Microbial utilization of cathodic hydrogen and related corrosion

MISSING IMAGE

Material Information

Title:
Microbial utilization of cathodic hydrogen and related corrosion
Physical Description:
xiii, 185 leaves : ill. ; 29 cm.
Language:
English
Creator:
Sifontes, Jose Rafael, 1949-
Publication Date:

Subjects

Subjects / Keywords:
Agricultural Engineering thesis Ph.D
Dissertations, Academic -- Agricultural Engineering -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 170-181).
Statement of Responsibility:
by Jose Rafael Sifontes.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002028116
oclc - 33032691
notis - AKL5727
System ID:
AA00004727:00001

Full Text











MICROBIAL UTILIZATION OF CATHODIC HYDROGEN
AND RELATED CORROSION



















By

JOSE RAFAEL SIFONTES


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1994


































To the memory of my father, Pedro Rafael Sifontes Lopez.















ACKNOWLEDGEMENTS


I wish to thank the following persons for their

contributions throughout my graduate studies at the

Agricultural Engineering Department. My sincere appreciation

goes to Dr. David P. Chynoweth, chairman of my supervisory

committee, for his friendship, concern, and economic support

throughout this work. I also would like to thank Dr. Henry

Aldrich, Dr. Edward Lincoln, Dr. Mark Orazem, Dr. Roger

Nordstedt, and Dr. Ellis Verink for always being there when I

needed them and for access to their laboratories.

My special thanks go to all my friends and departmental

staff who encouraged me, particularly to Dr. K.T. Shanmugam,

Jose Moratalla, Paul Lane, Brian Ferber, Emo Crews, Larry

Miller, and Hyung-Jib Lee for their help and encouragement

during my stay at the Agricultural Engineering Department.

This work was initially supported by the Gas Research

Institute.

And most important, I want to thank God in the name of my

Lord Jesus Christ for allowing this to happen. To HIM be all

the Glory!


iii















TABLE OF CONTENTS


ACKNOWLEDGEMENTS ....................................iii

LIST OF TABLES ...................................... vi

LIST OF FIGURES .....................................vii

ABSTRACT ................................ ........... x

CHAPTERS

1 INTRODUCTION .............. .................... 1

Objectives ...................................... 1
Justification ..................................... 2
Background ............................................... 7
The Defective and Nonuniform Nature of
Metallic Surfaces ........................ 10
The Molecular Fluctuations of the
Electrical Double Layer ................. 12
The Mass Transport Perturbation in the
Diffusion Layer .......................... 13
Statement of the Problem ....................... 14

2 REVIEW OF LITERATURE .......................... 17

Historical Overview ........................... 17
Other Aspects of Microbial Corrosion ........... 24
Microorganisms Involved in Microbial Corrosion.. 27
Mechanisms of Microbial Corrosion .............. 31
Theory of Cathodic Depolarization .............. 34
Previous Studies in Microbial Corrosion ........ 37
Microbial Corrosion Control ................... 42
Research Approach .............................. 47

3 MATERIALS AND METHODS .......................... 49

Introduction ................................... 49
Flowthrough Bioreactor ......................... 50
Description ....................... ....... 50
Operation ................................. 54
Batch Bioreactor ............................... 58
Description ............................... 58
Operation ............................... 59









Electrochemical Cells Description .............. 60
Electrochemical Cells Operations ............. 62
Single Flask Electrochemical Cell ......... 62
Dual Flask Electrochemical Cell ........... 64
Triad Flask Electrochemical Cell .......... 65
Preparation of Bacterial Suspensions ........... 67
Preparation of Mineral Solution Electrolyte..... 69
Metal Coupon .................................. 70
Parameter Setting ............................. 71
Measurement Procedures .................... 73
Analytical Procedures ..................... 75
Gases ..................................... 76
Scanning Electron Microscopy .............. 77
Light Microscopy .......................... 77

4 RESULTS AND DISCUSSION ....................... 78

Flowthrough Bioreactor ......................... 78
Batch Bioreactor .............................. 92
Effect of Head Space Gas Composition ...... 93
Effect of Bacteria Combination .......... 95
Selection of Media ....................... 95
Selection of a Reducing Agent ............. 96
Selection of a Sterilization Method ....... 97
Experimental Results ..................... 98
The Electrochemical Cell .......................106
Single Flask Electrochemical Cell ..............113
Dual Flask Electrochemical Cell ..............117
Triad Flask Electrochemical Cell ...............125
Final Discussion .......................... 137
New Proposed Microbial Corrosion Mechanism .....154
Capabilities of the Triad Cell ................156

5 SUMMARY AND CONCLUSIONS .......................161

Summary ........................................161
Conclusions ....................................165
Suggested Future Research......................168

LITERATURE CITED ....................................170

BIOGRAPHICAL SKETCH ............................... 182














LIST OF TABLES


Table page

2-1 Chemistry of the cathodic depolarization theory... 35

3-1 Protocol for metallic surface fixation for
scanning electron microscopy examinations ........ 55

3-2 Glucose Lactate Yeast Extract (GLYE)
Medium Composition ............................... 56

3-3 Trace Vitamins Solution Composition .............. 57

3-4 Mineral Solution Composition .................... 57

4-1 Batch bioreactor volatile fatty acids
measurements from experiments using steel
coupons and bacterial triculture in mineral
solution ......................................... 99

4-2 Batch bioreactor head-space pressure drop in
GLYE and mineral solution in the absence of
bacteria ........................ ....... .........102

4-3 Hydrogen production by filings of Mg, Zn,
and SA106 steel in mineral solution in the
absence of bacteria after 3 days ................. 107

4-4 Free corrosion potential of SA106 steel
coupons at different surface polishing grades ....108

4-5 Galvanic series of selected metals and alloys
in mineral solution at 30C vs Ag/AgCl
reference electrode in the absence of bacteria ...110

4-6 Effect of media composition on the hydrogen
uptake of E. coli at 300C during 6 hours .........112

4-7 Bacteria hydrogen uptake capabilities in
mineral solution at 300C during 6 hours ..........113

4-8 Effect of different bacteria on the free
corrosion potential of Mg, Cu, and SA106 steel
coupons in mineral solution at 30 C ..............115








4-9 Dual cell schedule of experiments and
corresponding corrosion rates, using E. coli
(JW111) and SA106 steel coupons ..................118

4-10 Triad cell schedule of runs and related
corrosion USING E. coli (JW111) and SA106
steel coupons ....................................127


vii














LIST OF FIGURES


Figure page

3-1 Flowthrough bioreactor general description of
components ........................... .......... 51
3-2 Metal coupon and fluorescence probe setup ........ 53
3-3 Metal coupon, fluorescence probe and reference
electrode setup ............................. ..... 53
3-4 Flowthrough bioreactor components ................. 54

3-5 Batch Bioreactor ................................ 59

3-6 Schematic of electrochemical cell and data
acquisition system .............................. 61

3-7 Single flask electrochemical cell ................ 62

3-8 Dual flask electrochemical cell ................ 64

3-9 Triad flask electrochemical cell ................. 65

4-1 Flowthrough bioreactor results on bacteria
attachment to carbon steel coupons.
a) SEM micrograph showing bacteria within
the outer and inner structure of the
biofilm ....................................... 81
b) SEM micrograph indicating the nature of the
outer biofilm structure ....................... 81
c) SEM micrograph showing bacteria on inner
biofilm ....................................... 82

4-2 SEM characterization of bacterial species from
triculture.
a) Enterobacter aerogenes ......................... 82
b) Clostridium acetobutylicum ................ 83
c) Desulfovibrio desulfuricans ................... 83

4-3 Flowthrough bioreactor SEM results of microbial
corrosion of carbon steel coupons.
a) Control surface of coupon unexposed to
bacteria and polished to 240 grid ............. 85
b) Localized pitting near biofilm after exposure
to bacteria triculture ........................ 85
c) SEM micrograph showing microbial corrosion
near biofilm .................................. 86


viii










d) SEM micrograph showing generalized metal
deterioration after exposure to bacteria ...... 86
e) SEM micrograph showing elongated pits ......... 87

4-4 Flowthrough bioreactor results of EDXA analysis.
a) Control EDXA spectrum of a steel coupon
unexposed to bacteria indicating the metallic
components of the SA106 alloy ................. 87
b) EDXA spectrum of a metal surface exposed to
bacteria showing the sulfur peak resulting
from metal sulfides deposition by
Desulfovibrio desulfuricans ................... 88

4-5 Flowthrough bioreactor biofilm appearance
under the stereo microscope.
a) Flat-side view .............................. 88
b) Edge-side view .............................. 89

4-6 Flowthrough bioreactor volatile fatty acids
and pH profile ................................... 89

4-7 Flowthrough bioreactor-second experiment.
a) SEM micrograph showing appearance of biofilm .. 90
b) SEM micrograph indicating massive pitting ..... 91

4-8 Flowthrough bioreactor EDXA spectrum of metal
coupon exposed to bacteria ...................... 91

4-9 Batch bioreactor biofilm formed within
24 hours .................................... ... 100

4-10 Batch bioreactor SEM and EDXA results
on bacterial attachment to carbon steel.
a) Bacteria colonization within 24 hours .........100
b) EDXA spectrum of metal coupon exposed
to bacteria for 24 hours ......................101

4-11 Batch bioreactor-formation of a crystalline
film on the metal surface after one week of
exposure to the mineral solution in the absence
of bacteria.
a) SEM micrograph showing characteristics of the
crystalline film ..............................104
b) EDXA spectrum showing elemental distribution
of crystalline film on metal surface ..........104

4-12 Batch bioreactor-formation of crystalline film
on metal surface after exposure to bacteria.
a) SEM micrograph showing a transition zone
from biofilm to crystalline film ..............105








b) SEM micrograph showing a deteriorated
polymeric biofilm structure and pitting
after bacteria have died ...................... 105

4-13 Potential-time curve for SA106 steel in the
absence of bacteria ..............................109

4-14 Growth curve for E. coli at 600 nm, 370C and
trypticase soy broth medium ......................111

4-15 Potential-time curves for SA106 steel in the
presence of E. coli ...........................114

4-16 Potential-time curves for Mg in the presence
of E. coli ......... ............................... 114

4-17 Dual cell run potential-time curves.
a) Effect of E. coli, 6 hr old ...................121
b) Effect of E. coli, 15 hr old .................122
c) Effect of E. coli, 23 hr old ..................123
d) Effect of E. coli, 27 hr old and fumarate .....124

4-18 Dual cell run dissolved iron profiles.
a) Effect of E. coli, 6 hr old ..................121
b) Effect of E. coli, 15 hr old ................. 122
c) Effect of E. coli, 23 hr old ..................123
d) Effect of E. coli, 27 hr old and fumarate .....124

4-19 Triad cell run potential-time curves.
a) Effect of E. coli, 8 hr old ...................129
b) Effect of E. coli, 20 hr old and fumarate .....130
c) Effect of E. coli, >6 days old ...............132

4-20 Triad cell run dissolved iron profiles.
a) Effect of E. coli, 8 hr old ...................129
b) Effect of E. coli, 20 hr old and fumarate .....130
c) Effect of E. coli, >6 days old ................132

4-21 Triad cell artificial Hydrogen uptake test .......134

4-22 Artificial hydrogen uptake in a triad cell
using Cu and Mg electrodes .....................136

4-23 Hydrogen embrittlement of SA106 steel in
mineral solution electrolyte.
a) SEM micrograph of crack at a
low magnification ...........................148
b) SEM micrograph of crack at a
high magnification ............................148
c) EDXA micrograph of corrosion products
inside the crack .............................149








d) EDXA micrograph of uncracked control
surface .................................... 149

4-24 Proposed anaerobic mechanism for the on-set of
microbial corrosion of steel by hydrogen
oxidizing bacteria .............................155














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

MICROBIAL UTILIZATION OF CATHODIC HYDROGEN
AND RELATED CORROSION

By

Jose Rafael Sifontes

August 1994

Chairman: David P. Chynoweth
Major Department: Agricultural Engineering

Microbial corrosion is the deterioration of a material by

corrosion processes that occur directly or indirectly as a

result of the activity of microorganisms. Although it is a

widely recognized phenomenon in many industrial processes

worldwide, the fundamental corrosion mechanisms are not

understood, it is not well defined, and the microbial effect

has not been quantified. The objective of this research was

to show a relationship between cathodic hydrogen utilization

by bacteria and corrosion at its onset for the purpose of

improving its understanding and attempting to quantify the

biological component of corrosion.

Two new experimental flowthrough and batch bioreactor

systems were evaluated to reproduce anaerobic microbial

corrosion. Both reactors were able to reproduce microbial

corrosion in 24 hr. However, the batch system offered the


xii








better alternative to study microbial corrosion at an

attractive economic operation. Since microbial corrosion is

difficult to separate from pure electrochemical corrosion and

electrochemical measurements offers a nondestructive technique

to measure corrosion, the batch bioreactor was instrumented

for electrochemical measurements and for physicochemical

analysis.

Results from potential-time curves, dissolved iron

profiles, and differential corrosion currents indicated that

metals were reactive to the presence of bacteria. However,

the results were not adequate to quantify the biological

component of corrosion. Significant increases in corrosion

rates, observed during the addition of a terminal electron

acceptor, suggested a catalytic effect on the bacterial

hydrogenase system. The observation of cracks on steel

samples suggested the occurrence of hydrogen embrittlement.

The results of this research are important because they

offer a new approach for the investigation of microbial

corrosion, the triad cell offers an opportunity to quantify

the biological component of corrosion, and a new mechanism was

proposed for the understanding of the onset of anaerobic

microbial corrosion.


xiii














CHAPTER 1
INTRODUCTION


Objectives

The overall objective of this research was to show a

direct relationship between corrosion of metals and hydrogen

utilization by bacteria during the onset of microbial

corrosion, for the purpose of improving our understanding of

the phenomenon. A new approach to the investigation of

microbial corrosion was selected to break ground in an

alternative direction to the one established over the past

century and for a better comprehension of the fundamental

microbial corrosion mechanisms.

The specific objectives of the study were to

1) evaluate new experimental flowthrough and batch bioreactors

to adequately reproduce anaerobic microbial corrosion,

2) quantify the biological component of corrosion at its onset

by measuring indirectly hydrogen uptake by bacteria and

correlating it to total metal corrosion, using nondisrupting

electrochemical techniques,

3) incorporate into a comprehensive microbial corrosion model

all concepts developed during the investigation.










Justification


The ability of microorganisms to induce or influence the

deterioration of metals has been known for over a century, and

now it is a widely recognized problem in most industrial

processes worldwide. The involvement of microorganisms in

metal deterioration has led to the question of how biological

agents affect the classical corrosion mechanisms.

The true economic impact of metals deterioration is very

difficult to determine. However, some estimates have been

made in the USA and the UK. In the USA, the cost of corrosion

estimated by Poff (1985) and a projection by the National

Bureau of Standards for the year 1985 was approximately $140

billion and $170 billion, respectively. As for microbial

metal deterioration, studies performed by Butlin et al.

(1952), Allred et al. (1959), Booth (1964), and the National

Corrosion Service (Wakerly, 1969) estimated that metallic

corrosion caused by microbial intervention in a wide range of

industrial cases in the UK was on the order of 70%, 77%, 50%,

and 10%, respectively, of the total corrosion problem. In the

USA, Paternaude (1985) indicated that 50% of the steel culvert

pipe corrosion in Wisconsin was due to sulfate reducing

bacteria. If the lowest 10% is selected, the cost of

microbial corrosion represents a $17 billion per year problem

in the USA alone. Although such calculations may be open to

objections on matters of detail, the magnitude is likely to








3

stay high whatever route is followed to arrive at a final cost

because the above costs include replacement, prevention, and

maintenance but exclude losses of time, money, natural

resources, human suffering, and death due to equipment

failure.

Metallic corrosion is a natural process in which metals

return to their natural, oxidized states. It is an

interdisciplinary field of engineering and science where a

knowledge of the metallurgy of the metal, the environmental

conditions, the chemical composition, and the electrochemistry

of the system, are essential to understanding the process.

For example, hydrogen embrittlement is too complex for one of

the above disciplines to explain, and despite several decades

of investigation it is still unclear why hydrogen embrittles

some metals and alloys and not others (Oriani, 1987; Wilde and

Kim, 1986). Most corrosion processes are essentially surface

electrochemical mechanisms common to most metals in aqueous,

or at least humid, environments. This conclusion was first

reached by Whitney (1903). During the manufacturing process

of metals from their ores, metals are converted to a reduced

state, which makes them thermodynamically unstable in the

presence of oxygen, except gold.

Microbial corrosion is defined as the deterioration of

materials by natural processes that are directly or indirectly

related to the activity of microorganisms. The microbial

corrosion phenomenon has been known for nearly a century. It









4
has been reported to exist almost anyplace that microorganisms

colonize. In the process industry, it can happen inside and

outside of equipment and in aerobic or anaerobic environments.

Microbial corrosion is a diverse and complex phenomenon, and

the literature associated with it tends to be just as diverse

and complex. Microbial corrosion was well recognized and

established as a serious problem in the 1970s (Tatnall, 1981).

Microbial corrosion is difficult to separate from pure

electrochemical corrosion. Microorganisms involved in

microbial corrosion generally do not lead to a new form of

corrosion, but to a stimulation of the normal electrochemical

corrosion process. Consequently, if the microbes interfere

with an electrode process, with their own metabolism or

presence, it is a direct effect, otherwise it is an indirect

effect.

To demonstrate microbial corrosion, the presence of

microorganisms must be shown to induce or influence metal loss

during corrosion. Unfortunately, microbial corrosion in real

cases is very difficult to reproduce, extremely complex, and

difficult to model (Tatnall, 1988). Laboratory-based

controlled experiments using defined media and characterized

pure cultures do not often yield the expected results because

a broad range of variables makes it difficult to simulate the

dynamic natural environment (Allsopp and Seal, 1986).

The study of microbial corrosion has had a tendency to

consider processes in isolation rather than actual microbial









5
corrosion cases. For instance, there are many studies dealing

with the effect of a single species of microorganisms on metal

corrosion (Tatnall, 1988). While such investigations are

valuable in elucidating mechanisms, they give little insight

into the wider biological/metal/fluid interactions. Although

recent reappraisal of the role of bacterial consortia together

with studies of the effect of such consortia have improved our

knowledge of the microbial corrosion process, few studies of

microbial corrosion look beyond bacterial consortia and

surface corrosion effects to wider biological/metal/fluid

interactions. Furthermore, any living or indeed dead organism

that becomes associated with a metal surface immersed in an

electrolyte has the potential to influence the corrosion of

that metal. Edyvean (1988) in his recent work discusses the

interaction of both bacteria and macro- and micro-algae in the

fouling community on steel substrata in sea water.

Most engineering materials in general use are susceptible

to some form of microbial corrosion. Several of the metals

and alloys reported as being susceptible to microbial

corrosion include iron, copper, aluminum, nickel, cobalt, and

zinc and their alloys (Gabrielli, 1988; Zamanzadeh et al.,

1989; Griffin et al., 1989). The industries affected by

microbial corrosion have been identified as wastewater

facilities, water flood control systems, petrochemical

equipment, cooling water systems, underground structures and

pipelines, aircraft fuel systems, ships and marine structures,








6

chemical process industries, power generation industries, and

paper mills.

From case histories of microbial corrosion, the forms of

corrosion that are stimulated by the interaction of

microorganisms with metals range from general pitting

corrosion, crevice corrosion, and stress corrosion cracking to

enhancement of corrosion fatigue, intergranular stress

cracking, and hydrogen embrittlement with cracking (Sifontes

and Block, 1991).

The fundamental mechanisms that drive microbial corrosion

are not properly understood nor have they been well defined.

Consequently, additional research to understand them is

required. Most work to date in microbial corrosion has dealt

with reports of case histories or observations incidental to

the main study, with poor documentation regarding the

physical, chemical, and microbiological conditions under which

it occurred (Tatnall, 1981, 1988).

Data from the literature surveyed suggest that

qualitative relationships exist between the surface properties

of the various metals and their alloys and the extent of

biological response in relation to corrosion. The fact is

that it is not well known why some metals are more susceptible

to microbial corrosion than are others and why the microbial

corrosion effect of a particular microorganism is different

from the microbial corrosion effect of similar or different

species.










Background



The literature indicates that the first time a mechanism

for microbial corrosion was proposed was in 1934 by von

Wolzogen Kuhr and van der Vlught. This theory initiated

systematic studies on microbial corrosion and identified

formally the components of the microbial corrosion phenomenon,

namely, the metal surface, the suspended fluid, and the

microorganisms. They proposed that sulfur reducing bacteria

accelerated the corrosion of ferrous metals by cathodic

depolarization, that is, by removing adsorbed hydrogen from

the cathodic surfaces of the metal.

The theory appeared to be relatively simple to confirm by

conventional electrochemical techniques. However, qualitative

confirmation has been obtained only for specific hydrogenase-

positive species of microorganisms such as sulfate reducing

bacteria (Booth and Tiller, 1960) and methanogens (Daniels et

al., 1987). Belay and Daniels (1990) reported that besides

iron, other metals (Al, Zn, Ni, and As) can also be oxidized

via cathodic depolarization.

In the work performed by Daniels and coworkers (1987), it

was demonstrated that methanogenic bacteria use either pure

elemental iron or iron from mild steel as a source of

electrons in the reduction of carbon dioxide to methane.

These bacteria use the oxidation of iron for energy generation

and growth. The mechanism of iron oxidation is cathodic








8

depolarization, in which electrons from iron and hydrogen ions

from water produce molecular hydrogen, which is then released

for use by the methanogens.

The extent of bacterial adhesion is determined by the

surface properties of the phases involved and the need of

bacteria to locate energy sources. Bacterial cells will stick

to most surfaces, whether the surfaces are those of other

cells or merely inert material such as metals. In nutrient-

poor environments (e.g., water transmission pipelines), most

bacteria grow attached to surfaces mainly due to

hydrophobicity.

It is the belief of some researchers that initially,

sessile bacteria adhere randomly to metal surfaces by means of

their production of extracellular polysaccharides. The

continued production of the polysaccharide and the

reproduction of the bacteria lead to the development of

biofilms in which a consortium of cells interact in a hydrated

matrix of anionic polysaccharide polymers that provide

protection from natural or synthetically produced

antimicrobial agents (Costerton and Geesey, 1985).

As a bacterium begins to proliferate within biofilms, its

metabolic products stimulate the growth of other organisms.

As the different microorganisms develop, molecular or proton

exchanges occur; consortia such as these have been detected

and associated with microbial corrosion of metallic surfaces.








9

In many neutral solutions the corrosion of the common

structural metals appears to be associated with the flow of

electric currents between various parts of the metal surface

at finite distances from one another (Mears and Brown, 1941).

This statement is supported by much evidence in the case of

steels, where the quantities of current flowing during

corrosion account for the amount of corrosion that occurs. In

other words, the corrosion of metals and their alloys in

neutral solutions is electrochemical in nature. Furthermore,

for corrosion to occur, it requires all four components of an

electrochemical cell which include: 1) electrodes, 2)

electrolyte, 3) potential difference, and 4) electrical

continuity. When one of these conditions is regulated, then

the corrosion process is controlled.

Among some of the nonbiological factors of microbial

corrosion that corrosion engineers and scientists have

associated with corrosion and have known for several decades

are 1) impurities in the corroding metal, 2) grain boundaries,

3) orientation of grains, 4) differential grain size, 5)

differential thermal treatment, 6) surface roughness, 7) local

scratches and abrasions, 8) difference in shape, 9)

differential strain, 10) differential pre-exposure to air or

oxygen, 11) differential concentration or composition of the

corroding solution, 12) differential aeration, 13)

differential heating, 14) differential illumination, 15)

differential agitation, 16) contact with dissimilar metals,








10

17) externally applied potential, and 18) complex cells (Mears

and Brown, 1941). Those factors are closely related with the

so-colled "eight forms" of corrosion, which include uniform,

galvanic, crevice, pitting, intergranular, dealloying,

erosion, and stress corrosion (Fontana and Green, 1987).

In addition to the above, there is always a lack of

homogeneity at the metal-electrolyte interface during the

microbial corrosion of metals. Consequently, several of the

following specific interactions need to be recognized (Sato,

1987).



The Defective and Nonuniform Nature of Metallic Surfaces

During the formation of a metal or an alloy, different

elements compose the molten metal. When solidification

starts, metal is formed of crystalline grains that grow in

size and they meet each other to form grain boundaries until

the metal is all solid. Once the grains are formed, the

overall energy of the individual metal atoms

lowers and the grain boundaries remain as sites of higher

energy or higher thermodynamic instability (Smith, 1986).

The process of electron transfer is suspected to take

place at metallic surface locations of higher thermodynamic

instability such as the grain boundaries. This fact may be

explained by the nature of the metallic bond where the metal

ions occupy positions in the crystal lattice of the grains and

the outer valence electrons are shared by the surrounding









11
metal atoms. These electrons are like a cloud that moves

freely throughout the lattice and binds the crystal together.

Consequently, microorganisms can take advantage of these

electrons as their energy source for growth and affect the

stability of the metal atoms.

In addition, the steel-making process and treatment

influences the microstructure of the alloy by producing

microcompositional differences that have different

electrochemical behavior, which leads to the formation of

localized corrosion. Other changes in the alloy's internal

structure, such as aging and welding, may occur after it has

been established during fabrication.

Among other metallurgical factors to be considered at the

onset of microbial corrosion are the following

a) Composition of alloying elements and impurities

(inclusions), cooling conditions and post heat treatments have

a marked influence on their size, shape, number, and

distribution. These constituents are never uniformly

distributed throughout the alloy and certain elements tend to

congregate in localized concentrations as second phase

particles in grain boundaries, as pure compounds or as

intermetallic compounds (Godard,1980). These particles play

a major role in the corrosion behavior of alloys, especially

in the case of pitting corrosion.

b) Surface contamination by mill scale during the rolling

of the steel causes severe localized corrosion at breaks and








12

imperfections in the surface and may need to be removed before

service. Steel can also be contaminated during forging with

particles of metal from the forming equipment and consequently

produce serious localized corrosion, too.



The Molecular Fluctuations of the Electrical Double Layer

For iron, there exists a spatial fluctuation of the

electrical double layer potential on the metal surface in the

order of 0.8 V for every 2-3 atomic distances (Sato, 1987).

One possible reason is that, in addition to the heterogeneous

metal surface, water dipoles tend to point nearly

perpendicular to the metal surface, while the two nearest

dipoles tend to align antiparallel to each other in order to

minimize their dipole interaction energy. Consequently,

reorientation of water dipoles in the electrical double layer

occur in a time scale much greater than the time scale

relevant to the electron fluctuation. In addition, water

dipoles require greater time to reorient at surface defects

sites where the adsorption energy of water dipoles is greater

(Sato, 1987). This fact causes the electron transfer to take

place unevenly on the metal surface. At certain surface

defect sites, local enhancement of the electrical field in the

electrical double layer may be maintained for a sufficiently

long time so that the electrode reaction takes place

preferentially at these sites, making them active reaction

sites. Other sources of the spatial non-uniform electrical








13

double layer potential are specific ion adsorption. In this

case, the ions divest of their hydration water molecules and

come into direct contact with the metal surface. This causes

a local enhancement (approximately 3 times) of the electrical

field in the inner part of the electrical double layer. This

phenomenon takes place preferentially at certain lattice or

defect sites and will generate active spots for some electrode

reactions.



The Mass Transport Perturbation in the Diffusion Layer

This dynamic perturbation causes the local ion

concentration and the local electric field to fluctuate in the

diffusion layer. In addition, when the corroding metal

surface is usually covered with a porous corrosion precipitate

film of hydrated metal oxides or insoluble salt, selective

mass transport occurs, and this film either accelerates or

decelerates further corrosion of the underlying metal. When

those interactions are favorable to corrosion, chemical and/or

electrochemical active surface sites occur that result in a

corrosion activity. The nonuniform nature of the metal-

electrolyte interface, whether passive or not, implies that

the corrosion process takes place preferentially at specific

sites that somehow differ energetically from other parts of

the metal surface. These specific sites may be permanently

localized or spatially fluctuate during the progress of

corrosion.








14

As a result of the above, the microbial corrosion

phenomenon is presumably initiated by the interaction of the

metal, the bacteria, and the electrolyte at the onset of the

process, followed by the formation of either differential

concentration cells or local electrochemical corrosion cells

within the biofilm, which further complicates the corrosion

mechanism.

Microorganisms selected for this study had to contain a

hydrogenase enzyme. Such microbes belong to the facultative

lithotrophs, sulfur oxidizing bacteria, phototrophics,

methanogens, and denitrifying bacteria. These bacteria are

all hydrogen oxidizing bacteria that use molecular hydrogen as

their electron donor. In anaerobic environments hydrogen is

generated from the processes of fermentation or metal

oxidation. Hydrogen oxidizing bacteria may grow

autotrophically on hydrogen carbon dioxide as an electron

acceptor.



Statement of the Problem



The cathodic depolarization theory of von Wolzegen Kuhr

and van der Vlugt (1934) has become the center of attention of

researchers in the field because it proposes a logical

separation of the components of corrosion: biological,

metallic, and fluid. As a result, much of the literature on








15

microbial corrosion has been influenced by this theory and has

referred to it, either to criticize or prove it.

In work performed by Daniels and coworkers (1987), it was

demonstrated that methanogenic bacteria used hydrogen produced

from iron oxidation of carbon dioxide to methane. The

mechanism of iron oxidation is cathodic depolarization, in

which electrons from iron, and hydrogen ions from water

produce molecular hydrogen that is then released and used by

methanogens.

Most important, according to the literature reviewed,

nobody has been able to determine how microorganisms induce or

influence the corrosion process. Inspired by the cathodic

depolarization theory and by Daniels and coworkers'

demonstration, this research indicated that it might be

feasible to measure and correlate total metal corrosion and

hydrogen uptake by bacteria.

The project was initiated while evaluating some new

bioreactor systems. In an attempt to understand and quantify

the biological component of corrosion, an experimental system

based on the batch bioreactor, which includes electrochemical

measurement devices to study the utilization of cathodic

hydrogen by bacteria and its relation to the corrosion at its

onset, was implemented. Further along, data acquisition

hardware and computer software were adapted to the system.

The hypothesis is as follows: The corrosion rate of a

metal, under anaerobic conditions, in the presence of








16

bacteria, is a function of its free corrosion potential and

the ability of microorganisms to utilize cathodic hydrogen.














CHAPTER 2
REVIEW OF LITERATURE

Historical Overview


Microorganisms can exist almost anywhere and so can

microbial corrosion. In process industries, it can occur

inside and outside of equipment. In soil and water, it can

happen in aerobic or anaerobic environments.

Microbial corrosion of steel by sulfate reducing bacteria

was probably the first area investigated in this field of

study and remains one of the most important. In an earlier

work (Sifontes and Block, 1991), the authors covered

extensively a historical review on the subject. Here some of

the main historical aspects are presented here along with a

fresh look of microbial corrosion.

Microbial corrosion was first reported before the turn of

the century by Garret (1891). He ascribed the corrosion of

the lead sheathed cable to the action of bacterial metabolites

(ammonia, nitrates, and nitrites). In 1910, Gaines defined

the problem more clearly, providing evidence that iron and

sulfur bacteria were involved in the corrosion of the inside

and the outside of water pipes by making evident the presence

of abnormally large quantities of sulfur. He showed that

Gallionella, Sphaerotilus, and sulfate reducing bacteria were








18

responsible for the corrosion of ferrous alloys buried in

soil.

In 1924, Bengough and May demonstrated the effect of

ammonia produced by bacteria on the corrosion of copper

alloys. Later, von Wolzogen Kuhr and van der Vlugt (1934)

reported on the anaerobic corrosion of ferrous metals by

sulfate reducers. They proposed for the first time a

mechanism for microbial corrosion, which actually initiated

the systematic studies on microbial corrosion and

differentiated the various components of the microbial

corrosion process (metal, liquid, and microorganisms).

Details on the theory are presented later in this chapter.

Evidence for this bacterially influenced corrosion

continued to accumulate from around the world and was reviewed

by Starkey and Wright (1945). In 1953, Uhlig reported that

the primary role of slime-forming microorganisms was the

production of differential aeration and concentration cells

type corrosion. The first studies that demonstrated cathodic

depolarization with sulfate reducing bacteria was conducted by

Booth and Tiller (1960). They added a new dimension to the

problem, indicating that depolarization occurred with a

hydrogenase positive strain of Desulfovibrio bulgaris and did

not occur with a pure strain of hydrogenase negative

Desulfovibrio orientis. Furthermore, they demonstrated an

additional phenomenon that complicated matters: 1)

depolarization was observed when the culture was in active








19

growth, 2) the stimulation of corrosion was approximately

similar for both microbes, 3) the FeS film that formed on the

iron samples had an apparent inhibitory effect on corrosion

rates, and 4) the corrosion rates reported were much lower

than reported rates for similar alloys in natural anaerobic

environments in the presence of sulfate reducing bacteria.

Booth and Tiller (1962) and Booth, Robb, and Wakerly

(1967) demonstrated also that cathodic depolarization was

affected by FeS precipitate, presented experimental evidence

that the structure of the FeS film was instrumental in the

corrosion process, and indicated that once formed its

depolarizing activity continued even in the absence of

bacteria.

Since the energy crisis of the 1970s, several reviews on

the subject have been published. The results reported by Mara

and Williams (1972), King, Miller, and Smith (1973), Smith and

Miller (1975), and King, Dittmer, and Miller (1976) suggest

that precipitated FeS may initially form a protective film on

a ferrous metal surface in the presence of sulfate reducing

bacteria. As the microbial corrosion process continues, the

film thickens and changes stoichiometrically. As the ratio

Fe/S in the film changes from a sulfur deficient to a sulfur

rich structure, the film becomes less protective and

eventually spalls. Once spalled, the film does not reform and

vigorous anodic activity proceeds at the exposed metal

surfaces. According to Smith and Miller (1975), the FeS film,








20

regardless of structure, is cathodic to iron and the corrosion

process continues galvanically. Smith (1980) indicated that

the FeS film would not remain permanently cathodic in the

absence of bacteria. The role of the sulphate reducing

bacteria, he suggested, could be either to depolarize the FeS,

enabling it to remain cathodic, or to produce more FeS by

their metabolism.

Iverson (1981) discounted the FeS argument in his paper.

He indicated corrosion rates were above 210 mpy for mild steel

specimens exposed to filtered media from actively growing

culture of Desulfovibrio (API strain). He also suggested that

SRB produced a highly corrosive compound in addition to

hydrogen sulfide. The process appeared to depend on whether

FeS formed a protective film before the highly corrosive

product contacted the metal surface. Thus, it was apparent

that a number of factors from the metal and the solution were

involved in the process of microbial corrosion by sulphate

reducing bacteria.

During the 1980s, researchers found evidence of more

factors involved during anaerobic microbial corrosion.

Volatile metabolites such as phosphine have been reported to

be responsible for microbial corrosion of steels in

environments free of sulfate and sulfide (Iverson, 1985).

Since iron phosphide was detected among the corrosion products

found, it seemed that the amount of the chemical reaction was

the result of the competence between the sulfide that would








21

passivate the metal and the volatile phosphorous (phosphine)

that would replace the sulfide. There was no experimental

evidence of the chemical nature of the corrosion product,

excluding the direct contact between the bacteria and the

metal. Postgate (1979) rather concisely accounted for the

variety of factors involved in the microbial corrosion

process: nature of metal surface, dissolved ions and/or

organic matter, biofilm formation, FeS precipitate forms, and

other ions (sodium, chlorine) present.

On the other hand, the role of microorganisms in aerobic

corrosion was postulated by Olsen and Szaybalski (1949) to be

due in part to the formation of tubercles in conjunction with

microbial growth, which initiates oxygen concentration cells.

This mechanism, along with others, was proposed as the cause

of the worldwide problem of microbiologically associated

aluminum aircraft wing-tank corrosion that surfaced in the

late 1950s and early 1960s. Both commercial and military

aircraft were affected. Many microorganisms were reported to

be present in significant number in the fuel tank sludge

(Churchill, 1963). The same year, Leathen and Kinsel obtained

184 isolates of microorganisms from jet fuel-storage tanks at

nine Air Force bases. Results indicated the presence of

bacteria and fungi. The predominant bacteria were species of

Pseudomonas, and the most prevalent fungi were species

Hormodendrum. This case indicated the presence of a variety

of organisms including fungi, bacteria, and yeast at a site








22

of microbial corrosion. The medium was also complex due to

the presence of two liquids water and fuel phases.

Later work suggested Cladosporium resinae was the most

important fungus encountered in wing tanks and considered it

responsible for filter blockage and metal corrosion (Berner

and Ahearn, 1977). In 1976, Hill indicated that Cladosporium

resinae is only found in subsonic aircraft. In supersonic

aircraft, a higher temperature prevails and the predominant

flora found were Aspergillus fumigatus, gram negative

bacteria, yeast, and other fungi.

Recent evidence appears to indicate that organic acids,

produced by fungi, were primarily involved in this corrosion

(Miller, 1981).

Other problems due to the activities of SRB have arisen

in offshore oil operations (Hamilton and Sanders, 1986). They

were identified in the legs and storage cells of offshore

structures and include the production of hydrogen sulfide,

which is a serious personnel hazard, and the production of

bacterial metabolites that give rise to accelerated concrete

deterioration (Wilkinson, 1983). Furthermore, internal and

external microbial corrosion of long large-capacity subsea

pipelines, which transport oil and gas from offshore

production fields to shore, are of a major concern due to the

high cost associated with their failures and the harm to their

surrounding environment (King et al., 1986).








23

A considerable amount of research on microbial corrosion

in relation to sulphate reducing bacteria has been done in the

UK, especially in the 1960s. Most of this research was

carried out at the National Chemical Laboratory at Teddington

(presently the National Physical Laboratory) by a group of

well-known researchers in the areas of microbiology and

corrosion science. Despite the disbandment of the group in

1968 as a result of a decision of the British Government that

microbiology should not be conducted at that institution, the

researchers have continued to dominate the literature on

microbial corrosion (Tiller, 1985). Among some of the

original members of the group are G.H. Booth, D.S. Wakerly,

A.K. Tiller, J.R. Postgate, R.A. Kim, W.A. Hamilton, J.A.

Hardy, E.C. Hill, and B.N. Herbert.

In the USA, interest in the field has increased in the

past 15 years, but the scope has been limited. The most

noteworthy efforts began 12 years ago when a microbial

corrosion program was initiated by the Materials Technology

Group and a separate symposium and technical committee on

microbial corrosion was created by the National Association

of Corrosion Engineers (NACE) for their annual meetings.

Among some of the principal investigators in the USA are D.C.

White, R. Mitchell, G. Geesey, W. Characklis, W. Lee, N.

Dowling, R. Tatnall, G. Licina, B. Little, and J. LeGall.

During the last 10 years interest in the area has spread

in other countries.










Other Aspects of Microbial Corrosion



Despite advancement in the area of microbial corrosion,

some researchers have detected some common mistakes in the

practice of this type of investigation. Bryant and Laishley

(1989, 1993) indicated that the result of uncertain

conclusions and overestimates of corrosion due to influence of

microorganisms is the failure to account for effects of media

constituents such as phosphate. They found that phosphate was

one factor among many that together contribute to the overall

corrosion rate and that corrosion studies should include

controls that account for the constituents of the growth

media. There was little or no information on the levels of

phosphates on corroding metals. However, sea water contains

normally a concentration of 3 pM of phosphate, and dead cells

may release some phosphate that may raise the concentration in

spacial locations such as crevices and pitts and influence

corrosion. Booth and Tiller (1968) did not recommend the use

of batch cultures of sulfate reducing bacteria directly to

study their effect on the cathodic reaction. They suggested

that batch bioreactors were undesirable because the

composition of the media would be changing continuously due to

the release of metabolites including sulfide, which was known

to have a cathodic effect on the corrosion of steel and in the

presence of dissolved iron results in the precipitation of

FeS.








25

Hardy (1983) indicated that many attempts of researchers

to demonstrate the utilization of cathodic hydrogen by

sulphate reducing bacteria and its stimulatory effect on the

cathodic reaction have failed because the evidence was based

exclusively on measured electrochemical effects. The observed

depolarization might have been due to the effect of sulfides

present and could not be attributed exclusively to the effect

of hydrogenase. He also indicated that there was some

evidence that the periplasmic hydrogenase enzyme that appears

to be the hydrogen uptake enzyme constituted a greater

proportion of the total hydrogenase present in hydrogen-grown

cells than of the hydrogenase in lactate-grown cells. This

feature, as explained earlier, is a characteristic of bacteria

adaptation that is seldom considered during microbial

corrosion research. Researchers most of the time use bacteria

that have been cultured in standard media rather than wild

type cultures. Wild type species of Desulfovibrio have been

shown to lose their adhesion properties when they are

transferred from the wild to laboratory and grow in standard

medium.

Rajagopal and LeGall (1989) suggested that the ability of

some sulphate reducing bacteria to reduce nitrate or nitrite

with hydrogen as sole energy source provides a better

experimental system to study the cathodic depolarization

phenomenon, since problems with measurements of dissolved iron









26
and growth in the presence of precipitated sulfides resulting

from the customary sulphate reduction can be avoided.

The presence of organic electron donors has been

considered during cathodic depolarization of steel. Widdel

(1988) indicated that most Desulfovibrio spp are able to grow

on acetate and carbon dioxide as carbon sources and hydrogen

as sole energy source. Cord-Ruwisch and Widdel (1986) found

that the availability of organic electron donors appears to be

an important factor that influences the removal of cathodic

hydrogen. If a favorable organic energy source was present,

sulphide was produced, indicating that iron alone did not

allow sulphate reduction. They suggested either that cathodic

hydrogen was preferentially oxidized with the organic

substrate or that sulphide from sulphate reduction with the

lactate reacts with the remaining ferrous ions or ferrous

hydroxide from the corrosion process to produce more FeS.

In addition to the above observations, some other

important scientific areas related to the microbial corrosion

phenomenon need to be addressed in order to envision the

complexity of the problem. Those areas include bacterial

adhesion, surface thermodynamics, interfacial chemistry,

hydrogen embrittlement, interspecies hydrogen, bioenergetics,

microbial ecology, and applied fundamentals of metallurgy,

electrochemistry and chemistry. Information on those subjects

is referred to in the cited literature. Specific topics

include the prevailing mass transfer conditions and the









27
surface properties of the phases involved, which include

molecular fluctuations at the electrical double layer (Liu,

1983), the adhering bacteria, the substrate, and the suspended

media (Absolom et al., 1983; David and Misra, 1985), chemical

interaction at metal surfaces and at the liquid interface

(Sato, 1987), and metallurgical aspects such as

microstructure, grain boundaries, inclusions, surface

contamination, and sources of corrosion currents (Godard,

1980; Mears and Brown, 1941).



Microorganisms Involved in Microbial Corrosion



The microorganisms that have been associated with

microbial corrosion include many genera and species.

Microorganisms are anatomically simple yet biochemically

complex. They may be divided into three groups: (1) algae and

fungi, (2) protozoa, and (3) monera (e.g., bacteria). Many of

these organisms have been firmly established in laboratories

and field sites as having roles in the corrosion process,

whereas others have merely been isolated from suspected

corrosion sites. Microorganisms are sustained by chemical

reactions by ingesting reactants and eliminating waste

products. Thus these processes can influence corrosion

general by 1) directly influencing anodic and cathodic

reactions, 2) influencing protective surface films,









28
3) producing deposits, 4) producing corrosive metabolites, and

5) feeding on corrosion inhibitors.

Several characteristics of microorganisms enhance their

involvement in corrosion. They are generally very small,

starting from less than 0.2 pm, which allows them to penetrate

crevices very easily, to several hundred pm in length. Some

of them are motile, which aids in their migration to more

favorable environmental conditions. Microorganisms are able

to establish in sites that encourage their growth. For

instance, microbes establish and colonize surfaces in cooling

water systems where food sources concentrate at metal surfaces

because of their hydrophobic properties. They can withstand

a wide range of environmental conditions: pH values from 0 to

11 (Brock and Madigan, 1991), temperatures from -30 to 1100C,

and oxygen concentrations from 0 to 100% (Morgan and Dow,

1986).

Microorganisms can adhere to a surface and form colonies

of different species. These consortia, once formed, can

sustain survival under adverse conditions. They can reproduce

themselves to a great number in a short time. This fact

allows them to bloom and take over an environment quickly.

They are easily dispersed in air, water, animals, etc., and

adapt to other environments in which it may be easier for them

to grow. Many can adapt easily to a wide variety of

substrates, such as the Pseudomonas, some species of which

can use well over 100 different kinds of food as sole carbon









29
and energy sources. Many can produce extracellular

polysaccharides or slime layers where a consortia of bacteria

can develop and consequently influence corrosion (White et

al., 1986). These layers attract food and other

microorganisms and cause several other well-known problems in

the process industry, such as poor heat transfer.

Some microbes produce spores that resist the most severe

environmental conditions and are capable of surviving for long

periods of time by remaining dormant. They can quickly

colonize surfaces when the environment changes to their

liking. Some resist antimicrobial agents by virtue of their

ability to degrade them or by resisting permeability through

the cell wall or by the extracellular polysaccharide

protection. Such resistance may be acquired by mutation or

acquisition of a plasmid. Many species produce a wide variety

of organic acids that may promote corrosion of many alloys

even at low concentrations (e.g., Clostridium acetobutylicum

produces acetic acid).

Some species produce mineral acids that are extremely

corrosive. Thiobacillus thiooxidans produces sulfuric acid,

which is of economic importance in biohydrometallurgy because

it makes possible the leaching of metal sulfide ores. In the

case of pyrite, it oxidizes both the sulfur and ferrous

moiety. Boes and Kuenen (1983) have recently reviewed the

sulfur oxidizing bacteria and their relationship to corrosion

and leaching. Several bacteria metabolize nitrate, sometimes









30
used as a corrosion inhibitor (e.g., Pseudomonas spp reduce

nitrate and nitrite to nitrogen gas). Other organisms convert

nitrate to nitrite or ammonia to nitrite (e.g., Nitrosomonas)

and others turn nitrite into nitrate (e.g., Nitrobacter).

Many organisms form ammonia from the metabolism of amino

acids. This forms ammonia ions in basic solution, which may

be corrosive to copper alloys.

Microorganisms such as sulfate reducing bacteria produce

enzymes, which may be excreted outside the cell and which can

act on substances outside the cells. For instance,

hydrogenase has been reported as being responsible for

depolarizing cathodic sites during the microbial corrosion of

iron and steel. Many organisms can produce carbon dioxide and

hydrogen as a result of their fermentative metabolism. Carbon

dioxide in acidic solutions becomes carbonic acid, which is

highly corrosive, and hydrogen can polarize metal surfaces of

stainless steel and may cause hydrogen embrittlement.

Many genera of bacteria that normally use organic

compounds as carbon and energy sources can use hydrogen gas as

their energy source and carbon dioxide as their carbon source

and live chemoautotrophically. This can cause depolarization

of cathodic sites on steel and promote corrosion (e.g.,

Methanogens).

Some bacteria can oxidize or reduce metals or metallic

ions directly. For example, the iron-oxidizing bacteria

(Gallionella, Sphaerotilus) oxidize ferrous ion to ferric ion.









31
The ferric compounds precipitate in a sheath around the cells

and form tubercles in pipes and cause plugging. Thus,

concentration of cells is easily formed under those deposits.

On the other hand, ferric ion can be reduced to ferrous ion by

Pseudomonas spp from oil wells and marine sediments. It has

been suggested that the ferric film, which normally stabilizes

the surface of mild steel from corrosion, are destroyed

leaving the surface susceptible to corrosion attack. Other

bacteria can oxidize or reduce metals such as manganese.

Microorganisms can form synergistic communities (e.g.,

algae and bacteria). These consortia can accomplish things

that individually would be difficult if not impossible, such

as the case of fungi and Desulfovibrio spp. The fungi break

down wood to organic acids and consume oxygen, thus providing

the food and anaerobic conditions for Desulfovibrio spp..

Communities providing protection for individuals can also

change structure, dominant species, etc. by genetic mutation,

and can adapt to environmental changes, even to deliberate

chemical changes intended to kill them. More details on this

subject may be found in Sifontes and Block (1991).



Mechanisms of Microbial Corrosion


For proper selection of methods to prevent or control

microbial corrosion, it is necessary to know the mechanisms by

which microbial activity affect the deterioration of metals.








32

The mechanisms of microbial corrosion in some cases are well

defined, but where environments encourage the activity of

bacteria, the corrosion processes are more complex and still

not fully understood. The mechanisms of microbial corrosion

can be subdivided in direct and indirect mechanisms. If a

microbe interlinks an electrode process with its own

metabolism or presence, it is a direct effect (i.e.,

differential aeration cell produced by sessile bacteria),

otherwise it is an indirect effect (i.e., corrosive

metabolites produced by planktonic bacteria).

Even though the mechanisms of microbial corrosion are not

well understood, microbes that cause or influence corrosion

have been classified previously (Kobrim, 1976) into the

following groups:

1) Acid production. Some microbes can oxidize sulfur

compounds to sulfuric acid. Very low pH has been reported in

places where sulfur oxidizing bacteria are active. Many

species produce a wide variety of organic acids (e.g. acetic,

butyric, succinic, and formic) which may promote corrosion of

many metals and their alloys.

2) Protective coating destruction. Protective coatings

ranging from polymeric materials to passive films can be

broken by the activity of microorganisms and corrosion of bare

metal starts rapidly.

3) Production of corrosion cells. Differential aeration and

ion concentration cells are notable examples such as the case









33
of a metallic surface that is accessible to oxygen. If a

deposit such as a biofilm or a corrosion product covers it,

the surface under the deposit is shielded from oxygen and the

surface outside of the deposit is not. This results in a

corrosion cell.

4) Sulfur reduction and oxidation. Sulfate reducing bacteria

are the most publicized class of corrosive microbes. They

reduce sulfates to sulfides and can depolarize cathodic sites

by consuming hydrogen. Most sulfur oxidizing bacteria known

fall in the category of acid producers.

5) Concentration of anions and/or cations. Iron and manganese

bacteria are examples of this category. They generally form

thick, bulky deposits which create concentration cells or

harbor other corrosive microbes. This group is also known as

metal ion oxidizers or reducers.

6) Hydrocarbon utilization. Certain microorganisms have been

observed that destroy organic coatings or linings in the

presence of hydrocarbon fuels. Some others destroy metals

such as aluminum and feed on hydrocarbon fuels.

7) Slime formation. Certain algae, yeast, bacteria, and fungi

may form deposits which foul heat transfer equipment and

produce concentration cells on metal surfaces.

The above mechanisms can be grouped in three general

modes of microbial attack based on their metabolism: corroded

material serves as substrate for microbial growth, microbes








34

colonize material surface but feed on something else, and

microbes produce metabolites that corrode material.



Theory of Cathodic Depolarization



The theory of cathodic depolarization was postulated in

1934 by von Walzogen Kuhr and van der Vlught. The idea of the

authors came from their interpretation from the

electrochemical point of view of the anaerobic corrosion of a

cast iron pipe in wet soils near Amsterdam. Since the pipe

was under clay at a near neutral pH, the severely corroded pipe

did not allow them to explain the phenomena using the

reduction of oxygen as the cathodic reaction. The alternative

cathodic reaction was the reduction of hydrogen that was

feasible under those conditions.

The theory proposes that when iron is immersed in water,

a natural equilibrium is set up between the ferrous cations

released at the anode, and the metal surface negatively

charged at the cathode by the remaining electrons. The

dissolving process continues only if the electrons are

removed. Under aerobic conditions oxygen serves as an

electron acceptor resulting in rust formation. Under

anaerobic conditions free protons from the dissociation of

water are reduced on the cathodic metallic surface by the

remaining electrons, to form a protective hydrogen polarized

envelope that protects the iron metal from further








35

dissociation. A dynamic equilibrium is established which

keeps the iron polarized. The theory suggests that the

principle mechanism of anaerobic corrosion is cathodic

depolarization of the iron surface by hydrogen oxidizing

microorganisms such as sulfate reducing bacteria and

methanogen. These organisms disturb the equilibrium by

oxidation of the cathodically formed hydrogen with sulphate

and carbon dioxide as electron acceptors respectively, via the

hydrogenase enzyme. In an attempt to re-establish the

anodic/cathodic equilibrium more iron is oxidized, the end

result of which is pitting formation.


Table 2-1. Chemistry of the cathodic depolarization theory.



component reaction


1)Metal dissolution 4Fe --> 4Fe+2 + 8e

2)Hydrogen reduction 8H+ + 8e --> 8H --> 4H2

3)Water dissociation 8H20 --> 8H+ + 80H

4)Microbial activity SO4-2 + 8H --> S-2 + 4H20

5)Corrosion product Fe+2 + S-2 -> FeS

6)Corrosion product 3Fe+2 + 60H --> 3Fe(OH)2

7)Total rxn 4Fe + SO4-2 + 4H20 --> 3Fe(OH)2 + FeS + 2(OH)


The importance of this theory lies in the fact that it

separates for the first time the three components of the








36

microbial corrosion system which include the microorganisms,

the metal, and the suspended medium.

A large volume of literature up to the last decade has

been influenced by the theory of cathodic depolarization and

has referred to it, either to prove or to disprove it.

However, the use of electrochemical techniques allowed Horvath

and Solti (1959) to discover an anodic effect in addition to

the cathodic effect of the theory. They studied this effect

as a function of pH, environmental regulatory conditions and

the concentration of FeS present. According to those results,

SRB had an indirect role, which would be the stabilization of

the sulfide compounds over the metal surface by modification

of the redox potential.

One of the most conclusive findings was reported by

Costello (1974) who indicated a cathodic effect due to the

hydrogen sulfide produced by the sulfate reducing bacteria,

suggesting that hydrogen utilization by the sulfate reducing

bacteria became secondary and so the participation of those

bacteria in the corrosion process. Furthermore, It has been

suggested as an amendment to the corrosion theory that solid

ferrous sulfide, in contact with iron, acts as a cathode

(Booth et al., 1968). Other authors proposed that reduced

phosphorous compounds, too, are involved in the anaerobic

corrosion process (Iverson and Olson, 1984). Recently it was

reported that phosphate and hydrogenase can affect the

corrosion of mild steel. Phosphate reacts on mild steel with








37

concomitant production of hydrogen gas and the formation of an

iron/phosphate complex (vivianite) Bryant and Laishley, 1990;

Weimer, et al., 1988). The enzyme hydrogenase was reported to

accelerate cathodic depolarization by oxidizing the hydrogen

produced (Bryant and Laishley, 1993). Although many studies

have clearly demonstrated the involvement of sulphate reducing

bacteria and methanogens in corrosion, only a few have shown

the influence of those microorganisms using the electrons from

the metal as energy source.

On the other hand, some authors have provided evidence in

support of the cathodic depolarization theory (Cord-Ruwisch

and Widdel, 1986; Hardy, 1983; Tiller and Booth, 1962; Booth

and Tiller, 1960, 1962). Few, however have actually

demonstrated that this phenomenon is coupled to microbial

growth (Belay and Daniels, 1990; Rajagopal and LeGall, 1989;

Daniels et al., 1987; Pankhania et al., 1986; and Tomei and

Mitchell, 1986). In conclusion, research indicates that the

described theoretical mechanism is not entirely correct.



Previous Studies in Microbial Corrosion



Despite the large amount of literature in microbial

corrosion, it is still not an easy subject to understand,

because of the multiplicity of factors at play. Some authors

have indicated that microbial corrosion is a newly discovered

problem or an emerging science, possibly still in its infancy









38
(Tatnall, 1988). Others have confirmed by surveys that the

continued incidence of microbial corrosion could be due to a

general lack of awareness of the problem (Wakerly, 1979). The

fact is that its scientific study began more than 80 years ago

(Tiller, 1982) and after the work of von Wolzogen Kuhr and van

der Vlugt in 1934, who established the classical mechanism of

anaerobic corrosion, the subject of microbial corrosion in

general and of anaerobic corrosion in particular has become

recognized of prime importance (Bessems, 1983; Hamilton,

1985).

Most of the literature on microbial corrosion, before

1960, was concerned basically with observations of the effect

of bacteria on the environment. During the 1960s, there was

considerable activity on the field and most of the research

was focused on gaining understanding at the different

mechanisms. Some notable papers of this period belong to

Postgate (1960), Horvath (1960), Sorokin (1966), Iverson

(1966), Booth, Elford and Wakerly (1968), Booth and Tiller

(1968), and Costello (1969).

During the 1970s the interest in the subject broadened

the scope of research involving scientists from other fields

including plant engineers, corrosion scientists,

microbiologists, and biochemists (Sequeira and Carrasquinho,

1988). This has brought improved understanding of the

ecology, nutrition and physiological requirements of the

microbes involved in microbial corrosion, which also improved








39

the biochemical techniques used in both the field and the

laboratory. In addition, the publication of case histories of

a diverse range of failed industrial equipment has enhance

understanding of the problem. This period is well described

in the articles by Iverson (1972), King and Miller (1973),

Mara and Williams (1972), Miller and King (1975), Kobrin

(1976), Widdel and Pfennig (1977), Jorgensen (1977, 1978,

1980), Wakerly (1979), and Postgate (1979).

It was only within the last decade that microbial

corrosion was recognized as a serious problem in the chemical

industry by the leading corrosion society in the world (NACE)

and substantial advancement has been achieved specially in the

areas of microbiology and corrosion. In microbiology, among

other findings, the biochemistry of dissimilatory sulfate

reduction in Desulfovibrio has revealed enzymes and electron

carriers of special character and structure whose function and

distribution within the cell are just beginning to be

revealed, indicating new discoveries in the peculiar energy

generation systems in these bacteria.

In corrosion, sophisticated methods to study it, have

evolved; such as new methods in electrochemistry,

metallography, macroanalysis, and microanalysis. A

significant part of the research in that period is described

in the articles by Hamilton (1985), Iverson (1987), White et.

al. (1986), Tatnal (1981), Odom and Peck (1984), Cord-Ruwisch

and Widdel (1986), Tiller (1982), Miller (1981), Hardy (1981,








40

1983), Rajagopal and LeGall (1989), Bryant and Laishley

(1989), Daniels et al. (1987), and Widdel and Pfennig (1981).

Today, microbial corrosion is well recognized as a

serious problem in most industries, particularly in oil and

gas, the power generating and the process industries. In the

chemical industry alone, multimillion dollar failures due to

microbial corrosion have been reported involving cooling water

systems (Felzin et al., 1988).

While the above studies have yielded a wealth of

information, the literature reveals that the corrosion

community still knows very little about how microorganisms

influence or induce corrosion, and what the role of

microorganisms is in the corrosion process. In general, most

of the research deals with observations related to the basic

problem, case histories, biological research, and reviews of

literature. It is important to notice that this was not

usually the fault of the investigator; in the majority of

instances the specialist lacked interdisciplinary help

required to tackle such complex studies.

Microbial corrosion is by definition an interdisciplinary

field that requires among others, the understanding of

microbiology, corrosion science, metallurgy, electrochemistry,

transport phenomena, and surface chemistry. Several factors

have cause a further hindrance in the development of this

microbial corrosion. Among some of the factors are: the lack

of awareness; the difficulty of growing, isolating, and








41

identifying anaerobic microorganisms; the adaptation of

techniques from clinical microbiology to handle the microbes

involved; the fact that to the average researcher microbial

corrosion spans the boundary of traditional specialties; and

the idea that many microbiologists have of microbial corrosion

as a rather specialized even esoteric field, and the lack of

acceptance of the role of bacteria in corrosion by corrosion

engineers.

Actually, many theories on how microorganisms influence

or induce corrosion have been proposed. However, none of them

are fully proven but rather equal number of existing papers on

the subject claim to support or refute these theories.

Consequently, it seems fair to say that there is little

uniform agreement among those working on this field about what

is really going on (Tatnall, 1988).

Customary practices, used during the investigation of

microbial corrosion, include separated studies of either the

microorganisms, the metal or the fluid but ignore their

interaction. In addition, most laboratory microbial corrosion

experiments use flowthrough type bioreactors which are run for

weeks or even months. The most common corrosion measurement

methods used are adopted from classical corrosion science such

as Tafel plots and weight loss techniques. In many cases,

these practices have failed to give reproducible results

because of the length of the experiments, nature of the metal

surface, and fluid chemistry in addition to microbial









42
ecological considerations of such a complex systems. The

application of external potentials to the system, a common

practice in microbial corrosion experiments, may produce

significant thermodynamic changes known to affect the

stability of the system, and constitute a source of

uncontrolled variables.

The practical separation of the three components of the

microbial corrosion systems as suggested by the cathodic

depolarization theory, and the work performed by Daniels et.

al. (1987), set the stage for studying the microbial

utilization of cathodic hydrogen during corrosion. If hydrogen

oxidizing bacteria can take up the hydrogen produced

cathodically from steel in well defined experimental systems

then there exists the possibility to further explore in this

field and increase our understanding of these microbes and

their mechanisms of attack on metals.



Microbial Corrosion Control


Prior to 1970 the control of microbial corrosion focused

on the use of cathodic protection and antimicrobial agents.

This practice was complemented by improvements in surface

coatings, in particular antifouling systems such as tapes and

wrappings. Because of the limited ability of those methods to

control bacterial activity and the inefficacy of the

methodology to asses antimicrobial agents; the main driving








43
force after 1975 has been the consolidation of several

methodologies and assessment procedures to improve the

standard practice (Tiller, 1985).

In 1964, Saleh and co-workers evaluated nearly 200

antimicrobial agents and concluded that laboratory evaluations

of those chemicals can only be considered as an introductory

sign and should be supplemented with trials on the field. In

1983 Bessems recommended the importance of assessment

procedures and Gaylarde and Johnston emphasized the

deficiencies in current methodology and the need for

improvement. During last decade a joint venture of the

Institute of Corrosion Science and Technology and the National

Association of Corrosion Engineers has reviewed and updated

the current recommended practice for monitoring bacterial

growth. Topics such as killing time, the importance between

planktonic and sessile bacterial consortia and their

attachment properties are now important issues in the control

of microbial corrosion (Tiller, 1985).

Due to the complexity of the environmental factors that

results in microbial corrosion, the success of a control

program depends more in our outlook on the available

information about the problem, and becomes a challenge to

identify the most practical and economic solution.

In general, the control of microbial corrosion requires

a sound strategy. This involves a diagnosis of the microbial

corrosion problem that begin with determining the cause and








44

the mechanisms associated with the corrosion problem. A sound

diagnosis include: comprehensive current system diagrams,

materials of construction, fabrication methods, operating

history, chemical analysis, site specific environmental

conditions, biological history, and historical and current

chemical treatment. Unfortunately, if this is not followed,

adverse consequences may result. For example, a strong

oxidant may have a good killing power but may produce severe

local corrosion if applied to a particular system.

Fellers (1989) suggested that the best strategy is the

product of a multi-disciplinary team which looks at all

aspects of the problem, examines root cases, and objectively

evaluates alternatives. He also indicated in a later paper

that if the main strategy is to maintain the system clean of

microbial growth, instead of looking for independent solutions

to several problems, several issues may be avoided including

microbial corrosion (Fellers, 1990).

The essential strategies to control microbial corrosion

include: detection, prevention, and mitigation. Detection

techniques are basic to prevention and mitigation. They

include electrochemical methods, microbiological methods,

physical and metallurgical methods, in system monitoring

techniques, and other laboratory techniques. Details on these

techniques are found in earlier work of the author (Sifontes

and Block, 1991). Additional detection techniques include

visual inspections to identify suspected indications of









45
microbial attack and side stream monitors for determination of

corrosion and fouling rates.

Prevention and mitigation techniques hold the highest

payback in plant protection and preservation. There is no

universal approach to the prevention and mitigation of

microbial corrosion since it is almost impossible to use a

single preventive method. The prevention and mitigation

techniques offered by the current technology include planning

considerations such as material selection, design, and non-

metallic materials; physical-chemical control such as

corrosion inhibitors, selection and control of the

environment, protective coatings, electrochemical protection,

maintenance cleaning, and chemical control; and biological

control such as antimicrobial agents, and environmental system

control. Details on above prevention techniques are found in

earlier work of the author (Sifontes and Block, 1991). Other

prevention and mitigation techniques include general corrosion

control such as prevention of scales which enhances

biofouling, suspended solids dispersion such as the use of

penetrants of other surface active agents which re-disperse

fouling materials (Fellers, 1990).

Finally, an effective control of microbial corrosion

involves a combination of selected techniques to solve an

specific problem. Numerous history cases that include the

results of the applied techniques are available throughout the

literature cited. A reported case history of successful








46

corrosion control in a cooling water system used a combination

of biodispersant/biocide. It employed an organic corrosion

inhibitor, a polyacrylate/phosphonate dispersant, and a

combination of two microbiocides used simultaneously

(Honneysett et al., 1985).

the use of surfactants penetrantss /biodispersants) has

improved the effectiveness of microbial corrosion control.

Chemically, these substances are composed of nontoxic organic

compounds with penetrating and dispersing properties. The

biodispersants allow the sessile colonies to be penetrated by

the antimicrobial agents, thus used at lower dosages with

improved effectiveness. They inhibit the biomass produced

from becoming so massive that antimicrobial agents can not

penetrate the consortia of microbes. Some penetrants are

hydrophobic to the extent that a film forms on the metal

surfaces allowing less deposition of sessile colonies

associate or not with microbial corrosion.

Other references that contain a variety of history cases

include: Microbial Corrosion proceedings by the Metals

Society of London, 1983; International Conference on

Biologically Induced Corrosion proceedings by the NACE, 1985;

Microbial Corrosion 1 proceedings by the First European

Federation of Corrosion workshop on Microbial Corrosion, 1988;

and Microbially Influenced Corrosion and Biodeterioration

proceedings by the International Congress on Microbially

Influenced Corrosion, 1990.










Research Approach



The following tasks were considered in order to comply

with the objective of this research. The first task included

the evaluation of some new experimental flowthrough and batch

bioreactors to reproduce and study the anaerobic microbial

corrosion phenomena. The initial task consisted of runs of

bioreactors with the object of reproducing microbial corrosion

in the laboratory. The second task involved the use of the

batch bioreactor to study the onset of microbial corrosion,

using conditions that resemble the inside of gas transmission

pipelines. The operation of the batch bioreactor was examined

concerning practical functioning for head space composition,

sterilization method, reducing agent use, metal coupon

preparation, and bacteria handling. Different metals and

their alloys were tested for electroactivity and their ability

to produce hydrogen; and different hydrogen oxidizing bacteria

were also tested to determine their ability to uptake

hydrogen. A third task involved the development of an

experimental system that included the redesign of the batch

bioreactor, development of a working electrode preparation of

bacterial suspensions, preparation of mineral solution

electrolyte, and the implementation of a data acquisition

system to accommodate electrochemical measurements, in an

attempt to quantify the biological component of corrosion.

The redesign of the batch bioreactor consisted of improvements









48
to the experimental system that resulted in the development of

the single, dual, and triad flask electrochemical cells. A

forth task involved the general setting of parameters to

assure simpler and reliable results such as the selection and

setting of fixed conditions and analytical procedures.














CHAPTER 3
MATERIALS AND METHODS


Introduction


The overall objective of this research was to study

anaerobic microbial corrosion at its on-set and to quantify

cathodic hydrogen utilization by bacteria. For this purpose,

a total of seven experimental set-ups were utilized. These

include: the 1) flowthrough bioreactor, 2) batch bioreactor,

3) single flask electrochemical cell or single cell, 4) dual

flask electrochemical cell or dual cell, 5) triad flask

electrochemical cell or triad cell, 6) artificial hydrogen

uptake triad cell, and 7) galvanic couple triad cell.

Initially, the flowthrough bioreactor was used to

replicate microbial corrosion and to get familiar with the

laboratory techniques and analyses required for the

investigation. This traditional bioreactor design used a

known bacteria triculture commonly found at microbial

corrosion sites and a glucose lactate yeast extract (GLYE)

medium that allowed the three bacteria to grow rapidly. Runs

of this bioreactor were performed using carbon steel coupons

in the presence of the bacteria triculture in GLYE medium.

Then, since the corrosion reaction and the biofilm formation

developed relatively fast in the flowthrough bioreactor, a









50
simpler batch bioreactor to study the on-set of the reaction

was used. This new bioreactor offered the best alternative to

study the initiation of the microbial corrosion process and

was able to reproduce microbial corrosion of carbon steel

within 24 hours in GLYE and in mineral solution media. The

batch bioreactor was redesigned and instrumented to include

electrochemical measurements and became the single flask

electrochemical cell. The single cell was used to measure

free corrosion potential in the presence and absence of

bacteria. It was improved later to the dual cell and the

triad cell. The dual cell allowed bacteria to avoid direct

contact with toxic metals and the triad cell provided for

measurement of a differential corrosion current. The last two

experimental set-ups were identical to the triad cell, except

they were used to study the effect of an artificial hydrogen

uptake and to determine the reliability of the electrochemical

measuring system respectively.

All electrochemical set-ups used the mineral electrolyte

solution developed for the purpose of studying microbial

corrosion at its on-set and hinder the formation of a biofilm



Flowthrough Bioreactor


Description

The 3-L flowthrough bioreactor system, used for the

preliminary work of this research was designed in house and









51

built out of 5 cm 4 PVC pipe components. It consisted of two

sections for the study of sessile and planktonic bacteria

activity and their effect on metal deterioration. A basic

diagram showing its components is illustrated in Figure 3-1.









FEED


INOCULUM PORT


A'


w


SESILE ACTIVITY
TESTING


NAD(P)H PROBE PORT







. '-- METAL SPECIMEN
PORT



x\ "
\ ,



'Ni'


PLANKTONIC
ACTIVITY
TESTING




CLEAR PLASTIC
FOR LEVEL
CONTROL


,,f





/

1L
u


Figure 3-1. Flowthrough bioreactor general description of
components


FLOW
CONTROL
VALVE









52
The first section, for sessile activity testing,

consisted of eight side screw cap ports that fit a

fluorescence probe, developed by Biochem Technology, Inc.,

King of Prussia, Pennsylvania, and four top screw caps, to

which metal coupons and reference electrodes are attached.

The second section, for planktonic activity testing, contained

two side screw caps ports for attachment of the fluorescence

probe, and a clear section at the end of the reactor for the

purpose of controlling the bioreactor head space volume.

The bioreactor was fed from a 4-L stoppered flask through

a tubing, Tygon R-3603, and an effluent flow of 125 mL/hr was

controlled using a hose clamp on a 30 cm tubing, Tygon R-3603,

at its outlet.

A known bacteria triculture composed of Entrobacter

aergenes, Desulfovibrio desulfuricans, and Clostridium

acetobutylicum were used as the inoculum. Bacteria were

obtained from the American Type Culture Collection (ATCC).

Steel coupons, 5 mm x 20 mm x 10 mm, polished to 240

grid, and tied plastic strips to a 1/4 0 teflon rods were

attached to the upper screw caps of section one. Figures 3-lb

and 3-1c illustrate the assembly combinations of metal

coupon/fluorescence probe and a metal coupon/fluorescence

probe/reference electrode, respectively. Figure 3-1d shows an

actual set-up of the flowthrough bioreactor.













SCREW CAP


METAL SPECIMEN


Figure 3-2. Flowthrough bioreactor metal coupon and
fluorescence probe setup.


SCREW CAP


IONIC BRIDGE


REFERENCE
ELECTRODE


METAL SPECIMEN


Figure 3-3. Flowthrough bioreactor metal coupon, fluorescence
probe and reference electrode setup.



































Figure 3-4. Flowthrough bioreactor components.


Operation

The flowthrough bioreactor system was operated at a 1-day

retention time (125 mL/hr). Medium was fed to the bioreactor

by gravity from the 4-L reservoir under a nitrogen atmosphere

to assure anoxic conditions. The feed, a glucose-lactate

yeast-extract (GLYE) media, was prepared fresh daily,

sterilized and neutralized prior to use.

Effluent samples for pH and volatile fatty acids were

taken directly from the effluent. Metal coupons samples for

scanning electron microscopy analysis were sampled under

aseptic conditions while sparging with N2 gas and placed









55
immediately into 10 mL vials containing a 0.5% glutaraldehyde,

2% formaldehyde, cacodylate buffer, pH 7.2. Metal samples

were further processed, according to protocol described in

Table 3-1, for scanning electron microscopy examinations.



Table 3-1. Protocol for metallic surface fixation for
scanning electron microscopy examinations.


1) Place metal coupon in a 0.5% glutaraldehyde, 2.0%

formaldehyde, cacodylate buffer, pH 7.2 solution for 5 min at

room temperature, then on ice for 10 min.

2) Wash coupon in ice-cold buffer two times for 5 min each.

3) Wash coupon in ice-cold ethanol solutions for 5 min each,

in the following ethanol concentrations: 25%, 50%, 75%, and

95%.

4) Wash coupon at room temperature in 100% ethanol.

5) Wash coupon in acetone for 15 min at room temperature.

6) Wash coupon in fresh acetone at room temperature for 30

min.

7) Treat coupon with solutions of epon-araldite 30%, 70%, and

100% for 1 hr each at room temperature, except the 100% which

is done at 60C in oven.



Bacteria inoculations were done as follows for initial

run. Ten mL of E. aerogenes were inoculated first in order to

reduce the media. After 24 hours, 15 mL each of D.

desulfuricans and C. acetobutylicum were inoculated. For








56

other runs, bacteria were inoculated within six hours in

amounts 10 times larger than the volumes used in the initial

run. All runs were temperature controlled at 35"C inside a

walk in incubator.

GLYE media was prepared using ingredients outlined in

Table 3-2, Table 3-3, and Table 3-4. Media was prepared in 4-

L feed reservoir flasks and autoclaved at 15 psig for 15 min

(Balch et al., 1979).



Table 3-2. Glucose Lactate Yeast Extract Medium Composition.



Yeast Extract 0.5 g

Glucose 0.5 g

Sodium Lactate 0.5 g

Mineral Solution 1 (table 3) 25.0 mL

Mineral Solution 2 (table 3) 25.0 mL

Trace Mineral Solution 3 (table 3) 5.0 mL

Trace Vitamins Solution 4 (table 2) 5.0 mL

L-Cysteine HC1.H20 0.5 g

Na2S.9H20 0.5 g

Resazurin (0.5 mg/mL) 1.0 mL

Complete to 1-L with distilled deionized water, adjust pH to

7.5 with 5N NaOH and autoclave for 15 min at 15 psig.














Table 3-3. Trace Vitamins Solution 4 Composition.

Biotin 2 mg

Folic Acid 2 mg

Pyridoxine HC1 10 mg

Thiamine HC1 5 mg

Riboflavin 5 mg

Nicotinic Acid 5 mg

DL-Calcium Pantothenate 5 mg

Vitamin B12 0.1 mg

p-Amino Benzoic Acid 5 mg

Lipoic Acid 5 mg

Complete with distilled and deionized water to 1-L


Table 3-4. Mineral Solutions Composition.

1) Mineral solution 1

K2HPO4 6.0 g/L

2) Mineral solution 2

KH2PO4 6.0 g/L

(NH) 2SO, 6.0 g/L

NaCl 2.0 g/L

MgSO4.7H20 2.6 g/L

CaC12.2H20 0.16 g/L


J










3) Trace Minerals solution 3

N (CH2CO2H)3

MgSO4. 7H20

MnSO4.2H20

NaCl

FeSO4 7H20

CoSO,

CaC12 2H20

ZnSO4

CuSO4 5H20

AlK(SO4)2

H3Bo

Na2MoO4.2H20


1.5 g/L

3.0 g/L

0.5 g/L

1.0 g/L

0.1 g/L

0.1 g/L

0.1 g/L

0.1 g/L

0.01 g/L

0.01 g/L

0.01 g/L

0.01 g/L


Batch Bioreactor



Description

The batch bioreactor system consists of a 250 mL bioassay

bottle with the metal coupon suspended in 150 mL of liquid

media by a nylon string attached to a crimp top butyl rubber

septum to preserve bottle anaerobically. A typical batch

bioreactor experiment is shown in Figure 3-2. This set-up was

used to study microbial corrosion under a known growth

environment, medium, metal coupon, and head space, using

different bacteria combinations.


















nylon string

-- metal coupon

suspended medium


Figure 3-5. Batch bioreactor.



Prior to the tests, the batch bioreactor was assembled

using a 250 mL bioassay bottle to which 150 mL of GLYE media

was added and a metal coupon is suspended in the media. The

metal coupon was fabricated from ASTM-ASME SA106 grade B1

steel pipe 2.5 cm 4, schedule 80, obtained from Texas Eastern,

Louisiana. Coupons were machine ground to a size of 1.5 cm x

2.0 cm x 3 mm with a side hole of 3mm 4 for the attachment of

the nylon string that holds them to the butyl rubber stoppers.

The inoculum used for the batch bioreactor was the same as

used for the flowthrough bioreactor and it is described

therein.



Operation

Batch bioreactors were intended for one day runs;

however, some experiments were run for several days. Their

100 mL head spaces were maintained with N2 at 7 psig to assure









60
anoxic conditions. Bioassay bottles, including medium and

metal coupon, were autoclaved at 15 psig for 15 min prior to

each run or inoculation. Bacteria were inoculated at room

temperature and placed in a Fisher low temperature incubator,

model 307, at 370C under aseptic conditions. One mL of E.

aerogenes was inoculated first, then after 12 hrs one mL of

each D. desulfuricans and C. acetobutylicum was added.

Samples for pH and volatile fatty acids analysis were taken

via the rubber stopper with a 5 mL syringe. Metal coupons

were sampled aseptically and placed in 10 mL vials containing

a cacodylate buffer, pH 7.2, to be further processed according

to the protocol described in Table 3-1 for scanning electron

microscopy examinations.

The GLYE media were prepared using ingredients outlined

in Table 3-2, Table 3-3, and Table 3-4.



Electrochemical Cells Description



The new design of the batch bioreactor, here called the

electrochemical cell, consisted of a modified 1-L erlenmeyer

flask including an extra side opening that fit #8 butyl rubber

stoppers. At the top opening port, an assembly of electrodes

was fitted that included the working electrode (metal

coupon), the pH combination Ag/AgCl reference electrode,

polymer body unit/gel filled by Fisher Scientific, Co., and









61
pressure transducer by Setra Systems, Inc., model 205-2 and

digital pressure indicator, model 300C.

The new system was instrumented for automated data

acquisition of free corrosion potential, pH, oxidation-

reduction potential, and head space pressure. Prior to

filling the electrochemical cell to the 1-L mark with the

medium electrolyte, the cell was outgased in an atmosphere of

80% N2/20% CO2. The flask was then sealed under a positive

head space pressure and placed in a precision water bath by

GCA corporation at 300C. A head space pressure in the range

of 20 to 40 mm of Hg above atmospheric pressure was used

throughout the investigation. This was a pressure that

allowed stoppers not to blow away, it was economical and

allowed an adequate gas supply to the head space. A schematic

of the electrochemical cell is detailed in Figure 3-3.



REDOX POTENTIAL

pH : METRABYTE'S
METRABUS
g A/D DATA
POTENTIAL C ACQUISITION
SYSTEM
PRESSURE






UQUID 1 L
SAMPLUNG 1
PORT

4 EFC E METAL
:ECpODE IBM PC
WATER BATH

Figure 3-6. Schematic of electrochemical cell and data
acquisition system.









62
The system set-up employed an IBM compatible computer

with a four channel MetraByte's MetraBus A/D data acquisition

system that interfaced the four parameters. Custom software,

written in Quickbasic, was used to acquire pH and the three

other parameters, free corrosion potential, oxidation-

reduction potential, and pressure in the form of voltages.

The liquid sampling port consisted of a Pasteur pipet

connected to a 10 mL syringe via a clear plastic tube with an

adjustable tube clamp.



Electrochemical Cells Operations



Single Flask Electrochemical Cell

The first modification or redesign of the batch

bioreactor was the single flask electrochemical cell. A

detailed schematic of the single cell is shown in Figure 3-4.

POTENTIAL


pH PRESSURE





S1L


REFERENCE
ELECTRODE
/ METAL
COUPON

Figure 3-7. Single flask electrochemical cell.








63

Experiments using the electrochemical cells were run at

constant temperature in a water bath at 30C, described

earlier. Anaerobic conditions were maintained throughout the

experiments using gases treated in a 5.0 cm 4 and 75 cm long

copper column that was electrically heated to approximately

3500C and reduced with pure hydrogen. A gas mixture

consisting of 80% N2/20% CO2 was used during outgasing,

preparation of the mineral electrolyte, and preparation of the

bacteria suspensions as described earlier in this chapter.

After outgasing the single cell, the electrochemical cell

was filled with the mineral solution electrolyte to the 1

liter mark. Then, it was sealed with the stopper containing

the metal coupon and electrode assembly, leaving a head space

of approximately 200 mL at a pressure in the range of 20 to 40

mm of Hg above atmospheric pressure. At this stage, the metal

coupon was held clear of the solution electrolyte and the

complete cell was placed in the water bath. Resting cells

harvested by centrifugation were supplied to the single cell

by injecting them through the rubber stopper. Once the data

acquisition system was connected to the SC and data started to

be collected, the metal coupon was lowered into the

electrolyte by pushing the insulated copper wire that holds

the epoxy-mounted coupon through the rubber stopper.

Experiments using this electrochemical cell allowed

fitting 4 single cells in the water bath for each run, one

control and triplicate samples. Gas and liquid samples were








64

removed under sterile conditions, using appropriate syringes

through the septums provided at the top and side flask

openings.



Dual Flask Electrochemical Cell

It consisted of the single cell (flask I) interconnected

to an additional flask II through their head spaces as

described in Figure 3-5. The objective of flask II was to

overcome some anticipated inconveniences experienced in the

single cell such as metal toxicity and/or any stress imposed

on bacteria during the harvesting procedure. Now flask II

could allow bacteria to grow freely and out of contact with

the metal to avoid metal toxicity. The head space connection

allowed cathodically produced hydrogen to be transferred from

flask I to flask II.
PRESSURE
REDOX
POTENTIAL -I

POTENTIAL
1 L VALVE


pH



1L
UQUID
SAMPLING
PORT
METAL
REFERENCE COUPON
ELECTRODE


Figure 3-8. Dual flasks electrochemical cell.









65

Dual cell experimental runs followed a procedure similar

to the setting and operation of the single cell and described

above, except that flask II could be used to grow bacteria in

their optimum media. The setting of the dual cell did not

allow more than one dual cell in the water bath, for which

only single experiments were run.



Triad Flask Electrochemical Cell

This final design consisted of a set-up similar to the

dual cell including an additional flask III, which is

interconnected to flask I via an ionic bridge made out of 5%

agar, KC1 saturated. Flask III was similar to flask II except

the latter was connected to flask I by the head space. A

detailed schematic is shown in Figure 3-6.


POTENTIAL
DIFFERENTIAL II
CORROSION
PRESSURE CURRENT
REDOX
POTENTIAL I
^ = POTENTIAL
1 L VALVE I 1 L

I|
pH


I L
II 1L III
LIUQUID 4 G
SAMPLNG GEL
PORT 1 IONIC
BRIDGE
METAL
REFERENCE COUPOt --
ELECTRODE


Figure 3-9. Triad flasks electrochemical cell.









66
This design was considered capable of satisfying the

objectives of this research because it allowed the measurement

of a differential corrosion current (DCC) induced by bacteria

in addition to other parameters mentioned earlier in the

single cell. A home-made zero resistance ammeter was

incorporated in the data acquisition system in order to

measure and acquire the new DCC parameter. The triad cell as

well as the dual cell allowed bacteria to be grown in flask II

or injected in as suspended cells.

The four available data acquisition system channels were

used to acquire data on the following parameters: pH in flask

I, free corrosion potential in flasks I and III, and DCC

between metal coupons exposed and unexposed to bacteria in

flasks I and III respectively. Triad cell runs used the

selected metal AS106 and the selected bacteria Escherichia

coli (JW11).

Liquid samples for dissolved iron determinations were

taken simultaneously from flask I and flask III, every ten

minutes. Only one triad cell fit the 300C water bath.

Other experiments using the triad cell include a run in

which a vacuum was applied to flask I and then to flask III in

order to exaggerate a case of hydrogen uptake. Near the end

of the test a positive hydrogen pressure was applied to flask

III in order to observe its effect on the potential and

differential corrosion current measurements. The last

experiment consisted of a galvanic metal couple of Mg working








67

electrode in flask I and a Cu working electrode in flask III,

to which vacuums were applied correspondingly to test the

triad cell response to hydrogen uptakes in both flasks. The

purpose of this test was to check the measurement capacity of

the electrochemical system since a significant differential

corrosion current was supposed to be developed.



Preparation of Bacterial Suspensions



The organisms used in this investigation were

Desulfovibrio desulfuricans (ATCC 7757), Entrobacter aerogenes

(ATCC 13048), Clostridium acetobutvlicum (ATCC 824),

Escherichia coli (ATCC 8739), Alcaliaenes eutrophus (ATCC

29597), and Escherichia coli (JW 111), provided by Dr. K. T.

Shanmugam. The first three bacteria have been found often in

real MC cases and all of them represented species of hydrogen

oxidizing bacteria.

The first three bacteria D. desulfuricans, E. aerogenes,

and C. acetobutylicum were grown in glucose lactate yeast

extract (GLYE) medium, as shown in Tables 3-2, Table 3-3, and

Table 3-4. The rest of the bacteria were grown in Trypticase

Soy Broth media. All chemicals used to prepare the media were

reagent grade and were obtained from Fisher Scientific or

Sigma Chemical Co. All media were prepared under an

atmosphere of 80% N/20% CO2. However, when microorganisms

used for MC experiments were inoculated in their respective








68

medium, an atmosphere of 80% Hz/20% CO2 was used to activate

their hydrogen uptake enzymatic systems.

Stock cultures from ATCC were prepared at 30C from

freeze-dried ampules of the organisms in their recommended

growth media indicated above. Cultures were maintained

inoculating agar slants (monthly) and liquid media (weekly).

20 mL serum vials were used for facultative bacteria, and 25

mL crimp-top tubes, sealed with butyl rubber septa, for

anaerobic bacteria.

Resting-cell suspensions were used regularly only in

experimental set-ups that included electrochemical

measurements; other experiments used bacteria inoculated

directly from their grow media. The batch bioreactor also used

washout from slants in some of its runs. Resting-cell

suspensions avoid electrolyte contamination with media used to

grow bacteria.

Cell suspensions were prepared as follows: 1 mL of

subculture of facultative microorganisms less than 1 week old

or 2 mL of subculture of anaerobic microorganisms were

inoculated in 400 mL of appropriate growth media and incubated

in a Fisher low temperature incubator, model 307, at 35C

under a 80% Hz/20% CO2 headspace for 1 day for facultative

bacteria and 3 days for anaerobic bacteria. Once cells were

grown, they were harvested by centrifugation in 50 mL screw

cap tubes for 10 min at 10,000 g.








69

Bacteria pellets from tubes were then resuspended in 5 mL

of the mineral solution electrolyte to give an approximate

concentration of the cell suspension of 65 mg (dry weight

basis).


Preparation of Mineral Solution Electrolyte



In order to provide the electrochemical cell with minimum

nutrient conditions as encountered inside gas transmission

pipelines and still allow bacteria to survive for few days, a

mineral solution electrolyte was developed to be used in the

experimental system to study the onset of microbial corrosion.

The ingredients of the mineral solution electrolyte consisted

of 10% of the concentration of the components of solutions

specified in Table 3-4 including 1 mg/L of resazurine and

deionized distilled water to complete the required volume

concentration. The ingredients were heated to the maximum mark

on the dial on a Fisher Scientific hot plate, model 210T and

stirred under an 80% N/20% CO2 atmosphere. After heating for

0.5 hr, the flask was cooled down to 300C surrounded by

crushed ice in a plastic container for approximately 1 hr. At

this time, the electrolyte was transferred to previously

cleaned electrochemical cells and filled to the l-L marks.

Remaining electrolyte was used to resuspend bacteria and other

routine tests.










Metal Coupon


From an original collection of 33 different metals and

alloys, only 18 different metal samples were cut and coupons

fabricated. Samples were prepared based on availability and

ease of cutting to the maximum size allowed in the molds used

for epoxy mounting. Metal coupons used in the flowthrough and

batch bioreactors were not epoxy mounted.

The availability of SA106 steel coupons, its non-toxic

effects on bacteria, its successful surface colonization and

observed surface deterioration, and the popular use of the

alloy in gas production pipelines suggested its selection for

the electrochemical experiments. In addition, it has a

moderate hydrogen production and it is relatively easy to

polish.

A technique was developed to fabricate the working

electrodes used in the electrochemical cells. It consisted in

cutting metal coupons from ASTM-ASME SA106 grade Bl steel, 2.5

cm diameter schedule 80 17-24 H.T. 2500 psi pipe, obtained

from Texas Eastern, Louisiana. The metal coupons were machine

ground to a size of 1 cm x 1.9 cm x 0.02 cm, and attached to

30 cm long 12 THWN insulated hard copper wire by Cerrowire

using nickel print paint (cat. No. 22-207) by GC Electronics.

Coupons were centrally mounted on epoxide resin 20-8130-128

and hardener 20-8132-032 from Buehler, leaving an area of 1.0

cm x 1.9 cm exposed for the microbial corrosion studies. The








71

hard wire served to support and to electrically connect the

coupon and was thereby insulated from contact with the

electrolyte in which the working electrode was immersed.

Prior to the start, the working electrode was held clear of

the solution contained in the flask and lowered into the

solution at will by pushing the hard copper wire through the

butyl rubber stopper that holds the electrode assembly.

After preparation, the working electrodes were polished

with 600 grit, rinsed with ethanol and blotted dried with

Kimwipes EX-L delicate task wipers, then stored in either a

desiccator glass container or a desiccator cabinet by Labconco

Co., model 55300, under vacuum until ready to be used. During

the polishing of the working electrodes, they were initially

sanded with a 240 emery grid paper to remove any

inconsistencies on the metal surface, then with a 400 grid,

and finally with a 600 grid. The working electrodes were

placed in an ethanol bath in the ultrasonic cleaner to ensure

the removal of metal particles left from previous polishing.

The polishing procedure was done using a Buehler rotary

polishing wheel and care was taken so that no cross-hatching

was left on the metal surface.



Parameters Setting



In an effort to assure consistency during the

electrochemical measurements, several parameters were fixed








72

based on equipment capabilities, typical inside conditions of

gas transmission pipelines, and other anaerobic stagnant

conditions to make measurements simpler and more reliable.

This task involved the use of the batch bioreactor and

consisted of the study of several determining factors that

formed the basis for the development of the electrochemical

system described herein. Experiments done included

determination of head space gas composition, selection of

reducing agent, selection of sterilization method, and other

related studies.

The head space gases used were methane, hydrogen,

nitrogen, carbon dioxide, and helium. The reducing agents

tested were sodium sulfide, cysteine, thyoglycolate/ascorbate,

and Enterobacter aerogenes as a natural reducing agent. Among

the sterilization methods used were: 1) all components of

batch bioreactor autoclaved at 15 psig during 15 min, 2) all

components of batch bioreactor autoclaved except metal-coupon-

rubber-stopper assembly that was oven treated at 105C for 12

hrs, 3) media autoclaved as before and metal-coupon-rubber-

stopper assembly sterilized in acetone and ethyl alcohol for

15 min then blotted dry prior the set-up in bioassay bottle,

and 4) media components autoclaved and metal coupon flamed.

The basic anaerobic and stagnant conditions included fixed

temperature of 35C and pH in the range of 6.0 and 6.5.










Measurement Procedures

All biocorrosion experiments, using electrochemical

cells, were performed at 30C. Flasks using electrode

assemblies include a rectangular metal coupon (working

electrode) with surface areas of 1.9 cm2 mounted in epoxy

resin and abraded to a 600 grit finish. Metal coupons were

connected to 30 cm of insulated hard copper wire that served

as electrical conductors and support in the stopper.

Electrode assemblies included Ag/AgC1 reference electrodes,

polymer body unit/gel filled, from Fisher Scientific.

Reference electrodes were continuously calibrated against each

other using the data acquisition system prior to the

experimental runs. The side opening of flasks contained

platinum electrodes for oxidation-reduction potential

measurement and Pasteur pipets attached to a 10 mL syringe for

liquid sampling.

The four data acquisition channels were connected to the

triad cell as follows: channel four was wired to a pH

reference electrode for pH readings. Channel one and channel

two were wired to the working electrode and reference

electrode of flask I for free corrosion potential

measurements. Channel two was wired to the working electrode

and reference electrode of flask III for free corrosion

potential measurements, and channel three was wired to the

working electrodes of flasks I and III for differential

corrosion current measurements. Current was measured with a









74
homemade zero resistance ammeter with a 10 nA sensitivity,

made of a #308 operational amplifier and a high precision

10,000 0 resistor. This device has the property of converting

current into voltage out with a gain determined by the high

precision resistor, allowing the data acquisition system to

measure and acquire data in volts. A digital pressure

indicator, model 300C by Setra Systems was also adapted to

monitor the headspace pressure in flask I. Once the data

acquisition system was started the working electrodes were

immersed in their respective electrolytes and data was stored

in computer floppy diskettes.

Samples of bacteria, in their growth medium, in their

resuspended medium, and in the liquid discarded after

centrifugation, were homogenized using a vortex Fisher brand,

model Genie 2, and taken to a spectrophotometer Spectronic 21D

by Milton Roy for determination of transmittance, absorbance,

and concentration at the start of each experiment. After

spectrophotometer determinations, samples were vacuum filtered

using 47 Mm pore size membrane filters, preweighted in a

Metler AE100 balance. Wet mounts were also prepared regularly

to check bacteria viability under a light microscope by Nikon,

model LABOPHOT-2. At intervals of 10 minutes, 10 mLs aliquots

were withdrawn from the electrolytes from the flasks I and III

under stagnant conditions and analyzed for dissolved Fe at the

IFAS Soil Science Analytical Research Laboratory. Last runs

were sampled only at the beginning and at the end, shaking








75

flasks to avoid stagnant conditions for dissolved iron. The

duration of each electrochemical experiment was approximately

1 hour. Experiments were continued until the corrosion

potentials leveled off.



Analytical Procedures

Dissolved iron analyses were performed at the IFAS Soil

Science Analytical Research Laboratory. Samples were

collected in 10 mL vials and preserved using 1 drop of

concentrated H2SO and analyzed on an atomic absorption unit

by Perkin Elmer.

Molecular hydrogen was measured on a gas chromatograph by

Gow-Mac Instruments Co., series 580, equipped with a thermal

conductivity detector. 50 pL samples were collected in a gas

tight syringe and injected onto a molecular sieve type 5A

column. Samples were injected at an inlet temperature of 400C

with column temperature of 35C and a detector temperature of

93C. The carrier gas used was nitrogen and the detector

current was 50 mA.

Transmittance, absorbance and concentration were measured

on a spectrophotometer Spectronic 21D by Milton Ray. The

samples were collected in 10 mL Hatch COD tubes and

homogenized in a vortex mixer by Fisher, model Gene 2, prior

to measurements.

Volatile fatty acids were measured on a FID gas

chromatograph by Shimadzu, model GC-9AM. The samples were









76
prepared by centrifugation after acidification with 20%

phosphoric acid. The samples were injected onto a 2m long by

2mm 0 glass column packed with 80/100 chromosorb 1200 WAW

coated with 3% H3PO, and carried with nitrogen gas. IpL

volumes were injected at an inlet temperature of 180C with

column temperature ramped from 130C to 170C over 5 min and

a detector temperature of 2000C.

Dry weights of resting cells were determined by vacuum

filtering known volumes of bacterial suspensions. Filtration

was performed in a 47 mm magnetic Gelman Filter funnel (cat.

09-735). The samples were filtered using 0.45 pm pore size

membrane filters, 47 mm plain Supor 450, by Gelman Sciences,

Co.. Vacuum was achieved with a precision belt-driven vacuum

pump by Fisher Scientific, model D-75. Membrane filters were

then oven dried at 105C for 12 hrs. Weights were measured in

a Metler balance, model AE100.



Gases

The gases used during the investigation were: Nitrogen

(UN1066), Helium (UN1046), Hydrogen (UN1049), Nitrogen/Carbon

Dioxide (80/20, UN1956), and Hydrogen/Carbon Dioxide (80/20,

UN1954) supplied by Liquid Air Corporation. For anaerobic

experiments all gases were passed through a heated copper

column (5 cm 4 and 75 cm long) to free gases from traces of

oxygen. The column was heated electrically to about 350C by

a coil of electrical wire wrapped around the column.








77

Reduction of the column was achieved by passing hydrogen gas

through the column.



Scanning Electron Microscopy

Metal coupons, after exposure to microorganisms, were

treated with a buffered glutaraldehyde solution to fix the

morphology of the bacteria, and then dehydrated through a

graded ethanol series of increasing concentration ranged from

25% to 100% using procedure outline in Table 3-1, page 93.

Samples were then dried and sputter coated with gold

palladium, then examined in a Hitachi S-450 SEM, that included

an Energy Dispersive X-ray Analyzer (EDXA). Coupons were

analyzed throughout and micrographs and spectra were taken at

sites of major interest.



Light Microscopy

Metal coupons were observed and photographed under a

Nikon stereomicroscope, model SMZ-2T and a Nikon microscope,

model LABOPHOT-2, for biofilm and corrosion examination. Pure

cultures and liquid samples from bioreactors were examined

periodically for contamination via wet mounts.














CHAPTER 4
RESULTS AND DISCUSSION


Flowthrough Bioreactor


Two experimental runs were performed in this bioreactor

in order to reproduce microbial corrosion and to study the

effect of a bacterial triculture on the corrosion of carbon

steel coupons and the formation of biofilm. Its set-up is

described in Figure 3-1. Both runs reproduced microbial

corrosion and their results are shown in Figures 4-1, 4-2, 4-

3, 4-5, and 4-6. Metal coupons exposed to a bacterial

triculture, grown in glucose lactate yeast extract (GLYE)

media under anaerobic conditions, were examined under the

light and the scanning electron microscope (SEM). The results

indicated extensive bacteria colonization as well as metal

deterioration. Careful observations of biofilm micrographs

have suggested a mechanism for understanding its formation.

The first experiment was run for a period of four weeks.

Metal coupons and liquid media were sampled weekly for routine

analyses. First week scanning electron microscopy (SEM)

micrograph results indicated that the bacterial triculture

developed a biofilm consisting of a separated polymeric double








79

layer structure; an inner structure attached to the metal

surface, and an outer structure that covers it. It appears

that Enterobacter aeroaenes, the bacterium inoculated first,

was responsible for the production of the inner structure and

perhaps the outer structure of the biofilm, see Figures 4-la

and 4-1b. This bacterium, besides reducing the media to

accommodate the strict anaerobes, synthesized a considerable

amount of extracellular polymeric material for attachment

purposes. This material is believed to overcome the natural

surface repulsion caused by the negatively charged bacteria

and metal surfaces (Beveridge and Doyle, 1989). Figure 4-1c

shows the characteristics of the abundant polymeric network-

like inner structure produced by the bacterial triculture. E.

aerogenes is probably responsible for building this structure

because it is the most prolific of the three species and

reaches the log phase of the growth curve in less than 6 hrs

as compared to the other two bacteria which require often more

than two days. Wachenheim and Patterson (1992) reported that

anaerobic production of extracellular polysacharides is

enhanced generally with any improvement of the conditions of

bacterial growth and that this material is only produced

during the log phase of growth where its production is also

exponential. In our work, the flowthrough bioreactor was run

using GLYE medium which must have enhanced the production of

the polymeric material. This fact, if compared to experiments

run in a medium with limited nutrient composition (mineral








80

solution only), indicates that the amount of biofilm formed on

the metal surface is proportional to the amount of organic

nutrient in the medium.

Desulfovibrio desulfuricans and Clostridium

acetobutylicum were inoculated one day after the initial

inoculation of Enterobacter aerogenes. The components of the

bacteria triculture were cultured separately. Figure 4-2

shows scanning electron microscopy micrographs of each species

for identification purposes. After inoculation, the strict

anaerobes probably displaced E. aerogenes from the inner

structure of the biofilm so it moved upward and started

producing the outer biofilm layer. This mechanism, suggested

in Figure 4-la, indicates the presence of D. desulfuricans and

C. acetobutylicum in between the inner and outer structure and

Figure 4-1c shows that most bacteria in the outside are

represented by E. aerogenes and C. acetobutylicum. These

figures also show that the outer biofilm structure is rougher

on the inside than on the outside. The inner surface seems

like a woven net of polymeric material attached to the metal

surface.

Metal coupon surfaces for this experiment were polished

to 240 grid. Figure 4-3a represents a micrograph of the

control surface, unexposed to bacteria and magnified 1200

times. Figure 4-4a shows the corresponding energy dispersive

x-ray analysis (EDXA) spectrum of the metal surface.








81
After anaerobic exposure of the metal coupons to the

bacterial triculture for a week, surface colonization

including pitting corrosion was observed. Figures 4-3b, 4-3c,

4-3d, and 4-3e show localized pitting and polymeric biofilm,










a)










Figure 4-1. Flowthrough bioreactor results on bacteria
attachment to carbon steel coupons. a) SEM micrograph showing
bacteria within the outer and the inner structures of the
biofilm. b) SEM micrograph indicating the nature of the outer
biofilm structure.


























c)










Figure 4-1.
biofilm.


c) SEM micrograph showing bacteria on inner


Figure 4-2. SEM characterization of bacterial species from
triculture. a) Enterobacter aeroqenes.

































Figure 4-2. b) Clostridium acetobutylicum.


Figure 4-2. c) Desulfovibrio desulfuricans.








84

metal deterioration near the edge of the biofilm, generalized

metal deterioration, and elongated pits respectively. Figure

4-4b shows the EDXA spectrum of a metal coupon after exposure

to the bacterial triculture. This analysis presents

additional evidence of microbial corrosion. It shows the

presence of an additional peak (if compared to the control),

corresponding to a sulfur compound, which is probably due to

the deposition of ferrous sulfide resulting from the activity

of D. desulfuricans.

Stereomicroscopic observation of coupon surfaces after

exposure to the bacterial triculture suggest the nature of the

biofilm to be slimy. Figure 4-5a and 4-5b are photographs of

the face and edge of a metal coupon respectively.

During the initial run of the flowthrough bioreactor,

volatile fatty acids and pH were measured weekly for one

month. Results, shown in Figure 4-6, indicate that the only

VFAs produced were acetic and butyric acids. During the first

week, only acetic acid was produced and its concentration

ranged between 300 and 400 ppm until the 3rd week and almost

double by the end of the fourth week. Butyric acid appeared

after the second week and increased at a rate of 200 ppm per

week to a concentration very close to 600 ppm at the end of

the 4th week. Figure 4-6 also shows that pH dropped to 6.5,

probably due to the accumulation of volatile fatty acids.

During the first three weeks the pH ranged between 7.0 and 7.5

then dropped to 6.5 during the 4th week.


































Figure 4-3. Flowthrough bioreactor SEM results of microbial
corrosion of carbon steel coupons. a) Control surface of
coupon unexposed to bacteria and polished to 240 grid.


Figure 4-3. b) Localized pitting near biofilm after exposure
to bacterial triculture.


































Figure 4-3.
near biofilm.


c) SEM micrograph showing microbial corrosion


Figure 4-3. d) SEM micrograph showing generalized metal
deterioration after exposure to bacteria.
































Figure 4-3. e) SEM micrograph showing elongated pits.


Figure 4-4. Flowthrough bioreactor results of EDXA analysis.
a) Control EDXA spectrum of a steel coupon unexposed to
bacteria indicating metallic components of the SA106 alloy.