MICROBIAL INFLUENCED CORROSION STUDIES ON AA2024 ALUMINUM ALLOYS

R. J. Racicot and M. E. Rauch

Department of Chemistry
United States Air Force Academy
USAFA, CO 80840

ABSTRACT

Jet fuel is frequently contaminated with microorganisms that can cause corrosion of fuel tanks.
The wing tank bottom and sides are affected by the growth of a biological sludge associated with the
condensed water. The behavior of AA2024 aluminum alloy in a culture of one of the principal
contaminates, the bacteria Bacillus Licheniformis (BL), is determined. Comparative results of cyclic
polarization and electrochemical impedance spectroscopy (EIS) studies are presented on sterile controls
and inoculated electrochemical cells. Cyclic polarization results show an electropositive shift in
corrosion potential for inoculated cells versus the sterile controls. EIS results show a steady increase in
the charge transfer resistance over several weeks for the inoculated cells versus the sterile controls.
Analysis suggests these electrochemically measured changes reflect the deposition of a protective
biofilm on the alloy’s surface which provides inhibition behavior. Surface changes on the alloy,
influenced from the formation of a biofilm, are shown from results of scanning electron microscopy with
an energy dispersive system (EDS) analysis.

Keywords: Microbiological Influenced Corrosion, MIC, Bacillus Licheniformis, AA2024 Aluminum

Alloy

INTRODUCTION

Microbiological influenced corrosion (MIC) of integral fuel tanks (IFT) was a severe problem
during the late 1950s1,2. The microbial community, primarily fungi, flourished in a combination of
kerosene and saltwater.2 Pitting corrosion was common and pit depths ranged from 0.06 to 0.125
inches. Corrosion attack was severe enough in some cases that aircraft actually leaked jet fuel while
parked on the flight line.3 The occurrence of MIC was drastically reduced by two concurrent actions.
Integral fuel tank interiors were required to be coated by aircraft manufacturers and biostatic additives

1
were added to USAF jet fuels. Following these actions, the threat of MIC has been managed for several
decades. Many of the planes built after MIC abatement, described above, are still in use (e.g., C-17, B-
52, KC-135, E-3, P-3, etc.). The coatings that have been intact for several decades are now beginning to
flake off of the IFT substrates. Most water contamination in jet fuels results from condensation that
occurs within underground fuel tanks. This water is removed from fuel by several filter and inspection
stages before loading onto fuel trucks. Condensation from fuel also occurs in above ground bulk storage
tanks, fuel trucks and aircraft.1,4,5,6 There are several filtering and sump stages to remove contaminants
on fuel trucks before aircraft loading. Filter and sump stages on fuel trucks are routinely sampled.
Thorough wing tank sumping practices also remove much of the condensate in the aircraft IFT. Yet all
these controls still do not remove the water that is soluble in fuel when it is pumped onto aircraft integral
fuel tanks. Ambient and operational temperature fluctuations condense both some of the water from the
fuel, and the moisture content from the air intake. Good maintenance practices minimize the water
content of the fuel tanks; however, some water inevitably remains. The presence of water in the fuel
promotes an environment that is a nutrient-rich breeding ground for the formation of a biofilm with
special distribution of growth and interfacial properties. Chemical contaminates are formed as by-
products of the microbial activities. The formation of a biofilm from microbiological activities may
cause acceleration of corrosion, an inhibiting effect, or not change in materials degradation. The
corrosive role of chemical and microbial fuel system contaminates present in naturally occurring IFT
condensates are not well known for present day systems. If corrosion of IFTs by chemical and microbial
activities is experienced, it reduces operational readiness due to increased maintenance down time.
Bacteria and fungi isolated from United States Air Force (USAF) aviation fuel samples have
been identified, by gas chromatography fatty acid methyl ester (GC-FAME) profiling and 16S and 18S
rRNA gene sequencing by the Air Force Research Laboratory Propulsion Directorate Fuels Branch at
Wright Patterson AFB, OH7. Thirty-six samples from 10 geographically separated USAF bases were
collected. At each base, an above ground storage tank, a refueling truck and an aircraft wing tank were
sampled at the lowest sample point, or sump, to investigate microbial diversity and dispersion within the
fuel distributing chain. Twelve genera, including four Bacillus species and two Staphylococcus species
were isolated and identified. Bacillus Licheniformis (BL), the most prevalent organism isolated, was
found at seven of the ten bases. The isolation of previously undocumented organisms suggests either
changes in aviation fuel microbial community compositions or simply improvements in the isolation and
identification techniques. Microbial community adaptation to changes in aviation fuel composition,
additives and biocide use could also account for the differences.

EXPERIMENTAL PROCEDURE

We selected the bacteria Bacillus Licheniformis (BL) as our representative microorganism test
sample, since this was the most prevalent bacteria found in the collection studies. The particular BL
sample used came from the wing tank of a Hurlburt AFB, FL AC-130 aircraft. The isolated BL was
grown in 500 ml of sterile Luria-Bertani broth (LB) (DIFCO, Sparks, MD). The culture was grown
overnight in a 37 oC incubator/shaker. The confluent cells were transferred to sterile centrifuge tubes,
centrifuged to pellet form and the supernatant removed. The cell pellets were then resuspended in a
0.5M NaCl buffer solution to remove residual LB from the cells. The cells were again centrifuged and
the supernatant removed. This wash procedure was repeated two additional times. The cells were then
resuspended in 500ml of a 0.5 N NaCl buffer. Approximate cell concentration was estimated by
measuring the optical density of the resulting suspension at 600 nm. The bacteria were grown to a
concentration of 95,000 cells per milliliter.
The bacteria salt solution mix was then placed into EG&G flat cells with a thin layer of JP-8 fuel
on the surface of the cell that acted as a food source for the bacteria. The bacteria inoculated and sterile

2
control cells were tested using cyclic polarization scans performed with an EG&G 273A potentiostat
with a saturated calomel reference electrode. Scans were run from -0 1.0V to +0.1V versus the open
circuit potential with a step height of 1.0 mV and a scan rate of 2 mV per second. The cyclic scans were
controlled using Princeton’s Applied Research’s Power Suite software package. Results were compared
with sterile control flat cells run with only the salt solution and a very thin JP-8 fuel layer on the top of
the cell.
Electrochemical impedance spectroscopy cells were constructed using hollow glass tubes with
square flat alloy samples attached to one side of the tube. The tubes were filled with the BL bacteria and
0.5 N NaCl salt solutions and a layer of JP-8 fuel on the top. EIS tests were performed with a PAR 1025
frequency response detector coupled with an EG&G 273A potentiostat controlled with EG&G M398
software. The EIS runs were set at a 5mV AC voltage and the frequency was swept from 10-3 Hz to 106
Hz.
Investigations of the metal substrates before and after exposure to chemical and microbial
condensate environments were performed using an Oxford scanning electron microscope coupled with
an energy dispersive system (EDS) controlled with software from Inca.

RESULTS

Figure 1 shows cyclic polarization results for an inoculated cell compared to the sterile control
cell after two weeks exposure to the test solutions. The inoculated cell shows an approximately 75 mV
electropositive shift of the rest potential over the sterile control. Extrapolation of the cathodic and
anodic Tafel curves shows the corrosion current densities are very similar for both inoculated cells and
the sterile control cell. There appears to be a sharp increase in current at a potential of approximately -
625 mV. This breakdown, or pitting potential, seems to occur at the same potential and at a similar
current density for both cells.
The anodic shift in potential from the exposure of 2024 alloys to fungi and bacteria has been
seen by other researchers in the past.8-10 This electropositive shift in potential seems to indicate the
bacteria attached on the metal surface is offering corrosion protection to the surface. Other researchers
have indicated the possible presence of a biosurfactant by-product being formed on the metal surface.
Ornek et al have shown that the bacteria Bacillus Licheniformis secretes the biofilm polymer γ-
Polyglutamate (γ-PGA).11 It is possible this bio-film is limiting the amount of oxygen diffusion to the
aluminum alloy surface and thereby shifting the corrosion potential of AA 2024 into a more anodic
potential region.
Figure 2 shows electrochemical impedance spectroscopy results for an inoculated cell at day one.
Analysis of the Bode plot of frequency versus the log impedance shows a value of approximately 103.2
for the low frequency impedance value. The leveling off behavior of the low frequency impedance
values allows for the accurate use of this value as the charge transfer resistance value for the sample.
This charge transfer resistance value is very low for a first day EIS test on an AA2024 aluminum alloy.
Values this low for AA2024 alloy are not typically seen except under low pH conditions. It is quite
possible that the bacteria’s presence on the alloy’s surface is changing the localized chemistry and
lowering the pH, thereby dissolving the natural oxide layer and increasing charge transfer from the
surface. However, the measured bulk pH of the test cell does not significantly change. Another
possibility is that the bacteria’s presence on the alloy’s surface is facilitating a more rapid electron
transfer from the interface between the alloy and the biofilm by acting as an electron acceptor or transfer
agent.
This first day EIS charge transfer resistance value is lower than the sterile control at day one.
Figure 3 shows the Bode plot for a sterile control cell containing only the 0.5M NaCl test solution on

3
bare AA2024 alloy. The charge transfer resistance value for this sample is 104, in fact, every bare or
sterile AA2024 alloy tested under EIS has a charge transfer value of this magnitude and this value stays
constant over time for months to years.
The interesting feature from the EIS results is seen when the charge transfer value for the
inoculated cell is tracked over time. Figure 4 shows the Bode plot for the same inoculated test cell
shown in Figure 2 after 14 days exposure to the bacteria solution. The charge transfer resistance value
has increased by an order of magnitude to a value of approximately 104.3. The charge transfer value for
this cell and other tested cells continues to increase over time and appears to level off after about 45
days. Figure 5 shows the plot of charge transfer resistance versus time for an inoculated cell. Every
inoculated cell shows very similar behavior as shown in Figure 5, with the charge transfer value initially
starting out low and within one day the value starts to increase until we observe a leveling off after
approximately 45 days by at least one order of magnitude. Additionally, the test cell’s corrosion
potential was monitored over time. Figure 6 shows the impedance test cell’s corrosion potential with
time. The data shows a correspondingly increase in the corrosion potential, again indicating an
increased inhibition of corrosion on the alloy surface for the inoculated cell. We do not believe this
change in corrosion potential to be associated with ennoblement, as commonly founding nickel based
alloys. The AA2024 alloy does not have any nickel present in its composition and the sterile control
sample does not show a change in corrosion potential as well. The only difference between the
inoculated test cells and the sterile control test cells is the presence of the Bacillus Licheniformis in the
inoculated test cell.
Ornek et al11 showed similar EIS results for engineering grown poly-glutamate from Bacillus
Licheniformis bacteria cultures exposed to AA2024 aluminum alloys. The tracked the relative corrosion
rates values over time and observed a decrease in the corrosion rate attributed to the presence of poly-
glutamate acting as an inhibitive bio-film on the alloy’s surface.
The proposed mechanism for inhibition of corrosion on AA22024 alloys from secreted γ-PGA
films is due to the chemical structure of the polymer. γ-PGA12 is an unusual anionic water-soluble
polyamide. The glutamic acid units are polymerized in a ribosomal-independent manner via amide
linkages between the α-amino and the γ-carboxylic acid functional groups.

O
H
N
* *

O OH

The acidified form of γ-polyglutamate

The carboxylic acid groups along the polymer backbone provide an inhibiting functional
group by chelating or coupling with the aluminum ions or aluminum oxides that are present at the metal-
solution interface via dipole-dipole interactions as proposed by Hefter et al.13 Mclean et al has shown
that the Bacillus Licheniformis capsule has a strong binding affinity to aluminum ions.14
The samples surfaces were investigated after 14 days exposure to the bacteria using the scanning
electron microscope’s EDS analysis. Figure 7a shows the EDS results for an inoculated sample after

4
cyclic polarization. Figure 7b shows a comparison with a sterile control sample with its corresponding
EDS results.
The EDS data offers a unique and quite interesting feature to the behavior of the bacteria on the
alloy surface. The EDS analysis shows twice as much copper on the bacteria exposed surfaces than the
sterile surfaces. Even more interesting is the data that consistently shows four times less magnesium
found on the bacteria-exposed surfaces than on the sterile surfaces. It appears the bacteria are possibly
using the magnesium found in the alloy as part of their biochemical processes, quite possibly as a co-
factor on enzyme or bio-film production. It was postulated in papers by Hedrick et al15 that corrosion of
aluminum alloys by bacterial isolated including Bacillus Licheniformis resulted from the removal of
certain metallic atoms from the alloy structure by extracellular enzyme activity. These results were
particularly prevalent on alloys with high magnesium content. A similar phenomenon is evidenced with
our EDS results showing the reduction of magnesium content on the alloy surface, as seen in an almost
three times reduction in magnesium percent for the inoculated cell versus the sterile control.

CONCLUSIONS

Our cyclic polarization results and the result of other researchers indicate the bacteria Bacillus
Licheniformis might be providing some protection through the deposition of a protective bio-film
composed of polyglutamate on the alloy’s surface. The effect of this biofilm is evidenced by an anodic
shift in the corrosion potential of the alloy’s surface. The EIS data seems to offer support to this
conclusion. The data shows a steady increase in the charge transfer resistance of the alloy’s surface over
time with a leveling off after two weeks. This steady increase in impedance over just a few days could
represent the kinetics of a biofilm formation on the surface.
An interesting feature found on the samples exposed to the bacteria is the decrease in the amount
of magnesium as evidenced from the EDS data and the increase in the amount of copper. It is possible
the bacteria are using the magnesium from the alloy as part of their biochemical processes.
It is premature at this point in the investigation to conclude on the long term effects the presence
of the bio-film will have on the alloy’s surface in terms of localized corrosion, corrosion prevention or
interaction with protective coatings applied to the alloy’s surface.

ACKNOWLEDGEMENTS

The authors wish to thank the United States Air Force Academy’s Research Office,
USAFA/DFER for funding this research project.

REFERENCES

1. F. J. Passman, B. F. McFarland, Understanding, Recognizing, and Controlling Microbial

Contamination in Fuels and Fuel Systems – A Primer. Chevron USA, August 1997.

2. R. C. Salvarezza, H. A. Videla. “Biological and Electrochemical Factors Involved in Microbial;

Corrosion of Aluminum.” Anales Asoc. Quim., Vol 66, pp 317-25, Argentina, 1978.

3. B. R. de Meybaum, E. R. de Schiapparelli. “A Corrosion Test for Determining the Quality of

Maintenance in Jet Fuel Storage.” National Association of Corrosion Engineers, International, Materials
Performance Vol. 19, No. 8, pp. 41-4, 1990

5
4. “Standard Guide for Microbial Contamination of Fuels and Fuel Systems.” Designation ASTM D-
6469. American Standard for Testing and Materials. Pennsylvania, 1999.

5. C. McNamara, T. Perry, N. Wolf, R. Mitchell, R. Leard, J. Dante, “Corrosion of Aluminum Alloy

2024 by Jet Fuel Degrading Microorganisms.” NACE 2003 Corrosion Conference Paper, National
Association of Corrosion Engineers, International, 2003.

6. K. Ferrier, R. Kelly. “Development of an Aircraft Lap Joint Stimulant Environment”, Corrosion, Vol.
58, No. 5, pp. 452-459, National Association of Corrosion Engineers, International, Houston, Texas,
May 2002.

7. M. Rauch, H. W. Graef, S. M. Rozenzhak, S. E. Jones, C. A. Bleckmann, R. L. Kruger, R. R. Naik,

and M. O. Stone, “Characterization of Microbial Contamination in United States Air Force Aviation
Fuel Tanks.” Accepted for publication in Journal of Industrial Microbiology and Technology, 2005.

8. E. S. Ayllon and B. M. Rosales, “Corrosion of AA7075 Aluminum Alloy in Media Contaminated

with Cladosporium resinae”, Corrosion Science, Vol. 44, No. 9, pp 638-643.

9. E. S. Ayllon and B. M. Rosales, “Electrochemical Test for Predicting Microbiological Influenced

Corrosion of Aluminum and AA7075 Alloy”, Corrosion, Vol 50, No. 8, pp571-575.

10. R. C. Salvarezza, A. F. L. de Mele and H. A. Videla, “The Use of Pitting Potential to Study the
Microbial Corrosion of 2024 Aluminum Alloy”, Int. Biodeterior. Bull. (ISSN 0020-6164) 15 (4) 1979,
pp125-132.

11. D. Ornek, A. Jayaraman, B. C. Syrett, C. H. Hsu, F. B. Mansfeld, T. K. Wood, “Pitting Corrosion

Inhibition of Aluminum 2024 by Bacillus Biofilms Secreting Polyaspartate or Gamma-Polyglutamate.”,
Appl. Microbiol. Biotechnol. (2002) 58, pp651-657.

12. G. Birrer, A. Cromwick and R. Gross, “gamma-Poly(glutamic acid) formation by Bacillus

licheniformis 9945a:physiological and biochemical studies.” , Int. J. Biol. Macromol. Vol 16 Number 5
1994, pp265-275.

13. G. T. Hefter, N. A. North and S. H. Tan, “Organic Corrosion Inhibitors in Neutral Solutions; Part 1-
Inhibition of Steel, Copper and Aluminum by Straight Chain Carboxylates”, Corrosion, Vol 53, No. 8,
1997, pp657-667

14. R. J. Mclean, D. Beauchemin, L. Clapham and T. J. Beveridge, “Metal-binding Characteristics of

the Gamma-Glutamyl Capsular Polymer of Bacillus Licheniformis ATCC 9945”, Applied and
Environmental Microbiology, Vol 56, No. 12, Dec. 1990, pp3671-3677.

15. H. G. Hedrick, R. J. Reynolds and M. G. Crum, “Microbiological Corrosion of Aircraft Metal

Alloys”, Developments in Industrial Microbiology, 10, 1969, pp228-233.

6
Figure 1. Cyclic polarization results for a Bacillus Licheniformis inoculated cell versus a
sterile control cell for AA2024 alloy after two weeks exposure to the respective test solutions.

7
 .
Figure 2. EIS Bode plot results for the first day of testing for an inoculated cell. Low
frequency impedance shows a low charge transfer resistance value of 103.2.

Figure 3. Electrochemical impedance spectroscopy results for AA2024 alloy for the sterile control cell
with no bacteria present in the test solution. The low frequency impedance shows a charge transfer
resistance value of 104.

8
 Figure 4. EIS results for the same inoculated cell shown in figure 2 after a two week exposure
to the bacteria solution. The low frequency impedance value shows an increase in charge
transfer resistance to 104.3.

Charge Transfer Resistance versus Time

35000
Charge Transfer Resistance (Ohms)

30000

25000

20000

15000

10000

5000

0
0 10 20 30 40 50
Time (Days)

Figure 5. Charge transfer resistance values for an EIS test cell inoculated with Bacillus Licheniformis
versus time. Diamond symbols show the inoculated cell charge transfer resistance values, the square
symbols show the sterile control cell.

9
 Corrosion Potential versus Time

-0.6

-0.65
Potential (volts)

-0.7

-0.75

-0.8

-0.85
0 10 20 30 40 50
Time (days)

Figure 6. Corrosion potential versus time for an EIS test cell inoculated with Bacillus Licheniformis.
Diamond symbols, show the corrosion potential fort he inoculated cell, the square symbols show the
sterile control cell corrosion potential values.

a. b.

Figure 7. 7a.) EDS data for a sample run under cycle polarization exposed to Bacillus Licheniformis for
two weeks. The data shows twice as much copper precipitate on the surface than a sterile control test
seen in Figure 7b and four times less magnesium than the sterile control test sample. The sample also
shows the presence of sulfur on the surface, not seen in the sterile control samples.

10