ASM Handbook, Volume 13A: Corrosion: Fundamentals, Testing, and Protection Copyright © 2003 ASM International®

S.D. Cramer, B.S. Covino, Jr., editors, p196-209 All rights reserved.
DOI: 10.1361/asmhba0003606 www.asminternational.org

Atmospheric Corrosion
Lucien Veleva, CINVESTAV-IPN, Yucatan, Mexico
Russell D. Kane, InterCorr International, Inc.

ATMOSPHERIC CORROSION is the degra- cally stable state. This fact drives metals to con- Cathodes are the metal sites with a lower en-
dation and destruction of metallic materials and vert into corrosion products having a chemical ergy state, for example, inert non-metallic inclu-
their structure and properties due to interaction composition similar to that of the original ores sions and lower active-metal phases or struc-
with the terrestrial atmosphere at its character- that are in a more thermodynamically stable tures. The cathodic reaction occurs on these sites
istic air temperature, humidity, air chemistry, and state. When metals come in contact with the at- and involves the reduction of an oxidizing agent,
climatic values (Ref 1–13). Atmospheric corro- mosphere (oxygen) and water (moisture) in the such as air, oxygen, or hydrogen ions.
sion is distinguished from the corrosion of met- presence of corrosive species such as chlorides An electrolyte, such as moisture, comes in
als exposed to high temperatures in an absence or sulfur dioxide, the corrosion process starts, contact with the metal surface. The moisture
of moisture on the metal surface (dry or gaseous and corrosion products such as oxides, hydrox- contains dissolved ionic species (atmospheric
corrosion), which does not correspond to the ter- ides, or oxyhydroxides are formed. Table 1 pollutants) and is a good ionic conductor that can
restrial atmospheric humidity and temperature. shows the tendency for corrosion of some metals sustain electrochemical reactions.
The mechanism of high-temperature corrosion is as a function of the energy required for their sep- An oxidizing agent, such as oxygen and hy-
a chemical corrosion, and it is quite different aration from ores. drogen ions (HⳭ), is necessary for accepting the
from atmospheric corrosion. (See the article electrons emitted from the metal in the anode
“High-Temperature Gaseous Corrosion” in this reaction (Eq 1):
Volume.) Elements of the Process
The majority of metal structures and equip- Oxy Ⳮ eⳮ ⇒ Redform (Eq 2)
ment are exposed to terrestrial air conditions to Atmospheric corrosion is an aqueous process,
some degree and therefore can suffer from at- and its mechanism is electrochemical. There is ⁄2O2 Ⳮ H2O Ⳮ 2eⳮ ⇒ 2OHⳮ
1
(Eq 3)
mospheric corrosion. In some severe cases, the not only a transfer of mass during the chemical
metal can be completely destroyed and con- reaction but also an interchange of charged par- 2HⳭ Ⳮ H2O Ⳮ 2eⳮ ⇒ H2 F (gas) (Eq 4)
verted to corrosion products. With background ticles (electrons and ions) at the interface of the
knowledge of the principal exposure conditions metal (an electronic conductor) and the electro- where Oxy is the oxidizing agent, and Red is the
and their influence on metal corrosion, most se- lyte (an ionic conductor). The transfer of elec- reduced species.
rious corrosion problems can be prevented. It is trons (flow of electric current) occurs because of Corrosion Reactions. Figures 1 and 2 illus-
recognized that several industries face difficult the formation of a galvanic corrosion cell on the trate the corrosion reactions (oxidation and re-
corrosion problems because of very aggressive metal surface. Three elements are necessary for duction) that occur on the metal surface at the
atmospheres, including the electrical power the cell operation: anode and cathode sites, an metal-electrolyte interface and the movement of
transmission and distribution industry, chemical/ electrolyte, and an oxidizing agent. electrons from the anodic to cathodic sites. The
petrochemical production plants and equipment, Anode and cathode sites form multiple cor- net electric current is zero, because the electrons
aircraft, automotive manufacture and associated rosion cells. liberated during the oxidation of the metal (Eq
components, transportation infrastructure, off- Anodes are the areas on the metal with a 1) are accepted by the oxidizing agent in the
shore structures and equipment, the construction higher energy state, due to various factors such cathodic (reduction) reaction (Eq 2). In Fig. 2,
industry, and electronic devices. as inhomogeneous metal composition, grain the oxidizing agent is the HⳭ ion (cation), a prin-
Atmospheric corrosion occurs spontaneously boundary, multiple metallurgical phases, local cipal ion in the moisture formed on the metal
but may be slowed, prevented, and controlled but metal defects, and nonuniform metal treatments. surface exposed to industrial and urban atmo-
never stopped. The reason is that the commonly The oxidation corrosion reaction is done on the spheres. Typically in these sites, the atmosphere
used metals are not in a pure state in the earth, anodic sites: may be contaminated by SO2, which can be con-
except for some noble metals. Metals usually are verted to H2SO3 and then to H2SO4 (sulfuric
in ores, chemical compounds that include oxy- Me ⳮ neⳮ ⇒ MenⳭ • (mH2O) (Eq 1) acid) in the presence of moist air. When this oc-
gen, hydrogen, and sulfur. These mineral com- curs in the atmosphere, acid rain is formed,
pounds are the thermodynamic steady state of where Me is the metal. The metal is dissolved at which is a severe environment for metal struc-
the metals, in which Gibbs free energy (DG) has the anode to form cations (positively charged tures. The movement of the electrons from an-
a minimum value. For the separation of the met- ions). These may originally appear as metal hy- odic to cathodic sites in a metal is a result of the
als from their ores and for metallurgical and drated ions (MenⳭ • (mH2O) but subsequently difference in Gibbs free energy between the an-
manufacturing processes, energy, in the form of convert into oxides and hydroxides, the metal ode (higher level) and the cathode (lower level).
heat, chemical, electrical, or mechanical, ele- corrosion products. For example, steel atmo- This results in a potential difference between
vates the metal to a higher energy level. The spheric corrosion products typically include ␣- both reaction metal sites, which yields a current
metal product is not in its most thermodynami- and c-FeOOH as main constituents. flow from the anodic to cathodic sites. Detailed
 Atmospheric Corrosion / 197

Table 1 Position of some metals according information about the possible corrosion reac- rine-coastal environments. Localized attack of
to their standard electrode potentials in tion (and their metal potential values) as a func- some aluminum alloys, such as those containing
aqueous solutions at 25 C (77 F) in V tion of aqueous electrolyte concentration and pH copper, can take the form of layered corrosion,
(versus NHE)(a) (acidity or alkalinity) in the presence of certain exfoliation, detachment, and deformation, of
Standard electrode potential
ions (atmospheric contamination) can be found thin layers within the metal surface when ex-
Metal at 25 C (77 F), V using Pourbaix diagrams (Ref 13). These dia- posed to coastal environments.
Higher excess of free energy (very high corrosion grams are a useful tool for any corrosion engi- Localized atmospheric corrosion can also be
tendency) neer and scientist in evaluating and understand- observed on the surface of brass and copper-zinc
Potassium ⳮ2.92 ing the conditions that lead to specific corrosion alloys due to the reaction of the distinct alloying
Magnesium ⳮ2.34 reactions and their associated corrosion prod- metals in contact with the environment. In this
Beryllium ⳮ1.70 ucts. case, the corrosion is referred to as selective cor-
Aluminum ⳮ1.67
Manganese ⳮ1.05
Types of Atmospheric Corrosion Attack. rosion (Fig. 3c). Some metals or alloys can be
Zinc ⳮ0.76 Atmospheric corrosion can occur in two basic susceptible to localized attack that forms at lo-
Chromium ⳮ0.71 forms: uniform (general) and non-uniform (lo- cations of distinct phases on the grain bound-
Iron ⳮ0.44 calized) attack. Uniform corrosion results at a aries. This corrosion is recognized as intergran-
Cadmium ⳮ0.40
Cobalt ⳮ0.34
similar corrosion rate over the metal surface and ular corrosion (Fig. 3d). An example is the
Nickel ⳮ0.27 has the same appearance throughout (Fig. 3a). corrosion in cast iron, which occurs around the
Tin ⳮ0.25 Uniform attack is typical for atmospheric cor- boundary of the ferritic phase or at carbides in
Lead ⳮ0.14 rosion of steel and copper. Localized corrosion grain boundaries of stainless steels. The atmo-
Copper 0.34
Silver 0.80
usually occurs at small and specific locations on spheric corrosion process can also be increased
Palladium 0.83 the metal surface where the corrosion process is when two or more different metals are in direct
Platinum 1.2 focused, resulting in local acceleration of the contact in a structure. This metal coupling allows
Gold 1.42 corrosion rate (Fig. 3b). This type of corrosion the formation of a galvanic corrosion cell having
Lower excess of free energy (low tendency for corrosion) attack is referred to as pitting corrosion and can different electromotive force (voltage), depend-
be observed on aluminum and its alloys, zinc ing on the potential values of the metals in con-
Note: The excess of free energy is related to the standard electrode
(metal) potential value. (Complete metal electrode potential values can
(hot dip zinc or electrodeposited zinc on steel), tact (Table 1) (Fig. 4).
be found in Tables of Standard Electrode Potentials, G. Milazzo and S. stainless steels, nickel, and other metals. It is of- A very dangerous type of atmospheric corro-
Caroli, Ed., Wiley-Interscience, 1977.). (a) NHE, normal hydrogen elec- ten induced by the presence of chloride ions, sion attack is metal cracking, which can occur
trode ⳱ SHE, standard hydrogen electrode with hydrogen ions at unity
activity/concentration (a ⳱ 1,aqueous) which can be found in airborne salinity in ma- when a metal structure such as a bridge is ex-

Fig. 1 Schematic presentation of corrosion metal cell

formed by anodic (A) and cathodic (C) sites. The
A sites (Me2) have a more negative potential (E) relative to
that of the C sites (Me1).

Fig. 2 Schematic presentation of the corrosion galvanic

cell created in a zinc-copper alloy in an acid en-
vironment. The cathode is the copper-rich phase and the
anode is the zinc-rich phase. The corrosion attack is selec- Fig. 3 Schematic presentation of cross sections of several forms of corrosion attacks. (a) Uniform. (b) Nonuniform
tive to the zinc-rich phase. (localized). (c) Selective. (d) Intergranular. C, cathodic areas (Me); A, anodic areas between the metal grains
198 / Forms of Corrosion

posed to a corrosive environment and continuous by a thin electrolyte layer containing air contam- face, that can act as a physical barrier for oxygen
or cyclic mechanical loading. This combination inants and during which the corrosion cell can diffusion to cathodic sites and that results in a
leads to surface or internal microcrevices, fis- operate (Ref 19, 22, 25–29). The TOW is usually decrease in the rate of corrosion.
sures, and cracks that result in stress-corrosion calculated in hours, according to International The environmental corrosion aggressiveness
cracking (under relatively constant loads) or fa- Organization for Standardization (ISO) 9223, category of an atmosphere can be assigned based
tigue corrosion (under cyclic deformation). “Corrosion of Metals and Alloys, Corrosivity of on the annual TOW value according to ISO
Atmospheres, Classification,” and includes the 9223. However, this procedure is adequate to use
daily temperature/relative humidity (T-RH) only in an atmosphere free from chloride. In the
Atmospheric Parameters complex, using 80% as a critical RH value for T presence of chlorides, the deposition of hygro-
and Their Influence ⱖ 0 C (32 F), when the condensation starts on scopic contaminants (for example, chloride salts
the metal (Ref 1). Above RH 90% and T  25 in marine-coastal regions) occurs on the metal
A variety of atmospheric factors, climatic con- C (77 F), the dewpoint is reached, and the surface. This lowers the critical relative humidity
ditions, and air-chemical pollutants determines moisture formed on the metal surface is visible. value (RHc), and corrosion can start at RH as low
the corrosiveness of the atmosphere and contrib- The wet layer is actually thicker than that formed as 40 to 50% (Ref 1). This fact implies that in
utes to the metal corrosion process in distinct by initial condensation. This change of the mois- a marine-coastal environment, the higher con-
ways (Ref 1, 7, 8, 12–33). ture layer thickness, in turn, induces an alteration centration of chlorides can increase the real
Climatic characteristics play a major role in in the metal corrosion rate. The thinner layer of TOW in a zone even far from the shore (Ref 25).
the atmospheric corrosion process. To fully un- moisture is a minor barrier for the diffusion of The development of corrosion and TOW has
derstand atmospheric corrosion, it is important molecular oxygen from the environment. The been detected on samples exposed to the open
to properly describe and characterize the envi- thin aqueous layer can be practically saturated in atmosphere in the Antarctic when the tempera-
ronment that causes metal degradation. Factors dissolved oxygen; thus, the corrosion rate of the ture is below 0 C (32 F). Reduction of the RHc
and the interaction between them that need to be metal is actually more rapid in the thinner layer value can be produced by deposition of ammo-
considered are sun radiation, air temperature, formed by first condensation than in the rela- nium sulfates on the metal surface, which are
relative humidity, air chemistry, precipitation, tively thick layer formed at higher RH. known to accelerate corrosion and provide sul-
winds, and the mechanical and chemical action The rain is a climatic factor that also contrib- fate ions. It was confirmed (Ref 29) by measur-
of natural forces such as sand and rock particles, utes to moisture formation on a metal surface, ing TOW, using a copper-gold sensor according
soil dust, volcanic dust, organic matter, and in- but it can have additional effects. These include to ASTM G 84, that the annual TOW value was
dustrial dust. Also, various physical, chemical, dilution and washing of the corrosive pollutants as large as 8500 h at a location 20 to 30 km (12
and biological factors, including manipulation of deposited on the metal surface. This situation re- to 18 miles) from the seashore in a marine-
the environment as may occur in many engi- sults in a decrease of the corrosion rate, even coastal atmosphere in a humid tropical climate.
neering applications, must be considered. Such when the TOW is prolonged. Precipitation can This value is twice the value of 4500 to 4800 h
factors can directly affect the corrosion rate of also dissolve some metal corrosion products sol- for a typical rural-urban environment. This fact
metals exposed in outdoor or indoor atmo- uble in water (zinc carbonate and hydroxide, for can explain the different corrosion rates ob-
spheres. The atmospheric corrosion process can example). A fresh metal surface will be in closer served for standard metals (low-carbon steel,
be further complicated and accelerated when contact with the atmosphere, resulting in an in- copper, zinc) when exposed to these atmo-
micro- and/or macroorganisms are present. In crease in the corrosion rate. This situation is in spheres. This difference in TOW values is due
humid tropical and subtropical climates, micro- contrast to the compact and well-adhered cor- mainly to specific changes in their daily T-RH
bial corrosion or biocorrosion is commonly ob- rosion product layer, formed on the metal sur- complex. Because of this, it is recommended that
served.
When studying the atmospheric corrosion of
engineering materials, the most important factors
related to the climate and its effect on that ma-
terial are represented by a combination of:
● Temperature (T) and relative humidity (RH),
often described as the temperature-humidity
complex (THC). Humidity is a measure of the
amount of water vapor in air, and relative hu-
midity is the ratio between absolute humidity
and its saturation value, expressed in per-
centage. This percentage is a reverse function
of the temperature (T); the RH increases
while the T is decreasing, and vice versa.
● Annual values of pluvial precipitation (PP)
● Time of wetness (TOW), during which mois-
ture exists on the metal surface, and corrosion
may occur. This moisture layer on the metal
surface can be generated by rain, fog, snow,
dew condensation, and capillary condensa-
tion.
Standards that are useful in characterizing the
environment, as well as atmospheric corrosion
test standards, are listed in Table 2.
Time of Metal Wetness. In recent years, this
parameter has received special attention, because
it is the fundamental parameter that relates to the
time during which the metal surface is covered Fig. 4 Schematic presentation of corrosion reaction in galvanic coupling of zinc and platinum
 Atmospheric Corrosion / 199

the daily T-RH complex be used rather than the It is recognized that chlorides (airborne salinity) on metals exposed to atmospheres free of chlo-
annual T and RH average. and sulfur dioxide (SO2) are the principal pol- ride include copper (Cu2O, copper patina), alu-
Figure 5 presents the daily T-RH complex of lutants that can accelerate the atmospheric cor- minum (Al2O3), and zinc (ZnO and Zn(OH)2).
two atmospheres, marine-coastal and rural-ur- rosion rate by several orders of magnitude. A second aggressive environmental pollutant
ban, both part of a tropical humid climate. It can The principal source of chlorides is aerosols, for metals is SO2 gas. It is found in urban and
be seen that the corrosion cell can work almost which are suspensions of small liquid or solid industrial atmospheres and, in the prescence of
all day in the marine-coastal environment at rela- particles in the atmosphere that come from salt oxygen, is easily converted to sulfuric acid in the
tively constant T and RH values (due to the sea spray and salt fog in the vicinity of the seashore condensed moisture layer on the metal surface.
thermodynamic buffer capacity), while in the ru- and from the contaminated environment around The sulfuric acid dissociates to give HⳭ ions
ral-urban atmosphere, the corrosion cell is inter- industrial plants producing hydrogen chloride (H2SO4 } 2HⳭ Ⳮ SO2ⳮ 4 ), which participate as
rupted during the daily hours and starts again and sodium hypochloride. Chloride ion (Clⳮ) is the oxidizing agent in the cathodic corrosion re-
when the RH reaches the critical value (ⱖ80%) one of the principal environmental agents that action (Eq 2, 4 and Fig. 2). Due to the presence
for formation of moisture on the metal surface. accelerates corrosion and, in particular, pitting of HⳭ ions, the moisture has a lower pH (often
The results (Fig. 5) also indicate that the metal attack. Chlorides do damage by penetrating and below 4.5). The addition of SO2 air contamina-
surface exposed in the rural-urban environment destroying the normally protective and passive tion, acid rain, results in a highly accelerated cor-
experiences wet/dry cycles. Such cycles affect layer of oxides and hydroxides formed under rosion rate. Some metals, such as aluminum and
the structure and morphology of corrosion prod- natural conditions. Such protective surface films zinc, are relatively resistant in pH-neutral at-
ucts and promote micro- and macrocracking, as
well as the detachment and exfoliation of inter-
nal corrosion layers. The difference in the daily Table 2 Standards for testing and characterizing atmospheric corrosion
T-RH complex of the marine-coastal and rural-
Designation Title
urban environments also has another important
effect on metal corrosion behavior. Because International Organization of Standardization (ISO), Geneva, Switzerland
TOW occurs in different temperature ranges ISO 8565 “Metals and Alloys, Atmospheric Corrosion Testing, General Requirements for Field Tests”
(Fig. 6, 7), this fact determines a distinct corro- ISO 9223 “Corrosion of Metals and Alloys, Corrosivity of Atmospheres, Classification”
ISO 9225 “Corrosion of Metals and Alloys, Corrosivity of Atmospheres, Measurement of Pollution”
sion rate. Following the 10 rule, a 10 C (18 F) ISO 9226 “Corrosion of Metals and Alloys, Corrosivity of Atmospheres, Method of Determination of Corrosion
temperature difference can roughly change the Rate of Standard Specimens for the Evaluation of Corrosivity”
corrosion rate by an order of magnitude. ISO 8407 “Corrosion of Metals and Alloys, Removal of Corrosion Products from Corrosion Test Specimens”
The nature and orientation of the metal surface ISO 11463 “Corrosion of Metals and Alloys, Evaluation of Pitting Corrosion”
ISO 7384 “Corrosion Tests in Artificial Atmospheres, General Requirements”
and its inclination to the horizon or exposure an- ISO 9227 “Corrosion Tests in Artificial Atmospheres, Salt Spray Tests”
gle also influence the real metal T and TOW val-
ASTM International, West Conshohocken, PA, USA
ues, due to the difference in solar absorbance,
emissivity, and conductivity of the metal. The ASTM G 50 “Standard Practice for Conducting Atmospheric Corrosion Tests on Metals”
ASTM G 4 “Standard Guide for Conducting Corrosion Coupon Tests in Field Applications”
surface condition and color of the metal and its ASTM G 1 “Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens”
corrosion products are factors. These specific ASTM G 92 “Standard Practice for Characterization of Atmospheric Test Sites”
characteristics contribute to surface T and TOW ASTM G 84 “Standard Practice for Measurement of Time-of-Wetness on Surfaces Exposed to Wetting Conditions
changes. Exposed metals with corrosion prod- as in Atmospheric Corrosion Testing”
ASTM G 91 “Standard Practice for Monitoring Atmospheric SO2 Using the Sulphation Plate Technique”
ucts on the surface can have a higher T compared ASTM G 140 “Standard Test Method for Determining Atmospheric Chloride Deposition Rate by Wet Candle
to that of the environment (Fig. 8, 9). In cold Method”
regions, this can result in the appearance of liq- ASTM G 33 “Standard Practice for Recording Data from Atmospheric Corrosion Tests of Metallic-Coated Steel
uid on the metal surface, even when ambient Specimens”
ASTM G 107 “Standard Guide for Formats for Collection and Compilation of Corrosion Data for Metals for
temperature is below 0 C (32 F). This explains Computerized Database Input”
why metals having a similar TOW period can ASTM G 135 “Standard Guide for Computerized Exchange of Corrosion Data for Metals”
corrode at different rates when they are exposed ASTM G 46 “Standard Guide for Examination and Evaluation of Pitting Corrosion”
in distinctly different climatic areas. Therefore, ASTM G 101 “Standard Guide for Estimating the Atmospheric Corrosion Resistance of Low-Alloy Steels”
ASTM G 48 “Standard Test Method for Pitting and Crevice Corrosion Resistance of Stainless Steels and Related
defining the T-RH complex is of primary impor- Alloys by Use of Ferric Chloride Solution”
tance in completely understanding the corrosion ASTM G 112 “Standard Guide for Conducting Exfoliation Corrosion Test in Aluminum Alloys”
process. Local TOW values can also be ex- ASTM G 66 “Standard Test Method for Visual Assessment of Exfoliation Corrosion Susceptibility of 5xxx Series
tended, due to the porous cavity structure of cor- Aluminum Alloys (ASSET Test)”
ASTM G 38 “Standard Practice for Making and Using C-Ring Stress-Corrosion Test Specimens”
rosion products on the metal surface. ASTM G 16 “Standard Guide for Applying Statistics to Analysis of Corrosion Data”
Metal temperature and TOW values are influ- ASTM G 31 “Standard Practice for Laboratory Immersed Corrosion Testing of Metals”
enced by winds and their predominate direction ASTM B 117 “Standard Practice for Operating Salt Spray (Fog) Apparatus”
(north or south, continental or onshore). This can ASTM G 85 “Standard Practice for Modified Salt Spray (Fog) Testing”
ASTM G 87 “Standard Practice for Conducting Moist SO2 Tests”
change the type and amount of atmospheric pol- ASTM G 60 “Standard Test Method for Conducting Cyclic Humidity Tests”
lutants that settle on the metal surface. In some ASTM G 3 “Conventions Applicable to Electrochemical Measurements in Corrosion Testing”
cases, winds can transport sand and other hard ASTM G 102 “Standard Practice for Calculation of Corrosion Rates and Related Information from Electrochemical
particles that provoke accelerated metal erosion Measurements”
ASTM G 100 “Standard Test for Conducting Cyclic Galvanostaircase Polarization”
or corrosion-erosion effects. ASTM G 59 “Standard Test Method for Conducting Potentiodynamic Polarization Resistance Measurements”
ASTM G 5 “Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic Polarization
Measurements”
ANSI/ASTM G 61(a) “Conducting Cyclic Potentiodynamic Polarization Measurements for Localized Corrosion”
Air Chemistry and Principal ASTM G 106 “Standard Practice for Verification of Algorithm and Equipment for Electrochemical Impedance
Measurements”
Pollutants Inducing Corrosion ASTM G 61 “Standard Test Method for Conducting Cyclic Potentiodynamic Polarization Measurements for
Localized Corrosion Susceptibility of Iron-, Nickel-, or Cobalt-Based Alloys”
Air chemistry is closely related to the corro- ASTM G 96 “Standard Guide for On-Line Monitoring of Corrosion in Plant Equipment (Electrical and
Electrochemical Methods)”
sion aggressiveness of the atmosphere and this
fact needs careful attention (Ref 1, 7–9, 11–12). (a) ANSI, American National Standards Institute
200 / Forms of Corrosion

mospheres (pH  6 to 7.5) but corrode rapidly mechanism process, which increases the corro- forms of the metal that are thermodynamically
in an acid environment (Ref 12). sion current from electrochemical (galvanic) stable over a range of pH and electrochemical
According to ISO 9223, the annual average cells, and the resultant corrosion rate. potential.
deposition rate (mg/m2/day) of chlorides (air- Most aerosol particles absorb water, leading Atmospheric Corrosion of Iron and Car-
borne salinity) and sulfur dioxide (SO2) com- to an increase in the TOW period and the cor- bon Steels. The most-used metal for construc-
pounds is used to classify the atmospheric cor- rosive process. tion of structures and equipment is iron, because
rosivity. The recommended methods for it is the main constituent of the carbon and alloy
measurement of chloride and sulfate levels are steels. The iron Pourbaix diagram (Fig. 10) dem-
the wet candle and sulfation plate sampling ap- Thermodynamics of onstrates the possibility for multiple states: cor-
paratuses, as cited in ISO 9225, “Corrosion of rosion (active state), passivity and immunity
Metals and Alloys, Corrosivity of Atmospheres, Atmospheric Corrosion (Ref 1, 13, 34–46).
Measurement of Pollution.” and Use of Pourbaix Diagrams In the region of potentials and pH values de-
Some atmospheric gases, such as carbon di- fined by Fe(OH)3 and Fe2O3 (the solid com-
oxide (CO2), nitrogen dioxide (NO2), ozone When the question is asked, “Under what con- pounds thermodynamically stable in these con-
(O3), ammonia (NH3), hydrogen sulfide (H2S), ditions can corrosion of metals occur in aqueous ditions), the initial corrosion process forms a
and hydrogen chloride (HCl), and organic acids, solutions?”, the answer can usually be found us- very dense and usually thin and impervious rust
such as formic (HCOOH) and acetic ing Pourbaix diagrams, which show regions of
(CH3COOH), are also known to be highly cor- metal stability and corrosion on axes of metal
rosive for several commonly used metals. After electrochemical potential as a function of pH of
being dissolved in the moisture layer on the the electrolyte (Ref 13). It is well known that
metal surface, these gases result in a number of metals occur in different states, depending
ions and ionic species, such as HⳭ, CO2ⳮ ⳮ
3 , Cl , mainly on the pH of the environment and their
NHⳭ 4 , NO ⳮ
3 , SO 2ⳮ
4 , COOH ⳮ
, and CH 3 COO ⳮ
, electrochemical potential (e.g., Gibbs free en-
that may have a major influence on the corrosion ergy, Table 1). Pourbaix diagrams show the

Fig. 6 Distribution of the annual time of wetness (%) in

different temperature intervals presented in the
rural-urban tropical humid environment of Merida, 30 km
(18 miles) from the Gulf of Mexico, in 1998

Fig. 7 Distribution of the annual time of wetness (%) in

different temperature intervals presented in the
Fig. 5 Variation in temperature (T) and relative humidity (RH) during 1998 in marine-coastal and rural-urban environ- marine-coastal tropical humid environment of the port of
ments with tropical humid climate (Gulf of Mexico) Progreso (Gulf of Mexico) in 1998
 Atmospheric Corrosion / 201

layer of iron oxide and oxyhydroxides that acts alloys is still not well explained, and several
as an effective physical barrier between the metal models have been proposed (Ref 11).
and corrosive atmosphere. Due to this physical Figure 10 shows two regions of corrosion (ac-
barrier, oxygen and water molecules cannot eas- tive) metal state, when the metal is corroded ei-
ily penetrate and reach the underlying metal sur- ther to Fe2Ⳮ or Fe3Ⳮ ions, depending on pH and
face. Therefore, the corrosion process is effec- potential values. In natural atmospheric condi-
tively stopped, and the metal is in a passive state. tions, the standard potential of iron is negative
Iron can also be observed in another passive (approximately ⳮ0.44 V, Table 1). This situa-
region, that of Fe3O4 and Fe(OH)2, when corro- tion indicates that in neutral and low-acid indus-
sion produces an oxide (magnetite, Fe3O4) that trial environments, the iron carbon steel and low-
is a very thin, dense, and almost transparent rust alloy steel will corrode whenever the pH of the
layer. This layer can also act as an effective environment is lower than 8.5 to 9. The corro-
physical barrier that stops corrosion. However, sion attack on iron and steels generally occurs
for this passive state to exist, the metal needs be uniformly, extending on all metal surfaces to the
exposed to an environment at an alkaline pH same extent.
higher than 8.5 to 9. This explains the passive An interesting case is the immune-metal state,
carbon steel state when it is embedded in an al- which is not present for all metals but is possible
kaline (pH  12 to 13) concrete environment, for iron. The corrosion of steel is not possible in
such as occurs in steel reinforcement. However, the region of pH and potentials where iron is in
any changes of pH (below the pH 8.5 to 9) re- the thermodynamically stable immune condi-
move the metal from the passive state, and cor- tion, as shown in Fig. 10. An external voltage
rosion can resume. This is what happens in the can be applied to take the iron potential from its
case of the phenomenon referred to as carbona- standard state (ⳮ0.44 V, Table 1) to more neg-
tion of concrete. Reinforcing steel suffers serious ative values (less than ⳮ0.6 to ⳮ0.7 V). This is
and accelerated corrosion due to the lowering of the basis for the cathodic protection applied to
concrete pH as a result of the penetration of CO2 underground metallic pipeline systems, steel-re-
gas from the atmosphere into the concrete pores. inforced concrete bridges, ships, offshore plat-
This will cause subsequent dissolution of the forms, and other metal structures.
steel in the moisture, filling the pores of the con- Atmospheric Corrosion of Aluminum. An-
crete. other widely used metal for construction is alu-
A very dangerous pollutant for the destruction minum. Thermodynamically, it is very active
of the passive oxide layer, even in a favorable and immediately corrodes when produced (Table
alkaline pH medium, is the chloride ion. The 1, high level of free energy and high negative
chloride ion has a relatively small ionic radius potential value) (Ref 1, 13, 47–53). The Pour-
and high mobility in aqueous solutions. It can baix diagram (Fig. 11) shows three states that
penetrate the oxide layer, resulting in its destruc- are possible for this metal (and its alloys): pas-
tion, increasing the corrosion rate, and often sivity, immunity, and corrosion (active) (Ref 13).
leading to localized corrosion (pitting). It should There is a region of pH and potentials where the
be noted that the complete mechanism of chlo- metal is passivated and well protected from at-
ride-induced corrosion and pit formation on iron mospheric corrosion, due to the formation of a

Fig. 8 Sensor system for measurement of time of wet-

ness (TOW). (a) Closeup view of sensor. (b) Sen-
sors for TOW and temperature measurements on the sur- Fig. 9 Variation of environmental and low-carbon steel (Fe) sample temperatures and time of wetness (TOW) measured
face of sample. (c) View of the sensor electronic system on the steel surface (an amplified potential of the gold-copper sensor). Registered over two days
202 / Forms of Corrosion

very thin, transparent, and adherent low-porosity solution of the protective oxide film on the alu- standard metal potential of Ⳮ0.34 V), due to the
hydrated oxide layer (Al2O3H2O). This layer, minum surface. Corrosion protection for formation of an oxide layer of Cu2O (cuprite).
however, can be destroyed by the presence of aluminum can be obtained through anodizing, Another passive state can be obtained when the
chloride ions in the environment, which will pro- which is achieved by growing a thicker oxide metal potential is shifted to more positive values,
duce pitting. The passive layer provides very film (approximately 20 lm, or 0.8 mil) under forming copper oxide (CuO). The immune state
good corrosion resistance when exposed to nor- anodic polarization in appropriate electrolytes. requires more negative potential values (using an
mal atmospheric conditions and lower resistance Pure aluminum is seldom used for structures, external electric source), where the metal does
to corrosion when structures are exposed in ma- because its alloys have better mechanical prop- not corrode regardless of pH value. The diagram
rine-coastal environments. Aluminum and its al- erties. However, alloy corrosion resistance varies in Fig. 12 shows two regions of the metal cor-
loys can also exhibit a layered corrosion exfoli- from less than that of pure aluminum when al- rosion state: one at low (acid) pH and the other
ation attack. loyed with magnesium and copper to much bet- at higher (alkaline) pH values, when metal ions
In natural conditions, aluminum has a high ter when combined with tungsten and tantalum. (Cu2Ⳮ) and complex anions (CuO2ⳮ 2 ), respec-
negative standard potential value (ⳮ1.67 V, Ta- One of the most widely used aluminum alloys is tively, are formed during the anodic corrosion
ble 1), but due to the immediate formation of a Duralumin, which contains 4% Cu and a small reaction (Eq 1).
passive oxide layer, it can be protected in envi- amount of other metals, such as iron. This alloy The atmospheric behavior and protection of
ronments where the pH is in the range of 2 to corrodes much more readily than pure alumi- copper against corrosion is of interest because it
14. However, the range of pH from 4 to 9 is the num. Localized pitting corrosion is usually ob- is a construction material in monuments,
practical range for many applications where sta- served as the failure mode, due to the breakdown churches, and architectural objects. The oxide
bility exists, and the highly passive layer forms of the passive oxide layer. passive layer (Cu2O, cuprite) formed during the
and remains protective. This results in generally Atmospheric Corrosion of Copper. Copper initial stages of the copper corrosion process is
good corrosion resistance and the wide use of is widely used because of its good corrosion re- called the patina, and extensive investigations
aluminum. The corrosion state is found in very sistance in a variety of atmospheres (Ref 1, 13, have been dedicated to this subject, especially
polluted, mainly SO2, industrial atmospheres, 54–67), high electric and thermal conductivities, for restoration of monuments (Ref 56–60, 65).
where pH is below 2, and also in alkaline envi- and attractive mechanical properties when ex- The patina layer is transparent (30 Å of thick-
ronments over a substantial range of potentials posed at low, moderate, and high temperatures ness at the beginning) and changes color from
(Fig. 11). In noncontaminated atmospheres, and (Ref 1). Its electrochemical potential value is orange to red-brown. The color deepens to dark
after a long exposure time, small pits (100 lm, positive (Ⳮ0.34 V, Table 1) in natural environ- brown when the thickness of this layer increases.
or 4 mil deep) can occur for aluminum in contact ments, as compared to values observed for iron Recent studies have revealed that copper patina
with water (high values of TOW). The presence and aluminum, and close to the potentials for consists of Cu2O/CuO (Cu(OH)2 or CuO • H2O).
of metal impurities (iron, copper) increases this noble nonreactive metals. Therefore, copper is In marine-coastal and industrial atmospheres, the
pitting, which is rate controlled by oxygen cath- not very active chemically, and its rate of oxi- color of the patina can be superficially changed
odic reduction on the surface inclusion, Hence, dation, when exposed to the atmosphere, is very to a more greenish hue because of the formation
the growth of the pit depth slows as the pit di- low. The corrosion open-circuit potential (OCP) of copper basic salts, such as sulfates in urban
ameter increases. Galvanic couple corrosion (bi- of copper is usually below the standard hydrogen or rural environments, chlorides in marine at-
metallic corrosion) occurs when a metal having reaction potential value (0.00 V, Table 1), and mospheres, nitrates, and carbonates. The knowl-
a less negative standard potential than that of due to this fact, there is no participation by hy- edge of patina formation and its development in
aluminum (Table 1) is brought into contact. Cop- drogen ion (HⳭ) as an oxidizing agent in the time is also used for the creation of an artificial
per is a cathode in the formed pair with alumi- copper patina. This type of surface treatment can
cathodic corrosion reaction (Eq 4) when the at-
be used to give a more antiquated appearance to
num. mosphere is very polluted (as in the acid pH
copper objects.
The metal potential of aluminum must be range). According to the Pourbaix diagram (Ref
Atmospheric Corrosion of Zinc. The ability
shifted to values more negative than ⳮ1.7, using 13) for copper (Fig. 12), three thermodynamic
of zinc to galvanically protect steel, because of
an external source of direct current to reach the states are possible: corrosion, passivity, and im-
immune state. Cathodic protection can be carried munity. It can be seen that in atmospheres with
out but may run into difficulties arising from lo- neutral and alkaline pH, copper is passive (at a
cal pH increases, which could lead to the dis- 1.0

0.8 CuO CuO 2–

2

Corrosion
1.4
0.6 Cu2+
1.0 AlO–2 Corrosion
0.4
0.6 Al3+
0.2
Potential, V

0.2 Al2O3·H2O
0
Potential, V

–0.2 Passivation
Corrosion
Cu2O
–0.6 Corrosion –0.2 Cu
–1.0
–0.4
–1.4 Passivation
–0.6
–1.8 Immune
Al –0.8
–2.2 Immune
–2.6 –1.0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 16
pH pH

Fig. 10 Pourbaix diagram (metal potential versus pH) Fig. 11 Pourbaix diagram (potential versus pH) for alu- Fig. 12 Pourbaix diagram (potential versus pH) for
for iron in aqueous (water) solution minum in water at 25 C (77 F) copper in water at 25 C (77 F)
 Atmospheric Corrosion / 203

A = a1 + a2 TOW + a3CSO2 + a4CCl−

its more negative standard electrochemical po- Models for Prediction (Eq 7)
tential, has given this metal a wide variety of of Atmospheric Corrosion
application (Ref 1, 13, 68–80). Although zinc by where a1, a2, a3, and a4 are coefficients, and CSO2
itself is available in sheet, plate, strip, and pres- and CCl− are the deposition rates (mg/m2/day) of
sure die castings, its principal uses for corrosion Models predicting the corrosion damage of these pollutants, measured according to ISO
control are as a coating on steel and as zinc an- metals exposed to atmosphere have been a recent 9225.
addition to understanding corrosion in atmo- A general model has been proposed that di-
odes for cathodic protection of steel on under-
spheric environments (Ref 12, 16, 18, 21, 23, 26, vides the total corrosion attack (K) into three
ground pipelines, oil rigs, offshore structures,
32, 79–84, 96). They are important and useful in dominating parts:
and bridges.
predicting the durability of metallic structures
Because of the low standard negative potential
and their degradation due to the corrosion pro-
value for zinc (ⳮ0.76, Table 1), it is a very ac-
cess. K ⳱ fdry(SO2) Ⳮ fdry(Cl) Ⳮ fwet(HⳭ) (Eq 8)
tive metal and tends to corrode in contact with
First-year atmospheric corrosion weight loss
air and moisture, Figure 13 presents the Pourbaix where fdry(SO2) is the effect of dry deposition of
is a parameter that allows the classification of a
diagram of zinc and shows that when a moisture SO2, fdry(Cl) is the effect of dry deposition of
given atmosphere into a corrosivity category.
layer forms on zinc with a neutral pH (pH  6
ISO 9223 uses the annual corrosion rate (weight chlorides, and fwet(HⳭ) is the effect of wet de-
to 7), corrosion occurs, and the main corrosion loss per year) of four standard metals (low-car- position of HⳭ (acid rain), which is not included
product is Zn5(CO3)2(OH)6 (hydrozincite). In ru- bon steel, copper, zinc, and aluminum) exposed in ISO 9223.
ral environments, this corrosion layer is dense according to ISO 9226. For this purpose, flat The terms of SO2 and chlorides can be ex-
and adherent; therefore, the metal is relatively samples are exposed on racks, usually at 45 in pressed as follows:
well protected. In highly polluted industrial at- Europe or 30 angle of inclination to the horizon
mospheres, the corrosion rate may increase sub- for a period of 1 year. The formed corrosion fdry(SO2) ⳱ A(SO2)B(TOW)C exp{g(T)} (Eq 9)
stantially with time of exposure. Strongly acid products on the metal surface are removed from
or basic environments tend to dissolve the stable the coupons in accordance with ISO 8407, and fdry(Cl) ⳱ D(Cl)E(TOW)F exp{kT)} (Eq 10)
corrosion film, leading to significantly higher the metal weight loss (g/m2/year) or corrosion
corrosion rates. In acidic atmospheres, no pro- penetration (lm/year) is determined. Based on where A, B, C, D, E, F, and k are constants; TOW
tective corrosion films form on zinc, leading to these results, corrosivity categories are assigned. is time of wetness; T is temperature, and g(T) is
very rapid metal dissolution. Certain ionic spe- Classification of the atmosphere is the basis for a temperature function.
cies, such as chlorides, also promote the disso- the design of good corrosion protection through Many dose/response relationships can be ob-
lution of zinc corrosion products and lead to material selection. Annual corrosion data are tained, and the significance of factors can be
higher corrosion rates. also used for prediction of longer-term service judged. The results of these efforts are not yet
When zinc is used as a coating on steel, the life of metal construction in given environments. suitable for general use at the international level.
steel is a cathode, and the zinc coating is the The corrosiveness of atmospheric sites can be The main reason is that the physical-chemical
anode. Corrosion protection in this situation is determined according to ISO 9223, based on an- background of the atmospheric corrosion is very
attributed to a combination of the corrosion re- nual deposition rate (ISO 9226) of the principal complicated, and the interpretation of its kinetics
sistance of zinc and the sacrificial protection that pollutants, such as SO2 and Clⳮ. However, the is limited. One serious complication is the phe-
is afforded by zinc to the steel. This cathodic evaluation of atmospheric corrosivity is more di- nomenon of runoff on the corroded metal sur-
protection is also provided when zinc anodes are rectly connected to metal performance when the face, which is difficult to involve in a general
electrically connected to a steel structure and aggressiveness category of a given atmosphere dose/response relationship because of its specific
both are immersed in the same conductive elec- is assigned based on the annual corrosion rates particularities of corrosion reactions and prod-
trolyte. of standard metal samples in combination with uct-formation chemistry. For example, in ma-
Note that Pourbaix diagrams do not give in- the annual deposition rate of the main contami- rine-coastal areas or regions with a calcareous
formation about the kinetics or rate of the cor- nants. soil, dust of calcium compounds is deposited on
rosion process. They indicate the thermody- The corrosion rate (C, in g/m2) of a given the metal surface. Sulfur compound pollutants
namic conditions for the development of metal after time (t, in years) depends directly on are absorbed in the surface, and a series of ad-
corrosion, the possibility for reaching other ox- its first-year atmospheric corrosion rate (A, in g/ ditional reactions begin, with gypsum (calcium
idation states, and the corrosion product com- m2 or lm) and its dependence with the time (n):
position. This information, in turn, indicates pos-
sible regimes for metal dissolution or protection. C ⳱ Atn (Eq 5) 1.4
Moreover, these diagrams correspond to metals
exposed in pure aqueous solutions that do not where A and exponent n are dependent on the 1.0
ⳮ ⳮ
include other ions, such as SO2ⳮ 4 , Cl , NO3 , type of metal and climatic parameters. Values of Zn2+
and COⳮ 3 contaminants. However, the effects of n typically range from 0.5 to 1, with most values 0.6
other species can be taken into account, using being close to unity.
ZnCO3

Zn(OH)2
more complex Pourbaix diagrams (Ref 13) or The bilogarithmic model of atmospheric cor- 0.2
Potential, V

through thermodynamic modeling software. For rosion gives a linear relationship between log C Corrosion
example, the presence of SO42ⳮ, Clⳮ, NO3ⳮ, and log t: –0.2
Corrosion
and CO3ⳮ ions as atmospheric contaminants in
the metal-moisture system can eliminate the im- log C ⳱ A Ⳮ n log t (Eq 6) –0.6
mune copper state (Ref 13), as predicted using
the Pourbaix diagram shown in Fig. 12. There- Parameter A depends on properties of the test –1.0 Zn Immune
fore, when investigating the influence of service site and suggests a correlation with climatic vari-
environments, it is necessary to analyze the at- ables and air chemistry (TOW, T-RH complex, –1.4
0 2 4 6 8 10 12 14
mosphere for all its parameters (T-RH complex and pollution level). For this reason, it is often pH
and air chemistry) and to use the Pourbaix dia- correlated to the pollutant level of SO2 and Clⳮ
grams as a guide to the thermodynamics of the and to meteorological parameters, leading to a Fig. 13 Pourbaix diagram (potential versus pH) for zinc
metal dissolution process. relationship such as the following: in water at 25 C (77 F)
204 / Forms of Corrosion

sulfate) as the end product. Gypsum, which is conducted to establish atmospheric corrosion ing released during first flush. The total annual
more soluble than calcium carbonate, is then and metal runoff processes, mainly on copper runoff rate of copper is significant for green-pa-
washed away by rain. An increase of volume and and zinc used for roofing applications. It has tinated samples, whereas it is negligible for
variations in temperatures can take place when been proven that the runoff rate of zinc is con- brown-patinated samples.
calcium carbonate reacts to form gypsum. siderably lower than its corrosion rate, varying Zinc commonly forms voluminous and highly
Results from a number of atmospheric corro- between a quotient of 50 to 90% for zinc and 20 porous atmospheric corrosion products. This
sion testing programs show that the bi-logarith- to 50% for copper during exposure of up to 5 may explain why no large differences in runoff
mic model (Eq 6) for atmospheric corrosion is and 2 years, respectively. Detailed studies have rate have been seen between new and aged zinc
applicable to a number of commonly used metals been performed to disclose the effect of various samples in the reported field investigations.
(carbon steels, low-alloyed steels, galvanized parameters on the runoff rate, including surface Rather, the reason may be related to a high sus-
steels, and aluminized steels) in many environ- orientation and inclination, natural patinated ceptibility for proton-induced dissolution of zinc
ments. Wide-ranging atmospheric corrosion tests copper, patina composition, rain duration and corrosion products by rain. During the dry pe-
of standard metals have been conducted to unify volume, rain pH, and length of dry periods in riod, neutral zinc salts with high solubility con-
operational procedures and to acquire metal cor- between rain events. Based on field exposures stants are frequently formed, such as ZnSO4 or
rosion data for modeling and predicting atmo- and literature data, a correlation has been estab- Zn(NO3)2. These are easily dissolved during the
spheric corrosion. Based on these efforts, cor- lished between runoff rate and the prevailing first-flush release; then, less soluble zinc salts are
rosivity maps of atmospheres in a number of SO2 environmental concentration. The runoff formed, including zinc hydrosulfates and zinc
countries have been created. The bilogarithmic rate of zinc and copper increases with increasing hydrocarbonates, which govern the dissolution
model is helpful in extrapolating short-term at- SO2 level for exposure sites of similar annual rate during the steady-state runoff.
mospheric corrosion data to longer time. When precipitation quantities (500 to 1000 mm/year, Lead ions may also be introduced into the en-
considering corrosion severity over the long or 20 to 40 in./year). vironment by the flow of precipitation runoff
term, changes in the environment may be more High metal concentrations have been found in from the surface of old lead structures, such as
significant than deviations from the model. the initial rain volume flushing the surface—the gutters, roofs, piping, siding, and sculptures.
first flush, during which the most easily soluble This metal is of particular concern to the public
and poorly adhesive corrosion products are because of the adverse effect of even very small
Atmospheric Corrosion washed off from the surface. The magnitude of amounts on human health. The corrosion rate of
first flush depends on the presence and amount lead depends on solubility and the physical char-
and Precipitation of soluble corrosion products and also is asso- acteristics of the corrosion products formed.
Runoff from Corroded Metals ciated with the capacity of the corrosion products Lead presents a high corrosion resistance in ex-
to absorb and retain water. In turn, this is related posure to the atmosphere and to water, due to the
Atmospheric corrosion of some metal struc- with their adherence, morphology, thickness, po- formation of insoluble lead salts deposited on its
tures, such as zinc and copper sheets commonly rosity, and presence of internal micro- and ma- surface. However, in natural and domestic wa-
used for roofing and drain water systems, zinc crocracks and defects. The precipitation volume ters, the corrosion rate depends on the degree of
anodes for cathodic protection, and zinc and is considered as the most important parameter water hardness (calcium and magnesium salts
zinc-aluminum coatings, involves the formation affecting the runoff quantity of copper and zinc. content), the content of dissolved oxygen, and
of protective oxide/hydroxide corrosion prod- Samples exposed in different environments ex- the CO2 concentration. In the absence of passi-
ucts that act as effective physical barriers be- hibit large differences in the magnitude of the vating substances (such as carbonates), any ox-
tween the metal and the aggressive atmosphere. first flush. For copper and zinc panels preexpo- idizing agent can cause lead to corrode. The
However, due to interaction with the environ- sed in Swedish urban, rural, and marine environ- presence of nitrate and chloride ions interferes
ment, the metal protective film could suffer mod- ments, yearly runoff rates were 1.2, 0.7, and 1.7 with the formation of a protective layer or pen-
ification into nonprotective corrosion products. g/m2/year of copper and 2.6, 1.6, and 3.7 g/m2/ etrates it, thus increasing the corrosion. Corro-
Physical removal from the metal surface through year of zinc, assuming an annual precipitation of sion film studies indicate that lead in the runoff
dissolution of soluble corrosion products in pre- 500 mm/year (20 in./year). is primarily from solubility of lead carbonate
cipitation runoff or by spalling could result (Ref The first-flush effect usually decreases to (cerrusite) and lead hydroxyl carbonate (hydro-
85–96). Precipitation runoff is water from rain, rather constant metal concentration during the cerrusite).
dew, snow, or fog that drains from a surface and subsequent rains. The metal concentration in While atmospheric corrosion rates of metals
contains air- or waterborne deposited reactants runoff increases with rain acidity, decreases with usually exhibit a continuous decrease with time,
and soluble ions from the metal surface. For that rain intensity, and increases with length of the the yearly runoff rates are fairly independent of
reason, traces of metals such as copper, zinc, dry period preceding a rain event. Drizzle (1 time. The runoff process can be presented at any
lead, and iron are commonly detected in roof mm/h, or 0.04 in./h), with the longest surface time t by the mass balance:
runoff water. contact period, results in an increased amount of
The use of copper as a roofing material has a released copper.
long tradition, and zinc sheets have been used It is considered by researchers that the copper C(t) ⳱ T(t) Ⳮ R(t) (Eq 11)
for over 200 years. During the last decade, a con- runoff is caused by proton- or ligand-induced
cern has been raised by legislators in Europe and water dissolution of the noncrystalline cupric where C(t) is the cumulative corrosion mass loss,
the United States on the quantity of metal re- sulfate as a part of copper corrosion products T(t) is the protective corrosion mass loss, and
moved from a roof during precipitation and the formed in environments polluted with SO2, HCl, R(t) is the cumulative nonprotective corrosion
potential effect that the released metal may have and NOx. The runoff effect is more pronounced product (runoff effect).
on the environment. Urban stormwater is rec- for aged copper (ⱖ40 years old). No significant In dry-season exposure of metals in polluted
ognized as a source of contaminants, including differences have been observed between epi- atmospheres (pH below 5), the environment re-
trace metals, and roof runoff is a contributor. It sodes of light rain (8 mm/h, or 0.3 in./h) and acts with the metal surface, forming neutral salts
has been reported that galvanized roofs can con- moderate rain (20 mm/h, or 0.8 in./h). (soluble products), but there is little or no runoff.
tribute zinc concentrations of between 1 and 44 In the case of copper patina formed in natural When the precipitation starts, these salts are dis-
g/m3, whereas tile roofs contribute between 0.01 environments, a higher wetting capacity and ab- solved in the presence of acid gases as pollutants
and 2.6 g/m3 of zinc in the runoff. sorption of rainwater have been found on green- of the atmosphere. In the absence of spalling or
An extensive investigation in the last decade, patinated copper compared to brown-patinated, significant accumulation of nonprotective cor-
with parallel field and laboratory exposures, was which explains a higher magnitude of copper be- rosion products, R in Eq 11 represents the cu-
 Atmospheric Corrosion / 205

mulative loss of soluble corrosion products in The corrosive microbial effect on metals can microbes. Microbiologically influenced corro-
precipitation runoff. The time derivative of Eq be attributed to the removal of electrons from the sion can be studied using field survey and dif-
11 shows that the corrosion rate is equal to the metal and the formation of corrosion products ferent techniques, such as electrochemical polar-
rate of protective film growth and the rate of cor- by: ization curves and polarization resistance,
rosion film loss in precipitation runoff: electrochemical impedance spectroscopy, thin-
● Direct chemical action of metabolic product,
film electrical resistance probe, galvanic current
such as sulfuric acid, inorganic or organic sul-
dC/dt ⳱ dT/dt Ⳮ dR/dt (Eq 12) measurement, scanning electron microscopy and
fides, and chelating agents such as organic
energy-dispersive spectroscopy surface analysis,
acid
After long exposure, when the corrosion layer Mössbauer spectroscopy and x-ray diffraction
● Changes in oxygen content, salt concentra-
is developed and further corrosion film grows at for analysis of corrosion products, and so on. See
tion, and pH, thus increasing the possibilities
a slower rate, dT/dt approaches 0, and the cor- the articles “Microbiologically Influenced Cor-
for differential diffusion, concentration gra-
rosion rate is equal to the rate of precipitation rosion” and “Microbiologically Influenced Cor-
dients, and so on, which establish local elec-
runoff loss. Recent results suggest that, follow- rosion Testing” in this volume.
trochemical corrosion cells
ing an inductive period when the corrosion film
is establishing itself on the metal surface, the cu- One predominant mechanism for mild steel is
mulative runoff from corroding metal surfaces the bacteria reduction of cathodic hydrogen, re- Trends in Atmospheric
(such as rolled zinc, thermal spray zinc, and zinc- sulting in the depolarization of the cathode, Corrosion Research and Methods
aluminum and aluminum-zinc-indium alloys) is which facilitates the cathodic corrosion reaction.
linear to time and precipitation volume and rela- The presence of hydrogenase in the SRB allows Atmospheric corrosion processes have been
tively insensitive to seasonal variations in pre- the use of hydrogen gas as an electron donor in studied by exposure tests, accelerated corrosion
cipitation chemistry, air chemistry, and meteor- place of carbohydrates as an energy source. In testing, as well as electrochemical, surface ana-
ology. this case, the levels of assimilable carbon (usu- lytical, and spectroscopic techniques (Ref 107–
ally organic) and energy sources (nitrogen, sul- 155). These methods often give only integral in-
fate, and phosphate) in the water layer to support formation on corrosion processes (anodic and
Biologically Influenced the growth and activity of SRB also need to be cathodic) occurring at the solid-liquid interface.
Atmospheric Corrosion considered. This mechanism is valid where the The development of local electrochemical tech-
cathodic reaction can involve the generation of niques, such as scanning vibrating electrode
Atmospheric corrosion can be accelerated in hydrogen. The other mechanism is the bacterial technique, localized electrochemical impedance
the presence of different types of bacteria and, production of sulfide, leading to the formation of spectroscopy, and scanning Kelvin probe, with
specifically, anaerobic bacteria, which may con- FeS deposits. These deposits are hypothesized to different lateral resolutions and the increased use
vert a noncorrosive environment to a very ag- act as large surface areas for the cathodic reac- of various scanning probe microscopes (scan-
gressive one. Therefore, microbiologically in- tion and also as cathodic areas in a galvanic cell ning tunneling microscope, atomic force micro-
duced corrosion (MIC) is given serious attention, with the steel. scope, electrochemical scanning tunneling mi-
because MIC failure in plant service systems can Rate and extent of corrosion are related di- croscope, electrochemical force atomic
also have an ecological impact (Ref 97–106). rectly to bacteria growth in contact with the me- microscope, magnetic force microscope, scan-
Microorganisms have been shown to play an im- tallic surface. Biofilm bacteria growing on a sur- ning electrochemical microscope, and scanning
portant role in the corrosion of mild steel, aus- face produce extracellular polymeric substances near-field optical microscope) in corrosion sci-
tenitic alloys, and copper-base alloys, despite the that promote sediment attachment, leading to the ence allow local processes (in situ) to be studied
fact that copper ions are toxic to most organisms. development of deposits and colonies of anaer- on an atomic scale.
Failures in these materials are generally mani- obic bacteria, specifically SRB, which have been Combining conventional electrochemical and
fested by pitting, erosion-corrosion, and occa- implicated in most MIC failures. Bacteria have surface analytical measurements with these local
sionally by stress-corrosion cracking. been found that are capable of growing on many techniques permits the treatment of corrosion as
The role of microbes or bacteria in the cor- kinds of coating materials (including hydrocar- a local phenomenon and contributes to a better
rosion of metals is due to their chemical activi- bons). understanding of the processes and the effects of
ties (metabolism) associated with microbial Monitoring is particularly important for early the controlling factors. Currently, local atomistic
growth and reproduction. For example, the MIC effective MIC control, because biofilms can events and effects of surface imperfections of
of metals by sulfate-reducing bacteria (SRB) is form on metal surfaces very rapidly. Once a ma- different dimensionality on the interfacial pro-
a recognized problem in pipelines. This bacteria ture biofilm is established, the slime layer pro- cesses can be directly analyzed. Since the de-
is anaerobic and also functions in poorly drained duced by the microorganisms, along with cor- velopment of in situ scanning probe microscopy
wet soil that has a pH from 6 to 8; contains sul- rosion products, makes the biofilm extremely techniques, the structures, thermodynamics, and
fate ions, organic compounds, and minerals in resistant to the effects of chemical treatments. kinetics of interfacial processes can be investi-
the absence of oxygen; and has a temperature The key symptom that usually indicates SRB in- gated directly in a real-time domain on an atomic
from 20 to 30 C (70 to 85 F). In well-aerated volvement in the corrosion process of ferrous al- scale at the solid-liquid interface. These techni-
local sites, the anaerobic bacteria cause no prob- loys is localized corrosion filled with black sul- cal achievements point out a new direction of
lem. For metabolism of the bacteria, oxygen is fide corrosion products. Tests for SRB include corrosion research aimed toward elaborate new
extracted from the sulfate ions, and this reaction counts of total viable sulfate reducers. Other tests corrosion models based on the submicroscopic
converts the soluble sulfates to sulfide (hydrogen include measurement of the hydrogen uptake of approach to the electrochemistry of corrosion
sulfide), which is a pitting activator and attacks a soil and the time for a soil sample to blacken metal processes. While these techniques are cur-
the metal surface, forming, for example, iron sul- a medium used to grow sulfate reducers. Chem- rently in the forefront of today’s research, they
fide (FeS). It has also been suggested that hy- ical analysis of the water layer on the metal is will likely yield large future benefits in terms of
drogen sulfide may be subsequently oxidized to helpful in determining the critical nutrients nec- better characterization of corrosion phenomena
thiosulfate, which is an even more aggressive essary to support microbiological activity. These and prediction of corrosive degradation.
pitting activator. The formed sulfides and other tests should be conducted under the direction of
reduced-sulfur compounds can catalyze the an- a microbiologist who has experience in MIC. REFERENCES
odic dissolution of stainless steels and lower Field test kits are currently available that employ
their repassivation and pitting potential, which an antibody tagging technique to identify and fa- 1. L.L. Shreir, R.A. Jarman, and G.T. Bur-
increases the metal susceptibility to pitting. cilitate the determination of the population of stein, Ed., Corrosion, Vol 1, Metal/Envi-
206 / Forms of Corrosion

ronment Reactions, 3rd ed., Butterworth- Ganther, The Rate of Drying of Moisture door Atmospheric Corrosion, STP 1421,
Heinemann, 1994 from a Metal Surface and Its Implication H.E. Townsend, Ed., American Society for
2. “Economic Effect of Metallic Corrosion in for Time of Wetness, Corros. Sci., Vol 37 Testing and Materials, 2002, p 3–18
the United States,” NBS Special Publica- (No. 3), 1995, p 455–465 31. R.D. Klassen, P.R. Roberge, D.R. Lenard,
tions 511-1 and 511-2, National Bureau of 20. A.A. Mikhailov, M.N. Suloeva, and E.G. and G.N. Blenkinsop, Corrosivity Patterns
Standards, May 1978 Vasileva, Environmental Aspects of At- Near Sources of Salt Aerosols, Outdoor
3. S.C. Cederquist, Cost of Corrosion Studies mospheric Corrosion, Water, Air and Soil Atmospheric Corrosion, STP 1421, H.E.
Underway, Mater. Perform., Vol 38 (No. Pollution, Vol 85, 1995, p 2673–2678 Townsend, Ed., American Society for Test-
5), 1999, p 26–28 21. D. Knotkova, P. Boschek, and K. Kreis- ing and Materials, 2002, p 19–33
4. G.H. Koch, M.P.J. Brongers, N.G. Thomp- lova, Results of ISO CORRAG Program: 32. J. Tidblad, V. Kucera, A. Mihhailov, and
son, Y.P. Virmani, and J.H. Payer, “Cor- Processing of One-Year Data in Respect to D. Knotkova, Improvement of the ISO
rosion Costs and Preventive Strategies in Corrosivity Classification, Atmospheric Classification System Based on Dose-Re-
the United States,” FHWA-RD-01-156, Corrosion, STP 1239, W.W. Kirk and H.H. sponse Functions Describing the Corrosiv-
Federal Highway Administration, U.S. De- Lawson, Ed., American Society for Testing ity of Outdoor Atmospheres, Outdoor At-
partment of Transportation, March 2002 and Materials, 1995, p 38–55 mospheric Corrosion, STP 1421, H.E.
5. B.C. Syrett, J.A. Gorman, M.L. Arey, G.H. 22. S.W. Dean and D.B. Reiser, Time of Wet- Townsend, Ed., American Society for Test-
Kosch, and G.A. Jacobson, Cost of Cor- ness and Dew Formation: A Model of At- ing and Materials, 2002, p 73–87
rosion, Mater. Perform., Vol 41 (No. 3), mospheric Heat Transfer, Atmospheric 33. H. Le Thi Lien and P. Thy San, The Effect
2002, p 18–22 Corrosion, STP 1239, W.W. Kirk and H.H. of Environmental Factors on Carbon Steel
6. U. Evans, Electrochemical Mechanism for Lawson, Ed., American Society for Testing Atmospheric Corrosion; The Prediction of
Atmospheric Rusting, Nature, Vol 206, and Materials, 1995, p 3–10 Corrosion, Outdoor Atmospheric Corro-
June 1965, p 980–982 23. J. Tidblad, A.A. Mikhailov, V. Kucera, Ap- sion, STP 1421, H.E. Townsend, Ed.,
7. I.L. Rozenfeld, Atmospheric Corrosion of plication of a Model for Prediction of At- American Society for Testing and Materi-
Metals, B.H. Tytel, Ed., National Associ- mospheric Corrosion in Tropical Environ- als, 2002, p 103–108
ation of Corrosion Engineers, 1973 ments, Marine Corrosion in Tropical 34. U.R. Evans, Mechanism of Atmospheric
8. V. Kucera and E. Mattsson, Atmospheric Environments, STP 1399, S.W. Dean, G. Rusting, Corros. Sci., Vol 12, 1972, p 227–
Corrosion, W.H. Ailor, Ed., John Wiley & Hernandez-Duque Delgadillo, and J.B. 246
Sons, 1982 Bushman, Ed., American Society for Test- 35. T. Misava, K. Asami, K. Hashimoto, and
9. Corrosion Basics, An Introduction, Na- ing and Materials, 1995, p 250–285 S. Shimodaira, The Mechanism of Atmo-
tional Association of Corrosion Engineers 24. M. Morcillo, J. Simancas, and S. Feliu, spheric Rusting and the Protective Amor-
1984 Long-Term Atmospheric Corrosion in phous Rust on Low Alloy Steels, Corros.
10. E. Mattsson, Corrosion: An Electrochem- Spain: Results after 13–16 Years of Ex- Sci., Vol 14, 1974, p 279–289
ical Problem, Chem. Technol., April 1985, posure and Comparison with Worldwide 36. J.B. Johnson, P. Elliott, M.A. Winterbot-
p 234–243 Data, Atmospheric Corrosion, STP 1239, tom, and G.C. Wood, Short-Term Atmo-
11. P. Marcus and J. Oudar, Ed., Corrosion W.W. Kirk and H.H. Lawson, Ed., Amer- spheric Corrosion of Mild Steel at Two
Mechanism in Theory and Practice, Mar- ican Society for Testing and Materials, Weather and Pollution Monitored Sites,
cel Dekker, Inc., 1995 1995, p 195–214 Corros. Sci., Vol 17, 1977, p 691–700
12. C. Leygraf and T. Graedel, Atmospheric 25. L. Veleva, G. Pérez, and M. Acosta, Sta- 37. E. Mattson, The Atmospheric Corrosion
Corrosion, Wiley-Interscience, 2000 tistical Analysis of the Temperature-Hu- Properties of Some Common Structural
13. M. Pourbaix, Atlas of Electrochemical midity Complex and Time of Wetness of a Metals: A Comparative Study, Mater. Per-
Equilibria in Aqueous Solutions, NACE- Tropical Climate in the Yucatan Peninsula form., Vol 21, July 1982, p 9–19
Cebelcor, 1974 in Mexico, Atmosph. Environ., Vol 31 (No. 38. E.A. Baker, Long-Term Corrosion Behav-
14. T.S. Lee and E.A. Baker, Calibration of At- 5), 1997, p 773–776 ior of Materials in the Marine Atmosphere,
mospheric Corrosion Test Sites, Atmo- 26. L. Veleva and L. Maldonado, Classifica- Degradation of Metals in the Atmosphere,
spheric Corrosion of Metals, STP 767, tion of the Atmosphere Corrosivity in the STP 965, S.W. Dean and T.S. Lee, Ed.,
S.W. Dean, Jr. and E.C. Rhea, Ed., Amer- Humid Tropical Climate, Br. Corros. J., American Society for Testing and Materi-
ican Society for Testing and Materials, Vol 33 (No. 1), 1998, p 53–57 als, 1988, p 125–144
1982, p 250–266 27. S. Feliu, M. Morcillo, and B. Chico, Effect 39. T.E. Graedel and R.P. Frankonthal, Cor-
15. S.D. Cramer and L.G. McDonald, Atmo- of Distance from Sea on Atmospheric Cor- rosion Mechanism for Iron and Low Alloy
spheric Factors Affecting the Corrosion of rosion Rate, Corrosion, Vol 55 (No. 6), Steel Exposed to the Atmosphere, J. Elec-
Zinc, Galvanized Steel and Copper, Cor- 1999, p 883–891 trochem. Soc., Vol 137 (No. 8), 1990, p
rosion Testing and Evaluation: Silver An- 28. J. Tidblad, A.A. Mikhailov, and V. Kucera, 2385–2394
niversary Volume, STP 1000, R. Baboian Model for the Prediction of the Time of 40. J.L. Crolet, Mechanism of Uniform Cor-
and S.W. Dean, Jr., Ed., American Society Wetness from Average Annual Data on rosion Under Corrosion Deposits, J. Mater.
for Testing and Materials, 1990, p 241–259 Relative Air Humidity and Air Tempera- Sci., Vol 28, 1993, p 2589–2606
16. S. Feliu, M. Morcillo, and S. Feliu, Jr., The ture, Prot. Met. (Russia), Vol 36 (No. 6), 41. S.H. Zhang and S.B. Lyon, The Electro-
Prediction of Atmospheric Corrosion from 2000, p 533–540 chemistry of Iron, Zinc and Copper in Thin
Meteorological and Pollution Parameters, 29. L. Veleva and M.A. Alpuche-Aviles, Time Layer of Electrolytes, Corros. Sci., Vol 35
Part I: Annual Corrosion; Part II: Long- of Wetness (TOW) and Surface Tempera- (N.S. 1–4), 1993, p 713–718
Term Forecasts, Corros. Sci., Vol 34, 1993, ture Characteristics of Corroded Metals in 42. J.O’M. Bockris and S.U. Khan, On the
p 403–414; 415–422 Humid Tropical Climate, Outdoor Atmo- Mechanism of the Dissolution of Iron: The
17. S.K. Roy and K.H. Ho, Corrosion of Steel spheric Corrosion, STP 1421, H.E. Town- Catalytic Mechanism, Surface Electro-
in Tropical Marine Atmospheres, Br. Cor- send, Ed., American Society for Testing chemistry: A Molecular Level Approach,
ros. J., Vol 29 (No. 4), 1994, p 287–291 and Materials, 2002, p 48–58 Plenum Press, 1994, p 756–771
18. D.H. Vu, Atmospheric Corrosion of Metals 30. S.W. Dean and D.B. Reiser, Analysis of 43. A. Cox and S.B. Lyon, An Electrochemical
in Tropics, Nauka, Moscow, Russia, 1994 Long-Term Atmospheric Corrosion Re- Study of the Atmospheric Corrosion of
19. I.S. Cole, R. Holgate, P. Kao, and W. sults from ISO CORRAG Program, Out- Mild Steel, Part III: The Effect of Sulfur
 Atmospheric Corrosion / 207

Dioxide, Corros. Sci., Vol 36 (No. 7), ment on the Patina of the Statue of Liberty, Atmospheric Hydrogen Chloride, Corros.
1994, p 1193–1199 Environ. Sci. Technol., Vol 25 (No. 8), Sci., Vol 34 (No. 2), 1993, p 233–247
44. S.K. Roy and K.H. Ho, Corrosion of Steel 1991, p 1400–1408 75. J.M. Costa and M. Vilarrasa, Effect of Air
in Tropical Marine Atmospheres, Br. Cor- 61. P. Eriksson, L.-G. Johansson, and H. Pollution on Atmospheric Corrosion of
ros. J., Vol 29 (No. 4), 1994, p 287–291 Strandberg, Initial Stages of Copper Cor- Zinc, Br. Corros. J., Vol 28 (No. 2), 1993,
45. L. Maldonado, L. Veleva, P. Quintana, O.T. rosion in Humid Air Containing SO2 and p 117–120
Rincon, A. Rincon, C. Corvo and Haces, NO2, J. Electrochem. Soc., Vol 140 (No. 76. J. Svensson and L.G. Johansson, A Labo-
Electrochemical, Gravimetric and X-Ray 1), 1993, p 53–58 ratory Study of the Initial Stages of the At-
Characterization of Low Carbon Steel Cor- 62. J.F. Dante and R.G. Kelly, The Evolution mospheric Corrosion of Zinc in the Pres-
rosion Rate and Products After Atmo- of the Adsorbed Solution Layer During At- ence of NaCl; Influence of SO2 and NO2,
spheric Exposure in the Caribbean Area, mospheric Corrosion and Its Effect on the Corros. Sci., Vol 34 (No. 5), 1993, p 721–
Corros. Rev., Vol 19, 5–6 Nov. 2001 Corrosion Rate of Copper, J. Electrochem. 740
46. L. Veleva, “Phase Transformation of Iron Soc., Vol 140 (No. 7), 1993, p 1890–1897 77. X.G. Zhang and H.Q. Tran, Effect of Cy-
Hydroxide in the Corrosion Products 63. A. Salnick, W. Faubel, H. Klewe-Neben- cling Wetting and Drying on Corrosion of
Formed in Humid Tropical Climate,” Pa- ius, A. Vend, and H.J. Ache, Photothermal Zinc and Steel, Cyclic Cabinet Corrosion
per 03602, 58th Annual Conference, Cor- Studies of Copper Patina Formed in the Testing, STP 1238, G.S. Haynes, Ed.,
rosion 2003, NACE, 2003 Atmosphere, Corros. Sci., Vol 37 (No. 5), American Society for Testing and Materi-
47. Z. Szklarska-Smialovwska, Review of Lit- 1995, p 741–767 als, 1995, p 125–135
erature on Pitting Corrosion Published 64. A.A. Ambler and A.A.J. Bain, Corrosion 78. P. Quintana, L. Veleva, W. Cauich, R.
Since 1960, Corrosion, Vol 27 (No. 6), of Metals in Tropics, J. Appl. Chem., Sept Pomes, and L. Peña, Study of the Com-
1971, p 223–233 1995, p 437–467 position and Morphology of Initial Stages
48. A.M. Beccaria, E.D. Mor, and G. Poggi, 65. L. Veleva, P. Quintana, R. Ramanauskas, of Corrosion Products Formed on Zn
Examination of Aluminum Corrosion R. Pomés, and L. Maldonado, Mechanism Plates Exposed to the Atmosphere of
Products in Marine or Industrial-Marine of Copper Patina Formation in Marine En- Southeast Mexico, Appl. Surf. Sci., Vol 99,
Atmosphere, Werkst. Korros., Vol 34 (No. vironments, Electrochim. Acta, Vol 41 1996, p 325–334
5), 1983, p 236–240 (No. 10), 1996, p 1641–1646 79. J.W. Spence, F.H. Haynie, F.W. Lipfert,
49. R.T. Foley, Localized Corrosion of Alu- 66. L. Veleva and M. Luja, SEM Characteriza- S.D. Cramer, and L.G. McDonald, Atmo-
minum Alloys—A Review, Corrosion, Vol tion of Copper Corrosion Products (Patina) spheric Corrosion Model for Galvanized
42 (No. 5), 1986, p 227–286 Formed in a Tropical Humid Climate, Br. Steel Structures, Corrosion, Vol 48 (No.
50. J.J. Friel, Atmospheric Corrosion Products Corros. J., Vol 34 (No. 1), 1999, p 34–36 12), 1992, p 1009–1019
on Al, Zn and Al-Zn Metallic Coatings, 67. S.D. Cramer, S.A. Matthes, B.S. Covino, 80. R.A. Legault and V.P. Pearson, Kinetics of
Corrosion, Vol 42 (No. 7), 1986, p 422– S.J. Bullard, and G.R. Holcomb, Environ- the Atmospheric Corrosion of Galvanized
426 mental Factors Affecting the Atmospheric Steel, Atmospheric Factors Affecting the
51. T.E. Graedel, Corrosion Mechanism for Corrosion of Copper, Outdoor Atmo- Corrosion of Engineering Metals, STP
Aluminum Exposed to the Atmosphere, J. spheric Corrosion, STP 1421, H.E. Town- 646, S.K. Coburn, Ed., American Society
Electrochem. Soc., Vol 136 (No. 4), 1989, send, Ed., American Society for Testing for Testing and Materials, 1978, p 83–96
p 204C–211C and Materials, 2002, p 245–264 81. M. Pourbaix, The Linear Bilogarithmic
52. J.O’M. Bockris and S.U. Khan, Aspect of 68. C.J. Slunder and W.K. Boyd, Zinc: Its Cor- Law for Atmospheric Corrosion, Atmo-
Aluminum Corrosion, Surface Electro- rosion Resistance, Zinc Institute Inc., 1971 spheric Corrosion, W.H. Ailor, Ed., Wiley,
Inc., 1982, p 107–121
chemistry: A Molecular Level Approach, 69. C.J. Slunder and W.K. Boyd, Zinc: Its Cor-
82. Yu.M. Panchenko and P.V. Strekalov, Cor-
Plenum Press, 1994, p 771–778 rosion Resistance, 2nd ed., T.K. Christman
relation between Corrosion Mass Losses
53. Z. Szklarska-Smialovwska, Pitting Corro- and J. Payer, Ed., Int. Lead Zinc Research
and Corrosion Product Quantity Retained
sion of Aluminum, Corros. Sci., Vol 41, Org., Inc., 1983
on Metals in a Cold, Moderate and Tropi-
1999, p 1743–1767 70. I. Suzuki, The Behavior of Corrosion
cal Climate, Prot. Met. (Russia), Vol 30
54. C.W. Hummer, Jr., Corrosion of Metals in Products on Zinc in Sodium Chloride So-
(No. 5), 1994, p 459–467
Tropic Environments—Copper and lution, Corros. Sci., Vol 25 (No. 11), 1985,
83. M. Morcillo, S. Feliu, and J. Simancas,
Wrought Copper Alloys, Mater. Prot., Jan p 1029–1034 Deviation from Bilogarithmic Law for At-
1968, p 41–47 71. S.D. Cramer, J.P. Carter, P.J. Linstrom, and mospheric Corrosion of Steel, Br. Corros.
55. E. Mattsson and R. Holm, Atmospheric D.R. Flinn, Environmental Effects in the J., Vol 28 (No. 1), 1993, p 50–52
Corrosion of Copper and Its Alloys, At- Atmospheric Corrosion of Zinc, Degra- 84. J.L. Crolet, The Electrochemistry of Cor-
mospheric Corrosion, W.H. Ailor, Ed., Wi- dation of Metals in the Atmosphere, STP rosion Beneath Corrosion Deposits, J. Ma-
ley, Inc., 1982, p 365–381 965, S.W. Dean and T.S. Lee, Ed., Amer- ter. Sci., Vol 28, 1993, p 2577–2588
56. T.E. Graedel, K. Nassau, and J.P. Franey, ican Society for Testing and Materials, 85. I. Odnevall, P. Verbiest, W. He, and C.
Copper Patinas Formed in the Atmo- 1988, p 229–247 Leygraf, The Influence of Patina Age and
sphere, Part I: Introduction, Corros. Sci., 72. F.H. Haynie, Environmental Factors Af- Pollutant Levels on the Runoff Rate of
Vol 27 (No. 7), 1987, p 639–657 fecting the Corrosion of Galvanized Steel, Zinc from Roofing Materials, Corros. Sci.,
57. T.E. Graedel, Copper Patinas Formed in Degradation of Metals in the Atmosphere, Vol 40, 1998, p 1977–1982
the Atmosphere, Part II: A Qualitative As- STP 965, S.W. Dean and T.S. Lee, Ed., 86. I. Odnevall, P. Verbiest, W. He, and C.
sessment of Mechanisms, Corros. Sci., Vol American Society for Testing and Materi- Leygraf, Effect of Exposure Direction and
27 (No. 7), 1987, p 721–740 als, 1988, p 282–289 Inclination on the Runoff Rates of Zinc
58. R.L. Opila, Copper Patinas: An Investi- 73. T.E. Graedel, Corrosion Mechanism for and Copper Roofs, Corros. Sci., Vol 42,
gation by Auger Electron Spectroscopy, Zinc Exposed to the Atmosphere, J. Elec- 2000, p 1471–1487
Corros. Sci., Vol 27, 1987, p 685–694 trochem. Soc., Vol 136 (No. 4), 1989, p 87. W. He, I. Odnevall, and C. Leygraf, A Lab-
59. R. Baboian, E.L. Bellante, and E.B. Cliver, 193C–203C oratory Study of Copper and Zinc Runoff
Ed., The Statue of Liberty Restoration, 74. A. Askey, S.B. Lyon, G.E. Thompson, J.B. during First Flush and Steady-State Con-
NACE, 1990 Johnson, G.C. Wood, M. Cooke, and P. ditions, Corros. Sci., Vol 43, 2001, p 127–
60. R.A. Livingston, Influence of the Environ- Sage, The Corrosion of Iron and Zinc by 146
208 / Forms of Corrosion

88. W. He, I. Odnevall, and C. Leygraf, A enced Corrosion Handbook, (NACE Item num, Galvanized Steel and Steel, Corro-
Comparison between Corrosion Rates and 37740) Industrial Press, 1994 sion, Vol 43 (No. 12), 1987, p 719–726
Runoff Rates from New and Aged Copper 100. H.A. Videla, Manual of Biocorrosion, 114. H. Grange, C. Bieth, H. Boucher, and G.
and Zinc as Roofing Material, Water, Air, (NACE Item 38214) CRC Press Inc., 1996 Delapierre, A Capacitive Humidity Sensor
and Soil Pollution: Focus 1, 2001, p 67– 101. B.J. Little, P.A. Wagner, and F. Mansfeld, with Very Fast Response Time and Very
82 CTME: Microbiologically Influenced Cor- Low Hysteresis, Sens. Actuators, No.12,
89. X. Zhang, W. He, I. Odnevall, J. Pan, and rosion, (Item 37546) NACE International, 1987, p 291–296
C. Leygraf, Determination of Instantane- 1997 115. D.M. Drazic and V. Vascic, The Correla-
ous Corrosion Rates and Runoff of Copper 102. V. Rainha and T. Fonesca, Kinetic Studies tion between Accelerated Laboratory Cor-
from Naturally Patinated Copper during on the SRB Influenced Corrosion on Steel: rosion Tests and Atmospheric Corrosion
Continuous Rain Events, Corros. Sci., Vol A First Approach, Corros. Sci., Vol 39 Station Tests on Steels, Corros. Sci., Vol
44, 2002, p 2131–2151 (No. 4), 1997, p 807–813 29 (No. 10), 1989, p 1197–1204
90. I. Odnevall and C. Leygraf, Environmental 103. H.A. Widela, A Practical Manual of Bio- 116. M. Stratmann and H. Streckel, On the At-
Effects of Metals Induced by Atmospheric corrosion and Biofouling for the Industry, mospheric Corrosion of Metals Which are
Corrosion, Outdoor Atmospheric Corro- (English, Spanish and Portuguese Edi- Covered with Thin Electrolyte Layers, Part
sion, STP 1421, H.E. Townsend, Ed., tions), (NACE Item 38251) CYTED Re- I: Verification of the Experimental Tech-
American Society for Testing and Materi- search Network XV BIOCORR, 1998 nique; Part II: Experimental Results; Part
als, 2002, p 185–199 104. C.A.C. Sequeira, Ed., “Microbial Corro- III: The Measurement of Polarization
91. S. Bertling, I. Odnevall, C. Leygraf, and sion,” (European Federation of Corrosion Curves on Metal Surfaces Which are Cov-
D. Berggren, Environmental Effects of Report 29) NACE Item 38331, IOM Com- ered by Thin Electrolyte Layers, Corros.
Zinc Runoff from Roofing Materials—A munications, Ltd., 2000 Sci., Vol 30 (No. 6/7), 1990, p 681–734
New Multidisciplinary Approach, Outdoor 105. P. Angell and K. Urbanic, Sulphate-Re- 117. J.W. Spence, E.O. Edney, F.H. Haynie,
Atmospheric Corrosion, STP 1421, H.E. ducing Bacteria Activity as a Parameter to C.D. Stiles, E.W. Corse, M.S. Wheeker,
Townsend, Ed., American Society for Test- Predict Localized Corrosion of Stainless and S.F. Cheek, Advanced Laboratory and
ing and Materials, 2002, p 200–215 Steel, Corros. Sci., Vol 42, 2000, p 897– Field Exposure Systems for Testing Ma-
92. W. He, W. Odnevall, and C. Leygraf, Run- 912 terials, Corrosion Testing and Evaluation:
off Rates of Zinc—A Four-Year Field and 106. T. Rap, T. Sairam, B. Viswanathan, and K. Silver Anniversary Volume, STP 1000, R.
Laboratory Study, Outdoor Atmospheric Nair, Carbon Steel Corrosion by Iron Ox- Baboian and S.W. Dean, Ed., American
Corrosion, STP 1421, H.E. Townsend, idizing and Sulfate Reducing Bacteria in a Society for Testing and Materials, 1900, p
Ed., American Society for Testing and Ma- Freshwater Cooling System, Corros. Sci., 191–207
terials, 2002, p 216–229 Vol 42, 2000, p 1417–1431 118. G. Nekoksa, Correlation of Failure Anal-
93. I. Odnevall, T. Korpinen, R. Sundberg, and 107. J.D. Sinclair, An Instrumental Gravimetric ysis Corrosion Testing, Corrosion Testing
C. Leygraf, Atmospheric Corrosion of Method for Indexing Materials, Contami- and Evaluation: Silver Anniversary Vol-
Natural and Pre-Patinated Copper Roofs in nants, and Corrosion Products According ume, STP 1000, R. Baboian and S.W.
Singapore and Stockholm—Runoff Rates to Their Hygroscopicity, J. Electrochem. Dean, Ed., American Society for Testing
and Corrosion Products Formed, Outdoor Soc., Vol 125 (No. 5), 1978, p 734–742 and Materials, 1900, p 66–79
Atmospheric Corrosion, STP 1421, H.E. 108. R.D. Smith, An Accelerated Atmospheric 119. A.M. Beccaria, Zinc Layer Characteriza-
Townsend, Ed., American Society for Test- Corrosion Test (AACT), Factor Affecting tion on Galvanized Steel by Chemical
the Corrosion of Engineering Metals, STP Methods, Corrosion, Vol 46 (No. 11),
ing and Materials, 2002, p 230–244
646, S.K. Coburn, Ed., American Society 1990, p 906–912
94. S. Matthes, S. Cramer, B. Covino, S. Bul-
for Testing and Materials, 1978, p 160–164 120. R.J. Cordner, Atmospheric Corrosion Sur-
lard, and G. Holcomb, Precipitation Run-
109. T.R. Shaw, Corrosion Map of the British vey of New Zealand, Br. Corros. J., Vol 25
off from Lead, Outdoor Atmospheric Cor-
Isles, Factor Affecting the Corrosion of En- (No. 2), 1990, p 115–118
rosion, STP 1421, H.E. Townsend, Ed.,
gineering Metals, STP 646, S.K. Coburn, 121. J. Scully, Electrochemical Methods for
American Society for Testing and Materi-
Ed., American Society for Testing and Ma- Laboratory Corrosion Testing, Corrosion
als, 2002, p 245–265–274
terials, 1978, p 204–215 Testing and Evaluation: Silver Anniver-
95. H. Michels, B. Boulanger, and N. Nikolai-
110. D. Kontkova and K. Barton, Corrosion sary Volume, STP 1000, R. Baboian and
dis, Environmental Impact of Stormwater Aggressivity of Atmospheres (Derivation S.W. Dean, Ed., American Society for
Runoff from a Copper Roof, Mater. Per- and Classification), Atmospheric Corro- Testing and Materials, 1900, p 351–378
form., Vol 42 (No. 2), 2003, p 70–74 sion of Metals, STP 767, S.W. Dean, Jr. 122. W.G. Clark, Jr. and M.J. Metala, The Non-
96. S. Matthes, S. Cramer, S. Bullard, B. Cov- and E.C. Rhea, Ed., American Society for destructive Characterization of Corrosion,
ino, and G. Holcomb, “Atmospheric Cor- Testing and Materials, 1982, p 225–249 Corrosion Testing and Evaluation: Silver
rosion and Precipitation Runoff from Zinc 111. A.B. Harker, F. Haynie, F. Mansfeld, D.R. Anniversary Volume, STP 1000, R. Ba-
and Zinc Alloy Surfaces,” Paper 03598, Strauss, and D.A. Landis, Measurement of boian and S.W. Dean, Ed., American So-
58th Annual Conference, Corrosion 2003, the Sulfur Dioxide and Sulfuric Acid ciety for Testing and Materials, 1900, p
NACE, 2003 Aerosol Induced Corrosion of Zinc in a 39–40
97. J.G. Stoecker, Guide for the Investigation Dynamic Flow System, Atmosph. Envi- 123. G. Haynes, Review of Laboratory Corro-
of Microbiological Induced Corrosion, ron., Vol 1 (No. 11), 1982, p 2691–2702 sion Tests and Standards, Corrosion Test-
Mater. Perform., Vol 20 (No. 9), 1981, p 112. C. Fiaud, M. Safavi, and J. Vedel, Identi- ing and Evaluation: Silver Anniversary
32 fication of the Corrosion Products Formed Volume, STP 1000, R. Baboian and S.W.
98. D.H. Pope, Discussion of Methods for the on Copper in Sulfur Containing Environ- Dean, Ed., American Society for Testing
Detection of Microorganisms Involved in ments, Werkst. Korros., Vol 35 (No. 8), and Materials, 1900, p 281–288
Microbiologically Influenced Corrosion, 1984, p 361–366 124. A. Sterling, A. Atrens, and I.O. Smith, Ac-
Biologically Induced Corrosion—Pro- 113. S.B. Lyon, G.E. Thompson, J.B. Johnson, celerated Atmospheric Corrosion of Cop-
ceedings of the International Conference, G.C. Wood, and J.M. Ferguson, Acceler- per and Copper Alloys, Br. Corros. J., Vol
NACE, 1986, p 275 ated Atmospheric Corrosion Testing Using 25 (No. 4), 1990, p 271–278
99. S.W. Borenstein, Microbiologically Influ- a Cyclic Wet/Dry Exposure Test: Alumi- 125. I. Odnevall and C. Leygraf, A Comparison
 Atmospheric Corrosion / 209

between Analytical Methods for Zinc Thick Film Sensor for Atmospheric Cor- Networks and a Deterministic Model,” Pa-
Specimens Exposed in a Rural Atmo- rosion Testing, Sens. Actuators B, No. 18– per 546, Proceedings of NACE’95 Corro-
sphere, J. Electrochem. Soc., Vol 138 (No. 19, 1994, p 558–561 sion Congress, NACE, 1995
7), 1991, p 1923–1928 136. S. Motoda, Y. Suzuki, T. Shinohara, Y. Ko- 146. A.O. Salnick and W. Faubel, Photoacous-
126. S.B. Lyon, G.E. Thompson, and J.B. John- jima, S. Tsujikawa, W. Oshikawa, S. Ito- tic FT-IR Spectroscopy of Natural Copper
son, Materials Evaluation Using Wet-Dry mura, T. Fukushima, and S. Izumo, ACM Patina, Appl. Spectrosc., Vol 49, 1995, p
Mixed Salt-Spray Tests, New Methods for (Atmospheric Corrosion Monitoring) Type 1516–1523
Corrosion Testing of Aluminum Alloys, Corrosion Sensor to Evaluate Corrosivity 147. L. Veleva, S. Thomas, E. Marrin, A. Cruz-
STP 1134, V.S. Agarwala and G.M. Ug- of Marine Atmosphere, Corros. Eng. Orea, I. Delgadillo, J. Alvarado-Gil, P.
iansky, Ed., American Society for Testing (Jpn.), Vol 43, 1994, p 538–594 Quintana, R. Pomes, F. Sanchez, H. Var-
and Materials, 1992, p 20–31 137. S. Motoda, Y. Suzuki, T. Shinohara, Y. Ko- gas, and L. Miranda, On the Use of Pho-
127. E.M. Rosen and D.C. Silverman, Corro- jima, S. Tsujikawa, W. Oshikawa, S. Ito- toacoustic Technique for Corrosion Moni-
sion Prediction from Polarization Scans mura, T. Fukushima, and S. Izumo, Cor- toring of Metals: Cu and Zn Oxides
Using an Artificial Neural Network Inte- rosive Factors of Marine Atmosphere Formed, Corros. Sci., Vol 39 (No. 9), 1997,
grated with an Expert System, Corrosion, Analyzed by ACM Sensor for 1 Year, Cor- p 1641–1655
Vol 48 (No. 9), 1992, p 734–744 ros. Eng. (Jpn.), Vol 44, 1995, p 253–265 148. J.W. Still and D.O. Wipf, Breakdown of
128. A.Y. Peng, S.B. Lyon, G.E. Thompson, 138. R. Alkire, and M. Verhoff, Electrochemical the Passive Layer by Use of the Scanning
J.B. Johnson, G.C. Wood, and J.M. Fer- Reaction Engineering in Materials Pro- Electrochemical Microscope, J. Electro-
guson, Comparison of Cross-Sectional Im- cessing, Chem. Eng. Sci., Vol 49 (No. chem. Soc., Vol 144 (No. 8), 1997, p 2657–
age Analysis with Weight Change Mea- 24A), 1994, p 4085–4093 2665
surements for Assessing Non-Uniform 139. R. Baboian, Ed., Corrosion Tests and Stan- 149. C. Leygraf and M. Forslund, Quartz Crys-
Attack during Corrosion Testing of Alu- dards: Application and Interpretation, tal Microbalance for In Situ Outdoor Cor-
minum, Br. Corros. J., Vol 28 (No. 2), ASTM Manual Series, American Society rosivity Studies, Paper 238, Proceedings of
1993, p 103–106 for Testing and Materials, 1995 NACE’98 Corrosion Congress, NACE,
129. W.S. Dean, Classifying Atmospheric Cor- 140. S.B. Lyon, C.W. Wong, and P. Ajiboye, An 1998
rosivity—A Challenge for ISO, Mater. Approach to the Modeling of Atmospheric 150. A. Garcia-Quiroz, S. Tomás, H. Vargas, A.
Perform., Vol 32 (No. 10), 1993, p 53–58 Corrosion, Atmospheric Corrosion, STP Cruz-Orea, J. Alvarado, L. Veleva, L. Mi-
130. D. Persson and C. Leygraf, In Situ Infrared 1239, W.W. Kirk and H.H. Lawson, Ed., randa, Photoacoustic Spectroscopy of Cor-
Reflection Absorption Spectroscopy for American Society for Testing and Materi- rosion Products of Copper Formed in
Studies of Atmospheric Corrosion, J. Elec- als, 1995, p 26–37 Tropical Environments, Instrum. Sci. Tech-
trochem. Soc., Vol 140 (No. 5), 1993, p 141. P.R. Roberge, What Is Accelerated in Ac- nol., Vol 26 (No. 2–3), 1998, p 241–260
1256–1260 celerated Testing: A Framework for Defi- 151. R. Ramanauskas, L. Muleshkova, L. Mal-
131. J. Stringer and A.J. Markworth, Applica- nition, Cyclic Cabinet Corrosion Testing, donado, and P. Dobrovolskis, Character-
tion of Deterministic Chaos Theory to Cor- STP 1238, G.S. Haynes, Ed., American ization of the Corrosion Behavior of Zn
rosion, Corros. Sci., Vol 35 (No. 1–4), Society for Testing and Materials, 1995, p and Zn Alloy Electrodeposits: Atmo-
1993, p 751–760 18–36 spheric and Accelerated Tests, Corros.
132. D.O. Wipf, Initiation and Study of Local- 142. G.A. King, Corrosivity Mapping: A Novel Sci., Vol 40 (No. 2–3), 1998, p 401–410
ized Corrosion by Scanning Electrochem- Tool for Materials Selection and Asset 152. E. Kálmán, Trends in Corrosion Research,
ical Microscopy, Colloids Surf. A: Physi- Management, Mater. Perform., Vol 34 (No. Electrochim. Acta, Vol 46, 2001, p 3607–
cochem. Eng. Aspects, Vol 93, 1994, p 1), 1995, p 7–9 3609
251–261 143. J. Tidblad and C. Leygraf, Atmospheric 153. W.T. Stephen, Increase Your Confidence in
133. A. Cox and S.B. Lyon, An Electrochemical Corrosion Effects of SO2 and NO2: A Corrosion Test Data, Mater. Perform., Vol
Study of the Atmospheric Corrosion of Comparison of Laboratory and Field-Ex- 40 (No. 3), 2001, p 58–61
Mild Steel, Part I: Experimental Method, posed Copper, J. Electrochem. Soc., Vol 154. H. Katayama, M. Yamamoto, and K. Ko-
Corros. Sci., Vol 36 (No. 7), 1994, p 1167– 142 (No. 3), 1995, p 749–756 dama, Atmospheric Corrosion Monitoring
1176 144. A. Nishikata, Y. Ichihara, and T. Tsuru, An Sensor in Outdoor Environment Using AC
134. A. Cox and S.B. Lyon, An Electrochemical Application of Electrochemical Impedance Impedance Technique, Outdoor Atmo-
Study of the Atmospheric Corrosion of Spectroscopy to Atmospheric Corrosion spheric Corrosion, STP 1421, H.E. Town-
Iron, Part II: Cathodic and Anodic Pro- Study, Corros. Sci., Vol 37 (No. 6), 1995, send, Ed., American Society for Testing
cesses on Uncorroded and Pre-Corroded p 897–911 and Materials, 2002, p 171–181
Iron, Corros. Sci., Vol 36 (No. 7), 1994, p 145. M. Urqudi and D. Van Voorthis, “Predic- 155. J. Repp, Accelerated Corrosion Testing—
1177–1192 tion of Single-Phase Erosion-Corrosion in Truths and Misconceptions, Mater. Per-
135. L. Fraigi, S.N. Gwirc, and D. Lupi, A Mild Steel Pipes Using Artificial Neural form., Vol 41 (No. 9), 2002, p 60–63