community. These efforts are largely con-
fined to monoliths. Two papers on the topic, both including transport
-------
TABLE 23
Summary of Catalytic NO Decomposition Kinetics
Tempera- Reac-
ture tion w/
Gas range, Pressure, respect
Authors
Catalyst
Reactor
Mixture
°C
Torr
to NO
Ref.
Fraser and Daniels
Metal oxides
Flow
107= NO
• «. TT _
740-1040
760
0
(2)
Yur'eva et al.
Transition metal
oxides
Recycle
m He
100% NO
250-750
100-380
2
(3)
Shelef et al.
Supported Pt and
oxides
Flow
4-1007=
279-938
760
~1
(4)
NO in He
Winter
Oxides
Recycle
1007= NO
330-870
50-400
1
(5)
Bachman and Taylor
Pt wire
Batch
1007= NO
1210
201-479
2
(6)
Zawadski & Perlinsky Pt-Rh wire
Batch
1007= NO
860-1060
100
1
(7)
Green & Hinshelwood
Pt wire
Batch
1007= NO
882-1450
200-500
1
(8)
Sakaida et al.
Supported Pt-Ni
Flow
0.404 &
427-538
1-15 atm
2
(9)
0.432%
NO in N2
References for table
2. Fraser, J.M. and Daniels, F., J. Phys. Chem., 62, 215 (1958).
3. Yur'eva, T.M., Popovskii, V.V., and Boreskov, G.K., Kinet. Ratal. 6, 941 (1965).
4. Shelef, M., Ot£o» K., and Gandhi, H., Atm. Environ. 3, 107 (1969).
5. Winter, E.R.S., J. Catal. 22, 158 (1971).
6. Bachman, P.W., and Taylor, G.B., J. Phys. Chem. 33, 447 (1929).
7. Zawadski, J., and Perlinsky, G., Compt. Rend. 198, 260 (1934).
8. Green, T.E., and Hinshelwood, C.N., J. Chem. Sco. 129, 1709 (1926).
9. Sakaida, R.R., Rinker, R.G., Wang, Y.L. and Corcoran, W.H. AIChE J. 7, 658 (1961).
REF. 87
-------
103
and kinetic effects and capable of describing thermal transients, are
to be presented at a forthcoming Symposium on Chemical Reactor Engi-
90 91
neering, ' and a review paper summarizing the bulk of work to date
will appear in "Catalysis Reviews - Science and Engineering" in the
92
fall of 1974. Apparently none of these mathematical models of mono-
lithic converters has been tested with experimental data, nor do they
include catalyst deactivation.
All the above citations to monolith modeling, however, pertain
to work that is not published at the present time. There are a few
efforts at modeling that have been publicly described, some of which
93
are summarized here. Young and Finlayson describe a monolith model,
based on CO oxidation alone, postulating laminar flow in each duct,
negligible axial diffusion or conduction, transient response controlled
by thermal response of the solid, net adiabatic operation, and no in-
ternal (pore) diffusion limitation; the model equations were solved
94
by numerical methods. Hegedus has analyzed the effects of channel
geometry on monolith performance, assuming the overall rate was
determined by the rate of mass transfer to the catalyst surface. He
found that channels of elongated rectangular cross section were
preferable in their mass transfer characteristics to other geometries
(i.e., hexagons, circles, squares, etc.). The preference criterion
was based on the minimum length of a monolith required for a specified
conversion level. However, it should be noted that, in mass-transfer
limited operation, facilitated mass transfer implies facilitated
poisoning as well. The implications of this have not been fully ex-
95
plored, although for pelleted catalysts Wei has suggested that an
"egg yolk" catalyst (one with the active component buried beneath the
surface) may be preferable to an "egg shell" catalyst with the active
component located on the external surface of the pellets.
96
Johnson and Chang have carried similar calculations to those
97
of Hegedus, and Carberry and Kulkarni have analyzed the intra-
interphase temperature gradients likely to occur under typical exhaust
operating conditions. Their results indicate that very large
-------
104
interphase (gas to catalytic surface) thermal gradients are possible,
and these may lead to burnout. None of these models incorporate
deactivation phenomena in their description of the converter system.
-------
REFERENCES
1. Furlong, L.E., E.L. Holt, and L.S. Burnstein, "Emission control
and fuel economy," a paper presented at the American Chemical
Society's National Meeting in Los Angeles, April 1974.
2. Data presented to the Panel of Consultants on Catalysts by Ford,
May 6, 1974.
3. Chrysler Corporation's "Application for Suspension of 1976 Motor
Vehicle Standards (Part I)," to EPA, May 1973.
4. Data presented to the Panel of Consultants on Catalysts by Engel-
hard Industries, June 10, 1974.
5. Chrysler Corporation, Material Standard MS 5747, March 12, 1974.
6. Data presented to the Panel of Consultants on Catalysts by Gulf
Research and Development, June 5, 1974.
7. Shelef, M. and H.S. Gandhi. Platinum Metals Rev. XVIII (1974),
2; Ind. Eng. Chem. Prod. Res. Dev., II, (1972), 2.
7a. Kobylinski, T.P., B.W. Taylor and J.E. Young, SAE Paper 74250,
February 1974.
8. McArthur, D.P. "Activity, selectivity, and degradation of auto
exhaust noble metal oxidation catalysts," paper presented at the
Third No. Am. Meeting of The Catalysis Soc., San Francisco,
February 1974.
9. Taylor, K.C., R.M. Sinkevich, and R.L. Klimisch, General Motors
Report GMR-1538, PCP-19, February 1974.
10. Klimisch, R.L. and K.C. Taylor, "Catalytic reduction of NO over
an oxygen-treated Ru catalyst," paper presented at the California
Catalysis Society Meeting, Pasadena, November 1973.
11. Data presented to the Panel of Consultants on Catalysts by Union
Oil Research Labs, April 2, 1974.
12. Mannion, W.A., K. Aykan, J.G. Cohn, C.E. Thompson, and J.M. Mooney,
"Catalytic reduction of NO* emissions in auto exhaust gas," paper
presented at the American Chemical Society's National Meeting in
Los Angeles, April 1974.
13. Data presented to the Panel of Consultants on Catalysts by American
Oil Company, March 28, 1974.
105
-------
106
14. Simpson, H.D. "Performance and durability data for Union Oil
stabilized ruthenium NOx catalysts," paper presented at American
Chemical Society's National Meeting, Los Angeles, April 1974.
15. Jackson, H.R., D.P. McArthur, and H.D. Simpson, "Catalytic N0X
reduction studies," SAE Paper No. 730568, Detroit, May 1973.
16. Burnstein, L.S., R.J. Long, R.S. Lunt, G.S. Meisser, and R.J.
Fedor, "Nickel-copper alloy N0X reduction catalysts for dual
catalyst systems," SAE Paper No. 730567, Detroit, May 1973.
17. Fedor, R.J., C.H. Lee, and M.P. Makowski, "Metallic catalysts,"
paper presented at the Am. Inst, of Chem. Engs. National Meeting,
New Orleans, March 1973.
18. Lunt, R.S., L.S. Burnstein, J.G. Hansel, and E.L. Holt, SAE Paper
No. 720209, 1972.
19. Data presented to the Panel of Consultants on Catalysts by Gould
Inc., June 13, 1974.
20. Bentley, D.R. and D.J. Schweibold, "Questor reverter emission con-
trol system--total vehicle concept," Fuels and Lubricants Div. of
Nat. Petro. Refiners Assn. Nat. Meeting, Houston, September 1973.
21. Questor Corporation Report to the National Academy of Sciences,
March 20, 1974.
22. Giordano, N., E. Cavaterra, and D. Zema. J. Cat. V (1966), 325.
23. Holbrook, L. and H. Wise. J. Cat. XXVII (1972), 322.
24. Haensel, V. and M.J. Sterba, "Energy and emission control,"
presented to EPA, June 25, 1973.
25. Data presented to the Panel of Consultants on Catalysts by General
Motors, May 7, 1974.
26. Wei, J., R.L. Burwell, J.W. Hightower, D.F. Ollis, and H. Wise,
"Evaluation of Catalysts as Automotive Exhaust Treatment Devices,"
Catalyst Panel Report to the NAS Committee on Motor Vehicle Emis-
sions, March 28, 1973.
27. Data presented to the Panel of Consultants on Catalysts by Volks-
wagenwerk AG, May 24, 1974.
-------
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
107
Data presented to the Panel of Consultants on Catalysts by Corning,
June 5, 1974.
Data presented to the Panel of Consultants on Catalysts by Nissan,
May 23, 1974.
Data presented to the Panel of Consults on Catalysts by Saab,
May 23, 1974.
Data presented to the Panel of Consultants on Catalysts by Exxon
Research and Engineering, June 3, 1974.
Data presented to the Panel of Consultants on Catalysts by Houdry
Division, Air Products and Chemicals, June 6, 1974.
Data received from General Motors, private communication.
McConnell, R.J. and J.F. McDonnell, SAE Paper No. 730597, 1973;
Bouffard and Waddey, SAE Paper No. 740135, 1974.
Data presented to the Panel of Consultants on Catalysts by
Universal Oil Produces Company, May 16, 1974.
Gagliardi, J.C., C.S. Smith, and E.E. Weaver, API Meeting, Div. of
Refining, New York, May 1972.
Hofer, L.J.E., J.F. Schultz, and J.J. Feenan, Bureau of Mines
Report No. RI-6243, 1963.
Data presented to the Panel of Consultants on Catalysts by Chrysler
Corporation, May 6, 1974.
Giacommazzi, R.A. and M.F. Holmfeld, "The effect of lead, phosphorus,
and sulfur on deterioration of two oxidizing bead-type catalysts,"
SAE Paper No. 730595, May 1973.
Klimisch, R.L., J.C. Summers, and J.C. Schlatter, "The chemistry
of degradation in automobile emissions control catalysts,"
GMR-1539, paper presented at Third No. Am Meeting of The
Catalysis Society, San Francisco, February 1974.
Bowditch, F.W. in letter to Eric 0. Stork, EPA, June 4, 1974.
Yarrington, R.M. and W.E. Bambrick, "Deactivation of Automobile
Exhaust Control Catalysts," J. Am. Poll. Control Assoc. XX
(1970), 398.
-------
108
43. Weaver, E.E. "Effects of Tetraethyllead on Catalyst Life in
Customer Type Vehicle Operation," SAE Transactions LXXVIII
(1969), Paper No. 690016.
44. Mooi, J., J.P. Kuebrick, M.F.L. Johnson, and F.L. Chloupek,
"Modes of deactivation of exhaust purification catalysts," API
Preprint No. 01-73, Philadelphia Meeting, 1973.
45. Gallopoulos, N.E., J.C. Summers, and R.L. Klimisch, "Effects of
engine oil consumption on the activity of exhaust oxidation
catalysts," SAE Meeting in Detroit, May 1973, (GMR-1371).
46. Shelef, M., R.S. Dalla Betta, J.A. Larson, et al., "Poisoning of
noble metal monolithic oxidation catalysts in automobile exhaust
environment," presented at the National AIChE Meeting, New Orleans,
March 1973.
47. Hetrick, S.S. and F.J. Hills, "Fuel lead and sulfur effects on
aging of exhaust emission control catalysts," SAE Paper No.
730596, May 1973.
48. Neal, A.H., E.E. Wigg, and E.L. Holt, "Fuel effects on oxidation
catalysts and catalyst-equipped vehicles," SAE Paper No. 730593,
Detroit, May 1973.
49. Holt, E.L., E.E. Wigg, and A.H. Neal, "Fuel effects on oxidation
catalyst and oxidation systems II," SAE Paper No. 740248,
Detroit, January 1974.
50. Data presented to the Panel of Consultants on Catalysts by
Daimler-Benz A.G., May 21, 1974.
51. Data presented to the Panel of Consultants on Catalysts by Matthey-
Bishop, June 27, 1974.
52. Data presented to J.W. Hightower by M. Gothelf, Gould Inc., in
telephone conversation July 9, 1974.
53. Data presented to the Panel of Consultants on Catalysts by Volvo,
May 23, 1974.
54. Southwest Research Institute, "A Literature Search and Analysis
of Information Regarding Sources, Uses, Production, Consumption,
Reported Medical Cases, and Toxicology of Pt and Pd," prepared
for EPA, Contract No. 68-02-1274, April 15, 1974.
-------
109
55. Data presented to the Panel of Consultants on Catalysts by Engelhard
Industries, June 21, 1974; Pt Metals Rev. XVIII (1974), 64.
56. Data presented to the Panel of Consultants on Catalysts by Ford,
May 29, 1974.
57. Voorhoeve, R.J.H., J.P. Remeika, P.E. Freeland, and B.T. Matthias,
Science, (July 28, 1972), 353.
58. Harris, W.J. and N.B. Hannay, "Trends in Usage of Rare Earths,"
NMAB Report 266, October 1970.
59. Gauguin, R. , M. Graulier, and D. Papee, "Properties required of
catalyst carriers for exhaust gas purification," Am. Chem. Soc.
National Meeting, Los Angeles, April 1974.
60. Data on U.S. production of alumina (8 million tons/year) supplied
by W. Hayden, Martin Marietta, July 9, 1974.
61. Data submitted to NAS Medical Biologic Effects of Environmental
Pollutants, Platinum Group Metals Panel, by General Motors,
March 8, 1974.
62. Beltzer, M., R.J. Campion, and W.L. Peterson, "Measurement of
vehicle particulate emissions," SAE Paper No. 740286, Detroit,
March 1974.
63. Malanchuk, M., N. Barkley, G. Contuer, M. Richards, R. Salter,
J. Burkart, and Y. Yang, "Exhaust Emissions from Catalyst-
Equipped Cars," preliminary EPA report, March 1974,
64. Balgord, W.D., "Fine Particles Produced from Automotive Emissions-
Control Catalysts," Science CLXXX (1969), 1168.
65. Roberts, A.E. "Platinosis - A Five Year Study of the Effects of
Soluble Pt Salts on Employees in a Pt Laboratory and Refinery,"
Arch. Indust. Hyg. & Occ. Health IV (1951), 549.
66. Freedroan, S.O. and J. Krupey, "Respiratory Allergy Caused by Pt
Salts," J. Allergy XLII (1968), 233.
67. Data presented by R.M. Clyne and W.V. Andresen, American Cyanamid
Company, March 5, 1974.
68. Wood, J.M. "Biological Cycles for Toxic Elements in the Environment,"
Science CLXXXIII (1974), 1049.
-------
110
69. Jensen, S. and A. Jernelov, Int. At. Energy Agency Tech. Report
Ser. 137 (chapter 4), 1972, p. 43.
70. Sax, N.I. Dangerous Properties of Industrial Materials. 2d Ed.
Reinhold Pub. Co., 1973.
71. Moran, J.B. "Assuring public health protection as a result of the
mobil source emissions control program," SAE Paper in Detroit,
February 1974.
72. Report to EPA by General Motors, "Regarding Automotive Sulfate
Emissions: A Status Report," May 7, 1974.
72a. Moran, John B. "Review of Comments on Automotive Sulfate; Federal
Register Notice, Second Draft," Sept. 5, 1974.
72b. Redmond, John Jr., (National Research Council) in letter to
Joe W. Hightower, August 2, 1974.
72c. Lewis, T.R., M.O. Amdur, M.D. Fritzhand, and K.I. Campbell,
"Toxicology of Atmospheric Sulfur Dioxide Decay Products," EPA
Publication AP-III, July 1972.
73. Phillips, P. British Patent 6,096, 1831.
74. Lee, S.D., V.N. Finelli, L. McMillian, and R.M. Danner, "Bio-
chemical Effects of Emissions from an Automobile Engine with and
without Catalytic Converter," Preliminary EPA report, March 1974.
74a. Treou, J.F. et al., AMA Arch. Ind. Hyg. Occup. Med. 2_, 716(1950).
75. Schlatter, J.C., R.L. Klimisch, and K.C. Taylor, Science CLXXIX
(1973), 798.
76. Skyarov, A.V., I.I. Tretyakov, B.R. Shub, and S.Z. Roginski,
Doklad. Akad Nauk. SSSR CLXXXIX (1969), 1302.
77. Bond, G.C. Catalysis by Metals, Academic Press, 1962, p. 461.
78. Carberry, J.J. and J.B. Butt, Catal. Revs. Sci. and Eng. (in
press).
79. Brittain, M.I., Ph.D. Dissertation, Yale University, New Haven,
Conn., 1967.
80. Shelef, M. and J.T. Kuxnmer, Chem. Eng. Progress Symposium Series
No. 115. LXVII (1971), 74.
-------
Ill
81. Peters, M.S. AEC TID-18423, 1963.
82. Ayen, R.J. and Y.-S. Ng. Inter. J. Air & Water Poll. X (1966), 1.
83. Ayen, R.J. and M.S. Peters, IEC Proc. Design Devel. I (1962), 205.
84. London, J.W. and A.T. Bell, J. Catal. XXXI (1973), 32.
85. Unland, M.L. J. Catal. XXXI (1973), 156.
86. Kobylinski, T.P. and J.B.W. Taylor, J. Catal. (in press).
87. Amirhazmi, A., J.E. Benson, and M. Boudart, J. Catal. XXX
(1973), 55.
88. Hegedus, L.L. GMR-1394, May 16, 1973; GMR-1571, March 29, 1974;
and Baron, GMR-1584.
89. Masamune, S. and J.M. Smith, AIChE J XII (1966), 384.
90. Heck, R.H., J. Wei, and J.R. Katzer, "The transient thermal
response of a monoloth catalyst support," Inter. Symp. in Chem.
Reaction Eng., Evanston, 111., August 1974.
91. Young, L.C. and B.A. Finlayson, "Mathematical modeling of the
monolith converter," Inter. Symp. in Chem. Reaction Eng.,
Evanston, 111., August 1974.
92. Finalyson, B.A. "Modeling Monolith Converters," Cat. Revs.,
Sci. and Eng. (in press).
93. Young, L.C. and B.A. Finlayson, Paper given at AIChE Philadelphia
Meeting, November 1973.
94. Hegedus, L.L. Paper presented at Am. Chem. Soc. National Meeting,
Chicago, August 1973.
95. Wei, J. and R. Becker, "The optimum distribution of catalytic
material on support layers in automotive catalysts," paper
presented at the Am. Chem. Soc. National Meeting, Los Angeles,
April 1974.
96. Johnson, W.C. and J.C. Chang, SAE Paper No. 740196, 1974.
97. Carberry, J.J. and A.A. Kulkarni, J. Catal. XXXI (1973), 41.
-------
APPENDIX A
Companies Site-Visited or Interviewed
1. Domestic
a. Ford Motor Company
Dearborn, Michigan
Chrysler Corporation
Detroit, Michigan
General Motors Corporation
Warren, Michigan
d. Engelhard Industries
Murray Hill, New Jersey
e. Air Products and Chemicals
Marcus Hook, Pennsylvania
f. Oxy-Catalyst
West Chester, Pennsylvania
g. Union Oil Co. of California
Brea, California
h. American Oil Company
Wh i t ing, Ind iana
i. E.I. duPont de Nemours
Wilmington, Delaware
j. Universal Oil Products
Des Plaines, Illinois
k. Clyde Engineering Service
(meeting in Evanston, 111.)
1. Questor Corporation
Toledo, Ohio
May 6, 1974
May 6, 1974
May 7, 1974
June 10, 1974
June 6, 1974
June 11, 1974
April 2, 1974
March 28, 1974
June 12, 1974
May 16, 1974
May 20, 1974
June 12, 1974
m. Gulf Research and Development June 5, 1974
Pittsburgh, Pennsylvania
*Hightower,
Ollis, Butt,
Wise
Hightower,
Ollis, Butt,
Wise
Hightower,
Ollis, Butt,
Wise
Hightower,
Ollis, Butt
Ollis
Ollis
Hightower,
Wise
Hightower
Ollis
Butt
Butt
Hightower,
Butt
Butt
*Last names of the members of the Panel of Consultants on Catalysts
making visits.
112
-------
113
APPENDIX A (continued)
1. Domestic (continued)
n. Bell Laboratories
Murray Hill, New Jersey
June 7, 1974
Ollis
o. Corning Glass Works
Corning, New York
June 5, 1974
Ollis
p. Exxon Research and Engineering June 3, 1974
Linden, New Jersey
Ollis
q. Gould, Incorporated
Cleveland, Ohio
June 13, 1974 Hightower,
Butt
2. Nondomestic Companies Interviewed at CMVE Meeting in Washington
May 21-24, 1974 by Butt or Ollis
a. Daimler-Benz AG
b. Fiat, S.p.A./Ferrari
c. Honda Motor Company
d. Nissan Motor Company, Ltd.
e. Adam Opel AG
f. Peugeot, Inc.
g. Regie Nationale Des Usines Renault
h. Saab-Scania Aktiebolag
i. Toyo Kogyo Company, Ltd.
j. Toyoto Motor Company, Ltd.
k. Volkswagenwerk AG
1. AB Volvo
-------
114
APPENDIX B
General Questions to Companies
from the
Members of the Panel of
Consultants on Catalysts
of the
Committee on Motor Vehicle Emissions
One or more members of the Panel of Consultants would like to visit your
company to discuss the current status of the technological feasibility
of using catalysts to control automobile emissions. Please be prepared
to discuss the following general areas, and supply data supporting your
statements. We realize that other panels of consultants may have
requested similar information; in such cases where you have nothing new
to add, please include a copy of or cite your previous responses. Also,
please feel free to provide additional information not specifically
mentioned that you think would help us fulfill this mission. If you
have no data in a specific area, so indicate.
System - What catalytic systems are being considered currently for
the removal of (a) HC and CO, and (b) NO^?
2. Composition - What are the chemical components and compositions of
your most promising oxidation and reduction catalysts? For various
engines sizes could you specify reactor volume, maximum space
velocities, and operating temperatures? What is the weight loading
of active components? What is the composition of the support material
and/or washcoat?
3. Physical Form - Are the emission-control catalysts in the form of
pellets, monoliths, expanded metal screens, metal sponges, etc.?
What advantages does the preferred physical form of the catalyst
have over the others?
4. Reactor - What is the configuration of the container, and what
measures are being taken to overcome attrition?
Durability - What are the durability limits under different
driving conditions? What external conditions cause catalyst
deactivation? Could you delineate the relative contribution of fuel
and lubricant additives, over-temperature, to deactivation? Is any
operating information available about effective operational windows
and the results of excursions from these windows (temperature,
space velocity, gas composition, etc.)? After exposure to a given
quantity of leaded gasoline, what conditions are required for
complete or partial recovery of a given catalyst? Is there
-------
115
APPENDIX B (continued)
5. Durability (continued)
any indication of synergistic effects caused by various catalyst
poisons?
6. Performance - Could a summary of data be obtained on the performance
of catalysts in terms of conversion, light-off temperature, and
physical deterioration with mileage?
7. Health Hazards - Discuss the effects of catalysts on emissions of
polynuclear aromatics, partially oxidized compounds, S0X, etc.
Has any catalyst loss been observed and if so, to what extent, and
in what physical and chemical form? Have any toxicological studies
been performed on particulate matter issuing from catalyst exhaust
systems? Provide information about the sulfate problem.
8. Activity Monitoring - What methods are being considered for
monitoring the activity of the emission-control catalysts; i.e.,
the degree of deactivation? Are accelerated aging tests available
for evaluation of catalyst performance? If catalysts are guaranteed
for a given mileage, what factors would void the guarantee and how
will these factors be assessed?
9. NOx Standard - What catalytic systems currently meet the 0.4 g/mi
NOx standard? What solutions are being considered to the problem
of ammonia formation? Can the reduction catalysts under considera-
tion withstand exposure to an oxidizing atmosphere? What is the
fuel economy penalty required to meet the current N0X standard?
All factors considered (system complexity, expense, durability,
performance, fuel economy, as well as emission control), wtlat in
your opinion is a "reasonable" NOx standard and why?
10. Three-Way Catalysts - What is the present development stage of a
single catalyst for simultaneous control of HC, CO, NOx? What
are the operational windows for such conversion? What additional
hardware is required to make these catalysts effective?
11. Mathematical Models - Do you have working mathematical models that
accurately describe the p erf ©finance of the catalytic systems (kinet-
ics, diffusion, temperature, poisoning, etc.)? How do the various
kinetic parameters change as the catalyst is poisoned?
12. Catalyst Supply - Do you foresee any problems associated with supply
of catalytic material? Where will the noble metals come from, and
is the mining capacity sufficient? Is recycling of deactivated
catalysts being considered feasible at this time?
-------
116
APPENDIX B (continued)
13. Anything Else Relative to Automobile Catalysts Not Already Covered.
------- |