EPA-650/3-74-004-a
June 1974
Ecological Research Series
This document has not been
submitted to NTIS, therefore it
should be retained.
3J
Ui
CD
W^^^^^^^^^^^^:
-------
EPA-650/3-74-004-Q
STUDY OF FACTORS
AFFECTING REACTIONS
IN ENVIRONMENTAL CHAMBERS
FINAL REPORT ON PHASE II
by
R. J. Jaffe, F. C. Smith, Jr., and K. W. Last
Lockheed Missiles and Space Company, Inc.
Sunnyvale, California 94088
Contract No. 68-02-0287
Project No. 21AKC-34
Program Element No. 1AA008
EPA Project Officer: B . Dimitriades
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
COORDINATING RESEARCH COUNCIL INC.
30 ROCKEFELLER PLAZA
NEW YORK, N. Y. 10020
and
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
April 1974
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
Phase I, Study of Factors Affecting Reactions in Environmental
Chambers, was issued as EPA-R3-72-016, under Contract No. 68-02-0038.
-------
LMSC-D401598
SUMMARY
An experimental study has been conducted of effects of materials, spectrum, surface/
volume ratio (S/V) and cleaning technique on the photochemical reactions observed
in a smog chamber. A unique chamber and lighting system was used, which permitted
independent variation in chamber materials and in light conditions. A xenon arc lamp
parabolic reflector combination provided a collimated light beam. By orienting plates
of materials parallel to the beam, it has been possible to independently vary light
conditions and materials.
The study included four materials — aluminum, Pyrex, Teflon, and stainless steel, and
two conditions each of spectrum, S/V, and cleaning. A complete factorial testing
sequence was performed. All photochemical runs were at k, of 0.3 min as deter-
mined by frequent NO0 in N0 photolysis tests. The propylene (3 ppm)/NO (1.5 ppm)
z z x
reaction system was used, at 95° F and 25% relative humidity. Initial NO0 content was
^j
nominally 10% of NO . Chamber background was < 0.1 ppm C.
X
Effects of the different materials and of the two levels of each parameter have been
determined. The time to NO2 maximum is shortest for stainless steel followed by
aluminum, Pyrex and Teflon, in order. Maximum ozone concentration increases in
the order: stainless steel, Pyrex, aluminum, Teflon. Stainless steel behaves in a
manner unlike the other three materials.
The cutoff spectrum (little energy below 340 nm wavelength) strikingly lowers reaction
rates compared to the full spectrum. Surf ace/volume ratio measurably affects the
reactions. The variations in the two cleaning techniques do not affect as many of the
run characteristics. The presence of this large spectral effect (at constant k,) was
not anticipated, and cannot be explained in a simple manner.
111
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
PROJECT PERSONNEL
Name
Raphael J. Jaffe
Frank C. Smith, Jr.
Ken W. Last
E. H. Kawasaki
R. C. Tuttle
Dr. H. S. Johnston, Consultant
Area of Contribution
Project Direction
Analytical Chemistry/Chamber Operation
Statistical Analysis
Analytical Chemistry
Analytical Chemistry
Photochemistry
PROJECT MONITORSHIP
COORDINATING RESEARCH COUNCIL
PROJECT CAPA 1-69
Member
Mr. D. B. Wimmer, Chairman
Mr. Frank Bonamassa
Mr. Basil Dimitriades
Dr. J. M. Heuss
Mr. Stanley Kopezynski*
Dr. Hiromi Niki
Dr. E. E. Wigg
Affiliation
Phillips Petroleum Company
California Air Resources Board
Environmental Protection Agency
General Motors Corporation
Environmental Protection Agency
Ford Motor Company
Esso Research & Engineering Company
*Until January 1973
IV
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
CONTENTS
Section Page
ILLUSTRATIONS vii
1 INTRODUCTION 1-1
2 EXPERIMENTAL METHOD 2-1
2.1 Test Conditions 2-1
2.2 Apparatus 2-2
2.2.1 Irradiation Chamber 2-2
2.2.2 Illuminator 2-8
2.2.3 Thermal Enclosure 2-9
2.2.4 Material and S/V Changes 2-11
2. 2.5 Spectral Distribution and Spectral Changes 2-13
2.2.6 Cleaning Technique 2-15
2.2.7 Chamber Charging Technique 2-17
2.3 Chemical Analysis Methodology 2-17
2.3.1 NO0-NO 2-17
^ X
2.3.2 Ozone 2-19
2.3.3 Total Hydrocarbons 2-21
2.3.4 Propylene 2-22
2.3.5 Acetaldehyde 2-22
2.3.6 Peroxyacetyl Nitrate (PAN) 2-22
2.3.7 Moisture (Water) 2-23
3 RESULTS 3-1
3.1 Material Differences 3-10
3.2 Effect of Factors 3-10
3.3 Ozone Decay Results 3-11
4 DISCUSSION 4-1
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Section Page
5 FUTURE WORK 5-1
5.1 Ongoing Work 5-1
5.2 Recommended Further Work 5-1
REFERENCES R-l
VI
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
ILLUSTRATIONS
Figure
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
3-1
3-2
3-3
3-4
4-1
Table
1-1
2-1
3-1
3-2
3-3
4-1
Smog Chamber Assembly
Stand Assembly
Environmental Chamber Showing Side Stream Mixer,
Charge Ports, and Clean-Out Port
Chamber Inside Thermal Enclosure
Arrangement of Surface Plates
Measured Spectral Irradiance Inside LMSC Smog Chamber -
Full and Cut Spectra
Typical Raw Data for Determining k,
Smog Chamber During Vacuum Off-gassing Cleaning
Typical Linearity Check of NO Instrument
Composite Photochemical Test Results for Teflon Film Surfaces
Composite Photochemical Test Results for Pyrex Surfaces
Composite Photochemical Test Results for Aluminum Surfaces
Composite Photochemical Test Results for Stainless Steel
Surfaces
Distribution of NO0 Photodisintegrations for Various Spectra
Lt
TABLES
Characteristics of Chambers Used for Previous Intercomparison
Investigations
Chamber Description
Photochemical Test Calculated Parameter Definitions
Effects by Material
Ozone Half-Life Study
"Mylar/Teflon" Spectral Effect
Page
2-4
2-6
2-7
2-10
2-12
2-14
2-15
2-16
2-20
3-5
3-5
3-5
3-5
4-3
Page
1-2
2-8
3-2
3-6
3-11
4-2
VII
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
Section 1
INTRODUCTION
Chambers in which systematic studies can be made of the reactions between hydro-
carbons and nitrogen oxide in the presence of simulated sunlight have been in use for
some twenty years. Such chambers have generally been successful in simulating the
gross features of photochemical smog, such as production of oxidants and eye irritants,
and haze. Intercomparison of the results obtained in these smog chambers has not
been extensively attempted until relatively recently (Ref. 1), at which time the interest
in individual hydrocarbon reactivity measurements led to an understanding of the need
to compare the various facilities. Intensive comparisons of results obtained in eleven
smog chambers have been performed by the Coordinating Research Council project
CAPI-6, Techniques for Irradiation Chamber Studies, and CAPA 1-69 (Factors Affect-
ing Reactions in Environmental Chambers). The range of physical characteristics of
these chambers is shown in Table 1-1. A group of round-robin tests was conducted
using these chambers as follows: (1) irradiation of seven different hydrocarbons with
nitrogen oxide; (2) replicate runs to establish reproducibility using the propylene-
nitrogen oxide system; (3) a reactant concentration study in which 3. 0 ppm propylene
was reacted with 3.0, 1.5, and 0. 5 ppm nitrogen oxides (Ref. 2).
The differences observed among the chambers could not be accounted for analytically,
and an experimental study of how various design and operational variables affect the
photochemical reactions observed in smog chambers was instituted. This is a report
of the results of this study, which are also the first reported results for a smog
chamber illuminated by a xenon arc lamp. The facility was developed and initial tests
performed under Phase I of the project, which has been previously reported (Ref. 3).
Additional experimental and analytical work is in progress (see Section 5.1) and will
be reported upon shortly.
1-1
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Table 1-1
CHARACTERISTICS OF CHAMBERS USED FOR PREVIOUS
INTERCOMPARISON INVESTIGATIONS
Volume (ft )
Surface/Volume Ratio (ft )
Surface Type as S/V
Stainless Steel
Aluminum
Glass
Plastic Film
Light Intensity
(kd, min"1)
Type of Lighting
2.9 to 1140
0.78 to 4.91
0 to 2.44
0 to 0.92
0 to 2.81
0 to 0.83
0.16 to 0.40
Fluorescent lamp combinations, of
sunlamps, black lamps, and blue
lamps (both internal and external)
1-2
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Section 2
EXPERIMENTAL METHOD
2.1 TEST CONDITIONS
The following combinations of factors have been studied in a full factorial test plan:
Materials S/V Spectrum Cleaning
Aluminum 2.7ft Full Vacuum Off-gas
Pyrex 1.3ft""1 Cutoff Purge at 110° F
Teflon
Stainless Steel
The testing sequence consisted of performing the photochemical tests for the aluminum
surfaces, followed by Pyrex, Teflon, stainless steel, and a retest of the aluminum.
This has allowed an analysis for time trend, to see if a systematic drift was present
in the experiment. A number of replicate tests were performed. These were dis-
tributed among immediate replicates, replicates with 1 to 15 intervening runs, and
the aluminum re-test replicates, which had greater than 50 intervening runs.
The propylene/NO system was used for all tests, at 3. 0 ppm propylene and 1. 5 ppm
X
NO . The initial NO content was nominally held at 10 percent. The chamber was
x &
held at 95 ± 3° F throughout all tests. Relative humidity was 25 ± 5 percent (49 to
59° F dew point). Chamber pressure was slightly above atmospheric (0.1 in. HO).
^
Zero air was used to maintain chamber pressure, to make up for sampling and leakage,
at about 3 percent/hour make-up rate.
2-1
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
2.2 APPARATUS
The apparatus used was specifically designed to meet the objectives of the study by
allowing independent variation in materials configuration, lighting conditions, and
cleaning technique. The illuminator produces a collimated light beam. Plates of
materials can be placed parallel to the beam axis without affecting the light conditions.
This decoupling of light and materials allows independent variation of the two factors.
Major apparatus items are the smog chamber, the xenon arc lamp illuminator, the
thermal enclosure for the chamber, and the gas analysis instrumentation.
2.2.1 Irradiation Chamber
The irradiation chamber is hexagonal in cross section, measuring 54 in. across the
diagonal of the cross section. The chamber configuration is shown in Fig. 2-1 and the
chamber support stand is shown in Fig. 2-2. Figure 2-3 represents a pictorial view
of the chamber. The chamber is constructed of six flat side panels that fit into an
extruded aluminum framework. The aluminum framework is coated with 5 to 8 mils
of FEP Teflon. The resultant panels are bolted together and are supported from cir-
cular rings on the stand.
The faces of the chamber are of tempered 1/4 in. Pyrex glass to admit the light and
pass the beam through the chamber with minimal reflection or absorption. The flat
sides of the chamber are also fabricated from 1/4 in. tempered Pyrex glass. The
working stress of the tempered Pyrex is 3600 psi, which results in an allowable pressure
differential of 7.1 in. of water. This is based on the hexagonal faces, which are the
weakest members. Sealing of the 1/4 in. Pyrex to the aluminum extrusion is accom-
_c
plished with silicon rubber gaskets that have been off-gassed at about 10 torr for
over 24 hours. The gasket is fitted between the aluminum channel and the Pyrex, and
the Pyrex is pressed into a channel section in the extrusion. A silicon rubber O-ring
is used to seal adjacent aluminum extrusions which are bolted together. Additional
sealing is accomplished with a coat of Dow Corning 30-121 RTV silicon rubber on all
external joints.
2-2
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Three 1-in. diameter ports are located in one of the lower side panels. The ports are
fabricated from brass and are coated with FEP Teflon. Sealing of these ports to the
Pyrex chamber walls is accomplished with silicon rubber gaskets. Sealing of tubes
inserted into the ports is accomplished with Teflon seals. Tubes can be fitted into
these ports and adjusted to various distances into the hexagon. The rear port is used
for a gas thermocouple, which is Teflon coated and shielded from direct illuminator
radiation. The front port has a 1-in. diameter Pyrex tube installed that is used for
venting the chamber.
The center port is used for gas sampling. A 1-in. diameter Pyrex tube containing a
concentric 7-mm tube is used. The sample enters at about the geometric center of
the chamber, and is drawn to a sample manifold through a 1/4-in. Teflon tube, and
thence to the analysis instrumentation. The sample contacts only Pyrex or Teflon.
before entering the instruments. Delay time from sample withdrawal to instrument
inlet is about 1-1/2 min.
A cleanout port is located in this same Pyrex panel. This port is 3-in. diameter and
is used for chamber outgassing during vacuum off-gas cleaning. A Pyrex disc with a
silicon rubber gasket is used to seal this port during normal operation. Metallic parts
are all coated with FEP Teflon. A Pyrex relief valve is also mounted on this panel.
This relief valve has a 2-in. diameter opening leading to the chamber. The valve
utilizes a water seal principle and the relief setting is adjusted by the addition or
withdrawal of water. Both negative and positive differential pressure protection are
provided. An overflow trap is included to prevent water from entering the chamber.
A side stream mixer is also located on this panel. This mixer consists of a 6-in.
diameter duct connected to the panel at a point 1/3 of the way from the front and 1/3
of the way from the rear of the chamber. Gas is circulated through the duct by means
of a Teflon-coated fan blade installed in the duct. The blade is connected to a shaft
that penetrates the duct through a rotary Teflon seal. The shaft is connected to a
motor via a pulley and belt which are located outside the chamber. The fan is rotated
2-3
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
UNCLASSIFIED
LMSC-D401598
,*?£/££ •sC'ftf
A
#£
-------
UNCLASSIFIED
LMSC-D401598
Jl
. eo- 7^? -
/OOOTX*—-j i— -T-fQ&tee, S/*ZK&$
A1
A-
o
l^1
O
o
^
**=A
&-
Fig. 2-1 Smog Chamber Assembly (Cont.)
UNCLASSIFIED
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
UNCLASSIFIED
LMSC-D401598
UNCLASSIFIED
0)
w
0!)
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Fig. 2-3 Environmental Chamber Showing Side Stream Mixer,
Charge Ports, and Clean-Out Port
2-7
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
at 250 rpm, producing a circulation rate of approximately 75 cfm. The inlet end of
the duct enters the chamber at a 45-deg angle to minimize short circuiting of the gas
circulation. All parts of the side stream recirculator are coated with Teflon. The
mixer is operated during charging of the chamber only. It is shut off during photo-
chemical tests.
The chamber is mounted at each hexagonal apex to a circular structural framework,
which in turn mounts to a dolly with casters for easy transport of the chamber. The
structural frame is configured to allow chamber assembly within the framework. A
locating jack system is used to adjust the height of the chamber to exactly match the
height of the light source.
Geometric characteristics of the chamber are summarized in Table 2-1.
Table 2-1
CHAMBER DESCRIPTION
Configuration Hexagonal Prism
Length 60 in.
Diagonal 54 in.
Volume 65.9ft3
Surface Area
End Plates 26.4ft2
Side Plates 67.5ft2
Total 93.9ft2
S/V 1.43ft""1
2.2.2 Illuminator
The irradiation source for the chamber is external to the chamber. This external
source consists of an arc lamp situated in front of a large parabolic reflector.
2-8
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Collimated energy from the reflector is directed toward the chamber. The collimated
beam used as the light source for the irradiation chamber is 5 ft in diameter. The
light source for the illumination system is an air cooled Osram compact xenon arc
lamp (6,500 watt nominal rating). The fireball for the lamp is 2.4 by 9 mm. The
optical system consists of a 5-ft diameter parabolic primary mirror and a spherical
back-collector mirror. The primary mirror is made of copper with a rhodium plating
with vacuum deposited aluminum over the rhodium. An SiO coating is used to protect
the aluminum. Reflectance is approximately 0.9 at a wavelength of 300 nm. The
spherical back reflector is similarly coated. The back reflector is utilized to capture
energy from the lamp that would normally not strike the parabola, and focus it back on
the parabolic reflector. The optical system collects about 35 percent of the lamp
radiated light and directs it as a collimated beam into the chamber front face. The
o
lampholder casts a shadow about 1.5 ft in area, which obscures about 8 percent of
the beam. This shadowed area has no appreciable effect on the experimental data.
For the cut spectrum configuration, a plane reflector is mounted behind the chamber
rear face. This reflector provides a second pass of the light through the chamber. It
is a front-surfaced, aluminized SiO -coated reflector, to maintain reflection down to
A.
300 nm wavelength.
The illuminator is integrated into a searchlight housing. The housing is on casters,
which allows the illuminator to be moved readily.
2.2.3 Thermal Enclosure
A thermal enclosure is used to control the chamber temperature during a photochemical
run. Design requirements for the thermal enclosure are maintenance of chamber gas
temperature at 95 ± 3°F. The enclosure, depicted pictorally in Fig. 2-4 with the
chamber installed, consists of a plywood housing insulated with fiberglass, with heated
air circulating throughout the enclosure. The enclosure has a hexagonal-shaped hole
in the rear to allow the chamber to protrude. The front end of the thermal enclosure
2-9
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Fig. 2-4 Chamber Inside Thermal Enclosure
2-10
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
has sliding doors that allow easy removal and installation of the chamber. These doors
also have a hexagonal-shaped hole that will allow the forward end of the irradiation
chamber to protrude. This allows the illuminator energy to enter the chamber.
Both the rear and forward hexagonal openings in the enclosure fit tightly to the chamber
to minimize gas leakage from the enclosure. The interior surfaces of the enclosure
are insulated with 3 in. of fiberglass.
A centrifugal blower is utilized to circulate approximately 1500 cfm of air over four
I. 3-kw heaters that are used for thermal control. Three of the heaters are manually
switched on or off and the fourth heater is thermostatically controlled to maintain the
enclosure air temperature at a fixed level. This system allows use of the fixed heaters
for purge cleaning, or warm-up, and coarse temperature control, with fine temperature
control being accomplished with the thermostatically controlled heater. The thermal
sensor is a West resistance bulb controller. This controls the air temperature in the
thermal enclosure. Air circulation is accomplished with a centrifugal blower that
passes air through plenums located at the bottom of the thermal enclosure. The heaters
are located in these plenums. Air passes out of the plenums, over the chamber and
is withdrawn out of the top of the thermal enclosure, where it is recirculated to the
heater plenum. The thermal enclosure is approximately 8 ft high by 8 ft wide by 6 ft
long.
2. 2.4 Material and S/V Changes
Plates of the material under study are inserted into the chamber, oriented parallel
to the light-beam axis of collimation, and vertically. These plates vary from 5-3/8
to 22-3/8 in. in width and are the height of the chamber. Plates are arranged in six
sets, with each set running the length of the chamber. Frequent open spaces interrupt
each set of plates, and form openings that are the height of the chamber and 2 to
4-1/2 in. wide, to allow the sampling and vent tubes to penetrate into the chamber,
and to promote mixing of the entire chamber contents. Figure 2-5 gives the plate
layout. Either six or three sets of plates are used, which gives either 2.7 or 1.3 ft
2-11
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
c
0
.21
"c
o
o
o
•a
OJ
o
E
a>
O
ccf
(/>
"S
CO
a>
|ro
o_
CO
aj
cc
0)
OJ
bJD
(Nl
be
2-12
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
of the material under study. The metallic materials are polished to a mirror finish.
The materials details are aluminum — 1100 H 14 alloy, stainless steel — type 304,
Pyrex-Corning 7740 plates, and Teflon- 5 mil FEP film.
2.2.5 Spectral Distribution and Spectral Changes
The spectral interval of interest in atmospheric photochemical simulation is 300 to
about 400 nm. The lower wavelength is the natural cutoff provided by the earth's
ozone layer, and the upper wavelength is set by the energy required to dissociate
NO9. It is generally recognized that compact xenon arc lamps provide the best
^
available match for this UV region.
Two spectral distributions were used in these experiments. The full spectrum con-
figuration is shown in Fig. 2-6. The second distribution is the cutoff spectrum, and
is also shown in the figure. This is obtained by placing a sheet of 3/16 in. thick
Plexiglas between the light source and the smog chamber. To avoid aging effects, a
fresh sheet of Plexiglas is used after about each fifth test. Total light intensity is
restored to the same value as used for the full spectrum runs by providing a second
pass of the light through the chamber. The front surfaced aluminized reflector is
used for this purpose.
Spectral data were taken with an Optronics Laboratory spectroradiometer. The radi-
ometer consists of a calibrated photovoltaic cell, a grating prism spectrometer, and
-9 -2 2
blocking filters. Digital readout is provided over a range of 10 to 10 watt/cm -nm.
Bandpass is 5 nm. The unit is calibrated against an NBS standard quartz iodide lamp,
over the wavelength interval of 250 to 1100 nm. Note that the characteristic xenon arc
lamp peak at 467 nm is easily seen by the radiometer. These data are in agreement
with earlier data taken using a photomultiplier tube spectroradiometer unit.
Measurements of the light beam were made at the plane corresponding to the front
window of the chamber, and of the beam at the exit of the chamber, after passing
2-13
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
150 400
WAVELENGTH (nm)
Fig. 2-6 Measured Spectral Irradiance Inside LMSC Smog Chamber —
Full and Cut Spectra
through the front and rear 1/4-in. Pyrex window. To measure spectral irradiance
inside the chamber, the rear window is removed, and the spectroradiometer entrance
slit positioned somewhat inside the chamber. These are the values shown in Fig. 2-6.
Total light intensity has been maintained at k, of 0.3 min~ throughout the test series.*
To measure k, the smog chamber is filled with pure Ng (< 100 ppm Cv, < O.I ppm HC)
and about 1.5 ppm NO0 added. The lamp is turned on, allowed to stabilize, and the
/i
chamber illuminated for three or four successive one-minute intervals. The data are
plotted on semi-log paper and usually show the expected upward deviation from linearity
after the third one-minute interval. Two or three NO instruments are used for each
k determination, and usually agree within 10 percent. Measurements of kd are per-
formed each time a new S/V configuration is established. Figure 2-7 shows a typical
plot.
*Lamp power was adjusted to maintain kj at a constant value throughout the study. A
variation of perhaps ±10 percent occurred due to uncompensated aging effects of the lamp,
reflector, and filter.
2-14
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
o
z
Fig. 2-7 Typical Raw Data for Determining k,
2.2.6 Cleaning Technique
The two cleaning techniques used are: purging at 110° F, or vacuum off-gassing. Purge
cleaning consists of holding the chamber overnight at a temperature of 110 ± 3° F while
4 to 6 chamber volumes of pure air are purged through the chamber. For vacuum off-
gassing, the smog chamber is moved to an adjoining 18 by 18 by 36-ft vacuum chamber.
The vacuum chamber is pumped by Roots Blowers and mechanical pumps which prevents
any back migration of pump oil. The chamber walls are 304 stainless steel, polished
to a No. 4 mill finish. The smog chamber is held at about 2 microns pressure (or less)
for at least 16 hours. During this time, the chamber is maintained at about 100° F.
After the off-gassing, the smog chamber is repressurized by bleeding charge gas into
the smog chamber while repressurizing the vacuum chamber. Figure 2-8 shows the
smog chamber inside the vacuum chamber.
2-15
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Fig. 2-8 Smog Chamber During Vacuum Off-Gassing Cleaning
2-16
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
2.2.7 Chamber Charging Technique
The chamber is charged with synthetic compressed air. The air is purchased from a
single vendor to a specification of < 0.1 ppm total hydrocarbons (as methane), and
< 10 ppm hydrogen. The air is produced by combining nitrogen gasified from liquid
nitrogen and electrolytic oxygen. The oxygen is made by electrolysis of distilled water.
It then is further purified by catalytic combustion of the trace hydrogen, followed by
molecular-sieve drying of the oxygen. A pre-charge determination of total hydro-
carbons verifies the air purity. The chamber is charged through a stainless steel
humidifier, packed with Rashig rings and filled with triple distilled water. Starting
dew point is adjusted to 54 ± 5° F by dilution with the dry pure air.
Reactants are added to the chamber from stock cylinders of about 350 ppm in nitrogen.
The reactant blends are transferred using separate 1/2 liter sampling cylinders. The
transfer cylinder is pressurized at about 200 psi, placed in the charge manifold, and
slowly bled to the smog chamber. In this technique, the transfer cylinder pressure
(which is read to 0. 5 psi) becomes the reproducibility limit for the initial reactant
charge.
2.3 CHEMICAL ANALYSIS METHODOLOGY
2.3.1 NO0-NO
£ X
Nitrogen dioxide is monitored by the modified Saltzman-Lyskow wet chemical technique,
utilizing a continuous sampling Technicon Autoanalyzer unit. The NO0 absorbing
&
solution is made from 2. 0-gm N-1-naphthylethylenediamine dihydrochloride, 100-gm
sulfanilic acid, 5-cc Kodak Photoflo, and 1 liter glacial acetic acid in 5 gallons of
water. The lifetime of the solution is greater than 1 month when stored in an aluminum
foil covered bottle and shielded from air exposure. The gas sampling system consists
of two 15-cm by 2.4-mm I.D. (28 turn) glass mixing coils in series where the gas
sample stream contacts the absorbing solution, an accumulator/liquid-gas separator,
2-17
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
a flowmeter to measure sampling rate, a chain driven peristaltic pump using variable
flow fluoroelastomer tubes, a colorimeter with a 50-mm flow cell, and an extended
range recorder. Gas sampling rate is 150 cc/min and liquid sampling rate is 1.5 cc/min.
When new absorbing solution is prepared, a static calibration using NaNO0 solutions
-3 -1
ranging from 1. 5 x 10 to 1. 5 x 10 jul NO2/ml is performed. NO2 gas concentra-
tions are determined from the formula
pphm N02 = /gjmx 10°
where:
A - microliters NO2 gas per milliliter of liquid standards
(mg/liter NaNO2) (24. 5 liters/mole)
(Mol wt. NaNO2) 0. 72 moles NaNO2/mole NO2
where 24. 5 is the molar volume at 25° C and 760 torr and 0. 72
is the Saltzman factor
B = flow rate of absorbing solution reagent (ml/min)
C = gas stream flow rate (liters/min)
D = column efficiency (expressed decimally)
Daily dynamic calibrations are performed using standard Metronics 4-cm constant
rate NO2 permeation tube. The permeation tube is placed inside a constant temperature
condenser. Low NO? concentrations are obtained by passing compressed air through a
calibrated flow meter and then sweeping the NO2 from the permeation tube into a 12. 5
liter dilution flask. The NO? stream enters the bottom of the flask and sampling is done
at the top using a single 4-way path to the gas stream. The dilution flask pressure is
measured with a -0. 5 to +0. 5 in. water Magnihelic gage and is adjusted with a vent line
constriction to get pressure resembling chamber run conditions (+0. 02 in. H2O). The
dilution system is all Teflon and glass except for small Tygon connections.
2-18
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Permeation tube NO2 concentrations are determined from the formula
(L) (K) (P )
C = ;; — X 100
where:
C = gas concentration (pphm)
L = length of permeation tube (cm)
K = molar volume/molecular weight = 0. 532 for NO0 at 25° C and 760 torr
&
(liter/gm)
P, = permeation rate of NO9 at temperature t (ng/min per cm of tube length)
3
F = carrier gas flow rate past tube (cm /min)
Nitric oxide is continuously measured with a Thermo Electron Corporation Model 12A
chemiluminescent gas monitor and Honeywell 18 recorder. This instrument measures
the chemiluminescent reaction of NO and (X. Gas flow to the instrument is 150 cc/min.
The NO mode of the instrument uses a stainless steel converter run at 800° C to reduce
X
NO2 to measurable NO. Converter efficiency has been established at 99+ percent. In-
strument zero, full scale and photomultiplier tube dark current are checked prior to
turning on the O~ generator. The instrument is calibrated daily by the dynamic NO2
dilution gases and by a stock 88 ppm NO in N« standard gas. Linearity of the instrument
is periodically confirmed by the exponential dilution technique. Precision is within 5
percent. Figure 2-9 shows an exponential dilution linearity check.
2.3.2 Ozone
Ozone is measured directly and continuously by a McMillan Model 1100 Ozone Meter.
This instrument measures the chemiluminescent reaction of ozone and ethylene. The
meter has four scales, with 0 to 2 ppm scale most commonly used. Daily meter cali-
bration is done at 1 ppm before and after a photochemical run by using a McMillan 1000
ozone generator. The output of the generator is regulated by sliding cover for the UV
lamp and output at 1.00 ppm (999 dial setting) is checked with a null meter. Periodic
2-19
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
1UU
80
60
40
20
10
1 8
£\ £
^— 0
1 4
i —
UJ
z 2
o
0
0
^ 1
0.8
0.6
0.4
0.2
n
;
-
.
-
-
-
\
\
SCALE C
i
\
V
\
-
:HANGE
- SCALE 1
:HANGE
0 5 10 15 20
TIMEtmin.)
Fig. 2-9 Typical Linearity Check of NO Instrument
25
2-20
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
calibrations of the generator and meter are done with the neutral buffered KI technique.
The 2-percent KI absorbing solution is calibrated with a stock I2 solution titrated
against standard Na S O . The absorbance of the resulting KI/I0 solution is measured
Z ^ o ^
on a Perkin-Elmer 202 spectrometer at 350 nm. Precision is about 4 percent. In the
first 20 months of usage of the ozone generator, the calibrations against neutral KI
showed no significant drift (<5%). The null meter-front panel adjustment potentiometer
technique, which is used to maintain the ozone generating lamp light output, was
evidently effective. At that time the generator malfunctioned and was returned for
repair and recalibration. The good stability observed may also reflect the constant
air flow system and the air pre-treatment. Room air is filtered through an MSA Type
N canister and a pre-ozonation UV lamp. This lamp produces ozone which converts
trace NO to NO . The air then goes through a molecular sieve 13X filter which destroys
&
the ozone. This scheme avoids loss of part of the calibrated ozone output by the
NO-OQ reaction.
o
Ozone concentration is corrected by checking 1. 00 ppm before and after a run. If a
discrepancy is noted, the initial and final readings at 1 ppm from the generator are
plotted linearly against time. Corrections are read from the line and applied to the
readings. Corrections are applied if the meter is more than 5 percent different from
the generator. The largest drift noted was about 18 percent.
2.3.3 Total Hydrocarbons
Chamber total hydrocarbons as CH. are monitored with a F&M Model 700 gas chroma-
tograph with a Model 810 electrometer using an O0-H flame ionization detector and
A Lt
an unpacked 1/8 in. O.D. stainless steel column. Calibration is done from a zero air
cylinder with a THC concentration of 0. 07 ± 0. 03 CH.. The cylinder concentration
was determined at the supplier and checked upon delivery. Chamber samples from the
glass manifold are drawn through a 1/8 in. O.D. Teflon line and then through a sliding
sampling valve by a small vacuum pump.
2-21
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
2.3.4 Propylene
Propylene is measured using the same detector and valve used for total hydrocarbons.
A Carle Microvolume Switching Valve is used for analysis selection. The column for
propylene is 0.6 ft x 1/8 in. 100/120 mesh Porapak S stainless steel. The oven
temperature is 105° C; the detector is set at 150°; and the injection port is set at 125° .
The sampling loop has a 5. 0 cc volume, which gives an LLD of 0. 03 ppm.
Calibration was done using a stock cylinder. The concentration and linearity from 0. 05
to 3. 00 ppm of the analysis were confirmed. Precision at 3. 00ppm is less than 5 percent.
2.3.5 Acetaldehyde
Acetaldehyde uses the other detector of Model 700 gas chromatograph and a separate
Model 810 electrometer. The column is a 20-percent FFAP Chromosorb W (DMSC
treated), 60/80 mesh, 20 ft x 1/8 in. stainless steel column. The 10. 0 cc sampling
loop and valve use the same Teflon line from the manifold and vacuum pump as do the
propylene and total hydrocarbon analyses. Oven, detector, and injection port temper-
atures are the same as for propylene. Calibration is done with a stock cylinder and
linearity has been established from 0. 08 to 1.50 ppm, with an LLD of 0. 03 ppm and
precision less than 10 percent at 1. 50 ppm. When the column is too noisy to give
useful data, the Porapak S column used for propylene can be used for acetaldehyde
analysis, although the acetaldehyde peak is broad and LLD is 0. 07 ppm. A comparison
of acetaldehyde concentration indicated by both columns shows quite good agreement.
2.3.6 Peroxyacetyl Nitrate (PAN)
Peroxyacetyl nitrate is measured on a Varian Model 600C gas chromatograph using a
tritium electron capture detector with an N2 carrier and a 5-percent Carbowax 600 on
60/100 mesh Chromosorb W (DMSC treated) 22 in. x 1/8 in. Teflon column run at
ambient temperature (25° C). The 0. 5 cc sampling loop is made of Teflon and gives
2-22
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
an LLD of about 0. 003 ppm PAN. Standards are prepared by irradiation of an ethyl
nitrate-oxygen mixture with a UV lamp. Concentration of PAN in this standard is
determined by infrared analysis, and the 10 cm cell containing the mixture is purged
into a Tedlar bag with air measured by a wet test meter. Extensive calibration has
shown good linearity from 0. 05 to 1. 0 ppm with precision estimated at 25 percent.
2.3.7 Moisture (Water)
Moisture content was monitored with a Cambridge System Model 992 Hygrometer.
Samples were drawn periodically from the chamber manifold through a two-way
sampling valve. Otherwise, room air was purged through the instrument allowing
continuous instrument readout. Stable instrument response at 250 cc/min sample
flow was less than 3 minutes for dew points ranging from 48° F to 56° F. Precision
was less than 0.1 percent, based on saturation pressure.
2-23
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Section 3
RESULTS
Behavior of photochemical reaction systems is usually characterized by a few chosen
calculated parameters. In this work, the parameters used are those defined by the
CAPI-6-69 Project, with the addition of several measures that appear interesting, for
a total of 23, as defined in Table 3-1.
Parameter U.7J was added to describe the NO disappearance as good NO data are avail-
able from the chemiluminescent instrument. For almost all runs, parameter (ITJ is
smaller than parameter Clj , the NO_ formation rate.
An induction period of several minutes is observed when the chamber lights are turned
on. During this period, the NO disappears slowly, whereas the NO2 is increasing
(after accounting for the instrument lag time). Such an induction period would make
parameter (l?) smaller than parameter MM . Further analysis of this behavior and of
the ratio of NO at maximum to initial NO will be conducted and reported upon in the
L* X
Phase III report. The NO/NO9 rate difference is largest for the stainless steel surfaces.
Parameters (18) and (19) are calculated to give additional insight into the NO and oxi-
dant dosage values, by normalizing them to a potential maximum dosage represented
by the denominator.
Parameter (20) describes the NO2 curve to some extent, by giving the full width at half
maximum of the curve. It h!
shape on a numerical basis.
maximum of the curve. It has been included to facilitate comparisons of the NO curve
Li
Parameters (2l) to (23) give various defined intervals in the photochemical run. The
crossover time is used because it is well demarked on the run graphs. Measuring
time from the crossover time yields system behavior characteristics that are less
dependent on initial conditions.
3-1
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
Table 3-1
PHOTOCHEMICAL TEST CALCULATED PARAMETER DEFINITIONS
NO9 Formation
i-i
©
NO.
Rate = 7j7= - where T , is the time to form an amount of NO0 equal to 1/2
1/2
the initial NO in addition to the NO present initially; and NO.
iL 1
is the initial NO concentration in ppb (ppb/min)
2 ) T = time, in minutes, to the maximum NO0 concentration (min)
x? 300
JBJ Dose = / NO2 dt where NO2 is NO ppm and t = minutes (ppm-min)
Oxidant Formation
(I)
Max. Rate = ' Qxidant where T and T, , . are the times to form 3/4
' 3/4 ~ 1/4' ' '
and 1/4 the maximum oxidant; and max. oxidant
is the maximum oxidant concentration (ppb/min)
c > A ID 4. Max. Oxidant , „ . ,, ,. , , . /0 ,,
5 ) Avg. Rate = r^ where T1 , is the time to form 1/2 the maximum
1/2 '
oxidant, and Max. Oxidant is the maximum oxidant
concentration (ppb/min)
6 J Max. Cone. = maximum oxidant concentration (ppm)
7 } T = time to the maximum oxidant concentration (min)
' max v '
8) Dose =/ Oxid. dt where Oxid. is oxidant and t = minutes (ppm-min)
•• — 0
Hydrocarbon Disappearance
9J Final Cone. = ppm hydrocarbon after 300 minutes irradiation (ppm)
the times required to reduce the hydrocar
tration to 3/4, 1/2, and 1/4 of the original (min)
(lo) T , T and T = the times required to reduce the hydrocarbon concen-
U • i O U.O U«^jO
ai
HC. - HC
[13) Max. Rate = -^^— =, . where T / and T / are times for the disappearance
- i ^ i£ 0/4 1/4
of 3/4 and 1/4, respectively, of the hydrocarbon
disappearing in 300 minutes; HC. is the initial
hydrocarbon concentration; and HCc is the final
hydrocarbon concentration (ppb/min)
3-2
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Table 3-1 (Cont.)
HC. - HCf
(14) Avg. Rate = — r= - where Tn , is time for the disappearance of 1/2 the
^^ 1/2 '
hydrocarbon disappearing in 300 minutes; HC. is the
initial hydrocarbon concentration; and HC,. is the final
hydrocarbon concentration (ppb/min)
(is) Max. Aid. = Maximum total aldehyde concentration (acetaldehyde for these runs)
(ppm)
(16) Max. PAN = Maximum peroxyacetylnitrate concentration (ppm)
NO.
NO Rate = „ 1 - where NO. = initial NO and T , = time to reduce NO to
1/2 l
half of original concentration (ppb/min)
NO Dose
NO0 Dose Factor = ,T_ — x 100 where NO = initial NO
£ oUU JN v_/ X. X
X. 1
^ T-. T-I t Ozone Dose _ 1nn ,n.
Ozone Dose Factor = NO - x 100 (%)
X .
1
(20) NO0 FWHM = Full width at half-maximum of NO curve (min)
£ &
Crossover Time = Time at which NO and NO2 curves cross (min)
2-21 = NO_ T - Crossover time (min)
2 max v '
7-21 = Ozone T - Crossover time (min)
3-3
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
The four materials affect the behavior of the propylene/NO reaction system differ-
X
ently. As a summary, Figs. 3-1 to 3-4 show the photochemical results for Teflon,
Pyrex, aluminum, and stainless steel, respectively. Each figure gives the average
for the two S/V ratios and the two cleaning techniques for each spectral distribution.
Run parameter data are given in Appendix A for each run.
Table 3-2 gives the effects of the three independent variables, for the four materials
separately. Notations for that table are:
m = mean value of the parameter
A = effect of changing from low S/V value (1.3 ft ) (represented by -1) to high
S/V value (represented by +1) (2.7 ft"1)
B = effect of changing from cutoff spectrum (-) to full spectrum (+)
C = effect of changing from purge cleaning technique (-) to vacuum offgassing
(+) cleaning technique
AB, AC, BC, ABC = interaction effects
s = estimate of the standard deviation of replicates
10% = value of parameter that would be exceeded by chance 1 time in 10 if
true value of effect were zero, i.e., significant at 10% level
1% = value of parameter that would be exceeded by chance 1 time in 100 if true
value of effect were zero, i.e., significant at 1% level
The table is calculated using standard methods for factorial tests (Ref. 4). For each
material, 2 levels of A x 2 levels of B x 2 levels of C = 8 runs are available.
From these we form 4 pairs that differ only with respect to one of the independent
variables (A, B, or C). The average of the four differences is then the estimated
effect of the independent variable. This procedure is repeated three times. For this
balanced experimental plan the estimated effects are uncorrelated with each other, a
decided advantage in interpretation. Interactions are determined from the intereffects
of the variables.
3-4
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
TEFLON
PYREX
50 100 150 200 250 300
TIME (MINI
0 50 100 150 200 250 300
Fig. 3-1
Composite Photochemical Test Fig. 3-2 Composite Photochemical Test
Results for Teflon Film Surfaces Results for Pyrex Surfaces
50 100 150 200 250 300
TIME (MINI
STAINLESS STEEL
™
—— FULL SPECTRUM
CUT SPECTRUM
50 100 150 200 250 300
Fig. 3-3 Composite Photochemical Test Fig. 3-4 Composite Photochemical Test
Results for Aluminum Surfaces Results for Stainless Steel Surfaces
3-5
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
(J
J5
w
H
°? Jj
00 bH
-2 g
oj in
H H
O
w
h
fc
W
|
§
fa
V
on
a
eg
i
©
0)
QQ
O
^
co if
S z
" ©
w
g
3
p.
1
i
o
N V
Q £5
Z HS
S HCM
C^V
.&
&
CM
O
/^\
1
SB
H
.
co
co'
c
C
£_,
S
£
|
^
,
^
CO
1
X
2
£
1
co'
c
»2
H
(U
H
1
•
CQ"
^
0
H
£
ft
^
,3
•K- *
toc-ocgegeat-"*
•^©os^rHoego
CO 1 1 1 II
* *
t-iOrHooegrHeOLO
00©COi-lTtCOOlOOt-O KJ^
52coocooiocMco
CO rH rH rH
1 1 1 1
Oiot-t-lOCOCOlO
»CMt-rHCMOOO
CO^flrH^^^COrH
IO -44 rH
rH 1 1 1 II
* *
Cnt~OCM^rHint- 1
3 t 7 , •"-'•' ,
". *. ". (n)
CO CM C- ^~S
rH rH
* * # ||
QOincOTfrHCnrHCO
t-OOt-CMTfCMdCO
vej/
+- # N.
CO^CMOOOinrH
t- rH crj -H rH
rH 1 1 1
* »
»o o to o m o o
OCMOOOrHOOOrH
^ CM CM CM
* * 2
rH 1/5 B
TfOSt-Ttoeoost-d t^
IO rH o ,-H rH ^^
<— ' 1 | 1 I*
to ^
CO O rH >
* # . . * O
egcgegc-t-ejcg^-^jfj g
rHt-^eoo^cooo P
3 " 7 , *? , , •" o
to o o io o to ^— ^
CO . 00 CO CO C- Tt< Tf
S ' T
* # *
•*lOcOCOrHC-CO-#
lOCOOSOCOrHOrH
CM rH ^
1 III
COCDD-rHCnoOCOCO
COCMt^c-OlOOrH g
OOOcMrHOOOO ffi
O t- CO f*1
co oo co ej
Ot-COinTllc'u' yr
OO lOOOrHrHrH ^^
rH 1 II /3\
k£3/
m CM co io o oo o
rHCOTj*COCOinCOrH
CO O^rHvHrHl-HO
rH II 1
S<;«o<:<:w< -H^ §
w
co' "* °°
•* CO
CO rH
c o en
o CM m
*§ *- <=>
H
„ in co
g in o
fc a' N'
PH 1
g rH CO
3 e- o
<* 00 O
^ 1
co' *
CO 00
CQ CO
e *
I
1 s ?
B «
H 00 CO
>> m CM
g t-
1 g -
*
o
CQ
• 00 **
CQ CO rH
1
§
C NO
U oo
g «
H co eg
>> t> rH
S t~
1 s 7
co' *-
• co en
CQ CO
a
fio en
CM rH
H rH 1
Q) 41
(V* ° "
g «
rH en co
<; 05
s <
CO
•*
+-
CM
#
rH
CO
*
o
CO
eg
#
rH
t
*
t-
LO
*
00
CO
1
•K-
rH
*
rH
1
w
CO
•t
*
to
CO
1
?!
rH
1
#
rH
CO
1
*
OS
f
en
1
«•
t-
CM
1
m
en
rH
CO
rH
O
1
CO
CM
rH
CO
rH
O
1
CO
*
t-
c-
m
c-
l
CM
CM
«
O
CO
rH
t-
eo
rH
io
"
eg
LO
rH
rH
*-
cn
OS
i
eg
rH
O
TP
o
rH
O
o
1
00
CO
o
s
o
1
rH
1
fr-
co
LO
O
1
eg
rH
O
eg
rH
to
OS
1
t-
*
rH
1
t-
rH
CO
eg
t-
0
CQ
CO
o
1
0
CO
o
1
eg
OS
0
1
CO
t-
o
rH
eg
CO
o
CO
1
t-
rH
1
LO
1
eg
t-
LO
rH
t-
10
1
rH
rH
rH
rH
rH
I
eg
LO
i
c^
t-
o
CO
o
o
1
CO
in
rH
CO
0
1
•*
1
CM
"
10
rH
(N
O
O
1
CM
O
1
in
CO
i
t-
CO
in
i
CM
CM
t-
O
8
CM
O
1
CO
CO
o
eg
rH
I
O)
O
o
rH
1
CM
rH
I
10
CO
rH
t-
rH
10
rH
eg
Tf
1
LO
CO
c-
CM
rH
1
CO
1
to
rH
IN
eg
£
<
eg os co
rH rH* CM*
rH CO C-
cg ^* LO
OO CO t-
rH CM IO
^ CO OO
CO O O
oo eg o>
TP t-" OS
m B^ ^
ffl 0 rH
rH
•a
II
3-6
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
~
•g
o
o^
°?
oo
*i\
QJ
rO
H
CO
PH
H
W
s
«
Q^
w
s
0
S2-
,
CO
CO*
(3
3 o
*5J fl)
H H
o
§ g
o >>
^^
|
5
CO
co'
3 a
o *""*
s S
s
o
X §j
A £
Tco)
a
<
CO
co'
5
S
o
bo ^
S
0
N X
O ^
@ &
S
r3
w
CO
-M
d (H
^ r2
I S
1 s
O H
©ft
s
^
"o
CD
4-1
* *
CO rH Tj<
rH 1 1
* #
O rH 01
CO CO 00
eg i i
* #
rH 1 1
*
c- m
rH 1
*
O CO O
O O O
CO CO 0
O 0 CM
rH O O
1
* #
CO CM Ol
OS rH 0
000
1
#
O rH O
rH O O
1
•M-
CO O CO
LO O CO
f
cxi o eg
rH ea en
CO O rH
* *
CO CO rH
CO rH CM
1
# *
OO CO rH
O CO CO
^ O rH
1
* #
cxi eg oo
rH 00 O
eg i rH
*
CO ^ CM
eg i rH
* *
CM T °°
* *
CM T ^
fi<»
(M
, V
4—
(N LO t- rH CO
i-H rH LQ
o o to
oo eg co
rH rH
T t ^ 7 s
oo eg Oi -^ co
i
-t—
LO co CM rH eg
o o o o o
1 1
# -i—
LO CD ^ m
rH 0 0 0
0 0 0 Cp CO Oi 00
§38
^ 0 0 0
CO iH CO iH CO
iH 0 0 0 0
o o o o o
1 1
*
oo oa
°. °. 1 1 1
o o
#
0 0 0 0 O
rH O O O O
1 1 1
tO r-l rH O 00
rH rH CM CM 0
o o o o o
oo in en
iti m co eg
eg co to
^ o o o
LO 0 0 CO LO
o o o o o
1 1
OO rH LO CO
o -^
•i— eg co 10
i i
01 t- 0 rH 03
1 1 1
B98BS-S«
g
o
cd
h
m
a
Q
O
(S)
0
ra
Q
Q
(V
c
o
N
O
(oo)
(S)
1
-^
fc)
II
(§)
u
§
O
1
PH
S
1
CO
o
Q)
S
I1
B
,3
^
.
CO*
CO
o
s
I1
s
<
•
,
CO
o
EfH
H
s
FM
g
r3
•
CO
co"
rt
5
G
S
6
IM
ft
S
<
rH ^
T^ CD
1
co en
QO" rH
eg
•K-
00 LO
t- eg
eg rH
#
CO -^
35 w
CO
t> Ol
rH °?
O
B ^
#
CD OO
#
O
CO N
LO CM
rH 1
CO ^*
0
rH 1
^.
10
oo o
rH 1
O
LO CD
rH
-H—
10
CD CO
rH
rH
CO CO
eg o
0 0
#
O5 O
eg rH
o o
1
rH O
O O
LO LO
CO 0
o o
1
a <
CO O
. 1
0 CO
rH
*
CO 00 CO
o* o" co"
CM (
* *
co eg o
CM •* eg
rH
*
en 01 c-
c- o cxi
i
co -^ eg
t i
#
LO LO O
Ol CSI
* *
rH CO LO
CO CM i-H
*
O LO O
OO m rH
CO rH
1
* *
rH rH
CO rH |
1
^f.
in o in
eg I-H Tf
• ^ rH CM* CO*
O CO
1 1
C- LO
i— 1 rH
1
00 CO
LO O
rH OO
^ rH
oo cxi t™*
rH rH
CO CO
rH
1
LO in
00 CO
CD CM
1
LO m
O T^
1
CO ^ CD
00 CM Ol
•^ t> Oi
LO lO
in o
eg
o o
rH rH
CO
o ,
o
co eg
o o
o o
i Oi oo o
CO LO OO
0 O 0
„ . o* o* o*
CO rH
o o
0 0
Tj1 CO
o o
0 0
« < w 3 ""
r-T 'CD
® J
^
?H 2
t3 15
11
ll
3-7
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
M
w
w
S
H
CO O
(D CL
i—H "
p" f^"
H
CO
CO
£
LO EM
C-J O
o
H
y s
(S)
g
J2
<
.
CO
CO*
c
10 ,2
o ^
H
£_}
K g
© &
g
r— <
CO*
.
CO
Irt
"•"". o
0 EH*
o
/o\ S
^-^ £
PH
§
,J5
CO
C) ^
O
r-l O
§ ^
E H
o
K
®Q)
fe
S
r3
<£
"o
W
*
CO QO
CO iH
rH 1
#
CO O>
iH rH
CO 1
*
C} in
T— 1
CO CO
CO
iH
t- CO
^f
o m
CO rH
rH 1
•X-
t- Tt<
in T^
rH
CO CO
CO
rH
•X-
CO CO
1
TH
CO 1
rH
#
t- o
rH CO
rH
H4 CO
'
in co
rH O
0 0
1
#
CO O
CO rH
O O
1
#
co in
|
CO 00
o
1 1
-K- -i—
CO CO
1
* *
O CO
CO CO
r
CO rH
CO rH
1
#
CO CO
1 1
^
CO CTi
CO
1
•x-
°. 1
0 '
1
*
••^ co
CO O
o o
1 1
o^ o
rH rH
o o
1 1
•3—
CD
°. |
O
1
CQ Q
Tf iH
rH
1
to t>
CO T^
CO rH
1 1
CO O
tH
1
t- t-
1
CO t-
m as
CO
i i
t> t-
i
Tf rH
rH
I
•^
|
1
CO rH
rH rH
1 1
C^ CO
1
o o
o o
1
o o
o o
CO O
O rH
o o
1 1
rH rH
O O
o o
1 1
PQ U
< <
rH
rH
1
t-
in
r-
Cft
i
-*
rH
1
t-
G>
1
rH
1
Tf
0}
CO
O
^>
OO
o
o
Tf
o
o
•^
o
o
s
m
ffi>
1 <£.
t- t-H Oi
CD O CO
to
Cd
IN
to
CO • 0
in
0 l>
<*> 2 ^ °'
, < H
m .j
00 IN O Q
» o ^ 1 * °
rH rH °
. t-
H CO
CO ^0
1 <-
CO
Cfi
t^ VJ. co
CO rH
-2 *
S ! N
i . ^ °°
CO Tj< rH > £~"
00 CO* t-"
iH iH Q
ffi X ^i
co /SN £j o
rH \*-4) ^* rH
\O/ p^ ^
g m
OO ,3 rH
CO
a
00 -
0 >< ° *
O ra *tt to
I *=d QJ .— i
OS 00 O ^ tH
« m co r,
0 0 O U
CD O O O
o /SJ\ g to
o
-------
LMSC-D401598
For example, consider the following data for NO? time-to-maximum for Teflon
surfaces:
Run No. A B C Time
38 + + + 121
34, 35 + + - 103
42 - + + 142
43, 44 - + - 115
36, 39 + - + 192
37 + 162
40 + 193
41 - 205
Effect A is [(121-142) + (103-115) + (192-193) + (162-205)]/4 = -19.1
Effect B is [(121-192) + (103-162) + (142-193) + (115-205)]/4 = -67.5, etc.
Note the efficiency with which each data point is utilized. This is characteristic of full
factorial experimental designs. Selected conditions were replicated, to provide data on
reproducibility to determine whether effects are "real" or are due to random deviation
in chamber behavior. In this preliminary analysis, these replicates are averaged to
allow the orthogonal data treatment just described.
Deviations between replicates were used to obtain s, the estimated standard deviation
of the parameter. For all four materials and 23 parameters, pooling of the deviations
is justified. The significance levels are then calculated from the t value for the num-
ber of degrees of freedom and s. The "effects" found are compared to the 1-percent
and 10-percent significance levels in data interpretation.
Practical considerations required that the experimental plan be conducted for each
material and each S/V level as a subgroup. A time trend analysis (Appendix B) was
thus conducted, to see if systematic drift was present in the experiment. Three sets
of replicates are available, immediate reruns, reruns differing by more than 1 and
less than 15 run numbers, and a complete rerun of the first material used in the experi-
ment (run number differing by more than 50). Both the Rank Sum Test and the Sign
Test show no compelling evidence of drift.
3-9
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
3.1 MATERIAL DIFFERENCES
The four materials may be grouped in terms of increasing NO? formation rate: Teflon,
Pyrex, aluminum, stainless steel. Pyrex and aluminum are similar in behavior for
most parameters. Other reactivity manifestations such as times to NCL maximum,
ozone maximum, 50 percent propylene destruction, and NO dose follow the same
order. Preliminary correlation analyses indicate that stainless steel behaves in a
different manner than do the other three materials.
3.2 EFFECT OF FACTORS
Of the three independent variables studied, the spectral change caused the largest
change in behavior. For all four materials, the cutoff spectrum consistently and
clearly slowed the reaction relative to the full spectrum. The following table shows
the ratio of cut to full spectrum for several "reactivity" measures.
Teflon Pyrex Aluminum Stainless Steel
N°2Tmax0 i'58 i'78 i'54 !-49
Ozone T 7 1.49 1.68 1.47 1.39
50% Propylene Destruction (n) 1.50 1.64 1.47 1.43
Cleaning technique appreciably affected several of the behavior characteristics for the
stainless steel system. S/V ratio measurably affected most parameters for the four
materials.
Further analysis and interpretation of the data accumulated is underway, and will be
reported upon in the Phase HI final report. The behavior of acetaldehyde and PAN is
being examined, and will be included in the Phase in work.
3-10
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
3.3 OZONE DECAY RESULTS
One parameter frequently used as a measure of chamber cleanliness is the ozone half-
life in the smog chamber. Measurements were made of the ozone half-life in the
chamber each time a new set of surface materials was installed, and when the S/V
was changed. The conditions were: temperature, 95°F; relative humidity, 55° F dew-
point; total HC as methane, < 0.1 ppm. The initial ozone concentration was 1 to 2 ppm.
These tests were usually conducted in conjunction with the vacuum chamber off-gassing
that was used as the final cleaning step after installing the new material. Ozone decay
was determined in the dark and for the illuminator (full spectrum configuration) at its
nominal 6, 500-W power output (decay in the light). Results are shown in Table 3-3.
Ozone decayed fastest in the presence of stainless steel surfaces and most slowly in
the presence of Pyrex. It should be recognized that ozone decay behavior is a func-
tion of history as well as material and configuration, and, by itself, has been the
subject of several research investigations (e.g. , Sobersky et al. , Environ. Sci. and
Tech., Vol. 7, 1973, p. 347). Correlation of these results with the photochemical
test results will be undertaken and reported in the Phase in report.
Table 3-3
OZONE HALF-LIFE STUDY
Half-Life in Half-Life in
Configuration Dark (min) Light (min)
Base Chamber 430 180
Aluminum High S/V 270 210
Aluminum Low S/V 340 215
Pyrex High S/V 360 300
Pyrex Low S/V 340 275
Teflon High S/V 295 200
Teflon Low S/V 350 270
Stainless Steel High S/V 160 100
Stainless Steel Low S/V 190 120
3-11
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
Section 4
DISCUSSION
A number of reactions in the hydrocarbon-oxides of the nitrogen photochemical system
are particularly sensitive to the low wavelength region of sunlight. These include:
1. a. OQ + h^ — O_ (»Ag) + O(1D) (below 313 nm)
o Z
b. O(1D) + H2O A 2HO'
2. HO + hv — 2HO' (below 340 nm)
Z Z
3. ROOH + hiv -* HO' + HO' (below 300 nm)
4. ECHO + hv — R' + HCO* (below 352 nm)
The absorption cross sections for these substances is very low. However, reactions
1 through 3 lead to the very reactive hydroxyl radical, and reaction 4 also leads to
reactive chain carriers.
The large effect of spectral distribution on photochemical reaction systems — at the
same intensity as measured by k, — has not previously been reported. However, some
literature data are available that suggest that such an effect is not only present but
may be general. Altshuler and Cohen (Ref. 5) reported a factor of 2 to 3 times higher
NO formation rates for tests in Teflon vs. Mylar containers. This difference was
-1 -1
attributed to the difference in k, of 0.35 to 0.4 min for Teflon vs. 0.25 to 0.3 min
d
for the Mylar. The substantial difference in light below 330 nm for the two materials
was noted, but not further discussed. This differential rate was observed for some 16
hydrocarbons, ranging in reactivity from 1,3,5 trimethylbenzene and 1, 2, 3, 5
tetramethylbenzene at the high reactivity end to ethylbenzene and toluene at the low
reactivity end.
Table 4-1 gives a "Mylar/Teflon" spectral effect for time to NO0 maximum calculated
z
from Altshuler's data, after normalizing by the ratio of 0.275/0.325 to account for
4-1
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
the difference in k . Glasson and Tuesday (Ref. 6) give experimental data that show
that NO formation rate is linear with k for a variety of hydrocarbons. Such a linear
factor is also suggested by Niki et al. (Ref. 7) as applicable to the early stages of the
photochemical reaction. The difference in transmission characteristics for the Teflon
vs. the Mylar containers results in a cut/full spectral distribution somewhat similar to
that used in the present study. The spectral effect factor for propylene was 1.65.
A similar treatment of that data for oxygenates yields the following "Mylar/Teflon"
spectral effect: formaldehyde 5.9, acetaldehyde 3.5, proprionaldehyde 3.9, acrolein
2.9, ethanol 1.6.
Table 4-1
"MYLAR/TEFLON" SPECTRAL EFFECT*
Hydrocarbon Ratio of NCL T
—* 2— max
Ethylene 2.93
Propylene 1.65
Isobutene 1.81
Toluene 1.76
Ethylbenzene > 2.7
1,2 -dimethylbenzene 1.46
1,3-dimethylbenzene 1.32
1,4-dimethylbenzene 2.20
1,2 -methylethylbenzene 1.74
1,3-methylethylbenzene 1.96
1,4 -methylethylbenzene 2.29
1,3,5-trimethylbenzene 2.00
1,2 -diethylbenzene 2. 05
1,3-diethylbenzene 1.67
1,4 -diethylbenzene 2. 06
1,2,3,5-tetramethylbenzene 1.92
*Calculated from Ref. 5.
4-2
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Bufalini et al. (Ref. 8) have reported that photooxidation of formaldehyde in the pres-
ence of NO0 proceeded more rapidly at k, of 0.14 min with sunlamps than at 0.32
-1
min with blacklamps. This is an obvious result of the photodisintegration by the
shorter wavelength light from the sunlamps. The time ratio for 50 percent consump-
tion of formaldehyde, corrected for the k, ratio, is 1.6 for the blacklamp/sunlamp
distribution. For an irradiation without NO0, the corrected time ratio for 37 percent
Li
consumption is 2.7.
The distribution of NO2 photolysis events versus wavelength has been calculated for
several spectral distributions. For this calculation, the wavelength interval between
290 and 410 nm is considered. Absorption coefficient (Ref. 9) and quantum yield (Ref. 10)
for NO2 are tabulated at 10-nm intervals in this interval. These multiplied by each
other and by the number of photons in the 10-nm wavelength interval gives the total NO0
Ll
photolysis rate. This total rate divided into the events in each 10-nm band gives the
fractional distribution of NO2 photolysis events, or shows how the same k, occurs for
different spectra. Figure 4-1 gives the results.
FRACTION OF NO2 PHOTO DISINTEGRATIONS
O p O O C
D S o Ch 8 a
.
SUNL
XENC
i
- - n An-
^
IGHT
)N-FULL
IN-CUT
A CHAMBE
/
f
/
/
/
1
I/
/
i
R ___
/'
/
/
/
/
/
/,
V'
/
(
\
V
^
/
/
1
^ — •
/
N
\
\
1
\
t
1
\
\
\
\
\
s
\
1
\
i
i
i
I
\
xNd
280 300 320 340 360 380
WAVELENGTH (nm)
400
420
Fig. 4-1 Distribution of NCL Photodisintegrations for Various Spectra
4-3
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
The sunlight curve is for high noon in Los Angeles or Stanford, Conn, (two distribu-
tions averaged) (Refs. 11,12); the xenon-full and xenon-cut are for the spectra used in
this study, as previously shown in Fig. 2-6. The NAPCA chamber curve is for the
spectral distribution given by Korth et al. (Ref. 13), which is one of the few reported
chamber spectra in the literature.
The xenon-full NOr, disintegration spectral distribution does indeed closely match the
sunlight curve, over the entire wavelength interval. The NAPCA chamber distribution
is probably typical of chambers illuminated by fluorescent tube combinations, and shows
that the wavelength band from 335 to 365 nm is overemphasized (relative to the sunlight
distribution), while the 385—410 nm band is underrepresented. This disparity is fre-
quently unrecognized in discussing the application of smog chamber data to atmospheric
simulation.
It is clear that the spectral distribution of light provided a smog chamber influences
the photochemical reaction observed. In particular, the NO0 photodisintegration rate
*L
(k, or its equivalent k..) dies not sufficiently characterize the light conditions.
4-4
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
Section 5
FUTURE WORK
5.1 ONGOING WORK
Further work on this project is being conducted. Photochemical runs for the butane/
NO system are underway. The question of relative humidity effects is under study,
X
by a series of runs at low relative humidity (5 to 10 percent) to yield comparison data
with the runs discussed, which were at 25 percent relative humidity.
Further analysis of the completed experimental program also is underway. A covari-
ance analysis is being conducted, to refine the linear extrapolation method used to
account for varying initial NO0 content in the NO . A multivariate analysis will also
^ x
be conducted, to group run parameters and allow better estimation of effects.
This effort will then be utilized to synthesize a model that may be used to account for
differences reported in the behavior of various environmental chambers.
5.2 RECOMMENDED FURTHER WORK
Following completion of the scheduled ongoing work discussed above, further investi-
gations are recommended.
a. Conduct a similar set of tests for another hydrocarbon/NO system, such
J\.
as m-xylene/NO . This will indicate whether the observed spectral effects
X
are general, as the range of organic species from unreactive aliphatic
(butane), reactive olefin (propylene), and reactive aromatic (m-xylene)
will then be available. Also use of a range of compounds from unreactive
gas to polar high-boiling liquid may show differing cleaning effects.
5-1
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
b. Perform further studies of the spectral effect, by varying the cutoff wave-
length. By use of Teflon rather than Pyrex chamber faces, the amount of
light at wavelengths below 320 nm can be substantially increased. Evidently
this lower wavelength light is disproportionately important in smog photo-
chemistry. A second new spectrum to investigate is one that has a cutoff
between the 320 and 350 nm spectra just investigated. Having available
spectral effect data for four cutoffs, it is then possible to obtain an impor-
tant function for the wavelength range.
c. Determine the spectral distribution for the various smog chambers compared
by the Project group, and in conjunction with the importance function gen-
erated in item b above, normalize chamber data. Spectral distribution de-
termination could be done on a calculational or experimental basis.
d. Investigate light intensity effects by a set of tests at 50 percent and 150 per-
cent of the light intensity previously used. It is fairly well established that
initial behavior of the photochemical system is linear with light intensity,
but how late smog manifestations, such as ozone maximum concentration
and PAN build-up, vary is not well known. Such data will also be helpful
in applying chamber data to the atmospheric diurnal intensity variation.
e. Searching for explanations of persistent anomalies in smog chamber be-
havior would be productive. Among such anomalies not well understood at
present are the occurrence of peak NO0 concentrations greater than initial
&
NO charged (for fast reacting systems such as propylene); the initial
X
induction period in NO disappearance; and the entire nitrogen balance. One
technique for such an investigation would be to utilize an alternative detec-
tion method for the nitrogen species to correlate with the Saltzman NO2 and
the chemiluminescent NO. Fourier Transform Spectroscopy is such a tech-
nique, and arrangements may be made for such a spectrometer.
5-2
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
REFERENCES
1. Coordinating Research Council, Individual Hydrocarbon Reactivity Measurements:
State-of-the-Art, CRC Report No. 398, New York, Jun 1966
2. D. B. Wimmer, "Factors Affecting Reactions in Environmental Chambers,"
Coordinating Research Council Inc., Air Pollution Research Advisory Committee
Symposium, Chicago, May 1971
3. R. J. Jaffe, Factors Affecting Reactions in Environmental Chambers, Phase I,
LMSC-A997745, 20 May 1972
4. O. L. Davies (ed.), The Design and Analysis of Industrial Experiments, Ch. 7,
"Factorial Experiments," p. 247, Hafner Publishing Co., New York 1956
5. A. P. Altshuler and I. R. Cohen, "Structural Effects on the Rate of Nitrogen
Dioxide Formulation in the Photooxidation of Organic Compound-Nitric Oxide
Mixtures in Air," Int. J. Air Wat. Poll. , Vol. 7, 1963, p. 787
6. W. A. Glasson and C. S. Tuesday, "Hydrocarbon Reactivity and the Kinetics
of the Atmospheric Photooxidation of Nitric Oxide," J. Air Pollution Control
Assoc., Vol. 20, 1970, p. 239
7. H. Niki, E. E. Daby, and B. Weinstock, "Mechanisms of Smog Reactions,"
Advan. Chem. , Vol. 13, 1972, p. 16
8. J. J. Bufalini, B. W. Gay, and K, L. Brubaker, "Hydrogen Peroxide Formation
From Formaldehyde Photo Oxidation and Its Presence in Urban Atmospheres,"
Env. Sci. Tech. , Vol. 6, 1972, p. 816
9. P. A. Leighton, Photochemistry of Air Pollution, Academic Press, 1961
10. R. J. Gordon, in National Air Pollution Control Administration Pub. 999-AP-38
11. R. C. Hirt et al. , Ultraviolet Spectral Energy Distributions of Natural Sunlight
and Accelerated Test Light Sources, J. Opt. Soc. Am., Vol. 50, 1960, p. 706
12. J. S. Nader, in National Air Pollution Control Administration Pub. 999-AP-38
13. M. W. Korth, A. H. Rose, and R. C. Stahman, "Effects of Hydrocarbon to Oxides
of Nitrogen Ratios on Irradiated Auto Exhaust," Part I, J. Air Pollution Control
Assoc., Vol. 14, 1964, p. 168
R-l
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Appendix A
PHOTOCHEMICAL RUN DATA
Photochemical run data are given in this appendix in two forms — a plot of NO, NO ,
£t
ozone, and propylene vs irradiation time, and the tabulation of the calculated run param-
eters as defined in Table 3-1. Data at 10 or 15 minute intervals are available for
acetaldehyde and PAN, but only the maximum for these species is reported in the tables.
The materials order is aluminum, Pyrex, Teflon, and stainless steel. Eight run graphs
are shown for each material. These are for the eight combinations of S/V, spectrum,
and cleaning investigated, and are arranged in sequence as follows:
S/V Spectrum Cleaning
High Full Vac
High Full Purge
Low Full Vac
Low Full Purge
High Cut Vac
High Cut Purge
Low Cut Vac
Low Cut Purge
For each material and variable combination, the average of the replicates of the indi-
vidual runs is given. Each run was separately analyzed, to obtain the 23 parameters
given in the tabulation.
Initial condition variations affect some of the run parameters rather strongly. For this
preliminary analysis of the data, the only initial condition accounted for is the initial
percentage of NO0 in the NO . As previously suggested, both on theoretical and experi-
^ x
mental grounds (Niki, Ref. 7, and B. Dimitriades, Bureau of Mines RI 7433), this is
accounted for by a linear extrapolation along the time axis to the standard reference
A-l
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
starting condition of 10 percent NO9 content in NO . The tabulated run data show this
^ X
adjustment as the column T ADJ. For the runs used in the effects analysis, the largest
value of T ADJ is 12 minutes. The covariance and multivariate analyses will further
refine initial condition adjustments, and will be discussed in the Phase HI report.
It will be noted that two complete sets of experiments were performed for aluminum
surfaces. Changes were made in the instrumentation after runs 3 through 6 in the test
sequence. These changes made the set of runs between run 69 and run 85 better suited
for the effects analysis. In addition, as experience in operating the smog chamber
accumulated, better control of initial conditions was obtained. The difference in
means between the early and late aluminum tests is given in Table B-3 (page B-8).
The average percent difference is 7.6 (excluding parameter 15). The effects as cal-
culated for the early and late sets of aluminum runs are similar. The data accumulated
in the late set of aluminum runs is preferable for the reasons just mentioned, and are
the ones reported.
A-2
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
This page is intentionally blank.
A-3
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
ALUMINUM
RUN S/V SPEC CHAN
71 HIGH FULL VAC
50
100 150 200
TIME (MINI
250
300
RUNS
69.85
ALUMINUM
S/V SPEC
HIGH FULL
CLEAN
PURGE
150 200
TIME (MINI
3.Or
RUN
73
ALUMINUM
S/V SPEC
LOW FULL
CLEAN
VAC
100 150 200
TIME (MINI
250 300
3.0>
RUN
75
ALUMINUM
S/V SPEC
LOW FULL
CLEAN
PURGE
50
100 150 200
TIME(MIN)
A-4
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
ALUMINUM
RUN S/V SPEC CLEAN
81
HIGH CUT
150 200
TIME (MINI
RUNS
ALUMINUM
S/V SPEC CLEAN
70.82.M HIGH CUT PURGE
150
TIME (Ml N)
3.0|>
= 2.0
i.o
ALUMINUM
RUN S/V SPEC
74 LOW CUT VAC
50
PROPYLENE
• — NO
N02
OZONE
100 150 200
TIME (MINI
250
300
V**
CD
(
KJ
CD
ONCENT
j— •
CD
50
ALUMINUM
RUN S/V SPEC CLEAN
72 LOW CUT PURGE
\
X
\
/
/
^\
/
y,
\
•\
N
\
f\
t
i
v/
^
PROPYLENE
NO
wn
OZONE
\
\\
\
*^~ •
^
— —
"~-~
1
100 150 200
TIME(MIN)
250
300
A-5
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
PYREX
RUN S/V_ SPEC CLEAN
22 HIGH FULL VAC
PYREX
RUN 5/V SPEC CLEAN
21 HIGH FULL PURGE
50
100 150 200
TIME (mini
100 150 200
TIME (mini
PYREX
RUN S/.V SPEC CLEAN
3~f " LOW FULL PURGE
PYREX
RUNS S/V SPEC CLEAN
27.33 LOW FULL VAC
100 150 200
TIME (mm)
A-6
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
PYREX
RUN S/V_ SPEC CLEAN
26 HIGH CUT VAC
RUN
25
PYREX
S/V SPEC CLEAN
HIGH CUT PURGE
50 100 150 200 250 300
50 100 150 200 250 300
PYREX
RUN _, S/V SPJC CLEAN
30 LOW CUT VAC
PYREX
RUN S/V_ SPEC CLEAN
32 LOW CUT PURGE
50 100 150 200
TIME (mm)
250 300
50
100 150 200
TIME (mini
250 300
A-7
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
TEFLON
S/V SPEC
HIGH FULL
CLEAN
VAC
TEFLON
RUNS S/V SPEC CLEAN
34,35 HIGH FULL PURGE
100 150 200
TIME (MINI
50 100
150
TIME (MINI
TEFLON
RUN S'V SPEC CLEAN
42 LOW FULL VAC
TEFLON
RUNS S/V SPEC CLEAN
43,44 LOW FULL PURGE
50 100 150 200 250
TIME (MINI
300
50 100 150 200
TIME (MINI
250 300
A-8
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
TEFLON
RUNS S/V SPtC CLEAN
36.39 HIGH CUT VAC
50
100 150 200 250 300
TIME (MlN)
3 5
3.0
< 2 0
1 0
TEFLON
RUN S/V SPEC CLEAN
37 HIGH CUT" PURGE
• PROPYLENE
NO
NO,
OZONE
\
50 100 150 200 250 300
TIME (MINI
3 5
3.0
20
1 0
TEFLON
RUN !W SPEC CLEAN
40 LOW CUT VAC"
\
\
PROPYLENE
NO
NO,
OZONE
\
50
100 150 200 250
TIME (MINI
300
3.5
3 0
2.0
1.0
TEFLON
RUN Sl\i_ SJ>EC CLEAN
41 " LOW CUT PURGE
50
PROPYLENE
NO
NO.
OZONE
100 150 200 250 300
TIME (MINI
A-9
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
STAINLESS STEEL
RUNS S/y SPEC CLEAN
55.56 HIGH FULL VAC
PROPYLENE
NO
NO,
OZONE
100 150
TIME (MINI
STAINLESS STEEL
RUNS S/V SPEC CLEAN
51,52,58 HIGH FULL PURGE
50
100 150 200 250 300
TIME (MINI
STAINLESS STEEL
RUN S
-------
LMSC-D401598
RUN
57
STAINLESS STEEL
S/V SPEC
HIGH CUT
CLEAN
VAC
3.0 -X
150
TIME (MINI
RUNS
53.54
STAINLESS STEEL
S
-------
LMSC-D401598
TITLE- ALUM
1#
2#
3#
4#
5#
6#
7#
8#
9#
I0#
1 l#
I2#
I3#
I4#
I5#
I6#
I7#
I8#
I9#
20#
2I#
22#
23#
RUN! NO STO\
3.0
71 .
4.
4.
69.
85.
10.
II.
73.
12.
12.
75.
6.
81.
5.
70.
82.
84.
9.
74.
7.
8.
72.
0
0
5
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
t SPEC CLEM
1
1
1
1
1
1
1
1
1
1
1
1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
J HCIN
3.
3.
3.
2.
3.
2.
2.
3.
3.
2.
2.
3.
2.
3.
2.
2.
3.
2.
2.
3.
2.
2.
3.
IT
17
10
05
90
04
94
96
03
00
98
83
06
81
04
98
88
07
96
94
02
84
92
00
NOX
.52
.52
.54
.41
.47
.49
.51
.43
.47
.51
.67
.45
.59
.49
.33
.47
.49
.41
.49
.46
.42
.36
.46
N02
.22
.25
.29
.29
.21
.13
.31
.22
.14
. 11
.36
. 19
.31
.14
.12
.17
.17
.14
.33
. 16
.22
. 19
. 15
PC NO 2
14.5
16.4
18.8
20.6
14.3
8.7
20.5
15.4
9.5
7.3
19.2
13.1
19.5
9.4
9.0
1 1.6
1 1 .4
8.5
22.1
12.3
15.5
14.0
10.3
TADJ
9
12
17
21
8
-2
20
1 1
0
-8
18
6
28
-2
0
4
4
0
35
6
16
12
0
HCNOX
2.09
2.00
1 .98
2.06
2.07
1 .97
1 .96
2.12
2.02
1 .97
1 .69
2. 1 1
2.04
2.04
2.24
1 .96
2.06
2.10
1 .97
2.07
2.00
2.15
2.15
N02R
15. 10
14.00
15.80
13.00
14.70
16.00
15.80
13.20
13.80
13.10
15.50
18. 10
10. 10
12.60
8.40
8.78
13.70
II . 10
7.79
9.23
9.28
8.03
10.60
_/T\
^r\
/TN
^^
/T-\
/^\
_^T\
/T\.
/r*.
'TTTLE- ' ALUM" (2J (& Vl> <^> (6}~~(2J~ W~^> US'
RUN2NO N02TM N02DOS 03MAXR 03AVGR 03MAXC 03TM 03DOSE HCFC HCT75
1#
2#
3#
4#
5#
6#
7#
8#
9#
10#
Jl#
12#
13#
14#
lb#
\6#
!7#
18#
I9#
20#
2I#
22#
23#
3.0
71 .0
4.0
4.5
69.0
85.0
10.0
1 1 .0
73.0
12.0
12.5
75.0
6.0
81 .0
5.0
70.0
82.0
84.0
9.0
74.0
7.0
8.0
72.0
79
94
77
81
83
88
81
86
— 9j— -
92
92
76
125
118
107
144
114
120
145
136
124
125
120
142
165
142
128
152
166
145
146
160
159
168
157
178
211
123
200
206
186
176
188
159
140
172
22.8
21.3
32.5
30.9
21 .9
22.1
30.3
37.3
35.6
31 .7
28.9
34.7
19.2
17.4
17.6
18.0
2271"
13.8
23.0
26.2
22.8
26.7
21.2
4.46
4.55 1
5.84 1
5.53
4.48
4.23
6.09
5.54
5.18
5.38
5.31
5.47
3.24
3.29
2.72
2.71
3.02
2.73
2.74
3.48
3.32
3.38
3.52
.82
.00
.04
.05
.94
.93
. 15
.12
. 13
. 14
.04
.04
.92
.94
.77
.90
.84
.83
.92
. 10
.91
.96
.02
129
152
122
121
143
148
140
131
149
137
138
131
190
198
165
219
189
210-
205
206
176
1 77
190
135
165
165
158
144
158
200
190
181
189
193
201
133
140
114
112
133
121
116
147
127
135
137
.05
.18
.10
.10
.14
.10
. 19
.15
.12
.13
.14
.10
.15
.18
.35
.31
.22
.27
.31
.16
.27
.28
.17
70
86
69
64
76
82
81
83
91
85
91
63
101
104
96
107
T03
88
148
II 1
128
132
1 16
A-12
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
TITLb- ALUM
l#
3#
5#
7#
9#
1 !#
ill
14*
17#
!»J
2I#
22#
23#
RUN3NU
3.0
71.0
4.0
4.5
69.0
85.0
10.0
1 1 .0
73.0
— T2YO"
12.5
"""75VO"
6.0
81 .0
5.0
70.0
82.0
84.0
9.0
74.0
7.0
8.0
72.0
Qj)
(12)
(13) (14)
(15)
(16)
HCT50 HCT25 HCMAXR HCAVGR ALDMAX FANMAX
100 127 27.4 15.80 .60 .56
122
99
105
108
118
109
1 1 1
120
1 19
1 18
92
148
153
137
165
144
147
180
161
159
F6"0"
151
152
131
132
143
149
141
143
146
142
145
T24"
184
185
190
21 1
186
194
210
195
204
203
182
24
25
23
22
22
24
21
25
21
26
27
18
1 /
17
13
18
14
23
1 7
21
21
23
.9 1 2
.1 15
.1 14
.0 13
.9 12
.8 13
.2 13
.7 12
.0 12
.3 11
.4 16
.3 10
.0 9
.8 10
.2 8
.2 10
.8 9
.6 7
.9 H
.0 8
.6 8
.7 10
.30
.40
.70
.60
.50
.20
.30
.20
.80
.50
.60
.30
.67
.70
.10
.10
.89
.54
.87
.37
.63
.10
.70
.55
.51
. 72
.83
.59
.60
1 .00
.50
.54
.86
.77
.86
. 73
.94
.92
.82
.49
.90
.52
.58
.98
.31
1 .10
1 .10
.38
.29
.13
. 1 /
.35
.30
.33
.20"
.23
.35
.32
.37
.33
.26
.18
.45
.16
.14
.50
Qz)
NORATE
12.50
10.10
14.20
12.20
10.50
8.70
12.80
11.10
9.56
If. 00
II .80
TO . 80
9.93
7.88
9.03
6.70
7. 70
6.90
7.36
7.04
8.77
"7V82"
7.62
(18)
N02DF
31 .1
35.5
31 .6
30.3
33.6
35.9
32. 1
34.0
35.5
35.1
33.5
34.7
37.3
46.0
30.9
45.6
45.2
43.1
39.4
42. 1
37.3
34.3
38.5
Q9)
03DF
28.5
35.5
36.7
37.3
31 .8
34.2
44.2
44.3
40.2
41 .7
38.5
44.4
27.9
30 . 5
28.5
25.6
29.2
28.0
26.0
32.9
29.8
33 . 1
30.6
TTT1
2J
4*
6#
8#
9 it
10#
12#
13#
14#
16#
1 7#
I8#
~| OjU
20#
22#
23#
L"E- "ALUM
RUN4NO
71 !o
4.0
4.5
69.0
85.0
FD.O
1 1 .0
73.0
12.0
12.5
75.0
6.0
81 .0
5.0
70.0
82.0
84.0
"9VO
74.0
/ .0
8.0
~"~~7~2~."0"
FWHM XTI
~"~7"3
86
66
70
89"""
85
-7-4-—
67
82
72
73
77
110
115
89
126
110
131
FIT
117
101
97
TOP
^f\
/£S\
21) \&-
ME N02XT
43"""
52
44
47
44
49
43
49
52
52
48
40
66
58
64
76
56
63
82
73
64
71
63
36
42
33
34
39
39
38
37
39
40
44
36
59
60
43
68
58
57
63
63
60
54
57
(2_3)
03XT
100
78
74
99
99
" 97
82
97
85
90
91
124
140
101
143
138
147
" T2 3"
133
1 1 2
106
T2~7~"
A-13
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
TITLE- PYRhX
\# 22 1
?# ?1 1
3#
6#
7#
9#
27
33-
31
25
in
32
-1
-1
L
1
-1
-1
TIT' ^- PYPPY (2)
2#
,1#
4#
6#
7#
8#
9#
TIT
?#
3#
6#
7#
R#
9#
RUN2NO
22
21
27_-
33
11
26
25-
30
32
riE- PYR
RUK3N.Q-
22
?l
27
, 33.
31
25
10
32
N02TM
99
95
- 78
79
in?
192
2-L2
147
J2J
IEX©
-HCIbO
130
88
98
120
- 2LD
249
17?
148
i^EC-CLE
1
1
1
L
1
-1
-1
-1
-1
(T)
NO 2 DOS
161
-11 A.
114
1 ??
225
-332-
161
±44
©
159
117
L28
150
295
183
IAN. H'
1
-1
1
J
-1
..±-
-1
1
-1
C1NLT . -
3.22
l.?0
2.94
3..QQ
2.84
2.95
3.00
1 .68
1 -S7
1.55
1.^52
1 .39
1^44
1 .54
i _Rn
1 .43
(ft (s)
.27
. 1 7
.18
»J-6
. 14
.±8
.21
• In
.20
(&)
03MAXR 03AVGR 03MAXC
2B-R 4.19 .9H
24
— _36
32
17
28
25
26.
34.
34^
36.
12.
20.
.7 3
.3. 6
.9 6
.R 4
.7 2
.& J
.5 3
.A 3
(B
8 12.
0 16.
6 J_5*
8 11.
1 5.
9 R.
6 9.
.85
.13
. 17
*4-S
.02
)
00
40
LO
70
61
20
9?
52
.84
L. L6
1.11
.98
.92
.7-3
.97
.91
(15)
2MAX-.
.70
.96
.77
.68
.67
.65
._P-CJ102-TADJ_JdC-LtKIX
16.1 12 1.92
10.R 0 ?.f]4
11 .6
m.5-
10.1
-I2.3
13.6
in. 7
14.0
Q
03TMAX
143
J-22
125
140
267
3JO
207
J_66
P-ANMAX-J
.15
. 11
.13
J-4-
.18
. 12
5 1
0 1
0 2
7 2
10 1
? ?
11 2
®
03DOSE
L52.9
127.5
±87.7
190.0
1 S6.O
80.7
4J ^5--
127.8
©
4UBAIE N
10.30
9.7?
12.70
14.20
9.19
5.00
7.10
7.81
.90
.97
.04
.92
. 1 R
.10
^
HCFC
.26
r\ Q
m \. J f]
. 1 1
. 1 1
.30
.26
..24
([§)
H2HF-
32.0
14.?
24.8
24.8
29. 1
5-2,-t .
72.0
1R.9
33.6
N02R
12.80
1 1 . SO
13.50
15. DO
10. 10
6.25
7.26
7.6R
8. 10
(jo)
HCT75
9R
89
7Q
81
R7
149
-185-
133
(J9)
03DJ-"
30.3
?7. 1
43.4
.41 .1
37.4
.J9-.9-
9.0
?R.4
29.5
TTTT F- PYRFX(2<3) (2l) (22)
1#
4#
6#
8#
RUN4NO
2.2.
21
27
33
26
25.
30
FWHM XT I ME N02XT
74 61 4?
77
.63
66
77
138
115
.--1DQ-.
55
50.
bO
106
±on
89
_— 73.—
40
28
29
4?
86
JL4
59
(23)
03XT
91
88
72
75
RO
161
210-
118
..__93—
A-14
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
TITLE- TEFLON
RUNINO STOV SPEC CLEAN HCINIT
Iff 38 1 1 3.00
2# 34 1 - 2.82
3# 35 1 - 2.86
4# 42 - 1 1 2.97
5# 43 - -1 2.98
6# 44 - -1 " 3.03
7# 36 - 1 2.99
8# 39 "" ""- T 2.96
9# 37 - -1 2.86
I0#
11*
40
41
_
TITLH- TEFLON^)
RUN2NO N02TM
"l/» 38 121
2# 34 102
3* 35 104
4# 42 142
5#
6#
~7#
8#
"9*
10#
43
44
36
39
37
40
1 l# 41
1 6
1 3
f80
203
162
193
205
-
N02DOS
176
151
16?
205
203
188
21 7
22H
223
267
303
1
-1
2.96
2.96
NOX N02
.52 .16
.38 .3
.50 .5
.54 . 8
.52 . 5
.47 . 5
.63 . 7
.45 . 7
.48 . 5
*
•
(4) ^
03MAXR 03AVGR
22
23
""24
31
32
26
" """20
19
T5
16
10
.5
.6
.8
.3
.9
.2
.2
.6
."7""""
.4
.8
4.04
4.33
4.09
4.08
4.08
3.90
2.42
2.25
2.28
2.33
1 .67
49 . 7
47 . 4
®
03MAXC
. 10
.04
.04
.25
. 2
. 0
.01
.98
"""" .85
.05
.80
PCN02
10.5
9.4
14. /
II ./
14.5
13.6
10.4
11.7
10. 1
1 1 .4
9.5
(2)
03TMAX
83
68
74
200
76
95
260
267
237
295
298
TADJ HCINOX
0 1.97
-2 2.04
14 1 .91
5 1 .93
13 1.96
10 2.06
0 1 .84
7 2.04
0 1 .93
5
-2
(9>
03DOSE
66.0
66.0
81.0
89.0
89.0
80.0
94.8
97.0
97.9
93.0
50.0
TITLE- TEFM)N (Q) O Q © © 0 O
l#
2#
3#
4#
5#
6#
1#
e#
9#
"10#
ll#
RUN3NO
38
34
35
42
43
44
36
39
37
40
41
HCT50
143
133
134
156
143
144
205
224
197
217
230
TITLE? TEFLON (g
RUN4NO FrtHM )
1#
2#
3#
4#
"5~#~"~
6#
7#
8#
W~~
10#
II*
38
34
35
42
43
44
36
39
T7
40
41
90
86
91
84
93
102
143
142
""T40" '
165
190"
HCT25
174
161
169
190
175
177
246
262
241
263
282
- -
HCMAXR HCAVGR
23.6 10.20
21 .
21 .
8.
3.
8.
2.
4.
3.
2.
8.
TIME N02XT
18
58
63
73
63
60
1 07
107
82
96
~ 94
43
42
41
70
53
53
85
85
""58"
91
111
3 10
7 10
8 9
4 10
7 10
7 6
0 6
1 6
0 6
5 5
03O
105
1 10
1 1 1
127
122
122
157
157
155
199
204
.50
.30
.05
.30
.20
.81
.1 1
./O'
.39
.80
-
ALDMAX
.70
. 74
.95
.82
. 19
.73
. 10
.72
.72
.73
.72
PANMAX
.31
.25
.23
.61
.39
.34
.25
.21
.18
.25
.12
----
-- -- ----
NORATE
8.61
8.86
8.77
1.99
3.61
/.86
6.24
5.33
6.65
5.92
5.41
1 .99
2.01
W
HCFC
.12
.09
.12
.22
.08
. 15
.32
.32
.30
.34
.56
O
N02DF
38.6
36.4
36.5
44.3
44.3
4 2". 4
44.4
52.9
50.1
59.8
68.6
N(S?R
8.72
1 1 .30
11.10
9.52
11 .60
10.50
6. 19
5.91
8. 1 1
6.81
7.56
HCT75
109
95
102
1 12
1 05
98
36
66
39
47
36
0
03DF
36.4
40.0
39.5
39.8
39.6
39.8
19.4
21 .0
22.0
19.8
11 .3
- - --
A-15
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
TITI
RUN
I*
2#
3#
4#
5#
6#
7#
8#
9#
I0#
II*
12#
13* _
14#
NO STOV SPEC CLEAN
55 1 1 1
56 I 1 1
51 J 1 -1
52
58
62
£1 -J
66
57
53
R4
63
6.0—.-
67
1 -1
1 -I
1 1
1 -J
1 -1
-J. J
-1 -1
-1 -1
-1 1
-1 -1
-1 -1
HCINIT
3.. 19
3.04
3-OS
3. 11
?.R6
3.02
2*9a
3.06
_ 3^JJ
3.05
1.?4
2.88
3-02 —
3.00
NOX
.48
.52
*42 _
.43
.49
.46
*50
.48
L,44__
.49
.5?
.44
L,.b5-
.53
N02
,-15.
.18
*15
.15
. 17
.15
*1A
. 16
^17
. 16
.17
.13
*ia
.18
PC NO 2
10*1
1 1 .8
___ ICL..6
10.5
1 1 .4
10.3
10,7
10.8
1J...8
10.7
11.?
9.0
1U6
11.8
TADJ
0
0
Q
0
0
0
... 0
0
Q
0
?
0
.3
3
HCNOX
2.16
2.00
2,ia
2.17
1 .9?
2.07
. 1,59 .
2.07
2. 16
2.05
?. 13
2.00
US 5.
1 .96
— w
N02R
33.5
29.1
49.2
4t>.7
41 . 1
17.8
-26.8
24.4
2UO- -
35.3
26.4
10.8
21*2.
19.2
TITLE- SSTEEL(2)
RUN2NO N02TM N
.1* 55- 51
2# 56 53
3* 51 38
4# 52 37
5# 58 43
6#
7#
9#
10#
12#
I3#
14*
62
61
66
57
53
54
63
60 .
67
73
63
86
53
65
1 11
81
80
i)2DOS 03MAXR 03AVGR 03MAXC
133 PT-? 6.77 .88
134
121
1 16
133
139
134
186
157
164
181
174
158
20.9
22.2
21 .6
L9,5
31.1
31.7
12.6
12.6
13.3
20.0
16.5
19.6
6.87
8.16
8.04
6.86
7"! 70
3.43
4.72
4.40
3.50
4,34
4.50
.91
.81
.82
1.18
__U12
1 . 15
.69
.68
.71
.96
,85
.89
03TMAX 03DOSE HC
- _JJi L62---
110 1 69
90 159
95 146
LJIL- 164_-_
1 30 200
.-. 1 10 - 194 __
120 210
163 118
135
137
190
153
143
130
133
136
147
147
© —Q
:FC HCT75
_LO. 5 1
1 1 49
10 44
15 39
-L4_ -44
10 71
Q8 50
12 58
14 71
17
2Q ___
27
J.L .
27
48
64
1 1 1
.61
64
.TITLE- ..SSTkE
RUN3NO h
l# 55
2#
3#
4#
6#
7#
9#
10#
!)#
56
51
52
58
62
61
66
57
53
54
63
60
:L O
19
0
r
4)
-------
LMSC-D401598
TTTI F-
- SSTI-R eo CD ea D^
RUN4NO
1 #
2ft
3rf
4rf
h#
f>#
Hf
8ft •
9 it
\0#
1 \#
\24
1 3#
I4#
55
56
51
52
SH
62
61
66
57
53
c,4
63
60
67
FriHJ.'rXTlMH N02XT
70
67
60
61
64
78
75
69
107
88
HW
1 13
100
97
25
26
17
17
I H
38
27
31
40
23
p>o
59
38
40
26
27
21
20
2^
35
31
32
46
30
^A
52
-43
40
03Xt
90 ... ....__
' 84
78
78
Q2
92
83
89
123
1 12
ine -
131
J15 - - -- - -
103
A-17
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Appendix B
STATISTICAL ANALYSIS
The hypothesis to be tested concerns the presence of a steady drift over time regard-
less of the variables of material, surface to volume ratio, spectrum, or cleaning. If
such a drift were present, then runs which are duplicates with respect to these four
variables might be expected to differ by an amount which grows steadily as a function
of the separation over time. Our attention is then focused upon an examination of the
differences as a function of the time separation.
Note that the time separations of duplicates tend to fall into three basic groups:
1. Nine sequence numbers differing by one (A, = 1), i.e., immediate reruns:
4- 4A (AL) 34-35 (T)
7-8 (AL) 43-44 (T)
10-11 (AL) 51-52 (SS)
12-12A (AL) 53-54 (SS)
55-56 (SS)
Note: AL = aluminum; P = Pyrex; T = Teflon; SS = stainless steel
2. Ten sequence numbers differing by more than 1 but less than 20:
27-33
36-39
51-58
52-58
60-67
(P)
(T)
(SS)
(SS)
(SS)
At
At
At
At
At
= 6
= 3
= 7
= 6
= 7
61-66
70-82
70-84
69-85
82-84
(SS)
(AL)
(AL)
(AL)
(AL)
At =
At =
At =
At =
At =
5
12
14
16
2
3. Long term replicates — a complete rerun of the aluminum experiment
(A? > 50).
B-l
LOCKHEED MISSILES 8e SPACE COMPANY, INC.
-------
LMSC-D401598
In the analysis, we concentrate on the differences between duplicate pairs, since this
provides a direct basis for comparison. In the absence of drift, these pairs should be
distributed around a central value of zero. If there is drift, then the pairs separated
further apart in time (or sequence) could be expected to center about a value, dif-
ferent from zero and increasing in absolute magnitude as the separation increases.
Note, however, that the long term replicates (group 3) all involve only one material,
aluminum, and that each set of runs (initial aluminum, replicate aluminum) was
fairly compact in time. In this case, one might anticipate a different kind of drift
from what might occur during a single series. Thus, we distinguish two types of
drift — drifts within a compact series of runs on a given material and drifts from
material to material, or series to series.
The drifts within materials can be studied using the shorter term replicates, groups 1
and 2, while the drifts across materials or series can be studied by comparing the
mean values across materials, especially between the initial aluminum set and the
final replicate aluminum set.
B. 1 DRIFTS WITHIN MATERIALS
If there is a drift within material, then the replicates with longer separations, group 2,
should be centered around a different value than those of group 1. Hence we test the
hypothesis that both sets are centered about the same central value. The test used is
the rank sum test using the value of T' defined as follows. Arrange two samples (of
differences) in order of size, and assign rank scores to the individual observations;
score 1 for the smallest, 2 for the second smallest, etc. Then TT is the sum of the
ranks of the observations in the smaller of the two sets. Using the Dixon and Massey
tables*, we reject the hypothesis if the calculated score is significantly large or
* * #
significantly small. From those tables we find that for a sample with N.. = 8 and N2 = 7,
*W. J. Dixon and F. J. Massey, Introduction to Statistical Analysis, McGraw Hill,
New York, 1957.
**One complication in application is the triple formed by runs 51, 52 and 58, and 70,
82 and 84. Clearly we are not justified in forming three pairs. We have avoided this
by dropping the middle run from consideration, thus leaving in the replicate differ-
ence involving the largest time separation, (51 minus 58) and (70 minus 84). The end
result is 8 short-term replicates in group 1 and 7 longer, term replicates in group 2.
B-2
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
the significant values are T^ = 41 and T^_a = 71 for a = 0.047. Thus, values of
TT < 41 and > 71 form a 9.4 percent critical region for the hypothesis.
The rank scores, calculated in the manner described for each of the 20 parameters,
are shown in Table B-2. From the table, it can be seen that only parameter 13 quali-
fies as significant.
Here, we are confronted with a matter of judgment; by selecting a slightly more
stringent criterion for significance, we would have found no significant scores. Using
what is believed to be a conservative significance level, only one parameter - and that
parameter not one that had previously stood out as particularly suspect of indicating a
trend — is in the significant zone. All in all, the conclusion must be that we do not
have compelling evidence of drift in this case.
Next, we turn our attention to a second hypothesis suggested by the data. Is the central
value of all differences significantly different from zero? For this we use the sign test;
that is, the number of positive and the number of negative differences are counted
(when the difference is zero, this is excluded and the sample size reduced). Letting
r represent the lesser of the two counts, and N the sample size, we obtain Table B-2.
The null hypothesis of no difference from zero is rejected if r is too small. In par-
ticular (from Dixon and Massey, Table A-lOa*), a value of r of 2 or less is significant
at the 5 percent (two-tailed) level, for N = 12, 13, or 14 and 3 for N = 15. Thus,
from Table B-2, parameters 11, 12, and 14 are significant. Parameters 1, 2, 5, 9,
and 13 are close. Of these 8 parameters, we can determine whether or not the long
term comparison on aluminum indicates a trend in the same direction. This leads to
the following:
Apparent Direction of Long Term Trend
Up Down
Apparent Direction of Up 1 5, 14
Short Term Trend Down 2, 11 9, 12, 13
*W. J. Dixon and F. J. Massey, Introduction to Statistical Analysis, McGraw Hill,
New York, 1957
B-3
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
PQ
•a
cn
H
o
w
PH
PH
O
03
H
O
fc
H
PH
rH
rH
PH
~J
.
1
^^
O>
00
t-
CD
LO
^
CO
(N
rH
t-t
£
0
r4
A
rH
1
CD
1
m
0
t-
rH
rH
O
0*
1
rH
CO
O
CO
rH
CO
rH
^
1
co
CM
rH
-------
LMSC-D401598
o
o
i
PQ
0)
CO
IN
-------
LMSC-D401598
Table B-2
RESULTS OF RANK SCORE AND SIGN TEST
Parameter
Rank Score
Sign Test l +
Results I
Parameter
Rank Score
Sign Test 1 +
Results \
Parameter
Rank Score
Sign Test ( +Q
Results 1
Parameter
Rank Score
, +
Sign Test 1 Q
Results 1 _
1
47
7/4 11
1/3 4
7
49
3/2 5
5/5 10
13
39.5
3/1 4
5/6 11
19
52
2/3 5
1 1
5/4 9
2
49.5
1/2 3
1 1
6/5 11
8
51
3/1 4
1 1
5/5 10
14
61.5
6/6 12
2/1 3
20
57
3/4 7
5/3 8
3
58
3/4 7
5/3 8
9
52.5
1/2 3
1/1 2
6/4 10
15
62
3/3 6
5/4 9
21
63.5
2/2 4
1 1
6/4 10
4
66
4/5 9
4/2 6
10
48.5
3/1 4
1 1
5/5 10
16
56
4/3 7
1 1
3/4 7
22
58.5
3/2 5
1/2 3
4/3 7
5
59
6/5 11
2/2 4
11
50.5
1/1 2
1 1
7/5 12
17
55.5
5/4 9
3/3 6
23
45
6
58
3/4 7
1 1
4/3 7
12
53
1/1 2
1
7/5 12
18
49
5/3 8
3/4 7
4/1 5
1/2 3
3/4 7
B-6
LOCKHEED MISSILES & SPACE COMPANY
-------
LMSC-D401598
Thus, the trend appears to continue in only half the cases.
Summarizing to this point; while the results of these tests are generally consistent
with the hypothesis of no trend within sets, they also suggest that further study is in
order. Thus, we intend to do one further analysis. When the general analysis of
covariance is completed, we will include a trend term, presumed to be the same,
within materials, and examine the results to see if this term materially affects the
analysis. This will also remove any possible systematic trend in initial conditions.
B. 2 LONG TERM TREND
The mean values for each material steadily decrease or increase for the first three
materials. To examine this, we consider the mean values as reported for the orthog-
onal analysis. These are summarized in Table B-3 (with AL repeats displayed sep-
arately). Only a cursory examination is needed to see that any trend which might
have been suggested by looking at AL, P, and T is not continued through steel.
The more interesting aspect of the data concerns a comparison between the initial
aluminum set and the replicate set.
B. 3 COMPARISON OF EARLY AND LATE AL RUNS
Early AL runs are compared with late AL runs using an analysis of variance, which
includes a block effect.
This model hypothesizes that a systematic shift has occurred between the initial set
of AL runs and the replicate set. Note that a large value of the lack of fit mean
square, compared against pure error from replicates within sets, indicates that the
model is inadequate. There are several potential factors which we will include in
future analyses, such as initial conditions and time trend within sets. In addition,
physical grounds exist which lead us to suspect that the very early runs (prior to run
B-7
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Table B-3
MEANS BY MATERIAL
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
AL (Early)
11.65
104.2
151.5
25.4
4.15
0.95
157
147
0.19
97.6
131.2
166.2
23.1
11.66
0.597
0.37
10.81
34.1
32.9
87.2
57.7
46.2
99.4
P
10
130.9
178
24.7
3.6
0.93
189
126
0.26
116.6
156.5
188.8
23.6
10.13
0.696
0.14
8.55
39.2
27.8
102.8
78.2
57.5
114.5
T
8.63
154.2
219
21.3
3.12
1.03
230
131
0.26
124.4
179.5
218.2
15.9
8.18
0.752
0.29
7.20
48.7
28.6
124.9
81.6
69.9
147.5
SS
25.38
70.20
152.8
21.2
5.68
0.90
137
157
0.15
64.2
96.8
133.4
23.9
16.16
0.758
0.23
14.36
34.5
34.1
85.8
38.2
35.8
102.6
AL (Late)
13.12
105.8
176.1
24.7
4.08
1.00
172
156
0.16
93.6
132.9
165.9
21.8
11.52
0.870
0.35
8.71
39.0
34.3
99.4
56.2
49.6
116.2
B-8
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
7), particularly run 5, may reflect differences in calibration or instrumentation.
Thus, these results are preliminary until the full-scale covariance analysis is
completed.
The format of the results in Table B-4 is as follows:
The components of variance are (1) the factorial model without a block effect; (2) the
additional reduction due to adding the block effect into the model; (3) the residual
variation after removing the pure error - this residual is a measure of lack of fit
of the model, and (4) the pure error from replicates within sets.
Degrees of Freedom Sum of Squares Mean Square
Factorial 8 —
Block After Factors 1 -
Residual (lack of fit) 7
Pure Error 7 —
Total 23
The lack of fit (LOF) is deemed significant if the ratio:
Mean Square LOF ,., „ „„
Mean Square Pure Error 7, 7,. 95
If LOF is not significant, then the BLOCK effect is deemed significant if the ratio:
Mean Square Blocks after Factors _ = 4 «n
(Mean Square LQF + Mean Square Error) 1,14 ,.95 ~ '
2
If LOF is significant, then the model is inadequate and no test for drift is made. Note,
however, that if the ratio of
B-9
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Mean Square Blocks after Factors
Mean Square LOF 1,7,. 95 ~ '
there is reason to believe that a block effect is present.
Following are comments about Table B-4:
1. The results presented in Table B-4 have not yet been checked and hence
should be considered preliminary.
2. Of the 23 parameters, 8 (1, 2, 4, 8, 12, 13, 21, and 22) exhibit no sig-
nificant drift, while 5 (7, 15, 17, 20, 23) do exhibit significant drift. In
particular, parameter 15 stands out in this regard. This is to be expected,
since a basic change in the method of measuring 15 was made early in the
program. Consequently, the early set, with respect to this parameter,
should be discarded.
3. In the remaining 10 cases (3, 5, 6, 9, 10, 11, 14, 16, 18 and 19) there is
significant lack of fit, which may be due to a variety of causes, as already
discussed. We hope to resolve these cases when the more detailed analysis
is completed. Note that among these 10 cases the mean square for block'
effect is generally small when compared to the residual, and one might
anticipate that trend is not present in most cases.
4. Note that no long term drift is exhibited with respect to parameter 13,
where previously a short term drift had been indicated. These two results
are in conflict and our tentative conclusion is in favor of the NO drift
hypothesis.
B.4 OVERALL CONCLUSIONS
We have found some evidence of trend or drift; however, no clear cut pattern nor
explanation has yet emerged (except in the case of parameter 15). We will continue
to examine the data in an effort to achieve a definite result.
B-10
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Table B-4
VARIANCE ANALYSIS
Parameter
1
2
3
4
Source
of
Variation
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Degree
of
Freedom
8
1
7
7
23
8
1
7
7
23
8
I
7
7
23
8
1
7
7
23
Sum
of
Squares
3751.6805
11.8225
8.0692
23.9225
3799.4947
259944.3333
25.5569
486.6098
537.5
261194.0
623107.9167
34240 4457
2682.4709
628.1667
269843.0
15268.6142
11.0260
112.5382
71.8616
15464. 04
Mean
Square
11.8225 NS
1.1527 NS
3.4175
25.5569 NS
69.5156 NS
76.7857
3424.4457
383.2101 *
89.7381
11.0260 NS
16.0769 NS
10.2659
NS = Not Significant
* - Significant
B-ll
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Table B-4 (Cont.)
Parameter
5
6
7
8
9
Source
of
Variation
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Degree
of
Freedom.
8
1
7
7
23
8
1
7
7
23
8
1
7
7
23
8
1
7
7
23
8
1
7
7
23
Sum
of
Squares
429.3611
0.2067
2.1473
0.2950
432.0101
22.2290
0.0052
0.0053
0.0080
22.2955
635802.5833
1357.3664
1047. 5503
528.5
638736.0
547148.9167
181.0846
1153.4987
434.5
54819.0
0.85934
0.00513
0.03047
0.00576
0.9007
Mean
Square
0.2067
0.3067 *
0.0421
0.0052
0.00761
0.00114 *
1357.3664 *
149.6500 NS
75.5
181.0846 NS
164.7855 NS
62.0714
0.00513
0.00435 *
0.00082
NS = Not Significant
* = Significant
B-12
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Table B-4 (Cont.)
Parameter
10
11
12
13
14
Source
of
Variation
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Degree
of
Freedom
8
I
1
1
23
8
I
1
1
23
8
1
7
7
23
8
1
7
7
23
8
1
7
7
23
Sum
of
Squares
214540.9167
37.3333
1545.5833
259.1667
216383.0
410467.25
17.4934
1290.2566
329.0
412104.0
652365.0
0.0119
1255.4880
351.5
653972.0
11691.208
7.560
23.972
18.75
11741.49
3196.2622
0.2177
27.9736
4. 1498
3228.6033
Mean
Square
37.3333
220.7976 *
37.0238
17.4934
184.3224 *
47.0
0.0119 NS
179.3554 NS
50.2142
7.560 NS
3.4245 NS
2.6786
0.2177
3.9962 *
0.5928
NS = Not Significant
* = Significant
B-13
LOCKHEED MISSILES & SPACE COMPANY. INC.
-------
LMSC-D401598
Table B-4 (Cont.)
Parameter
15
16
17
18
19
Source
of
Variation
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Degree
of
Freedom
8
1
7
7
23
8
1
7
7
23
8
I
7
7
23
8
1
7
7
23
8
1
7
7
23
Sum
of
Squares
12.1617
0.4291
0.0677
0.0178
12.0673
3.7977
0.0131
0.7460
0.0117
4.5867
2345. 9945
29.1736
25.4453
9.5113
2410.1247
31159.8633
126.4612
89.9039
14.6816
31390.91
27095.1742
1.9201
110.1157
20.35
27207.56
Mean
Square
0.4291 *
Conir
0.00967 NS
0.00254
0.0131
0.10657 *
0.00167
29.1736 *
3.6350 NS
1.3588
126.4612
12.8434 *
2.0974
1.9201
15.7308
2.9071
(See
nent 2)
NS = Not Significant
* = Significant
B-14
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
LMSC-D401598
Table B-4 (Cont.)
Parameter
20
21
22
23
Source
of
Variation
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Factorial Model
Blocks After Factorial
Residual
Error
Total
Degree
of
Freedom
8
1
7
7
23
8
1
7
7
23
8
1
7
7
23
8
1
7
7
23
Sum
of
Squares
205760.6667
942.2433
517.4233
289.6667
207510.
76628.92
17.49
220.59
251.
77118.00
72.43
244. 82
101.
55383.00
274151.56
1668.16
700.59
191.67
276712.00
Mean
Square
942.2433 *
73.9176 NS
41.3810
17.49 NS
31.5128 NS
35.8571
72.43 NS
34.9743 NS
14.4286
1668.16 *
100.0843 NS
27.3814
NS = Not Significant
* = Significant
B-15
LOCKHEED MISSILES & SPACE COMPANY, INC.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
EPA-650/3-74-004a
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Study of Factors Affecting Reactions in Environmental
Chambers Phase II.
5. REPORT D.ATE
April 1974
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R. J. Jaffe, F. C. Smith, Jr., and K. W. Last
8. PERFORMING ORGANIZATION REPORT NO.
LMSC-D401598
9 PERFORMING ORG-VN I ZATI ON NAME AND ADDRESS
Lockheed Missiles & Space Company, Inc.
Sunnyvale, Calif. 94088
10. PROGRAM ELEMENT NO.
1AA008 - 21AKC-34
11. CONTRACT/GRANT NO.
68-02-0287
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
NERC-RTP, Chemistry and Physics Laboratory
Research Triangle Park, N. C. 27711 and
Coordinating Research Council, New York,N.Y.
13. TYPE OF REPORT AND PERIOD COVERED
Yearly - 1973-74
14. SPONSORING AGENCY CODE
10020
15. SUPPLEMENTARY NOTES
Phase I was issued as EPA-R3-72-016.
16. ABSTRACT
An experimental study has been conducted of effects of materials, spectrum, surface/volume
ratio (S/V), and cleaning technique on the photochemical reactions observed in a smog cham-
ber. Use of a unique chamber and lighting system permitted independent variation in chamber
materials and in light conditions. A xenon arc lamp-parabolic reflector combination provided
a collimated light beam. The study included four materials — aluminum, Pyrex, Teflon, and
stainless steel, and two conditions each of spectrum, S/V, and cleaning. All photochemical
runs were at kd of 0.3 min~'. The propylene (3 ppm)/NOx (1.5 ppm) reaction system was
used, at 95°F and 25-percent relative humidity. Initial NO2 was 10 percent of NOX« Cham-
ber background was <0.1 ppmC. A complete factorial testing sequence was performed.
Effects of the different materials and of the two levels of each parameter have been determined.
The time to NO2 maximum is shortest for stainless steel, followed by aluminum, Pyrex, and
Teflon. The cutoff spectrum (little energy below 340-nm wavelength) strikingly lowers reaction
rates compared to the full spectrum. Surface/volume ratio measurably affects reactions. The
cleaning technique does not cause large changes. The presence of this large spectral effect (at
constant k
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Leave blank.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COS ATI Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. &20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reverse)
------- |