You are on page 1of 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/26264096

Assessing bacterial adhesion using DLVO and XDLVO theories and the jet
impingement technique

Article  in  Colloids and surfaces B: Biointerfaces · June 2009


DOI: 10.1016/j.colsurfb.2009.04.030 · Source: PubMed

CITATIONS READS

134 1,205

4 authors:

Sonia Bayoudh Ali Othmane


University of Monastir University of Monastir
9 PUBLICATIONS   548 CITATIONS    100 PUBLICATIONS   1,429 CITATIONS   

SEE PROFILE SEE PROFILE

Laudy Mora Hafedh Ben Ouada


Nueva Granada Military University University of Monastir
27 PUBLICATIONS   326 CITATIONS    208 PUBLICATIONS   3,462 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

FP7, Marie Curie IRSES project PIRSES_GA_2012-318053: SMARTCANCERSENS” View project

biosourced polymer syntheses View project

All content following this page was uploaded by Sonia Bayoudh on 28 June 2018.

The user has requested enhancement of the downloaded file.


Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces


journal homepage: www.elsevier.com/locate/colsurfb

Assessing bacterial adhesion using DLVO and XDLVO theories and the jet
impingement technique
Sonia Bayoudh a,b,∗ , Ali Othmane b , Laurence Mora c , Hafedh Ben Ouada a
a
Laboratoire de Physique et Chimie des Interfaces, Faculté des Sciences de Monastir, Monastir, Tunisia
b
Laboratoire de Biophysique, Faculté de Médecine de Monastir, Monastir, Tunisia
c
INSERM, U 698, Labo. de Bio-Ingenierie de Polymères Cardiovasculaires, Institut Galilée, Université Paris 13, 99 av JB Clement 93430, Villetaneuse, France

a r t i c l e i n f o a b s t r a c t

Article history: In this study, the adhesion of two bacterial strains (Pseudomonas stutzeri PS, and Staphylococcus epider-
Received 29 October 2008 midis, SE) to the glass and the indium tin oxide (ITO)-coated glass surfaces was examined qualitatively
Received in revised form 5 April 2009 and quantitatively using the theoretical approaches and the jet impingement technique. A comparison
Accepted 21 April 2009
between the DLVO and the extended DLVO (XDLVO) theories showed that the XDLVO predictions of bac-
Available online 14 May 2009
terial adhesion and its reversibility are more accurate than DLVO predictions. The adhesion tests revealed
that PS bacteria has much better adhesion rate than SE bacteria to both material surfaces, as predicted by
Keywords:
XDLVO approach. Also both bacterial strains adhered better to the hydrophobic ITO-coated glass than to
Bacterial adhesion
Contact angle
the hydrophilic glass surface, as predicted theoretically. Moreover, the microjet impingement technique
Adhesion energy was used not only to assess the bacterial adhesion strength on both materials, but also to verify the adhe-
Jet impingement sion reversibility. The detachment stress values demonstrated that PS bacterial cells adhered strongly and
Shear stress irreversibly in the primary energy minimum, while SE bacterial cells adhered weakly and reversibly in
the secondary energy minimum on both substrata surfaces. Also, the adhesion of both bacterial strains
was found better and stronger on the hydrophobic ITO-coated glass surface than on the hydrophilic glass
surface.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction interactions are involved, and progress to a more strongly attached


irreversible state which is governed by both non specific and spe-
During the past decade, understanding and controlling bacte- cific interactions [5]. The non specific interaction energies (forces)
rial adhesion to material surfaces are the primary interests for that govern the initial phase of the bacterial adhesion mechanism
many researchers in various disciplines: biomedical, environmen- are basically the attractive Lifshitz van der Waals (LW) and the
tal and industrial. Microbial adhesion may lead to the formation electrostatic double layer (EL) interactions, which can be either
of an infectious biofilm that may cause infection on biomaterials attractive or repulsive depending on the surface charge. These inter-
and implanted medical devices, contamination of water resources action energies are well understood and described generally in the
and biofouling in food-processing equipment and in many engi- classical DLVO (Derjaguin, Landau, Verwey, Overbeek) theory of
neered and marine systems. Microbial adhesion is the process of colloid stability [6,7]. The DLVO theory has been widely used as
bringing a bacterial cell or cell aggregates from an unbound state a theoretical model not only qualitatively but also quantitatively to
in the bulk liquid to a more or less firm attached state at a substra- calculate the actual adhesion energy variations involved in bacterial
tum/bacterium interface. Initial microbial adhesion to surfaces is (or colloidal) adhesion and aggregation as a function of separation
a complicated process that is affected by various physicochemical distance between the interacting surfaces [8–12].
properties of both bacteria and substratum surfaces, most notably However, in the classical DLVO theory, both the substratum
their hydrophobicity/hydrophilicity and surface charge [1–4]. and the colloidal particle surfaces are assumed to be chemically
The standard model for bacterial adhesion implies that bacteria inert. This is not valid for the bacterium and substratum surfaces
start from a weakly attached reversible state, where non specific where hydrogen and chemical bonds are involved in the adhesion
mechanism. Van Oss et al. suggested an additional term called the
short-range Lewis acid–base (AB) interactions to account for hydro-
gen bonding on close approach of bacteria and substrate surfaces,
∗ Corresponding author at: Laboratoire de Physique et Chimie des Interfaces, Fac-
in an extended XDLVO theory [13,14]. These AB interactions are
ulté des Sciences de Monastir, Monastir, Tunisia. Tel.: +216 7350 0279;
accounted for, in addition to LW and EL interactions, to explain
fax: +216 7350 0278.
E-mail address: soniab2j@yahoo.fr (S. Bayoudh). the discrepancies between the DLVO predictions and experimen-

0927-7765/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.colsurfb.2009.04.030
2 S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9

tal observations [15,16]. The AB interactions are based on electron hydrophobic indium tin oxide (ITO)-coated glass surfaces, using the
acceptor/electron donor interactions between polar moieties in DLVO theory and the theoretical approach of XDLVO theory. The
polar media (e.g., water). In addition, depending on the hydropho- DLVO and XDLVO theories were used to qualitatively and quantita-
bic/hydrophilic property of both microbial cells and substrate tively predict the non specific interaction energies between bacteria
surfaces, these polar interactions could be attractive (hydrophobic and substratum surfaces at long and short distances. In addition, the
attraction) or repulsive (hydrophilic repulsion or hydration effects), jet impingement technique was used to reveal the reversibility of
and may be up to 10–100 orders of magnitude greater than EL and bacterial adhesion based on the values of the detachment strength.
LW interactions [17]. The addition of the polar interactions has
resulted in an extended DLVO approach (XDLVO) for quantifying 2. Extended DLVO theory
the interaction energy in order to predict the adhesion [14–16].
It was claimed that the XDLVO approach may be the promising As described in the classical DLVO theory, the net interaction
model to explain the experimental results of bacterial adhesion energy (GDLVO ) needed to bring a bacterium (b) into contact with a
since it combines both the thermodynamic approach and DLVO flat substratum surface (s) immersed in aqueous medium (l) is the
theory [18]. However, the validation of this approach as a predic- balance between two additive interaction energies: the attractive
tive physicochemical model to study the bacterial adhesion is still Lifshitz van der Waals energy (GLW ) and the repulsive or attrac-
under investigation. tive electrostatic double layer energy (GEL ). The total interaction
Comparison between the DLVO and XDLVO predictions has or adhesion energy as a function of the separation distance (d)
been previously investigated by many researchers. Brant and between a bacterium (sphere) and a substratum (flat plane) sur-
Childress have investigated the interaction energies for different faces is therefore can be written as:
membrane–colloid combinations [19]. They found that the XDLVO
DLVO LW EL
approach predicts considerably different short-range (separation Gslb (d) = Gslb (d) + Gslb (d) (1)
distance <10 nm) interaction energies when compared with DLVO
Later, Van Oss et al. [24] suggested that the acid–base energy
predictions, particularly for hydrophilic membrane–colloid combi-
(GAB ), arising from hydrogen bonding between two surfaces
nation. The hydrophilic repulsion resulted in much larger energy
immersed in a polar solvent (e.g., water), must be accounted in
barrier at short-range, while the hydrophobic attraction resulted
addition to LW and EL interaction energies. The inclusion of the
in much attractive energy profile. Meinders et al. investigated the
polar interaction energy (GAB ) resulted in the XDLVO approach, in
deposition and reversibility of bacterial adhesion on various sub-
which the total interaction energy (GXDLVO ) may be written as:
strate surfaces, and they have found that XDLVO model explain
XDLVO LW EL AB
more accurately the bacterial adhesion than the DLVO theory for Gslb (d) = Gslb (d) + Gslb (d) + Gslb (d) (2)
a hydrophobic substratum surface [9]. Azeredo et al. showed, on
one hand, that the adhesion in phosphate buffer saline of bacte- From a thermodynamic point of view, adhesion or attraction
rial mutants to glass is mainly explained by the DLVO theory [20]. between two interacting surfaces occurs when the total energy
On the other hand, they found that the XDLVO theory enabled the GXDLVO is negative, and repulsion occurs when GXDLVO is positive.
interpretation of the adhesion of some of these mutants to glass Since the total interaction energy is evaluated as a function of the
in presence of the exopolymers, where hydrophobic interactions separation distance (d) between the interacting surfaces, there-
played an important role in the irreversible adhesion. fore the interaction energy profile illustrate the type of interaction
In addition, the XDLVO approach prediction has been (attraction or repulsion) as the microbial particle approaches a sub-
tested in comparison with other approaches and methods. For strate surface.
instance, Brant and Childress demonstrated that for hydrophilic
membrane–colloid pairs, the AFM force measurement curves are 2.1. Surface thermodynamics and hydrophobicity/hydrophilicity
similar to the interaction energy profiles predicted by the XDLVO
approach [21]. Also the XDLVO approach predicts substantially Application of the XDLVO approach requires that the surface
repulsive short-range interaction energies when compared to the energies of microbial particle and solid surfaces have to be deter-
attractive DLVO interaction energies. However, for hydrophobic mined. The surface energy of the solid and bacterial surfaces can be
membrane–colloid pairs where the XDLVO approach predicts an obtained through the contact angle measurements and using the
interaction energy profile similar to that predicted by the DLVO Lifshitz van der Waals acid–base approach. Van Oss et al. [24] sug-
theory, the measured force curves agree with the interaction energy gested that the total surface energy () of a pure substance is the
profiles observed in both DLVO and XDLVO predictions. Sharma and sum of a dispersive ( LW ) and polar ( AB ) components, yielding:
Hanumantha Rao [22] have compared the predictions of XDLVO
 =  LW +  AB (3)
and the surface thermodynamics approaches, when studying the
bacterial adhesion to minerals for different physicochemical con- The polar AB component comprises two non-additive electron
ditions (ionic strength and pH). They showed that in general acceptor ( + ) and electron donor ( − ) parameters, and is given by:
the XDLVO approach explains more effectively the adhesion of 
bacteria–bacteria, mineral–mineral or mineral–bacteria systems  AB = 2  + − (4)
than the thermodynamic approach, which does not account for
The surface energy components of a solid surface (sLW , s+ , s− )
the electrostatic interactions. In a recently published work, Kang
and bacterial surface (bLW , b+ , b− ) can be determined by measuring
and Choi reported on the deposition/detachment and reversibility
the contact angles [25] using three different probe liquids, with
of bacterial adhesion on various ionomers of different hydropho-
known surface tension parameters (lLW , l+ , l− ) and employing
bicity [23]. When comparing the XDLVO theory predictions with
the extended Young’s equation [26,27]:
the results of bacterial detachment experiments using high flow
   
velocity, they found that the adhesion is reversible in the secondary
minima on less hydrophobic substrates and irreversible in the pri- l (1 + cos ) = 2 sLW lLW + s+ l− + s− l+ (5)
mary minima on more hydrophobic substrates.
The aim of the present paper is to thoroughly examine the adhe- where  is the measured contact angle. Using the acid–base
sion behavior of two bacterial strains (Pseudomonas stutzeri PS, approach and from the calculated surface energies of solid and bac-
and Staphylococcus epidermidis, SE) to the hydrophilic glass and terial surfaces, the cohesive and adhesive free energies per unit area
S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9 3

can be determined. The free energy of cohesion (Gsls ) is the inter- area between two infinite planar surfaces brought into close contact
action free energy per unit area when two identical particles or with each other (and to a distance h) must be used. According to
surfaces (s) are immersed in polar liquid (l, such as water as a model Derjaguin approximation, the interaction energy is calculated as
solvent) and brought into contact. The free energy of cohesion is the sum integral of the corresponding plate–plate free energy per
proportional to the interfacial tension ( sl ) between two phases, unit area along the separation distance, d, and the actual LW, AB
solid and liquid, and are related by [28,29]: and EL interaction energies between a sphere (bacteria) and a flat
  2 surface immersed in a liquid are expressed as [31–33]:
Gsls = −2sl = −2 sLW − lLW LW Aa
Gslb (d) = − (10)
6d
    
d − d
−4 s+ s− + l+ l− − s+ l− − s− l+ (6) AB
Gslb (d) = 2a GdAB exp
0
(11)
0

The first term in this equation represents the LW component and   


the second term represents the AB component. Also, one can note EL 1 + e−d
Gslb (d) = εr ε0 a 2 b s ln
that the negative terms indicate attraction, while positive terms sig- 1 − e−d
nify repulsion between two identical surfaces immersed in water.

2 2

−2d

Based on Van Oss [28,29], the hydrophobicity or hydrophilicity of a + b
+ s ln 1 − e (12)
substrate and bacteria surfaces can be evaluated using Gsws (when
water is the suspending liquid). Positive values of the cohesive
where A is the un-retarded sphere (bacterium)–substrate Hamaker
energy indicate hydrophilic surfaces, while negative values denote
constant in water, (=0.6 nm) is the characteristic decay length
hydrophobic surfaces.
of AB interactions in water, and a is the sphere radius. The total
Similarly, surface tensions for bacterial and substrate surfaces
XDLVO adhesion energy is therefore the sum of the three compo-
calculated using the acid–base approach can be used to evaluate
nents at a given small separation distance. As it can be concluded
the free energies of adhesion per unit area. According to Van Oss,
from Eqs. (10)–(12), as the separation distance between the two
the expressions for the LW, AB and EL adhesion free energies per
surfaces of a sphere and flat plane increases, the adhesion energy
unit area, GdLW , GdAB and GdEL are given by Eqs. (7), (8) and (9)
0 0 0 components decrease rapidly from their corresponding adhesion
respectively, based on plate–plate interactions [26]: energy at contact. The LW component decays with the inverse of
      the distance, while the AB and EL components decay exponentially
GdLW = −2 sLW LW
w bLW − LW
w (7) with the separation distance.
0

     3. Materials and methods


GdAB = 2 l+ b− + s− − l−
0
3.1. Bacterial strain culture and conditions
      
+ l− b+ + s+ − l+ − b− s+ − b+ s− Two bacterial strains used in this study were P. stutzeri (PS) and S.
epidermidis (SE). PS strain was collected from surgical wards (hôpi-
(8) tal Sahloul, Sousse, Tunisie) and SE strain from catheters used in
haemodialysis wards. These strains were identified by a standard-
ised and miniaturised biochemical test system strips (API 20.NE and

ε0 εr 
API ID 32 Staph, bioMérieux, France). These bacterial strains used
GdEL (h) = 2
b
+ 2
s 1 − coth(d0 ) in this study are known to cause nosocomial infections through
0 2
medical instruments, particularly catheters, in immunosupressed
patients with respiratory tract infections.
2 PS was cultured overnight on a nutrient agar plate (Bio-RAD,
+
2
b s
2
csch(d0 ) (9)
France), while SE was cultured for 24–48 h on a Tryptone soya
b
+ s
caseine agar plate (Pronadisa, Spain). For each strain, the plate was
where ε0 (=8.854 × 10−12 C V−1 m−1 ) dielectric permittivity in stored at 4 ◦ C for a maximum duration of 4 weeks. Several colonies
the vacuum, εr (=79) is the relative permittivity of water,  from each strain were streaked to grow a preculture for 24 h at 37 ◦ C.
(=3.28 × 109 I1/2 m−1 , where I is the ionic strength of the electrolyte A second culture was inoculated from the preculture at 37 ◦ C for
in terms of molarity) is the inverse Debye screening length, and b 24 h. The bacterial cells were harvested from the second culture in
and s are the surface potentials of the bacterium and the substrate. the stationary growth phase and then were suspended in a sterile
The total free energy of adhesion is the sum of the LW, AB and EL phosphate buffer saline PBS of pH 7.0 and ionic strength 150 mM
components and it implies the interaction energy between two flat (g l−1 : KH2 PO4 , 0.68; K2 HPO4 , 0.86; NaCl, 8.76). For adhesion test,
surfaces physically touching each other at a minimum equilibrium the bacterial cells were resuspended in PBS at a concentration of
cut-off distance d0 (=0.158 nm). The minimum separation distance 4 × 107 cells ml−1 .
may be considered as the distance between the outer electron shells
(van der Waals boundaries) of adjoining non-covalently interacting 3.2. Substrate surface preparation
molecules [26,30].
Glass microscope coverslips (24 mm × 24 mm, Deckglaser, Ger-
2.2. Calculation of the XDLVO adhesion energy using the many) were used as a hydrophilic substratum surfaces. The glass
Derjaguin approximation surfaces were cleaned in chromium sulphuric acid for 24 h at 4 ◦ C
then rinsed thoroughly with distilled water. The clean coverslips
In order to compute the total XDLVO interaction energy between were stored in Millipore-Q water to keep the surface hydrophilic.
the surfaces of a sphere and a flat plane, the free energy per unit Prior to each experiment, the slides were dried under a nitrogen
4 S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9

flow and sterilized at 120 ◦ C for 2 h. The indium tin oxide (ITO)- field of view to obtain the density of bacteria per cm2 . All adhe-
coated glass plates (10 mm × 20 mm, Merck, Germany) were used sion experiments were performed in triplicate with suspensions
as hydrophobic substratum. ITO films had a sheet resistance per of separate bacterial suspensions and newly prepared substrate
surface unit of 30
/cm2 . The ITO-coated glass plates were ultra- samples.
sonically cleaned with pure acetone followed by methanol for
20 min each. The clean slides were then washed thoroughly with
3.5. Jet impingement experiments
Millipore-Q water and dried under a nitrogen flow. ITO-coated glass
slides were then sterilized using autoclave at 120 ◦ C and 1.5 bars for
The jet system was described in previous work [36]. It consisted
20 min.
of a peristaltic pump (Gilson Minipuls 3, France) with a regulated
flow rate and a nozzle with an internal diameter of 0.15 mm. The
3.3. Zeta potential measurements
fluid flowing through the system was PBS at room temperature
(25 ◦ C). In order to get a layer of adhered bacterial cells onto the
The measurements of the zeta potentials ( ) and the hydro-
substrate surface, the bacterial suspensions were formed to be con-
dynamic diameters (ac ) of bacterial cells were performed using a
centrated by suspending bacterial cells in PBS. The optical density
Zetasier Nano ZS electrophoretic light scattering spectrophotome-
of bacterial suspensions were optimised and adjusted close to 0.38
ter (Malvern Instruments, UK). The bacterial cells were suspended
and 0.76 at 600 nm of wavelength for PS and SE bacteria, respec-
in PBS at a concentration of 107 –108 cells ml−1 . The electrophoretic
tively. The bacteria cells were then allowed to static adhere to the
mobilities were measured in triplicate at an applied voltage of 150 V
substratum surface for 2 h at room temperature. The substrate sam-
and at temperature of 22 ◦ C, then were converted to zeta potentials
ple was then carefully washed three times with PBS to remove the
using the Helmoltz Von Smoluchowski relation [34]:
loosely adhered bacteria.
εε0 The jet impingement test was conducted as follows: the rele-
e = (13)
vant material sample, which was covered with a bacterial layer, was
At high a, surface potential ( ) could be greater than zeta transferred to a Petri dish containing PBS. The Petri dish was placed
potential, so it can be calculated by Debye–Huckel approximation on the inverted microscope stage and positioned directly below the
with the slipping distance (z) of 0.3 nm [35]: jet needle. The end of the needle was positioned perpendicular to
 z
 the material surface and the height was adjusted to 0.3 mm (that
 = 1+ ez (14) is four times the internal radius of the needle). A submerged lami-
a nar jet was then directed at the bacterial layer’s surface to create a
lesion. In this study, we used the values of Reynolds number 500 and
3.4. Flow chamber and adhesion experiments
1500 at exposure times of 10 and 30 s. The choice of the exposure
times was based on the findings of Bundy et al. [37].
Adhesion experiments were carried out in parallel plate
A digital image of the lesion was taken with a Canon camera
flow chamber (internal dimensions: l × w × h = 34 mm ×
that was mounted on the phase contrast inverted microscope. An
16 mm × 0.5 mm) and at room temperature. The first plate of the
image analysis and measurement software program (ImageJ, 1.32)
chamber (65 mm × 43 mm) is made of thin glass coverslip (Gold
was used to enhance the lesion image if necessary and to measure
Scal, 3335). The second plate of the chamber is made of biocompat-
the lesion radius. The program was used to determine the number of
ible Plexiglas (dimensions: l × w × h = 67 mm × 42 mm × 12 mm),
pixels comprising the lesion. Each processed image was calibrated
on which two holes in and out for fluid circulation were created on
to the dimensions of a bar scale so that pixels could be converted to
two sides. A third hole was created for air vacuum between two
square millimetres. The lesion was considered to be circular, so its
plates which were separated by two silicon membranes (thickness
radius was measured. The non-dimensional radial distance at the
of each is 0.25 mm). A rectangular groove (20 mm × 10 mm) was
lesion border can be then determined [36]. The non-dimensional
created centrally on the Plexiglas plate to affix the ITO plate. The
wall shear stress at the lesion border could be determined directly
depth of the groove was adapted to the thickness of the ITO plate
from the calibration curves provided by Deshpande and Vaishnav
in such a way that the ITO surface was at the same height as the
[38].
surface of the Plexiglas plate.
A peristaltic pump (Gilson Minipuls 3, France) with a regulated
flow rate was used for fluid circulation through the chamber under 4. Results and discussion
the influence of hydrostatic pressure. Prior to each experiment, all
tubes and the flow chamber were filled with PBS, taking care to 4.1. DLVO and XDLVO predictions at contact and adhesion
remove all air bubbles from the flow system. The PBS solution was experiments
first perfused through the flow system for 15 min. Then a bacterial
suspension of 4 × 107 cells ml−1 was allowed to flow through the Even though microbial adhesion does not depend on only a
system for 2 h at a shear rate of almost 6 s−1 , yielding to a lami- single parameter, much research has focused on the importance
nar flow. For bacterial deposition and enumeration of the adhering of hydrophobicity and hydrophilicity of surfaces. The hydropho-
cells, the flow chamber was mounted on the stage of an inverted bicity/hydrophilicity of the cell or the substrate surface can be
microscope (Zeiss, Axiovert 25, Germany) equipped with a 32× characterized by either the water contact angle or by the cohe-
objective (Zeiss, Germany) that was connected to a digital cam- sion free energy per unit area (Gsws ) of two identical bacterial
era (Canon G5, EU). Deposition images were taken at different time cells or solid surfaces immersed in water at contact. The sur-
intervals from the glass plate of the flow chamber, and were then face tensions components of the substrate and bacterial surfaces
analyzed later with an image analysis and measurement software were already published in previous work [25]. Based on the values
program (ImageJ, 1.32). This latter was used also to calibrate the of Gsws illustrated in Table 1, the positive sign of the cohe-
size of the field of view of each image to a bar scale so that pix- sion free energy indicates that both bacterial strain surfaces are
els could be converted to square millimeters. For each instant, the hydrophilic. Comparing the hydrophilicity of both strains, PS sur-
bacteria cells were counted for at least three fields of view for each face was found less hydrophilic than SE surface (16.3 mJ/m2 for PS is
image. The average value of the number of bacteria cells of dif- less than 30.4 mJ/m2 for SE). Moreover, when comparing the sub-
ferent fields of view was then divided by the surface area of the strate hydrophobicity/hydrophilicity surfaces, our results showed
S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9 5

Table 1 Eq. (14). The values of the surface potentials of the substrata ( s ,)
The values of the free energy of cohesion between two identical surfaces immersed in
were considered as those of the zeta potentials, which were taken
water, Gsws (in mJ/m2 ), or in PBS, Gsls (in mJ/m2 ), and the values of zeta potential
(in mV) of bacteria and substrates, the surface potentials (in mV) and the radius from Refs. [40,41].
a (in ␮m) of the bacterial cells. The DLVO predictions show that the adhesion is favourable
for all bacteria–substrate combinations. The negative values of
Bacteria/materials Gsws a Gsls a a
the free energy imply a weak attraction ranging from −1.8 to
PS 16.3 20.9 −40.4 −59.1 0.55 −4 mJ/m2 , which means that the adhesion of both bacterial strains
SE 30.4 31.5 −23.7 −34.7 0.92
is favourable on both substrates. Comparing the free energy val-
Glass 27.4 30.4 −22b – –
ITO −58.6 −34.8 −12c – – ues for each bacteria–substrate combination, PS adhesion on ITO
a
surface is predicted better on glass surface. Also, SE adhesion is pre-
The surface tensions of suspending media and of bacteria and materials surfaces
were used from Ref. [25].
dicted slightly better on ITO surface, since the free energy values are
b
Value of zeta potential was taken from Ref. [41]. close. For both materials, the adhesion of PS bacteria is better than
c
Value of zeta potential was taken from Ref. [40]. the adhesion of SE bacteria.
However, for the XDLVO approach, the inclusion of AB interac-
tions results in remarkably different prediction from that of DLVO.
that the glass surface is hydrophilic (27.4 mJ/m2 ) while the ITO glass
Due to the hydrophilic nature of bacterial cell and glass surfaces
surface is strongly hydrophobic (−58.6 mJ/m2 ).
(positive values of Gsws ), XDLVO approach predicts hydrophilic
In order to investigate on the influence of the suspending XDLVO =
repulsion at contact between the interacting surfaces (Gadh
medium (PBS) on the hydrophobicity/hydrophilicity of bacterial 2
24.7 and 32.1 mJ/m for PS and SE cells, respectively). For ITO
cell and substrate surfaces, the cohesion free energy per unit area
substrate though, and because of its hydrophobic surface, XDLVO
(Gsls , l indicates the PBS liquid) was evaluated (see Table 1). The
prediction shows strong attractive interaction with PS cell surface
obtained values of cohesion free energy of two identical surfaces
because of the negative free energy of adhesion. While for SE cell
immersed in PBS has increased compared to the values of cohesion
surface, and because it is more hydrophilic than PS surface, the
free energy of two identical surfaces immersed in water. This indi-
free adhesion at contact predicted by XDLVO approach is posi-
cates that bacterial and glass surfaces became more hydrophilic,
tive (14.6 mJ/m2 ) implicating repulsive interaction. Hence, bacterial
while the hydrophobicity of ITO glass surface has remarkably
adhesion is predicted energetically unfavourable on the glass sur-
decreased. Also, one can note that SE cell surface was not as much
face, while the adhesion on ITO surface is favourable for PS bacteria
affected by the suspending medium as PS cell surface. This is due
and unfavourable for SE bacteria.
to the fact that microbial surface thermodynamics is mainly deter-
In order to verify the theoretical predictions at contact, adhesion
mined by the outer surface of the bacterial membrane [16]. In fact,
experiments are conducted using a parallel plate flow chamber.
on one hand, SE strain is a Gram-positive bacteria and its outer
Our results revealed that despite the positive values of the adhe-
surface of the membrane is high in peptidoglycan. This makes it
sion free energies predicted by XDLVO approach for SE bacteria on
resistive to the changes in physicochemical properties (such as pH
both materials and for PS bacteria on glass, these bacterial cells
and ionic strength) of the suspending medium and therefore its sur-
showed their ability to attach to the material surfaces. Fig. 1 illus-
face hydrophobicity/hydrophilicity is relatively more stable. On the
trates the deposition and adhesion kinetics of PS (a) and SE (b)
other hand PS strain is Gram-negative bacteria and its outer sur-
bacteria on glass and ITO surfaces, expressed as the number of
face membrane is high in lipid contents and low in peptidoglycan
adhering/deposited bacterial cells per unit area n(t) recorded as
contents, which makes it easily influenced by the changes of the
a function of time by image sequence analysis during 2 h. For all
suspending medium [39].
bacteria–substrate combinations, the variation of n(t) in time is
Knowing the hydrophobicity/hydrophilicity parameter of the
fitted as an exponential rise in time and expressed as
interacting surfaces (bacteria and substrate) may only provide some
qualitative insight into potential interactions between bacteria and
n(t) = n∞ (1 − e−t/ ) (15)
solid surfaces. However, a quantitative assessment could be pro-
vided through the quantification of the free energy of adhesion per
where n∞ is the number of bacteria adhering per unit area over
unit area at contact. Table 2 compares DLVO and XDLVO predictions
the full duration of an experiment and  is the so-called charac-
of the free energy of adhesion at contact for each bacteria–substrate
teristic adhesion time. The initial deposition rate j0 represents the
combinations. The values of DLVO free energy of adhesion were
initial increase of n(t) with time and it was calculated directly by
calculated as the sum of LW and EL components at the minimum
linear regression analysis from the initial increase of the numbers
separation distance. Also, the values of XDLVO free energy of adhe-
of adhering bacteria [42,43].
sion were calculated as the sum of LW, AB and EL components. The
Table 2 summarises the fitting results of the initial deposition
values of LW and AB components were taken from our previous
rate j0 expressed in number of adhering cells per unit area (cm2 )
work in Ref. [25]. The values of EL component were calculated using
per second, n∞ expressed in number of adhering bacteria (×105 )
Eq. (9). In this equation, the values of the surface potentials of the
and the adhesion time  expressed in minutes. The initial depo-
bacterial cells ( b are indicated in Table 1) were deduced from the
sition rate represents the initial rate of arrival of bacterial cells at
measured values of zeta potentials and bacteria cell radius using
the ITO electrode surface per unit area and time, while the charac-
teristic deposition time represents the rapidity of bacterial cells to
Table 2 adhere to glass and ITO surfaces [43]. For PS bacteria, the values of
The values of the DLVO and XDLVO free energies of adhesion at contact for each
bacteria–substrate pair (kT), the fitting values of the initial deposition rate j0
j0 were found 227 and 407 cells/cm2 /s for glass and ITO surfaces,
expressed in number of adhering cells per unit area (cm2 ) per second, n∞ expressed respectively. For SE bacteria however, the values of j0 were found
in number of adhering bacteria (×105 ) and the adhesion time  expressed in minutes. 194 and 276 cells/cm2 /s for glass and ITO surfaces, respectively.
DLVO XDLVO Comparing these results, PS adhesion was initially better and
Bacteria Substrate Gadh Gadh j0  n∞
faster than SE adhesion on both substrata surfaces, as predicted
PS Glass −2.9 24.7 227 47 12
in DLVO and XDLVO approaches. Also, when comparing the fit-
ITO −4.0 −6.5 407 38 17
ting values of  and n∞ for each bacteria–substrate combination,
SE Glass −1.8 32.1 194 64 8 our results revealed that PS bacterial strain adhered much better
ITO −2.0 14.6 276 43 10
(larger values of n∞ ) and more rapidly (smaller values of ) than SE
6 S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9

surfaces. Our results showed as well that the adhesion of both bac-
terial strains was better on the hydrophobic ITO surface than on the
hydrophilic glass surface, which is in agreement with DLVO predic-
tions. In addition, when examining Fig. 1a and b, the deposition rate
(dn/dt) decreased over deposition time and became not important
as the initial rate j0 . This decrease can be explained by the fact that
since the repulsive hydrophilic interactions increase due to the ionic
strength of the medium, as discussed above, hence the adhesion
rate should decrease.
In conclusion, the DLVO predictions at close contact are found
more accurate than the XDLVO predictions. However, regardless of
the sign of the value of free energy of adhesion, the values of j0 were
well correlated in linear relationship with the values of the XDLVO
free energy of adhesion, as indicated in Fig. 2. This plot demon-
strates that the lower values of free energy of adhesion correspond
to the higher initial adhesion rates, which is in agreement with
thermodynamic laws.

4.2. Comparison between DLVO and XDLVO interaction energy


profiles

The discrepancy between the XDLVO predictions at contact and


the adhesion experiments findings could be explained by the fact
that the adhesion process should not be looked only at contact, but
in fact it starts from a distance between the interacting surfaces.
Therefore, the adhesion energy should be plotted as a function of
the separating distance. Figs. 3 and 4 show the DLVO and XDLVO
interaction energy profiles as a function of the separation distance
between the bacterial cell and substratum surface for PS and SE
bacteria, respectively. Eqs. (10), (11) and (12) were used to calcu-
late the LW, AB, and EL components of the interaction energies, for
each bacteria–substratum combination. By adding these compo-
nents, DLVO and XDLVO energy profiles were obtained according
to Eq. (1) and Eq. (2), respectively.
As it can be noted from those figures, the DLVO and XDLVO
Fig. 1. Comparison of the adhesion kinetics of PS (a) and SE (b) bacteria on the predictions are similar at large separation distances (>5 nm). At
hydrophilic glass and the hydrophobic ITO glass surfaces during 2 h of deposition large distance, the bacteria cells are attracted towards the substrate
time. surface by mainly the LW attraction. In Figs. 3a and 4 (correspond-
ing to PS–glass, SE–glass and SE–ITO combinations), the DLVO and
bacteria on both substrata surfaces over the full deposition period. XDLVO theories predict a shallow attractive secondary minimum
Besides, during deposition time, PS bacteria were observed under prior to entering a strongly repulsive region. The DLVO and XDLVO
microscope to be more mobile and energetic than SE bacteria cells, energy profiles are replotted and inserted in the bottom-right in
when approaching the substratum surface. This movement is more Figs. 3a and 4 to emphasize the magnitude of this secondary energy
energetic than the Brownian motion and hydrophilic interactions, minimum. The values of the secondary minimum magnitude and
therefore promoting the adhesion of PS bacteria to the substrata position are presented in Table 3, for all bacteria–substratum com-
binations. The secondary minimum depths predicted by DLVO
theory are slightly larger than those predicted by XDLVO approach,
whereas their positions are slightly lower. Because the thermal
energy is on the order of 1 kT, the values of the secondary energy
minimum calculated by DLVO and XDLVO approaches would be
sufficient to attach bacterial cells to the surface reversibly [11].
Upon close approach, deviation between DLVO and XDLVO
energy profiles appears. After experiencing the attraction in the
secondary energy minimum, bacteria cells encounter a substan-
tial repulsive energy barrier, as indicated in the inserted graphs
in Figs. 3a and 4 (inset). The magnitude values of the DLVO and
XDLVO energy barrier are shown in Table 3. The energy barrier
in XDLVO energy profiles is considerably higher than the DLVO
energy barrier. The repulsion in DLVO predictions is due to the
electrostatic interaction, whereas the repulsion in XDLVO predic-
tions is owing to the electrostatic and the hydrophilic repulsions.
The inclusion of AB interactions resulted in significantly different
predictions of XDLVO theory from that of DLVO, at short-range dis-
tances (<5 nm). Here, the hydrophilic repulsion resulted in much
larger energy barrier. The same results have been found by other
Fig. 2. The initial deposition rate as a function of the free energy of adhesion. researchers [19–21]. In fact, AB interactions are commonly thought
S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9 7

Fig. 4. Comparison between the DLVO and XDLVO interaction energy profiles
Fig. 3. Comparison between the DLVO and XDLVO interaction energy profiles
between SE bacteria cells and glass (a) or ITO-coated glass (b) surfaces as a function
between PS bacteria cells and glass (a) or ITO-coated glass (b) surfaces as a function
of the separation distance.
of the separating distance.

to be as short-range interactions since they arise from the electron barrier magnitudes (see Table 2). For PS-ITO pair, the XDLVO
donor/electron acceptor interactions [15,27]. approach predicts attraction at long and short distances (Fig. 3b).
Bacteria cells that are able to overcome this energy barrier can Therefore, PS bacteria cells are predicted to adhere irreversibly to
fall into a deep primary energy minimum at close contact and ITO surface. The strong attraction predicted by XDLVO approach is
adhere irreversibly. The primary minimum could not be shown in due to the strong hydrophobic attraction, which in turn is due to
these figures. However, if bacteria cells could not escape the energy the strong hydrophobic nature of ITO-coated glass surface. How-
barrier, they would remain associated with the substrate surface ever, the DLVO theory predicts a reversible or irreversible adhesion
within the secondary energy minimum. In this case, the adhesion of to ITO surface depending on the ability of PS bacteria to overcome
bacteria cells is reversible and very likely that some cells could leave the smaller energy barrier (133 kT, as in Table 3). Here also, DLVO
the substratum surface under any variation of the conditions (such prediction is not certain about PS adhesion reversibility to ITO sur-
as fluid flow, cell motility, solution chemistry). The XDLVO predic- face.
tions demonstrate that because of the large energy barriers, it is In addition, the comparison of the DLVO and XDLVO energy pro-
unlikely that the bacterial cells will deposit in the primary energy files shows that PS and SE adhesion on ITO surface is better than on
minimum at the substratum surface. Consequently, the adhesion glass surface which is proved by the kinetic adhesion tests. Com-
of SE bacteria cells to both substrata surfaces would be reversible paring also the XDLVO energy profiles for both bacterial cells, the
in the secondary minimum. The DLVO theory predicts though a adhesion of PS bacteria on both substrate surfaces is predicted to
reversible adhesion of SE bacterial cells on glass surface and a be much better than the adhesion of SE bacteria cells. This predic-
reversible or a possible irreversible adhesion to ITO surface depend- tion is also confirmed by the results of the adhesion tests, where
ing on the ability of these cells to overcome the smaller energy the values of j0 and n∞ for PS bacteria are larger than those for SE
barrier magnitude (148 kT). Here, the reversibility of SE adhesion bacteria. However, the comparison of DLVO energy profiles demon-
on ITO surface is not as certain as predicted by XDLVO theory. strates that PS adhesion is lightly better than SE adhesion on both
For PS bacteria, the DLVO and XDLVO approaches predict a substrata. Therefore, XDLVO approach predicts bacterial adhesion
reversible adhesion on the glass surface because of the large energy more accurately than DLVO theory.
8 S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9

Table 3
The values of the energy barriers (kT), the secondary minimum depth (kT) and the secondary minimum separation distance (nm) as predicted by DLVO and XDLVO theories
for PS–glass, PS-ITO, SE–glass and SE–ITO pairs.

Bacteria–substratum Energy Secondary minimum Secondary minimum


barrier (kT) depth (kT) separation (nm)

PS–glass DLVO 459 −4.9 6


XDLVO 16400 −4.4 6.4

PS-ITO DLVO 133 −4.2 5.4


XDLVO – −4.4 5.1

SE–glass DLVO 485 −13 5


XDLVO 34330 −10.9 6.5

SE–ITO DLVO 148 −11.5 4.7


XDLVO 15750 −9.4 6

4.3. Comparison between the adhesion strength of both bacterial bacteria only the shear stress represented the detachment strength
strains and theoretical predictions [36]. Obviously, Fig. 5 demonstrates that PS bacterial cells need
more strength to be detached than SE bacterial cells from both sub-
In order to investigate the adhesion strength and reversibility of strata. Therefore, PS bacteria adhere more strongly than SE bacteria
PS and SE bacteria cells on glass and ITO-coated glass surfaces, we on both substrata. The PS detachment stress was found to be 7–15
used the jet impingement technique to detach the adhered bacteria times higher than SE detachment stress for the glass surface, and
from the substratum surface. The fluid jet applies shear and normal 6–20 times higher for ITO-coated glass surface.
forces to the bacterial layer to cause detachment. These findings demonstrate that the stronger adhesion strength
Fig. 5 compares the detachment strength of both bacterial (i.e., higher detachment strength) of PS bacterial strain implies
strains from the glass and ITO-coated glass surfaces for the that this bacterial strain adheres irreversibly on these surfaces.
four Reynolds/exposure time combinations. For PS bacteria cells Accordingly, this is in agreement with XDLVO predictions on the
detached at the lesion periphery, the mixed stresses (shear and nor- reversibility of PS adhesion to ITO surface. Therefore, as found by
mal) were responsible of the detachment strength, whereas for SE Meinders et al. and Azeredo et al. [9,20], XDLVO approach explain
more accurately the bacterial adhesion than the DLVO theory for
a hydrophobic substratum surface. But for the reversibility of PS
adhesion to glass surface, this result is in contradiction to theoret-
ical predictions because these bacterial cells are able to escape the
energy barrier using their polymeric surface structures (or adhesion
molecules), and also because they are more motile and energetic.
Also, since the adhesion strength of SE bacterial cells is much
weaker than that of PS bacteria, hence its adhesion on both surfaces
is reversible. These findings are in good agreement with XDLVO
predictions. As a consequence, it can be concluded that the adhesion
of SE bacteria was not enforced by the specific interactions, while
PS bacteria had adhered more strongly on both substrata which
mean that specific interactions are involved in the second step of
adhesion process.

5. Conclusion

In this paper, the DLVO and XDLVO theories were used to qual-
itatively and quantitatively predict the non specific interaction
energies between bacteria and substratum surfaces at long and
short distances. Comparing the DLVO and XDLVO energy profiles,
the AB interaction energy has resulted in a significant difference
between DLVO and XDLVO predictions at close approach. Our
results revealed that the XDLVO approach predicts the adhesion
and its reversibility more accurately than DLVO theory. The adhe-
sion tests conducted to verify the theoretical predictions showed
that PS adhesion to both substrata is much better than SE adhesion
as predicted by XDLVO approach. Also, the adhesion of both bacte-
rial strains to ITO surface is better than to glass surface as predicted
by DLVO and XDLVO theories.
To improve knowledge of cell adhesion, the present work utilizes
a submerged laminar microjet impingement technique, as a simple
and promising technique, to investigate the adhesion strength and
its reversibility. Our results showed that for both bacterial strains
the detachment strength from the hydrophilic glass surface was
lower than from ITO-coated glass, implying that these strains have
Fig. 5. Comparison between the detachment stresses of PS and SE bacterial strains
from (a) glass surface and (b) ITO-coated glass surface for four Reynolds/exposure lower affinity to glass surface as predicted by XDLVO approach.
time combinations (Re/t (seconds)). Our results also demonstrated that PS bacterial cells needed more
S. Bayoudh et al. / Colloids and Surfaces B: Biointerfaces 73 (2009) 1–9 9

strength to be detached than SE bacterial cells do from both sub- [12] S.E. Truesdail, J. Lukasik, S.R. Farrah, D.O. Shah, R.B. Dickinson, J. Colloid Interface
strata. Therefore, PS bacterial cells used their surface structures to Sci. 203 (1998) 369.
[13] C.J. Van Oss, R.J. Good, M.K. Chaudhury, J. Colloid Interface Sci. 111 (1986)
adhere more strongly and irreversibly on both substrata, while SE 378–390.
bacterial cells adhere reversibly and were not able to escape the [14] C.J. Van Oss, Colloids Surf. B: Biointerfaces 5 (1995) 91.
large energy barrier. [15] C.J. Van Oss, M.K. Chaudhury, R.J. Good, Chem. Rev. 88 (1988) 927.
[16] C.J. Van Oss, Interfacial Forces in Aqueous Media, Marcel Dekker, New York,
1994.
Acknowledgements [17] C.J. Van Oss, M.K. Chaudhury, R.J. Good, Sep. Sci. Technol. 23 (1989) 15.
[18] M. Katsikogianni, Y.F. Missirlis, Eur. Cells Mater. 8 (2004) 37.
[19] J.A. Brant, A.E. Childress, J. Mem. Sci. 203 (2002) 257.
The authors acknowledge the financial support received from [20] J. Azeredo, J. Visser, R. Oliveira, Colloids Surf. B: Biointerfaces 14 (1999) 141.
MIRA 2003, CMCU 05 S 0812. We also thank Prof. Amina Bakhrouf, [21] J.A. Brant, A.E. Childress, Environ. Eng. Sci. 19 (6) (2002) 413.
[22] P.K. Sharma, K. Hanumantha Rao, Colloids Surf. B: Biointerfaces 29 (2003) 21.
Head of the «Laboratoire d’Analyse et de Contrôle des Polluants
[23] S. Kang, H. Choi, Colloids Surf. B: Biointerfaces 46 (2005) 70.
Chimique et Biologique de l’Environnement», Faculté de Pharma- [24] C.J. Van Oss, R.J. Good, M.K. Chaudhury, J. Colloid Interface Sci. 111 (1986) 378.
cie de Monastir, for providing kindly the bacterial strains. We also [25] S. Bayoudh, A. Othmane, F. Bettaib, A. Bakhrouf, H. Ben Ouada, L. Ponsonnet,
thank Prof. Norman Richard Heckenberg (Centre for laser science, Mater. Sci. Eng. C 26 (2006) 300.
[26] C.J. Van Oss, Colloids Surf. A: Physicochem. Eng. Aspects 78 (1993) 1.
University of Queensland, Australia) for his helpful suggestions for [27] C.J. Van Oss, R.J. Good, M.K. Chaudhury, Langmuir 4 (1988) 884.
improving this manuscript. [28] C.J. Van Oss, J. Mol. Recognit. 16 (2003) 177.
[29] C.J. Van Oss, R.F. Giese, J. Dispersion Sci. Technol. 25 (5) (2004) 631.
[30] A. Grabbe, R.G. Horn, J. Colloid Interf. Sci 157 (1993) 375.
References [31] M. Elimelech, J. Gregory, X. Jia, R.A. Williams, Particle deposition & aggrega-
tion: measurement, in: Modeling and Simulation, Butterworth–Heinemann,
[1] M.N. Bellon-Fontaine, N. Mozes, H.C. Van der Mei, J. Sjollema, O. Cerf, P.G. Roux- Woburn, MA, 1995.
het, H.J. Busscher, Cell Biophys. 17 (1990) 93. [32] S. Bhattacharjee, M. Elimelech, J. Colloid Interface Sci. 193 (1997) 273.
[2] A.M. Gallardo-Moreno, M.L. Gonzalez-Martin, C. Pérez-Giraldo, E. Garduno, J.M. [33] E.M.V. Hoek, G.K. Agarwal, J. Colloid Interface Sci. 298 (2006) 50.
Bruque, A.C. Gomez-Garcia, Appl. Environ. Microbiol. 68 (5) (2002) 2610. [34] D.J. Shaw, Introduction to Colloid and Surface Chemistry, Butterworth, London,
[3] A.M. Gallardo-Moreno, M.L. Gonzalez-Martin, C. Pérez-Giraldo, E. Garduno, J.M. United Kingdom, 1980.
Bruque, A.C. Gomez-Garcia, Appl. Environ. Microbiol. 68 (11) (2002) 5784. [35] K.C. Marshall, J.A. Breznak, G.B. Calleja, G.A. Mcfeters, P.R. Rutter, Microbial
[4] J. Wang, N. Huang, C.J. Pan, S.C.H. Kwok, P. Yang, Y.X. Leng, J.Y. Chen, H. Sun, G.J. Adhesion and Aggregation, Springer-Verlag, New York, 1984.
Wan, Z.Y. Liu, P.K. Chu, Surf. Coat. Technol. 186 (2004) 299. [36] S. Bayoudh, L. Ponsonnet, H. Ben Ouada, A. Bakhrouf, A. Othmane, Colloids Surf.
[5] M. Hermansson, Colloids Surf. B: Biointerfaces 14 (1999) 105. A: Physicochem., Eng. Aspects 266 (2005) 160.
[6] B.V. Derjaguin, L. Landau, Acta Physicochim. URSS 14 (1941) 633. [37] K.J. Bundy, L.G. Harris, B.A. Rahn, R.G. Richards, Cell Biol. Int. 25 (4) (2001) 289.
[7] E.J.W. Verwey, J.Th.G. Overbeek, Theory of the Stability of Lyophobic Colloids, [38] M.D. Deshpande, R.N. Vaishnav, J. Eng. Mech. 109 (2) (1983) 479.
Elsevier Publishing Company, Inc., Amsterdam, 1948. [39] I. Hancock, I. Poxton, Bacterial Cell Surface Techniques, Wiley, New York, 1988.
[8] K.C. Marshall, R. Stout, R. Mitchell, J. Gen. Microbiol. 68 (1971) 337. [40] J. Sun, B.V. Velamakanni, W.W. Gerberich, L.F. Francis, J. Colloid Interface Sci.
[9] J.M. Meinders, H.C. van der Mei, H.J. Busscher, J. Colloid Interface Sci. 176 (1995) 280 (2004) 387.
329. [41] C. Gómez-Suárez, H.J. Busscher, H.C. van der Mei, Appl. Environ. Microbiol. 67
[10] H.H.M. Rijnaarts, W. Norde, E.J. Bouwer, J. Lyklema, A.J.B. Zehnder, Colloids Surf. (2001) 2531.
B: Biointerfaces 4 (1995) 5. [42] D.P. Bakker, H.J. Busscher, H.C. van der Mei, Microbiology 148 (2002) 597.
[11] J.A. Redman, S. Walker, M. Elimelech, Environ. Sci. Technol. 38 (2004) 1777. [43] R. Bos, H.C. van der Mei, H.J. Busscher, FEMS Microbiol. Rev. 23 (1999) 179.

View publication stats

You might also like