SCIENCE
186 CORROSION–MARCH 19930010-9312/93/000045/$5.00+$0.50/0© 1993, National Association of Corrosion Engineers
K Submitted for publication September 1991; in revised form, September
1992.
* Center for Interfacial Microbial Process Engineer, Montana State Univer-
sity, Bozeman, MT 59717.
KEY WORDS: carbon steel, depolarization, microbial
corrosion, pitting potential, polarization
INTRODUCTION
Under abiotic conditions, there is very little, if any,
corrosion of steel in the absence of oxygen under
neutral or alkaline conditions because the cathode
becomes polarized by the build-up of a layer of atomic
hydrogen.1 However, anaerobic corrosion may occur
in the presence of sulfide due to cathodic
depolarization (2 H2S + 2 e– —> 2 HS– + H2). The
attack was not only by dissolved sulfide but also by
iron sulfide.2 Iron sulfide causes cathodic
depolarization by decreasing the hydrogen
overpotential, but the site of the hydrogen reduction
reaction is unknown.
Biotic production of sulfide may lead to corrosion
of mild steel in anaerobic environments. For example,
chemostat (planktonic cells) investigations indicate
that microbial sulfide production leads to cathodic
depolarization primarily due to the iron sulfide
reaction.3 Booth et al.4,5 reported that no correlation
exists between hydrogenase activity and corrosion
rate when experiments were undertaken in
semicontinuous and continuous culture in high- or low-
ferrous medium. The corrosion rate was controlled by
the stability of an iron sulfide film on the steel surface.
In field investigations, Starkey reported that only slight
corrosion of a completely buried steel pipe was
ABSTRACT
Corrosion of mild steel under completely anaerobic
conditions in the presence of a mixed population biofilm,
including sulfate-reducing bacteria (SRB), has been studied
in a continuous flow system. The closed channel flow reactor
was continuously fed with low concentration substrate at
different dilution rates that influenced biofilm accumulation.
No direct correlation was observed between corrosion and
SRB activity in the absence of ferrous iron. Furthermore,
corrosion of mild steel in the SRB environment was
determined by the nature of the metal and environmental
conditions such as dissolved iron concentration. When
formation of an iron sulfide film on mild steel was prevented
before the biofilm accumulated, the metal surface retained its
scratch lines after a 21-day experiment (SRB at 2.6 x 109/
cm2). However, when the iron sulfide film was formed before
the accumulation of biofilm, visible localized corrosion
appeared after 14 days and increased up to 21 days.
Intergranular and pitting attack was found in the localized
corrosion area. Inclusions (Al, Mn, and Fe) and grain
boundary triple points were also found in the localized
corrosion area. At high iron concentration (approximately 60
mg/L in the bulk water), all biogenic sulfide was precipitated
and corrosion had significantly enhanced. Intergranular
attack was found over the entire metal surface.
Corrosion of Mild Steel Under Anaerobic
Biofilm K
W. Lee and W.G. Characklis*
SCIENCE
187CORROSION–Vol. 49, No. 3
detected in anaerobic mud supporting an active
community of sulfate-reducing bacteria (SRB).6
V.W. Kühr and van der Vlugt7 proposed the
cathodic depolarization theory for microbial (SRB)
corrosion-based on batch (closed) reactor
investigations. They suggested that hydrogenase
catalyzed the hydrogen reduction reaction on the iron
surface. King and Miller8 supported the hydrogenase
theory but proposed that cathodic depolarization
occurred on the iron sulfide instead of the iron surface.
Costello9 reports that desulfovibrio desulfuricans in a
batch reactor produces H2S, which directly depolarizes
the cathode. In summary, there is evidence for
cathodic depolarization as an important mechanism in
the anaerobic corrosion of iron and mild steel.
However, the exact mechanism, including the role of
hydrogenase, is not definitively understood.
Most microbial corrosion in nature and in
technological systems occurs in association with
microbial aggregates or biofilms. Hamilton10 has
developed a qualitative biofilm model of anaerobic
microbial corrosion where the growth of biofilm SRB
requires that the appropriate physicochemical
conditions be supplied by other facultative and
anaerobic organisms within the biofilm. To address the
anaerobic microbial corrosion problems in natural
environments, the activities and behavior of biofilm or
sessile SRB, not planktonic SRB, must be considered.
There is very little quantitative information
available related to the rate and extent of microbial
sulfide production in biofilms. Nielsen11 reports kinetic
data from a continuous flow SRB biofilm reactor,
which indicates that sulfate diffusion in the biofilm is
rate-limiting for SRB biofilm activity.
The relation between SRB activity and the
corrosion rate of mild steel is still not well established
despite decades of research. Progress in microbial
corrosion research has been limited by at least two
factors: (1) lack of experimental methods that simulate
biofilm processes in continuous systems and (2) lack
of analytical techniques that can be used in a biofilm
system while still providing adequate characterization
of the corroding metal surface.
This investigation was conducted with a mixed
population biofilm containing SRB in a continuous flow
closed channel reactor that simulates some of the
important characteristics of relevant industrial
systems. The objectives of the study were to deter-
mine (1) the effect of substrate loading rate on the
corrosion of mild steel in the absence of ferrous iron,
(2) the correlation of SRB activity with the cathodic
depolarization reaction, (3) the effect of the iron sulfide
film on the corrosion of mild steel in the presence of
biofilm, and (4) the effect of suspended iron sulfide on
the corrosion of mild steel in the presence of biofilm.
Substrate loading rate is the product of the dilution
rate and influent lactate concentration divided by the
surface area to volume ratio.
EXPERIMENTAL
Experimental Reactor
All experiments were conducted in a continuous
closed channel flow reactor with dimensions 1.00 by
0.15 by 0.30 m, containing a working volume of 4.5 L
with a surface area of 0.35 m2 (Figure 1). The reactor
had a recirculation loop to provide sufficient mixing
and a flow velocity of approximately 0.3 m/s. The
experimental system was purged with nitrogen to
minimize air permeation. Before entering the reactor,
the purge gas was passed through hot copper filings
to remove traces of oxygen. The reactor contained 24
corrosion coupons that were connected to a
potentiostat. The dilution rate (influent flow rate
divided by working volume) was the inverse of
residence time and was varied without influencing flow
velocity in the channel.
Media and Biofilm Growth Conditions
Bacteria were grown in 1/10 strength artificial
seawater, 200 mg/L lactic acid, 50 mg/L yeast extract,
50 mg/L NH4Cl, and 5 mg/L Na2HPO4. Ferrous iron
was absent in the media except in the last experiment.
The pH was controlled at 8 with 0.1 N NaOH and 0.1
N HCl; they were both flushed with nitrogen and
reduced by adding 0.013% Na2SO3. Temperature was
controlled at 30° C by a heat exchanger and cooler.
Desulfovibrio desulfuricans was inoculated into the
open-channel flow reactor, which, for the first week,
was operated as a batch reactor until surface
colonization occurred. Thereafter, substrate was
continuously fed to the reactor. The experiment was
not conducted aseptically in order to simulate natural
marine environments in which the SRB grow.
Coupons
Disks of AISI 1018 (UNS(1) G10180) mild steel,
15.9 mm in diameter, were cast in acrylic resin to
avoid the risk of crevice corrosion. The coupons were
polished with a series of grit papers (120, 200, 320,
400, 600) followed by ultrasonic cleaning in 100%
alcohol for one minute and dried in air at room
temperature.
Corrosion Measurements
Corrosion potential, cathodic polarization, pitting
potential, alternating current (AC) impedance, and
weight loss were measured in the open channel flow
(1) UNS numbers are listed in Metals and Alloys in the Unified Numbering
System, published by the Society of Automotive Engineers (SAE) and
cosponsored by ASTM.
SCIENCE
188 CORROSION–MARCH 1993
FIGURE 1. Continuous open channel flow reactor with dimensions 1.00  by 0.15  by 0.30 m and with a working
volume of 4.5 L.
both influent and effluent to complete material
balances.
Lactate was determined by an enzymatic method.
Acetate was measured with a gas chromatograph.
Sulfate concentration was determined using the
barium chloride turbidimetric method.12 Total dissolved
sulfide was determined by adding zinc acetate to
precipitate sulfide followed by the methylene blue
method.13 The redox potential was determined by
measuring the potential of a platinum electrode using
the potentiostat/galvanostat against a saturated
calomel electrode.
Biological Analysis
SRB numbers in the biofilm were determined by
the five tube multiple dilution most probable number
(MPN) method using Postgate B broth. A metal
coupon covered with biofilm was removed from the
reactor and transferred to an anaerobic tent in which
the biofilm was removed by brushing into 30 mL of
sterile 1/10 artificial seawater and homogenized to
disperse the biofilm. A dilution series to 10-10 was
prepared from 1 mL of the suspended biofilm
solutions. SRB numbers in the bulk water were
determined by the same MPN method for a 1 mL
sample. Numbers of general anaerobic bacteria (GAB)
both in the biofilm and bulk water were determined by
the same procedure as that for SRB except that fluid
thioglycollate medium was used.
reactor periodically for up to 21 days. Direct current
(DC) measurements were made using a potentiostat/
galvanostat and AC measurements were made using
a frequency response analyzer through an electrical
interface.
The potential of the working electrode was
measured against a saturated calomel electrode
(SCE). Two graphite rods were used as counter
electrodes. Cathodic polarization measurements were
performed from E
corr
 to 400 mV in the cathodic
direction at a scan rate 0.2 mV/s. Pitting potential
measurements were performed beginning at –800 mV
in the anodic direction with the same scan rate as
cathodic polarization. AC impedance measurements
were made using a sine wave voltage with an
amplitude of 7.5 mV at frequencies between 2 mHz
and 10 kHz. Ten frequencies were examined per
decade. At the end of each experiment, coupons were
removed from the reactor and cleaned in an ultrasonic
bath with an inhibited acid solution for 30 s prior to the
weight loss measurement.
Water Chemistry
Desulfovibrio desulfuricans uses lactate as a
carbon and energy source and oxidizes it to acetate
coupled with the reduction of sulfate to sulfide. As a
consequence, the chemical analysis of the reactor
fluid included lactate, acetate, sulfate, total dissolved
sulfide, and redox potential. Samples were taken from
SCIENCE
189CORROSION–Vol. 49, No. 3
Surface Analysis of Biofilm and Corrosion
Products
A biofilm sample was transferred anaerobically to
the anaerobic tent and fixed for 8 h in 2%
glutaraldehyde diluted with filter-sterilized artificial
seawater. Then, the sample was dehydrated in a
graded series of alcohol treatments (15 min in alcohol
30% to 95% and 30 min in 100%). The sample was
then dried using a critical point dryer for scanning
electron microscopy (SEM). Finally, the sample was
coated with Au-Pd alloy and examined with a
scanning electron microscope. The composition of the
corrosion products was analyzed with the SEM and
energy dispersive x-ray analysis (EDXA).
RESULTS AND DISCUSSION
Stoichiometry of a Mixed Population SRB
Biofilm on Corrosion Coupon
Stoichiometric determinations were conducted in
batch conditions after biofilm had accumulated in the
open channel flow reactor. The exhaustion of lactate
corresponded to plateau values for sulfate reduction
and sulfide production (Figure 2). The decrease in
sulfide concentration was due to the continuous
purging of nitrogen in the reactor. Both sulfate
reduction and sulfide production agreed with the
theoretical calculations based on the stoichiometry of
lactate oxidation:
 2 CH3CHOCOO– + SO4–2 —> 2 CH3COO–
 + 2 CO2 + S–2 + 2 H2O (1)
The stoichiometric ratios provide strong evidence that
SRB dominated the biofilm under the experimental
conditions.
Effect of Substrate Loading Rates
on Corrosion of Mild Steel
Biofilm accumulation was strongly dependent on
substrate loading rate. At constant influent substrate
(lactate) concentration and at low dilution rate (0.04
h–1), there was enough residence time (25 h) for the
organisms to grow in the bulk liquid. This resulted in a
thin biofilm. It appears that bulk liquid cells could
compete effectively with biofilm cells for substrate.
However, at high dilution rate (0.75 h–1), there was not
enough time for organisms to grow in the bulk liquid
and a thick biofilm accumulated. The formation of iron
sulfide precipitates on the mild steel surface was
purposefully prevented before the biofilm
accumulation to reduce the interference of iron sulfide
with the cathodic depolarization reaction. Dissolved
sulfide in the bulk liquid and hydrogen sulfide in the
gas phase were flushed away after some initial biofilm
had accumulated in the open channel flow reactor.
Then, corrosion experiments were initiated by
insertion of clean metal coupons into the reactor
followed by normal substrate input. The continuous
flow experiments were conducted for 21 days.
Biofilm accumulation was a function of substrate
loading rate (Figures 3 and 4), which was manipulated
by changing influent flow rate (not influent substrate
concentration). Biofilm thickness after 21 days
reached approximately 1,000 m m at the high substrate
loading rate while accumulating to approximately 5 m m
at the low substrate loading rate. At the high loading
rate, the biofilm was mainly composed of cells with a
small amount of extracellular polymeric substances
(EPS) (Figure 5). Chemical analysis of the bulk water
showed similar sulfide concentration, acetate
concentration, and redox potential for the two loading
rates. However, there was residual lactate in the high
substrate loading rate experiment even after 21 days
(Tables 1 and 2). At the low substrate loading rate, the
ratio of SRB in the biofilm to that in bulk water was
approximately 0.1. However at high substrate loading
rate, the ratio of SRB in the biofilm to that in bulk water
was greater than 100 (Figure 6 and 7), which indicate
that the increased substrate loading rate (i.e., higher
dilution rate) creates an environment more suitable for
SRB biofilm accumulation.
Specific SRB activity in bulk water or within biofilm
can be monitored indirectly by specific substrate
removal rate (substrate removed per cell per unit time)
or specific product formation rate (product produced
per cell per unit time). The specific substrate removal
rate for lactate is not a good indicator of SRB activity
in a mixed culture experiment since GAB may also be
removing lactate. The specific sulfide formation rate is
not a good indicator of SRB activity due to its volatile
FIGURE 2. Sulfide production, lactate consumption, and
sulfate degradation in mixed culture containing SRB. At-
tached picture (x 50) indicates no iron sulfide corrosion
products at the end of experiment.
SCIENCE
190 CORROSION–MARCH 1993
FIGURE 3. Scanning electron micrographs of biofilm grown
on a mild steel surface for (a) 3 days, x 1,300, (b) 7 days,
x 1,300, and (c) 21 days, x 1,300 at low substrate loading rate.
FIGURE 4. Scanning electron micrographs of biofilm accu-
mulated on a mild steel surface for (a) 3 days, x 30, (b) 7 days,
x 30, and (c) 21 days, x 30 at high substrate loading rate.
(c) (c)
(b)
(a)
(b)
(a)
SCIENCE
191CORROSION–Vol. 49, No. 3
sulfide film formed on the coupon (Figure 8). An x-ray
dot map (Figure 9) indicated no iron in the biofilm. No
cathodic depolarization occurred either at low (Figure
10) or high (Figure 11) substrate loading rates in the
absence of ferrous iron. At high substrate loading rate,
the cathode was slightly polarized as a thick biofilm
accumulated and was depolarized as the biofilm was
removed (Figure 12). The difference is attributed to
polarization by the thick biofilm. Presumably, a
diffusion barrier formed between the bulk water and
the metal surface as a thick biofilm accumulated. The
diffusion barrier prevented the diffusion of OH– or H2(cathodic reaction products) away from the metal
TABLE 2
Sulfate and Lactate Consumption and Sulfide and Acetate Production
in Bulk Water Containing Mixed Culture Anaerobic Bacteria and SRB
at High Substrate Loading Rate
Chemicals Sulfate Lactate Sulfide Acetate Redox
Time Consumed Consumed Produced Produced Potential
3 days 90 ppm 110 ppm 46 ppm 40 ppm –0.440 V
7 days 120 ppm 120 ppm 46 ppm 43 ppm –0.420 V
21 days 100 ppm 127 ppm 47 ppm 41 ppm –0.483 V
TABLE 1
Sulfate and Lactate Consumption and Sulfide and Acetate Production
in Bulk Water Containing Mixed Culture Anaerobic Bacteria and SRB
at Low Substrate Loading Rate
Chemicals Sulfate Lactate Sulfide Acetate Redox
Time Consumed Consumed Produced Produced Potential
3 days 125 ppm 182 ppm — — –0.495 V
7 days 115 ppm 202 ppm — — —
21 days 120 ppm 230 ppm 58 ppm 50 ppm –0.465 V
nature. However, the specific sulfate removal rate can
indicate the specific SRB activity. Specific sulfate
removal rate is defined as follows:
 qs
 
=  
D(Si – S)V
X
 (2)
where
q
s
= specific sulfate removal rate (M
s
M
x
t–1)
D = dilution rate (Q/V) (t–1)
Si = influent sulfate concentration (MsL–3)
S = effluent sulfate concentration (M
s
L–3 )
V = working volume in the reactor (L3)
X = total cells in the biofilm or bulk liquid (M
x
)
Q = influent flow rate (L3t–1)
The SRB activity can be obtained by following:
 SRB activity = q
s
X = D(Si–S)V (3)
According to Equation (3), SRB activity is 21.6 mg/h at
low substrate loading rate and 338 mg/h at high
substrate loading rate at 21 days. However, the
increased SRB activity did not enhance the corrosion
of mild steel. There was no detectable difference in
weight loss due to change in loading rates and no
pitting or localized corrosion observed on the metal
surface under open-circuit conditions after 21 days.
Metal surface analysis and electrochemical
measurements provide further evidence that corrosion
was insignificant. Scratch lines on the coupon were
observed after 21 days under the biofilm and no iron
FIGURE 5. Scanning electron micrograph of a biofilm accu-
mulated on a mild steel surface for 21 days at high substrate
loading rate (x 20,000).
SCIENCE
192 CORROSION–MARCH 1993
(a) (b)
FIGURE 6. MPN for GAB and SRB at (a) low substrate loading rate and (b) high substrate loading rate in the
bulk water at different exposure times (cell numbers represent the total cell numbers).
surface during polarization and reduced cathodic
current density.14 The same result was observed when
a coupon was coated with a similar thickness of 1.5%
agar instead of a SRB biofilm (Figure 13).
In a batch culture experiment, Salvarezza and
Videla15 used potentiostatic polarization techniques to
show that SRB influence the metal in a similar manner
to nonbiogenic sulfide, which shifts the pitting potential
to a more active value. The pitting potential is defined
as the potential where current increases dramatically.
The pitting potential becomes more noble as the
biofilm becomes thicker both at low (Figure 14) and
high (Figure 15) substrate loading rates. This indicates
that the pitting tendency of mild steel becomes more
difficult with time. Pits were detected on the metal
surface after anodic polarization at both substrate
loading rates (Figure 16).
The probable role of biofilms in the protection of
metal against corrosion in SRB environments has
been proposed by Gaylarde and Videla.16 This
investigation suggests that the biofilms possibly
(a) (b)
FIGURE 7. MPN for GAB and SRB at (a) low substrate loading rate and (b) high substrate loading rate in the
biofilm at different exposure times (cell numbers represent the total cell numbers).
increase the activation energy for hydrogen reduction
and/or recombination on the steel surface.
Corrosion of Mild Steel on Precoated
Iron Sulfide Film Followed by Biofilm
Accumulation
The steel surface was precoated with a layer of
iron sulfide film followed by biofilm accumulation at the
high substrate loading rate. The precoating procedure
consisted of flushing the mild steel coupons with 30
mg/L sulfide solution under anaerobic, abiotic
conditions at pH 8 for one day.
Localized corrosion under open-circuit conditions
was observed when biofilm developed on a precoated
coupon. The pitting potential shifted to a more active
value (150 mV, [Figure 17]), which is consistent with
observation after the biofilm and iron sulfide film was
removed. There was little change in cathodic
polarization curves as the experiment proceeded
(Figure 18). Microscopic examination revealed
intergranular attack at inclusions (Al, Mn, and Fe) and
SCIENCE
193CORROSION–Vol. 49, No. 3
FIGURE 9. X-ray dot map indicating the elemental distribution of iron, sulfur, and phosphorous on the top of biofilm and
the cross section of biofilm and metal surface.
grain boundary triple points found in the localized
corrosion area (Figure 19). Aggregates of bacteria and
iron sulfide crystals (confirmed by EDAX) were found
in the localized corrosion area underneath the iron
sulfide film and biofilm (Figure 20). The accumulation
of bacteria within the localized corrosion area may
accelerate corrosion by enhancing the cathodic
depolarization within the area and may lead to
incipient open pits or through-wall corrosion, which is
commonly found in industry and supposedly due to
microbial processes.
In general, biofilm accumulation appears to
reduce the spalling of iron sulfide film as compared to
the abiotic conditions under similar chemical
environments.17 However, the risk of localized
corrosion is significant. Initiation of the localized
corrosion process is believed to be strongly related to
metallurgical heterogeneities in the metal alloy matrix.
Following the initiation of localized corrosion, the
attack may be propagated if there are active SRB
within the localized corrosion area.
Effect of Ferrous Ion on Corrosion
of Mild Steel with a Precoated Biofilm
Mild steel coupons were precoated with biofilm at
a high substrate loading rate for one week in the
FIGURE 8. Scanning electron micrograph of a mild steel
surface after biofilm has been removed at 21 days (x 700).
absence of ferrous ions followed by weekly step
increases in ferrous ion to 1, 10, 60 mg/L.
The accumulation of iron sulfide in the biofilm at
different ferrous ion concentrations is illustrated in
Figure 21. Very little corrosion occurred at zero or low
ferrous ion concentrations (0, 1, and 10 mg/L), but
SCIENCE
194 CORROSION–MARCH 1993
FIGURE 11. Cathodic polarization curves of mild steel at high
substrate loading rate at different exposure times.
FIGURE 15. Anodic polarization curves of mild steel at high
substrate loading rate at different exposure times.
FIGURE 14. Anodic polarization curves of mild steel at low
substrate loading rate at different exposure times.
FIGURE 12. Cathodic polarization curves of mild steel before
and after biofilm has been removed at high substrate loading
rate.
FIGURE 13. Cathodic polarization curves of mild steel before
and after coupon is coated with 1.5% agar.
FIGURE 10. Cathodic polarization curves of mild steel at low
substrate loading rate at different exposure times.
SCIENCE
195CORROSION–Vol. 49, No. 3
FIGURE 17. Anodic polarization curves of mild steel on a
precoated iron sulfide film followed by biofilm accumulation at
different exposure times.
FIGURE 20. (a) Accumulation of bacteria and iron sulfide crystals in the localized corrosion area (x 100). (b)
Detached iron sulfide film above localized corrosion area (x 5,300).
(b)(a)
FIGURE 16. Scanning electron micrograph of a mild steel
surface indicating pitting attack after anodic polarization
(x 750).
FIGURE 18. Cathodic polarization curves of mild steel on a
precoated iron sulfide film followed by biofilm accumulation at
different exposure times.
FIGURE 19. Scanning electron micrograph of a mild steel
surface indicating intergranular attack in the localized corro-
sion area; inclusions (Al, Mn, and Fe) and grain boundary
triple points were found in this area (x 1,200).
SCIENCE
196 CORROSION–MARCH 1993
accumulation of iron sulfide particles increased in the
biofilm. There was no direct contact between iron
sulfide particles and the mild steel surface. However,
iron-rich media (60 mg/L) resulted in the precipitation
of all the biogenic sulfide (Figure 22). An x-ray dot
map indicates that the iron sulfide particles penetrate
through the biofilm and contact the metal surface
(Figure 23). Cathodic and anodic currents increased
dramatically as ferrous ions increased from 10 to 60
mg/L (Figures 24 and 25).
AC impedance measurements cannot accurately
detect localized corrosion. However, AC impedance
can assess the rate of generalized corrosion. Nyquist
plots indicated that the charge transfer resistance (R
ct,
the diameter of the semicircle, which is inversely
proportional to the corrosion rate) decreased
dramatically as ferrous ion concentration increased
from 10 to 60 mg/L (Figure 26). There was no
protective coating of iron sulfide film formed under the
biofilm in the iron-rich media. Intergranular attack was
observed over the entire metal surface after
precipitated iron sulfide and biofilm was removed
(Figure 27). Thus, corrosion of mild steel was
enhanced by the accumulated iron sulfide due to
cathodic and anodic depolarization.
In the anaerobic biofilms containing SRB,
corrosion rate can be accelerated through two
mechanisms: (1) corrosion rate is accelerated by
cathodic depolarization but is limited by diffusion of
hydrogen sulfide through the corrosion product layer
(concentration polarization), as observed under abiotic
conditions.18 (2) Corrosion rate is accelerated by
cathodic depolarization but is not limited by
concentration polarization. In this case, accumulation
of sessile SRB in the loose iron sulfide accelerates
corrosion presumably by reducing the diffusion
distance for the biogenic sulfide to reach the metal
surface.
Gaylarde and Johnson19 report that biofilm
(sessile) SRB on the metal surface results in a higher
corrosion rate as compared to the corrosion rate due
to planktonic bacteria alone. These observations
support the second mechanism. However, the
corrosion rate may be reduced when sulfate or lactate
diffusion within the biofilm limits the rate of microbial
activity.
The loose iron sulfide particles apparently play a
more important role than the bacteria in the anaerobic
corrosion process under the conditions of the reported
experiments. The iron sulfide particles are cathodic to
the mild steel but do not act as permanent cathodes
unless dissolved sulfide or SRB (which produce
dissolved sulfide) are present. The role of the mixed
SRB biofilm on the anaerobic corrosion of mild steel is
to continuously supply hydrogen sulfide to keep loose
iron sulfide particles cathodically active. This
(a)
FIGURE 21. Accumulation of iron sulfide on a precoated
biofilm at different ferrous ion concentrations. (a) 1 mg/L, x 50,
(b) 10 mg/L, x 50, and (c) 60 mg/L, x 50.
(c)
(b)
SCIENCE
197CORROSION–Vol. 49, No. 3
v The combination of biofilm process analysis with
electrochemical measurements and surface analysis
offers a useful tool for investigating microbial corrosion
problems.
v Corrosion cannot be initiated by SRB on mild steel
in the absence of ferrous ion, regardless of dilution
rate.
FIGURE 23. X-ray dot map indicating the elemental distribution of iron, sulfur, and phosphor on the top and cross
section of biofilm and metal substratum.
interpretation of the experimental results is consistent
with proposals by Miller.20
SUMMARY/CONCLUSIONS
The experimental investigations reported herein
provide strong evidence that corrosion of mild steel
was insignificant even in the presence of an active
SRB biofilm. This may be attributed to the high
activation energy for hydrogen reduction and/or
recombination. However, in the combined presence of
an SRB biofilm with dissolved oxygen, dissolved iron,
and iron sulfide, corrosion is enhanced. Hardy and
Bown21 demonstrated that high corrosion rates and
severe pitting attack on mild steel, sometimes
observed in the field, occurs when oxygen is
introduced into the system. Under completely
anaerobic conditions, iron sulfide may play a similar
key role in the corrosion process. The physical forms
of iron sulfide determined the type of corrosion (e.g.,
localized corrosion vs uniformly corrosion) in the
presence of SRB biofilm. Generalized attack was
observed where there was loose iron sulfide
precipitation under the SRB biofilm, while localized
attack was observed at the defect sites of protective
iron sulfide film.
FIGURE 22. Scanning electron micrograph of biofilm when
the media contain 60 mg/L ferrous ion concentration
(x 4,000).
SCIENCE
198 CORROSION–MARCH 1993
8420785, “An Investigation of Mechanisms for
Microbially Assisted Corrosion,” and the NSF Center
for Interfacial Microbial Process Engineering. The
authors thank V. Griffths, Montana College of Mineral
Science and Technology, for making SEM/EDAX
instrumentation available. A. Camper and D. Davies
are acknowledged for their valuable suggestions. The
IPA Industrial Associates are acknowledged for their
support of this work.
REFERENCES
1. P.F. Sanders, W.A. Hamilton, Proceedings of the International
Conference on Biologically Induced Corrosion (Houston, TX: NACE,
1985), p. 47.
2. J.S. Smith, J.D.A. Miller, Brit. Corros. J. 10 (1975): p. 136.
3. G.H. Booth, L. Elford, D.S. Warkerley, Brit. Corros. J. 3 (1988): p. 242.
4. G.H. Booth, P. M. Cooper, D. S. Wakerley, Brit. Corros. J. 1 (1988):
p. 345.
FIGURE 24. Cathodic polarization curves of mild steel on a
precoated biofilm followed by step increase in ferrous ion
concentrations.
FIGURE 25. Anodic polarization curves of mild steel on a
precoated biofilm followed by step increase in ferrous ion
concentrations.
v There is no cathodic depolarization and no
correlation between corrosion and SRB activity in the
absence of ferrous ion in a biofilm reactor at pH = 8.
v Accumulation of bacteria on a precoated iron
sulfide film reduces the possibility of spalling of sulfide
film but does not eliminate the potential for localized
corrosion.
v Corrosion of mild steel in the presence of biofilm
will be enhanced by suspended iron sulfide once the
iron sulfide particle contacts the metal surface
v Intergranular attack is always associated with loose
iron sulfide deposition.
ACKNOWLEDGMENTS
The authors acknowledge partial support from the
National Science Foundation project no. CTB-
FIGURE 27. Scanning electron micrograph of a mild steel
surface indicating the nonprotective iron sulfide film and
intergranular attack over the entire metal surface (x 550).
FIGURE 26. Nyquist plots for mild steel on a precoated biofilm
at different ferrous ion concentrations.
SCIENCE
199CORROSION–Vol. 49, No. 3
5. G.H. Booth, J.A. Robb, D.S. Wakerley, 3rd Int. Congress on Metallic
Corrosion, Moscow, 2 (1967), p. 542.
6. R.L. Starkey, Proceedings of the International Conference on
Biologically Induced Corrosion (Houston, TX: NACE, 1985), p. 3.
7. V.W. Kühr, van der Vlugt, Water 18 (1934): p. 147.
8. R.H. King, J.D.A. Miller, J.S. Smith, Brit. Corros. J. 8 (1973): p. 137.
9. J.A. Costello, South African Journal of Science 70 (1974): p. 202.
10. W.A. Hamilton, S. Maxwell, Proceedings of the International
Conference on Biologically Induced Corrosion (Houston, TX: NACE,
1985), p. 131.
11. P.H. Nielsen, Appl. Environ. Microbiol. 53 (1987): p. 27.
12. Standard Methods for the Examination of Water and Waste Water, 13th
ed. (Washington, D.C.: American Public Health Association, 1971).
13. J.D. Cline, Limnol Oceanogr 14 (1969): p. 454.
14. Z. Lewandowski, W.C. Lee, W.G. Characklis, B.J. Little, Corrosion 45
(1989): p. 92.
15. R.C. Salvarezza, H.A. Videla, Corrosion 36 (1980): p. 550.
16. C.C. Gaylarde,  H.A. Videla, International Biodeterioration 23 (1987):
p. 91.
17. W.C. Lee, W.G. Characklis, 8th International Symposium on
Biodeterioration and Biodegradation, Windsor, Canada (Essex, United
Kingdom: Elsevier, 1991), p. 89.
18. R.L. Martin, R.R. Annand, Corrosion 36 (1981): p. 297.
19. C.C. Gaylarde, J.M. Johnson, Microbial Adhesion to Surface, Berkeley,
R.C.W. Lynch, J.W., P.R. Rutter, B. Vincent, eds. (Chichester, United
Kingdom: Ellis Harwood Ltd., 1980), p. 511.
20. J.D.A. Miller, Metals, Microbial Biodeterioration, A.H. Rose, ed. (New
York, NY: Academic Press, 1981), p. 149.
21. J.A. Hardy, J.L. Bown, Corrosion 40 (1984): p. 650.