Final  Report
                                 July 1971
                  DEVELOPMENT OF PARTICULATE
                  EMISSION CONTROL TECHNIQUES
                  FOR SPARK-IGNITION ENGINES
         John B.
Moran, Otto J.
   and Michael
Manary, Russell H. Fay
J. Baldwin
                 Organic Chemicals Department
                 The Dow Chemical  Company
                 Midland,  Michigan  48640
         Prepared  for:
       Office of Air Programs
       Environmental Protection Agency
       2565 Plymouth Road
       Ann Arbor,  Michigan  48105
         Attention:
    Mr.  Charles  L.  Gray,  Jr.
    Project Officer
                                            Contract  EHS  70-101

-------
              EHS 70-101
      DEVELOPMENT OF PARTICULATE
      EMISSION CONTROL TECHNIQUES
      FOR SPARK-IGNITION ENGINES
             John B.  Moran
             Otto J.  Manary
             Russell  H. Fay
             Michael  J. Baldwin
               Contributors
               J. C
F. J. Bartell
L. B. Crummett
H. L. Garrett
J. D. McLean
P. N. North
H. W. Rinn
L. P. Schloemann
Val
enta
L.
S.
J.
W.
P.
C.
L.
A.
N.
C.
B.
A.
E.
B.
Settlemeyer
Sharp
Tou
Tower
Traylor
Van Hall
Westover
           The Dow Chemical Company
           Midland, Michigan  48640
           Final Technical Report
            May 1970 - July 1971
For: Office of Air Programs
     Environmental Protection Agency
     2565 Plymouth Road
     Ann Arbor, Michigan  48105

Attention:  Mr. Charles L. Gray, Jr.
           Project Officer

-------
1
FOREWORD
This report was prepared by the Transportation Research Group,
Organic Chemicals Department, The Dow Chemical Company, Midland,
Michigan, under Contract EHS 70-101. The work reported herein
was administered under the direction of the Office of Air Programs,
Environmental Protection Agency, with Mr. Charles L. Gray, Jr.,
as Project Officer.
The report covers work performed from May 1, 1970, to July 1, 1971.
The authors of this report are John B. Moran, Otto J. Manary,
Dr. Russell H. Fay, and Dr. Michael J. Baldwin.
The authors wish to acknowledge the significant contributions of
the following individuals:
F. J. Bartell
L. B. Crummett
H. L. Garrett
3. D. McLean
P. N. North
H. W. Rinn
L. P. Schloemann
Settl emeyer
Sharp
Tou
Tower
Trayl or
Val enta
Van Hall
Westover
L. A.
S. N.
J. C.
W. B.
P. A.
J. C.
C. E.
L. B.
This report was submitted by the authors on August 5, 1971.

-------
11
ABSTRACT
This report describes work carried out on a research program
to characterize and trap particulate emissions from automotive
power plants. The work is presented in two separate phases
necessitated by a Governmental initiated change in the scope
of the work made in mid-contract year.
Phase I describes the characterization of particulate matter
emitted from automotive power plants operating on leaded gasoline.
These were sampled at four locations in the exhaust system ranging
from near the exhaust manifold to the tail pipe, and after air
dilution. Particle characterization was carried out with a
dynamometer controlled engine operating at the equivalent of
60 mph cruise, 30 mph cruise, and under mild cycling conditions.
The development of an exhaust particle-trap device is also
described. This device was found to be effective at 60 mph cruise
conditions but performed poorly at 30 mph cruise and under mild
cyclic conditions. Loss of the trapping medium, a molten salt mix,
was also evident under cycling conditions. Optimization of the
trap was not achieved because of a change in the contract scope.
Phase II describes the characterization of particulate emissions
from automotive power plants operating on non-leaded fuels. Changes
in the particulate emissions were determined at a rich and lean
air/fuel ratio and in the presence of two catalytic convertors using
the same engine operating conditions as in Phase I. Particulate
sampling was conducted after air dilution. Particle mass emission
rates, particle mass-size distributions, and total aldehyde
emissions were measured.

-------
1•11
TABLE OF CONTENTS
Page
FOREWORD
ABSTRACT ii
LIST OF TABLES v
LIST OF FIGURES vii
I. INTRODUCTION 1
PHASE I 4
II. CONCLUSIONS 5
III. EXPERIMENTAL PROCEDURE 7
A. Determination of Exhaust System Temperature
Profile on the Vehicle 7
B. Determination of Exhaust Temperature Profile
in Laboratory 9
C. Particle Generation 13
D. Sample Collection 16
1. Particulate Matter 16
2. Exhaust Gas Sampling 26
E. Sample Analysis 26
1 . Particulate Matter Analysis 26
2. Gas Analysis 28
3. Aldehyde Determination 32
IV. EXPERIMENTAL RESULTS 33
A. Determination of Exhaust System Temperature
Profile - Vehicle Tests 33
B. Determination of Exhaust System Temperature
Profile - Dynamonieter Tests 40
C. Exhaust Stream Analysis 40
1 . Direct Exhaust Stream Analysis 40
2. Dilute Exhaust Stream Analysis 91

-------
iv
D. Trapping Medium Selection
E. Design of Trap
V. DISCUSSION OF RESULTS
PHASE II 112
VI. CONCLUSIONS . 113
VII. EXPERIMENTAL 115
VIII. EXPERIMENTAL 120
IX. DISCUSSION OF 133
X. FUTURE 136
XI. REFERENCES . . 138
APPENDIX A . . 139
APPENDIX B .. 156
APPENDIX C . . 177
1 00
103
105
PROCEDURE
RESULTS
RESULTS

-------
V
LIST OF TABLES
Page
I. Physical Analysis of Test Fuel
II. Physical Properties of Lubricating Oils .
III . Trace Metal Analysis of Lubricating Oil .
IV. Particulate Sampling Device Location by
Test Series Number 14
V. Engine Conditioning Cycle 13
VI. Particulate Sampling Device Location by
Test Series Number 16A
VII. Cut-Off Values for Andersen Stack Sampler . . 20
VIII. Exhaust Temperature Profile - Road Test Car 33
IX. Exhaust Temperature Profile - Road Test Car 34
X. Exhaust Temperature Profile - Road Test Car 34
XI. Exhaust Temperature Profile - Road Test Car 35
XII. Exhaust Temperature Profile - Road Test Car 35
XIII. Chevrolet Exhaust System Temperature
Profile - Dynamorneter Tests 39
XIV. Sampling Temperature and Particulate Sample Weights 40
XV. Summary of Observations on Particulates Collected
at Position I 42
XVI. Summary of Observations on Particulates Collected
at Position II 43
XVII. Summary of Observations on Particulates Collected
at Position III 44
XVIII. Summary of Observations on Particulates Collected
at Position IV 45
XIX. Crystalline Species of Lead vs. Decreasing Exhaust
Gas Stream Temperature 46
XX. Emission Spectroscopic Analysis of Cold Trap Condensate . . . 67
XXI. Iron and Lead Content of Cold Trap Condensate 68

-------
vi
XXII. Examination of Particulates from Test II 72
XXIII. Particulates on Probes in Test IV 86
XXIV. Crystalline Particulate Samples Collected
from Test IV 88
XXV. Air/Fuel Ratios by Exhaust Analysis 91
XXVI. Effect of Lead Trap on Mass of Particulate
Matter Emitted in Air Diluted Exhaust 91
XXVII. Particle Mass Emissions (4 cfm Filter) 92
XXVIII. Analysis of Salt from Lead Trap After Dynamometer
Testing 99
XXIX. Weight of Salt in Lead Trap Muffler 99
XXX. Aldehyde Analysis of Cold Trap Condensate 100
XXXI. Analysis of fuel for Phase II 116
XXXII. Pontiac Engine Sampling Devices for Phase II 117
XXXIII. Air/Fuel Ratios for Phase II 120
XXXIV. Particulate Samples for Phas& II 121
XXXV. Effect of Air/Fuel Ratio on Particulate Emissions 121
XXXVI. Particulate Samples for Phase II 122
XXXVII. Effect of Noble Metal Catalytic Device on
Particulate Emissions 122
XXXVIII. Effect of Noble Metal Catalytic Muffler on
Hydrocarbon Emissions 125
XXXIX. Particulate Samples for Phase II -
Tests 8, 9, 10 129
XL. Effect of Proprietary Packed Bed Catalytic
Device on Particulate Emissions 129

-------
vii
LIST OF FIGURES
Page
1. Location of Thermocouple for Road Tests g
2. Sampling Locations for Dynamometer Tests 12
3. Exhaust System for Lead Trap Tests 17
4. Installation of Particulate Sampling Probe
and Thermocouple 19
5. Dilution Tube 23
6. Flow Diagram for Engine Exhaust Particulate
Collection . 24
7. Computer Format for Gas Analysis 30
8. Road Tests - Exhaust Temperatures . 36
9. Miles per Hour vs. Engine RPM 37
10. Exhaust Temperature Profile in
Dynamometer Tests 38
11. Particulate Sampling Devices for Test 1 41
12. The Effect of Engine Speed on the Particulate
Matter Collected at Position I (Near Exhaust
Manifold) 47
13. SEM Study of Particulate Matter Collected at
Four Locations in the Exhaust System at 870 RPM 48
14. SEM Study of Particulate Material Collected
at Position I at 870 RPM 49
15. SEM Study of Particulate Material Collected
at Four Locations in the Exhaust System at
1070 RPM 50
16. SEM Study of Particulate Material Collected
at Position I at 1070 RPM 51
17. SEM Study of Particulate Material Collected
at Four Locations in the Exhaust System at
1615 RPM 52
18. SEM Study of Particulate Material Collected
at Position I at 1615 RPM 53

-------
viii
19. SEM Study of Particulate Material Collected
at Four Locations in the Exhaust System at
2125 RPM 54
20. SEN Study of Particulate Material Collected at
Position I at 2125 RPM 55
21 . SEM Study of Particulate Material Collected at
Position IV at 2125 RPM 56
22. Transmission Electron Microscopy Study of
Particulate Material Collected at Position I
at 870 RPM and 1070 RPM 57
23. Transmission Electron Microscopy Study of
Particulate Material Collected at Position I
at 1615 RPM and 2125 RPM 58
24. Transmission Electron Microscopy Study of
Particulate Material Collected at Position II
at 870 RPM and 1070 RPM 59
25. Transmission Electron Microscopy Study of
Particulate Material Collected at Position II
at 1615 RPM and 2125 RPM 60
26. Transmission Electron Microscopy Study of
Particulate Material Collected at Position II
at 870 RPM and 1070 RPM 61
27. Transmission Electron Microscopy Study of
Particulate Material Collected at Position II
at 1615 RPM and 2125 RPM 62
28. Transmission Electron Microscopy Study of
Particulate Material Collected at Position IV
Test 1 at 870 RPM and 1070 RPM 63
29. Transmission Electron Microscopy Study of
Particulate Material Collected at Position IV
Test 1 at 1615 RPM and 2125 RPM 64
30. Mass of Lead in Cold Trap Condensate - Test 1 69
31. Particulate Sampling Devices for Test II 70
32. Distribution of Particulate Material on Andersen
Sampler Plates at Position IV - Test II 73
33. SEM Study of Particulate Material from Sampling
Probe at Position IV - Test II 74

-------
ix
34. SEM Study of Particulate Material from Plate 0
of the Andersen Sampler at Position IV - Test II 75
35. SEM Study of Particulate Material from Plate 0
of the Andersen Sampler at Position IV - Test II 76
36. SEM Study of Particulate Material from Plate 0
of the Andersen Sampler at Position IV - Test II 77
37. SEM Study of Particulate Material from Plate 0
of the Andersen Sampler at Position IV - Test II 78
38. TEM Study of Particulate Material Collected at
Sampling Position IV 81
39. TEM Study of Particulate Material Collected on
Andersen Sampler at Position IV - Plate 0 and
Plate 2 . 82
40. TEM Study of Particulate Material Collected on
Andersen Sampler at Position IV - Plate 5 and
Plate 8 83
41. Cumulative Mass Distribution Profile for
Sample Collected at Position IV (End of Tail
Pipe) Chevrolet V8 84
42. Particulate Sampling Devices for Test IV 85
43. Cumulative Mass Distribution Profile for
Sample Collected at Position III (Out of
Muffler) Chevrolet V8 . 89
44. Degradation of Andersen Sampler Plate at
Position II (Ahead of Muffler) 90
45. Cumulative Mass Distribution Profile from
Dilution Tube - Chevrolet V8 - Standard Muffler 93
46. Cumulative Mass Distribution Profile from
Dilution Tube - Chevrolet V8 - Lead Trap Muffler 94
47. Cumulative Mass Distribution Profile from
Dilution Tube - Chevrolet V8 - Standard Muffler . 95
48. Cumulative Mass Distribution Profile from
Dilution Tube - Chevrolet V8 - Lead Trap Muffler 96
49. Cumulative Mass Distribution Profile from
Dilution Tube - Chevrolet V8 - Standard Muffler 97
50. Cumulative Mass Distribution Profile from
Dilution Tube - Chevrolet V8 - Lead Trap Muffler 98

-------
x
51. Lead Trap Design . 104
52. Vapor Pressure of Lead Halides 109
53. Cumulative Mass Distribution Profile at 1200 RPM 123
54. Cumulative Mass Distribution Profile at 2250 RPM 124
55. Cumulative Mass Distribution Profile for Noble Metal
Catalytic Muffler at 1200 RPM 126
56. Cumulative Mass Distribution Profile for Noble Metal
Catalytic Muffler at 2250 RPM . 127
57. Cumulative Mass Distribution Profile for Noble Metal
Catalytic Muffler at Dow Cycle 128
58. Cumulative Mass Distribution Profile with
Catalytic Device 131
59. Cumulative Mass Distribution Profile with
Standard Exhaust System 132

-------
—1—
I. INTRODUCTION
Increased public concern over pollution of the atmosphere has
resulted in the identification of exhaust emissions from the
internal combustion engine as a major contributor to this
problem. Although a significant amount of work has been directed
towards achieving reductions in the levels of emitted carbon
monoxide, unburnt hydrocarbons, and oxides of nitrogen from this
source, a lesser effort has been expended on the identification
and removal of those potentially toxic combustion products which
are present at lower concentrations in the exhaust stream. These
include particulate matter.
Although considerable technology has been developed for the
removal of suspended particulate matter from gas streams, the
exhaust effluent of the internal combustion engine poses special
problems. Large variations in both temperature and gas velocity
can occur within the exhaust system depending on engine operating
conditions. In addition, particle size varies between wide limits
from those visible to the naked eye to those approaching molecular
dimensions. Particles in the 1 to O.lp range can be retained in
the lungs upon inhalation and pose the most significant threat to
human health. Because of their small size, these are among the most
difficult to remove from the exhaust stream. The limited space
and power available in the automobile also place constraints on
the design of a practical particle trapping device.
This report describes two pieces of work, subsequently referred
to as Phase I and Phase II, directed towards the characterization
and potential removal of particulate emissions.
Phase I of the program was directed towards the design and
construction of a trapping device to remove particulate matter
from the exhaust stream of an automobile engine operating on
leaded gasoline. The following sequence of tasks was defined as
necessary to the attainment of this objective.

-------
-2-
1. The temperature profile along the exhaust system of a
vehicle operating at a variety of highway cruise
conditions should be determined.
2. Correlation of this profile to that of a similar
engine operating on a laboratory dynamometer test
stand should be made.
3. Characterization of the emitted particulate matter
as it traverses the engine exhaust system should be
made. This would involve the sampling of particulates
at several points along such a system and their
subsequent analysis to determine particle mass size,
concentration, and chemical composition.
4. Based on the data generated above, selection of a
suitable trapping medium would be made and the design
of a trapping device undertaken.
5. Construction, evaluation of the device under a variety
of engine operating conditions and subsequent
optimization would be performed.
The purpose of Tasks 1 and 2 was to ascertain the degree of
correlation between the temperature profile of a vehicle exhaust
system operating under roadway driving conditions and that of a
similar exhaust system fitted to an engine operating on a
laboratory dynamometer. A close correlation would assure that
subsequent laboratory studies of exhaust particulate matter
within the exhaust system would relate directly to vehicles
driven on the roadway.
The objective of Task 3 was to define a profile for emitted
particulate matter within the exhaust system so that the most
appropriate design of a trapping device could be made in Task 4.
In Task 5 this device would be constructed and evaluated to
determine its efficiency and the need for further optimization.

-------
-3-
A Government-initiated change in the scope of work mid-way
through the program permitted the successful completion of
Tasks 1-4. However, only limited testing of a prototype device
in Task 5 was possible.
Phase II of the program was directed towards characterization of
the particulate matter emitted from an internal combustion engine
operating on non-leaded gasoline. n particular, it was necessary
to determine the effects of the following variab1es on such emissions.
1. Air/fuel ratio - one rich and one lean.
2. The presence in the system of catalytic reactors
designed to reduce regulated automotive emissions.
Two such reactors were evaluated.
In addition, aldehyde emissions from the engine were monitored.
Throughout this phase of the work, particulate sampling was made
after air dilution of the exhaust stream under controlled conditions
using the system reported by Moran and Manary 1 and described in
detail on page 22 of this report.

-------
-4-
PHASE I

-------
-5-
II. CONCLUSIONS
Phase I of this report covers a study of the chemical and physical
nature of particulate emissions as they traversed the exhaust
system of a 350 CID Chevrolet V8 engine. A particulate trapping
device was constructed and tested to determine its effect on
particulate emissions at 1200 rpm, 2250 rpm, and under mild cycling
conditions. The following conclusions were reached:
1. The temperature profile for the exhaust system of the
Chevrolet test vehicle was nearly linear from the
manifold to the end of the tail pipe; the temperature
profile for different engine speeds forms a family of
essentially parallel curves. Temperatures varied between
900—1375ฐF at the exhaust manifold to 325—810ฐF at the
tail pipe (30-80 mph).
2. The temperature profile for the exhaust system of the same
type engine on a dynamometer stand forms a similar
family of parallel curves at different speeds. The
rate at which the temperature drops is less than in the
moving vehicle. Temperatures varied between 890-1130ฐF
at the exhaust manifold to 430-750ฐF at the tail pipe
(30-60 mph).
3. Collection of particulate matter directly from the
exhaust system on heated Andersen Samplers identified
that the bulk of the particulate matter was formed in
the muffler. Prior to the muffler (“p900ฐF) the exhaust
effluent was mainly gaseous in nature. This conclusion
is supported by the identification of crystal growth on
non—heated filters located ahead of the muffler. Subsequent
to the muffler, there was little precipitation or growth
of particulate matter, but considerable shattering and
disintegration. The D 70 value of particles collected

-------
-6-
at the tail pipe was 1 micron while that of particles
collected immediately after the muffler under identical
conditions was 8.8 microns.
4. The larger particles (‘l j) (hence, more easily trapped) were
shown to be primarily lead halides. Particles smaller than
l i were again mainly inorganic but their major components
were lead sulfates and chlorophosphates. No discreet
particles of organic matter were observed. However, many
of the smaller (
-------
—7-
III. EXPERIMENTAL PROCEDURE
A. DETERMINATION OF EXHAUST SYSTEM TEMPERATURE PROFILE
ON THE VEHICLE
A 1970 Chevrolet Impala equipped with a 350 CID, 2—barrel
carburetor, V8 engine was procured and instrumented to allow
the determination of the temperature profile of the exhaust
system. Selection of the Chevrolet car was based on its
extensive field usage. Chromel-alumel thermocouples were
brazed to standard hose clamps and inserted into the exhaust
system at the locations shown in Figure 1. The thermocouple
locations were identified as follows:
Position I - immediately after the junction
of the two sides of the exhaust
manifold.
Position II - immediately ahead of the muffler.
Position III - immediately after the muffler.
Position IV - near the end of the tail pipe.
The output of each thermocouple was determined via a Thermo
Electric Minimite Potentiometer - Model 80200. The engine
revolutions per minute (rpm) was measured with a Sun
Tachometer - Model P1-4, the intake manifold vacuum by a Sun
Electric Model No. FPT-1 vacuum gauge connected into the distributor
vacuum line, and the vehicle road speed via the car speedometer.
Data was collected by having a driver keep the car a constant
speed on a nearly level stretch of expressway while an assistant
recorded the various instrument readings. Tests were made over
a variety of road speeds and during several days so that the
effect of changes in ambient air temperature could be assessed.
On each day of testing runs were made in both directions on the
expressway to compensate for wind effects.

-------
FIGURE 1
EXHAUST THERMOCOUPLE LOCATIONS - VEHICLE TESTS
II III IV
- - 108” 29 ” 78” - ___
MUFFLER

-------
—9—
B. DETERMINATION OF EXHAUST TEMPERATURE PROFILE IN LABORATORY
A new 1970 Chevrolet V8, 2-barrel carburetor, 350 CID engine
was obtained and mounted on a General Electric - Model 1G614
dynamometer test stand. The engine had the following
specifications:
Displacement 350 cubic inches
Horsepower 255 at 4800 rpm
Carburetor 2-barrel Rochester
Compression ratio 9.0:1
Bore 4.0 inches
Stroke 3.48 inches
Spark plugs AC R45S
Plug gap 0.035 inches
Point dwell 28-32ฐ
Timing 4ฐ B.T.C.
A Meriam Laminar Flow Element- Model 5OMC-2-45F Air Flow
Measurement unit was attached to the carburetor via a
flexible rubber hose to monitor air flow rate. An AC
paper filter element was attached to the Meriam unit to
filter incoming air.
The engine dynamometer was completely instrumented to monitor
and/or control coolant temperature, oil temperature, manifold
vacuum, fuel flow rate, air flow rate, engine rpm, load, etc.
Indolene HO 30 fuel was used throughout the experimental
work. A physical and chemical analysis of the fuel is shown
in Table I. Amoco 200, SAE 30 lubricating engine oil was
used in all laboratory runs. This oil was chosen as minimum
acceptable oil with a low additive package content so that
any contribution of the oil to the emitted particulate
matter would be minimized. The physical properties of the
oil are shown in Table II and its trace metal analysis in
Table III.

-------
-10-
PHYSICAL
TABLE I
ANALYSIS OF TEST FUEL
Distillation ASTM D86
(ฐF) IBP
5
10
20
30
40
50
60
70
80
90
95
EP
% Recovery
% Residue
% Loss
Octanes: D
MO N
RON
FIA: D
% Saturates
% Olefins
% Aromatics
cc/gal TEL
Indolene 110 30
90
118
133
162
188
208
222
234
249
273
320
368
394
96.5
0.5
3.0
95.9
102.9
68 . 1
3.0
28.9
3.0
TRACE METAL ANALYSIS
Lead
Iron
Nickel
Copper
Magnes i urn
Zinc
<1.0 ppm
<0.5 ppm
<0.2 ppm
<0.5 ppm
<3.0 ppm
Aluminum
Calcium
Manganese
Tin
Ti tani urn
<1 .0 ppm
<1.0 ppm
<1.0 ppm
<1.0 ppm
<1.0 ppm

-------
—11—
TABLE II
PHYSICAL PROPERTIES OF LUBRICATING OIL
Gravity 26.4
Flash, ฐF 450
Pour, ฐF 0
Viscosity 100ฐF 555 SUS
210ฐF 66 SUS
V.1. 95
Color, ASTM 5
Sulfated Ash, % wt. 1 .0
Carbon Residue, % wt. 1.3
TABLE III
TRACE METAL ANALYSIS OF LUBRICATING OIL
(Weight %)
Fe <.0005 Mn .0004
Ni <.0005 Pb <.0005
Cu <.0001 Cr <.0005
Al <.0005 Sn <.0005
Ca .0017 Zn .092
Si .0021 Ti <.0005
Mg .13
The engine exhaust system consisted of a 2-1/2 inch ID
exhaust pipe from the manifold to the muffler. The muffler
was a Walker Model No. 21530, W-5798 and was 21 inches long.
The tail pipe was 1-3/4 inches ID. A schematic of this
system is shown in Figure 2. Thermocouples were introduced
into the exhaust system through 1/4 inch pipe nipples welded
into the piping and positioned in such a way as to be
equivalent to those used in the vehicle road tests. Because
of differences in the configuration of the exhaust systems

-------
FIGURE 2
EXHAUST THERMOCOUPLE LOCATIONS - DYNAMOMETER TESTS
_____________ I II III Iv
I’— 28”— I
____________ 85 ” 1Jc 61” — -
______________ ___ - L
GA MUFFLER
ENGINE
GA - sample probe for exhaust gas analysis

-------
-13-
used on the vehicle and in the laboratory work, the distances
between successive thermocouples were slightly different.
However, the same relative positioning was maintained. At
points 2 inches ahead of each thermocouple, a second 1/4-inch
nipple was welded to the system to accommodate sampling probes
to be used in the later stages of the project. A final nipple
was located 12 inches ahead of Position II (Figure 2) so that
a probe could be introduced to obtain samples for exhaust
gas analysis.
Temperature profiles of the exhaust system were thus generated
in several engine dynamometer runs. These runs are outlined
in Table IV.
C. PARTICLE GENERATION
Exhaust particulate matter was generated by running the
Chevrolet engine on the dynamometer test stand under a variety
of controlled road load operating conditions. The engine wa;
first run on a conditioning cycle for 75 hours to stabilize
engine deposits, emission levels, and engine condition. This
cycle is shown in Table V.
TABLE V
ENGINE CONDITIONING CYCLE
Equivalent
Observed Time Vacuum Cruise Speed Decay
Cycle RPM H.P. ( mm) ( in. Hg) ( mph ) (min_ j
1 800 1.8 2.0 18.8 — -
2 1070 15.0 13.0 16.4 27 0.5
3 1615 21.5 20.0 17.2 45 0.5
4 2125 42.3 13.0 14.3 57.5 0.5
5 1070 15.0 12.0 16.4 27 0.5
The sequence repeats after cycle 5. Previous work reported by
Moran and Manary 1 has shown that repetition of such a cycle
over the 75-hour period is satisfactory for the engine to reach
quasi-stabilized operation.

-------
TABLE IV
EXHAUST STREAM SAMPLING
PARTICULATE SAMPLING DEVICE LOCATION BY TEST SERIES NUMBER
Gelman Filter
(Glass Fiber)
(Position)
I, II, III, IV
I, II, III, IV
I, II, III, IV
I, II, III, IV
IV (2)*
Cold Trap
(Ice/Water)
( Position )
I, II, III,
I, II, III,
I, II, III,
I, II, III,
IV (3)*
Sampi i ng
Time
IV 30 minutes
IV 30 minutes
IV 30 minutes
IV 30 minutes
III 2250
Iv 2250
V 2250
I, II, III, IV
I, II, III, IV
I, II, III, IV
5 hours
Dry ice!
5 hours acetone cold
trap @ ice!
water trap @
Position IV
5 hours
Note: See Figure 2 for sampling position location.
*Numbers in parentheses indicate the number of sampling devices of the type
indicated placed in series at the positions shown.
Test No. RPM
I
870
1070
1615
2125
Andersen Stack Sampler
(Position)
No
No
No
No
IV
5 hours
IV I, II, III, IV
III I, II, III, IV
II I, II, III, IV

-------
—15—
The complete set of test runs performed is outlined in
Tables IV and VI.
The first series of tests (Numbers I-V, Table IV) was designed
to allow for characterization of particles along the length
of the exhaust system.
During Test I, steady state sampling of particle was conducted
at four different engine operating conditions. The engine was
first allowed to equilibrate at 870 rpm for 1 hour. Particulate
sampling was then initiated for a 30-minute period after
which time the various sampling devices, described later,
were cleaned, and the engine speed was increased to 1070 rpm.
A second set of samples was collected for 30 minutes under
these conditions. This procedure was repeated at 1615 and
2125 rpm.
At the start of Tests II through V, the engine was held at
steady state operation identical to that under which subsequent
sampling was performed for a period of at least one hour.
Sampling of particulate matter was then made over a period of
five hours. This time period allowed sufficient matter to be
collected at each of the particulate sampling devices for
comprehensive analytical evaluations to be made (see page 26).
In the second series of dynamometer test runs (VI-XI, Table VI),
the exhaust system effluent was fed into an air dilution tube,
which is described later. In order to perform this set of
tests, it was necessary to relocate the Chevrolet engine on
a Dynaniatic Model lOl4DG dynamometer test stand, again
instrumented to control and/or monitor relevant engine
va r I a bl e s
Of the six tests made (Tests VI-XI, Table VI), Tests VI—Vill
utilized the same muffler as in the previous runs (Tests I—V)
in an exhaust system of similar overall length. However,

-------
-16-
a change in exhaust system configuration was necessary to
provide for convenient hook-up to the dilution tube. The
experimental set-up is shown in Figure 3. Tests VI-VIlI were
thus used to establish base-line particulate characterization
for the dilution tube work at 60 mph cruise, 30 mph cruise,
and under mild cycling conditions.
In order to maintain temperatures at the sampling zone of
the dilution tube at 75-80ฐF, only one-half of the engine
exhaust (left bank of cylinders) was fed to the tube. The
validity of this approach was established by Moran and
Manary
Prior to Test VI, the engine was conditioned in the Dow 5
mode cycle for a period of 26 hours. The exhaust system
was then connected to the dilution tube and particulate
matter sampled over a 6-hour period. For Tests VII and VIII,
the engine was allowed to warm up for at least 1 hour under
the conditions specified (Table VI) and the diluted exhaust
stream again sampled for 6-hour periods.
In Tests IX-XI, the conventional exhaust muffler was replaced
by the prototype lead trap muffler described in Section E.
D. SAMPLE COLLECTION
1 . Particulate Matter
For clarity of presentation, two distinct sets of particulate
sampling procedures will be discussed: those utilized for
sampling direct from the exhaust stream and those for
sampling from air-diluted exhaust.
Direct Exhaust Sampling — In the first set of runs
(Tests I-V, Table IV), particle collection was effected
via sampling probes located in the exhaust system at
various points along its length. Their position has

-------
TABLE VI
AIR DILUTIO TUBE SAMPLING
PARTICULATE SAMPLING DEVICE LOCATION BY TEST SERIES NUMBER
Dilution Tube
After
Exhaust System Before Muffler Muffler 1 cf i 4 cfni
Gelman Glass Cold Trap Cold Andersen Mifl pore Glass Glass Glass
Test RPM Fiber Filter Ice/Water Dry Ice/Acetone Trap Sampler Filter Fiber Fiber Fiber
vi 2250 Yes Yes Yes - Yes Yes Yes Yes Yes
vii Cycle Yes Yes Yes Yes Yes Yes Yes Yes
VIII 1200 Yes Yes Yes Yes Yes Yes Yes Yes
IX 1200 Yes Yes Yes Yes Yes Yes Yes Yes Yes
x 2250 Yes Yes Yes Yes Yes Yes Yes Yes Yes
XI Cycle Yes Yes Yes Yes Yes Yes Yes Yes Yes

-------
FIGURE 3
EXHAUST SYSTEM FOR LEAD TRAP EVALUATION
____ -- —-— -- 111” --
ENGINE
70”
MUFFLER tube
to di1 tion
o CD rr
ox :r
I— :r m
0.. 0 . )
I
rt (1) 0
rt 0
0.) 0
ag
0.)
(I ) U) I—’
0.) CD
C l)
0.3
CD

-------
-18-
already been described (page 11 and Figure 2). Each probe
was constructed of stainless steel tubing and sized to
assure isokinetic sampling. Such sizing was based on
calculated exhaust gas flow rates at various sampling points
along the length of the system when sampling at 1 cubic foot
per minute (cfrn). The probes were introduced into the
system in such a way that their inlet axes lay parallel to
and facing the exhaust gas stream and mid-way across the
cross-section of the exhaust piping (see Figure 4).
A variety of devices were used to collect particulate
samples from these probes. Exhaust effluent was drawn
through these devices by means of vacuum pumps. The
devices are outlined below:
a. Andersen Stack Sampler
Particle fractionation by aerodynamic size was achieved by
the use of an Andersen 2 eight-stage stack sampler operated at
a flow rate of 1 cfm. This was designed to function at
temperatures up to 1500ฐF. The sampler was heated by
a 230-volt band-electric heater controlled by a Variac
so that the temperature of the sampling device could be
maintained at that of the exhaust gas stream at the
sampling point. This technique was aimed at avoiding
precipitation of particulate matter due to thermal
quenching. In the Andersen Sampler the incoming gas
stream is forced through successive plates, each having
a number of jet orifices of the same size. The size
of the orifices decreased on each successive plate and
particle aerodynamic separation is thus achieved. The
Andersen unit used allowed separation of particles into
the aerodynamic size fractions shown in Table VII.

-------
-19—
FIGURE 4
INSTALLATION OF EXHAUST PARTICULATE
SAMPLING PROBE AND THERMOCOUPLE
thermocouple
2”
F--i
exhaust gas flow—- -—

-------
-20-
TABLE VII
CUT-OFF VALUES FOR ANDERSEN STACK SAMPLER
Stage Value (microns )
0 9.5
1 6.1
2 4.0
3 2.8
4 1.75
5 0.9
6 0.54
7 0.36
*D 50 = mass median equivalent diameter
b. Gelman Filter
Gelnian filters were used for sampling both directly
from the exhaust stream and as a back-up to the Andersen
unit. This filter unit consisted of a 2-5/16 inch
diameter stainless steel filter holder (Model 2220)
containing a Gelman Glass Fiber Filter Pad, Type A of
2-inch diameter. This filter system has an efficiency
rating of greater than 98 percent for particles as small
as 0.05 microns. These filters were not heated as was the
Andersen unit.
The glass fiber filter pads were held in a dessicator and
weighed to the nearest 0.1 milligram (my) immediately
prior to use. After sampling they were returned to the
dessicator and weighed periodically until two successive
weighings showed no change. This technique assured that
the sample was free from moisture and that a reliable
determination of the mass of collected particulate matter
could be made.

-------
-21-
Since the Gelman filters were not heated in use, there
exists the real possibility that condensation of volatile
organic matter from the exhaust effluent could occur at
the Gelman filter pads. Therefore, the only true indication
of the presence of actual organic particulate matter within
the exhaust system would result from analysis of those
particulate samples collected at the heated Andersen stack
sampler.
c. Cold Traps
As a final back-up for the Andersen and Gelman filters,
the exhaust effluent sample was drawn through one or
more cold traps. The cold traps were immersed in an
ice/water cooling mixture although the effect of using
a dry-ice/acetone cooling medium was also evaluated.
In some instances, samples were also obtained from the
surfaces of the sampling probes by sweeping the latter
with a pipe cleaner.
The dispostion of the various sampling devices described
above is outlined for Tests I-V in Table IV.
The major objective of the sampling work undertaken in
Tests I-V was to establish a reliable profile and analysis
for particulate matter along the length of the exhaust
system. This profile was then used as a basis for selection
of a suitable trapping medium for removal of particulate
products in the exhaust system.
In order to evaluate the effect of a candidate trapping
medium on the nature of the residual particles emitted
from the exhaust into ambient air, it was necessary in
subsequent testing to subject the exhaust system
effluent to air dilution as described in the subsequent
section and in Reference 1.

-------
-22-
Sampling of Air-Diluted Exhaust Effluent - In the second
set of tests (VI-XI, Table VI), the effluent from the
engine exhaust system was passed into an air dilution
tube prior to sampling. The dilution tube is an extruded
pipe made of polyvinyl chloride. It is 16 inches in
diameter with a 1/4 inch wall thickness. In total it is
27 feet long which includes a 90ฐ bend section, as shown
in Figure 5. A 7-foot air induction head which houses the
air filter assembly, exhaust inlet elbow, and mixing baffle
is mounted in the wall between the instrument room and the
engine test cell. The mixing baffle consists of a sheet
metal donut attached to the inside wall of the air induction
head which acts to force the incoming diluent air through
the hole in the baffle at the center of the tube. The exhaust
inlet elbow enters at 90ฐ to the tube axis and is bent
90ฐ so that the flow axis of the exhaust gas parallels the
axis of the dilution tube. The exit end of the exhaust
elbow is in the same plane as the baffle. A schematic
diagram of this apparatus is shown in Figure 6.
An exhaust fan is located at the exit end of the dilution
tube. A throttle plate is located in the dilution tube
exit assembly just following the fan in order to allow
control of air flow volumes through the tube. The
dilution tube is of several sections with butt joints
which are taped during assembly. This construction
allows for easy removal, cleaning, and inspection of
the complete dilution tube after each run.
Several small slits have been cut in the bottom of the
tube along its length. Special glass collecting plates
were fabricated which are attached to the outside of the
tube under each slit to collect particulate samples.
Such samples are referred to herein as “slit samples.”
Slit locations are shown as short, dashed lines perpen-
dicular to the tube axis in Figure 6.

-------
V
2’l”
/
n
V ____
Point of dilution
—
A
8’
‘v B
X 6’
19 t1p L
- _ 6 1O ’__ -—---— --
-It ’—
‘3,
FIGURE 5
PARTICULATE SAMPLING TUBE
5’
4,
- --
10’ 7’
.CandC’ E
ID
, ) , ,t,
(A)
Andersen
sampler
________-— --7 !_ -
A through G indicate positions of slit samplers

-------
fl Air out
1’
Engine Test Cell — ----- --4 Instrument Room __
flow
Particulate gravimetric control
Fi ter fallout
- Mixing
- ;1
i
>in
- —— - - D - - - - —‘
-> 1 __/••
L -‘
sampling slits Air
pump
I’ f
Tailpipe
Engine I ’
Dyno
Andersen —3
Separator
Standard Muffler
- Millipore
—- ,Scott NO and NO 2 filter J
analyzer
V
Engine - - Lfrnanomete
-- —— Fisher gas Partitioner
CO, C0 2 , N 2 , 02 Flow
- —7 Beckman lO9A meter
Total Hydrocarbon
/ - - Analyzer
FIGURE 6
ExiLust FLOW DIAGRAM FOR PARTICLE COLLECTION I Vacuum
FROM AIR DILUTED EXHAUST pump
pipe

-------
-25-
The diluent air coming into the tube is filtered by means
of a Dri-Pak Series 1100 Class II PIN 114-110-020 untreated
cotton filter assembly. This filter assembly is 24” x 24”
and has 36 filter socks which extend to 36 inches in length.
This filter will pass particles O. 3 p in size and smaller.
Pressure drop at 600 cfm flow rate is minimal.
The particulate sampling zone for particles smaller than
l 5 p is located at the exhaust end of the dilution tube.
Three sample probe elbows are located in the exhaust-air
stream. One probe is connected to an Andersen Impact
Sampler Model 0203, a filter assembly, and a vacuum pump,
in that sequence. The probes are 0.754 inch ID aluminum
tubes. These sample tubes are located as shown in Figure 6.
A mercury manometer is connected between the dilution tube
probe and the exhaust side of the filter assembly, and was
used to monitor and regulate the flow through the Andersen
Sampler during the course of each run. The remaining sample
probes were both connected to 4 cfm Millipore filter holders
(142 mm) fitted with Gelman Type A fiberglass filter pads and
vacuum pumps. The Andersen Sampler collection plates were
weighed before and after the run, the difference being
sample weight.
Prior to use, the fiberglass filters were stored in the
instrument room which is temperature and humidity controlled.
The filters were placed on the tray in a Mettler Analytical
balance, allowed to reach equilibrium, and then weighed
out to 0.1 milligram.
After the test, the filters were removed from the holders
and again allowed to reach equilibrium, noted by no further
change in weight, and then weighed to 0.1 milligram. This
was done in the same room in which the papers are stored.
The Millipore filter pads used were 142 mm Type AAWP 0.8 .
The fiberglass filter pads used were Gelman O. 3 p Type A.

-------
—26-
2. Exhaust Gas Sampling
Throughout Tests I-V (Table IV), samples for gaseous
analysis were collected through the stainless steel
sampling probe described earlier (page 16 and Figure 2)
into a 2 cu ft saran bag via a Neptune dyna-pump. The
bag was filled with exhaust effluent and evacuated three
times prior to a sample being collected for analysis. A
water-cooled copper condenser fitted with a trap was
introduced between the exhaust sample port of the dyna-pump
to remove water from the gas stream.
In Tests VI-XI (Table VI), stainless steel sample probes
were also introduced into the exhaust system both ahead
of and after the muffler (Figure 3). Gas samples were
drawn from these ports through a cooled copper condenser
by a dyna-pump and fed directly to the gas analysis
equipment. A Gelman glass fiber filter fitted with a
back-up cold trap was also periodically connected
to the system via these same probes. This permitted
aldehyde analysis to be performed on the condensates
collected.
E. SAMPLE ANALYSIS
Particulate Matter Analysis
In order to characterize and define the various samples
of particulate matter collected throughout the course
of this investigation, a wide variety of analytical
procedures was employed.
One of the principal approaches used was that of microscopic
examination. This included the use of optical, scanning
electron, and transmission electron microscopy. The
application of these techniques to the characterization
of particulate emissions from the internal combustion engine
has been previously described by Moran and Manary 1 .

-------
-27-
Optical microscopy was used at magnifications from 3X
to 400X. Lower magnifications were used to determine
particle distribution and impingement points on the plates
of the Andersen Sampler. The higher magnification served
to determine particle size ranges and to study the
birefringent properties of the particulate matter under
polarized light. Particle samples examined by these
techniques were also submit.ted for X-ray diffraction
analysis.
Use of the scanning electron microscope at magnifications
of 1000X to lO,000X allowed examination of individual
particles. This instrument was also fitted with an X-ray
fluorescence spectrometer probe which afforded elemental
analysis of both single particles and particle aggregates.
Single element scans were also made to determine the
concentration profile of specific elements within a single
particle.
Particles of less than l i size were examined under the
transmission electron microscope at magnifications of
lO OOOX and 60,000X. The opacity of the particle image
on a photomicrograph affords a measure of the electron
density within the particle. The darker images indicate
high electron density resulting from the presence of
inorganic elements such as the heavy metals. The less
opaque images result from particles having a high content
of organics.
X-ray diffraction afforded a powerful tool for the
identification of specific crystalline species present
at greater than 5 percent concentration in the
particulate samplers. The resulting X-ray diffraction
patterns could be compared to those of over 10,000
known compounds. The quantitative accuracy of this
procedure is about ฑ10 percent based on the amount of
a specific moiety present.

-------
-28-
Samples of cold-trap condensate obtained in Test I-V
were analyzed by emission spectroscopy to afford a
qualitative and quantitative determination of the various
cations present.
A limited number of particulate samples were analyzed
by atomic absorption spectroscopy to determine the
relative amounts of lead and other inorganics present.
Detailed accounts of the above analytical procedures
can be found in Appendix A.
2. Gas Analysis
The engine exhaust gas was analyzed for oxygen, nitrogen,
carbon monoxide, carbon dioxide, total and unsaturated
volatile hydrocarbons. The analysis is done by gas
chromatography, chemical absorption, and a total hydrocarbon
analyzer. Data reduction is by an IBM 1800 computer through
a Bell Telephone ASR 33 Teletype interface.
Analytical Equipment
A Fisher Gas Partitioner is used for the analysis of oxygen,
nitrogen, carbon monoxide, and carbon dioxide. The column
system consists of a 6-foot hexamethyl phosphoramide
and a 6-1/2 foot, 13x molecular sieve in series.
Total hydrocarbons are obtained from a Beckman Model 109A
Total Hydrocarbon Analyzer. The concentration of unsaturated
hydrocarbons is determined by passing the sample through
an absorption tube 8” x 12” filled with 30—60 mesh pink
Chromosorbฎ impregnated with 50 percent mercuric perchiorate.

-------
-29-
Standard ization
A gas mixture containing known concentrations of oxygen,
nitrogen, argon, carbon monoxide, carbon dioxide, and
n-hexane is used as a reference standard for the total
hydrocarbon analyzer and the Fisher Gas Partitioner.
Component identification is done by peak retention time.
Shifts in retention time and changes in detector
response are compensated for by standardization
preceding each series of samples.
Operation
The operator types the proper computer code and program
number on the teletypewriter, injects the reference
standard, and presses the integrator start button. As
the peaks emerge, the time and area information is
encoded and stored on punched paper tape. Each succeeding
exhaust gas is identified along with the total hydrocarbon
level, and run in the same manner as the standard. When
the series is finished, the punched tape is sent to the
computer by teletype over regular telephone lines.
Data Reduction
A typical output format for the gas analysis is shown in
Figure 7. Identification of the components in the standard
is based upon each peak size in descending order. Estimated
retention time is the up-dated time of each peak in the
standard. Retention time windows are 4 seconds plus
2 percent of the retention time. Actual percent is a
direct ratio of the area counts in the unknown sample to
the area counts in the standard times the volume percent
in the standard. The total percent actual will normally
be 97-98 percent since water is removed from the saturated
sample after the sampling valve.

-------
-30-
FIGURE 7
G. C. ANALYSIS - TECHNICAL DATA -
GOV RUM S 1778-3 OCT 20 1970
TIME 8.0 MRS
MC 670. PPM
8?
10•
8.
18
5p
0.9
82.0
2.8
3.62
10.7
COMPOUND
IDENTIFICATION
100.000 TOTALS
BALANCE BY DIFFERENCE
TOTAL CONTAMINATION LEVEL
100.0 TOTAL
FR TI W CARBON IN FUEL 0.8625
T๘T ’L w’DROCARBON CONTENT 670. PPM.
PEAI( TIME
P40. ACT.
10-20-70
PCI. VOL.
EST. ACTUAL NORM.
1
22.
22.
0.000
0.o00
COMPOSITE
2
58.
58.
10.302
10.661
CARBON
3
82.
82.
2.723
0.900
2.817
0.931
OXYGEN
ARGON
m
103.
103.
79.210
83.973
NITROGEN
MONOXIDE
S
182.
182.
3.494
3.615
CARBON
96 • 628
3.371
3.371
EXHAUST GAS ANALYSIS 10-20-70
Ggv ‘UN 5 1778-3 OCT 20 1970
TIME .0 MRS
HC 6 0. PPM
TIP ERCEN1 iDENTIFICATION
ARGON
NiTROGEN
OXYGEN
CARBON MONOXIDE
CARBON DIOXIDE
AI.P FU RATIO 14.8

-------
-31-
A correction for the unresolved argon in oxygen is made
based upon response factors and the amount of argon found
in a number of exhaust gas samples by mass spectroscopy.
The actual percent is normalized to 100 percent in the
next column on a moisture-free basis, and an Exhaust
Gas analysis report is issued as shown in Figure 7. The
air—to-fuel ratio is calculated from this analysis, the
total hydrocarbon content, and the percent carbon in the
fuel
Oxides of Nitrogen
Equipment: Beckman Ultraviolet Analyzer
Beckman Infrared Analyzer
Recorder — Texas Instrument Company
The above pieces of equipment are in a single, self-contained
unit built by Scotts Research Labs, Inc., San Bernadino,
California.
Calibrating Gases:
Nitric oxide (3545 ppm in nitrogen)
Nitrogen dioxide (862 ppm in nitrogen)
These standard gases are furnished by Scotts Research Labs,
Inc. Nitrogen is used as zero calibrating gas.
Procedure:
Before making NO, NO 2 measurements, the paper filters
(Whatman #3) to each analyzer are changed and the Drierite
dryer in the exhaust sample line is replaced.
Both analyzers are standardized using the appropriate
calibrating gas at a constant flow by use of a flow
meter. The zero standardizing is done using nitrogen as
the calibrating gas using the same flow rate.

-------
—32-
As the instrument is standardized, the exhaust gas is then
passed through the analyzer using the same constant flow rate
as in the standardization step. The NO, NO 2 values are
recorded by the dual pen Servo-riter recorder.
3. Aldehyde Determination
Samples of cold trap condensate collected both ahead
of and behind the muffler in Tests VI—XI were analyzed
for their aldehyde content by the polarographic procedure
described in Appendix A.

-------
-33-
IV. EXPERIMENTAL RESULTS
A. DETERMINATION OF EXHAUST SYSTEM TEMPERATURE PROFILE -
VEHICLE TESTS
Temperatures measured for the Chevrolet exhaust system during
the period of ten days of road testing are presented in
Tables VIII-XII. This data is summarized in Figure 8 which
shows a plot of the exhaust system temperature profile as a
function of road speed. Figure 9 shows a plot of vehicle
road speed against engine rpm’s.
TABLE VIII
EXHAUST TEMPERATURE PROFILE - ROAD TEST CAR
Temperature Profile (ฐF)
Manifold Ahead of After End of
_____ ______ ________ Outlet Muffler Muffler Tail Pipe
943 529 406 328
988 575 461 367
1047 623 470 373
1057 652 515 392
1180 772 618 501
1165 726 605 508
1321 968 802 *
1321 942 802 *
Weather Conditions: *Thermocouple malfunctioning
Temp - 75ฐF **Car direction
Wind - SW 7 mph
Humidity - 37%
Clear
Intake
Road
Engine
Manifold
Speed
RPM
Vacuum
30 mph
1200
19”
**
E
w
E
W
E
W
E
W
40 mph 1575 19”
50 mph 1860 -
70 mph 2250 14”
15
Barometer - 30.18” steady

-------
-34-
TABLE IX
EXHAUST TEMPERATURE PROFILE - ROAD TEST CAR
Wind - 5-10 mph
Humidity - Intermittent drizzle
Barometer - 30.11” steady
TABLE X
EXHAUST TEMPERATURE PROFILE - ROAD TEST CAR
1200
1 200
- E 1086
- W 1107
- E 1187
- W 1210
623
720
799
782
382
440
454
472
Weather Conditions:
Temp - 58ฐF
Winds - Variable
Humidity - 87%
Cloudy
**Car direction
Road
Speed
50 mph
60
70
80
70
Engine
RPM
1875
2150
2150
2525
2525
2875
2850
End of
Tail Pipe
480
558
560
Intake
Manifold
Temperature
Profile (ฐF)
Manifold
Ahead of
After
.
Vacuum
Outlet
Muffler
Muffler
17.5
E
1150
755
610
16.25
W
1235
880
690
16.5
E
1240
850
705
15.0
W
1315
955
818
680
14.5
E
1285
920
780
645
12.5
W
1380
1070
900
760
13.75
E
1375
1060
910
760
2500
15.5
W
1310
930
790
655
2500
15.5
E
1315
945
790
640
Weather Conditions:
Temp - 77ฐF
**Car direction
Engine
RPM
Road
Speed
Warm up
30 mph
Intake Temperature Profile (ฐF )
Manifold Manifold Ahead of After End of
Vacuum Outlet Muffler Muffler Tail Pipe
206 166 154 132
— E 900 576 - 325
W 985 564 - 348
40 mph 1475
1550
50 mph 1900
1 900
Barometer - 30.22” decreasing

-------
-35—
TABLE XI
EXHAUST TEMPERATURE PROFILE - ROAD TEST CAR
Temperature Profile (ฐF)
Manifold Ahead of After End of
Outlet Muffler Muffler Tail Pipe
E 1230 745 680 540
W 1320 800 670 650
E - 970 930 800
880 500 460 360
W 980 580 570 400
E 1160 780 640 560
W 1160 880 640 510
Barometer - 3O.O7
TABLE X II
EXHAUST TEMPERATURE PROFILE - ROAD TEST CAR
Road
Speed
60 mph
70
80
Idle
50
60
70
Engine
RPM
2150
2500
2900
800
1850
2200
2550
Intake
Manifold
Vacuum
17.75
19.00
10.00
19.5
18.25
16.5
14.5
Weather Conditions:
Temp - 69ฐF
Winds - Westerly at 5 mph
Humidity - 65%
Cloudy
**Car direction
Engine
RPM
1875
2200
2150
Road
Speed
50 mph
60
70
80
70
Intake
Manifold
Temperature
Profile (ฐF)
Manifold
Ahead of
After
End of
Vacuum
Outlet
Muffler
Muffler
Tail
Pipe
17.5
E
1150
755
610
480
15.0
W
1235
880
690
558
16.5
E
1240
850
705
560
2550
12.5
W
1315
955
818
680
2525
15.0
E
1285
920
780
645
2875
12.5
W
1380
1070
900
760
2850
13.75
E
1375
1060
910
760
2500
15.5
W
1310
930
790
655
2500
15.5
E
1315
945
790
640
Weather Conditions:
Temp - 90ฐF
Winds - SSW at 11 mph
Humidity - 76%
Partly cloudy
**Car direction
Barometer - 29.80” steady

-------
-36-
1
•
4
jJ11 J 1 j i 1-t4
4ti t ; 1 1 i ‘
: ,Ll : 4 ii i:: 1 r
1 i r
i i :!
$ i itT tU
I r qi1 1UUU
f *t
J . . 4 f:i 4 Ti 1

- r t E 9 E E
q + :1 4 -
-:- _ =4=z-
ti- - 1 -k1:ttrn :;1:1t t-::
Ft
i j H’ ! 1 .
T






r=
L :
1

H
f HTr 4
t t4 ‘ t
Ill
ii—

j• :: :
Eฑ
=-i•=::- ::= — =::t-- :: . .
: ::: :::t :: : :_:: :
:

41]
. =i 2:4= I

, N t f
1 - -f - -•
: ::::
-
_i I-
-t4 +1
-::-4: :::
m
___1.___4= ::::i :. :
:: . EiE: E
F -:4 :: : t: T:
_.:1 ::: : 1 : .::E i:: : 4 ::
f + H+H+H+ F*ฑ
J . tE r :
: :
r.
)- e,
U
U
,- w
9:;
x x
o,
L J
x
::
-
rt Th
t

ii_L [ L i

___
Ii : iv

Lifl

—

4i !
: T1 t :::
;-
-
..
c fJ :5:

E :
::— -

. : : .: ::

E L: -f
H- i
.44 . -
V
.
- iii
44+
H-;
u 4::: ::::1 j:...
:: -: -=I: - :: . :t: ::: T:
__. —. -; - —4.-—--- -4 i —i-• - —i
- i+-:r -t-
-• •1
II
I
-I-
t
IlL
- 1 Lj
[ II
01
c4 ‘P
t .
,:, .
bW e•I •I i .

-------
-37-
fIGURE 9
ENGINE SI’EED vs ROAD SPEED
(197Q Chev o1et Car)
44... Idle - 8OO RPt+
//
/
,/
I i I i I i I
4
Ve iic1e Spee i (mph)
60 80
30O0-
p.
: ‘
200O
Engine
RFM
1000
2D

-------
-38-
FIGURE 10
EXHAUST SYSTEM TEMPERATURE PROFILE
DYNM4OMETER TESTS
(197& Chevrolet V8_350 CID)
Manifold
I
Muffler
.1
Ta ilplpe
ENGINE
2125
2250
1615
1070
800
I I
0
100
200
1500.
1000
I - ’
0
4-)
500
RPM
e
300
Distance along exhaust system (inches)

-------
—39-
B. DETERMINATION OF EXHAUST SYSTEM TEMPERATURE PROFILE -
DYNAHOMETER TESTS
Data generated in laboratory Tests I and II at five different
engine operating speeds is presented in Table XIII and
graphically in Figure 10.
TABLE XIII
CHEVROLET EXHAUST SYSTEM TEMPERATURE PROFILE
DYNAMOMETER TESTS
Temperature by
Thermocouple Pos i tion
Observed Fuel Vacuum (ฐF)
RPM H.P. Lbs/hr Inches Hg I II III IV
800 1.8 8.4 20.3 855 430 325 234
1070 15.0 11.3 16.2 883 657 552 426
1615 21.5 18.0 17.2 996 780 685 560
2125 42.3 31.0 14.4 1120 950 870 740
2250 27.5 29.2 17.2 1109 921 826 690
C. EXHAUST STREAM ANALYSIS
For clarity of presentation the data generated from exhaust
stream sampling will be described in two sections. The first
deals with those samples collected directly from the exhaust
system in Test I-V. The second presents similar data for the
analysis of air diluted exhaust stream samples (Tests VI-XI).
Data is presented on a run by run basis.
Direct Exhaust Stream Analysis
a. Particulate Mass - Weights of particulate matter
collected via the four exhaust system sampling probes are
shown for a variety of engine operating conditions in
Table XIV. These are discussed in the “Discussion of
Results” section.
b. Particulate Analysis
Test I - The objective of Test I was to evaluate the
sampling procedures chosen and to determine the effect of
variations in engine operating conditions on the exhaust
system particulate profile.

-------
-40-
Position
I
II
III
IV
I
II
III
IV
I
II
III
IV
I
II
III
IV
II
III
IV
I
II
III
IV
I
II
III
IV
I
II
III
IV
Temperature
(ฐF)
655
430
325
234
883
657
552
426
996
780
685
560
fl20
950
870
740
1109
921
826
690
1125
925
825
690
1125
945
850
725
1100
905
810
675
Andersen Sampler
and/or
Gelman Filter
(gm/mile)
0.0394
.0403
.0357
.0475
.0727
.0342
.0161
.0346
.0976
.07 09
.0601
.0565
.245
.316
.477
.513
.0298*
.0765
.0448
.00474
.0720*
.0859
.0590
.0737*
0059
•l65
.0155*
• 0091 6
.0155
Cold Trap
Condensate
(gm)
21.02
26.55
17.38
37.46
19.80
29.07
24.78
38.72
19.18
28.50
23.38
27.12
24.12
34.19
27.37
24.02
66.63
89.73
13.16
111.45
94.35
205.54
1.13
31.48
63.27
11.91
1.82
13.54
(a) Sample contaminated by metal from back-up plate
*Andprcpn and Galman (b Thraa t n1d trans in carias
TABLE XIV
SAMPLING TEMPERATURE AND PARTICULATE SAMPLE WEIGHTS
Test RPM
I 870
1070
1615
2125
II 2250
III 2250
IV 2250
V 2250
242 -
30.6899 (b)
4. 0762)

-------
FIGURE 11
PARTICULATE SAMPLING DEVICES - TEST I
I II III IV
85” - 28” 61”
I I
- - _____J_
- MUFFLER —.
ENGINE
Glass fiber
filter
ice/water
cold trap
Vacuum Vacuum
pump pump

-------
TABLE XV
SUMMARY OF OBSERVATIOIiS ON PARTICULATES COLLECTED AT POSITION I
(Exhaust Manifold Outlet)
Overall
appearance
Crystal growth
on glass fibers
Appearance with
polarized light
Fine, light colored
particles only.
Very evident
Supports above
conclusions.
1070 RPM
Heaviest at center
and outer edge with
sparse concentration
in between.
Fine, light colored
particles plus a few
larger dark particles.
Very evident
Shows the presence
of some dark, opaque
non-birefri ngent
particles in addition
to the transparent
birefringent
crystals.
1615 RPM
Heaviest at center
and outer edge with
sparse concentration
in between.
Fine, light colored
particles plus a few
larger dark particles.
Very evident (some
of growth is
abnormally dark up
where the deposit
is light.
Shows the presence
of some dark, opaque
non-b irefri ngent
particles in addition
to the transparent
birefri ngent
crystals.
2125 RPM
Heaviest at center
and outer edge with
sparse concentra tion
in between.
Fine, light colored
particles plus a few
larger dark particles.
Very evident (some
of growth is
abnormally dark up
where the deposit
Is light.
Shows the presence
of some dark, opaque
non-b irefringent
particles plus some
dark red birefringent
particles (probably
Fe ,O ) as is the
tr n parent
birefringent
crystal s.
Particle size
X-ray diffraction
identification
Pb(Br,Cl )2
Br:Cl “1:1
Pb(Br,Cl ).,
Br: Cl ‘ l l with
lesser amount of
unidentified phase
and 5% metallic
lead. (This is the
first time that
metallic lead has
been detected as such.
Pb(Br,Cl )2
Br:Cl ‘ l:l
Pb(Br,Cl )2
Br:Cl ‘ i:l with
lesser amount of
unidentified phase
and 5% metallIc
lead.
Distribution on
filter
870 RPM
Uni form
<1 —2O

-------
TABLE XVI
SUMMARY OF OBSERVATIONS
ON PARTICULATES COLLECTED AT POSITION II
(Muffler Inlet)
Distribution on
filter
870 RPM
Uniform, appears to
be continuous
fused deposits.
1070 RPM
Uniform and does not
have fused
appearance.
1615 RPM
Uniform, does not
have fused
appearance, but is
a heavier deposit
than found for
1070 rpm.
2125 RPM
Uniform, does not
have fused
appearance, but is
a heavier deposit
than found for
1615 rpm.
Overall
appearance
Only a few foreign
particles observed.
Only a few foreign
particles observed.
Only a few large
dark particles
observed.
Only a few large
dark particles
observed.
Crystal growth
on glass fibers
Evident but not as
evident as at
manifold outlet.
Relatively small.
Moderate on exposed
fibers.
Heavy on exposed
fibers, but most
fibers are covered
by deposit.
Appearance with
polarized light
Particle size
Lower birefringent
than observed on
Series I and only a
trace of foreign
particles.

-------
TABLE XVII
SUMMARY OF OBSERVATIONS ON PARTICULATES COLLECTED AT POSITION III
(Muffler Outlet)
Distribution on
fi 1 ter
Overall
appearance
870 RPM
Non—uniform with
certain small areas
having no deposit.
Matted appearance
moderate number of
larger dark particles
found only in
certain areas.
1070 RPM
Non-uniform with
certain small isol-
ated areas having
almost no deposit.
Lesser deposit than
at 870 rpm.
A few larger dark
particles are
present.
1615 RPM
Uniform deposit except
for several Isolated
points which have no
deposit. Heavier
deposit similar to
that at 870 rpm.
Closely matted
deposit. A few larger
dark particles are
present.
Uniform deposit except
for several isolated points
which have no deposit
Heaviest deposit of
the III series.
Matted appearance
surface pitted due to
presence of many larger
particles — some light
and some large dark
particles.
Crystal growth
on glass fibers
Appearance
Particle size
X-ray diffraction
identification
Crystal growth on
glass fibers varied
from light to
moderate depending
upon area observed.
Low bi refringence.
<1 —5
Pb(Br,Cl )2
Br:Cl 1:l
Unidentified phase
which is different
than found in 1070 rpm
and 2125 rpm.
Light crystal
growth on glass
fibers.
Low birefringence.

-------
TABLE XVIII
SUMMARY OF OBSERVATIONS ON PARTICULATES COLLECTED AT POSITION IV
(End of Tail Pipe)
Distribution on
filter
Overall
a ppearance
870 RPM
Fairly uniform deposit
except for a few
irregular, nearly
void spots.
Matted surface with
numerous larger light
particles as well as
larger dark particles
embedded in the
matted surface.
1070 RPM
Fairly uniform deposit
but less than at
870 rpm.
Not sufficient
deposit to have matted
appearance. Dark
particles are smaller
than on other filters
but they are numerous.
1615 RPM
Fairly uniform deposit
similar to that at
1070 rpm.
Dark particles and
other larger agglo-
merates are larger
than found after the
muffler at 1070 rpm.
2125 RPM
Heaviest deposit of
the series.
Matter surface with a
heavy deposit of large
dark particles as well
as a moderate amount of
medium sized light
particles.
Crystal growth
on glass fibers
Appearance with
polarized light
Particl esi ze
X-ray diffraction
identification
Most glass fibers are
covered by deposit
but a few exposed
fibers do show light
crystal growth.
In addition to low
bi refri ngent
particles, there is
a trace of high
bi refri ngent
particles and up to
5% of reddish or
blackish particles.
<1 — 8 p
Pb(Br,C1 )2
Br:Cl %l:1
plus 5-10% of some
unidentified phase as
found after the
muffler at 870 rpm.
Slight crystal growth
on fibers.
In addition to low
birefri ngent
particles, there is
a trace of high
bi refri ngent
particles and up to
5% of reddish or
blackish particles.
<1—1 2 u
Pb(Br,Cl )2
Br:Cl ‘ .l:l
plus 10—15% of some
unidentified phase as
found after the
muffler at 870 rpm.
More extensive than
observed at 1070 rpm.
In addition to low
birefri ngent
particles, there are
1-5% highly bire-
fringent particles
and about 5% of
reddish and blackish
particles.
<1 —8 i
Pb(Br,CI )2
Br:C1 1;l
Moderate crystal growth
on exposed fibers.
In addition to low
bi refri ngent
particles, there are
1-5% highly bire-
fringent particles
and up to 5% of reddish
and blackish particles.

-------
-46-
TABLE XIX
OBSERVATIONS OF CRYSTAL GROWTH AGAINST DECREASING EXHAUST
GAS STREAM TEMPERATURES
X-ray Diffraction
Species
Pb(Br,Cl) 2 11
<5% Pb, 5-10% unknown
(Compound “A”)
Pb(Br,C1) 2 “1:1
Pb(Br,C1) 2 ‘ 1.5:1
(M. Pt. PbC1 2 )
Pb(Br,Cl) 2 1:1
Pb(Br,C1) 2 ‘ 1.5:1
Fe 2 0 3
Pb(Br,C1) 2 ‘ 1:1
Pb(Br,C1) 7 ‘ 1:1
(M. Pt. PBBr 2 )
Pb(Br,C1) 2 1.5:1
Pb(Br,C1) 2 1:1
Pb(Br,C1) 2 ‘ 1:1
Pb(Br,C1) 2 ‘ 1:1
Pb(Br,C1) 2 1:1
Pb(Br,C1) ‘ 1:1
10—15% un nown
(Compound “B”)
Pb(Br,C1) 2 ‘ l:1
Pb(Br,C1) ‘ 1:1
10—15% un nown
(Compound “B”)
Pb(Br,C1) 9 l:1
5-10% unknown
(Compound “B”)
Dendri tic
Crystal Growth
Very evident
Very evident
Heavy on exposed fibers
Very evident
Very evident on
exposed fibers
Moderate
Moderate
Minor
Relatively small
Very evident
Moderate
Light
Slight
Evident
Light to moderate
Light
Position
I
RPM
2125
Tern
(ฐF
1120
I 1615 995
II 2125 952
934
I 1070 880
III 2125 870
II
IV
III
II
I
IV
I II
IV
II
III
1615
2125
1615
1070
870
1615
1070
1070
870
870
780
740
702
685
660
655
560
555
430
430
325
IV 870 234

-------
Figure 12
THE EFFECT OF ENGINE SPEED ON THE
AT POSITION I (NEAR
PARTICULATE MATTER COLLECTED
EXHAUST MANIFOLD)
1070 RPM
870 RPM
400X
1615 RPM
2125 RPM

-------
-48 -
Figure
13
SEM STUDY OF PARTICULATE
AT FOUR LOCATIONS IN THE
870 RPM
1,000x
MATTER COLLECTED
EXHAUST SYSTEM AT
I , , ’
r
ii 4ax ‘.
Near Exhaust Manifold
A- series
Ahead of Muffler
After Muffler
End of Tail Pipe

-------
-49-
Figure 14
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT POSITION I AT 870 RPM
5000x
A
2 ,000x
10,000 x

-------
-50-
Figure 15
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT FOUR LOCATIONS IN THE EXHAUST SYSTEM AT
1070 RPM
Near Exhaust Manifold 1,000x Ahead of Muffler
B- series
After Muffler
End of Tail Pipe

-------
-51 -
Figure 16
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT POSITION I AT 1070 RPM
B
I f•
I . • —
‘ II
2,000x
.5,OO OX
D

-------
-52-
Figure 17
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT FOUR LOCATIONS IN THE EXHAUST SYSTEM AT
1615 RPM
Near Exhaust Manifold 1,000x Ahead of Muffler
c-series
After Muffler
End of Tail Pipe

-------
-53-
Figure 18
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT POSITION I AT 1615 RPM
1,000 x
2.000 x

-------
-54
Figure 19
SEM STUDY OF PARTI
AT FOUR LOCATIONS
CULATE
IN THE
2125
1,000x
MATERIAL COLLECTED
EXHAUST SYSTEM AT
RPM
Near Exhaust Manifold
Ahead of Muffler
D- series
After Muffler
End of Tail Pipe

-------
-55-
Figure 20
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT POSITION I AT 2125 RPM
2,00 Ox
2,0 00 x
10,0 0 Ox

-------
-56-
Figure 21
SEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT POSITION IV AT 2125 RPM
SEM STUDY OF UNUSED GLASS FIBER FILTER
1000 Ox
1 ,000x 200x
bIank
1 ,000x

-------
-57-
Figure 22
TRANSMISSION ELECTRON MICROSCOPY STUDY OF
PARTICULATE MATERIAL COLLECTED AT POSITION
• p ‘ %
.4 ..
‘ .4 .. .4 .
•1 •
.‘ 6 ..
/••, s• 7
ltf ’
ID K
a
a
Is, —
..
870
,1
4
4, “

I c
: . :i
RPM
* S IS.
I
I
I ,
. , “.4*
m
4
I
b
I. *
3
‘‘,? ‘.•
- ,
•1 4
S
S
60 K
a
V
*
4, 5
I
a

p j
1070 RPM

-------
-58-
Figure 23
TRANSMISSION ELECTRON MICROSCOPY STUDY OF
PARTICULATE MATERIAL COLLECTED AT POSITION I
a
‘I
I
4..
I
10 K
vw
‘-a..
V SI
I
4 . ,
41
1615 RPM
60 K
—I
•1 •
?
.
‘SI.
p.
S
. 5
mull
0
2125 RPM

-------
-59-
Figure 24
TRANSMISSION ELECTRON MICROSCOPY STUDY OF
PARTICULATE MATERIAL COLLECTED AT POSITION II
p.
< wa T
11 l . •.
• p 4
1 .’
4..
0
•1
.4
. .. ,
4
.4
‘4
I
Is,
(0 K
870 RPM
S. ft
I,
4
60 K
I
1070
RPM

-------
-60-
Figure 25
TRANSMISSION ELECTRON MICROSCOPY STUDY OF
PARTICULATE MATERIAL COLLECTED AT POSITION II
I .
a
L
2125 RPM
I0
K
60 K
1,
i
I /
F . •
I
4
4
ft
‘p

-------
r
870
‘a.
60K
0
.4’
-61-
Figure
26
i... all *
TRANSMISSION ELECTRON MICROSCOPY STUDY OF
PARTICULATE MATERIAL COLLECTED AT POSITION III
-a
.. iul ,
. ‘,
a.d ‘
I
• l
p
10K
I,
RPM
4
so
‘S
I.
$
• •
1c7o RPMb

-------
1
4’
1615 RPM
-62-
Figure 27
TRANSMISSION ELECTRON MICROSCOPY STUDY OF
PARTICULATE MATERIAL COLLECTED AT POSITION III
•
*
10 K
/ • 4
q 4
1 . •
-
‘I
b ‘
.g L
4,
60 K
4
4 p
. - .- ii ..
2125 RPM

-------
-63-
Figure 28
TEM STUDY OF PARTICULATE MATERIAL COLLECTED
AT POSITION
IV - TEST I
0
0
4 .
I
: i C 4
t e l I
— 1 ‘0
0 .
JO K
4
0
4 V
J 4w
870 RPM
:.
. .
It. • ‘ I
.,q -
• 0 .
‘ I
C ...
• .
7 .
— I .
. 1 -
• 60 I
• I
I - . .1 .
0 • ’
• .
— .
‘• .-:
V
‘ -I ’ :
S i t . • t • S a
- .
. -. .
1070 RPM
4
4
0
I ’
60 K
ผ
r
S
A
S.
V
V
S
t t
S
4
.4
•
0*
.0
I - • -
4 , -,
::
I
0 •
S.
•0*
0
0* ’
.0 *
4
a
S.
at
p

-------
-64 -
Figure 29
OF PARTICULATE
AT POSITION
i
MATERIAL COLLECTED
IV - TEST I
TEM STUDY
p
&
I
1615 RPM
tOM
60 K

-------
-65-
Samples collected at the four glass fiber filters (see
Figure 11) were examined by several analytical techniques.
Tabl s X through XVIII summarize the results of studies
on these samples by low power microscopy and X-ray
diffraction. The bulk of the crystalline material present
was found to be lead halides. Figure 12 shows typical
photomicrographs (400x magnification) of the Gelman filter
pads after exposure at sampling Position I for 30 minutes.
At the four different engine speeds the presence of uncombined
lead in low concentrations at the exhaust manifold outlet
during operation at 1070 and 2125 engine rpm should be noted
(Table XV). In Table XIX the crystalline species observed
are tabulated against decreasing exhaust gas stream
temperatures.
The collected particulate matter was next examined by
Scanning Electron Microscopy. Figures 13 through 21 show
photomicrographs of selected samples at magnifications
from l000x to lO,000x. In general those samples collected
at sample Position I, immediately after the exhaust
manifold, show crystal growth from nuclei on the
glass fiber filters. Samples taken at Positions II,
III, and IV along the exhaust system appear more matted and
in some instances, particle fusion can be observed (see
Figure 13). Use of the X-ray fluorescence probe confirmed
the previous X-ray diffraction studies and indicated lead
chlorobromides as the major particulate components. At 2125
engine rpm, several particles of high iron content were
observed. These probably originated from the exhaust system
itself rather than from the engine since they were detected
mainly on the tail pipe side of the muffler.
Transmission electron photomicrographs of the particulate
samples were obtained at 10,000x and 60,000x magnification.
These are shown in Figures 22 through 29. The particles
were dispersed in ethylene glycol prior to examination to

-------
-66-
eliminate matting. Many of the larger particles appear to
be composed of very fine inorganic matter of high electron
density in an organic matrix of low electron density (see
Figures 22 and 28).
The cold trap condensates collected at all four sampling
locations were examined by emission spectroscopy. The
data is summarized in Table XX. A large proportion of
the mass of condensed particles in these traps is
associated with lead and iron. The actual weight of
these elements present in the condensates in shown in
Table XXI. A profile for the mass of condensed lead
collected along the exhaust system under the different
engine operating conditions is shown in Figure 30. This
data illustrates the high level of lead compounds present
in the vapor state immediately after the exhaust manifold.
A surprising result is the relatively higher vapor phase
concentration of lead compounds at engine idling speed
(870 rpm) compared to those at 1070, 1615, and 2125 rpm.
Test II — Test II was designed to evaluate the efficiency
of the Andersen Stack Sampler for determining the particle
mass-size profile at sampling Position IV of the exhaust
system. Position IV is located close to the end of the
tail pipe and is thus subjected to temperatures lower than
those encountered in the remainder of the exhaust system.
Isokinetic (1 cfm), isothermal sampling was conducted at
an engine speed of 2250 rpm. This speed was selected so
the data generated could be compared to that of similar
tests performed under Government Contract CPA-22—69-145’.
Samples of particulate matter were examined from the sampling
probe, from Stages 0, 2, 5, 7, and 8 of the Andersen
sampler, and from the back-up Gelman filter (Figure 31).

-------
TABLE XX
EMISSION SPECTROSCOPIC ANALYSIS OF COLD TRAP CONDENSATE - TEST SERIES
Position
RPM
ppm:
Fe
Cu Al
Ca
Mn
Pb
Cr
Sn
Zn
Ti
I
870
5
4
1 0.5
3
0.3
0.9
0.4
257
0.4
<0.2
2
<0.2
I
1070
3
5
0.9 <0.2
3
<0.2
0.3
0.2
113
0.2
<0.2
3
<0.2
I
1615
3
0.7
0.6 <0.2
4
<0.2
<0.2
0.07
77
<0.2
<0.2
0.5
<0.2
I
2125
1
<0.2
0.2 <0.2
3
<0.2
<0.2
<0.05
88
<0.2
<0.2
<0.5
<0.2
II
870
18
10
0.05 <0.2
4
<0.2
0.8
0.5
29
1
<0.2
7
<0.2
II
1070
2
0.7
0.08 <0.2
3
<0.2
0.3
<0.05
10
0.2
<0.2
4
<0.2
II
1615
2
0.2
0.4 <0.2
4
<0.2
0.4
<0.05
11
0.4
<0.2
12
<0.2
II
2125
3
<0.2
0.5 <0.2
5
<0.2
4
<0.05
7
0.5
<0.2
18
<0.2
III
870
0.3
0.2
<0.05 <0.2
5
<0.2
<0.2
<0.05
3
<0.2
<0.2
0.5
<0.2
III
1070
<0.2
0.2
<0.05 <0.2
4
<0.2
<0.2
<0.05
0.6
<0.2
<0.2
0.7
<0.2
III
1615
0.2
0.2
<0.05 <0.2
4
<0.2
0.5
<0.05
3
<0.2
<0.2
16
<0.2
III
2125
<0.2
0.2
<0.05 <0.2
3
<0.2
<0.2
<0.05
3
<0.2
<0.2
3
<0.2
IV
870
0.2
0.2
<0.05 <0.2
5
0.2
0.5
<0.05
5
O.2
<0.2
<0.5
<0.2
IV
1070
0.2
0.2
<0.05 <0.2
3
<0.2
0.2
<0.05
8
<0.2
<0.2
0.5
<0.2
IV
1615
0.7
0.2
<0.05 <0.2
4
0.3
<0.2
<0.05
10
<0.2
<0.2
0.7
<0.2
—4
4 <0.2 <0.2 <0.5 <0.2
IV 2125
0.5 0.2 <0.05 <0.2 3 0.5 0.2 <0.05

-------
-68-
RPM
TABLE XXI
IRON AND LEAD CONTENT OF COLD TRAP CONDENSATE - TEST I
Sampling Position
I
870
Iron (ppm)
5
18
0.3
0.2
Condensate
(g)
21.02
26.55
17.38
37.46
Iron (mg)
0.105
0.4779
0.0052
0.00749
1070
Iron (ppm)
3
2
<0.2
0.2
Condensate
(g)
19.80
29.07
24.78
38.72
Iron (mg)
0.0594
0.0581
<0.004956
0.00774
1615
Iron (ppm)
3
2
0.2
0.7
Condensate
(g)
19.18
28.05
23.38
27.12
Iron (mg)
0.0575
0.0561
0.00468
0.01898
2125
Iron (ppm)
1
3
<0.2
0.5
Condensate
(9)
24.12
34.19
27.37
24.02
Iron (mg)
0.0241
0.1026
<0.0055
0.0120
870
Lead (ppm)
257
29
3
5
Condensate
(g)
21.02
26.55
17.38
37.46
Lead (mg)
5.702
0.76995
0.052
0.187
1070
Lead (ppm)
113
10
0.6
8
Condensate
(g)
19.80
29.07
24.78
38.72
Lead (mg)
2.237
0.2907
0.01487
0.3098
1615
Lead (ppm)
77
11
3
10
Condensate
(g)
19.18
28.05
23.38
27.12
Lead (mg)
1.477
0.3086
0.0701
0.271
2125
Lead (ppm)
88
7
3
4
Condensate
(g)
27.12
34.19
27.37
24.80
Lead (mg)
2.12
0.239
0.082
0.096

-------
-69-
FIGIJRE 30
LEAD CONTENT O COLD TRAP CONDENSATES
(TEST I)
6.0 -
ENGINE
RPM.
870
4.0
-o
Ll 1
0
I 1O7 -
2125h
2.0 - -
161
I -
I • - — _. ——% — —— —
0.0
I —. . _ I I
I 100 200
Distance along exhaust system (inches)

-------
FIGURE 31
PARTICULATE SAMPLING DEVICES-TEST II
Vacuum Pump
II III
Mu ff1 e r
Rotometer Glassfiber
Filter
Andersen
Filter
Glassfiber
Filter
U U
Cold Traps
I
Engine
85”
28””
-IV
61”
-- - - -—- ---.
Valve
H

-------
—71—
The results of an examination of these samples by optical
microscopy and X-ray diffraction are summarized in
Table XXII. Figure 32 shows photomicrographs at 15x
magnification of particulate matter collected on the
Andersen plates. The above analyses indicated that in
general, as particle size decreased through the Andersen,
the concentration of lead halide and particle birefringence
decreased. Optical microscopy indicated that particle
size varied from <1 to 20 microns on Plate 0 of the Andersen
to ‘ lp on the glass fiber filter.
The same samples were also examined by scanning electron
microscopy. Figure 33 shows photomicrographs at l000x
and 2000x magnification of particles collected from the
surface of the sampling probe. A single element X-ray
fluorescent scan for iron on Particle D shows clearly
that the distribution of this metal is non-uniform
(photomicrograph C). Multi-element scans indicated the
presence of lead, chlorine, bromine, and iron in all
the particles shown.
Figures 34-37 show Scanning Electron Microscopy (SEM)
photoniicrographs of particles collected on Plate 0
of the Andersen sampler.
Figure 34 is representative of the hard black deposits
encountered at Plate 0 of the Andersen. Multi-element
X-ray scans of the particle shown in “B” indicated high
concentrations of lead, chlorine and bromine at
Position R. Position G had low concentrations of these
elements but was high in iron. Zinc was also detected.
These same two types of elemental composition were observed
in particles scraped from Andersen Plate 0 (Figure 35).
Pictures B and D show particles which on X-ray scan
proved to be composed mainly of lead chiorobromide.

-------
TABLE XXII
EXAMINATION OF PARTICULJkTES FROM TEST II
Sampl e
Location Birefringence Size X-ray Diffraction
Sampling probe 75 wt. % highly birefrigence <1% below 1 Pb(Br,Cl) 2 - Br:Cl “1:1
l-l5 size; 10-20% black opaque
Andersen Sampler
Plate 0 25—50% highly birefringence <1% below li.’ Pb(Br,Cl)., Br:Cl “'1:1
Lighter l—2Oii size “R” form nd PbO•Pb(Br,Cl) 2
portion
Darker 25-50% black opaque Same except higher in
portion 5—2O size Pb(Br,Cl) ; no crystalline
fused, very hard, impossible to be iron comp und.
representative; high in dark reddish
particles of low transparency (possibly
iron compound).
Plate 2 <5% highly birefringent Major - PbSO 4
1—lOu size
“50% black opaque Medium - NiO
5-20 size
“ '5% dark reddish, transparent
probably iron containing
Remaining irregular agglomerates 
-------
—73-
Figure 32
DISTRIBUTION OF PARTICULATE MATERIAL ON ANDERSEN
SAMPLER PLATES AT POSITION IV - TEST II
050 6 .lp
B Plate 0 Plate 2
15X
050 >95
D
J
‘Plate 8
D 50
D 50 O. 36

-------
-74-
Figure 33
SEM STUDY OF PARTICULATE MATERIAL FROM
PROBE AT POSITION IV - TEST II
SAMPLING
2,000 X
1,000 X
C
L)

-------
-75-
Figure 34
SEM STUDY OF PARTICULATE MATERIAL FROM PLATE 0
OF THE ANDERSEN SAMPLER AT POSITION IV - TEST jj
B
1,000x
2,000 x
Fe
D

-------
-76-
Figure 35
SEM STUDY OF PARTICULATE MATERIAL FROM PLATE 0
OF THE ANDERSEN SAMPLER AT POSITION IV - TEST II
B
5,000 x
2,00 Ox

-------
-77-
Figure 36
SEM STUDY OF PARTICULATE MATERIAL FROM PLATE 0
OF THE ANDERSEN SAMPLER AT POSITION IV - TEST II
5 ,000x
B
.c.r -
:.,‘.
L 1 1-4
Plate•O
• 7 1
D
C

-------
-78-
Figure 37
SEM STUDY OF PARTICULATE MATERIAL FROM PLATE 0
OF THE ANDERSEN SAMPLER AT POSITION IV - TEST II
A B
L T 1 1
PIate O
2,0 00 x
HIL ,
D

-------
-79-
Particle R in Picture C was also lead chlorobromide.
Particle G in the same picture shows much lesser amounts
of this compound but showed a high concentration of iron
with smaller levels of nickel, copper, and zinc.
Figure 37 shows more examples of particles collected at
Plate 0. Of special interest are the particles shown in
Pictures B and 0. Picture B is again representative of
the hard black deposits which accumulated. Analysis proved
this material to contain lead, chlorine, and bromine.
Picture D is representative of the most commonly observed
particle type. An X-ray scan showed the presence of lead,
chlorine, and bromine.
Multi-element X-ray scans of the particle shown in
Figure 36-B indicate a non-uniform distribution of lead,
chlorine, bromine, and iron throughout the particle
(see Figure 36-C and D).
SEM analysis of the material collected on Plate 2 of the
Andersen separator again indicated the presence of lead
halides in combination with nickel and lower concentrations
of iron and zinc. Again single element scans indicated a
non-uniform distribution. flo photomicrographs were made
for particles collected on Plates 5 and 8 because of their
small size and mass. X-ray fluorescence analysis, however,
indicated high concentrations of lead, nickel, and iron
with lesser amounts of copper and zinc. Halide concen-
tration was low.
Particulate matter smaller than 1 micron was studied by
Transmission Electron Microscopy (TEN). Samples were
obtained from the Gelman filters following both the
Andersen and the cold trap (Figure 31) and from the
sample probe. Under 60,000x magnification these particles

-------
-80-
appeared as an organic matrix with inorganic matter
embedded on their surface. Typical photomicrographs are
shown in Figures 38-40. Some of the same types of
particle structure were observed as in previous studies
under Contract CPA-22-45-145’. These include the
“log-shaped” particle shown in Figure 38.
No particulate mass-size distribution was obtained in
this test since thermal decomposition of the asbestos
0-rings in the Andersen sampler led to erroneous
weights of the various Andersen plates.
Test III - This test was performed in an indentical
manner to Test II to establish a particle mass-size
distribution from the heated Andersen separator located
at Position IV of the exhaust system. However, in order
to avoid thermal decomposition of the asbestos plate separators
used in the Andersen, these were preheated to a temperature
some 50ฐF above that encountered at this sampling location
for a period of 3 hours before -sampling. The equivalent
mean mass diameter for particles collected on the Andersen
is plotted in Figure 41.
Test IV - The major objectives of this test were to
establish a mass-size distribution and characterization
for particles collected on the Andersen Impactor at
Position III of the exhaust system. Samples were also
obtained at the three remaining sample ports both on
Gelman filters fitted with back-up cold traps and by
sweeping the probes at the end of the test (Figure 42).
X-ray diffraction and X—ray fluorescent analyses of the
probe sweepings are shown in Table XXIII. X-ray diffraction
analysis of the material collected on the Gelman filters
and on Plates 0, 5, and 8 of the Andersen is presented

-------
-81-
Figure 38
TEM STUDY
COLLECTED
tO K
4
‘tp* I
I•t*.•\ t*
‘0
* - t
Sampling
OF PARTICULATE MATERIAL
AT SAMPLING POSITION IV
I .
F
p
L i
S .
I.,
so
ads
s .c
Glass
Fiber Filter
60K
we’
4%
R
4 ’
*
a
t o
*
4
I ’
I
Probe

-------
39
MATERIAL
AT POSIT
COLLECTED
ION IV
I
1”,-
1
-82-
Figure
TEM STUDY OF PARTICULATE
ON ANDERSEN SAMPLER
I ,
10 K
eO
I
60 K
S
—S
Plate 2

-------
-83-
Figure 40
TEM STUDY OF PARTICULATE MATERIAL COLLECTED
ON ANDERSEN SAMPLER AT POSITION IV
I
I
10 K
a
4
$1
P
•14”-
4. p
‘p
4 $
/
.4
I
Plate 5
60 K
• . . •
b
b •r
p a, ,
f
U
4
I ’:.
4
0
I
. ‘a_a
I
Plate 8

-------
LI’ f’I.l( RAI II Ii V 44 4O4 .1
X ‘I ’) ’.aV(tV ’-
99 9M
9999 599 99.8
‘I’, ‘ 1 ‘O it) 1 0 0 40 30 ‘() 10
-- .
-HH Fi<
u i
5
—I ::t ‘ I
CT
:41)4
I 0.5 0.2 0.1 005 001
_______________ ____________________ 8
T E -”EE t _J 1 i
Ji t:H
J 1,
‘1
8
7
0
5
4
2 - -
9_
8
7
3
2
c
--
4
I T
— - - H- -+ h
1.4 ::. .
H

e I - I
;L .1 T T 1fl

- H
I
-
—
itl
T


- i--- - - j- - - - -t
- -
LI
J
!:. . ‘I , . - .- - -- --‘, —
. : : —
3
T ‘-—t
ft iJ -
I - - T
I J f 1 1 ’
1L V ——
f [ _ EEt T
I I
- .
- - a
i__

- --i --i, -l - 1 --—t-— - -—:
- 44
U

t IL TIL ฑ
igi
ii i I,.
P
IrCE
nt Par ic
I’.
Les
P

-------
FIGURE 42
PARTICULATE SAMPLING DEVICES - TEST IV
II GA II [ II IV
1 ’ —
FFLER
ENGINE
U,
condensqr Glass fiber Andersen
filter —- separator
I cold-trap j
GA - gas analysis probe J

-------
-86-
TABLE XXIII
PARTICULATES ON PROBES IN TEST IV
X-ray Fluorescence
Position X-ray Diffraction ( Ratios only )
Br/Cl Zn/Fe/Pb
I Pb(Br,Cl) 2 Br:Cl ‘ l:l 0.9 1:9.8:92
Residual
Hydrocarbon Pb(Br,Cl) 2 Br:Cl l.5:l 0.6 1:2:3
Analysis Probe
II Pb(Br,Cl) 2 Br:Cl ‘ l:1 1.1 1:14:147
III Pb(Br,Cl) 2 Br:Cl l.5:l 0.86 1:3.3:2.9
Fe 2 O 3 5% or less
ZnO 5% or less
IV Pb(Br,Cl) 2 Sample too small
PbSO 4
Fe 2 0 3

-------
-87-
in Table XXIV. Similar data is also included for particles
collected on the glass fiber filter used as a back—up to
the Andersen.
Transmission Electron Microscopy (TEM) of these samples
revealed the same type of crystal structure reported in
Tests I and II.
The cumulative mass distribution for particles collected
on the Andersen is shown in Figure 43. Comparison of
Figure 43 to Figure 41 (Andersen at sampling Position IV)
shows a shift in the size of particulate matter to a
smaller equivalent mass median diameter for Position IV.
Test V - In Test V an attempt was made to establish
a particle mass-size profile for position II in the
exhaust system by means of the Andersen Sampler. However,
on examining the unit at the end of the test it was found
that not only had the sampler plates collected negligible
amounts of particulate matter, but they all had also lost
weight. In this test the Andersen was heated to 905ฐF,
equivalent to the temperature at exhaust Probe II. Passage
of exhaust effluent through the sampler at this
temperature resulted in extensive damage to the sampling
stages. Figure 44 illustrates the destruction which
took place at Plate 0 of this device. In view of these
findings, no attempt was made to locate the Andersen unit
at sampling Position I (near the exhaust manifold) where
the exhaust stream temperature was even higher.
c. Gaseous Analysis
Analysis of exhaust stream gases in Tests IV and V was
used for the calculation of engine air/fuel ratios. This
data is presented in Table XXV.

-------
-88-
TABLE XXIV
CRYSTALLINE PARTICULATE SAMPLES COLLECTED
FROM TEST IV
Position Crystalline Particulates
Pb(Br,Cl) 2 Br:Cl ‘ 1:1
II Pb(Br,Cl) 2 Br:C1 l:1
III (Andersen Sampler)
Plate 0 (light particulates) Pb(Br,Cl) 2 Br:Cl i l:2
Pb
Possibly Fe 2 (S0 4 ) 3 5-10%
(dark particulates) Pb(Br,Cl) 2 Br:Cl ‘ l:l 40-50%
ZnO 40—50%
possibly ZnC1•4Zn(OH) 2 .lH 2 O 5-10%
Plate 5 PbSO 4 40-50%
ZnO 20-30%
Fe 2 0 3 10-15%
possibly NiO 5-10%
possibly Fe 3 0 4 5-10%
Plate 8 PbSO 4 30-40%
ZnO 30-40%
PbC1(P0 4 ) 3 10-15%
Fe 2 0 3 10-15%
possibly NiO <5
possibly Fe 3 0 4 5-10%
Plate 8 (fused) PbSO 4
ZnO
Glass fiber filter Fe 2 0 3 40-60%
PbSO 4 20-30%
PbS 20-30%
IV PbSO 4
Fe 2 0 3

-------
LJ ’ : PROBABILITY 46 5043
‘ x 2 LC CYCI F ..,.. . *
in
99.9 99.8
99 9
95 rn R() 70 60 )O 4() 30
.o in 5
- : ฑ T
- -

9
q
-
7
6
5-
• -TrT- .:•r
4 —h -
• - - • •
H
-


-—
. i
----- — - -1 -

. *
: i: :i
H
p
::
‘H


‘;
—
:
- , -- -- ---- —
I _ _
t 1
r LII I U I
Ei Tf

-
1

;
:‘
Hi
: :
i _____
.:.;:.
FI _41
j jrj ri1
—-- -
.::::.,
_ i -
_o.5_ 0.2 0.1 0.05 9 1
UT
-
— -- -
-w fl :f
-1-
-t- r.t iiiEr:t tt
.ft F+ - -• —— — - I • I
9
. 1 t: H ::i t:
, I
Ii
-:1:

2
1ฑ
i- H
rf
! I4
, ‘ - t_t!I I-n-
7
..- -+... , ,
, . I I J
Gi -
U
I-i
Ii
- t -t1
--- -
H Th
H -
WI
I : ;
1
,f
t _ _4 --tn -
S

i_ : -
uJ I

ittt:
ill f
4
:: Tฑ-itiT- if
=1
F
g
2
to
3
T I I
I I
ฑ t
j4-
II
if

I
‘

7
6
5—-
4_ -
I_j f:E_: . i ฑi
T
t I
J . -
ij I
-
i

—‘E 4
I __ : 4::: ; : .:= = = = —
E E — — - E
-__:z - : - : - -_ - :::- =::: _ :
--- -+
I I I
. r
tT 1L
—‘-F-- -
I I
I I
i - :---E-
:
I


-
T ETT 4f
1
:-- : - - ; T:T =L:i= - i:::—:
ฑ . .
I

,_-
--
-

.1 I - : :
- - - -- - — ---
: ,
-
2_
I - i ii I _
g
t: - - — I I _
- , I —- —
4 : L-
- * _ Yi I ==
r r — T:i E:-- :
—-
E t
7
6
_ L
H - - —
- -iฑ T-
II
- I
4-1 - --------
L J - - -- _ - -.--
- - ----
! I
— -- -f--h
I I I I iI_ 1I III
-
Ii - :I
-Hฑ :LH:H
I I I I
f--fl I
tfr l-
I I I I
4_i ! _.
I I I I
- ----
—
0.01 0.05 0.1 0.2 0.5 1
—— I
I I I
--
I I
I I
4 4L
I I
L f

T:= II I 1i i


:
I 1tfI

:if ifi:i; iji•
--- -H-
i
10 20 30 40 5(1 50 70
9
99.8 ถ19.I
9 9_9 9
1__ -_
80 90 95 98 9

-------
-90-
Figure 44
DEGRADATION OF ANDERSEN SAMPLER PLATE
AT POSITION II (AHEAD OF MUFFLER)

-------
-91 -
TABLE XXV
AIR/FUEL RATIOS BY EXHAUST ANALYSIS
Test IV
8 hours
Test V
3 hours
4.75
Residual HC
670 ppm
670
555
555
575
575
575
Air/Fuel
13.4
14.8
15.0
14.4
15.0
15.0
15.0
2. Diluted Exhaust Stream Analysis
a. Particulate Mass
Throughout Tests VI-XI particulate matter in the air
diluted exhaust stream was collected on one or more 4 cfm
filters and at 1 cfm on an Andersen Impactor fitted with a
back-up Millipore filter. This sampling procedure has been
described earlier. Particulate mass measurements were made
from both sampling devices and the data is presented in
Table XXVI. A discussion of the disparity of results
obtained at the 1 cfm and 4 cfm filters is presented on
page 110.
TABLE XXVI
EFFECT OF LEAD TRAP ON MASS OF PARTICULATE MATTER
EMITTED IN AIR DILUTED EXHAUST
Test
No.
VIII
Engine
Condition Muffler
1200 Conventional
4 cfm Filter
(gm/mile)
Andersen and
1 cfrn Filter
(gm/mile)
0.096
0.125
VI
2250
Conventional
0.104
0.117
VII
Cycle
Conventional
0.140
0.150
IX
1200
Lead trap
0.109
0.143
X
2250
Lead trap
0.122
0.118
XI
Cycle
Lead trap
0.125
0.135

-------
-92-
At the start of Tests IX- .XI, the lead trap was allowed
to warm up by operating the engine for 30 minutes under
the specific test conditions prior to connecting the tail
pipe to the dilution tube. This insured that the contents
of the trap were molten.
b. Particle Analysis
Samples of particulate matter collected on the stages
of the Andersen Impactor were analyzed by atomic absorption
spectroscopy for their lead content. This allowed mass-
distribution profiles to be calculated for both the total
particulate matter collected in this way and for the leaded
particulates. The data are plotted in Figures 45 through 50.
Material collected on the 4 cfm Millipore filter was also
analyzed by atomic absorption spectroscopy for its lead,
potassium, and lithium content. The results of these
determinations are presented in Table XXVII on a gm/mile
ba s I s.
No analyses were performed on particulate matter which
accumulated within the dilution tube as a result of
gravitational fallout although the tube was cleaned between
runs.
TABLE XXVII
PARTICLE MASS EMISSIONS (4 CFM FILTER)
Engine Particle Mass (gm/mile)
Test RPM Muffler Total K Li Pb
VIII 1200 Conventional 0.096 0.0004 <0.00002 0.038
VI 2250 Conventional 0.104 0.00021 <0.00002 0.035
VII Cycle Conventional 0.140 0.00029 <0.00003 0.055
IX 1200 Trap 0.109 0.0016 0.00014 0.038
X 2250 Trap 0.122 0.0032 0.00045 0.019
XI Cycle Trap 0.122 0.0026 0.00058 0.043

-------
LJ PROSABILITY 46 8043
r x 2 LOG CYCL. i . . .
urri i. p (I
_
-.-“ ——----4---- - - H-.-
E :H T L’
H4 -
. ;- i-r •
f-—t —
70 60 50 40 30
L 4 L
Lu
o____
9--
8
7
99.9 99.8 99 9R
F
9S 90
c
I I
I I I I JIjI I JUJII1I I I 11 1iLI_ ) hIl 11 :‘T i1
F I I
t1 -T 1:Vt T 1”
20 10
r
1
•L-
F
t It
I 0.5 0.2 0.1 0.05
I -4-
0.01
_i
9
r: 4 Ui.
-* t41
—. — I
-- i_
- - + -
Li:L
4: J t
I : : :
H -- -- - --
uJ
. -4 , . -
-,- -t -
- - - -
- - - - - - + - -
::=:..: :
:: -TI
_i i 1_i , i
U
-‘p4 - - -
-. -
- - - - - - -
- 1--
----- -- --
h--

----- -
ฑL
- :I
I! I
- -!
-
ฑ - :t: 1 t
— --t -
_L -
--4-
8..
7-—
5--
4--
3-—
9
t I
7
5
4
3
2
3
7
.,
4
3
)
0
------ -H

I I I I III III IllIllIll
I I
II
I
I I I I
ill-_li
[
IiIIIHI]
W JU 1JJ LII LI; HT.
1 1 I
— — I --
T1 1*N’frI I ri’t
-.
4 .
r-t --•
- :
- I I
- -f -FI i!- - - - -
A L JL . L -‘
— —
, I
v-. - - I
I -
- I
nfl nn ni n, n i

-------
‘JOE PROBABILITY 48 8043
n x a LOG CYCLES ot IN U IA.
KEUFFEL S rnrp C o
In 99.99
9- ---
- *4 . -
- C l .:‘:h
•r+ ’ - -
_ = I_ r4 tt
:i4 :L 4
iLl : tfltt fl Tfl
i :
4 4— 1-- -- ฃ
- I -
99.9 99 3 99 98
-
Lz-: :2 =— . : - : : :
;--3- tttt :i4:-: :.n;4 : ; ; 4
I iiL :. 3 tt b E 1 E’: H L • ฑ :
95 90 80

5
I I !
a 1 :I4 -fi 1
70 60 50 40 30
T i . Ti ii FIT
- - 1
: -
4-
20
n-rr r
:: t
En1Tn4ntL t t4
T T n ฑ* 4 : : 4 :
—,-— -{ I T J
•iiit JJ IL I 4 S
i 1 iui iig
ILฑIiฑ:
! :
:Tc.
y r1 .
-4
t 1
r
I I .
I
- - -4 -4 --
ath ฑft
c iz n : : ‘
U S T I I
r
tiMft:E R1
j
in a
0_s 0.2 0.1 0.05 0.01
;. - = - 4 10
9
e = _
I 1
L 1
-I-F-:{ F.P Tf- ff
i i i - L i i
r
‘ I .
! L
u:.
- h-+-4 ‘-I-U-—f
t :
n a
t—t i:1
rt
:fl
-H ฑ4—
=1 t
4---
‘ri
F---
-L-L-i- L44-
:Iti 1 - - -
t
I

- -T : :th:
I + - - -‘
- - . -t
, -t : :ftt LTZI
I
I
f EE E E
E E E E E
E E
I
LF
I
nt- ฑ :r ,
9 —
8_ -
7---
5_...
4---
3—-
I
-U
2
- - 1 -j i t s
4:4
4-
1:
ll1 W I EEEEE E E

HEEE HE
3
tEn:
- 1 --t-- —i----
— 4---
-- 4
t0
a
b—
a
-U
i i
* I — I —
— I
I I f
I F F
i F
—k- T-
Th iiT
-t—-
- -
1__
- F
—--— 4- F-
F Ft
i 1 __ i I ; -
F F
I F F
- (p i Jj1T L1
F I
I I
I I I
! I
I

-------
E E ฑfFFE
E
EEE
:1i ____
L ฑJI I
— - — -j
t1 E
! t 1ฑf —--f
—r--—-r--t- ti-i- — —1
ถ
E LEEF4
I I I I I ITI
T 4 1
t
1 IilI L
fi-t U1 4
K ฐE PROSAS LITY 46 8043
K 2 LOG CYCLES . N U 1.1.
ICtUPPYL & tiStR CC )
99_99 99_9 99.8 99 98
‘E i 1 i: L
7 4T 1 N
6.
S
Ji L l
95 90 80 70 60 50 40 30
T [ 11TT{T L •j: 11
1 : j L fl
&
20 10
I U 1 1 1 L
f t h 1 m i t f t 1 4 ? . i i . 1 tif
I ll u11 tWT = -
rn ’
Lr S S ! T
T IE . ! , 1’ R 1-ฑ4 nL:t. .
03 0.2 0.1 0.05 0.01
T: E T ThE E 9
- 1=- ;;t= n — - - - - 8
-SF— tฑi •H—tฑ t rr S
2 .
I . -
-
a).
I i
TTE. t I
I:
I .t_ia 4 i-IEII
S
lft t fth :ft
: 4 .
I I I I I I I H
9 —
A
:E?t:i - 3:44t tt
7-.—.
6-..
5....
- -S .
: 11 - - -
-
-e- --- 4
—
2
t0
0 1
1 I
I
HH
T 1 I
EE E
T 1E
I - . E - iU

‘ ;
: : : t ’i Tt T.
H -
—
LL :: =
ff 1
tt-ri
: -
I : : ‘ :
I I 1 I I I
9
a
1
6
S
3
H -V -- R ฑ- HbLT—c?
u
I - .
. 1 1
I I
I I I I III
liIIIIIIIIII IIIlI!IIihuIh I I I I Ihh I I Ih I I _
F
I F rr-st’ t trn-rrr I.Nt I1 I 11 ImTII I I IF i I I I I I I
t I I I I I i li lill I I I I I I I I I I
Ill 1 IIdtIiLLLLkU.LILH1-ttItIII11 I I I
I
f iLL LLLL
I I I I
ffH -H-F
t
H
99.8 99 -9 99 9 9
. - - -. - , ‘ tn en n fin 41 ) q 98 99

-------
Lu x: PRJBAHII ii V 46 P O43
rv- t_,, x 2 IA) ( Y( I I ;
t
I
99 9 99.8 9g
1o______
7
6
.

5-—-
9I Hfl if.) r;o f,I 40 O
-— .
. : F1 URE
:i L f

&
1-
i( 1 T M.A Z’ n T rf T1’rrc r P
t-
. fl 0) 2
I • !
I . [ ‘
t I )
(W l
$L c L p
0.5 0.2 0.1 0.05 0.01
g
‘:
fff9
q I:
1 H :
,1
2
t- +-$•
t - -t
I
I
I(I - - - 6
5
I
t
:4 ::
r --
4
3
a
R
7
6
5-
I —
— t T -r-_-4--f -nT !
H-1- 1 - - t
3 - I - -L-- --- - i’ —+f -- :
. .. -- :;L:t: .. . ;1 •t - I ft ::t : H ; - t

S T jt Lt ftit + i I tH l : : t
- --- -; I Lr :. ::T :::E:
If i t_ r i t t: LEE
ZT ii f:: I • - :..:i iI
ii _ I /
I

- - —‘-.; : :t ; ;:; -E U — j :: : :I: ;i; ; ;;; ; _ :;:
-
v - I -.-- - E E E — —— = -‘--- .- E EE ET -
:

‘.O
Q
7
c
2
80 90 95 98 99
99.8 999 99.99
fin’ nn Al fl nc i 9
in 2fl fl 40 50 60 70

-------
LI PROBAB$LITV 46 8043
r * x 2 LOG CYCLCS , ia s. .
KEUFFEL & ESSER Co.
C’
-
: E I F:
- __ _j 4 _j_ _____ 4 __ —+
. .
11
T
I
I
-
-4-
ui.OJ O O5 O 1 0.2 0.5 1
. -
: ::: :::: :::-
i
- -ELk
--i - -- H-
.- — I
—— ,——
II-J-!
-F-H 1
. - 4
Th
—t
‘f
-—- -,

-
. * ,--
‘ -.
1
- -
I-
-ฑ
----L
--4—-
:
-f- -
-
-
— I
.4- . - -
:-t : 4 i:t ;t : TT 1i i
;: :-T- :T3I: i:i- j. T L i: -
3--
. -t-. :- ---
i Ji
I:1 ; ..:_::: -L:T_—-
-
t
I I
. .
I

,-:-:- -
- - --: :- -::
—= ..
- - =
-= -. -- -
,w
HI
.,--
. I
-4— LH 4 -_. _
.;:: :::1T
4 — —
.. - t-—
MO ‘H) ) .8 99.8 99.9
9’) .9’)
Q
7
6
5
9—
3
6
4--
- T
--- 1 -_--

- H —V--
--- . — .II-
. .f-, H------ -- --
.4 . -4
-r 4
ฑ
—I:
J
2_
g
.4
7
!: TTtj H
-
A
I II
I
Ia ,u sn 40 u bU /1)

-------
F< PROBABILITY 46 8043
x 2 LOG CYCLES o . ii . .
K UFFEL 8. S ER Cr ).
U)
-i- ; 1 T
70 60 50 40 30 20 10 5 2 1 0.5 0.2 0.1 0.05 0.01
— - — — — — . . . —. -- - .4--.—4 10
ฑฑฑ:i_4..TI::Z: = =3=:L := .. i::-: i4
•-1 - r *
I- 4 - i
; .
.—---- .-—— ,-. -t i—r
TTT: :: i T =: T EI : t: I =-::: -:::::: ฑ= -t:- - - ft•
:-.-: .:: Li::E= ==- =:z-_— ::•-; - — - - ---= :_ . _ :;:_ .: _-= ::;: = ; :::: = t: =_ :E:
—.- ___ _tฑ l--
99.9 99.8
7.
99 98 9590 ____
in
-
,

-- -i---
a_
-— 4 -: -fTh ฑ ฑ
:!E I
1 1!!: :
UL
1 T T
--.--- ---- -- .
- .‘ ---. --
-:::tr= : j
: = __
- . ..
E: T::
--o----———-——.--—— — —•— ——4-. - --—- - --———- —..—-—-—-.
- . -.--— . —.--- •
1 JEt
=
7
4-1 - -
2.
I II i;
L
i U 4 4-
:
_!
II :4n:1 I
j Jj L
I- —I.—
-F
—-1—
1IIi LI
4
E
fH
g
‘I
.i. T
R
I
I I
I I
-- i
II . 2
—
—;::- 4-H-’ - : -,- -r- ‘ —‘H-
I I E’
4-
I
j . [
. :
‘-. 4 .
-
:j : _ 1ii!1 .
:
I1 LL JJi]I1i II

H-
co
i I - ฑ -I4H - L:
HE t 4 . L -IrJiL fl H-
--.
5
.
._}: i
_1-
E =- =- 4 -: -::- T - - - - : — -r-



. — - —. -- —,—.-- - -
- - f -F
T-____-t


t—
.. : :1:
::::

— — — I -
-

— .—. .4 1
I. ฑ
ฑ+
- -----
—

- -4-’ —
1 H
- -
;‘ :
JLi
I
h r
-f -—- -
—-——

—- -I- -—-— - —
I
.: • ‘ :
4-4 - .4-
.. IT
4 t j
+ . t
-- - - - -- -
---— - r——

— -+- -. I -
I 1I

.1 ,I ; P :1
..LLi L 4
j . I
t rtr - — — — — r- - t I
-----— --=-- —i-— j— --- —---—n - - -—
————
I
—— — — — I
I It r
- -—-- —
:, ‘ . ::ti . :
H 4i L LLLI
5
4
G
2__
—
H IH
II

:
—.--—,---+.
..
f
J _
5 0.1 0.’ 0.5 1
I fH
L ceTtL i:rt
2 S tO ‘(I It) 4t 1 ‘ () III
II
S
I -
/ it l() • i I ’ 9
9 ) j),)

-------
-99—
Analysis of the salt trapping medium was made at the
completion of Tests IX and X, and again after Test XI. In
the latter case analyses were performed on salt from both the
exhaust inlet and outlet halves of the muffler. Data are
shown in Table XXVIII.
TABLE XXVIII
ANALYSIS OF SALT FROM LEAD TRAP AFTER
DYNAMOMETER TESTING
Weight %
Potassium Lithium Lead
After Tests IX and X 24.0 3.12 1.80
After Test XI (Inlet half) 22.6 3.10 0.45
After Test XI (Outlet half) 24.6 2.77 0.71
At the completion of Tests IX and X, the salt trap was
examined visually to determine if substantial quantities
of the molten salt trapping medium were being lost. At the
completion of Test XI, the trap was weighed to determine
salt loss. The data are shown in Table XXIX. No attempt
was made to add salt to the trap between runs.
TABLE XXIX
WEIGHT OF SALT IN LEAD TRAP MUFFLER
Initial - 6089 gm
After Tests IX and X - Little visual change
After Test XI - 1476 gm
The bulk of the salt lost from the trap in Tests IX-XI
accumulated within the tail pipe section of the exhaust
system. The pressure drop across the lead trap was found
to be ‘ l cm of mercury in Tests XI and X, and between 0.5
and 5.0 cm of mercury under the cycling conditions of
Test XI.

-------
-10ฐ-
c. Aldehyde Analysis
Data from the polarographic determination of aldehydes for
the cold trap condensates collected both ahead of and behind
the lead trap muffler in Tests IX-XI is presented in Table XXX.
For comparison purposes, similar analyses are reported for
runs VI-VIlI in the presence of the conventional muffler. In
these runs only a single analysis was obtained ahead of the
muffler.
These analyses cannot be related directly to the concentration
of aldehyde in the exhaust stream since, as has been discussed
earlier, the sample probes feeding the cold traps were not
used exclusively for this purpose but were also used for
obtaining exhaust gas analyses. Nevertheless, the relative
orders of magnitude of the numbers are significant.
TABLE XXX
ALDEHYDE ANALYSIS OF COLD TRAP CONDENSATES
Test Sampling Location* Engine RPM Aldehyde (ppm HCHO )
VIII 1 1200 950
VI 1 2250 1100
VII 1 Cycle 2000 (30 benzylphenone
IX 1 1200 910
2 660
X 1 2250 420
2 1.3 (2.9 CH 3 CHO)
XI 1 Cycle 1500
*1 - ahead of muffler
2 - behind muffler
0. TRAPPING MEDIUM SELECTION
Considerable technology has been developed for the removal
of particles from gas strea’ms; however, few systems afford
the diversity of conditions that are encountered in the exhaust
system of an internal combustion engine. Temperatures may range

-------
-1 01-
from over 1500ฐF in the combustion zone to those of ambient air
(e.g., -30ฐF or lower). The volume of gas within the system
varies from very low values at engine idle to several hundred
SCFM at accelerated vehicle speeds. The exhaust gas stream
velocity has been calculated to be of the order of 15,000 fpm
at about 60 mph. Particle size distribution is known to vary
over wide limits, from the visible corrosion products of the
exhaust system to the products of the combustion process that
could approach molecular dimensions’. The smallest particles
(e.g. <5 microns) represent the greatest inhalation hazard
(hence potentially the most dangerous to health) and are the
most difficult to remove. In designing a device to remove
particulate matter from automotive exhaust, there are additional
constraints due to the space and power available in the vehicle.
In considering the design of a device to remove particulate
matter from automobile exhaust, one would expect that a
gravitational separator would be impractical because of the space
it would occupy and its increasing inefficiency in removing the
smaller particles. Inertial devices would amplify the inherent
separating tendencies of the particulates by increasing gas
velocities and provide rapid changes in direction. However,
these devices can only be designed for narrow ranges of gas
flow rates and may introduce considerable pressure drops in
the system.
Electrostatic precipitators are very effective for removing
particles from gas streams but must be designed for the maximum
use condition; hence, in the automobile they would have very
large power requirements.
The approach chosen for design of a trapping device was to contact
the particles with a suitable liquid medium so that they would
be removed by chemical reaction or inertial separation. The
liquid medium would facilitate inertial separation by increasing
the mass of the particles by wetting.

-------
-1 02-
The primary criterion in the selection of the trapping medium
was its fluid vs. temperature range. A literature survey was
made to determine compositions of matter which had melting
points in a range that appeared comparable to that expected
in the vehicle exhaust stream. It was anticipated that
a liquid would be more efficient than a solid as a trapping
medium; therefore, it would be desirable to have the melting
point as low as practical. An arbitrary maximum melting point
of 1000ฐF was selected as the cut-off point for the purposes
of the literature search.
Once a tabulation had been made (Appendix B) based on melting
point, other characteristics were considered. Obviously the
vapor pressure would have to be negligible at the temperature
of operation to prevent excessive loss of the trapping medium
during use. In addition, a practical medium would have to
be sufficiently non-corrosive so it could be contained for the
useful life of the device, e.g., halides are known to be very
corrosive.’ The medium must not present a toxic hazard to the
atmosphere, e.g., cyanides or chromates would be undesirable.
The possibility of detonation or explosion was also considered,
e.g., ammonium nitrite would be undesirable from this standpoint.
Once a selection had been made, based on these criteria, other
characteristics of the medium could be considered. For example,
an economical, readily available medium would be preferable.
Chemical reaction with the exhaust emissions would be desirable
if it improved the efficiency of the device. However, reaction
might lead to decreased efficiency such as the formation of
carbonates in an alkaline trapping medium from reaction with
the carbon dioxide in the exhaust stream. Carbonates have
greatly increased melting points which are less desirable.

-------
-1 03-
The trapping medium of choice was a 70 percent potassium
nitrate—30 percent lithium nitrate mixture. It was selected
on the basis of:
1) A relatively low melting point (284ฐF).
2) Low corrosivity and high thermal stability — Nitrate
salts are used in the metal treating industry at
temperatures over 900ฐF and are contained in
mild steel.
3) Potential for acting as an oxidizing agent - In
addition to the mechanical entrapment of particulates,
oxidation of the organic fraction of the exhaust stream
can be postulated.
A potential disadvantage was the resultant emission of reduced
nitrogen oxide.
E. DESIGN OF TRAP
The effective trapping of exhaust particles depends on wetting
them so that they become aerodynamically unstable. This is
a function of proper trap design. The particles must first
be contacted with the trapping medium and the wetted particles
then removed from the exhaust stream. To better study this
process and the effect of trapping medium viscosity on trap
efficiency, two prototype mufflers were constructed of transparent
methyl methacrylate. An exhaust stream was simulated with
compressed air and exhaust particles by expanded polymeric
ml crospheres.
Initially particles were introduced with a syringe but this
caused them to pass through the muffler in “spurts” which
could not be witnessed or photographed. Subsequently an
aspirator was used which permitted control of flow through the
muffler so that photographs could be taken. Due to the
contractor requesting a change in the emphasis of the contract,
the effect of trapping medium viscosity was not investigated.
The muffler design selected was based on a device previously
used in a proprietary research program. The device is shown
schematically in Figure 51.

-------
FICURE 51
PROTOTYPE LEAD- TRAP
2” OD 3”
12”
3U
,
exhaus -L
outlet [ 1 [ G Sk t
II
Side View
Side
removeab le
3,’
1f
‘F
14”
I-
exhaust
inlet
I ’
A
1\
ฑ
F
II
I —
6”
-4
End View
10”

-------
-105-
V. DISCUSSION OF RESULTS
The data from the road tests show that the temperature profile
of the exhaust system forms a family of essentially parallel
curves at different speeds from 30 to 80 mph. If allowance
is made for the increased path of travel of the gas through
the muffler, the temperature drop of the gas is essentially
linear throughout the exhaust system. The temperature drop tends
to be greater from the exhaust manifold to the muffler than in
the remainder of the system and proportionately this initial
drop is greater at the lower speeds. Temperatures ranged from
900-1400ฐF at the exhaust manifold to 300-800ฐF at the end of
the tail pipe.
The temperature profiles generated in laboratory Tests I and II
again show that the temperature drop along the exhaust system
is linearly related to distance from the exhaust manifold.
However, the temperature drop along the system is somewhat
less than was experienced in the road tests. This is undoubtedly
due to the fact that in the laboratory tests the exhaust
system is not exposed to the same air cooling which is
encountered in the moving vehicle. Nevertheless, these
temperature profiles were considered close enough to those
experienced in the road tests to allow meaningful data on
particulate matter to be developed in the dynamometer tests.
The mass data generated on particulate matter collected at the
heated Andersen Sampler in Tests Il-V allows the determination of
a particle mass profile for the exhaust system at 2250 engine
rpm.
In Test I the Gelman filters used at sampling Positions I-IV
were not maintained under isothermal conditions or at
temperatures equivalent to those of the exhaust system at
their respective locations, thus there exists the possibility

-------
-l 06-
of condensation of organic matter which may in fact be volatile
at the various sampling point temperatures of the exhaust system.
Particulate matter collected on these filters and their back-up
cold traps was nevertheless of real value in defining the nature
of the exhaust particles present along the length of the exhaust
system as will be shown later in the discussion.
In Tests II-V, an effort was made to define a mass distribution
profile for the exhaust system under 60 mph steady state engine
operation by locating the heated Andersen stack sampler at
successively different sampling positions along this system.
This approach was successful at sampling Positions III and IV
(immediately after the muffler and close to the end of the
tail pipe) but failed at sampling Position II (ahead of the
muffler) due to degradation of the Andersen plates. It is
nevertheless significant that at this sampling position,
virtually no particulate matter was detected in the Andersen
unit, suggesting that at this point in the system the exhaust
effluent is almost exclusively gaseous in nature.
It is again significant that the weights of particulate matter
collected at the various unheated Gelman filters used in
Tests I l-V does not necessarily relate to the actual concentration
of particulate matter present within the exhaust system since
thermal quenching at these filters undoubtedly occurred.
Comparison of the particle mass-size distribution profiles
for Positions III and IV of the exhaust system shows that there
is a shift in the size of the particulate matter to a smaller
equivalent mass median diameter as the exhaust effluent moves
towards the end of the tail pipe. This indicates that particle
shattering, rather than agglomeration, is occurring in this
section of the exhaust system. Such a phenomenon must result
from the highly turbulent motion of the exhaust effluent within
the tail pipe.

-------
-l 07-
Analysis of particulate matter collected on the plates of the
Andersen sampler in sampling Positions III and IV of the exhaust
system confirmed that lead halides were the major components
of the larger particles collected. As particle size decreased,
however, the concentration of halide diminished and the levels
of lead sulfate and lead chlorophosphate increased. Iron, zinc,
and nickel were also detected. Salts of the first two of these
elements probably result from degradation of the exhaust system
itself. Nickel compounds could be a result of reaction of the
exhaust effluent with the nickel-plated stages of the Andersen
unit. In general, all the particulate matter analyzed by scanning
electron microscopy showed non-homogeneity, especially toward
the tail pipe end of the exhaust system.
Transmission electron photomicrographs of the smaller particles
collected (
-------
-108-
at the fiber g1ass filter with resultant dendritic crystal growth.
This type of crystal growth and particle birefringence decreases
considerably as one moves towards the tail pipe. This is
interpreted to mean that the exhaust effluent is essentially
gaseous in nature between the exhaust manifold and the muffler.
The bulk of particle precipitation and agglomeration occurs within
the muffler to afford particulate matter having random orientation
and thus low birefringence.
The principal component of the particulate matter collected
during direct sampling of the exhaust system was found to be
lead halide, specifically lead chiorobromide. Examination of
the vapor pressure curves for lead chloride and lead bromide
shown in Figure 52 suggests that these salts would not be expected
to precipitate from the exhaust stream prior to the muffler because
of the high temperatures which exist in this section of the
exhaust system. Also, since lead bromide has a higher vapor
pressure than lead chloride at any given temperature, one would
expect lower concentrations of the latter salt nearer the manifold
end of the exhaust. This was found to be the case. It should
also be noted, however, that the concentration of ethylene
dichioride scavenger in the fuel is twice that of ethylene dibromide.
This fact would account for the relatively higher concentrations
of chloride at the end of the tail pipe.
Lead determinations on the cold trap condensates sampled from
the exhaust system substantiated the hypothesis that the exhaust
stream is essentially gaseous in nature prior to the muffler.
The concentration of lead in these condensates dropped substantially
from exhaust manifold to tail pipe.
The identification of uncombined lead during Test I is interesting.
Its presence, especially at engine idling conditions (870 rpm),
must result from incomplete fuel combustion.

-------
0
U
$-i
0. .
— ----r
1400
1300
1
800
700
‘)APOR
H
[ J Pb
600
Melting
Point
10
20 30
Vapor Pres้ure (mm Hg)
-109-
—
FIGURE
52
vs TE PERKI URE
4nd L AD B 0MIDq
PRES URE
LEAD LCHLOI}IDE
/
P112
Pb r 2
1200-
1100

-L
I - L .
1000
I —
900
— I
-
40

-------
-110-
The mass distribution profiles plotted for the total air diluted
particulate matter collected in Tests VI, VII, and VIII (conventional
muffler) were very similar and in close agreement with that for
particles collected at the end of the tail pipe under 2250 engine
rpm. Comparison of the masses of air diluted particulate matter
collected in Tests VIII and VI (1200 and 2250 engine rpm) shows
them to be 0.096 gm/mile and 0.104 gm/mile, respectively, very
similar to that obtained at 2250 rpm from the Andersen unit positioned
close to the end of the tail pipe (0.072 gm/mile) (Test III).
From this data one would conclude that for the leaded fuel tested
and under steady state conditions employed, the bulk of particle
precipitation occurs within the exhaust system and that any further
precipitation which does take place in the dilution tube results in
very little change in particle mass median equivalent diameter.
The data also substantiates the fact that the bulk of particulate
matter present in both the exhaust system and the air diluted
exhaust effluent is associated with lead salts.
The difference in the gram/mile of particulate matter collected
in Tests VI-IX on the 4 cfm filter compared to that collected
on the Andersen sampler and back-up 1 cfm filter is significant
and merits discussion. In all cases except Test X, the data
obtained at the 1 cfm filtering system is higher than that obtained
at the 4 cfm filter (Table XXVI). This is believed to result from
the longer residence time of particulate matter in the 1 cfm sampling
system caused by the lower pumping rate. Also the adiabactic
expansion of the diluted exhaust effluent as it passes through the
Andersen unit could result in thermal quenching and the condensation
of volatiles within the unit.
Before any further work is undertaken in this area it is imperative
that the above sampling anomalies be resolved.

-------
—111—
At 2250 engine rpm and under mild cycling conditions, the mass-
distribution profiles for total air diluted particulate matter
and lead associated particles are almost identical. At 1200
engine rpm, however, the lead associated particles are smaller
than the total particulate matter collected.
Introduction of the prototype lead trap muffler into the exhaust
system showed little effect on either the total mass of air
diluted particulate matter collected or on its mass median
equivalent diameter at 1200 rpm. The amount of lead detected
was unchanged. There is, however, a decrease in the mass median
equivalent diameter for lead particulates.
The lead trap muffler was most effective at 2250 engine rpm when
the total mass of lead in the particles collected dropped by over
50 percent. The mass of total filtered particulate matter, however,
increased slightly, undoubtedly a result of nitrate salts being
carried over from the trap itself. In this instance the mass
median equivalent diameter for both total and lead associated
particles decreased. Under these conditions, the trap appears
to be most efficient for the removal of the large lead particles.
Particle data generated from the lead trap test under cycling
conditions is considered to be unrealistic because of the
massive amount of salt lost from the trapping device.
A Government-initiated change in the scope of the work precluded
further optimization of the lead trap unit. However, based on
the performance of this device at 2250 engine rpm, it is felt
that such a unit would form a good base for further activity
in this area.

-------
-112-
PHASE II

-------
-113-
VI. CONCLUSIONS
Phase II of this study covers an investigation of the particulate
emissions from an automotive engine operating on non—leaded fuel.
The effect on such emissions of changes in air-to-fuel ratio,
and of two emission control devices were evaluated.
The tests on the effect of the air/fuel ratio lead to the
following conclusions:
1. Over the range tested (13.3—17.1), the air/fuel ratio had
little effect on the mass of particles emitted per mile.
At 1200 rpm the ratio of lean/rich emissions was 1.3 by
the 4 cfm filter, 1.1 by the 1 cfm filter. At 2250 rpm
the ratio was 1.0 by the 4 cfm and 1.1 by the 1 cfm. The
cumulative mass distribution profiles are very similar at
a given speed.
2. The mass of particles/mile were smaller at 2250 rpm
than at 1200 rpm; the ratio was about 0.5 with the
4 cfm filter, 0.7 with the 1 cfm filter.
3. The 1 cfm filter collected a proportionately greater
mass of material than the 4 cfm filter due to the
greater residence time in the filter system and lower
temperature at the filter interface.
The test of the noble metal catalytic device led to
the following conclusions.
a. The emissions of an engine fitted with the catalytic
device as compared to the same engine without the device
1) had lower residual hydrocarbons (about one—half
at 2250 rpm, one-quarter at 1200 rpm).
2) had lower aldehydes (about one-third at 2250 rpm).

-------
-114-
3) had increased mass of particles per mile, using
the 4 cfm filter 6.5x at 1200 rpm, 22x at 2250 rpm;
using the 1 cfm filter l.3x at 1200 rpm, 4x at
2250 rpm.
4) had decreased particle size, from 20 percent less
than iji to 85 ‘percent at 1200 rpm, from about 50
percent less than lji to 90 percent at 2250 rpm.
b. Considering the Pontiac engine outfitted with a noble
metal catalyst device, the mass of particle per mile
increased 60 percent (with 4 cfm filter) and 140 percent
(with 1 cfm filter) when the speed was increased from
1200 to 2250 rpm.
The test of the packed bed catalytic muffler led to the
following conclusions:
a. The emissions of an engine fitted with the catalytic
device as compared to the same engine without the device
1) had lower residual hydrocarbons (about 45 percent
at 1200 rpm, 75 percent at 2250 rpm).
2) had about one-half reduction in aldehydes.
3) had about 156 percent increase in the mass of
emissions (at 2250 rpm — 1 cfni filter).
4) had decreased particle size (from 68 percent less
than 1 micron to 82 percent less than 1 micron).
b. Considering the Chevrolet engine outfitted with the
packed bed catalyst device, the mass of particles per
mile increased 461 percent (with 4 cfm filter) or
535 percent (with 1 cfm filter) when the speed was
increased from 1200 to 2250 rpm.

-------
—115-
VII. EXPERIMENTAL PROCEDURE
Studies under Phase II of the program were aimed at establishing
baseline exhaust emission data for an engine operating on Amoco 91
octane non-leaded gasoline and subsequently evaluating the effect
on such emissions of two different catalytic reactors. A fuel
analysis is shown in Table XXXI. No analyses of the particulate
matter collected were made.
In order to perform these tasks, a 1971 Pontiac V8, 400 CID engine
was set up on the Dynamatic dynamonieter test stand. The exhaust
stream from the left bank of cylinders after passage through a
conventional exhaust system was fed to the air dilution tube
described under Phase I of this report.
The series of tests performed are outlined in Table XXXII. In
Tests 2 and 3, the engine was operated at 1200 rpm at two different
air/fuel ratios - lean and rich. Tests 1 and 4 were conducted at
2250 engine rpm with the same two air/fuel ratios.
Particulate matter and condensate was collected through a Gelman
glass fiber filter with a backup cold trap via a sample probe
located immediately ahead of the muffler. Additional particulate
samples were collected from the various sections of the dilution
tube by sweeping at the end of the test, and at an Andersen Model 0203
impactor located at the end of the tube, sampled isokinetically at
1 cfm and fitted with a backup Millipore filter (142 mm Type
AAWP O. 8 u). Other samples of particulate matter were collected
isokinetically (4 cfm) through two separate Millipore filter holders
fitted with fiberglass filter pads via probes located at the end
of the dilution tube. This sampling system has been previously
described in detail by Moran and Manary 1 . In Runs 5-10, an
additional Gelman filter and back-up cold trap were introduced
immediately after the catalytic units tested. Because of the
limited amount of time available to complete Phase II of the research

-------
-116-
TABLE
ANALYSIS OF FUEL
XXXI
FOR PHASE II
Physical Analysis
ASTM D
Boiling Range:
% Recovery
% Residue
% Loss
IBP
5
10
20
30
40
50
60
70
80
90
95
EP
Research octane no.
Motor octane no.
Reid Vapor Pressure, psi
API Gravity
FIA: % Saturates
% Olefins
% Aromatics
American Unleaded
93
119
129
150
169
188
204
219
234
258
294
329
372
98.5
0.2
1.3
91.0
82.7
8.8
61.9
73.0
5.4
21 .6
Fe
Ni
Cu
Al
Ca
Si
Mg
Mn
0. 00005
<.00005
<.00003
00005
<.0001
<.0001
<.00005
<.00003
Pb
Cr
Sn
Zn
Ti
Sb
p
<0.0001
<.00005
<.00005
<.0001
<.00005
<.00005
<.0005
Trace Metal Analysis

-------
TABL [ XXXII
PONTIAC ENGINE SAMPLING DEVICES FOR PHASE II
91 RON - 0 TEL FUEL
EXHAUST SYSTEM SAMPLES
DILUTION TUBE SAMPLES
Ahead of Muffler
After
Catalytic
Muffler
Glass
Cold
Trap
Glass
Fiber
Filter
Cold
Trap
Slit
Tube
Sweepings
1
cfm
Fiber
Filter
Ice/Water
Dry Ice/
Acetone
Ice/Water
Dry Ice/
Acetone
Andersen
Sampler
Millipore
Filter
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Test
2250 rpm
A/F 15.5-16.0
1200 rpm Yes
A/F 15.5-16.0
1200 rpm Yes
A/F 13.5-14.0
2250 rpm ‘es
A/F 13.5-14.0
Noble Metal
Catalytic Muffler
2250 rpm Yes
1200 rpm Yes
Dow cycle Yes
Proprietary
Catalytic Muffler
Dow cycle Yes
1200 rpm Yes
2250 rpm Yes
4 cfm
Glass
Fiber
Yes
Glass
F I ber
Yes
Yes
Ye
Combined
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
-I
—.1

-------
-118-
program, no analytical work was conducted on the particulate samples
obtained at the Gelman filters located ahead of and after the
muffler. Since such samples were not collected isothermally,
the determination of gram/mile particulate matter from these filters
was not considered to be of value.
At the start of the test series, the Pontiac engine, which had
been operated only on non-leaded gasoline, was checked and conditioned
for 48 hours on the Dow cycle. At the beginning of each of the
tests in this phase of the program, the engine was held under the
desired test condition for 1 hour prior to introduction of the
exhaust effluent to the dilution tube. Sampling was then conducted
over a six hour period.
An unsuccessful attempt was made to get baseline data for the engine
operating in the Dow cycling mode with rich and lean air/fuel ratios.
However, it proved impossible to maintain this variable within
the prescribed limits over the various modes of the cycle. This
was not unexpected due to the idle and low rpm conditions involved.
Air/fuel ratios were calculated from analysis of exhaust gases
as described in Phase I of this program. No data was collected
on this run.
Tests 5-7 were conducted to evaluate the effect of a noble metal
catalyst on particulate emissions, residual hydrocarbons, and
aldehyde concentrations, under three different engine operating
conditions: 2250 rpm cruise, 1200 rpm cruise, and the Dow cycle.
The catalyst unit was introduced into the exhaust system in place
of the conventional muffler (and located 14 inches after the Y
junction normally present below the mainfold). During the above
test sequences, it operated at from 680 to 870ฐF. The proper
operation of this device required the introduction of air into
the exhaust system ahead of the unit. This was accomplished by
means of pumping air into the exhaust stream in such a way that the

-------
-119—
exhaust effluent entering the catalyst unit contained 1.5 percent
of the oxygen by analysis. Thus, calculation of engine air/fuel
ratio by exhaust gas analysis was not possible. Data obtained
in Runs 1-4, however, indicated that variations in this parameter
had little effect on the mass-size profile of emitted particles.
Air/fuel ratios were calculated from air and fuel flow measurements.
A similar series of tests (8—10, Table XXXII) were conducted in
the presence of a third party proprietary catalytic device
designed to reduce emissions of residual hydrocarbons and carbon
monoxide. Tests 8-10 were again run on a 350 CID Chevrolet V8
engine. The change in engine was necessitated by the fact that
the required air injection system for the second catalytic device
studied could not be readily adapted to the Pontiac exhaust manifold.
The 350 CID Chevrolet engine was cleaned, checked, and conditioned
for 24 hours on the Dow cycle using Amoco 91 gasoline (non-leaded)
prior to starting these tests. This catalytic unit was physically
located 14 inches from the normal V junction below the exhaust
manifold to insure adequate heating. Again, to insure effective
operation of the device, air was pumped into the exhaust system
ahead of the catalyst so as to maintain a free oxygen content of
2.1 percent by analysis.
The exhaust stream was sampled before and after the catalytic
device at 1 cfm. The exhaust gas sampled at each position was passed
through a Gelman filter fitted with a glass fiber filter pad, an
ice/water cold trap, and a dry-ice/acetone cold trap. The ice/water
cold trap condensate was analyzed for carbonyl content. The dilution
tube was sampled as before. The grams per mile and the mass
distribution profile for the particular emissions were calculated
from the dilution chamber data.

-------
-120-
VIII. EXPERIMENTAL RESULTS
Air/fuel ratios calculated from exhaust gas analyses and air and
fuel flow rates are shown in Table XXXIII.
TABLE XXXIII
AIR/FUEL RATIOS FOR PHASE II
Test 1 2250 rpm 15.2 - 15.31
Test 2 1200 rpm 15.0 - 15.21
Test 3 1200 rpm 13.8 - 14.61
Test 4 2250 rpm 13.3 - 15.01
Test 5 2250 rpm 16.4 - 17.12
Test 6 1200 rpm 15.7 - 15.82
Test 7 1070 rpm ]4•72
1615 rpm 16.02
2125 rpm 16.92
Test 8 1070 rpm 14.32
1615 rpm 15.02
2125 rpm 14.02
Test 9 1200 rpm 14.22
Test 10 2250 rpm 13.82
1 Determined by exhaust analysis
2 Determined by air/fuel measurement
Particle mass data for Tests 1-4 is shown in Table XXXIV. One
one-half of the total exhaust was fed into the dilution tube
(left bank of cylinders) and only a limited volume of the diluted
exhaust was passed through the filters (1 cfm or 4 cfm). This
data has, therefore, been corrected to the equivalent mass for the
total engine exhaust flow.

-------
-121-
________________________ 4 cfm
Glass Glass
____ _________ ________ ___________ Fiber Fiber
2.08 g 1.55 g
2.3 2.32
1.95 1.68
1.98 1.9
The resultant calculated values for the particulate emissions in
grams per mile are listed in Table XXXV.
TABLE XXXV
EFFECT OF AIR/FUEL RATIO ON PARTICULATE EMISSIONS
Filter 1 cfm Andersen Filter
Temp. and Filter Temp.
____________ ( ฐC) ( gm/mile) ( ฐC )
23 0.0406 23
45 .0308 30
.00972
00538
*Average of two filters
2250 rpm
Lean A/F
1200 rpm
Lean A/F
1200 rpm
Rich A/F
22.50 rpm
Rich A/F
TABLE XXXIV
PARTICULATE SAMPLES FOR PHASE II
TESTS 1, 2, 3, 4,
(Dilution Tube Samples)
1 cfm
Tube Andersen Mflhipore
Slit Sweepings Sampler Filter
O.0102g 0.1748 g 5.50 g 5.6 g
Combined 0.0204 7.3 0.2
0.0020 0.0236 6.2 0.30
0.0144 0.1824 6.7 3.0
Test Period - 6 Hours
Lean Mixtures:
1200 rpm
2250
4 cfm Filter*
(gm/mile)
0.0128
.00555
Rich Mixtures:
1200 rpm
2250
23
.0361
23
41
.0269
29

-------
-122-
The differences in data obtained at the 1 cfm and 4 cfni filter
systems are discussed on page 134.
The mass distribution profiles for these particulate emissions
are illustrated in Figure 53 for the lean and rich air/fuel
mixture at 1200 rpm, and in Figure 54 for the two air/fuel ratios
at 2250 rpm.
series of tests employed the same Pontiac 400 CID engine
standard muffler replaced by a noble metal catalytic
The data, again corrected for the total exhaust gas flow,
in Table XXXVI, with the calculated grams/mile particulate
emissions in Table XXXVII.
TABLE XXXVI
PARTICULATE SAMPLES FOR PHASE II
TESTS 5, 6, 7
NOBLE METAL CATALYST
(Dilution Tube Samples)
1 cfm
Test Period - 6 Hours
Eli ter
4 cfm Filter* Temp.
( gm/mile) ( ฐC )
0.0737 26
.120 29
.0978 24-44
1 cfm Andersen
and Filter
(gm/mile)
0.0489
.117
.0918
Filter
Temp.
(ฐC)
23
42
24-26
The next
with the
device.
is shown
2250
rpm
Slit
0.0118g
Tube
Sweepings
g
4 cfm
Andersen
Sampler
4.6 g
Millipore
Filter
Glass
Fiber
46 g
Glass
Fiber
48 g
0.1118
37.4
g
1200
rpm
.0016
.0254
1.9
6.9
12.46
14.08
Dow
cycle
.0172
.1736
3.8
18.5
23.5
24.0
TABLE XXXVII
EFFECT OF NOBLE METAL CATALYTIC DEVICE ON PARTICULATE
1200 rpm
2250
Dow cycle
EMISSIONS
*Average of two filters

-------
PROBABILITY 46 8043
-%‘ ; x 2 LOG CYCLES - ‘.
(I%IFFEL a •
99.9 99.8
— ::
99 ถ 9
1o______ .
9
8
7
6
5-——
:
:
95 90 Hf) 70 ( O 40
.F] UR :-$3
.
. ‘
j___ 4 __ 4 _ T —; --— -- -- -- - - - - . - ------ — --;- —. 4-- - - • L . ,
L n? riM -r1 fl 12( P
30 ?0 10 5 2
I I
., : _ ::.
I . _- ; . __ . . : . . ::
. . II,
I ,
i :t I i I iI1LL
I
. --W iฑ:
-:
; r
0.5 0.2 0.1 0.05
-L .. : r -:- - :.: t: : : : U
t4 r1 - - - ____ Ji JI
1T i;t-:-
- :H tJ
0.01
tt i- -tt
ฑ , -- I:
: ;-_;:T . . ::_TJ: T . ;
10
0
:H. f.
-
--- - - --- - - -
: r
i- __ -
: : :_ - :-— -- - : :
H- - - 1- —4-i-- -i--+- -
E
_ .:f 1 .L _ 1 i
T1 f!1 THJ
trn 11 -
Ti: I_4 : .
:
I _li _i
IL th
n !U
I4J
iTh 1
1 -
-t
4
7
6
5
4
=-- :: -_ : . =- = =:: :
-
1-
-f
—
r
-ii

-- -
--
-
t- -
.
.. - -
— -
L1
‘ .
. -.
. —
— :—1- ----- —
-t . -t -1-: p:- - :t:t ฑ
l- - tฑ f : ฑ T
‘ t H - -
Hr r r ฑ
-- — —--!-- -
ti — - - - ii i-
ฑ.
-h-I--- -
-,-
———
——
—t—— — .


:


---- --
--- ---
9--
8
7-
-
— r__ I
j r— J
TIi_j_ - :

T: _

i E1-::

- WI
--- - --
4 i T T ;
TTT t:-
5--
:
:4 T_ i r
: Z
— :: -

:1
—- -
- I
I
it
-xi: Tฑ
f t 4ii:
4_.
- - --- -
.: t T i
liL
T
-- j- -ฑ
::: :
—---- I! 4J --4
- T _ ——
—H- 1 -- ::
-— - - , - j - I I’I I + 4
-
- 14T- --r-i -
-I
-H-
- 1
-
T
- -4--
- -
7
—
I I
-- -- - -
— ---
---h
I -
f
= = = = -4:::=.;:
1 : :
— -----_--
-;
---Lv -
I
— - --—---
- -, -
. j
—
- -
‘I.,

“ - i
I --- T-r -

. . H- -
rI
;L ii - - i
- -
-1 - - r
-H-
L
--I
- -I-
--

-------
IJ*’ PROBABII fly 46 8043
r i x 2 LO( . c;YCLI S . . *
1(1 III ri i I I
In
99.99 99.9 99.8 99 98
95 90 80 70 O 50 40 30
20 10
2
.
-::::!T:::= :Lt;:
L ’ฑ!-
: :
:
1 0.5 0.2 0.1 0.05 0.01
8
7
6
5
4.
.;:-_;็:L4 L [ i 7 i :i:


—
FCWE54
7
.
:
-

! t
. :
tI

-nilE f t
J UJJ
L ‘

1t -r H 1 1 T
yr(
; i II

‘ F


.
i : __
1 T i2!


10
2.
-r
.- -
2
. 4
-I
7
1-
.
:!
L•
L
•1—
0
A
T j ::
L

2
=
7
I
- —
.+. .
c
A
3-
.

-
____#___ _ _

_ i j

::: -::::
IE: :- E::
- -
- : —::_
---- - - -

-.-- T: :H ; :=: .
— --: -:=tT:L
_ ::; -+-: - ,

---
— .
--
t _ t t I T i: fl•r V T ) FrrTT T
L- 1 it i it 1
- I [ : L
I : H - r:; ff ‘i -4 1 t

, L i1 1 • .:‘
:H ii:::: L JI I 1.
- - ; f - .- ----——— IT -
‘ TT IL f I H
$
—t-:: , —j- —4- - • •4. - - - - : —: -
T :: L ..
- - T ! -r + J-; - - i- - -
— - - - - - - —
- — - - - “ ---
i — : • - - -- : +_
I - —r • -.- - ..
t - - —- - - --- - :i—-
— ;- - - - -: i +4:t — -
- _1 t- -• : - - --.-t —t--
- -
- .. - -, - - - - -- -
— - — - - - t + ‘
-- .. . , —
9
8

6
5
3
LE ;L
—-—
—-.
:
:
—

. -
,


;

------—-

= —
——

.
i—


--—

—
VVVV
V
— I
- H
VVt
: i -i
- -i i 4
—




. LE
t
V
_V :

- L
V


4
,
. V
4 -


_ H -
. I . . •

• V
II ‘ J—
ฑ
V

:
I

:•
I
I
:
.
-
1

I

V
iT TI
I
-1

I

T

- t L H.
J 1
I
H-!


H 1
rr
t
I L
H
-
N)
2
IV
ini nnc fli A2 AS 1
in 20 30 dfl SQ 6 I 70 2 I 90 95 02 99
99.8 99.9 99.99

-------
-125-
The effect of this noble metal catalytic device on the residual
hydrocarbons in the exhaust gas stream is presented in Table XXXVIII.
The mass distribution profiles for 1200 rpm, 2250 rpm, and the Dow
cycle follow in Figures 55, 56, and 57.
TABLE XXXVIII
EFFECT OF NOBLE METAL CATALYTIC MUFFLER ON HYDROCARBON EMISSIONS
Ahead of Muffler After Muffler
( ppm) ( ppm )
2250 rpm 130 30
250 135
80 45
125 30
1200 130 25
170 45
Dow cycle Mode #2 195
#3 50, 50 Not available
#4 160, 135, 135
#5 180, 190, 190
The cold trap condensates collected before and after the noble
metal catalytic device at 2250 rpm were analyzed for their
aldehyde content. The aldehyde content decreased from 330 ppm
formaldehyde in the condensate ahead of the device to 99 ppm
after the device.
The final series of tests (8-10) necessitated the use of the
Chevrolet 350 CID engine to accommodate the auxiliary air injection
equipment for the proprietary catalytic device. The particulate
sample weights from the dilution tube corrected for the complete
exhaust gas flow are given in Table XXXIX. The particulate emissions,
calculated in grams per mile, are shown in Table XL. The corresponding

-------
U ! PROBABILITY 46 8043
x 2 LOG CYCLES -‘ .
XE(IF FEL a r ctp ‘
ILFJ:!:l:E t E : :::
rF :d 1 - =
- -.-- - - -- .
• :
‘ TT - ::: :
-

-w
if--
— t- -14-
tf
— :r- :: — ==t— -
_ _:-็:-: :I 4 - :::::: = -:1 - : J -
H
; -ฑ : -
: :4JF - - —ฑ- - --t---- :L:L - 4- -
-r-ti - 1
i_: : 4 =:: -== :::=็
::— :: t: 4 :: -; :- =:I :::::-:t:•:- =:::: :
E
-

E
Li

TE

•
_
IIILITI ]= I IIrn iT
-4
1
- — —
LJ
— —‘ -—-- — - --- --—— a
_
- - - - j- ____
99.9 99.8 99 98
r :iii-1 E:TT:I ::.: TTT• —
i
o_
9
8
7-.
5—
!
95 90 70 (0 O 4fl 4(1
.. 4 kr4 :
:
I —
- I
I —
—
-:7 i * Iv- ; w. tr fp :
I I
. .
: . : I
70 10 5 ________ 2 1 0.5 0.2 0.1 0.05 0.01
fiTiTTTL L
I - ii
I , I
.. I
-—- — -_ L - - -
ฑ iฑ-
S
I
-1_i-
Itt

!
+ft
3--
2_
I
—
1 E4 -
I .
;i: fl :t.
-
:
H L—-
tt TT t: f iฑ 1--1i4
4 .• _ _
JI4 4j U-
v sUTI:
- --a-
ITTSfP I
. , 4 5 t-ii
1 I I : 1
S :
S S

I : .r t Lt :
--fFI--f ------
t -ti- till - — —
: j

; — TE
ฑff
-

LT
t S
-iii
I t II I Hft
ft
ft 1 T
-F l

- : I
i_ _ ii i
S St
I
H
8
7-
6
10
9
8
7
6
5
4
2
t
- - -
/
IT-i
‘ --- I H
2-1::
T - -
-1
-I--t—r-i—1-1-• -i--i -ฑ I t:
I It
I t It
-4- -1
4--
S -I-
-_ _L
S I
- -
-j: _ ..
+ j s II
:ft- 4 -L
I t T I
I • I
: _ ฑ - .ฑ: : II
----- .
.: :-i::: :;i ; -

__
T

S : —
:_ __ _ LI
:-i- t1: 4 T
•T -f -ฑT
.7
TF 4-
.T T
-H-
c
H
1__
T 4i -4-L
T
t -
kt
5555
Tff
‘EFfI
Hj-
4
3
-
4--t-- 1-
T : L4 T
1
-I -
-1 - t-i-
I I
99-99
n OQ OQ qq g gq .9

-------
in
99.9 99.8
I< * PROBABILITY 48 8043
XZLOGCYCLFV
kU i.rrrI i P
I !- !L
99 98
9--
8 _
7 - -
6_ _
5-——
qo 80
- - -f
4-- ;:
70 ;o 50 40 30
--
1
1-
jc
‘
JL 1 1
10
- - :-
_4_4 , -_- - 4- _____j______ ___ —4- - -1
j : ::
iii
i 1 4
I —
-J
— -.I - - —
H
1
F -il
I
-
ถ t1 t
i—:t - - .jIiLi
LIt I.I1I11IitT1 - T - :
— - — t- -H- t ii 4---
-w - - -
- -t t f --• t ‘
IT
E
1—
-
— —+-— -—
t it
Jil I4 -l — —i-— i—
4 4 — --
-— . .-i -4 - __-__4 --_ —— ; — —

-—---
. -I
t—
fT.
; t +4 _j L -t - -1
___
0.01
=- 10
— :i:— - — -
— -- - - 8
7
6
5
4
Ut t - - 1 --
trT1L -;
0.5 0.2 0 1 0.05
: = = =
- EEE EE
: :
---- — -
= ::
- - - - - -
- -
:ki T !j
- H-1 -
1’--t1 -
: : : 4 r
- T
-H
4--
R
- iI F
I
I L
1
T —
- -
T 9
----- T —
6.
5---.
-
i 11
-- H-
- -
11__1_I
-
T --r T-i-i--rr-r-
I
T

I
liii
tT-
E -
:-;:
a
: t .:i :1 :‘ .L_, -_
- .-.
-1
g
7
ii T !I 1IJL:1 L!
.
- H
___
-
III
i I
fH -Hฑ
4L
- -+
S
4
3
9
- - --T - - --
H-f-
I I I !-
T
I I
I
LLL
I I
I — I
L L V
I I
- 1 -7 I
i ILt I
- -r—1 - - 4
- I
— in ,fl fl Afl 0 0 70
I I
Tt
f } Hi T - F
1i i
-h
I I
80 90 9b 98 99
99.8 99.
99.99

-------
LI PRO A ILITY 46 8043
r :; ,c 2 LOG CVCL B ‘. .
I,rpLI & I e FR •.
99.9 99.8 99 g
_f - - -- - - - :- t-:t - -
— - : ; _ :
ฑ
:
: : : j : :
ฑj
I I
I -
20 10 5 2 1 0.5 0.2 0.1 0.05 0.01
!_ : -: r I rr E
:
t — ฑ - h--f--
-4--- L
T- t - :: :, _ _
- 4-+
t t EEEE fEEE
.-- -. - - -
- - -,- - - --4 -— —
R
7
in
i TTiri -- - -J-I
9_
8.
7
6_.
5------
ii I [ 1
95 90 RO 70 60 ! fl 40 30
T - r- T -H
1
TL :
L
.
T T

— —
:0:
:: :: :
- --- T
- -
---1 -
—4--- --I-
Thur - -
10
0
+ tI L 1 t L UI1 ’ ifl 7 :L
I
IT -
FL Fr- ‘ I j
I —t-t-r-t- t
tI i—t-i i-• it I I
: -11 1_4
--H- T

H
- —4-
I _I
1
=1
h
—-4--— _4—_ - 4- -— - -

t ’:: 1 L- t t -
w II II


-a:.

-
- -f: fl
I I I I I I
I I I
Q
t: H
— ฑ + -
7-
6
1 I I-E-T
4 -t -
- --- 1 - t.
:
-: :
4
3
2
8
7
S
5
A
03
---
- -__
! !
I I I
I I I
! I
ifi
II
T
—4
lilt I I ; i I I ! I I I I I
I
,T -
3
I I I
- i_ - ffl J
T t fft t I
I I I
I I
iwiJLu:EIii -L1IIJi t
I I I ! I
L
I I I
I I I
H
,A 01% An n GO 00
aQQ G 0000

-------
-129-
TABLE XXXIX
PARTICULATE SAMPLES FOR PHASE II, TESTS 8, 9, 10
(Dilution Tube Samples)
1 cfm 4 cfm
Tube Andersen Millipore Glass Glass
Slit Sweepings Sampler Filter Fiber Fiber
Dow cycle
3.3 hrs 0.000 g 1.11 g 1.05 g
7 1.55 2.40 3.54
7 2.1 3.45 3.19
Total 17.7 0.1922 g 2.8136 g 3.65 6.96 7.78
1200 rpm
24 hrs 0.0471 0.8261 3.35 4.20 4.18 4.14
2250 rpm
24 hrs 0.0515 2.1551 4.7 37.3 46.6 46.7
TABLE XL
EFFECT OF PROPRIETARY PACKED BED CATALYTIC DEVICE
ON PARTICULATE EMISSIONS
Filter 1 cfm Andersen Filter
4 cfm Filter Temp. and Filter Temp.
( gm/mile) ( ฐC) ( gm/mile) ( ฐC )
Dow cycle
3.3 hrs 0.0089 37-46 0.0084 Thermocouple
7 .0094 52-74 .0338 failed
7 .0117 40-72 .0074
1200 rpm .00578 35 .0046 28
2250 rpm .0324 68-73 (70) .0292 38-42 (40)

-------
-l 30-
mass distribution profile curves are shown in Figure 58. A curve
for the same engine and fuel without the proprietary catalytic
device is shown in Figure 59. The mass of emission amounted to
0.0114 gm/mile for the Andersen sampler and Millipore filter (1 cfm
filter); no 4 cfm filters were used in this baseline test. This
baseline data was developed under Contract CPA-22-69-145’.

-------
LJ 4 PROBABILITY 46 8043
I ’L X 2 LOG CVCLES
K UFFFL Th srru .
1o___ 99.99
9 .
8
7
6
4J ._. - -
- ---
L_
C I
ถfl
• Hฑ - -
... --
20 10 5
- __-_ [ _
.
- 1
.-.
99.9 99.8 99 98 95
x0 10 60 50 40 30
5 .
4-----
- I _=
- — -
I ;
I , •
.
-rT-1-
L
.
I
T ::f _fl_fl 1T I fljfl T
— — —-.--_4--- - ____ _
: ,
H

ftt
‘ ‘ ‘ J1’J i 1:i:
-i-LL ...
Ht I

—‘!•-—
. - .
.
;L1
: t
3--
- iTI: T: -:- -‘-E -t1-t
.4 4:4 .: - . . :::: : i; : i$ :: :t T
0.5 0.2 0.1 0.0 0.01
-- -, - . .— - - -—-—+-- -—-------— —-, - -, .- --4-4.—-4 .--—4— --4-’--l --
.. -- —. - - - —. —.——- —. . -. -- .. - — —. -——-. - .f—- i —.- I
L T:: :;: .-.:.:T::--T t :,
9
F’
. i- f-;-
I 1;I1Ii1JIi i 4 JI Ifl T -;: r i -
4
L - -i -4::w
i.
f-H-f; : j
EI L1
I!
u_,
- . I - t 4 • 1 k1: -
10
9
8
7
6
a
-- 4 - t - - .
LJ
L ... — —
r
4- - -
- - --:- ;ฑ-:-..
: : :
.: .=: :L:
:
T:
: : ::
II

144 .
- -- - —
uv
(
—.p—
t- j
-i- . - -i .- — — — —
Tn
!

U
— - - -
-- - - - - — -
. - —-. - -- - - -
, — --
1
E E E

4--.
3---
2.
C.’)
a
6
.5
t r
I - 4
- -t H-mi tmrr
II
I I -
I . I
WI 1 ;J i
I I I
: iT T: iiii: ’
_LJ_1 II LI I LI I I I II LI I
I I I I
I N M
I I I
H
I I I I
F Ii
I I I
I I
I I
I I
I I I
I I I I
I I I I
I I I I I I I I
III I I I I
! I I I
I I
I! I I I I
99-99
0.01 0.05 0.1 0.2 0.5 1
2 5 10 20 30 40 50 60 70 EU 90 9S 98 99
qq_R 99 9

-------
I < PROBAeILITY 46 8043
x 2 LOG CYCLES i. ‘. ‘ ...
K UFFEL & Es R
-a-
-- --
---
-t -
—
- - - -
--
- - - -
— - -
- -
I--
- - - -
_ _
t:— ::ETr : :r
— — - -- - -- — =4_,__ - -==.= . - : __= _;—;— - 1 -
I =E: : E:I: a ii i T: : :

- :E-i:i =- -
- IIi -
70 f0 50 40 30

I I I 1_
in 99.9 99.8
9___ -
8
7-—,-
6
99 98 95 90 $0
:
:_ _ ,_—
4
:
20 10 5
th H i rL
I
L -; :
i ฑ ;
:
,
. —, ..
0.5 0.2 0.1 0.05 0.01
-T— —r. T .. _ •T : - I 0
: ; E EE T ‘ E E+ Tฑ= E :
—-—-- -. - - -- -i-- —-—1 - --
_: L t. Ji LiL-
— —๗ -— -;
: i :
j L Ji j
.t1- -- -----.—--r:-T=.:. ::-- .-t.-—

•‘:.
., - I
ti
H
.- -4-
--t : : r::.
1• -
; i — E_-Ti-: :: tt -;- - -
L
—
I ’! 1-
:i: : :
, . ‘ :. ,i j
R
—t: I I
— I .
: , . • - - -1
: - ;
- ii -i- -
= = - :4:
I L
:t; - ;r - - - E -
i-i
:== : :
G)
- .
,- -
-t 1
tttt

I 1
I il --I ---
--. --..t
.1
7-
..-- :i

t:-

- H---
P - +
- -
, t!-I-
IiL I !
------ F
:EEEEE [ tE
6
5
4
3
2
1
9
8
7
6
3
7
--
:

-T T T -:
-a
(A)
,i

H - ’ - E+
— --i -
-i-
-
. .-.- -.-
::
= :.i : :
: 11 III :Ii I
, .
- - , -.

- -
I
-L
- -
ft
-—-. r- r
. , -.
:
:—j - i - I :n :
4 -4- -.+ .
rL
I I
ru i
I I tI I
- : L:L TE : : ‘ : : T T
III I I I I
! I • I : I
__t Li I j I I ‘ I _ - L
— — — —F
— — -1 — — — I -— ——--—-
--
I ‘
Nt-H
rn-H--rH--
I I
I I ‘ I
L
-1 I
L_ n-H--i
L I_1

-------
-133-
IX. DISCUSSION OF RESULTS
A comparison of the particle mass data generated in Tests 1-4
indicates that variations in air/fuel ratio had little effect on
the grams/mile of particulate matter collected at 1200 or 2250 engine
rpm. This engine variable had a similarly small effect on the
particle cumulative mass distribution profile.
At each engine speed, there was a slightly higher mass of particulates
from the lean air/fuel mixture and the mass of particulates was
greater (3 to 5 times) with the 1 cfm than with the 4 cfm filter.
The cumulative mass distribution profiles for the 1200 rpm tests
had a D 50 value of 3 .2p. The curves were identical for particles
larger than this median value. However, for smaller particles the
particulate matter from the lean mixture was slightly larger than
those from the rich air/fuel (A/F) mixture. At 2250 rpm the
cumulative mass distribution profiles were again similar but
not as nearly identical as in the 1200 rpm run. For the 2250 rpm
run, the median particle size (D 50 ) was l.5p for the rich fuel
mixture and O.S i for the lean fuel mixture; the particles from the
latter run averaged a smaller equivalent mass over the entire
profile.
Engine speed had a significant influence on both the total grams/mile
of particulate matter collected and on the particle cumulative
mass distribution profile. At 2250 rpm there was a significant
decrease in the grams/mile, compared to 1200 rpm, particularly with
the 4 cfm filter where the decrease was about one-half. The marked
difference in the mass of particulate matter (grams/mile) collected
on the 4 cfm filter compared to that collected on the Andersen
sampler and back-up 1 cfni filter is believed to result from the
longer residence time of particles in the 1 cfm sampling system
caused by the lower pumping rate. Another contributing factor
could be the cooling effect of the expanding gases after being

-------
-l 34-
forced through the Andersen sampler resulting in lower filtering
temperature. The latter factor is more noticeable at higher engine
speeds. The subject is discussed further in papers by Moran,
Manary, Herling, Karches, and Wagman 3 ’ 4 .
Introduction of the noble metal catalyst into the exhaust system
resulted in the emission of up to 22 times greater grams/mile
of particulate matter than in the baseline series of tests. At
the same time a sharp decrease in the mass median equivalent
diameter occurred at both 1200 and 2250 rpm. During this series
of tests, the mass of particles increased with increasing engine
rpm. At both 2250 rpm and under Dow cycling conditions, there
was good agreement in the mass of particles per mile as measured
by the 1 cfm and 4 cfm filters. The data on residual hydrocarbons
measured ahead of and after the device indicates that the catalyst
was 50-75 percent efficient in their removal. The concentration
of formaldehyde was also reduced by 67 percent over the catalyst.
The implication of this data is that the particulate material
collected in this latter series of tests is of a different nature
than that which was observed in the baseline tests (1—4). Since
there is a very significant decrease in the hydrocarbon content
of the exhaust stream, one would suspect that the increase in
grams/mile of particulate matter must be due to inorganic material.
This hypothesis is supported by the close correlation of 1 cfm and
4 cfm filter data in the tests with this catalytic reactor.
With the second catalytic device there was again better agreement
between the 1 cfm system and the 4 cfm system than was observed
without the catalytic device. There was also an increase in the
mass of emissions (grams/mile) when the engine speed was increased
from 1200 to 2250 rpm. A direct comparison of the two devices would
not be appropriate because of the difference in engines.

-------
-135-
A basis for the comparison of the Chevrolet engine at 2250 rpm
without the catalytic device is available from Contract CPA-22-69—145.
There was nearly a threefold increase in grams/mile of emissions with
the catalytic device in place. In contrast, there was a decrease
of the particle size in the cumulative mass distribution profile
when the engine speed increased from 1200 to 2250 rpm. A comparison
with the cumulative mass profile from the engine operating with a
conventional muffler indicates that the equivalent mass of the
particles is larger without the device. The effect of this device
was again to (1) decrease the average particle size, (2) decrease
the residual hydrocarbons and aldehydes, and (3) increase the mass
of particles per mile.

-------
-136-
X. FUTURE
Under the Clean Air Act Amendments of 1970, stringent regulation
of hydrocarbons, carbon monoxide, and nitrogen oxide emissions
from automotive exhaust systems is prescribed. Consequently,
an intensive program is underway to develop devices and procedures
to accomplish the required reduction of these emissions, and also
to develop rapid and reliable measuring devices for monitoring
compliance with the regulations.
While it is of major importance to the public health and welfare
that these regulated emissions be substantially reduced, it is
recognized that there are also other constituents of automotive
exhaust gases which may adversely affect health and welfare,
and further that the nature and quantity of these constituents
may be substantially and perhaps adversely affected by the
control devices or procedures which are proposed. Typical of
the constituents which are being identified are polynuclear
aromatic compounds (other than the oxides) and metallic compounds
with hazard potential.
During the course of the work described in this report, procedures
for quantitative sampling and for physical and chemical characterization
have been developed and successfully applied to the characterization
of exhaust particulates, aldehydes and other non-regulated emissions
from internal combustion engines operating under a number of selected
conditions and in several configurations. This program should be
extended to provide information on the character and quantities
of the non-regulated constituents in exhausts over a wider
representation of operating conditions as well as a wider range
of emissions control systems and of light-duty vehicle power plants.
In order to properly assess the performance of proposed emission
control devices and technology, it is essential to have information
concerning the potential hazard of the emissions from the
controlled systems. Potentially hazardous constituents have

-------
-137-
already been identified in some of these emissions, and the quantities
of these materials emitted by the proposed systems under a variety
of operating conditions need to be determined in order that their
true (quantitative) hazard potential can be judged. Concurrently,
there is also a need for refinement of the techniques of identification
and measurement both to provide more complete characterization of
the constituents, and to identify other constituents which may be
present.
To be properly effective, such information must be developed
concurrent with the development of the control technology for
the regulated emissions.

-------
-l 38-
XI. REFERENCES
1. “Effect of Fuel Additives on the Chemical and Physical
Characteristics of Particulate Emissions in Automotive
Exhaust,” J. B 0 Moran and 0. J. Manary, Federal Clearing-
house Report No. TV 196783, July 1970.
2. “New Sampler for the Collection, Sizing, and Enumeration
of Viable Airborne Particles,” A. A. Andersen, J. Bacteriol .
76, 471 (1958).
3. “A Comparison of Automotive Particle Mass Emission
Measurement Techniques,” 0. J.. Manary, J. B. Moran,
R. H. Herling, W. E. Karches, and J. Wagman, presented
at Combustion Institute 1971 Technical Session, Ann Arbor,
Mich., March 23-24, 1971.
4. “The Nature of Automotive Aerosols,” J. B. Moran, 0. J. Manary,
R. J. Herling, W. E. Karches, and J. Wagman, presented at the
45th National Colloid Symposium, ACS Division of Colloid and
Surface Chemistry, Atlanta, Georgia, June 21-23, 1971.

-------
-139-
APPENDIX A
ATOMIC ABSORPTION METHODS FOR ENGINE PARTICULATES AND DEPOSITS
Method for Lead Determination
Following nitric acid digestion, the samples are washed into 50-ml
volumetric flasks and diluted to mark. This normally puts the
concentration of lead in the flasks between 20 and 200 jig Pb/mi.
If the concentration is higher than 200 jig Pb/mi, redilute the
sample. The samples are analyzed on an atomic absorption spectro-
photometer (Perkin-Elmer Model 303) using a hollow cathode lamp
with a lead cathode filament. Operating conditions are as follows:
10 milliamps tube current, light path slit opening - 4, ultraviolet
light range, acetylene-air oxidizing flame, one-slot burner head,
wavelength - 2170 angstroms. The sample solution is aspirated
into the flame where lead atoms present absorb the light from the
lead cathode filament. The amount of absorbed light is proportional
to the concentration of lead. The samples are analyzed in
conjunction with the following series of lead standards: 10,
20, 40, 60, 80, 100, 150, and 200 jig Pb/mi. The concentration of
the standards is plotted versus their absorbance values giving a
standard curve. With the absorbance values for the samples and
the standard curve, we can determine the concentration of lead
in the samples. The sensitivity for the lead determination in
an air-acetylene flame is about 0.25 pg Pb/mi at 1 percent
absorption. The detection limit is about 0.1 g Pb/mi.
Determination of Lead and Other Metals in Fibergass Filters
The fiberglass filters cannot be digested completely with nitric
acid. They are cooked with concentrated nitric acid for two hours
to leach out the metals. The pulp is filtered and washed and the
filtrate analyzed by atomic absorption for lead, and by emission
spectroscopy for other metals.

-------
-140-
Determination of Lithium and Potassium
Lithium and potassium were determined for some LiNO 3 /KNO 3 catalysts
and some fiberglass filters. This procedure consists of dissolving
or leaching the sample and then analyzing by atomic absorption
(similar to lead method).
MICROSCOPY
The responsibility of the Microscopy Laboratory is to apply
light, transmission, and scanning electron microscopes in
characterization of the automotive exhaust particulates relative
to:
1. Color
2. Size
3. Shape, form
4. Fusion characteristics
a. Melting point
b. Sublimation
c. Crystallinity
5. Electron diffraction
6. Solubility
7. Organic-inorganic composition
8. Elemental analysis of X-ray fluorescence while in the
scanning electron microscope.
The work on the exhaust particle was divided into four parts.
Submitted here are summary reports of what has been accomplished.
SCANNING ELECTRON MICROSCOPY AND X-RAY FLUORESCENCE ANALYSIS
A. Purpose
To characterize (SEM) and identify (X-ray spectrometer) the
particulates in exhaust emission collected by the Andersen sampler
or collected on the Millipore filter following the sampler.

-------
-141-
B. Instrumentation
Cambridge Stereoscan Mark 2A
Ortec Non-dispersive X—ray Detector
Nuclear Data Analyzer
Varian Vacuum Evaporator
Kinney Vacuum Evaporator
C. Work Outline
1. Particle characterization (SEM) on plates
of the Andersen sampler.
2. Particle identification (X-ray).
3. Single element X-ray scan.
4. X-ray spectra on impingement area of plates
and spectra on final filter.
D. Techniques and Methods
1. Substrates for sample collection. Most satisfactory for
photomicrography were micro cover glasses, while where X-ray
analysis was to be done ultra pure carbon strips proved best.
Silica interference from the micro cover slips, halogens in
epoxy, and thermal instability in mylar film reduced the
desirability for using these materials as substrates where X-ray
analysis was to be done.
2. Storage and sample preparation. All samples were maintained
in a dry atmosphere from collection to examination. Both the
glass cover slip and the carbon strip substratum were attached
to SEM sample stubs with conducting silver paint. Samples for
SEM characterization were made conductive with a thin layer
(%200 ) of gold or gold-palladium evaporated onto under vacuum
(5 x lO Torr). Graphite carbon was sputtered on the samples
used for X-ray diffraction.

-------
-142-
3. Normal operation for the Stereoscan.
(a) Gun potential - 20 to 30 kV (depending on degrading
of sample and resolution needed).
(b) Vacuum - ‘ l0 Torr maintained.
(c) Sample angle — 200.
(d) Working distance - 11 mm
(e) Polaroid P/N Type 55 film with 100 sec exposure.
4. Normal operations for X-ray Spectrometer (warranted 215 ev
FWHM resolution)
(a) Gun potential - 30 kV.
(b) 1024 channel - Series 2100 Nuclear Data Multichannel
Analyzer.
(c) Collection time — 200 sec.
(d) Count rate - 600 c.p.s.
(e) Spectra recorded on Moseley 7035B X-Y Recorder.
(f) Single channel recording
Polaroid P/N Type 55 film
400 sec or 800 sec exposure depending on concentration.
E. Analysis
1. Particle characterization and photomicrographical
documentation was done with the scanning electron microscope
employing standard operational procedures.
2. Particle identification involved elemental analysis using
the X-ray spectrometer on the scanning electron microscope.
This included, for multiple particles, full spectrum elemental
scan, and single element scan. Spot scans were carried out on
single particles or in specific regions of particles.

-------
-l 43-
APPLICATION OF LIGHT MICROSCOPY TO AUTO EXHAUST ANALYSIS
1. Low Power Photomacrography (15x )
This documentation is done with oblique reflected light
showing the sample as received against a black background.
2. Polarized Light Photoniicrography (400x) and X-ray Diffraction
The particles are scraped from the glass plate using a disecting
knife and transferred to a clean microscope slide and covered
with a clean cover slip. The slide is then placed on a hot
plate ( l4OฐC) and a drop of Arochior 5442 (also heated to
“ .140ฐC) is applied at one edge of the cover slip. Capillarity
draws the Arochior across the preparation. The particles are
dispersed by using the eraser end of a pencil to move the
cover glass around a little. The excess Arochior is removed
with absorbent paper tissue. When the slide is removed from
the hot p1ate, a permanent mount is obtained.
A like sample to that used for polarized light microscopy was
impregnated into a small amount of rubber cement and worked
into a ball and then mounted on the end of a 500 p diameter
glass rod. Such samples were taken for an X-ray diffraction
identification.
TRANSMISSION ELECTRON MICROSCOPY OF AUTO PARTICULATES
The purpose of this work of using the transmission electron
microscope (TEM) was to characterize and photographically document
those particles which are smaller than one micron and to obtain
crystallographic analysis of them by means of electron diffraction.
The sampling for this work was done by attaching transmission
electron microscope grids, which had been previously given a carbon
reinforced collodion substrate, direct1y onto the Andersen sample
collection plates and on the Millipore filter after the sampler.
After the sampling, the grids were removed and stored in a desiccator
to await analysis.

-------
-144-
The TEM used was the Philips EM300 where standard instrumental
operating conditions were employed including a 100 kV operating
potential to give the best penetration of the electron dense
particles.
The particles under consideration here were those ranging in size
from 100 to 1000 A. Photomicrographic documentation was carried
out recording similarities and differences as these were observed.
POLAROGRAPHIC ANALYSIS
The analytical method for the determination of carbonyl compounds
in automotive exhaust emissions employed polarographic techniques.
A Princeton Applied Research Model 170 Electrochemistry System
was used as the monitoring device. The derivative pulse polarographic
mode yielded the best combination of resolution and sensitivity
for the classification of carbonyl compounds. A dropping mercury
electrode with a Princeton Model 172 Drop Timer was employed as
the working electrode.
Hydrazine derivatives (hydrazones) were employed for the determination
of the carbonyl compounds, since hydrazones are easier to reduce
than the free compounds, thus eliminating many possible interferences.
An acetate buffer of approximately pH 4 (an equimolar mixture of
acetic acid and sodium acetate, 0.lM in water) was used to control
pH for hydrazone formation and also acted as supporting electrolyte.
Hydrazine was added as a 2 percent aqueous solution. In this system
formaldehyde gave a peak potential (half-wave potential) of -O.92v
vs. a saturated calomel reference electrode. A platinum wire was
employed as the auxiliary electrode.
With the above system, it is possible to distinguish between and
simultaneously determine aromatic aldehydes, formaldehyde, higher
aliphatic aldehydes, and aliphatic ketones as shown in Figure A.

-------
-145-
L 1
: L4 - -- - -
L Li
!
- -t
flT
j i
1.
r
i3 t
-tf44T 4
tt-4-ฑ 4 t
-1- -T 4
1 — *— -
-- -- ---4 . ,
t 1 • f
- -4- -
- *1
- -4 .
!! t’ ‘II r Trf1 ‘
I: 4ipiTi
1T I I I..
ฑ . H- -Hi
tt
- U: fli
__i_J_
-r-r- TH
tt1. j T tf

—., 1
L
1
—I. — .
-Li J-
1 + ‘ -H -
- -

- .—. - -
:1i J
-- :i
:EI iL E Tt

—. —--
.Q -.- - -—--
jฑiE:
. —--
- - .
w
-
. ::T:::
o-
- . . - -
i :

::
)t ;J 1 J Jt

-. -
—
• •_
--- .


ฑ
-:i: L r:T:




I 14
-
—----..
- .-—-———— - .- -.
-——-----.---—
:ฑ
.--—-.
- “-

—.- —--. . . ..
:
- . .. .j
. ..J
. .

. .- —‘
. .
::
:
-: ,


I


—+ ,
.

-‘- . .
T E j;I tB
I J’F I IE
: k::=i:
J: . :t:i:
iii :1H
: ‘
- +- -k+- ฑ - 4 -t: -Tฑ
4I1I -+- -- & t _’
- —1• • - - - -
_____
. .‘- -+ --—--
.__i_:__ .
- 1’
I
-i - - -—- - - - -,- —1- - -—
fj • . . f . . I - t- -
.
.L.::: .J
r j
4 _
TL
:
4 . L EjII 444 .
1
I

L. +fmH -H
- i
- tr1—n-i -
.- -- s4-.
. .. L. 1 4 .. . . ..L.

JI
- - —
:


1 ฑii
p :I
{ 1”
t I 3 ;
jill
. H’ 4 - - --H-1
-“--F--
- - - - -H- -4 H-

4 .J44 L


ft- -

t -tt-t iH - - - -
i 4 t —t—f -- 4 -- — -:f-


.fff44f .L 4 j
- ฃ_4_ ‘ ‘ . L.

.Lf4 / TTT .T I TLT


t tt - L
fft 1 4,..

L_ . . Lt 41 .+4 . . ..4 .

J . L.J. Li J .


-

E 4J:

Hr

ฑhJif*


- -t -


f tm r’-i-
f- Hii”

H-’ 4 f -f- -
.


I 4 L 4.r
Iฑ+
tiJฑ
L. H -1 -


-- . .•

.
I t LL
-i--r-----
t j tiฑ

H
-


. 4 1


‘ .
p iฑ t:
-

4 t - - -ft
-
i- -- --f H
.4 , 4. ..
r
ii j.jj 4

it-fl t-
L J fif

4.4
,
. . I_
. •
1 T
A i---- —‘--S - -_4_-
—t--r +-4 4-
-ft-1--H t H1
- I TT

4I I
4-
4 t _L 1
—I
IL
4 t

-------
-146-
Since aromatic ketones, e.g. benzophenone, give polarographic
response in pH 4 buffer without hydrazine, it is also possible
to detect aromatic ketones. Lead and zinc could also be determined
from the samples under these conditions.
Since formaldehyde was the main carhonyl component of the condensate
samples, all results were calibrated against and reported as
formaldehyde. The upper curve in Figure B shows an actual
sample without hydrazine present and demonstrates the lack of
interference in the carbonyl region. The lower curve shows the
same sample after the addition of hydrazine. Figure C shows the
same solution after the addition of a formaldehyde standard.
These two figures clearly establish the presence of formaldehyde
in the exhaust samples.
SPECTROSCOPIC DETERMINATION OF METALLIC IMPURITIES IN ORGANIC
COMPOUNDS
1 . Scope
This method provides for the determination of ten elements in the
range of 0.00001 to 0.01% in organic compounds. Although the
method specifically describes the analysis of organic materials,
the same method can be applied to the analysis of many inorganic
samples. Nor is the technique limited to the ten elements listed,
but it may be extended to cover practically all of the metallic
elements.
2. Principle
The organic matter is destroyed by wet ashing in sulfuric, nitric
and perchioric acids. The resulting solution is taken to dryness
and the residue is taken up in a spectroscopic buffer solution
containing the internal reference element, palladium. A portion
of the solution is dried on pure graphite electrodes. The electrodes
thus prepared are excited in an a.c. arc discharge and the spectrum
is photographed. The intensity ratios of selected lines are

-------
: ;
t Ht
I_f—I •f—1——r 4
;—t—— f——f•-f I +
1t1rr
-i--t---t--f t- 1- 1
•
f .7
: . . : ..
i —,—
+ —-—t—.• .
tT
ry•-r-r-T-
-.
1
—3. l
. —..
.‘ ,
‘
, ‘
• . :

--- —--- ------- . — - - --
1mic ro inp e E1i11
I
4 . . r t
. .• J . 0 ‘
. , c,.. , . . ;
I , , 4 t . .
t I, I
- -- - -- -

‘ t t
‘
.
! I • ‘
* . I ‘ • I
r’
,

t t -
.
,_,,; .

‘ • - t
I
4
-
t
[
;- ,
.i
t-t
I,,
t-,
- I ‘
; .
j l_. _i :
t- t -i t
- 4 ti t
— t— -
...



,
- -;--- -
.

. -.-. . . --
[ II T_ J : : , . _ . T__.;.__;. _ —: : : : : : :

i L c L * t

‘ฑ: ::: : :T:; ฑr T t
—f - 1 r iiLJJi : .i rtH 44 .4liH!
-
-+ •‘- -- •
- t 4 r 1 -’
-iL -
- tJ:jr 1 jr FJJ E :
-
.i 1- —r r —r--i—-

1 ft tTT : :i i :IT. : i 4 LL .


4

it ii ++

-— - -- - .- - t

- - --t —
- - - - H- - - 4-

• . : ;

l
+ i
L; tt

1• j
-I-
.
4 — i--


- -* - -



- f I I ‘

- - -

, :
4t -
i 4ฑ
1 J1
-.


- -
L --


4 T I1I

-

4

+
I
4
i .
T ฑ _
i T4!:a
i t t : ;
t I :t ฑ :1_-i -4 44

-
- ++-- Hi
-
tT I ’ 1Li
- -4 - 4
••••4•••• t
+


y; h
t i i
- - .t- -
I - Ii
- - - - - -
-, t-- - - I 1 -
I
i- f-f 1 f - -
- +- f
— 4 - -ฑ-
-- -- -
— -




-- - l -
—4-
-* -
- i: ; ‘t’
t-
4. . -- - - . tt
- - - -- - -r —- -- -
- - - - - - T t - f 4 1
--4 -L .
4 t’ 4 L L
r - b t
- -4f -
t -’ 4 - 1 -
,1
z144t4L
4 t t-t
- F- ฑ + — -

t T - 4- - - 4
-4- 1—-4------------
1 4
-
L [ T

-------
: : : : ii’ :•:i.Ti 1 I1ii.4
1 I
it i 1 1 4:
: ti ti
.-
II
iT E
i1 ‘I•
1— -r .-4--t- --.. ‘ —

r •
lt 1it1-t tt t:i ft
jijH 1 r 4 J
tii T iTt
L’i : I I
T 1 t • ‘ H
) •S t— 1 I - ‘ t
.ti—r1• ‘ .
+ttt j
•f 1 1 Ifl
-t . L

I j •‘
I • I
I !
k :
$ I ,
.
t—. .
4
4-. .
I

. .....
‘ j 1e&
_ i •H

H
I

1 1
1 ‘
::tiii

j•••
-4- 1 H- -
-H -H- I๗
1
— I •
-i- - !
. ‘ - -4 -
- 1 -t• I 1 - - -
11 :t. ‘r
: 4 I1 [
tI • ii 4 : _
i -f.-, - ‘
1 t • • I
1 : :
l....II
i- H-t .
-. -
- --
I “ — • -L
..—
I -t- ฑ -
I-”# +
4 v-

- t-; -- --t-- - ---4-—
- .-- . --.
- J tt : ac i:
- -— - “ ---- -

—-
:: : :

- — - -
-- r
-t --.---- - - -. -y
.

--

:

: : i i
- -
-H
-
-

-- --- --- - - -H
T -; - ;- - ‘1 TF ! -I
t - - -tI- r T i 1 r f-f - -1- -
t 144 T — ff- iT4 ‘ i’ I i 11 1 T I t
t t H T -
4 ็ 4
. i 1 - 1.1 1111!! . I tI - T- — - -
H t t i
. . i +.
—— -H ! -- - - ‘ f- f - - - t r
-
•-t --- —f- -f - f-fฑhHit H —H-+ j t
T 1tf1 ti —t- ‘ T — -t -t1 - it t

-L- - - - -- 4 -F
4 IiTf
iL i
- -H- 4 - 4 -- -i-- - .f 41
-ti’ iฑLT 4 1 t 1 T ฑ t ti -ฑI TI
J 4 H - ‘---i - - - I
4 *I
ฑ 4
I r I

Ir L - I, t JH -
- I
44 44 ii

-------
-149-
determined photometrically and the concentration of each element
is read from an analytical curve relating intensity ratio to
concentration.
3. Apparatus
(a) Excitation. Excitation is obtained by the use of a 2400 volt
a.c. arc discharge. Jarrel-Ash Custom Varisource, Jarrel-Ash Co.,
26 Farwell St., Newtonville 60, Mass., or equivalent.
(b) Spectrograph. The instrument used is a Baird 3 meter grating
spectrograph. Reciprocal dispersion is 5.55 A/mm in the first order.
Cc) Developing equipment. Jarrel-Ash Company. Plates are developed
in a thermostatically controlled developing machine, washed and
dried over heat in a stream of air.
(d) Densitometer. Spectral lines are measured with a non-
recording projection type densitometer. Densitometer Comparator,
Baird Associates Inc., Cambridge, Mass., or equivalent.
(e) Calculating equipment. A calculating board is employed
to covert densitometer readings to log intensity ratios. Jarrel—
Ash Co.
(f) Wet ashing equipment. A micro Kjeldahl digestion rack is
used for wet ashing the organic solvents.
4. Reagents and Materials
(a) Distilled nitric and perchioric acids. CAUTION! Perchloric
acid is an intense oxidizing agent. Do not permit organic matter
to be heated in perchioric acid unless in the presence of sulfuric
or nitric acid.
(b) Sodium nitrate, reagent grade (NaNO 3 ).
(c) Palladium dianine nitrite, Pd(NH 3 ) 2 (N0 2 ) 2 .
(d) Water soluble salts of the elements Al, Ca, Cu, Fe, Mg,
Mn, Ni, Pb, Sn, and Zn.
(e) Electrodes, high purity graphite, 1/4” diameter by 3/4” length.
Ultra Carbon Corporation, Bay City, Michigan.
(f) Photographic plates, Eastman Spectrum Analysis No. 3.
(g) Kjeldahl flasks, 10 ml.

-------
-l 50-
5. Calibration
(a) Dissolve 0.2182 gm of palladium diammine nitrite, Pd(NH 3 ) 2 (N0 2 ) 2 ,
in water, add 10 ml of concentrated reagent grade nitric acid,
and dilute to volume with water in a 100 ml volumetric flask. This
solution contains 1 mg Pd per ml.
(b) Prepare the following buffer solution. Dissolve 20 gm of
sodium nitrate in water, add 5.0 ml of the palladium solution (5a),
7.5 ml of concentrated reagent grade nitric acid, and dilute to
100 ml. Multiples of these numbers can be used to provide larger
amounts.
(c) Prepare a stock solution containing 0.01% (0.1 mg/ml) each
of the elements Al, Ca, Cu, Fe, Mg, Mn, Ni, Pb, Sn, and Zn. Dilute
two aliquots of this solution ten-fold and one hundred-fold to provide
0.001% and 0.0001% solutions.
(d) Make standard additions of the impurity elements to Kjeldahl
flasks as shown in Table I.
(e) Add 0.5 ml of concentrated reagent grade sulfuric acid to the
Kjeldahl flask and evaporate to dryness. Allow the flask to cool,
add 1 ml of concentrated nitric acid and evaporate to dryness again.
The residue is taken up in 5 ml of buffer solution, warming, if
necessary, to put the salts into solution.
(f) Polish the ends of the 3/4” graphite electrodes on filter paper
and place in a stainless steel drying tray. Place a drop of kerosene
on the top of each electrode to seal the porosity and allow to dry.
Prepare one pair of electrodes for each of the standard addition
solutions by pipetting 0.03 ml of the solution onto the end of each
electrode. Dry the electrodes slowly over micro burners in a gas
drying oven and store in a desiccator until run.
(g) Excite the samples in water cooled electrode holders using
the following conditions:
(1) Current, 4.0 amps, a.c. arc.
(2) Spectral region, 2150-3550 A.
(3) Slit width, 5 Op.
(4) Electrode gap, 2 mm.
(5) Pre-burn period, 10 seconds.
(6) Exposure period, 90 seconds.

-------
-151-
(h) Calibrate the emulsion by use of a stepped filter or by other
recommended methods described in the ฐReconimended Practice of
Photographic Photometry in Spectrochemical Analysis” A.S.T.M.
Designation: E116, Methods for Emission Spectrochemical Analysis,
(1964).
(I) Process the emulsion according to the following conditions:
(1) Developer (Dig, 20.5ฐC), 3 1/2 minutes.
(2) Stop bath (SB-4), 1 minute.
(3) Fixing bath (Kodak Rapid Fixer), 2 minutes.
(4) Washing, 3 minutes.
(5) Drying, in a stream of warm air.
(j) Select the analytical line pairs from Table II. Measure the
relative transmittances of the internal standard line and each
analytical line with a densitometer. Convert the transmittance
measurements of the analytical line pairs to intensity ratios by
the use of an emulsion calibration curve and a calculating board.
(k) Construct analytical curves by plotting concentration as a
function of intensity ratio on log-log graph paper. For best
results, plot the average of at least four determinations recorded
on two plates.
6. Procedure
(a) Weigh a 0.4 gm or available sample directly into a Kjeldahl
flask. A 0.4 gm sample size is seldom ever used because the amount
of material available is much less than this. A fraction of a
milligram may at times be the entire sample. Sulfuric acid is
not used in the wet ash procedure because the sample usually
contains a large amount of lead which would form the insoluble
sulfate. Wet oxidation is carried out with nitric and perchioric
acid only. Extreme caution should be exercised in the use of
this technique. Add concentrated nitric acid dropwise, a few
tenths ml at a time, to the hot mixture to aid in oxidation. A
few drops of concentrated perchioric acid may be added to the hot
solution after most of the free carbon has been destroyed, to
hasten complete oxidation. When the solution becomes water clear,

-------
-152-
evaporate to dryness. Allow the tube to cool , add 0.5 ml of
nitric acid and evaporate the solution to dryness. Repeat the
addition of 0.5 ml of nitric acid and evaporate the solution to
dryness again. The inorganic residue is dissolved in dilute
nitric acid and the volume is adjusted to a known concentration,
usually 10 mg/mi. If the original sample size is below 30 mg,
a less concentrated solution is usually made up. Aliquots of
this solution are taken to dryness and then the buffer solution (5b)
is added in an amount to give a dilution factor of lOOx. One
sample is analyzed by the direct reader while another one is
examined photographically. A sample may have to be run at
factors larger than lOOx in order to get the concentration for
some elements to fall within the range of the analytical curves.
By varying the sample to buffer ratio any number of concentration
or dilution factors can be achieved. Carry through a blank of the
acids used in the same manner as the sample.
(b) Proceed as in 5(f), (g), (h), (i), and (j) of the calibration
procedure. Duplicate spectra are recorded for each sample.
7. Calculations
Convert the intensity ratios to concentration by use of the
analytical curves.
8. Precision and Accuracy
Representative precision and accuracy of the method are given in
Table III. Each of the twelve samples A 1 , A 2 , A 3 , B 1 , B 2 , B 3 ,
C 1 , C 2 , C 3 , D 1 , D 2 , D 3 , was analyzed by means of duplicate
excitation.
9. Literature References
(a) Hess, 1. M., Owens, J. S. and Reinhardt, L. G., - md. Eng. Chem.
Anal. Ed., 11, 646 (1939).
(b) American Society for Testing Materials, 1916 Race St.,
Philadelphia 3, Pa., “Recommended Practice of Photographic Photometry
in Spectrochemical Analysis,” ASTM Designation E1l6 (1964).
******************

-------
—153-
The analytical procedures given herein have been adapted from
literature sources or developed upon the basis of experimental data
which are believed to be reliable. In the hands of a qualified
analyst they are expected to yield results of sufficient accuracy
for their intended purposes. However, The Dow Chemical Company
makes no representation or warranty whatsoever concerning the
procedures or results to be obtained and assumes no liability in
connection with their use. Users are cautioned to confirm the
suitability of the methods by appropriate tests.

-------
-154—
TABLE I
Internal Standard
Line A
Concentrat ion
Range S
3027.91 Pd
‘I
ml. of standard addition impurity solution
Concentration
Blank
0.00001%
0.000025%
0.00005%
0.0001%
0.00025%
0.0005%
0.00075%
0.001%
0.0025%
0.005%
0.01 5
Element Analytical
Line A
0.5 ml.
0.0001%
solution
125m1
“
“
0.25 ml.
0.001%
“
0.5 ml.
“
“
1.25 ml.
“
“
2.5 ml.
“
“
0.375 ml.
0.01%
“
0.5 ml.
“
“
1.23 ml.
“
“
25ml
“
“
5.0
‘
II
TABLE II
Analytical Line Pairs
Al
3092.71
0.000025-0.0010
Ca
3179.33
0.00025-0.010
Cu
3273.96
0.00001-0.00025
Fe
3021.07
“
0.0001-0.010
Fe
3020.64
“
0.000025—0.0050
Mg
802.69
“
0.000025-0.0010
Mg
2779.83
“
0.0005—0.010
Mn
2933.06
I’
0.0005-0.010
Mn
2794.82
“
0.00001—0.0010
Ni
3414.77
“
0.000025—0.0010
Ni
3037.94
“
0.0005-0.010
Pb
2873.32
0.0010-0.010
Pb
2833.07
fl
0.00005-0.0050
Sn
3175.02
“
0.00005—0.0050
Sn
2863.33
“
0.00075-0.010
Zn
3345.02
Background
0.0001-0.010

-------
T’ALJTt iTT
ฃA1L 4. J ฃ .&,
Representative Precision and Accuracy
SAl %ca
%Fe 5Mg %Mn Ni
%Pb %Sn %Zn
%Cu
A 1 0.000044
0.000052
0.00043 0.000048 0.00043 0.00049 0.00046 0.00047
0.00050 0.000054 0.00055 0.00052 0.00057 0.00055
0.00056 0.00052 0.00040
0.00059 0.00059 0.00045
A 2 0.000045
0.00043 0.000046 0.00044 0.00047 0.00051 0.00045
0.00050 0.00053 0.00054
0.000052
0.00037 0.000047 0.00043 0.00050 0.00050 0.00051
0.00051 0.00050 0.00040
0.00004
0.00043 0.000050 0.00046 0.00053 0.00049 0.00047
0.00052 0.00050 0.00052
0.000052
0.00050 0.000049 0.00046 0.00049 0.00046 0.00048
0.00053 0.00046 0.00042
B 1 0.00012
0.000097
0.00105 0.00012 0.0010 0.00105 0.0010 0.0010
0.00093 0.00010 0.00094 0.00095 0.0012 0.00096
0.00105 0.0011 0.00094
0.00098 0.00094 0.0012
2 0.000097
0.00096 0.000099 0.00000 0.00092 0.0011 0.0010
0.0010 0.00105 0.00125
0.000094
0.00069 0.000095 0.00105 0.00091 0.00066 0.00105 -
0.00105 0.00105 0.0010
B 3 0.000082
0.00085 0.000095 0.0010 0.0010 0.00086 0.0010
0.0010 0.00099 0.00096
0.00011
0.00074 0.000096 0.0010 0.00090 0.00092 0.00105
0.0010 0.0010 0.00115
C 1 0.00028
0.00030
0.0023 0.00023 0.0025 0.0023 0.00265 0.00245
0.0018 0.00028 0.0030 0.0023 0.00195 0.00265
0.00235 0.00255 0.0014
0.00255 0.0027 0.00215
C 2 0.00020
0.00225 0.00023 0.0023 0.0023 0.00265 0.0023
0.00245 0.00215 0.00225
0.00023
0.00285 0.00025 0.00235 0.0024 0.00275 0.00245
0.0026 0.0023 0.0030
C 0 00024
0.0025 0.00026 0.00275 0.0023 0.00245 0.0026
0.0025 0.0025 0.0030
0.00028
0.00275 0.00028 0.00285 0 0024 0.0025 0.00255
0.00245 0.00265 0.0020
0.00074
0.00064
0.0070 -- 0.0065 0.0057 0.(X)59 0.0065
0.0064 -- 0.0063 0.0051 0.0058 0.0058
0.00 6 0.0064 0.0058
0.0045 0.0059 0.0050
D 2 0 00059
0.0049 -- 0.0057 0.0048 0.0045 0.0056
0.004 0.0053 0.0050
0 0006
0.0057 -- 0.0059 0.0047 0.0048 0.0057
0.0048 0.0057 0.0060
D 3 0.00059
0.0048 0.0050 0.0045 0.0047 0.0050
0.0043 0.0054 0.0037
0.00053
0.0060 -- O.0O $ 0.0055 0.0054 0.0055
0.0049 0.0049 0.0041
A 1 , A 2 , and A 3
0.0001% of Al
and 0.0025% of
element.
contaj . 0.00005% of Al a 4 Cu, aLd 0.0005% of each ether
and Cu, and 0.0010% of each ether element. C , C, and C 3
each other element. D 1 , D 2 and D 3 contain & 0005% of Al
element. B , B 2 , and 83 contain
contain 0.&)025% of Al and Cu
and Cu and 0.0050% of each other
U,

-------
-156—
APPENDIX B
TABULATION OF CANDIDATE TRAPPING MEDIUM
BY INCREASING MELTING POINT
Temperature
ฐC ฐF
7 7% KNO 3 — 93% HCN
1 3)4 YF 3
17 63 MOF 6
30 77 Ga
36 97 NH CN
57 135 9% KC1 — 91% SbC1 3
70 158 2% KBr — 98% SbC1 3
73 163 SbC1 3
77 171 GaC1 3
86 187 2)4% LiNO 3 — 5% NFLC1 — 71% N1-LNO 3
87 188 SbBr 3
89 192 13% KBr — 87% A1Br 3
90 19)4 8% NaBr — 92% A1Br 3
92 198 41% KNH 2 — 59% NaNH
97 207 0.7% MgBr 2 — 99.3% A1Br 3
98 208 A1Br 3
100 212 56% NH CNS — 4)4% (NH 2 ) 2 CS 3
113 235 9% NaC1 — 8% NaNO 3 — 83% NH NO 3
18% NaNO 3 — 5% NH C1 — 77% NH 4 NO 3
11)4 237 17% LIC1 — 83% Aid 3
i8 2)4)4 20% NaNO 3 — 2% Na 2 SO — 78% NH NO 3
119 2)46 23% NaNO 3 — 71% NH N0 3 — 6% (NH ) 2 SO
121 250 21% NaNO 3 — 79% NJ-LNO 3
12)4 255 7)4% KCNS — 26% NaCNS
28% NaC1 — 72% A1C1 3
8% NaC1 — 92% NH NO 3
131 268 27% KNO 3 — 73% AgNO 3
135 275 23% A1C1 3 — 77% SnC1 3
11% KC1 — 1% NH C1 — 88% NH N0 3
13% KNO 3 — 9% NH C1 — 78% NH NO 3

-------
—157—
Temperature
ฐC ฐF
136 277 26% KC1 — 714% Cad
137 279 12% KC1 — 88% NH NO 3
2% NaC1 — 89% NaNO 3 — 9% NH C1
139 282 70% KNO 3 — 30% L1NO 3
1 42 288 (NH 2 ) 2 cs
9% KBr — 91% AgNO 3
1143 289 10% KC1 — 14% KNO 3 — 86% NH N0 3
148 298 36% RbC1 — 6’4% CaC1
150 302 NF1 CNS
157 315 114% KNO 3 — 86% NH NO 3
158 316 23% KC1 — 77% Aid 3
159 318 5% KC1 — 95% AgNO 3
160 320 17% KNO 3 — 8i% NH NO 3 — 2% (NH ) 2 S0
162 3214 7% NaNO 3 — 93% T1NO 3
165 329 50% KCNS — 50% RbCNS
170 338 NH NO 3
172 3142 33% Cu 2 C1 2 — 67% SnC1 2
173 343 KCNS
175 3 147 80% KHF 2 — 20% NaHF 2
176 3149 12% K 2 S0 4 — 88% NH NO 3
178 352 76% SnCl 2 — 214% TeC1
180 356 20% KC1 — 80% SnC1 2
66% SnCl 2 — 314% ZnC1 2
182 360 28% KBr — 72% CuBr
114% KNO 3 — 86% T1NO 3
183 361 13% NaC1 — 87% SnC1 2
95% Na 2 H 2 S 2 0 8 — 5% Na 2 S 2 0 7
185 365 58% KOH — 142% NaOH
186 367 NaH 2 S 2 O 0
11% MgC1 2 — 89% A1C1 3
38% KNO 3 — 19% NaNO 3 — 143% Pb(N0 3 ) 2
190 3714 A1C1 3

-------
-158-.
Temperature
ฐC ฐF
195 383 RbCNS
12% CaBr 2 — 88% A1Br
30% KC1 — 9% NaC1 — 61% ZnC1 2
201 3914 88% K 2 H 2 S 2 0 6 — 12% K 2 S 2 0 7
206 1403 NaNH 2
Ti NO
149% LINO 3 — 51% NaNO 3
207 1405 141% KNOB — 59% Pb(N0 3 ) 2
208 1406 53% KNOB — 37% NaNO 3 — 10% Sr(NO ) 2
210 1410 KHSO
211 1412 68% KNO 3 — 32% Ca(N0 3 ) 2
212 14114 AgNO 3
2114 417 K 2 H 2 S 2 0 6
27% FeC1 3 — 73% ZnC1 2
215 1419 9% KBr — 8% NaBr — 83% ZnBr 2
216 1421 33% CsC1 — 67% CuC1
96% NaNO 2 — 14% Ba(N0 3 ) 2
218 14214 11% NaNO 3 — 89% AgNO 3
219 1426 140% KNO 2 — 60% NaNO 2
221 1430 145% NaNO 2 — 55% NaNO 9
227 14140 KHF 2
228 ‘ 1142 3% KBr — 97% HgBr
29% RbBr — 71% AgBr
230 AgC1 O 3
5% MnC1 2 — 95% SnC1
232 1450 27% KC1 — 73% ZnC1 2
233 1451 10% CdC1 2 — 90% SnC1 2
236 1457 HgBr 2
K 2 Cr 2 O 7
237 l58 59% NaNO 3 — 41% Ca(N0 3 ) 2
2141 466 50% NaOH — 50% RbOH
2 1 12 1468 18% Cu 2 C1 2 — 82% ZnC1 2
2145 L 173 0.5% MgC1 2 — 99.5% SnCl 2
2146 1475 SnC 1 2
2 18 1478 NaC1O 3

-------
—159—
2% Li 2 CO 3 — 98% LINO 3
16% NaC1 — 814% ZnC1 2
14% KC10 3 — 96% NaC1O 3
8% NaC1O 3 — 92% AgC1O 3
92% NaC1O 3 — 8% BaC1O 3
1% KC1 — 99% NaC1O 3
1% KNO 3 — 99% NaC1O 3
97% L1NO 3 — 3% L1 2 30 4
14% NaBr — 96% NaC1O 3
96% NaC1O 3 — 14% Na 2 Cr0
98% NaC1O 3 — 2% NaNO 2
1% NaC1 — 99% NaC10
98% NaC1O 3 — 2% Na 2 CO 3
36% RbC1 — 614% AgC1
0.2% CsC1O 3 — 99.8% NaC1O 3
99% NaC1O 3 — 1% NaF
LINO 3
31% CsC1
67% KNO 3
6% KC1 —
39% NaBr
ZnC1 3
ZnC1 2
1% CdC1 2 — 99% ZnC1
1% RbC1 — 99% HgC1
1% CsC1 — 99% HgC1 2
0.3% KC1 — 99.7% HgC1 2
0.2% NaC1 — 99.8% HgC1 2
21% LIC1 — 79% NH 4 C1
114% K 2 CO — 13% Na 2 CO 3 — 73% NaOH
57% NaNO 9 — 143% Pb(N0 3 ) 2
NaNO 2
0.7% MgC1 2 — 99.3% ZnC1
Temperature
ฐF
250 1482
251
252
253
2514
255
258
14814
1486
1487
1489
1491
1496
500
502
50 14
505
507
509
511
5114
520
— 69% AgC1
— 33% Sr(NO 3 )
5% NaC1 — 89% ZnC1 2
— 61% NaOH
260
261
262
263
2614
265
266
268
271

-------
—160-
Temperature
ฐC ฐF
275 527 HgC1 2
82% LIBr — 18% LIOH
1% MnC1 2 — 99% ZnC1 2
277 531 62% KCN — 38% Cu(CN) 2
278 532 6% KC1 — 914% ZnC1 2
280 536 12% Na 2 CO 3 — 88% NaOH
282 5140 FeC1 3
9% KCN — 91% KCNO
285 5145 21% KBr — 39% CdBr—NaBr — 30% FeC1 3
288 550 31% KEr — 114% NaBr — 55% CdBr 2
290 554 314% KBr — 66% AgBr
51% LIC1 — 149% LIOH
291 556 149% KCN — 51% AgCN
295 563 14% K 2 CrO 4 — 96% KNO 3
297 567 KNO 2
86% NaNO 3 — 114% Ba(N0 ) 2
299 570 114% KNO 3 — 70% LINO 3 — 16% Ba(N0 3 ) 2
300 572 RbOH
0.14% BaCl 2 — 99.6% SbC1 3
53% KBr — 147% KOH
0.3% LIC1 — 99.7% SbCl
92% NaNO 3 — 8% Na 2 SO 4
3014 579 142% KBr — 58% CdBr 2
5% NaC1 — 95% NaNO 3
4% Na CO 3 — 96% NaNO 3
306 583 143% KC1 — 57% AgC1
50% KOH — 50% RbOH
307 5814 4% NaBr — 96% NaNO 2
308 586 NaNO 3
312 5914 76% KNO 3 — 214% Ba(N0 3 ) 2
30% LIC1 — 70% RbC1
316 601 17% NaC1 — 83% CuC1
318 6o ’4 NaOH
32% LIC1 — 2% NaC1 — 66% RbC1

-------
-161—
Temperature
ฐC ฐF
320 608 AgCN
17% KNO 2 — 83% KNO,
32% LIC1 — 2% NaC1 — 66% CsC1
323 613 BaC1O 3
NaCNS
326 619 5% K CO 3 — 95% KNO 3
329 6214 KNH 2
332 628 97% CsC1 — 3% LIC1
6% KC1 — 94% KNO 3
95% KNO 3 — 5% K 2 S0
3314 633 KNO 3
55% KBr — 45% MgBr 2
335 637 9% NaC1 — 91% NO 2
3142 648 10% LIC1 — 90% T1C1
3143 6149 10% BaC1 2 — 90% ZnBr 2
3 145 653 HgBr
17% NaCN — 83% Cu 2 (CN) 2
3148 658 148% KBr — 52% LIBr
350 662 NH C1
351 6614 146% KC1 — 5% KF — 149% LIC1
352 666 55% KC1 — 145% LIC1
3514 669 10% BaBr 2 — 90% ZnBr 2
357 675 Pb(N0 3 ) 2
360 680 29% NaC1 — 71% NaOH
361 682 12% MgC1 2 — 88% T1C1
362 6814 72% KC1 — 19% MCi — 9% NaC1
363 685 51% MCi — 149% NaC1
3614 687 56% KC1 — 144% LIC1
366 691 9% KC1 — 91% K 2 Cr 2 0 7
367 693 31% NaBr — 69% CdBr 2
368 6914 KC1O 3
3714 705 31% KC1 — 13% NaC1 — 56% CdC1 2
314% KC1 — 5% KF — 50% LiC1 — 11% NaC1

-------
—162-
Temperature
ฐC ฐF
380 716 KOH
383 721 17% KC1 — 83% CdC1 2
385 725 76% K 2 Cr 2 O 7 — 214% K 2 W 2 0 7
53% KC1 — 12% BaC1 2 — 1% CaF 2 — 314% MgC1 2
390 7314 2% CsC1 — 98% T1C1
392 738 21% NaC1 — 79% CdC1
3914 7141 ZnBr 2
396 7145 11% KC1 — 9% NaC1 — 80% MgC1 2
15% KC1 — 26% NaC1 — 16% MgC1 2 — 143% ZnC1 2
397 7146 17% KC1 — 23% NaC1 — 60% MnC1 2
1400 752 11% CaC1 2 — 89% CuC1
401 7514 Na 2 S 2 O 7
1406 763 21% KC1 — 79% PbC1 2
1407 765 22% RbC1 — 78% PbC1 2
58% KC1 — 1% NaF — 1% CaF 2 — 18% MgC1 2 — 22% MnC1 2
1408 766 10% LIC1 — 90% CuC1
1% MgC1 2 — 99% CuC1
1409 768 18% KC1 — 2 4% NaC1 — 58% NgC1 2
1410 770 13% CdC1 2 — 87% Cu 2 C1 2
11% LiC1 — 89% PbC1 2
8% NaC1 — 92% PbC1 2
17% KC1 — 22% NaC1 — 7% BaC1 2 — 514% MgC1 2
1412 7714 14% NaC1 — 96% T1C1
14114 777 K 2 S 2 0 7
1416 781 8% SrC1 2 — 92% T1C1
1119 786 3% Cad 2 — 97% T1C1
56% KC1 — 6% BaC1 2 — 1% CaF 2 — 37% MgC1 2
1422 792 CuC1
57% KC1 — 3% BaC1 — 1% CaF 2 — 39% MgC1
141% KC1 — 1 4% NaC1 — 3% BaC1 2 — 1% Cad 2 — 141% MgC1 2
112 )4 795 Cu 2 C1 2
147% NaCN — 53% AgCN

-------
-163-
Temperature
oc
1425 797
426 799
428 802
1430 806
32%
62%
52%
TiC 1
1% BaC1 2 —
141% KC1 —
21% LIF —
1414% NaC1
145% NaBr
8% NaC1 — 1% CaC1 2 — 140% MgC1
12% NaC1 — 5% BaC1 2 — 1% CaP 2 — 35% MgC1 2 —
143% NaC1 — 3% CaF 2 — 514% MgF 2
L IOH
145% BaC1 2 — 55% CdC1
15% CaC1 2 — 85% AgC1
8% MgC1 2 — 92% AgC1
53%KF_25%LIF —12%NaF—7%BaF 2 —1%CaF 2 —2%MgF
97% LIBr — 3% LIP
142% LIBr — 58% SrBr 2
514% KC1 — 13% BaC1 2 — 33% MgC1
58% KC1 — 14% NaC1 — 3% CaC1 — 35% MgC1
AgC1
75% KF — 16% LIP — 9% NaF
NaC1 — 68% MnC1 2
KC1 — 38% MgC1 2
KC1 — 148% MnC1 2
99% T1C1
59% KOH
79% LIOH
— 56% MgC1 2
— 55% MgBr
1431 808
1432
1433
810
811
51%
145%
2%
KC1 —
KC1 —
NgO
143 1J 813
1437 819
14140 8211
A gBr
148% xci —
145% LiNO 3
59% KC1 —
Ba d 2
38% KC1 —
62% KC1 —
LII
32% MgC1 2
6% NaC1 — 111% BaC1 2 —
— 55% Ba(NO 3 )
5% CaF — 36% MgC1 2
15% Bad 2 — 147% NgC1 2
38% MgC1 2
14111
14146
14148
1450
826
835
838
8112
8146
1453 8117
145)4
1 )55
8149
851

-------
-164-
55% KC1 — 14% NaC1 — 3% BaC1 2 — 14% CaC1 2 — 314% MgC1 2
59% KF — 28% LIF — 13% NaF
25% KC1 — 14% NaC1 — 3% CaP 2 — 68% MgC1 2
7% MgC1 2 — 93% PbC1 2
10% KC1 — 90% KNO 2
38% NaC1 — 5% Bad 2 — 57% MgC1
50% KC1 — 12% BaC1 2 — 38% NgC1 2
35% NaC1 — 14% BaC1 2 — 9% CaF 2 — 52% MgC1 2
8% CaC1 2 — 92% PbC1 2
9% L1C1 — 91% AgC1
9% KC1 — 25% NaC1 — 14% BaC1 2 — 62% rvlgCl 2
Cu(CN) 2
23% LIC1 — 77% SrC1 2
9% KC1 — 26% NaC1 — 65% MgC1 2
29% KC1 — 2% BaC1 2 — 2% CaF 2 — 67% MgC1 2
38% KC1 — 3% NaC1 — 3% BaC1 2 — 1% CaC1 2 — 55% MgC1
Cu 2 (CN) 2
90% Na 2 B 0 7 — 10% Ca 3 (PO ) 2
68% KF — 32% LIF
117% LIBr — 53% BaBr
87% LIC1 — 13% LiF
21% NaBr — 79% SrBr 2
69% KCN — 31% ZN(CN) 2
6% KC1 — 32% Bad 2 — 58% CaC1 2 — 14% CaF 2
37% KC1 — 12% Bad 2 — 11% CaF 2 — 147% MgC1 2
8% KBr — 11% KC1 — 81% KI
BeBr 2
5% KC1 — 95% XI
52% L1 2 WO — 148% Na 2 WO
12% NaC1 — 88% SrBr 2
110% KC1 — 140% Na 2 C0 — 20% NaOH
10% KF — 6 4% MCi — 26% NaC1
6% KC1 — 31% Bad 2 — 59% CaCi — 11% CaF 2
79% K2C0 3 — 21% LICO 3
85% C0C1 — 15% NaC1
Temperature
1457 855
1458 856
1459
1460
1166
1167
1468
1472
Li7 3
1475
1482
14814
1485
1486
1487
1488
1490
1191
1492
1493
858
860
871
873
8711
882
883
887
900
903
905
907
909
910
9114
916
918
919

-------
—165-
Temperature
ฐC ฐF
1495 923 81% Na 2 Mo0 — 19% MoO 3
1496 925 7% KAsO 3 — 93% NaAsO 3
38% LIC1 — 62% CaC1 2
1499 930 61% SrC1 2 — 39% MnC1
501 93 4 PbC1 2
25% NaC1 — 5 Bad 2 — 65% CaC1 2 — 5% CaF 2
502 936 147% KCN — 53% NaCN
81% KI — 19% NaC1
16% KBr — 5% KC1 — 79% KI
14% KC1 — 6% NaC1 — 90% MnC1 2
503 937 37% SrC1 2 — 63% CdC1 2
33% KC1 — 114% KI — 53% NaT
5014 939 CuBr
119% BaC1 2 — 51% MnC1 2
5% KC1 — 28% NaC1 — 67% CaC1 2
23% KC1 — 22% NaC1 — 55% SrC1
505 9141 33% NaC1 — 67% CaC1 2
148% Na 2 S — 52% Cu 2 S
9% KC1 — 5% NaC1 — 71% KF — 15% LIF
508 9146 32% LiF — 68% NaF
510 950 27% LIC1 — 73% Bad 2
512 9514 16% KC1 — 814% KI
52% LiBr — 148% MgBr 2
513 955 (NH ) 2 sO
114% KC1 — 614% KI — 22% NaC1
5114 957 26% NaBr — 714% CaBr 2
10% KBr — 18% KC1 0 72% Nal
30% KC1 — 19% KI — 14% NaBr — 147% Nal
26% KC1 — 214% KI — 8% NaBr — 42% Nal
515 960 2% KC1 — 81% KI — 17% NaC1
69% KI — 19% NaC1 — 12% NaT
516 961 6% KBr — 11% KC1 — 32% KI — 8% NaC1 — 143% Nal
517 962 14% KC1 0 27% NaCT — 69% Cad 2
521 970 23% NaC1 — 5% Bad 2 — 68% Cad 2 — 14% CaF 2

-------
—166-
Temperature
OC
522 972
523
52 ) 4
525
528
530
89% LIBr — 11% LIC1
3% CaF — 36% NgC1 2 — 61% SrC1 2
20% K 2 Cr0 — 14% KF — 76% Nal
56%KBr—8%NaBr—36%NaI
5% KC1 — 57% BaC1 2 — 38% CaC1 2
18% KC1 — 82% Nal
81% KI — 19% NaC1
6% KC1 — 314% KI — 9% NaC1 — 51% Nal
77% L1Br — 23% NaBr
72% KC1 — 19% LIC1 — 9% NaC1
53% KC1 — 5% KF — 12% NaC1 — 30% CaC1 2
147% KC1 — 20% LIC1 — 2% LiF — 31% NaC1 (Dow 1 45—B)
147% KC1 — 23% LIC1 — 26% NaC1 — 14% NaF (Dow 145_C)
61% KC1 — 10% LIC1 — 3% BaC1 — 214% Cad 2 — 2% CaF 2
140% L1 2 C0 3 — 60% L1SO
9% KBr — 18% KC1 — 8% KI — 65% Nal
9714
976
977
982
986
988
990
992
993
997
531
532
533
53)4
535
140%
KC1 —
214% LiC1 — 31% NaC1
— 5% CaF
31%
KOl —
7% BaC1 — 214% LIC1 3
— 31% NaC1
—
7% CaCh
140%
KC1 —
18% LiC1 — 3% LIF —
32% NaC1 —
7%
CaC1 2
140%
KC1 —
2 4% LiC1 — 2 4% NaC1
— 5% NaF —
7%
CaC1 2
66%
XI —
25% NaBr — 9% NaC1
37%
KBr —
5% NaC1 — 58% NaT
32%
KBr —
68% SrBr 2
61%
K 2 S0
— 39% CoSO
28%
K 2 S0
— 72% L1 2 SO
37%
KEr —
6% KC1 — 57% Nal

-------
—167-
APPENDIX C
THERMAL DECOMPOSITION OF PVC DILUTION TUBE
Tests were run to determine if thermal decomposition of the
polyvinyl chloride (PVC) dilution tube was contributing in any
way to the halide analysis of particulate emissions collected
in the samplers. Shavings were taken from samples of the tube.
These were gradually heated (about 10ฐC per 7 minutes) to 250ฐC.
Decomposition of the material was measured by mass spectroscopic
monitoring of the peaks at m/e = 36 and 38 which would indicate
the presence of HC1 . As can be seen in the following graph,
decomposition did not occur at temperatures below 200ฐC.
A second test was made to determine the effect of prolonged
heating at lower temperatures. PVC shavings were heated at
150ฐC for 100 minutes. No significant evolution of HC1 was
detected.
Since the dilution tube cannot be practically used at temperatures
over 50ฐC because of excessive softening and loss of rigidity, it
could not impart any significant quantity of halides to the
particulate emissions analyzed.

-------
-168-
•
2_.
- -
—4 - -
- 1
:t
— —
-
:1
: -
----—
—- -
- -
t --
;
L -•I t E
-J - :t
4

i_
Eฑi : ::

- -
1::
—
_
— r4
------—

— I
:::
,—.4
.
t
.--

-r
.:
- .
:
-
:
-
I
H.
1 :
f
-H
fl’
- — — L
1 i 1
— + — —
- —-k- T
------h- - - -
1 -
I
:..
-
t I t 4
•i- - - -1 -4-ฑ t’ - --Li H --. I-
E : Er:, 11 :i;:
- - -L - — L. - -r -J —. i
1T-f 4
— — — — 1 — — — • — —-— — — — —1— — — — -4.-
--- — - - h ——
--- -- --- — - - - - --4-h- I
- - - - - - -4- - - — — - -
- - 41 —4 - - — —
I
— ---r- ’
19
t-4
__
3.-
2...
- -
:. • -iL-
j.
-,
-i -i
:L
4 -
•
H
H L L
4 -t :
-Hf-ฑ --
Z:
Th-
-
- -

+i
-
:
-.
- - -H
_____
-.-
-4-
—
I
ttj .Jit
-1 - -
— I
4L
I
::
I:
- -
:.T:
R
7
4—.-—
5—
•-t ;.t --
--i i i I
-
I-I
--4 -- - - -
- 4—.—-— I—
TI
H-.-
I—
150
170 /
Temperature ฐC

-------