United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
EPA-600/3-80-018
January 1980
Research and Development
Mechanism of
SO? and H2S(X
Aerosol
Corrosion
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1 Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-80-018
January 1980
MECHANISM OF S02 AND H2$04 AEROSOL
ZINC CORROSION
Alan B. Marker
Florian B. Mansfeld
Dennis R. Strauss
Dwight D. Landis
Rockwell International Science Center
Thousand Oaks, CA 91360
Contract No. 68-02-2944
Project Officer
Fred Haynie
Environmental Sciences Research Laboratory
Research Triangle Park, N.C. 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
DISCLAIMER
I his report has been reviewed by the Environmental Sciences Research
laboratory, ll.'.. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
v1t>w. and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
ri'commendat1on for usr.
11
-------
ABSTRACT
A twelve-month experimental study ha-, been conducted to establish the
physical variable1, controlling the S02 (gas) and H-SO. (aerosol) Induced
corrosion of zinc. The study was carried out using a (>.6 m aerosol llow
reactor 1n which relative humidity, temperature, air flow velocity, flow
turbulence, aerosol size range, and pollutant concentration were controlled.
Corrosion measurements wert made through the use of an atmospheric corrosion
monitor previously developed 1n this laboratory. The- results ol the study
showed that the principal factors controlling pollutant Induced corrosion
are relative humidity (time of surface wetness), the rate of pollutant, flux
to the surface, and the chemical form of the pollutant, while temperature was
not observed to be a controlling factor within the experimental ramie (12 to
20C). S02 was observed to Induce a higher corrosion rate 1n the /Inc than
H2S04 on a molecule for molecule basis. Flow dynamic measurements provided
bulk and s1/e detailed deposition velocities for two different accumulation
mode HoSC^ aerosol s1/e distributions as a (unction of frlctlonal velocity.
and a deposition velocity for S02 gas. The overall results Indicate that
under most ambient conditions $03 Induced corrosion damage will dominate over
I^SO^ effects. This study demonstrated the capability of the experimental
technique to observe the physical and chemical processes controlling pollu-
tant Induced corrosion and offers a means of quantitatively describing these
effects through further Investigation.
This report was submitted 1n fulfillment of (contract No. 68-02-2944) by
(Rockwell International Science Center) under the sponsorship of the U.S.
Environmental Protection Agency. This report covers a period from 5/9/78 to
6/9/79, and work was completed as of 6/9/79.
111
-------
CONTENTS
Abstract 111
Figures v1
Tables vii
1. Introduction 1
2. Experimental 4
Aerosol flow system 4
Gas/aerosol flow generation 4
Turbulent flow development 6
Test section - 8
Experimental approach 10
3. Results 12
Flow dynamic measurements 12
Aerosol size distribution 15
Deposition velocity determination 15
Corrosion measurements 20
High-low condition tests 23
NH4HS04 aeros°l generation 25
Corrosion chemistry 25
4. Discussion of Results 30
References 32
-------
FIGURES
Number Page
1 Size distribution of atmospheric particles^ ' ........... 2
2 Aerosol flow system ........................ 5
3 Aerosol generation system ..................... 7
4 Atmospheric corrosion monitor ................... 9
5 Velocity profiles observed as a function of the distance
from the wall for the smooth tube and the tube with per-
turbance barriers 20 cm from the ACM position. (R=R0 at
the wall ...................... ....... 13
6 HpSO/i aerosol particle size and mass distribution without
trie Cyclone separator in the flow stream ............. 16
7 h^SO^ aerosol particle size and mass distribution with
cyclone separator in the flow system ............... 17
8
Ratio of integrated XPS Sip s^9nal to integrated XPS Zngn
signal for H2S04~treated Zh discs (0.242 cm2 surface area) .... 21
9 Comparison of experimental aerosol deposition velocities
with those predicted by Sehmel (9), for a frictional velo-
city of 35 cm/sec 22
10 Comparison of the humidity response of the atmospheric
corrosion monitor with various surface pre-treatments 24
11 Output current of atmospheric corrosion monitor as a
function of S02 concentration 26
12 Comparison of observed ACM current response to H^SO.
aerosol deposition at two frictional velocities.
hLSO. concentration = 3.2 mg/m3 27
VI
-------
TABLES
Number Page
1 Frictional velocities and roughness heights 14
2 Experimental H^SO. aerosol deposition velocities 19
vn
-------
SECTION 1
INTRODUCTION
The degradation of materials through corrosion, accelerated by airborne
gaseous and particulate pollutants, is one of the major economic effects of
the increase in atmospheric pollution in urban areas. Primary concern in
this area is focused upon the role of sulfur species in the corrosion process
due to the increased use of high sulfur fuel oil and coal fuels in power
generation. This twelve-month laboratory study was conducted to provide
kinetic data required to develop a materials damage model to describe the
effects of pollutants upon atmospheric corrosion rates.
Industrial sources by far supply the bulk of pollutant sulfur in the
urban United States, with the sulfur being primarily in the form of gaseous
S02 with only about 1.25% of the sulfur being produced as ^$04 mist.U)
However, by direct and indirect photochemical as well as heterogeneous oxida-
tion processes, the gaseous S02 is converted to sulfates and incorporated
into atmospheric aerosols. The chemical form of the sulfate is to a large
extent controlled by the age of the aerosols and the gaseous concentrations
of neutralizing species such as ammonia. Newly-formed sulfate aerosols,
either from a primary source or from photochemical oxidation of $03, are
principally h^SOa, while aged aerosols have a greater sulfate salt content
with ammonium salts dominating.(2) Urban sulfate levels in the United States
are regional; however, levels from 3 to 25 #g/nr are typical except during
episode conditions, where levels of 80 Mg/nr and higher are reached.(3'
The size distribution of sulfate in aerosols determines the mechanism by
which they reach the surface of materials. Typically, ambient aerosols have
a size distribution similar to that shown in Fig. lv*»J with three distinct
particle size modes; the Aitken nuclei, accumulation range, and mechanically
generated particles. The bulk of aerosol sulfates (over 80%) is associated
with the accumulation mode aerosols - those particles with diameters between
0.1 and 1.0 micron.(5) These accumulation mode aerosols, formed primarily
from coagulation of and condensation on Aitken nuclei, are those with which
this study is concerned. This size range of aerosols is primarily removed
from the atmosphere by dry deposition on surfaces, incorporation into cloud
droplets (rainout), and removal by falling precipitation (washout).
Electrochemical corrosion occurs when pollutant sulfur species are
deposited on wet metal surfaces. Typically, surfaces are wet from condensa-
tion rather than precipitation, with condensation occurring when the relative
humidity adjacent to the surface exceeds the equilibrium value for a satur-
ated solution of the corrosion products or hygroscopic contaminants on the
1
-------
/ *
y ,,
\ \—
11 mi ill i ihn
0.001 0.01 O.I I 10
PARTICLE DIAMETER, microns
100
V "*-a*.
MOST C MASS
NUCLEATION
COAGULATION AND
CONDENSATION
SEDIMENTATION
Figure 1. Size distribution of atmospheric particles.
(4)
-------
surface. When the sulfur pollutants react with a surface, they can increase
the rate of corrosion through a number of processes. The initial chemical
reaction itself damages the outer passivated surface of the material and the
products formed can be hygroscopic which will increase the time of wetness of
the corroding surface. The reaction products can also increase the corrosion
rate by making the surface water layer more conductive and they can partici-
pate in corrosion reactions as catalytic reaction intermediates. The pre-
sence of acid pollutants also lowers the pH of the surface solution and in-
creases the solubility of the corrosion products.
In this investigation, research was carried out to determine the physi-
cal variables controlling the corrosion rate of zinc caused by exposure to
gaseous sulfur dioxide (S02) and aerosol particles containing sul fates in
fyHSC^- To provide the necessary experimental control
the form of IJ2S04 and
for observation of the corrosion process and to determine the rates of depo-
sition of both gaseous and parti oil ate sulfur pollutants, an aerosol flow
reactor was designed and developed specifically for this investigation. This
apparatus described in the next section provides a fully developed turbulent
air flow at a constant velocity, temperature, and relative humidity in which
both S02 and sul fate aerosols can be supplied to a test section containing
flow dynamic, aerosol, and corrosion monitoring instrumentation. The corro-
sion measurements were made through the use of an atmospheric corrosion mpn
tor (ACM) developed at the Science Center by F. Mansfeld and co-workers. (v-
The variables found to have the greatest effect upon the corrosion process
were relative humidity, pollutant concentration, wind fractional velocity,
and aerosol size range. The direct relationship observed between the flow
dynamic parameters and the corrosion rates supports the hypothesis that the
corrosion process was limited by flux of pollutants to the surface.
-------
SECTION 2
EXPERIMENTAL
**
AEROSOL FLOW SYSTEM
The flow reactor, designed and constructed during this program, is shown
schematically in Figure 2. The flow reactor was constructed in a 12'x20'
steel shed under the Science Center so that the main air supply system of the
building could be utilized to provide a high flow volume of conditioned air
and so that exhausting the test air would provide no health problems. The
flow system had four main sections; the gas/aerosol flow generation section,
the turbulent flow development pipe, the experimental test section, and the
aerosol scrubber and exhaust assembly. These sections will be described in
detail in the following.
GAS/AEROSOL FLOW GENERATION
The air inlet to the flow system as shown in Figure 2 was attached to
the main air conditioning supply for the laboratory with a high pressure
blower pushing the required 0.5 to 8.5 nr/min of air through the experimental
system. The air flow rate is controlled by an adjustable damper mounted on
the blower inlet and can be regulated at average flow velocities from 0.5 to
8 m sec"'. The average flow velocity was continuously monitored at the inlet
to the turbulent flow development pipe section by a pitot tube, and the abso-
lute flow velocity profile was determined across the diameter of the pipe by
a Thermosystems model 1610-2 velocity probe in the test section.
In the 30x30 cm ducting bringing the air into the blower are mounted
two water aerosol generators and six 500 watt radiant heaters to provide
initial humidity and temperature control. The flow rate through the water
aerosol generators can be controlled and the heaters can be individually
turned on or off. Two additional 600 watt heater units and a third water
aerosol generator were placed after the blower unit to provide further humid-
ity control. The entire ducting system was wrapped in 6-inch batts of fiber-
glass insulation to provide temperature stability.
Through the thermostatic control on the laboratory air conditioning
supply and the eight separate radiant heaters, it is possible to control the
air temperature in the flow system over the range 12 to 28 C with a tempera-
ture drift of less than ±0.5°C per hour. The relative humidity can be con-
trolled over the range 30 to 90% RH with stability of ±0.5% RH per hour. At
RH levels above 95% small fluctuations could cause rainout to occur in the
flow lines, hence RH values in the main pipe section were kept below 90%.
In the test section higher RH values were obtained by cooling the surface of
-------
AIR CONDITIONING DUCT
cn
ELECTROSTATIC
PRECIPITATOR
TEMP
TEST
SECTION
OPC
EAA
TOTAL FILTER
—SURFACE FLUX-
EM. OM and ESCA
-CORROSION DETECTOR
I—VELOCITY
• TEMP
VELOCITY
Figure 2. Aerosol flow system.
-------
the corrosion monitor by pumping refrigerated water through cooling lines
wrapped around the ACM base.
After the humidifying section the air was passed through a high effi-
ciency HEPA filter to remove any water aerosols and particulates in the air.
The conditioned air exiting from the HEPA filter then passed into a section
of ducting containing the mixing ports for test aerosols and gaseous SOo as
well as a Phys. Chemical Corp. Humeter, a Pi tot flow meter, and air tempera-
ture monitoring thermistor.
The SOo for the experiments was from a 0.5% Matheson Certified Standard
mixture of 502 in Ultra High Purity nitrogen. The gas was metered into the
main air flow through a calibrated rotameter.
The test aerosols were produced by an Environmental Research Corp. Model
7330 Fluid Atomization Aerosol Generator. The atomizer source provides a
poly-disperse aerosol distribution with a mean diameter of about 0.05 micron.
To provide larger aerosols with a narrower distribution, the output of the
atomizer was passed into a 22 liter cylindrical glass reservoir with a resi-
dence time of 3 minutes as shown in Figure 3. The reservoir provided enough
time for the aerosols to coagulate into the accumulation mode with the
resulting size distribution having a mean diameter of about 0.2 microns. The
output from the aerosol reservoir was passed through a cyclone separator
designed to remove particles with an aerodynamic diameter greater than 1 mic-
ron, and then into a beta radiation source to remove any residual static
charge the particles may have developed in the flow system. The particles
were mixed with the main flow stream to the apparatus at the outlet of the
HEPA filter.
The aerosol size distribution, total mass and number concentration were
the subject of extensive analysis with measurements being taken by three
types of particle analyzers and through microscopic examination of filter
samples. The aerosol analyzers used in the experiments were a TSI Model 3030
Electrical Aerosol Analyzer, a Royco Optical Particle Counter and a Califor-
nia Measurements Corp. Model C1004 Cascade Impactor equipped with Piezo-
electric microbalance particle mass monitors. Total mass concentration and
deposition rates were determined by weighing collected filters on a Cahn
microbalance and deposition rates were measured by transmission microscopic
examination of samples collected on TEM analysis grids and XPS analysis of
zinc samples. The results of these aerosol analyses will be discussed in
detail in the results section.
TURBULENT FLOW DEVELOPMENT
After passing through the HEPA filter and being mixed with the aerosol
flow, the air enters a 3.3 meter section of smooth 16.7 cm ID PVC pipe. In
this flow length (about 20 pipe diameters) turbulent flow is developed for
all velocities in excess of 0.2 m/sec using the Reynolds equation for criti-
cal velocities in a smooth pipe to describe the flow. To increase the air-
flow turbulence and give a range of frictional velocities for the experiments
perturbance barriers were placed in the pipe in some of the experiments. The
perturbances are 2 cm high Teflon rings with 1 cm square serrations cut into
-------
TO MA/A/
\
t.
(cvr TV/XT 0.
Figure 3. Aerosol generation system
-------
them. Up to three rings were used in the system with the serrations offset
to create maximum turbulence. The rings were spaced 6 cm apart and were
placed at distances varying from 2 to 20 cm upstream from the corrosion moni-
tor.
TEST SECTION
The test section of the apparatus is a removable aluminum pipe (60 cm
length, 16.4 cm O.D.,15.2 cml.D.) which is inserted into and joins the two
longer sections of PVC pipe. 0-ring-sealed ports in this section accommodate
velocity, temperature, aerosol, and corrosion monitoring detectors. The test
section is well grounded to reduce surface charging and prevent electrostatic
deposition of aerosol particles. It is wrapped with heating tape and insu-
lated with fiberglass batting. A YSI Model 72 temperature controller main-
tains constant wall temperature by controlling the current through the
heating tape.
Corrosion measurements were made in the test section through the use of
the atmospheric corrosion monitor (ACM) developed at the Science Center by
Mansfeld and co-workers(6,7) Tne ACM was mde up of twenty 0.62 mm zinc
plates separated by 0.025 mm Mylar film with alternate plates electrically
connected to make a capacitor circuit as shown in Figure 4. The assembly was
potted in epoxy and machined to fit into a gas-tight opening in the test sec-
tion. The ACM was AC biased to ±30 mv at a 1 cycle per minute while the
resulting current was continuously monitored. The temperature of the ACM was
controlled by the heating unit on the test section and cooling coils carrying
water from a Haake refrigeration system wrapped around the ACM's base. With
this cooling capability, the ACM's temperature, as recorded by a thermistor,
could be kept slightly lower than the air temperature to induce condensation
on the surface, if desired.
The YSI model 1610-2 velocity probe can be inserted into the test sec-
tion through either of two 90 degree offset, o-ring sealed ports and trans-
lated across the pipe to provide flow velocity profiles.
The test section is also fitted with o-ring mounted electron microscope
sample grid stages. These stages are flush to the wall of the pipe and
introduce minimal flow disturbance. Transmission electron microscope aero-
sol collection grids and zinc x-ray photoelectron spectroscopy samples were
mounted on these stages with silver paint.
Following the test section the flow pipe extends another 3.3 meters to
prevent interruption of the turbulent flow. Aerosol sample ports were
machined into the walls of this section to provide a continuous flow of
aerosols for the aerosol monitoring equipment and total filter collection.
At the end of this section the flow stream passes through a commercial
electrostatic precipitator to remove the sulfate aerosols before venting to
the ambient air.
-------
Figure 4. Atmospheric corrosion monitor.
-------
EXPERIMENTAL APPROACH
The nature and number of the experimental measurements being made during
each test run required that the flow system be maintained in a steady-state
with respect to flow rate, relative humidity, and aerosol concentration for
periods of 4 to 8 hours duration. To obtain this stability was difficult,
due to the high volume of air being passed through the system; however, as
experience was gained with the technique, the required stability of operation
was achieved. With steady-state conditions established, the following mea-
surements were performed for each complete test.
I) Environmental Measurements
a) % RH and temperature monitored in two points
b) Average flow velocity (Pitot tube)
c) Flow velocity profile recorded when steady-state established
II) Aerosol Measurements
a) Aerosol size distribution and number concentration determined at
intervals during test by TSI 3030 analyzer.
b) 2 total mass filter samples were collected
c) TEM deposition grid samples collected continuously
d) X-ray photoelectron spectroscopy samples (both zinc plate and alum-
inum foil) collected continuously during experiment.
Ill) Corrosion Rate Measurements were recorded continuously from the ACM
detector throughout each experiment.
Typically experimental conditions were selected over the following ranges
for the controllable variables.
Temperature 12 - 20 C
% RH 65 - 100%
Average Flow Velocity 0.5-8 m/s
SO,, concentration - 46 to 216 ppb by volume
3
SO. = aerosol mass concentration 1.2 mg/m
Aerosol size distribution 0.01 - 1.0 micron diameter
ACM pretreated with 0.1 N H2S04 or NH4HS04
The experimental procedure consisted of running the flow apparatus with
only humidified air to establish a steady-state RH and temperature and to
obtain a background current from the ACM. Upon obtaining a stable background,
S02 would be introduced to the gas flow and the ACM response recorded over a
two-hour period at which time the S02 would be turned off and the ACM allowed
to again reach a stable baseline. With a baseline again established, aero-
sols would be introduced into the flow and continued for 2 to 4 hours.
10
-------
At the completion of a test, the total filter samples, TEM grids, and
XPS samples would be removed from the test section and analyzed by the appro-
priate techniques.
11
-------
SECTION 3
RESULTS
The data from the flow dynamics, aerosol characterization, and corrosion
measurements will be discussed separately with the combined results being
treated later in this report.
FLOW DYNAMIC MEASUREMENTS
The initial concern in the design of the experimental apparatus was to
provide a well-character!'zed turbulent air flow to the measurement section.
Hence a long flow tube was utilized (6.6 meters) providing 22 diameters of
length both before and after the test section, and in all test cases the air
flow velocity was set above the critical velocity for turbulent flow (0.2
m/sec at 20°C) which is defined as:
V crit = 2000JL (I)
2
In this expression, v is the kinematic coefficient,of viscosity (1.5x10 m
sec'1 for air at 20°C) and d is the pipe diameter.w
The extent of the turbulence in the pipe is a function of the resistance
of the pipe wall to the flow. As the friction between the flowing material
and the pipe wall is increased (due to surface roughness or barriers) the tur-
bulence increases. This effect was observed in the experimental apparatus by
placing the serrated barriers described in the experimental section in the
pipe. Figure 5 shows two velocity profiles, one taken at 3.96 m/sec in the
smooth pipe, the other taken at 4.65 m/sec in the pipe with 3 barriers in
place, the closest one 21 cm from the test point.
To provide a measure of the turbulence, "frictional velocity" and
"roughness height" are defined for the flow. The frictional velocity (v*)
is related to the velocity profile by the expression
V(y) = v* In /y+y0\ (II)
where V(y) is the observed velocity at distance y from the pipe wall, K*" is
Von Karman's constant (0.417, ref 9) and y is the roughness height of the
pipe. The frictional velocity and roughness height parameters were obtained
for the flow conditions of the test runs by carrying out a least squares fit
of the observed velocity profiles to equation (II). As the results in Table
1 show, frictional velocities of 3.5 to 75 cm/sec were obtained in the flow
system with roughness heights of < 10"^ cm and ~ 0.2 cm respectively. The
correlation coefficients for the fit of the observed data to eq (II) were
12
-------
1.0
.8
.6
.2
o o
O te 3.96 n/sec
-Smooth Tube
.4 _
U,a=x4.65m/sec
Barriers in Flow
1 1
U.L
1
.2
1 1
.4
1 1
.6
1 I
.8
1 1
AXIS
1 - R/R,
Figure 5. Velocity profiles observed as a function of the distance from
the wall for the smooth tube and the tube with perturbance barriers 20 cm
from the ACM position. (R = R at the wall)
13
-------
TABLE 1. FRICTIONAL VELOCITIES AND ROUGHNESS HEIGHTS OF SMOOTH FLOW TUBE
AND TUBE WITH THREE SERRATED PERTURBANCES PLACED IN FLOW WITH THE FIRST PER-
TURBANCE BARRIER 20 CM FROM THE TEST POSITION. TABULATED VALUES ARE FROM A
LEAST SQUARES FIT OF THE EXPERIMENTAL VELOCITY PROFILE DATA TO EQUATION (II).
Maximum Flow Velocity
Vmax (cm/sec)
(Smooth 146
Pipe)
238
305
760
1,036
(Barrier 700
in flow)
425
289
155
Frictional
Velocity
v* (cm/sec)
3.5
8.6
9.1
22.3
39.3
75
50
33.8
18.8
Roughness
Height
Y0 (cm)
<10"3
_
< 10"
<10"3
<10"3
<10"3
0.16
0.26
0.21
0.18
Correlation
Coefficient
r2
0.93
0.94
0.95
0.96
0.92
0.82
0.87
0.81
0.97
14
-------
typically around 0.94 for the smooth pipe flow and between 0.80 to 0.90 for
the tests with the perturbance barriers in place. The fit between equation
(II) and the velocity profile data is shown in Figure 5 where the solid
curves are those predicted by the least squares fitting procedures.
AEROSOL SIZE DISTRIBUTION
The sulfate aerosols produced by the ERC fluid atomization aerosol gen-
erator were monitored by a number of aerosol analyzers during the experimen-
tal program to provide a size distribution and number concentration for the
airborne particulates. The ^504 aerosol size distribution of the direct
output of the ERC aerosol generator, as determined by the Thermosysterns
Electrical Aerosol Analyzer (EAA) and CMC Model C1004 Cascade Impactor (PPM),
showed about 90% of the particles to be below 0.1 micron in diameter with
only a small portion of the particles haying diameters as great as 1 M.
Since accumulation mode aerosols are typically 0.1 M to 1.0 M in diameter, a
22 liter coagulation chamber was added to increase the size of the small
particles. Figure 6 shows the effect of the addition of the coagulation
chamber as determined by EAA and PPM analysis. The particle size and mass
distribution data show the maximum of the number distribution to occur at
0.17 to 0.2 microns diameter with > 90% of the particles being in the accumu-
lation mode size range. Assuming the aerosol particles to be spherical, a
mass concentration distribution can be calculated and this is shown by the
filled circles in Figure 6 for the aerosol from the coagulation chamber. As
can be seen, the mass maximum is at 0.6 micron with a significant portion of
the aerosol mass being contained in the particles with diameters greater than
1.0 micron. To reduce the mass contribution of the particlesof>l M diameter
a cyclone separator was included in the aerosol flow system after the coagu-
lation chamber. The cyclone separator was designed to have a 50% cut point
at 0.96 Mm for a flow rate of 1.0 scfm. The result of adding the cyclone is
shown in Figure 7. The maximum of the number concentration is at 0.2 microns
with the maximum mass concentration occurring at 0.4 microns. For this dis-
tribution particles of diameter greater than 1 micron comprise less than 10%
of the total mass. The cyclone also lowers the total number concentration
significantly, bringing the total aerosol mass down about a factor of five
from the case without the separator.
To obtain gravimetric determinations of the total aerosol mass, known
volumes of air from the flow system were passed through preweighed Mi Hi pore
Spectro Grade A glass fiber filters for particle collection. The collected
samples were weighed on a Cahn microbalance at known humidity to determine
the aerosol mass.
DEPOSITION VELOCITY DETERMINATION
The deposition velocity K of gaseous and particulate pollutants is
defined by the relationship
2
I/ _ amount deposited/cm of surface-sec
airborne particle concentration above the surface (III)
This deposition velocity results from the combined effect of all the pro-
.15
-------
10"
105
104
•X
103
o
' ?
i
-------
10 r-
10J -
10H -
O
O
S-
0)
(-> Y>
ZJ O
Q.
a
o>
2- 0
<
. •>
OTJCO
1 (J
a.
a
< o
<
ioj
102
10
1.0
0.1
0.01
io-3
io-4
io-5
io-6
o
o
-
—
-
-
• •
••
-
1 1 1
0.1 1.0 10
Particle size (microns)
Figure 7. H^SO^ aerosol particle number and mass distribution
with cyclone separator in the flow system.
O- Number Distribution
• - Mass Distribution
17
-------
cesses contributing to deposition. To determine this value for the aerosol
and gaseous sulfur species in this study a combination of spectroscopic and
gravimetric measurements was used.
The flux of aerosols depositing onto the ACM surface at the bottom of
the flow reactor was observed by two methods; analyzing deposited samples
with a transmission electron microscope (TEM) and carrying out x-ray photo-
electron spectroscopy (XPS) analyses of sulfate deposition on zinc discs.
Gaseous $03 deposition was also characterized by XPS sulfate analyses.
The TEM measurements used an instrument equipped with a computer auto-
mated program for particle counting and sizing on standard grids. In the
experimental test section, graphite-coated TEM sample grids were mounted in
the plane of the ACM detector (bottom of the flow tube). After each test run
the grids were Pt-Pd coated at a fixed 30 degree deposition angle for study
in the TEM. The use of angular Pt-Pd deposition produces a "shadow" of the
individual particles which can be seen by the TEM to aid in calculating par-
ticle volumes. The primary goal of the TEM analyses was to establish the
dependence of the deposition velocities on the diameter of the aerosol parti-
cles and the frictional velocity of the flow system. Due to the complexity
of the analysis procedures TEM samples were not obtained over the range of
conditions used in this study. However, for the case of ^$04 aerosols depos-
ited at frictional velocities of about 35 cm/sec, adequate data was obtained
to describe the deposition velocities for the two aerosol distributions shown
in Figures 6 and 7.
The particle size dependence of the deposition velocities was calculated
from the TEM data by determining the total particle mass in each incremental
size range per unit area on the collection surface and dividing by the collec-
tion time and the corresponding airborne aerosol mass as shown by,
K_ = (Aerosol mass per unit area on surface in size range Dp) / (IV)
p (Aerosol mass airborne per unit volume in size range Dp)/
The total mass of aerosol particles observed on the surface and total
airborne aerosol mass can also be used with equation (III) to calculate a
bulk deposition velocity for the aerosol distribution. The TEM data from
three similar test runs for the aerosol distribution formed with the cyclone
separator (Fig. 7) were used to calculate the KDP at a flow velocity of Vmax=
305 cm/sec with the barriers in the flow (»>*=35 em/sec). Two similar test
runs were combined for the case with the cyclone separator removed from the
flow system (Fig. 6) again at Vmax=305 cm/sec and v*=35 cm/sec. The calcu-
lated values for K and KDQ for these distributions are shown in Table 2. The
bulk deposition velocity for the aerosol distribution without the cyclone is
significantly larger due to the contribution of the larger particles.
The XPS analyses of collected aerosol particles provided a second means
of determining the bulk deposition velocities. This was accomplished by com-
paring the sulfur signal amplitude observed on 7/32 inch zinc discs mounted
in the flow tube with sulfate standards prepared by spraying polished Zn discs
with 1.0 Ml aliquots of dilute HoSO/i solutions of known concentration (0.0028M
to 0.028M). The samples were allowed to dry overnight in room air, then
18
-------
TABLE 2. AEROSOL DEPOSITION VELOCITIES DETERMINED FROM XPS AND TEM ANALYSES
FOR THE TWO EXPERIMENTAL ACCUMULATION MODE PARTICLE SIZE DISTRIBUTIONS.
DISTRIBUTION I - FIG. 6 (NO CYCLONE SEPARATOR)
DISTRIBUTION II - FIG. 7 (CYCLONE SEPARATOR IN FLOW)
Deposition Velocities From XPS Measurements
Aerosol Distribution I
KH?SO/I = 0.07 ± 0.02 (cm/sec) for Vmav = 305 (cm/sec)
£ T1 max
v* = 35 (cm/sec)
KH2S04 = 0.10 ± 0.04 (cm/sec) for Vmax = 700 (on/sec)
v* = 75 (cm/sec)
Gaseous S02
KSO/i = °-93 * °-2 (cm/sec) V = 400 (cm/sec)
t * I *^ "
v* = 50 (cm/sec)
Deposition Velocities From TEM Measurements
Due to the small surface area examined in each TEM analysis (100 to
5000 A*m2) and the uncertainties in aerosol size distribution measurements,
an experimental uncertainty of a factor of 3 to 5 is set on the TEM based
results.
Particle Size Detailed Deposition Velocities
V = 305- cm/sec, v* = 35 cm/sec
1IICtA
Particle Diameter KQ (cm/sec) KQ (cm/sec)
Dp (microns) Distribution p
I II
0.1 0.002 0.001 0.002
0.2 0.0008 0.0008 0.0008
0.4 0.002 0.002 0.002
0.8 0.01 0.03 0.02
1.6 0.03 0.04 0.04
3.2 0.1 — 0.1
19
-------
analyzed by XPS. A standardization curve (Fig. 8) was constructed by plotting
the XPS peak area ratio $2p/Zn2p versus moles of h^SCty for each standardiza-
tion sample. Identical Zn discs were mounted flush with the walls of the
test section of the aerosol flow system and were exposed to measured concen-
trations of aerosol h^SC^ for a measured length of time. The XPS spectra of
each of these samples were measured and the S?n/Zn2p peak area ratio was cal-
culated. The number of moles of deposited h^stty for each sample was taken
off the standardization curve. The range where the deposited aerosol test
samples fell on the standard curve is also shown in Figure 8. The bulk flux
determinations combined with the total filter aerosol mass data were used to
calculate the deposition velocities for two frictional velocities. These
results, shown in Table 2, were for the aerosol size distribution without the
cyclone separator.
The potential for experimental error in determining the size distribu-
tion of the particles on the TEM grids, and the small surface area involved
in the analyses, leads to an estimated uncertainty in the deposition veloci-
ties calculated from the TEM data of about a factor of 3. The bulk deposi-
tion velocities obtained from the XPS analyses did involve greater sample
sizes and a means of quantitative calibration and provide a more accurate
determination. Comparison of the two determinations for KHpStty from Distribu-
tion I (Fig. 6) shows the XPS based value to be about 3.5 times greater than
that from the TEM data. A comparison between the TEM based detailed deposi-
tion velocities and the theoretical values calculated by G. A. Sehmel in
1972(9) from wind tunnel data on uranine particles (density 1.5 gr/cm^) is
shown in Fig. 9. The magnitude of the values for KDP agree quite well consi-
dering the uncertainty in the results, and the shape of the profile supports
Sehmel's model which predicts a minimum in the deposition velocity for par-
ticles in the.accumulation mode size range.
The deposition velocities for S02 on zinc as determined from XPS measure-
ments were obtained for concentrations of S02 between 100 and 200 ppb by
volume at flow velocities of 400 cm/sec, corresponding to a frictional velo-
city of about 50 cm/sec and a roughness height of about 0.2 cm. The results
for these conditions gave:
KSQ = 0.93 ± 0.2 cm/sec
Typical ambient grassland studies show S02 deposition velocities varying from
0.1 to 1.5 cm/sec for frictional velocities 20 to 40 cm/secno), and Owen and
Powell (1973) report that a K$02 value of 0.8 cm/sec is an average value over
land(H) for the British Isles.
CORROSION MEASUREMENTS
All corrosion measurements were made through the use of the atmospheric
corrosion monitor (ACM) described in the experimental section of this report.
This unit consisted of 10, zinc-to-zinc anode-cathode couples across which
current, (Ig), was monitored by a zero resistance ammeter as a function of
time. The polarity of the 30 mv bias voltage was reversed at 1 cycle per
minute. As described by Mansfeldv1Z), the relationship of Ig with the rate
20
-------
0.6
Zn
2p
0.5
0.4
0.3
0.2
0.1
AND S0
DEPOSITION MEASUREMENT
RANGE
I
0.01 0.02
M MOLES OF H2S04
0.03
0.04
FIGURE 8.. Ratio of integrated XPS S2 signal to integrated XPS
Zn2 signal for H2S04-treated Zn discs (0.242 cm surface area)
21
-------
o
UJ
o
o-
10'
10
10
-1
10
-2
10
-3
10
-4
10
-4
10
-3
10
-2
10
-1
10
PARTICLE DIAMETER (MICRONS)
Figure 9. Comparison of experimental aerosol deposition velocities
with those predicted by Sehmel (9).
• - Experimental points at frictional velocity
v*-35 cm/s
Smooth curves - Theoretical values at v*-72.6 cm/s
and at 34.1 cm/s
22
-------
of the corrosion process for a theoretical case in which the cathodic reac-
tion is under pure diffusion control is given by
Ic = kc T , (V)
where 7 is the corrosion rate and kc a constant. The observed ACM current Ig
is related to the actual corrosion current by Ic = K Ig (V), where k is a
constant composed of the ratio of Tafel slopes and the polarization resis-
tance. Hence the observed current Ig under controlled conditions should be
directly related to the instantaneous corrosion rate. Before each experiment
the surface of the ACM unit was mechanically polished on 600 grit corborundum
paper and the resistance of the background current in the absence of applied
electrolyte determined. The ACM was only used in a test if the observed
resistance was greater than 10 M ft. The freshly polished ACM surface was
then uniformly wetted with 0.7 ml of 0.01 N F^SOA electrolyte and was allowed
to dry overnight in an airflow at 60% RH to develop a uniform layer of corro-
sion products.
Various pretreatment electrolytes were investigated to determine their
relative sensitivity to humidity and sulfate deposition. Pretreatment elec-
trolytes studied included deionized water, 0.01 N HoSO^ 0.01 N NaoSO/p and
mixtures of 0.01 N HC1 and 0.01 N ^804. The results of these tests showed
that except for the case of deionized water, the surface pretreatment solu-
tions produce equivalent sensitivities when dried under equivalent conditions.
The deionized water produced very little corrosion products on the surface of
the ACM and very low (10~7 amp) current response when cycled by the potentio-
stat. The other surface pretreatments produced stable current levels at a
given level of relative humidity, with the humidity dependence of the current
being shown in Figure 10. As can be seen, the data on the log Ig versus % RH
has scatter but gives a fairly linear relationship for all of the pretreat-
ment electrolytes.(14) Since no advantage was found to having Na+ or Cl~
ions on the ACM surface, 0.01 N H2$04 was selected as the standard pretreat-
ment solution. Under standard test conditions where RH was normally main-
tained between 85 and 95%, the background corrosion currents due to the sur-
face pretreatment were about 1 microamp.
HIGH-LOW CONDITION TESTS
The experimental design called for a series of "High-Low" tests to be
conducted to determine the experimental parameters controlling the Zn corro-
sion rate. Tests with high and low levels for flow rate, % RH, pollutant
concentration, and turbulence established that the primary controlling para-
meters were relative humidity and pollutant flux.
Relative humidity was, as expected, a dominant variable. At RH levels
below the surface wetness point, where the surface corrosion products have
established an equilibrium with gaseous H20 producing a thin electrolyte film
on the ACM, no pollutant related current effects were observed. Aerosol
droplets falling on the ACM surface were no doubt causing point corrosion
effects while drying on the surface, but the absence of an electrolyte film
prevented any significant current changes from being observed.
-------
QC
OC
10
-6
10
-7
10
-8
PRETREATMENT
• - 0.0 IN H2S04
D-0.01 IN H2S04/0.01N HCL ||
A - 0.01 N Na2S04
40
50
60
80
90
100
Figure 10. Comparison of the humidity response of the atmospheric
corrosion monitor with various surface pre-treatments.
24
-------
The second controlling feature was the flux of pollutants to the surface.
Pollutant concentration and frlctlonal velocity are the two parameters des-
cribing pollutant flux, and both were found to have an effect upon the observ-
ed corrosion rate.
Pollutant concentration was obviously a controlling factor 1n that at
low pollutant levels (<10ppb of SO?), no enhancement of the corrosion
current was observed, while at higher levels of SO? and H2S04 the corrosion
current Increased with pollutant concentration. This can be seen in the Ig
versus time profiles observed for S02 in Figure 11, where both the rate of
Increase and the magnitude of the current are controlled by the S02 level.
That frictlonal velocity was affecting the corrosion current was more
difficult to clearly demonstrate 1n that over the range of frlctional veloci-
ties used 1n the corrosion tests (10 to 75 cm/sec) the total deposition velo-
city for the accumulation mode aerosols did not change greatly (Table 2).
However, as can be seen in Figure 12, there is an observable increase 1n the
corrosion current between the two test runs at v* = 30 and v* = 50 for com-
parable aerosol concentrations and % RH.
NH4HS04 AEROSOL GENERATION:
The ACM measurements demonstrated that under RH conditions where the ACM
surface had established a thin layer of electrolyte corrosion rate enhance-
ment due to SO? and ^SQ^ surface deposition could be observed. The test
runs using HN4HSOA aerosol did not produce any corrosion current increase.
Investigation of the aerosol size distribution and concentration showed that
the saturated solution of NH^SO*, though generating a large Initial concen-
tration of aerosols, was not producing high aerosol levels in the flow system.
A check of the flow system showed that the aerosols were rapidly agglomerat-
ing and forming salt crystals in the aerosol inlet tubes, allowing only small
concentrations of aerosols to enter the flow tube.
Examining the literature on generation methods for NH^HSO/i aerosols
revealed that other studies had also encountered difficulties in generating
the ammonium add sulfate aerosols. An extensive study by R. Zygmunt'")
showed that the blsulfate aerosols react in air to form other sulfate forms
Including letoviclte, (HN^oHfSO,^)?. and ammonium sulfate, (NH^SC*. The
combination of the agglomeration phenomena and the neutralization of the
aerosols while airborne discouraged any further tests with these aerosols.
CORROSION CHEMISTRY
The ACM current, Ig, observed during the S02 and ^SO^ exposure experi-
ments showed different time profiles for the two pollutants as can be seen
1n Figures 11 and 12. The S0£ exposure experiments showed the current to
build up towards a maximum limit, while H2$04 exposure produced a more linear
monotonlc increase 1n Ig as a function or time. In both cases when the
pollutant flow was discontinued the ACM current dropped back towards a base-
line value Indicating that the sulfur pollutants reacted to completion. XPS
analysis of the chemical composition of the surface corrosion layer produced
25
-------
CTi
15
.0
o
o
ex:
on
o 5
s:
o
-------
.ro
UJ
V 2
O
<
a:
oc
o
o
AEROSOL OFF
\
p*=50cm/sec
O
AEROSOL ON
v*=30cm/sec O
O OO
* • o°
000°
I I I I I I
I
I I I 1
100
200
TIME (MINUTES)
Figure 12. Comparison of observed ACM current response to
velocities. ^$04 concentration = 3.2 mg/m3.
300
aerosol deposition at two frictional
-------
by exposure to S02 and H2SC>4 snowed that in both cases the surface reactions
lead to the formation of zinc sulfate. The XPS spectra showed all the sur-
face Zn to be in the Zn2+ state with S042~ being the only form of sulfur on
the surface.
Since in the examples shown in Figure 11 and 12 the deposition rate of
S02 is a factor of two greater than that for H2S04 and Ig for S02 was ap-
proaching a steady-state limit more rapidly than the l^Stty, it is reasonable
to assume that the SO? was reacting more rapidly in the surface electrolyte
than the sulfuric acid. The steady increase in the ^$04 induced current
with time indicates that the reaction was proceeding at a rate slow enough
such that the concentration of sulfuric acid in the surface electrolyte was
continuously building up. Further insight into the rates of the two reaction
mechanisms can be obtained from the half-life of the decay of the observed
ACM current when the pollutant flow was turned off. The average of all the
observed cases showed the S02 induced current to drop to one half of its
maximum value above the base line in 15 ± 1 minutes, while the H2S04 case
required 30 ± 9 minutes.
If as suggested by equation (V) the observed ACM current is directly
related to the instantaneous corrosion rate, the area under the Ig versus
time profiles normalized for the deposition of pollutants should provide a
means of comparing the relative number of electrons involved in the SC"2 and
H2S04 induced corrosion processes. The average of this total charge value
for two S02 tests and three HoSC^ tests run under identical flow dynamic
conditions, using the bulk XPb deposition velocities for normalization,
yielded
Total Charge SCL = 3.64 ± 0.8 x 10 coulombs/mole
Total Charge H2S04 = 4.6 ± 1.5 x 10 coulombs/mole.
These total charge values are consistent with the two reaction mechanisms
involving the transfer of the same number of electrons between the zinc
metal and the electrolyte; however the experimental uncertainty in the depo-
sition velocity values prevents any conclusive statement.
The S02 adsorbed into the surface electrolyte was ultimately converted
to sulfate, hence the net reaction would be
(1) S02 + 02 + 2e~ -» S042" ,
where the zinc metal could supply the electrons and the observed ACM current
through
(2) Zn -* Zn2+ + 2e~ .
The reaction of the sulfuric acid aerosols with the zinc would be expected
to proceed through the liberation of gaseous hydrogen as shown by
(3) H2S04 + H20 -» 2H30* + S042"
28
-------
(4)
2H/ +
Zn -» H2 (gas)
Zn
2+
29
-------
SECTION 4
DISCUSSION OF RESULTS
The results of this study confirm that the principal factors controlling
the rate of sulfur pollutant induced corrosion of zinc are relative humidity,
pollutant flux to the surface, and the chemical form of the pollutant. No
corrosion was observed to proceed on the zinc surfaces at relative humidities
below that required to establish a layer of electrolyte, while at RH levels
high enough to produce a "wet" surface the corrosion rate was primarily a
function of the concentration of the sulfur pollutant in the thin electrolyte
layer and the chemical nature of the pollutant. The results indicate that
S02 induced corrosion of zinc proceeds at a rate approximately a factor of
two greater than that for the equivalent amount of deposited sulfuric acid
aerosol while ammonium bi-sulfate induced corrosion was not observed for the
levels of aerosols generated in this study. The experimental data also
supports chemical mechanisms for the S02 and ^SO^ induced corrosion of zinc
which involve the same number of electrons being transferred from the zinc
metal per molecule of pollutant.
The flow dynamics data provided a measurement of the particle size de-
pendence of the accumulation mode aerosols at 35 cm/sec frictiona! velocity
which support the general theoretical treatment of Sehrnel^) showing atmos-
pheric deposition velocities to be at a minimum for aerosol particles in this
size range (0.1 to 1.0 micron diameter). The aerosol deposition measurements
showed the two experimental accumulation mode aerosol size distributions to
have bulk deposition velocities of less than 10~1 cm sec~l at a v* of 35 cm
sec'1 versus the 0.9 cm sec'l value observed for S02 at a frictional velocity
of v* = 50 cm sec~'. These deposition velocities can be used with ambient
levels of SOo and ^SO^ aerosols to obtain an approximate assessment of their
relative contribution to zinc corrosion.
As an example, the Los Angeles basin in the summer and early fall months
has S02 levels ranging from 1 to 100 ppb by volume with a typical level for
an hourly average being around 20 ppb. During the same period, background
levels of 10 to 20 Mg/nr of S042~ are normal. Using 20 ppb of S02 and 15
jug/m3 of 50^2- as ^$04 (though much of the S042~ would be in the form of
ammonium sulfate) the experimental deposition velocities predict deposition
rates of 0.18 Mg cm'2 hr'1 for S02 and <0.006Mg cm'2 hr~' for ^$04 aerosols
on a horizontal surface. The actual relative contribution of these pollu-
tants to enhanced materials damage will of course be a direct function of the
ambient conditions in a given area. However, it appears that SOg induced
damage will dominate over f^SO^ effects in most urban areas.
In order to quantitatively relate the observed ACM current to the actual
30
-------
corrosion rates and to assess the combined effects of various pollutants on
the corrosion of ambient materials, further experimental efforts will be
required. The experimental technique developed throughout this study has
shown itself capable of observing both the physical and chemical variables
controlling pollutant induced corrosion and offers a unique means of deve-
loping an in-depth understanding of this area through continued research.
31
-------
REFERENCES
1. Robinson, E., and R. C. Robbins, Sources, Abundance, and Fate of Gaseous
Atmospheric Pollutants. Stanford Research Institute, Menlo Park, CA,
Project PR-6755, prepared for American Petroleum Institute, New York,
N.Y. (Feb. 1968).
2. Charlson, R. J., D. S. Covert, T. V. Larson and A. P. Waggoner. Chemical
Properties of Troposheric Sulfur Aerosols. Atm. Environ. 12 pp 39-53
(1978). ~
3. Friend, J. P. The Global Sulfur Cycle. Chemistry of the Lower Atmos-
phere. Ed. S. I. Rasool (Plenum Press, N.Y., N.Y., 1973) pp 177-201.
4. Whitby, K. T. The Aerosol Size Distribution of Los Angeles Smog. J.
Colloid Interface Sci. 39, pp 177-204 (1972).
5. Position Paper on Regulation of Atmospheric Sulfates. EPA-450/2-75-007.
U.S. Environmental Protection Agency, Research Triangle Park, N.C. pp 2,
13, 22-24, 27-34, 36 (Sept. 1975).
6. Mansfeld, F. and J. V. Kenkel. Electrochemical Monitoring of Atmospheric
Corrosion Phenomena. Corrosion Science. 16, pp 111-122 (1976).
7. Mansfeld, F. B. Electrochemical Studies of Atmospheric Corrosion. Final
Report Contract No. N00014-75-C-0788 prepared for Office of Naval Re-
search (1979).
8. Owers, E. and R. C. Pankhurst. The Measurement of Air Flow. Pergamon
Press, N.Y., N.Y. 1977) pp 77.
9. Sehmel, G. A. Particle Eddy Diffusivities and Deposition Velocities For
Isothermal Flow and Smooth Surfaces. Aerosol Science. 4., pp 125-138
(1973).
10. Proceedings of the International Symposium on Sulfur in the Atmosphere.
Held in Dubrovnik, Yugoslavia September 7-14, 1977. Published in Atmos.
Environ. 12., pp 14 (1978).
11. Owers, M. J. and A. VI. Powell. Deposition Velocity of Sulfur Dioxide on
Land and Water Surfaces Using a 35S Tracer Method. Atmos. Environ. 8_,
pp 63-67 (1974).
12. Mansfeld, F. and J. V. Kenkel. Electrochemical Measurements of Time-of-
Wetness and Atmospheric Corrosion Rates. Corrosion 33, pp 13-16 (1977).
13. Zygmunt, R. W. A Study of the Production of Well-Character!zed Aerosols
Using a Flow Reactor. Argonne National Lube report #ANL-77-90, (1977).
14. Mansfeld, F. and J. V. Kenkel. Atmospheric Corrosion Rates, Time-of-
Wetness and Relative Humidity. Werkstoffe and Korrosion. 30, 38 (1979).
32
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-80-018
3. RECIPIENT'S ACCESSION- NO.
4. TITLE AND SUBTITLE
MECHANISM OF S02 AND H2S04 AEROSOL ZINC CORROSION
5. REPORT DATE
January 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Alan B. Marker, Florian B. Mansfeld, Dennis
R. Strauss and Dwight D. Landis
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Rockwell International Science Center
Thousand Oaks, CA 91360
10. PROGRAM ELEMENT NO.
1AA6Q3A AE-Q29 (FY-79)
11. CONTRACT/GRANT NO.'
68^02-2944
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory—RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This study established the physical variables controlling the S02 and H2S04 induced
corrosion of zinc. Relative humidity, temperature, air flow velocity, flow turbu-
lence, aerosol size range, and pollutant concentration were controlled. Corrosion
measurements were made through the use of an atmospheric corrosion monitor. The
results showed that the principal factors controlling pollutant induced corrosion are
relative humidity, the rate of pollutant flux to the surface, and the chemical form
of the pollutant. S02 was observed to induce a higher corrosion rate in the zinc
than h^SOa on a molecule for molecule basis. Flow dynamic measurements provided bulk
and size detailed deposition velocities for two different accumulation mode H2S04
aerosol size distributions as a function of frictional velocity, and a deposition
velocity for S02 gas. The overall results indicate that under most ambient condi-
tions S02 induced corrosion damage will dominate over H2S04 effects.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*Air pollution
*Aerosols
*Sulfur dioxide
*Sulfuric acid
*Zinc
*Corrosion mechanisms
13B
0.7D
07B
11M
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report!
UNCLASSIFIED
21. NO. OF PAGES
41
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
33
------- |