PB85-161222
Combustor Study of the Deactivation of a
Three-Way Catalyst by Lead and Manganese
Northrop Services, Inc.
Research Triangle Park, NC
Prepared for
Environmental Protection Agency
Research Triangle Park, NC
Jan 85
U.S. Department of Commerce
National Technical Information Service
-------
PB85-161222
EPA/600/D-85/006
January 1985
COMBUSTOR STUDY OF THE DEACTIVATION OF A
THREE-WAY CATALYST BY LEAD AND MANGANESE
by
John Duncan
Northrop Services, Inc.
Environmental Chemistry and Emissions Research Laboratory
Research Triangle Park, NC
and
James N. Braddock
Atmospheric Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
ATMOSPHERIC SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
TECHNICAL REPORT DATA
(Please recJ Ir.ttrjtnont on the reterse before completing!
1. REPORT NO.
EPA/600/D-85/006
3. RECIPIENT;
4. TITLE AND SUBTITLE
Conbustor Study of the Deactivation of a
Three-Way Catalyst by Lead and Manganese
5. REPORT DATE
January 1985
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
John Duncan and James N. Braddock
B. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Northrop Services, Inc.
Post Office Box 12313
Research Triangle Park, NC 27709
10. PROGRAM ELEMENT NO.
C9YA1C/01-407A (FY-G4)
1V CONTRACT/GRANT NO.
68-02-2566
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Lab-RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The activity and durability of a platinum-rhodium automotive three-v/ay catalyst
v/ere investigated as a function of lead and manganese fuel levels using a pulse-
.flame combustor. Total hydrocarbons, carbon monoxide, and nitric oxide conversions
and three-viay (HC/CO/NO) conversion efficiency windows were determined for approxi-
mately 24,000 combustor-siniulated miles. The vfindow for 80?; HC/CO/NO efficiency
disappeared at approximately 9,000 miles, 13,500 miles, and 4,500 miles for 0.5 g
of lead per gallon of fuel, 0.0625 g of manganese per gallon of fuel, and a combined
manganese and lead misfu°ling study, respectively. The catalyst's nitric oxide
reduction activity displayed the greatest sensitivity to catalytic poisoning with
both lead and manganese fuels.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLA.-.S (This Report/
Unclassified
21. NO. OF PAGES
11
20. SECURITY CLASS (This page I
Unclassified
22. PRICE
EPA Form 2220-1 ("•». 4-77) PREVIOUS EDITION is OBSOLETE
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recoranendaticn for use.
ii
-------
ABSTRACT
The activity and durability of a platinum-
rhodium automotive three-way catalyst were
investigated as a function of lead and
manganese fuel levels using a pulse-flame
combustor. Total hydrocarbons, carbon
monoxide, and nitric oxide conversions and'-'
three-way (UC/CO/NO) conversion efficiency
windows were determined for approximately
24,000 combustor-simulated miles. The window
for SOZ UC/CO/NO efficiency disappeared at
approximately 9,000 miles, 13,500 miles, and
4,500 miles for 0.5 g of lead per gallon of
fuel, 0.0625 g of manganese per gallon of fuel,
and a combined manganese and lead misfuel ing
study, respectively. The catalyst's nitric
oxide reduction activity displayed tbe greatest
sensitivity to catalytic poisoning with both
lead and manganese fuels.
THE WIDESPREAD USE OP CATALYTIC CONVERTERS on
automobiles and tbe subsequent need for a
laboratory method that permits realistic
screening of these catalysts prompted tbe
development of the pulse-flame combustor
(1,2).* The combustor used in this study was
developed by Amoco Oil Company under contrac-
tual agreement with the U.S. Environmental
Protection Agency (EPA) (3). The apparatus was
subsequently modified by the Mobile Source
Emissions Research Branch of EPA. The output
of the combustor closely resembles engine
exhaust with respect to gaseous composition,
contaminants, and space velocity. The
durability of a three-way monolithic catalyst
was evaluated by exposure to simulated auto
exhaust. Catalyst activity was measured as a
result of poisoning with antiknock additives.
Poisoning of the platinum/rhodium (Pt/Rh)
catalyst by fuels containing low concentrations
of manganese (Hn) in the form of methylcyclo—
pentadienyl manganese tricarbonyl (MKT) (4-8)
and lead (Pb) in the form of tetraethyl lead
(TEL) (9-12) was studied with this coabustor.
The three-way catalyst (TWO emission
control system controls exhaust levels of the
three regulated automotive pollutants:
unburned hydrocarbons (UC), carbon monoxide
(CO), and oxides of nitrogen (NOK). This
system controls the exhaust levels by ' ' •
simultaneously oxidizing unburned HC aad CO to
carbon dioxide (CO2> and water (U2O) and by
reducing nitric oxide (NO) to nitrogen (Nj).
The TWC works properly in a narrow "window" of
exhaust gas composition. The exact composition
is achieved only when tbe engine (or combustor)
gets the ideal (stoichiometric) mixture of air
and fuel, a 14.7:1 ratio. For example, a TWC
conversion efficiency profile, depicted in
Figure 1, indicates that the SOZ conversion
efficiency window for tbe three pollutants of
interest (UC, CO, and NO) occurs in a specific
region of air/fuel ratio(s) (see shaded area),
or redox potential. The redox potential (R) is
the measure of the gas mixture stoichiometry of
reducing components to oxidizing components.
Therefore, at R — I, a stoichiometric gas
mixture is present. In the lean region
(R < 1), oxidation is favored, and the NO
conversion efficiency is reduced. In the rich
region (R > 1), reduction is favored, and the
UC and CO conversion efficiencies are reduced.
Therefore, for the TWC to act properly, .
the exhaust gas composition must be in the SOZ
conversion window. If the gas mixture is not
in this range, excess HC, CO, or HOX emissions
will result. Even when the proper exhaust gas
composition is reaching the catalyst, excess
emissions may result because of damage (e.g.,
thermal sintering) or poisoning of the active
oxidation and/or reduction sites on the
catalyst.
eXPERlKENTAL
APPARATUS - A pulse-flame combustor was
used to study the poisoning of a commercially
available Pt/Rh TWC by Pb and Hn derived from
TEL and KMT antiknock additives, respectively.
The combustor consisted of a fuel delivery
-------
80% CONVERSION WINDOW
*
100
90
8°
o 70
UJ
y eo
u.
u.
UJ
z 50
UJ
Z
40
30
20
10
0
NO
\
13:1
RICHER
14:1
AIR-FUEL RATIO
15:1
LEANER
r—i 1—i-
4.0 2.0 1.0 0.5
REDOX POTENTIAL, R
R =
CO
+ 3 (HC)
2O, + NO
0.25
Fig. 1 - Three-way catalyst efficiency profile
system that incorporated primary and auxiliary
fuel cells, a carburetor and mixer assembly, an
ignition furnace, and a controlled-temperature
catalyst furnace (Figure 2).
The primary fuel cell used electroysis of
a saturated barium hydroxide solution to
deliver fuel to the combustor carburetor. A
90-volt potential applied across two electrodes
immersed in the electrolyte solution caused
hydrogen (Ha) and oxygen (02) to be released at
a constant race. These gases were forced into
the water cell, where they displaced water into
the fuel cell. The water then displaced the
fuel at a rate that was directly proportional
to the current flowing through the electrolyte
cell, thereby delivering fuel to the carburetor
at & constant rate. The Pyrex* glass
carburetor, which was heated to 60°C, vaporized
the liquid fuel and mixed it with N£ for
delivery to the mixer. Three bubbles were
blown into the main body of the carburetor to
provide better mixing and to allow for
expansion of the fuel vapors. The mixer
assembly blended the fuel vapor/N2 mixture with
02 before combustion. The mixer assembly also
housed a flame adjuster that quenched the flame
at the mixer tip after ignition.
The fuel delivery system for the auxiliary
(additive) fuel was used only during the
catalyst aging mode to supply nonvolatile fuel
additives to the flame. The auxiliary fuel
system was similar in design to the primary
fuel delivery system. The fuel from the
auxiliary fuel cell was delivered to an
ultrasonic atomizer where it was dispersed
directly into the carburetor tube-as a very
fine mist. When th^ combustor was used in the
aging mode, the primary fuel was delivered to
the carburetor at a flow rate of 0.102 al/ain.
Fuel additives, such as TEL or MMT, were added
to the iso-octane in the auxiliary fuel cell
and delivered via the ultrasonic atomizer into
the combustion tube at a rate of 0.034 ml/rain.
The total fuel flow entering the combustion
furnace (i.e., the primary flow at 0.102 ml/rain
and the auxiliary flow at 0.034 ml/min) was
0.136 ml/min. Combustion of this fuel resulted
in an exhaust gas flow rate, of approximately
750 cmVmin. During the aging mode, all of she
exhaust gases,were allowed to flow across the-
catalyst at a space velocity of approximately
90,000 h-l.
When the combustor was used to evaluate
catalyst activity (shown by dotted lines in
-------
9-..M»_Kx«0
xo~_
««->• - r~\ . ........~%-.
IIPOH«44W—1 » >' "* ' ~ '"' " "
MI uu\ V "J
~£C\ OL
Pig. 2 - Pulse-flame combustor schematic
Figure 2), only 500 cmVmin of the combustion
exhaust gases were allowed to flow through the
catalyst at a space velocity of about
60,000 h~l, and the remainder of the exhaust
gas was vented to the atmosphere. During the
evaluation mode, a separator/ funnel was
provided between the combustion tube and the
catalyst tube to function as a water trap and
to allow thorough mixing of the combustion
exhaust gases with dilutive Hj. Exhaust gases
frora the combustor were pulled through
Penna-Pure" dryers by a Teflon'* bellows sample
puop to reduce their moisture content, thereby
reducing the chance of condensation (13). A
five-way stainless steel ball valve allowed
selection of the sample to be analyzed and
provided delivery of calibration gases to the
analytical instrumentation.
ANALYTICAL SCHEME - The exhaust gas
sampling system (Figure 3) was configured to
provide both real—time and sample loop
•easurements of the exhaust gas. The exhaust
gas was delivered by a Teflon bellows puap to a
Monitor Labs Model 8430H NO/NO* analyzer and a
Perkin-Blmer 3920 gas chromatograph. The NO
sample flow was regulated with a flow
controller and delivered to a diluter to be
diluted to a 10:1 mixture. The
cheniluminescence detector and 10:1 diluter
permitted real time determination of NO at a
minimum detection limit of I ppm. Total HC
were determined by a flame ionization detector
(PID). The FID inlet was connected to the
outlet of the sample pump manifold and a
constant 15 cm3/min flow was maintained with a
digital flow controller. The constant flow
rate sample permitted accurate determination of
total HC at the 2 ppm detection limit.
Gas chromi.ography with thermal
conductivity detection was used to determine
H2, Ojj and CO levels. The chromatographic
column was 6 ft x 1/8 in. stainless steel
packed with 60/80 mesh.13X molecular sieve.
The column oven was maintained at 60°C and was
programmed to reach 90°C in 10 rain; the helium
carrier gas flow rate was maintained at
20 cm^/min through the column. The l-CB>3
sample loop permitted accurate determinations
of O.IZ H2, 0.05102, and 0.05Z CO.
CATALYST DEACTIVATION STUDIES - The pulse-
flame cofflbustor was used to study the
deactivatlon of a commercially available TWC
for 1980 vehicles. Test conditions are
summarized in Table 1. The neat fuel used in
the studies contained 70Z iso-octane, 30Z
toluene, and 0.025Z sulfur. Study A aged the
catalyst with neat .fuel alone, thereby
simulating approximately 22,500 miles of
vehicle operation. Evaluation of the catalyst
activity was performed using only neat fuel
(studies A, B, C, and D). Study B aged the
catalyst with neat fuel containing 0.5 g of Pb
per gallon (simulating 16,000 miles). Study C
aged the catalyst with 0.0625 g of Mn per
gallon of neat fuel (simulating 24,000 miles).
Study D, which simulated over 24,000 miles,
aged the catalyst with repetitions of the
following sequence: three simulated fill-ups
of 0.0625 g of Kn per gallon, followed by one
simulated fill-up of 0.80 g of Pb per gallon.
The catalyst used in this study was
derived from a commercially available TWC for
1980 vehicles, with a Pt/Rb ratio of 11:1 and
40 g/ft3 of precious metal loading. The
catalyst used a Corning'* substrate with
400 cells/in^. The catalyst buttons used in
-------
Pig. 3 — Cas anclysis schematic
Table 1 - Comparison of Studies A, B, C, and D
Aging
Evaluation'
Study
A
B
C
Db
Fuel
Neat: 70Z Iso-octane
30Z toluene
0.025Z Sulfur
Neat * 0.5 g of Pb
per gallon of fuel
Neat + 0.062S g of Mn
per gallon of fuel
Neat + 0.0625 g of Mn
per gallon of fuel
(fill-ups one
through three)
Neat » 0.80 g of Pb
Conditions
65
-------
Che cofflbustor studies were all hand-cue from
the commercial TUC. These buttons, which were
approximately 1.0 cm by C.8 cm (diameter), fit
snugly inside the quartz tube in the catalyst
furnace.
The experimental conditions 'used in this
study were 24,900 miles of vehicle operation
simulated in 160 h of combustor aging. The KMT
and TEL concentrations, which were based on a
vehicle industry average fuel economy of
20 mi/gal, were twice the national average
antiknock concentrations for automobile fuel.
The increased concentrations of the antiknock
additives that flow through the catalyst in the
24,000-mile simulation were assumed to be
directly related to that of the national
average concentrations in a 48,000-mile
simulation. For example, during part of 1979,
EPA suspended enforcement of its ban against
use of the gasoline additive KMT, thus allowing
refiners to use up to 1/32 g of KMT per gallon
of fuel (0.031 g/gal) (14). In this study,
1/16 g of HUT per gallon of fuel (0.0625 g/gal)
was used in two of the catalyst deactivation
studies over 24,000 simulated miles.
RESULTS
In Study A, the catalyst button was aged
for 150 h and evaluated with neat fuel
containing 70Z iso-octane, 30Z toluene, and
0.25Z sulfur. The catalyst was aged at 650°C
and a space velocity of about 90,000 h""', and
was evaluated at 550 C and a space velocity of
approximately 60,000 h'1 aC 30-h intervals.
(The lower evaluation temperature was chosen to
minimize the stress on the catalyst and to
reduce the amount of aging while evaluating.)
As shown in Figure 4, the catalyst activity or
efficiency was slightly reduced throughout the
simulation. Although the efficiency increased
slightly between the 60-h and the 90-h
evaluation, the slight downward trend continued
beyond the 90-h evaluation. Table 1 describes
the fuel content and test interval parameters
for Studies A, B, C, and D. The catalyst
button in Study B was aged for 120 h with 0.5 g
of Pb per gallon added to the neat fuel. TEL
caused a decrease in conversion efficiencies of
HC and NO, as depicted in Figure 5. By the end
of the 120-h simulation, HC efficiency was
reduced from 90Z to 64Z, and NO efficiency was
reduced from 951 to 48Z.
In Study C, the catalyst button was aged
for 160 h with 0.0625 g of Hn per gallon added
to the neat fuel. Figure 6 shows that MMT
primarily affected NO conversion efficiency; NO
was reduced quickly after 60 h of aging to 26Z
efficiency by the end of the simulation.
Study D was planned to determine the
effect of casual misfueling on the catalyst's
NO reduction efficiency (15,16). This study
aged the catalyst button with repetitions of
three simulated fill-ups of fuel containing
0.0625 g of Mn per gallon of fuel followed by
one simulated fill-up of 0.80 g of Pb per
gallon of fuel. The aging fuel contents are
AGING WITH NEAT FUEL
100-
<
80-
60-
40-
20-
0
840-
§20
{£.
UJ
o- 0
100
80
60
40
20
HC
NO
30
Fig. 4 - Study A:
with neat fuel
60 90
AGING, h
120
150
Three-way catalyst aged
shown in Table 1. The catalyst was aged at
750°C and a space velocity f about 90,000 h~l
for 150 h. At 150 h of aging, the catalyst was
aged for an additional 10 h with neat fuel to
investigate the possibility of reactivation of
the TWC in a poison-free atmosphere (i.e., no
Mn or Pb present). The catalyst was evaluated
at 30-h intervals at 600°C and a space velocity
of about 60,000 h~l (temperatures recommended
by the Automotive Catalyst Section of Rngelhard
Industries). Figure 7 indicates that the HC
and NO conversion efficiencies were reduced at
150 h of aging; HC was reduced to 40Z
efficiency and NO was reduced to 18Z
efficiency. The HC and NO efficiencies
increased slightly after aging in the poison-
free atmosphere, but catalytic activity
continued to be below 60Z efficiency.
------- |