Previous Article | Next Article 
Applied and Environmental Microbiology, October 1998, p. 3690-3697, Vol. 64, No. 10
0099-2240/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.
Ellipsometric Measurement of Bacterial Films at
Metal-Electrolyte Interfaces
J. P.
Busalmen,1
S. R.
de
Sánchez,1 and
D. J.
Schiffrin2,*
INTEMA, Facultad de Ingeniería,
Universidad Nacional de Mar del Plata, 7600 Mar del Plata,
Argentina,1 and
Chemistry Department,
University of Liverpool, Liverpool L69 7ZD, United
Kingdom2
Received 7 May 1998/Accepted 10 August 1998
 |
ABSTRACT |
Ellipsometric measurements were used to monitor the formation of a
bacterial cell film on polarized metal surfaces (Al-brass and Ti).
Under cathodic polarization bacterial attachment was measured from
changes in the ellipsometric angles. These were fitted to an effective
medium model for a nonabsorbing bacterial film with an effective
refractive index (nf) of 1.38 and a thickness (df) of 160 ± 10 nm. From the optical
measurements a surface coverage of 17% was estimated, in agreement
with direct microscopic observations. The influence of bacteria on the
formation of oxide films was monitored by ellipsometry following the
film growth in situ. A strong inhibition of metal oxide film formation
was observed, which was assigned to the decrease in oxygen
concentration due to the presence of bacteria. It is shown that the
irreversible adhesion of bacteria to the surface can be monitored
ellipsometrically. Electrophoretic mobility is proposed as one of the
factors determining bacterial attachment. The high sensitivity of
ellipsometry and its usefulness for the determination of growth of
interfacial bacterial films is demonstrated.
| |
INTRODUCTION |
There is a wide interest in
bacterial attachment to surfaces since this strongly influences many
natural and industrial processes. Besides optical microscopic
observations (4) important new advances have been made in
the investigation of bacterial attachment, employing modern optical
techniques. Evanescent field Fourier-transform infrared spectroscopy
has been extensively used for monitoring bacterial colonization of
surfaces (36, 37), for instance, by following an amide band
as a marker for biofilm biomass (38). Importantly, Suci et
al. (43) have demonstrated that complementary data can be
obtained from attenuated total reflection-Fourier-transform infrared
spectroscopy and reflected differential interference contrast
microscopy for the in situ observation of hydrodynamic effects on the
chemistry and architecture of biofilms.
Bacterial adhesion to metal surfaces and formation of biofilms is known
to lead to the enhancement of corrosion rates (15, 17, 29, 30,
33), and localized corrosion is usually observed as a consequence
of bacterial surface colonization. The mechanisms by which bacteria
recognize and approach metal surfaces are not clear, but positive
chemotaxis in the presence of heavy metal ion gradients from a
corroding metal has been demonstrated (16). Once bacteria
reach a solid surface both reversible and time-dependent irreversible
adhesion processes take place (32). Primary events in
substrate-bacteria interaction are well predicted by the Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory of colloidal systems on
hydrophobic low-energy surfaces, for instance, on polymers. In this
case, long-range interactions (van der Waals) seem to play a major role
in surface recognition (6, 45). Also, interactions with hydrophilic high-energy surfaces, such as those of metals, are
significantly influenced by short-range double-layer effects (20) since most bacteria bear a negative surface
charge.
Ellipsometry is a form of reflectance spectroscopy which measures the
changes in amplitude and in degree of polarization of light reflected
from a surface. These changes are expressed in terms of two so-called
ellipsometric angles, 
and 
(3, 22, 35). gives the
difference in the phase shift experienced by the reflected light beam
for p and s optical polarization (p
and s polarization are parallel and perpendicular,
respectively, to the plane of light incidence), whereas tan
is the ratio of the attenuation of the amplitude on reflection for
p and s polarized light. Recent developments in
ellipsometric equipment and software (25) have greatly
simplified the measurement of these angles and the establishing of
their relationship to the properties of the surface under
investigation. In particular, since ellipsometry is a very sensitive
surface technique, it can be conveniently used for the analysis of
films on substrates. In the present work the effects of bacterial
attachment at polarized and nonpolarized metal surfaces have been
monitored by this technique and theoretical models have been used to
determine interfacial optical properties of bacterial and oxide films.
| |
MATERIALS AND METHODS |
Equipment and cells.
Ellipsometric measurements were
conducted with a computer-controlled Gaertner L126 ellipsometer with a
helium-neon laser (
= 632.8 nm) at an angle of incidence of 70°.
The electrochemical cell used has been previously described
(8); it had appropriate entrances for counter and reference
electrodes. The working electrode was placed at the bottom of the cell,
and the cell body was secured to the cell base by side screws. Flat
working electrodes of 2-cm2 exposed area were first abraded
with 600 grit carborundum paper, polished with 0.05-µm alumina (type
B; Buehler Ltd.), and finally rinsed with distilled water. The metal
surfaces investigated were from materials commonly used in seawater
heat exchangers: aluminum brass (ASTM B111) and titanium (99.6+%;
Goodfellow, T1000431). All the experiments were performed in a 3.5%
NaCl (BDH; AnalaR) solution, close to the chloride concentration of
seawater. A KCl-saturated calomel electrode (SCE; Radiometer) was used
as the reference electrode; the counterelectrode was made of a platinum
wire immersed in the solution under study.
Biological material.
A pure culture of a
Pseudomonas sp. strain isolated from a copper base alloy
heat exchanger tube (5) was used. Cultures were grown at
32°C until the mid-exponential phase, as determined by the absorbance
at 600 nm (A600) (12). A nutrient
broth containing 0.1% meat extract (Difco), 0.2% yeast extract (
Lab), and 0.5% Bacto Peptone MC24 (Lab M) in artificial seawater (pH
8) was employed (9). The cells were harvested by
centrifugation for 10 min at 9,300 rpm in a Sorvall RC 5B refrigerated
centrifuge, washed with 3.5% NaCl at a pH of 7.2, centrifuged again,
and resuspended in the same solution. In order to ensure an equal
bacterial concentration for each experiment the absorbance of the test
solution was set to an A600 of 0.150 (approximately 105 bacteria cm
3) by
appropriate dilutions with 3.5% NaCl.
For the experiments to determine the influence of possible charge
reversal of the cell wall by adsorption of Cu2+ ions
(9) on attachment rate, CuCl2 was added to
bacterial suspensions to a final concentration of 1 mM. After an
incubation period of 10 min the cells were harvested by centrifugation
at 9,300 rpm in the Sorvall centrifuge and resuspended in 3.5% NaCl at
a pH of 7.2.
Ellipsometric measurements.
The ellipsometric angles and
were measured at constant time intervals, and the data were stored
in a microcomputer (24). Before each experiment, the
electrodes were prereduced by keeping them for 10 min at 
0.6 V in the
NaCl solution in order to obtain a reproducible bare substrate as a
starting point for the measurements. A total of 2 ml of bacterial
suspension was added to the cell, and measurements were taken every
30 s. In the charge reversal experiments 2 ml of the
Cu2+-treated bacterial suspension was used. For the oxide
growth experiments, the electrode was first kept at 0.6 V for
approximately 30 min in the bacterial suspension. After this time, the
potentiostat was disconnected and the electrode was allowed to reach
the corrosion potential where oxide growth takes place. Measurements of
the ellipsometric angles were taken at least for an extra 20 min and all ellipsometric calculations were carried out with commercial software (23).
| |
RESULTS AND DISCUSSION |
Optical properties of the substrate.
Ellipsometric
measurements of thin films require a well-characterized substrate
(39). For copper and its alloys the rapid growth of oxide
films in air after sample polishing results in a scatter of the initial
measured values of and . For this reason, it is better to use
changes of the ellipsometric angles (
and ), relative
either to the bare substrate or to a reproducible surface condition,
for calculating film parameters (8).
and
values for blank experiments on Al-brass surfaces
polarized at
0.6 V are shown in Fig.
1.
Both
and
rapidly
reach constant values, but significant changes
in
with time
were observed after approximately 500 s from the
time of applying
the potential. The properties of corroding
surfaces are variable
during the initial phase of exposure to the
corrosive medium.
For this reason, the values of
and
were
different for different
samples during the first ~500 s of exposure.
After this period,
steady-state properties were observed. The complex
refractive
index,
ns = n ksi, for the metal in contact with the solution
was
computed from the data for which
and
were constant. Taking
the
refractive index of the solution as
nsol = 1.378, and using
commercial software (
22), a value for
ns of 0.645
3.35
i was
calculated. A full description of the ellipsometric equations
used in
these calculations can be found in references
3 and
35.
View larger version (18K):
[in this window]
[in a new window]
|
FIG. 1.
Variations of (a) and (b) during cathodic
polarization of Al-brass in 3.5% NaCl in the absence ( ) and
presence ( ) of bacteria. The potential was kept constant at E = 0.6 V. The vertical lines show the times at which bacteria were
added.
|
|
The real component of the index is very sensitive to the surface
composition of the alloy and/or to the presence of surface
oxides. However, the decrease observed cannot be due to surface
dealloying since the value of
would increase if a layer of copper
is present on the alloy surface as a consequence of minor surface
dealloying. For example, it was calculated that a 1-nm layer of
copper
with
ncu = 0.14
3.45
i
(
8) increases
by 0.1 deg (
23).
In contrast,
the same calculations based on the assumption of
the presence of
Cu
2O (
8) or ZnO (
48) films lead to a
strong
decrease of
with coverage. Therefore, the observed long-term
decrease of
with time (Fig.
1a) probably results from the formation
of an oxide submonolayer, as previously observed for copper
(
8).
The value of
ns obtained at
times between 200 and 500 s was used
in the following
calculations. As will be seen further on, the
main conclusions are not
affected by this choice.
The corresponding changes of
and
with time for titanium can be
seen in Fig.
2. The refractive index of
the metal was calculated
in the same way as for Al-brass, and a complex
refractive index
for the substrate of
ns = 2.49
3.09
i was obtained. An
ns value
of 3.23
3.62
i has
been previously obtained at 632.8 nm for electropolished
titanium in a
sulfuric acid solution (
39). The difference in
the
optical constants found in the present work could be accounted
for by
the surface coverage of the metal by TiO
2.
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 2.
Variations of (a) and (b) during cathodic
polarization of titanium in 3.5% NaCl in the absence () and
presence () of bacteria. E = 0.6 V. The vertical lines show
the times at which bacteria were added.
|
|
Formation of a bacterial film.
As can be seen in Fig. 1, reaches an almost constant value when bacteria are present, whereas the
slow formation of an oxide submonolayer is observed for the blank. The
addition of bacteria to the solution causes changes in and of
approximately ~0.15 and ~0.05°, respectively, for Al-brass (Fig.
1). These changes fitted well a single-film model for a transparent
film (with a value of kf, the imaginary
component of the film refractive index [3] of 0 at
632.8 nm) (34) with an effective thickness for a film
(df) of 160 nm and an effective refractive index
for a film (nf) of 1.38. The sensitivities of
the calculated values of df and
nf to the optical constants of the substrate
used in the calculations are shown in Fig.
3 for different values of
ns. No significant changes in the calculated
values of nf or df for the bacterial film are observed for a wide range of optical constants of the substrate. Only ks has a minor effect on
the calculated film thickness, and it can be concluded that the
bacterial film thickness is 160 ± 10 nm. Similar results were
obtained with titanium.
View larger version (12K):
[in this window]
[in a new window]
|
FIG. 3.
Influence of the substrate optical constants on the
calculated bacterial film refractive index and thickness for
ks = 3.35 (a) and ns = 0.645 (b).
|
|
It is recognized that the initial step of bacterial adhesion to solid
surfaces is a reversible adsorption phase (
6,
21,
31,
32,
45). The DLVO theory predicts a secondary energy
minimum suitable
for reversible adsorption (
6,
32,
45)
which is the result of
the balance between long-range attractive
forces and the repulsive
overlap between the ionic double layers
of the charged interacting
surfaces. In these experiments the
metal surface is negatively charged
since the potential of zero
charge for copper in chloride solutions is
0.26 V (SCE), i.e.,
it is very positive with respect to the
potential of
0.6 V maintained
during the adsorption experiments
(
42). The bacteria also carry
a negative charge on the
cell wall (
9) and therefore will experience
an electrostatic
repulsion on approaching the metal surface. However,
the high-salt
concentration used places the secondary energy minimum
very close
(~0.5 nm) to the surface, thus allowing bacterial adsorption
to
take place in spite of short-range electrostatic repulsion
(
45). These phenomena are well known from colloidal
chemistry
(
1,
46). Figure
1 and subsequent analysis
(see below) also
indicate that bacterial adsorption occurs
on a submonolayer oxidized
surface (
8).
Optical model for the bacterial films.
It is proposed that the
observed changes in and are a consequence of the adsorption of
a bacterial layer, and in order to validate this assumption a model
describing the changes in the interfacial optical properties due to
bacterial attachment was considered. Coverage of the metal surface can
be calculated assuming a layer of constant thickness and variable
refractive index (44). The cells of the
Pseudomonas sp. strain used are rod shaped and can be
regarded as cylinders of constant width, capped with hemispherical
poles (11). Previous studies of Pseudomonas adsorption have shown that the cells are attached with an orientation either perpendicular or parallel to the surface, with no intermediate angles (18). Optical microscopy revealed a parallel
orientation and the cell width was found to be approximately 1 µm. This was taken as the thickness of the bacterial film layer, and
its refractive index was calculated by using an effective medium
theory. Since the filling factor, i.e., the volume fraction of bacteria
in the interfacial film volume, as observed by optical microscopy (see below) is very low, the simple Maxwell-Garnett approximation
corresponding to noninteracting particles can be used for modeling the
optical properties of the film. The nf for a
film consisting of particles immersed in a medium of refractive index
nm is given as follows (2):
|
(1)
|
where
f is the filling factor, i.e., the volume
fraction of particles with a refractive index for suspended bacteria
(
nb)
present in the film.
nm represents the refractive index of the
other
film component, which in the present case has been taken
as the bulk
solution.
Morel et al. (
34) showed that the real component of
nb is 1.05 times the refractive index of the
surrounding solution. A
more recent study by Walthman et al.
(
47) gave a value for
nb of
(1.064 ± 0.015)
nm, from which an average
nb of 1.47 ± 0.02
was estimated. The
imaginary component of the index to be considered
is negligible
(
34). From this and the
nf value of
1.38 fitted
to the data of Fig.
1, a filling factor corresponding to a
coverage
of 17% ± 4% was obtained from equation 1. This is
equivalent to
an average film thickness of 174 nm, a result in good
agreement
with the previous calculation of d
f = (160 ± 10) nm obtained assuming
a uniform film. The difference between the
two calculations is
that a more detailed model has been considered in
the former,
using an independent estimate of the refractive index of
the bacteria.
To confirm these values, the coverage was measured at the
end
of the ellipsometric experiments using photographs (×1,000) from
an optical microscope. Coverages ranging from 12 to 16% were obtained,
which are in good agreement with the filling factors calculated
from
the ellipsometric data. Thus, the three approaches used,
simple average
film thickness considerations directly fitted from
ellipsometric data,
an effective medium model, and direct microscopical
observations, lead
to similar results, giving confidence in the
use of ellipsometry to
monitor bacterial attachment to metal surfaces.
Effect of bacteria on oxide formation.
In order to investigate
the influence of the presence of bacteria at the metal-solution
interface on oxide film formation, bacterial growth on Al-brass was
monitored at the end of each experiment by allowing the sample to reach
the corrosion potential where oxide film growth is known to occur
(41). The results of these experiments are shown in Fig.
4. In the absence of bacteria a strong
decrease in accompanied also by an increase in are indicative
of the rapid formation of an oxide film at the corrosion potential
(12). The dependence of on could not be fitted to a
single-film model with a constant nf value.
Figure 5 shows a comparison between the
observed changes of and (filled points) and theoretical
calculations obtained using a one-oxide film model; the effect of
variations of nf and df
on the - signature are also shown. This is a convenient way of
visualizing the variations of ellipsometric angles for different
refractive indices and film thicknesses. It can be seen in this figure
that the experimental - signature is comprised within an envelope
of values for nf between 2.3 and 3 and an oxide
film thickness from 0 to 5 nm. kf was taken as
zero (26). The optical constants for the Al-brass substrate
were the same as those used above. It is noteworthy that an increment
in the value of nf was observed during the
growth of the film (Fig. 5). This can be related to different
time-dependent growth rates of the oxides of the alloying metals.
During the initial oxide growth period zinc is preferentially oxidized
(28) and the variations of and are primarily due to
the formation of a ZnO layer (nf = 2.019)
(48). The later growth of Cu2O
(n = 2.8) (8) into this oxide film leads to
the observed increase in the value of nf.
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 4.
Variations of (a) and (b) with the growth of an
oxide film on Al-brass in 3.5% NaCl at the free corrosion potential in
the absence () and presence () of bacteria. Arrows indicate the
times when the potential was allowed to reach its free corrosion value.
The vertical line in b shows the time at which bacteria were added.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
FIG. 5.
Variations of and with the growth of an oxide
film on Al-brass in 3.5% NaCl. , experimental data (from Fig. 4)
from the initial phase of oxide growth at the free potential;  ,
calculated data (16) for nf values of
2.3 to 3.0 in 7 steps (the arrow indicates the direction of increase)
and for df values of 0 to 5 nm in 10 steps.
kf = 0.
|
|
Figure
4 shows that there are two distinct oxide growth rates at the
corrosion potential in the presence of adsorbed bacteria.
During an
initial oxide growth phase of ~170 s at the corrosion
potential, a
fast decrease of
is observed; this is followed
by a further slow
decrease at longer times. The rate of decrease
of
is proportional
to the oxide film growth kinetics (
8),
and from a comparison
of the oxide growth with and without bacteria,
it is apparent (Fig.
4a)
that oxide growth is inhibited by the
presence of bacteria. The rate of
oxide formation at the free
corrosion potential is controlled by the
concentration of electroactive
species in solution constituting the
cathodic reaction of the
corrosion couple. For copper alloys in
chloride solutions the
latter is the reduction of oxygen (
14,
15), and the lower
oxide growth rate observed can therefore
result from the lower
oxygen concentration present at the surface as a
consequence of
cell respiration both in the solution and within the
absorbed
layer. Both effects should be considered; the first would
certainly
decrease the amount of O
2 available for the
corrosion reaction.
The second proposal relates to the well-known
behavior of ultramicroelectrodes
and reaction centers at surfaces
in electrochemical systems (
19,
40). Although the
coverage is less than 1/5 of the total surface,
the nature of the
diffusional fields of O
2 must be taken into
account. The
bacteria adsorbed on the surface behave like ultramicroelectrodes
for which hemispherical diffusional fields prevail. For this
diffusional
geometry the distance for which deprivation of
electroactive reagent
occurs is approximately five times the
ultramicroelectrode diameter
(
40). Therefore, oxygen
deprivation of the surface will take
place with the consequent
decrease of the rate of metal oxidation.
In addition, local
acidification can result from the metabolic
activity of the cells,
which would also lead to a lower oxide
thickness.
The practical consequences of the above require further consideration.
A model proposed for the structure of aerobic biofilms
(
13)
describes a complex structure, consisting of microbial
cell clusters
and interstitial voids, which strongly influences
oxygen distribution.
In particular, the cathodic reaction becomes
diffusionally
controlled in this case (
15). In addition, low-oxygen
concentrations are present underneath the cell clusters
(
13).
For this situation, the results in Fig.
4a and b
suggest that
the thickness of the passivating film under the cell
clusters
of a biofilm will be less than under the voids in the biofilm
structure.
Significant changes were observed in the
values during oxide film
growth (Fig.
4b) in the presence of bacteria. In contrast
with the
increase in
for the sterile control, a fast decrease
in this
angle was observed during the first 120 s at the corrosion
potential, followed by a slower decrease, which is similar to
that
observed for
. It is important to note that, in general,
increases and
decreases with the growth of a transparent oxide
film
(
22). The decrease observed in
is unusual considering
that the calculations in Fig.
5 clearly show the growth of an
oxide
with the characteristic increase in
and a slope
d/d as predicted for
kf = 0.
Irreversible bacterial adhesion.
It is proposed that the
changes in the time dependence of and during the
growth of the oxide film in the presence of bacteria (Fig. 4) are
consequences of the change from reversible to irreversible in the
adhesion of the cells to the surface undergoing oxidation. Ellipsometry
is a very sensitive technique for the study of interfacial films and
this behavior can be related to a small change in the number of
bacteria present in the film, as calculated below. The signature
for nonabsorbing films is determined by the refractive index of the
films present, and variations in Fig. 4b must be related to a
decrease in the value of nf for the bacterial
film and/or that of the oxide layer. The influence of bacterial
attachment and of oxide composition changes, or growth, will be
analyzed separately since all these processes can take place.
The refractive index of the bacterial film is very sensitive to the
value of the filling factor (
f), as is shown in Fig.
6 where a linear decrease of the
calculated values of
nf (equation
1) with an
increase in
f can be observed. A minor increase in
the
number of attached bacteria of less than 1% in
f can
produce
a large decrease of the refractive index of the film due to the
large film thickness being considered. The significance of such
variations can be appreciated from the calculations shown in Fig.
7 where each point (open circles)
corresponds to a different value
of the bacterial film thickness and
refractive index. Superimposed
on these data are experimental
and
values obtained during
the first 150 s after the metal
is placed at the free corrosion
potential in the presence of
bacteria (Fig.
4). Considering only
at first bacterial film changes,
the experimental results could
be an indication of a change in
nf of ~0.01 and a progressive
increase in the
average bacterial film thickness from the initial
adsorption value of
160 to 190 nm. In this case, the changes in
interfacial conditions
(i.e., decreased electrostatic repulsion)
when the metal is placed at
the free potential induce an increase
in the number of bacteria present
in the film.
View larger version (10K):
[in this window]
[in a new window]
|
FIG. 6.
Variation of the effective refractive index for a
bacterial film with the filling factor corresponding to the interfacial
model in equation 1.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
FIG. 7.
- signature corresponding to a one-film model on
Al-brass with nf values of 1.37 to 1.382 in 0.001 steps and
df values of 100 to 250 nm in 10-nm steps () and
experimental and values from the first 150 s at the
free corrosion potential (). The arrows indicate the direction of
increase.
|
|
The decrease of
(Fig.
4b) when bacteria are present can also be
related to changes in the oxide layer. The presence of a
Cu
2O submonolayer during polarization at
0.6 V
(
8) was observed
(Fig.
1a), but as was previously discussed,
zinc oxide is preferentially
formed during the initial oxide growth
when the metal is allowed
to reach its free corrosion potential
(
28). Changes in oxide
layer composition after allowing the
material to reach the free
corrosion potential can account for the
decrease in
nf from the
value corresponding to
that of Cu
2O (2.8) to a value that approaches
that of ZnO
(2.02). Figure
8 shows the
-
signature for a two-film
model when the refractive index of the oxide
is decreased from
2.8 to 2.0, simultaneously with the optical changes
introduced
by an increase of the bacterial film thickness from 160 to
190
nm. The rationale for these calculations is the need to account
for
both a smaller decrease in
and a decrease, instead of an
increase,
in
when comparing oxide growth in the presence and
absence of
bacteria. Although changes in oxide thickness will
occur, the main
features of the
-
dependence can only be modeled
on the basis of
an oxide film with a bacterial layer attached
on top of it. Any other
combination leads to an actual increase
in
against experimental
observations. The influence of an increase
in the thickness of the
oxide layer on the changes of
and
is also considered in the
analysis in Fig.
8 (filled circles).
From these results it is proposed
that the variations in
and
when the metal is allowed to reach
its corrosion potential (Fig.
4) are the product of at least two
combined processes, the increase
of the filling factor of the bacterial
film (Fig.
6 and
7) and
the decrease in the refractive index of the
oxide layer due to
its enrichment with ZnO (Fig.
8).
View larger version (14K):
[in this window]
[in a new window]
|
FIG. 8.
Changes in the - signature for a two-film model
composed of a bacterial film (refractive index of 1.38, as previously
found) over an oxide film on Al-brass. The values of
nf for the oxide film are changed from 2.8 to
2.0 in eight steps and the calculations were carried out for bacterial
film thicknesses of 160 and 190 nm. , 1-nm oxide film thickness;
, 1.5-nm oxide film thickness. The arrow indicates the direction of
change of giving values comparable with the experimental results.
|
|
The full analysis of the unusual response of
and
when the
material is allowed to reach its free corrosion potential is
obviously
extremely complex and, with only two measured parameters,
only general
features can be depicted. It is quite clear, however,
that the simple
one-film model implied in the
-
plane representation
given in
Fig.
7 is an oversimplification. A more physically realistic
two-film
model (Fig.
8) gives a better interpretation of the results,
considering that the optical constants of all the film compounds
and of
the substrate are known. The calculated bacterial film
thickness can
only be regarded as tentative, although the microscopic
observations
give support to the values quoted.
Electrophoretic migration.
Irreversible adsorption of bacteria
as predicted by the DLVO theory is only possible if the energy barrier
determined by electrostatic repulsion does not exceed a few kT units
(45). Four possible situations for irreversible adhesion
have been described by van Loosdrecht et al. (45): (i) the
surface is positively charged, (ii) both the metal and the bacterial
surfaces are hydrophobic and carry a low surface charge density, (iii)
a high concentration of electrolyte is present, and (iv) bacteria have
special surface appendages that can bridge the distance between the
cell wall and the metal surface.
When the electrode is allowed to reach the free corrosion potential,
the shift to positive potentials reduces the negative
surface charge
density of the metal. In addition, both the high
ionic strength
employed, which reduces the width of the electrical
double layer, and a
reversal of bacterial wall surface charge
density, due to the presence
of Cu(I) in the diffusional field
close to the metal surface, can
enhance irreversible adhesion.
Collins et al. (
9) reported a
change in bacterial electrophoretic
mobility to positive values in the
presence of heavy metal ions
such as Cu
2+,
Zn
2+, and Ni
2+ at neutral pH. The ability of
bacteria to sense metal ions at
considerable distances from a corroding
electrode is very intriguing
(
16). Besides a biological
chemotaxis mechanism previously discussed
(
16), adsorption
of metal ions on the bacterial cell walls through
coordination to
surface groups can lead to charge reversal. To
check this possibility
cells were treated with 1 mM CuCl
2 in 3.5%
NaCl solution
for 10 min and their behavior on titanium was studied
by ellipsometry.
Titanium was used in preference to the Cu alloy
to avoid the presence
of copper metal ions due to metal corrosion.
It was expected that by
using this approach the influence of bacterial
charge effects on the
attachment rate could be observed. Cells
extracted after treatment with
CuCl
2 showed a similar viability
to that of untreated
cells. Similar survival resistances to surface
coordination by
heavy metal ions leading to charge reversal have
been observed for
other bacteria (
10).
Figure
9 shows the results of these
experiments after addition of bacteria at
0.6 V. These results have
to be compared with
those obtained in the absence of bacteria (Fig.
2).
A large linear
decrease in
and a corresponding increase in
were
observed.
In these Cu(II) surface-derivatized bacterial preparations
the
cell walls bear a net positive charge (
9). Although a
full
analysis is not possible at present and a biological process might
be operative, the increased rate of adsorption of Cu-treated bacteria
on the surface indicates that electrophoretic migration can
represent
an important mechanism by which bacteria are
attached to the metal
surface. Similar charge effects have been
proposed by Jucker et
al. for the adhesion of
Stenotrophomonas
maltophilia 70401 to
negatively charged glass surfaces
(
27).
View larger version (18K):
[in this window]
[in a new window]
|
FIG. 9.
Variations of (a) and (b) during cathodic
polarization of titanium in 3.5% NaCl in the absence () and
presence () of Cu-treated bacteria. E = 0.6 V. The vertical
lines show the times at which bacteria were added.
|
|
Conclusions.
The usefulness of ellipsometric techniques
applied to the study of the formation of interfacial bacterial films
has been highlighted. It has been shown that bacterial adhesion can be
followed by this method and reliable information can be obtained by the
use of an appropriate interfacial model. The experimental evidence
indicates that bacterial adsorption occurs at polarized metal surfaces. From the ellipsometric results a coverage of adsorbed bacteria, in an
orientation parallel to the surface, of 17% has been calculated, and
these results have been confirmed from direct microscopic observations.
The kinetics of oxide growth is inhibited by the presence of bacteria.
This is probably due either to local acidification
at the bacterial
film-metal solution interface or to the restricted
diffusion of
oxygen through the bacterial film, which results
in a lower oxide film
thickness. It is important to note that
in practical applications
differences in the thickness of the
passivating film as a product of
the unequal distribution of oxygen
through a biofilm structure can lead
to severe corrosion effects
resulting from localized differential
aeration of the corroding
surface (
13).
The rate of irreversible attachment of bacteria to metal surfaces can
be increased by adsorption of metal ions on their surface
walls, which
results in charge reversal which both decreases the
energy barrier for
attachment and introduces an electrophoretic
mobility component to the
rate of surface colonization. These
effects also occur during oxide
growth at the free potential by
the establishment of a metal ion
gradient in the proximity of
the corroding surface.
| |
ACKNOWLEDGMENTS |
The financial support from The British Council (British
Council-Fundación Antorchas program) for staff exchanges is
gratefully acknowledged.
We thank Clementina Gomis Bas for help with the ellipsometric data
analysis.
| |
FOOTNOTES |
*
Corresponding author. Mailing address: Chemistry
Department, University of Liverpool, Liverpool L69 7ZD, United Kingdom.
Phone: 44 151 794 3574. Fax: 44 151 794 3588. E-mail:
d.j.schiffrin{at}liv.ac.uk.
| |
REFERENCES |
| 1.
|
Albery, W. J.,
R. A. Fredlain,
G. J. O'Shea, and A. L. Smith.
1990.
Colloidal deposition under conditions of controlled potential.
Faraday Discuss. Chem. Soc.
90:223-234.
|
| 2.
|
Aspens, D. E., and J. B. Theeten.
1979.
Investigation of effective-medium models of microscopic surface roughness by spectroscopic ellipsometry.
Phys. Rev. B
20:3292-3302.
|
| 3.
|
Azzan, R. M. A., and N. M. Bashara.
1987.
Ellipsometry and polarized light.
North Holland, Amsterdam, The Netherlands.
|
| 4.
|
Bakke, R., and P. Q. Olsson.
1986.
Biofilm thickness measurements by light microscopy.
J. Microbiol. Methods
5:93-98.
|
| 5.
| Busalmen, J. P., M. A. Frontini, and S. R. de Sánchez. Microbial corrosion: effect of the microbial
catalase on the oxygen reduction. In F. C. Walsh and
S. A. Campbell (ed.), Recent developments in marine corrosion.
Royal Society of Chemistry, London, United Kingdom, in press.
|
| 6.
|
Busscher, H. J., and A. H. Weerkamp.
1987.
Specific and non-specific interactions in bacterial adhesion to solid substrata.
FEMS Microbiol. Rev.
46:165-173.
|
| 7.
|
Butcher, E. C.,
A. J. Dyer, and N. E. Gilbert.
1968.
Ellipsometric measurements of oxide films on copper.
Brit. J. Appl. Phys.
1:1673-1678.
|
| 8.
|
Ceré, S.,
S. R. de Sánchez, and D. J. Schiffrin.
1995.
Potential dependence of the optical properties of the cooper/aqueous borax interface.
J. Electroanal. Chem.
386:165-171.
|
| 9.
|
Collins, Y. E., and G. Stotzky.
1992.
Heavy metals alter the electrokinetic properties of bacteria, yeasts, and clay minerals.
Appl. Environ. Microbiol.
58:1592-1600[Abstract/Free Full Text].
|
| 10.
|
Collins, Y. E., and G. Stotzky.
1996.
Changes in the surface charge of bacteria caused by heavy metals do not affect survival.
Can. J. Microbiol.
42:621-627[Medline].
|
| 11.
|
Cooper, S.
1991.
Bacterial growth and division, p. 147.
Academic Press, San Diego, Calif.
|
| 12.
|
Davis, B. D.,
R. Dulbecco,
H. N. Eisen, and H. S. Ginsberg (ed.).
1980.
Microbiology 1980, p. 64.
Harper & Row, Philadelphia, Pa.
|
| 13.
|
de Beer, D.,
P. Stoodley,
F. Roe, and Z. Lewandowski.
1994.
Effects of biofilm structures on oxygen distribution and mass transport.
Biotechnol. Bioeng.
43:1131-1138.
|
| 14.
|
de Sánchez, S. R., and D. J. Schiffrin.
1982.
Flow corrosion mechanism of copper alloys in sea water in the presence of sulphide contamination.
Corros. Sci.
22:585-607.
|
| 15.
|
de Sánchez, S. R., and D. J. Schiffrin.
1985.
The effect of pollutants in the corrosion of copper base alloys by sea water.
Corrosion
4:1-31.
|
| 16.
|
de Sánchez, S. R., and D. J. Schiffrin.
1996.
Bacterial chemo-attractant properties of metal ions from dissolving electrode surfaces.
J. Electroanal. Chem.
409:9-14.
|
| 17.
|
Dexter, S. C.
1996.
Biofouling and biocorrosion.
Bull. Electrochem.
12:1-7.
|
| 18.
|
Escher, A., and W. C. Characklis.
1990.
Modelling of initial events in biofilm accumulation, p. 445-486.
In
W. C. Characklis, and K. C. Marshall (ed.), Biofilms. Wiley Press, New York, N.Y.
|
| 19.
|
Fleischmann, M.,
S. Pons,
D. R. Rolison, and P. P. Schmidt (ed.).
1987.
Ultramicroelectrodes.
Datatech Systems Inc., Morganton, N.C.
|
| 20.
|
Fletcher, M., and G. I. Loeb.
1979.
Influence of substratum characteristics on the attachment of a marine pseudomonad to solid surfaces.
Appl. Environ. Microbiol.
37:67-72[Abstract/Free Full Text].
|
| 21.
|
Frymier, P. L.,
R. M. Ford,
H. C. Berg, and P. T. Cummings.
1995.
Three-dimensional tracking of motile bacteria near a solid planar surface.
Proc. Natl. Acad. Sci. USA
92:6195-6199[Abstract/Free Full Text].
|
| 22.
|
Greef, R.
1980.
Ellipsometry, p. 339-369.
In
R. E. White, J. O'M. Bockris, B. E. Conway, and E. Yeager (ed.), Comprehensive treatise of electrochemistry, vol. 8. Plenum Press, New York, N.Y.
|
| 23.
|
Greef, R.
1992.
ELLGRAPH software.
Optochem, Southampton, United Kingdom.
|
| 24.
|
Greef, R.
1992.
TDELPSI software.
Optochem, Southampton, United Kingdom.
|
| 25.
|
Greef, R.
1993.
FILMFIT software.
Optochem, Southampton, United Kingdom.
|
| 26.
|
Habraken, F. H. P. M.,
O. L. J. Gijzeman, and G. A. Bootsma.
1980.
Ellipsometry of clean surfaces, submonolayer and monolayer films.
Surf. Sci.
96:482-507.
|
| 27.
|
Jucker, B. A.,
H. Harms, and A. J. B. Zehnder.
1996.
Adhesion of the positively charged bacterium Stenotrophomonas (Xanthomonas) maltophilia 70401 to glass and teflon.
J. Bacteriol.
178:5472-5479[Abstract/Free Full Text].
|
| 28.
|
Kim, B. S.,
S. N. Hoier, and S. M. Park.
1995.
In situ spectro-electrochemical studies on the oxidation mechanism of brass.
Corros. Sci.
22:557-570.
|
| 29.
|
Little, B.,
P. Wagner, and F. Mansfeld.
1992.
An overview of microbiologically influenced corrosion.
Electrochim. Acta
37:2185-2194.
|
| 30.
|
Little, B. J.,
P. A. Wagner,
W. G. Characklis, and W. Lee.
1990.
Microbial corrosion, p. 635-670.
In
W. C. Characklis, and K. C. Marshall (ed.), Biofilms. Wiley Press, New York, N.Y.
|
| 31.
|
Marshall, K. C., and R. H. Cruickshank.
1973.
Cell surface hydrophobicity and the orientation of certain bacteria at interfaces.
Arch. Mikrobiol.
91:29-40[Medline].
|
| 32.
|
Marshall, K. C.,
R. Stout, and R. Mitchell.
1971.
Mechanism of the initial events in the sorption of marine bacteria to surfaces.
J. Gen. Microbiol.
68:337-348.
|
| 33.
|
Mollica, A.,
G. Ventura,
E. Traverso, and V. Scotto.
1988.
Cathodic behavior of nickel and titanium in natural seawater.
Int. Biodeterior. Bull.
24:221-230.
|
| 34.
|
Morel, A., and Y. H. Ahn.
1990.
Optical efficiency factors of free-living marine bacteria: influence of bacterioplankton upon the optical properties and particulate organic carbon in oceanic waters.
J. Mar. Res.
48:145-175.
|
| 35.
|
Muller, R. H.
1973.
Optical techniques in electrochemistry.
Adv. Electrochem. Electrochem. Eng.
9:167-226.
|
| 36.
|
Naumann, D.,
D. Helm, and H. Labischinski.
1991.
Microbiological characterization by FT-IR spectroscopy.
Nature
351:81-82[Medline].
|
| 37.
|
Nichols, P. D.,
J. M. Henson,
J. B. Guckert,
D. E. Nivens, and D. C. White.
1985.
Fourier transform-infrared spectroscopic methods for microbial ecology: analysis of bacteria, bacteria-polymer mixtures and biofilms.
J. Microbiol. Methods
4:79-94.
[Medline] |
| 38.
|
Nivens, D. E.,
J. Q. Chambers,
T. R. Anderson,
A. Tunlid,
J. Smit, and D. C. White.
1993.
Monitoring microbial adhesion and biofilm formation by attenuated total reflection/Fourier transform infrared spectroscopy.
J. Microbiol. Methods
17:199-213.
|
| 39.
|
Ord, J. L.,
D. J. De Smet, and D. J. Beckstead.
1989.
Electrochemical and optical properties of anodic oxide films on titanium.
J. Electrochem. Soc.
136:2178-2184.
|
| 40.
|
Scharifker, B. R., and J. Mostany.
1984.
Three dimensional nucleation with diffusion controlled growth.
J. Electroanal. Chem.
177:13-23.
|
| 41.
|
Shanley, C. W.,
R. E. Hummel, and E. D. Verink, Jr.
1980.
Differential reflectometrya new optical technique to study corrosion phenomena.
Corros. Sci.
20:467-480.
|
| 42.
|
Sokolowski, J.,
J. M. Lzajkowski, and M. Turowska.
1990.
Zero charge potential measurements of solid electrodes by inversion immersion methods.
Electrochim. Acta
35:1393-1398.
|
| 43.
|
Suci, P. A.,
K. J. Siedlecki,
R. J. Palmer, Jr.,
D. C. White, and G. G. Geesey.
1997.
Combined light microscopy and attenuated total reflection Fourier transform spectroscopy for integration of biofilm structure, distribution, and chemistry at solid-liquid interfaces.
Appl. Environ. Microbiol.
63:4600-4603[Abstract].
|
| 44.
|
Tadjeddine, A.
1984.
Optical-response of the ideally polarizable electrodeapplication to the quantitative determination of the interactions in the interface.
J. Electroanal. Chem.
169:129-144.
|
| 45.
|
Van Loosdrecht, M. C. M.,
J. Lyklema,
W. Norde, and A. J. B. Zehnder.
1989.
Bacterial adhesion: a physicochemical approach.
Microb. Ecol.
17:1-15.
|
| 46.
|
Vold, R. D., and M. J. Vold.
1983.
Colloid and interface chemistry.
Addison-Wesley, London, United Kingdom.
|
| 47.
|
Walthman, C.,
J. Boyle,
B. Ramey, and J. Smit.
1994.
Light scattering and adsorption caused by bacterial activity in water.
Appl. Optics
33:7536-7540.
|
| 48.
|
Weast, R. C. (ed.).
1987.
CRC handbook of chemistry and physics, 68th ed.
CRC Press, Inc., Boca Raton, Fla.
|
Applied and Environmental Microbiology, October 1998, p. 3690-3697, Vol. 64, No. 10
0099-2240/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.
This article has been cited by other articles:
-
Busalmen, J. P., de Sanchez, S. R.
(2001). Adhesion of Pseudomonas fluorescens (ATCC 17552) to Nonpolarized and Polarized Thin Films of Gold. Appl. Environ. Microbiol.
67: 3188-3194
[Abstract]
[Full Text]
Top
Abstract
Introduction
